HomeMy WebLinkAboutAPA4158MANAGING WATER RESOURCES
FOR ALASKA'S DEVELOPMENT
IWR-105
INSTITUTE OF WATER RESOURCES
University of Alaskci
Fairbanks, Alaska 99701
Report IWR-105
MANAGING WATER RESOURCES
FOR ALASKA'S DEVELOPMENT
PROCEEDINGS
James W. Aldrich, Chairman
Alaska Section
American Water Resources Association
Institute of l~ater Resources
University of Alaska
Fairbanks, Alaska 99701
November 1983
-ii-
Aldrich, J.W., Chairman. 1983. Managing water resources for Alaska's
development. Proceedings. Alaska Section, American Water Resources
Association. Institute of Water Resources, University of Alaska,
Fairbanks. Report IWR-105. 404 pp.
-iii-
ACKNOI~LEDGMENTS
The Alaska Section of the American Water Resources Association would
like to acknowledge the following for their assistance in organizing the
1983 annual meeting which was held November 10-11 at the Chena Hot
Springs Resort near Fairbanks, Alaska.
Conference Chairman
Preparation of instructions to
authors and author assistance
Preparation of the brochure
announcing the annual meeting
General assistance with
preparations for the conference
James W. Aldrich
David r~. Hoch
Brent Petrie
Steven Mack
Judith Strohmaier
and the entire staff at
Conferences and Institutes,
University of Alaska
The Alaska Section would also like to thank all those members who
contributed their time and energy to performing the many thankless jobs
needed to conduct our first annual meeting.
-iv-
TABLE OF CONTENTS
Paper No. Title
1 Influence of Temperate Glaciers on Flood Events in Maritime
Alaska
J.H. Humphrey, C.J. Newton and R.D. Black
2 Sea Ice Characteristics in the Nearshore Environment
D.M. Hoch and B.T. Drage
3 Groundwater Occurrence in Eagle River, Alaska: With
Recommendations for Water Managers
J. A. Munter
4 Are We Ready to Manage Groundwater Resources In Alaska?
L.L. Dearborn
5 A l~ater Balance for Two Subarctic Watersheds
R.E. Gieck, D.L. Kane and J. Stein
6 Data Generated from Alaskan Hydropower Development
S.R. Bredthauer, J.H. Coffin and E.A. Machegiani
7 Trophic Status of Susitna River Impoundments
G. Nichols and L.A. Peterson
8 Environmental Effects of Ice Processes on the Susitna River
G.C. Schoch and S.R. Bredthauer
9 Alaskan Hydropower: Balancing the Long Run Advantages with
the Short Run Problems
J. S. Whitehead
10 Historical Development of Alaska Water Law
R.E. Miller
11 Asbestos Levels in Alaskan Drinking Water: A Preliminary
Study
W. C. Leitch
12 Effects of Gold Placer Mining on Interior Alaskan Stream
Ecosystems
J.D. LaPerriere, D.M. Bjerklie, E.V. Niewenhuyse,
R.C. Simmons, S.M. Wagner and J.B. Reynolds
-v-
TABLE OF CONTENTS (Continued)
Paper No. Title
13 Non-Solar Influences on Temperatures of South Coastal Alaskan
Streams
D.M. Bishop
14 A Comparison of Velocity Measurements Between Cup-Type and
Electromagnetic Current Meters
P.M. Wellen and D.L. Kane
15 Development and Use of a Resource Atlas for the Chugach
National Forest
D. Blanchet
16 The Aquatic Portion of the Integrated Resource Inventory,
Tongas National Forest -Chatham Area
D.A. Marion and S.J. Paustian
17 An Aquatic Value Rating Procedure for Fisheries and Water
Resource Management in Southeast Alaska
S.J. Paustian, D. Perkinson, D.A. Marion and P. Hunsicker
18 Water Quality Protection Program for Agriculture in Alaska
B.W. Rummel and W.C. Leitch
-vi-
by
Abstract
INFLUENCE OF TEMPERATE GLACIERS
ON FLOOD EVENTS IN MARITIME ALASKA
John H. Humphreyl, Carole J. Newton2,
and R. David Black3
Many stream basins in maritime Alaska have significant glacier-covered
areas. A methodology is required for determining design floods for facil-
ities in flood plains by glaciated streams. Published techniques for
estimating the magnitude of design floods for ungaged streams in maritime
Alaska were not applicable if glaciers were present. Available meteoro-
logic and hydrologic data were reviewed to select rain-flood events for
analysis. The largest floods occurred in glaciated basins in late summer
or fall after glacial snowpacks became melted at lower elevations and
saturated at higher elevations. Stream basins studied were located in
south coastal Alaska near Seward and Southeast Alaska near Juneau. Meteo-
rologic data, when not available in the basins, were estimated by correla-
tion of surface weather observations to upper air data. Unit hydrograph
and hydraulic modeling parameters were selected which were appropriate for
forest, rock slopes, snowpacks, firn, glacier ice, englacial tunnels, and
stream channels. Attenuation of rain-floods was related to glacier hydro-
logic controls. Snythetic unit hydrograph and routing parameters were
suggested for determining design floods from glaciated basins. Adjustment
factors to adapt regional non-glacial peak flow regression models to
glaciated terrain were derived.
Introduction
In 1980 an estimate of the 100-year recurrence annual peak flood was
required for an ungaged stream near Seward which had over 50 percent
glaciation. Existing regional regression methods were known to over-esti-
mate flood peaks. Their use would have substantially increased the cost of
flood protection for facilities adjacent to the stream. Efforts to esti-
mate the attenuation of the flood discharge by the glacier pointed out the
need for a justifiable methodology. This report determines techniques for
lHydrologist/Meteorologist, Ott Water Engineers, 2334 Washington Avenue,
Redding, CA 96001
2water Resources Specialist, Ott Water Engineers, 2334 Washington Avenue,
Redding, CA 96001
3water Resources Engineer, Ott Water Engineers, 4790 Business Park Blvd.,
Bldg. D, Suite 1, Anchorage, AK 99503
1-1
incorporating the effects of the glaciers and snowfields on the delay and
attenuation of runoff.
Methodology
Annual peak flow frequency analyses of streamflow records on stream basins
with significant glacier cover (defined as over 10 percent) were excluded
from the regression models in the U.S. Forest Service Water Resources Atlas
(Ott Water Engineers, 1979) because they were few in number and some
glaciated basins were notorious for catastrophic floods known as jokul-
hlaups or glacier outburst floods. A review of the literature indicated
that glacier stream basins subject to this type of flooding can be readily
identified and treated as a separate category to which the methodology
later described in this report would not apply. A report by Post and Mayo
(1971) inventoried streams known in Alaska to be subject to jokulhlaups.
Other papers by Richardson (1968), Young (1980), and Clarke (1982)
described circumstances in which jokulhlaups have occurred and described
theories regarding their origin. It is clear that nearly all jokulhlaups
are caused by movement of glacier ice blocking drainage, creating a dam and
backing up a lake in a valley, embankment, or sub-glacial basin. A jokul-
hlaup occurs periodically when hydraulic head becomes sufficient to
partially float the ice dam or to enlarge existing conduits by melting.
Lakes which are capable of causing jokulhlaups can be identified on aerial
photography. However, most glaciers have been in a state of retreat or
equilibrium in recent years. Glaciers in Alaska have been generally
retreating with only minor advances since 1760, the end of the Little Ice
Age (Field, 1975). Historic and geologic evidence of jokulhlaups or
glacier lake damming must also be considered since a glacier can advance
and renew the threat of jokulhlaups during the life of a project.
1-2
A much rarer type of jokulhlaup is associated with volcanoes. Subsurface
melting due to volcanic heat can create underground lakes with no outlet
until catastrophic release occurs. Glaciers on volcanoes with loose cinder
and ash material also have jokulhlaups due to sub-glacial landslides which
temporarily block flow before failing. This type of jokulhlaup is usually
a mud or debris flow which does not travel a large distance downstream.
Peak flow estimates for glaciated stream basins on volcanoes should be
treated with extreme caution.
In conclusion, jokulhlaups only occur on a small minority of glaciated
stream basins. Nearly all major floods in glaciated basins result from
combined rain and snowmelt events occurring late in the ablation season
(August-October) when snow cover on the glacier is at a minimum and ice
exposure is at a maximum. Even at this time, runoff delay and attenuation
due to the presence of the glacier is significant.
Criteria for selection of study stream basins were as follows:
1. Continuous stream flow records with peak flow events were avail-
able.
2. Glaciated and non-glaciated stream basins were located in closely
related geographic and climatic areas to facilitate comparisons
of storm events.
3. Meteorological data in the basins were available or it was likely
the data could be synthesized from representative surface weather
observations at nearby locations.
1-3
A review of glaciological research in Alaska, (Hubley, 1957; Marcus, 1963;
Miller, 1963; Wendler and Streten, 1969; Muir et al, 1971; and Tangborn et
al, 1977) indicated that extensive research appropriate to this study had
been done on Wolverine Glacier near Seward and on the Juneau Ice Field.
Research data from other areas, (Sharp, 1951; Field, 1975; Larson, 1978;
and Alaska Geographic Society, 1982) were missing continuous flow records
or meteorological data essential to runoff modeling.
All available U.S. Geological Survey streamflow records (1964, 1971, 1976,
1972-1982) were examined for stream basin records appropriate for use in
this study. Stream basins selected for study are shown in Table 1. Loca-
tions of these bas ins are shown in Figures la and lb. The base map for
these figures comes from Tangborn et al (1977).
TABLE 1
STREAMFLOW RECORDS
USGS Glacier
Gage Period Area Control
Number Name of Record (kul) (%)
15052000 Lemon Creek near Aug 51-Sep 73 33.0 86
Juneau
15052500 Mendenhall River May 65-Sep 82 221.4 77
near Auke Bay
15052800 Montana Creek near Aug 65-Sep 75 37.2 3
Auke Bay
15054200 Herbert River near Oct 66-Sep 71 148.0 71
Auke Bay
15101500 Greens Creek near Oct 78-Sep 82 59.3 0
Juneau
15109000 Fish Creek near Oct 58-Sep 78 35.4 0
Auke Bay
15236900 Wolverine Creek Oct 66-Sep 78 24.7 99
near Lawing
15238600 Spruce Creek near Sep 67-Sep 82 24.1 8
Seward
15238990 Upper Bradley River Oct 79-Sep 82 26.0 80
near Homer (Nuka Glacier)
1-4
' f\!
SPRUCE CREEK
O~~~liiiiiiiiiiiiiiiiiiii1 OOkm
Scale: 1-3,000,000
FIGURE 1A
, I
j
I
I
SOUTH COASTAL ALASKA
SHOWING STUDY AREAS AND
GLACIER EQUILIBRIUM LINE ELEVATIONS
1-5
HERBERT GLACIER
~-MENDENHALL GLACIER
~~-c------LEMON GLACIER
MONTANA CREEK~==~~~~~~~c---JUINEi~U
0 100km
Scale: 1-3,000,000
FIGURE 18
SOUTHEAST ALASKA SHOWING
STUDY AREAS AND GLACIER
EQUILIBRIUM LINE ELEVATIONS
1-6
Annual peak runoff events were examined for the locations listed in
Table l. The largest flood events from the glaciated basins were selected
if data for the same event was available on a nearby non-glaciated basin.
The opposite procedure, selecting the largest event on non-glaciated
basins, was not done since in many cases, glacier snow cover runoff atten-
uation or a low freezing level resulted in no corresponding significant
flood event on the glaciated basin. Table 2 lists runoff events selected
for analysis.
TABLE 2
FLOOD EVENTS
Aug Sep Oct Sep Sep Sep Sep Aug Sep Oct Oct Sep Sep
Basins '66 '67 '69 '70 '72 '74 '76 '77 '78 '78 '79 '81 '82
Lemon X X X X
Mendenhall X X X X X X X X X
Montana X X X X X
Herbert X X
Greens X X X
Fish X X X X X
Wolverine X X X X X
Spruce X X X X X
u. Bradley X
Computer simulation models capable of converting rainfall and snowmelt to
runoff incorporating suitable runoff delay and attenuation due to basin
hydrologic conditions were reviewed.
Single event models suitable for this purpose were described by Feldman
(1979) and Perrier et al (1977). The U.S. Army Corps of Engineers HEC-1
model was clearly superior due to its flexibility and ease of application.
The description and use of the HEC-1 model were found in U.S. Army reports
(1971, 198la, l98lb, l98lc, 1982).
1-7
Application of snowmelt equations in the HEC-1 model (U.S. Army, 1960;
Anderson, 1968) is simplified during storm events s1nce cloud cover
produces low solar radiation and since saturated air occurs at higher
elevations on the glacier. Meteorological input data includes air tempera-
ture, wind speed, and precipitation. Relationships between sea level and
upper altitude meteorologic data were available (Meier et al, 1971;
Tangborn et al, 1977) for Wolverine Glacier and for the Juneau Ice Field
(Miller, 1963). Information on precipitation increases with elevation were
also available (Ott Water Engineers, 1979). Comparisons of sea level
meteorologic data at Seward with Anchorage upper air data and of sea level
meteorologic data at Juneau with Yakutat upper air data facilitated esti-
mates of air temperature and windspeed at upper elevations on the glaciers.
Details on the storm event modeling will be given in a paper which may be
presented at the Third International Cold Regions Specialty Conference 1n
April 1984 at Edmonton, Alberta.
Delay and attenuation of glacier runoff due to snow, firn, and the
englacial conduit system were estimated from literature values. The
effects of seasonal snow cover on runoff are well documented due to
interest in this subject in non-glaciated regions. Snow cover is important
to glacial runoff processes in the early part of the ablation season but
becomes less important in the later part of the ablation season as the
snowline retreats to the accumulation area of the glacier. The delay and
attenuation of water infiltrating snowpacks were described (Denoth el al,
1978; Ebaugh and Dewalle, 1978; Wankiewicz, 1978; Anderson, 1973; Male and
Gray, 1981; Jordan, 1983a, 1983b) especially in a series of papers by
Colbeck (1972, 1974, 1975, 1977). Delay times for homogeneous wet snow
1-8
were l to 2 hours per meter of depth. Typical attenuation of peak inflows
was 15 percent per meter of snow} depending on inflow rates.
Snow 1mich has been subject to the passage of meltwater for some time tends
to reduce to rounded, coarse (l to 2 mm) grains, and further metamorphism
proceeds very slowly (Colbeck, 1982). By definition, firn is wetted snow
that has survived one summer without being transformed to ice. Firn
exhibits increasing density and gra~n s~ze with age. Firn accumulations on
the upper parts of glaciers can reach depths exceeding 50 meters before
having such a large grain size (2 to 10 mm) and density (0.8 g/cc) as to be
indistinguishable from ice. The exposed contact between firn and ice (the
firn line) can be found on the glacier when seasonal snow cover has melted.
But the depth of firn and the contact between firn and ice becomes increas-
ingly hard to define as the head of the glacier is approached. As new firn
is buried and compacted by additional years of snow it has increasingly
higher density and lower porosity. Delay and attenuation times for water
percolating firn have been found directly proportional to depth (Derikx,
1969; Stenberg, 1970; Krimmel et al, 1972; Meier, 1972; Colbeck, 1977;
Ambach et al, 1981; Oerter et al, 1981; and Paterson, 1981). Water perco-
lated firn at 0.16 to 0.31 meters per hour with an average of 0.22 meters
per hour. Runoff delay and attenuation is also caused by the inability of
englacial conduits to immediately convey runoff to the glacier terminus.
The size of englacial conduits is adjusted to meltwater peaks or previous
flood events during the ablation season. An unusual event exceeds their
capacity and causes water to back up internally in glacier cavities and
crevasses (Collins, 1979; Meier, 1972; and Oerter et al, 1981). Flow
conduit capacity can increase under the influence of hydraulic head and can
increase slowly as boundary friction dissipation of potential energy in
1-9
flowing water is converted into heat for enlarging channels. This melt
process is relatively slow compared to the time frame of most storm runoff
events on the lower glacier (6 to 12 hours). Theoretical calculations show
that enlargement rates of 50 to 100 percent in 24 hours would be typical.
This slow response would also serve to further delay and attenuate runoff
flowing through the glacier.
Most of the papers referenced regarding firn travel times also gave typical
transit times for rain/meltwater from various parts of the glacier to its
terminus. Golubev (1969) related time of travel to glacier area (lag time
in days is equal to four times the log of the area in square kilometers).
His paper also showed that unit hydrographs for discharge were closely
related to elevation area curves since low elevations have short lag times
and low attenuation and since high elevations have long lag times and high
attenuation. Campbell and Rasmussen (1969) stated glacier runoff delay is
probably related to glacier length.
Travel times and attenuation factors in the literature for glacier runoff
were generally given for regions of the glacier defined as above the firn
line (accumulation area), below the firn line (ablation area), and glacier
tongue (ablation area where glacier narrows and has numerous moulins and
crevasses). In some papers the term equilibrium line is given as a synonym
for firn line, but the equilibrium line is actually defined as the boundary
between accumulation and ablation zones where there is no net gain or loss
to the glacier surface. The firn line can be located up to a few hundred
meters above or below the equilibrium line depending on glacier topography
and whether the glacier is advancing or retreating. The firn line can
1-10
usually be determined from late ablation season aerial photographs while
the equilibrium line must be determined from comprehensive field studies.
Firn lines for the study glaciers were given by Marcus (1963), Heusser and
Marcus (1964), and Field (1975). Equilibrium lines for glaciers in Alaska
were g1ven by Tangborn et al (1977) and these lines are shown in Figures la
and lb.
Delay times, attenuation, and recession rates for glaciers were found 10
Behrens et al (1975), Larson (1978), Derikx (1969), Elliston (1969),
Colbeck (1978), Collins (1979), Ambach et al (1981), Oerter et al (1981),
and Paterson (1981). These factors are summarized as follows:
Accumulation area: Delay 4 to 10 days, recession 10 to 100 days
Ablation area: Delay 20 to 30 hours, recession 1 to 10 days
Tongue area: Delay l to 6 hours, recession 6 to 12 hours
Attenuation of flows from the accumulation area is so great that these
flows can be considered a base flow that does not increase significantly
during a storm event.
Runoff from the glacier was delayed, attenuated, and receded using unit
hydrographs. Description and use of synthetic unit hydrographs was given
in Linsley and Franzini (1979) and U.S. Army (1978). In the HEC-1 model
assigned lag times were: accumulation area--5 days, ablation area--
30 hours, and tongue area--5 hours. The time of maximum precipitation was
also compared with peak flow for all events listed in Table 2 in order to
estimate characteristic basin lag times. These times were consistent with
the tongue and non-glaciated areas.
1-11
Glaciated stream basins were divided into sub-basins according to areas of
accumulation, ablation, glacier tongues, and non-glaciated area. Non-
glaciated basins were divided into upper and lower zones based upon
differences in precipitation and runoff routing. Stream basins are shown
in Figures 2, 3, 4, 5, and 6.
An annual peak flow probability analysis was performed for those stream
gage records with sufficient length to extrapolate to 100-year recur-
rence. These included Lemon Creek, Mendenhall River, Montana Creek,
Herbert River, Fish Creek, Wolverine Creek, and Spruce Creek. Annual peak
flows were ranked, assigned a plotting position, and plotted on log-normal
probability paper. A best-fit straight line (zero skew) was used to extra-
polate to the 100-year recurrence.
One hundred-year-recurrence peak flows for the same stream basins were also
done using the regression equations in Ott Water Engineers (1979) and Lamke
(1979).
Results
Comparison of simulated and observed flood hydrographs gave strong support
to the theoretical lag times and synthetic unit hydrographs assigned to the
stream basin models. In glaciated stream basins, runoff contributions from
the accumulation and ablation areas were not a significant component of the
peak flow. In some cases recession flow from upper glacier areas from
earlier storm events represented a significant base flow to the modeled
storm event. For major events, base flow from preceding events was up to
10 to 20 percent of the peak runoff rate. The model was most sensitive to
1-12
JUNEAU
AIRPORT
FISH CREEK
GAGING
STATION
'< \
/ LOWER \
L/\ \
) \
[/" \ .. (/'--.,,. -i~J
2.7km.
~~iiiiiil
0
Scale: 1-880011.
GAGING STATION
LEMON CREEK
1-13
JUNEAU·
FIGURE 2 cc-
LEMON CREEK AND
FISH CREEK BASINS ·
HERBERT
RIVER
I
I
\
\ GAGING
STATION LOCAL \ • . . '\ .. ,
o~~iiiiiiiiiiiiiiiiE~4.0 km
Scale: 1-13,2001!.
Q: \ ACCUMULATION
lil"-. 0 ·~ "l" .
._e I
Q: • /fj ABLATION
If :r:
_/ "')
'--
ABLATION { \
~---~.~.·
~
0
5
C!l
TONGUE
LOCAL
ACCUMULATION
FIGURE 3
MENDENHALL RIVER, MONT ANA CREEK
AND HERBERT RIVER BASINS
1-14
0 2.2km
~liiiiiiiiiiiiii
Scale: 1-720011.
1-15
UPPER BASIN
FIGURE 4
GREENS CREEK BASIN
o 1.6 km
~~
Scale: 1-528011.
ACCUMULATION
')\-_
i i''' 1: !{
/-----/ ..........
TONGUE AREA
1-16
ABLATION
FIGURE 5
WOLVERINE CREEK BASIN
0 1.6 km
~'!"!iiiiiiiiiiiiiiiiil
Scale: 1-528011.
NON-FORESTED /
/
/
/
_/ .
..,....., ~"' ...
_....; ~0~
( . / " . -:/ ---
1-17
/
/
GAGING STATION
FIGURE 6
SPRUCE CREEK BASIN
the area assigned to the glacier tongue and to the precipitation multiplier
used to estimate precipitation on the lower glacier from the sea level
index station. Both of these values were varied within reasonable limits
as part of the calibration process until consistent results were obtained
from all storm events on each stream basin. Modeled peak flows and runoff
volumes had errors less than 30 percent which were quite acceptable consid-
ering the lack of in-basin meteorological data and the purpose of the
study.
One hundred-year-recurrence annual peak flows from a probability analysis
are shown in Table 3. Average runoff and precipitation data which are
derived from the average runoff data are also shown. Mean annual precipi-
tation and basin area are the most important inputs to regional peak flow
regression methods. Predicted 100-year recurrence based on regression
equations from Lamke (1979) and Ott Water Engineers (1979) are shown for
comparison. The final column on the table shows predicted flows using Ott
Water Engineers (1979) when the accumulation and upper ablation areas are
excluded.
TABLE 3
100-YEAR RECURRENCE FLOOD PEAKS
100-Year Flood
Average Annual Excluding
Prec ip Lamke, OTT, Area
Runoff itation Observed 1979 1979 Above
cm/yr cm/yr m 3; sec m 3; sec m 3; sec Tongue
Lemon Creek 439 465 68.0 153. 126. 62.3
Mendenhall Rv. 460 483 368 0 795. 745. 380.
Montana Creek 234 264 85.0 124. 85.0
Herbert River 356 381 234. 503,0 411. 241.
Fish Creek 198 229 85.0 111.3 79.3
Wolverine Ck. 315 340 59.5 98.0 90.7* 60.9
Spruce Creek 284 318 113. 89.8 80.7
Juneau Airport 140
Juneau 234
Seward 160
*120. -if based on correlation with Spruce Creek
1-18
On Table 3, note that the predicted flow for Spruce Creek was significantly
lower than the observed 100-year recurrence peak flow. Using the Ott Water
Engineers (1979) method gave an underprediction of 30 percent. The origi-
nal regression model had 11 years of record on this creek and gave an under
prediction of 17 percent. These errors are well within the 90 percent
confidence limits of the regression equations.
Conclusions
The HEC-1 runoff simulation model results showed that runoff from accumula-
tion areas of the glacier (above the firn line or equilibrium line) is not
a significant contributor to annual peak flood events. A portion of the
ablation zone 1.s also non-contributing to peak events. The lower boundary
of this ablation area is above the heavily crevassed ice tongue area 100 to
200 meters below the equilibrium line shown in Figures la and lb. Runoff
from non-glaciated or rock areas which reaches the glacier above this
elevation also does not contribute to peak flows. Table 3 shows that if
all runoff area above the glacier tongue is excluded from the regression
model, then reasonable estimates of flood peaks from glaciated basins are
obtained. The last column in Table 3 closely agrees with observed 100-year
flood peaks. Suggested application of the method follows:
Step 1. Determine applicability. This method cannot be used on any basins
with a jokulhlaup history or potential threat. The method may also be
applied to glaciated basins with continental climate (that is, the glacier
interior stays below O"C) but will probably overestimate flows since runoff
attenuation on continental glaciers whould usually be greater than on mari-
time temperate glaciers.
1-19
Step 2. Determine the area below accumulation and ablation areas but
include area controlled by the glacier tongue. The boundary between the
areas is 100 to 200 meters below the equilibrium line.
Step 3. Calculate peak flow using Lamke (1979) or Ott Water Engineers
(1979) regression equations and compare to results using correlation with a
representative non-glaciated basin.
Step 4. Check results against a peak flow regime which would be consistent
with channel morphology, channel bank full capacity, and high water marks.
Note that although glaciated basins have lower peak flows per unit area
compared to non-glaciated basin, high flows have much longer duration. The
relatively long duration of high flows should be considered in flood
control design applications.
A comparison of methods was made for Fourth of July Creek near Seward. The
stream basin is shown in Figure 7. The excluded area where runoff is
controlled by snow fields and the upper zone of the glacier 1s shown. The
elevation of this boundary is approximately 850 meters, which is 150 meters
below the equilibrium line shown in Figure la. The total basin area was
66.1 km2, the lower area was 36.7 km2, and mean annual precipitation
was 330 em. The mean annual basin precipitation is for the entire basin
including the glacier runoff controlled area.
An example of each of the six methods applied to Fourth of July Creek
follows:
1. 100-year recurrence peak using Lamke (1979): 220 m3 jsec.
1-20
o~~~iiiiiiiiiiiiiiiiiii2.2km
Scale: 1-720011.
IG·.·L .. ACIER CONTROLLED · \ ABOVE TONGUE
J'\. ...... . ~"-
) ··~····· \ \
.\ ' l
I
I
/ -_.;.-/
PERMANENT
FIGURE 7
FOURTH OF JULY CREEK BASIN
1-21
2. 100-year recurrence peak using Ott Water Engineers (1979):
230m3/sec.
3. 100-year recurrence peak using correlation with nearby Spruce
Creek: 310 m3/sec.
4. 100-year recurrence using Lamke (1979) corrected for glaciated
area: 132m3/sec.
5. 100-year recurrence using Ott Water Engineers (1979) corrected
for glaciated area: 129 m3/sec.
6. 100-year recurrence using Spruce Creek correlation corrected for
glaciated area: 165 m3/sec.
The result of Method 6 is recommended since a nearby representative
non-glaciated stream record existed. Otherwise for other basins, Method 5
in south coastal and Southeast Alaska and Method 4 elsewhere in Alaska
might be appropriate.
1-22
REFERENCES
Alaska Geographic Society, 1982, Alaska's Glaciers, Alaska Geographic,
9(1), 144 pp.
Ambach, W., M. Blumthaler, and P. Kerchlechner, 1981, Application of the
Gravity Flow Theory to the Percolation of Melt Water Through Firn, Journal
of Glaciology, 27(95), pp. 67-75.
Anderson, E. A., 1968, Development and Testing of Snow Pack Energy Balance
Equations, Water Resources Research, 4(1), pp. 19-37.
Anderson, E. A., 1973, National Weather Service River Forecast System-
Snow Accumulation and Ablation Model, NOAA Technical Memorandum NWS
Hydro 17.
Behrens, H., H. Bergmann, H. Moser, 1975, On the Water Channels of the
Internal Drainage System of the Hintereisferner, Otztal Alps, Austria,
Journal of Glaciology, 14(72), pp. 375-382
Clarke, Garry K., 1982, "Glacier Outburst Floods from 'Hazard Lake', Yukon
Territory, and the Problem of Flood Magnitude Prediction", Journal of
Glaciology, 28(98), pp. 3-21.
Campbell, William J., and Lowell A. Rasmussen, 1969, The Production, Flow,
and Distribution of Melt Water in A Glacier Treated as a Porous Medium,
Symposium on the Hydrology of Glaciers, Cambridge, Publication No. 95,
International Association of Scientific Hydrology, pp. ll-28.
Colbeck, S. C., 1972, A Theory of Water Percolation in Snow, Journal of
Glaciology, 11(63), pp. 369-385.
Colbeck, S. C., 1974, On Predicting Water Runoff from a Snow Cover,
Advanced Concepts and Techniques in the Study of Snow and Ice Resources,
National Academy of Sciences, Washington, DC, pp. 55-66.
Colbeck, S. C., 1975, Analysis of Hydrologic Response to Rain-on-Snow,
Research Report 340, Cold Regions Research and Engineering Laboratory,
Hanover, NH, 9 pp.
Colbeck, S. C., 1977, Short-Term Forecasting of Water Run-Off from Snow and
Ice, Journal of Glaciology, 19(81), pp. 571-588.
Colbeck, S. C., 1982, An Overview of Seasonal Snow Metamorphism, Reviews of
Geophysics and Space Physics, 20(1), pp. 45-41.
Collins, David N., 1979, Quantitative Determination of the Subglacial
Hydrology of Two Alpine Glaciers, Journal of Glaciology, Vol. 23, No. 89.
Denoth, A., W. Seidenbusch, M. Blumthaler, and P Kirchlectner, 1978, Some
Experimental Data on Water Percolation Through Homogeneous Snow, Proceed-
ings Modeling of Snow Cover Runoff, U.S. Army Cold Regions Research and
Engineering Laboratory, Hanover, NH, pp. 222-253.
1-23
Derikx, Leo, 1969, Glacier Discharge Simulation by Groundwater Analogue,
Symposium on the Hydrology of Glaciers, Cambridge, Publication No. 95,
International Association of Scientific Hydrology, pp. 29-40.
Elliston, G. R., 1969, Water Movement Through the Gonergletscher, Symposium
on the Hydrology of Glaciers, Cambridge, Publication No. 95, International
Association of Scientific Hydrology, pp. 79-84.
Feldman, D., 1979, Flood Hydrograph and Peak Flow Frequency Analysis,
Preceedings of the Engineering Foundation Conference on Improved Hydrologic
Forecasting, Why, and How, ASCE, New York, pp. l 19.
Field, William 0., ed., 1975, Mountain Glaciers of the Northern Hemisphere,
Volume 2, Chapter l, Coast Mountains: Boundary Ranges of Alaska, Volume 2,
Chapter 4, Kenai Mountains, Cold Regions Research and Engineering Labora-
tory, Hanover, NH.
Golubev, G. N., 1969, Analysis of the Run-Off and Flow Routing for a Moun-
tain Glacier Basin, Symposium on the Hydrology of Glaciers, Cambridge,
Publication No. 95, International Association of Scientific Hydrology,
pp. 41-52.
Heusser, Calvin J., and Melvin G. Marcus, 1964, Surface Movement, Hydrolog-
ical Change and Equilibrium Flow on Lemon Creek Glacier, Alaska, Journal of
Glaciology, 5(37), pp. 61-75.
Hubley, Richard C., 1957, An Analysis of Surface Energy During the Ablation
Season on Lemon Creek Glacier, Alaska, Trans AGU 38(1).
Jordan, Peter, 1983, Meltwater Movement in a Deep Snowpack, l; Field Obser-
vations, Water Resources Research 19(4), pp. 971-978.
Jordan, Peter, 1983, Meltwater Movement in a Deep Snowpack, 2; Simulation
Model, Water Resources Research 19(4), pp. 979-985.
Krimmel, R. M., W. V. Tangborn, and M. F. Meier, 1972, "Water Flow Through
a Temperate Glacier", Proceedings of the Banff Symposium on the Role of
Snow and Ice in Hydrology, IAHS Publication No. 107, pp. 401-416.
Lamke, R. D., 1979, Flood Characteristics of Alaskan Streams, Water
Resources Investigations 78-129, U.S. Geological Survey, Anchorage, Al<,
61 pp.
Larson, Graham J., 1978, Meltwater Storage in a Temperate Glacier,
Borroughs Glacier, Southeast Alaska, Institute of Polar Studies, Report
No. 66, The Ohio State University, Columbus, OH, 56 pp.
Linsley, Ray K., and Joseph B. Franzini, 1979, Water Resources Engineering,
McGraw-Hill Book Company, 716 pp.
Male, D. H., and D. M. Gray, 1981, Snowcover Ablation and Runoff, Handbook
of Snow, Pergamon Press, Willowdale, Ontario, Canada, Chapter 9,
pp. 360 436.
1-24
Marcus, Melvin G., 1963, Climate-Glacier Studies in the Juneau Ice Field
Region, Alaska, University of Chicago, Department of Geography, Paper
No. 88, 128 pp.
Mathews, W. H., 1969, Record of Two Jokulhlaups, Symposium on the Hydrology
of Glaciers, Cambridge, September 1969, Publication No. 95, International
Association of Scientific Hydrology, pp. 99-110.
Meier, Mark F., Wendell V. Tangborn, Lawrence R. Mayo, and Austin Post,
1971, Combined Ice and Water Balances of Gulkana and Wolverine Glaciers,
Alaska and South Cascade Glacier, Washington, 1965 and 1966 Hydrologic
Years, Geological Survey Professional Paper 715 A, U.S. Government Printing
Office, Washington, DC, 23 pp, 6 plates.
Meier, Mark F., 1972, Hydraulics and Hydrology of Glaciers, The Role of
Snow and Ice in Hydrology, lASH Publication No. 107, Proceedings of the
Banff Symposium, pp. 353 370.
Miller, Maynard M., 1963, Taku Glacier Evaluation Study for State of Alaska
Department of Highways, Alaska Dept. of Highways, Juneau, AK.
Murphy, Thomas D., and Seymour Schamach, 1960, Mountain Versus Sea Level
Rainfall Measurements During Storms at Juneau, Alaska, Journal Hydrology 4,
pp. 12-21.
Oerter, H., D. Baker, H. Moser, and 0. Reinwarth, 1981, Glacial-Hydrolog-
ical Investigations at the Vernagtferner ~lacier as a Basis for a Discharge
Model, Nordic Hydrology, 12, pp. 335-348.
Ott Water Engineers, 1979, Water Resources Atlas, U.S. Forest Service,
Region 10, Juneau, N<, 90 pp.
Paterson, W.S.B., 1981, The Physics of Glaciers, Pergamon Press, Elms-
ford, NY, Second Edition, 380 pp.
Perrier, E. R., J. Harris, and W. B. Ford, 1977, A Comparison of Deter-
ministic Mathematical Watershed Models, Paper No. 77-2047, American
Society of Agricultural Engineers, St. Joseph, MI, 115 pp.
Post, Austin and Lawrence R. Mayo, 1971, Glacier Dammed Lakes and Outburst
Floods in Alaska, Hydrologic Investigations Atlas HA 455, U.S. Geological
Survey, Washington, DC, 10 pp, 3 maps.
Richardson, Donald, 1968, Glacier Outburst Floods in the Pacific North-
west, Geological Survey Research 1968, Chapter D, U.S. Geological Survey
Professional Paper 600-D, U.S. Government Printing Office, Washington, DC,
pp. D79-D86.
Sharp, Robert P., 1951, Accumulation and Ablation on the Seward-Malaspina
Glacier System, Canada-Alaska, Journal of Geology, 62, pp. 725-744.
Stenberg, Thorsten, 1970, Delay of Run-Off from a Glacier Basin, Geogra-
fiska Annaler 52A(l), pp. 1-30.
1-25
Tangborn, Wendall V., Lawrence R. Mayo, David R. Scully, and Robert M.
Krimmel, 1977, Combined Ice and Water Balances of Maclure Glacier,
California, South Cascade Glacier, Washington, and Wolverine and Gulkana
Glaciers, Alaska, 1967 Hydrologic Year, Geological Survey Professional
Paper 715-B, U.S. Government Printing Office, Washington, DC, 20 pp,
4 plates.
U.S. Army Corps of Engineers, 1960, Runoff from Snowmelt, EM 1110-2-1406,
66 pp.
U.S. Army Corps of Engineers, 1971, Hydrologic Engineering Methods for
Water Resources Development, Volume 1, Requirements and General Proce
dures, The Hydrologic Engineering Center, Davis, CA.
U.S. Army Corps of Engineers, 1978, HYDPAR (Hydrologic Parameters), Users
Manual, Hydrologic Engineering Center, Davis, CA, 60 pp.
U.S. Army Corps of Engineers, 1981, The New HEC-l Flood Hydrograph Pack-
age, Technical Paper No. 82, Hydrologic Engineering Center, Davis, CA,
zs-pp.
U.S. Army Corps of Engineers, 1981, HEC-1 Flood Hydrograph Package, Users
Manual, Hydrologic Engineering Center, Davis, CA, 192 pp.
U.S. Army Corps of Engineers, 1981, HEC-1 Flood Hydrograph Package, Users
Manual, Section 16: Input Description, Hydrologic Engineering Center,
Davis, CA, 110 pp.
U.S. Army Corps of Engineers, 1982, Hydrologic Analysis of Ungaged Water-
sheds Using HEC-1, Training Document No. 15, The Hydrologic Engineering
Center, Davis, CA, 160 pp.
U.S. Geological Survey, 1964, Compilation of Records of Surface Waters of
Alaska, October 1950 to September 1960, Water Supply Paper 1740, 86 pp.
U.S. Geological Survey, 1971, Surface Water Supply of the United States,
1961-65, Part 15, Alaska, Water-Supply Paper 1936, 342 pp.
U.S. Geological Survey, 1976, Surface Water Supply of the United States,
1966-70, Part 15, Alaska, 425 pp.
U.S. Geological Survey, 1972 through 1983, Water Resources Data for Alaska,
Water Year 1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980,
1981, 1982, Anchorage, AK.
Wankiewicz, Anthony, 1978, A Review of Water Movement in Snow, Proceedings
Modeling of Snow Cover Runoff U.S. Army Cold Regions Research and Engineer
ing Laboratory, Hanover, NH, pp. 222-253.
Wendler, G., and A. Streten, 1969, A Short-Term Heat Balance Study on a
Coast Range Glacier, Pure and Applied Geophysics 77, pp. 68-77.
Young, G. L., 1980, Monitoring Glacier Outburst Floods, Nordic Hydrology
11(5), pp. 285-300.
1-26
SEA ICE CHARACTERISTICS IN THE NEARSHORE ENVIRONMENT
David M. Hoch 1 and Brent T. Drage 2
Abstract
The development of large-scale marine structures in arctic waters is a
relatively recent phenomenon. In the Alaskan experience, the design of
such offshore and coastal structures has been predominantly associated
with North Slope oil field development. Engineering design for such
hostile ice environments is still in its infancy, and many unanswered
questions await further analysis and testing. With these constraints in
mind, contemporary design philosophy often utilizes broad generalizations
regarding sea ice strength in an attempt to insure conservative
estimates.
A specific case history is presented in which a generalized analysis
results in an excessively high value for sea ice strength. A field
sampling program and a subsequent laboratory analysis suggests that
substantially lower values may, in fact, be the case. By utilizing the
knowledge gained from such research, significant savings in construction
costs may be realized for ice-affected structures. It is the opinion of
the authors that sea ice strength is a more "site-specific" phenomenon
than is commonly thought. A defense of this argument is presented, and
its "real world" applicability is discussed. The economic benefits
realized by this approach may be substantial 1-1hen viet-~ed in the context of
the potential offshore and coastal development in the Alaskan Beaufort
Sea.
Introduction
The discovery of world-class oil reserves in the North American arctic
during the 1960's has spurred the development of large-scale marine
structures for the arctic environment. As oil exploration efforts begin
to move into the true offshore regions of the Arctic Ocean, the
requirements for marine design become more and more demanding and
complex. During the early days of arctic oil field development, design
1Hydrologist, Peratrovich, Nottingham
Avenue, Suite 101, Anchorage, Alaska
& Drage, Inc., 1506 West 36th
99503
2 Vice President, Peratruvich, Nottingham & Drage, Inc., 1506 West 36th
Avenue, Suite 101, Anchorage, Alaska 99503
2-1
requirements were limited tu nearshore structures such as barge uffluading
ducks which were used tu support land-based activities. Subsequent
exploration, however, revealed the existence uf massive uil reserves in
the true offshore regions uf both the Canadian and Alaskan arctic. With
these discoveries came the need fur sophisticated marine design techniques
which could be applied tu offshore drilling and production structures in
such ice impacted environments.
The tremendous natural forces affecting facilities in such an environment,
combined with the extreme oust uf construction (and repair) in these
remote regions, have led tu the widespread use uf conservative estimates
for design parameters. Empirical data fur such design situations are
still relatively scarce, although data collection efforts are increasing
each year. Faced with these limitations, engineers must often utilize
simplistic generalizations when faced with a design problem. It is
generally felt that a conservative estimate must be made until a more
substantial data base is developed.
Unfortunately, the extreme cost uf fabrication and construction makes
"over-design" an especially expensive process in the arctic regions. In
many cases, a well-executed data collection effort could provide site-
specific information which would justify the relaxation uf some uf the
accepted design criteria currently in use. The following paper describes
such a data-collection effort. The engineering firm in question was faced
with the task uf developing ice loading criteria fur a marine facility un
the shu res uf Alaska • s Beaufort Sea. It was felt that the currently
utilized values fur ice strength would result in a potential over-design
if applied tu this specific project. In order tu verify this belief, a
2-2
concise data-collection program was initiated tu provide the necessary
data. Calculations based un these results did indeed lead tu the
relaxation uf initial design specifications. Project construction ousts
were reduced considerably. It is the opinion uf the authors that similar
situations will be encountered in many future offshore design projects.
This will become a significant economic factor as arctic offshore
development becomes more intensive.
Project Location
The location uf the proposed design is shown in Figure 1, Of primary
interest is the protective chain uf barrier islands located to the north
uf OU.ktuk Point. The Junes Islands/Thetis Isla'1ds gruup furms a natural
line uf separatiun between the offshore environment uf the Beaufurt Sea
and nearshore Simpsun Lagoun regiun. The Beaufort Sea is, uf course,
subjected tu all the large-scale ice furces commonly associated with
arctic waters: massive pressure ridges, multi-year ice floes, long
fetches, and generally extreme ice cunditiuns. Such an environment can be
considered typical uf must uf the North American arctic coast, and as such
it has received must uf the experimental attention during the past several
decades.
The project location under discussion, however, is pusitiuned shoreward uf
a chain uf barrier islands, and as such is relatively protected from the
large driving forces fuund offshore. It was felt that the commonly
applied values fur sea ice strength and driving forces could perhaps be
reduced fur a structure being placed in this area. The question, uf
course, is: How much uf a reduction can be justified? Assuming that the
project site depicted in Figure is subjected tu site-specific ice
2-3
conditions, what sort of values can be realistically expected, and how can
this be verified in the field?
FIGURE 1 LOCATION MAP
Experimental Design
The primary site-specific value to be discussed in this paper concerns sea
ice strength. It was felt that the unique geographical location of
Simpson Lagoon contributes to the formation of sea ice which is quite
different in character to that found offshore. If a field sampling
program could be devised which would provide empirical data on this issue,
a relaxation of the stringent design criteria currently in use could
perhaps be justified. The overall oust savings to the project cuuld be
significant.
2-4
Fortunately, an un-guing data collection pruject had been conducted in the
Simp sun Laguun regiun fur the twu winters preceding this study. Munthly
measurements uf ice thickness, cuntinuuus air temperature, ice prufile
temperatures, sea water salinity/ cunducti vity, and seawater temperature
had been recurded fur this periud uf recurd. With this data base tu build
upun, a sampling plan was devised in which several cures wuuld be
retrieved frum the ice cuver. These samples wuuld then be examined in
the laburatury, and an analysis made uf the crystalline structure,
salinity cuncentratiuns, and brine vuid ratius uf the ice.
Field sampling was performed un March 15, 1983. Three cures were cut frum
the 5.4 ft. thick ice cuver using a standard CRREL/SIPRE curing tuul. The
cures were cut by hand and packed in an insulated chest fur shipment tu
the laburatury.
Parameters Hhich were sampled, analyzed, and/ur measured during this
experiment consisted uf:
u Ice Thickness
u Ice Crystalline Structure
u Ambient Air Temperature
u Ice Temperature Prufile
u Ice Salinity
u Ice Brine Vuid Ratiu
u Seawater Salinity and Temperature
With these site specific data, definitive ice strengths cuuld be
determined frum standard relationships. Fur the purpuse uf this study,
the relationships published in API Bulletin 2N "Planning, Designing, and
Constructing Fixed Offshure Structures in Ice Environments," (API, 1982)
were used primarily.
2-5
Ice Crys•alline S•ructure
The three cores were retrieved in mid-March at a location approximately
200 feet offshore from Oliktol{ Point, Simpson Lagoon. Ice thickness at
this location was recorded at 5.4 feet. Each core had about 0.2 feet of
granular/ snow ice on the surface; the remaining ice column consisted of
columnar crystalline structure. Due to Simpson Lagoon's protection by the
barrier islands and mainland, this result was to be expected. During the
autumn freezeup period, frazil ice may form on the surface if a proper
!-lind/temperature regime occurs. But after an ice cover forms and reaches
a thickness of about 8 inches, calculations suggest that there will not be
sufficient environmental driving force available to fracture and move the
ice sheet. Ice would therefore grow with a columnar crystalline pattern
throughout the winter under these stable conditions.
0
-I
,,
I ' ,--
1 . '-./ I
/ " /-1 I / '. \
//~ I / ,/ \
,-,. ~ 1 I \t--tt>->:-
,, I ,' ~ I
I ' ' I . ' I . '· I I \ ,-i I I I ..... : I I , , I' . ' ;'.\J\ ,_: '.\ ,,, t:\v _, J,.
1
.,,: ~' ,
I ' • I I : ','\ \.,j / \ ', \ !\
,: ./'• / ~I 1\1, .... : : \ I I I / t-, I : ' I 1 I I ,, ,'
I ~ ·, ' ' f ··;\ I
I .. , ; I 1 •. -
\ / I ~ \
I / ... ·~-/ •\ ,, ' }' ..._,, ..... ~\ .. :
' ,-
•/ ,.
FIGURE 2 DAILY MAX/MIN/MEAN TEMPERATURES, 1983
2-6
Ambien~ Air Tempera~ure
The 1982/83 winter was considered to be somewhat colder than average.
Figure 2 presents recorded daily air temperatures measured at a nearby
point fur the preceding 30 days. The prevailing air temperature was
relatively cold during the 2-week period preceding the March 15, 1983
sampling period,
T"'-'-1A+-""1lirze. "F r;;. 1.0 1? zo
~
-I
-z
-)
FIGURE 3 ICE TEMPERATURE PROFILE, MARCH 15, 1983
Ice Tempera~ure Profile
Two thermistor s~rings were installed in the ice cover in January of
1983. One string was located in the natural ice and another was located
beneath a 40 ft. x 40 ft, slab of 4 1/2 in. thick styrofoam insulation
which had been in place for approximately four months. The March 15, 1983
temperature profiles are shown in Figure 3. The ice temperature on this
2-7
date is considered to be cold, with an average temperature of +9°F at the
time of measurement. Note the warm temperatures that exist in the bottom
0.5 to 1.0 feet of ice. The cores that were retrieved revealed a soft
texture on this lower portion of ice.
Ice Salinity
Following retrieval, each sample core was divided into thirds and the
entrapped salinity was measured in the laboratory utilizing techniques
recommended by CRREL (Cox & Weeks, 1975). Figure 4 summarizes the
measured salinities for each core segment. Salinities averaged about 6
parts per thousand, a figure which correlates well with published findings
on this tonic.
With the salinity data and ice temperature profile discussed above, sea
ice brine void ratios could be computed.
Ice Brine Void Ratio
Most relationships that define ice strength properties utilize the brine
void ratio as a prime parameter. Brine void ratio can be defined as the
square root of the brine volume. Figure 4 summarizes the brine void
ratios for each segment of the three ice columns. The upper third of each
core has a brine void ratio similar to those that are reported in the
published literature. It can be seen, however, that the ratio increases
with ice depth as the relative warmth of the seawater is approached.
Sea ice loading calculations traditionally use the brine void ratio
characteristic of the upper 1/3 of the ice column to define ice strength
parameters. The results of this experiment --in particular the warm ice
temperature and high brine void ratios lead to a theory that an
2-8
-I
-2
~' \.[\
~ I
..L -3
',-/
d;
L
B:
-1
BRINE VOID RATIO & SALINITY [0 /aa]
OF ICE CORES
FIGURE 4
2-9
"effective ice thickness" shuuld be cunsidered when defining ice
strengths. This "effective thickness" wuuld be defined by a temperature
threshuld differentiating between guud bunded crystalline ice and slushy
warm ice. Fur this particular case, it appears that the effective ice
thickness shuuld be 0.5 tu 1.0 feet thinner than the measured ice
thickness.
Seawater Salinity and Temperature
Nearshure seawater salinity increases as the ice thickness increases. As
the ice cuver gruws, it excretes brine intu the parent water. Since there
is little circulatiun and dilutiun in this shalluw nearshure envirunment,
the salinity is able tu increase tu abuve nurmal levels. These higher
salinity values, cumbined with the relatively warm seawater temperatures,
tend tu retard ice gruwth rates during late winter. The brine excretiun
prucess is likewise retarded, alluwing higher brine vuid ratius, and
hence, weaker ice.
As was mentiuned abuve, must existing literature correlates ice strength
pruperties with ice brine void ratiu. Figures 5 thruugh 8 are taken frum
API Bulletin 2N "Planning, Designing, and Constructing Fixed Offshure
Structures in Ice Envirunments," (API, 1982). Based upun the wurk uf
researchers cited in these figures, various mechanical pruperties uf sea
ice have been plutted as a functiun uf brine vuid ratiu. We have uverlain
the results uf the Simp sun Laguun experiment un these figures by way uf
cumparisun. The dashed lines represent pruperties derived from the
upper one-third uf the cures; the solid lines represent the luwer third.
In Figure 5, the darker stippling is used fur the lower third of the
cures.
2-10
250
ALL ttSTS @ C'ONSTlUIT
SALINITY or 1·9 PPT
ISE:E: E:QUA.TIOil l•l)
.......... ot tn Utlb
11 ........... ·-· ... o ..... ,CA>. .... ......
•~'""""'-· .... .. o-..-. ,.,,.,.
1 '" .._, ............
.0
SEA ICE TENSILE STRENGTH AS A FUNCTION
OF THE SQUARE ROOT OF BRINE VOLUME
(DYKINS (1971))
160
FIGURE 5
IU.L 't[STS ' cotlttA.tn'
SA.LlNITY or 7-9 PPT
{S"U: EQUA.TtOil J•l)
... <P ........ c..to, ...... ..
.6.0• ..... -....... .._ .......... , ____ ... , .. , ....
~
300 I I I ' 2.0
-1.5 -200 ~
5 --~~ ~ I •• I ••• 1.0 0 • z . ~~ ~ I • •• •• -100 I , • • • • • • • ~ I • ...... • .5 ~
I • • I
Vb I"' Brine Volume ,
0~-U--~~~~~-L--~~
0.10 0.20 0.30 0.40 vv;;
SEA ICI' SHEAR STRENGTH AS A FUNCTION
OF THE SQUARE ROOT OF BRINE VOLUME
(PAIGE (1967))
FIGURE 6
II LA,OOMT'ORY au.H T:.STS
o ru:.LD au.~~ nSTS
• ~
"'
.. -.,..,,.,-.,,_,.., I "' ,_.,....., .....,.,.
1.0
AU. ttsTS f COIS'r.o.lll'
So\LDII'I'T Or l•' PP'r
tSU: tQU.\'rlOil J.-U I t5t e~rtl>t!Ct LIMITS
'.s U ~ lfl). or SA.XI'LES ll:~D 5000 -···""'"' .
~ ~ ' ___ ._.,_ , .. •coo
.s i ••• •• JOOO
SEA ICE FLEXURAL STRENGTH AS A
FUNCTION OF THE.SQUARE ROOT OF BRINE
VOLUME (VAUDREY (1977))
FIGURE 7
lOOO -• 1.5 ~" 1000
j 0 -
0 o. o.z O.J '·' v.;
APPARENT MODULUS OF ELASTICITY OF SEA
ICE AS A FUNCTION OF THE SQUARE ROOT OF
BRINE VOLUME (VAUDREY (1977))
FIGURE 8
· ~ :::.136 UPPER f/3 OF ICE CORE
~ /vb =.274 LOWER 1/3 OF ICE CORE
---~ = .136 UPPER 1/3 OF ICE CORE
--f))b = .274 LOWER 1/3 OF ICE CORE
Frum Figures 5 thruugh 8, the fulluwing mechanical pruperties have been
extracted utilizing the previuusly discussed data:
2-11
Mechanical
Property
Tensile Strength
Flexural Strength
Shear Strength
Apparent Elastic Modules
Upper 1/3
uf Ice Column
75-170 psi
105 psi
190 psi
5. 5 x 105 psi
Expected Ice Loading in the Nearshore Environment
Lower 1/3
uf Ice Column
40-120 psi
60 psi
90 psi
2. 5 x 105 psi
As was stated in the Introduction tu this paper, the present experiment
was applied tu a "real wur ld" design problem: in particular, the
calculated ice loads on a vertical wall fabricated near Oliktok Point.
Calulations utilizing the traditionally recognized ice property values had
already been performed fur this situation by applying the following
relationship (Ralston, 1979):
where:
F/H = Ice load per unit width of the structure,
fc = Contact factor between ice and structure,
IG, Ic = Indentation factors for granular and columnar ice,
G crc Unconfined crushing strength of granular and columnar crc, c =
ice,
hG, he = Thickness uf granular and columnar ice.
Since our field data indicated a columnar crystalline structure throughout
the ice column (with the exception uf the upper 0, 2 feet), we chose tu
eliminate the portion uf the relationship pertaining to granular ice.
2-12
Accepting the effective ice thickness theory, but still utilizing a
conservative value, an assigned thickness of 5 feet was applied.
All other parameters within this relationship were retained as if we were
studying a location outside the barrier islands. The resultant load
expected against a vertical structure was calculated to be:
F f I a h w = c c c
F 0.8 X 3 X 91 lbs. 5 ft. 1 kiE 144 sg.in ..
w = sq.in. X X 1000 lbs. X sq.ft.
F 157 kips w = ft.
This computed ice loading of 157 kips per unit width of structure is
significantly less than the values computed using "offshore" design
values.
Summary
Based upon the results of a field data collection program, revised values
for sea ice strength were developed for a specific location on the shores
of Alaska's Beaufort Sea. These values were significantly less than those
initially calculated. A downgrading of design specifications was thus
justified, resulting in substantial potential savings in construction
costs. Field investigations verified that sea ice strength can often be a
site-specific parameter; ice strength values applicable to the true
offshore environment are not necessarily valid for the more protected
nearshore regions.
2-13
The resulcs of che salinicy and brine void raciu analyses agree fairly
closely wich che published resulcs of ocher researchers. The primary
differ'ence becween Simpson Lagoon ice and che ice r'efer'enced by these
author's lies in the cryscalline str'uctur'e of the material. Due to its
protection by the barr'ier' islands, Simpson Lagoon ice is allowed to gr'uH
in a relatively stable, calm envir'onment. When insulated fr'um the
tur'bulent conditions offshor'e, crystalline ice dominates a major'ity of the
ice column. Ice of this configur'ation maincains a significantly differenc
str'engch fr'um che mor'e gr'anular ice which fur'ms under mor'e unstable,
cUr'bulent conditions.
Much of the published liter'atur'e tends to assign a single value to the
entire ice-thiclmess column. Based upon che r'esul ts uf t.his exper'iment.,
it is suggest.ed that the Simpson Laguun ice pr'ufile is sufficiently
str'atified to modify this simplified view. Fur' such situations, the
"effective ice t.hickness" concept provides a mor'e r'ealistic view of
nat.ur'al conditions.
This exer'cise was not undertaken to pr'uvide a rationalization fur' reducing
design specifications. Rather', it suggests that sea ice st.r'ength is a
mor'e site-specific phenomenon than is gener'ally thought. Cost benefits in
futur'e design/constr'UCt projects may be r'ealized by analyzing these
char'acteristics for' the specific site in quest.ion. When placed in the
context of potential development in t.he Alaskan and Canadian arctic, the
economic impact. uf these concepcs cannot be ignored.
2-14
Ref'erences
American Petruleum Institute (1982), Planning, Designing, and Cunstructing
Fixed Offshure Structures in Ice Envirunments: API Bul 2N. API,
Washingtun, D.C.
Cux, G.N.F., and 'ti.F. Weeks (1975), Brine Drainage and Initial
Entrapment in Sudium Chluride Ice: Research Repurt 345, U.s.
CRREL, December 1975.
Salt
Army
Dykins, J.E. (1971), Ice Engineering-Material Pruperties uf Saline Ice
fur a Limited Range uf Cunditiuns, Naval Civil Engineering Laburatury
TR-720, Purt Huenene, Califurnia,
Paige, R.A. and c.w. Lee (1967), Preliminary Studies uf Sea Ice in McMurdu
Suund, Antarctica! During "Deep Freeze 65," Juurnal uf Glaciulugy,
Vul. 6, Nu. 46.
Ralstun, T .D. ( 1979), Sea Ice Luads. In: Technical Seminar un Alaskan
Beaufurt Sea Gravel Island Design. Anchurage, Alaska/Huustun, Texas,
Octuber 18, 1979.
Vaudrey, K.D. (1977),
Sea Ice Fluating
Analyses,
Califurnia.
Civil
Ice Engineering -Study uf Related Pruperties uf
Sheets and Summary uf Elastic and Viscuelastic
Engineering Laburatury, TR860, Purt Huenene,
2-15
GROUNDWATER OCCURRENCE IN EAGLE RIVER, ALASIU\
WITH RECOMMENDATIONS FOR WATER MANAGERS
By James A. Munter 1
Abstract
Five distinct three-dimensional hydrogeologic terranes in Eagle River are
mapped. A near-surface alluvial fan aquifer encompasses much of the area
bounded by Meadow Creek, the Glenn Highway, and the Eagle River Loop Road.
A two square mile area is underlain by four confined aquifers arranged
tier-like (ascending from east to west) proceeding south and eastward for
three miles along the Eagle River Valley Road from the Glenn Highway.
A survey of 99 private domestic wells was done that included water-level
measurements in 91 wells. Comparison of 81 of these water-level
measurements with levels reported by drillers at the time of well
construction shows that water levels were higher in 1983 than were
initially reported by drillers in 89 percent of the wells. The average
water-level increase was 7.2 feet with a standard deviation of 5.8 feet.
The water level increase is attributed to several recent years of
above-average precipitation. An estimate 53 percent of water currently
used in Eagle River is pumped from the confined aquifer system, 30 percent
is withdrawn from shallow alluvial fans, and 17 percent is obtained from
bedrock or miscellaneous glacial deposits. The alluvial fan aquifer at the
present time is only lightly stressed. Many water-supply problems in Eagle
River are directly attributable to an inadequate water storage and
distribution system. Further development of groundwater in Eagle River
should focus on the near-surface alluvial fan aquifer, and should proceed
with caution in the confined aquifers system.
Introduction
With a rapidly increasing population and a scattered distribution of
productive aquifers, the community of Eagle River, Alaska, has always had
water-supply problems. Inquiries from developers, planners, water
managers, and the public regarding the location and extent of developable
groundwater, and the effects of existing and anticipated groundwater
1Hydrogeologist, Alaska Department of Natural Resources, Division of
Geological and Geophysical Surveys, Water Resources Section, P.O. Box
772116, Eagle River, Alaska 99577.
3-1
pumpage, have never been adequately answered. In view of escalating
pressures on existing resources, a detailed hydrogeologic examination of
the area was undertaken in the spring of 1982. Pending formal publication
of the study results, this paper discusses issues that are most pertinent
to current management concerns. Data not presented in this paper are
available for inspection at the offices of the Alaska Division of
Geological and Geophysical Surveys (DGGS) on Fish Hatchery Road in Eagle
River.
Hydrogeologic setting
The community of Eagle River is situated at the junction of the Chugach
Mountains, the valleys of Meadow Creek and Eagle River, and the glaciated
lowlands of Knik Arm of Cook Inlet. The close proximity of such diverse
geologic features has resulted in a correspondingly diverse assemblage of
aquifers. As a guide for evaluating groundwater conditions, a map of
hydrogeologic terranes has been prepared for the study area (fig. 1).
Hydrogeologic terranes are defined in this paper as three-dimensional
geologic units with distinctive water-bearing characteristics. Detailed
descriptions of the map units are presented in Table 1. A surficial
geologic map (Schmoll and other, 1971) and data from approximately 420
water-well drillers logs were used to construct the map. Drillers' logs
typically consist of well-construction information such as casing depths,
static water levels, and well yields, and drillers descriptions of geologic
materials encountered during drilling.
Test drilling was conducted at a central location in the study area (fig.
1). The results of the drilling and of the analysis of other well logs is
presented in two cross sections through the area (figs. 2 and 3).
3-2
~ Alluvial Fan Aquifers
~ Miscellaneous Glacial Deposits
~ Confined Aquifers
~ Bedrock-Kenai Group(sedimentary rocks)
ALASKA
STUDY AREA
19 ~ Bedrock-Metamorphic rocks(chiefly McHugh Complex)
Lin-es of cross sections
----·· ·· Boundaries of map units(dotted where concealed)
Well log data point used for cross sections
149°30'
Base from Municipality
of Anchorage
.... ?60 .... ,
APPROXIMATE MEAN
DECL!NATION, 1979
Figure 1. Hydrogeologic
explanation of
terranes in Eagle
map units).
River (see Table 1 for further
3-3
18'
Table 1. Description of map units used in figure 1.
Map Unit Description
Alluvial fan
aquifers
Miscellaneous
glacial
deposits
Confined
aquifers
Bedrock-
Kenai Group
Bedrock-
metamorphic
rocks
Quaternary-age near-surface unit consisting of gravel and
sand, with cobbles and boulders common, and minor silt.
Thickness ranges from a few feet to about 70 feet. Fan
sediment source areas probably include both adjacent bedrock
areas and Pleistocene glaciers in Knik Arm. Most
groundwater occurs under water-table conditions, although
the surficial fan deposits are underlain in some areas by
interbedded till and alluvial deposits, resulting in local
confined or semiconfined aquifers. Yields to wells range up
to 500 gpm. Unit boundaries are modified from Schmoll and
others (1971).
Quaternary-age unit includes: till; thin alluvium overlying
glacial deposits; glaciolacustrine deposits; colluvium
derived from glacial deposits or interbedded with glacial
deposits. Water-bearing sand and gravel deposits are
typically thin, shallow, and discontinuous. Reported well
yields are commonly not more than a few gallons per minute.
Some wells obtain water from the underlying bedrock.
Portions of map boundaries are modified from Schmoll and
others (1971).
Quaternary-age unit represents an area where wells may
penetrate one or more confined aquifers consisting of sand
and gravel with small to moderate amounts of silt. The
aquifers may be old, buried alluvial fans, or glacial
outwash, or both. Thickness of aquifers ranges from a few
feet to about 90 feet. Silty interbeds commonly occur in
thick aquifer sections. Aquifers are confined by silty till
and glaciolacustrine or glaciomarine deposits. Several
wells yield about 300 gpm.
Sedimentary rocks of the Kenai Group are within about 50
feet of the land surface. The Kenai Group rocks are of
Tertiary (Oligocene or Miocene) age (Wolfe and others,
1966). The Kenai Group rocks in the study area consist of
relatively flat-lying beds of siltstone, sandstone, and
coal. Lithification of the clastic rocks varies from very
friable to well cemented. Reported well yields in the study
area are typically a few gallons per minute or less.
Rocks of the McHugh Complex and another small, unnamed
formation (Zenone and others, 1974) are within about 50 feet
of the land surface. The McHugh Complex consists of
deformed and chaotically juxtaposed sequences of metaclastic
and metavolcanic rocks of Jurassic and/or Cretaceous age
(Clark, 1973). Nearly all wells obtain water from fracture
systems in bedrock. Reported well yields are a few gallons
per minute or less.
3-4
\.N
I
\Jl
South
A
Feet
1200
1000
BOO
Sea
Meadow Creek Alluvial Fan Aquifer
Miscellaneous Glacial Deposits
Confined Aquifers
Bedrock-Metamorphic rocks(McHugh Complex)
-----Geologic contact, dashed where approximate
"' i
Schematic groundwater flow lines
<t
ci z
~
5
M
60
30
0
Ft
200
100
1000 2000 Ft
0 300 600 M
Vertical Exaggeration= 6.25
N
I
"'
c
0
" u • "' .s
Figure 2. Hydrogeologic cross section A-A 1 (see Figure l for l'ocation of line of section) .
North
A'
FeeT
Sea Level
\.N
I a--
Feet
300
200
100
Ed
l:c:::l
t;;::d
~ ~
~
G
West
B
Alluvial Fan Aquifers
Miscellaneous Glacial Deposits
Confined Aquifers
Bedrock-Kenai Group(sedimentary rocks)
Bedrock-Metamorphic rocks( chiefly McHugh Complex)
Geologic contact, dashed where approximate
Potentiometric surface of upper aquifer
DGGS water-level measurement
" 0 ;:
u
~
.s
• " • >
M F'
30-~c:o
O~OOF<
a 3oo Goo M
Vertical Exaggeration = 12.5
" 0
" u • "' .s
Figure 3. Hydrogeologic cross section B-B'(see Figure l for location of line of section)
East
B'
Feet
BOO
600
500
400
300
200
100
Sea Level
Cross section A-A' shows the three thickest and most extensive confined
aquifers identified in the area. Detailed data on the middle aquifer at
the site of the DGGS test well and on the lower aquifer at the site of a
well (fig. 2, AWWU well A-1) drilled by the Anchorage Water and Wastewater
Utility (AWWU) shows that the aquifers, although relatively thick at these
locations, consist of alternating layers of silty glacial sediments and
relatively silt-free sands and gravels. The task of constructing an
efficient well in the confined aquifers is significantly complicated by the
depth below land surface (up to 450 feet) and the interlayered character of
the fine and coarse-grained sediments.
Water-level data show that the potentiometric surfaces for the three
lowermost aquifers are generally within 10 to 20 feet of each other, and
that groundwater gradients have a downward component of flow throughout
most of the area mapped. A small upward gradient exists from the lower
aquifer to the middle aquifer near well AWWU A-1, probably resulting from
heavy pumping from the middle aquifer, to be discussed subsequently. The
small gradients bet,<een the confined aquifers provide indirect evidence
that the confining beds between aquifers are leaky, to some extent, or that
the aquifers are physically connected by coarse-grained glacial sediments.
Aquifer test data by a local engineering firm (F. Damron, CH2M Hill,
written comm., 1983) indicates that pumping a well in the middle aquifer at
300 gpm for 24 hours results in a two-foot decline in water levels in the
upper aquifer. This provides direct evidence that the confining beds
separating the confined aquifers are leaky.
Cross section B-B' illustrates the fact that the potentiometric surface of
the upper aquifer is approximately 20 feet or less above the top of the
3-7
aquifer at some locations. Wells constructed in these areas have a low
tolerance for water-level declines, whether naturally occurring or
artificially induced. Ten to fifteen feet of water-level decline from
current levels could significantly reduce the ability of the wells to
provide water. Deepening of these wells should not be considered as a
routine solution to the problem because the thickness of the upper aquifers
is limited, and the extent of the deeper confined aquifers is questionable.
Indications are that similar conditions exist in the western part of the
study area where some wells drilled into the Ravenwood aquifer (fig. 3)
have less than 20 feet of available drawdown.
The alluvial fan aquifer includes both the surficial alluvial fan deposits
and older, buried alluvial sands and gravels that occur in the same area
(Cross section B-B', fig. 3). The alluvial fan aquifer is surrounded by
till and related glacial deposits in moraines and kames. The eastern
portion of the aquifer is underlain by silty glacial sediments and the
western portion is underlain by siltstones, sandstones, and lignites of the
Kenai Group. Deeper sands and gravels were encountered in the Sunny Slopes
well and the Lazy Mountain Trailer Court well (fig. 3). As a result of the
irregular bedrock topography in the area, and the irregular distribution of
glacial sediments, the deeper sands and gravels are not mappable with the
existing data base. The potential for water production from these
deposits, however, appears to be quite significant, probably because of
their close proximity to the surficial water-table aquifer.
Water use in Eagle River
Average continuous-supply rates of estimated and metered water-use data in
Eagle River for the period May -July, 1983, are presented in figure 4.
3-8
149°35'
T.14N.
0
I
0
6
3000 feet
I
I
900 meters
5
Creek
@
Sunny Slopes*
~gle River Height~ North**
('];) 4J'!Norfolk Utilities*
Eagle River Heights South @
Eagle Crest
EAGLE RIVER ROAD
14 17
WATER USE DATA
8
18
149°30'
Base from Municipality
Alluvial Fan Confined of Anchorage
0
@
~
~
@ 10-30 gpm
® 30-BOgpm
• 60-100gpm
100-200 gpm
*Water use is estimated from number of households served x 400 gpd.
**Breakdown between confined and alluvial fan sources is estimated from total production data.
Figure 4. Netered and estimated water use by major
community water systems in Eagle River.
3-9
Only water distribution systems delivering more than an average of 10
gallons per minute are included. Water-use data for systems for which
pumping records are not available was estimated by multiplying the number
of residential household consumers by 400 gallons per day per household.
Based on the short period of record, 400 gpd appears to be a reasonably
close approximation to annual average household water consumption in
residential areas of Eagle River.
The population of the Eagle River study area is rapidly changing and is
difficult to specify precisely. With dwelling unit counts for 1981 and
1982, and 1981 through 1985 population projections from the Municipality of
Anchorage (A. Van Domelen, written comm. 1983), a current (August, 1983)
population of approximately 9,000 people is estimated. This makes use of
the 1980 U.S. Census estimate of 3.2 persons per household in the Eagle
River area. Using a per capita use figure of 140 gallons per capita per
day (gpcd), the total average water usage in the study area is about 900
gpm. The figure of 140 gpcd is somewhat higher than is currently used in
the residential areas of Eagle River, but includes water used for
commercial purposes in the town center. The per capita use figure for the
AWWU in 1982 was estimated to be 140 gpcd ( Mar. 24, 1983 Memorandum from
J. Munter and L. Dearborn to State Rep. M. Szymanski).
The water-use data shown in figure 4 totals about 550 gpm, or about 60 per
cent of the total estimated usage. The remaining 350 gpm is withdrawn from
an estimated 1000 to 1500 private wells, which includes a few small
neighborhood water systems. Based on the distribution of known wells in
the study area and the location of major water-supply wells, an average of
approximately 53 percent of the estimated water consumption in the study
3-10
area, or 480 gpm, is currently withdrawn from the confined aquifer system.
An average of approximately 30 percent, or 270 gpm is currently withdrawn
from the alluvial fan aquifers. The remainder, or an average of
approximately 17 percent (150 gpm), is withdrawn from areas mapped as
shallow bedrock or miscellaneous glacial deposits. Considering only the
water pumped from the confined aquifer system, it is roughly estimated that
70 percent of the water is pumped from the middle aquifer, 30 percent from
the upper and Ravenwood aquifers, and 1 to 2 percent is pumped from the
lower aquifer.
Potential water supply
This analysis of water-supply potential from the alluvial fan aquifers
considers only the potential of several existing well systems. The
limiting factor in producing water at alluvial fan aquifer well sites in
Eagle River is drawdown in the individual production wells because:
1. The wells are typically shallow; with at most several tens of
feet of available drawdown, and
2. The drawdown from pumping generally does not affect other wells
because of the irregular aquifer geometry, the hydraulics of
water-table aquifers in general, and the relatively small number
of wells in the alluvial fan aquifer.
A test of the alluvial fan aquifer was performed in August, 1982, at the
Eagle River Heights North well field by DGGS and AWWU. The author
described the test results in a letter to the AWWU, in which he concluded
that a pumping rate of 200 to 300 gpm could be sustained by the aquifer for
an extended period of time. Unfortunately, severe well clogging or
encrustation prevents effective utilization of the resource. Based on
these results and existing water-use data, it appears that a conservative
3-11
estimate of aquifer yield at this site is about 200 gpm greater than the
current rate of extraction.
Another test of the alluvial fan aquifer was conducted during June, 1982,
at the Eagle Glenn well field (Munter and Dearborn, in press). Although
single-well pumping rates in excess of 500 gpm were attained, aquifer
boundaries limit long-term aquifer yield to rates lower than 500 gpm. Data
collected subsequent to the test and discussed in a memorandum from the
author to the Alaska Division of Land and Water Management (Aug. 25, 1983)
suggests that a conservative estimate for long-term aquifer yield at Eagle
Glenn under current conditions is 150 gpm, or approximately double the
current rate of withdrawal.
Two separate aquifer evaluations were conducted by local engineering firms
at the Sunny Slopes Water system: one in 1970 by Dickinson, Oswald and
Partners, Consulting Engineers, and the other in 1981 by Beyer Engineering.
They projected sustainable yields of 350 and 500 gpm, respectively, based
on low-rate (170-190 gpm) aquifer tests. The reported static water level
was one foot higher in 1981 than in 1970, despite continuous supply of the
Sunny Slopes subdivision in the interim. At this time, owing to the short
duration and low rate of previous aquifer tests, it appears that a
conservative estimate of long term aquifer yield at Sunny Slopes is about
250 gpm, or about 200 gpm in excess of current production (fig. 4).
Summarizing the preceding discussion, alluvial fan aquifers appear to be
capable of producing a minimum of 475 gpm in excess of current withdrawal
from existing well developments, for an estimated total yield of about 750
gpm. Additional water supplies from the alluvial fan aquifers could be
obtained by developing new well sites, constructing infiltration galleries,
3-12
or pumping existing wells at higher rates during wetter times of the year
(May through October). The ultimate potential of the aquifer, of course,
is unknown.
Methods of evaluating the potential yield of the confined aquifer system
are necessarily different than methods used to evaluate the alluvial fan
aquifers. The effects of large groundwater withdrawals from the confined
system are propagated relatively large distances because: 1) the confined
aquifers extend over larger areas, 2) the general hydraulic behavior of
confined aquifers favors widespread drawdown propagation, and 3) the
confining units separating individual aquifers are "leaky" enough to
transmit the effects of pumping vertically from one aquifer to another.
The large potential for propagation of drawdown forces consideration of the
impacts of development on other users of the aquifers, or more
specifically, prior water rights holders. The potential yield of the
confined aquifer system is dependent on the amount of water-level decline
that is deemed acceptable throughout the aquifers, rather than on drawdown
limitations of individual production wells.
To determine the present hydrologic status of the confined aquifers, a
water-level survey of 99 private domestic wells was conducted from February
through July, 1983. Water levels were successfully measured in 91 wells,
none of which were perforated, screened, or left with open ends in more
than one of the confined aquifers. Of these 91 water levels, it was
possible to compare 81 water levels with levels measured and reported by
drillers at the time of well construction. In 89 percent of these
comparisons, the water level measured during 1983 was higher. The average
water-level increase was 7.2 feet, with a standard deviation of 5.8 feet.
3-13
In general, the shallowest aquifers showed the most change, and the most
net increase in water level. Water-level data collected in the vicinity of
the Eagle River Loop Road indicate that the non-pumping water levels of the
middle aquifer may be about 10 feet lower than they were in the 1970's.
This is attributed to large rates of groundwater extraction from the middle
aquifer.
In determining the reason that water levels appeared to be higher in early
1983 than in the past, several factors were considered:
1. Uncertainty is commonly present as to whether the datum for a
water-level measurement from a driller is ground surface, or the
top of the well casing, which is commonly about 2 feet above the
ground.
2. Water-level measurements taken shortly after well drilling ma.y
not reflect static conditions because of drilling, developing,
and testing techniques commonly used by drillers.
3. The existing data set of drillers' logs was collected over a
period of many years and during all seasons of the year.
4. The accuracy of driller's water-level measuring techniques and
equipment used in the past are poorly documented.
5. Statistically the survey results may reflect an actual increase
in groundwater levels caused by an increase in the rate of
recharge subsequent to construction of most of the wells.
Figure 5 shows annual precipitation at Anchorage International Airport,
along with a histogram of the year of well construction of the 81 wells for
which comparisons were made. Although precipitation events in Eagle River
are somewhat different than in Anchorage, the annual trends are probably
similar. Figure 5b shows that most wells considered in this analysis were
drilled between 1975 and 1977, near the end of a long period of below
average precipitation in Anchorage. Figure 5a shows that the years 1979 to
1982 were abnormally wet in Anchorage.
3-14
2
0
i=
j:':
0:: u w
0:
"-
LL
0
(/)
w
I
(.)
2
..J
<(
::J
2
2
<(
(/)
..J
..J w s:
LL
0
0: w
CD :;;;
::J
2
0
YEAR
Fig. SA Annual precipitation at Anchorage Weather Service
Meteorological Office -Airport.
YEAR OF WELL CONSTRUCTION
Fig. 5B Histrogram showing year of well construction of wells for
which water levels were remeasured in 1983 and were compared
with levels reported by drillers.
3-15
Thus, the interpretation favored in this paper is that the statistical
results of the water-level survey reflect a real rise in water levels in
the confined aquifers caused by an increase in precipitation subsequent to
the drilling of most of the wells (option no. 5, listed above). An
accompanying interpretation is that natural fluctuations in precipitation
affect water levels much more so than does groundwater withdrawal from the
confined aquifer system at current rates of withdrawal. The exception to
this is in the local vicinity of the Eagle River Loop Road, where pumping
is heaviest.
As previously discussed, several wells in the upper aquifer may be
significantly impacted if water levels in the upper or Ravenwood aquifers
drop 10 to 15 feet from present levels. The data suggest that natural
fluctuations in water levels could be the most significant contributor to
such an occurrence. Thus, a major issue facing water managers considering
an expansion of existing water use from the confined aquifers is the degree
to which water levels in the confined aquifer system should be maintained
during droughts.
Under current conditions, approximately 10 feet of drawdown appears to have
occurred near existing pumping centers, with imperceptible effects
elsewhere in the aquifer system. Assuming that a simple, direct
proportionality exists between pumping rate and drawdown, a doubling of
existing withdrawal would result in approximately 20 feet of total drawdown
near existing pumping centers, and small drawdowns elsewhere in the aquifer
system. With available well construction data, it does not appear as
though this would, by itself, unduly affect existing users.
3-16
Until further data are compiled and analyzed, a doubling of current use
from an estimated 480 gpm to 960 gpm could be proposed as a reasonably
conservative lower limit of potential aquifer supply. This assumes that
future pumpage is areally distributed somewhat similarly to existing
pumpage. Therefore, the sum of the potential yield of the confined aquifer
system, the alluvial fan aquifers, and the current estimated pumpage from
miscellaneous glacial and bedrock sources is 1860 gpm.
Future water-demand and water-supply options
Table 2 presents population projections for the study area through 2005 (A.
Van Domelen, Municipality of Anchorage, written comm., 1983) and average
water-demand projections. Population projections were not extrapolated
beyond 2005 because the unrestricted zoning in most of the study area
precludes estimation of ultimate housing densities.
Table 2. Projections of populations and water demand in Eagle River
Projected 1
Estimated avzrage
water demand
Year population (gallons per minute)
1981 7,461 720
1982 7' 779 760
1983 9,251 900
1984 10,453 1000
1985 10' 855 llOO
1990 14,068 1400
1995 17,732 1700
2000 22,953 2200
2005 25,123 2400
1 Using transportation districts 46, 47, and 48, which encompass an area
slightly larger than the study area shown in figure 1.
2 Using 140 gallons per capita per day
3-17
A comparison of the average water-demand figures listed in Table 2 with the
estimates of minimum available groundwater resources presented previously
in this paper~ reveals that groundwater resources are adequate to serve as
a source of water at least until 1996. This conclusion assumes that an
adequate water-distribution system and water-storage facilities are
constructed, Providing water for daily or seasonal peak demands, fire
protection, or emergency service does not alter projections of average
water demand, which is based on an an annual time span. To meet short-term
peak demand, preliminary plans call for a five million gallon (mg) water
storage facility for Eagle River (R. Illian, AWWU, oral comm., 1983).
It is common belief that importation of water to Eagle River is necessary
in the near future because of the lack of adequate quantities of local
groundwater. Current plans by the Municipality of Anchorage under the
Eklutna Water Project call for construction of a pipeline from Anchorage to
Eagle River to deliver treated Ship Creek water to Eagle River as early as
1985 or 1986. Extension of the pipeline to Eklutna Lake and the
construction of a treatment plant is planned to provide for delivery of
water to both Eagle River (and neighboring communities) and Anchorage as
early as 1988 or 1989. This study indicates that a decision to import
water into Eagle River in the 1980's should be based on engineering,
economic, or other considerations, and not on the lack of available
groundwater in Eagle River.
Current water-supply problems
Several problems related to the water-supply situation in Eagle River have
not changed since they were first recognized and reported in 1977 (Quadra,
3-18
1977). For example, some existing wells are too small to efficiently
utilize available groundwater. Also, as previously mentioned, shallow
wells at Eagle River Heights North are clogged. A number of existing water
distribution systems, such as Sunny Slopes, have inadequately sized mains
for a fully integrated water distribution system (Quadra, 1977). Another
problem is that adequate storage does not exist in Eagle River for fire
protection or emergency water supply. A 0.5 mg storage facility currently
under construction near the AWWU well A-1 by the AWWU will be sufficient to
provide only for short term variations in residential water demand, but
will not be sufficient for fire protection (Bob Smith, AWWU, oral cornrn.
Aug., 1983). Although work is progressing on integrating several existing
water distribution systems by the AWWU, the task of obtaining a fully
integrated water distribution system of acceptable quality in Eagle River
should be viewed as a long-term project.
Conclusions
A system of alluvial fan and confined aquifers have been identified in
Eagle River that, under current conditions, are lightly stressed in most
areas. The impacts of further development of the alluvial fan aquifers are
anticipated to be minor. The impacts of large-scale development of the
confined aquifer system are likely to be significant and widespread,
particularly during multi-year periods of below-average precipitation.
A minimum, long-term, average rate of potential aquifer yield of 1860 gpm
(2.68 million gallons per day) of water is projected. With economic
incentives, additional quantities of water could be developed. Current
water-use data and population projections indicate that 1860 gpm is
sufficient to supply the water requirements of the study area until at
3-19
least 1996, provided adequate storage facilities and water-transmission
systems are constructed. Significant problems currently exist with clogged
wells, inadequate storage facilities, and substandard and unconnected water
distribution systems in the Eagle River area, causing certain areas to be
without adequate water service. A decision to import water into Eagle
River in the late 1980's should be based on engineering, economic, or other
considerations, and not on the lack of available groundwater in Eagle
River.
Recommendations for water managers
1. Adequate funding should be supported and efforts should be continued
to construct an adequate system of water storage and distribution in
Eagle River. This would include rehabilitation or redrilling of wells
and acquisition of private water-distribution systems by the AWWU.
2. Expanded development of groundwater should focus on further
utilization and development of the alluvial fan aquifer.
3. Protection against water shortages resulting from seasonal or
prolonged dry spells or periods of peak demand should be based on
wells in the two lowermost aquifers in the confined aquifer system.
Further development of the confined aquifers should be done with
caution.
4. Policy should be established concerning the maintenance of water
levels in the confined aquifer system. Consideration should be given
to both the costs to domestic well owners of over-development of
groundwater during periods of below-average precipitation, and the
costs to water utilities (and ultimately to the utility customers or
3-20
the public) of over-protection of the aquifer system during periods of
average or above-average precipitation.
Acknowledgements
The author would like to thank numerous employees of the Municipality of
Anchorage for providing several opportunities for data collection, for
allowing ready access to previously collected data, and for engaging in
rewarding dialogues concerning Eagle River's water supply. Assistance from
the engineering firms cited in the text is also appreciated. Significant
contributions were made by Roger Allely and Larry Dearborn of DGGS. Larry
Dearborn, Bill Long (DGGS), and Rick Illian (AWWU) reviewed the manuscript.
References
Clark, S.H.B., 1973, The McHugh Complex of South-central Alaska: U.S.
Geological Survey Bull. 1372-D, 11p.
Munter, J.A., and Dearborn, L.D., 1983, Evaluation of a shallow sand
and gravel aquifer at Eagle River, Alaska, in Short Notes on Alaskan
Geology, 1982: Alaska Division of Geological and Geophysical Surveys
Professional Report [in press].
Quadra Engineering, Inc., 1977, The Eagle River Community water supply and
distribution plan: Unpublished report prepared for the Alaska
Department of Environmental Conservation, 124 p.
Schmoll, H.R., Dobrovolny, E., and Zenone, C., 1971, Generalized geologic
map of the Eagle River-Birchwood Area, Greater Anchorage Area Borough,
Alaska: U.S. Geological Survey Open-file Map 71-248, 1 sheet.
Wolfe, J.A., Hopkins, D.M., and Leopofd, E.B., 1966, Tertiary stratigraphy
and paleobotany of the Cook Inlet region, Alaska: U.S. Geological
Survey Prof. Paper 398-A, 29 p.
Zenone, C., Schmoll, H.R., and Dobrovolny, E., 1974, Geology and ground
water for land use planning in the Eagle River Area, Alaska: U.S.
Geological Survey Open-File report 74-57, 25 p.
3-21
ARE WE READY TO MANAGE GROUNDWATER RESOURCES IN ALASKA?
by: Larry L. Dearborn 1
Abstract
The Alaska Department of Natural Resources is mandated the responsibility
of identifying the amount of ground water that is available for development
and of assuring, through a process of water appropriation, that aquifers
are not overpumped to the detriment of public interests. Increasingly
intense groundwater development is occurring in many Alaskan communities.
Safe appropriation of withdrawals is difficult in most communities because
of a lack adequate geohydrologic information for quantitative assessments.
Consequently, a four-step procedure was designed to allow recognition of
significantly low and significantly high groundwater-development
potentials. Geologic, groundwater-level, hydrologic boundary, and climatic
data are used to give a general characterization of the hydrologic
environment containing the supply aquifer. Inferences drawn from the
composited data may be further interpreted to provide guidelines ultimately
used to make management decisions. A status review of data availability
for six Alaskan communities, where keen competition for ground water seems
imminent, shows that geohydrologic data for the Anchorage Bowl, Eagle
River, and North Kenai are adequate for cursory-reconnaissance assessments.
But, data deficiencies for the Fairbanks uplands, Chugiak-Peters Creek
area, and the Mendenhall Peninsula-Auke Bay area are too great to make
useful resource assessments.
INTRODUCTION
The utilization of ground water in a few densely populated areas in Alaska
has increased to the point where resource management is essential. The
need to be able to arrive at an elementary understanding of aquifer
potential to benefit managers prior to completion of comprehensive
hydraulically-oriented studies is obvious. Hydrologists and water managers
should carefully consider what impacts new wells or future pumping schemes
have on prior appropriators and the hydrologic environment.
1Geohydrologist, Alaska Department of Natural Resources, Division of
Geological and Geophysical Surveys, Water Resources Section, P.O. Box
772116, Eagle River, Alaska 99577.
~1
Alaska's water-rights doctrine, which provides for the appropriated,
beneficial use of this public resource, is the basis for governing
withdrawals of ground water. In recent years state water-management
officers occasionally have not issued an appropriation that would have
certified a substantial increase in local pumping. In most instances, the
denial resulted from projections of intolerable drawdown, which presumably
would interfere with the yield of neighboring wells. In reality, an issue
of equal importance --the health of the entire aquifer --was not given
much, if any, consideration.
The purpose of this paper is to present a methodology that can be used to
identify the gross supply potential of aquifers in a specified area, and
thereby inform water managers and community planners early on about the
assets of the resource or the possibilities for conflicting use and
depletion of a low-yielding aquifer. Critical geohydrologic information
that must be compiled and analyzed to accomplish the identification of
aquifer potential are reviewed. Additionally, the status of geohydrologic
information categories for six major development areas in Alaska are
presented. The confined aquifer system in Eagle River described by Munter
(1983) is given as an example of the application of the methodology.
Other authors have addressed methodologies for estimating groundwater-
supply potentials where insufficient amounts of geohydrologic knowledge
exists to examine the potential quantitatively. Davis (1982) proposed a
"risk analysis procedure" using cumulative probability curves for making
preliminary assessments in areas that are undeveloped or are poorly
explored. Heath (1982) listed five "general geohydrologic criteria" as
characterizing groundwater systems, and then rated 14 groundwater regions
4-2
of the United States on these criteria. Peters (1972) summarized
California's "experience with using the hydrologic balance as a method of
determining safe yield and overdraft" by stating that "ground water is
primarily a storage resource". She presented numerous criteria for
designing water-level collection networks, which along with well-rounded
geologic data, were considered the main inputs in preparing an early
estimate of the groundwater-supply potential.
CENTRAL CONCEPTS IN MANAGEMENT OF GROUNDWATER WITHDRAWALS
The meanings of the terms "safe yield", "perennial yield", and "optimum
yield" have been widely debated by many authors, yet despite the lack of
universal acceptance of these or any yield expression, the basic concept
must be applied whenever the use of an aquifer is planned or managed
(Fetter, 1980). Although each expression has a slightly different connota-
tion, the intent, safeguarding against over-development of the resource, is
present in all definitions. In this report, 'maximum permissible pumping'
will be used, as the author believes it reflects a variable limit that may
be determined for any specified set of development conditions at any point
in time. Development conditions of importance include well density, well
distribution, depths of shallowest wells, pumping schedules, and location
of major centers of pumping in relation to aquifer boundaries.
The engineering, or planning, of groundwater withdrawals is of concern on a
local scale that addresses individual wells, and on a broad scale that
includes entire groundwater basins that may underlie thousands of square
miles. This paper focuses primarily on reconnaissance-level determinations
of the supply potential of discrete aquifers or of a single confined
aquifer system.
4-3
The general information requirements relating to groundwater management can
be grouped into three types of input, or knowledge, as follows:
1. Management objectives must be defined that reflect permissible
aquifer conditions for a specified period of time.
2. The pattern and nature of groundwater extractions (well
constructions, pumping schedules, etc.) are defined, or are
predictable.
3. The physical nature and behavior of the aquifer(s) are adequately
understood.
Management objectives
McCleskey (1971) lists the following general management objectives:
1. to limit withdrawals to a level such that the life of aquifer
utility is significantly extended;
2. to protect the basin from water-quality deterioration brought on
by man's activities;
3. to provide water at a minimum cost;
4. to avoid land subsidence resulting from excessive drawdown of
water levels.
Because of water-resource statutes, the most important management objective
in Alaska might be the protection of the water rights of prior appropri-
ators as resource development increases. This objective is not as well
defined as the above objectives, because each water-rights case commonly
requires unique technical and socioeconomic criteria for establishing the
degree of protection to be afforded prior appropriators.
4-4
Pattern of groundwater extraction
To achieve the maximum yield from an aquifer, optimization in the locations
and constructional features of production wells is required. Otherwise,
severe artificial limitations may not allow even half the supply potential
to be developed. However, rarely do the extraction facilities approach
optimization; therefore, the degree of deviation is important for the
resource analyst to consider. Such advantageous practices as locating the
heaviest withdrawal stress in recharge areas, distributing pumpage as
uniformly as possible among many wells versus a few wells, and screening of
only the lower section of aquifers need to be considered. Another means of
maximizing aquifer yield is to use surface-storage reservoirs to furnish
ground water for periods of peak water demand, thus promoting a more
constant rate of extraction from the aquifer.
CHARACTERIZATION OF AQUIFER POTENTIAL
The elements, or steps, of a simplified approach the author considers
necessary to characterize the production potential of an aquifer are shown
in figure 1. The term "cursory-reconnaissance assessment 11 (CRA) was chosen
as a result of combining Peters (1972) study levels A and B. Modification
of the definitions of her study levels gives an assessment scope that
encompasses a rough delineation of aquifer boundaries, aquifer thickness,
aquifer storage capacity, present status of storage, aquifer thru-flow,
annual recharge, and sources of recharge.
The objective of the final step (4) is to assign a qualitative, and perhaps
relative, rating to the maximum permissible pumping of an aquifer, such as
low, moderate, or high. An underlying assumption is that if a given
aquifer were thus characterized, groundwater managers can judge the
4-5
sensitivity of the aquifer to existing or imminent development pressures in
lieu of awaiting the outcome of many years of detailed hydrologic study.
STEP 1
collect and compile required data
(see Table 1)
STEP 2
draw hydrologic inferences for seven aquifer parameters
(see Table 2)
STEP 3
interpret favorability of five primary geohydrologic variables
(see Table 3)
STEP 4
assess combined favorability of variables
against existing development restrictions
Figure 1. Basic elements of a cursory-reconnaissance assessment of
aquifer-supply potential
Data requirements
The common observable manifestations of critical geohydrologic factors are
listed under three data categories in Table l. Step 1 in initiating a CRA,
is to note manifestations that apply to a given basin or aquifer, and
subjectively assigned a qualifier such as low, high, pronounced, subtle, or
some other suitable descriptor. The data must allow for a rudimentary
understanding of the influence of the basin-wide groundwater system on the
given aquifer and should be sufficient to recognize the nature of
hydrologic interactions between ground water and surface waters.
4-6
TABLE 1. Data categories, associated geohydrologic factors, and their
common manifestations.
Data category Geohydrologic factor
dominant structure
basin/aquifer rocks
aquifer stratigraphy
areal
geology
aquifer size/occurrence
aquifer thickness
aquifer water-levels
areal
hydraulics
system boundaries
climate
basin
physical features
environment
4-7
Observable manifestations
horizontal stratification
sloped/folded stratification
anticlinal or synclinal
jointed, fractured
unconsolidated sediments
mildly indurated
strongly indurated
crystalline
single layer
multiple layers
interbedded zone
massive (non-stratified)
areally small
areally extensive
network of stringers
locally patchy
relatively thin
individually thick
compositely thick
depth below ground surface
magnitude of seasonal
fluctuation
existence of multi-year decline
daily cyclic pattern
sensitivity to barometric changes
response to rainstorms
slope of major rises in level
slope of seasonal recession
response to glacier-melt runoff
proximity to streams
bordered by impermeable rock
bordered by faults
thickness and composition of
confining layers
overlain/underlain by aquifers
amount and intensity of rainfall
water equivalent of snowpack
average temperatures and ranges
wind and humidity
basin exposure (orientation)
topography
vegetation types
thermokarst terrain
The necessary data collection and compilation in each of the categories in
Table 1 to achieve the above objectives is usually straightforward.
Normally, well logs and geologic maps will be relied on to provide areal
geologic and system boundary data, however, some assessments might be made
with strong dependence on results from geophysical interpretations.
Aquifers for which considerable data are available may require several
months to accumulate, organize, and review enough information on the
spacial variations of subsurface geology. In regards to water-level data,
Peters (1972) suggests one year of data to define the annual cycle of
fluctuation for a cursory study level, and about five years for an areal
reconnaissance study. To satisfy the purposes of the proposed methodology,
the author believes one year will normally be adequate. This time frame
seems equally applicable to climatic data.
Hydrologic inferences
In step 2, seven aquifer parameters are depicted (Table 2) for which
inferences as to each's magnitude must be drawn from geohydrologic
information represented in Table 1. These parameters are similar to the
criteria that Heath (1982) identified. The need for inferences (a) through
(e) relate to determining the magnitude of aquifer thru-flow and
storativity --major technical components of the assessment step. The last
two inferences are important for broader assessment of the favorability of
aquifer replenishment. Characterization of recharge requires two
complimentary parameters as neither inference alone is sufficient, because
these parameters may represent contrasting favorability. An example is an
intermittent stream that is capable of losing large amounts of water to
shallow aquifers, but seldom does because its channel rarely receives
runoff.
TABLE 2. Common choices for generalized inferences regarding aquifer
parameters.
Aquifer Parameter
a) confinement of aquifer
b) typical hydraulic gradient
c) aquifer permeability
(hydraulic conductivity)
d) porosity of aquifer
e) aquifer extent and volume
f) continuity of recharge
Inference Descriptors
---unconfined, weakly confined,
tightly confined
----slight, moderate, steep, variable
----low, moderate, high
highly variable
----low, moderate, high, fractured,
solution openings
-- --small, medium, large
----seasonal pattern (areal
infiltration), sporadic/variable
(streambed perc.), vertical leak-
age (relatively constant)
g) availability of recharge water
(topographic, geographic, and
climatic considerations)
very limited, moderate, plentiful,
seasonally variable
In selecting the appropriate inference descriptor for the parameters named
in Table 2, choices other than the common ones shown exist and should be
considered. It is also feasible for two descriptors to apply to a single
parameter. Considering glacial outwash, for instance, the naquifer extent"
could be large but the "aquifer volume" may be small if the sands and
gravels are compositely thin.
Assessment of aquifer potential
The third step is to weigh the pertinent inferences for each aquifer, or
aquifer system, under consideration and interpret the net favorability of
each primary geohydrologic variable shown in Table 3.
4-9
TABLE 3. Primary geohydrologic variables governing aquifer yield and their
ranges of favorability.
Geohydrologic variable Favor ability range
a) ground water in storage small voluminous
b) aquifer thru-flow low voluminous
c) aquifer replenishment ineffectively slow rapid filling of
storage capacity
d) available drawdown slight great
e) boundary effects total restriction inexhaustible
of water movement recharge
A number of techniques described extensively in the literature may be
employed during in-depth groundwater studies to quantify the transmission
of water through an aquifer, and to quantify the volume of water stored
within an aquifer and its confining layers, if appropriate. The efforts
required are too data intensive and time-consuming to justify undertaking
for CRA's. Instead, the hydrologist can make judgement calls on the
impacts of all the geohydrologic variables by drawing both data and
inference comparisons with detailed study results for similar environments.
An analogous subjective procedure was recommended to develop estimates of
groundwater storage capacity (Davis, 1982). Certainly, the principles of
groundwater occurrence and hydraulic behavior should be applied to guide
judgement calls in the interpretation of favorability. A review of the
commonly-applied principles is outside the scope of this paper.
Final assessment
In step 4, an overall assessment is made by considering the favorability of
the geohydrologic variables in relation to the severity of limitations or
4-10
restrictions imposed by existing water-supply development and by other land
uses, The assessment should qualitatively identify the supply potential
with respect to maximum permissible pumping, but also may describe the
problems or limitations directly related to one or more management
objectives named earlier. The artificial limitations are:
1) shallowness of uppermost well openings {screens, perforations) in
relation to the depth to the bottom of the aquifer or confined
aquifer system;
2) large drawdown in individual wells relative to areal drawdown,
due to low specific capacity resulting from inefficient well
construction;
3) excessive drawdown in one part of the aquifer caused by
concentrated pumping from closely-spaced wells with pumping rates
that are maintained too high;
4) location of major withdrawals unnecessarily close to non-recharge
boundaries of the aquifer;
5) the placement of potential sources of pollutants, such as
landfills, within areas critical to aquifer recharge;
6) appropriations of water extractions that are hydraulically
inefficient or incompatible with one another; and,
7) social, economic, engineering, legal, or political restrictions
preventing siting wells at favorable geologic locations,
Further explanation is required to deal properly with drawdown. The
"available drawdown" variable, listed in Table 3, is that which exists
without consideration of depths of wells currently in use. That is, wells
are assumed to tap the lower third or half of the aquifer, and thus, the
available drawdown is the distance from the current water table {or
potentiometric surface) to just above the deepest practical pump setting.
Therefore, the available drawdown is the maximum possible permitted by the
physical limitations of the aquifer, and is not determined by existing well
constructions. In step 4 the available drawdown is compared with actual
4-11
well constructions for any point in time to determine the severity of this
particular limitation.
Because of the number of possible combinations of favorability ratings due
to seven geohydrologic variables and because of the necessity to maintain
flexibility in weighing their significance, a standard scheme to arrive at
the final assessment appears unworkable. Careful consideration of any
extremes in favorability of the variables should allow the analyst to
correctly categorize maximum permissible pumping as either low, moderate,
or high. In making the assessment, a low favorability rating for any one
of the five variables does not require that the overall assessed potential
be rated low. But, unfavorable ratings for two or more variables would in
most cases be cause to expect that less-than-abundant water supplies could
be developed from the aquifer in question.
DATA STATUS OF SELECTED ALASKAN COMMUNITIES
Full-fledged evaluations that will provide managers with desired control
over impacts caused by issuance of water-extraction rights require detailed
technical studies lasting many years. For example, in the Raymond Basin
northeast of Los Angeles, California (Mann, Jr., 1969), two 4-year studies
were separated by 7 years of unacceptable groundwater allocations.
Locally, the Anchorage aquifers have been investigated for three decades,
and yet groundwater assessments to guide Bowl-wide aquifer management are
not available.
The preceding pages outlined one methodology for gaining early insights of
managerial importance. Step 1 was described as collecting and assembling
certain data belonging to three basic categories. With these data,
4-12
appropriately collected both temporally and spacially, a cursory-
reconnaissance assessment of aquifer potential can be attained as outlined~
The intent of Table 4 is to indicate deficiencies of existing data for
selected Alaskan communities. An "A" status means that the data are
sufficient for a CRA. Assessments may also be possible, although not as
reliable, if "M" status data are used. The tabulation, which is based on
published reports and data on file at the USGS and the DGGS, shows that
geohydrologic information for the Anchorage Bowl, Eagle River, and North
Kenai is sufficient for CRA's. However, the Fairbanks uplands,
Chugiak-Peters Creek, and the Mendenhall Peninsula-Auke Bay areas have too
many data deficiencies at the present to make realistic assessments.
TABLE 4. Status of critical geohydrologic information for selected Alaskan
communities for initial groundwater-supply assessments.
COMMUNITY
GEOHYDROLOGIC Anchorage Chugiak-Eagle Fairbanks Mendenhall North
DATA CATEGORY Bowl Peters Cr. River Uplands Peninsula Kenai
areal geology A M A D-M M-D M-A
areal hydraulics A D M M-D D A-M
basin environment A M A M M M
A adequate M = marginally deficient D deficient in many respects
The communities listed in Table 4 are those for which important groundwater
management decisions are occurring or appear imminent. Each of these
rapidly developing areas have geohydrologic environments that are unique
from one another, except for some similarities between the Eagle River and
4-13
Chugiak-Peters Creek areas, which lie adjacent to each other. Three
communities have environments bounded by saline coastal waters. Geologic
settings and climatic characteristics differ widely among the areas listed,
causing groundwater occurrence and supply potential to vary markedly. As a
result, water managers should expect little transfer value of resource
experiences in one community to another. Therefore, a thorough evaluation
of favorable groundwater conditions versus natural and man-imposed
limitations to development is needed for each community to facilitate wise
management of its ground water.
APPLICATION OF THE CRA METHOD
To illustrate the procedure outlined in this paper, a CRA is performed
(below) for the system of confined aquifers underlying the community of
Eagle River. Data collection and compilation (step 1) have been
accomplished by Zenone and others (1974), Johnson (1979), and Munter
(1983). These works have provided sufficient data for all geohydrologic
factors listed in Table 1. The following manifestations of the aquifer
system have been identified {presented in descending order as in Table 1).
nearly horizontal stratification
-unconsolidated sediments; fine silty sand to clean sand and gravel
-multiple layers with interbedded water-yielding beds
-areally small (spanning about two square miles)
-compositely moderate system thickness
individual aquifers thin in places
-water levels 50 to 350 feet below ground surface
5 ft of seasonal fluctuation; 10 ft between wet and dry years
-water levels fluctuate some with barometer
-response to rainstorms much delayed
-water-level rises are gradual
-aquifers may extend to or under Eagle River (stream)
-metaclastic or metavolcanic rock bounds system upgradient
-layers of silty glacial sediments confine individual aquifers
-confined system is overlain by a shallow, discontinuous, unconfined
aquifer
-rainfall totals about 20 in. annually; few intense storms
4-14
-snowpack in recharge area contains small water equivalent
-temperature range: winter -20°to 30°F, summer 50 to 70°F
-frequent winds and generally low humidity
-southwest exposure of moderately-sloping undulating topography
The following hydrologic inferences are drawn from the geohydrologic
manifestations listed above, based on the format of Table 2:
a) Aquifer system is weakly to moderately confined.
b) The general hydraulic gradient is moderately low.
c) Aquifer permeability varies from low to high.
d) Aquifer porosity is probably moderately high.
e) Aquifer is relatively small in extent and volume.
f) Most recharge occurs largely from slow, fairly-constant,
downward, vertical leakage.
g) Moderate to high availability of water for vertical leakage
exists.
Next, the favorability of each geohydrologic variable listed in Table 3 is
interpreted from the above hydrologic inferences, with brief explanations,
as follows:
a) Relatively small volume of groundwater storage is due to small
volume of aquifer system, and the geologic evidence that the
porosity of the confining layers are not much different than that
of the aquifers. Thus, dewatering of the upper part of the
system will not result in large volumes of pumpage over the long
term.
b) The transmission of water through the aquifers is low to
moderate, because the hydraulic gradient is naturally low and the
aquifers contain appreciable quantities of silt.
c) Aquifer replenishment at relatively fixed rates of vertical
leakage occurs. Recharge probably cannot be induced to increase
substantially with increased pumping stress, due to low
permeability and 100 ft-plus thickness of confining layers.
d) The naturally occurring (and current) available drawdown is
slight; less than 20 ft above the top of the upper and confined
aquifer in places. However, the deepest aquifer has over 100 ft
of available drawdown. Thus, exclusive use of this aquifer would
result in a high favorability for this variable.
4-15
e) The movement of ground water across an upgradient boundary is
greatly restricted by the truncation of the aquifers by dense
bedrock. The aquifers may also abut bedrock in their discharge
areas near the stream of Eagle River. Inducing recharge from
Eagle River appears unlikely because the river is much lower in
elevation than the potentiometric surface.
In the final assessment of supply potential, the unfavorable inter-
pretations made for (a), (b), and (e) above indicate that large supplies of
water are not available from the confined aquifer system. In addition,
small available drawdowns of the upper confined aquifer, which is tapped by
many wells, present a potential for serious water-rights conflicts. The
combination of these circumstances leads to a CRA rating of moderately low.
Therefore, planners, water managers, and well developers should procede
with considerable caution when additional large withdrawals of water from
the confined system are desired.
CONCLUSIONS
Because water managers in Alaska are faced with allocations of water rights
early in the groundwater exploitation of an aquifer, or basin, a
rudimentary understanding of aquifer-supply potential and limitations is
crucial. Groundwater experts are in general agreement that a reliable
quantification of safe yield (rephrased as maximum permissible pumping in
this paper) is not possible in the early stages of pumping stress.
For these reasons, a fundamental methodology for providing a
cursory-reconnaissance assessment (CRA) of aquifers was presented to guide
resource management decisions before substantial pumping stress occurs, or
is analyzed. Using this methodology, aquifers in any stage of exploitation
can be rated; thus, a historical perspective of aquifer performance is not
4-16
a mandatory calibration requirement as it is for the quantitative
mathematical-modeling approach.
Data requirements of the four-step procedure were classified under general
categories labeled: areal geology, areal l1ydraulics, and basin
environment. The adequacy of existing data for six Alaskan con@unities
were subjectively rated. Data for the Anchorage Bowl, Eagle River, and
North Kenai are judged to be sufficient for CRA's, whereas data for
Chugiak-Peters Creek, the Fairbanks uplands, and the Mendenhall
Peninsula-Auke Bay area are not sufficient. Therefore, in answer to the
title question, we are not as ready as we should be to manage ground water
in our state.
An example CRA made for the confined aquifer system at Eagle River
indicated a moderately low development potential, due primarily to the
proximity of unfavorable boundaries and a small tolerance for drawdown. A
severe restriction to full development is imposed by numerous, certified
water-rights granted wells that tap the upper confined aquifer.
Although this assessment method was primarily designed to allow recognition
of significantly low and significantly high water-development potentials,
its application to some communities will result in a qualitative
determination that lies in between low and high. If so, the value to early
management decisions will be an awareness to proceed with caution.
Acknowledgements
I wish to thank Jim Munter, colleague with the Water Resources Section of
DGGS, for his thoughtful suggestions upon review of this paper.
4-17
References
Davis, G.H., 1982, Prospect risk analysis applied to ground-water reservoir
evaluation: Ground Water, v. 20, no. 6, p. 657-662.
Fetter, C.W., Jr., 1980, Applied hydrogeology: Columbus, Ohio, Charles E.
Merrill Co., 488 p.
Heath, R.C., 1982, Classification of ground-water systems of the United
States: Ground Water, v.20, no.4, p. 393-401.
Johnson, P.,
Alaska:
17 p.
1979, Hydrogeologic data for the Eagle River-Chugiak area,
U.S. Geological Survey Water Resources Investigation 79-59,
Mann, J.F., Jr.,
California:
61-74.
1969, Ground water management in the Raymond Basin,
Geol. Soc. Amer. Engr., geol. case studies, no. 7, p.
McCleskey, G.W., 1972, Problems and benefits in ground-water management:
Ground Water v. 10, no. 2, p. 2-5.
Munter, J.A., 1983, Groundwater occurrence in Eagle River, Alaska, with
management implications: Proceedings of the 1983 annual meeting of
Alaska Chapter of the American Water Resources Association, Chena Hot
Springs, Alaska (this volume).
Peters, H.J., 1972, Criteria for groundwater level data networks for
hydrologic and modeling purposes: Water Resour. Res., v. 8, no. 1, p.
194-200.
Zenone, C., Schmoll, H.R., and Dobrovolny, E., 1974, Geology and
ground water for land-use planning in the Eagle River-Chugiak area,
Alaska: U.S. Geological Survey Open-file Report 74-57, 25 p.
4-18
A WATER BALANCE FOR THO SUBARCTIC HATERSHEDS
By Robert E. Gieck Jr.*, Douglas L. Kane** and Jean Stein***
Abstract
The hydrology of the interior Alaska uplands near Fairbanks, Alaska is
not well known. Previous studies have focused on unpopulated
watersheds, such as the Caribou-Poker Creeks Research Hatershed and
Glenn Creek. Recent water-use conflicts in the Ester Dome area have
emphasized the need to study a populated area to help determine the
available water resources for industrial, residential, and natural uses.
The Ester Dome area was studied to obtain a water balance for the Ester
Creek and Happy Creek watersheds. Precipitation volumes were calculated
from isohyetal maps that were constructed from a network of rain gages.
Spring snowpacl< volumes were obtained from snow surveys at each 200-foot
elevation zone. Evaporation during the snowmelt period was calculated
using evaporation pan data. Runoff quantities were determined by stream
gaging and by utilizing data collected at runoff plots. Summer
evapotranspiration was estimated from evaporation pan data.
Precipitation distribution, total annual precipitation, streamflow, and
evapotranspiration were calculated from these data. Groundwater
recharge for the Ester Creek and Happy Creek watersheds was calculated
by solving for the groundwater term in the water balance equation.
Introduction
The Ester Dome area west of Fairbanks, Alaska, is an area which has
grown rapidly in recent years. ~1ining has historically played an
important role in the area's economy and supported a small community.
Recently the area has become a popular location for residential
development and has also been identified as a potential site for
industrial development. As this development occurs, use of the limited
groundwater resources in the area may exceed the recharge
capacity of the watershed, especially at higher elevations.
* Graduate Research Assistant, Institute of Hater Resources,
University of Alaska, Fairbanks, Alaska 99701
** Associ ate Professor, Institute of Hater Resources, University of
Alaska, Fairbanks, Alaska 99701
*** Presently, Assistant Professor, Laval University, School of
Forestry, Quebec City, Quebec, Canada G1K 7P4
5-1
To determine the groundwater recharge potential of the area a study was
carried out during the 1981-1982 and 1982-1983 water years. The Ester
Dome area was instrumented to provide data for a water balance in the
Ester Creek and Happy Creek watersheds. Precipitation, soil moisture,
runoff, and evaporation data were collected to estimate groundwater
recharge by solving for the groundwater term of the water balance
equation.
This paper is the first attempt at a simple water balance using data
that were just collected. A more detailed analysis will be completed
after the latest water year's data has been tabulated and reduced.
Related Research
Numerous watershed studies have been carried out in the Fairbanks area.
Glenn Creek, a second order stream located 7 miles north of Fairbanks
and 13 miles northeast of Ester Dome, has been the site of several
related hydrologic studies. The Glenn Creek watershed area is 0.70
square miles. Dingman (1971) estimated a water balance for the stream
based on precipitation and runoff data. Kane et al. (1981) studied
snowmelt runoff generation using lysimeters. They determined
premafrost-free areas of the watershed contributed little to the
snowmelt runoff, while areas with permafrost contributed most to surface
runoff. A study by Eaton and Wendler (1982) of the snowmelt process was
done using an energy balance. They determined 78% of the solar energy
was used to evaporate water from the melting snowpack. Chacho and
Bredthauer (1983) studied the precipitation-runoff ratios of Glenn Creek
and found the watershed had a very fast response time with long
recessions and subsurface runoff before overlying organic soils were
5-2
saturated. They also found little runoff was generated from
nonpermafrost areas in the watershed.
In Goldstream Valley to the northeast, Slaughter and Kane (1973) studied
the hydrology of a small lake in permafrost terrain. They determined
the lake was recharged, at least in part, from below the lake by
subpermafrost groundwater. Nearby, Hartman and Carlson (1973) studied a
small thaw lake and determined it was isolated from groundwater recharge
by permafrost. Stein and Kane (1983) studied groundwater recharge using
runoff plots and infiltrometers. They concluded the spring snowmelt
season was the principle period for groundwater recharge.
The Caribou-Poker Creeks Research Watershed has been the site of several
studies including vegetation (Vogel and Slaughter 1972),
precipitation-runoff characteristics (Ford 1973), geology (Kautz and
Slaughter 1974), drainage network analysis (Bredthauer and Hoch 1979),
and hydrology and climatology (Haugen et al 1982). Lotspeich and
Slaughter (1981) studied the watershed on an ecosystem basis, relating
the biotic, climatic, and physical aspects of the watershed.
In the Ester area, several hydrologic studies have been done. Barsdate
(1967) studied the limnology of Ace lake, near the Lasonsky site in this
paper, to determine the concentration of trace elements and aeration of
the stratified waters in the lake. Hawkins et al. (1982) studied the
geohydrology of the Ester area in terms of the source and distribution
of dissolved arsenic in the area's waters. Presently, personnel at the
Institute of Water Resources, University of Alaska, are carrying out a
groundwater geochemisty study of the Ester Dome area.
5-3
Study Area
Ester Dome is located within the Yukon-Tanana Upland of interior Alaska,
at latitude 64° 53' north and longitude 148° 02' west. The area lies
north of the George Parks Highway and approximately seven miles west of
Fairbanks within the Fairbanks Mining District. While much of the area
remains undeveloped, several sections have been extensively modified by
human activities.
Initial development of the area occurred during the early part of this
century with the discovery and subsequent placer and lode mining of
gold. Mining has continued to the present time, but the level of
activity has varied. More recently, the neighboring communities of
College and Fairbanks have grown, and the area has become a popular
location for residential development.
Ester Creek and Happy Creek watersheds lie on the south and east slopes
of Ester Dome. The residential development and much of the mining
development are contained in these two watersheds. The Ester Creek
watershed has an area of 10.13 square miles and is adjacent to the 9.65
square mile Happy Creek watershed (see Figures 1 and 2). Both drainage
areas were calculated above the location of our stream gaging stations.
A large portion of the valley bottom in the Ester Creek watershed is
composed of mine tailings. Residential areas lie along Old Nenana Road,
in the town of Ester, and on south-facing slopes above Ester. There is
little evidence of placer mining in the Happy Creek watershed, but there
are several lode mines in the drainage. The residentially developed
areas in the Happy Creek watershed lie along the Parks Highway, Old
Nenana Road, Henderson Road, Sheep Creek Road, and St. Patrick Creek
5-4
\J1
I
\J1
ANNUAL PRECIPITATION OCT81-SEPT82
MlLES
Figure 1. Isohyetal map of Ester Dome area (Oct. 1981-Sept. 1982).
\Jl
I
0"
ANNUAL PRECIPITATION OCT82-16 SEPT 83
/
/
?
J
,, ..
a ro
MILES
Figure 2. Isohyetal map of Ester Dome area (IOct. 1982-16 Sept. 1983).
N
Road. There are currently 78 wells in the Happy Creek watershed and 57
wells in the Ester Creek watershed according to the records of the U.S.
Geological Survey. Many more unrecorded wells may exist.
Vegetation The vegetation of the area is typical of the Yukon-Tanana
Uplands. The well-drained south-facing slopes support forests of Paper
Birch (Betula papyrifera), Hhite Spruce (Picea glauca), and Quaking
Aspen (Populus tremuloides). The relatively sparse undergrowth consists
of shrubs and forbs. The valley bottoms have shallow slopes of poorly
drained soils, and the north-facing slopes support forests of Black
Spruce (Picea mariana), with occasional Paper Birch, Green Alder (Alnus
cripus), willow (Salix spp.), and Larch (Larix Larecena). The
undergrowth consists of a thick mat of mosses, lichens, tussock grasses,
and shrubs.
Geology According to Forbes (1983), the bedrock of Ester Dome is
primarily crystalline schists of the Yukon-Tanana metamorphic complex.
These metamorphic formations are highly folded and jointed. The schists
and associated quartzite lenses have experienced at least three folding
events. Ester Dome is at the southwest end of a mineralized belt known
as the Fairbanks Mining District. The primary economic phases {gold,
lead and zinc sulfides) occur within the mineralized quartzite lenses.
Soils The mineral soils of the area are composed of layers of micaceous
loess originating from the glacial outwash plains of the Tanana Valley
to the south. Soil thickness varies from a few inches to over 180 feet.
Typically, the organic soils of the well-drained, south-facing slopes
are well developed (four-inches thick) and overlain by two inches of
organic litter. The poorly drained soils of the north-facing or shallow
slopes have up to 12 inches of living moss atop 10 inches of slightly
5-7
decomposed moss and roots. Large areas of mine tailing consisting of
coarse rock fragments remain in Ester Creek valley {USDA Soil
Conservation Service 1963).
Permafrost Permafrost is absent on the steep, south-facing slopes.
Discontinuous permafrost is encountered over much of the valley floor
and on north-facing slopes. Well logs of the area show permafrost
begins as shallow as two feet and extends to depths beyond 150 feet.
Massive ice occurs in the area. Ice wedges were found beneath fields
north of the University of Alaska's Agricultural Experiment Station near
Smith Lake, and homes along Henderson Road have been damaged due to
uneven settling {Pewe 1982).
Climate Ester Dome lies within an area of continental climate
characterized by warm summers and winters of severe cold. The extremes
of seasonal variation are illustrated by a record high temperature of
99°F and a record low of -66°F which were recorded at the Fairbanks
International Airport several miles south of Ester Dome. The average
annual temperature at the airport is 26.1°F with an average
precipitation of 11.7 inches and an average annual snowfall of 66.6
inches. The duration of daylight also varies seasonally due to the high
latitude of the area. There are nearly 22 hours of daylight during the
summer solstice and less than four hours of daylight during the winter
solstice.
Methods
The water balance equation was used to estimate the groundwater recharge
of each basin:
P = R + E + S
5-8
Solving for S,
s = p -E - R
where
s = change in groundwater storage
p = precipitation
E = evapotranspiration
R = runoff.
The evapotranspiration term includes interception losses. The runoff
term represents overland flow, interflow, and baseflow.
During the 1981-1982 water year, eight sites were established in the
Ester Dome area to collect precipitation data, data were also obtained
for three additional sites from the National Heather Service. These
sites were the Agricultural Experiment Station (1), Coutts (2), Ester
Dome Road (3), Gedney (4), International Airport (5), Rice (6), Stone
(7), Swainbank (8), College Observatory (9), Ester Dome Observatory
(10), and Nugget Creek (13) (Figure 1).
The following year four additional sites were added: Quartz (11),
Lasonsky (12), \liillow Creek (14), and St. Patrick Creek Road (15)
(Figure 2). The College Observatory site was not included in the 1983
calculations. Summer precipitation data were obtained using standard
eight-inch dipstick and tipping bucket raingages (Table 1).
Snowpack water equivalents were determined at snow courses designated by
elevation along the existing road network. Snowpack water equivalents
were obtained using an Adirondak sampler and averaging the amount
obtained in eight to ten trails at each course. Snowmelt volumes were
5-9
TABLE 1. ESTER Dot~E SUMMER PRECIPITATION (in. )
1982
Site May June July Aug. Sept Total
1. Ag. Stat. 0.68 2.26 3.87 1.83 0.62 9.26
2. Coutts 0.94 2.49 5.05 2.48 1. 07 12.03
3. E.D. Road 0.67 2.31 3.03 1.77 0. 73 8.51
4. Gedney 0.68 2.25 3.02 1. 90 0.78 8.63
5. Int. Arpt. 0.96 1. 96 2.33 1.67 0. 77 7.69
6. Rice 0.69 2.49 3.11 1. 71 0.76 8.76
7. Stone 0. 71 2.13 3.38 1. 97 0.87 9.06
8. Swa i nbank 0.84 2.48 4.16 2.37 0.88 10.73
9. Co ll. Obs. 0.80 2.19 3.97 1. 78 0.82 9.56
1983
Site May June July Aug. Sept. 01-16 Total
1. Ag. Stat. 0.09 1.11 0.87 2.87 0.32 5.26
2. Coutts 0.30 1. 25 1. 64 5.18 0.86 9.23
3. E. D. Road 0.40 1.11 0.89 6.04 1.08 8.44
4. Gedney 0.24 0.84 1.10 4.68 0.59 7.45
5. Int. Arpt. 0.14 0.57 1.71 3.33 0. 71 6.46
6. Rice 0.34 1.10 0.73 4.05 0.48 6.70
7. Stone 0.29 1. 34 1. 02 5.24 0.76 8.65
8. Swainbank 0.48 1. 38 1. 31 6.44 1.17 10.78
10. E.D. Obs. 0.48 1.36 1.30 6.39 1.13 10.66
11. Quartz 0.33 1.15 1.40 6.64 0.92 10.44
12. Lasonsky 0.24 1. 07 0.90 3.73 0.54 6.48
13. Nugget Crk 0.30 0.92 1.00 5.61 1.01 8.84
14. Hillow Crk 0.30 1.16 1. 41 5. 77 0.98 9.62
15. St. Pat. 0.30 1. 35 0. 77 4.34 0.55 7.31
obtained by applying an average water equivalent to the area of each
200-foot elevation zone (Table 2).
Annual precipitation was obtained by adding the summer precipitation to
the average snowpack water equivalent at each site. Precipitation
volumes (Table 3) were obtained by plotting isohyetal maps (Figures 1
and 2) and applying the mean of the isohyets to each area within each
watershed between the isohyets (Table 3). Reported areas were the mean
of three calculations using a 9847A Hewlett Packard digitizer.
5-10
TABLE 2. AVERAGE SNOW HATER EQUIVALENT BY ELEVATION.
Water Eq. (in ) ~latershed Area ( s q mi )
Elev. Zone (ft)
4/22/82 4/06/83 Ester Happy
500-700 3.76 5.21 0.80 5.78
700-900 3.44 5.07 1.45 1.64
900-1100 3.83 5.29 1.72 1.04
1100-1300 3.87 5.64 l. 79 0.64
1300-1500 4.24 5.80 1. 87 0.27
1500-1700 4.73 6.66 1.00 0.21
1700-1900 5.51 7.84 0.62 0.06
1900-2100 5.72 7.94 0.52 0.01
2100-2364 5.94 9.35 0.36 0.00
10. 13 9.65
TABLE 3. TOTAL PRECIPITATON DURING THE WATER YEAR
SITE 1981 -1982 1982 -1983*
l. Ag. Station 13.02 10.47
2. Coutts 16.76 15.89
3. E. D. Road 12.75 14.24
4. Gedney 12.07 12.52
5. Int. Arpt. 11.89 10.41
6. Rice 12.52 11.91
7. Stone 12.93 14.45
8. Swainbank 16.45 18.72
9. Co ll. Obs. 13.32
10. E.D. Obs. 20.01
11. Quartz 15.51
12. Lasonsky 11.69
13. Nugget Crk 14.48
14. Hillow Crk 17.56
15. St. Pat 12.38
* Ending September 16th, 1983.
Stream gaging stations were established on Happy Creek where it crosses
the Old Nenana Road and on Ester Creek approximately one mile downstream
of the George Parks Highway crossing. Discharge measurements were taken
using Gurley and pygmy current meters. Runoff volumes were obtained by
multiplying the average discharge for each time period by the length of
the time period. Additional snowmelt runoff data were collected at two
5-11
BOO-square-foot runoff plots in Goldstream Valley approximately five
miles northeast of Ester Dome.
Soil moisture data were collected at the Stone (7) and Gedney (4) sites.
Soil moisture data were obtained with tensiometers (Table 4) during the
summer of 1982 and by time domain reflectometry during the summer of
1983. Due to the unavailability of an instrument, soil moistures were
only obtained for April and May of 1983.
We monitored three unpumped wells on Ester Dome: the Swainbank site
(8), and the upper and lower wells at the St. ,Joseph American Hine on
Henderson Road. The depth to the piezometric surface in the wells was
measured using an acoustic well probe and by a well tape. The wells
were monitored during the 1981-1982 water year by Northern Testing
Laboratory for the Alaska Department of Natural Resources.
Evaporation data were collected at the the Ester Dome Observatory (10)
using a standard four-foot-diameter by ten-inch-deep evaporation pan.
Additional evaporation data were obtained from the Agricultural
Experiment Station (1).
Results and Discussion
lve divided our study into two distinct intervals, the spring snowmelt
period (mid-April to mid-May) and the summer-fall period (mid-May to
October 1). Conditions during these periods were substantially
different. Winter precipitation was temporarily stored as snow, so it
did not affect the water balance until spring snowmelt. The spring
snowmelt period is characterized by low evaporation rates, low
interception, and high soil moisture. The summer-fall period has high
rates of evapotranspiration and interception, and low soil moisture.
5-12
The interception losses are lower during the spring snowmelt period
because the deciduous canopy is open, lacking its leaves.
TABLE 4. 1982 SOIL MOISTURE
CENTIBARS OF SUCTION
GEDNEY STONE
DATE 20 em* 40 em* 60 em* 20 em* 40 em* 60 em*
6/25 22.0 36.0 41.5
6/28 18.0 15.0 29.0
7/02 39.0 42.0 38.0 22.0 28.0 30.0
7/06 47.0 46.0 49.5 25.0 22.0 35.0
7/13 59.0 52.0 54.0 31.0 28.5 40.0
7/20 69.0 62.0 59.5 32.0 30.0 45.0
7/27 70.0 66.5 64.5 24.0 26.0 47.0
8/03 59.5 70.0 69.5 26.0 24.0 47.5
8/06 55.0 68.0 65.0 34.5 32.0 51.0
8/09 63.0 72.0 64.5
8/10 66.5 73.5 66.5 42.5 40.5 55.5
8/17 72.0 74.5 57.5 39.0 46.0 59.0
8/24 74.0 76.0 70 + 47.5 48.5 62.5
8/31 74.0 74.5 70 + 40.0 58.0 66.5
* depth of tensiometer's porous cup
5-13
Spring Snowmelt Period The 1981-1982 maximum snow pack volume for the
Ester Creek watershed was 2,330 acre-feet, and the average snow water
equivalent for the watershed was 4.31 inches. The snowpack volume for
the Happy Creek watershed was 1,940 acre-feet, and the average snow
water equivalent was 3.77 inches. The runoff in Ester Creek and Happy
Creek was measured once in 1982 to establish the baseflow.
The Ester Creek snowpack during the 1983 period was equivalent to 3,230
acre feet of water. The average snow water equivalent for the watershed
was 6.00 inches. Evaporation following the snow survey and during the
time of actual snowmelt to May 12 was 1.16 inches (626 acre-feet). The
runoff volume for the snowmelt period, including the falling limb of the
hydrograph, was 818 acre-feet. By placing these figures in the water
balance equation, we estimated that recharge for the period was 1,790
acre feet. Snowmelt was distributed as 55.3% potential groundwater
recharge, 25.3% runoff, and 19.4% as evaporation.
The Happy Creek snowmelt during the period was 2,720 acre-feet. The
average snow water equivalent for the watershed was 5.28 inches. The
evaporation losses were 626 acre-feet. The runoff volume was 994
acre-feet. The estimated potential groundwater recharge was 1,100
acre-feet. The melt was distributed as 40.5% groundwater recharge,
36.5% runoff, and 30.0% evaporation.
The two runoff plots that we monitored extensively during the snowmelt
period are used for comparison. For 1983, the snowpack on the control
plot was equivalent to 385 cubic feet of water. Runoff was 125 cubic
feet (Figure 3), and evaporation was 77.3 cubic feet. The estimated
groundwater recharge was 183 cubic feet. The control plot snowmelt was
distributed as 47.5% potential groundwater recharge, 32.4% runoff, and
5-14
20.1% evaporation. The snowpack on the east plot was equivalent to 323
cubic feet of water. Runoff was 78.7 cubic feet (Figure 3), and
evaporation was 77.3 cubic feet. The estimated groundwater recharge was
171 cubic feet. Snowmelt from the east plot was distributed as 53.3%
potential groundwater recharge, 24.1% runoff, and 23.6% evaporation.
The snowmelt ended on April 28 at the east plot and on April 29 at the
control plot. Peak runoff occurred on April 23 at both plots
(Figure 4).
The lower groundwater recharge estimate and higher runoff percentage in
the Happy Creek basin may be a result of reduced infiltration rates in
the watershed. The Happy Creek watershed is dominated by relatively low
terrain compared to Ester Creek, and contains more permafrost which
could reduce the infiltration capacity. The Ester Creek watershed is
dominated by relatively steep south-facing slopes and mine tailings
which are well drained and have little or no permafrost.
Note that this water balance represents a single year of data, and soil
moisture conditions vary from year to year. The soils were dry in the
fall of 1982. Had the soils been wet when they froze, the infiltration
rates could have been much lower for the nonpermafrost soils, and the
groundwater recharge could have been much less.
Summer-fall Period The summer-fall period is one of showery light
precipitation and high potential evapotranspiration. Significant
groundwater recharge only occurs 1vhen the surface soi 1 s have moderately
high levels of soil moisture. Evapotranspiration by plants uses much of
the soil water. Groundwater does move in the vadose (unsaturated) zone
of soils along pressure gradients, but the volumes are small compared to
saturated flow. Following a precipitation event of short duration, any
5-15
(f)
lLJ :r: u
2
u_-
lL
0
2 ;::, n:
lLJ >
1-
<[
...J ;::,
::';;
;::, u
2 0
I 80
I 60
I 40
I 20
I 0
0 80
0 60
0 40
0 20
EAST OF EAST PLOT
18 19 20 21 22 23 24 25 26 27 28 29 30
APRIL-1983
Figure 3. Cumulative runoff for the two runoff plots.
5-16
1\ I,
i
0.50
0.45
0.40
0.35
0.30
"-"-0
2
::) 0.25 ct:
"-0
(f)
w 0.20 :I: u
2
0.15
0.10
0.05
0
18 20 21 22 23 24 25
APRIL, 1983
CONTROL PLOT -
EAST OF EAST f!1Zl!3
PLOT
26 27 28 29
Figure 4. Daily runoff increments for the two runoff plots.
5-17
water infiltrating into the soil is quickly lost to the atmosphere
(unless it migrates past the rooting zone). The soil moisture data tend
to indicate that most soil water is lost by evapotranspiration (Table
4). Soil moisture levels decreased throughout the summer of 1982 as
indicated by increasing soil tensions.
The pan evaporation was higher in 1983 than in 1982 at the Agricultural
Experiment Station (Figure 5). The total pan evaporation for the period
of June 1 to September 15 was 12.03 inches for 1982 and 14.64 inches for
1983. Precipitation in 1982 was more evenly distributed. Average
ambient air temperatures in June and July of 1983 were above normal.
The monthly average ambient temperatures for Fairbanks, as reported by
the National Weather Service, in June and July of 1983 were 3.0°F and
2.7°F above normal respectively. During the summer of 1982, the monthly
averages were 0.5°F below normal for June and 1.2°F above normal for
July. During both years, less pan evaporation was recorded at the Ester
Dome site (Figure 6). From July 8 to September 14, 1982, 5.55 inches
were recorded at Ester Dome versus 6.09 inches at the Agricultural
Experiment Station. From June 7 to September 15, 1983, 10.41 inches
evaporation were recorded at Ester Dome versus 13.61 inches at the
Agricultural Experiment Station. For these time periods, Ester Dome pan
evaporation was 8.9% lower in 1982 and 26.0% lower in 1983.
vlell data (Figures 7 and 8) showed no large responses to precipitation
events, with the exception of the S1~ainbank well. This site showed a
response to precipitation in late August 1981 and 1983; no data were
available for August 1982. Evaporation rates were also lower during
August of both years (Figures 5 and 6). The precipitation from August
20 to September 1, 1983, was 5.15 inches at the Swainbank site. One
5-18
<f) w r u
2
"--u w a:
"-
a:
0
"-q > w
<f)
w
J: u
'=
"-
u w a:
"-
a:
0
"-q > w
AGRICULTURAL EXPERIMENT STATION
25
24
23
22
21
20
19
18
17
16
15
14 CUMULATIVE EVAPORATION
13
12
/I \
10
9
8
7
6 \
5
4 CUMULATIVE PRECIPITATION
3
2
I
0
0 10 20 10 20 10 20
JULY AUGUST SEPTEMBER
1982
AGRICULTURAL EXPERIMENT STATION
25
24
23
22
21
20
19
18
17 CUMULATIVE
16
15
14
13
12
II
10
9
8
7
6
5
4 CUMULATIVE
3
2
I
0
10 20 10 20
APRIL MAY
10 20
JUNE JULY
1983
10 20
AUGUST SEPTEMBER
Figure 5. Cummulative evaporation and precipitation at the
Agricultural Experiment Station (1982 and 1983).
5-19
Ul w :r u
2
a:
[cl a: a.
a:
0
a.
"' > w
Ul w
I u
2
a. u w a: a.
a:
0
a.
§
w
25
24
23
22
21
20
19
18
17
16
15
14
13
12
I I
10
9
8
7
6
5
4
3
2
I
0
0 10 20
JUNE
25
24
23
22
21
20
19
18
17
16
15
14
13
12
I I
10
9
8
7
6
5
4
3
2
I
0
10 20
ESTER DOME OBSERVATORY
1982
10
CUMULATIVE PRECIPITATION ; _ __..--
\
CUMULATIVE EVAPORATION
20 10 20
AUGUST SEP~EMBER
ESTER DOME OBSERVATORY
CUMULATIVE EVAPORATION
\
CUMULATIVE
"-PRECIPITATION
APRIL MAY
10 20
JULY
10 20
AUGUST
10 20
SEPTEMBER
1983
Figure 6. Cummulative evaporation and precipitation at Ester
Dome Obsrvatory (1982-16 September 1983).
5-20
Vl
I
1\J
20
40
6c
f--80 "-
w u 100 Lt a:
::0
(J) 12(
a: w
f--140 ~
0
f--/
I 310
f--a. 320 w ~ 0 330
340
350
360
370
JUN
6
SWAINBANK'S WELL .........
ST. JOE LOWER (#5) ~
ST JOE UPPER (#I) IS-D.
~----------------------~<'> 6 ~
JAN FEB MAR APR MAY JUN
1981-1982
JUL
Figure 7. Well level fluctuations for 1981 and 1982 (data collected by Northern Testing Laboratories, Inc.).
1-
u_
w
u
c;:
0::
:::>
(f)
0:: w
\Jl 1-
I ~ [\)
[\) 0 1-
I
1-
Q
w
0
10
20
30
40
50
60
70
80
90
SWAINBANK's WELL --
ST. JOE LOWER (#5) x---x
ST. JOE UPPER I'll) t>--c,
X-·-·-·-·-·-·-·-·-·-·
10 20 /0 20 /0 20
J/\N FEB MAR
10 20 10 20 /0 20 10 20
JUN JUL AUG SEPT
Figure 8. Well level fluctuations for 1983.
precipitation event beginning on August 20 lasted 28 hours and accounted
for 1.88 inches of precipitation. This event seemed to initiate the
97-foot rise in the well between August 19 and 25. The well continued
to rise slowly until September 9, when it attained a level of 39 feet.
The other wells, located at lower elevations, showed little response
over the period even though the precipitation amounts were similar to
those at the Swainbank site. Unfortunately, there are no well logs
available for these three unpumped wells.
During the period that the Swainbank well rose dramatically, increases
in streamflow at both creeks were also noted (Figure 9). It should be
pointed out that this was the only significant rise in the flow rates,
except during the snowmelt period.
Potential groundwater recharge during this period was estimated by
assuming the soils were relatively dry. Dry soil has pore space which
can store water. Kane et al. (1978) reports that the porosity of a
Fairbanks silt loam is 50% by volume. They also report that a dry
Fairbanks silt loam is approximately 12% moisture by volume, leaving
fillable porosity of 38% by volume (equal to 3.04 inches of
precipitation in the eight-inch deep rooting zone). The precipitation
remaining after interception and evapotranspiration losses is free to
move through the soil to the water table. Water in deeper soils is not
directly affected by transpiration.
The average pan evaporation for the two sites was 0.54 inches. Assuming
the pan coefficient is between 0.7 and 0.85 for the Ester Dome area, the
evapotranspiration was between 0.38 and 0.46 inches during this period.
5-23
Average precipitation in the Ester Creek and Happy Creek watersheds was
4.36 and 4.06 inches, respectively, during this period. Dingman (1971)
estimated interception was 22% by deciduous forests and 38% by
coniferous forests in the Glenn Creek watershed. We used the 22%
interception factor in the Ester Creek watershed since it is dominated
by deciduous forest and the 38% interception factor in the Happy Creek
watershed since it is dominated by coniferous forest. After
interception and evapotranspiration losses were subtracted,
precipitation was between 2.9 to 3.0 inches for Ester Creek watershed,
and between 2.0 to 2.1 inches for Happy Creek watershed. On the basis
of average watershed infiltration, there was not sufficient
precipitation to saturate the upper soils. Much of the soil water
stored in the rooting zone would be lost as evapotranspiration.
These crude estimates would suggest that recharge from rainfall would be
minimal. However, watershed conditions are never average. For example,
higher elevation vegetation is sparse, soil cover thinner, more
precipitation falls and bedrock is closer to the surface. This may
account for the response of the Swainbank well, and apparent lack of
response at the lower 1·1ells. The Swainbank site r·eceived more
precipitation (5.15 inches).
The 1983 summer base flow for Ester Creek ranged between 1.0 and 3.0
cfs, and remained fairly constant from mid-May through late July. Our
estimate of base flow for Ester Creek from mid-May to mid-September was
570.0 acre-feet. The base flow for Happy Creek declined gradually from
3.08 cfs on May 12 to a minimum of 0.028 cfs on July 14 (Figure 9), when
the flow rate began to rise reaching 3.15 cfs on September 9. Our
estimate of base flow for Happy Creek from mid-May to mid-September was
5-24
CJ)
"-u .
w
l9 a:
<:{
:r: u
CJ) -0
100
80
60
40
20
10
8
6
4
2
1.0
0.8
0.6
0.4
0 2
01
0 05
DISCHARGE HAPPY CREEK --
DISCHARGE ESTER CREEK o--o
o.olr-~~--------,---~--r------.------,-------r-------APR MAY JUN JUL AUG SEP
TIME, 1983
Figure 9. Variations in streamflow for Happy and Ester Creeks, 1983.
5-25
85.2 acre-feet. Precipitation from late April through mid-July was
light over the entire area and extremely light at lower elevations.
The extremely low flows in Happy Creek may be due to the relatively low
boggy terrain and permafrost. The limnological study of Ace Lake by
Barsdate (1967) discusses the anaerobic conditions of the deep water in
the lake. This may be due to groundwater entering the lake from beneath
the permafrost. The upper reaches of Happy Creek and its tributaries
above the lakes were observed flowing only dul'ing the spring snowmelt
period and after the extended August precipitation period. In a study
of a similar lake in permafrost terrain, Kane and Slaughter (1973)
showed water moved upward through the thawed sediments beneath the lake
and into the lake. This recharge of the lake water can only occur if
permafrost does not exist beneath the lake. In a small thaw lake near
Ester Dome, Hartman and Carlson (1973) found that the lake was isolated
from the groundwater table and recharged by spring snowmelt.
Groundwater flow and subsequent recharge of Ace Lake, which is over
26-feet deep and covers 14 acres, and other similar lakes in the
watershed is probably the source of the Happy Creek base flow.
Annual Relationships Average annual precipitation for Ester Creek
watershed was 14.95 inches during the 1981-82 water year which is
equivalent to 8,070 acre-feet. Average annual precipitation for the
Happy Creek watershed was 12.67 inches equivalent to 6,520 acre-feet.
The precipitation fell as 72% rain and 28% snow in the Ester Creek
watershed, and as 70% rain and 30% snow in the Happy Creek watershed.
Average annual precipitation for the Ester Creek watershed during the
1982-1983 water year was 17.14 inches, equivalent to 9,260 acre-feet.
Average annual precipitation for the Happy Creek watershed was
5-26
12.14 inches, equivalent to 6248 acre-feet. The annual precipitation
fell as 60% rain and 40% snow in the Ester Creek
watershed, and as 56% rain and 44% snow in the Happy Creek watershed.
Conclusfons
High evapotranspiration rates and light showery precipitation that occur
in eastern interior Alaska prevent nearly all of the summer
precipitation from infiltrating to the water table and providing any
groundwater recharge. Only long-duration, high-volume summer
precipitation events (such as the one in late August of 1983) saturate
the soil to sufficient depth to provide some recharge. Therefore, the
significant period for groundwater recharge is the spring snowmelt
season, when a large volume of water is steadily released and
evapotranspiration rates are low. During the snowmelt period of 1983,
40% to 55% of the snowpack water equivalent infiltrated to provide
potential recharge to the groundwater. Some additional groundwater
recharge may have occured in late August, but probably only at higher
elevations.
Acknowlegments
This study was made possible through funding provided by the Alaska
Department of Natural Resources, Division of Geological and Geophysical
Surveys. The cooperation of the residents of the Ester Dome area
facilitated the collection of data. Special thanks go to David Stone,
Harold Coutts, Larry Gedney, Richard Swainbank, Gerry Bergraf, Bobbi
Rice, Steve Lasonsky, and the St. Joseph American Mining Company for
providing data collection sites. Personal thanks go to Catherine Egan
for assisting in data collection and reduction, and to Paula Wellen for
assistance in the preparation of the paper.
5-27
References
Barsdate, R. J. 1967. Pathways of Trace Elements in Arctic Lake
Ecosystems. Annual Progress Report to the U.S. Atomic Energy. Contract
No. AT(D4-3)310PA4.
Bredthauer, S. R., and D. Hoch. 1978. Drainage network analysis of a
subarctic watershed. U.S. Army Cold Regions Research and Engineering
Laboratory, Hanover, New Hampshire. Special Report 129.
Chacho, E. F., and S. R. Bredthauer. 1983. Runoff from a small subarctic
watershed, Alaska. Proceedings: Fourth International Conference on
Permafrost, Fairbanks, Alaska. (in press).
Dingman, S. L. 1971. Hydrology of the Glenn Creek Vlatershed, Tanana
River basin, central Alaska. U.S. Army Cold Regions Research and
Engineering Laboratory. Research Report 297.
Eaton, F. and Wendler. 1982. The heat balance during the snow melt
season for a permafrost watershed in interior Alaska. Arch. Met. Geoph.
Biokl. Ser.A(31) pp 19-33.
Forbes, R. B. 1982. Bedrock geology and petrology of the Fairbanks
Mining District, Alaska. Alaska Department of Natural Reources, Division
of Geological and Geophysical Surveys. Open-file Report 169.
Ford, T. R. 1973. Precipitation-runoff characteristics of the Caribou
Creek Research Watershed, near Fairbanks, Alaska. University of Alaska,
Fairbanks. MS Thesis.
Hartman, C. W. and R.
a permafrost region.
Report 42.
F. Carlson. 1973. Water balance of a small lake in
University of Alaska, Institute of Water Resoures
Hawkins, D. B., R. B. Forbes, C. I. Hok, and D. Dinkle. 1982. Arenic in
the water, soil, bedrock and plants of the Ester Dome area of Alaska.
University of Alaska, Institute of Water Resources Report 103.
Haugen, R. K., C. H. Slaughter, K. E. Howe, and S. L. Dingman. 1982.
Hydrolgy and climatology of the Caribou-Poker Creeks Research Watershed,
Alaska. U.S. Army, Cold Regions Research and Engineering Laboratory
Report 82-26.
Kane, D. L., R. D. Seifert, and G. S. Taylor. 1978. Hydrologic
properties of subarctic organic soils. University of Alaska, Institute
of Hater Resources Report 88.
Kane, D. L., S. R. Bredthauer, and J. Stein. 1981. Subarctic snowmelt
runoff generation. T. S. Vinson, ed. The Northern Community: A Search
For a Qua 1 i ty Environment. American Society of Civil Engineers, New
York. pp 591-601.
5-28
Kane, D. L., and J. Stein. 1983. Physics of snowmelt into seasonally
frozen soils. Proceedings: Advances in Infiltraion, American Society of
Agricultural Engineers (in press).
Kautz, F. R., and C. W. Slaughter. 1972. Geologic setting of the
Caribou-Poker Creeks Reseaerch Watershed, interior Alaska. U.S. Army
Cold Regions Research and Engineering Labroratory. Technical Note.
Lotspeich, F. B. and C. vJ. Slaughter. 1981. Preliminary results on the
structure and functioning of a taiga watershed. , University of Alaska,
Institute of Water Resources Report 101.
Pewe, T. L. 1982. Geologic Hazards of the Fairbanks Area, Alaska. State
of Alaska Department of Natural Resources, Division of Geological and
Geophysical Surveys. Special Report 15.
Slaughter, C. W. ,and D. L. Kane. 1973. Recharge of a central Alaska
lake by subpermafrost groundwater. Permafrost. Proceedings of the 2nd
International Conference, Yakutsk, U.S.S.R., National Academy of
Sciences. pP 458-462.
U.S. Department of Agriculture, Soil Conservation Service. 1963. Soil
survey Fairbanks area, Alaska, Series 159, No. 25. U.S. Government
Printing Office, Washington, D.C.
Vogel, T. C., and C. W. Slaughter. 1972. A preliminary vegetation map of
the Caribou-Poker Creek Research Watershed, interior Alaska. U.S. Army
Cold Regions Research and Engineering Laboratory, Hanover, New
Hampshire. Technical Note.
5-29
DATA GENERATED FROH ALASKAN HYDROPO\vER DEVELOPNENT
By Stephen 1 R. Bredthauer , Jeffrey H. Coffin2 ,
and Eric A. tlarchegiani3
Abstract
Hydroelectric development in Alaska is usually hampered by lack of an
adequate data base. This has often been the case for the numerous
hydropower studies being conducted by the Alaska Power Authority.
Significant efforts have been expended in each project to obtain the
required information. This has led to the creation of a water resources
data base previously unavailable for much of Alaska. This report
summarizes the data base developed by the Pm<er Authority.
Nost hydroelectric projects have required basic data on basin hydrology,
climatology and geology. Instream flow studies have required data on
seasonal variation in fisheries populations and streamflmv, stream
temperature and sediment transport regimes. In addition, local
groundwater, \Vater quality and river and lake ice conditions have been
documented. The influence of glacial melt on basin water yield, reservoir
thermal regimes and sediment regimes have also been investigated. The
availability of these data has contributed to the knowledge of hydrologic
processes influencing hydroelectric development in Alaska.
Introduction
The Alaska Power Authority was established by the 1976 state legislature as
a public corporation of the State of Alaska. The stated purpose of the
Power Authority is to identify, evaluate, and develop electrical potver
generation facilities, using the most appropriate commercial technologies
1. Chief, Hydrology Department, RM! Consultants, Inc., Anchorage, Alaska.
2. Senior Civil Engineer, R&H Consultants, Inc., Anchorage, Alaska.
3. Project ilanager, Alaska Po~<er Authority, Anchorage, Alaska.
6-1
(except for nuclear). \vith this purpose in mind, the Pm<er Authority has
identified as its goal the development of Alaska 1 s energy resources in a
manner that strengthens and diversifies the economy and improves the
standard of living. This is to be accomplished by providing energy and
capacity at the lo~vest reasonable economic, social, and environmental
costs, with emphasis on energy from rene~vable and local resources.
The Po\ifer Authority's purpose and goal require that a variety of electrical
technologies be evaluated. These include diesel or gas turbine, tvaste
heat, wood ~vaste, geothermal, coal, 1vind., and hydroelectric generation.
Each type of generation requires certain data collection programs, but the
successful development of hydroelectric projects is perhaps the most
dependent upon an adequate tvater resources data base. The importance of
water resources data collection for hydroelectric development has required
the Pmver Authority to invest significant funds in the ~vater resources
field.
This emerging \Vater resources data base is the emphasis of this paper. The
paper will illustrate and describe a matrix of projects and data types.
Detailed studies have been conducted at some projects to better understand
the interaction of the project with the natural environment. These are
briefly described in the section on special studies. The paper also
describes available reports, by project, and where the reports may be
located.
6-2
Description of Matrix
The location of each hydroelectric project site investigated by the Alaska
Power Authority is illus-trated in Figure 1. The numbers on the map relate
to the numbers with the projects in the matrix in Table 1, and Table 1
shows the types of \cater resources field data that have been collected at
each project site. Not all sites investigated by the Power Authority have
been entered in the matrix. Projects have only had
reconnaissance-level studies to date (October 1983), and thus have had no
or minimal field data collection efforts, have not been included. All
other projects are noted 1 however, including a few where most of the field
data were collected by agencies other than the Power Authority (such as the
Army Corps of Engineers or local elec-tric utilities).
Each of the thirteen field parameters in the matrix is briefly described
belm;. Also discussed are the columns defining the projects 1 names,
locations, capacity and status.
1. Project Name -The name of the project, as used by the Alaska Pm;er
Authority (though abbreviated in some cases). In the regional
studies, primary emphasis J,..IBS placed on a few specific sites, as
follows:
Bethel Area Power Plan Feasibility Assessment (Chikuminuk Lake)
Bristol Bay Regional Power Plan (Newhalen River, Tazimina River)
Cordova Power Supply (Silver Lake, Power Creek, Allison Lake)
6-3
0\
I
+:-
!68 16'1
oBETHEL
160 "'
·.-·-Q-FAIRBANKS
@
' I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
I
I
I
I
I
' I
I
I
,,~o....,:o~--=="!~50 1.11
@4 .•
• .. }[··.@ •. • •. ~· ·~····A····.·.·.C.Ht.·.·O·.fR.'A.· .. ~~.·.,.·'·'··· ..... ·.·.·.·.•. \ ·· • .. ·.· \. · ..... . "-.-~--(. · .. ·· •• ···· .... ~~' .·· .. --c..~--£..A :-,, -<"'-g --. . ' . .r--. '--· . _'..., / ' . ,d __ ·
y " -.O~EAU
' .. ,.
3
.n /_~1
~.
4
-" f.?<..o,
i
\52
'
144 '"
' ,.
·~~
~\ '"
FIGURE 1
HYDROELECTRIC PROJECT LOCATION MAP
6]
J
1 ]
l
CJ\
I
\J1
TAOLE 1
WATER RESOURCES DATA COLLECTED AT
ALASKA POWER AUTHORITY HYDROELECTRIC PROJECTS
/'fW,JECT /1/11-1[
3. Orad ley lal<e
11. nristul !l<Jy Region
~-Cllahac/\amnn
l. Cordova I'O\.IJ~r Supply
Ill. ll;r inn~.;Sh<~!J""ilY
We~;L Ct'l~t!k
11, 1\ i IHJ CIJVI!
12. 1.<11<.e Elvll
13. Lart,en lliiY
lll. OILI llar!Jor
15. l't: 1 i can
16. Sui0/11[)11 Gulctl
1-f. Sus 1 tna
18. S'W;1r1 talw
l'J. Tt:l"fOI' Laku
21. lyel~ Lake
/IOHS:
(rO" 13'
•,'j'' J 3'
Gl ,, 12 I
(if)''5 7 I
')')"21'
'iOJ" 30'
'i'l'' 1?'
';6" 11 1
·o,
c
0
_j
JJJ "JJ I
162"19'
1 !:J3" 1 B 1
1 36" 13'
153"0()1
J6(}'1 2/J I
JJJ 0 2{J I
f(CfiH' tO lf:.:t f'Clf' e:.:pl,1fliii.IOII nr par;rmetOfS.
l·l.ttrl :< curn:nt <b or OctulH~r 1'JU3.
.£ u ro
0. ro u
3-6
60-lJ:i
.2-16
330
2.5
15
7
2.5
fi
0.6
1. 5
lJ. 3
0. 3
O.B
12
1620
22
20
0.11
20
IF
L
IF
If
F
F
0
l
c
G
" c
L
-~ u ro
X
X
X
3
0 c
V1
X
X
X
X
X
X
X
X
X
X
X
3
0 ;:;::
E ro
" L
~
V1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
)(
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
>-rn
0
0 c
E
_j
X
X
)(
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
" c
0 :;o
u • V1
' X
L • .":
a:
X
X
X
X
• u
X
X
u.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2. Latitude -Approximate north latitude in degrees and minutes, at the
primary lake or site where much of the data collection has been
concentrated (e.g. Chikuminuk Lake for the Bethel Regional Study,
Newhalen River for the Bristol Bay Study, Silver Lake for the Cordova
Power Supply Study, and the Susitna River at Gold Creek for the
Susitna Project).
3. Longitude -Approximate west longitude in degrees and minutes, at the
same point as in Item 2.
4. Power Capacity (Niv) -Approximate size of the project, in megmcatts.
Size given is the one recommended or is a range for projects proposed
to be built in phases. In regional studies, the capacity for several
alternatives is specified.
5. Project Status -Current status of the project. The most recent data
reports may have been completed under this phase or under the prior
phase. The project phases, with the codes used to identify them in
the matrix, are as follows:
R = reconnaissance study
IF = interim feasibility assessment
F = feasibility study
L = license (FERC) applied for
C = construction
0 = operation
6-6
6. Glacier -In-depth study of and data collection on glaciers in the
drainage basin.
7. Snmv -Systema-tic snow surveys of depth and water content, conducted
8.
in the basin supplying the project. Conducted on a monthly bas is
either by or for the U.S. Soil Conservation Service. Data are
reported monthly from February to June in the 11 Snot.;r Surveys and \'later
Supply Outlook for Alaska" series (SCS, annual).
Neteorological Climatic data (air temperature, wind speed and
direction, and precipitation) that have been collected specifically
for the project or are unpublished, systematic data collected in the
immediate vicinity (e.g. G.O. Balding report for Pelican [Balding,
1974]). NOAA data are readily available in agency publications.
Since they have not been collected specifically for hydroelectric
projects, they have not been considered in the ma·trix.
9. Streamflm; -Continuous streamflm; records collected for and at the
project. Collected and reported by U.S. Geological Survey and/or by a
private contractor. U.S.G.S. data are reported annually in the "\~ater
Resources Data for Alaska" series (U.S.G.S., annual).
10. \~ater Quality -Data collected on stream water quality. Noted in
matrix if one or more measurements were made of any parameters besides
temperature (e.g. King Cove, Larson Bay, Old Harbor, and Togiak each
have data for one date only).
6-7
11. Water Temperature Continuous records of stream temperature,
collected either by U.S.G.S., by another agency, or by a private
contractor.
12. Sediment -Stream sediment data collected for analysis related to
sediment deposition in the project reservoir. Bedload and suspended
sediment data have also been collected for the Susitna projec·t for
analysis of do~vnstream impacts.
10. Limnology -Data-collection and analysis of lake characteristics and
processes, generally as related to project operation or environmental
impacts. Data are for water quality parameters on an existing lake in
the project or on an off-project lake studied as an analogy
(e.g. Eklutna Lake as an analogy to Watana Reservoir in the Susitna
project).
14. Bathymetric Data (Lake) -Lake-depth data collected for development of
an area-capacity curve -reported in the matrix if such a curve is
available. Old (1950's) U.S.G.S. plan-and-profile maps of potential
project sites have also been considered.
J.S. River X-Sectio11s -Cross-sections surveyed through a significant reach
of river for hydraulic modeling purposes. Does not include local
surveys made at damsites or other sites strictly for design purposes.
6-8
16. Ice -Collection of field data on river or lake ice conditions and
processes, but only if collected systematically and if analyzed to
some extent (i.e. not considered if only random measurements of ice
thickness were made).
17. Fisheries -Data collection on anadromous species, resident species,
and habitat characteristics in river or lake environments.
18. Photogrammetry -Aerial photography taken for water-resource studies,
such as for measurement of wetted areas in sloughs or changes in
glaciers. Does not include photography made strictly for topographic
mapping.
Special Studies
A few of the parameters have received special attention in several Power
Authority projects. These four areas of study (glaciers, limnology, ice
and instream flmv) and the projects \\1hich addressed them are explained in
greater detail in this section.
Glacier Studies -l·Jany of the proposed hydroelectric developments in
southeast and south-central Alaska are located on glaciated basins.
Glaciers significantly alter the timing of the hydrologic cycle in a
basin, storing water as glacier ice during cool, wet years, and
releasing water from glacial melt during dry years. This
characteristic dampens the annual variations in summer streamflow.
6-9
\{ater re,source engineers must recognize this shift in streamflotv when
analyzing the economic feasibility of a project.
The impact of glaciers on the \Vater supply of a project is most
evident where long-term discharge records exist. If the glaciers in
the basin significantly receded during the streamflow period of
record, additional flow from glacial melt is being recorded.
Conversely, if the glaciers have been expanding, the gage would be
recording less flo\v than normal. The flmv frequencies estimated from
streamflmv records could change significantly if the glaciers 1
hydrologic regime were to shift due to a climate change. A project
could be oversized or undersized due to these shifts.
Glaciers may have several other significant impacts on hydroelectric
development. Lakes either on the glacier, or else behind a glacier
blocking a tributary_\ may suddenly break out, causing the equivalent
of a dam-break flood. Glacier surges may form lakes where none
previously existed. Glaciers also contribute a large amount of
sediment to the stream. The volume and size distribution of the
sediment affect both the economic life of the project and the
environm<:cmtal conditions in the rest:rvoir and dmvnstream. Finally,
the glaciers have a significant role in the operation of the project,
providing a steady water supply under conditions tvhich may cause
drought in non-glacierized basins.
6-10
Reconnaissance-level studies have been conducted for the Susitna and
Bradley Lake projects. Photogramrnetric techniques were used to obtain
rough estimates of long-term mass balance changes in some of the
glaciers in both basins, and estimates ~vere made of the impacts on
«ater supply. Hass balance studies have been initiated at both
projects. A reconnaissance of the Susitna basin was also conducted to
determine the presence of glacial lakes with potential for significant
outburst floods. In addition, investigations and observations have
been made by the U.S.G.S. of the terminal zone of Barrier Glacier at
the outlet of Chakachamna Lake (Giles, 1967).
Limnology Studies As the matrix indicates, many of Alaska's
hydroelectric projects have had limnological data collection and
analysis. The level of effort has varied considerably from project to
project. In most cases, data have been collected to identify baseline
water quality conditions in existing lakes. In a few of the projects,
fisheries in the lakes were also well-documented (Black Bear and
Chakachamna). Numerous minnow traps v . .1ere used in Terror Lake to
search for resident fish, but none were found.
vertical plankton tows were also made in Terror Lake.
Horizonta 1 and
An exception to the general rule of mere baseline data collection_,
however, was the Susitna Project. There is not an existing lake
which will be part of the Susitna Project, but detailed observations
were made of an off-project lake (Eklutna), in order to calibrate a
numerical model of lake processes. The model (DYRESH) (Imberger and
Patterson, 1981) is currently being used to analyze project effects on
6-11
temperature in the reservoir and in the river dotvnstream. Field data
collected included vertical profiles of temperature, turbidity,
conductivity, and light extinction at several stations on the lake.
Data were collected at least monthly through the open-water season for
two consecutive summers. Occasional lake samples were also analyzed
for concentration and size of suspended sediment. Additional data
collected on a continuous basis t\rere meteorological conditions at the
lake, streamflow into the lake from two major tribu-taries, and water
temperature of the two tributaries.
In the studies where data were collected to document baseline
conditions, measurements most commonly made t\rere profiles of
temperature. Dissolved oxygen, pH, nutrients, and other water qualitY
parameters tvere also commonly measured. Spatial intensity tvas usually
limited to one or sometimes two sites on the lake. Frequency of
measurement varied widely in intensi·ty: seasonal profiles were
obtained in Black Bear Lake, Bradley Lake, and Grant Lake; monthly
summer profiles were obtained in Tyee Lake; winter and summer profiles
t\'ere measured in Chakachamna Lake and in Cordova (Silver Lake); and
single summer profiles were observed in the Bristol Bay (Sixmile and
Tazimina Lakes), Lake Elva, and Swan Lake projects.
Ice Studies Construction of a hydroelectric power project on a
northern river significantly alters the winter flow and thermal
regimes of the river, subsequently modifying the rivers ice regime.
Flows higher than the normal winter flows are released in winter to
6-12
meet the high power demands. These flows will generally have
temperatures of 2 °-4 °C. The combination of higher temperatures and
large volumes of t\'ater creates a large heat mass tvhich must be
dissipated before ice formation can again occur downstream of a dam.
In addition, the dam and reservoir block the dDh1 nstream flow of ice
fanned in the upstream stretches of the river.
Intensive field studies have been conducted on the Susitna River to
document freeze-up and breakup processes, to document the
environmental impacts of these processes, and to provide data for
mathematical modeling of the pre-and post-project ice conditions.
Ice formation on lakes and reservoirs is also important, both from
engineering and envirorunental viewpoints. Forces exerted on
structures during freeze-up may be significant. Frazil ice may cause
blockage of intake structures. Reservoir drmvdown during the winter
may affect wildlife migration. Ice cover formation influences the
reservoir temperature. Data are being collected at Eklutna Lake to
calibrate the DYRESH reservoir temperature model in support of the
Susitna studies. This model is being modified to include ice cover
formation. Limited ice data have also been collected at Grant Lake
and Bradley Lake.
6-13
Instream Flmv Studies -The objective of completing an instream flow
study is to determine the relationship between different discharges
and the effects on various instream uses and resources. The natural
conditions of streams fluctuate due to rainfall events and also due to
seasonal effects such as snowmelt. A reservoir system has the effect
of attenuating the wide fluctuations of natural flow so that a
hydroelectric development could enhance certain aspects of the fishery
resource, depending upon its operation. The completion of an instream
flow analysis is the means of determing the options available and the
ramifications of each of the options.
There are a numbers of methods (lvesche & Rechard, 1980) which might be
employed to complete an insteam flow analysis. A fe1v of the more
commonly used methods are the Montana Method (Tennant, 1975 & 1976),
the Oregon t!ethod (Thompson, 1972 & 1974), and the Instream Flow
Incremental t!ethodology (IFH!) (Bovee, 1982). Each of these methods
has its o'vn limitations and assumptions and must be applied with the
appropriate judgment. The various methods are tools for evaluating
the potential changes in the nattn-al flow regime. A specific method
may not always be applicable to a particular situation. Therefore,
the best utilization of a method may be to either combine. it \•.dth
another method, or else ·to modify one method such that it will provide
better information concerning any proposed changes to the flow regime.
6-14
The level of a particular study influences the type of instream flow
methodology selected to evaluate the potential changes in flow
regime. The instre.am flow study could be divided into two levels.
Level 1 might be considered as a reconnaissance inst1-eam flow study,
h1hile Level 2 ~.;auld be a comprehensive instream flmv study.
The Level 1 study would determine if the project is feasible from a
biological perspective. This analysis would insure that there were no
unacceptable environmental conditions which ~vould preclude development
of a project. Generally, the data requirements would be seasonal
flmv, temperatnre.s .• and use of habitats by various species, along ~vith
some assessment of channel stability.
The Level 2 study \could be a detailed instream flo1v analysis which
t .. ,ould include impact evaluations and mitigation measures. The
analysis would provide a biological perspective which would be
utilized in the development of the project design and operation. A
large amount of site-specific data would be needed in order to
complete this analysis. These data tvould include reach-specific
hydraulics, tvater quality, water temperature, and sediment transport,
along wj th the necessary fishery data. A Level 2 analysis would
produce a document with a delineation of pre-a11d post-project,
reach-specific conditions which would enable one to define impacts
and develop the necessary mitigation plans.
6-15
A good example of a reconnaissance level (Phase 1) study ~;as completed
in conjunction "ith the Bristol Bay Regional Study (Baldrige and
Trihey, 1982). The analysis completed on the Terror Lake
Hydroelectric Project (\iilson et .§l, 1981) is a good example of a
Phase 2 analysis. These analyses are on two completely separate
projects, which enables one to view two different local conditions and
review the approaches. The Po\ver Authority is presently pursuing a
Phase 2 type analysis on the Bradley Lake Hydroelectric Project
through its contractors. This study is unden"'ay, and results should
be available within three months. The largest Phase 2 analysis being
conducted is the one for the Susitna Project. Many of the available
water resource papers listed for Susitna were writ-ten in support of
the instream flow studies. These studies are in progress, and results
"ill be available in the future.
Availability of Data
The availability of different types of data is directly related to the
level of the study being conducted, as 1vas discussed above in Instream F,lo«
Studie.s. The Authority generally has nvo study levels:
reconnaissance studies and detailed feasibility studies. On occasion, the
Power Authority also completes an interim feasibility assessment which is
an analysis between the reconnaissance study and a detailed feasibility
study. At the reconnaissance level, many alternative sites are being
revim<ed, and it IVDuld not be cost effective to collect detailed data on
all sites. Instead, office studies are conducted with a limited amount of
6-16
field work being completed on the best alternatives. The detailed
feasibility level involves an intensive field data collection program t;ith
detailed analysis of the data. Therefore, the level of detail for field
data collection in a reconnaissance study is substantially less than that
compiled for a detailed feasibility study.
Once a reconnaissance study or a detailed feasibility study has been
finalized, the Pmver Authority submits twenty copies of the report to the
state library in Juneau, Alaska. Individuals or libraries interested in
borrowing Alaska State publications should contact their nearest depository
library. Additional inquiries or difficulties in obtaining publications
should be referred to the Alaska State Publications Distribution Center,
Alaska State Library, Pouch G, Juneau, Alaska 99811 (907 465-2942).
6-17
The follo.ving is a list of the twenty depository libraries.
Alaska State Library,
JuneauJ Alaska
Alaska Historical Library,
Juneau, Alaska
University of Alaska Library,
Juneau, Alaska
Sheldon Jackson College Library,
Sitka, Alaska
Ketchikan Public Library,
Ketchikan, Alaska
Z.J. Loussac Public Library,
Anchorage, Alaska
Alaska Resources Library,
~1chorage, Alaska
University of Alaska Library,
Anchorage, Alaska
Kenai Community Library,
Kenai, Alaska
A. Holmes Johnson Public Library,
Kodiak, Alaska
Noel Wien Memorial Library
Fairbanks, Alaska
University of Alaska
Rasmusson Library, Fairbanks
Kegoayah Kozga Public Library,
Nome, Alaska
Kuskokwim Consortium Library,
Bethel, Alaska
Seattle Public Library,
Seattle, Washington
University of \iashington Library,
Seattle, Washington
Washington State Library,
Olympia, Washington
Center for Research Libraries,
Chicago, Illinois
Library of Congress,
Washington, D.C.
National Library of Canada,
Ottawa, Ontario
In addition to this distribution list, there are a number of other places
v,rhere one can locate information generated by the Power Authority. The
Pmver Au-thority deposits two copies of all final reports in its mvn library
(334 West 5th Avenue, Anchorage). The project manager will usually deposit
additional copies of reports in a local library near the project if it is
not on the Alaska State Library depository list. The primary contractor
for the project might also have additional data which have not been
published. The Power Authority also funds data collection efforts in
conjunction with hydroelectric projects through cooperative agreements with
6-18
the U.S. Geological Survey (USGS). The majority of this data is published
in their water supply papers on an annual basis. The same type of
arrangement exists with the U.S. Soil Conservation Service (SCS). The SCS
collects snmcpack data which is published for the months of February
through June on an annual basis.
It should be noted that the Alaska Power Authority has become involved ~;ith
a number of projects which were initiated by other entities. These
projects have been transferred to the Power Authority due to its ability to
finance large hydroelectric projects. Specifically, these projects are
Solomon Gulch, Terror Lake, Swan Lake, Susitna, and Bradley Lake. Host of
the project reports containing these data are available in the Power
Authority library.
Summary
The purpose of this paper has been to present a consolidated list of the
types of water resources data available from hydroelectric projects
undertaken by the Alaska Power Authority. Since the data, in general, are
widely scattered and di.fficult to locate. it was felt that such a summary
would be baneficial to describe the availability and locations of the data
for engineers, scientists, and planners working in the water resources
field.
6-19
The matrix of projects and parameters in Table 1 indicates that project
data from the field have primarily been collected to determine ;vater
availability (streamflo1') and conditions related to instream flo;v. Host
projects have addressed water quality and fisheries on at least a
preliminary basis to identify baseline conditions in existing streams and
lakes. The number of parameters measured (and which ones) gives an
indication of the level of detail of the instream flm' study.
Revim.,r of the existing data base is an important part of the management
process of deciding «hat additional data collection is required. Thj.s step
is necessary, whether the concern of a project is feasibility assessment 1
impact assessment., project design, or project operation. It is hoped that
this presentation will assist in the process of revie\ving previous studies
in a particular geographic area and will facilitate decision-making
relative to water resources development in Alaska.
Available Reports
This section has been prepared on a project-by-project basis to compile
specific reports in which hydrologic data are available. The project name
is listed first, followC>.d by the name of the prime contractor in
parentheses. Reports in ~d1ich field data are reported are then listed.
Additit1nal supporting refP.rences can be found in the bibliographies of
these reports.
Bethel Region (Harza Engineering)
Harza Engineering. 1982. Bethel Area Po;ver Plan Feasibility
Assessment, Regional Report, App. B, D.
6-20
Black Bear Lake (Harza Engineering)
Federal Energy Regulatory Commission. 1983. Black Bear Lake, Project
No. 5714, Draft Environmental Impact Statement.
Harza Engineering. 1981. Application for License before the Federal
Energy Regulatory Commission, Black Bear Lake Hydroelectric
Project on Prince of \lales Island, Alaska.
1979. Black Bear Lake Project, A Reconnaissance Report.
Black Bear Lake (Harza Engineering/CH2 N Hill Northwest)
Bishop, Daniel ~!. 1982. "Late Summer and Fall Observation in Upper
Black Bear Creek and Black Bear Lake." Environaid.
Environaid. 1982. Biological-Ecological Investigations on the Black
Bear Creek System near Klmvock, Alaska. Environaid. Juneau,
Alaska. Prepared for Alaska Power Authority.
Harza Engineering Company and CH 2H Hill Northwest. 1981. "Black Bear
Lake Project Feasibility Report, Volume 2, Appendices."
Bradley Lake (Stone & \vebster)
Colonell, J .H. 1980. Circulation and Dispersion of Bradley River
Water in Upper Kachemak Bay. \vood~<ard-Clyde Consultants.
Prepared for U.S. Army Corps of Engineers, Alaska District.
Gatto, L.\1. 1981. Ice Distribution and \linter Surface Circulation
Patterns, Kachemak Bay, Alaska. U.S. Army Cold Regions Research
& Engineering Laboratory, Hanover, N.H.
Gosink, J.P. and T.E. Osterkamp. 1981. A Theoretical Investigation
of the Potential rlodification of Ice Formation in Kachemak Bay by
the Bradley Lake Hydroelectric Project. U.S. Army Cold Regions
Research & Engineering Laboratory, Hanover, N.H.
Stone & \lebster. 1983. Interim Report on Feasibility Evaluation for
Bradley Lake Hydroelectric Project.
U.S. Army Corps of Engineers, Alaska District. 1981. Bradley Lake
Hydroelectric Project, Design ~lemorandum. No. 1 -Hydrology.
1982. Bradley Lake Hydroelectric Project, Alaska. Final
Environmental Impact Statement.
Bristol Bay Region (Stone & \lebster)
Stone & lvebster.
Feasibility
Volumes 1-4.
1982. Bristol Bay Regional Power Plan, Detailed
Analysis, Interim Feasibility Assessment,
6-21
Baldrige, J.E. and E.\{. Trihey. 1982. Potential Effects of Tl4o
Alternative Hydroelectric Developments on the Fishery Resources
of the Lmo~er Tazimina River, Alaska. AEIDC., in cooperation with
Dames & Noore.
Chakachamna (Bechtel Civil and Hinerals, Inc.)
Bechtel. 1983. Chakachamna Hydroelectric Project, Interim
Feasibility Assessment Report, Volumes I-IV.
BechteljWoodt4ard-Clyde Consultants. 1982. A Summary of Fish Passage
Facility Design Concepts and Preliminary Results of FY 1982-83
Fish Studies.
Chester Lake (Harza Engineering)
Harza Engineering. 1982. Chester Lake Project Feasibility Report.
Cordova Pmver Supply (Stone & \vebster)
Stone & \lebster. 1982. Cordova Pmver Supply, Interim Feasibility
Assessment, Vols. 1-2.
1982. Cordova Po14er Supply, Interim Feasibility
Assessment, Technical Data, June 1982.
1982. Cordova Po14er Supply, Interim Feasibility
Assessment, Addendum I.
1982. Cordova Pmver Supply, Silver Lake Hydroelectric
Site, Field Data Collection, April -October 1982.
1982. Draft Environmental Field Study Plan. 1983-1984.
Silver Lake Alternative, Cordova Pmver Supply Feasibility
Analysis: Phase II, March 1983.
Grant Lake (Ebasco Services, Incorporated)
Alaska Department of Fish and Game. 1981.
(unpublished, on file at Alaska Department
Soldotna, Alaska).
Grant Lake Survey
of Fish and Game,
Arctic Environmental Information and Data Cen-ter. 1982. Summary of
Environmental Knowledge of the Proposed Grant Lake Hydroelectric
Project Area.
Ebasco Services, Inc. 1983. Grant Lake Hydroelectric Project,
Detailed Feasibility Analysis, Vol. 1 & 2 -(1981). Grant Lake
Hydroelectric Project, Interim Report.
R&M Consultants, Inc.
(unpublished).
1982. Grant Lake Hydrological Data Report
6-22
Haines/Skagway -Dayebas (R.\1'. Beck & Associates)
R.\1'. Beck and Associates. 1981. Addendum to Reconnaissance Report on
Alternatives for the Haines-Skagway Region.
Haines/Skag'''ay -\vest Creek (R.\v. Beck & Associates)
R.\1'. Beck and Associates. 1982. Haines-Skagway Region Feasibility
Study, Volumes 1-3.
King Cove (DOKL Engineers)
DOIVL Engineers. 1982. Volume B, Final Report, Feasibility Study for
King Cove Hydroelectric Project.
1983. Feasibility Study
Project, Supplemental Data Report:
Lake Elva (R.\1'. Beck & Associates)
for King
Hydrology.
Cove Hydroelectric
R.\1'. Beck & Associates, Inc. 1981. Lake Elva Project, Detailed
Feasibility Analysis, Volume 1 -Report.
1981. Lake Elva Project, Detailed Feasibility Analysis,
Volume 2 -Appendices.
Lake Elva (Robert IV. Retherford Associates)
Robert IV. Retherford Associates. 1980. Reconnaissance Study of the
Lake Elva and other Hydroelectric Power Potentials in the
Dillingham Area.
Larsen Bay (DOII'L Engineers)
DO\iL Engineers. 1982. VolumeD, Report, Feasibility Study for Larsen
Bay Hydroelectric Project.
1983. Feasibility Study for Larsen Bay Hydroelectric
Project, Supplemental Data Report: Hydrology.
Qld Harbor (DDIVL Engineers)
DD\VL Engineers. 1982. Volume C, Final Report, Feasibility Study for
Old Harbor Hydroelectric Project.
1983. Feasibility Study for Old Harbor Hydroelectric
Project, Supplemental Data Report: Hydrology.
Pelican (USKH -Engineering Science)
Balding, G.O. 1974. \iater Reconnaissance Study of Pelican, Alaska.
U.S. Geological Survey.
6-23
USKH -Engineering Science. 1982. Pelican Pm;er Alternatives,
Phase I -Reconnaissance Assessment.
1983a. Pelican Power Alternatives, Phase II
Feasibility Study.
Solomon Gulch (Robert W. Retherford Associates)
Robert 1-1. Retherford Associates. 1974 -Revised 1976. Exhibit -\i,
Environmental Report for Solomon Gulch Hydroelectric Project, FPC
Project No. 2742. Prepared for Copper Valley Electric
Association, Inc.
Federal Pohrer Commission Bureau of Power. 1977. Solomon Gulch
Project/No. 2742-Alaska, Draft Environmental Impact Statement.
Susitna (Acres American, 1980-1983; Harza Ebasco, 1983)
Acres American, Inc. 1982a. Susitna Hydroelectric Project,
Feasibility Report. Final Draft. Volumes II, IV. Alaska Power
Authority, Anchorage, Alaska.
1982b. Susitna Hydroelectric Project: Fish and Wildlife
Mitigation Policy. Alaska Power Authority. Anchorage, Alaska.
llpp.
l983a. Application for License before the Federal Energy
Regulatory Commission, Susitna Hydroelectric Project, Alaska.
1983b. Slough Hydrogeology Report. Draft.
Alaska Department of
Status Report.
Alaska.
Fish and Game. 1972. Cook Inlet King Salmon
Alaska Department of Fish and Game. Juneau,
1974. An Assessment of Anadromous Fish Populations in
the Upper Susitna River \vatershed between Devil Canyon and the
Chulitna River. Alaska Department of Fish and Game. Anchorage,
Alaska.
_____ lgj6. Fish and \Vi ldl ife Studies Re.lated to the Corps of
Engineers Devil Canyon, Watana Reservoir Hydroelectric Project.
Alaska Department of Fish and Game. Alaska.
1977. Preauthorization Assessment
Susitna Hydroelectric Projects: Preliminary
\later Quality and Aquatic Species Composition.
6-24
of the Proposed
Investigations of
Alaska.
1978. Preliminary Environmental
Hydroelectric Development on the Susitna
Assessment of
River. Alaska
for U.S. Fish and Department of Fish and Game. Alaska. Prepared
Wildlife Service.
1980a. Inventory and Cataloging of Sport Fish Waters of
the Lower Susitna River and Central Cook Inlet Drainages. Alaska
Department of Fish and Game. Anchorage, Alaska.
1980b. Inventory and Cataloging of the Sport Fish and
Sport Fish \vaters in the Upper Cook Inlet. Alaska Department of
Fish and Game. Anchorage, Alaska.
Report.
Fish and
1981a. Adult Anadromous Phase 1
Susitna Hydro Aquatic Studies.
Game. Anchorage, Alaska.
Final Species/Subject
Alaska Department of
1981b. Phase 1 Final Draft Report Adult Anadromous
Fisheries Project. Susitna Hydro Aquatic Studies. 1981. Alaska
Department of Fish and Game. Anchorage, Alaska. Prepared for
Alaska Power Authority.
1981c.
Ins tream Flmv
Game. Alaska.
Phase 1 Final Draft Report Aquatic Habitat
Project. Volume I. Alaska Department of Fish
Prepared for Acres American, Incorporated.
and
and
198ld. Phase 1 Final Draft Report Juvenile Anadromous
Fish Study on the Lo«er Susitna River. Susitna Hydro Aquatic
Studies. 1981. Alaska Department of Fish and Game. Alaska.
Prepared for Acres American, Incorporated.
198le. Phase 1 Final Draft Report Resident Fish
Investigation on the Lower Susitna River. Susitna Hydro Aquatic
Studie.s. 1981. Alaska De.partment of Fish and Game. Alaska.
Prepared for Acres American, Incorporated.
l981f. Phase 1 Final Draft Report Resident Fish
Investigation on the Upper Susitna River. Susitna Hydro Aquatic
Studies. Alaska Department of Fish and Game. Alaska. Prepared
for Acres American, Incorporated.
l982a. Phase l Final Draft Report Aquatic Studies
Program. Sus i tna Hydro Aquatic Studies. Alaska Department of
Fish and Game. Alaska. Prepared for Acres American,
Incorporated.
1982b. Phase 1 Final Draft Stock Separation Feasibility
Report Adult Anadromous Fisheries Project. Susitna Hydro Aquatic
Studies. Alaska Department of Fish and Game. Alaska. Prepared
for Alaska Power Authority.
1983. Phase
Studies, Volumes 1-5.
II Final Report, Susitna Hydro Aquatic
Prepared for Alaska Power Authority.
6-25
Arctic Environmental Information and Data Center. 1982.
Chart for the Susitna River Salmon. Blueline.
Periodicity
Prepared for
Alaska Po1;er Authority, Anchorage, Alaska.
1983. Effects of the Proposed Susitna Hydroelectric
Project on Fishery Resources in the Susitna River Basin
Demonstration Impact Analysis. Draft Report. Submitted to
Harza-Ebasco Susitna Joint Venture.
Atkinson, S.C. 1982. Susitna Intergravel Temperature Report. Draft
Report. University of Alaska, Arctic Environmental Information
and Data Center. Anchorage, Alaska. Prepared for Acres
American, Incorporated.
Barrett, Bruce M. 1974. An Assessment Study of the Anadromous Fish
Populations in the Upper Susitna River \l'atershed between Devil
Canyon and the Chulitna River: Alaska Department of Fish and
Game, Division of Commercial Fisheries, 56pp.
1975a. December Investigations
River \\ratershed Bett\reen Devil Canyon and
Department of Fish and Game. Unpublished.
on the Upper
Chulitna River.
Susitna
Alaska
197Sb. January Investigations in the Upper Susitna River
\l'atershed Between Devil Canyon and Chulitna River. Alaska
Department of Fish and Game. Unpublished.
1975c. February Investigations
River Watershed Betv:een Devil Canyon and
Department of Fish and Game. Unpublished.
in the Upper
Chulitna River.
Susitna
Alaska
Bilello, Michael A. 1980.
Drainage Basin of the
Report 80-19, 30 pp.
A \Vinter Environmental Data Survey of the
Upper Susitna River, Alaska: CRREL Special
Bishop, Dan. 1974.
Belm.,r Devil 1 s
69 pp.
A Hydrologic Reconnaissance of the Susitna River
Canyon: for NOAA, U.S. Department of Commerce,
Cole, Terrence. 1979. The History of the Use of the Upper Susitna
River: Indian River to the Head1caters. July 1979.
Friese~ Nancy V. 1975. Pre-Authorization Assessment of Anadromous
Fish Populations of the Upper Susitna River \vatershed in the
Vicinity of the Proposed Devil Canyon Hydroelectric Project:
Alaska Department of Fish and Game, Division of Commercial
Fisheries, 121 pp.
Gatto, 1.\v., C.J. Harry, H.L. flcKim, and D.E. Lal<son.
Environmental Analysis of the Upper Susitna River Basin
Landsat Imagery. U.S. Army Corps of Engineers
Report 80-4.
6-26
1980.
Using
CRREL,
R&il Consultants, Inc., 198la.
1982). Prepared for Acres
Hydrographic Surveys.
American, Incorporated.
(2nd Report,
1981b. Ice
Reports: 1981-82,
Observations 1980-1981. (Subsequent Annual
1982-83). Prepared for Acres American,
Incorporated.
1981c. Preliminary Channel Geometry, Velocity, and Water
Level Data for the Susitna River at Devil Canyon. April.
Prepared for Acres American, Incorporated.
1981d. Processed Climatic Data Volumes 1 -8.
Reports for 6 Stations 1980-81; and for 8 Stations.
Prepared for Acres American, Incorporated.
Annual
1982.
1981e. Regional Flood Studies, December. Prepared for
Acres American, Incorporated.
1981f. Susitna River flile Index: rlouth to Susitna
Glacier. December. Prepared for Acres American, Incorporated.
198lg.
Annual Reports
Incorporated.
1982a.
(Supplement 1,
Incorporated.
1982b.
Acres American,
1982c.
Acres American,
Quality Annual Report 1980. (Subsequent
Acres American, 1982). Prepared for
Field Data Collection and Processing Volumes 1-3.
1982. Data). Prepared for Acres American,
Reservoir Evaporation. January. Prepared for
Incorporated.
Reservoir Sedimentation. January. Prepared for
Incorporated.
1982d. River ilorphology. January. Prepared for Acres
American, Incorporated.
1982e. Tributary Stability Analysis. Prepared for Acres
American, Incorporated.
1982f. Water Quality Interpretation 1981. February.
Prepared for Acres American_, Incorporated.
1983a. Field Data Index. (updated annually). Prepared
for Acres American, Incorporated.
1983b. Glacial Lake Studies Interim Report. Prepared
for Acres American, Incorporated.
R&i-1 Consultants, Inc. and Acres American, Inc. 1982. Hydraulic and
Ice Studies. rtarch.
6-27
R&N Consultants, Inc. and W.D. Harrison, 1981. Glacier Studies. (2nd
Report, 1982). Prepared for Acres American, Incorporated.
RMJ Consultants, Inc. and L.A. Peterson and Associates, 1982. \later
Quality Effects Resulting From Impoundment of the Sus i tna River.
December. Prepared for Acres American, Incorporated.
Riis, James C. 1975. Pre-Authorization Assessment of the Susitna
River Hydroelectric Projects: Preliminary Evaluation of \later
Quality and Aquatic Species Compositions: Alaska Department of
Fish & Game, Sport Fish Division, 61 pp.
1977. Pre-authorization Assessment of the Proposed
Susitna River Hydroelectric Projects: Preliminary Investigations
of \vater Quality and Aquatic Species Composition: Alaska
Department of Fish and Game, Sport Fish Division, 91 pp.
Riis, James C., and Nancy V. Friese. 1978.
Investigations of the Susitna River
Potential Impacts of the Devils Canyon
Projects, Alaska Department of Fish
Division, 116 pp.
Fisheries and Habitat
A Preliminary Study of
& \latana Hydroelectric
and Game, Sport Fish
Trihey, E.W., 1982a.
Salmon to Side
Preliminary Assessment of Access
Slough Habitat above Talkeetna.
Acres American, Incorporated.
by Spa\,ning
Prepared for
1982b. \;inter Temperature Study. Open File Report.
Prepared for Acres American, Incorporated.
1983. Preliminary Assessment of
Salmon into Portage Creek and Indian River.
Power Authority.
Access by Spmming
Prepared for Alaska
U.S. Army Corps of Engineers, Alaska District. 1975. Southcentral
Railbelt Area, Alaska. Upper Susitna Basin. Hydropm,er and
Related Purposes. Interim Feasibility Report 1975.
1978. Southcentral Railbelt Area, Alaska.
Basin. Hydropm,rer and Related Purposes.
Feasibility Reporr 1978.
Swan Lake (R.W. Beck & Associates)
Upper Sus itna
Supplemental
Federal Energy Regulatory Commission. 1980. Final Environmental
Impact Statement, S"an Lake Project No. 2911 -Alaska.
R.\L Beck. 1979. Application for License for the Swan Lake Project,
F.E.R.C. Project No. 2911, Ketchikan Public Utilities.
S\,Bll Lake Project, Evaluation Report, Volumes I & II.
Prepared for Ketchikan Public Utilities.
6-28
Terror Lake (International Engineering Company, Inc.)
Arctic Environmental Information and Data Center. 1979. An
Assessment of Environmental Effects of Construction of
Lake Hydroelectric Facility, Kodiak Island, Alaska.
Kodiak Electric Association.
the Terror
Report for
1980. An Assessment of Environmental Effects of
Construction and Operation of the Proposed Terror Lake
Hydroelectric Facility, Kodiak Island, Alaska; Raptor Studies,
Intragravel Water Temperature Studies.
1981. An Assessment of Environmental Effects of
Construction and Operation of the Proposed Terror Lake
Hydroelectric Facility, Kodiak Island, Alaska, Instream Flow
Studies. IV. J. lvilson et al.
Baldrige, J.R. and E.\v. Trihey. 1982. General Characteristics of
Surface \iater Temperatures in the Terror River Basin. AEIDC.
Prepared for Kodiak Electric Association.
Kodiak Electric Association.
Project -Kodiak, Alaska,
rlapping and Engineering.
1967. Terror Lake Hydroelectric
Definite Project Report, Vol. 1,
Robert W. Retherford Associates and International Engineering Company,
Inc. 1978. Terror Lake Hydroelectric Project, Kodiak Island,
Alaska. Definite Project Report.
Simons, Li & Associates, Inc. 1980a. Analysis
Sedimentation and rlorphological Changes in the
Terror Rivers Associated with the Terror Lake
Project. Report for Kodiak Electric Association.
of Hydraulic,
Kizhuyak and
Hydroelectric
1980b. Analysis of Thermal Changes in the Kizhuyak and
Terror Rivers Associated tvith Terror Lake Hydroelectric Project.
Report for Kodiak Electric Association.
Simons, D. B., .R.N. Li and J.R. Kinzey.
Hydrologic Analysis for Terror River
for Kodiak Electric Association.
Togiak (DOWL Engineers)
1980. Review and Extended
and Kizhnyak River. Report
D0\,71 Engineers.
for Togiak
1982. Volume E, Final
Hydroelectric Project.
Report. Reconnaissance Study
DD\IL Engineers. 1983. Feasibility Study for Togiak Hydroelectric
Project, Supplemental Data Report: Hydrology.
6-29
Tyee Lake (International Engineering Company, Inc.)
Arctic Environmental Information and Data Center. 1980. An Assessment
of Environmental Effects of Construction and Operation of the
Proposed Tyee Lake Hydroelectric Project, Petersburg and
Wrangell, Alaska.
1983. Tyee Lake Hydroelectric Project, Revised Fisheries
Hitigation Plan.
Federal Energy Regulatory Commission. 1981. Tyee Lake Hydroelectric
Project, FERC No. 3015, Alaska. Final Environmental Impact
Statement.
6-30
Acknowledgments
The authors wish to express their gratitude to the Alaska Power Authority
for ready access to reports in the Poh•er Authority library. The Power
Authority should be commended for its contribution in the development of a
broader ~,1a·ter resource data base for Alaska. The tvater resource data base
must continue to expand in order for professionals to complete tvater
resources analyses. In addition) the authors wish to thank R&M Consultants
for the clerical and drafting support provided for this paper. The authors
would appreciate being informed of any errors or omissions noted within the
paper.
6-31
References Cited
Balding, G.O. 1974.
U.S. Geological
\\1ater Reconnaissance Study of Pelican,
Survey.
Alaska.
Baldrige, J.E. and E.\V. Trihey. 1982. Potential Effects of T1w
Alternative Hydroelectric Developments on the Fishery Resources
of the Lower Tazimina River, Alaska. AEIDC, in cooperation ~.;ith
Dames & r!oore.
Bovee, Ken D. 1982. A guide to Stream Habitat Analysis Using the
Instream Flow Incremental r!ethodology. Instream Fl01v Information
Paper No. 12. \Ves tern Energy and Land Use Team, Fish and
liildlife Service, U.S. Dept. of the Interior.
Imberger, J. and J. Patterson. 1981. A Dynamic Reservoir Simulation
Hodel: DYRESI!5. In Transport Hodels for Inland and Coastal
\Vaters, H. B. Fischer, ed. Academic Press, New York.
Soil Conservation Service. Annual. Snmv Survey Heasurements for
Alaska. Anchorage, Alaska.
Tennant, D.L. 1975. Instream Flow Regimens for Fish,
Recreation and Related Environmental Resources. U.S.
Wildlife Service Report. Billings, r!ont. 18 pp.
\Vildlife,
Fish and
Tennant, D.L. 1976. Instream Fl01v Regimens for Fish, Wildlife,
Recreation, and Related Environment a 1 Resources. In Proc. Symp.
and Spec. Conf. on Instream Flow Needs, ed., J.F. Orsborn and
C.H. Allman. Vol. II, pp. 359-373. Amer. Fish. Soc., Bethesda,
i!d.
Thompson, K.E. 1972. Dete-rmining Stre.amflmvs
Proc. Instream Flmv Requirement Workshop,
Basins Comm., Portland, Oregon. pp. 31-50.
for Fish Life. In
Pacific N. \v. River
Thompson, K.E. 1974. Salmonids--Chap. 7. In Bayha, K. (ed.), The
Anatomy of a River. Report of the Hells Canyon Task Force,
Pacific N.IL River Basins Comm. pp. 85-103.
U.S. Geological Survey. Annual. \vater Resources Data for Alaska.
Anchorage, Alaska.
Wesche, T.A.
Hethods
and P.A. Rechard. 1980. A Summary of Instream Flow
for Fisheries and Related Research Needs. Eisenhower
Consortium Bulletin.
Wilson, IV.J., E.\V. Trihey, J.E. Baldrige, C.D. Evans, J.G. Thiele, and
D.E. Trudgen. 1981. An Assessment of Environmental Effects of
Construction and Operation of the Proposed Terror Lake
Hydroelectric Facility, Kodiak Island, Alaska. Prepared by AEIDC
for Kodiak Electric Association.
6-32
TROPHIC STATUS OF SUSITNA RIVER IMPOUNDMENTS
By Gary Nichols1 and Laurence A. Peterson2
Abstract
Summer inflow concentrations of carbon, silica, nitrogen and phosphorus may
be used to ~uantitatively predict the trophic status of lakes and reservoirs.
Among these nutrients, biologically available. nitrogen and phosphorus most
often control the eutrophication process depending on which of these nutrients
occurs in shortest supply. Several eutrophication models have been developed
for the prediction of lake trophic status based on phosphorus and nitrogen
loading values. The most widely recognized model is the Vollenweider--OECD
Program model which was developed from numerous data collected in a diversity
of limnological settings around the world.
The nitrogen:phosphorus (N:P) ratio calculated from water samples collected
at Vee Canyon ranged between 22:1 and 46:1, indicating that phosphorus is the
limiting nutrient in the Susitna River. Application of the Vollenweider--
OECD Program model to the Susitna Hydroelectric Project was accomplished by
incorporating measured phosphorus values with the mean depth and hydraulic
residence time at each reservoir. Based solely on nutrient enrichment, Watana
and Devil Canyon Reservoirs will be oligotrophic under natural conditions.
Additionally, high suspended solids concentrations and turbidity levels in
the Susitna River may limit the eutrophication process to a greater extent
than phosphorus concentrations. Furthermore, Watana and Devil Canyon will
maintain oligotrophic status if provided with a maximum additional phosphorus
load eg_uivalent to niore than 100:,uoo. permanent residents and 40 ,ooo·. permanent
residents, respectively. Additional loading from a 3000. person construction
camp would amount to a small fraction of the maximum permissible artificial
phosphorus load at each reservoir.
Introduction
The process of eutrophication is defined as the increase in nutrient enrichment
that causes increased productivity in lakes (Welch, 1980). This enrichment is
expressed in terms of nutrient supply or load. Nutrient supply is the concen-
tration of a nutrient per unit volume of water received by a lake expressed
in terms of mg/m 3 . Nutrient load on the other hand is the concentration of a
2 nutrient per unit of lake surface area expressed in terms of mg/m .
1
2
Principal, Nichols Consulting Service, 85 Pepperdine, Fairbanks, Alaska.
President, L. A. Peterson & Associates, Inc., 118 Slater Drive West,
Fairbanks, Alaska.
7-1
Lake trophic status is an expression of the degree to which the eutrophication
process has proceeded in a particular lake of a known mean depth, hydraulic
residence time, and annual inflow volume. The major characteristics used to
~uantify the trophic status of clearwater lakes are nutrient concentration,
algal biomass, and Secchi disc transparency.
Background
The ~uantitative prediction of reservoir trophic status at Watana and Devil
Canyon, resulting from the impoundment of the Susitna River, is based on the
following rationale.
The primary nutrients controlling algal growth include carbon, silica, nitro-
gen, and phosphorus. Among these nutrients, nitrogen and phosphorus most
often limit algal growth in freshwater systems (Shindler, 1977, Rast and Lee,
1978; Shaffner and Oglesby, 1978; Lee et al., 1978, Jones and Lee, 1982;
OECD, 1982; Smith, 1982). Since algal growth occurs rapidly over a short
period of time (Rast and Lee, 1978) and because certain forms of nitrogen
and phosphorus are unavailable to algal uptake (Lean, 1973, Shaffner and
Oglesby, 1978; Lee et al., 1978, OECD, 1982), it is more meaningful to con-
sider the biologically available forms of these nutrients than ~uantities
of total nitrogen and total phosphorus. The predominant form of phosphorus
which is readily available for algal growth consists of the dissolved ortho-
phosphate fraction (St. John et al., 1976; Lee et al., 1978, Welch, 1980;: Lee
et al., 1980; OECD, 1982). However, solubilization and mineralization reactions
in a waterbody may result in the formation of dissolved orthophosphate from
other phosphorus species. In many instances, the potentially bio-available
phosphorus concentration in clearwater lakes is a ~uantity that lies between
the concentrations of soluble orthophosphate and total phosphorus. It was
7-2
found (Lee et al., 1978, 1980) that the bio-available fraction in these lakes
could be approximated by adding the dissolved orthophosphate concentration
to 0 .'2 times the difference between total P and Po 4 . The form of nitrogen
which is readily available to algae consists of the inorganic (mineral)
fraction (Lee et al., 1978; Rast and Lee, 1978; OECD, 1982). The growth rate
of algae in natural fresh water is regulated by the single nutrient occurring
in shortest supply (OECD, 1982, Smith, 1982). A judgement on whether nitrogen
or phosphorus will limit algal growth can be made from considering the nitro-
gen:phosphorus (N:P) ratio. However, because nutrients are required in the
inorganic form for the purposes of algal growth, the use of inorganic nitro-
gen and dissolved orthophosphate ratios is often considered more meaningful
in determining thelimi ting nutrient, than total N and P ratios (OECD, 1982).
On the average, algal tissues contain nitrogen and phosphorus atoms in the
proportion of 16N:lP and require these nutrients in this proportion for growth.
When the N:P atomic ratio is greater than 16:1, phosphorus atoms are insuffic-
ient for algal growth, and algal biomass is limited by the quantity of phos-
phorus present. If on the other hand, the N:P atomic ratio is less than 16:1,
nitrogen becomes the limiting nutrient (Lee et al., 1978, OECD, 1982). More
important than the N:P ratio alone is whether nitrogen or phosphorus in a
water body are reduced to growth-limiting levels during the summer period of
eutrophication-related water quality concern (Rast et al., 1983). In some
water bodies, inorganic suspended solids and turbidity may have a greater
effect on algal growth than the N:P ratio (Smith, 1982).
Extensive eutrophication research in the last decade has resulted in the
development of several models for predicting the trophic state of phosphorus
limited lakes, solely on the basis of phosphorus enrichment. The phosphorus
concentration in a proposed impoundment may be predicted by applying external
7-3
(inflow) nutrient concentration data to the appropriate model. Through his
work with the Organization for Economic Cooperation and Development (OECD),
Vollenweider (1976) developed a generalmodel which quantitatively describes
the empirical relationship between the average areal load of phosphorus and
inlake phosphorus concentrations, from amongseverallakes in North America
and Europe. This relationship is expressed as:
L {Tw)
[P] = z(l+iTw)
where [P] =Average phosphorus concentration contained in a waterbody (Mg/m 3 ),
L =Areal phosphorus load (mg/m2 /time)
Tw =Hydraulic residence time (years)
z = Mean depth of a water body (meters).
The loading term (L) in this equation can be calculated by multiplying the
average inflow phosphorus concentration by tbe total volume of water received
by a lake during a specified period of time (i.e., m3 /yr), and dividing the
product by the surface area of the lake or reservoir.
This equation can also be expressed in terms of the average inflow concentra-
tion of phosphorus:
[Pl = Pi
( 1 + ~Tw )
where Pi 3 = Average inflow concentration of phosphorus (mg/m ) .
Based on the data from U.S. OECD study lakes, Lee et al. (1978) and Rast
and Lee (1978) substantiated this general relationship and defined it for a
number of waterbodies. Most recently, the OECD (1982) found that approxi-
mately 200'. waterbodies in 22 countries around the world follow the same
general relationship.
7-4
Jones and Lee (1982) concluded that this model has a wide-spread applicability
to most waterbodies--those in northern latitudes as well as southern latitudes,
lakes as well as impoundments. Furthermore, this model was successfully used
to estimate the annual total phosphorus input at Crescent Lake in south-cen-
tral Alaska (Koenings and Kyle, 1982). The U.S. Environmental Protection
Agency has suggested that this model be used as a basis for establishing
nutrient load criteria to U.S. waterbodies (EPA, 1976)
Statistical models developed by Dillon and Rigler (1974) and Larsen and
Mercier (1976) have the same technical foundation as the Vollenweider-OECD
models. However, the correlations for these models were not developed from
as broad a data base as those of the OECD load-response models.
Algal biomass is often a visible symptom of eutrophication, and it is
usually the cause of eutrophication-related water ~uality problems. Chloro-
phyll "a" concentration is widely recognized as the best expression of algal
biomass in lakes, and was selected as a principle trophic state indicator
by the OECD (1982). Several e~uations have been developed which express the
statistical relationship between phosphorus concentrations and chlorophyll
"a" concentrations in clearwater lakes. The most recent of these e~uations
are presented by Smith and Shapiro (1981), Jones and Lee (1982), OECD (1982),
and Rast et al. (1983).
' Lake trophic status can be generally classified on the basis of "fixed
boundary" phosphorus and chlorophyll "a" concentrations. Average in-lake
phosphorus concentrations of 0-'J.O mg/m 3 are indicative of oligotrophic
conditions, 10-20 ing/m 3 are in the mesotrophic range, and levels above 20
mg/m3 are considered eutrophic (Vollenweider, 1976). These conditions
correspond to average in-lake chlorophyll "a" concentrations of 0-'2 mg/m 3 ,
7-5
2-6 mg/m 3 , and greater than 6 mg/m 3 , respectively, in clearwater lakes.
Subsequently, Vollenweider determined that the critical surface load which
will result in oligotrophic conditions may be calculated by using the
maximum oligotrophic phosphorus concentration (10 mg/m 3 ) in the following
equation:
Lc =
where: Lc = Critical areal phosphorus load (mg/m 2 /time),
10 = Maximum inlake phosphorus concentration resulting in oligotrophic
status 3 (mg/m ) ,
z = Mean depth
Tw = Hydraulic residence time
Methods
Because of the wide-spread applicability of Vollenweider's (1976) model,
and its successful use in south-central Alaska, this model was selected for
use in the Susitna Project. At this time there are no koown limitations
to the application of Vollenweider's model in Alaska which are not common
to all models. Average inlake concentrations of phosphorus at Watana and
Devil Canyon were predicted by applying Vollenweider's equation to the
average summer phosphorus concentration in river water, the maximum atmos-
pheric phosphorus concentration at Fairbanks, Alaska, and the hydraulic
residence time at each reservoir. Subsequently, the average phosphorus
concentration in river water was multiplied by the average inflow volume of
water at each damsite (7.0148 X 109 m3 /yr at Watana and 7.9965 X 10 9 m3 /yr
at Devil Canyon) to derive the total phosphorus supply at each reservoir.
Upon dividing the supply by the surface area of each reservoir (153,786,00'0:
m2 at Watana and 31,566,600 m2 at Devil Canyon), areal phosphorus loading
7-6
(L) from the land was obtained.
Smith and Shapiro (1981) strongly recommend that the predictive use of
nutrient models be based on data collected during the summer months when
algal biomass is most closely related to nutrient concentrations. Accord-
ingly, nutrient concentrations in the Susitna River were determined from
samples collected during the summers of 1980 and 1981 at Vee Canyon by
R & M Consultants (1981). The Susitna Project location and Vee Canyon
sample station appear in Figure 1. Total nitrogen, inorganic nitrogen,
total phosphorus, and dissolved orthophosphate were analyzed according to
APHA (1981) methodology. The input of atmospheric phosphorus at Watana and
Devil Canyon was assumed to be approximately e~ual to the maximum phosphorus
concentration contained in rain and snow samples collected in Fairbanks,
Alaska by Peterson (1973). Conse~uently, the total natural phosphorus
load at each reservoir e~uals the sum of phosphorus from the land and from
precipitation.
All measured summer phosphorus concentrations at Vee Canyon were below the
detectable limit. However, "worst case" concentrations e~ual to the phos-
phorus detection limit were assumed because it was felt that values of zero
are inappropriate for the Susitna Project area. Conse~uently, the total
phosphorus and dissolved orthophosphate values assumed in this paper may
over-estimate actual concentrations contained in the Susitna River. The
biologically available fraction of nitrogen was determined by summing
concentrations of ammonia, nitrate, and nitrite nitrogen contained in each
sample. Dissolved orthophosphate was considered to be the biologically
available phosphorus fraction in the Susitna River.
7-7
FIGURE I
SUSITNA HYDROELECTRIC PROJECT AREA
t 0 10 20
N Scale in Miles
The hydraulic residence time of a unit volume of water (Tw) represents the
reservoir's water budget expressed as the total reservoir water volume (m 3 ),
divided by the annual inflow volume (m 3 /yr). Estimates of hydraulic residence
time at Watana and Devil Canyon were provided by R & N Consultants (1982a).
The mean depth (z) was calculated as the "full pool" volume divided by the
surface area at each reservoir. This is the same method used to determine
mean depth at Crescent Lake by Koenings and Kyle (1982).
Any additional phosphorus loading to Watana and Devil Canyon will cause a
subse~uent increase in the steady-state phosphorus concentration which may
result in a change in water ~uality. Therefore, artificial loading in the
form of domestic phosphorus inputs must be incorporated into the phosphorus
model if the capacity for residential dwelling or summer cottage development
is to be determined. Results from 13 studies in North America and Europe
concluded that the average per capita contribution of phosphorus (excrement
plus household waste) is 800,000 mg/yr from domestic sources (Dillon and
Rigler, 1975). By dividing the average per capita supply by the surface
area of each reservoir, average per capita surface phosphorus loading was
obtained. The maximum permissible artificial load was calculated as the
difference between the critical surface load and the natural surface load
at each reservoir. The permissible number of permanent (year-round) residents
at each reservoir was obtained by dividing the permissible artificial load
at each reservoir by the corresponding average per capita surface phosphorus
load. The maximum number of "permanent" dwelling unit e~uivalents around
each reservoir was calculated by dividing the number of permissible residents
by the number of residents at each dwelling unit. In the event that dwelling
units will be (on the average) occupied for less than 365 days per year
7-9
(i.e., summer cottages), the permissible number of "seasonal" units will
e~ual the number of permanent dwelling units multiplied by 365 days, divided
by the average number of days spent at each unit per year.
Chlorophyll "a" and Secchi disc transparency data from the Susitna River
are unavailable. However, high suspended sediment and turbidity levels at
Vee Canyon indicate the Chlorophyll "a" concentrations will be low in Watana
and Devil Canyon. Conse~uently, prediction of chlorophyll "a" concentrations
and Secchi depth transparencies following impoundment have been disregarded
in this paper.
Results
During the summer of 1980 and 1981, the N:P atomic ratio at Vee Canyon ranged
between 22:1 and 46:1. Thus, among the nutrients considered to be important
to algal growth, it is apparent that phosphorus is the limiting nutrient.
Average summer total phosphorus and dissolved orthophosphate concentrations
measured at Vee Canyon were below the detection limit (0.05 mg/1 and O.Ol mg/1,
respectively) of the analytical method used. Upon conversion of these values,
the "worst case" average total phosphorus concentration is 50 :mg/m3 , and the
worst case average dissolved orthophosphate concentration is 10 mg/m 3 . The
maximum phosphorus concentration measured in precipitation at Fairbanks,
Alaska was 0.03 mg/1 (Peterson, 1973). 3 Conversion of this value (30 mg/m ) ,
in combination with the normal average precipitation at Talkeetna, Alaska,
2 indicates that the natural phosphorus load from precipitation will be 22 mg/m
at both reservoirs. Reservoir mean depths were determined to be 76 meters
at Watana and 43 meters at Devil Canyon (R & M Consultants, 1982b). Sub-
se~uently, the hydraulic residence time was estimated to be 1.64 years at
Watana and 0.16 year at Devil Canyon (R & M Consultants, 1982a).
7-10
With respect to the bio-available phosphorus fraction (dissolved orthophos-
phate), the predicted summer inlake phosphorus concentration [P] e~uals 4.5
3 3 mg/m at Watana and 6.8 mg/m at Devil Canyon. These values correspond to
phosphorus loading values (L) of 478 mg/m 2 /yr and 2555 mg/m 2 /yr, respec-
tively. Inlake concentrations of bio-available phosphorus [P] at Watana
and Devil Canyon plot in the same area as oligotrophic waterbodies with
similar areal phosphorus loads, mean depths, and hydraulic residence times
(Figure 2). Thus, assuming that dissolved orthophosphate represents the
only bio-available phosphorus fraction in the Susitna River, both of the
proposed reservoirs will be oligotrophic.
Although both reservoirs initially will be oligotrophic, artificial loading
from domestic sources could cause a shift in trophic status at one or both
reservoirs at some future time. Because of this concern, an analysis of
artificial loading was made. The average per capita artificial phosphorus
load will be 0.005 mg/m 2 /yr at Watana and 0.02.5 mg/m 2 /yr at Devil Canyon.
The maximum permissible artificial load was calculated to be 579 mg/m 2 /yr at
~iatana and 1208 mg/m2 /yr at Devil Canyon with one dam in place. The loading
at Devil Canyon could be higher if Watana is in place because Watana may act
as a nutrient trap. Upon dividing the permissible artificial load by the
average per capita load at each reservoir, Watana will accomodate 115,800
and Devil Canyon 48, 300'. permanent residents, respectively. If an average
of three individuals occupy each dwelling unit for the entire year, the
maximum permissible number of dwelling units will be 38,600. ·at Watana and
16 ,100' ·at Devil Canyon. If permanent and seasonal dwellings are con-
structed, the domestic load should not exceed the amount generated by
115,800: ·permanent residents at Watana, or 48,300: 'permanent residents at Devil
7-11
IO,OOOr~------------r-1 -----------.1 ------/------,
/ /
1000 ;:-
~ -r
f-
EUTROPHIC
/ /
/ /
/ / / v /
/ ~/. / o« / / "'-'~:>/ Devil
/ '?0 / Co nyon // _ ... <v /
/ "'" / / /
/ /
/ /-
/ / "' / / --/ Wotan a _... / ---/ _... .....
/ -
IOOF ----
OLIGOTROPHIC
-
-
I I I II II " I I I .II II I I I I II II I lo.o~--~~~~~--~~~~~~--~~~-U~
...
>,
' C\1
E
' a.
"' E
"' c:
"' c
0
....1
"' ::l ...
0
.t::; a.
"' 0
.t::;
a.
1.0 10 100 1000
(Mean Depth, z) / (Hydraulic Residence Time 1 Tw) 1 m I y r
FIGURE 2
SUSITNA PROJECT DATA APPLIED TO VOLLENWEIDER'S MODEL
7-12
Canyon, if oligotrophic conditions are to be maintained.
Artificial loading from a 300'0' :person construction camp would amount to 15
2 2 mg/m /yr at Watana and 75 mg/m /yr at Devil Canyon. These loading levels
represent about 3 percent (Watana) and 6 percent (Devil Canyon) of the
maximum permissible artificial loading re~uired to maintain oligotrophic
conditions .
Discussion
The aforementioned trophic status predictions are dependent upon several
assumptions that cannot be ~uantified on the basis of existing information.
These assumptions include:
(1) the N:P ratio does not fluctuate in subse~uent years to the extent that
a nutrient other than phosphorus becomes limiting,
(2) no appreciable amount of bio-available phosphorus is released from soil
upon filling of the reservoir (i.e., internal loading),
(3) estimates of land and precipitation phosphorus concentrations are accurate,
(4) phosphorus input levels are constant throughout the algal growth period
(summer),
(5) phosphorus concentrations measured at Vee Canyon correspond to the time
of peak algal productivity,
(6) an appreciable fraction of the total phosphorus pool is not converted
to dissolved orthophosphate,
(7) hydraulic residence times are constant,
(8) phosphorus losses occur only through sedimentation and the outlet,
(9) the net loss of phosphorus to sediments is proportional to the amount
of phosphorus in each reservoir, .and,
(10}. steady-state conditions prevail in both reservoirs.
7-13
Among the models 1<hich predict lake trophic status, the Vollen1feider ( 1976)
model appears to be the most reliable and most 1fidely applied. Ho1<ever, this
and other models 1<ere developed from data collected in clear1<ater lakes 1<here
nutrient concentrations and chlorophyll "a" concentrations are statistically
related. For the proper application of the Vollen1feider-OECD models, only
a moderate amount of non-algal turbidity should be present (Rast et al., 1983).
Haterbodies containing a high inorganic particulate load or large amounts of
suspended solids should be expected to contain chlorophyll "a" concentrations
lo1<er than 1fhat these models predict (Jones and Lee, 1982). As suspended
sediment concentrations increase, conversion of initially available phosphorus
to unavailable forms increases (Lee et· al., 1978). Koenings and Kyle (1982)
report that during summer, soluble inorganic phosphorus 1<as largely converted
to particulate phosphorus in the epilimnion of Crescent Lake. They concluded
that the nutrient dynamics in glacially influenced lakes having high silt inputs
are different from those of clear1<ater lakes. ConseQUently, trophic status
cannot be predicted solely by the nutrient concentration in a 1fater body --
phosphorus does not adversely affect 1<ater QUality unless it produces un-
desirable aQuatic plant gro1fth (Rast et al. , 1983).
In the event that only one of the Susitna Project reservoirs is constructed,
lo1< light penetration levels o1<ing to high concentrations of suspended glacial
flour 1<ill likely limit algal biomass to a greater extent than phosphorus
concentrations. Furthermore, high suspended sediment concentrations may
convert available phosphorus to unavailable forms. If both reservoirs are
constructed, a large fraction of the suspended sediment load entering Hatana
1<ill settle to the bottom resulting in higher light penetration levels do1fn-
stream at Devil Canyon. In this instance, light may not become a limiting
7-14
factor to algal growth at Devil Canyon. However, a significant portion of
phosphorus may also be trapped at Watana resulting in lower phosphorus
concentrations downstream at Devil Canyon. Typically, 80-'90 percent of the
phosphorus entering lakes and impoundments is incorporated into their sedi-
ments, and will not be available for stimulation of algal growth in downstream
waters (Lee et al. , 1978).
Because suspended solids concentrations are high during the open-water season
at Vee Canyon, and because actual phosphorus concentrations will be less
than the worst case concentration used in our calculations, trophic status
(oligotrophic) predictions at each reservoir probably overestimate the
degree to which the eutrophication process will proceed under natural conditions.
Conse~uently, the maximum allowable number of permanent residents at each
reservoir may be greater than those determined on the basis of trophic status.
Summary
Reservoir trophic status is determined in part by relative amounts of nitrogen
and phosphorus present in a system as well as the ~uality and ~uantity of
light penetration. The N:P ratio indicates which nutrient limits algal
productivity. The nutrient which is least abundant will be limiting. On
this basis, it was concluded that phosphorus is limiting in the Susitna
impoundments. Vollenweider's (1976) model was considered to be the most
reliable in determining phosphorus concentrations in the Watana and Devil
Canyon impoundments. However, because the validity of this model is based
on phosphorus data from clearwater lakes, predicting trophic status of
silt-laden waterbodies with reduced light conditions and high inorganic
phosphorus levels may over-estimate actual trophic status. The summer
7-15
phosphorus concentration is considered the best estimate of trophic status
in phosphorus limited lakes. Bio-available phosphorus is the fraction of
the total phosphorus pool which controls algae growth in a particular lruce.
The measured dissolved orthophosphate concentration at Vee Canyon was con-
sidered to be the best estimate of the bio-available phosphorus fraction in
the Susitna River. Accordingly, average summer dissolved orthophosphate
was multiplied by the average inflow at each reservoir to calculate summer
phosphorus supplies from the land. These values were in turn combined with
atmospheric phosphorus values and divided by the surface area of each impound-
ment. The resultant summer phosphorus loading values at Watana and Devil
Canyon were below the maximum loading levels required for the maintenance
of oligotrophic conditions. Likewise, upon incorporating summer loading
values into Vollenweider's (1976) phosphorus model, the volumetric spring
phosphorus concentration at both reservoirs fell into the same range as
oligotrophic lakes with similar mean depths, detention times, and phosphorus
loading values.
Acknowledgements
This work was performed under contract toR & M Consultants, Inc. and
Acres American Incorporated, as part of the Susitna Hydroelectric Project
feasibility studies for the Alaska Power Authority.
7-16
References
APRA, 1981. Standard Methods for the Examination of Water and Wastewater.
15th Edition, American Public Health Association, Wash. D.C. 1134 pp.
Dillon, P.J., and F.H. Rigler, 1974. A test of a simple nutrient budget
model predicting the phosphorus concentration in lake water. Journal
Fish. Res. Board Canada, Vol. 31, No. 11, pp. 1771-1778.
____
7
, 1975. A simple method for predicting the capacity of a lake for
development based on lake trophic status. Journal Fish. Res. Board
Canada, Vol. 32, No. 9, pp. 1519-1531.
EPA, 1976. Quality criteria for Water. U.S. Environmental Protection
Agency, Washington, D.C., 256 pp.
Jones, R.A., and G.F. Lee, 1982. Recent advances in assessing impact of
phosphorus loads on eutrophication-related water ~uality. Journal Water
Research, Vol. 16, pp. 503-515.
Koenings, J.P., and G.B. Kyle, 1982. Limnology and fisheries investigations
at Crescent Lake (1979-1982), Part I: Crescent Lake limnology data summary.
Alaska Dept. of Fish and Game, F.R.E.D. Div., Soldotna, Alaska, 54 pp.
Larsen, D.P., and H.T. Mercier, 1976. Phosphorus retention capacity in lakes.
Journal Fish. Res. Board Canada, Vol. 33, pp. 1742-1750 •.
Lean, D.R.S., 1973. Movements of phosphorus between its biologically important
forms in lake water. Journal Fish. Res. Board of Canada, Vol. 30,. pp.
1525-1536.
Lee, G.F., W. Rast, R.A. Jones, 1978. Eutrophication of water bodies: insights
for an age-old problem. Environmental Science and Technology, Vol. 12,
pp. 900,..:908..
Lee, G.F., R.A. Jones, and W. Rast, 1980. Availability of phosphorus to
phytoplankton and its implications for phosphorus management strategies,
In: phosphorus management strategies for lakes, R.C. Loehr, C.S. Martin,
and W. Rast (eds.). Ann Arbor Science Publ. Inc., pp. 259-308.
OECD, 1982. Eutrophication of waters: monitoring, assessment, and control.
Organization for Economic Co-operation and Development, Paris, France,
154 pp.
Peterson, L.A. , 1973.
characteristics of
Alaska, Fairbanks ,
An investigation of selected
two subarctic streams. M.S.
185 pp.
physical and chemical
Thesis, University of
R & M Consultants, 1981. Alaska Power Authority Susitna Hydroelectric
Project task 3 -hydrology subtask 3.03 -water ~uality 1981 annual
report prepared by R & M Consultants Inc. for Acres American Inc.
_____ , 1982a. Reservoir sedimentation studies. Susitna Hydroelectric Project
7-17
feasibility report, Anchorage, Alaska, January.
_____ , 1982b. Personal communication to Larry Peterson, L. A. Peterson &
Associates, Inc., Fairbanks, Alaska.
Rast, W., and F. Lee, 1978. Summary analysis of the North American (U.S.
portion) OECD eutrophication project: nutrient loading -lake response
relationships and trophic state indices. Env. Research Lab., Office
of Research and Development, U.S. EPA, Corvallis, Oregon, 455 pp.
Rast, W., R.A. Jones, and G.F. Lee, 1983. Predictive capability of U.S.
OECD phorphorus loading-eutrophication response models. Journal
HPCF, Vol. 55, No. 7, pp. 990 -1003.
Shaffner, W.R., and R.T. Oglesby, 1978. Phosphorus loading to lakes and
some of their responses. Part I. A new calculation of phosphorus
loading and its application to 13 New York lakes. Limnology Oceano-
graphy, Vol. 23, No. 1, pp. 120 '-134.
Shindler, D.W., 1977. Evolution of phosphorus limitation in lakes. Science,
Vol. 195, pp. 260 -262.
Smith, V.H., 1982. The nitrogen and phosphorus dependence of algal biomass
in lakes: an empirical and theoretical analysis. Limnology Oceano-
graphy, Vol. 27, No. 6, pp. 1101. -1112.
Smith, V.H., and J. Shapiro, 1981. Chlorophyll-phosphorus relations in
individual lakes. Their importance to lake restoration strategies.
Environmental Science and Technology, Vol. 15, No. 4, pp. 444 -451.
St. John, B.E., E.C. Carmack, R.J. Daley, C.B.J. Gray, and C. H. Pharo, 1976.
The limnology of Kamloops Lake, B.C. Dept. of Environment, Inland
Waters Directorate, Pacific and Yukon Region, Vancouver, B.C., 167 pp.
Vollenweider, R.A., 1976. Advances in defining critical loading levels for
phosphorus in lake eutrophication. Mem. Ist. Ital. Idrobiol., 33,
pp. 53 -83.
Welch, E.B., 1980 .. Ecological effects of waste water. Cambridge University
Press, New York, 337 pp.
7-18
El\VIRONNENTAL EFFECTS OF ICE PROCESSES
ON THE SUS ITNA RIVER
1 ? G. Carl Schoch , and Stephen R. Bredthauer-
Observations of ice processes on the Susitna River have been conducted
since 1980 in conjunc-tion with planning of the proposed Susitna
Hydroelectric Project. The observations describe baseline conditions, and
are providing data to better assess post-project conditions. The processes
of ice generation, staging, ice jamming; and breakup on the Susitna River
are presented, as well as the effect of these processes on turbidity, river
morphology, nquatic habitat_, 1dldlife, and vege·tation.
Construction of the proposed \Vatana and Devil Canyon dams \Vill
significantly alter floh7 , thermal, and ice conditions on ·the Susitna River.
Frazil ice generated in the upper basin will be trapped by the \Vatana
Reservoir. Relatively t-.:arm ~vater (2°C -4°C) will be released in large
volumes from the reservoirs during ~vinter. Ice formation dmvnstream of the
project will be significantly delayed. Breakup processes will be altered
by the reservoirs ·trapping ice from upstrean1, by the changes in downstream
ice conditions, and by reduced flm.;s during breakup. The projected changes
and their environmental impacts are also described.
Introduction
Tbe proposed Susitna Hydroelectric Project \oou1d be one of the largest
hydroelectric projects ever located on a northern river, consisting t\.;o
dams wi·th a total capacity of 1620 ffl'J. \Vat ana Reservoir, the upper
project, \~'auld be 48 miles long, cover 38,000 acres, and have a storage
capacity of 9.5 million acre-feet. The lm~'er reservoir, Devil Canyon,
\\'Ould b0. 26 mile.s long, cover 7,800 acres, and store about a 1.1 million
acre-foot.
1 Hydrologist, R&~I Consultants, Inc.
2 Chief, Hydrology Department, R&N Consultants, Inc.
8-1
Because of the proposed projects, the study of ice on the Susitna River has
been ongoing since the winter of 1980. Initially, the intent was to target
locations of specific ice processes such as frazil ice generation, shore
ice constric-tions, ice bridges, and ice jams (R&H 198lb, 1982d). Renewed
emphasis by environmental concerns on potential modifications to the river
ice regime by hydroelectric power development resulted in a more refined
ice program for 1982-1983, directed towards answering specific problems of
the Susitna River. Staging, ice cover development and ice jams, together
with impacts of ice on river morphology and aquatic habitat, are among the
topics discussed in this paper. General processes will be described, as
will events observed in 1982-1983. Comparisons will be made to processes
observed in earlier years.
Description of Basin
The Susitna River drainage basin, sixth largest in Alaska, is located in
the Cook Inlet subregion of southcentral Alaska (Figure 1). The drainage
basin covers 19,600 square miles. It is bordered on the west and north by
the Alaska Range, on the east by the Talkeetna Hountains and the Copper
River lowlands, and on the south by Cook Inlet. The river is 320 miles
long from the mouth at Cook Inlet to tbe headwaters at Susitna Glacier.
Major tributaries include the Chulitna, Talkeetna, and Yentna Rivers, all
located downstream of the proposed project. Extensive glaciers in the
headwaters contribute substantial suspended sediment loads during summer
months. Streamflo<< is characterized by high flows between flay and
September and low flows from October to April.
8-2
ALASKA
LOCATION MAP
a DENALI
0 TALKEETNA
PRINCE WILLIAM
SOUND
20 0 20 60
SCALE IN MILES
SUSITNA HYDROELECTRIC PROJECT LOCATION MAP
FIGURE 1
8-3
The head~~·aters of the Susitna River and the major upper basin tributaries
are characterized by brood, braided, gravel floodplains below the glaciers
of the Alaska Range. Belo\v the \Ve-_st Fork confluence, the river develops a
split-channel configuration with numerous gravel bars, flowing south
between narrow bluffs for about 55 miles. Below the confluence with the
Tyone River, the Susitna River flo~vs \\·est for 96 miles through steep-,\ralled
canyons before reaching the mouth of Devil Canyon. This reach contains the
\~1 a·tana and Devil Canyon damsites at River Hiles (R}1) 184.4 and 151.6,
respectively_, measured from Cook Inlet. River gradients are quite high,
averaging nearly 1.4 feet/mile in the 54 miles above Watana damsite,
10.4 feet/mile from Watana downstream to Devil Creek, and 31 feet/mile in
the 12-mile stretch bet\veen Devil Creek and Devil Canyon. Below Devil
Canyon, the gradient decreases from about 14 feet/mile to 8 feet/mile above
TaH.eetna. The river in this reach is generally characterized by a
split-channel configuration, \v·ith numerous side-channels and sloughs.
About 4 miles above the confluence wi·th the Chulitna River, the Susitna
River begins to braid, and remains braided the remainder of its length to
Cook Inlet. Numerous islands and side channels appear. The gradient
continues to decrease, ranging from 5. 5 feet/mile for the 34-mile reach
below Talkeetna to 1.6 feet/mile for the last 42 miles (Figure 2).
Basin Climate
The Susitna River originates in the continental climatic zone, flowing
south into the transitional climatic zone. Due to the maritime influence
and the lower elevations 1 temperatures are more moderate in the lower basin
8-4
~
" ~
"' w
~
" "' w
E
en ~ I >
\n 0
.Q
"' ~ w
!
z
0
1-
< > w
-1 w
BOO
600
400
200
CD
"' u
0
SUSJTNA-YENTNA\
KASHWITNA CREEK\
CONFLUENCE •
CD
"' ;;
0
"' > 0> 0
"' z u
0
2'.1 ml/d
......-1 mild~
.....
>
0 z
>
0
0> z
>
0 z
1.2 mild
2.2 ml/d
"' "' >
0 z
.6 mild
\GOLD CREEK
"'--SHERMAN
'-..sLOUGH 9
\.__SLOUGH a
'\_CURRY
RABIOEAUX CREEK~
\
' \CHUUTNA-SUSITNA CONFLUENCE
"" BIRCH CREEK SLOUGH
'-sUNSHINE
\ " ~ONT ANA CREEK
\..GOOSE CREEK SLOUGH
CONFLUENCE \
0-·~~~~~~~~~~~~~
20 40 60 80
RIVER MILE
100 120 140
SUSITNA RIVER ICE LEADING EDGE PROGRESSION RATES (miles/day) RELATIVE
TO THE THALWEG PROFILE FROM RIVER MILE 0 (Cook Inlet) TO RIVER MILE 155
FIGURE 2
160
than in the upper basin. Freezing temperatures occur in the upper basin by
mid-September, with frazil ice generated in the reach from Denali through
Vee Canyon by car ly October.
Several meteorological stations have been installed along the river since
1980. Records from these stations, loca·ted at Susitna Glacier, Denali,
Kosina Creek (bet~\reen Vee Canyon and \\rat ana), \{atana, Devil Canyon and
Sherman, together \Vith records from the National \\Teather Service at
Talkeetna, illustrate the sharp difference in freezing degree-days along
the length of the river (Figure 3). In general, the meteorology within the
upper Susitna River basin is highly variable between weather station sites.
This is due. in part, to the movement of storm systems, the topographic
variance, and the change in latitude, but the major reason for the
temperature variance between Denali and Talkeetna is the 2,400-foot
elevation difference.
Freeze-Up Processes
Development of an ice cover on the Susitna River is a complex process
influenced by many variables and mechanisms that are not fully understood.
The ice on this ri\"er is primarily a continuous accnmulation of frazil
slush and sno~v slush called 11 hununocked . " 1ce (Nichel, 1971). It is
therefore important to understand the relationship and significance of air
temperature, water temperature, turbulence, snowfall and suspended sediment
to frazil ice generation. These relationships will first be discussed,
8-6
()
U)
>
3000
«!: 2000
!:;!
i'<
UL w 1000
a: u.
500
0 SEP
AVERAGE HISTORICAl
ACCUMULATED FREEZING DEGREE DAYS
fOR SUSITNA ROVER BASIN
METEOROLOGICAL STATIONS
1980-1983
NOV DEC JAN FEB
MONTH
FIGURE 3
8-7
(NWS l
MAR APR
followed by a description of freeze-up characteristics of specific reaches
of the river from Cook Inlet to Devil Canyon.
a. Frazil Ice Generation
Frazil ice crystals are formed when water becomes supercooled (Ashton,
1978; rlichel, 1971; Ne~Cbury, 1968; Oste.rkamp, 1978). Supercooling is
a pl1enomena by which water remains in a liquid state at temperatures
below 0°C. Foreign particles are associated tvith the nucleation of
ice crystals (Osterkamp, 1978). The Susitna River discharges
tremendous volun1es of silt and clay size particles prior to freeze-up.
There is an apparent correlation bet\~reen the first occurrence of
frazil ice and a sudden reduction of turbidity in the river water,
indicating that the fine suspended sediments may initiate the
nucleation of ice (R&~J, 1983). Once the river is at the freezing
point, snowfall also contributes to the total slush ice discharge.
With sustained air temperatures below 0°C, a thin layer of water will
be cooled to the freezing point and ice crystals tvill fol"m. Linder
quiescent conditions, the ice crystals will form on the water surface,
eventually bonding together into a sheet of black icE, i:J11d continuing
to grow vertically a 1 ong t be. thPrma l gradient. However, laboratory
experiments have determined that flow velocities of only 0.79 ft.jsec.
are necessary to mix the surface layer sufficiently to produce frazil
(Osterkamp, 1978). These velocities are exceeded on the Susitna
mainstem through most reaches so that the water body is continually
being mixed. Under these conditions, the water can be supercooled to
8-8
several hundredths of a degree below 0°C throughout the water column,
and crystals of frazil ice form in suspension beneath the water's
surface. Once the frazil ice forms, it has a tendency to rise to the
surface. However, during the initial ice formation, frazil particles
are so small that they remain entrained in the river due to
turbulence.
Channel morphology can play an important role in concentrating frazil
ice, as indicated by ice plumes. These plumes are an early indicator
of frazil ice and have been observed at several locations bet\\1een
Talkeetna and Vee Carryon when otherwise no ice ~vas seen. Nest sites
occur at sharp river bends caused by outcrops protruding into the
channel. The rock outcrops often create a slight backh'ater effect on
the upstream side. Suspended frazil floes are swep·t into these areas
and S\.Jirl about, increasing in density and ice concentration until
sufficient buoyancy is obtained so that the ice rises to the surface
as slush. The slush floats past the outcrop in a long narrmv stream
which is rapidly dissipated by the river. Any subsequent turbulence
can re-entrain the slush, once again making it difficult to observe.
In September these ice plumes are often observed near Gold Creek and
Sherman. The. f1ow patterns are such tha-t the.sc~ sites concentrate ice
t]Jroughout freeze-up.
After November, the majority of frazil ice is generated in the rapids
of Devil Canyon, \\7atana Canyon and Vee Canyon. However, during the
initial freeze-up period in October 1982, the difference in the number
of freezing degree days between Denali (370) and Talkeetna (170)
8-9
suggests that the majority of the slush accumulating against the
leading edge downstream of Talkeetna originates either as snowfall or
as frazil in the upper river from Vee Canyon on upstream. This
appeared to be verified during a flight on October 21, 1982.
Estimates at various locations from Talkeetna to Watana Creek showed a
consistent ice discharge in this reach, indicating that no frazil ice
was being generated at ·tlte rapids at Devil Canyon and Watana on this
date.
b. Cook Inlet to Chulitna Confluence
Temperatures are usually not cold enough to cause significant shore
ice develop1nent in this reach prior to the relatively rapid advance of
the ice cover. The initiation of ice cover formation in this reach
usually occurs when trPmendous volumes of slush ice fail to pass
through a channel constrict ion near the river mouth at Cook Inlet.
Bet\~een October 22 and October 26, 1982, slush ice jammed and
accumulated upstream for 57 miles. Daily ice discharge estimates from
Talkeetna showed a sudden increase in ice concentra-tions during this
ppri._od. The ice. discharge on October 21 tvas estimated at 1.3 x 10 5
cu ft/l!r and rose steadily to . 5 5.8 X 10 cu ft:/hr on October 26
follmving several snmv storms. Assuming that the ice cover began
progressing upstream on October 22., then the progression rate was 11.5
miles per d11y.
8-10
As the ice cov·er moved upstream in 1982, staging rarely exceeded
2 feet in this reach. (Staging is a process by ~d1ich the ice cover
thickens_, restricting flo\\r and causing increased stages upstream of
the ice front. This lowers the upstream velocity so that incoming ice
may accumulate against the leading edge instead of being S"l\'ept under
the ice cover.) Large open \~ater areas appeared frequently in the ice
pack. Surprisingly little consolidation of the ice pack had taken
place by October 26, 1982. This could be due to the shallm< gradient
of the channel through this reach. In lm.;r velocity areas, the ice
front continued to advance by juxtaposition of ice floes at a rate
proportional to the ice discharge and channel configuration. Slush
ice observed at the le.ading edge \.;ras not submerging under ·the existing
ice cover. All of the major tributaries to the Susitna below
Talkeetna \~ere still flowing and remained ice-free during this period.
The dischi:lrge from these tributaries kept large areas at their
confluences free of ice. Near Talkeetna, the river remained free of
shore ice even though a large volume of slush ice from the Susitna
River \vas continually drifting dow·nstream. No telescoping of the ice
cover was evident. The ice pack remained in the narrmv thalweg
channel, which in nlost areas constitutes only 20 percent of the flat,
broad river clwnnel. Gravel islands rt.~mained above the '\"ater surface,
although the staging did divert water into some side channels.
Staging effects were larger in channels near Talkeetna. On November 2
a staff gage at Talkee-tna was dry, with the nearest open water more
than 1 foot below the gage. After consolidation and freezing of the
ice pack about November 17, the gage had a reading of 3. 6 feet, a
8-11
stage increase of over 4 feet at Talkeetna due to the ice cover
advtmce.
After the initial ice cover formation, the remainder of the lower
river free.ze-up process requires considerably more time. fJany of the
side channels that are flooded by the increased stage in the mainstem
gradually become narro\\7 er as shore ice layers build up along the
chartnel banks and the flow discharge decreases. The gradual reduction
of flow during winter causes the ice cover to settle. Where the
sagging ice beco1nes stranded, it conforms to the configuration of the
channel bottom and create.s an undulating ice surface. Open water
areas persist through ~larch in high velocity zones. Some
side-channels and sloughs may also receive a thermal influx from
ground~,·ater upwelling sufficient to keep these channels ice-free.
c. Ch11litna Confluence
A slush ice bridge which forms at the confluence of the Chulitna and
Susitna Rivers initiates the ice cover progression on the Susitna
River above tbjs point. This bridge has been observed each \\;inter
during t.he :ice observation program, occurring on November 2 in 1982,
but at later dates in 1980 and 1981. The processes described in this
section were observed in 1982, but similar processes have occured in
other years.
8-12
The Susitna River contributes approximately 80 percent of the slush
ice at the confluence area near Talkeetna_, t\rhile the Chulitna and
Talkeetna Rive1·s combined produce the remaining 20 percent. The high
velocities (4-5 ft/sec) of the Susitna keep the river channel open and
push the slush ice downstream. After entering the confluence area,
the masses of slush ice slot.: dm..,rn and begin to pile up at the south
bend of the Susi·tna adjacent to the entering east channel of the
Chulitna. On October 18, 1982, slush \Vas still moving easily through
this area but \VBS covering all of the open \Vater for about 600 feet
with a translucent sheet of slush ice. By October 29_, the compressed
slush h'BS still moving. The ice through this area was nol' wl1ite
instead of translucent, since the slush had consolidated sufficiently
to rise higher out of the tvater and partially drain.
A snmv storm immediately preceded the formation of the stable ice
bridge at the Sus i tna and Chu 1 itna confluence on November 2, 1982,
This storm caused a substantial local increase in ice discharge which
could not pass through the channel. The result was a sudden
consolida-tion of the ice cover, compacting the slush tvhich at some
point became shore-fas-t. The cover remained stable long enough to
freeze and increase in thickness. The majority of the incoming slush
ice floes tbt::n accumu 1 i:Jted against the leading edge, causing the cover
to begin advancing upstream. Approximately 10-20 percent of the
incoming slush ice submerged on contact tvith the upstream edge, either
adhering under the ice cover or continuing downstream. Ice discharge
estimates at Talkeetna were substantially lower after November 2. The
mast dramatic effect of the ice consolidation at ·the confluence t~~as
8-13
the flooding on the Susitna just upstream of the confluence. The flmv
capacity of the ice-choked main channel was greatly reduced. \\1ater
spilled from underneath the cover, flowing laterally across the river
channel to\\rards the opposite (north) bank. \Vater '\ras also diverted
from upstret-lm of the ice jam, flowing into the nmv channel. These
diverted flm\7 S combined and entered the Chulitna east channel
approximately 1,500 feet upstream of the original confluence. The
total estimated discharge of the diverted flow was 700-1000 cfs, about
15-20 percent of the total flow. Substantial channel erosion ~\ras
caused by these diverted flows, as subsequent depth measurements
through the ice located a isolated channel about 700 feet from the
left bank.
After the confluence bridge formed the ice pack advanced slowly up the
Susitna River. The river gradient begins to increase at this point.
Slush ice can no longer accttmulate by simple juxtaposition, as the
high flmv velocities submerge the slush on co1rtact with the leading
edge. Staging levels of 2-4 feet are necessary before ice can
continue accumulating against the upstream edge.
The processes of ice cover telescoping, sagging, open lead development
and secondary ice cover progre_ssion are important characteristics
through this reach. Telescoping occurs during consolidation of the
ice cover. When the velocity at the leading edge is subcritical for
ice progression, ice floes drifting do~\7 nstream will contact the edge,
remain on the surface, and accumulate upstream by juxtaposition at a
rate propor·tional to the concentration of slush ice and to the channel
8-14
width. This accumulation zone can be extremely long, generally being
governed by the local chartnel gradient, amount of staging and extent
of the resulting backl\1 ater. This buildup will continue until a
critical velocity is encountered, causing the leading edge to become
unstable with ice floes submerging under the ice cover. The pressure
on the thin initial ice cover increases as the upstream ice mass
builds up and higher velocities are reached in conjunction with
upstream advance. At an ttndetermined critical pressure, the ice cover
becomes unstable and fails. This sets off a chain reaction, and
\Vithin seconds the entire ice sheet is moving doh1nstream. Several
miles of ice cover below the leading edge can be affected by this
consolidation. This process results in ice cover stabilization due to
a shorte.ning of the ice cover, substclntial thickening as the ice is
compressed, a stage increase, and telescoping. The telescoping occurs
only during each consolidation. As the ice compresses dmvnstream,
tremendous pressures are exerted on the ice cover below the
accumulation zone. Here the ice mass tvill shift to relieve the
stresses exerted on it by the upstream cover, often becoming thicker
in the process. This 1.:ill tend to further constrict the flow,
resulting in an increase in stage. As the stage increases, the entire
icc cover lifts. Any additional pressures witltin the ice cover will
then be relieved by ]atPral expansion of the ice across the river
channel. This process can continue until the ice cover has either
expanded from bank to bank or else has encountered some other
obstruction (such as gravel islands) on ~d1ich the ice becomes
stranded.
8-15
The ice cover over ~~rater-filled channels h1ill continue to float.
Because of constant contact 1vith the flmving \\rater, the ice cover
erodes rapidly, sagging 21nd eventually collapsing. This process
usually occurs within days after the initial ice cover formation. In
some reaches these open leads can extend for several hundred yards. A
secondary ice cover generally accumulates in the open leads, and
completely closes the open water by the end of ~1arch. The process is
similar to ttte initial progression except on a smaller scale. Slush
ice begins accumulating against the doh7 nstream end of the leads and
progresses upstream.
complete closure.
Generally it takes several ~,reeks to effect a
The ice cover continues to move up the Susitna River, altl1ough at a
steadily decreasing rate as the channel gradient increases. Since the
gradient and the river velocities are increasing, staging levels must
increase in orde-r ·to create sufficient backwater to slot.; velocities to
allm~· ice juxtaposition. Although flows are only in the range of
3]000-5,000 cfs at this ti1ne] the water rises to levels equivalent to
open water flm\'s of up to 45,000 cfs. This often causes breaching of
upstream berms on many of the sloughs and side-channels. Significant
quant-ities of slush ice are S\\iept into these channels~ en-cering the
bac~(aler 8rea caused by the downstream staging. The slush ice then
consolidates and freezes in the side-channel, resulting in ice
thicknesses of up to 5-6 feet. This process occurs at different
levels in different years and at different locations on the river.
8-16
Hany of the sloughs have groundtvater seeps which persist through .. the
1<1inter. This groundwater is relatively h7 arm, with tvinter temperatures
of l 0 -3°C (R&M, 1982d). This is sufficiently warm to prevent a stable
ice cover from forming in those areas not filled tvith slush ice. This
thennal influence is evident as long, narrow, open leads extending
thousands of feet dotvn the sloughs.
d. Gold Creek to Devil Canyon
The reach from Gold Creek to Devil Canyon freezes over gradually, with
complete ice cover occurring much later than on the river belot-.r it.
The delay can be e.xplained by the relatively high velocities
encountered due to the steep gradient and single channel, and to the
absence of a continuous ice pack progression past Gold Creek. Another
factor for the slow ice cover development is tha·t by the time the ice
cover reaches this reach, there is less frazil ice from upstream, due
to the upper river having already frozen over.
The mast significant feature,s of freeze-up between Gold Creek and
Devil Canyon are t\'ide bordpr ice layers, ice build-up on rocks and
fol~mation of ice covprs over eddies. Ice sills have been identified
at several locations Portage Creek. Generally, the
constrictions form when the rocks to which the frazil ice adheres are
located near the water surface. When air temperatures are cold (less
than -l0°C), tl1e ice covered rocks will continue accumulating
additional layers of frazil until they break the water surface. The
8-17
ice-covered rocks effectively increase the water turbulence,
stimulating frazil production and accelerating ice formation. The ice
sills are often at sites constricted by border ice. This creates a
backwater area by restricting the streamflow, subsequently causing
extensive overflmv on·to the border ice. The overflo~v bypasses the ice
sills and re-enters the channel at a point further downstream. Within
the back\vater area, slush ice accumulates in a thin layer from bank to
bank and eye.ntuully freezes. A number of individual ice bridges form
in this reach by the process of borde.r ice grot~rth.
Since the ice formation process in this reach is primarily due to
border ice gro~th, the processes described for the Talkeetna to Gold
Creek reach do not occur. There is only minimal staging. Sloughs and
side-cllannels are not breached at the upper end 1 and remain open all
winter due to groundwater inflow. Open leads exist, but are primarily
in high-velocity areas between ice bridges.
e. Devil Canyon
Ice procE:.sses in De\'il Canyon create the thickest: ice along the
Susitna River, t>Jith measured thicknesses of up to 23 feet: (R&H,
1981c). The canyon has a narrow 1 confined channel 1-Jith high flow
velocities and extreme turbulence} making direct observations
difficult. Consequently, in 1982 a time-lapse ca1nera was mounted on
the soutl1 rim of tl1e canyon to document the pr?cesses causing these
great ice thicknesses.
8-18
The time-lapse camera provided docume.ntation that the ice formation
through Devil Canyon is primarily a staging process. Large volumes of
slttsh ice enter the canyon, and additional frazil ice is generated in
the canyon. The slush ice jams up in the lower canyon, and the ice
cover progresses up the canyon through large staging processes.
However, the slttsh ice has little strength, and the center of the ice
cover rapidly collapses afte.r the do\vnstream jam disappears and the
'"ater drai11S from beneath the ice. The slush ice bonds to the canyon
walls, increasing in thickness each time the staging process occurs.
1'he ice cover foruts and erodes several times during the winter.
f. t·Jid-\Vinter
The ice cover on the Susitna River is extremely dynamic. From the
moment that the initial cover forms, it is either thickening or
eroding. Slush ice will adhere to the underside of an ice cover in
low-velocity areas. Cold temperatures ~vill subsequently bond this ne~v
layer to the sur face ice.
If the ice cover conld eve.r he consid(-!red stable, it would be at the
be.ight of its maturity jn Nurch. During this period, sno\\'falls bt'Come
less frequent and very little frazil slush is generated. The only air
h1 Bter intet·faces are at the numerous open leads which persist over
turbulent reaches or ground\\1 ater see.ps. These are usually of short
length with insufficient heat exchange taking place to generate
significant amounts of frazil ice.
8-19
Discharges in ~larch are generally at the annual minimum, reducing the
flmving water to a shallow and narrow thalweg channel, indicated by a
depression in the ice cover. The depressions form shortly after ice
cover formation 1d1en the compacted slush ice is flexible and porous.
h7ater le\·els decrease through Harch., resulting in the floating ice
cover grounding on the river bottom. Water gradually percolates out
of the cover. Alternating layers of bonded and unconsolidated ice
crystals form within the ice pack when the receding level of saturated
slush freezes at extreme air ten~eratures. The result is the
formation of rigid layers at random levels, the layers representing
the frequency of critically cold periods.
Breakup processes on the Susitna River are similar to those described for
othor northern ri\Ters_, with a pre-breakup period, a drive, and a \\1ash.
Olichel, 1971). The general processes involved in breakup will be
described, together with specific examples from the Susitna River.
The. pre-breakup pE.,riod occurs as snowmelt begins due to incre.asE.~d solar
rJdintion in early April. This process gt_-.me.rally begins at the lower
elevations near the mouth of the Susitna River, working its way north. By
late April j the sn0\\1 has generally disappeared from the river south of
Talkeetna and has started to melt along the river above Talkeetna. Snow on
the river ice generally disappears before that along the banks, either due
to overflmv or because the snm\'pack is simply thinner on the river due to
8-20
exposure to winds. As the river discharge increases, the ice cover begins
to lift, causing fractures at various points. On the Susitna River, long,
narrow leads bPgin to form. Sma 11 jams of fragmented ice form at the
d0\11nstremn ends against the solid ice cover. These ice jams often resemble
a U-or V-shaped wedge, with the apex of the wedge corresponding to the
hi.ghest velocities in tl1e flow distribution. The constant pressure exerted
by these \\'edge-shaped ice jams effectively lengthens and \Videns many open
leads~ reducing the potential for major jams at these points.
The drive, or the actual downstream breakup of the ice cover, occurs ~1en
the discharge is high enough to break and move the ice sheet. The
inteiiSity and duration is dependent on meteorological conditions during the
pre-breakup period. Both \\'eak and strong ice drives have been observed on
the SusiLna River during the last 3 years. In 1981) there ~~ras a minimal
snowpack and only li~It precipitation during spring. Air temperatures were
\\Bnner than normal in early spring, but returned to normal in April,
resulting in slmv melting of \d1at snm.;r there was. Consequently, there was
not a sufficient increase in flow to develop strong forces on the ice
cover, and the ice tended to slowly disintegrate in place. Although some
ice jams did occur during the drive, they did not tend to last long, a11d
the breakup was general Jy mild.
Conditions were reversed in 1982. There was a significant snowpack still
remaining in l.3.te April, and temperatures t~'ere slightly cooler than normal.
The snowpack on the ice prevented weakening of the ice) which remained
strong. The final ice push did not begin until Nay 10. The ice ~vas
sufficiently strong to cause jams more severe than normal. Near RN 128
8-21
bel0\~1 Shennan, a dry jam formed t~rhich diverted practically all the flow out
of the mainstem into side channels. Closer to Talkeetna, a jam formed at
RM 107 that lasted for 3 days, jamming ice for over a mile and damaging
sections of the Alaska Railroad track.
Jam sites generally have similar channel configurations, consisting of a
broad channel with gravel is lands or bars, and a narrmv, deep thalweg
coJ·Ifined along one of the banks. Sharp bends in the river are also good
jnm sites. The presence of s laughs on a river reach may indicate the
locations of frequently recurring ice jams. Nany of the sloughs on the
Strsilna River between Curry ~nd Devil Canyon were carved through terrace
plains by some extreme flood. Summer floods, although frequently flowing
through sloughs_, do not generally result in ~vater levels high enough to
ove.rtop the river bank. During breakup, however, ice jams commonly cause
rapid, local stage increases that continue rising until either the jam
releases or the sloughs are flooded. While the jam holds, channel capacity
is greatly reduced, and flow is diverted into the trees and side-channels,
carrying large amo11nts of ice. The ice has tremendous erosive force, and
can rapidly remove large sections of bank. Old ice scars up to 10 feet
above the bunk level have been noted along side-channels near this reach.
It appears that thrc,se sloughs arH an indicator of frequent ice jams on the
adjacent mainstem, influencing the stability arrd longevity of these jams by
relieving the stage increases and subsequent water pressures acting against
the ice.
Stable ice jams are sometimes created when massive ice sheets snap loose
from shore-fast ice and pivot out into the mains tern flow. This occurred
8-22
near Indian River in 1982, resulting in an ice jam that lasted for several
days. The ice sheet was approximately 300 feet in diameter and probably
bet,,·een 3 and 4 feet thick. The upstream end pivoted around unt.il it
contacted the right bank of the mainstem. The ice sheet was then in a very
stable position, jammed against the steep ri.ght bank and grounded in
shallow 11·ater along a gravel island on the left bank. Several small ice
jElms upstreElm had released and were accumulating against this ice sheet,
extending the jam for about one-half mile. The water level rose, with an
estimated 2_,000 cfs flmving around the upstream end of a gravel island into
a side channel, overtopping the entrance berm to an adjacent slough.
Although the estimated discharge at Gold Creek was less than 6,000 cfs, the
normal summer flows required to breach this berm exceed 20,000 cfs. This
illustrates the extre1ne ~ater level changes caused by jams. Many ice floes
also drifted through this narrow access channel and grounded in the slough
as the flow dissipated over a wider area.
As drifting ice floes accumulated against the upstream edge of this jam,
tl1e floating layer beca1ne increasingly unstable. At some critical pressure
within this cover, the shear resistance between floes was exceeded,
resulting in a chain reaction of collisions that rapidly caused the entire
cover to fail. At tl1is point, severJl hundred feet of ice cover
conso]idated s inm1 tarwous ly. These consolidation phases occurred
frequently during a 4 hour observation period. The frequency ~vas dependent
on tlte volu1ne of incoming ice floes. With each consolidation, a surge wave
resulted. During one particular consolidation of the entire half-mile ice
jam, a surge wave broke loose all the shorefast ice along the left bank and
pushed it onto an adjacent gravel island. These blocks of shore ice were
8-23
up to 4 feet thick and 30 feet wide. The zone affected was almost 100 feet
longl with tl1e event lasting only a few seconds. This process is
essentially the same as telescoping during freeze-up except that the ice is
in massive rigid blocks instead of fine frazil slush, and is thus capable
of eroding substantial volumes of material in a very short time. The ease
with h'hich these ice blocks h·ere shoved over the river bank indicates the
tremendous pressures that build within major ice jams.
During all of the observed consolidations, the large ice sheet forming tlte
key of the jam 118\'er appeared to move or shift. The surge ~.;raves would
occasionally overtop the ice sheet, sending smaller ice fragments rushing
O\Ter the surface of t~he sheet. Towards the end of the day, the ice sheet
began to deform. Incident solar radiation, erosion and shear stresses ~~ere
rapidly deteriorating this massive ice block. Final observations showed it
to have buckled in an undulating wave and frnctured in places.
In general, the final destruction of the ice cover is accomplished by a
series of ice jams which break in succession and are added to the next jam.
This mass of ice continues building as it moves downstream. Upstream from
this accumulation, the river channel is commonly ice-free except for
:strand('d icc:' floes e:md some drifting ice coming from above Devil Canyon.
Near the Chulitna confluence the final ice release leaves accumulations of
ice c1nd debris stranded on the river banks. \1hen ice jams on this river
reach rele.ase, the ice floes piled up along the banks do not move, probably
due to strong frictional forces against the boulder strewn shoreline. This
creates a fracture line parallel to the £lo1v vector where shear stresses
8-24
were relieved. The main body of the ice jam flows dm-.rnstrearn, leaving
stranded ice deposits tvith smooth vertical \Valls at the edge of water.
Shear ~\'alls up to 16 feet high have been measured. In this case., the
extreme height of the water surface within the ice jam was demarcated by a
difference in color. A dark brown layer represented the area through which
h:ater had flo\\·ed and deposited sedime.nt in the ice pack. A white layer
near the surface ~as free of sediment and probably was not inundated by
flmving water.
Environtnental Effects
~-~--~-----~--
Ice processes are a major environmental force on the Susitna River,
affecting channel morphology l vegetation, and aquatic and terrestial
habitats. The impacts vary along the length of the river. The
environmental impacts of ice processes will be sununarized in the following
paragraphs. This "ill be follDI,·ed by a brief discussion of potential
modifications to tl1e ice processes of the Susitna River caused by operation
of the proposed hydroelectric development, and the subsequent changes in
environmental processes.
Ice processes appear to be a major factor controlling morphology of the
river beLv.lee.n the Chulitna confluence. and Portage Creek. Areas w i_·th
frequent jams have numerous side-channels and sloughs. The size and
configuration of existing sloughs appear to be dependent on the frequency
of ice jamming in the adjacent mainstem.
8-25
Najar ice events probably formed the sloughs tvhen ice floes surmounted the
river banks. The size and configuration of existing sloughs is dependent
on the frequency of ice jamming in the adjacent mainstem. Ice floes can
easily move the bed material, substantially modifying the elevation of
en t ranee berms to the s laughs. In Nay., 1983, a surge h:ave overtopped a
shallow gravel bar that isolated a side channel near Gold Creek. The surge
also created enough lifting force to shift large ice floes. These floes
barely floated but were carried into the side channel by the onrush of
\\1 3 ter, dragging against the bottom for several hundred feet, scouring
trou.ghs in the bed mate.ria 1. This same process will also enlarge the
s laughs. WhtHl staging is extreme in the mains tem and a large volume of
water spills over the berms, then ice floes drift into the side channel.
These ice floes scour the banks and move bed material, expanding the slough
perin1eter. This scouring action by ice can therefore drastically alter the
aquatic habitat.
Ice processes do not appear to play as important a role in the morphology
of the Susitna River below the Cliulitna confluence. This river reach below
the confluence regularly experiences extensive flooding during summer
storms. These seem to have significan-tly more effect on the riverine
~~nviroitmcnt than processes associated with ice cover formation (R&M, 1982a,
1982c·). This reach is characterized by a broad, multichannel configuration
tvith distances between ve.getated banks often exceeding 1 mile. The thahveg
is represented by a relatively deep meandering channel that usually
occupies less than 20 percent of the total bank to bank ,;idth. At lo~<
tvinter flot..-:s the thah\1 eg is bordered by an expanse of sand and gravel (R&r-1,
l982c). Although ice cover progression frequently increases the stage
8-26
about 1-2 fee·t above normal October water levels, no significant flooding
takes place, although some sloughs and the mouths of some tributaries do
receive some overflm.;. The ice cover belm.;r Talkeetna is usually confined
to the thalh'eg_, and surface profiles do not approach the vegetation trim
li11e along the hanks.
The erosive force of ice effects vegetation along the river. The frequency
of major ice jam events is often indicated by the age or condition of
vegetation on the upstream fmd of islands in the mainstem. Islands that
are annually subjected to large jams usually show a stand of ice scarred
JJJature trees ending abruptly at a steep and often undercut bank. A stand
of yot1ng trees occupying Lhe upstream end of islands probably represents
second generation growth after a major ice jam event destroyed the original
vegetation. Vegetation is prevented £ro1n re-establishing by ice jams that
completely override these islands.
Ice processes ha\Te several impacts on aquatic habitat. The sloughs may
fill with slush ice, which then forms a ice cover up to 5-6 feet thick.
This \\1ould p1·olong colder than normal \Vater temperatut~es in the slough.
(It could also cause problems for any beavers with lodges in the slough by
filling poo1s \\'i th ice). Divr!rs_ion of flow and ice into the sloughs may
cause large change.s in cl1r-nmel morphology. Large amounts of silt may be
deposited in the system at breakup, migra-ting doh'nstream during high flows
in the summer and covering good spawning habitat.
8-27
Operation of the \Vatana and Devil Canyon projects t.Jould significantly
modify the ice regime of the river below Devil Canyon. Flow rates will be
2-4 times grea-ter than natural t~rinter flow rates, \•iith t•;rater temperatures
of 2°-4°C immediately below the dams. The frazil ice generated in the
upper basin in early winter t~ill be trapped by the upper reservoir. Once
Devil Canyon Dam is bttilt, the major rapids in the system will be flooded,
further reducing frazil ice generation. These major changes in the
pl1ysical system and in the hydrologic and thermal regimes will combine to
greatly delay ice formation below the project.
Progression of the ice cover on the lower Susitna is notv due to rapid
j ux tapas it ion of ice floes from the upper river, tv i th the Sus i tna River
contributing 70-80 percent of the ice. cluch of this ice will not be
available under post-project conditions. Consequently, the ice cover tvill
not form until significantly more border ice has formed on the lower river.
Full ice cover developme.nt will be delayed for several weeks.
\Vater temperature belmv the project will not decay to the freezing level
for many miles. It is more likely that an ice cover will form on the river
above the Chulitna confluence when Watana is the only project in operation,
Lltan when Devil Canyon is also in opcra~ion. The ice cover now progresses
upstrc~rn from the conf]_uence when slush ice bridges a narrow channel at the
confluence. One question now under study is ·the ice formation process at
this point, and whether sufficient ice will be generated under post-project
conditions to cause this bridge to form. If ice cover does progress
8-28
upstream of the Cltulitna confluence, staging levels will likely be to a
higher level, as flo~v len~.ls and velocities ~.,:rill be greater than under
na·tural conditions.
Breakup patterns h1ill chcmge on the river belmv the project. The 1.;arm
water released from the r<:>servoirs, combined ~vith the increased air
temperatuL~es and solar 1:adiation in spring, ~vill cause the mainstem ice
cover to decay earlier in ·the season. Flow levels 1vill be significantly
lower in May as the reservoir stores flow from upstream. No ice will reach
t:he river above the Chulitna confluence from its upper reaches. This will
result in the breakup processe.s no\v occurring above the Chulitna confluence
being greatly attenuated or eliminated. Below the Chulitna confluence,
breakup impacts will probably also be reduced, although ice thicknesses may
be increased due to the increased winter flow levels.
Development of the Susitna Hydroelectric Project will significant change
the ice processes on the Susitna River. Below Talkeetna, ice formation
\.;iJ.l be signific::mtly de]aye.d due -co the re.duction of frazil aiJd slush ice
from the Sus .-i t.na R i v.::.·.r abo\'e. Ta lkPetna. Above the Chulitna cuHfluence., ice
formation will be delayed or eliminated due to the large volumes of tvarm
~vate.r released from the project and to the trapping of ice from the
upstream reaches. Both field studies and modelling are attempting to
define post-project conditions at specific sites. The results of these
8-29
studies t\1ill be utilized in developing the environmental impact statement
for the project.
Refere.nces
~~~---
Alaska Department of Fish & Game. 1982. Susitna Hydro Aquatic Studies
Phase II Basic Data Report. Anchorage, Alaska. 5 vol.
Ashton, George D. 1978. River Ice. Annual Reviews on Fluid l'Iechanics.
Vol. 10. pp. 369-392.
Benson, Carl S. 1973.
Fairbanks, Alaska.
A Study of the Freezing Cycle in an Alaskan Stream.
Institute of Water Resources. 25 pp.
(fichel, Bernard. 1971. \Vinter Re.gime of Rivers
Corps of Engineers, Cold Regions Research and
1-Iano\·er_, :\ew Hampshire. 130 pp.
and Lakes. U.S. Army
Engineering Laboratory}
Nct\1bury 1 Robert \<l. 1968. The Nelson River: A Study of Subarctic River
Proce_sses. University r·licrofllms, Inc.J Ann Arbor) :achigan. 319 pp.
Osterkamp) Tom E. 1978. frazil Ice Formation: A Review. Journal of the
Hydraulics Division. Proceedings of the American Society of Civil
Engineers. September, pp. 1239-1255.
f{~£~1 Consultants I rnc. 198la. Ice Observa-tions 1980-1981, Anchorage,
Hydroelectric Project. Alaska. Alaska PoKer Authority. Susitna
Report for Acres American_, Inc. 1 vol.
198Jb.
Data for the
Alaska Power
Acres American}
Preliminary Channel Geometry, Velocity and Water Level
Sus.i.tna River at Devil Canyon. Anchorage, Alaska.
Authority Susitna Hydroelectric Project. Report for
Inc. 1 vol.
1982a. Hydraulic and Ice Studies. Anchorage, Alaska. Alaska
Power Authority. Susitna Hydroelectric Project. Report for Acres
American, Inc. 1 vo1.
19b2b. Ice Observations 1981-82.
-·----~--~-
Anchorage 1 Alaska. AJ ask a
Pmver Au-thority. Susitna Hydroelectric Project. [{eport for Acres
American} Inc. 1 vol.
1982c. River Horphology. Anchorage, Alaska. Alaska Po~<er
Authority. Susit11a Hydroelectric Project. Report for Acres American,
Inc. 1 vol.
1982d. Slough Hydrology.
Authority. Susitna Hydroelectric
Inc. 1 vol.
8-30
Anchorage, Alaska. Alaska Pmver
Project. Report for Acres American,
1983. Susitna River Ice Study 1982-1983. Anchorage, Alaska.
Alaska Pm,-er Authority. Susitna Hydroelectric Project. Report for
Harza/Ebasco Joint VEnture. 1 val.
U.S. Geological Survey. 1982. \Vater Resources Data, Alaska, Water Year
1981. Anchorage 1 Alaska. \\rater H.esources Division, U.S. Geological
Survey. United States Department of the Interior.
8-31
Abstract
ALASKAN HYDROPOWER: BALANCING THE LONG RUN ADVANTAGES
WITH THE SHORT RUN PROBLEMS
By John S. Whitehead
Hydroelectric facilities have been operating in Alaska since the turn
of the twentieth century. Through the use of historical documents drawn
from twelve hydro installations, this paper looks at the historical
performance record of Alaskan hydropower. The analysis compares the
advantages of hydropower with its disadvantages in terms of electric power
prices, operational reliability, capital financing, power demand growth
projections, and legislative intervention in the operation of the
installations. The advantages and disadvantages are analyzed in terms of
short run and long run time frames.
Introduction
Over the last decade the promotion of new hydroelectric power projects
has been particularly strong in Alaska. Much debate has taken place in the
public media both for and against this expanded use of Alaska 1 s water
resources. The debate has become particularly heated since 1981 when the
Alaska legislature authorized $460 million for energy related projects
including funds for the construction of seven medium-sized hydroelectric
projects as well as feasibility and reconnaissance studies of a dozen
potential projects ranging in size from a few thousand kilowatts to the
mammoth 1.6 million KW Susitna project (SLA 1981, Chap. 90).
Advocates of hydropower often point to the use of a renewable energy
source, water, which would free the state from the use of fossil fuels with
ever escalating costs. Hydro is also claimed to provide stable and pre-
dictable power prices. Opponents often cite such disadvantages as runaway
capital costs, environmental hazards and cheaper kilowatt hour costs coming
from alternative sources such as natural gas. In the debate, as it appears
in the media, there is rarely any systematic reference to Alaska's actual
1Associate Professor, Department of History, University of Alaska,
Fairbanks, 99701
9-1
experience with hydropower. At best, selected statistics from particular
projects, sometimes from projects in other states, are brought forward.
In order to compile a systematic account of Alaska's actual experience
with hydroelectric power I examined the records and operational histories
of 12 hydroelectric facilities which were operational or under construction
in the summer of 1981 (see Table 1). The selection covers plants built
between the early 1900s and the present day and ranging in capacity from
1600 ICW to 47,160 KW. It includes plants in both southeast and
southcentral Alaska--the only areas of the state with major hydroelectric
facilities. The survey reveals that hydropower has had definite long run
advantages in terms of power price and operational reliability over periods
of 30-50 years. On the other hand definite short run problems in terms of
power price and operational reliability have occurred over periods of less
than 10 years. Such problems have been great enough, in some cases, to
jeopardize the financial viability and continued operation of certain
projects. The principal tool used to balance the short run problems with
the long run advantages has been legislative intervention in the operation
of the projects (Whitehead, 1983).
Long Run Advantages
In general the histories reveal that in the long run (i.e. 30-50
years) hydroelectric projects have fulfilled and exceeded the expectations
of their builders for three primary reasons. 1) Hydroelectric projects
were responsible for bringing reasonably priced--and in some cases very low
priced--electric power to Alaskan communities from the turn of the century
to the early 1960s--and into the 1980s in southeastern Alaska. 2) The
operation of Alaska's hydro projects has been extraordinarily reliable with
examples of plants in continuous operation from 1913 to the present day.
9-2
1
2
3
4
5
6
7
8
9
10
11
12
Table 1
Hydroelectric Facilities Surveyed
Operating Authority
Alaska Elec. Light and Power
Alaska Elec. Light and Power
Alaska Elec. Light and Power
Alaska Elec. Light and Power
Sitka Public Utilities
Sitka Public Utilities
Ketchikan Public Utilities
Ketchikan Public Utilities
Ketchikan Public Utilities
Ketchikan Public Utilities
Plant Name
Gold Creek
Annex Creek
Salmon Creek (Upper)
Salmon Creek (Lower)
Blue Lake
Green Lake
Ketchikan Lakes
Beaver Falls
Silvis
Swan Lake
Chugach Electric Association Cooper Lake
Alaska Power Administration Snettisham
Alaska Power Administration Eklutna
Location
Juneau
Juneau
Juneau
Juneau
Sitka
Sitka
Ketchikan
Ketchikan
Ketchikan
Ketchikan
Cooper Landing
Juneau
Anchorage
Capacity Ownership Date of Initial
KW Operation
1,600
3,500
2,800
2,800
6,000
16,500
4,200
5,000
2,100
22,000
15,000
47,160
30,000
Private
Private
Private
Private
Municipal
Municipal
Municipal
Municipal
Municipal
Municipal
REA
Federal
Federal
1904
1916
1913
1914
1961
1982
1923
1947
1968
1983 (anticipated)
1961
1973
1955
3) The long operational life of some plants has led to decreasing costs
over time.
Low Priced Power.
Throughout the history of Alaska in the 20th century, the high cost of
living has been a constant and recurring theme. One element in that high
price has been electricity generated by imported fossil fuels--primarily
diesel generation. Before the discovery of natural gas on the Kenai
Peninsula and in Cook Inlet in the late 1950s and early 1960s, the lowest
priced power in Alaska was hydropower.
Ketchikan was the first city in Alaska to have low priced power.
Ketchikan's first developed water resource, Ketchikan Creek, was placed in
service as early as 1903. By 1922 Ketchikan, with a population of approxi-
mately 2, 500, had a utility capacity of 2, 600 KW and a power price of a
little over 2¢ per kwh (Dart, 1924). Steady growth in the capacity of the
Ketchikan Lake facility from 1923 to 1957 and construction of the Beaver
Falls facility in 1946-47 gave Ketchikan a system capacity of 10,000 KW in
1957 with a power price under 2¢ per kwh--less than the U.S. national
average. In that year 500 kwh cost $9.88 in Ketchikan versus $10.81 in
Juneau, $14.50 in Anchorage, and $27.50 in Fairbanks where there was no
hydropower. Ketchikan's low prices resulted in an average annual power use
of 5,800 kwh per residential customer compared to 3,780 kwh in Juneau,
3,759 kwh in Anchorage, and 2,800 kwh in Fairbanks (USFPC, 1960).
Hydropower was responsible for making Ketchikan Alaska's most electrified
city in the first half of the 20th century.
Juneau's experience with hydropower was similar to Ketchikan's.
Several hydroelectric plants were constructed in Juneau before World War I
by various private industrial corporations to power stamp mills in the gold
9-4
mining industry. Surplus power was then sold to a private utility, Alaska
Electric Light and Power, for distribution to utility customers (Stone,
1980). Juneau's electric rates, while not as low as Ketchikan's, were
nonetheless reasonable. In 1922 AEL&P 1 s rates varied from 3-6¢ per kwh
depending on use {Dart, 1924). In 1957 Juneau still offered power at an
average of 3¢ per kwh {USFPC, 1960).
Juneau and Ketchikan both had utility systems based on modern hydro-
electric plants before the Second World War. Reasonably priced electricity
was the norm in these cities. More dramatic illustrations of the effect of
hydropower on electric prices can be seen in areas that began utility
production with less efficient power systems and later switched to modern
hydro facilities.
From 1912 to 1961 Sitka relied on an antiquated utility system which
was composed at varying times of two 160 KW hydro generators, an ineffi-
cient steam electric plant, and diesel generators. In 1950 Sitka's system
had a capacity of 2, 000-3,000 KW, depending on the season, and produced
power at 7-8¢ per kwh (USER, 1954). In 1957 500 kwh sold for $26.50 (5.3¢
per kwh). But few customers could get this price for such large
consumption as the average annual use was only 17 50 kwh per customer or
only 150 kwh a month (USFPC, 1960). In 1961 the modern Blue Lake hydro
project went on line. By 1968 power prices had dropped to $19 per 500 kwh
with a rise in annual customer consumption to 6,516 kwh (USFPC, 1969).
In Anchorage a similar scenario took place. The city's first hydro
installation, the 1,000 KW Eklutna Creek project, began production in 1929.
Unfortunately, hydro development did not keep pace with Anchorage 1 s post
World War II growth. In 1947-48 Anchorage with an estimated population of
19,000 had a system capacity of 6,800-7,700 KW, including 2,000 KW in
9-5
hydro, 1, 300 KW in diesel generators, and the remainder in a makeshift
steam electric system salvaged from a beached naval vessel. Power was
priced at $17.08 per 500 kwh (3.4¢ per kwh), but the production cost of the
steam and diesel power was 1¢ per kwh above that price. The low cost of
the hydropower subsidized the non-hydropower to create the relatively
reasonable price of 3.4¢ per kwh (USER, 1948). With the completion of the
30,000 KW Eklutna hydroelectric plant in 1955 prices dropped to $14.50 per
500 kwh (2.9¢ per kwh) by 1957 (USFPC, 1960).
The experiences of Ketchikan, Sitka, Juneau, and Anchorage certainly
indicate that hydropower was the key to bringing the first reasonably
priced electricity to Alaska.
Long Run Dependable Operation
Hydroelectric plants in Alaska have compiled a record of long term
reliable operation reaching decades beyond the term in which it takes to
amortize their capital costs. (Federal projects are scheduled to payout in
50 years. Municipally financed projects payout in shorter periods of
approximately 30 years.) The Ketchikan Lakes facility has been operating
continuously since 1923, though its capacity has been increased from 2,600
KW (1923) to 4,200 KW (1957). The 1923 facility was in fact a
refurbishment of a 1912 plant. So the date of reliable continuous opera-
tion can be increased by a decade.
Two particularly striking instances of long, reliable operational
lives are the Annex Creek and the Salmon Creek projects in Juneau. Con-
structed in 1913-14 and 1915-16 respectively, they were Juneau 1 s basic
source of electricity until 1973 (Stone, 1980). The plants were owned
until 1972 by a California firm, A-J Industries, which sold wholesale power
to the local utility, AEL&P, for retail distribution. A-J Industries kept
9-6
the facilities in poor physical repair after the Alaska-Juneau mine closed
in 1944 and also made no public disclosures of the financial aspects of its
hydroelectric operations. As a result the U.S. Bureau of Reclamation as
><ell as AEL&P considered both Annex and Salmon Creek outmoded and ineffi-
cient facilities. They both assumed that these plants would be closed
after the Snettisham plant began operation. In 1972 AEL&P purchased the
entire power system of A-J Industries with the expectation that it would
use only the company's transmission and distribution lines, not its
operating facilities (Whitehead, 1983).
In 1973 the new 47,160 KW Snettisham project went on line. Problems
with its transmission system, however, led to repeated power outages in it
first years of operation, thus forcing AEL&P to continue to use Annex Creek
and Salmon Creek for base load power production. The utility discovered
that the operation, both physical and financial, of these plants was so
reliable that it has continued to run them 365 days a year after the
transmission problems at Snettisham were corrected (Whitehead, 1983). In
fact, the continued reliable and economical operation of these plants has
caused an underconsumption of Snettisham power (see Short Run Prob-
lems--Surplus Capacity).
Rather than being junked as outdated projects, both Annex Creek and
Salmon Creek are being refitted for automatic control operation which will
further reduce their operating cost. The generating capacity of both
plants is also being increased with loans from the Alaska Power Authority
(SLA, 1981, Chap. 90).
Decreasing Costs Over Time
Hydroelectric projects have high capital costs compared to their
operating costs; the price of electricity produced is thus composed of a
9-7
substantial cost, from 50-90% in some cases, to amortize the capital and a
smaller amount for operation and maintenance. If the capital component of
the project remains operational after its initial cost has been amortized,
the price of power production will obviously drop to the operation and
maintenance costs--unless a large new infusion of capital is required to
rehabilitate the project. Such decreasing costs over time have been
acknowledged by utility operators in Juneau and Ketchikan--though reliable
historic cost data in these locations is hard to come by. It appears, for
example, that in 1981 the Annex Creel< and Salmon Creek facilities could
produce power for less than 20 mills per kwh compared to 22.5 mills per kwh
charged by A-J Industries in 1962.
Possibly the most reliable data to illustrate the decreasing cost
phenomenon can be found in the Eklutna plant in Anchorage, operated by the
Alaska Power Administration, Eklutna went on line in 1955 and is now more
than halfway into its 50 year payout schedule which will terminate in 2005.
In that year the price of Eklutna power should fall dramatically. A few
figures will help illustrate this. In 1979 the wholesale power rate at
Eklutna was 12.5 mills per kwh. More than half of the price, however,
included interest and amortization expenses. The operation and maintenance
costs at Eklutna for FY 1979 were $693, 928; if the allowance for plant
depreciation is added the costs rise to $882,496. These costs divided by
the firm annual energy generation of 153 million kwh would yield a price
for Eklutna power of 5. 8 mills per kwh, including depreciation, or 4, 5
mills per kwh, excluding depreciation. It is possible that operation and
maintenance expenses may rise over the years. In fact, APA announced a 21%
price increase in January 1980. This, however, may be offset by increased
production through rewinding the generators and upping their capacity by
9-8
15%. Soon after the turn of the 21st century, it is definitely possible
that Eklutna will be producing power for less than 10 mills per kwh in 2005
prices. Few other known sources of power offer such possibilities (APA,
1980).
Short Run Problems
While the long run advantages cited above make a convincing case for
hydropower in Alaska, the histories of the twelve facilities in my study
revealed a number of short run problems which in some cases called the
continued use of hydropower into question and in others produced a re-
markably high price for power. The principal short run problems were 1)
high power prices resulting from the debt service costs of new projects, 2)
substantial variations in the annual water flow--and consequently of the
annual power production--in some projects, 3) competition from natural gas,
and 4) underconsumption of power.
High Power Prices Resulting From Debt Service Costs
The completion of Sitka's Blue Lake project in 1961 brought reasonably
priced power to that community. By 1969 Blue Lake was beginning to reach
its installed capacity, based on a low reservoir level, of 6,000 KW. To
prepare for future demand the city purchased a 2,000 KW diesel generator in
addition to 1,100 KW in diesel units that it already owned. Several good
water years after 1969 staved off the need to generate substantial quan-
tities of diesel power. But by 1978-79 Sitka was generating 10-15% of its
powers needs through diesel production. Consequently, the price of 500 kwh
of power, which had risen from only $19 in 1968 to $20.90 in 1976, rose to
$25.60 in 1979. Diesel generation ~<as eroding Sitka's reputation of
low-priced electricity. (Official Statement $54,000,000, 1979).
9-9
To re-establish total hydropower generation the city embarked on plans
to construct the 16,500 KW Green Lake project with a $54 million bond sale.
The city was able to market the bonds at 7 5/8% interest in 1979, but under
conditions which were far from ideal. The bond underwriters, Dillon Read
and Co. required Sitka to refinance its outstanding utility debt as a
portion of the new bond issue. Thus the city was forced to pay 7 5/8%
interest on some of the Blue Lake bonds it had sold in 1961 for 4%. The
utility was also required to raise its electric rates so that revenues
would bring in 1. 25 times the amount required for debt service.
translated into an overall 45% increase in Sitka 1 s electric rates.
This
That
500 kwh of power which cost $25. 60 in 1979 rose to $38 in November 1980.
(Official Statement $54,000,000, 1979).
The debt service requirements to build Green Lake raised Sitka's power
price in the short run far beyond what it would have cost to add small
annual increments of diesel generation. The city was willing to accept a
substantial, though predictable, rate increase from hydropower to prevent
the potentially uncontrollable rate rise which might come from ever in-
creasing diesel generation in the long run. Sitka had to pay now for what
it hoped would be cheaper pm;er in the future.
Annual Waterflow Variation
Substantial variations in annual waterflow and a consequent variation
in annual power production have occurred at two hydroelectric facilities in
southcentral Alaska--Eklutna and Cooper Lake. While the average energy
production over any decade has been reliable, the peaks and valleys in
individual years require closer examination as potential problem areas.
Before Eklutna was constructed, the Bureau of Reclamation noted that
it did not have sufficient streamflow data to make accurate predictions for
9-10
Eklutna's firm annual energy production. The Bureau set a target in 1948
of 100 million kwh of critical year firm energy and 43.6 million kwh of
non-firm energy (USER, 1948). More streamflow data was accumulated during
the years of construction, and the Bureau revised the critical year esti-
mate to 137 million kwh in 1955.
million kwh.
Later the figure was raised to 153
In the first decade of Eklutna's operation water flow was sufficient
to maintain a level of generating capacity substantially above the critical
year estimates, The good years, however, came to an end in 1969. From
1969 to 1976 a period of poor water years severely lowered Eklutna' s power
production. The Alaska Power Administration, the operator of Eklutna, drew
down the reservoir for a number of years to maintain capacity, but in 1973
even this option was no longer viable. In FY 1974 Eklutna produced only
86.5 million kwh of power--less than 57% of its estimated firm annual
production. Low power production continued in FY 1975. Exceptionally good
water years, however, came after 1976, and in FY 1980 Eklutna produced
198,864 kwh or 130% of its firm annual supply. Table 2 illustrates the
power variation at Eklutna (APA, 1980).
A similar water flow problem has been encountered at the Cooper Lake
hydro project, operated by the Chugach Electric Association. Cooper Lake's
annual firm energy output is approximately 41 million kwh. Chugach rep-
resentative Tom Kolasinski noted in 1981 that annual generation has fluctu-
ated between 24 and 60 million kwh. As a result of this fluctuating water
flow, Chugach did not deem it feasible to raise Cooper Lake 1 s original
installed capacity of 15,000 KW to the anticipated 30,000 KW (Whitehead,
1983).
9-11
a
b
c
d
Table 2
Annual Generation of Eklutna Power Projecta
FY
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1976 (Third Quarter)
1977
1978
1979
1980
1981
Million kwh
b 43.8b
119. 3b
136.7
164.5
165.8
188.2
198.8
150.5
156.5
159. 1
135.3c
138.9
184.2
164.3
168.0
160.8
127.3
159.2
142.8
86.6
120.9
160.2d
24.7
174.4
193.6
153.0
198.9
196.3
·Source: Alaska Power Administration, March 1982.
Project capability exceeded demand in early years of operation.
Low production mainly due to draw down of reservoir in 1964 to permit
repairs to earthquake damage.
After FY 1976 the federal fiscal year changed from July 1-June 30, to
October 1-September 30. This entry covers July 1, 1976, to September
30, 1976.
9-12
Annual water flow variation and a resulting variation in power produc-
tion are expected in all hydroelectric projects. But the variation in
Anchorage seems high. At Eklutna, production has fluctuated between 199
million kwh and 87 million kwh--a drop of 57% from the high to the low.
Similar figures hold for Cooper Lake. By comparison, power production in
Ketchikan has fluctuated between 68 million kwh and 57 million kwh for all
three plants in its municipal system--a drop of 16% from the high to the
low. One may well wonder if such wide variations as those in Anchorage
indicate that hydropower in certain locations is an unreliable power
source. What would have happened if low water years had come 10 to 15
years earlier when Anchorage was more dependent on Eklutna 1 s production?
In 1957, for example, the energy demand in Anchorage was 154 million kwh.
If Eklutna 1 s production had dropped from 140-150 million kwh to 86.6
million kwh, Anchorage would have faced a power crisis. The two utilities
with operating capacity, Chugach and the Anchorage Municipal Light and
Power Department, would have been hard pressed to fill the gap from their
steam and diesel plants since their combined capacity was little more than
half of Eklutna 1 s 30,000 KW.
Alaska Power Administration head Bob Cross has noted that the varia-
tion in Eklutna 1 s production requires closer scrutiny. Before 1968 APA
operated Eklutna on a "critical year" mode. Water in the reservoir was
conserved in good water years so that the firm target of 137 million kwh
could be met in poor water years. After 1968, when hydro was no longer the
major source of power in Anchorage, APA shifted its mode of operation to
"maximum annual energy production." Under this mode all the available
reservoir capacity was used for energy production in good years rather than
stored for poor years. According to Cross a severe drop in power
9-13
production would not have occurred if poor water years had come earlier.
He estimated that under critical year operation Eklutna could still have
produced 130 million kwh annually under drought conditions (Whitehead,
1983).
Cross' explanation is helpful. But let us look at the figures again.
Even under "critical year" operation, the variation in Eklutna' s power
production would have been substantial if a drought had occurred. From
1958 to 1968 Eklutna produced substantially more than 137 million kwh,
except in the earthquake year of 1964. If a drought had come in the late
1950s or early 1960s, Eklutna's production could have fallen by as much as
65-70 million kwh from a high of 199 million (1961) to an estimated low of
130 million kwh--a drop of 35% between the high and the low. Chugach and
AML&P would not have been as hard pressed to generate the difference with
diesel and steam, but the price of electricity would certainly have risen
in the days before cheap natural gas became an alternative fuel (Whitehead,
1983).
Much of my above concern is hypothetical. The poor water years came
after Eklutna had acquired a reputation for good service to Anchorage and
at a time when alternate energy production from natural gas was cheaper
than hydropower. But what about such variations in future projects?
Consumers who have enjoyed an abundance of cheap hydropower for a series of
good water years may react negatively to a drop in hydro production and a
consequent rise in electric rates, if power must be generated from a more
expensive source. Such a short term public reaction could cause problems
in Alaska where positive public opinion is often critical in securing state
legislation and approving local bond proposals for a new hydro facilities.
In future hydro developments it may be wise to make the potential
9-14
fluctuations in production known to consumers. It might even be advisable
to include an allowance for alternative fuel generation in the rate
structure to smooth out any variation in power prices between good and poor
water years.
Competition From Natural Gas in Southcentral Alaska
The opening of the Eklutna plant in 1955 established hydropower as the
preferred form of electrical generation in the Anchorage load area. Six
years later Chugach Electric Association opened its 15,000 KW Cooper Lake
plant on the Kenai Peninsula. In the late 1950s and early 1960s plans were
proposed by the U.S. Corps of Engineers to build the 46,000 KW Bradley Lake
project; Chugach also obtained a federal license to build the 10,000 KW
Grant Lake plant. Hydro advocates also pushed for federal construction of
the 580,000 KW Devil Canyon (Susitna) project 150 miles north of Anchorage.
By 1964 most of the enthusiasm for new hydro construction in the
state's largest load area was over. The Corps of Engineers announced that
there would be no demand for Bradley Lake power, even though the project
was authorized for construction in the Flood Control Act of 1962. Chugach
abandoned it plans for Grant Lake. Since 1962 not one kilowatt of
hydropower has been added to the Anchorage system (Whitehead, 1983). What
happened?
The answer is simple. Discoveries of natural gas on the Kenai Penin-
sula in 1957 and later at the Beluga Field in Cook Inlet undercut the cost
of hydro product ion by a half. Electricity from combustion turbines could
be generated for less than 5 mills per kwh compared to 11 mills for Eklutna
power and a projected 9-10 mills per kwh for Bradley Lake hydro. Chugach
opened its first combustion turbine plant at Bernice Lake in 1963 and
installed its first gas facility at the Beluga field in 1968. The price of
9-15
Chugach gas power dropped to $12.95 per 500 kwh in 1968 compared to the
$14.50 per 500 kwh it charged for hydropower in 1957 (USFPC, 1960, 1968).
By 1976 Chugach had installed 316,000 KW in gas power compared to 15,000 KW
in hydro (USFPC, 1976). Gas turbine electricity effectively stopped the
construction of new hydroelectric facilities in the Anchorage load area.
What effect did it have on the existing facilities?
The purchasers of Eklutna power--Chugach, the Anchorage Municipal
Light and Power Department, and the Matanuska Electric Association--~<ere
tied to 25 year contracts. Chugach also continued operation of Cooper
Lake. So no immediate move to discontinue existing production developed.
However, after the 1964 Anchorage earthquake concern mounted that the
long-term contracts for Eklutna power might not be renewed when they
expired, The cause for concern lay in the cost of reparing earthquake
damage at Eklutna.
On the day of the earthquake, March 27, 1964, both Eklutna and Cooper
Lake sustained little visible damage. Both facilities were able to gener-
ate power within a few hours after minor repairs. Later investigations at
Eklutna in July of 1964 revealed that there had been settling at the base
of the dam and a general weakening of the structure. It soon became
evident that substantial rebuilding of the dam, particularly of the
spillway, would have to take place (USER, 1966).
The repairs t;ere completed at a cost of $2,885,415. Under the terms
of the original Eklutna Act of 1950, this cost ~<Ould have to be fully
reimbursable through power rates--an effective 1 mill per kwh increase. By
the late 1960s, the increasing use of natural gas for electrical generation
led the Department of Interior to be concerned over the potential effect of
the 1 mill increase. In 1968 an assistant secretary in the department told
9-16
Congress that "this rate differential. .. will add to the problem created by
current competitive natural gas prices in future contract negotiations for
Eklutna Power." (U.S. Congress, 1968). In response Congress intervened in
September 1968 and passed Public Law 90-523 making all but $80,000 of the
repairs non-reimbursable. This legislation, coupled with the fact that
Eklutna had generated more revenue in power sales prior to 1968 than had
originally projected, allowed the Alaska Power Administration to lower
Eklutna's prices by 10% in 1968 (APA, 1969).
When the time came to renew the power contracts in the late 1970s (the
contracts would expire in 1980), the Alaska Power Administration had no
problem finding purchasers for Eklutna power at 12.5 mills. The rising
price of natural gas and Anchorage's ever increasing demand for power made
Eklutna' s electricity fully competitive, It does not appear that the
legislation of 1968 was particularly important a decade later in contract
negotiations. The long run stable price and availability of Eklutna power
were its selling points.
The legislation of 1968 did, however, have a more important effect of
the future development of hydroelectric power. It set a precedent for
legislative intervention in the financial operation of a facility.
Alaskans would not forget it. They l.rould use the 1968 law as a precedent
in asking the federal government to intervene in the financial operation of
the Snettisham plant in 1976 for reasons much less dramatic than earthquake
damage.
Surplus Capacity
The Eklutna project was built to meet an acute shortage of power for
utility customers in a rapidly growing load area. Three years after going
on line, Eklutna was selling more than its annual firm energy capacity (see
9-17
Table 2). In contrast, the substantially larger Snettisham plant near
Juneau was built with the assumption, really the hope, that a full demand
for its power would develop in 2-3 decades. If such hopes failed to
materialize, or if the power growth was considerably off schedule, the
project would have surplus capacity. The price of power per kwh would
obviously have to rise to higher than projected levels to pay off the fixed
capital costs. Depending on how much surplus capacity existed, the price
rise could be minimal or it could be substantial. Surplus capacity could
have the effect of making hydropower one of the most expensive forms of
electricity. Why was the federal government willing to take such a risk in
building Snettisham?
Snettisham was not originally planned with surplus capacity in mind.
When the project was first designed in the late 1950s by the U.S. Bureau of
Reclamation, it was to be a supplier of industrial power. Specifically,
Snettisham would provide power for a pulp and newsprint mill to be built by
the Georgia-Pacific Corporation. The hydro facility would thus promote the
economic development of the timber industry in southeastern Alaska. Of the
facility's projected annual energy production of 292 million kwh, 230
million kwh would go to Georgia-Pacific and only 47.4 million kwh would go
for utility use. The remaining 14. 6 million kwh would be absorbed in
transmission losses. Based on these assumptions the Bureau recommended in
1959 that Snettisham be constructed (USBR, 1959).
The planning for Snettisham changed abruptly in June 1961 when
Georgia-Pacific Corporation announced that it would not build its newsprint
plant. On the surface of things, it would appear that there was no longer
any justification for building Snettisham. But by 1961 Juneau residents,
Alaska's new congressional delegation, and the Bureau of Reclamation itself
9-18
were so committed to seeing Snettisham built that the project had almost
taken on a life of its own. In November 1961 the Bureau of Reclamation
revised its estimates of Juneau 1 s potential utility growth over the next
two decades and concluded that if Snettisham were built in stages, it would
be feasible for utility production alone. It would take approximately a
decade longer for utility demand to reach the level originally proposed for
industrial demand. According to the Bureau, a rise in the price of power
produced from 6.1 mills per kwh to 7.47 mills per kwh would make Snettisham
feasible (USBR, 1961).
These new planning estimates assumed that the existing hydro facil-
ities in Juneau (Annex Creek and Salmon Creek) would be retired when
Snettisham came on line. The projections also assumed a certain surplus
capacity or underconsumption of power in the early years of operation. But
at 7.47 mills per kwh, enough revenue would be generated in later years to
offset initial deficits and hence to pay out the project in the standard 50
year period for federally financed installations. In essence, Snettisham's
new payout schedule resembled a "balloon mortgage" for a home. The deci-
sion to take the risk with such a forecast of initial surplus capacity was
not the original plan; it was one which developed to save the project in
mid-stream.
Snettisham was authorized for construction by Congress in the Flood
Control Act of 1962 (P.L. 87-874). After many delays in receiving
appropriations, the Long Lake stage was completed in 1972-73 at a cost
roughly 50% greater than the amount authorized in 1962. As a result, the
price of Snettisham power rose from the projected 7.47 mills per kwh to
15.6 mills per kwh. This was still lower than the price A-J Industries had
9-19
charged for its hydropower. The price rise resulting from escalating
construction costs was the least of Snettisham's problems.
During its first three years of operation (1973-76), Snettisham 1 s
transmission line was constantly problem-prone~ As a result, Snettisham
was out of service for months at a time. Repairs were made, but finally
the Alaska Power Administration relocated the line in 1976. The total cost
for repairs and relocation was $11 million--all of which was required by
law to be reimbursable through increased power rates.
The failure of Snettisham 1 s transmission line was only part of the
facility's problem. By 1976 it was evident that Snettisham was simply not
selling as much power as had been projected. As late as 1979 Snettisham
sold only 80.45 million kwh or less than half of its 168 million kwh of
firm annual energy. What caused such underconsumption? (APA, 1980)
As noted earlier, Snettisham's transmission line failures led AEL&P to
depend on hydro power from its older facilities (Annex Creek and Salmon
Creek) and to continue using them after Snettisham went back into service.
The permanent operation of Annex Creek and Salmon Creek thus took an annual
40-50 million kwh of the market away from Snettisham. In addition, the
1961 estimates of Juneau's projected utility demand had been too optimis-
tic. From 1960 to 1973 growth in demand had been closer to 7.6-7.8% rather
than the "conservative" 9. 3% estimated by the Bureau of Reclamation.
(Table 3 gives the original 1959 estimate of utility growth in Juneau, the
revised 1961 estimate of utility growth in Juneau, and the actual utility
generation in Juneau from 1960 to 1982.)
The combination of competition from the older hydro plants and the
slower than anticipated growth of the Juneau power market resulted in a
surplus of power at Snettisham. If the price of electricity had to reflect
9-20
Table 3 A-C
A 1959 U.S. Bureau of Reclamation Feasibility Report of Utility Load
Growth in Juneau.
1952 (actual)
1958 (actual)
1960 (projected)
1962
1965
1970
1975
(USER, 1959)
Peak
(thousand KH)
4.1
5.1
6.6
7.6
10.9
15.3
20.4
Annual Generation
(million kwh)
16.70
24.40
29.20
33.64
47.90
67.61
89.72
B 1961 U.S. Bureau of Reclamation Reappraisal of Utility Load Growth
1958 (actual)
1960 (actual)
1962 (projected)
1965
1970
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
(USER, 1961)
Peak
(thousand KW)
5.1
5.8
7.2
9.4
15.2
24.3
26.5
28.9
31.4
34.1
37.0
40.0
43.2
46.6
50.4
54.4
58.8
63.5
9-21
Annual Generation
(million kwh)
24.4
29.2
34.9
45.5
73.4
116.9
127.6
139.1
151.3
164.3
178. 1
192.7
208.1
224.7
242.7
262.1
283.1
305.7
Table 3 cont.
C Actual Generation of Power in the Juneau Area, 1960-1982
1960 (Calendar Year)
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970 (Fiscal Year)
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
Peak
(thousand KW)
5.8
7.8
7. 1
9.0
9.4
10.0
10.9
10.5
11. 1
11.8
12.4
13.8
14.9
15.5
16.2
17.8
19.8
20.4
23.4
23.1
26.2
32.2
42.2a
(Alaska Power Administration, March 1982)
a January 1982
Annual Generation
(million kwh)
29.2
32.3
34.7
37.2
41.5
43.5
48.3
49.3
52.8
56.0
58.3
63.8
70.3
75.8
83.1
94.6
106.3
112.2
122.2
133.5
143.1
160.7
Note: As a rough rule of thumb, Snettisham's generation for any one
year would be 50 million kwh less than the annual generation figure.
9-22
the costs involved with the transmission line as well as amortize the
project's full capital costs, Snettisham's power rate would rise to a much
higher level. (To my knowledge, projections of those rates have never been
published.) Such potential price increases were forestalled in 1976 by
federal legislation which resembled in many ways the Eklutna legislation of
1968.
In the Water Resources Development Act of 1976 (P.L. 94.-587, Sec.
201) Congress provided that the cost of relocating the transmission line
($5. 6 million), though not the cost of line repairs, would be
non-reimbursable. To alleviate the problem of surplus capacity the act
extended the payout schedule for 10 years and froze the price of power at
the rate of 15. 6 mills per kwh until 1986. During this 10 year "load
development period" the project would not be required to cover its full
amortization costs, but would actually increase its overall capital indebt-
edness. In effect, the "balloon" aspects of the payout schedule were
simply extended another ten years. In 1986 the price of power will rise to
generate sufficient revenues to complete the 60 year payout schedule. The
Alaska Power Administration predicts that the 1986 price will be 25.8 mills
per kwh. What chance of success does Snettisham have to develop a full
load for its power?
The current policy of the Alaska Power Administration for utilizing
Snettisham' s surplus capacity is the development of new markets for elec-
tricity in Juneau. The principal new market is residential electric
heating. According to APA estimates made in 1980, this could provide a
full demand for Snettisham power by 1983; without heating the full utility
demand would not develop until 1995 or 2000. And in the event that the
9-23
capital of Alaska moved from Juneau, Snettisham would never reach a full
demand without residential heating (APA, 1980).
Oddly enough both the "heat" and the "no heat" strategies present
problems. If the residential heating strategy is successful, Snettisham
could reach capacity rather quickly. Then additional electricity may have
to be generated by diesel fuel thus raising the price of power. Or new
hydro facilities could be built with the potential debt servicing costs we
have already noted in Sitka. If the heating strategy does not work,
Snettisham will continue to have surplus capacity for at least another
decade. The price of power will have to rise beyond the projected 1986
rate unless a new round of political intervention occurs. The most likely
form of intervention would be a state purchase of Snettisham from the
federal government. The capital costs of the project could then be
absorbed by the state and an arbitrary price for power could be set.
The dilemma of surplus capacity in many ways defies a simple solution.
It is particularly exaggerated in Juneau because Snettisham is not connect-
ed with another power market. Thus Snettisham' s short run surplus cannot
be sold to another area and saved in the long run for Juneau 1 s potential
growth. In an isolated load center surplus capacity in hydropower can
cause the price of electricity to be as unstable as that generated by a
fossil fuel. In such a situation, hydropower loses its advantage of stable
and predictable power rates.
Conclusions
The historical survey of twelve of Alaska's hydroelectric instal-
lations provides evidence that hydropower has been successful in the long
run in bringing reasonably priced electricity to Alaska. The operational
lives of some of the facilities have exceeded the expectations of their
9-24
builders. Hydroelectric generation has presented few operational problems.
No hydro installations in the survey have declined in their ability to
produce power over the long run. In fact, a number of the water power
sites have had their capacity increased. Even in the Anchorage area, where
water flow has varied substantially from year to year, the long run average
power production has been quite reliable--even exceeding the original
estimate for Eklutna.
Despite these long run advantages we have seen that in the short run
communities may have to pay a substantial price for hydropower. This has
come from debt servicing costs in Sitka, from earthquake damage in
Anchorage, from transmission line failures and surplus capacity in Juneau.
A community may also have to pay a higher price for hydropower in certain
periods when an alternative fuel, natural gas in Anchorage's case, can
provide a lower price. And it may be necessary to provide stand-by sources
of power--and absorb the cost of power rates--in places where the annual
waterflow of a project causes power production to fluctuate substantially
between years.
The case histories also indicate that the Alaskan public has felt at
times that the costs of the short run problems should not be borne by power
consumers alone. Attempts to balance or smooth out the short run costs
through legislative intervention have occurred. In the case of Eklutna the
legislation was probably justified on the ground of disaster relief. In
the long run the legislation has actually proved unnecessary for keeping
Eklutna's price competitive.
The 1976 legislation in regard to Snettisham, however, is more
problematic. It provided relief from the cost of operational failures (the
transmission line problems) which had nothing to do with a natural
9-25
disaster. The transmission line risks were w""ll known. Equally risky were
the planning assumptions for Juneau's electric power growth. There is an
inherent risk in any project built on long-range growth projections. The
Water Resources Development Act of 1976 essentially absorbed the costs of
those risks to maintain reasonably priced hydropower. Thus the 1976
legislation set the precedent that consumers may not have to absorb the
risks involved in constructing and operating hydro projects in their
communities. If we consider the construction of a hydropower project as
partially an economic enterprise and partially a political enterprise, the
1976 legislation clearly pushed Snettisham toward the political end of that
scale. If future government intervention, state or federal, at Snettisham
or other installations continues in this direction, the Alaskan public may
well come to view the development of hydropower as a game played by
politicians in which the public purse absorbs the economic risks. Such a
negative public view could do serious damage to the image of hydropower and
jeopardize the future development of one of the state's most valuable
natural resources.
Hydropower's short run cost problems definitely pose a dilemma for its
future development. The balancing act through political intervention is a
delicate one which must be handled with extreme care.
References
Alaska Power Administration, 1969, First Annual Report 1968.
Alaska Power Administration, 1980a, 1979 Annual Report.
Alaska Power Administration, 1980b, Juneau Area Power Market Analysis.
Dort, J. C., 1924, Report to the Federal Power Commission on the Water
Powers of Southeastern Alaska.
9-26
Official Statement $54,000,000 City and Borough of Sitka, Alaska, Municipal
Utilities Revenue Bonds, 1979.
Stone, David, 1980, Hard Rock Gold.
Session Laws of Alaska 1981, Chapter 90.
U.S. Bureau of Reclamation, 1948, Eklutna Project Alaska.
U.S. Bureau of Reclamation, 1954, A Report on the Blue Lake Project.
U.S. Bureau of Reclamation, 1959, Snettisham Project Crater-Long Lakes
Division Alaska.
U.S. Bureau of Reclamation, 1961, Reappraisal of the Crater-Long Lakes
Division, Snettisham Project, Alaska.
U.S. Bureau of Reclamation, 1966, Eklutna and the Alaska Earthquake.
u.s. Congress, House, 1968, Providing for the Rehabilitation of the Eklutna
Project Alaska, H. Report 1852, 90th Congress, 2nd Session.
u.s. Federal Power Commission, 1960, Alaska Power Market Survey 1960.
u.s. Federal Power Commission, 1969, Alaska Power Survey 1969.
u.s. Federal Power Commission, 1979, Alaska Power Survez 1976, Vol. l.
Whitehead, John S., 1983, Hzdroelectric Power in 20th Centurz Alaska:
Anchorage, Juneau, Ketchikan, and Sitka. This paper is an adaptation and
condensation of research reported in this work. Readers interested in
greater detail and explanation of the issues noted in this paper should
consult the above work. In citing references in this paper, I have cited
(Whitehead, 1983) when the information was received through interviews or
unpublished sources. When the information or data originally appeared in
another published source, that source is cited.
9-27
HISTORICAL DEVELOPMENT OF ALASKA WATERLAW
By Robert E. Miller
Abstract
There are two major doctrines of water law in use in the United States.
The first is the riparian doctrine which holds that private water rights
exist as an incidence of ownership of land bordered or crossed by a natural
water course. The second doctrine is of prior appropriation, which
provides that the earliest appropriator in point of time has the exclusive
right to use the water to the extent of his appropriation without
diminution of quantity or deterioration of quality. Alaska's Water Use Act
of 1966 provided the most recent evolution of this doctrine. The Act
recognizes the unity of the hydrologic cycle by putting all water in one
class. Alaska thus avoids the legal difficulties that have resulted from
other states attempts to divide water into legal classes and to apply
different rules of law to different types of water occurrence.
Introduction
The term "water law" refers to all those rules which have been established
to resolve water use decisions toward the goal of maximizing benefits to
society. Water law thus includes such topics as the allocation of supply
among competing users, quality management, flood control, drainage,
ins tream use, and wetlands management. Although these topics are
functionally related, applicable law covering each topic has often
developed independently. Thus, water law does not consist of an integrated
body of legal principles for managing the resource and problems of
coordination between different bodies of law remains a difficulty (Cox &
Miller 1982).
Water Law Systems
The most developed area of water law is the allocation of supply. Two
major doctrines of water law are in use in the United States. Before the
1Associate Professor, Civil Engineering, University of Alaska, Anchorage,
3211 Providence Drive, Anchorage, Alaska.
10-1
settlement of the West, the right to use the water of natural streams
accrued to the owners of land along the streams. This is the common law
doctrine of riparian rights. Under a riparian right the owner of land
adjacent to a stream is entitled to use the full natural flow, undiminished
in quantity and unchanged in quality. The next downstream owner of
riparian land has the same right. In applying this doctrine to groundwater
a "quasi-riparian" right is based on ownership of the land overlying a
water-bearing formation. Riparian rights cannot be lost by non-use; the
water can be used only on riparian land; and there are no requirements as
to beneficial use. With some notable exceptions, it is found that the
riparian doctrine is followed by the humid East. In the humid states, as
noted by Thornthwaite (1948), precipitation is generally greater than the
potential evapotranspiration. Thus, precipitation is usually more than
that necessary to support agriculture.
The arid and semi-arid areas of the West generally find that precipitation
is less than potential evapotranspiration.! The climate, the topography,
and an economy based primarily on agriculture and mining required a more
flexible doctrine for water allocation. As more people settled in the West
it became evident that a method had to be devised which would offer
protection to existing water users and also conserve water. Mining law
presented the precedent that the first person to stake out a mine had a
claim which was superior to those that came after. Since many of the early
water diversions were for mining, it followed that water claims should also
adhere to the principle "first in time gives first in right." Although
1 The dividing line can be taken as the 97th meridian, i.e., about 200 to
400 miles west of the Mississippi River.
10-2
initially developed as a custom, this appropriation doctrine soon became
formalized by constitutional and statutory enactments. The appropriation
doctrine has no land ownership requirements, and water use is not
restricted to riparian land.
It was customary for the early appropriators to post a notice of their
water use. Later such notices were filed with the district courts, and
since many streams flowed through several districts, it was necessary to
make the filings statewide. The states which followed the appropriation
doctrine also realized the necessity for some form of regulation. The
overwhelming majority elected to adopt the permit or Wyoming system (a few
Eastern states have also recently adopted permit systems -Sax, 1965).
States which have initiated a permit system require as a prerequisite to
taking water that an application for a permit to proceed be made with some
state administrative agency, usually the State Engineer or a water control
board. In Alaska the right to use water is obtained by making an
application to the Commissioner of the Department of Natural Resources.
Most of the permit system states have legislation empowering the
administrative agency to deny the permits if there is no longer any
appropriable water in the stream or if the appropriation is not in the
public interest.
Of all the appropriation states, Colorado and Montana remained alone in
allowing water to be appropriated without any administrative intervention.
However, developments in Colorado water law suggest a trend toward a permit
system. The Ground Water Management Act and the Water Right Determination
and Administration Act of 1969 both provide for procedures somewhat similar
to those in permit states. (Colo. Rev. Stat. Ann. 148-18-6 (Supp. 1966)
10-3
Apparently the miners of Alaska had become accustomed to having a water
claim as part of a placer claim and in 1917 the territorial legislature
gave the locator of any mining claim that included both banks of a stream
the right to use as much water as was needed to work the claim.
Riparianism in early groundwater law was evident in Trillingham v. Alaska
Housing Authority. A District Judge stated that a complaint seeking to
enjoin a defendant from diminishing plaintiff's supply of groundwater did
not state a claim for relief "because percolating waters may be used by the
owner as he sees fit.'' This, of course, is a statement of the common law
rule of absolute ownership. To this limited extent there were early
remnants of riparianism in Alaska. However, courts with the exception of
the Trillingham case have had little interest in riparianism, and the
mining law has been repealed.
State Constitution:
Alaska was eager for statehood and a state constitution was ratified by the
people of the territory in 1956. The constitution contains 18 sections on
natural resources. The constitution provides that all surface and
subsurface waters except "mineral and medicinal waters" are reserved to the
people for common use and are subject to appropriation. Priority of
appropriation shall give prior right. Statehood came on January 3, 1959.
Legislation to implement the constitutional provisions did not come until
1966.
Water Code:
In 1961, Governor William Egan called for a comprehensive water code
covering all aspects of water problems, before the problem arose. The
Commissioners of the Alaska Departments of Health and Welfare, Natural
10-7
The administrative systems for water rights are often quite different
depending on the nature of the water source. Although not the case in
Alaska, it is quite common for a state to adhere to some variation of one
doctrine for surface water and to apply the other doctrine to its
groundwater. There are four basic doctrines governing the regulation of
groundwater rights; Prior Appropriation Doctrine, Absolute Ownership
Doctrine, Reasonable Use Doctrine, and Correlative Rights Doctrine. The
Prior Appropriation Doctrine for groundwater is analogous to that of
surface water .. It appears that this doctrine for percolating groundwater
is applicable only in those states which have adopted it by statute. The
1966 Alaska Water Use Act best illustrates the pure prior appropriation
doctrine as applied to groundwater. Alaska makes no distinction between
ground and surface water and provides that such waters are reserved to the
people for appropriation and beneficial use. Thus, Alaska provides a
pristine form of groundwater appropriation. The remaining three
groundwater doctrines are all variations of general riparian concepts where
by virtue of land ownership, an owner has some rights in the water under
his land.
The Absolute Ownership Doctrine allows the landowner complete freedom to
drain an entire groundwater source as long as the pumping takes place on
his own land. This rule is based on the Latin maxim, cujus est solum, ejus
est usque ad coelum ad inferos. There are almost no legal restrictions on
the removal, even if the water is drained from beneath adjacent land.
Under this doctrine the pumper is not even required to use the water on his
own land.
10-4
Under the Reasonable Use Doctrine, water rights are dependent on a
reasonable and non-wasteful use of water, generally, on the overlying land.
This doctrine is widely followed and is quite similar to the reasonable use
rule applied to riparian users of surface water. As under the absolute
ownership doctrine the landowner may freely take water from under adjacent
land, but his own use must be a reasonable one.
The Correlative Rights Doctrine has been developed by the California
courts. This doctrine recognizes the finite limit of groundwater
resources, and in time of shortage holds that the available water is to be
equitably apportioned among the overlying owners.
As Sax (1968) noted these four doctrines are not rigidly followed by any
one state, but rather, the individual states have developed regulations
based on varying combinations of the four doctrines.
Alaska Water Supply Problems:
In an area with a seeming abundance of water, it may seem inappropriate to
talk of water supply problems. However, there are periods of the year
where water may become almost unavailable. The reason for the shortage
varies with location. In the Southeast where stream flow is the only
feasible source of water supply, a short drought can result in the drying
up of streams not receiving lake outflows because there is such a lack of
alluvium for natural storage. What used to be Southeastern Alaska's river
valleys are now fjords.
In the interior, during the long cold winters small streams and shallow
lakes freeze solid and water is made available only by melting snow or ice.
Even if interior lakes do not freeze solid there is often a water quality
10-5
problem due to the increase in dissolved solids in the unfrozen fraction.
The long winters also cause a decline in stream flow and then at spring
breakup there is a large runoff peak. Thus, seasonal variation means
storage is a requisite for any power development. There are few, if any,
streams in Alaska on which a run-of-river installation would be feasible.
In some areas of Alaska, there is a shortage of good quality water. In
many places the groundwater is high in iron, organic matter, or both. Most
of the shallow wells around Anchorage and also in other areas yield water
with high iron content.
History of Alaska Law:
Alaska did not become a territory with the power to enact its own laws
until 1912. The territorial law was the doctrine of prior appropriation in
almost pristine form. In Alaska, as in California, miners competed for
water needed to wash gold from alluvial deposits of their placer claims.
The prior appropriation doctrine arrived in Alaska via Oregon whose laws
relating to real estate were made applicable to the "District of Alaska" in
a mining case decided in 1905.
Alaska did not adopt elaborate procedures for acquiring and recording
rights like other western states. Several mining cases at the turn of the
century hinted that Alaska might become a "California doctrine"
jurisdiction by giving effect to both appropriative and riparian rights.
However, a decision by the Circuit Court of Appeals in 1910 (Van Dyke v.
Midnight Sun Mining and Ditch Co.) held that riparian rights were
inapplicable to Alaska and the territory was added to the list of "Colorado
doctrine" jurisdictions.
10-6
Resources, Fish and Game, and Public Works employed Frank J. Trelease as a
consultant to draft a water code. Trelease, at that time, was a law
professor at the University of Wyoming. He apparently enjoyed his grand
tour of Alaska and visited state leaders and submitted a code in January
1962.
Trelease's code was packaged as a bill and submitted to the legislature.
The code had six articles:
(1) Organization
(2) Appropriation
(3) Water Pollution
(4) Conservation of Public Waters
(5) Drainage and Flood Control
(6) Water Conservancy Service Areas
The code failed to pass. Trelease believed that it failed because it was
too comprehensive and addressed problems that were too distant. Repeated
attempts at passage failed until a scaled down bill dealing mostly with
appropriation was passed in 1966. The Water Use Act gave statutory
definition to the doctrine of prior appropriation mandated by the
constitution. The Act applied to all state waters, ground and surface not
subject to superior federal rights.
Definition of Water Course:
The Act recognizes the unity of the hydrologic cycle by putting all water
in one class. Alaska thus avoids the endless legal tangles that have
resulted from other states attempts to divide water into legal classes and
to apply different rules of law to different types of water occurrence. In
fairness, most of these classes (water courses, diffused surface waters,
10-8
springs, percolating water, underground streams, seepage, subflow) were
invented a century ago and had some practical or pseudoscientific reason
for their existence. Today, they are worse than useless. They require
different rules of law to what is basically the same thing.
For example, in many states there is a distinction between "diffused
surface water" and "water in a water course." In many states the latter
can be appropriated while the former is the land owner's property.
Trelease does an interesting thing with a Wyoming case, State v. Hiber, in
which the court reviewed a number of defintions of a ··water course" and
their exceptions. When these are paralleled, the resulting definition goes
something like this: A water course is a stream of water (except that the
water need not always flow) in a definite channel, having a bed and banks
(except that sometimes it may lack banks), usually flowing in a particular
direction (but lo, a slough is a water course, though it connects two
rivers and changes its direction according to which is higher) and
discharging itself into some other stream or body of water (except for
creeks which disappear into sand dunes). In Alaska a water source is
defined simply as a "substantial quantity of water capable of being put to
beneficial use."
Acquiring ~Water Right:
No person can acquire a water right other than by complying with the Act
and obtaining a permit. In fact, all diversions of a significant amount of
water without compliance are made criminal acts. Alaskas thus will not
follow Idaho's mistake of allowing a parallel system of unrecorded
appropriations develop by allowing rights to develop by diversion and
application to use.
10-9
The Alaska Act does not allow water rights to be obtained by adverse use or
possession. Utah, Nevade, Idaho, and Montana have all had problems with
adverse possession at one time or another.
Permits:
The right to appropriate water is obtained by application to the
Commissioner of the Department of Natural Resources. The permit system was
invented in Wyoming in 1890. It gives the state the power to protect
itself from undesirable uses. It allows for underdevelopment of water
resources to protect in place, non-appropriate uses such as recreation,
fish habitat, or waste disposal.
The permit requires four conditions be met:
1) Rights of prior appropriators must not be unduly affected.
2) Proposed means of diversion must be adequate.
3) Proposed use is beneficial.
4) Appropriation must be in public interest.
These provisions seek to protect the public, but also to permit the
development of resources for individual benefit but without undesirable
detriments to society as a whole.
Different Viewpoints:
The real strength of the Act is in setting up procedures which allow all
viewpoints to be brought together. Although the Department of Natural
Resources (DNR) is given the authority to adjudicate water rights, the DNR
must also give notice of the application to the Department of Fish and
Game, the Department of Health, and at DNR discretion any agency,
organization or person. Any person denied a permit may appeal to the
superior court.
10-10
Problems:
After the passage of the Water Act, the Water Management Section of DNR was
inundated with declarations of appropriation, some of which have still not
been processed. The number of applications has continued to grow. With
this large backlog it is impossible to determine how much water is being
appropriated and what effect an appropriation would have on water and other
resources. Oil development and the resulting population growth aggravated
these problems, and the effects of limited management have begun to
manifest themselves.
The DNR is seeking an amendment to the Water Use Act for the maintenance of
minimum stream flows. The Act now requires that there be a diversion,
withdrawal, or impounding of water before water can be appropriated.
Conclusion:
From the position of Western Water La«, the Alaska Water Use Act is the
most recent stage in the evolution of prior appropriation which saw its
last major mutation in 1890 in Wyoming. It would be impossible to export
the Act but it may have some value as a model. Alaska was fortunate that
it did not need ala« to stop undesirable practices and activities. The
Act was not enacted in an emergency to correct a bad situation. The State
of Alaska was indeed fortunate to have expert advice in drafting a forward-
looking law designed to prevent emergencies and to protect present and
future water uses.
10-1 1
REFERENCES
Cox, w., Miller, R.E., et als, "Water Law Primer," Journal of WRP&M
Division, ASCE, vor:-108, No. WR1, March 1982, P• 107.
Curran, H. J. and Dwight, Linda Perry, Analysis of Alaska's Water Use
Act and its Interaction with Federal Reserved Water Rights,
lrlStitute Of Water Resources, University of Alaska, IWR 98,
February, 1979.
Dewsnut, R. L., Jensen, D. w., and Swenson, R. w., A Summary Digest of
State Water Laws, National Water Commission, 1973.
Sax, J. L., "Water Law Cases and Commentary," (unpublished), 1965, p. 37.
Sax, J. L., Water Law Planning and Policy, 1968, p.463.
Trelease, F.J., "A Water Code for Alaska,'' Report to State of Alaska, 1962.
Trelease, F. J., "Alaska's New Water Use Act," Land and Water Law Review,
Vol. II, 1967, 1-49.
Trelease, F. J., Water Law: Cases and Materials, Third Edition, West
Publishing Co., 197~
10-12
ASBESTOS LEVELS IN ALASKAN DRINKING WATER: A PRELIMINARY STUDY
By Helen A. Myers 1 and Edwin S. Boatman2
Abstract
Thirty-five samples of water from rivers and other drinking water sources
were collected for analysis of asbestos content. Water from wells or public
treatment plants generally did not contain significant amounts of asbestos,
even if the water source or a nearby river contained asbestos. Water content
of asbestos is discussed in relation to the mineral terranes of the river
watersheds. There is some likelihood that the levels of asbestos found could
increase the incidence of cancer, but toxic levels of ingested asbestos are
not well defined at this date.
Introduction
Since Alaska is a heavily mineralized state, it is not surprising that
these minerals are found in rivers and other drinking water sources. Asbestos
is one of several toxic compounds that appear, as a result of both natural
erosion and mining activity.
Few measurements of asbestos levels in Alaskan rivers were found prior to
the present study, all in the Yukon River. These measurements are given in
Table I. Variation between amounts can reflect differences in analytical
techniques, seasonal variations, or the extent of mining activity. (In 1977
the Clinton Creek Mine on the lower Forty Mile River in Canada was
operational; activity has ceased by 1980). Both of the major types of
asbestos, chrysotile and amphibole, were detected. Chrysotile asbestos is a
long fiber; the term amphibole refers to several different subtypes of
asbestos, all shorter than chrysotile. Although the length of the fiber has
been postulated to contribute to the toxicity of asbestos, both types are
considered to be very hazardous. Thus, the question arises whether asbestos
1Assis tant Professor, \VAMI Program in Medical Education, University of
Alaska, Fairbanks, Alaska, 99701.
2Professor, Department of Environmental Health SC-34, School of Public
Health, University of llashington, Seattle, l<ashington 98195.
11-1
in the Yukon River, and possibly other Alaskan rivers, is present in amounts
that would have adverse health effects.
l~hile a link between the inhalation of asbestos fibers and the later
appearance of asbestosis or various types of cancers has been well
established, the link between asbestos ingestion and disease has not
(Becklake, 1976; Millette, 1981). In 1979 the estimate was made that the
ingestion of 0.3 million fibers per liter of water over a seventy year
lifetime would result in one additional death per 100,000 people (EPA, 1979;
Millette, 1981). This estimate was made from several studies of occupational
exposure, largely by inhalation, making assumptions regarding the amount of
asbestos that would have been cleared from the lung and subsequently
swallowed. Epidemiological studies have looked at populations exposed to
drinking water containing asbestos levels ranging as high as 1,800 million
fibers per liter with varying results (Boatman and Polissar, 1982; Kanarek,
1980; Levy, 1976; Millette, 1981; Toft, 1981). A study of the population
exposed to San Francisco area drinking water did report a statistical
correlation between several types of cancer and the levels of asbestos in the
drinking water. The asbestos levels in this study approximate the range of
asbestos found in the Yukon both as reported in Table I and the present study.
Table I:
ASBESTOS LEVELS REPORTED IN THE YUKON RIVER
(Prior to present study)
Location, Date
Eagle, 1977
5 samples
u.s. border, 1977
Mouth of Fortymile
1980
Eagle, 1980
Group
Chrysotile
Amphibole
Unspecified
Unspecified
Chrysotile
Amphibole
11-2
Amount, million
fibers per liter
126 -674
2150 -6230
200
0.5 -327.2
14.7
Reference
Metsker, 1981
EVS Consultants,
1981
Justice, 1982
A survey of mining claims and the geological formations that might contain
asbestos indicated that asbestos might well be present in more Alaskan
drinking water than that in the region of the upper Yukon. Samples were
therefore collected from several villages and rivers to provide a preliminary
study of the extent of the potential for exposure to asbestos via ingestion.
Methods
Sample Collection:
The United States Fish and Wildlife Service collected samples for the
study from the lower Yukon near the mouth of the Andreafsky River, from the
upper Noatak River (two sites), the Kobuk River (two sites), and the Kuskokwim
River.
Because of other research interests concerned with the possibility of a
link between asbestos and the appearance of primary hepatocellular carcinoma
(a usually rare cancer of the liver) in several western Alaskan villages, most
of the drinking water sources were tested in these villages. A letter was
written to the mayor and/or village health aide of several villages,
explaining the project and inviting participation, at no cost to the
village. Of the respondents, villages were chosen primarily for ease of
access by one collector (Myers) in one trip, so that samples would be
representative of the same period in time. A person specified by the village
met the collector and helped collect the samples from sources identified as
those from which people often obtained water.
Water Sample Preparative Procedure:
All samples were treated with one milliliter of 2% mercuric chloride to
prevent bacterial growth. Samples received from Fish and Wildlife were
11-3
treated when received in Fairbanl<s; the rest of the samples were treated at
time of collection.
The preparative procedures for the analysis of asbestos fibers in water as
detected by transmission electron microscopy are those described under "The
Interim Method for Asbestos in Water" by Anderson and Long {1980). For
asbestos fiber analyses by transmission electron microscopy the preferred
preparation based upon a variety of interlaboratory comparisons is the Carbon
Coated Nuclepore Jaffe Wick method modified from the original technique of
Jaffe.
The sample water was well shaken and filtered as soon as possible after
arrival; if this could not be accomplished the sample was refrigerated at 4°C
to minimize bacterial and algal growth.
Depending upon the turbidity of the sample, a sui table aliquot from 10 to
500 ml was filtered by suction through a 0.1 ~m pore size Nuclepore, 47 mm
diameter membrane supported by a 2.0 ~m pore size Hillipore backing filter.
The backing filter served to ensure a uniform deposition of the particulate
material on the surface of the Nuclepore membrane. After filtration, and
while the Nuclepore membrane was still wet, an equatorial strip was cut out of
the membrane, attached to a glass slide, dried and coated with carbon by
rotation in a vacuum evaporator (the carbon coating serves to trap and retain
the particulate matter during the subsequent dissolution procedure). Random
portions ( -2mm 2 ) were cut from the strip and placed individually on 200 mesh
formvar coated copper rhodium electron microscopy grids. The grids were
exposed to chloroform vapor in a biohazard cabinet and the filter matrix
dissolved. This usually took between 18-48 hours. After dissolution of the
filter, the grids were observed by a transmission electron microscope
operating at 80-100 kv which had the capability of selected area electron
11-4
diffraction (SAED). Up to this point, the chance errors had been associated
with the manipulative skills, Sources of error could include settling of
particulates in the water with time; clumping of the fibers during filtration;
a non-uniform deposition of particulates on the Nuclepore membrane surface;
loss of fibers during carbon coating and/or dissolution in chloroform; and
fibers obscured by other organic or non-organic particulates. Beyond this
point, possible sources of error involved those of counting and
identification.
Counting and Identification of Asbestos Fibers:
Grids were first observed at low magnification to see if the deposited
particulates were uniformly distributed, If the distribution were poor or
there were either too much deposit (enough to mask fibers) or too little, then
the whole procedure was repeated using a smaller or larger volume of water for
filtration, In rarer circumstances, where the turbidity of the water sample
was high enough to preclude immediate use, an aliquot was diluted with
asbestos-free water and filtrated again, If the grids were satisfactory, 20
to 30 grid squares were randomly selected from 3 to 4 grids of each sample and
observed at 2l,OOOX for particulates suggestive of asbestos fibers.
Fibers were then rated according to aspect ratio (i.e., parallel sides and
a length/width ratio of greater than or equal to 3:1), morphology and crystal
structure by SAED, After these assessments, the fibers were classified as (l)
chrysotile asbestos (2) amphibole asbestos (3) ambiguous or (4) non-
asbestiform. Periodically throughout analyses the overall instrumentation was
checked by use of preparations of reference asbestos fibers (UICC),
11-5
Results and Discussion
The levels of asbestos found in river water and other drinking water
sources are given in Tables II and III. It should be emphasized that every
locality visited used the rivers as a source of drinking water. Sometimes an
effort was made to avoid excess silt by drawing water from mid-channel, or, as
in Selawik, by travel to a clearer river (in this case the Fish River).
However, some people remarked that silt "was good for you." It is likely that
the asbestos fibers travel with the silt; an experiment in progress is testing
this possibility.
Figure l provides information regarding the possibility that asbestos
deposits might be present in the watershed of a river. The shaded areas
indicate regions in which rock units with which asbestos is typically
associated might be found. These rocks are mafic and ultramafic rocks of
plutonic or mixed volcanic and sedimentary environments (AEIDC, 1979; Levine,
1978). Also included are regions in which asbestos claims are located (Sims,
1982). The areas illustrated on the map are neither all-inclusive nor, at the
scale of the map, very precise. The intent of the figure is to illustrate how
widespread in Alaska are the terranes which are associated with ashes tos.
However, the terranes illustrated do not necessarily all contain asbestos;
they only indicate known geological areas in which asbestos might be found.
The map also indicates approximately the areas from which the water
samples were collected.
The area to the north of the Kobuk River was an important source of
asbestos in 1944-1945, particularly of tremolite, a type of amphibole asbestos
(Bundtzen, 1982), It was thus surprising that the amount of asbestos in the
Kobuk was so much lower than the amounts in the Yukon, and that only
chrysotile was present. It is possible that tributaries from watersheds not
11-6
containing asbestos had a dilution effect; the samples were taken near the
mouth of the river.
The Noatak is also regarded as draining areas containing asbestos. A
claim has been staked on a tri bu ta ry be tween sample sites 1 and 2; this region
may therefore be the source for the asbestos in the sample drawn at the
village of Noatak.
The Kuskokwim watershed also contained a producing mine in the NeGra th
region. Deposits of asbestos have been sighted in the region of the lower
Kuskokwim as well. However, none of the samples drawn from the Kuskokwim
(just above the town of Bethel) contained asbestos.
The town of Selawik is located in a region of islands surrounded by a
network of water channels. It is possible that the asbestos found in the
sample taken from the Selawik River in town originated in the terrane to the
east drained by the Selawik. The Fish River drains the Waring !1ountains,
which are separated by another valley from the mountains in the known asbestos
terrane to the north. The presence of asbestos in the Fish River may indicate
that the terrane is more extensive than previously thought.
The asbestos detected in the North River to the northwest of Unalakleet
also cannot be related to any known asbestos related formations.
The largest amounts of asbestos were found in the Yukon River. The Yukon
already contains asbestos as it leaves Canada (see Table I) both from the
contribution of the Fortymile River and regions further upstream. A "world
class" desposit of asbestos is located in the Slate Creek area of the
Fortymile river system. It was during the operation of the Clinton Creek
mine, near the mouth of the Fortymile, that the asbestos levels were recorded
at the highest levels (1977) recorded in Table I; the reasons behind the
differences in the values for this year are not clear.
11-7
I
CP
NOATAK RIVER
1. Noatak
2. Noatak Canyon
3. Cutler River Junction
KOBUK RIVER -----
NORTH RIVER
Unalakleet
Stebbins
YUKON RIVER
4.
5. Hountain
6. Goose Island
7. Pilot Station
8. Circle
KUSKOKWIH RIVER
Bethel ------
'
" 6A ~Figure
(/'Ou
Potential Asbestos
Bearing Terranes
1; Site of Water Sample
Collection
Potential Asbestos Bearing
Terranes
Table II
ASBESTOS LEVELS IN DRINKING WATER SOURCES
(Excluding rivers, see Table III)
Source, Date
Noatak, 6 August, 1982
PHS hose; well in Noatak River
Health Clinic tap, another well in river
Unalakleet, 9, 10 August 1982
City Hall tap, public water from well
Me Too Creek, culvert between Musk Ox Farm
and White Alice Station
Stebbins, 10 August 1982
Elementary school, own well
Public water supply in laundry,
source in volcanic lake
Rain \~ater, off metal roof into metal can
Alakanuk, 12 August 1982
Clear Lake, at end of path from Yukon channel
Safe water plant, source in another channel
passing village
11ountain Village, 12 August 1982
Health Clinic tap, main well
High School tap, different well
Elementary school tap, different well
Spring on bank of Yukon, downstream
from village
Pilot Station, 13 August 1982
Spring on side of hill
Pump house tank, public water from well
Bethel, 20 August 1982
Yukon Health Clinic tap, own well
Hospital tap, own well
Circle, 29 August 1982
McDonald's we 11, on bank of Yukon
11-9
ASBESTOS CONTENT
Million fibers per liter
Chrysotile Amphibole
0.1
1.3
1.0 0.1
1.2
5.3
o.s
0.3
Table III
ASBESTOS CONTENT OF SELECTED ALASKAN RIVERS
Source, Date
ASBESTOS CONTENT
Million fibers per liter
Chrysotile Amphibole
----~-------------------------------------------------------------------------
Selawik River, in Selawik, 5 August 1982
Fish River, near Selawik, 5 August 1982
North River, at bridge near Unalakleet
Noatak River, 19 July, 6 August 1982
Above Noatak Canyon, at game warden cabin
1000 ft. Noatak, where drinking water
collected
Kobuk River, 16 July 1982
Above junction of Riley Channel
At mouth of Riley Channel
Kuskokwim River, upstream from Bethel
11 November 1981 (three samples)
20 August 1982
Yukon River
Circle, 29 August 1982
Slough where drinking water collected
Bank below slough, main channel
Goose Island (off Pitka's Point)
11 November, 1981 (two samples)
3 June 1982 (three samples
just after breakup)
Delta, main channel near Alakanuk
12 August 1982
16.4
2.0
2.6
1.9
1.9
1.2
32.7
42.9
33.2
41.8
5.7
3.7
8.9
22.5
6.1
22.5
13.5
8.4
1.4
1.0
21.0
------------------------------------------------------------------------------
Samples taken in 1982 (Table III) did not show much difference between the
amounts of asbestos in the upper Yukon (Circle) and the lower Yukon (delta,
near Alakanuk). While several large and many small tributaries between the
border and the delta might be expected to dilute the asbestos from the upper
region deposits, it can be seen from Figure 1 that these tributaries may, for
the most part, also contain asbestos.
1 1-10
If the asbestos in the rivers comes from natural erosion of the
watershed, a seasonal variation might be expected. Such a variation is seen
in Table I: the high value of 327 million fibers per liter was obtained in
July, while only 0.5 million fibers were detected in November. The variation
seen at Goose Island in the lower Yukon did not follow the same time pattern,
however. It is possible that the lower Yukon contains asbestos eroded by the
Yukon itself, and that in winter this asbestos is not diluted by rainwater or
water from tributaries. More data regarding seasonal variation is needed. To
estimate the total exposure of the population to ingested asbestos it is
necessary to know this variation. The differences between the variation in
the upper and lower Yukon point out the need for data rather than assumptions
in accessing year-round exposure.
It would appear that knowledge of the geology of river watersheds at the
level presented in Figure 1 is not enough to predict the presence or the
amount of asbestos in the water of major rivers draining a large area.
However, in most instances the presence of asbestos in the water could be
related to presence of appropriate geological formations. The presence of
asbestos in the Fish and North Rivers could represent a sampling error or
indicate the presence of as yet undetected sources of asbestos.
In two locations, Circle and Noatak, water from wells either in the river
(Noatak) or nearby on the bank (Circle) contained considerably less asbestos
than did the river water. It seems quite probable that the ground filters out
the asbestos before the river water percolates through to the well.
The safe water plant in Alakanuk drew its water supply directly from the
Yukon. The treatment system seemed quite effective in removing asbestos,
since the level dropped from about 20 million fibers each of chrysotile and
amphibole forms in raw river water to levels that were undetectable in treated
water.
1 1-1 1
It should be kept in mind that the asbestos levels reported in this study
cannot be taken as the representative asbestos level for the sources, since
generally only one sample was taken from each source, Experimental
variability due to the factors of sampling and measuring are thus not
assessable. In addition, it is clear that the levels could have a seasonal
variation.
It is difficult at this point to make a statement regarding the possible
health effects of the asbestos found in Alaskan drinking water. Firstly, the
results just reported are preliminary and do not adequately define the
exposure level, which must consider year-round variation. Secondly, neither
animal nor epidemiological studies have produced conclusive results clarifying
the question of what levels of asbestos in water are definitely toxic. The
estimation of 0,3 million fibers per liter by the Environmental Protection
Agency has not been adequately supported by animal and epidemiological
studies, The San Francisco study, which did report a correlation between
asbestos ingestion (at levels found in the Yukon and Selawik Rivers) and
cancer development has been criticized for its statistical methods, The lack
of a statistical correlation in other studies can be due to the long latency
{decades) between exposure and appearance of cancer, to the impossibility of
maintaining a stable population to study, and/or to the difficulty in
establishing the actual exposure of individual cases. A more definite
statement regarding the toxicity of ingested asbestos must await replication
by other studies of the studies to date, in a time frame that satisfies
concerns that the expected cancers had enough time to develop.
11-12
Conclusions
The preliminary data presented indicate that asbestos is present in
varying amounts in rivers as well as other drinking water sources. Seasonal
variations may not be qualitatively similar from site to site. Research to
date has not provided a firm figure for levels of asbestos ingestion that
would be toxic. However, the amounts of asbestos measured in the drinking water
would, according to some of the current experiments and estimates, present
some risk for the development of cancer.
References
Anderson, C.H. and Long, J.M., 1980, Interim Method for Determining Asbestos
in I< a ter. EPA-ORD Research Report IlEPA 600/4-80-005, p. 34.
Arctic Environmental Information and Data Center, 1979, Mineral Terranes of
Alaska (a map series). Anchorage, Alaska.
Becklake, M.R., 1976, Asbestos Related Diseases of the Lung and Other
Organs: Their Epidemiology and Implications for Clinical Practice. American
Review of Respiratory Disease 114:187-227.
Boatman, E.S. and Polissar, L. 1983, Unpublished research.
Bundtzen, T.K. and Smith, T.E., 1982, Alaska's Industrial Minerals. Alaska
Geographic 9:64-73.
EVS Consultants, 1981, Assessment of the Effects of the Clinton Creek Mine
l<aste Dump and Trailings, Yukon Territory. For Cassiar Resources Ltd.,
Vancouver, B.C. Canada.
Environmental Protection Agency (U.S.), 1979, Water Quality Criteria:
Asbestos. Federal Register 44:56632-56635.
Justice, S. (Alaska Department of Environmental Conservation, Fairbanks,
Alaska) 1982. Personal communication regarding analysis by E.S. Boatman.
Kanarek, M.S. et al, 1980, Asbestos in Drinking l<ater and Cancer Incidence in
the San Francisco Bay Area. American Journal of Epidemiology 112:54-72.
Levine, R.J., ed., 1978, Asbestos: An Information Resource. DHEW Publication
Number (NIH) 79-1681.
Levy, B.S., et all, 1976, Investigating Possible Effects of Asbestos in City
l<ater: Surveillance of Gastrointestinal Cancer Incidence in Duluth,
Minnesota. American Journal of Epidemiology 103:362-368.
11-13
Metsker, H., (United States Fish and Wildlife Service, Anchorage, Alaska),
1981, Personal communication.
Millette, J.R., 1981, The Need to Control Asbestos Fibers in Potable Water
Supply Systems. Science of the Total Environment 18:91-102.
Sims, J. (Office of Mineral Development, State of Alaska, Fairbanks), 1982,
Personal communication of computerized search of U.S. Bureau of !1ines records
for status of asbestos claims.
Toft, P. et al, 1981, Asbestos and Drinking l~ater in Canada. Science of the
Total Environment 18:77-89.
11-14
EFFECTS OF GOLD PLACER MINING ON INTERIOR ALASKAN
STREAM ECOSYSTEMS
Jacqueline D. LaFerriere 1, David
Simmons3, Erwin Van Nieuwenhuyse4,
James B. Reynolds 0
Abstract
M. Bjerklie2, Rodney C.
S. Mitchell Wagener5, and
During the summers of 1982 and 1983, we evaluated the effects
of placer mining for gold on the water quality and ecology of
streams in interior Alaska, northeast of Fairbanks. Our field
studies involved two sets of paired watersheds, each having
one with placer mines and one without. Increased suspended
sediments that settled out downstream cemented the streambed,
causing the surface flow and the groundwater flow to be
hydraulically isolated from each other. As a result the water
chemistry of the mined streams was different than that of
their unmined partners. The mined streams were often lower in
hardness, alkalinity, and specific conductance. Due to mining
activity, the mined streams were higher in settleable solids
and turbidity. Total heavy metals were also elevated in the
mined streams.
Benthic algae, the base of the food chain, were severely
reduced in heavily mined streams. Whether this was due to
increased turbidity cutting off light for photosynthesis,
scouring, or toxic heavy metals was not determined. Benthic
macroinvertebrates, mostly aquatic insects, were also reduced
in numbers and species; these animals are the critical link
between algae and fish in the food chain. Although the unmined
streams contained many Arctic grayling (.'!.£1.mallus arcticus),
no grayling were found in the mined streams when mining
produced heavy loads of suspended materials. The only
exception occurred during fall out-migration, when grayling
were running to overwintering grounds in the large glacial
rivers. Cage bioassays demonstrated that if grayling could not
1Assistant Leader, Alaska Cooperative Fishery Research Unit,
U of Ak., Fairbanks 99701 (The Unit is jointly sponsored by
the U.S. Fish and Wildlife Service, Alaska Department of Fish
and Game, and University of Alaska, Fairbanks.)
2 Research Assistant, Institute of Water Resources, U of Ak.,
Fairbanks 99701
3, 4, 5Research Assistant, ACFRU
6Leader, ACFRU
12-1
escape from streams carrying mining sediments, they would
suffer physiological harm including reduced feeding, slowed
maturation, and gill damage. We conclude that placer mining
sedimentation severely reduces the aquatic life in heavily
mined interior streams, and results in reduced biological
carrying capacity of the affected watersheds.
Introduction
In 1980 ·the u.s. Environmental Protection Agency published a
request for proposals, soliciting responses to study the
effects of placer mining in Alaska on salmonid fishes. This
request was issued from a new program called "Partners in
Research" which funds academic institutions to conduct some of
the research needed by the EPA to address its mission. The
University of Alaska-Fairbanks, through the Institute of Water
Resources, and the Alaska Cooperative Fishery Research Unit,
proposed a comprehensive study of the stream-ecosystem effects
of placer mining.
We proposed an ecosystem approach because we felt that
important effects of mining on fishes might be not only direct
(e.g. gill damage), but indirect (e.g. through reductions
along their food chain). Fish are mobile enough to avoid
localized increases of stream sedimentation, but their food
organisms, the benthic invertebrates, are less mobile. The
algae on which the macroinvertebrates feed are mostly benthic
and sessile in streams (only large rivers have true
phytoplankton), and thus are very vulnerable to increased
suspended sediment.
Our proposal was funded in August, 1981. Because of
limitations on available funds, some of the work we originally
12-2
proposed could not be conducted, Detailed engineering studies
of mining practices and water use could not be included.
Studies of the mobilization of heavy metals associated with
this activity also had to be cut from this project. However,
those mobilization studies have been conducted by Dr. Edward
Brown of the Institute of Water Resources with other funding.
The data presented in this paper should be considered
preliminary at this time, since data analysis for the 1983
field season has just begun. We hope that the dialogue
generated by this manuscript at the meetings will improve our
analysis and subsequent journal publications.
Study site description
This study was carried out at two major sites accessed from
the Steese Highway northeast of Fairbanks (Figure 1). The
first site surrounded the confluence of McManus Creek with
Faith Creek to form the Chatanika River, which then flows
southward, ultimately entering the Yukon River after joining
the Tolovana. This site is near milepost 69 on the Steese
Highway (69 miles from Fairbanks). At least four mines are
located on Faith Creek and its tributaries while the McManus
Creek watershed has no mines.
Our second site was located around the confluence of
Twelvemile Creek with Birch Creek near milepost 94 of the
Steese Highway. Twelvemile Creek has no active mines while
Birch Creek above the confluence has on the order of ten
mines. Below the confluence, Birch Creek is classified as a
Wild and Scenic River under the Alaska Native Interest Land
12-3
1\)
I
.p:-
<v":J'v CLEARY
"-..(() SUMMIT
":J
FAIRBANKS
Elev. 436'
Precip. II"
40 yrs Records
%
I
(NO SCALE)
" ~
<::
"' "" .E ( EAGLE
~ SU,MMIT
-~
CENTRAL
Elevation 960'
Precip. 10"
10 yr Record
CIRCLE
HOT SPRINGS
Elev. 935'
Preci p. II"
30 yrs Records
Figure l. The Study Area with Annual Precipitation Data for Three Stations.
Claims Act for approximately 160 river miles until it passes
under the Steese Highway at milepost 138.
A secondary study site was established at Ptarmigan Creek, a
clearwater tributary of the Eagle Fork of Birch Creek, where
some algal productivity work was conducted in 1982. In August
of 1983 mining was started for the first time on this creek,
and algal productivity above and below mining was measured for
a three day period. Secondary sites were also established in
1983 for invertebrate sampling at Ptarmigan, Mammoth, Boulder
and Ketchem Creeks. The latter three streams are tributaries
of Crooked Creek which flows into Birch Creek before it enters
the Yukon River.
Methods
The description of methods which follows is somewhat detailed.
We have presented this detail in anticipation of questions
about our methodology. The reader may prefer to refer to this
section only as necessary.
Precipitation
Twelve rain gauges were placed along the Steese Highway.
Terrain along the highway includes low valleys with forest
vegetation (birch, white spruce, and black spruce) and higher
elevations with muskeg and tundra vegetation. The network of
gauges was designed to provide a representation of surface
distribution of rainfall. In order to obtain a representative
elevational distribution, some gauges were placed off the
highway at higher elevations. The rain gauges used were non-
12-5
standard. In order to obtain a check on the accuracy of the
measurements, two brass standard Weather Bureau gauges were
placed alongside one of our gauges. Our gauges were
constructed with graduated cylinders and funnels, and
calibrated to be read with a dipstick. Volume measurements
were also taken to check the dipstick readings. The standard
gauges not only helped to verify the results obtained from the
homemade gauges, but also allowed for catching an overflow
which our gauges do not. Oil was placed in each gauge to
reduce evaporation. They were read and emptied weekly.
Stream flow measurement
Stream flow was measured using a Marsh-McBirney meter to
determine velocities across a section of each river. Seven to
ten flow measurements were taken over the period of the summer
and correlated with staff-gauge heights in order to develop a
stage-discharge relationship for each stream. Gauge heights
were read daily while at each camp and periodically at the
other sites when visited. Flows not measured directly were
calculated from the rating curves.
Groundwater
Well points were driven in the stream beds in pools at Birch,
Twelvemile, Faith and McManus Creeks to obtain weekly water
levels and water samples. The chemistry of the groundwater
just beneath the streams and the peizometric surface of the
gr6undwater were determined. Each well point was driven into
the streambed at least 6 inches past the screening. The wells
were hand pumped to obtain samples for measuring water
12-6
quality constituents other than dissolved oxygen. Dissolved
oxygen samples were pulled under J:iand-pump-generated vacuum
through 1/4 inch tubing into a 1000 ml Erlenmeyer flask, after
the well had "rested" for at least 24 hours. Samples were then
siphoned to the bottom of 300 ml BOD bottles, overflowing the
bottle at least twice before obtaining the sample. Samples
were fixed with dry chemicals (Hach Chemical Company) and
titrated with phenylarsine oxide which had been standardized
against an iodate-iodide standard. Percent saturation was
calculated according to Mortimer (1981).
Water Quality
Stream water samples were collected by grab sampling. Ground-
water samples were collected by pumping the wells for several
minutes, to replace the water standing in the well several
times, before sampling. Both types of samples were collected
weekly. Rain water samples were collected on tent canvas to
obtain adequate surface area to collect large enough samples.
The canvas material was checked for possible contamination,
and none was found. In addition, rain was collected with
funnels for comparison and no differences in water chemistry
were found.
Accepted procedures for measuring pH, temperature, conduc-
tivity, alkalinity, total and calcium hardnesses, iron,
copper, silica, manganese, ammonia-N, nitrate-N, nitrite-N and
color were used. Our primary tool for water quality
measurements was a Hach field test kit containing a
spectrophotometer, conductivity meter and pH meter. The
12-7
method of standard additions was used periodically to check
the accuracy of quantitative analyses. Turbidity was measured
using a Hach portable turbidity meter. Suspended sediment
samples were taken with a hand-held suspended-sediment
sampler, and residues were determined in the lab using
procedures from Standard Methods (A.P.H.A. 1980). Settleable
solids were determined in the field using Imhoff cones.
Chemical oxygen demand was measured by the reactor digestion
method (Jirka and Carter 1975). All constituents were
reported as mg/1 except temperature (°C), pH, conductivity
(pmhos), turbidity (NTU) and settleable solids (ml/1).
Algae
In this two-year study, stream algal productivity was
estimated using the
method (Odum 1956).
now standard Odum diel oxygen curve
Two YSI Model 56 dissolved oxygen (D.O.)
monitors, calibrated against five Winkler titrations prior to
each measurement period, continuously recorded D.O. concentra-
tion and temperature. These data were digitized, computer-
ized, then subjected to a FORTRAN program which calculates
values for change in D.O. concentration (corrected for
diffusion), plots them against time, and then solves for gross
daily production and respiration using the trapezoid rule.
Daytime respiration was accounted for by connecting the pre-
dawn and post-sunset minimums as suggested by Hall and Moll
(1975). We used the tables of Mortimer (1981) to calculate
oxygen saturation and the formula of O'Connor and Dobbins
(1956) to derive the reaeration coefficient. Wilcock ( 1982)
12-8
showed that the O'Connor-Dobbins formula is best suited for
use in relatively small headwater streams. The reaeration
coefficient was corrected for temperature according to Elmore
& West (1961). All calculations were based on mean
temperature and D.O. values of the appropriate time interval
(either 1 hr or 0.5 hr). The single station method was used
during the 1982 field season, but was verified as being
equivalent to the dual station method during the 1983 season
when two D.O. monitors became available. The amount of
photosynthetically active radiation (PAR) reaching a study
site was recorded continuously during each study period using
a LI-COR 190SB deck sensor and a LI-COR 550B printing
integrator. The deck sensor remained in an unshaded area
throughout the study period.
Data used to calculate the light extinction coefficient were
collected with a LI-COR 188B integrating quantum radiometer
photometer and two sensors; one 190SB sensor for surface
readings and one 192SB sensor for underwater readings. A LI-
COR SS-3 sensor selector permitted readings to be taken
alternately at the surface and at depth. This was done at 20
second intervals using an integration time of 10 seconds and
repeated at least six times for each depth. The entire
process required an average of one hour. The deck sensor was
positioned in an unshaded area on shore, while the underwater
sensor was mounted on a #5 rebar section with a clamp.
Measurements were taken at 5 em intervals beginning at 5 em
beneath the surface. The three key plant nutrients we
12-9
measured were inorganic carbon (from alkalinity, pH and
temperature), nitrogen (as ammonia,
phosphorus (as total phosphorous).
nitrite and nitrate), and
Total phosphorus samples
were stored frozen and analyzed within three months according
to the method presented in Eisenreich, et al (1975).
Invertebrates
Stream benthic invertebrates were sampled over the course of
two summers. In 1982, four streams, Faith Creek, McManus
Creek, Birch Creek, and Twelvemile Creek were sampled using a
O.lm2 box sampler. Twenty five non-random benthos samples
each were taken from Faith Creek and McManus Creek. Fifty
non-random benthos samples each were taken from Twelvemile and
Birch Creeks. All the sampling done in 1982 occurred in July
and August.
In 1983, nine streams were sampled for benthic invertebrates.
These streams were the four above and, Ptarmigan Creek,
Ketchem Creek, Boulder Creek, Mammoth Creek, and the Chatanika
River downstream from the confluence of Faith and McManus
Creeks. Ketchem, Mammoth, Birch, and Faith Creek were mined.
Ptarmigan Creek was clear and unmined until the middle of
August when mining on this creek began. The Chatanika River
is formed from the confluence of a mined creek and an unmined
creek. The remaining sampled creeks were unmined. These
streams were sampled randomly during each of six sampling
periods evenly spaced from mid-June through September. Five
box samples were taken from each stream during each sampling
period for a total of thirty random samples from each stream
12-10
for the year. Substrate particle size composition and percent
embeddedness were estimated at the point at which each sample
was taken. Concurrently with benthos sampling, turbidity,
settleable solids, suspended solids, alkalinity, calcium and
total hardnesses, pH, water temperature, conductivity, color,
and chemical oxygen demand were measured once for each stream
during each of the six sampling periods.
sampled once for each stream.
Heavy metals were
Since all of the 1982 samples were taken non-randomly within
the streams, the tests for significant differences in numbers
of benthic organisms between streams were done using non-
parametric statistics. The 1983 samples were taken randomly
and the date will be analyzed using param~tric methods.
Fish
Mined and unmined streams were seined occasionally using fine-
meshed seines to attempt to capture grayling residing in them.
Approximately 150 seine-hauls were made in each of the four
major streams.
To assess the impacts of mining sediment on grayling, two life
stages of these fish were used for experimentation: young-of-
the-year (YOY) fish from the Twel vemile Creek, and juvenile
fish (Age II) transported from Pile Driver Slough, a backwater
of the Tanana River near Fairbanks. These year classes were
chosen for testing because they could be captured in
sufficient numbers and would likely be more sensitive to
sediment exposure than adult grayling.
12-11
YOY grayling were held in 18 x 8 inch cylindrical cages with
1/8-inch mesh size screening,
placed in the mined streams
10 fish per cage. Cages were
and exposed to the existing
sediment levels. Exposure periods varied from 24 hours to 10
days. Control cages were kept in the clearwater unmined
streams for equal durations. The fish were monitored for
mortalities and later sacrificed for examination of gill
tissue, stomach contents, and external and internal appearance
after exposure. Gill-tissue samples were fixed in Bouin's
solution, paraffin embedded, sectioned at 0.7 micron
thickness, and stained with hematoxylin-eosin solution.
Stomach contents were examined using a low power (15x)
compound dissecting scope and analyzed by percent volume and
number of individual organisms in each taxonomic order.
To study stress responses in grayling, Age II fish captured
from Pile Driver Slough were anesthetized with MS-222 at a
dosage of 10 ppm; transported to the two study sites; and held
in a 10 x 30 foot pen in the control streams to acclimate for
48 hours. These fish were then held in 2 x 4 foot cylindrical
cages with 1/4 inch mesh size screening, with 10 fish per
cage. Exposure periods varied from 6-96 hours. After each
exposure period, each fish was weighed to the nearest 0.01 gm,
measured <:t. 1 mm), and two blood samples taken in heprinized
microhematocrit tubes (O.S:t_ 0.05 mm I.D.) from the sev~red
caudal peduncle. Leucocrit and hematocrit values were
determined using the method of Wedemeyer and McLeay (1981).
During all caged bioassay experiments four turbidity and
12-12
settleable solid samples, and one suspended sediment sample
were taken daily.
Results and Discussion
Precipitation
There are no long term precipitation records for the immediate
vicinity of the study area. However, records from Fairbanks,
Gilmore Creek, Central and Circle Hot Springs are available.
The more detailed study of precipitation made for the study
area during the summer of 1983 has revealed distinct patterns
(Figure 2). Ten rain gauges, distributed along a 35 mile
stretch of the Steese Highway, from Faith Creek to Eagle
Summit showed that rainfall is concentrated on Twelvemile
Summit, dropping off to the east and west; and that, over the
summer greater precipitation occurred at greater elevations,
but on a weekly basis elevational trends were much less
apparent (Table 1).
Observations taken over the course of the summer showed storms
moving from the north, west and south, and that frontal storms
produced longer duration rains and were more prone to be
affected by elevation, whereas thunder storms often produced
heavy rain but occurred more randomly.
In addition to the rain gauge depths, pH, conductivity,
alkalinity and hardness. measurements were taken on the rain
water, occasionally. It was interesting to find that pH's
were often below 5.00, and on several occasions below 4.00
(Table 2). With consecutive storms, the pH increased,
. 12-13
'
II",
To Fairbanks
70 Miles _ .......
:.,;A
'-"-
Finure 2 b •
c uununer Preci . plta tion _ 1983
13"
I
I
I
12"
I
I
I
\
\
\
\
\
\
\
\
\
\
10" To Central
I 30 Miles ;r----~ ~
\
\
10"
~--L---1--'--j_-0 -___ j
5 10
MILES
TABLE 1. Precipitation
Dates Faith Me Faith Idaho Twelve-Twelve-Reed Birch Ptar-Eagle
Creek Manus Rd mile mile migan
Le:f't Right
5/23-6/1 0.15" 0. 1 2 0.08 0. 1 3 0.06 0. 12 0.07 0.09
6/1-6/6 0.42 11 0.44 0.56 0.62 0.62 0.47 0.37 0. 2 1 0. 13 0. 1 5
6/6-6/14 0.10" 0. 13 0.04 0.28 0.17 o. 19 0. 1 4 0.29 0.36 0.48
6/14-6/19 0.09 0. 1 0 0.18 0.14 o. 1 8 0. 21 0. 19 0.15 0. 1 4 0.18
6/19-6/29 0.78" 0.63 "'1.50 1.0 5 "-'1.50 "-'1.50 ~ 1.50 'V 1.50 1. 3 5 'V 1.50
6/29-7/3 0.40 11 0.38 0.27 0.22 0.43 0.45 0.52 0.68 0.40 0.45
f\J
I 7/3-7/13 1.39" 1. 17 1.33 -v1.50 ·'--1.50 1 .a 5 "-' 1.50 1.28 1.3 1 1.08
\.11
7/13-7/18 1.25 11 1 .3 2 "'1.50 -v1.50 "-1.50 1 .a 2 1. 4 0 0.9 6 "-'1.50 1.26
7/18-7/28 0.12 11 0. 15 0.12 0.49 0. 61 0.42 0.34 0.3 4 0.31 0.28
7/28-8/1 0.03 11 0.02 o.o 1 o.oo 0.02 0.03 0.02 0. 11 0.03 0.06
8/1-8/10 0.98 11 1.00 0.60 0.67 1. 12 1. 02 0. 7 1 0.47 0.52 0. 3 1
8/10-8/15 0.57" 0.60 0.85 1 .27 1 .o 4 1.05 1. 07 1.29 1.28
8/15-8/24 1 .2 7" 1.35 ~ 1.50 ~ 1.50 ~ 1 :so 1.55 -v1.50 1 .69 1 .0 8 1 .08
8/24-9/2 '-'!.50" -v1.50 .~1.50 "-1.50 ..v1.50 2.00 N 1.50 1. 3 7 1.13 ....... 1. 50
.f ·* -f
TOTAL: 9.05" 8.91" 10.04 11 10.86 11 11.82 11 13.12" 10.86 11 10.04 11 9.62" 9. 70"
rv -rain gauge was :full.
,.., -adding average o:r two nearest rain gauges to make up missing data.
1\J
I
()'\
TABLE 2. Rainfall Chemistry
pH
7/2 3.60
7/2 4.95
7/3 4.02
7/3 4.50
7/13 4.43
7/13 5.03
7/14 4.03
7/15 3.70
7/29 4.5
8/2 3.78
8/2 4.76
8/14 5.02
8/14 4. 72
8/15 4.88
8/15 4.80
8/16 6 • 1 1
8/16 4.87
8/16 5.70
8/24 5.45
8/24 6.02
8/25 5.85
8/26 5. 1 4
8/26 5. 19
Conductivity
(pmhos)
14
9
20
1 6
8.5
5.0
3
Alkalinity
(mg/1)
o.o
Hardness
(mg/1)
4.0
2.0
Comments
1st storm; complete overcast
2nd storm
1st thunder storm
2nd storm, light rain
2:00 pm storm; overcast
4:00 pm storm
frontal system, drizzle
frontal system, showers
short thunder showers
heavy, brief
2nd storm, drizzle
9:00 am, night before
8:00 pm day-long light rain
all night rain
5:00 pm
intermittent showers
from 2 directions
snowing on summits
frontal storms
frontal storms
frontal systems
showers from west
indicating a washing-out of acidic constituents in the
atmosphere.
Hydrology
Streamflow measurements were taken and rating curves developed
for 5 streams in the study area: Birch Creek, Twelvemile
Creek, Faith Creek, McManus Creek and Ptarmigan Creek.
Watershed areas, discharge per unit areas and high and low
flows for the summer of 1983 are recorded in Table 3. The
highest flows for all streams were recorded early, and were
the result of intense thunder storms. The lowest flows for
all streams occurred in the beginning of August. The end of
August showed high flows again, similar to the early highs,
but were the result of an extended period of rain. Flows for
the summer of 1982 were high in spring and low in August.
High rainfalls and subsequent high flows were not as dramatic
in 1982 as in 1983. Birch Creek, in both summers, exhibited
rapid changes in discharge because of draining of settling
ponds. Stage changes of several tenths of a foot were
observed on Birch Creek over a half-hour period.
The streambeds of the mined streams were heavily embedded and
compacted with silt, Birch more so than Faith. Ptarmigan
became noticeably embedded and compacted within weeks of the
start of mining on that stream. The impact of the
embeddedness on groundwater was investigated by driving in
well points in each stream bed except Ptarmigan. We found
that over the entire summer, 1983, the peizometric surface of
the groundwater at the unmined streams was essentially at the
12-17
f\J
I
OJ
TABLE 3. Hydrology
Stream
Birch
Twelve-
mile
Ptar-
migan
Faith
McManus
High Flow Date
(cfs)
490.30 6/27
338.80 6/27
89.08 7/18
332.74 7/17
465.51 7/17
Flow per Area Low Flow Date
(cfs/sq. mi.) (cfs)
5.50 45.40 8/11
7. 18 20. 13 8/12
4.57 11.36 8/10
5.51 39.83 8 I 11
5.98 38.71 8/13
Flow per Area Watershed Area
(cfs/sq. mi.) (sq. mi.)
0. 51 89.1
0.43 47.2
0.58 19. 5
0.66 60.3
0.50 77.8
same level as the stream. The peizometric surface below the
mined streams was lower compared with the unmined streams.
The increased siltation and subsequent embeddedness in the
mined streams acts to partially seal the streambed and thereby
decrease recharge to the groundwaters. This hypothesis is
supported by the lack of any artesian conditions in the wells
at the mined streams and the different chemistry of their
water as compared to their stream
unmined streams was chemically
water. The groundwater of
similar to their surface
waters. In August, three well points were placed at the Birch
Creek site and at the Twelvemile site to investigate the water
level differences at different locations in the streams and to
profile the peizometric surfa~es downstream and across (Figure
3 and 4). It is clear that the siltation has acted to depress
groundwater levels beneath the mined streams and surrounding
them, and reduced the hydraulic contact between the
groundwater and surface water.
Water Chemistry
Two streams, Ptarmigan and Faith, showed low levels of
alkalinity, (20 to 30 mg/1), total hardness (30 to 40 mg/1),
calcium hardness (20 to 30 mg/1), and conductivity (50-80
Jlmhos). It was felt that these values reflected less
groundwater contribution to the flows, Ptarmigan due to its
high elevation and Faith for this reason as well as a
partially sealed streambed. Birch, Twelvemile and McManus all
showed higher levels of these constituents. Alkalinity ranged
from 40 to 60 mg/1, total hardness 50 to 70 mg/1, calciuim
12-19
~
ru
I
ru
0
A WPI
0.33'....,_ 5 .33• (PEIZOMETRIC SURFACE
Sl ~~ WP2 S3
5.00' ~ 5.65' 5.86' c 002 ,/I='==S2=:===~--, ~ WP3
0.01
STREAM
SURFACE
8 WPI
. 5.70' 5.87'
WP =Well Point Levels
S =Stream Levels
A
PEIZOMETRIC SURFACE
Figure 3. Well Positions (left) and Downstream (right upper) and Across-stream (right lower) Profiles of Groundwater
and Surface Water Levels, Twelvemile Creek, September, 1983.
-1\J
I
1\J
A o~56'-Sl
B WPI
(;j
~
0
()) =t ~
0::
---Q)
WP~ , 0 -61 -~
S2 32'
c &wP3
f'--S3
A I Sl 4.95 • [STREAM SURFACE
1.9~6:-· --...J~ ------JI ~s2:_5.96' c
WP 1 6.01'1 I 08' j ?3 6.23'
/
·, 1.12
WP2 7.04'1 l WP3 7.35'
PEIZOMETRIC SURFACE
A
WP= Well Point Levels
S =Stream Levels
-----IWP-i ~:gTRIC ::2 _B_
SURFACE
Figure 4. Well Positions (left) and Downstream (right upper) and Across Stream (right lower) Profiles of Groundwater
and Surface Water Levels, Birch Creek, September, 1983.
hardness 35 to 50 mg/1 and conductivity 110 to 180 umhos.
Recent rains caused concentrations to be lower in Birch than
Twelvemile, but the reverse was true during dry periods. All
the streams showed high silicia (6 to 10 mg/1) and low ammonia
and nitrate (less than 1 mg/1 as nitrogen), low copper,
manganese and nitrite (as nitrogen)(all less than 0.5 mg/1).
The pH ranged from 6.00 to 7.50 over the summer for all
streams and the groundwater. Temperatures in the streams
ranged from 2 to 3 C in May and early June to highs of 10 to
12 C in August. The temperatures dropped off quickly at the
end of August into September.
Iron was found to be much higher in the mined streams than in
the unmined streams. Iron ranged from less than 0.10 mg/1 for
the unmined streams to 1.5 mg/1 for the mined streams. When
Ptarmigan began to be mined on August 15th, iron went from
less than 0.1 mg/1 to 0.35 mg/1.
Settleable solids were higher in Birch than the other streams
due to the mining and values ran between trace amounts and 2.0
ml/1. Faith showed values that were, on occasion, as high as
0.2 ml/1, but were substantially lower than those at Birch.
Twelvemile and McManus showed no detectable settleable solids.
Ptarmigan showed no settleables until after mining, when they
rose as high as 3.5 ml/1. Turbidities were higher in the
mined streams, as would have been expected. Suspended
sediments in Birch Creek were an order of magnitude higher
than the other streams in 1982. However, during 1983, the
suspended sediment loads in Birch dropped considerably to
12-22
average several times that of Twelvemile. This change may
have been brought about by a change in mining practices or the
wetter weather. Total residues in Ptarmigan Creek increased
an order of magnitude with mining (from less than 100 to the
1000's mg/1). The total residues in Faith were never as high
as the other mined streams and often totals were no higher
than in McManus (less than 100 mg/1). Total residues,
however, include all dissolved solids as well as suspended
sediment.
Groundwater temperatures were lower than stream temperatures
over the summer until September, when the situation reversed.
The groundwater chemistry was similar to the surface water for
McManus and Twelvemile, the unmined streams, and was
significantly different than the surface water at the mined
streams. Comparisons of chemistry between the different
groundwater sites reveals similar levels of all constituents
in Twelvemile, McManus and Faith with the exception of iron,
which was higher in Faith. Birch Creek groundwater was higher
in conductivity, alkalinity, total hardness, calcium hardness
and iron than any of the other groundwaters. The
concentrations of iron in the groundwaters of Birch and Faith
were both lower than their respective surface waters, however.
The dissolved oxygen conditions of the groundwaters under the
streams also indicated that the groundwaters of the mined
streams are isolated from the surface waters. Table 4
displays data obtained from wells under pools of five streams
in 1983. Unmined streams are seen to be near saturation and
12-23
N
I
N
.j::-
TABLE 4. Temporal Distribition of Dissolved Oxygen (mg/1 and % saturation)
in Groundwater under Pools of Selected Streams
Stream Type
Date Clearwater Mined Glacial
8/12/83 Twelvemile Creek Eagle Fork-Birch Creek
5.90+0.30(n=3) 0.25+0.05 (n=3)
54% 2%
8/13/83 McManus Creek Faith Creek
3.76+0.13 (n=3) 2.90+0.27 (n=3)
33% 25%
8/31/83 Phelan Creek
9.31+0.29 (n=5)
82%
9/7/83 McManus Creek Faith Creek
10.95+0.07 (n=4) 2.99+0.83 (n=5)
86% 25%
9/14/83 Twelvemile Creek Eagle Fork-Birch Creek
7.75+0.99 (n=3) 0.41+0.04 (n=5)
63% 3%
mined streams depleted. Comparing the August and September
percent saturation for the two unmined streams, there appears
little change over time for Twelvemile Creek, but considerable
increase in McManus Creek. We believe that this increase in
dissolved oxygen is associated with high flows. Surface flow
is known to correlate with dissolved oxygen conditions in the
groundwater, probably because of increased recharge (McNeil
1962). One glacial stream was also successfully sampled, and
it is interesting to note that the dissolved oxygen is near
saturation. We will be examining the size distribution of the
substrate to determine if glacial flour has different
characteristics than the clays that are sealing Birch and
Faith Creeks.
Table 5 presents D.O. data obtained from wells placed one each
in a pool, riffle and on the bank of each Birch Creek and
Twelvemile Creek. This table shows that the low D.O.
conditions in the mined streams are not localized. It is
interesting to note that even under the banks, which in
ordinary streams provide recharge head, Birch Creek
groundwaters have low D.O. content as well as low elevation.
These low dissolved oxygen conditions have important
ecological implications. Salmon eggs that are buried in nests
called redds could be suffocated if the streambed becomes
sealed after they are laid,
macroinvertebrates
Algae
as could benthic
In both systems we studied, undisturbed streams displayed
12-25
rv
I rv
CJ'\
TABLE 5. Spatial Distribution of Dissolved Oxygen (mg/1 and % saturatio~, of Ground-
water under a Steam-pair (unmined and mined) in early September, 1983.
Location
Pool
Riffle
Bank
Twelvemile Creek (unmined)
7,75+0,99 (n=3)
63%
10.13+1.47 (n=3)
82%
4 ,43+0 .11 (n=3)
36%
Eagle Fork of Birch Creek (mined)
0.41+0.04 (n=5)
3%
1.15+0.26 (n=5)
9%
1.99+0.01 (n=3)
15%
higher productivity than mined streams. Birch Creek, which
supports intensive mining throughout the summer, averaged only
0.1 g-0 2 /m 2 /d with a summer peak of 0.49, while unmined
Twelvemile averaged 1.09 with a peak in July of 2.57.
Ptarmigan Creek, which was unmined throughout the 1982 field
season and most of the 1983 season, displayed an average daily
gross production rate of 0.65. For three days in August 1983,
productivity was measured simultaneously both above and below
a mining operation. During this period, which was mostly
cloudy or raining, productivity above the mine averaged 0.38,
while below the mining it averaged 0.13.
Faith Creek, which supports only four operations, faired
slightly better than Birch with a mean daily gross production
rate of 0.39 for the two field seasons, but the unmined stream
receiving it (McManus Creek) displayed a seasonal average of
0. 87.
Daily gross production rates for both mined and unmined
streams apparently increased with the amount of available
light as defined by incident PAR levels, mean depth, and
turbidity. This relationship has yet to be defined pending
further analysis. However, the relationship between turbidity
and its effect on the ability of water to transmit PAR to the
substrate has been defined on a regionl basis as: n(t) =
0.00022(T) + 0.011 (r2 = 0.98) where n(t) is the total
extinction coefficient and Tis turbidity (NTU's).
Total phosphorus concentrations averaged 598 ug/1 in Birch
12-27
Creek., 34.4 in Twelvemile, and 36.8 in Ptarmigan before
mining. Faith Creek averaged 225 and McManus 23.4 pg/1. The
considerably higher levels found in mined streams probably
results from the organic matter they carry (from removal of
over burden) and/or the adhesion of P-containing molecules to
the sediment particles. This is likely since the total P
concentrations recorded tended to increase with suspended
sediment concentrations. Another contributing factor could be
the scouring effect of suspended sediment which would tend to
increase the amount of algae found in the water column.
Macroinvertebrates
The differences in invertebrate densities between streams is
summarized in Table 6. Mann-Whitney tests for significant
differences in medians between streams showed that each stream
was significantly different from every other stream. The
unmined streams showed significantly higher invertebrate
densities than did mined streams.
From the data analyzed so far it appeared that mining in Birch
and Faith Creeks significantly lowered the density of benthic
invertebrates with some alteration of community structure. As
the degree of mining impact increased, certain taxa, such as
stoneflies, mayflies, caddisflies, and blackflies, made up a
decreasing proportion of the invertebrates found, while other
taxa, such as chironomid midges occupied a larger percentage
of the invertebrate community. With a high degree of mining
activity, like that found at Birch Creek, most taxa became
very rare or disappeared completely. It has been hypothesized
12-28
that sediment input like that associated with placer mining
eliminates those invertebrates that make a living by filter-
feeding and has a relatively less effect on collector-gatherer
organisms. This hypothesis will be tested using the 1983
data.
It seemed apparent from examination of the relatively scant
data of 1982 that those factors that did the most to determine
invertebrate densities in mined streams were settleable solids
and substrate embeddedness. Analysis of the 1983 data will
attempt to relate these factors and others to invertebrate
density, biomass, and community structure.
Table 6. Invertebrate densities in sampled creeks.
Mean Std. Median
no,/m2 Dev. no./m 2
Faith (mined) 206 140 210
McManus (unmined) 460 296 365
Birch (mined) B 1 1 0
Twelvemile (unmined) 693 285 680
Fish
Periodic beach seining of mined and unmined streams resulted
in many fish being caught in unmined streams, but none in
mined ones. The only exception to this was during the spring
spawning migration of adults and fall out-migration of
juvenile and adult grayling. Apparently, adult and juvenile
12-29
grayling avoided the mined streams in preference for the
clearwater tributaries. Mined streams supported neither
reproductive nor feeding areas.
From the caged YOY grayling experiments we concluded that the
sediment levels of Birch Creek were not of sufficient
concentration to cause any mortalities (up to ten days
exposure periods). Although there were no deaths caused by
sedimentation, the external appearance of the fish caged in
Birch Creek was noticeably different from normal healthy
individuals. Spots and parr marks were nearly absent in the
Birch Creek fish as well as their having a more pale brown
coloration on the dorsal surface. Examination of internal
organs showed that the fat bodies surrounding viscera in the
Birch Creek caged fish were nearly absent which is a likely
sign of starvation. Stomach analysis supported this idea in
that the Birch Creek fish were not capable of locating many
food items. Fish caged in Twelvemile Creek were able to
locate aquatic insects and looked normal in appearance when
compared to free-living grayling in Twelvemile Creek.
Gross microscopic examination of gill tissue of Birch Creek
YOY grayling showed mucus secretions with embedded sediment
particles in as short a time period as 12 hours when suspended
solids were> BOO mg/1.
Short-term exposure of mining sediments on Age II grayling did
not cause a consistent acute stress response (Table 7). The
mean leucocrit and hematocrit values of tested individuals did
12-30
not differ significantly (p:0.5) from control groups. It
appears that Arctic grayling in interior Alaska are more
tolerant to short-term exposure to sediment than other
salmonids such as rainbow trout (Herbert and Merkens, 1961).
A recent study completed by McLeay ( 1983) on the effects of
sediment on grayling in the Yukon showed a decrease in mean
leucocrit value and/or more variable glucose levels, but no
difference in hematocrit values. It would be difficult to
determine why leucocrit values were significantly lower in
test fish in the Yukon study, but not in ours. Several
possibilities exist: there may be a difference in the
tolerance to stress of Yukon and interior Alaska grayling;
this study conducted on-site field testing while the Yukon
study was mostly laboratory research; or, there is of course,
the chance of differences in experimental techniques. It
seems evident that further investigations are in order to
obtain more baseline data.
Although individual Arctic grayling were tolerant to short
term exposure to sediment, they were likely affected at the
population level. Heavily mined streams degraded the spawning
and rearing habitat of grayling by the filling-in of
interstitial gravel spaces 1 forming a cement-like substrate.
Also, highly turbid waters made slight-feeding impossible for
grayling. As a result of these disturbances, grayling were
forced into clearwater tributaries for the majority of the
summer.
12-31
1\J
I
\.N
[\)
TABLE 7. Acute Stress Bioassay. Leucocrit and hematocrit values of grayling held in
Birch Creek and Twelvemile Creek (control)
Exposure Suspended Settleable Turbidity Leucocrit
Period Solids {mg/1) Solids (mg/1) (NTU) value (%)
Stream (hr) X S.D. X S.D. X S.D.
N • Fork 6 124 N.D 0.4
Birch 624 0.5 2250.
N. Fork 24 11 0 1 4. 1 N.D 0.6 0.23
Birch 1388 212.1 0.8 0.53 1855. 490.6
N. Fork 24 152 5.7 N.D 1.1 0.31
Birch 462 93-3 0.4 0.55 813. 927.0
N. Fork 36 11 3 6 7. 1 N.D 1.1 0.29
Birch 527 130.0 0.4 0.43 721. 7332.0
N. Fork 48 1 1 4 9.5 N.D 0.6 0. 1 8
Birch 1205 219.5 0.8 0.54 21 2 1 . 1058.2
N. Fork 96 11 5 8.7 N.D 0.6 0. 1 6
Birch 1158 216.7 0.4 0.55 1931. 1024.0
* Leucocrit value significantly (p=0.5) lower than control
** Hematocrit value significantly {p:0.5) lower than control
N.D Not Detectable
X
1.16
1. 06
1. 12
1. 12
1. 00
1. 18
1. 37
1. 20
1. 23
*0.96
1. 13
1. 18
Hematocrit
value (%)
S.D. X
0.38 42.7
0. 1 9 41.9
0.24 3 8. 1
0.34 43.5
0. 1 4 37.8
0.31 39.9
0.25 39.7
0. 1 6 40.8
0.20 38.2
0. 1 7 40.9
0. 3 1 36.9
0.45 **41.8
S.D.
2.7
3.7
4.4
2.7
2.0
3.8
3.3
3.5
4.0
2.3
3:6
3.2
The ultimate effect, then, of mining on streams in a watershed
is to render some of them--the mined ones--partially or wholly
incapable of supporting fish and their food base. In other
words, the biological carrying capacity for aquatic life in
the watershed is reduced in direct proportion to the number
and length of streams that receive sediment from placer mines.
References
American Public Health Association. 1980. Standard Methods
for the Examination of Water and Wastewater. 15th Edition.
1134pp.
Eisenreich, S.J., R.T. Bannerman, and D.E. Armstrong. 1975.
A simplified phosphorous analysis technique. Environmental
Letters 9(1): 43-53.
Elmore, H.L., and W.F. West. 1961. Effect of water
temperature on stream reaeration. Journal of the Sanitary
Engineering Division ASCE, 87 (No. SA6):59-71.
Hall, C.A., and R. Moll. 1975. Methods of assessing aquatic
primary productivity. In Primary Productivity of the
B i o s ph e r e • H • L i e t h , an d-R • H • W h i t t a k e r ( e d s • ) S p r in g e r-
Verlag. New York.
Herbert, D.W.M., and J.C. Merkens. 1961. The effect of
suspended mineral solids on the survival of trout. Int. J,
Air. Wat. Poll. 5(1):46-55.
Jirka, A.M., and M.J. Carter. 1975. Anal. Chern. 47(8):1397.
McLeay, D.J. 1983. Effect on Arctic grayling of short-term
exposure to Yukon placer mining sediments: laboratory and
field studies. Can. Tech. Rpt. Fish. Aq. Sci. No. 1171.
McNeil, W.J. 1962. Variations in the dissolved oxygen
content of intragravel water in four spawning streams of
Southeastern Alaska. United States Fish and Wildlife Service.
Special Scientific Report--Fisheries No. 402. Washington,
D.C. 15pp.
Mortimer, C.H. 1981. The oxygen content
fresh waters over ranges of temperature
pressure of limnological interest. Mitt.
Limnol. 22:23pp.
of air-saturated
and atmospheric
Internal. Verein.
O'Connor, D.J., and W. Dobbins.
reaeration in natural streams.
1956. The mechanisms of
J. Sanit. Eng. Div., Proc.
12-33
ASCE, 82, SA6.
Odum, H.T. 1956. Primary production in flowing waters.
Limnol. and Oceanogr. 1:102-117.
Wedemeyeer, G.A., and D.J. McLeay. 1981. Methods for
determining the tolerance of fishes to environmental
stressors. pp. 247-275. !!! Stress and Fish. A.D. Pickering,
ed. Academic Press. London.
Wilcock, R.J. 1982. Simple predictive equations for
calculating stream reaeration rate coefficients. New Zealand
Journal of Science 25:53-56.
12-34
NON-SOLAR INFLUENCES ON TEMPERATURES OF SOUTH COASTAL ALASKAN STREAMS
Abstract
Solar radiation plays a key role in the temperature regime of south coastal
Alaskan streams, but is less dominating than in sunnier climes. Other factors
wind, precipitation, lakes, glaciers,falls and rapids, soils and aquifer rout-
ings -also have important effects, and are the subject of this work.
Conduction and evaporation-condensation processes of heat exchange both pro-
ceed at rates linearly responsive to wind speed. These processes may be more
dramatic in cooling exposed waters in fall-winter than in warming during sum-
mer. Wind produces important heat losses and mixing action in lake waters.
Freeze-up is likely to occur in clear, quiet weather following wind chill
conditions. Rainfall seldom provides heat to surface waters, but snowfall can
be a significant cooling agent.In addition to the role of lakes as solar heat
sinks, mechanisms of wind mixing, wind chilling, and protective ice cover pro-
duce temperature effects in the outlet flows of lakes. Glacial streams, though
cold in spring to fall months, often produce sustained, relatively warm spring-
fed flows in winter. Falls and rapids act like wind in producing changes as
large as 4° C. through extended falls. Groundwater aquifer routes are import-
ant temperature moderators, cooling in summer and warming in winter. Deeper
soils probably also moderate temperature, while muskeg dominated watersheds
favor warm stream temperatures.
The sequence or combination of these influencing factors produce the temper-
ature character of a stream, which may be variable from head to mouth or may
be dominated by only few factors according to the conditions of local climate
and physiography.
Introduction
The most prominent variable influence on stream temperatures in mid-latitude
drainages is generally recognized as sunshine. Presence or absence of shade
along a stream is often a critical habitat feature in plans for logging or
other activities which may affect streamside margins. While solar radiation
remains an important influence on south-coastal Alaskan streams, other factors
can also have major, even dominating, effects on stream temperatures and are
associated with varied specific climatic and physiographic conditions within
watersheds. These factors --wind, precipitation, lakes, glaciers, falls and
rapids, soils and aquifer conditions --are the subject of this discussion.
13-1
The process or mechanism by which factors affect temperature is described,
and the magnitude of effects are illustrated when information is available.
Examples of the effect of specific factors upon fishery habitats are described.
Finally, a discussion is provided of how the several temperature influences
of a drainage may be integrated into patterns of variable complexity.
Wind alters temperatures of streams and tributary lakes and ponds by influ-
encing the rates of transfer of sensible heat by conduction, or of evapora-
tion-condensation. The conduction or transfer of sensible heat is linearly
dependent upon (a) wind speed, and, (b) difference between air and water
temperatures. In a study of the heat and water balance of Lynn Canal, south-
east Alaska, oceanographer D.R. Mclain (1969) notes that transfer of sensible
heat "was greatest during the winter (November to March) at Eldred Rock, due
again to the strong cold winds blowing over relatively warmer water." The
rate of evaporation or condensation is linearly dependent upon (a) wind
speed, and upon (b) difference between saturation vapor pressure of water at
surface water temperature and the vapor pressure at dewpoint temperature con-
dition.Mclain's analysis(l969) indicates that evaporation is a significant
factor in heat loss from Lynn Canal during winter. This, he notes, is partic-
ularly true in upper Lynn Canal, where strong winds are common. His work
further suggests that heat gain in a southeast Alaskan water body due to con-
densation is only likely in summer, and then only when the water body is quite
cold compared to the air temperature. He found little or no heat gain by con-
densation in his investigation of Lynn Canal.
Measurements of water temperatures at two locations in Black Bear Creek near
Klawock, Alaska on Prince of Wales Island, were examined for the possible role
of wind on lake outlet water temperature. Figure 1 shows average water temp-
13-2
eratures at the mouth of Black Bear Lake, elevation 1685 ft., and at the
base of the falls and rapids at elevation about 200ft., about a half mile
below the lake mouth.
"" "' "' "3"c-u
...
"' "' "' "" u
"' ..-<
,outlet, Black Bear Lake, elevation 1685 feet
"'2°C
q half mile below Black Bear Lake and
falls, rapids; elevation ca. 200 feet
. /
"g
• . • •
17
. . ·, :
o ..... .(). ....... (). ... /
18 19
~-··0·--.. ·-0 ..... 0----... o ..... Ci
20 21 22
November 1982
23 24
I
25
. .
' .
.'
26
Figure 1: Comparison of Black Bear Lake outlet temperatures with temper-
atures a half-mile below lak~ outlet and falls-rapids, used to
illustrate fall wind-chill, lake mixing and freeze-up behavior.
This figure is believed to show: (a) initial brief period of conductive
cooling;(b)mixing of lake surface and sub-surface waters; and,(c)probable
freeze-up of lake surface as winds and mixing cease under clear weather.
The role of exposed versus ice-free lake conditions is further indicated
in Figure 2. The effect of wind on lake surface temperatures is probably an
13-3
tor when 1 ted fac t re-re a tempera u
important over.
freezes the lake ing when pear
1.• ce is not lake
BLACK BEAR 1685 ft.) LAKE (elev.
BLACK LAKE
disap-present,
. e·absent 1980~81. :tC
. intermittent or
1980-81 ice-free t
Figure 2: 80-81 and inters 19 Data rature, w utlets. f daily tempe Black Lake o xtremes o Lake and Monthly e Black Bear 82 for 1982 1981- ' 1 . et.a ., from Bishop'
13-4
As shown in Figure 2, Black Bear Lake was never fully ice-covered in winter,
1980-81, and Black Lake (at elevation 80 ft. about a mile and a half downstream)
remained ice-free. In the winter of 1981-82 both lakes were ice covered.
Strong, local mountain winds are common in the vicinity of these lakes,
with evident potential for mixing . The access of wind to the water surfaces
is believed to have been an important factor in producing the differences
in water temperature ranges seen in Figure 2.
Wind can also directly cool water within a stream channel. Bishop (1974)
made the following observation of wind cooling of a stream tributary to
Becharof Lake, Alaska Peninsula. "A graphic example of cooling was seen
on (stream) E-130.0,At the mouth on Sept. 30 this stream measured 6.0°C.
with air temperature of 7.4°C, while 20 minutes later and 1/3 mile upstream
0 above the falls (about 20-30 ft. high) the stream was 6.4 C. The strong
wind (10-20 mph at the surface) provided a powerful cooling agent as it
struck the turbulent water at the falls."
In summary, the effects of wind on water temperature are to:
a. increase rates of sensible heat transfer or evapor~tion-condensation
in proportion to wind velocity, and to,
b. induce mixing in water bodies.
Wind on shallow lakes or exposed streams can produce rapid temperature
change. Even on deeper lakes it appears that surface water temperature
change, particularly from winter chilling, can occur rapidly in addition
to mixing action.
13-5
Precipitation effects: In Mclain's (1969) heat balance study of Lynn Canal,
he concluded that rainfall is likely to have a relatively small impact upon
water temperatures. He states, m The transfer·· of heat by precipitation,Q , p
between the atmosphere and the water of Lynn Canal was primarily by the
mechanism of melting of the winter snowfall. Thus the heat transfer was
predominantly a negative heat exchange --a heat loss by the water. The
only times when the heat exchange might be positive would be when warm rain-
water fell on cooler water. This would only occur during summer months
when snowmelt is zero. If positive heat exchange by precipitation ever
occurred, it would be of small magnitude since during summer the air temp-
erature (and hence the rainwater temperature) is almost equal to the sea
surface temperature."
The validity of Mclain's conclusion, as it might be extended in applica-
tion to fresh waters, was briefly examined using observations made in the
Black Lake basin of Black Bear Creek, near Klawock. Figure 3 compares mean
daily temperatures taken at the mouth of Black Lake with rainfall measure-
ments made at the lake.These plots, for spring and for late summer suggest
that rainy periods are associated with dropping water temperatures through
these seasons. Exceptions to this behavior seem to 6ccur in the late March
and mid-September plot relations.
There are occasions when precipitation plays an important water temper-
ature role. In August of 1971 the author observed sockeye salmon holding
at the saltwater mouth of Shrode Creek in Prince William Sound. Biologists
attributed this delay of salmon migration to the exceptionally low temper-
atures in Shrode Creek -about 4°C. The low stream temperatures during
that August were caused by the very heavy winter smowpack and the late
13-6
Lower Black Lake Thermograph Record
9°C ~~5------~--~LW~LLO-~L-~4UW4~UU~~··
ugust
r
Match
August
1981
Lower Black Lake Thermograph Record
30
April
1983
Rainfall
2 11
1 "
Rainfall
2 "
1 "
May
Figure 3: Simultaneous plottings of Black Lake outlet temperatures and of rainfall
measurements taken at Black Lake.
13-7
spring-summer melt, even at tidewater. Such depression of stream temper-
ature is not uncommon in areas which may receive heavy snowfall. Noeren-
berg (1957) also reported depressed temperatures in Prince William Sound
during the summer of 1956, after a winter with heavy snow followed by a
slow melt season.
Rain on snow accelerates melt both because of the rain's heat content and
because the action of rain on snow reduces the snow-surface albedo and
thereby increases absorption of solar radiation. These effects have been
observed by the author to produce increased runoff with low water temper-
atures in streams having sizeable sub-alpine areas.
Lakes are widely recognized to profoundly influence temperatures of the
associated stream system. Lakes provide the most important of physiographic
influences upon water temperatures in many drainages and exert control
not only on means and extremes of water temperature, but also upon the
timing and the duration of temperature conditions.
The climatic elements of air temperature, solar radiation, back radiation,
wind, precipitation,act individually or together on lake surface waters
to affect temperatures. The physical characteristics of a lake including
water quality, area, shape, depth, elevation, orientation with regard to
surrounding terrain, and exposure to wind, control the degree of tempera-
ture response to climatic elements.
Bishop (1974) performed a small experiment on the rate of solar heating of
four kinds of water held in respective containers on a sunny day. The re-
spective waters were Chulitna River(glacial sediment), Talkeetna River
13-8
(glacial sediment), Auke Lake(stained with iron-organics), and tapwater
(clear).An electronic thermometer with a highly sensitive probe was used
to measure the changes in profiles of temperatures for the respective
containers of water. The following conclusions were reached:
a. In sunny conditions, with air temperatures greater than water temper-
atures, sedimented waters warmed fastest. Iron-organic stained water
warmed faster than tap water.
b. Under conditions when air temperatures were less than water temperatures,
heat loss of sedimented water evidently exceeded that of stained or
tap water.
c. the more heavily sedimented Chulitna River water developed a stronger
temperature stratification than Talkeetna River water. Both sedimented
waters stratified more than the clearer waters.
These water quality characteristics undoubtedly affect lake heating.
Further information describing this effect is not available.
The most important lake characteristic is area, since each of the climatic
elements act in proportion to the area, though the respective roles of direct
solar radiation or effective back radiation will be diminished according
to the lake's shading or to lakeside obstructions which influence the in-
tensity of back radiation heat loss.
Shape and depth of lake act together to influence rates of shoreline versus
offshore heating, development of lake temperature stratification, and impacts
of winds.
Lake elevation affects potential solar radiation with significant increases
developing above 3,000 ft. elevation. Mean air temperature drops with in-
13-9
creasing elevation, while wind speed effects on temperature will be larger
as elevation increases.
The compass orientation of a lake and its relation to surrounding terrain
may influence the solar radiation it receives. In some cases a lake's orient-
tation or position will strongly influence its exposure to wind, as for
example, mainland lakes which may be exposed to strong gradient winds moving
off mountains. Island lakes are sometimes located in positions exposed to
williwaw winds, while other lakes are not vulnerable to these frontal dis-
turbances.
Wind action also has strong effects upon the thermal regime of a lake and
its outlet stream's temperature, as discussed earlier. Differences in ex-
tremes of lake outlet water temperatures under ice-covered versus open
winter contions were shown earlier in Figure 2.
The timing of weather events acting on a lake can have strong influence on
its thermal characteristics. In the fall and winter of 1981-82,Black Bear
Lake near Klawock froze over about December 25. In the fall and winter of
1982-83, the lake froze over about November 18-25. lf This difference of
a month in the time of freezing may be the cause of the range of values shown
in lake temperature profiles, January 1982 versus January 1983 (see Figure 4).
The reduced amount of heat stored in Black Bear lake in January, 1982 as
compared with heat remaining in January, 1983 (Figure4) is believed to be
due to the cooling action of the wind on the lake surface throughout December,
1981, compared with the early freeze-up in November, 1982. The comparative
spring responses have not been evaluated, but there i~ no doubt, a basis
lltiates were derived from interpreting lake outlet temperature records.
13-10
-20 ft.
1-21-83
BLACK BEAR LAKE
1-20-83
1-20-82
BLACK LAKE
-40 ft. -60 ft.
Ice Condition
1982 surfac~
12"opaque ice 6" hard snow
-4°w;t;r----6~s;o;-ic;-
-------------"'" 4"clear ice 16"water@0.4 C.
-------4-6"-cle;r-i~e
Ice Condition
1982 surfac~
4-6"opaque ice ice-free
-1:·2~-;t;r---
-80 ft. -100 ft.
Figure 4:Temperature profiles for Black Bear Lake and Black Lake, near Klawock;
January 1982 and January 1983. Lake conditions are also shown.
13-11
for a significantly different temperature regime below Black Bear Lake
in spring 1983, as compared with the previous year. The most notable fea-
ture of this comparative set of observations is that differences in heat-
temperature regime appear to have resulted primarily from differences in
timing and pattern of freezing of the lake during the two years.
Glacial streams found along the mainland and in some instances on larger
islands with glacial headwaters, have special water temperature regimes.
In the vicinity of Lituya Bay, meltwaters immediately below glacial sources
were found to be in the range of 0°C to 4 or 5°C during May through October
(Bishop,l977). In larger glacial rivers with considerable distance of flow
below the glacial origin of flow, downstream surface water temperatures
0 may reach or exceed 10 C. Glacial lakes act to elevate temperatures, and
glacial sediment probably affects the nature of this process.
A somewhat unexpected feature of glacial streams is their winter flow
regime. L.R. May, USGS Glaciologist, notes that glaciers provide good
winter low flows (personal communication, 1981) These flows may derive
from relatively large groundwater aquifers common in glacial vallies. The
effect of sustained winter low flows on river water temperatures just down-
0 stream of the glacial origin , is to elevate temperatures above near-0 C
levels, in some cases even during the coldest periods of winter. Thus,
Bishop (1981,1982) found that glacially-fed West Creek, near Skagway, Alaska,
seldom fell below 1°C, while the receiving water of the nearby Taiya River,
0 flowing from more distant glacial and other sources,dropped to near 0 C.
The role of groundwater flows in glacial streams during fall to spring
months is a key factor in late runs of chum and coho salmon into many large
and small glacial rivers of Alaska. These runs of salmon have adapted timing
13-12
suitable for the temperature regime found ingmundwater upwelling areas,
and locate these areas by using temperature sensors on their skin.
Waterfalls and rapids favor extremes of temperature and associated flow
conditions. In summer, south-facing falls and rapids are particularly
favorable conditions for rapid warming. Even the north-facing falls below
Black Bear Lake (vertical fall about 1500 ft.) produced a temperature in-
crease on June 20, 1982, from 1°C above the falls to 5.5°C immediately
below the falls.(Bishop, et.al.,l982)
In fall and winter, falls and rapids may lose heat rapidly and when stream-
flows are already at or very near freezing may produce heavy volumes of
needle or frazil ice. Figure 5 provides a record of stream temperatures
of uppe~ Black Bear Creek, __ at the outlet of Black Bear Lake, at the base
of the falls -also the location of entry of much flow into groundwater,
and, downstream 1500 to 2000 ft., where much of the stream re-appears as
springflow. Comparison of the temperatures above and below the falls
demonstrates fall-winter temperature losses of up to 4 1 12°C. along the
course of the falls.
An important result of falls or rapids demonstrated in Figure 5 is the
fluctuating character of temperature immediately below the falls as corn-
pared with the entry temperatures found above the falls at the outlet of
the lake. The loss of thermal stability in a system due to falls or rapids
is apt to work against the quality of fishery habitats.
Soil and groundwater aquifer routes exert strong influence on water temp-
eratures. As a general rule, deeper mineral soils provide greater flow
13-13
Black Bear Lake ................................ =
At base of 1500 ft. falls below Black Bear Lake
(also location of flow entry into aquifer •····
At head of Spring Fork, where springs.upwell
(about 1500 -2000 ft. below base of falls) • • · • ·
Oo::t. 31 Jot~. 31 root~. :ra
·~·
Figure 5: Average daily temperatures for three sites below Black Bear Lake,
September 1982 through May 1983.
=
=
lbn:tl 31 ll.oJ ll
regulation and less surface heating of tributary water sources than soils
which are shallow to bedrock. Deeper soils are also highly correlated with
better forest growing conditions, indicating that heavily timbered watersheds
are likely to have well regulated water temperatures in their natural state.
In contrast, open muskeg soils provide direct access of solar radiation to
the soil-groundwater profile. Thus, although the albedos of wet muskegs and
northern conifer forests are probably similar, the energy actually received
at the soil level in a muskeg is likely to be higher, since a significant
part of a forest's incoming radiation is absorbed by the forest canopy and
only part of this direct radiation received by the forest canopy is re-
transmitted as long-wave radiation reaching the soil surface. In effect,
such open, wet muskegs may be strong heat sinks, once the winter's snows
are melted, producing dramatic reduction in heat ( radiation) reflected
up from the surface. Sheridan and Bloom (1975) reported maximum expected
temperatures for muskeg-forest tributaries at 55-60°F. and muskeg tributaries
0 at 65-70 F.
Groundwater entry into surface waters may dominate temperature conditions
prevailing at specific locations on a drainage. In some smaller systems,
and during periods of either summer or winter low flow, the groundwaater
contribution may control stream temperatures. The temperature of emergent
groundwater is conditioned by the depth of its flow path, the time spent in
groundwater flow, and by the mean annual temperature of the geographic
region.Examples of groundwater temperatures at various locations along the
Gulf of Alaska, shown in Figure 6, suggest the likely annual range of temp-
eratures of waters emerging after a significant period of groundwater flow.
13-15
l0°C
8oc-
Figure 6: Plotting of miscellaneous groundwater temperature measure-
ments made by D.M. Bishop at varied locations and times on
fresh waters along the Gulf of Alaska. A speculative curve
of likely groundwater temperatures is shown.
A=
B =
B = t
B t •
Aleutians, Umnak Island, Sheep Creek drainage, 1977.
spring near Becharof Lake, Alaska Peninsula, 1974.
upwelling near Becharof Lake, Alaska Peninsula, 1974;
water temperature may be geothermally influenced.
C = Clear Creek, near Chilkat Lake near Klukwan, 1981.
K = groundwater from Klehini River near Klukwan, 1981.
L =groundwater in dug pit, 1/4 mile from ocean near Lituya Bay, 1977.
S = groundwater headwaters of Switzer Creek near Juneau, 1971 -72.
T = six groundwater observation wells at confluence of Tsirku River
with Chilkat River, near Klukwan, 1981.
13-16
Bishop, et.al.(l982) measured inflow and outflow water temperatures assoc-
iated with a very fast flowing groundwater route, estimated by dye tracing
work to flow at a rate of 100 ft per hour through an aquifer about 1500 to
2000 ft. in length. In summer, when water flowing into the groundwater route
0 0 was 12 C.,the temperature of the outflow springs ranged from 9.5 to 12 C.
Thermographs installed in early fall, 1982 immediately above and below this
groundwater route yielded the results also shown in Figure 5. Fall-winter
observations of water temperature show increases of as much as 3lf2°c.
through the groundwater route. Volume of this flow was in the magnitude of
12 to 25 c.f.s.
In general, the entry of water into groundwater flow produces modulated
temperatures at the springs below. The direction of heat flow varies sea-
sonally. This process is likely to be strongest when largesttemperature
differences occur between inflowing water and aquifer materials. In the
fall-winter period larger differences between inflow and outflow tempera-
tures may be expected than in the winter-spring period, because the heat
reserve and temperature differential required to produce change in temper-
ature may be reduced or depleted. This reduction in magnitude of change is
suggested upon careful examination of Figure 5.
Groundwater flows are critical to the spawning habitat of chum and coho
salmon, as noted earlier in the discussion of glacial waters.Springflow
temperatures are in a useful range and are dependable; stable flows are
also highly beneficial. In Japan, only springfed streams are selected as.·
sites suitable for establishment of chum salmon hatcheries.
Important geologic and landform situations producing springflows include
glacial and alluvial gravel formations,limestone areas, uplifted beaches,
13-17
Landslide deposits, fault zones, and volcanic ash deposits. All of these
occur along south coastal Alaska.
In conclusion, the immediate function of this work is to assist the field
ecologist or hydrologist in identifying likely stream temperature regimes
within a drainage. An expansion of the work is in the recognition that
many streams contain several distinct, linked temperature regimes. The
nature of these respective regimes and the sequence of their occurrence
is vital to a stream's character. Some examples of linked regimes are:
a. a lake in the headwaters of a stream has more impact on stream temp-
erature downstream if the lake does not have a waterfall or rapids
at its mouth;
b. the fluctuating temperature effect of rapids or falls is rapidly
mitigated by a groundwater flow route downstream and conversely, the
temperature stabilizing quality of groundwater flow may be largely
eliminated by a waterfall downstream and,
c. lakes with little exposure to wind are likely to have different outlet
temperature regime than those exposed to winds.
An understanding of these relationships between streams and their associated
terrain and climatic conditions may help to enlarge our background of knowl-
edge and improve our ability to predict or to understand the hydrologic
and biologic performance of coastal Alaskan streams.
13-18
CITATIONS
Bishop, D.M. 1974. A Hydrologic Reconnaissance of the Susitna River
Below Devil's Canyon. A report for the National Marine Fisheries
Service, Juneau, Alaska. Environaid Juneau, Alaska.
1974. Investigation of Selected Streams of Recharof and
Ugashik Lakes. Environaid. October 1974 in The Bristol Bay
Rehabilitation and Enhancement Opportunity Program, A Progress
Report. Dec. 17, 1974. Alaska Dept. of Fish & Game, Div. of Comm.
Fish, Anchorage, Alaska.
1977. Hydrology of Coastal Streams Near Lituya Bay.
Environaid. Dec. 1, 1977, A report for U.S. Park Service, Juneau
Alaska.
1982. Letter to R.W. Beck & Associates, Seattle, summarizing
a year's water temperature records on West Creek & Taiya River riear
Skagway, Alaska. Environaid, Oct. 21, 1982.
Bishop, D.M., A. Milner & L. Smith 1982. Biological-Ecological
Investigations on the Black Bear Creek System·Near Klawock, Alaska.
A report to Harza Engineering Company for Alaska Power Authority,
Anchorage, Alaska. August, 1982.
Bishop, D.M., O.C. Wallmo, & T. Moore. Environmental Investigation of the
West Creek Hydroelectric Project. A report to R.W. Beck & Associates
for Alaska Power Authority, Anchprage, Alaska. December 15, 1981.
Me. Lain, D.R. 1969. Heat and Water Balance of Lynn Canal, Alaska.
Doctoral dissertation, University of Michigan, 1969. 143 pp.
Mayo, L.R., 1981. Personal Communication with D.M. Bishop re Bishop &
Havrilak's work (1981) cited earlier.
Noerenberg, W. 1957. In Fisheries Research Institute Archives, University
of Washington, Seattle, Washington.
Sheridan, W.L. & A.M. Bloom, 1975. Effects of Canopy Removal on Temper-
atures of Some Small Streams·in Southeast.Alaska. USDA Forest
Service Administration Report, R-10 Juneau, Alaska, June, 1975.
13-19
Abstract
A COMPARISON OF VELOCITY MEASUREMENTS BETWEEN CUP-TYPE
AND ELECTROMAGNETIC CURRENT METERS
By Paula M. Wellen* and Douglas L. Kane**
Several methods for determining the velocity in an open channel stream
are in use today, two of which are the cup-type current meter (such as
the Price AA, Gurley, or pygmy) and the electromagnetic flowmeter. The
cup-type current meter is the most widely accepted and is used as a
standard by the U.S. Geological Survey. Use of the electromagnetic
flowmeter by field workers is increasing due to its ease of use and the
direct readout of the velocity measurement.
Velocity measurements were taken with a Price AA current meter, a pygmy
current meter, and two Marsh McBirney Model 201 flowmeters. Velocity
profiles were made in numerous streams with a wide range of water
velocities; measurements were also taken in metal culverts.
Differences were detected in the velocity measurements from the
different current meters. A strong linear relationship was found
between the readings from the electromagnetic flowmeters and the
cup-type current meters at lower velocities. For water velocities
greater than 5.0 fps (152.4 em/sec), the electromagnetic meter
registered a lower velocity than the Price AA, but no quantifiable
relationship was found.
Introduction
Velocity determination in open channels is an important design parameter
used by a variety of engineers and scientists. Biologists, foresters,
agricultural engineers, hydrologists, environmental engineers and civil
engineers all gather and use stream velocity data. Because of the
diversity of people and possible end uses, the data must be accurate,
precise and independent of the method by which they were gathered.
Several methods for determining open channel stream velocities are
presently used: cup-type current meters; pitot tubes; floats;
* Graduate Research Assistant, Institute of Water Resources, University
of Alaska, Fairbanks, Alaska 99701
** Associate Professor, Institute of Water Resources, University
of Alaska, Fairbanks, Alaska 99701
14-1
propeller-type current meters; and electromagnetic flowmeters. The
cup-type current meter is the most widely accepted method and is used by
the U.S. Geological Survey as a standard. The mechanical simplicity,
ease of checking proper functioning and relatively low initial and
maintenance costs of this device contribute to its acceptance. The
electromagnetic flowmeter is becoming popular due to its ease of use in
the field and the direct readout of the velocity measurement. This
paper compares velocity measurements using these two types of
instruments in a range of field conditions.
Cup-type current meters are manufactured under several different names
and come in two different sizes (for example, Gurley and Price AA, pygmy
and mini-current meter). All of these instruments work on the same
principle. A set of six conical cups revolve about a vertical axis;
when placed in the current, the water impinges on the cups causing them
to rotate. Tail vanes or fins are used to keep the current meter facing
directly into the current. A small cam inside the current meter driven
by the rotating cups closes an electrical circuit with each revolution
(or each fifth revolution) producing a click in the headphones worn by
the operator. The clicks are counted and timed, and are directly
proportional to the water velocity (Brater and King, 1976).
The electromagnetic flowmeter is based on Faraday's law of induction
(Grzenda, 1981). This law states that a conductor moving through a
magnetic field (at right angles to that field) produces a voltage that
is proportional to the velocity of the conductor (see Figure 1). In
1832 Faraday performed an experiment to measure the discharge of the
river Thames. By placing two large metal electrodes in the river, he
14-2
Water Velocity
v
Magnetic
/8
Field
J-F
v
--F= q(Vx8)
V = Water Velocity -8 = Magnetic Field Intensity -F = Electromotive Force
Figure 1. A schematic drawing of the ~1arsh McBirney electromagentic current meter probe.
tried to measure the induced voltage produced by the water flowing
through the earth's magnetic field. Although Faraday's original
experiment failed, the principle behind his experiment set the stage for
the development of the electromagnetic flowmeter (Springer, 1980).
Water is the conductor when using the electromagnetic current meter to
measure stream velocities. A probe containing an electromagnet and two
electrodes is lowered into the stream. Tail vanes or fins are used to
keep the probe facing directly into the current, ensuring the conductor
and the magnetic field are at right angles. The water flowing through
the magnetic field produced by the electromagnet creates a voltage drop
between the two electrodes and the voltage drop is linearly proportional
to the water velocity. Direct readout of this velocity is available in
English (feet per second) or SI (centimeters per second) units.
Field Methods
Twenty stream sites were visited resulting in 145 comparisons of
velocity measurements using both electromagnetic and Price AA meters.
Velocity profile measurements were taken with a Price AA cup-type
current meter and two Marsh McBirney Model 201 electromagnetic
flowmeters. At streams with low water velocities, a pygmy current meter
was used. A Marsh McBirney (in English units) was used for most of the
electromagnetic flowmeter work. At selected field sites, the accuracy
of this unit was confirmed by comparing it with a second Marsh McBirney
(in SI units). In all cases, the electromagnetic probe and the rotating
cups were affixed to a top setting rod.
Velocity measurements were started 0.10 ft (3.0 em) or 0.15 ft (4.6 em)
fron1 the streambed. Near the stream bottom, readings were taken at 0.10
14-4
ft (3.0 em) increments. This spacing was maintained to the stream
surface in relatively shallow streams. For deeper streams the readings
were spaced progressively further apart as the depth of the reading
approached the surface of the stream. This variable spacing was used to
better define the velocity profile close to the streambed where the
greatest velocity change occurs, while economically reducing the number
of readings required in deeper streams.
Measurements were performed over a wide range of water velocities,
ranging from 0.40 to 12.09 fps (12.3 to 368.5 em/sec). The Marsh
McBirney was calibrated to read velocities from 0 to 10 fps (0 to 305
em/sec). The Price AA meter had rating limits of 0.25 to 8.0 fps (7.6
to 243.8 em/sec). The pygmy current meter was rated for velocities
ranging from 0.25 to 3.0 fps (7.6 to 91.4 em/sec). Profiles were taken
in natural stream channels, near bridges and in culverts to cover the
widest possible range of velocities.
Results
Ideally, the velocities measured by cup-type and electromagnetic current
meters should be the same. A plot of the Price AA velocities versus the
Marsh McBirney velocities indicates a high correlation between the
readings from the two instruments at lower velocities (see Figure 2).
At higher velocities the relationship (if any exists) is not readily
apparent.
Figure 2 clearly illustrates the linear relationship between
measurements from the two meters at lower velocities. A linear
regression of the Marsh McBirney velocity readings on the Price AA
velocity readings through the origin for velocities less than 5.0 fps
14-5
9 •
8
•
(f)
Q •
lL. 7 0 •
>-• •
1--
8 6 ' _.J . ~~ w • > o« •
0:: c,"
w 5 1--• w ::;: • • • • 0 •
1--• •
+=-2 4 • • w • • I 0:: •
CJ\ 0:: ::::> • • •
u ..
u 3 i= w
2
<.:)
~ 2 • 0
0
0:: , •
1--o• • u w •
_.J • • • w 0 • 0 •
00 2 3 4 5 6 7 8 9 10 II 12
PRICE AA VELOCITY (FPS)
Figure 2. Comparison of measured velocities from Price AA and electromagnetic current meters.
(152.4 em/sec) results in a straight line with a slope of 1.006. The
results of a standard t-test (t = 0.4736, a = 0.05, n = 121, degrees of
freedom = 120, P-value = 0.637) show there is no statistical difference
between this line and a line through the origin with a slope of 1. The
close agreement in readings between the two types of instruments is
exemplified in Figure 3: velocity profiles are illustrated for two
sites in this study. The water velocity (as measured by the Price AA
meter) never exceeds 5.0 fps (152.4 em/sec) at these two sites. The
Marsh McBirney velocity readings are within ten percent of the Price AA
velocity readings for all but four measurements at these two locations
and are within twenty percent for all but one measurement.
The linear relationship in Figure 2 does not hold true at higher
velocities. In fact, the readings from the two meters are negatively
correlated (r = -0.435, n = 24) for velocities greater than 5.0 fps
(152.4 em/sec). Sites 009 and 020 both have water velocities in excess
of 5.0 fps (152.4 em/sec). The velocity profiles for these two
locations are illustrated in Figure 4. Even with the compressed
velocity axis (as opposed to Figure 3), the difference in readings from
the two types of meters is apparent. Two Marsh McBirney instruments
were used at site 020, producing comparable readings to each other as
opposed to the Price AA meter.
Discussion
One plausible explanation for the erroneous velocity readings given by
the electromagnetic current meter lies in the design shape of the probe
(see Figure 1). At higher velocities, separation of water from the
14-7
-!O-
J
()J
1.2
1.0
0.8
f--
LL 0 6
:r:
f--
Q
w
0 0.4
0.2
6 •
0
41> 6
€>6 ..
• ..
..,
dill
f--
lL.
:r:
f--
Q
w
0
SITE 010
6 PRICE AA CURRENT METER
• ELECTROMAGNETIC CURRENT
METER (I)
0 PYGMY CURRENT METER
0~-----L------L-----~
6 0 2 4
VELOCITY (FPS)
1.2
1.0
08
06
0.4
0.2
o•
SITE Oil
6 PRICE AA CURRENT METER
• ELECTROMAGNETIC CURRENT
METER ( i)
0 PYGMY CURRENT METER
VELOCITY ( FPS)
Figure 3. Velocity profiles for two sites showing close agreement of current meter readings for water
velocities less than 5.0 fps (152.4 em/sec).
/.50
1.25
eo e:.
1.00 e 0 e:. 1.00
}:::: f-
lL lL
:I: 0.75 eo :I: 0.75 e 6
+-f-f-
I Q Q e 6
'.!) w w
0 0
e 0 6 e 6
0.50 eo 6 0.50
Lit
6e
eo 6 0.25 c. 6 SITE 020 0.25 6 e
6 PRICE AA CURRENT METER SITE 009
e ELECTROMAGNETIC CURRENT 6e 6 PRICE AA CURRENT METER
METER (I) e ELECTROMAGNETIC CURRENT
0 ELECTROMAGNETIC CURRENT METER (I)
0 METER 2) 0
0 4 8 12 0 4 8 12
VELOCITY ( FPS) VELOCITY (FPS)
-Figure 4. Velocity profiles for two sites showing variations in current meter readings for water velocities
greater than 5.0 fps (152.4 em/sec).
probe was noticed during several measurements. Separation was more
readily apparent at depths closer to the water surface. In some cases,
readings were not obtainable because of pronounced separation and wake.
Improved streamlining of the probe might eliminate the separation or
move it further back along the probe (behind the sensing electrodes),
thus decreasing its interference with the velocity reading. Roughening
the leading edge of the probe might also move the separation point
further back, possibly behind the sensing electrodes. The effect of
inducing a turbulent boundary layer on the sensing electrodes is
unknown.
Electromagnetic current meters are increasingly being used by field
workers. Liithout a backup velocity reading by another method, the field
worker does not know if the Marsh McBirney velocity reading is correct.
For example, the Marsh McBirney current meter might incorrectly read 4
fps (122 em/sec) at a given stream flowing 6 fps (183 em/sec). Yet a
stream actually flowing at 4 fps (122 em/sec) might result in a correct
Marsh l~cBirney reading of 4 fps (122 em/sec). Thus the Marsh McBirney
current meter should not be used indiscriminately without knowing the
expected range of velocities.
The cutoff velocity of 5.0 fps (152.4 em/sec) at which the Marsh
~lcBirney starts giving erroneous readings is rather arbitrary. This
value was selected by an iterative process, finding the maximum velocity
where the slope of the line in Figure 2 was closest to 1 with the
highest coefficient of correlation (r). Thus, the cutoff velocity is
defined by this particular data set. The reported P-value (0.637) is
14-10
subject to the iterative value obtained for the cutoff velocity.
Further study may result in a slight adjustment of the cutoff velocity.
Conclusions
The Marsh ~lcBi rney current meter can be used by field workers with
confidence where expected velocities are less than 5.0 fps (152.4
em/sec). Results of this study show the Marsh McBirney electromagnetic
current meter gives erroneously low velocity readings above this cutoff
velocity. Published electromagnetic velocity readings should be used
with caution unless additional velocity measurements have been made by
another method.
Three areas need further study to better define the relationship between
the velocity readings from the two instruments. First, the cutoff
velocity at which the Marsh McBirney no longer gives accurate results
needs to be better defined. Second, the relationship for velocities
above 5.0 fps (152.4 em/sec), or some other specified cutoff velocity,
needs to be examined and quantified. Finally, the shape of the probe
needs to be experimentally altered to see if separation is the major
cause of the erroneous readings and, if so, to eliminate the separation.
Acknowledgements
The State of Alaska Department of Transportation and Public Facilities
provided funding for this study (Project No. TA-17-82 F26172). The
authors would like to acknowledge Dana Thomas for his help with the
statistics in this study and Cathy Egan for her help with the velocity
profile measurements. He would also like to thank the Alaska
Cooperative Fisheries Research Unit for the use of a Marsh r~cBirney
current meter.
14-1 1
References
Brater, E.F., and H.W. King. 1976. Handbook of Hydraulics. McGraw
Hill, Inc. New York, N.Y.
Grzenda, J. H.
Engineering.
Springer, E.K.
Applications.
pp. 66-83.
1981. Understanding Magnetic Fl owmeters. Plant
Vol. 35, April. pp. 67-70.
1980. The Electromagnetic Flowmeter -its Theory and
The South African Mechanical Engineer. Vol. 30, March.
14-12
DEVELWN~iH ANIJ USE OF A KESOUKCi:S ATLAS FOR THE CHUGACH NAT llmAL FlJKEST
Hy Davia blanchet
Abstract
The Chuyach Nat ion a I Forest encompasses the eastern h a If of the Kenai
Peninsula, Prince William Sauna, and the Copper River Delta Region. The
Forest is manageu for recreation use, f1sn and wildl1fe habitat,
potential wilderness, timber, minerals, and power development. The
"Chugach National forest Environmental Atlas" was developed to aio in
planning for Forest management activities. The Atlas focuses on climate,
water resources, geology, and availaole maps and aerial photography for
the Forest. The intended audience is land managers and resource
planners from a variety of aisclplines.
Development of the Atlas involved inventorying climatic, hydrologic and
geologic data and information collec~ed from tne Forest ana presenting it
in map form with accompanying text. Several of the maps were createa
through new interprecations of existing data. The Atlas can accmnmodate
additions of new maps or changes in the existing maps.
Introuuction
Purpose and Scope. rnroughout the 1 as t 80 years, a great aea 1 of
environmental data and information has oeen collected on the Chugach
National Forest oy numerous agencies ana inaividua·ls. T11e "Cnugacn
Nationa·l Forest Environmental Atlas" was published in an attempt to
compile much of this information into one document. Tile Atlas provides a
reference to a large variety of climatic, hydrologic, geologic, and
aer i a 1 photographic information. Its intended audience is botn resource
specialists and lana managers, and its intended use is for long-range
Forest resource planning and for proJect-·level worK. Information from
the Atlas has oeen used by the author in evaluating a wide variety of
Forest 5ervice proJects. Tne Atlas can aid also resource specialists
I Forest Hydro·lugist, Chugacn National Forest, ~a I E. Northern Lights
Blvd. Suite 238, Anchorage, Alaska.
15-1
investigating other parts of Alaska by identifying types of available
1nformation and by demonstrating a possible presentation format.
The Atlas is comprised of maps, figures, and text which interpret data
conecteo on the Forest. It aoes not oisplay an actual listing of data
records (for example: daily stream aischarge records), but does list
where tne oat a can be touna ana in some cases how it can be best used.
Information in the Environmental Atlas is divided into four primary
climate, water resources, geology, and map ana aerial catagories:
pr10tography resources. This paper focuses on the climate and water
resources sections of the Atlas, and explains the processes used in
developing these section.
Setting. The Cnugacn National Forest presently covers approximately ~.Y
million
Copper
acres
kiver
of the Kenai
Delta east
Peninsula, Prince Wil"liam Sound, and the
to Cape Suck I i ng. There are numerous
non-National Forest land tracts within the Forest boundary ownea by
private inaividuals, Native corporatiorrs, ana tne State or Alaska.
Thirty-five percent af the Forest ·lana ·1 ies within the Greater
l'runicipal ity of Anchorage ana tne Kena1 and f~atanuska-Susitna Boruu9ns.
The remaining 6S% is not within any specific governmental unit other than
the State of Alaska. Communities lying within or aojctcent to tne Forest
boundaries include Cordova, Valdez, Whittier, Girdwood, Seward, Moose
Pass, and Cooper Landing. Anchorage, AlasKa's largest city (population
230,000), lies approximately 35 miles west of the Forest boundary.
The Forest 1s contained witnin tne Kenai-Cnugach ~iountain System. Tne
majority of the landscape is rugged mountainous terrain which arcs the
15-2
Gulf of Alaska. Mountain peaks on the Forest range in elevation from
l , OCJO to 13,000 feet. All areas on the Forest (excepcing some mountain
peaks ana ri ages) have been glaciatea within the last 200,000 years, ana
most areas within the last lU-15 'ouo years. Approximately one tniro of
the Forest is still ice covered. The entire region is also very
seismically active and has experienced episoaes of ooth upl1tting ana
suosiaence in recent geologic time.
Tne Forest is an area of sharp contrasts in climate, runoff, vegetation,
l anascapes, land use, and access. Annual prec i pi tat ion ranges from very
nigh along the Gulf of Alaska to quite low at son1e inlana sites.
Topographies vary from the vast, leve·l plains of the Copper River Uelta
to the ruygeo, glaciated mountains of the Chugach Range. rnese
contrasts, along with a limited data base, add togetner to make
forecasting of climatic and hyarologic events on tne Forest a complex and
often inexact science.
Applying the Atlas for Agency Goals. The U~DI\ Forest )ervice is a
multiple-use land management agency. Management on tne Chugach is
directed pr1marily towards recreation (dispersed and non-aisperseo), fish
ana wildlife habitat improvements, timber production, proposed
wilderness, m1 nera l extract ion, ana power deve ·1 opment. Typ 1 ca ·1 proJec;ts
on the Forest include: campground, trail and winter sports developments;
salmon steeppass construction ana cnannE:I ennancements; moose haDitat
improvement through prescribed burns; long-range Forest planning; timber
sale layout; and review of mining plans of operations. Environmental
information provided in the Atlas is intenaea for use in project
preplanning, site evaluation and selection, and project desiyn, and
monitoring.
15-3
Methods ana Discussion
~laps were used as the primary tool for aisplayiny information in tile
"Chugach Environmental Atlas". Text was adaed to furtner explain the
purpose and use of the maps. A orown, ·1:500,000 (approximately tl rni les
to the inch) Forest map was used as the base for the majority of the maps
in the Atlas. Environmental information was over·layed onto tnese maps in
black. The 1:~00,000 scale proved very satisfactory for displayiny
information on most of tne maps in the At.las. For several maps tne
density of information was too great for resolution at a this scale and
base maps at a 1:.:'~0,000 and l:D3,360 sca·le became neccessary.
The Atlas also disp"lays information using figures ana graphs witn
exp 1 ana tory text. ~1any of the figures have oeen taKen from otner
resource documents ana moo if i ed to the specific i nf ormation a l needs on
the Chugach National Forest.
Maps ana fiyures presented in tnese sections incluae botn original
presentations ana reformatting of existing maps and figures. Development
of the "Chugach National Forest Environmental Atlas" has relieu heavily
on information and formatting used in other resource atlases on Alaska.
The "Alaska Regional Profiles, ::.outhcentral Region" (::,ell<regg, 1974) nas
been particularly useful in terms of providing information, ana for
guidence in searching for more specific informat10n concerning the
Chugach National Forest.
Most of tne meteorologic and hydrologic data stations referenceu in tne
Atlas are located at elevations of 1,500 feet anu ae·low (where tne
n1ajority of Forest anivities ana aeve·lupmencs tal<e place.) Tile Atlas
presents information for botrl low and high elevation areas, however,
15-4
higher elevation information has frequently been extrapolated from data
taken at lower elevation sites. Fo I lowing is a aiscussion of information
displayed in the climate ana water resources sections and how that
information was derivea.
Climate
Introduction. The Cnugach National Forest has extremes of both
precipitation ana temperature. Changes in the weatner can occur rapialy
and may be difficult ta predict. Several locations on the Forest have
annual precipitation and/or snowfall values w11ich are among the highest
in the world, while other areas have very moderate (semi-arid) annual
precipitation. Temperatures vary from moderate and steaay in coastal
regions, to cooler and more highly fluctuating in inland areas, to
continuously cola in high altitude, mountainous regions. The extreme
variation in daylight hours and tne low sun angles (as compared to tne
continental United States) aaas another degree of variaoi I ity to tne
Forest climate. In the conventiona·l climate zone system (temperate,
subarctic, and arctic), most of the Forest lies within tne suoarctic zone
(one to four montns with mean montnly temperatures grea~er than 50"f ana
at least one month with a mean temperature 32"F or coloer.) Forest areas
above 3,000 fee~ in elevation fit tne classification for arctic climate
(no months wnh mean montn·ly temperatures greater tnan 50"F and at least
one month with a mean temperature 32"F or colder.)
This section of the Atlas displays a variety of maps and figures campi ied
from climate data co 11 ected on the Fore st. This data can De used as a
working tool for project design, a11alysis, and implementation.
15-5
1. Hydrometeorologic Stations ~lap. This map displays the location,
station number, and dates of operation for meteorologic and hydrologic
data stations on the Cnugach National Forest and surrounding areas for
which published data is available. Data collectea from tnese stations is
of three different types: 1.) weather records (primarily temperature ana
precipitation) collected by the Environmental Data ~erv1ce of the
National Oceanic and Atmospheric Administration (NUAA), 2.) depth and
water content of the seasonal snowpack col lectea uy tne u.~. ~oil
Conservation Service (SCS) and, 3.) surface water quantity and quality
records collected by tne u.::,. Geological Survey, Water Kesourses l.livision
(USGS).
weather records for a few sites on the Forest date back as tar as the
early 1920's, ana USGS stream aata to 1910. Collection of snowpack data
on ana near the Forest first started in the early ·1950's. Host data
records for stations on and aajacent to the Forest are relatively short
in duration and are sometimes only intermittent over the period of
record. Data gaps exists for many parts of the Forest. Nonetheless, a
remarkable amount of data is available ana can prove very usely fur
project ana·lysis and development.
Accurate identification of data sources is often the first step in
project analysis and may involve aigging through a variety of records to
locate ana date the stations. Tnis map can De very useful for quickiJ
assessing what climatic and hydrologic data sources are available and in
turn, iJOw they may be used.
2. ~lean 1\nnua·l Temperature 1v1ap. This map displays the averctge yearly
temperature at sites throughout the Forest. This value is useful for
15-6
gaining initial insignts into: predictable annual heating costs ana
energy needs, likelihood for development of permafrost soi Is, type of
climate, wildlife habitat areas, ground and groundwater temperatures, ana
environmental stresses to vegetation ana/or structures. The map aisplays
a series of isotherms or contour lines of equal mean annual temperature
for the Forest. Also displayed are the mean annual temperatures from tne
NOAA weather stations on and near the Forest. Isotherms for the map were
extrapolated from the existing NOAA weather stations. The primary
consideration used in developing the isotherms was that of elevation.
Using an average lapse rate of -3.b°F per "1000 feet gain in elevation,
isotherms were drawn along elevation contour lines using existing weather
stations as base points. Factors such as sun exposure, wino patterns,
and topography, whicn in some cases can significantly effect mean annual
temperatures, were not used in developing this ma!J.
3. Mean January and July Temperatures !'lap. For most areas on tt1e Forest,
January is the coldest month of the year and Ju.ly is the warmest. This
map lists four temperatures at each NOAA weather station on and near tne
Forest. These temperatures are as fo.llows: January mean 111ontnly
temperature, January mean daily minimum temperature, July mean monthly
temperature, and July mean daily maximum temperature. These temperature
values are helpful for project design, for energy costing, and for
understanding access problems ana winter icing potentials.
4. ~lap of Primary Storm Tracks by Month. ~lost storm tracks reaching the
Forest come from tne west southwest out of tne l:iulf ot A·lasKa or more
infrequently out of the west from the 8ering ~ea. A map is presented for
each month of the year snowing tne generalized pattern of storm tracK
15-7
movements across southern A 1 ask a. These maps were taken from "C 1 imat i c
Altas of the Outer Continenta-l Shelf Waters and Coastal Regions uf
Alaska, VoL 1 -Gulf of AlasKa" (Brower, 1977). They help to explain
weather system benavior on the Forese.
5. !•lean Annual Precipitation fvlap. This map displays mean annual
precipitation values across the Forest. Tn1s information can De usea in
determining flood flows, mean annual flow, ana low flows for a given
stream. Mean annual precipitatior1 is a useful desiyn criteria for
numerous developments on the Forest sucn as roads, bridges, campgrounds,
timber sales, fish passes, and sKi areas.
The mean annual precipitation map was taken from the ''Region lU, Water
Resources Atlas" (Ott water Engineers, 197~). ~lean annua·l precipitation
includes both inches of rainfall and water equivalent inches of snow.
Tne contour lines on tne map ( isohyetsj snow mean annual precipitat1on
througnout the Forest. These isohyetal lines were developea from: l.J
weather station data, ~.J runoff values ot U~GS gagea watersneu& mir1us
evapotranspiration, ana 3.) elevation contours (increased precipitation
at higher elevations.)
b. ~lean Annual !~aximum SnowpacK (By Lawrence K. ~iayo and Davia
Blanchet). This map shows the average maximum depth of snow on the
ground during tne winter. Displayed art: contours of equal snow deptn
(isopleths) and the actual measurement sites used in development of the
map. This information has a vanety of applications incluuing
determination of: structural loading on buildings and briages, river
runoff and flooa potential&, mode of surface travel, winter range
limitations for wildlife, ana avalanche potential.
15-8
Relative to the large size and complex terrain of the Forest, only few
measurer11ents of snow depth hdve been maae. However, these measurements
are widely scattered ana sample forests, mountains, and glaciers to as
high as 8,400 feet. The following elements were usea in aeveloping the
mean annual maximum snowpack map:
a. Snow course data (U.S. Soil Conservation Service)
b. Weather station data (National Oceanic and Atmospheric
Administration).
c. Snow data on glaciers (U.S. Geological Survey).
d. Snow data from Turnagain Pass (USDA Forest Service).
e. Mean annual precipitation map of the Forese (Ott water Engineers).
f. Location and size of existing glaciers (U.S. Geological Survey).
g. Terrain altitude (U.S. Geological Survey).
h. ~iaJOr storm patterns (National Oceanic and Atmospheric
Administration).
i. ~iajor snow rearistribution by winu ac rliyher altitudes (authors).
j. Slope steepness ana avalanching. Steeper slopes usually have less
snowpack tnan nearoy flat areas due to wino redistribution ctnd snow
sliding (authors).
Tne measurements reported on the map are re.ldtively accuriite. The
contours on the map are extrapo 1 a ted us i ny the criteria mentioned above
ana are less accurate. The snow depths reported an the map are a
genera 1 i zat ion of the rea 1 situation. An accurate map of snow depths
measured in any small area wou 1 d revea 1 deta i 1 s of the cornp 1 ex icy of snow
depth variability that cannot be shown at the scale of this map.
7. Selected Diagrams Relating to ~leteorological Data. This page or the
Atlas reproduces a series of diagrams from other publications and tailors
15-9
them to the Chugach National Forest. These diagrams are as follows:
a. Sun Path Diagram For bD 0 North Latitude. This dia~ram was tal<en from
"A :.alar Design ~Janual for Alaska" {Seifert, 1981) and plots the patn
ot the sun across the sky (solar elliptic) tnrouynout tne year.
Using the diagram, the location of the sun in the sky can be
determ i nea for any t irne of the day ana year. Tn is i nf ormation is
valuable for locating sites or
input. Heating costs for many
reduced oy optimizing solar input.
structures which depend on solar
structures can be siynificantly
The diagram displayed is for t>0°
north ·latitude. This is we latituoe at ~ewaru, AlasKa, nowever, the
aiagram may be applied to the entire Forest witn only a srna"ll margin
of error.
b Sunlight ana Twilight Hours. This diagram, taKen from "AlasKa
Regional Profiles, Southcentra·l Region" (Selkregg, 1974), displays
the numoer of lighted hours per day tnroughout the year. Tne ·1 i ytlteo
hours of tne day include both sunlight and twilight hours. The
diagram can be read for different northern I at ituaes. This
information is valuable for both schedule planning and design.
c. Growing Degree Days. This cnart was cornp1lea by using both tne
"Alaska l{egional Profiles, Southcentral Region" (Selkregg, 1974) ana
add it i una 1 NOAA weather aata for other se I ecteu l ocdt ions on tne
Forest. The chart lists the average growing degree days and length
of growing season for seven cornrnun it i es l oc a teo w i tn HI or near the
Forest. This cnart is useful for planning revegetation or planting
projects.
a. Heating Deyree Days. Tnis chart was developea us1ny tt1e "AlasKa
Regional Profiles, Southcentral Region" (Selkregg, 1974) ana
15-10
additional NOAA weather data. It displays heating days by month for
nine communities within or near the Forest. Heating degree day
values are useful in determining energy needs for different areas and
for different times of the year.
e. Ground Temperature. Tnis chart (~elkregg, 1~74) snows average
monthly soil temperatures by depths for a site located in Anchorage,
Alaska. The chart gives an indication of: wnat Kino of temperature
fluctuations may be seen at various depths, now deep and how
1 ong-1 ast i ng the frost 1 eve I is, ana how temperature fluctuations
decrease witn aeptn. Tnese concepts may be applied (along with
additional weatner and soils data) to most lower elevation areas on
the Forest.
f. Freeze-up and Break-up Data for ~elected Lakes and kivers. Tnis
chart {Selkregg, IY74) shows the average annual dates for freezing
and breaking up of several lal<es and rivers on or near the forest.
The chart indicates the average date when the ice first becomes safe
for walking and vehicular travel, and tnen when it becomes unsafe in
the spring. Lake freeze-up and break-up dates may be extrapolated to
other lakes on tne Forest using caution and additional temperature
and wind data. Extrapolation of freeze-up and break-up to otner
rivers is difficu It because of the number of variables involvea ( ie.
gradient, elevation, local climate, ana drainage characteristics.)
g. Spring and fall Freeze Dates for Selectea Locations. Tnis cnart was
compilea using "Alaska Kegional Profiles, Soutncentral Region"
(::.elKregg, 1974) and aaditional IIUAA weather data. It snows the
average date of both the last spring frost and the first fall frost
for seven communities within or near the Cnugach National Forest.
These dates are useful for planning revegetation or planting projects.
15-1 1
Water Resources.
Introduction. The Chugach National Forest has an abundant water supply
resulting from the heavy precipitation it receives. Fresh water laKes
and streams on the Forest were used in the past by Native cultures for
locating settlements, water supply, and transportation routes. Within
the last century, new uses of water on the Forest have included metal
separation in lode and placer mining operations, aomestic water supply,
industrial water supply (such as canneries), hydropower, and small scale
agriculture. water plays a role in almost all Forest projects. This
section of the Atlas explains water resource characteristics on the
Forest from a variety of perspectives.
Surface runoff, groundwater, and water quality (physical, chemical, ana
biological) are discussed in a narrative in the "Chugach National Forest
Environmental Atlas". A series of maps and figures are used in this
section to display different aspects of the water resource. Also, a
bib 1 iography of water resources reports concerning the Forest is given.
' Development of the maps displayed in tnis section and tneir use is
aiscussea below.
1 . Map of ~Jean Annual Runoff per Square ~li 1 e. This map ae 1 i neates the
mean annua 1 runoff of Forest watersheds in cubic feet per secona per
square mile of drainage area. Mean annual runoff can be a useful tool in
determining watershed storage capabilities for such uses as hydropower,
drinking water, flood control, and hatcheries. Nean annual runoff may
also be used for water rights determinations and a variety of other land
management applications.
15-12
The mean annual runoff displayed on the map was derived by using an
equation from the "k-10 Water Resources Atlas" (Ott Water Engineers,
1979). The equation was developed by regressing a variety of drainage
basin characteristics against the actual mean annual runoff of U~G~ gagea
watersheds on and near the Forest. For this particular regression
equation, mean annual precipitation and drainage basin area turned out to
be the only significant parameters in estimating mean annual runoff. The
equation (for the Forest) is as follows:
~lean annual runoff = .0283 p l.lb A 1.02
Where P = mean annual precipitation ana A = tne drainage area In square
miles. For determining mean annual flow per square mile, A= l.u, ana
A 1 •0 ~ Is aroppea frum the equation. This yielas:
Mean annual runoff per square mile = .0283 p l.lG
Thus, the contour lines of equal annual runoff on tne map become a close
reflection of the mean annual precipitation map (discussed in the climate
section of this paper.) By determining the drainage area of a basin ana
taking it times the mean annual flow per square mile value from the map,
an estimate of mean annual flow for the basin may be derlvea. The ninety
percent confidence Interval for mean annual runoff may be determined from
a graph In the "R-10 Water Resources Atlas• (Ott Water Engineers, 1979).
2. Fifty-Year Peak Flows ~lap. This map shows flow values in cubic feet
per second per square mile for a fifty-year recurrence Interval flood.
The fifty-year flood value Is useful for designing any instream structure
and in determining wnat areas a·long a stream are enaangerea during a
flood. It is a neccessary design criteria for determining what forces
must be witnstood uuring a flood. Tnis map was again proauced by using a
regression equation from the "R-10 Water Resources Atlas• (Ott Water
15-1.3
Engineers, 1979). This equation for Chugach streams is as follows:
Fifty-year peak flow= .63c p 1 •26 A"Y~4 L-.J44
Where P = precipitation in inches, A = drainaye area in square miles, ana
L = + the percent of the drainage basin area which is lake surface. (L
= when there is no lake area). Thus, for a one square mile drainage
witt1 no lakes:
Fifty-year peak flow= .632 pl-26
The flood value contour lines on the map were developed from this
equation, and are again a reflection of the mean annual precipitation
map. By multiplying the drainage area of a oasin times the ~0-year flow
per square mile value from the map, an estimate of the 50-year flow value
is derived for tne given basin (assuming there is not s1gnificant lake
area within the drainage). If there are lakes of any significant size
located within a orainage, the values displayed on the map will be
anomalously hign. The ninety percent confidence interval for the
fifty-year peal< f"Jow may De determined from a graph in tne "K-lU Water
Resources Atlas" (Ott Water Engineers, l97Y).
3. Glaciers ana Glacial Streams ~lap. This map snows the location of
glaciers ana glacial streams on the Forest and surrounaing areas. From
this map it may be seen tnat approximately 30 percent of tne Forest area
is covered by glacial ice. USGS contour maps (1:63,360) were used as a
reference for mapping onto the base map for the Atlas. The map gives a
strong visual aisplay of the extent of present glaciation on and near the
Forest.
4. Mean and Extreme Tides ~lap. This map snows mean and extreme tides for
a number of coastal locations around the Forest for the year 1982. "Tide
15-14
Tables for 1982, West Coast of North and Soutt1 America" (NuAA, 1~81)
served as the aata source for this map. Average daily tidal fluctuations
on the Forest vary from about ten feet for exposed areas along tne Gulf
or Alaska to over 3.J feet along portions of Turnaga1n Arm. The extreme
highs and lows are the maximum and minimum levels of the tide reached in
l%2 ana will differ slightly in other years. The tidal fluctuations
displayed are useful for determining access problems for a given stretch
of coastline, and for consideration ir1 oesigning coastal structures (sucn
as boat ramps, docks, and cabins.)
Water Resource Reports Concerning the Forest. Tne Atlas lists a
bibliography of water resource reports (1904-1~81) relating to the Forest
which have oeen published in a variety of formats oy the U.S. Geological
Survey, and the Alaska Division of Geological and Geophysica·l Surveys.
These reports were accessed through sever a I bib I i ograph i es published by
the USGS (Cobb 1~74-82).
Summary
This paper has only focused on the development of tne climate and water
resources section of the "Chugach National Forest Environmental Atlas".
Uata and information for this document were taken entirely from existing
sources with no new data being collected specifically for the Atlas.
Information displayea in tne Atlas involves bottJ original presentations
and rep'lications from existing documents.
The Atlas also covers a variety of geological and map and aerial
photographic data ana information concerning tne Forest, but these
sections are not discussed in thus paper. Tne information aisplayea in
the Atlas is used by the author in planning and investigation of Forest
15-15
projects. The Atlas is designed in a loose leaf format so that it may
accommodate additional maps or revisions to tne existing maps. Th1s
format also allows for maps and figures to be removed for copying.
The Atlas may be used in application to both Forest projects and proJects
on State, Native, community, and private lands within the Forest. Land
management agencies interested in developing similar map displays for
other parts of Alaska may wish to reference the Atlas for information
sources.
15-16
References
Blanchet, David. "Cnugach National Forest Environmental Atlas'',
Anchorage, AK: USUA-F orest Service, Region 10, I ~83.
Brower, William A., Jr., Oiaz, Henry F., Precntel, Anton :,., Searoy,
Harold W., Wise, James L., "Climatic Altas of the Outer Continenta·l
Shelt Waters ana Coastal Regions of AlasKa, Vol. I -Gulf or AlasKa".
Anchorage, AK: Arctic Environmental Information and Data Center, 1~77.
CobtJ, E.H., "keports of tne Alaska Division of beologica·l ana Geophysical
Surveys and Predecessor Agencies, 1913-1973, Indexed oy Quadrangle".
Ancnorage, AK: U.S. Geulo~ical Survey Open File Report 74-209, 1974.
Cobb, E.H., "Geological Survey and Selected U.S. Bureau of Mines and
Alaska Division of Geological ana GeophysJcal Surveys Reports ana
Maps on Alaska keleased during 1974". Anchorage, AK: U.S. Geological
Survey Open File Reports, 1975 throuyh 1982.
Environmental Data and Information Service,
Atmospheric Aaministration, "CI imatoloyica·l
U.S. Dept. of Commerce, 1900-1981.
Nation a I Oceanic ana
Data''. Ashville, N.C.:
Hartman, Charles w., Johnson, Pni lip 1{., "Environmental AU as of Alaska".
Fairbanks, AK: University of Alaska, 1978.
Ott Water engineers, Inc.,"Water kesources Atlas". Juneau, AK:
USDA-Forest Service, Region 10, 1979.
National Oceanic and Acmosplleric Administration, U.S. Uepartment of
Commerce, "Tide Tab I es for ·1982, west Coast of North and Soutn
America". kockville, JvilJ: 1981.
Searby, H.w., ''CI1mates of tne States: Alaska". Climatology of tne United
States No. 60-49, Environmental lJata Service, ESSA, 1968.
Seifert, kichard D., "A So.lar Design 1'/anual for AlasKa". FairbanKs, AK:
University of Alaska, 1~81.
Selkregg, Lydia L., "Alaska Regional Profi1es, Soutllcentral Region".
Anchorage, AK: Environmental Information and Data Center, University
of Alaska, 1974.
Soi I Conservation :,ervice, "Snow Surveys ana Water Supply Outlool< for
Alaska". Anchorage, AK: U.S. Department of Agriculture, 1951-198~.
Still, Patsy J., ''Inaex of Streamflow and Water Quality Records".
Open-File Report 80-600, Anchorage, AK: U.S. Geological Survey, 1978.
U.S. Geo·Jogical Survey, "List of Geological Survey Geo·rogic and
Water-Supply Reports and fvlaps for Alaska". Washington, D.C.: U.S.
Government Printing Office: 1~81.
15-17
U.S. Geological Survey, ''Water Resources Data for AlasKa''. Anchorage, AK:
1966-80.
U.S. Geological Survey, ''Surface Water ~upply of the United States
1961-65, Part 15. Alaska''. water-Supply Paper 1936, Washington, D.C.:
U.S. Government Printing Office, 1~71.
U.S. Geological Survey, "Compilation of Records of Surface Waters of
AlasKa, October 1950 to September .l9b0". Water-supply Paper 174U,
Washington, D.C.: U.S. Government Printing Office, 1!:164.
U.S. Geological Survey, "Compilation ot kecoros of Quantity and ljuality
of Surface Waters of A I ask a through September 1950". water-Supply
Paper U7~, Wasnington, D.C.: U.~. Government Printing Office, l~b7.
U.S. Geological Survey, "List of Geological
Water-Supply keports and i'Japs tor Alaska".
Government Printing Office: 1981.
Survey Geo I og i c
Washington, D.C.:
and
u.s.
U.S. Geological Survey, 1:63,360 and 1:<::50,000 quaarangle maps, (various).
15-18
Abstract
THE AQUATIC PORTION OF THE INTEGRATED RESOURCE INVENTORY,
TONGASS NATIONAL FOREST -CHATHAM AREA
by Daniel A. Marionl and Steven J. Paustian2
The development, implementation, and applications of the Aquatic portion
of the Integrated Resource Inventory (A-IRI) for the Tongass National
Forest-Chatham Area are described in this paper. The A-IRI is a
characterization of water and fisheries habitat designed to provide
aquatic resource information and evaluation techniques for use in forest
planning and management. Essential to A-IRI applications is the use of a
channel type classification system. Watershed drainage networks are
stratified into stream segments with homogeneous aquatic resource
characteristics. These characteristics are then used to determine
aquatic resource conditions and sensitivity. Use of the channel type
classification system promotes resource discipline integration while
dividing the drainage network complex into a workable number of resource
management areas. In addition, it provides an efficient resource
accounting and data extrapolation system. Other notable developments
during A-IRI implementation have been a channel type mapping procedure
and computer data base.
The A-IRI is designed so that aquatic resource information and
evaluations can be applied at several planning levels. Recent A-IRI
applications have clearly demonstrated its utility and value in making
multiple use resource management decisions.
A hydrologic and fish habitat resource inventory is currently being
conducted on the Tongass National Forest -Chatham Area (TNF-CA) which is
noteworthy for several reasons. It is the first complete, systematic
inventory of those resources within the intensively managed areas of the
TNF-CA. Fisheries and hydrology disciplines are combined in this
inventory to produce one, integrated effort driven by identified resource
information needs. A hierarchial stream classification system is
lHydrologist, Tongass National Forest -Chatham Area, P.O. Box
1980, Sitka, Alaska 99835.
2supervisory Hydrologist, Tongass National Forest -Chatham Area,
P.O. Box 1980, Sitka, Alaska 99835
16-1
employed to delineate meaningful resource management areas and provide a
tool for data stratification and extrapolation. The development,
implementation, and applications of this inventory are described in this
paper.
Background
The legal direction to collect and interpret resource information is
contained in several national laws. The Forest and Rangelands Renewable
Resources Planning Act of 1974 as amended by the Nationa·l Forest
Management Act of 1976 requires the Forest Service to supply basic
distribution, capability, limitation, and condition information on soil,
water, and geologic materials on all National Forest System lands. In
addition, the Soil and Water Resources Conservation Act of 1977 and the
Federal Land Policy and Management Act of 1976 both require that soil,
water, and related resource inventories be conducted periodically.
Federal laws also strongly encourage the integration of these natural
resource inventories and information. The Multiple Use -Sustained Yield
Act of 1960 requires a joint consideration of all major outputs from the
national forests. The interdisciplinary planning process stipulated
within the National Environmental Policy Act of 1969 would be greatly
facilitated by an existing integrated information base. The Forest and
Rangelands Renewable Resources Planning Act of 1974 as amended by the
National Forest Management Act of 1976 directs that all Forest Service
planning activities recognize and consider the interrelationships which
exist between resources.
Hamilton (1978) offers several reasons for integrating separate resource
16-2
inventory efforts. Among others he lists increased data collection
efficiency, providing for data compatibility, improving resource
assessment precision, and providing a single, best estimate of the
resource situation. However, he concludes that the most direct reason
for integration is that because resources are interrelated and
interdependent in nature, our inventory approach must allow us to
describe these interactions or interaction products in a particular time
and place. Dixon (1981) supports these reasons, adding that separate
resource inventories cannot meet current multiresource information needs
because of the high costs involved and the difficulty in assessing
resource interrelationships.
Integration is also necessary to produce a resource classification system
in which all concerned disciplines relate their information and
evaluations to common geographic areas. This system would incorporate
the principle of resource interrelationships while providing a means for
delineating resource management areas in which environmental conditions
and potential responses to development activities are similar. Most
importantly, it would reduce the natural complexity of the biosphere to a
manageable number of different situations that can be communicated to and
understood by land managers, planners, and technical specialists (Bailey
et al., 1978).
The TNF-CA Integrated Resource Inventory project is being implemented as
a means of collecting and evaluating multiresource information within a
framework of integrated land and stream classification systems. The
project is divided into two parts based on an aquatic versus a land
emphasis. Timber, soils, ana wildlife specialists have adopted a
16-3
Land Systems Classification approach (Wertz and Arnold, 1973) to
integrate land based resource data collection and interpretation. In the
aquatic portion of the Integrated Resource Inventory 3 (A-IRI)
hydrologists and fisheries biologists are using a Channel Type
Classification System for integrating stream based resource data
collection and interpretation. The two systems are designed to be
complementary in that each fills a void inherent in the other. They are
linked through the use of a common landform classification legend. In
the Landsystems approach landforms are subdivided using geology,
vegetation cover, slope class, and soil type to produce lanatypes. These
landtypes satisfy the needs of timber, soils, and wildlife specialists
but are inadequate for characterizing and evaluating aquatic resources.
In the Channel Type Classification System landforms are subdivided using
channel morphologic features to produce channel types. These channel
types suit the needs of hydrologists and fisheries biologists but are
insufficient for timber, soils, and wildlife specialist needs. Either
system alone is inadequate to address all potential natural resource
concerns for a given landform. However, used together they provide a
comprehensive means for addressing both the lana and aquatic resource
situation for all TNF-CA landforms.
The A-IRI is designed to provide water and fish habitat resource
information and evaluation techniques for forest management and
planning. The Channel Type Classification System is employed to stratify
drainage networks into stream areas with different hydrologic and fish
3Former-ly called the Aquatic Resource Inventory (e.g., Paustian et
al., in preparation).
16-4
habitat conditions and responses. This system provides the common stream
based geographic area in which fisheries biologists and hydrologists
collect resource data and make evaluations. Using this integrated
approach the A-IRI accomplishes the objectives of separate Level IV
Forest Service Fisheries Survey (USDA Forest Service Alaska Region, 1981)
and Order 2 Water Resource Inventory (USDA Forest Service Intermountain
Region, 1979) projects while avoiding the redundant data collection and
potential incompatibility of separate inventory efforts.
Geographic Setting
The TNF-CA covers almost 28,000km 2 of the northern Southeast Alaska
panhandle. Approximately 39% or 11,000km 2, of this area is allocated
for intensive, multiple use management under the Tongass Land Management
Plan (USDA Forest Service Alaska Region, 1979). The A-IRI effort is
currently focused on this latter area, with the remaining TNF-CA area to
be considered later (see Figure 1).
The inventory area falls within the Sitka Spruce -Cedar -Hemlock forest
section of the Pacific Forest Province as defined by Bailey (1980). The
area climate is maritime with mean annual temperatures ranging between
2°C and l0°C, and annual precipitation ranging between 1500mm and
6600mm. Vegetation is predominantly coniferous forest types with
interspersed moss sedge marshes (after Viereck and Dyrness, 1980).
Alpine tundra and shrub types occur at elevations above 500m.
Major streams draining this area occur within glaciated valleys of
varying width. Watershed size typically varies between 25km 2 to
200km 2 for these streams. Runoff commonly originates within alpine
16-5
Primary
Integrated
Resource
Inventory Area
___ Chatham Area
Boundary
Map Area
-Scale 1:3,640,000 rr-, ~
I <:., \..__ 0"1
J ~ .... • .. 1-'<o (."'<)..,
~~';.,\ ..,,.. ........
~<l' \
'
\
\
1
N
,r\
\ , .,
I ..
Figure l. Area covered by the aquatic portion of the Integrated
Resource Inventory.
snowfields or cirque basins on the islands, or from alpine glaciers on
the mainland. High unit runoff volumes coupled with rapid elevation
drops over short distances result in high energy streamflows occurring
within the inventory area.
Five commercially important anadromous fish species utilize TNF-CA
streams for spawning ana rearing. The majority of this activity occurs
within watersheos which are mostly undeveloped. Ex.1sting development is
16-6
Identify r Pre-Map
Forest Issues Inventory Area
~ ~
Develop Resource I Collect Data I
Evaluation Procedures ...
~ Develop
Data Base
Identify Information/
j_ Inventory Data Needs
Inventory
Development l Implementation
Analyze Data
Develop Channel Type
Channel Type Classification System Correalation
~ Perform Final
'-Document Inventory
Methodology
Resource Evaluations
~
I Final Map
'\. Inventory Area
Publish
Inventory Report
Figure 2. Flow chart of the process used for the aquatic portion of the
Integrated Resource Inventory.
largely restricted to forest road systems, timber harvest activities, and
small scattered fishing villages and logging camps.
Inventory Development
The A-IRI process discussed below is illustrated in the flow chart in
Figure 2. This process can be divided into a development phase and an
implementation phase. During the development phase information needs are
determined and procedures are developed to satisfy these needs. In the
latter phase these proceoures are implemented to obtain the required
data, information, and map products. The A-IRI process began in 1981 and
is scheduled to be completed by 1985.
Forest resource issues are addresseo by information and evaluation
16-7
techniques developed from the A-IRI. They were compiled from previously
identified TNF-CA management issues (USDA Forest Service Alaska Region,
1979, 1980, and l983a) and from consultation with regional, forest, and
research specialists. An information needs assessment was then performed
to identify what evaluation techniques, information, and data were
required to address these issues, and what information was currently
available. The assessment results are summarized in Table 1. They
indicated that very little was currently available to address these
issues. Several resource evaluation procedures (often called management
interpretations) were recognized as being needed to meet the A-IRI
objective. In order to use these procedures successfully it was
necessary to control the variation in environmental characteristics
within each area they were applied. Therefore, a stream classification
system for stratifying drainage networks into discrete channel areas with
consistent physical characteristics was also identified as being
essential.
Management interpretations are systematic methods for assessing resource
conditions (e.g., fish habitat quality) and resource sensitivity to
disruption by forest development activities. They are designed to answer
the questions posed by forest resource issues. The management
interpretations developed for the A-IRI are listed as column headings in
Table 2. They are generally qualitative rating procedures due to the
lack of quantified cause-and-effect knowledge. Each procedure is based
on measurements or evaluations of physical channel characterisitcs
determined as being important in interpreting natural resource conditions
or resource sensitivities. These determinations are based on resource
16-8
0'\
I
\!)
Forest PlannlrQ
Issue
Potential irn-
P'~cts to "ln~d
ramous or resi-
dent fish habi-
tat.
Potential
chames in
water quality,
Potential
failure of
in-channel
structures
(nonconsurrptive
water uses).
Potential
cansumptl ve
water uses for
water supply,
aau:::~culture,
recre<~tian,
minerals, or
energy devel-
opment.
Potential lm-
p8cts to wet-
laMs and
flood plains.
TAble 1. To~ ass National Forest-D"latham Area Aquatic Resoun::e Data Needs Assessment SulMlary,
Activities
Involved
Timber harvesting
Road construction
and maintenance
Fisheries enhance-
ment projects
Timber harvestinq
Road construction
and maintenance
Fisheries enhance-
ment projects
Stream crossit"Q
construction and
maintemmce
Fish pass con-
structlon and
maintenance
Timber harvest
Road construction
and maintenance
Inchannel struc-
ture construe-
tian and main-
tenance
Water withdrawls
Timber harvest
Road construction
and maintenance
Info rmatlon/Resoun:e
Interpretation
Required
Location, quRnt!ty,
quality, and sensi-
tivity of exist!~
and potential fish
habitat.
Natural and post-
treatment water
quality charac-
teristics.
DischaiQe varl-
at ion
Site stability
Typical end area
Oischaxge vari-
11tlon
51 te stablllty
Ins tream flow
needs
I)Jantlty, loca-
tion, and sensi-
tivity of eds-
tlng wetlands
and flood plains
Factors
of Concern
Sediment
Tf!rrperature
Debris
Riparian VeQe-
tation
Olannel stab-
ility
Sediment
Terrper11ture
All chemical
p~:~r.<:~meters
reaulAted
by Alaska
Stl'Jte Water
0J11llty Stan-
dards
Dischame var-
iation
Flow contain-
ment
D"lannel mi-
gration
D!scharqe var-
iation
Flow contain-
ment
O"lannel m!Qra-
tlon
Veaetatlon
water quality
Sall erosion
Olannel erosion
Existinq Informa-
tlon/Dai:a Sources
ADF&G escapement re-
cords and anadromaus
fish stream maps.
USGS water quality
data for SE Alaska
streams
Various research
publications
R-10 Water Resourees
AITaSTOft W11Eer
Enolneers, 1979)
Existin:~ instream flow
determination
methods
R-10 Water Resoun:es A£1,-,---
water Use Penni ts
Addltlon;Jl
---------~~~t~·~~~.e~d~'----------------
Channel type classification
Sediment input, storaa~, and trans-
port charar.teristlcs
O:!bris loadino char:::~cteristics
Stream temoerature ch:::~r:::~cteristics
Olannel stability char11cteristtcs
Habitat au.::~ltty
~ssaqe barrier locations
Olannel type classific~Jt!on
Water quality characteristics
Fbtential for post-tre"ltment water qu.::~lity
chaf'Qes
~bsectlon class!fic.::~tion
Channel type classification
S.,nk and bed corrposl tlon
Bank and bed stability
StreRm power
D"lannel morpholoay chal'i'lc-
teristlcs
Channel type cl:::~sstficatlon
Olannel st11bility ch.::~r
acteristlcs
Stream power
Channel morpholoqy charac-
teristics
Landtype classification
Olrmnel type cl~:~sslflc~:~tlon
Veoetation ch.::~r:::~cteristics
Channel stabillty ch:::~rar.teristics
Soil characteristics
Sediment irput, storaae, and transport
characterlstlcs
Stream terrperature characteristics
Table 2. Manaaement Interpretations for Selected Tonaass ~tlonal ~orest-Ch~tham ArP.~ Cha~l Tyoes
(revised 9-15-93).
Naturnl Condition Inte!Eretations Resource Sensitivity Interoretatlons
Natur~l Flood Stream eros-Stream eros-End Are::~ Rlonrian IIQ.Jatlc
Olannel Sediment Sedl~nt FrequencY sino JPoroach siroo 51 te Sta Recomenda t ton Zone V.:.lue
~ F\::ltential OJnve:z:arce <•l H:izard b!lltv H=.!zard -.J!J._2_)-Sensitlvttv ~
AI Hloh Hlah High Hiqh Mod
A2 High High IO Hlah Low 15-20 Low Low
A4 Low Hlqh Mod-Hioh Low Low
A5 Hl~h HIC)h Low Low Low
BIA Mod Low 10 Low--Mod Mod 15-20 Hloh Hloh
AlB Low Low 10 Low Mod 15-20 Low ">d
82 Mod Mod 18 Mod-H!Qh Mod 30 Hloh '<od
83A Hiqh Mod "'od-High Hl<t> Hiqh Hloh
838 Hlqh High Mod Hiah Mod Low
84 Low Mod 10 Low Mod 15-20 Low Mod
CIA Hlah Low 35 Mod Hloh 60 Hloh Hloh
ClB Mod Mod Ia low ..Mad Low 30 Low Mod
C2A Mod Low 35 Mod Hloh 100 Mod Mod
C28 Hloh Mod 35 HIC)h Hloh 100 Mod tow
C34 Mod Mod 35 V. Hlqh Low 60 Mod tow
C38 Hl9h Mod 35 V. High tow 60 Mod tow
01 Mod Mod Low Mod tow
E Mod .Mod Hloh Mod Low ..,d
NOTF.S: Ol:tmel typP.s ::~nd manaaement interpretations ror the Tonaass National Forest-O'l::ltham ArP-R. ~re rlesr.rlhP.d ln
the AQu>J.tlt: ftlrtlon or the Inteqr.~terl Resource Inventory ltlrv1hook (USDA Forest ServlcP. Alaska RerJion, l'lEIJb).
The symbol "---" tndlclites that the lnterpretatton is not applicable to that pi'lrtlcuhr chamel tyee.
interrelationships recognized in the literature, successful procedures
used elsewhere, and the professional judgment of TNF-CA resource
specialists. Each qualitative interpretation procedure produces an index
number that is then assigned a relative ranking. For resource condition
interpretations these rankings are based on the range of values founo for
all TNF-CA channel types. For resource sensitivity interpretations these
rankings indicate the probability of severe limitations to forest
development activities being present. The current management
interpretation rankings for selected TNF-CA channel types are also shown
in Table 2.
The Channel Type Classification System (CTCS) is the means by which
drainage networks are divided into channel areas having relatively
16-10
homogeneous management interpretations. It satisfies the need for a
manageable framework, reducing natural complexity (Bailey et al., 1978)
while strengthening the ecological foundation of land management (Nelson
et al., 1978). In addition:
1. It provides a framework for an aquatic resource accounting system.
2. It serves as the initial stratification step in a data sampling
design method.
3. It provides a means of data extrapolation to unsampled areas.
A review of existing stream classification methodologies found that no
previously documented system suited the A-IRI requirements. The stream
reach pre-typing methods of Aquatic Biophysical Inventory of British
Columbia (Chamberlain, 1980) are poorly defined and seem intended for
single issue interpretation rather than several issues as required by the
A-IRI. Subdivisions by landform alone as done in Cole (1972), Collotzi
(1974), and Harding (1981) produce too general a stratification for
intended A-IRI uses. In contrast, habitat type classification schemes
such as Bisson et al. (1981) or methods using first order watersheds
(Lotspeich and Platts, 1981) are too site specific or large in scale to
be appropriate. Therefore, the CTCS is largely a unique product,
although it makes use of many concepts put forward in these earlier works.
Three concepts form the basis of the CTCS (Paustian et al., in
preparation). First, geomorphic processes that are independent of
in-channel processes affect stream channel characteristics. Second,
in-channel fluvial processes affect channel characteristics and fish
habitat quality. Third, abiotic processes within the streamside riparian
zone affect in-channel fish habitat quality. Implicit in all three of
16-11
these concepts is the assumption that physical channel characteristics
determine aquatic resource conditions and sensitivity.
Based on these assumptions then, physical characteristics which
distinguish areas of differing out-channel geomorphic processes,
in-channel fluvial processes, and riparian processes can be used to
produce the CTCS. The physical characteristics used to distinguish
selected TNF-CA channel types are shown in Table 3. Figure 3 is included
to show the relative size and landscape positions of these channel
types. At present 27 channel types have been defined for the TNF-CA
although this number may change before the A-IRI completion.
Completing the A-IRI development phase is the preparation of the A-IRI
Handbook (USDA Forest Service Alaska Region, 1983b). Documentation of
all concepts, techniques, and procedures used for classifying and mapping
channel types, making management interpretations, collecting data, and
using the A-IRI computer data base are contained within this report. It
is a working draft at present, being revised as preliminary data analyses
and increasing experience result in procedural modifications. It will be
in final form at the A-IRI completion in 1985.
Inventory Implementation
As shown in Figure 2, the A-IRI implementation phase consists of five
steps: channel type pre-mapping; data collection; preliminary data base
development; final data analysis and channel type correlation; and final
report publication. Highlights of each of these steps are discussed
below.
The channel type pre-mapping procedure is a technique for locating and
16-12
Table 3. Awsic~l 01aracteristics used to Distinouish Selected TOnq::!SS
National Forest -Dlatham Area D"lannel Types.
Olannel Adjacent 01annel Basin C<inooy ,
Type Land form( s) 1 Morpholoov2 Area3 Caver"
Al (Steep, SJb:<ilp ine side-Very hioh gradient Sm:<ill Moderate
mountain slope, slopes Deep ircision to hictl
forest channel) Mountain slopes Narrow width
Hillslopes Sinqle channel
A2 (Hilfl grad-SJbalpine side-Hiqh gradient Moderate Moder:3.te
lent, upper slopes ModerF~te to to hictl
valley, forest Mountain slopes deep incision
channel) Hill slopes Narrow width
S!rqle channel
BlA (Low grad-Undissected foot-Low gr<~dient Sm<~.ll to Moderate
lent, lowland, slopes Shallow incision troder:::~te to hioh
forest channel) Flood plains Narrow width
Slopinq plateaus Sirqle channel
82 (M::Jderate Mountain slopes Moderate gradient "'ode rate Moderate
gradient, mid-H.illslopes Shallow to moder-to larqe
dle valley, Undissected foot-ate irclsion
forest channe 1) slcoes Moderate width
Alluvial fans Single channel
CIA (Low grad-Flood plains Low gradient "'!oderate Low to
lent, valley Shallow incision to lame moderate
bottom, forest "'klderate width
channel) Multiple channels
C3A (Moderate Mountain slQJes Low to moderate Lame to Model"<<te
gradient, nar-Hillslooes gr3dient very hme
row valley, lhdissected foot-Moderate incision
fOrest channel) slopes Moderate width
Sln;~!e channel
01 (Moderate SJbalplne side-Low to moderate Sm:<ill to Low
gradient, upper slopes gradient moder:::~te
valley, g!a-~.fountain slopes Shallow incision
cia! outwash Cirque basins Moderate width
channel) Braided channel
E (l.Dw grad!-Estuaries Low gradient Laroe to Low
ent, estuarine Shallow incision very lame
channel) Moderate to broad
width
Multiple channels
NOTE: All feature interpretations are made usinq 1977 1:15,840 color ::~erial
~oto9raD1s.
!Landforms are defined accordina to the Landfonn Descriptive Leoend of the
Olatham Area (USDA Forest Service Alaska Reoion 1982).
2Qhannel incision r:::~nks: Shallow = less than 3m; moderate = 3m to !Om; deeo
= !Om to 30m; very deeo = gre11ter than JOn. D"lannel orad!ent ranks: Low
less than 3%; moderatP. = ~ to ~i hid"l = 6% to l!l'l:; very hloh = gre<:~ter than
lilt. D"lrmnel oat tern classes: Slro:Jie ch<:~nnels have one continuous main
chF~nnel bed; multiole channels hav~ a mRln channel bed that ls freouently
broken by overflow r.hnnnels or lsi11nds; br:::~lded channels h<lve numerous,
lnterl::!Ced channels and extensive or:::~vel bar develop~nt. Oh:::~nnel width
ranks: t~rrow = less tMn 10m; moderate = 10m to 30m; broad = oreater thRn
30n. -
3Basin A.reR classes: Small = less th<ln 3.2 km2; moderate = 3.2 krn2 to
a.o km2; laiQe = a.o km2 to 24 krn2; very l::uqe = gre<ltP.r th:m 24 km2.
4CAnooy cover cl:::~sses: Low = less than 2~; moderatP. = 2~ to srn;; h~d1
greater than SO%.
16-13
Figure 3. Typical channel type distribution in Tongass
National Forest -Chatham Area watersheds.
delineating channel types on 1:15840 color aerial photographs. It was
developed by TNF-CA hydrologists and fisheries biologists using standard
aerial photograph interpretation techniques (Paustian et al., in
preparation) found to be effective in consistently differentiating
channel types. In addition to aerial photograph interpretation
techniques, methods were also developed for delineating the extent of the
drainage network to be mapped, and for assigning each stream order
segment a unique identification number. We have found this pre-mapping
procedure to be most accurate and precise when interpreters visit several
of the areas they have mapped. This allows them to calibrate their
aerial photograph perceptions with the actual channel features.
Field data acquisition is accomplished using a representative sampling
design method oaseo on channel types. Channel areas representative of
the channel types within a watershed are ioentifieo during pre-mapping
16-14
and visited by field crews. Emphasis is given to sampling within those
channe 1 types with expected high fisheries va·l ue because management
interpretation accuracy is most important there. Field crews composed of
one hydrologist and one to two fisheries biologists are able to
accurately and efficiently accomplish this data collection thus
minimizing personnel needs.
A computer data base is used to facilitate A-IRI data storage, analysis,
and retrieval. At present, it is a two dimensional data array stored on
the USDA Fort Collins Computer Center system. This arrangement
facilitates preliminary data analysis using the extensive software
statistical programs available at Fort Collins. Such a computer data
base has proved essential for the efficient handling and analysis of a
data file that already exceeds 2500 eighty column lines, and will
probably be twice this size when field work is completed. A-IRI data
will be converted to a more flexible, multidimensional data base system
in the near future to facilitate future data storage, updates, and ad-hoc
retrievals.
Once all data are accumulated and analyzed, final channel type
correlation will occur. The correlation process will eliminate those
channel types covering insufficient area to be significant. New channel
types may be created where the data indicate substantial variation exists
within a single channel type, and such variation can be discriminated in
a systematic manner. Data analysis completion will also permit the final
manayement interpretations to be determined for all channel types
retained after the correlation.
16-15
A final A-IRI report will be published along with the channel type maps
by 1985. The final map product will be published on a 1:31680 scale
topographic map base. This report will contain the final channel type
descriptions and management interpretations. A User's Manual for
non-resource specialists will be included to explain the CTCS, the
interpretations, and their applications. This will insure understanding
and the appropriate usage of the A-IRI information by administrators,
planners, and project engineers and foresters.
Inventory Applications
The A-IRI is primarily intended to supply resource information and
interpretations for forest project planning. Individual channel types
are designed to be of a size most applicable for this purpose. However,
the CTCS is also designed to be a hierarchial classification system in
which channel types can be aggregated into larger hydrologic groupings
(Paustian et al., in preparation). This hierarchy and possible
interpretations associated with each level are shown in Table 4. A-IRI
information can be generalized through successive hierarchy divisions to
produce management interpretations at several organizational or
information resolution levels. This CTCS feature provides for
coordination between different Forest Service management levels and
maximizes A-IRI information usage. Both of these attributes are highly
important in deciding the adequacy and value of a classification system
(Frayer et al., 1978).
Although it is still not completed, the A-IRI information and
interpretations have been used recently in developing the Alaska Lumber
and Pulp Co. 1986-90 Operating Plan (USDA Forest Service Alaska Region,
16-16
Table 4. Possible Management Interpretations for Different Hierarchy
Levels of the Tongass Nati anal Forest-Chatham Area Channel
Type Classification System.
Classification
Level Size
Subsection 50,000 -250,000ha
Watershed 1,000-lO,OOOha
Channel type 1,000-5,000m
association
Channel type 500 -lOOOm
Primary Interpretation Value
Regional fish habitat amount and
value patterns
Regional/Forest chemical water
quality patterns
Forest fish habitat value patterns
Land use allocations
Reconnaissance level projects
(e.g., wilderness plans)
Project planning (e.g., timber
sales)
l983a). Inventory information was used as the basis for the hydrologic and
fisheries resource reports prepared for each of the 48 affected watersheds on
the TNF-CA. The availability of A-IRI data collected prior to the project
proposal eliminated much of the need for additional field work. In addition,
use of the preliminary channel type maps and descriptions maximized resource
specialist field time efficiency by directing them to those channel areas most
requiring onsite inspections. Areas indicated as having highly sensitive or
valuable stream resources could be thoroughly checked to avoid or reduce
impacts. Resistent or less valuable areas could be visited briefly or, in
certain cases, safely ignored.
An A-IRI management interpretation procedure called the Aquatic Value
Rating (Paustian et al., this volume) was also used in the project plan
proposal. Using this procedure stream value classes were developed based
on fish habitat quality and sensitivity to adverse impacts from timber
harvest activities. The relative impacts of each operating plan
16-17
alternative were then evaluated based on the amount each stream value
class would be affected by each alternative. This approach proved an
efficient, rational, and defensible means of judging the potential
resource impacts between the plan alternatives.
A-IRI utility has also been demonstrated recently in resolving two site
specific resource problems. In each case, knowledge of typical channel
type characteristics and processes gained through the A-IRI, and
extrapolation of this knowledge to other areas using the CTCS, greatly
aided specialist ability to evaluate specific resource questions.
The first situation involved a proposed municipal water intake
development for the city of Hoonah on Chichagof Island. Available
anadromous and resident fish habitat possibily affected by this project
were quickly identified without fielu work using the channel type
pre-mapping. Potential resource impacts were efficiently analyzed using
the appropriate channel type management interpretations and verifying
their accuracy during a brief field visit. In addition, the assessment
of potential in-channel structure stability problems did not have to rely
on site observations alone. knowledge gained from numerous observations
of streamflow, channel migration, ana bank ana bed stability
characteristics made on the same channel types in different areas could
also be drawn upon for this evaluation. Therefore, the confidence placed
in the accuracy of this assessment was greatly increased.
The second case involved a boundary dispute between the TNF-CA and a
private lana owner. A change in the stream course used as the legal
boundary between these two parties resulted in the private owner claiming
16-18
the additional land which now lay on his side of the boundary stream.
Once again, A-IRI information on the channel types in question, plus
onsite hydrologic observations, lended very strong support to the TNF-CA
position that the stream course change was an avulsive action, and
therefore the original boundary location was still enforce, regardless of
the present stream location. The case was settled out-of-court in favor
of the TNF -CA.
Summary and Conclusions
Managing Alaskan water resources is a complex task due to the diversity
and high value of the resource, and the numerous development interests
involved. It is a task made more difficult by the lack of basic aquatic
resource information and evaluation techniques for Alaskan streams. The
aquatic portion of the Integrated Resource Inventory is a particularly
useful means of gaining this information on the Tongass National Forest
Chatham Area. To date, approximately 8000km 2 have been mapped and
inventoried. Information and resource evaluation methods developed
through this inventory have already proved valuable in assessing forest
management related problems. Two key features make this inventory a
sucessful tool. First, all data collection is driven by identified
management interpretation needs. Second, use of the Channel Type
Classification System allows for the delineation of integrated aquatic
resource management areas; systematic resource accounting and data
extrapolation; and a hierarchial structure for applying resource
information and evaluations at several management levels.
Acknowledgments
The authors are grateful to Steve Brink, Charles M. Holstine, Norma Jo
16-19
Sorgman, and Denise P. Narion for their critical review of the
manuscript. We also wish to thank Connie Adams who typed the manuscript,
and Patricia Pierce and Chery"l ~loody who helped with preparation of the
tables and figures.
References
Bailey, Robert G. 1980. Description of the Ecoregions of the United
States. USDA Forest Service Intermountain Region Misc. Publication
No. 1391, Ogden, Utah, 77 pp.
Bailey, Robert G., Robert D. Pfister, and Jan A. Henderson.
Nature of Land and Resource Classification - A Review.
Forestry, val, 76, no. 10, pp. 650-655.
1978.
Journal of
Bisson, Peter A., Jenifer L. Nielsen, Ray A. Palmason, and Larry E. Grove.
1981. A System of Naming Habitat Types in Small Streams, with
Examples of Habitat Utilization by Salmonids during Low Streamflow.
In: Acquisition and Utilization of Aquatic Habitat Inventory
Trlformat ion: Proceedings of the Symposium, Oct. 28-30, Portland,
Oregon, pp. 62-73.
Chamberlin, T. W. 1980. Aquatic System Inventory (Biophysical Stream
Surveys). APD Technical Paper 1, B.C. Ministry of Environment,
Victoria, British Columbia, 33 pp.
Cole, Gene F. 1972. Valley Types, an Extension of the Landsystems
Inventory to Valleys. Unpublished report on file at the Boise
National Forest, Boise, Idaho.
Collotzi, Albert W. 1974. A Systematic Approach to the Stratification
of the Valley Bottom and the Relationship to Land Use Planning.
Unpublished report on file at the Bridger -Teton National Forest,
Jackson, Wyoming.
Dixon, Gary E. 1981. Multiresource Inventories: Meeting Challenging
New Information Needs. In: In-Place Resource Inventories:
Principles and Practices--, Society of American Foresters National
Workshop, Univ. of Maine, Orono, Maine, Aug. 9-14, 1981, pp. 383-388.
Frayer, W. E., L. s. Davis, and Paul G. Risser. 1978. Uses of Land
Classification. Journal of Forestry, val. 76, no. 10, pp. 647-649.
Hamilton, Thomas E. 1978. National Integrated Inventories -Is What You
Need What You Do? In: Integrated Inventories of Renewable Natural
Resources: Proceedings of the Workshop, Tucson, Arizona, Jan. 8-12,
1978. USDA Forest Service Rocky Mountain Forest and Range Experiment
Station General Technical Report RM-55, Fort Collins, Colorado, pp.
136-139.
16-20
Harris, Arland S., o. Keith Hutchison, William R. Meehan, Douglas N.
Swanston, Austin E. Helmers, John C. Hendee, and Thomas M. Collins.
1974. The Forest Ecosystem of Southeast Alaska: The Setting. USDA
Forest Service Pacific Northwest Forest and Range Experiment Station
General Technical Report PNW-12, Portland, Oregon, 40 pp.
Lotspeich, Frederick B., and Williams. Platts. 1981. An Integrated
Land -Aquatic Classification. In: Acquisition and Utilization of
Aquatic Habitat Inventory Information: Proceedings of the Symposium,
Oct. 28-30, Port 1 and, Oregon, pp. 103-108.
Nelson, DeVon, Grant A. Harris, and Thomas E. Hamilton. 1978.
Resource Classification-Who Cares? Journal of Forestry,
no. 10, pp. 644-646.
Land and
vo 1 . 7 6,
Ott Water Engineers, Inc.
Service Region X[lO].
Alaska, 7 pp.
1979. Water Resources Atlas for USDA Forest
USDA Forest Service Alaska Region, Juneau,
Paustian, Steven J., Daniel A. Marion, And Daniel F. Kelliher. In
preparation. Stream Classification using Large Scale Aerial
Photography for Southeast Alaska Watershed ~lanagement. In:
Proceedings of the RNRF Symposium on the Application of Remote
Sensing to Resource Management, May 22-27, 1983, Seattle, Washington.
Paustian, Steven J., Douglas Perkinson, Daniel A. Marion, and Philip
Hunsicker. In preparation. An Aquatic Value Rating Procedure for
Fisheries and Water Resource ~lanagement in Southeast Alaska. In:
Managing Water Resources for Alaska's Development: Proceedings-of
the ~Jeeting, Nov. 10-11, 1983, Chena Hot Springs, Alaska.
USDA Forest Service Alaska Region. 1979. Tongass
Final Environmental Impact Statement, Part 1.
Alaska Region, Juneau, Alaska, 312 pp.
Land Management Plan:
USUA Forest Service
1980. The ALP 1981-86 Timber Sale Operating Plan:
Final Environmental Impact Statement for the Chatham and Stikine
Areas, Part 1. USDA Forest Service Alaska Region Report No. 100,
Sitka, Alaska, 311 pp.
1982. Integrated Resource Inventory:
I I. Unpub 1 i shed report on file at Tong ass National
Area Supervisor's Office, Sitka, Alaska.
Legends Handbook
Forest-Chatham
1983a. Planning Update for the ALP 1986-90 Operating
Plan. USDA Forest Service Tongass National Forest -Chatham and
-Stikine Areas, Sitka, Alaska, 25pp.
1983b. Aquatic Portion of the Integrated Resource
Inventory: Draft. Unpubl ishea report on file at Tongass National
Forest-Chatham Area Supervisor's Office, Sitka, Alaska.
USDA Forest Service Intermountain Region. 1979. Guidelines for Water
16-21
Resources Inventory. USDA Forest Service Intermountain Region,
Ogden, Utah, 9 pp.
Viereck, Leslie. A., and C. T. Dyrness. 1980. A Preliminary
Classification System for Vegetation of Alaska. USDA Forest Service
Pacific Northwest Forest and Range Experiment Station General
Technical Report PNW-106, Portland, Oregon, 38 pp.
Wertz, W. A., and J. A. Arnold. 1973. Land Systems Inventory. USDA
Forest Service Intermountain Region, Ogden, Utah, 12 pp.
16-22
AN AQUATIC VALUE RATING PROCEDURE FOR FISHERIES AND WATER RESOURCE
MANAGEMENT IN SOUTHEAST ALASKA
By Steven J. Paustianl, Douglas Perkinson2,Daniel A. Marion3,
and Philip Hunsicker4
Abstract
The Aquatic Value Rating (AVR) is a tool for evaluating forest management
alternatives on the Tongass National Forest-Chatham Area (TNF-CA). The
AVR is a qualitative rating method designed to assess stream ecosystem
suitability for streamside timber harvesting activities. Stream
ecosystems that have significantly different characteristics are
identified using the TNF-CA channel type classification system. Channel
types are readily discernible stream segments with relatively homogeneous
physical, biological, and resource management properties.
An AVR is developed for each channel type from fish habitat, channel
stability, and morphology data. The fish habitat component is derived
from empirical relationships between measured habitat features and fish
utilization estimates. These values are used as a relative index of
habitat quality. The channel stability and morphology component is
determined from quantitative and qualitative estimates of channel
conditions. It is assumed that the channel stability parameters are
useful indices for assessing stream ecosystem sensitivity to management
activities. The channel type AVR procedure provides resource managers
and planners with an efficient, flexible, rational, and defensible
methodology for planning management activities affecting fisheries and
water resources.
Introduction
The Aquatic Value Rating (AVR) is an outgrowth of the Tongass National
Forest-Chatham Area (TNF-CA) Integrated Resource Inventory effort. The
AVR approach was developed in response to a need for better assimilation
of fisheries and water resource concerns and opportunities into the
lsupervisory Hydrologist, Tongass National Forest-Chatham Area, P. 0.
Box 1980, Sitka, AK 99835.
2District Fisheries Biologist, Hoonah Ranger District, P. 0. Box 135,
Hoonah, AK 99829. 3Hydrologist, Tongass National Forest-Chatham Area, P. 0. Box 1980,
Sitka, AK 99835. 4Fish Culturist, Alaska Department of Fish and Game, P. o. Box 499,
Sitka, AK 99835.
17-1
Forest Service land management planning process. The concepts and
applications discussed in this paper have been used successfully in
evaluating resource management alternatives for the proposed Alaska
Lumber and Pulp Co. 1986-90 Operating Plan (USDA Forest Service Alaska
Region, l983b).
The AVR is an initial attempt to depict stream or watershed fish
production capability, and to account for relative fish habitat
sensitivity to timber harvest induced impacts. Important attributes of
the AVR approach include:
l. The AVR is based on distinct stream channel mapping units called
channel types, that allow for the efficient compilation,
verification, and extrapolation of resource inventory and research
data.
2. The AVR incorporates biotic and abiotic factors that are known to
influence fish habitat quality and sensitivity to management induced
perturbations.
3. The AVR model provides a foundation for a more precise quantitative
approach to assess fish production and management induced habitat
changes and recovery trends.
Aquatic Value Rating.
The AVR approach is a fish habitat Classification scheme with sufficient
precision to address site specific stream and fish habitat management
issues. It also has sufficient generality to expand data applications to
large watershed or land use planning units. The AVR incorporates
quantifiable aquatic and riparian habitat features that correlate with
fish habitat quality and sensitivity as evidenced by fish production and
17-2
observations of timber harvest responses.
The TNF-CA AVR has three fundamental components: Channel Type mapping
units (CT); a Habitat Quality Index (HQI); and a Riparian Sensitivity
Index (RSI). Channel type measurements provide the spatial framework for
quantification of fish habitat characteristics. The habitat quality
index is a relative measure of fish production potential. The riparian
sensitivity index is a qualitative measure of stream dependence on
streamside characteristics for regulating inchannel structure and
functional processes. These three AVR components are combined
algebraically to form the following expression:
CT x HQI x RSI = AVR
The product of these three parameters (AVR) is a dimensionless,
qualitative value rating of aquatic resource potential. The following
discussion will examine each AVR component in detail and present some
examples for AVR applications in forest land use planning problems.
Channel Type Mapping Units. Channel types are discernable stream
channel mapping units having relatively homogeneous morphological,
biological, and resource management characteristics (Paustian et al., in
preparation). Channel types are delineated on aerial photographs using
differentiating criteria that include: adjacent landforms, channel
incision depth, channel gradient, riparian vegetation, and contributing
drainage area. This channel type classification provides a framework
within which fish habitat and channel stability features can be measured
and quantified in a systematic manner. Limited field samples of
representative channel type segments can be efficiently compiled and
17-3
extrapolated to unsurveyed areas using channel types. A limited fish
habitat quality and sensitivity data base can thus be extended over
relatively large geographic land use planning areas with some degree of
confidence.
Habitat Quality Index. Numerous fish habitat studies have focused on
geologic and morphologic characteristics of watersheds, but few have had
land and water management value (Platts, 1974). These habitat
classifications typically correlate fish habitat, stream channel, and
watershed characteristics with wome measure or index of fish distribution
or production potential. Dunham and Collotzi's (1975) transect survey
used stream channel morphologic variables to rate channel reaches by a
percentage of "optimum" habitat. Heller, Maxwell, and Parsons (1983)
developed a relative index for watershed hab~tat quality using fluvial
geomorphic parameters. Platts (1974, 1979) examined geomorphic and
aquatic variables as to how they relate to fish distribution and
abundance, but restricted his classification to large landform units.
Four basic objectives: realism, generality, efficiency, and precision were
set for the TNF-CA Habitat Quality Index (HQI). The goal of realism demands
that the HQI should portray stream habitat values in a way that allows for
easy comparison with other resource values. A relative measure of fish
production potential was chosen that with further refinement could be used to
put a dollar value on the fish habitat resource. The HQI also had to be
applicable to a wide range of geomorphic and fluvial conditions, and account
for the varied life phases of most of the important fish species found in SE
Alaska. The actual inventory methodology for collecting HQI data needed to be
efficient and reliable in measuring habitat variables amenable to modeling
17-4
fish production. Finally, index variables had to be precise enough to have
information value for land and water management decisions.
Barber et al. (1981) examined two inventory methods in use by fisheries
biologists in Southeast Alaska. The two methods were a subjective
estimate of quality (transect method), and an objective measurement of
quantity (area method)(USOA Forest Service Alaska Region, 1981). Barber
et al. evaluated the statistical abilities of the two methods to predict
fish populations residing in the inventoried stream sections.
Examination of their multivariate predictive equations suggested that the
area method could be adapted to meet our criteria for a habitat quality
index.
The area method requires objective measurements of stream features and
construction of diagramatic maps of 30m segments of active stream
channels. Prior to our field work, 300m long representative stream
reaches were selected on 1:15840 aerial photographs. Within the
representative reach 30m samples segments were selected and mapped.
Samples were either single or matched pairs stratified by channel type.
Seventeen habitat variables: gradient, available spawning area, water
types, and three cover classes among others, were measured and mapped to
scale.
Calculations from the Barber et al. equations resulted in a population
estimate for a 30m sample station. For data comparison purposes we
reduced that value to an estimate for a meter long channel segment. As
an index of habitat quality for stream rearing anadromous species we
chose the authors' equations for coho (Onchorhynchus kisutch) age 0 and
17-5
1+. The index for stream rearing resident species used the Dolly Varden
(Salvelinus malma) equation. Because the majority of fish production
comes from pink (Onchorhynchus gorbuscha) and chum (Onchorhynchus keta)
salmon that rear in saltwater, we felt it necessary to develop and use an
index to rate habitat quality for saltwater rearing species. Standard
Alaska Region (USDA Forest Service Alaska Region, 1981) values for
production (harvestable adults produced per square meter of spawning
habitat) of pink and chum were averaged (l.5l/m2 [pink] + l.l2/m2
[chum] I 2 = l.3l/m2 ) and multiplied by the amount of spawning habitat
measured in the surveys. This better accounts for situations in which
one or the other species dominates watershed production. Table l
illustrates the multivariate regression equations used in the indices,
and their respective correlation coefficients (Rand R2 ). Also
included is the index for saltwater rearing species.
The final form of the three HQI indices were:
Freshwater Rearing HQI = 0.033(log-1coho D + log-1coho l+)
Saltwater Rearing HQI = 0.033)pinklchum)
Resident HQI = 0.033(log-1Dolly Varden)
The mean HQI scores and standard error of the estimates are listed for
selected channel types in Table 2. Scores are the mean of HQI's for all
sample sites stratified by channel type.
The HQI is a realistic index of fish habitat quality. After sampling
nearly 2DO major watersheds, participants in this inventory are confident
that it tracks observed habitat use. Generality is maintained due to
17-6
Table 1. Barber et al. Multivariate Regression Equations used in the
Tongass National Forest -Chatham Area Habitat Quality Index.
Fish R R2 Predictive Eguationl
coho 0 0.87 0.76 logY = 0.871 + 1.011 logASA
+0.010RV -0.009S (1)
coho 1+ 0.70 0.49 logY = 0.249 -0.073G + 0.416logSS
+ 0.006RV + 0.260logUB (1)
Dolly Varden 0.70 0.49 logY = 3.223 -l.llOlogiA +
0.344logFDR (l)
pink/chum y = (1.36/m2)ASA/1000 (2)
SOURCES: (l)Barber et al., 1981, p. 19; (2)USDA Forest Service
Alaska Region production estimate used in cost/benefit analyses,
multiplied by available spawning area of the 30m sample. Production
equals mean of pink (1.51 adults/m2) and chum (1.12 adults/m2).
lEquation abbreviations: ASA =available spawning area; RV =riparian
vegetation; S = season; G = gradient; SS = shallow slow water type; UB =
undercut banks; IA = intensive area; FOR = forest debris riffle.
use of explicit and consistent channel type and inventory standards over
broad geographic areas. This confirms the wisdom of stratifying
watersheds into channel types based on aquatic and riparian structure and
function. However, confidence limits frequently overlap between two or
more channel type HQI's. This is probably attributable to variability in
habitat characteristics and inventory sample design. Narrower confidence
bands are anticipated after stratification of the HQI data within land
units (TNF-CA physiographic sections) having relatively homogeneous
climatic, lithologic, and geomorphic characteristics. Channel type
delineations are not based solely on fish habitat considerations but also
account for timber harvesting and road design concerns. A certain degree
of precision is sacrificed by this multifunctional stream mapping
approach. It should be noted that at present the HQI rates fish habitat
capability in relative terms and not in numbers of fish. We have applied
17-7
Table 2. Habitat Quality Index Mean Scores and Standard Error of the
Estimates for Selected Tongass National Forest -Chatham Area
Channel Types.
Channel Sample Freshwater Saltwater
Type Size Rearing Rearing Resident
E (Low gradient, 23 5.12 (1.32) 13.96 0.02 (0.005)
esturarine channel
ClA (Low gradient, 54 2.14* (0.47) 5.03 0.08 (0.02)
valley bottom,
forest channel)
ClB (Low gradient, 35 1.56 (0.36) 5.36 0.16 (0.13)
incised valley bot-
tom, forest/muskeg
channel)
BlA (Low gradient, 23 1.58* (0.25) 3.03 0.80 (0.26)
lowland, forest
cllannel)
82 (Moderate grad-43 0.36 (0.09) 0.97 0.77 (0.56)
ient, middle val-
ley, forest channel)
84 (Moderate grad-26 0.24 (0.04) 0.84 0.27 (0.10)
ient, incised low-
land, muskeg'for-
est channel)
NOTE: Values are mean Habitat Quality Index per meter length of
channel type. Standard error estimated by taking square root of standard
error of the mean for a 30m sample.
*Freshwater rearing scores were adjusted upward by 50% to account for
large amounts of off-channel flood plain rearing sites typically found in
association with these channel types.
the Barber et al. equations in a broader spectrum of stream channels than
the orginal population samples used to develop the multivariate
equations, thus reducing statistical accuracy. In addition, fish
production rates are generally lower in northern panhandle streams versus
streams in the southern panhandle due to lower water temperatures.
17-8
Furthermore, important habitat influences such as upwelling ground water,
lakes, and off-channel flood plain rearing areas were not taken into
account in the orginal validation. Therefore, in our judgement there is
presently insufficient validation of the Barber et al. equations on the
TNF-CA to permit their use for precise estimates of fish populations.
However, we feel strongly that use of these equations as a relative index
of habitat quality is justifiable.
Riparian Zone Sensitivity
A major problem facing aquatic resource managers is not only defining a
means of assessing aquatic habitat quality. In addition, they must make
predictions about how activities such as timber harvesting will affect or
impact aquatic habitat. In the AVR approach we apply the term riparian
zone sensitivity to describe the probability for timber harvesting
induced impacts to aquatic habitats.
The interactions between fluvial processes, stream channel morphology,
and riparian factors directly shape aquatic habitat conditions (Swanson
et al., 1982, 1983; Trisca el al., 1983). A riparian or stream
sensitivity index must adequately describe the balance between these
biotic and abiotic factors as they affect the stability and quality of
aquatic habitats. Most attempts at such a classification are necessarily
qualitative. The Stream Reach Stability Classification by Pfankuch
(1975) is a widely used approach to predicting stream sensitivity.
Platts, Megahan, and Minshall (1983) iaentify several important
parameters including: stream bank soil stability, stream bank vegetation
stability, stream bank undercut, stream bank slope, channel elevation,
channel gradient, channel sinuosity, substate size, substrate
17-9
embeddedness, channel debris and sediment storage, vegetation cover, that
have proven useful in evaluating stream channel and riparian zone
conditions. We have incorporated many of these qualitative ano
quantitative channel assessment procedures (USDA Forest Service Alaska
Region 1983) into developing the riparian sensitivity index of the AVR.
Various aquatic ecosystems functions of riparian zone vegetation that may
be influenced by timber harvesting activities are displayed in Figure 1.
For the AVR model these riparian zone functions have been combined into
three categories referred to as: sediment, LOD, and energy factors.
Each of these three factors are combined to form the riparian zone
sensitivity index (RSI). The RSI is a qualitative means of assessing the
positive or negative influences of streamside timber harvesting on fish
habitat in TNF-CA.
In developing this approach we recognized the futility of attempting to
quantify all the many complex interrelationships between riparian factors
and instream processes. These include fluv~al and biotic processes such
as: sediment input and routing; stream temperature regimes; the
processing and flushing of organic detritus; photosynthesis and nutrient
uptake; and stream bank cutting, channel scouring, or sediment
deposition. Instead, two major assumptions are made in the determining
riparian sensitivity factors:
1. Riparian functional relationships can be adequately described for
TNF-CA streams using three simple functional categories (sediment,
LOD, and energy) and that these factors reflect stream dependence on
riparian vegetation for long term aquatic environment stability.
2. The relative importance of each of these functional components is
17-10
AQUATIC ECOSYSTEM FUNCTIONS
OF RIPARIAN VEGETATION
-CONTROLS STREAM -ROUTES FLOW AND SEDIMENT -SHADES STREAM
CHANNEL EROSION
SEDIMENT
FACTOR
• • • • •
-SHAPES FISH HABITAT -PROVIDES
-PROVIDES BIOLOGICAL SUBSTRATE
LOD
FACTOR
• • • •
DITRITUS INPUTS
ENERGY
FACTOR
• •
•
• •
RIPARIAN ZONE SENSITIVITY INDEX
Figure l. The relationships between qualitative sensitivity factors
(Seo1ment, LOU, Energy) used to derive the Tongass National Forest -
Chatham Area Riparian Zone Sensitivity Inaex, ana the functional roles of
riparian vegetation that influence aquatic ecosystem stability ana
responses to timber harvesting activity.
17-11
equal and is therefore given equal weight in qualitative assessments
of riparian zone sensitivity.
Sediment Factor. The sediment factor in the riparian zone
sensitivity index reflects the role of riparian vegetation in controlling
the process of stream channel erosion (Figure 1). Timber harvesting can
exacerbate natural channel erosion through the breakdown and destruction
of stream banks by falling and yarding operations, and through loss of
root binding strength due to decay of conifer root mats after cutting
(Chamberlin, 1982). A number of important fish habitat features can be
affected by such alterations to stream channel erosion processes (Reiser
and Bjornn, 1979). For example, the quality of spawning gravel can be
altered through shifts in substrate particle size distribution or reduced
intergravel permeability. Mechanical disturbance to the stream channel
can reduce available cover associated with undercut banks and shallow
pools. Filling of substrate voids with fine sediments eroded from
channel banks can eliminate important habitat for bentic organizms.
Sediment sensitivity factors are determined from a combination of
sediment input and sediment conveyance ratings for TNF-CA channel types.
Sediment input potential is determined from channel stability ratings
for: upper bank stability, lower bank stability, and streambed
stability. Sediment conveyance is based upon bankfull unit stream power
which is used as a relative predictor of stream sediment load transport
ability (USDA Forest Service Alaska Region, l983a). Bull (1979) has
demonstrated that channel sensitivity to sediment induced impacts can be
expressed in terms of the balance between sediment input and conveyance.
A qualitative comparison between sediment input and conveyance potentials
17-12
Table 3.
Riparian Zone Sensitivity Ratings for Selected Tongass National Forest -
Chatham Area Channel Types.
Channel
Type
BlA (Low gradient,
lowland, forest
channel)
B2 (Moderate grad-
ient, middle valley,
forest channel)
B4 (Moderate grad-
ient, incised lowland
muskeg/forest channel)
ClA (Low gradient,
valley bottom,
forest channel)
ClB (Low gradient,
incised valley bot-
tom, forest channel)
E (Low gradient,
estuarine channel)
Sediment
Factor
Mod
Low
Low
High
Low
Low
LOD
Factor
Mod
High
Low
High
Low
Low
Energy
Factor
Mod
High
Low
Mod
Low
Low
Riparian
Sensitivity
Index
1.98
2.31
0.99
:<' .64
0.99
0.99
NOTE: Sediment, LOD, and energy ratings: Low = 0.33; Moderate(Mod)
= 0.66; and High = 0.99.
was therefore used to derive the overall sensitivity ranking for the
sediment factor in column one of Table 3.
Rating classes of high, moderate, and low were determined for the
sediment input and sediment conveyance potentials. If the sediment input
class is the same rank or lower than the conveyance class (e.g., low
17-13
versus moderate), than the sediment sensitivity is rated low (see Table
3). In this situation it is assumed that increases in sediment due to
timber harvesting activities will not significantly exceed the ability of
the stream to cope with the new sediment load. However, if the sediment
input class is greater than the conveyance class rating, increased
sediment loads will likely change channel conditions. In the latter
situation sediment sensitivity is moderate if input exceeds conveyance by
one class (e.g., high versus moderate), and high if input exceeds
conveyance by two classes (e.g., high versus low).
This sediment rating approach addresses only potential onsite impacts
from streamside timber harvesting. The effects of hillslope erosion and
sediment transported from upstream channels are not accounted for by the
sediment factor of the riparian zone sensitivity index.
LOD Factor. LOD input potential is the second major factor used in
rating riparian zone sensitivity. LOD is one of the more important links
between terrestrial and aquatic components of northern coniferous forest
streams (Swanson et al., 1982). The primary functions of LOD in stream
ecosystems are: controlling the routing of streamflow and sediment;
shaping fish habitat; and providing a substrate for biological activity
(Figure 1). These functions are highly time dependent in nature ana are
a result of balances between input and output that occurs over decaaes
(Swanson et al., 1982)
Logging can shift this equilibrium between the rate of processing or
breakdown of LOD and the rate of LOD input causing changes in stream
habitat and fish populations (Everest and Meehan 1981). The cumulative
17-14
effects of LOO inputs from natural or management related causes may be
both positive and negative. Positive aspects of LOD are that it
contributes cover and enhances habitat diversity for stream biota. Once
encorporated into streambed and banks LDD improves channel stability.
Organic debris dams slow the movement of sediment and fine particulate
organic matter and thus contribute to the availability of allocthonous
food sources and spawning gravels. However, large scale debris inputs
such as those resulting from hillslope debris torrents or major blowdown
certainly have negative short term impacts on fish habitat. These large
scale debris jams commonly impede fish migration and are closely
associated with large sediment inputs and channel migration. Heavy
accumulations of needles and branches also associated with massive debris
inputs may completely clog small streams and may significantly reduce
intergravel dissolved uxygen concentrations in larger channels (Everest
and Meehan, 1981; Chamberlin, 1982).
Although judicious inputs of organic debris and slash during logging
activities will likely improve fish habitat on some stream reaches, the
long term effects of loyging on the stability of LOO related habitat is a
concern. Rapid liquidation of old growth conifer stands along extensive
reaches of channel prevents an even flow of LOD inputs to the channel.
Intensively managed second growth stands will be denser and more wind
firm resulting in fewer natural debris input events. The smaller stems
and rootwads of second growth trees will also be smaller providing less
fish cover, and will be less stable features in larger channels.
The LOD sensitivity factor (Table 3, column 2) for TNF-CA channel types
was determined from field observations of LOD loading in stream
17-15
channels. LOD includes all woody debris greater than lOcm in diameter
(Swanson et al., 1982). The class ranks of high, moderate, and low
sensitivity corresponds to the relative channel area typically affected
by LOD accumulations: less than 10%, 10% to 20%, and greater than 20%,
respectively. The area or zone of LOD influence was defined by the
percentage of sample stream segment within which LOD controlled the
development or maintenance of habitat features such as bank and bebris
cover, scour pools, and gravel bars.
In assessing LOD influences on riparian zone sensitivity it was assumed
that those stream habitats typically having the highest natural debris
loading are also the most dependent on LOD for maintenance of habitat
integrity. Therefore, in Table 3 ClA and 82 channel types are assumed to
have a high LOD sensitivity because frequent organic debris accumulations
are consistently associated with them and the removal of streamside
conifers can limit the long term supply of LOD. It should also be
recognized that LOD input rates are controlled by a large number of
environmental factors including: riparian canopy successional stages,
local wind patterns; and stream channel dynamics. Therefore, a high
degree of natural spatial variability in LOO accumulations can be
expected to occur within a given channel type segment.
Energy Factor. The energy factor is the last major component of the
riparian sensitivity index. The functions of riparian vegetation in
stream shading and fine organic detritus input are accounted for by this
parameter (Figure 1). The quantity and quality of stream detrital food
sources are major factors in controlling the distribution of various
functional groups of aquatic invertebrates (Cummins, 1974). The
17-16
diversity and abundance of these aquatic invertebrates in turn directly
aff'ects the productivity of anadromous and resident fisheries (Swanson et
al., 1982). The effects of clearcutting on stream energy budgets are
twofold. First, the rate of fine particulate organic matter inputs to
the stream are altered. Second, the rate of aquatic photosynthetic
activity can increase significantly following riparian canopy removal
(Chamberlin, 1982). This potential increase in autotrophic food supplies
may counteract the loss of the stable allocthonous fooo supply provided
by the undisturbed riparian canopy. However, this may be only a
temporary enhancement of stream productivity, and may result in long term
reductions in overall biological production due to intense shading by
second growth stands (Trisca et al., 1982).
Riparian canopies also regulate stream temperature, a second major factor
influencing the growth and survival of aquatic biota. The effects of
canopy removal in elevating summer stream temperatures can be substantial
in forest streams (Brown and Krygier, 1967). However, the moderating
influence of riparian forest canopies on winter stream temperature could
potentially be a more important management consideration in TNF-CA
(Sheridan and Bloom, 1975).
Management implications relating to maintenance of a stable stream energy
regime are somewhat ill-defined in TNF-CA. Canopy openings in some low
gradient floodplain channels (ClA, BlA) will likely enhance summertime
aquatic productivity by elevating stream temperatures and increasing
photosynthesis. However, these benefits may be totally negated by
increased winter mortality of incubating eggs or rearing juveniles with
the removal of temperature moderation effects by the canopy. Data are
17-17
not presently available to indicate the long term responses of the
various functional groups of aquatic invertebrates to manipulation of
riparian vegetation. Therefore, potential impacts to anadromous
fisheries from riparian clearcutting is not clear (Swanson et al.,
1982). However, a general riparian management criteria for the long term
maintenance of a stable and productive stream energy base would suggest
that manageable portions of important channel types should have a variety
of canopy successional stages associated with them.
Energy factor ratings in Table 3 are based upon the size, composition,
and structure of riparian canopy types generally associated with a given
channel type (USDA Forest Service Alaska Region, l983a). The high,
moderate, and low energy sensitivity ratings correspond to the following
riparian canopy closure classes determined from aerial photographs:
greater than 50%, 25% to 50%, and less than 25%. This rating criteria
assumes that stream ecosystems associated with dense riparian canopies
must rely on sources of fine particulate organic matter for a significant
portion of their aquatic food base. Aquatic production in these streams
may also be sensitive to changes in water temperature resulting from
canopy removal. An example is the 82 channel type in Table 3. Streams
typically having a greater range in canopy closure (e.g., ClA and BlA
channel types) are rated as moderately sensitive to energy factors
because a better natural balance between primary production and detrital
processing would tend to buffer the effects of canopy removal.
Temperature sensitivity may still be a concern for channels in this
category. Streams with sparse or no forest canopies recieve a low energy
sensitivity rating.
17-18
The summation of ratings for the sediment, LOO, and energy factors yields
the overall riparian sensitivity index (RSI) in Table 3. The RSI
represents a qualitative measure of channel type response to timber
harvesting. It should be reiterated that this numerical index is
conceptually sound but the rating classes were not derived from
quantitative analyses. This approach is undoubtebly an
oversimplification of the complex interactions between biotic and abiotic
factors of riparian and aquatic ecosystems. However, the RSI is useful
in comparing the relative sensitivity of distinct channels to timber
harvesting. Hopefully with this preliminary foundation the RSI will
evolve into a more precise predictor of aquatic ecosystem response to
timber harvesting.
Assumptions/limitations
l.
2.
3.
4.
The correlations between habitat and fish utilization that
Barber et al. (1981) found are generally applicable
throughout the TNF-CA, and it is only the magnituoe of the
population estimate that varies due to local conditions.
Pink and chum habitat quality can be adequately measured and
modeled using available spawning area.
Sediment, LOD, and ener·gy factors are the riparian habitat
characteristics most likely to be impacted.
In the development of riparian sensitivity ratings the use of
standard timber harvesting practices are assumed. These are
clearcutting of all merchantable streamside timber within
defined unit boundaries using high-lead cable yarding with
partial log suspension.
17-19
5. In its present form, channel typing and the A.R.V. model applies only
to the geomorphic zone thus far incorporated into the inventory.
AVR Applications
The AVR approach was developed specifically to facilitate forest resource
planning efforts and management decisions pertaining to fisheries and
watershed management on the TNF-CA. The AVR information base can be
applied to several levels or types of Forest resource planning. Examples
for two levels of forest land management planning (general allocative
resource planning and area specific project planning) are presented in
the following discussion.
Allocative planning in the Forest Service multiple resource management
scheme determines management emphasis for large management units. This
allocation may shift management objectives in favor of intrinsic
wilderness values on one extreme or intensive management of commodity
resources such as timber products at the other end of the spectrum. At
this point, the AVR does not display fisheries habitat as a dollar
commodity but can be useful in displaying and comparing the relative
value of fisheries resource between two or more management units. Two
watershed planning units, "Finger Creek" and "Bent Prop Creek," on
Chicagof Island are shown in Figure 2. They have many similarities
including drainage area, elevation, drainage pattern, total channel
length and length of useable anadromous fish habitat. However, the
watershed aquatic value rating for "Bent Prop" Creek is 2.5 times greater
than the total for "Finger Creek'' (see Table 4). This difference is due
primarily to the occurrence of a long segment of potentially very high
quality ClA habitat on "Bent Prop Creek." This example illustrates that
17-20
~
J \·..... FINGER
·... CREEK ~\, ············-... ) \ ~~--...... ,. ' ...... .,
············ "" .. / B2 ............... ... ............_
BASIN AREA 19kJ
··~~MEAN ELEVATION 290m
/) TOTAL CHANNEL
........................... / 1 LENGTH 30km
134 ../ \
\ ANAOROMOUS CHANNEL
C L\ ....... ······-..... ... ··· l LENGTH O.Skm .. ·············\ . ... ·· \ ....
.. ·· ...... -.. ... .
·~ ..... """" .... ._._;;_ ...... "" ................
Scale: [ 1 1 ]
0 .5 1.0 1.5 (km)
BASIN AREA 15kal
MEAN: EI...EVATlON _ t8Clzn
TOTAL CHANNEL
LENGTH 2!JkJD
ANADROMOUS CHANNEL
lENGTH 7 .7km
Scale: !-[ ----.-..-. --J]
0 .5 1.0 1.5Ckm)
Fi9ure 2. Distribution of channel types havino anadromous fish habitat in two typical Tonoass National
Forest -Chatham Area 3rd order watersheds.
(\j
I
['-
fish habitat management opportunities between two apparently similar
watersheds can be differentiated in a simple, straight forward manner.
The AVR approach has not been applied in allocative planning efforts to
date. It is hoped that in future allocation planning a version of the
AVR will aid in developing improved land management allocations
compatable with fisheries, water, timber and recreation management goals.
The AVR can also be used to assist project planning efforts by
identifying site specific management issues, concerns, and opportunities
that will result from various resource development alternatives. The
AVR-channel type information base allows for consistent, reproducible and
efficient assessment of the potential effects of individual management
alternatives on fish habitat and stream stability. These assets are
particularly important for scheduling multiple entries for timber
harvesting where the long term impacts of management activities are a
concern. An examination of a proposed project timber harvest plan for
"Finger Creek" illustrates the utility of AVR data in this type of
application (Figure 3). In this timber harvesting and road construction
alternative (treatment l and 2), four cutting units are proposed
immediately adjacent to the upper valley 82 channel type, impacting about
60% of the riparian zone. Although this channel type is rated as having
only moderate fish habitat quality, it has a high riparian sensitivity
(Tables 2 and 3). Fisheries management and water quality objectives for
this watershed may require that special stream management prescriptions
be considered, or another cutting unit alternative be implemented. These
prescriptions may include a number of options: deferring harvest of a
17-22
Scale: [
0
I
.5
I
1.0
Fll"\'GER
CREEK
TOTAL CHANNEL
LENGTH 30km
ANADROMOUS CHANNEL
LENGTH 6.5km
LEGEND
Nonanadromous
Anadromous ]
1.5(km) Road Location
\\htcrshcd Boundary
0 Bridge Location
1m Tn:ut mcnt 2
EJ Treatment 1
Figure 3. Example of a timber harvest ana road construction planning
alternative for a Tongass National Forest -Chatham Area
watershed.
17-23
unit until the second entry (treatment 2, Figure 3); partial retention of
a narrow strip of riparian timber; or a selective cutting treatment
Table 4. Aquatic Value Ratings for Two Watersheds on Chichagof Island,
Southeast Alaska.
Channel Type
Habitat Quality
Index
Riparian Sen-
sitivity
Index
Aquatic Value
Rating
"Bent Prop Creek"
ClA E ClB BlA
25,89:<: 16.304 11,050 3,776
2.64 0.99 0.99 1.98
68,354 16' 141 lO ,939 7,476
watershed Aquatic Value Rating = 103,989
"Finger Creek"
Channel Type ClA 82 84
Habitat Quality 9,877 4,662 1,831
Index
Riparian Sen-2.64 2.31 0.99
sitivity Index
Aquatic Value 26,075 lG, 769 1,813
Rating
Watershed Aquatic Value Rating = 38,657
17-24
84
1,090
0.99
1,079
rather than clearcutting. Through proper distribution of cutting units
and timing of entries habitat maintenance and possibly enhancement of
fish productivity can be achieved. Canopy openings can be spaced out
along high value stream segments with some short term retention of old
growth riparian zones. This approach to managing riparian vegetation
would provide a stable detrital energy base. It would also provide a
stable source for LOD inputs.
AVR information can also be useful to in indicating the need for impact
mitigation measures by identifying potential problem areas. In Figure 3,
potential conflicts exist between habitat protection objectives and
timber harvesting and road construction plans associated with the ClA
floodplain channel segment near the mouth of "Finger Creek." The nature
of the proposed flood plain bridge crossing (circled) is such that
stabilization of bridge abuttments and channel banks and special
provisions for passing flood flows will require careful onsite
investigation by hydrology and fisheries resource specialists prior to
project implementation. The clearcut unit (treatment 2, Figure 3)
adjacent to this ClA channel segment may also pose a hazard to fish
habitat and stream channel integrity. Practices to help mitigate
potential timber harvesting impacts might also be considered here.
The final outgrowth of the AVR is that it improves the efficiency of
future resource inventory, monitoring, and research efforts. The AVR
allows for prioritization of these studies to specific streams based upon
the relative project importance to fish production and the potential
project sensitivity to natural or management induced perturbations.
17-25
Opportunities
We have already realized the utility of channel typing and the AVR model
as tools appropriate for large scale land and water management planning.
Channel type field data were extrapolated to unsurveyed areas with a
reasonable degree of confidence, and facilitated a cost efficient
interdisciplinary decision in the Alaska Lumber and Pulp 1986-1990
Operating Plan Draft Environmental Impact Statement (USDA Forest Service
Alaska Region, l983b).
Future developments and data base management goals we have identified are:
1. Further data stratification for identified geomorphic zones.
2. Area-specific validation of the Barber et al. regression equations.
3. Development of regression equations specific to channel types.
4. Replace the RSI with quantifiable values.
5. Refine the static AVR with dynamic model components that track land
management activities through time.
6. Develop area method validation procedures to account for the
influence of the following habitat features: lakes, springs, beaver
ponds, upwelling groundwater, water quality, and the presence of
unmappable first order floodplain streams.
7. Examine the potential for incorporating species and life-phase
habitat preferences and seasonal use for all fish species encountered.
8. Encourage the application of the indices, AVR, and data to guide
site-specific research, inventory, and project implementation.
Summary.
Numerous management and resource planning applications for the Aquatic
17-26
Value Rating exist. The Aquatic Value Rating is a communication tool
that succinctly describes fisheries habitat conditions in terms of a
habitat quality index; sensitivity to streamside timber harvesting in
terms of a riparian sensitivity index; and habitat distribution in terms
of channel type mapping units. As an approach to presenting aquatic
resource information, the Aquatic Value Rating allows for greater
flexibility in applications requiring varying levels of planning. This
permits relative value comparisons between watersheds, watershed
subunits, or individual stream channel segments.
The Aquatic Value Rating rates the potential for stream and riparian zone
multiple resource management conflicts and opportunities. It aids in
developing management plans that optimize prescription, mitigation, and
enhancement strategies. It can also aid in optimization of aquatic
resource inventory, monitoring, and research efforts by focusing
attention on critical habitat areas.
Acknowledgments
The authors would like to acknowledge the contributions of numerous
individuals of the TNF-CA Integrated Resource Inventory Team who
participated in the extensive data collection and analysis effort
required to implement the Aquatic Value Rating procedure. Graphics for
this paper were done by Cheryl Moody.
References
Barber, W.E., M.W. Oswood, and S. Oeschermeir. 1981. Validation of Two
Habitat Fish Stream Survey Techniques: The Area and Transect Methods.
Unpublished Final Report to USDA Forest Service, Contract 53-0109-0-00054,
Juneau, Alaska, 64 pp.
Bull, W. B. 1979. Threshold of Critical Power in Streams. Geol. Soc. of
17-27
Amer. Bull. Part I, val. 90, pp. 453-464.
Chamberlin, T. W. 1982. Influence of Forest and Rangeland Management on
Anadromous Fish Habitat in Western North America: Timber Harvest. USDA
Forest Service, General Technical Report PNW-136, 30 pp.
Cummins, D. W. 1974. Structure and Function in Stream Ecosystems. Bio
Science val. 24, pp. 631-641.
Dunham, D. K., and A. Collotzi. 1975. The Transect Method of Stream Habitat
Inventory: Guidelines and Applications. USDA Forest Service
Intermountain Region, Ogden, Utah, 98 pp.
Everest, Fred H., William R. Meehan. 1981. Forest Management and Anadromous
Fish Habitat Productivity. In: Transactions of the 46th North American
Wildlife and Natural Resources Conference, Wildlife Management Institute
Publication, Washington D.C., pp. 521-530.
Heller, D. A., J. R. Maxwell, and M. Parsons. 1983. Modeling the Effects of
Forest Management on Salmonid Habitat. USDA Forest Service, Siuslaw
National Forest, 38 pp.
Paustian, Steven J., Daniel A. Marion, and Daniel F. Kelliher. In:
preparation. Stream Classification using Large Scale Aerial-photography
for Southeast Alaska Watershed Management. In: Proceedings of the RNRF
Symposium of the Application of Remote Sensing to Resource Management, May
22-27, 1983, Seattle, Washington.
Pfankuch, D. J. 1975. Stream Reach Inventory and Channel Stability
Evaluation: A Watershed Management Procedure of the U.S. Forest Service,
USDA Forest Service Northern Region, 26 pp.
Platts, W. S. 1974. Geomorphic and Aquatic Conditions Influencing Salmonids
and Stream Classification with Application to Ecosystem Management. USDA
Forest Service, SEAM Program, Billings, Montana, 199 pp.
1979. Relationships Among Stream Order,
and Aquatic Geomorphology in an Idaho River Drainage.
No. 2, pp. 5-9.
Fish Populations,
Fsiheries, val. 4,
Platts, W. S., Walter F. Megahan, G. Wayne Minshall. 1983. Methods for
Evaluating Stream, Riparian, and Biotic Conditions. USDA Forest Service,
General Technical Report INT-138, 70 pp.
Reiser, D W., and T. c. Bjornn. 1979. Habitat Requirements of Anadromous
Salmonids. USDA Forest Service General Technical Report PNW-96.
Sheridan, w. L., and A.M. Bloom. 1975. Effects of Canopy
Temperatures of Some Small Streams in Southeast Alaska.
Service Alaska Region, 13 pp.
Removal on
USDA Forest
Swanson, F. J., S. V. Gregory, J. R. Seaell, and A. G. Campbell. 1982. Land
-Water Interaction: The Riparian Zone. US/IBP Synthesis Series 14,
17-28
Hutchinson Ross Publishers, pp. 267-291.
Trisca, F. J., J. R. Sedell, and S. V. Gregory. 1982. Coniferous Forest
Streams. US/IBP Synthesis Series 14, Hutchison Ross Publishers, pp.
292-331.
USDA Forest Service Alaska Region. 1981. Fisheries Survey Handbook -FSH
26D9.23. Region 10 Amendment No. l, Jan. 1981.
l983a. Aquatic Portion of the Integrated Resource
Inventory: Draft. Unpublished report on file at Tongass National Forest
-Chatham Area Supervisor's Office, Sitka, Alaska.
l983b. Alaska Lumber and Pulp 1986-90 Operating Plan Draft
Environmental Impact Statement.
17-29
HATER QUALITY PROTECTION PROGRAM FOR AGRICULTURE IN ALASKA
Bruce W Rummel and William C Leitch 1
Abstract
A water quality protection program for agriculture contains four major
elements: 1) technical information, including a set of agricultural best
management practices, 2) a set of institutional mechanisms to deliver the
technical information and implement the program, 3) agriculturalists,
including farmers and ranchers, and 4) an assessment activity to monitor,
verify, and evaluate program operations. In Alaska, technical information
is available in published literature and is delivered to farmers by the
Alaska Soil Conservation District. Institutional mechanisms to implement
the program are being established by the Alaska Department of Environmental
Conservation. ADEC is developing a cooperative relationship with ASCD to
provide for mutual assistance in meeting the goal of water quality
protection with agricultural development. In addition, ADEC is seeking
participation of other resource and regulatory agencies to promote an
effective and efficient program and where advantageous, to coordinate
activities.
Introduction
Water quality protection for agricultural development in Alaska has evolved
in two steps: 1) preparation of a background report and recommendations and
2) implementation. In addition to being a program designed in the
preventive, rather than remedial mode, this program exemplifies a smooth and
effective transition from a consultant's report to an institutional program.
BACKGROUND REPORT
Summary
1) Water Quality Protection Program. A water quality protection program
should maintain Alaska Water Quality Standards (18AAC70) without having to
1 Principal,
Ecologist,
Juneau AK
Great !Vater Associates,
Alaska Department of
99811.
Box A-475H, Anchorage Al( 99507, and
Environmental Conservation, Pouch 0,
18-1
enforce those standards directly. The program should respond to identified
needs and should be flexible. The program recommended in the background
report has four basic elements related to agricultural activities: 1)
recommended practices: agricultural best management practices, 2) means of
disseminating and implementing those practices: delivery and implementation,
3) agriculturalists who use those practices: farmers and ranchers, and 4)
means of verifying and evaluating applied practices: program verification
and assessment (see Figure 1).
2) History of Agriculture in Alaska. Agriculture in Alaska probably began
in 1784 with the introduction of livestock on Kodiak Island. Throughout
most of its history, agriculture has been a subsistence activity, sometimes
undertaken to support development of other natural resources or commercial
activities. In this century, commercial agriculture in Alaska developed
first in the Tanana Valley around Fairbanks, then around Palmer, and most
recently in the area near Delta Junction as a result of State agricultural
disposals.
3) Location of Future Agricultural Development. Based on agricultural
potential, land ownership, and proposed development schedules, future agri-
cultural development will most likely occur in areas with existing agricul-
tural activities, on Pt. MacKenzie, or in the Nenana or Yukon Flats areas.
4) Water Quality Effects. The water quality effects of agricultural
development in Alaska are similar to those of other activities that cause
large-scale changes in vegetation and land use, e.g. timber harvest and
18-2
CP
I
\.N
FIGURE 1. WATER QUALITY PROTECTION PROGRAM FOR GENERAL AGRICULTURAL ACTIVITIES
,/_7 RESOURCE AGENCY ASSESSMENT 1
~7
ALASKA DEPARTMENT OF ALASKA
ENVIRONMENTAL CONSERVATION DEPARTMENT OF
SOIL ENVIRONMENTAL .. CONSERVATION -Paatlclde Regulallomll CONSERVATION
w u SERVICE -Animal Waota Dlopaaal Polley ffi j: ~ ill -Enlorcement of Alnka ~ u ~ Warar Quality Standards RANCHERS :1! ..
If ALASKA "' DEPARTMENT OF "' DIVISION OF w
'"
ENVIRONMENTAL nl w nl .. AGAICUL TURE
CONSERVATION :1! "' w ~ ..
:1! .. .. ... .. w l!l AGRICULTURAL FARM !!l ~ " COoPERATIVE z .. CONSERVATION CONSERVATION ... z EXTENSION PROGRAM PLAN .. 0 .. SERVICE .. ~ ALASKA :1! => ... nl u DEPARTMENT OF ... ... e; FISH AND GAME .. =>
ill u ..
ALASKA SOIL a: w > ... CONSERVATION " .. DISTRICT .. :1! .. .. => I~ .. ... ~ FARMERS " OTHER RESOURCE ... => > ALASKA 0 AGENCIES
~ AGRICULTURAL ALASKA SOIL .. .. DEPARTMENT OF ..
EXPERIMENT ~ CONSERVATION
STATION ENVIRONMENTAL DISTRICT AND .. ~ :::; CONSERVATION w SUBDISmiCTS a ASSISTANCE LOCAL GOVERNMENT
) )J L] ~
PARTICIPATION IN CONSERVATION DISTRICT
Source: Rummel 1982
residential development; in fact, only a few water quality effects are
expected to be unique to agriculture. Most of the effects occur as a
result of soil erosion: the transport, processing, and relocation of soil
particles. Agricultural chemicals and wastes may also alter natural water
quality.
5) Alaska Environmental Information. Very generally, the environmental
factors that contribute to high agricultural potential, e.g. adequate soil
depth and drainage, low slope, high number of degree days, tend to promote
processes beneficial to protecting water quality, e. g. infiltration, low
soil erosion, high rate of pesticide weathering.
6) Agricultural Best Management Practices. Agricultural best management
practices (BMPs) consist of three kinds of practices: l) practices designed
to protect water quality directly and inherently; 2) practices that are
components of general agricultural operations but are conducted in ways to
minimize adverse water quality effects; and 3) agricultural practices that
combine l) and 2). BMPs that have been adapted to Alaskan conditions are
available from the Soil Conservation Service of the U.S. Department of
Agriculture and from the Cooperative Extension Service of the University of
Alaska and USDA. B~Ws can be incorporated into field activities on
cropland, pasture and hay land, and rangeland, and can provide guidance for
proper animal waste disposal (see Table l for examples).
7) Institutional Mechanisms. Existing institutional mechanisms for water
quality protection form a patchwork of individual programs aimed at specific
agricultural activities. Nonetheless, existing statutory and regulatory
18-4
CP
I
\n
Table 1. Examples of Agricultural Best Management Practices (BMPs)
Practice
Code
Number
Name of
Practice
(Units) Definition
t. BMP's for Field Activities
A. CROPLAND
56D
31D
324
328
Source:
ACCESS ROAD
(Ft.)
BEDDING
(Ac.)
CHISELING
AND
SUBSOlllNG
CONSERVATION
CROPPING
SVSTEM
(Ac.)
Rummel 1982
A travelway constructed as
part of a conservation plan.
Plowing, blading, or otherwise
elevating the surface of flat
land into a series of broad,
low ridges separated by
shallow, parallel dead
furrows.
Loosening the soil, without
inverting and with a minimum
of mixing of the surface soil,
to shatter restrictive layers
below normal plow depth that
inhibit water movement or root
development.
Growing crops in combination
with needed cultural and
management measures. Cropping
systems include the use of
rotations that contain grasses
and legumes, as well as
rotations in which the desired
benefits are achieved without
the use of such crops.
'General Agricultural Purpose
Provides a route for travel
for moving equipment and
supplies and provides access
for proper operation and
maintenance of conservation
enterprises.
Provides improved surface
drainage at relatively low
cost by establishing adjoining
parallel beds or lands
running in the direction of
available natural slope. This
purpose is accomplished by
moving soil toward centers of
beds to form a series of
ridges and dead furrows
(troughs) which will
accomplish one or more of the
following: minimize water
pondage, provide gradients for
moving runoff, and permit
efficient operation of tillage
and harvesting equipment, or
eliminate sources for mosquito
production.
Improves water and root
penetration and aeration
and breaks up subsurface
compaction to improve internal
soil drainage.
Improves or maintains good
physical condition of the
soil; protects the soil during
periods when erosion usually
occurs; helps control weeds,
insects, and diseases; and
meets the needs and desires of
the farmer for an economic
return.
Potential
Primary Water
Qua 1 ity Effects
erosional
effects
erosional
effects
erosional
effects
Water Quality
Protection Measures
Stabilize cut and fill
slopes and install cross
drains for road ditches;
treat surface if required
to limit erosion.
Limit slope (combination of
natural land slope and cross
slope); assure adequate root
zone.
Avoid inversion and minimize
mixing; chisel along contour
on lands subject to water
erosion.
This practice generally
protects water quality (see
Table 7 for examples of
cropping system alterna-
tives).
authorities are probably sufficient to implement a comprehensive water
quality protection program. Institutional mechanisms for soil
conservation are extremely well developed, as are state regulations
governing activities in the agricultural land disposal program. Pesticide
application is governed by explicit regulations. Less well developed are
institutional mechanisms for management of fertilizer use or animal wastes.
8) Monitoring Activities. Four types of programs for field collection of
water quality data can be identified: 1) programs for collecting baseline
and follow-up data, 2) ·experimental programs, 3) programs for monitoring
compliance, and 4) programs for monitoring trends. Baseline and follow-up
studies seek to measure pre-and post-project conditions to assess changes
in water quality. Experimental studies seek to assess water quality effects
of agricultural practices in controlled conditions for eventual broader
application. Compliance monitoring checks for maintenance of Alaska water
quality standards, and trend monitoring assesses long-term, large-scale
changes in water quality conditions.
9) Environmental Assessment. The program should produce environmental
benefits.
Recommendations
To institute a water quality protection program for agriculture, the Alaska
Department of Environmental Conservation should:
18-6
o Adopt the program structure described in the background report
o Establish a working relationship with the Alaska Soil Conservation
District and subdistricts
o Formulate a policy on animal waste disposal
o Participate in the land use planning process of the Alaska Depart-
ment of Natural Resources at all three planning levels: 1) state-
wide, 2) area, and 3) management
o Establish an interagency review group for program verification and
assessment
o Establish a field program to extend existing baseline studies and
to incorporate field data acquisition in the disposal process for
agricultural lands
o Encourage the Agricultural Experiment Station, the Institute of
Water Resources, and other research institutions to study the
water quality effects of agricultural practices (Rummel 1982).
IMPLEMENTATION
The current program, also called the Agricultural Phase of the State Water
Quality Management Plan, has been developed from the information and recom-
18-7
mendations of the background report. In particular, a number of milestone
accomplishments can be identified: 1) Memorandum of Understanding between
ADEC and the soil conservation district, 2) Checklist of Agricultural Best
Management Practices, 3) Senate Bill 120, 4) 205(j) proposal.
Memorandum of Understanding
After a series of preliminary meetings between staff of the Alaska
Department of Environmental Conservation and the Alaska Soil Conservation
District, the two organizations decided to draft a Memorandum of
Understanding (MOU) that would formalize their relationship and establish a
written basis for cooperation on activities related to water quality and
agriculture. The MOU provided for increased contact between staff of the
two organizations, provided for a policy and procedures document that
spelled out how ADEC field staff would cooperate with subdistricts in
resolving water quality problems, and authorized ADEC to grant $10,000 to
ASCD to assist that organization in undertaking activities related to water
quality concerns associated with agricultural operations. In addition, the
MOU established the use of "Checklist of Agricultural Best Management
Practices" in reviewing farm conservation plans submitted during
applications for agricultural leases and required that subdistricts provide
DEC with summary reviews of the checklists.
The MOU, signed in April 1983, set two important precedents with respect to
the farming community, ADEC, and water quality problems:
18-8
l) it established a formal cooperative relationship between ADEC--a
regulatory agency--and an organization representing the Alaskan
farming community, and
2) it cast ADEC field staff into the role of cooperators rather than
enforcers with respect to farmers and water quality problems.
Checklist
The Checklist of Agricultural Best Management Practices (ADEC 1983) was
developed for three purposes:
l) to bring pertinent agricultural BMPs to the attention of officers
and cooperators of the Alaska Soil Conservation District and
Subdistricts,
2) to facilitate the efforts of Subdistricts to include water quality
considerations in their reviews of farm conservation plans, and
3) to assist the Department of Environmental Conservation to identify
potential water quality problems associated with agricultural
development.
The BMPs described in the checklist identify specific water quality
conservation practices, as well as methods for conducting general
agricultural operations so as to minimize adverse water quality effects (see
Table 2 for example). To keep the checklist brief, the BMPs were
18-9
Table 2. Example of Checklist
CHECKLIST OF AGRICULTURAL BEST MANAGEMENT PRACTICE
Agricultural Activity
A. CROPLAND
Access Road(34)
Bedding{34)
Chiseling and
Subsoiling(34)
Conservation
Cropping
System( 34)
Cover and Green
Manure Crop {35)
Critical Area
Planting {35)
Source: ADEC 1983
Yes
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
No
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
18-10
N/A
[]
[]
[]
[]
[J
[]
[]
[]
[]
[]
[]
Water Quality Protection Measures
Stabilize s 1 opes
Install cross drains
Treat surface, if required
Limit slope
Preserve adequate root zone
Avoid inversion of soil
Minimize mixing
Chisel along topographic contour
This practice protects water quality
This practice protects water quality
This practice protects water quality
summarized; detailed descriptions of each BMP are found in the background
report and in publications of the Soil Conservation Service and the
Cooperative Extension Service. The BMPs in these publications are
specifically adapted for application in Alaska.
Like the MOU, the Checklist, distributed to Subdistricts in August 1983,
also set important precedents:
1) in effect, the checklist puts a significant share of
responsibility for protection of water quality into the hands of
the farming community itself,
2) designed for repeated use, and referenced by page to the original
report, the checklist serves as an ongoing educational tool for
farmers.
Senate Bill 120
This act of the Alaskan Legistature, "An Act relating to soil and water
conservation," (committee substitute for Senate Bill 120) modifies the
structure and purposes of the state soil conservation district. Effective
July 1, 1983:
1) the number of Governor-appointed members of the District Board is
increased from three to five, providing for broader geographic
representation,
18-11
2) the organization's name is changed from Alaska Soil Conservation
District to Alaska Soil and Water Conservation District, a subtle,
but important change, and
3) the jurisdiction of the District is increased so as to include
land areas rather than farming areas, and users of land rather
than occupiers of land.
These changes broaden the conservation district program in Alaska.
205 (j) Proposal
In June 1983, staff of four organizations--(Department of Environmental
Conservation, U.S. Environmental Protection Agency, Alaska Soil and Water
Conservation District, and Soil Conservation Service)--met to discuss
Idaho's water quality program related to agriculture, a program widely
acknowledged as the model of such projects. Their program is distinguished
by careful identification of priority problem areas and maintenance of close
contact with the Idaho Legislature in order to secure implementation funds
from the state for studies completed with funds made available through
Section 208 of the Clean Water Act.
It was clear that many of the successful features of the Idaho program could
be applied in Alaska, although on a much reduced scale. As a result, staff
of ASWCD and SCS applied for federal funding to identify priority problem
areas in the state and carry out programs to ensure that adverse water
18-12
quality effects associated with agriculture would be prevented or minimized
in those areasa
The final result of the June 1983 meeting was a 2-year proposal to complete
both a comprehensive drainage and erosion control plan for the Delta
Project, and a dairy waste management plan for the Point MacKenzie Project.
The plan, submitted in August 1983, was drafted by the ASWCD, and will be a
cooperative effort with the SCS. The proposal is still under study, but is
likely to be accepted and funded.
References
Rummel, B.W.
Prepared
82pp+vi,
1982. Agricultural
for Alaska Department
2 Appendixes.
Practices and Water Quality Effects.
of Environmental Conservation, Juneau.
Alaska Department of Environmental Conservation 1983. Checklist of
Agricultural Best Management Practices. MS 12pp.
18-13