HomeMy WebLinkAboutAPA90TK.
1425
.S8
A23
no.90
EIDC 707 A STREET. ANCHORAGE. ALASKA.99501; (907)27114523
NETHODOLOGICAL APPROACH TO QUANTITATIVE
IMPACT ASSESSMENT FOR THE PROPOSED
SUSITNA HYDROELECTRIC PROJECT
RCTIC ENVIRONMENTAL INFORMATION AND DATA CENTER
Report Prepared by:
Arctic F.nvironmental Information
and Data Center
University of Alaska
707 A Street
Anchorage, Alaska 99501
Submitted to:
Harza/Ebasco Susitna Joint Venture
8740 Hartzell Road
Anchorage, Alaska 99507
For:
The Alaska Power Authority
334 W. Fifth Avenue, Second Floor
Anchorage, Alaska 99501
METHODOLOGICAL APPROACH TO QUANTITATIVE
IMPACT ASSESSMENT FOR THE PROPOSED
SUSITNA HYDROELECTRIC PROJECT
March 12, 1983
ARLIS
Alaska Resources
Library & Information SerVices
AnchoraJ;!e. Ala~k::1
Arctfc F.nvironmental Information and Data Center
Aauatic Assessment Staff
Hilliam J. Wilson, 'Principal Investigator
Charles G. "Mike" Prewitt, Ph.D., Fishery Biologist
Michael D. Kelly, Fishery Biologist
Ken A. Voos, Ph.D., Hydrologist
Paul R. Meyer, Hydrologist
Joseph C. LaBelle, Geomorphologist
James L. Wise, Climatologist
TABLE OF CONTENTS
AEIDC SUSITNA AQUATIC IMPACT ASSESSMENT APPROACH
Introduction • • • • • • • • • • • • • • •
Susitna Aquatic Studies Program • • • •
Study Program Members and Tasks ••
Aquatic Studies Current and Future • •
The Susitna River Basin Habitat Simulation Model.
The Susitna Aquatic System Simulation Model (SUSIM) •••
The Linked Model System • • • •
Reservoir Operation Model • • • • • • • • •
General Description • • • • • • • • •
Uses in Susitna Aquatic Assessments •
Available Models and Data Requirements
Water Balance Model
General Description • • • • • • • • • •
Uses in Susitna Aquatic Simulation
Available Models and Data Requirements
Reservoir Temperature Model • • • • • •
General Description • • • • • • • • • •
Uses in Susitna Aquatic Simulation
Available Models and Data Requirements
Instream Temperature Model • • • • • • •
General Description • • • • • • • • • • •
Uses in Susitna Aquatic Simulation
Available Models and Data Requirements
Mainstem Temperature
Slough Temperature
Hydraulic Simulation Models
General Description • • • •
Uses in Susitna Aquatic Simulation
Available Models and Data Requirements
The Unlinked Models -Ice and Sediment • • • • •
Ice
The Need for Ice Simulation Modeling in the
Susitna Assessment Process
Description of Present and Future
Ice Processes •••••••
Background • • • • • • •
Freezeup •
Breakup. • • •
Effects of Ice Processes on the
Stability of Side Channels and Sloughs
Freezeup . . . . . . . . . .
Breakup. • • • • • . • • •
Potential Aquatic Impact Issues •
Tnstream • • • • • • • •
Staging-Overtopping • •
Breakup • • • • • •
1
1
1
2
2
6
10
10
10
11
12
12
12
13
13
13
13
14
14
15
15
16
16
16
17
18
18
18
19
21
21
21
21
21
22
23
23
23
24
25
25
25
26
The
TA~LE OF CONTENTS (Continued)
Tributary Mouth Morphologic Changes.
In-Reservoir • • • • • • • • •
Tee Models -Capabilities and Cost
Channel Morphology and Sediment Issues •
General Description • • • • • • • • •
Application in Susitna Aquatic Assessment •
Streambed Elevation Changes ••••••
Channel Configuration-Substrate Effects.
Available Models and Data Requirements
Suspended Sediment Only • • • • •
Aggradation/Degradation Studies
Susitna Instream Impact Assessment Approach ••
SUSIM Applications. • • • • •
Sensitivity Testing ••••••••••
Ice and Sediment Studies • • • • • • • • • • •
Linked Model Runs and Impact Assessments
The Susitna Instream Fishery Impact Assessment Approach:
Coniunctive Use of the Simulation Models •
Quantitative Assessment Tools • • • • • •
Hydrologic Methodologies . • • • •
Hydraulic Simulation Methodologies •••••
Hydraulic Simulation Methodologies with
Habitat Preferences • • •
Habitat Index Evaluations • •
Miscellaneous Approaches
Assessment Applic8.tion.
Mainstem-Side Channel • • •
Hydrologic Methodologies
Hydraulic Simulation Approaches.
Hydraulic Simulation Plus Species
Preference Approach • • • • • • • • • • •
Habitat Evaluation Indices •
Miscellaneous Methodologies.
Summary. • • • . • . • •
Tributaries • • • • • • • • • . •
Hydrologic Methodologies • • •
Rydraulic Simulation Methodologies
Habitat Evaluation Procedures.
Miscellaneous Approaches
Summary. • • • •
S laughs . . . . . . . . . . . . . .
Fixed-Percentage Methodologies • • •
Hydraulic Simulation Approaches
Hydraulic Simulation with Habitat
Preferences • • • • • • • • • • • • . • •
Habitat Evaluation Indices • •
Miscellaneous Approaches -
Proposed Slough Impact Assessment Designs •
Habitat Dynamics Displays and Comparisons ••
Population Trend Models •.•••••••••
26
27
28
29
29
32
32
33
34
34
34
36
36
36
38
38
39
39
39
40
40
40
41
41
42
42
43
43
44
44
44
44
45
45
45
46
46
46
47
47
48
48
48
49
52
TABLE OF CONTENTS (Continued)
Other Construction-Related Effects
Introduction • • • • • • • . • • • • • • •
Potential Impact Areas and Related Aquatic Resources •
Access Corridors •••
Transmission Corridor •
Gravel Removal Sites • • • •
Habitation Areas. • • • •
Spills ••••••••••••••
Clearing. • • • • •
Sport Fishing • • •
General Impact Problems. • •••
Erosion and Sedimentation • •
Access Corridors and Habitation Areas
Transmission Line Corridor • • • • • •
Gravel Removal Sites •
Pollution • • • • • • • • • • • •
Water Use.
Spills • •
Clearing • •
Indirect Impacts.
Impact
Bibliography • •
Sport Fishing ••
Analysis Approach
58
58
58
58
59
59
61
61
61
61
62
62
62
63
64
65
65
65
66
66
66
67
69
INTRODUCTION
In accordance with guidelines from the Federal Energy Regulatory
Commission (FERC), licensing of the Susitna Hydroelectric Project
(Suhydro) must be based upon a specified environmental assessment
process. This assessment process must include (1) description of the
present aquatic resources, (2) assessment of the project impacts upon
those resources, and (3) mitigation of impacts where possible. Through
subcontract with Acres American (ACRES) Corporation and Harza-Ebasco
Susitna Joint Venture, the Arctic Environmental Information and Data
Center (AETDC) of the University of Alaska has been charged with
assessment of Suhydro aquatic impacts. This paper presents the approach
as of March 1983 toward quantitative impact assessment and gives a
detailed description of study plans, conceptual relationships, and the
identified computer models or model systems. Receipt of final data from
other study grouo members will allow for assessment completion at a
later date.
SUSITNA AOUATIC STUDIES PROGRAM
STUDY PROGRAM HEHBERS AND TASKS
The Susitna aquatic studies group consists of the following
members.
1. Alaska Department of Fish and Game (ADF&G) to provide baseline
information on the aquatic habitat and resources of the
Susitna River and its tributaries.
2. E. Woody Trihey, P.E., to supervise the field phase of the
aquatic habitat and instream flow study.
3. AEIDC to assess Suhydro aquatic impacts.
4. v!oodward-Clyde Consultants to prepare Exhibit E for the FERC
licensing application and to develop · the Suhydro aquatic
mitigation plan.
5. R&M Consultants to provide data collection and retrieval
services in hydrologv, meteorology, and related areas.
-1-
AC.RES, the major feasibility study contractor, also has directly
subcontracted for studies concerning ice and groundwater. All of these
contractors provide data for use in simulation models and quantitative
impact assessment. Harza-Ebasco will provide future engineering data.
AQUATIC. STUDIES CURRENT AND FUTlTRE
Exhibit E of the FERC license application and the ADF&G completion
reports summarize results of biologic and aquatic habitat studies
conducted since 1979 (ACRES 1982; ADF&G 1982, 1983a,b,c). The aquatic
studies program includes investigation of mainstem, tributary, and
slough habitats in all segments of the Susitna River. Project emphasis
to date has been on side-slough habitats associated with the mainstem
above the Chulitna-Talkeetna River confluences. Information has been
gathered on species distribution, slough accessibility, substrate
distribution, and extent of acceptable spawning and rearing conditions
at various discharges. Emphasis will shift to other river reaches and
habitats, primarily in the lower Susitna River. The objective is to
provide the basis for some quantitative impact assessment in all habitat
types in all reaches of the river system likely to be affected by
project operations.
THE SUStTNA RIVER BASIN HABITAT SIMULATION MODEL
As described in detail in the following section, AEIDC will assess
impacts in two major areas: (J) instream--emphasis on predicting stream
habitat change with respect to changing flow, temperature, sediment, and
ice and (2) peripheral--those construction-related impacts on aquatic
habitats not in the Susitna downstream from the impoundment(s) but which
may be impacted by some project activity. This division is obviously
somewhat synthetic but allows separation of emphasis \vhile providing
analysis of all areas of potential impact. The instream impact analysis
w:Ul rely on results of simulation models to predict changes in aquatic
physical habitat (accounting for project effects on ice formation,
sediment transport, and streamflow and temperature regimes; to interpret
these changes in terms of fishery impacts over long-term project
-?-
operations; and to suggest feasible changes in project operations to
minimize negative effects on the fishery.
Impact assessment results should be directly usable in the
mitigation planning process. Design of the final impact assessment
should correspond with the needs of the mitigation contractor to the
greatest extent possible. To meet these needs, AEIDC has developed the
Susitna Aquatic System Simulation Model (SUSIM). This model system
resulted from consideration of the special aquatic habitat relationships
in the Susitna River basin and the need to account for ice, sediment,
and temperature changes which will accompany construction, filling, and
operation of the dam complex. The proposed modeling system includes
data-model linkages necessary to generate and analyze effects of each
potential project operation (Figure 1).
Reservoir operation parameters are considered to be fixed within
economic feasibility limits, whereas the release schedule component
represents the avenue through which reservoir operations respond to
fishery or other streamflow requirements downriver. The reservoir
operations model serves to integrate operating parameters, inflow (from
the historical streamflow record), and desired fishery release flows
into a time-series of monthly (or other frequency) flows immediately
below the dam. An in-reservoir temperature model provides temperatures
associated with the flows. To route the reservoir release discharge
downstream, a water balance model utilizes basin streamflow data
(measured and synthesized) to account for tributary and groundwater
inflow and to more accurately predict discharge patterns at various
points of fishery interest. The basinwide streamflow and temperature
data base serves as the input to hydraulic simulation models that are
the basis for habitat analysis. (The specific interplays among physical
process models and habitat evaluation are described in the next
section.)
flescriptions of systemwide aquatic habitat effects will result from
interpretation of both long-and short-term habitat variations
associated with each operational schedule. As fishery information
accumulates, it should be possible to link postproject characteristics
of fish habitats with consequent population effects and determine
connnercial and sport fishery consequences of each proposed operating
-3-
Figure 1. General Susitna system modeling and analysis approach.
-----------~--Historical
Basin
Stream flows
-------------Tributary and
Mainstem Data
----data files
models
--------------Reservoir Operation
Parameters
Reservoir Operations
Model
------~--.---.--Reservoir Release
Flows and Temperatures
• Basin Water
Balance Model
.-------'----.-.-.-... Basin-wide Streamflow
and Temperature
Regimes
Hydraulic Simulation
Model(s)
Ice and Sediment Models
(Assumption Checking)
Physical Habitat
Simulation
-------~--------System Habitat
Effects
-------------· Release .,__ Schedule
In-reservoir
+--Temperature
Model
lnstream
~ Temperature
Model
-------------· Slough and .,__ Side Channel
Data
Flow vs. habitat
~ relationships
Mitigation and
~ Optimization
r.--
-
regime. The most important feature of the system model is the
capability to iteratively change proposed flow regimes through
incremental changes in the release schedule based on feedback of habitat
and population effects.
-5-
TRE SUSITNA AQUATIC SYSTEM SI~ruLATION MODEL (SUSIM)
Development of AEIDC' s SUSIM has progressed to the point that
certain specific models, computer programs, and data sources have been
identified, and functional relationships between models and certain
assessment needs have become clearer. This paper gives details about
models, programs, linkages, and outputs to more clearly define (1) the
function of the simulation system, (2) data sources and deficiencies,
and (3) the need for additional information.
The general assessment process involves two major steps. First,
the physical changes expected to result from the dams are predicted
throughout the reaches to be impacted. This process includes
simulations of streamflow, temperature, ice dynamics, and hydraulic
geometry for as many operating schedules as might be feasible. Next,
these physical conditions are interpreted in terms of fishery habitat or
population effects to be evaluated in the impact assessment process.
The distinguishing feature of the model system approach is in the utili-
zation of many process models normally involved in project design
(reservoir operation, water routing, stage prediction, etc.) to credibly
predict the environmental changes likely to result from dam
construction, filling, and operation. Use of these models in close
cooperation with project engineers provides (1) a view of project
effects at the level of resolution required to do biological assessment,
(2) the ability to generate a sufficient range of project operations to
bound most potential impacts, and (3) the basis for planning and
mitigation by iteration of desired fishery flows through the reservoir
operations model.
The models will be discussed as t,.;ro basic types--those in the
linked iterative subsystem and those which may be run independently to
provide interpretative support or to insure that other modeling assump-
tions have been met. For each model type introduced in the previous
section a specific model has been selected for incorporation in SUSIM.
Figures 2 and 3 show specific models and the sources of their computer
programs and input data. Model categories mav represent assemblages of
similarly functioning models.
-6-
Figure 2.
Model category
Reservoir
operations
Reservoir
temperature
Instream
temperature
Selected linked SUSHi component models, computer programs,
identified data sources, and requirements as of February 1983.
Source, computer
program or model
concept (computer
program name)
ACRES (Susitna
energy simulation
models, one and
two reservoir,
monthly and
weekly)
ACRES (DYRESM)
U.S. Fish and
Wildlife, Instream
Flow Group (SNTEMP)
-7-
Data source
ACRES
R&M
Harza-Ebasco
ACRFS
Harza-Ebasco
R&M
ACRES
U.S. Geological
Survey
ADF&G Aquatic.
Fabitat and
Instream Flow
( AHIF)
NOAA and AEIDC
AEIDC
Data to be provided
Initial runs, oper-
ation, logic, inflow,
Revised raw inflow
data
Revised energy demand
estimates, revised
operations logic,
refined inflow, revised
project design spec-
ifications.
Results of initial
DYRESM model runs
Revised DYRESM runs,
ice cover additions
Revised operation
schedules, inflow
and outflow estimates
Streamflow, temp-
erature simulations
and compilations
Streamflow and temp-
erature records
Recent streamflow and
temperature records
Slough and tributary
temperature records
Meteorologic records
Shading, tributarv
contribution estimates
Model category
Hydraulic
simulation
Habitat
Figure ?
Computer program
or model concept
(computer program
name)
U.S. Fish and
Wildlife Service
(IFG)
(IFG-4)
(WSP)
ADF&G (stage-
discharge model)
U.S. Army Corps
of Engineers
(HEC-2)
AEIDC, ADF&G
(flow-habitat
relationships
to be developed)
U.S. Fish and
Hildlife Service
( IFG, HABTAT).
-8-
(continued)
Data source
ADF&G (AHIF)
ADF&G (AHIF)
ADF&G ( AHIF)
Woody Trihey
R&M Consultants
ADF&G Adult
Anadromous (AA)
ADF&G Resident
Juvenile (RJ)
ARIF
ADF&G
AA, RJ, AHIF
Data to be provided
Slough cross-section
measurements,
discharge, depth,
velocity, substrate
Slough cross-section
measurements,
discharge, depth,
velocity, substrate
Main channel stage
vs. Gold Creek dis-
charge at several
locations
Nain channel cross-
sections, calibrated
model results at
more than 50 locations
Passage needs,
relative salmon
habitat utilization
Relative juvenile
salmon utilization of
zones in sloughs,
habitat relationships
for mainstem resident
species (burbot,
rainbow trout,
grayling)
Slough and tributary
hydraulic model
results
Habitat preference
curves for selected
assessment species
Figure 3.
Model category
Slough ground-
water and temp-
erature
Sediment-
Channel
Morphology
Ice
Selected unlinked SUSIM component models, computer programs,
identified data sources, and requirements as of February 1983.
Computer program Data source
or model concept
(computer program
name)
Tony Burgess, ACRES ACRES
Consultants (Susitna
slough groundwater
finite element
models)
Models not yet
se1ected
Thomas Lavender,
ACRES Consultants
ADF&G (AHIF)
U.S. Geological
Survey
ACRES
R&M Consultants
-9-
Data to be provided
Initial model results,
current model, revised
data input
Slough discharge
measurements
1982 bed and
suspended sediment
data for Chulitna,
Talkeetna, Susitna
Results of Lavender's
study
1982-83 ice
observation data
THE LINKED MODE~ SYSTEM
Computer linkage of the models in this
input/output processing and time efficiency when
system facilitate
analyzing several
iterations of operating schedules but are not necessary to complete an
assessment. The models in the linked system are conceptually related
because they function sequentially; that is, output from the first model
is directly usable as input to the second and so on. Results of models
outside the linked system require interpretation and are not directly
usable by other models unless they are refined or in some way altered.
Each model in the linked system will be discussed in the following
sequence: (1) general description of the model or type of model; (2)
proposed function of the model in the impact assessment process; and
(3) a brief description of the model operation, available computer
software, and data input requirements. The exact details of certain
highly complex models are referenced in basic texts, user-related
material, or program documentation. Also, the models described are
those selected as of Februarv 1983. Other or additional models or
computer programs may be selected as the project continues.
RESERVOIR OPERATION MODEL
General Description
Reservoir operation models have become important in hydroelectric
aquatic impact assessment because they provide the link between project
operation and both the reservoir and streamflow conditions upon which
the impact assessment is based. Properlv utilized, such models can
serve to RUide project design and evaluation of environmental and
mitigation planning.
Generally, hydroelectric project feasibility is initially
determined by use of models which account for available water supply,
storage characteristics of the reservoir, power generating capabilities
at various storage levels, and specified downstream demands such as
municipal water needs, irrigation, interstate water compacts, or
instream flow requirements. The computer model accounts for the complex
interactions among these factors to help delineate what size ( 5_n terms
-10-
of power production) hydroelectric project might be constructed within
the constraints of available inflow and required releases.
Current water allocations and demands in the Susitna River basin
are negligible (Dwight 1981); therefore, accounting for water rights,
interstate compacts, and municipal demand is unnecessary, greatly
simplifying the process of integrating fishery flow requirements into
the Susitna project design. The Susitna reservoir operation model,
then, is essentially an energy production simulation model which
predicts electrical power output relative to reservoir storage over a
32-year forecast period described as having the same streamflow pattern
as the recorded past. Power generation is convertible to streamflow
(monthly or weekly, and the streamflow simulations serve as the basis
for aquatic impact assessment.
The reservoir operation model can be used to check whether releases
which meet power needs also meet downstream demands for instream and
offstream water uses. AEIDC's assessment approach requires the
capability to iteratively change the reservoir operation to insure that
downstream demands are met relative to their priorities.
model allows for this.
The ACRES
Results of reservoir operation models usually include reservoir
elevation and head tables (useful in reservoir temperature and fishery
assessments), power production tables, and monthly release tables for a
period of record similar to that used in the inflow.
Uses in Susitna Aquatic Assessments
The tables of reservoir releases provide the basis for all stream-
flow assessments below the proposed dam. Thev depict stream discharge
both before (using the inflow record only) and after dam construction,
and they represent the monthly or weekly flow patterns expected during
postpro_iect time periods, thereby serving as the basis for sophisticated
fishery popuJ at ion assessments. Above a] 1, reservoir operation model
results represent a common ground upon which both project design
engineers and environmental analvsts may exercise planning flexibility
within a rigorous framework. Reservoir operation models have not
generally been widely developed or utilized as integral parts of aquatic
impact assessments; thus, many are not entirely compatihle in terms of
-11-
resolution, ease of operations, and flexibility. It has been
demonstrated, however, that cooperation between biologists and design
engineers can almost always lead to development of models whose features
improve not only environmental analytic capabilities but prolect design
flexibility as well.
Available Models and Data Requirements
AEIDC has reviewed the ACRES energy simulation models and found
them to be suitable for many parts of the aquatic impact assessment.
First, :i.nput and output time units may be either weekly or monthly, for
a 32-year forecast period.
assessment of maximum and
Weekly model results can greatly facilitate
minimum flow events and allow detailed
evaluation of releases during the critical salmon migration and spawning
period. Second, the downstream demand input (at this time used solely
for fishery flow requirements) allows changes in the required flow
regime and resulting ease in initial feasibility assessment. Further,
the reservoir rule curve and drawdown levels can be easily changed in
accordance with accepted changes in project design or load forecast.
Finally, the model provides a table of predicted monthly reservoir
elevations necessary in determining in-reservoir temperature patterns
and inundation areas. Data for the ACRES reservoir operation model are
currently available, and the computer soft,vare and associated user
material (in the form of personal instruction) have been provided.
WATER RALANCE MODEL
General Description
The water balance model is a functional component process which
accounts for gains or losses in streamflow throughout the river network.
In regions with complex water demand structures, the water balance model
must account for tributary and groundwater, evaporation, appropriated
water rights, consumptive use, and return flow iag. However, in the
Susitna svstem, only tributary, ground, and surface inflow are
considered significant, and water balancing becomes a process of simply
addin~ in water from those sources. The process is made less reliable
in the Susitna system bv the paucity of both surface and groundwater
data.
-12-
Uses in Susitna Aquatic Simulation
The water balance model provides streamflow patterns at any desired
point on the Susitna River. It accounts for all significant inflow
between the dam and the point of interest. Because this component adds
in tributary flow instead of simply incrementing Susitna discharge by a
constant downstream factor, more precise analysis of the effects of the
variation attributed to tributaries is possible.
Available Models and Data Requirements
AEIDC has computerized the mechanics of water balancing, and a
water balance component based upon incremental discharge from all
. significant tributaries is available. The tributary streamflow
simulation process, however, has necessarily been simplistic (1) because
of the short or nonexistent gage records for both the tributaries and
some mainstem Susitna sites and (2) because tributary discharge has been
simulated based on tributary watershed area alone. This has resulted
from a lack of precipitation and other meteorological data for each
tributary watershed. Availability of such data would allow a more
reliable estimate of monthly discharge patterns for ungaged tributaries.
The problem is not considered crucial at this time but will ultimately
reduce the confidence in predictions of postproject streamflows,
especially in the middle and lower Susitna basins.
RESERVOIR TEMPERATURE MODEL
General Description
Reservoir temperature models are used primarily to predict the
thermal characteristics of lakes, ponds, and reservoirs. Unlike other
components of the assessment system, reservoir temperature and the
stream temperature modeling capabilities described later arose from the
need to address environmental and water quality issues and are not
strictly necessary in the design and operation of hydroelectric
projects. Typically, reservoir temperature models utilize the physical
relationships between meteorologic conditions and lake or reservoir
dimensions (primarily surface area and depth) to estimate the lake's
heat transfer characteristics and yearly stratification pattern. Most
reservoir temperature models predict monthly patterns of the depths of
-13-
the thermal stratification layers ( epilimnion, metalimnion, or
thermocline and hypolimnion) and their respective temperatures.
Uses in Susitna Aquatic Simulation
Reservoir temperature models have two principal applications: (1)
to assess conditions within the impoundment and (2) to provide the
outlet water temperature which together with outlet discharge serve to
simulate initial postproject downstream conditions. Questions regarding
temperature within the Susitna impoundments have been limited to those
addressing littoral zone conditions and mean reservoir temperatures.
AEIDC believes it to be more important at this time to determine the
temperature patterns at the dam face to predict stream temperatures
immediately below the dam.
Available Models and Data Requirements
Prediction of Watana or Devil Canyon reservoir temperatures during
the open water season is currently possible using the ACRES DYRESM
(Figure 1) reservoir temperature model. DYRESM is a one-dimensional
dynamic model which predicts daily salinity and temperature variations
with depth in a reservoir. Though input data are required on a daily
basis, the model may use smaller time steps to more accurately model
meteorological influences and mixing processes among the thermal zones.
The model approximates a set of horjzontal reservoir layers of different
thicknesses. Rased on variations in input meteorology, jnflow, outflow,
or reservoir water surface elevatjon, daily changes in the vertical
location, volume, temperature, and salinity of the layers can be pre-
dicted.
Although the model operates on one-day time intervals, changes in
certain variables occur on a much shorter time scale and are calculated
in quarter-hour time steps at the beginning of each daily simulation.
In subdaily stmulation the effects of surface heating (heat budget),
epilimnetic mixing, and hypolimnetic thermal and saline diffusion
effects are calculated. Following this the effects of the daily inflow
and outflow are determined and a final daily stratification pattern
predicted.
-14-
Reservoir temperature simulations for winter conditions require
predictions of ice cover (extended thickness) which are not currently
available. When these predictions become available, it will be feasible
to estimate the temperature stratification patterns in either Watana or
Devil Canyon reservoir based on normal meteorologic conditions and
steady-state operating patterns for each month. Because of the high
cost of each simulation run, however, it will probably not be
cost-effective to run the DYRESM model in a linked fashion.
INSTREAM TEMPERATURE MODEL
General Description
InstreaM temperature simulations are important in any stream impact
assessment but particularly so when dealing with salmon species whose
life history patterns are so closely tied with temperature cues or cumu-
lative temperature effects. When possible, it has been very valuable to
predict temperature downstream from the reservoir at least to the point
at which project-induced temperature changes are no longer apparent.
Instream temperature models predict downstream temperatures using the
structure, hydrology, and meteorology of the stream network.
Instream temperature models which have been app1ied in aquatic
studies are usually steady state; that is, the conditions of the inde-
pendent variables are assumed not to change during the selected model
time period. For example, a mean predicted monthly temperature at a
certain point downstream from a reservoir would have been calculated on
the basis of single, mean monthly values for the climatologic and
hydrologic variables. If mean weekly values were input, temperatures
could not be predicted at locations greater than one week travel time
downstream since these locations would have been subjected to conditions
not represented by the weekly average. To achieve finer resolution it
would be necessary to measure travel times at such closely spaced
intervals as to be prohibitive within the cost-time framework of most
studies. Therefore, stream temperature models are useful to predict
mean monthly or perhaps mean weekly temperatures but not downstream
daily temperature patterns resulting from peaking power generation or
other pulsed inflow.
-15-
Uses in Susitna Aquatic Simulation
Clearly, instream temperature modeling capabilities are necessary
to determine main channel water temperatures downstream from the
reservoir. The predicted temperatures are necessary to assess
suitability for inmigration, spawning, rearing, and outmigration of all
fish species which utilize the main channel for these purposes. Also,
main channel temperature modeling is necessary to predict the time and
location at which water temperature reaches 0° C during winter, which is
in turn necessary to predict ice formation. A major concern exists
regarding the effects of main channel temperature upon inmigration and
passage through possible temperature barriers created at tributary
mouths by the gradient between altered Susitna temperatures and those
from unaltered tributaries.
The most consequential biological effects of water temperatures are
upon spawning, incubation, emergence, and rearing of salmon species.
Temperature is a primary spawning stimulus, and changes in stream
temperature also affect development rates of embryos and may change
mortality rates during incubation, emergence and outmigration. Most
Susi tna River spawning occurs either within tributaries, beyond the
influence of Susitna River temperatures, or in side sloughs where
temperatures appear to be remotely related to those in the adjacent
Susitna mainstem. Recent studies have indicated that slough water tem-
peratures are relatively stable and equal to the seasonal mean tempera-
ture of the Susitna River. It is not known precisely how and at which
rate Susitna temperatures influence those in sloughs, however, making it
especially difficult to predict monthly or weeklv slough temperatures
based upon similar data for the mainstem Susitna.
Available Models and Data Requirements
Mainstem Temperature. The Stream Network Temperature Simulation model
(SNTEMP), developed by the U.S. Fish & Wildlife Service (Theurer and
Voos 1983), has been selected for use in SUSIM. SNTF.MP was designed to
predict average daily and daily extreme water temperatures at selected
points within a river network. The model requires meteorologic,
hydrologic, and stream geometrv data to compute heat flux relationships
and to transport the heat content through the stream network.
-16-
The following SNTEMP features make it particularly applicable for
use in the Susitna system.
1. A temperature regression technique which allows use of
incomplete or noncontinuous input temperature data.
2. Solar shading by vegetation and topographic features (e.g.,
canyon walls) throughout the year.
3. A calibration technique which provides the ability to adjust
low-confidence input parameters to obtain minimum error when
matching observed vs. historical temperature predictions.
4. The ability to predict daily average, maximum, and minimum
water temperatures for periods ranging from as short as one
day to as long as one year (continuously variable in one-day
increments). Thus, short yet critical river reaches can be
modeled in daily detail, and the full length of the system is
simulated with longer averaging periods.
For the Susitna system, SNTEMP will be configured to simulate
monthly and weekly average temperatures at any location between the
Hatana dam site and Cook Inlet. Historical data at Gold Creek and
recent temperature data collected at various locations by ADF&G will be
available for validating and calibrating the model. The model utilizes
either historical mean weekly or monthly hydrology and meteorology or
hydrologic and meteorologic data from a specific year or period. In
this latter mode of simulation, historical variability is used as an
approximation of future conditions.
Slough Temperature. The study sloughs of the Susitna system will be
modeled as separate physical systems and perhaps necessarily included in
the unlinked model svstem. The thermal properties of these sloughs must
be defined by the ongoing slough/groundwater data collection efforts and
modeling efforts by ACRES, ADF&G, and R&M Consultants. These study
results will support selection or development· of the models and
techniques which will he used to determine whether or not slough
temperatures will varv significantly under project stream-temperature
regimes,
-17-
HYDRAULIC SIMULATION MODELS
General Description
Hydraulic simulation models have become increasingly valuable in
instream assessments. When changes in the discharge pattern of a river
are expected, the aquatic biologist needs to know the associated changes
in stream depth, velocity, and wetted width. Changes in these variables
as well as substrate and temperature result in changes in stream
physical habitat, which represent the most definable impact of altered
discharge regimes.
Applications of hydraulic simulation models in instream assessments
have increased in recent years, primarily due to interest in the field
of instream flow methodologies. Stalnaker and Arnette (1976) and Wesche
and Rechard (1980) provide good summaries of the instream flow methods
which utilize hydraulic simulation models.
The Instream Flow and Aquatic Systems Group (IFG) of the U.S. Fish
and Wildlife Service has made the most notable contribution to hydraulic
simulation in instream flow assessments. The IFG Instream Flow
Incremental Methodology (IFIM) (Bovee 1982) is a conceptual framework
for resolution of streamflow regulation and allocation problems. The
IFIM is largely based upon use of the PHABSIH (Physical HABitat
SIMulation) computer system (Bovee and Milhous 1978; Milhous, Wegner,
and Waddle 1981). Use of PHABSIM requires extensive hydraulic
simulation capabilities based upon measurements taken at selected
representative or critical fish habitat reaches. The PHABSIM computer
program system automates the process of coupling predicted depths,
velocities, and substrates with habitat preference curves for those
variables for a given fish species or life stage. The resulting value,
Weighted Usable Area (WUA), is a measure of the areal extent of physical
habitat for the species/life stage at the modeled discharge.
Uses in Susitna Aauatic Simulation
Though there are extensive hydraulic simulation field efforts, the
Susitna assessment process will probably not be based extensively upon
WUA as calculated using the PHABSIH system. The use of hydraulic
simulation at this time 'appears to be limited to prediction of stage
(water surface elevation) in the main Susitna for use in (1) determining
-18-
c1epth and '.J'etted area in an adjacent slough or side channel, (2)
prediction of slough wetted area and available substrate relative to
slough discharge, and (3) prediction of main channel depth at various
discharges. Presently, for sloughs 8A, 9, 11, 21, Rabidoux Creek and
Chum Channel (ADF&G 1q83c), hydraulic simulation for the above purposes
is possible.
Available Models and Data Requirements
Data for four documented hydraulic simulation models have been
collected for the Susitna assessment. These are HEC-2 (USCOE 1976),
IFG-4 (Main 1978), Water Surface Profile (WSP) (US Bureau of Reclamation
1968; Spence 1975) and a stage regression model utilizing available
linear regression techniques. These models are of two types--backwater
models and empirical models.
Backwater or energy balance models utilize measured dimensions at
multiple cross sections (width, cross-section geometry, water surface
elevation at a known stream discharge) to predict stage and other
hydraulic characteristics at unmeasured flows. The process usually
requires extensive use of large computer models because of the detail of
the input data and the iterative calculation process employed. Of the
available Susitna models, WSP and HEC-2 are of this type. Backwater
models require measurements taken at only one discharge but gain
considerably in reliability if calibration information from several
discharges is gathered.
Empirical hydraulic models are based on measurements of target
variables (stage or velocity) at various discharges and development of
predictive discharge-velocity or discharge-stage regression models.
This modeling approach requires that measurements be taken at a minimum
of three different discharges. For further detailed discussions of
either backwater or empirical models, the reader is referred to Chow
(1965) and Bovee and Milhous (1978). Of the current Susitna assessment
models, IFG-4 and the stage regression model are of this type
(Figure 4).
Both the lJSP and IFG-4 models have been modified to couple, through
a program called HABTAT, with habitat preference curves to produce WUA
values. Because of a lack of specific habitat preference curves for the
Fi~ure 4. Hydraulic simulation models available for Susitna aquatic
assessment with specific estimates of predictive accuracy.
Hydraulic Model Major Category
HEC-2 backwater
IFG-4 (alone) empirical
Stage regression empirical
Data Available
51 main channel
6 slough and
side channel
sites
3 sloughs,
9 mainstem sites
(ADF&G 1983c)
-20-
Output (ace.)
mean depth
velocity, stage
(± 0.6 ft.)
depth
velocity
substrate
stage
(+ 0.1 ft)
Susitna assessment, however, any mention of IFG-4 or WSP refers to their
use in prediction of stage, velocity, or wetted area.
THE UNLINKED MODELS -ICE AND SEDIMENT
ICE
The Need for Ice Simulation Modeling in the Susitna Assessment Process
The need to simulate and otherwise account for ice processes in the
Susitna analysis is one common to all arctic assessments. Whereas river
icing is not even considered in assessment of impacts in most temperate
climates, ice processes in areas such as the Susitna Basin may be the
single most important determinant of stream channel morphology, riparian
vegetation distribution, and, perhaps, yearly mortality of some aquatic
species. Also, ice processes are unlike other instream mechanical
processes in that they are difficult to predict in a site-specific
manner. Finally, though the environmental effects of ice in large
northern rivers such as the Susitna are expected to be great, little
study is available detailing the ways in which ice, its movements and
presence, affects aquatic organisms.
The following section on ice, then, instead of presenting a
finalized program of study, presents the problem in three steps. First,
historical and expected ice processes in the Susitna are presented,
emphasizing ice effects upon habitat suitability and stability; second,
the expected sources of ice impact upon aquatic resources are presented;
and third, the capabilities attainable through ice modeling are
presented and evaluated in terms of required cost and time.
Description of Present and Future Ice Processes
Background. The data base pertaining to ice processes on the Susitna
River is extremely limited. Observations for the last few years have
been made at various gaging stations by the Water ·Resources Division of
the USGS as well as by R&M Consultants of Anchorage. The Alaska Rail-
road has informal records dating back roughly 20 years which describe
ice i amming and flooding events that affected the rail. This section
summarizes the available information concerning freezeup and breakup
patterns on the river.
-21-
Freezeup. The upper, t!liddle, and lower basins of the Susitna are
characterized by a stream temperature gradient that results from
differences in elevation and latitude and from the initial cold tempera-
tures of glacial melt. This gradient affects the sequence and timing of
ice cover events in that there is a period during late October and early
November when temperatures in the upper basin are below freezing while
those in the lower basin are still above freezing (R&M 19R2b). During
this early stage of freezeup (second or third week of October), frazil
ice (individual ice crystals) is generated only in the upper basin,
particularly in the colder turbulent reaches such as Vee Canyon, Watana,
and Devil Canyon. The areal coverage and strength of the ice from then
until early December is determined by local climatic conditions. Frazil
ice generation usually continues for three to five weeks, and sheet ice
develops simultaneouslv in areas of slower water. Slush ice floes may
form, and anchor ice may appear in shallow (4 to 5 ft) but fast water as
a result of frazil contact with the streambed. Toward the end of this
period, the .iamming of frazil ice pans or sheets causes the formation of
a solid ice cover in the lower river. Ice accumulates above the leading
edge of a jam, and the ice cover progresses upstream. Water elevations
at and above the ice front are often raised, or staged, by as much as 2
to 4ft (Bredthauer and Drage 19R2; R&M 1982a). A continuous ice cover
has usually formed by early December.
Available data show that ice seems to jam in the same places every
year--near constrictions due to bedrock outcrops, channel configuration,
and border ice (Bredthauer and Drage 1982). The solid ice cover
progresses from the confluence with the Chulitna River upstream to Devil
Canyon within about two to three weeks. Leads may still occur even
after this cover has developed, and some side channels and sloughs above
Talkeetna never freeze. The thickness of the ice increases throughout
the winter, and though it averages more than 4 ft by breakup,
thicknesses of more than lO ft have been measured near Vee Canyon. The
upper Susitna mav contrihute roughly 75 to 85 percent of the ice load of
the combined Susitna, Chulitna, and Talkeetna rivers (R&M 198la, 1982b).
Freezeup on the Chulitna and Talkeetna normally begins several weeks
after freezeup on the middle and upper Susitna River.
-22-
Breakup. Breakup is the process whereby the ice cover of a river
fractures due to increasing temperature and hydraulic forces in the
river. Once the ice has fractured and is afloat, breakup is complete.
The process is often rapid and awesome on large rivers such as the
Susitna. Because the lower basin experiences warmer temperatures
earlier than the upper basin, the snowpack at lower elevations melts
first, increasing river discharge and causing the ice cover to fracture.
Generally, breakup on the Susitna commences close to the mouth in
mid-May, progressing upstream over about a week, thus causing little
severe ice jamming. The severity of breakup is influenced by the
snowpack depth and melt rate as well as by the amount of rainfall. For
example, in 1981 breakup was mild because of the shallow snowpack, warm
spring air temperatures, and limited precipitation, which caused the ice
to slowly disintegrate in situ and leads to develop gradually
(R&M 1981). In 1982, however, severe ice jamming and resultant erosion
and flooding occurred because melting of the deep snowpack and late but
rapid rise in air temperature caused a sudden increase in water level
and, in turn, ice movement (R&M 1982b). While the severity of the
breakup iams mav vary, they tend to recur in the same places every year,
often the same areas where ice accumulates during freezeup.
Effects of Tee Processes on the Stability of Side Channels and Sloughs
Freezeup. As the ice front forms between Talkeetna and Devil Canyon and
progresses upriver, the water level (stage) in the main channel rises
due to the ice cover. Water and ice may flow into side channels and
sloughs tvhich were previously isolated from the main channel due to low
winter stage (Bredthauer and Drage 1982). Upstream progress of the main
channel ice cover then slows while frazil floes accumulate and thicken
in the side channels.
Tee fails to form in substantial portions of some side channels and
sloughs above Talkeetna due to upwelling of groundwater. Groundwater
h 2 and 4 ° c d ib h h temperatures vary etween an contr. ute cnoug eat to
prevent formation of :ice and to open leads. These areas are often
salmonid egg incubation areas (Trihey 1982a). F.ven when air
temperatures become cold enough to form an ice cover over these sloughs,
the substrates are not expected to freeze (Trihey 1982b).
-23-
River ice cover is important in maintaining groundwater flow into
the sloughs. Increased stage main channel ice cover causes a hydraulic
head (pressure differential) between the mainstem and nearby sloughs
during low winter flows. This differential is similar to that which
exists during much higher normal late summer flows (Trihey l982b) and
maintains groundwater upwelling into the sloughs throughout much of the
winter.
Breakup. River breakup processes, such as flooding and erosion associ-
ated with high runoff flows and ice jams, are considered the primary
factors influencing river morphology in the reach between Devil Canyon
and Talkeetna (Bredthauer and Drage 1982). During this period very
large short-duration flows pass through side channels and overtop the
berms separating sloughs from the main channel. These flows remove fine
sediments that may have accumulated in the sloughs during low winter
flows. Periodically flows are large enough to redistribute streambed
gravels, remove debris and beaver dams, and at times alter the thalweg
profile or alignment of a slough (Trihey 1982b). Ice blocks carried in
these flows probably exacerbate this effect.
The most severe flooding events appear to be caused by dry ice jams
(ice jams that become grounded). These usually occur at constrictions
or sharp bends in the river. High-water velocities cause ice blocks
from upriver to be submerged mid-channel under the ice cover. Some
submerged blocks become grounded, preventing the passage of additional
ice, and newly arriving ice blocks accrete to the upstream edge of the
;am. Passage of water through the ice jam is restricted, and water
upstream rises rapidly until it overflows into existing side channels or
creates new channels. Slough 11 below the Gold Creek bridge on the
Susitna River was apparently formed this way within the past 30 years
(R&M l982a). The berm at the head of this slough is unusually high and
apparently overtopped only at very high main channel stages.
Breakup flooding events frequently result in the deposition of
unconsolidated cobbles, sand, and silt upon berms and river bars that
are above norMal high-water levels and even well up into the forests of
vegetated islands and riverbanks. Extensive damage due to water and ice
block erosion has been reported both in channels and overbank
-24-
ve~etation. At slough 21 scarring of cottonwood trees to heights of
5 ft above ground was observed (R&M 1982a). These trees were well away
from the normal channel.
Ten ice jam sites have been observed in the Susitna River between
Talkeetna and Devil Canyon. Jamming apparently occurs there nearly
every year, causing various degrees of flooding and erosion, depending
on breakup conditions (Bredthauer and Drage 1982). Ice jam induced
flooding and erosion events appear to be the principal causes of change
and evolution in side channel and slough morphology.
Potential Aquatic Impact Issues
Instream. Potential instream impacts related to ice processes are ex-
pected to be: (1) staging-overtopping, (2) breakup timing, and (3)
tributary mouth and slough morphological changes. Although not limited
to any one section of the river, the impacts are expected to be greater
above Talkeetna.
Stagin~-Overtopping. Staging is the process whereby the surface
elevation (stage) of a stream becomes higher at some dlscharge due to
increased channel roughness, instream structures, or ice. Staging due
to ice may result from flow impediment caused by surface ice or anchor
ice. Susitna River staging, due primarily to surface ice, may exceed
4 ft, that is, the stage for a given discharge may be as much as 4 ft
higher after formation of an ice cover than during the ice-free period.
Increases in winter discharge would range between 115 percent in
the lower river to more than 600 percent in the upper river under
postproject conditions (ACRES 1982a). Areas with an ice cover might
experience staging due both to increased discharge and ice staging.
This would almost certainly lead to the phenomenon of overtopping in
certain side sloughs. Overtopping occurs when the main channel stage
exceeds the elevation of the berm at the upstr~am end of a slough,
causing main channel water to flow through the slough. This results in
(1) increased frazil production, (2) local anchor icing, (3) reduced
temperatures, and (4) increased ice cover in the sloughs. These
conditions are clearly unfavorable for successful salmon egg
incubation. Slough overtopping due to staging occurs naturally as the
-25-
ice front passes a certain slough during freezeup. Slough 8A was
overtopped during the winter of 1982-83, and a thick ice cover and 0° C
substrate temperatures were observed. Postproject operations may make
slough overtopping a common rather than an isolated occurrence.
Breakup. As discussed in the previous section, increased ice
thickness and reduced postproject spring and summer flows could change
the timing and magnitude of breakup. If breakup were delayed due to
increased ice thickness and longer melting times, delays in fish
migration would occur. Spring spawning fish, such as grayling and
longnose sucker, lvould be delayed access to their spawning grounds if
ice masses remained near tributary mouths. Early chinook salmon runs
would have difficulty reaching spawning tributaries if breakup and ice
flows were still in progress. Increases in the amount of ice in sloughs
and side channels could delay the outmigration of salmon smolts.
Changes in the timing patterns of fish that have evolved over long
periods of time could result in substantial mortality.
Tributary Mouth Morphologic Changes. Because the Susitna River
tributaries are such important fish habitats, it is vitally important to
assess potential impacts upon the fish populations which utilize them.
Clearly, conditions in the tributaries will not change with
project-induced changes in the main Susitna, but structural conditions
at tributary mouths might change either directly with Susitna discharge
changes or indirectly through effects of altered ice and/or sediment
dynamics.
Ice effects at tributary mouths result from combinations of the
?rocesses described earlier and are generally expected to arise as
follows. (1) Increased winter discharges and staging would increase the
lateral extent of the ice cover, the extent of which would vary with the
degree of discharge change and the slope of the iristream and nearstream
topography. Therefore, effects would probably differ in the upper,
middle, and lower Susitna basins. (2) If lateral ice extension reached
tributarv mouths, especially in the lower and middle Susitna reaches, it
would remain until breakup. (3) If breakup discharge levels were not
sufficient to carry the ice out, large quantities of ice would degrade
-26-
thermally and remain at the tributary mouths. If spring flows are high
enough to transport the ice, the additional ice volume might scour or
otherwise disrupt the tributary mouth areas, again causing potential
access problems to immigrating fish.
In-Reservoir. The filling of Watana Reservoir would inundate 54 mi of
Susitna River mainstem and 28 mi of tributary habitat, converting it
from a lotic to a lentic system. Habitat development in the reservoir
would be limited as the continuous filling and drawdown cycle would
inhibit development of a productive littoral zone. This would be com-
pounded by the continuous formation of ice ledges along the drawdown
zone. These layers of ice might further erode and scour the shoreline,
disrupting the littoral zone and preventing the establishment of a pro-
ductive habitat.
Grayling and longnose suckers could be expected to use the
reservoir for overwintering, and water level fluctuations probably would
affect them adversely. Both grayling and longnose suckers spawn in
tributary habitats during late spring. The reservoir would be rapidly
filling during this time of year, and their spawning areas would be
i.nundated. This inundation along with increased sediment deposition
from the tributaries could result in increased mortalities to the devel-
oping embryos. Any surviving fry would probably have a low survival
rate because of the lack of a productive rearing habitat in the
reservoir. The addition of shoreline and in-channel ice to this tribu-
tary area would compound the detrimental effects on the habitat by in-
creasing erosion and scouring and producing potential fish migration
barriers.
An increase in the amount of tributary ice in the inundation area
could cause larger-than-normal ice accumulation, These larger ice jams
would take longer to melt and wash out and could block the upstream
passage of fish to their spawning grounds. The incubation and rearing
success of fish utilizing tributary habitats above the inundation zone
would not be affected by water fluctuations or sediment deposition, but
increased ice accumulation could affect the timing of their access from
overwintering areas to spawning grounds.
-27-
Ice Models -Capabilities and Cost
Answers to ice questions probably cannot be provided by model
output alone. Regardless of the degree of ice modeling sophistication,
actual assessments of ice impacts would be based on combinations of
model results and professional opinion. The kinds of ice models
available are discussed here to familiarize the reader.
There are four levels of ice modeling efforts:
Level 1. Statistical analysis of historical air temperature and
ice formation and/or breakup data.
Level 2. Empirical/physical process models for heat loss and
frazil ice prediction.
Level 3. Empirical/physical process models for ice cover, forma-
tion, stability, progression, and breakup.
Level 4. Physical analog (scale model).
The first level of modeling determines the most probable date of
the first appearance of ice and the first shore-to-shore cover at
previously observed locations. The date of the start of breakup can
also be predicted. This level is data intensive and assumes that flow
and climatic conditions will not change beyond historical variations.
Level 2 models predict heat loss and frazil ice production.
Varying degrees of sophistication exist with respect to heat budget
computations. A low level of heat budget analysis would involve
approximating the total heat loss from the water by estimating
atmospheric conduction. Under winter conditions, conductive heat loss
to the atmosphere is the major source of heat exchange. A more detailed
approach would consider solar and atmospheric radiation, water back
radiation, streambed conduction, convection, evaporation, and
groundwater exchanges as sources/sinks of heat. The heat transferred to
and from the water determines both the temperature of the water and the
amount of frazil ice which can be produced.
The third level of modeling uses empirical or physical process
equations to predict ice cover stability and advancement as determined
by structural and hydrodynamic forces. This level of modeling requires
level 2 estimates of the frazil ice supply.
Level 4 modeling involves construction of a scale model of a river
reach and observation of synthetic ice flows. This kind of modeling can
-2R-
be used to predict ice cover stability and advancement as well as the
extent and location of ice .i ams. Level 4 also requires level 2
estimates of the ice supply.
Various ice processes which are of interest are related to the
capabilities of these modeling levels in Figure 5.
Several conclusions can be drawn from this table.
1. Level 1 modeling is generally inappropriate for systems where
the flow or temperature regime would deviate from the histor-
ical conditions.
2. Ice processes which involve deviation from historical con-
ditions require level 2 effort.
3. Ice processes w·hich involve simulating the mechanical and
hydrodynamic forces are expensive to model.
4. Level 4 analyses are expensive and require site specific mod-
els; level 3 is relatively expensive but can be performed for
a large stream system in one model setup.
Based on available data and expected future field activities, AEIDC
can provide level 1 and 2 modeling capabilities. Higher-level efforts
would require considerable additional time and funding. To answer the
questions posed on the fishery assessment, levels 2 and 3 at least will
be required to the level at which certain events can be excluded from
further consideration.
CHANNEL MORPHOLOGY AND SEDIMENT ISSUES
General Description
In most aquatic impact assessments the term sediment has meant fine
particles which, when suspended, cause increased turbidity, reduced fish
vision, feeding efficiency or gill function and, when deposited, cover
spawning and incubation areas or limit aquatic invertebrate production.
Most hydroelectric projects reduce suspended sediment concentrations
and, therefore, ameliorate conditions which may have been degraded by
high suspended sediment concentration. The extent to which reservoirs
reduce downstream turbidity is predictable if sediment storage rates
within the reservoirs are known. A competent estimate of
postimpoundment suspended sediment conditions is normally possible
-29-
I w
0
I
Modeling
Process
frazil ice formation
anchor ice
initiai ice cover
bridging
ice cover advancement
ice cover thickening
Figure 5.
Level
2
2
l
1+3
2+4
2+3
2+4
2+3
?+4
Predictive capabiJ5tier provided by various combinations
of ice Modeling levels with predictive reliability and
estimated relative cost.
Prediction Reliability
high for both location
and quantity
unknown
moderate
moderate
high
moderate
high
moderate
high
Relative Cost 2
(1 to 10) Comments
1 Wide range of sophistication among
different models.
1 Applications not recorded in
literature. A relatively large
assumption would have to be made
as to areas affected. Pro-
fessional judgment required.
1
4
10
4
10
4
10
Site specific. Large historical
data base required; cannot be
applied to altered flow or water
temperature conditjons.
Requires professional judgment and
historical data for determining
potential lodgement sites.
Separate model necessary for each
potential site.
Requires model output plus pro-
fessional opinion
Separate model necessary for each
reach of interest.
Thickening by juxtaposition, frazi]
accumulation, and static growth.
Thickening by juxtaposition only;
separate model necessary for each
reach of interest.
I w
I-'
I
aufeis
Hodel ing
Process
ice .iamming
staging
timing of breakup
breakup water levels
Level
2+4
2+4
2+3
2+4
1
2+3
2+3
2+3+4
Figure 5. (cont.)
Prediction Reliability
?
high
moderate
high
moderate
low
low
low
1. See preceding text for description of modeling levels.
2. Cost based on relative scale of 1 to 10.
Relative Cost2
(l to 10)
?
10
4
10
1
4
4
1
Comments
New application, reliability, and
costs unknown.
Separate model necessary for each
potential site.
Requires model output plus pro-
fessional opinion
Separate model necessary for each
reach of interest; site specific
Large historical data base required;
cannot be applied to altered flow
or water temperature conditions.
Requires running level 3 for entire
winter season resulting in low
reliability of ice cover
estimates.
Requires running level 3 for entire
winter season resulting in low
reliability of ice cover
estimates,
Confidence in estimate of water
levels would be higher than above.
Separate model necessary for each
reach of interest.
without the need for extensive field work or sophisticated computer
modeling.
Large impoundments such as the Susitna project reservoirs alter
downstream sediment dynamics through transport of streambed material
that can ultimately change the configuration of the stream channel and
substrate division. Though the impacts of changes in suspended sediment
concentration.are usually assessable, effects of substrate distribution
and channel configuration changes are often difficult to evaluate. Even
more difficult to predict are structural changes which could take place.
Therefore, not only are channel and substrate changes harder to predict
in themselves, but their effects upon the aquatic environment are also
less certain.
Application in Susitna Aquatic Assessment
Impoundments as large as \-Jatana and Devil Canyon reservoirs would
undoubtedly change both the discharge and sediment regimes of the
Susi.tna River. Relatively small changes in suspended sediment
concentrations could be expected, primarily because the dams are located
in the upper reaches of the Susitna River where natural sediment
concentrations are low or moderate (50 to 500 ppm) (R&M 198lb). The
Susitna River acquires most of its sediment load from the Chulitna and
Talkeetna rivers which would be unaffected by the Susitna dams. The
reduction in upper Susitna River suspended sediment concentration could
have a positive effect upon the aquatic habitat in that section of
river. Because gravel-cobble substrate predominates in the reaches
above Talkeetna, little channel change could be expected other than
establishment of a local area of scour immediately belm.r the dam (ACRES
19R2b).
All major channel configuration and substrate distribution changes
would occur below the Chulitna-Talkeetna confluences and are discussed
below as either streambed elevation changes or channe 1 configuration
substrate changes.
Streambed lUevation Changes. Streambed elevation changes are a widely
recognized result of operation of large reservoirs. The process
popularly invoked is that of scour below the tmpoundment resulting from
-32-
releases of clear water with high transport competency.
reduces the bed elevation, which results in degradation.
Scouring
Scour or
degradation is often predictable immediately below an impoundment, but
predictions are less reliable further downstream, where it is difficult
even to predict whether the change will be degradation or aggradation.
For example, hydroelectric darns usually reduce peak discharge levels and
raise base flows. Because most sediment is transported by peak flows,
the downstream effect is to reduce sediment transport to the extent that
the bedload normally carried away during runoff peaks remains in the
river reach of interest. Aggradation or raising of the bed elevation
has, in fact, been the major effect of many large reservoirs, especially
when tributaries below the reservoir were major sources of the impounded
river's sediment load.
Aggradation in the lower Susitna could have three effects. First,
coupled with reduced summer discharges, the aggraded bed might cause
passage problems. As the lower Susitna mainstem is a very important
migration corridor (ADF&G 1982; ACRES 1982a) any impediment to passage
might affect very large numbers of fish. Second, aggradation near
tributary mouths might cause access problems or changes in tributary
mouth habitats. Finally, significant aggradation would violate the
assumptions of fixed bed-elevations upon which any hydraulic simulation
models would be based.
The extent of postproj ect aggradation in the Susitna below the
Chulitna-Talkeetna confluence is probably not predictable without
results of specific sediment studies. Studies suitable to provide data
for this analysis are discussed in the next section.
Channel Configuration-Substrate Effects. Long term changes in sediment-
discharge equilibrium also result in changes in channel configuration.
Channel changes are always to varying degrees associated with streambed
elevation changes and relate to the previously discussed concerns.
Channel configuration changes are more difficult to predict than
aggradation/ degradation process changes nnd their effects more subtle
and less known to fishery biologists.
Channel changes which are positively associated with fishery
characteristics usually involve increasing stream habitat diversity in
-33-
terms of depth, velocity, substrate, and cover. Because significant
channel changes are not expected in the upper Susitna, and because they
would be difficult or impossible to predict in the lower Susitna,
channel change modeling will probably not be employed.
Available Models and Data Requirements
Simons et al. (1981) presented a review of available technologies
to predict either aggradation/degradation or measurable channel change
in river systems. In general only qualitative or semiquantitative
results are possible within present engineering capabilities.
Quantitative channel-change modeling appears to require field data
collection and computer modeling beyond the scope of most environmental
assessments. Suspended sediment and aggradation-degradation studies are
feasible within most project scopes and are discussed here in terms of
data requirements and expected capabilities.
Suspended Sediment Only. Only a detailed compilation of suspended
sediment gage records is required, along with the expected storage of
such particles in the reservoir. The reservoir storage study is usually
accomplished during the feasibility phase of the proiect but may not
offer enough resolution to provide particle-size distinctions required
by fisherv biologists. The approach is usually quantitative but unso-
phisticated; suspended sediment concentrations below the dam are comput-
ed on the basis of the difference between inflow concentrations and
total storage. Qualitative estimates of postproject Susitna River
suspended sediment concentrations are available (R&M 1982b) and will be
refined as project specifications change.
Aggradation/Degradation Studies. This approach normally involves
determination of a preproj ect sediment budget for all stream reaches
(mainstem and major tributaries) and categorization of the reaches in
terms of their equilibrium status. From the present discharge/sediment
equilibrium re)ationships, the postproject relationships are assumed or
calculated based on pro_; ected sedirn.ent storage, water discharge, and
tributarv sediment inflow. Normally, a fairly extensive water-sediment
discharge record is required for all mainstem sites and tributaries of
-34-
interest. In-channel data (cross sections with bed particle sizes) and
available stage rating curves from USGS gages are required at certain
locations. Computations and analyses utilize one-dimensional computer
models which predict bed elevations along the length of the river, al-
though some generalized, preprogrammed statistical packages are quite
helpful. Results are both site-and river reach-specific and are useful
to define the reaches where aggradation or degradation might occur and
to estjmate relative magnitude of the process. Results are valuable in
sensitivity testing to determine whether the processes will be
significant under various reservoir operating regimes.
-35-
TFF SUSITNA INSTREAM IMPACT ASSESSMENT APPROACH
Because the aquatic studies program and associated field data
collection efforts are presently at an intermediate stage, it is
difficult to describe an exact impact assessment approach. Because
consensus has been reached on certain major aquatic impact concerns, the
approach described at this time emphasizes near-term (fiscal years
1983-84) activities. The assessment impact approach is expected to be a
combination of those herein presented, and alternative avenues as
necessary.
SUSIM APPLICATIONS
SENSITIVITY TESTING
The SUSIM system, as described in the foregoing sections,
incorporates components which should address the ma_i ority of Suhydro
aquatic impacts. The system is clearly quite complex and would be very
cumbersome to run if all of the linked models were required for analysis
of each operating regime. The separation between linked and unlinked
models results from the necessity to run as few models as possible
during the task of iteratively "tailoring" flow regimes to minimize
impacts.
As initial runs of certain models become available, we find that
some parameters might not change significantly under project operations.
As an initial project analysis we will perform sensitivity testing on
all parameters prior to structuring SUS 1M in the assessment
configuration (Figure 6). For example, if through application of a
reservoir temperature model it is determined , that winter release
temperatures will range only between 2° and 4° C with high likelihood of
3° C, it might be unnecessary to run the reservoir temperature model in
a costly linked fashion. Similarly, if slough temperatures seem to
respond little to mainstem temperature changes, the highly complex
groundwater temperature model will not be linked to the instream
temperature model --slough temperatures wlll be fixed at some long-term
average for all conditions except when overtopping occurs, at which time
they would be predicted as main channel temperatures. In each
-36-
Figure 6. Expected sensitivity testing of SUSIM models and components
Model or component
Reservoir operation
model
Reservoir temperature
model
Instream temperature
Slough temperature
Independent variables
to be varied
Inflow
Reservoir stage,
inflow temperature
Reservoir release
temperature, discharge
Main channel
temperature, discharge
1. To be obtained from Acres American DYRESM results.
-37-
Output whose variation
is to be checked
Release discharge
1 Outlet temperature
Downstream temperature
variation
Slough temperature
discharge
sensitivity test, the independent variables will be input under the
range of expected natural or project-induced conditions, whichever is
greater. For example, the instream temperature model will be run under
the range of expected temperatures and reservoir outflows for each
month. Variation of the predicted instream temperature will be
evaluated in terms of significance both at the dam site and downstream.
If total temperature variation during critical periods is within bounds
considered to be biologically acceptable, the reservoir temperature
model will not be run for each successive reservoir operation, and a
seasonal mean temperature will be assumed.
ICE AND SEDIMENT STUDIES
Similar to model sensitivity testing, ice and sediment studies
should result in conclusions regarding potential effects of these
components. Because the sediment issue is expected to center around
aggradation in the lower river, a credible prediction of that process
may assist in concluding that aggradation would not be a problem. If,
however, it is predicted that aggradation would be problematical, this
suggests additional effort to either reduce effects (through changes in
reservoir operations) or to more clearly define their magnitude.
Similarly, if ice study results do not suggest significant occurrence of
any of the three expected ice problems, it will not be necessary to
continue ice modeling efforts. Ice process predictions can be expected
to be much more difficult and subjective than those from any other
component and dismissal of the ice issue likely will not be achieved in
the near future.
LINKED MODEL RUNS AND IMPACT ASSESSNENTS
Actual impact assessment runs of SUSIM will not be accomplished
until sensitivity and ice-sediment studies are complete enough to insure
SUSIN reliability; however, the general process of quantitative impact
assessment is known at this time. Though a variety of approaches
exists, only those which pertain to the current aquatic studies emphasis
on quantitative instream flow assessments will be presented.
-38-
THE SUSITNA INSTREAM FISHERY IMPACT ASSESSMENT APPROACH:
CONJUNCTIVE USE OF THE SIMULATION MODELS
The instream impact assessment will be accomplished in three
phases. First, a sensitivity testing and preliminary analysis will be
used to determine which parameters are likely to change significantly
throughout the Susitna River. Second, through use of the simulation
models described in the following section, these changes will be
predicted at all selected fish habitats in terms of habitat structure
resulting from changes in streamflow, temperature, ice, and channel
morphology. Third, these predicted changes will be interpreted in terms
of fish habitat and/or population impacts.
QUANTITATIVE ASSESSMENT TOOLS
Recent developments in instream flow assessment have led to a vari-
ety of standardized methodologies, each of which might have some appli-
cability to the Susitna assessment. Each requires output from the simu-
lation model system. Under each habitat type (main and side channel,
tributary, and slough), the following potential methodologies (or combi-
nation thereof) will be discussed with respect to (1) applicability, (2)
data requirements and current or expected sufficiency of data, and (3)
need for interpretive refinement. Details about methodologies follow.
Hydrologic Methodologies
Evaluation of postproject streamflow is based on frequency of
preproject flow events and knowledge of the habitat or population ef-
fects of those events. For example, the Tennant method (Tennant 1976)
suggests that flo~~s equaling 60, 40, and 20 percent of mean annual dis-
charge levels provide "optimum," "maintenance, 11 and "minimum" habitat
levels for trout species in certain Rocky Mountain streams. Hoppe and
Finnel (1970) found certain high-flow recurrence intervals to be
necessary for gravel flushing to insure salmonid incubation success.
Data required are pre-and postproject streamflow records, available
from the reservoir operation model. Such hydrologic methodologies have
not been considered for use on the Susitna pro.i ect in general because
(1) the evaluation criteria have not been applied or verified in Alaska
-39-
or other Pacific salmon habitat tyves and (2) the overall Susitna
assessment apvroach has been toward comvrehensive and quantitative
studies which preclude the simpler "office" approaches. Such methods
may be of value on this project, however, especially in reaches for
which detailed data might be unavailable.
Hydraulic Simulation Hethodologies
These methodologies base their impact evaluation on changes in
certain hydraulic variables (depth, velocity, substrate, etc.) at sites
of known importance. Use of such methodologies requires considerable
hydraulic data but often only limited or general knowledge of fish
habitat preferences. Many states have used either single or multiple
cross-section applications of these methodologies (U.S. Forest Service
1973; Spence 1975), and interpretations are based upon calculations of
predicted depths or velocities or passage requirements related to the
body depth of the target species. The Susitna slough studies are
actually a form of this methodology in their calculation of depth and
access into sloughs at given flows but without dependence on detailed,
species-specific habitat requirements or incremental calculation of
habitat suitability or area.
Hydraulic Simulation Methodologies with Habitat Preferences
Exemplified bv the U.S. Fish and Wildlife Service (USFWS) instream
flow incremental methodology fRovee 1975, 1982), these studies provide
quantified habitat output based upon prediction of hydraulic habitat
variables (depth, velocity, and substrate) and interpretation of those
variables in terms of habitat preferences defined by suitability-of-use
or habitat utilization curves (Bovee and Cochnauer 1Q77). The output,
usually surface area of preferred physical habitat at a given flow
level, supports a highly quantified impact evaluation.
Habitat Tndex F.valuations
Associated primarilv with habitat evaluation procedures (HEP)
developed bv USFWS Pro.i ect Impact Evaluation Group, this method is pri-
marily for habitat accounting. It may have predictive value in aquatic
-40-
impact assessment if enough is known of the postproject streamflow and
hydraulic conditions (USFWS 1980). Quantified results from some
hydraulic-based assessment usually are used as habitat units in the REP
analysis; however, HEP analysis might be valuable in project areas to be
inundated or where hydraulic information (from physical habitat
analysis) provides estimates of postproj ect riparian or other
terrestrial habitat.
Miscellaneous Approaches
Most standardized instream or reservoir analysis methodologies do
not deal directly with many aspects of project construction or operation
which must be addressed in a formal impact evaluation process. For
example, building and maintenance of roads and construction camps may
not directly affect aquatic habitat but often result in increased access
and corresponding fishing pressure. Similarly, quantifiable effects of
dam construction may be difficult to factor into a prestructured
methodological approach but must be considered among potential impacts.
Therefore, any truly comprehensive impact evaluation must avoid
restriction by structured approaches to assess all possible impacts.
ASSESSMENT APPLICATTON
Any impact assessment in a large complex river system such as the
Susitna should provide accounting for postproject watershed effects,
such as multiple operational schedules, variable release temperatures,
ice processes, and postproj ect channel changes and sediment dynamics.
The methods for accounting for such watershed effects using simulation
models have been presented in a previous section, but no single approach
for the Susitna impact assessment can be recommended.
Data and interpretation limitations and unique Susitna River
habitat relat:f_onships preclude a comprehensive analysis in terms of
geographic scope and species using any sing"le or prestructured
assessment methodology. A number of potential impact assessment
approaches could be used, and this section describes several of the most
applicable. The final assessment method will be a mixture of
approaches. The emphasis will remain on quantitative evaluations where
possible.
-41-
Potential assessment methods seem to vary more by habitat and
geography than by target species. For example, data are probably
complete enough to allow quantification of habitat values in certain
sloughs based on detailed hydraulic and habitat measurements; however,
for the mainstem above Talkeetna data are not sufficient for a similarly
quantitative assessment. Data concerning lower Susitna sloughs are even
scarcer. Therefore, the kinds of assessment approaches will be
presented under mainstem-side channel, slough, and tributary headings.
Slough habitats above Talkeetna are presented last because of their
relative abundance of data and potential assessment approaches.
Mainstem-Side Channel
The small numbers of salmon so far observed spawning in main
channel areas suggest limited utilization, but the amount of mainstem
salmon spawning is not known at this time. Mainstem areas, especially
those near channel margins, may provice access for
between sloughs and between tributaries and sloughs.
juvenile salmon
Overwintering of
fry or juvenile salmon may be a significant mainstem activity. Far less
information exists for fish utilization in the mainstem belmv Talkeetna
than in the upstream reach.
Hydrologic Methodologies. Though hydrologic methods have not been con-
sidered, their most relevant use might be to evaluate impacts in the
lower Susitna for resident populations and for such main-channel
spawners as eulachon and Bering cisco. This is because (1) there are
presently no quantitative habitat preference data for either species,
(2) there is no cross-section information upon which to base spe-
cies-specific physical habitat predictions in this reach, and (3) life
history dynamics probably are not well enough understood to support a
population predictive approach. A similar lack of data in the upper
Susitna may increase the attractiveness of fixed-percentage approaches.
The actual means of evaluation using this approach would involve
the following steps.
1. Determine the present flow regime from measured data at Gold
Creek, Chulitna, and Talkeetna gages and determine mean annu-
al, 20th, and 80th percentile exceedence flows.
-42-
2. Determine from literature sources percentages of mean annual
flow associated with the population levels of the lower
Susitna mainstem fish species and in similar situations else-
where.
3. Establish flow ranges for maintenance of present population
and for preservation of minimum population levels.
4. Evaluate project flows with respect to established ranges and
determine the likely postproject population condition.
As stated earlier, this approach (or any other like it) may not be
appropriate for the Susitna assessment because of the lack of relativity
to Alaska situations. Its ultimate weakness may lie in the scarcity or
absence of similar streamflow-climate situations upon which to base
development of Alaska criteria.
Hydraulic Simulation Approaches. Few usable cross-section data exist
for the lower Susitna mainstem, but some of the available single
cross-section work (Bredthauer and Drage 1982) may help simulate passage
conditions and determine expected channel morphology changes. Salmon
passage assessments using single and multiple cross-section approaches
have become quite common in Oregon and WRshington. These assessments
require predicted depth and velocity information with which to predict
habitat conditions for the species in question. More complete
cross-section data are available for upper Susitna reaches, but passage
is likely to be less problematic there.
Hydraulic Simulation Plus Species Preference Approach. Lack of multiple
cross-section hydraulic simulation data in the lower Susitna mainstem
again precludes this approach entirely. In the upper Susitna, some
cross sections ;ue available, but substrate data are lacking and the
general low level of specific life history data and mainstem habitat
utilization makes the value of multiple cross-section data collection
questionable. If greater utilization of this habitat w·ere demonstrated,
especially for critical activities such as spawning, considerable field
effort might be justified.
-43-
Habitat Evaluation Indices. Instream habitat changes will probably be
predicted using a hydraulic-based quantitative approach (PHABSIM,
R2-cross, HEC-2), thus precluding use of the aquatic habitat evaluation
procedures. Riparian habitat changes predicted using results of the
hydraulic simulation may be used in a terrestrial impact assessment or
in conjunction with the streamflow assessment if mainstem habitat units
are calculated.
Miscellaneous Methodologies. Because effects other than flow or
temperature alterations are expected in the Susitna mainstem,
miscellaneous methodologies will probably not be employed.
Summary. Mainstern impact evaluations may be based on fixed-percentage
approaches for resident species throughout the river, but relevant
criteria will be a problem. Some passage evaluation will be possible
using available hydraulic data and passage requirements from Alaska or
other regions. More sophisticated habitat predictions will depend upon
provision of mainstem cross-section data for both the upper and lower
Susitna.
Tributaries
Two types of tributaries exist in the Susitna basin. These are
defined as major tributaries (the Yentna, Chulitna and Talkeetna rivers)
and minor tributaries (all other tributaries to either the Susitna or
major tributaries). Tributary streams support a variety of fish species
and life stages throughout the Susitna basin. Recent surveys indicate
that all resident species utilize tributaries at all phases of their
life history. Bering cisco are probably the only anadromous fishes
which do not utilize tributaries at some time. Of the Pacific salmon,
only sockeye appear to have ] imited involvement in minor tributaries,
hut sockeye habitat uti] ization is poorly known in areas other than
upper river sloughs. Of the remaining salmon species, chinook, coho,
churn, and pink salmon spawn primarily in minor tributaries. Chinook and
coho mav use tributaries, especially the mouths, as rearing or
overwintering locations.
-44-
Because no tributaries will actually be dammed by this project, the
expected impacts will be those associated with inundation, access and
road building, or changes in mainstem Susitna River discharge patterns.
Inundation effects are expected both during filling and seasonal
Discharge effects are expected to result
mainstem lvater level and perching of
fluctuation of the reservoir.
from decreased postproject
tributary sediment deltas. In this case failure or delay of the tribu-
tary to erode through the perched delta would eliminate access into the
tributary during periods when mainstem discharge fell below critical
levels.
Hydrologic Methodologies.
known.
No previous use of such methodologies is
Hydraulic Simulation Methodologies. Some quantitative evaluations of
tributary access may be made using sediment delta simulations based on
cross sections measured near the tributarv mouths. Unless there is some
knowledge of postproject mainstem sediment dynamics, however, the degree
of perching will not be well known and the predictive benefits of
cross-section measurements limited. Questions regarding sedimentation
on eggs deposited in tributary channels during low flow might be
addressed using detailed cross-section information for affected
tributaries plus sediment transport measurements for those tributaries.
Such data are not currently available but may be collected in the near
future.
Habitat Evaluation Procedures. Effects on tributaries in the impound-
ment area may be quantified by map-based estimates of losses of riverine
habitat at various reservoir elevations. Habitat accounting might be
applied through such indices as RSU' s to evaluate tradeoffs between
inundated tributary and main channel fisheries and those expected to
arise as a result of reservoir existence and increased access near the
access corridor. Regardless of the actual methodology used, it would be
advisable to place habitat values in impoundment area tributaries and in
the reservoir itself on an equal scale to facilitate evaluations of
tradeoffs.
-45-
Miscellaneous Approaches. Perhaps the most valuable approach to tribu-
tary analyses would result from combining quantitative methodology main-
stem information with tributary population and recruitment predictions.
Changes in Susitna River discharges could affect the tributary fisheries
in three ways: (1) through reductions in adult mains tern passage the
number of tributary inmigrants might change; (2) if significant perching
occurs, tributary inmigrants will again be reduced; and (3) changes in
main channel flow might alter the efficacy of travel between tributary
mouths and nearby sloughs thought to be important for juvenile tribu-
tary-spawning salmon species.
If mainstem-tributary mouth channel relationships were known, re-
ductions (or additions) to normal tributary inmigration could be cal-
culated and applied directly to escapement counts, which could form the
basis for new population sequences. Time-series of project flows could
then be evaluated with respect to long-term escapement and return trends
in the tributaries and the various project operations compared in terms
of tributary fish production. Such an analytic approach would require
extensive hydraulic and fish population information with strong
predictive capabilities in ice and sediment dynamics under project flow
conditions.
Sunnnarv. Tributary impact assessments depend heavily on quantitative
hydraulic information currently not available for the lower Susitna.
Upper Susitna applications are possible.
Sloughs
The greatest amount of quantitative project-related habitat data is
available for sloughs between Talkeetna and Devil Canyon; however, these
data are not necessarily in proportion to the percentage of the total
Susitna basin fish stocks that utilize sloughs. Based on recent ADF&G
survey results, a total of fewer than 7,000 saimon (mostly chum and
sockeye) utilize these habitats for spawning each year in the upper
Susitna. The emphasis on upper river slough data collection has
resulted primarily from the assumption that project effects will be most
severe and definable on access, ~pawni.ng, and incubation in the sloughs.
-46-
All salmon species utilize sloughs at some time in their life cy-
cle, but the primary use is by chum and sockeye salmon for spawning,
incubation, and early rearing. Sockeye salmon in the upper Susitna
appear to utilize sloughs almost exclusively for spawning because no
significant sockeye spawning has been demonstrated in the upper Susitna
tributaries or main channel. Most Susitna basin chum salmon spawn in
tributaries, and they are also the most numerous slough spawners. Pink,
chinook, and coho salmon seldom spawn in sloughs but use them to various
degrees for overwintering and juvenile rearing. The hydraulics
associated with habitat conditions at certain sloughs are quite well
studied and provide (for those sloughs) a credible and comprehensive
basis for predictive modeling.
At the intensively studied sloughs it is possible to predict
mainstem water elevation (stage) for any discharge and to relate that
stage to an estimate of access to the slough by salmon. Wetted surface
area within the slough can similarly be determined from mainstem
discharge. Relative seasonal utilization by different salmon species is
known for a number of sloughs. Because of the detailed physical-bio-
logical data base available and because project effects are likely to be
great in the upper Susitna, these sloughs currently offer the best
opportunity for quantified impact assessment.
Fixed-Percentage Methodologies. Little or no basis exists, either of
necessity or bv supporting technolo~y. for use of "office approaches" to
assess slough impacts.
Hvdraulic Simulation Approaches. Most quantitative efforts to date have
involved use of hydraulic simulation results related to mainstem
discharge with access into certain sloughs, Through predictions of
depth at slough mouths and observations of fish access at certain flows,
it is possible to determine relative access efficiency (degree to which
depth limits fish passage) for mainstem discharges from 8,000 to about
20,000 cfs. Because depth at the slough mouth is a function of
discharges in both the slough and the main channel and because slough
discharge mav not be directly related to main channel discharge, the
predictive modeling of access is not strictly deterministic.
-47-
Certain sloughs have been characterized by multiple cross-section
measurements at several flows. From these it will be possible to pre-
dict surface area and spatial distribution of depth, velocity, and
substrate at a range of slough discharges. These hydraulic
simulation-based assessments and associated studies for slough discharge
dynamics with respect to groundwater, upwelling, ice formation, and
surface runoff have provided the major basis for simulation modeling in
the Susitna basin.
Hydraulic Simulation with Habitat Preferences. At six sloughs for which
multiple cross-section (IFG-4) data sets exist, it will be possible to
calculate actual spawning, incubation, and rearing habitat using the IFG
HABTAT program. This modeling process requires habitat preference or
utilization curves which, though available through IFG for all Pacific
salmon species, will probably not be acceptable for Susitna River
populations. Calculation of WUA or a similar area-weighted habitat
index will be relatively simple once decisions are made of acceptability
of currently available habitat preference curves. A conceptual popula-
tion model using WUA and/or slough surface area vs. mainstem flow
relationships is presented later in this paper.
Habitat Evaluation Indices. The general guidelines of aquatic REP are
probably not applicable to slough assessments because of their highly
unique and variable habitat characteristics for which REP weighting
factors and model formulations have not been developed. As in the other
habitat types mentioned, however, quantification of habitat or popu-
lation gains and losses in sloughs should be expressed in terms compati-
ble with those developed for the mainstem and tributaries.
Miscellaneous Approaches -Proposed Slough Impact Assessment Designs.
Because the hydraulics, dynamics, and fish utilization of sloughs are
highly unique and because they cannot be considered simply as small
streams or ephemeral lakes, the approach to assessment of slough impacts
will necessarily be a combination of the above approaches and in many
respects a collection of singular processes utilized on a
slough-by-slough basis. Depending on requirements for analytic
-48-
comprehensiveness and available information, numerous assessment
approaches are possible for sloughs. Those presented are examples of
potential levels of activity, listed with associated data needs and
output limitations.
As mentioned earlier, two distinct analytic approaches exist where
streamflow is the primary variable--those which analyze impacts based on
changes in aquatic habitat and those \vhich do so based on either inter-
pretation of habitat in terms of fish population effects or by direct
determination of population effects. Projects are often evaluated in
terms of fish habitat changes alone without knowledge relating changes
in habitat for a given life stage to effects upon fish population
strength, commercial fishery economics, or sport fishery success. The
following potential approaches range from those dealing only with
habitat to those which might predict trends in salmon populations. The
latter may seem unfeasible because of data limitations but may offer the
most reliable assessments upon which to base decisions on project
operations.
Habitat Dynamics Displays and Comparisons. This approach would require
provision of mainstem discharge vs. slough habitat by ADF&G, based on
observations and interpretation of quantitative material. For each
species at each slough or slough type, ADF&G would provide a curve simi-
lar to the following. In this case, surface area of standing water
would be related to mainstem discharge under the assumption that
standing water surface area is equivalent to usable habitat (Figure 7).
At each slough, postproject discharges in the form of mean monthly flows
expected for a 32-year forecast period would be available for each
potential project operational schedule. This discharge time series
would be in the form given in Figure 8. A similar flow-time matrix is
available for pre-project flows.
Habitat assessment of any project operation s~hedule could be done
quickly by determining the habitat suitability value for each discharge
in the matrix and then constructing a 32 x 12 matrix of habitat suita-
bility values. Both pre-and postproject matrices could be constructed
and comparisons made for each target species in pre-and postproject
spawning habitats, for example. Comparisons could be made in a
-49-
Figure 7. Susitna River discharge (Q) vs. surface area (in 1000 sq. ft)
for Slough 8A (extrapolated from ADF&G 1983c).
STANDING H20 SURF ACE ARf.lt
r t-
200 L
L ~
f.. I r
lSIIJ L
!
t-~
t
I
lllll I-
~ l
L..-.J-.._J_--l---L--l~l..-..L-_J_.--l---L-----'-'----'--L_j_'----'---J-~---L-l--L-L~-.<--JC-.1-------L . ..l.._j
B l01l01l 15000
Q (Cf5) SUSITNA MAlNS1EH
-50-
Figure S. ~lean September pre-and postproject Susitna River flow at Gold Creek.
YEAR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Surface area is assumed equal to usable habitat.
PRE-PROJECT
FLOiol
(cfs)
S301
21240
14480
15270
12920
14290
18330
19800
7550
16920
20510
13370
15890
12320
9571
19350
11750
16870
8816
9776
9121
14440
12400
9074
12250
16310
6881
12640
8607
10770
13280
13171
SLOUGH SA
SURFACE AREA
(standing water)
57.1
1S6.0
162.6
165.3
157.2
162.0
176.0
1Sl. 0
42.5
171.1
1S3.5
15S.S
167.5
151.6
S6.9
179.5
138.2
170.9
69.2
9l.S
76.4
162.5
153.4
75.3
149.9
169.0
31.4
156.3
64.3
115.1
158.5
158.1
-51-
POSTPROJECT
FLOW
(cfs)
9300
9300
9300
9300
9300
10444
1S330
10173
9300
14603
9300
9300
15S90
11551
9300
10645
9300
16S70
9300
9300
9300
10053
9300
9300
9300
9300
9300
9300
9300
9300
9300
13171
SLOUGH SA
SURFACE AREA
(standing water)
S0.6
S0.6
S0.6
S0.6
S0.6
107.5
176.0
101.1
S0.6
163.0
S0.6
80.6
167.5
133.5
80.6
112.2
S0.6
170.9
S0.6
S0.6
80.6
98.3
80.6
80.6
80.6
80.6
80.6
80.6
80.6
80.6
80.6
158. 1
strictly numerical way with means and t-tests, summation or integration
processes, or cumulative frequency comparisons (Figure 9). In any case,
project assessments would be made on a species-by-species, life
stage-by-life stage, and slough-by-slough basis. Impact assessment
could be based on weighting of the individual factors or could consider
all species, life stages, and sloughs of equal value.
This process could be employed using WUA as the dependent habitat
variable in sloughs, tributaries, or main channel locations if suitable
cross-section data and habitat preference information were available.
Since such data are not currently available, the spawning surface area
dependent variable was used.
The described habitat comparison approach has certain advantages in
the Susitna assessment. First, specific biologic data requirements may
be low. Second, the resultant comparisons are relative] y simple and
demonstrable using computer graphics or summary statistics. Third, the
time-series concept promotes consideration of the project in terms of
monthly streamflow variation and not simply a single mean value likely
never actually to occur (Trihey 1981).
As discussed in Bovee (1982), however, such habitat comparisons do
not evaluate the effects of habitat changes, merely the fact that they
have changed. Habitat changes affect life stages differently. Actual
population effects range from zero to total mortality. Further, even if
changes in habitat for all life stages were quantified, it would be
difficult to evaluate those changes by simply assuming some linear
relationship between habitat and abundance; numeric strength of fry
requires knowledge of both fry habitat and numbers of fry available
which, in turn, requires knowledge of numeric strength of eggs and both
natural-and habitat-related reductions in that strength.
Population Trend Models. If a certain life stage were known to be
especially sensitive or critical to maintenance of a salmon population,
it might follow that provision of enough water for that life stage would
minimize impacts while also reducing constraints on project operations
because of the limited time frame for which flows must be provided. A
good example involves slough access and spawning and the need for a
certain minimum main channel discharge during August of each year to
-52-
Figure 9. Percent exceedance of Slough 8A surface area (in 1000 sq. ft
for September) under 32-year pre-and postproject flow conditions.
PERCEljT EXCEEDNICE <SEPT>
100
PRE ?ROJECT
!'CST i'K'uJECT
B Hl 2ll 311 4~ 50 6ll 71! Bll 00 !Ill! i Ill l2ll 13ll 141! !50 16~ 170 i8!l Hlll 2llll
STANDING fQ1l SURF AC£ AREJ.
-53-
provide access of sockeye and chum salmon to sloughs prior to spawning.
Flow-related egg mortal5.ty is a function both of water elevations below
critical levels (dessication) and other factors relating to upwelling,
gravel permeability, and substrate stability.
Actual numbers of available inmigrating salmon, fecundity, or
spawning densities are generally available but not necessary to predict
project-induced trends because the primary objective is to predict
population index trends, not actual populations. Therefore, any assumed
value of inmigrants could serve as an initial index to be reduced or
augmented by various flow and nonflow related factors.
A chum salmon trend model would sequence through the following
steps using the indicated relationships.
Honth Time-step
8 i
9 i+l
10-3 i+2
4 i+3
Life stage sequence and reduction factors
N ps
N as
N eggs
N se
N out
N. * A.E.F. * M 1m p
d * A redds slough
if
N ps
A slough
> d -redd
N * X fee as
(fee 3,000 egg/fish)
N * H eggs e if WSE > WSE c
N * f(Q) * M if WSE < ¥SE eggs e - c
Mf * N * f(T) se
-54-
where: N = number
subscript ps potential spawners
im inmigrants
as actual spawners
eggs = spawnable eggs
se surviving eggs
out outmigrating smolts
d = density of redds in a slough redds
fee fecundity
c = critical
WSE = water surface elevation
M = mortality
subscript p = passage
f(Q)
f(T)
e = eggs
f fry
n = non-riverine (oceanic)
function of streamflow
function of stream temperature
1.0
A.E.F -~
.2
·08L-o:::::::::::....---:';:----!:,4;-----;,;-s ----;.2o
Q(cfs)
U• thousand$)
A. E. F. Access Efficiency Factor from the relationship
30,000
S~~trfoee
Ana
{112)
A slough
16 18 2S
Q (cfs)
ltn thotaands)
Surface area (S.A.) of slough vs Qi+l
-55-
%
ndd~
CO'Itred
O(chl
(in tho~o~sands)
10
0 = Critical Q minimum to incubate all eggs deposited
c at Qi from relationship
Following is an example evaluating effects of pre-and postproject
mean monthly flows with an imposed minimum of 12,000 cfs in August. For
each project operation at a given slough either a number of potential
spawners (N ) is assumed or the maximum number based on greatest ps
surface area times redd densitv (A 1 h * redds/ Area) is calculated. -s oug
The Slough Access F.fficiency (S.A.E.) max is calculated for the predict-
ed mean August discharge (time step i) from function 1. This Access
Efficiency is multiplied times either the number of potential spawners
(if known) or by the maximum effective spawning number to produce number
of actual spawners, or the number likely to successfully enter the
slough and be available for spawning. The number of actual spawners
reflects correction by a sex ratio factor to more accurately indicate
spawning pairs. At this point, any August discharge in excess of that
offering Access Efficiency of 100 percent would become available for
project operation nr storage. Next, the surface area of the slough
would be determined using function 2. If IFG-4 data are available, the
September discharge will be converted to spawning WUA. Again referring
to the redd density ratio, the number of actual spawners would then be
compared with the maximum number of spawners at the September discharge
(Qi+l). Again, discharge providing greater surface area than could be
utilized by the maximum number of spawners would be made available for
storage or later accounting. Discharges creating surface area limita-
tions would result in a recluction in the number of redds in direct
proportion to losses in surface area. The resulting number of redds
would then be multiplied by the average number of eggs per female (fee)
to determine the number of potentially deposited eggs (Ne). At this
-56-
point, all natural (non-flow-related) mortality on eggs could be
accumulated and one egg mortality factor applied to determine the number
of surviving eggs (N ) • se The discharge table for the months October
through May would then be searched to locate months when flows are at or
below a critical minimum determined from function 3. If all flows
during this period turn out to be above the minimum, only natural
mortality factors would apply. Below this minimum, an additional egg
mortality directly proportioned to loss of wetted area would be applied,
reducing the number of surviving eggs.
Also during this period, the locations of the slough(s) under study
would be checked for proximity to the predicted ice front. Water
surface elevations for sloughs above the ice front would be determined
from summer rating curves. Water surface elevations near or below the
ice front would be derived from winter rating curves which reflected
staging. Those at or quite near the ice front would be determined from
rating curves which reflect maximum staging conditions. For each
discharge, staged water surface elevations should be known and compared
to the minimum elevation necessary to breech the head of the slough. If
breeching occurs during incubation, total egg mortality could be
assumed.
If breeching does not occur, N could be modified bv fry mortality se -
factors. If overtopping occurs during emergence, total mortality again
could be assumed. If overtopping occurs when fry are free-swimming (a
highly unlikely event), mortality could be estimated, based on predicted
slough velocities and tolerance limits of salmon fry for velocities
expected in the overtopped slough.
The final number of potential outmigrants (N )
0
reflects all ex-
pected natural mortality factors plus effects of those attributed to
changes in mainstem or slough discharge. Under preproject conditions,
little flow-related limitation is implied, especially near the median or
modal flows for each month. This does not mean that limiting factors do
not now exist but that the current condition should be considered a base
line upon which project effects might either improve or degrade fish
populations. The extreme variation in preproject flows might easily be
among the greater population limiting factors. Changes in discharge
during the critical August through September period, and to a lesser
-57-
extent during the rest of the year, should cause changes in the number
of potential outmigrants, which after predictable oceanic mortality will
return three or four years into the project operation. This reduced
number of inmigrants will serve as a new initial number subject to re-
duction during August of the project's fifth year by the Q. flow and
]._
associated access efficiency.
OTHER CONSTRUCTION-RELATED EFFECTS
INTRODUCTION
There are a number of potential impacts that may be related to
peripheral construction activities. These activities take place away
from the Susitna River course itself. These construction activities
include service and access roads and corridors, transmission line
routes, construction camps and other habitation areas, gravel removal
sites, reservoir vegetation clearing, and other related human effects
such as sport fishing along project access corridors. Effects caused by
erosion/sedimentation and pollutant spills may impact aquatic habitats
such as tributary streams and lakes within the Susitna basin. At this
time, proposed peripheral construction activities are described only
generally; specific details of routing or siting and specific
construction techniques would be required before specific analyses can
be completed. Potential impact areas and the related aquatic resources
are identified in the following discussion.
POTENTIAL IMPACT AREA~ AND RELATED AQUATIC RESOURCES
Access Corridors
The proposed access road to the dam sites would depart from the
Denali Highway near Seattle Creek and proceed south to the Susitna River
below the Hatana Creek confluence, then traverse either the north or
south side of the Susitna River to the Devil Canyon darn site. A rail
extension from Gold Creek would be added for construction of Devil
Canyon facilities (ACRES 1982c).
The access road corridor from Denali to Watana would be approxi-
mately 40 miles long and crosses or parallels at least 37 streams and
rivers in both the Nenana and Susitna river drainages. Major streams
-5R-
crossed or paralleled in the Nenana drainage would be Lily Creek,
Seattle Creek, and Brushkana Creek (ACRES 1982a). These streams support
populations of grayling, northern pike, whitefish, burbot, and sculpin
(Figure 10). Deadman Creek is the major system in the Susitna drainage
that would be affected by the Watana access road. It is considered
prime grayling habitat and also contains populations of longnose sucker,
sculpin, and burbot. Between the Watana and Devil Canyon dam sites, the
access road would cross Tsusena and Devil creeks. Tsusena Creek
contains grayling, whitefish, burbot, and sculpin. The 16-mile-long
railroad line between Devil Canyon and Gold Creek would cross or
parallel six streams, including Jack Long Creek and Gold Creek. Jack
Long Creek contains small populations of chinook, silver, chum, and pink
salmon. Gold creek contains small populations of chinook, silver, and
pink salmon.
Transmission Corridor
Transmission lines would be built from Watana and Devil Canyon
Gold Creek and from there connect into the power houses to
Anchorage-Fairbanks intertie. From Watana to Gold Creek the
transmission line route would straddle the south side of the Susitna
River. This recommended route is approximately 40 miles long and would
cross the Susitna River and 17 small tributaries including Fog Creek,
Jack Long Creek, and Gold Creek (ACRES 1982a). Fog Creek contains
grayling, burbot, sculpin, and Dolly Varden. Jack Long and Gold creeks
contain small populations of salmon.
Gravel Removal Sites
Floodplain and upland gravel mining has the potential to adversely
affect aquatic habitats from related erosion and sedimentation problems.
The extent of this effect would depend on the location of these sites
(Figure 10). The alluvial fans at the mouth of Tsusena Creek, Cheechako
Creek, and mainstem Susitna River have been proposed as material sites.
Tsusena Creek would be rehabilitated but not the Cheechako Creek and
Susitna River sites because they would be inundated by the reservoir
(ACRES 1982a). Tsusena Creek contains grayling, whitefish, sculpin, and
-59-
Figure 10. Aquatic resources present in potentially affected areas from construction
activities for the Susitna Hydroelectric Project.
Potential
Construction Access Transmission Gravel Habitation Pollutant Sport
Species Present 1 Activity Corridors Corridor Removal Sites Areas Clearing Spills Fishing
WATER BODY
Susitna River X X X X X X X HR, SU, H'F, GR, DV
Lily Creek X X X X GR,RR,WF,NP,SC
Seattle Creek X X X X GR,RR,WF,NP,SC
Brushkana Creek X X X X GR, RR ,"HF, NP, SC
Deadman Creek X X X X X r,R,SU,BB,SC
Tsusena Creek X X X X X X GR,WF,BB,SC
Devil Creek X X X X 2
I Fog Creek X X X X GR,BB,SC,DV
0'\ Gold Creek X X X X X KS,SS,PS 0
I Jack Long Creek X X X X X KS,SS,CS,PS
Cheechako Creek X X X X X X KS,GR,DV
Watana Creek X GR,BR,WF,SU,SC
Kosina Creek X GR,BR,WF,SU,SC
Jay Creek X GR,BB,WF,SU
Goose Creek X GR,RR,SU,SC
Oshetna River X GR,BB, 't-7F, SU, SC
1. Chinook (king) salmon KS Arctic grayling GR Slimy sculpin sc
Coho (silver) salmon ss Northern pike NP Whitefish HF
Chum (dog) salmon cs Longnose sucker su Dolly Varden DV
Pink (humpback) salmon PS Burbot BR
2. No data available.
burbot. Chinook salmon, grayling, and Dolly Varden are found in the
lower portion of Cheechako Creek.
Habitation Areas
During construction of Watana Dam a construction camp and permanent
village would be located near the dam site. Each development would
occupy approximately 170 acres. The water source for both camps and
villages would be Tsusena Creek. Wastewater effluent would be
discharged into Deadman Creek. During construction of Devil Canyon Dam,
both a construction camp and village would be located about a mile from
the dam site. Water would be drawn from the Susitna River and the
effluent from a biological lagoon system discharged back into the river
(ACRES 1982a). Burbot, sculpin, and longnose sucker may occupy the
Susitna River in these areas.
Spills
Toxic pollutants could be spilled into any water body along access
corridors, near camps, fuel depots, and related facilities.
Clearing
In addition to the vegetation clearing activities to take place
along the access and transmission corridors, a major clearing operation
is proposed for the reservoirs. The Watana Reservoir would inundate 54
miles of mainstem and 28 miles of tributary habitat. Portions of six
major tributaries would be cleared of timber including Deadman, Watana,
Kosina, Jay, and Goose creeks and the Osetna River. These tributaries
are prime grayling habitat and also contain populations of burbot,
whitefish, longnose sucker, and sculpin (Figure 10). Impacts from
vegetation clearing on
actual inundation from
these tributaries would be secondary
the reservoir. Areas and methods of
removal may affect water quality in local stream courses.
Sport Fishing
to the
debris
Operation of the camps and villages would increase access to waters
that previously experienced little sport fishing pressure. Potentially
affected would be Deadman, Tsusena, Jack Long creeks, and stretches of
-61-
the mainstem Susitna River. Deadman and Tsusena creeks contain
substantial grayling populations as well as longnose sucker, burbot, and
sculpin. Jack Long Creek has small populations of chinook, coho, chum,
and pink salmon. Major species of mainstem Susitna are burbot, longnose
sucker, and whitefish.
GENERAL IMPACT PROBLEMS
Construction activities associated with the Susitna hydroelectric
project could result in the introduction of sediment or pollution
products into aquatic systems within the basin. These products could
directly affect the fisheries resources present in these aquatic
systems. The potential for erosion or pollution would vary with the
types of construction techniques, the nature of local surficial
materials, the topography at and surrounding specific construction
sites, and the timing of the activities. The follm.ring potential
impacts could result from construction and operation of the project and
will be addressed by AEIDC when specific site and methodological
information become available.
Erosion and Sedimentation
Wherever soil erosion takes place, the soil material breaks up and
is carried away by water runoff. Coarse sediments may not be carried
far before being deposited again, but fine-grained sediments,
principally silt or clay particles, are carried in suspension for long
distances and usually end up in local runnels and brooks that feed major
streams. Thus, silt often finds its way into anadromous fish streams as
far as several mi]es away from the erosional source. Sedimentation can
affect development of fish egRs and benthic organisms as well as causing
changes in species composition.
the literature (ACRES 1982a).
These effects are well documented in
The following construction activities have the potential for
causing erosion and subsequent stream sedimentation.
Access Corridors and Habitation Areas. Construction
areas, construction camps, and habitation sites
removal of some surface vegetation, cutting and
-62-
of roads, staging
generally requires
grading of slopes,
filling of depressions, and sometimes a surface gravel pad or roadway.
Areas underlain by permafrost are generally covered with an insulating
surface gravel layer to prevent thaw slumping, and culverts can be
placed to route runoff through the area.
Removal of the organic surface layer exposes mineral soils to
erosion from surface runoff and wind. Desiccation of exposed soils can
increase erosion potential. Proper culvert placement is important to
prevent local runoff from crossing and eroding the surface. Exposed cut
faces become prone to erosion unless stabilized by vegetation. Areas
with permafrost, especially ice-rich fine grained soils, are subject to
severe subsidence and erosion once the organic surface cover is removed
unless insulated by gravel pads. Exposed fill slopes are highly
erodable unless protected by vegetation.
Transmission Line l.orridor. Transmission line corridors are generally
cleared of any timber tall enough to fall on the lines. Felled timber
is either left in place or skidded off the corridor. Roads or trails
are usually constructed to provide access for vehicles and equipment for
timber removal and other construction activities. These can vary from
gravel roads that meet secondary road standards, to rough roads graded
through the surface organic mat, to surface trails wj th no grading
involved and low vegetation left in place. Surface trails often require
all-terrain vehicles with large tires or tracks. In some instances
helicopter access is required to reduce surface disturbance.
Clearing of vegetation, log skidding, and movement of vehicles on
slopes could induce long-term processes that would eventually trigger
severe erosion and mass wasting. Root systems of trees and other vege-
tation serve as cohesive binders within the soil, providing about 25
percent of the strength of the soil mass. If roots penetrate completely
through the soil zone, they often anchor directly into cracks in the
rock substrate, increasing their stabilizing influence.
Removal or destruction of surface vegetation exposes bare mineral
soil to the direct effects of surface runoff and destroys the mechanical
stabilizing effect of root reinforcing and anchoring within the soil
mantle. Soils then become more susceptible to soil mass movement and
gullying. A marked decrease in soil stability may not become noticeable
-63-
for several years after vegetation clearing because roots of destroyed
surface vegetation progressively deteriorate over time.
Clearing of trees from the transmission line corridor may cause
wind channeling and windthrow problems during storms. Channeled winds
can cause blowdown of trees along the margins of the corridor,
increasing the corridor width and increasing soil erosion through
further destruction of root binding in the soil. Where the organic mat
has been removed or damaged, exposing bare mineral soil, wind channeling
also increases soil desiccation in the corridor and dry soils are more
prone to erosion.
Removal of streambank vegetation cover can also affect water tem-
perature by exposing streams to direct sunlight. The insulating effect
of the riparian vegetation is of primary importance in maintaining
acceptable stream temperatures in small streams that serve as nursery
areas for small fish. Streams with south-facing drainage basins are
more likely to experience stream temperature increases and possible
dewatering during the periods of high solar energy input. Removal of
bank cover could also increase the exposure of fish to predators and
lead to a change in the population (Joyce, Rundquist, and Moulton 1980).
Gravel Removal Sites. Gravel will be required for many purposes in
construction of the Susitna hydroelectric project. For example, gravel
is used for surface fill material in the construction of habitation
areas, building pads, and construction camp and staging area pads.
Roadbuilding requires gravel to fill depressions, insulate ice-rich
frozen ground, and construct the surface course roadway.
Gravel is generally obtained from pj_t mines in river floodplains
and upland gravel deposits. Bulldozers and backhoes extract the gravel,
which is then transported by dump trucks. Floodplain gravel is usually
removed from areas away from active river channels, but these areas may
be inundated during floods. In upland areas surface soil is usually
removed and either disposed of or stored for eventual reclamation of the
site.
Gravel mining in floodplains can cause
bidity in the river system resulting fr
transportation of the resource. The p
-64-
direct inducement of tur-
-,ining operation and
gravel pits within
floodplains can alter sedimentation patterns within the floodplain
during flood periods and alter stream channel morphology. Pending of
water in borrow pits can trap fish.
Gravel mining in upland areas can induce erosion of exposed mineral
soils in the removal site and in the bounding walls. Erosion of
bounding walls can cause slumping and extension of the site,
exacerbating erosion problems. If an eroding borrow site should
intercept local stream courses, fish could become trapped in the flooded
site. If top soil removed from the site is not properly disposed of as
spoil or stored for eventual reclamation, it becomes susceptible to
erosion.
Pollution
Water Use. Construction and operation of camps and related facilities
could impact aquatic resources in several ways. As part of these
activities, water would be diverted from area streams or lakes for dust
control, concrete hatching, and gravel washing among other construction
uses as well as for domestic use in the camps and villages.
Potential impacts would primarily be caused by increased turbidity
due to erosion and discharge of effluent from concrete hatching and
gravel washing operations. Prolonged turbidity can reduce the producti-
vity of a system and cause emigration of fish populations. Fish could
also be impinged or entrained by improperly designed water intakes.
The extent of any potential impact from domestic water use depends
on the treatment of sanitary waste. v1astewater effluents can affect the
water quality of fish habitat by changing the BOD; however, point of
discharge and type of treatment control the extent of impact. The
effluent is not expected to cause any degradation of water quality as a
secondary wastewater treatment facility is proposed to treat all
wastewater prior to its discharge (ACRES 1982b).
Spills. Contamination of water courses from accidental spills of
hazardous materials is a major concern. Spills during major
construction projects commonly occur as a result of equipment repair,
refueling, and accidents~ Substances used in large quantities, like
fuels and oils, would be most likely to be involved and then other
-65-
materials such as solvents, antifreeze, hydraulic oil, grease, and
paints. If more than 10,000 gallons are stored at a site (common for
large projects) the contractor would be required to file a spill
prevention, containment, and countermeasure (SPCC) plan with the U.S.
Environmental Protection Agency.
Spills are generally short-term events but can have severe impact,
depending on the substance spilled, quantity spilled, the season,
species and life stages present, and the clean-up capabilities
available. Incubating eggs and alevins are the fish life stages most
likely to be affected by spills because adults and juveniles usually
leave the affected area. Aromatic compounds in oils are particularly
toxic, and there is a great deal of literature describing the
deleterious effects on aquatic life caused by petroleum products
(ACRES 1982a).
Clearing. Within the dam, spillway, and impoundment areas for Watana
reservoir more than 12,000 ha of vegetation would be removed by
construction and clearing operations, and more than 2,000 ha would be
removed for the Devil Canyon reservoir (ACRES 1982a). An additional
loss of approximately 300 ha of mixed vegetation types would be lost to
access roads as well as about 2,000 ha to the camps and related
facilities (ACRES 1982a). Unforested or sparsely forested locations
would be utilized as much as possible for features located outside the
impoundment area.
Vegetation/timber sJash and debris from the reservoir would be
stockpiled and burned over a three-year period during winter. Clearing
would be confined to the area to be inundated during each following
year. (Merchantable timber is not believed to occur in sufficient
quantities to remove for sale.) Depending on the location of the
burning, a potential impact on water quality is the possibility of large
quantities of ash entering the lake or river system, especially during
breakup.
Indirect Impacts
Sport Fishing. Operation of the camps would result in increased access
to an area that has previously experienced little human pressure. Those
-66-
portions of streams and lakes that are easily accessible from camps and
roads would be subjected to increased fishing pressure. Studies on the
streams in the proposed construction area indicate a relatively high
percentage of older age class grayling (ADF&G 1981). Sport fishing may
remove these larger older fish, resulting in a change in the age and
size distribution of the population.
IMPACT ANALYSIS APPROACH
Analysis of the potential effects of erosion or pollution products
into aquatic systems of the Susitna River basin is necessary to predict
how aquatic resources in the basin might be affected. Heretofore, these
topics have been addressed only in generic, qualitative terms since
specific sites and methods of construction for access roads, gravel
removal sites, transmission line corridors, habitation areas, and spoil
disposal sites have not been identified. Descriptions and discussions
of such anticipated impacts have so far addressed only the types of
impacts that have occurred in similar projects or are likely to occur
under the various developmental stages of this project. Those
discussions have represented the collective understanding of the
physical processes, habitat relationships, and likely response of
fishery resources but have necessarily been speculative in nature
(ACRES 1982a).
Once details of construction sites, routes, and methods are known,
AEIDC will perform more site-specific analyses in an attempt to
determine potential impacts to specific watercourses or portions thereof
and their effects on fisheries habitat suitability and change. These
analyses will include the following components.
1. Hork with the engineers during the final design stages to
become thoroughly familiar with planned project facilities and
structure, their siting, and proposed construction methods and
techniques. Consider design changes or modifications as they
occur and incorporate them into the analysis.
2. Perform a comprehensive review of published and unpublished
information and data on the effects of erosion and pollution
resulting from similar engineering projects on aquatic and
fisheries resources.
-67-
3. Consult with experts who have knowledge and experience related
to specific erosion and pollution problems. Work closely with
other agencies involved in data collection for the Susitna
hydroelectric project in order to collect necessary data and
to survey other professional opinion.
4. Perform site-specific field reconnaissance of all construction
sites to define specific potential erosion and pollution
problem areas. Determine types of local soils, slopes,
aspects, vegetation cover, and surface runoff patterns to
estimate the vulnerability to erosion and potential routing of
erosion and pollution products toward watercourses. Determine
specific watercourses and fisheries resources that might be
impacted by specific erosion and pollution problems.
5. Perform site-specific analysis of potential erosion and pollu-
tion problems, quantifying these effects where possible. Make
a sensitivity matrix to illustrate:
a. potential site-specific or method-specific erosion or
pollution problems with each potential problem ranked as
low, medium, or high probability of occurrence.
b. local watercourses or portions thereof that might be
affected by erosion or pollution problems identified
above, and rank potential effects as low, medium, or high
severity.
c. fisheries resources in local watercourses that might be
impacted and rank potential impacts as low, medium, or
high.
6. Assess the effects of human fishing pressure on project area
streams that will result from increased access to the area.
Identify the most likely areas to be impacted by increased
fishing pressure and the fisheries resources present. It will
be difficult to quantify these effects. It may be necessary
to estimate preproject fish populations (presently there is a
paucity of data on fish population estimates in streams that
may be affected) and compare to a possible postproject
population scenario in the absence of any altered fishing
regulations.
-68-
BJP.LTOGRAPHY
Acres American, Inc. 1982a. FERC license application, exhibit E. -Vol.
2, Chap. 3. Fish, wildlife, and botanical resources. Draft Report.
Alaska Power Authority. Susitna Hydroelectric Project. 1 vol.
1982b. FERC license application, exhibit E -Vol. 1, Chap. 1.
General description of the locale; Vol. 1., Chap. 2. Water use and
quality. Draft Report. Alaska Power Authority. Susitna Hydroelectric
Project. 1 vol.
1982c. Access plan recommendation report. Alaska Power Authority.
Susitna Hydroelectric Pro~ect. 1 vol.
Alaska Dept. of Fish & Game. 1983a. Susitna hydro aquatic studies, phase
2 basic data report. Vol. 3. Resident and juvenile anadromous fish
studies below Devil Canyon, 1982. Preliminary Draft Report.
Anchorage, AK. Alaska Power Authority. Susitna Hydro Aquatic
Studies. Report for Acres American, Inc. 277 pp.
1983b. Susitna hydro aquatic studies, phase 2 basic data report.
Vol. 2. Adult anadromous fisheries studies, 1982. First Draft
Report. Anchorage, AK. Alaska Power Authority. Susitna Hydro
Aquatic Studies. Report for Acres American, Inc. 2 vols.
1983c. Susitna hydro aquatic studies, phase 2 basic data report.
Vol 4. Aquatic habitat and instream flow studies, 1982. Preliminary
Draft Report. Anchorage, AK. Alaska Power Authority. Susitna Hydro
Aquatic Studies. Report for Acres American, Inc. 7 vols.
1982. Aquatic studies program. Phase T. Final Draft Report.
Anchorage, AK. Alaska Power Authority. Susitna Hydro Aquatic Studies.
Report for Acres American, Inc. 1 vol.
1981. Resident fish investigation on the upper Susitna River.
Final Draft Report. Anchorage, AK. Alaska Power Authority. Susitna
Hydro Aquatic Studies. Report for Acres American, Inc. 1 val.
Rovee, K.D. 1982. User's guide to
Methodology. Office of Biological
Service. Fort Collins, CO. 1 val .
the Instream
Services, U.S.
Flow Incremental
Fish & Wildlife
. 1975. The determination, assessment, and design of instream value ---studies for northern Great Plains regions. University of Montana,
Missoula, Mt. 109 pp.
Rovee, K.D., and T. Cochnauer. 1977. Development and
weighted criteria, probability-of-use curves for
assessments, fisheries. Cooperative Instream Flow
Office of Rio logical Service, U.S. Fish & Wildlife
Collins, co. 4Q pp.
69
evaluation of
instream flow
Service Group,
Service. Fort
~ovee, K.D., and R.T. Milhous. 1978. Hydraulic simulation in instream
flow studies: theory and techniques, Instream Flow Information Paper
5. Fish & Wildlife Service. Fort Collins, CO. FWS/OBS 78/33.
Bredthauer, S., and B. Drage. 198?.. River morphology. R & M Consultants
Inc., Anchorage, AK. Alaska Power Authority. Susitna Hydroelectric
Project. Report for Acres American, Inc. 1 vol.
Chow, V. T. 1965. Open channel hydraulics. McGraw Hill, New York, NY.
1000 pp.
Dwight, L.P. 1981. Susitna hydroelectric project--review of existing
water rights in the Susitna River basin. Report for Acres American,
Inc. 1 vol.
Hoppe, R. A. ,
Colorado
Studies.
and LM.
1969-70.
12 pp.
Finnel 1970. Aquatic studies in Frywipan River,
Bureau of Sport Fish. Division of River Basin
Joyce, M.R., L.A. Rundquist, and L.L. Moulton. 1980. Gravel removal
guidelines manual for arctic-subarctic floodplains. U.S. Fish and
Wildlife Service, Biological Services Program, FWS/OBS-80/09. 169 pp.
}fain, R.B. 1978. IFG-4 program user's manual. Unpublished manuscript.
U.S. Fish & Wildlife Service. Fort Collins, CO. 1 vol.
Milhous, R.T., D.T. Wegner, and T. Waddle. 1981. User's guide to the
physical habitat simulation system (PHARSIM). Cooperative Instream
Flow Service Group, Office of Biological Services, U.S. Fish &
Wildlife Service. Fort Collins, CO. Instream Flow Information Paper
11. 254 pp.
R & M Consultants, Inc. 1982a. Ice observations 1981-82. Anchorage, AK.
Alaska Power Authority. Susitna Hydroelectric Project. Report for
Acres American, Inc. 1 vol.
1982b. Hydraulic and ice studies. Anchorage, AK. Alaska Power
Authority. Susitna Hydroelectric Pro.;ect. Report for Acres American,
Inc. 1 vol.
198la. Ice observations 1980-81. Anchorage, AK. Alaska Power
Authority. Susitna Hydroelectric Project. Report for Acres American,
Inc. 1 vol.
1981b. Hater quality annual report 1981_. Anchorage,
Power Authority. Susitna Hydroelectric Project. Report
American, Inc. 1 vol.
AK. Alaska
for Acres
Simons, D. B., et al., eds. 1981. Proceedings of the workshop on downstream
river channel change·s resulting from diversions on reservoir
construction. Offic~ of Biological Services, U.S. Fish & Wildlife
Service, Washington, D.C. 346 pp.
70
Spence, L.E. 1975. Guidelines for using Water Service Profile program to
determine in stream flow needs for aquatic life. Preliminary Draft.
Montana Fish & Game Dept. (Environmental and Information Di), Helena,
MT. 22 pp.
Stalnaker, G.B., and J.L. Arnette. 1976. Methodologies for the
determination of stream resource flow requirements: an assessment.
U.S. Fish & l.Jildlife Service. Logan, UT. 199 pp.
Tennant, D.L. 1976. Tnstream flow regimes for fish, wildlife, recreation,
and related environmental resources. Pages 359-373 in J. F. Orsborn
and C.H. Allman, eds. Instream Flow needs. Western Division of
American Fisheries Society and Power Division of American Society of
Civil Engineers.
Theurer, F.D., and K.A. Voos. 1983. Instream water temperature model.
Draft. Cooperative Instream Flow Service Group, Office of Biological
Services, U.S. Fish & l·Jildlife Service. Fort Collins, CO. In stream
Flow Information Paper 16. 263 pp.
Trihey, E. W. 1982 a. 1982 Winter Temperature Study, February 24-25 trip
report. Unpublished. 24 pp.
1982b. Preliminary assessment of access by spawning salmon to side
slough habitat above Talkeetna. Draft Report. Report for Acres
American, Inc. Anchorage, AK. 24 pp.
1981. Using time series streamflow data to determine pro.i ect
effects on physical habitat for spawning and incubating pink salmon.
Paper prsented at Symposium on the acquisition and utilization of
aquatic habitat inventory information, Portland, OR. October 28-30.
Amer. Fish Soc. 9 pp.
U.S. Bureau of Reclamation. 1968. Guide for application of water surface
profile computer program. Sedimentation Sect., Hydrology Branch.
U.S. Bureau of Reclamation, Denver, CO. 31 pp.
U.S. Corps of Engineers. 1976. HEC-2 water surface profiles. Hydrology
Engineering Center, U.S. Army Corps of Engineers. Davis, CA. 59 pp.
U.S. Fish and Wildlife Service. 1980. Habitat evaluation procedures
(REP). Division of Ecological Services, Washington, D.C. ESM102.
n.s. Forest Service. 1973. Method for recommending stream flows to meet
national forests needs (critical area method). U.S. Forest Servfce.
6 pp.
Wesche, T.A., and P.A. Rechard. 1980. A summary of instream flow methods
for fisheries and related research needs. Laramie, WY. Eisenhower
Consortium Rulletin no. 9. 122 pp.
71