Loading...
HomeMy WebLinkAboutAPA862SUSITNA HYDROELECTRIC PROJECT FEDERAL ENERGY REGULATORY COMMISSION PROJECT No. 7114 STREAM FLOW AND TeMPERATURE MODELING IN THE SUSITNA BASIN, ALASKA UNIVERSITY OF ALASKA ARCTIC ENVIRONMENTAL INFORMATION AND DATA CENTER UNDER CONTRACT TO [}{]~~~~=~[ID~@©@ SUSITNA JOINT VENTURE FINAL REPORT SEPTEMBER 1983 DOCUMENT No. 862 .____ALASKA POWER AUTHORITY_-J SUSITNA HYDROELECTRIC PROJECT STREAM FLOW AND TEMPERATURE MODELING IH THE SUSITNA BASIN r ALASKA Report by Arctic Environmental Information and Data Center Under Contract to Harza-Ebasco Susitna Joint Venture Prepared for Alaska Power Authority Final Report September 1983 Document No. 862 - - .... -I I r .... .... NOTICE ANY QUESTIONS OR COMMENTS CONCERNING THIS REPORT SHOULD BE DIRECTED TO THE ALASKA POWER AUTHORITY SUSITNA PROJECT OFFICE i I r I r I I .... i .... .... - This report was prepared by the following AEIDC staff William J. Wilson, Principal Investigator Ken A. Voos, Ph.D., Environmental Engineer Paul R. Meyer, Hydrologist Lynn D. Leslie, Climatologist Beverly J. Valdez, Information Coordinator ,... r - """' .... r TABLE OF CONTENTS PAGE NO. LIST OF FIGURES. • . • • • . . • . . • • . . • • • . • • • • • • • . . • • • • • • • • • • • • • • • . • • • • • • • • • vi AC'KNOWLEDGEM.ENTS •••••••••••••••••••••••••••••••••••••••••••••••••••• viii INTRODUCTION ••••••••••••••••••••••••• ·-• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1 WATER BALANCE ACCOUNTING FOR THE SUSITNA BASIN...................... 3 N ('t') \1') r-~ ({.) ('I) 0 0 0 I!) I!) ~ ('t') INTRODUCTION •••••••••••••••••••••••••••••• DESCRIPTION OF THE WATER BALANCE MODEL •••• METHODS TO APPORTION SUB-BASIN WATER •••••• Method I. Method II. Linear Watershed Area Contribution •• Areal Precipitation Weighting. Method III. Water-Yield Weighting •••• TESTING THE C COEFFICIENTS •••• C COEFFICIENTS TEST RESULTS ••• USE OF RELATIVE PRECIPITATION WEIGHTING •••• May through September •••• October through April •••• RESULTS AND DISCUSSION •••••••• 3 3 5 7 7 9 10 11 12 12 13 13 STREAM NETWORK TEMPERATURE SIMULATION MODEL ••••••••••••••••••••••••• 17 INTR.ODUCTION ••••••••••••••••••••••••••••••••••••••••••••••.••••• DESCRIPTION OF THE STREAM TEMPERATURE MODEL ••••• MODIFICATIONS ...••••.••••••••.•••••••.•.••••••••• STREAM. NETWORK ••••••••••••••••••••• STREAM. STRUCTURE •••••••••••••••• Topographic Shading •••.•••• Stream Widths •••••••••••••• Hydraulic Retardance ••• Tributary Assumptions •• HYDROLOGY •••••••••••• Flows ........... . Stream Temperatures •••••••••••• Observed Temperatures •••••••••••••••••••••• Synthetic Temperatures ••••••••••••••••••••• USGS Cantwell gage on the Susitna River 17 18 21 21 24 24 27 29 29 29 29 30 30 30 (RM 223. 7)........................ .. . . . . . . . . . . . . 30 USGS gage data collected on the Chulitna and Talkeetna rivers •••••••••••••••••••• Temperatures of Distributed Flow. 'METEOROLOGY ••••••••••••••••••••••••••••••••••••• Selection of Meteorologic Data ••••••••••• Ground Reflectivity and Atmospheric Dust. Meteorologic Predictions •••.•••.••.••••••••••••••• V ~!DATION •••••••••••••••••••••••••••.••••••••••••••• CALIBRATION ••••••••••••••••••••• RESULTS .AN'D DISCUSSION •••••••••••••••••••••.•••••••••••••••.•.• 33 33 39 39 40 40 46 51 54 ('t') --------~- iv TABLE OF CONTENTS (Continued) PAGE NO. FUTURE APPLICATIONS AND ENHANCEMENTS •••••••••••••••••••••••••••••••• 55 BIBLIOGRAPHY. • . . • • • • • • • • • • • • . • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 57 r APPENDICES A. TOPOGRAPHIC SHADING ••.••.•••••••••••••••• ' ••••••••••.•••..•• A-1 B. WIDTH/FLOW FUNCTIONS. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • B-1 C. LONGITUDINAL TEMPERATURE PROFILES, JUNE TO SEPT'EM:BER 19 81-19 82 . • • • • • • • • • • • • • • • • • . • • • • • • • • • • • • . • • • • • • • C-1 - - - r ! r - - v r ·I i' I I !""' - r - FIGURE NUMBER 1. 2. 3. 4. 5. LIST OF FIGURES PAGE NO. Flow balance sub-basins, Cantwell gage to Sunshine gage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 4 Period of record for gage stations used in H20BAL......... 6 Calculated mean annual precipitation and water- yield values, Cantwell to Sunshine Basin.................. 9 C coefficients, Cantwell to Gold Creek Basin •••••••••••••• 11 Mean pre-and postproject June flow profiles, Watana to Sunshine Station, using precipitation-weighting water balance method....................................... 14 6. Mean preproject June flows, Watana to Sunshine Station, using three weighting methods.................... 15 7. Vertical air temperature distribution. Anchorage and Fairbanks 1968, 1969, 1970, 1980, 1981, 1982 •••••••••••••• 22 8. Vertical relative humidity distribution. Anchorage and Fairbanks 1968, 1969, 1970, 1980, 1981, 1982.......... 23 9. Stream network from Watana to Sunshine. • • • • • • • • • • • • • • • • • • • 25 10. Susitna mainstem reach definitions •••••••••••••••••••••••• 26 11. Tabular values of width function parameters ••••••••••••••• 28 12. Monthly stream temperatures, available data June to Sept. 1980, 1981, 1982 •••••••••••••••••••••••••••••••••••• 31 13. Monthly stream temperatures, usable data June to Sept. 1980' 1981' 198_2.......................................... 32 14. Temperature regression for Susitna River at Cantwell gage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 15. Temperature regression for Chulitna River at USGS gage •••• 35 16. Temperature regression for Talkeetna River at USGS gage ••• 36 17. Ground reflectivity and atmospheric dust coefficients, Matanuska Agricultural Experiment Station, Palmer .•••••••• 41 18. Susitna Basin observed air temperatures vs. tempera- tures predicted from Talkeetna data ••••••••••••••••••••••• 42 vi -' r I r - - r FIGURE NUMBER 19. LIST OF FIGURES (Continued) PAGE NO. Susitna observed humidities vs. humidities predicted from Talkeetna data.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 20. Average monthly wind speeds (M/S), 1980, 1981, 1982 ••••.•• 47 21. Predicted vs. observed solar predictions, 1980, 1981, 1982...................................................... 48 22. Tributary temperatures; 3 C groundwater inflow assumed •••• 49 23. Tributary temperatures; postcalibration, including distributed flow temperature model •••••••••••••••••••••••• 52 24. Temperature model calibration statistics for tributary predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 vii -I ACKNOWLEDGEMENTS This report is the result of a cumulative effort by a number of people. Special appreciation is extended to Gail Heineman (ADF&G) for facilitating necessary data transfers; Tim Quane and Jay Sauntner (both of ADF&G) and Carl Schoch (R&M) for sharing their on-site knowledge; Fred Theurer (IFG) and Woody Trihey for their technical guidance; and Jeff Coffin (R&M) for his continued assistance throughout the project. At AEIDC~ we would like to thank the Resource and Sciences staff for their ongoing support in all phases of this report; Information Services staff for library and bibliographic services; Judy Brogan for her editorial assistance; and Nancy Walters and Sherrie Cornett for the production of this report. viii r r : r r - r -' -I -i ! -I .... r -l ' INTRODUCTION The Alaska Power Authority, through Harza-Ebasco Joint Venture, contracted with the University of Alaska's Arctic Environmental Informa- tion and Data Center (AEIDC) to simulate postproject physical habitat conditions in the Susitna River drainage with a computerized model system. Water balance and stream temperature models permit the simu- lation of unmeasured water discharges and temperatures at various locations downstream from the proposed Watana or Devil Canyon dams (AEIDC 1983). These predictions are necessary for the analysis of project impacts on downstream fishery populations and habitats and will allow identification of appropriate streamflow regimes to minimize negative effects and aid mitigation efforts. Determination of stream temperatures requires flow data at various mainstem and tributary locations. This is the main purpose of the Susitna water balance model. Water temperature is important because it has various effects on fish behavior, including habitat selection, migration, movement patterns, food selection, and the physiological functions associated with growth and metabolism. It has a direct effect on the time required for salmonid egg development. Many studies have illustrated the relationship between small temperature change over long periods of time and salmonid egg incubation (Reiser and Bjornn 1979). Temperature has also been implicated as a factor affecting the timing of outmigration of smolts and inmigration by adult spawners (Brett 1971; Coutant 1970; Cherry, et al. 1975; Reiser and Bjornn 1979). These physiological and behavioral functions may be altered by temperature changes of as little as 0.5 to 1.0 C. For these reasons, it is important to predict downstream tempera- tures accurately and at the specific locations where fishery habitat may be affected. Tributary flows and temperatures also should be simulated so that the dilution or buffering effect of tributaries on the mainstem can be understood. Water balance and temperature predictions will also be critical to the river ice modeling efforts of Harza-Ebasco. This report is organized into three major sections. The first section describes the water balance model and hydrologic data synthesis. The second section provides a description of the stream temperature 1 - - P- I - ~ I model and how it was modified to more accurately reflect Alaska environ- mental conditions. It also includes an analysis of the temperature model's performance to date. The last section is a discussion of the future applications and enhancements of both the water balance and stream temperature models, including how they will be applied for esti- mating project effects. This report does not include actual estimates of project impacts. 2 r - - .... WATER BALANCE ACCOUNTING FOR THE SUSITNA BASIN INTRODUCTION The task of water balance accounting in the Susitna Basin is one of defining the methodology to assign inflows between known flows at mainstem gage stations. The lack of hydrometeorologic data in this region makes this a difficult task, subject to a number of gross assumptions. Three basin water apportionment methods have been explored and are discussed in this section. AEIDC developed a computer program to employ these apportionment methods, generating time series of flows at a number of mainstem and tributary locations within the Susitna Basin. Output files containing these flows are directly usable as input to the stream network temperature model (SNTEMP). DESCRIPTION OF THE WATER BALANCE MODEL The water balance accounting program, H20BAL, was designed to operate on the Susitna Basin between the USGS gages at Cantwell (Vee Canyon) and Susitna Station. AEIDC's initial modeling efforts focus on the reach from the Watana dam site to the USGS gaging station near the Parks Highway bridge at Sunshine. The Chulitna and Talkeetna river flows are incorporated into the system at the gage station on each river near Talkeetna. The basin between Cantwell and Sunshine Station was divided into 16 sub-basins (excluding the Chulitna and Talkeetna basins above their respective USGS gages) for the purpose of water apportionment. These basins center around the larger tributaries and are defined by drainage divides (Figure 1). They do not necessarily follow the watershed boundaries of any single stream, often including drainages of three or more streams. In most of the sub-basins, a node location on the mainstem river was chosen, representing the point source for all inflow to the mainstem. For the few sub-basins without a dominant tributary, inflow is linearly distributed along the adjacent mainstem reach. The accuracy associated with assigning flow within a basin between gage stations increases as the distance between gage stations decreases. Thus, it is advantageous to use as many data stations within the basin as are available. Gaps in historical data records exist at some 3 ] J (!) USGS GOQI Stat lon A Node locot ion Sub-Basin ldenllflcatlon Sub-Bosln Bliuodliry Dam 5ill FLOW BALANCE SUB-BASINS, Cantwell Gage to Sunshine Gage Scale I"' 8 miles 1 1 OQ Ill en OQ c (I) c" I c" Ill en 1-'• ::I en C"l Ill ::I ~ (I) ..... ..... OQ Ill OQ (I) ... - - - stations within the basin (see Figure 2 for historical flow data periods). Rather than discarding all data at a gage with occasional gaps, we used linear regression to fill them. The H20BAL program requires input data for the following USGS gage stations: Susitna River near Cantwell (Vee Canyon) Susitna River at Gold Creek Susitna River at Susitna Station Chulitna River near Talkeetna Talkeetna River near Talkeetna We used flows at the Yentna River gage for the period that they are available. For the present extent of simulation, flow data at Watana are preferable to those at Cantwell, and flows at Sunshine are used instead of those at Susitna Station. These additional stations provide for greater accuracy by effectively reducing the size of the basin under consideration. Usable statistically-filled 32-year data sets are available for the Cantwell, Watana, Devil Canyon, Chulitna, Talkeetna and Susitna Station sites (Acres 1983a). A filled data record is also available for Sunshine gage but was not used in H20BAL because of resulting flow deficits in the Gold Creek to Sunshine reach. These deficits occur when the sum of flows at the Gold Creek, Chulitna, and Talkeetna gages exceed the synthesized flow at Sunshine Station. The alternate method used to assign flows at Sunshine was to assume that the flow-per-unit-drainage-area contribution to the mainstem was the same for the Gold Creek to Sunshine basin as it was for the summed Gold Creek, Chulitna, and Talkeetna drainages. The limited accuracy of this method is acceptable considering this sub-basin comprises only 3.3 percent of the total drainage area defined at Sunshine. METHODS TO APPORTION SUB-BASIN WATER Once data records are collected or filled for the skeletal gage station network, interstation flows incrementally increase downstream by the following relationship: (1) 5 -~ 1 1 1 ] Figure 2. Period of record for gage stations used in H20BAL. Gage Station Water Years Beginning 1 70 '71 '72 1 73 '74 '75 1 76 '77 1 78 1 79 '80 '81 '82 Susitna near cantwell 5/61 6/80 Susitna at Watana 7/80 Susitna at Gold Creek 10/49 chulitna River 2/58 5/80 Talkeetna River 6/64 Susitna at Sunshine 5/81 Yentna River 10/80 Susitna at Susitna 10/74 Station I""" I - - ..... -I - .... .... where: and Q is the mean flow for the given period (L 3 /t), C is a fractional constant determined from some combination of watershed area, areal precipitation, and water yield esti- mates (decimal), s, s - 1 are subscripts referring to mainstem locations, numerically increasing for each sub-basin downstream, 1, 2 are subscripts referring to mainstem gage locations, numerically increasing downstream, L, t refer to dimensions of length and time respectively The node structure defining the network of sub-basins is fixed (nonvariable) within the water balance model. The different values of the C coefficients are selected as input options. We developed three different methods for determining values of the C coefficients. Method I. Linear Watershed Area Contribution--Acres ( 1982) used this method to determine flow series at proposed dam sites between the USGS gages at Cantwell (Vee Canyon) and Gold Creek. A sub-basin that drains 10 percent of the basin area between gage stations is consistently assigned 10 percent of the diffe~ence in flow between these two sites. The C coefficients are defined by:. c = s where: A s A is the planimetered area (L 2), and subscripts refer to sub-basin, s, and the total basin, b. (2) Method II. Areal Precipitation Weighting--The purpose of this method is to incorporate the weight of relative sub-basin precipitation into the C coefficients. These coefficients are now defined by: 7 - - - - .... .... 8 c = s where: p A s s P is the mean annual precipitation (L), and the subscripts and other variables remain as previously defined. (3) The methods employed to determine the mean annual precipitati·on for each sub-basin are important to note, since a great amount of subjectiveness is involved. The primary data source for the precipitation distribution was a statewide, annual precipitation isohyetal map prepared by James Wise (1977), Alaska state climatologist. This map is contoured in 10-in intervals for the 10-to 40-in annual precipitation range, and 20-in intervals above 40-in annual precipitation. These isohyetals were redrawn on a 1: 250,00Q-scale map of the Susitna Basin. Additional isohyetals were interpolated between each of the existing ones, resulting in 5-in contour intervals in the 10-to 40-in range, and 10-in intervals in areas with over 40-in of annual precipitation. To find average precipitation for each basin, we assumed that the total precipitation between two isohyetals could be estimated as the product of the area between the isohyetals (found by polar planimetry) and the average of the two isohyetal values. These products were summed for all of the intercontour areas within a sub-basin and then divided by the sub-basin area to determine the average annual precipitation. The same process was used to find the mean annual precipi- tation for the entire basin (Figure 3) • - Figure 3. Calculated mean annual precipitation and water-yield values, Cantwell to Sunshine Basin. Method III. Water-Yield Weighting--A report by Evan Merrell of the U.S. Soil Conservation Service (1982) suggests a third method for determining the C coefficients. In this report Merrell uses precipitation and evapotranspiration estimates to develop a mean annual water-yield map of the Susitna Basin. To incorporate the relative weights of sub-basin water yield estimates, the C coefficients are defined as: c = s _y A s s (4) 9 r where: Y refers to the mean annual water yield (L), and the remaining variables are as defined previously. The mean annual water-yield values for each sub-basin were determined in the same manner as the mean annual precipitation values. The water yield isopleths were redrawn on a base map of the basin, along with the sub-basin outline. The exception to note is that no isopleths were interpolated between those given by Merrell. Once again, polar planimetry was used to determine the areally-weighted basin water-yield values (refer to Figure 3). TESTING THE C COEFFICIENTS The C coefficients determined for any of the methods will sum to the value 1. 0 over the basin defined by two gage stations. A variety of basins can be defined within the area of concern by using different pairs of gaging stations. As previously discussed, increased accuracy results from using data at all available gage stations. The applicability of each method was tested by determining the three sets of C coefficients for the Cantwell to Gold Creek basin and applying these methods to the period for which historical records are available at the Watana dam site. The predicted values were then compared to the historical record at Watana. Figure 4 gives the C coefficients for the Cantwell to Gold Creek basin. Predicted flow at Watana is given by: (5) where: 10 subscripts w, c and g refer to Watana, Cantwell and Gold Creek respectively, and Q and C are as previously defined. The calculated C for each of the three methods is: w Method I Method II Method III 0.5104 0.6759 0.4636 .... .... .... ,.... I .... -I .... Figure 4 • C coefficients, Cantwell to Gold Creek Basin. c Coefficients Meth I Meth II Meth III Sub-Basin Name Area p y A p A y A s s s s s s s (mi 2) pb yb ~ pb~ yb~ Clarence 76.8 1.3660 .4741 .0383 .0524 .0181 Kosina 485.1 1. 4049 .8954 .2421 .3401 .2153 t•Jiatana 242.4 1.4439 .8837 • 1210 .1747 .1090 Deadman 218.4 .9979 1. 0971 .1090 .1087 .1212 Tsusena 191.5 .7248 1.2529 .0996 .0693 .1175 F'og 175.0 .9394 1.0428 .0873 .0820 .0891 Devil 174.5 .5603 1.0310 .0871 .0488 .0896 Chin-Chee 94.2 .4739 .8425 .0470 .0223 • 0399 Portage 186.4 .5018 1.2548 .0930 .0467 .1150 Indian 159.4 .6913 1.0358 .0796 .0550 .0852 The mean observed value of C for the 13 months of record when data were w collected at all three stations (Cantwell, Watana, and Gold Creek) was 0.6034, with a standard deviation of 0.1119. It is important to note that these data were collected during the June through November period, and may not be representative for the entire year. However, since approximately 82 percent of the annual flow occurs during this period (based on the 1950 through 1979 flow record at Gold Creek), this period of record appears adequate. C COEFFICIENTS TEST RESULTS One conclusion that can be drawn from this simple test is that none of the three methods show clear superiority. Based on the estimates of C , we preferred Method II, the relative precipitation weighting scheme, w for determining C coefticients; however, a couple of points concerning these three methods should be mentioned. One concerns the differences resulting from use of the linear drainage area method and the observed 11 r r ..... - .... - flows at Watana. Acres (1982), using a drainage area-based C value of w 0.515, calculated a synthesized mean annual flow at Watana of 8023 cfs (Acres 1983a). The observed C value of 0. 6034 applied to the same w . 32-year period results in an annual flow of 8338 cfs. This constitutes a 3. 9 percent increase in available water for the Watana reservoir. Though the magnitude of this increase seems insignificant, it indicates that any error would probably be on the side of underestimating water supply at Watana. Second, the wat.er-yield map used for Method III was developed to consider the smaller topographic features of the Susitna Basin, while the precipitation map used for Method II has considerably less topographic resolution. Consequently, greater utility would be expected from the increased sophistication of Method III. The water-yield map, however, apparently underestimates the contributions of the upper basin (Cantwell to Watana) substantially. In calibrating the map, Merrell was restricted to the available gage data at Cantwell and Gold Creek. If used on a small scale sub-basin such as Cantwell to Watana, Method III might prove to be much more accurate than Method II. However, the lack of flow data for the smaller tributaries presently makes this assumption untenable. USE OF RELATIVE PRECIPITATION WEIGHTING Method II accept~ the premise that the sub-basin watersheds contribute to mainstem flow in amounts relative to the distribution of mean annual precipitation. However, actual watershed conditions exhibit strong seasonal influences which must be considered. Consequently, the year was divided into two periods for application of this method • 12 May through September. Flow in the early part of this period (May through June) is dominated by the melt of winter precipitation. During July through September, when storm events contribute a large amount to tributary flow, the accuracy of this method depends on the matching of storm precipitation with average annual precipitation patterns. - - I""' ' October through April. Most tributary flow during this period is generated by groundwater baseflow; very little is a direct result of precipitation or of snowpack melting. Consequently, annual precipitation patterns are not used to weight relative basin contributions. For this period we have continued to use linear drainage area weighting (Method I). RESULTS AND DISCUSSION The water balance accounting model is largely a support program, providing input flows to other component models. As such, it operates on a specific scenario, generating an output flow time series for each nodal location in the system. To generate the postproject flow time series, H20BAL runs through two cycles. A time series at each node is first determined based on the natural input flows. Tributary contributions are determined in this step. The next cycle reassigns postproject output flows to the dam node and flows at the remaining mainstem nodes are re-adjusted. Figure 5, longitudinal profiles of the pre-and postproj ect mean June flow regimes, provides graphic representation of H20BAL output. Figure 6 gives tabular comparison of the three apportionment methods for the same preproject mean June flows. Since ·filled flow records for the 32-year period of simulation exist for the Talkeetna and Chulitna rivers, flow from these systems can be treated as point source inputs to the mainstem basin. The Yentna River, however, cannot. be treated in this way, except when simulating the period covered by the two-year gage record. When extending the water accounting system downstream from Sunshine, Yentna River flow must be apportioned as a fraction of the difference between Susitna Station and Sunshine Station gage flows. The decision to use the area weighting procedure in the lower basin, regardless of the method used in the upper basin (i.e., upstream from Sunshine Station), was based on the following considerations: 1. The large size of the Yentna Basin (6180 mi 2 ) makes the task of developing C coefficients for water-yield or precipitation weighting formidable. 13 Preproject Flows Postproject Flows ... -1 .l 1 -~ Flow (cfs) 70000 r--------------------------------------------------------------~ 60000 50000 40000 30000 20000 10000 190 f I I ~--~--~----~1 I I ----- 170 150 130 110 River Mile 90 ~ rt ::s: ,.. 0 CD rt Ill ~ til tj §"d O"Ulti Ill ::1' CD 1-' 1-'· I Ill tj t:' CD Ill n ;::I CDtllP... rt a ~» "d CD rt 0 rt 1-'• [/l P'Ort O:;l"d p..... ti • 0 ~ w. Ul CD 1-'· n tj rt OQ "d ~ ti tj CD CD n 1-'• l-1l "dl-' I-'• 0 rt ~ Ill rt"d 1-'• t1 0 0 tj l-1l I 1-'• ~ .... CD CD 1-'• [/l OQ .. ::r rt=<:l 1-'• Ill ;::I rt OQ ~ Ill r I i I I""' r I""' - ... Figure 6. Mean preproject June flows, Watana to Sunshine Station, using three weighting methods. --------Preproject Flows (cfs)-------- Location River Area Precip. Water-Yield Name Mile Weighting Weighting Weighting Watana 184.4 23034 23034 23034 Tsusena 181.3 23999 24056 24081 Fog 176.0 24844 25266 24876 Devil Canyon 161.3 25688 25986 25675 Chinchee 154.6 26143 26314 26031 Portage 148.8 27044 27003 27056 Indian 138.6 27815 27815 27815 Mckenzie 116.8 28543 28655 28687 Whiskers 101.4 28787 28917 28952 Chulitna 98.6 52359 52526 52589 Talkeetna 97.2 63916 64064 64103 Trapper 91.2 64117 64280 64291 Sunshine 83.8 64555 64555 64555 2. The lack of gage data for the Yentna River with which to calibrate makes any selection of a weighting scheme somewhat arbitrary. 3. The confluence of the Yentna River is at the downstream end of the Susitna Basin, far from the dam sites. Consequently, this is the region least sensitive to differences in flow apportionment methods. Enhancement of the apportionment methodologies might be undertaken in a number of ways. The relative precipitation weighting method could be improved by using monthly or seasonal precipitation distribution maps. Presently, however, these maps are not available. Kilday (1974) developed mean monthly precipitation maps for the State, but they do not have the resolution necessary to be used on the Susitna Basin • Sub-basin water yield would be determined most directly using Method III, relative water-yield weighting. Improvement of the present water-yield map is possible as additional precipitation and streamflow data become available. Continued enhancement could lead to monthly or seasonal water-yield estimates. 15 r r F i -I ..... -16 -I -! ..... I - STREAM NETWORK TEMPERATURE SIMULATION MODEL INTRODUCTION AEIDC selected the Stream Network Temperature Simulation model (SNTEMP), developed by the U.S. Fish & Wildlife Service (Theurer et al. 1983), for use in the Susitna simulation model. SNTEMP predicts average daily and daily minimum and maximum water temperatures at selected points within a river network. The model requires meteorologic, hydrologic, and stream geometry data to compute heat flux relationships and to transport heat through the system. Several features of SNTEMP make it particularly applicable for use in the Susitna system. 1. SNTEMP contains a temperature regression technique which allows use of incomplete or noncontinuous input temperature data. Much of the Susitna water temperature data are point measurements or incomplete records. SNTEMP contains a calibration technique which provides the ability to adjust low-confidence input parameters to obtain minimum prediction error. Daily average, maximum, and minimum water temperatures can be predicted for periods ranging from as short as one day to as long as one year (continuously variable in one-day incre- ments). Thus, short yet critical river reaches could be modeled in daily detail, but the full length of the system is simulated with longer averaging periods. For the Susitna system, SNTEMP has been configured to simulate mean monthly temperatures at any location between the Watana dam site and the Parks Highway bridge at Sunshine Station. The model utilizes either historical or synthetic hydrologic and meteorologic data. In this latter mode of simulation, referred to as "gaming," synthetic data are used to approximate temperatures during the construction and postproject phases of the proposed project. 17 r l r r - - - DESCRIPTION OF THE STREAM TEMPERATURE MODEL SNTEMP is a collection of several submodels: 1. a solar model which predicts solar radiation based on the latitude of the stream basin~ time of year, basin topographic characteristics, and prevailing meteorologic conditions; 2. 3. a meteorologic correction model accounting for changes in air temperature, relative humidity, and atmospheric pressure with elevation; a heat.flux model accounting for all significant heat sources and sinks; 4. a heat transport model to move the water and its associated heat content downstream; 5. a regression model for smoothing or completing observed water temperature data; and 6. a flow mixing model for merging tributary flows and heat content with those of the mainstem. A complete description for each of these components is provided in the model description/documentation available from the U.S. Fish and Wildlife Service (USFWS) (Theurer et al. 1983). A brief description of the heat transport model will be provided since it is this component~ more than any other, which determines the model's limitations. The heat transport model used in SNTEMP is based on the following dynamic temperature-steady flow equation (Theurer et al. 1983): (A/Q) (aT/at) + aT/ax= (qd/Q) (Td -T) + (BEH)/(Qpcp) !<-dynamic term-->1<------steady state equation---------->1 !<-----dynamic temperature-steady flow equation-------->1 where: A = flow area, L2 Q = flow, L 3 /t T = temperature, T t = time, t X = distance, L qd = distributed inflow, L2 /t Td = distributed inflow temperature, T 18 (6) r r ' r - - .... I - B = stream top width, L EH = net heat flux, (E/L 2 ) /t p = water density, M/L 3 c = specific heat of water, (E/M)/T p and dimensions are: M -mass T temperature L -length t -time E -energy The assumption of steady state (3T/at = 0) can be used to reduce the order of Equation (6) when 24-hour average temperature predictions are sufficient, resulting in: (7) It is significant that this equation does not contain a stream velocity term. SNTE'MP does not require stream velocities for prediction of average daily temperatures downstream from a known temperature. Dynamic temperature predictions are possible if steady state is not assumed. Equation (6) can also be solved by the method of characteristics (Theurer et al. 1983) which results in a solution identical in form to Equation (7). Dynamic temperature predictions require Equation (7) to be solved along the characteristic line equation as follows: dx = (Q/A) dt (8) The factor Q/ A is stream velocity. Dynamic temperature predictions require an estimate of stream velocity which SNTE'MP computes using Manning's equation. Closed form solutions of Equation (7) are obtained by assuming that 1) the flow is uniform within a reach and 2) a second order approximation of the heat flux is valid. This heat flux approximation can be expressed mathematically: 19 F l .... I I .... r I where: T = equilibrium temperature, T e (9) K1 =first order thermal exchange coefficient, [((E/L 2 )/t)/T] K2 =second order thermal exchange coefficient, [((E/L2 )/t)/T2 ] The equilibrium temperature is the theoretical temperature the stream would approach if all heat transfer processes were held constant with time. If the water reached equilibrium temperature, the rate of heat input to the water would equal the rate of heat loss (LH = 0). Equilibrium temperature and steady flow assumptions constrain the methods used to average input data. The input hydrologic and meteorologic conditions must be representative throughout the travel time from the initial to final points of the model network. If the travel time from the most upstream point to the downstream end of the network becomes significant compared to the data averaging time, then model prediction becomes less reliable. For example, assume that a 30-day meteorologic data averaging period has been selected and that it takes 30 days for water to travel from point A to point B. Water passing point B on the first day of this 30-day period left point A 30 days earlier. Therefore, the meteorologic conditions which determine the daily average water temperature at point B on the first day are not included in the time period averages. Only the last day's water column can be considered to have been influenced by the 30-day average meteorologic conditions. This data averaging versus travel time dilemma can be overcome either by 1) selecting averaging periods greater than the network travel time or 2) dividing the network into serially connected subnetworks, or reaches, and using moving average input conditions. The first technique is the standard way of operating SNTEMP. If short-term average water temperature predictions are necessary, the second technique can be accomplished with SNTEMP by simulating an upstream reach with appropriate average input data, and using this simulation's output as· input to the next downstream reach. 20 - - .... r MODIFICATIONS AEIDC modified SNTEMP to more accurately simulate conditions specific to Alaska and the Susitna Basin, including techniques to approximate the seasonal variation in canyon wall shading and winter air temperature inversions which normally occur in the Susitna River basin. The original design of SNTEMP assumed topographic shading to be constant. Since solar altitude angles are so acute in Alaska, resulting in extreme shading during the winter months, SNTEMP was modified to accept a monthly topographic shading parameter. SNTEMP originally featured a constant lapse rate to simulate air temperature and humidity change at elevations other than those where data were recorded. Radiosonde data from Fairbanks ana Anchorage indicated this approximation to be a poor predictor of actual conditions, especially in the colder months (U.S. National Weather Service 1968, 1969, 1970, 1980; World Meteorological Organization 1981, 1982). AEIDC modified SNTEMP to accept monthly, nonconstant lapse rates. Local monthly temperature lapse rates were determined by regressing temperature on elevation using data recorded above Anchorage and Fairbanks (1968 through 1970; 1980 through 1982) by U.S. National Weather Service balloons. The temperature lapse rate curves for June, July, August, and September are shown in Figure 7. Piece-wise linear humidity lapse rate curves were also determined from the balloon data and are presented in Figure 8. In addition, we also adjusted the normal SNTEMP operating method to accommodate the limited water temperature data available throughout the study area. Typically, a built-in regression model provides missing water temperature data and smooths the data but we had to bypass this feature since it required more data than were available at any of the water temperature collection sites. This will be discussed further in the section entitled "Synthetic Temperatures." STREAM NETWORK The stream network as defined for SNTEMP is designed to allow easy manipulation of flows and water temperatures at specific locations. This network can be used for simulations with either or both Watana and Devil Canyon reservoirs. Using expected water temperatures and outflows 21 ·-····· 1 l'%j JUNE f-'• N l!lfllAIUiE LJIS( &<It AUGUST ~ N 1£1QOI IM I(IIJS TEIIPERA TLIR£ LAPSE RATE t'i m HEIGHT IN METERS (D 31100 -..! -.. . ' 2Silll ••• •• >":Ej<l Ill (D ..... t'i IJIIll , 2BBB ..... t'i rt ... r:t f-'• Ill () Ilia •• ,jl. 15011 [~ IP D Ill -f-'· , 10011 \Ot'i ID D o 0'\ OOrt Sllll 1 a a a .. (D * D D -~ oo nl D. \O(D a 0'\t'i ... \Oil) ·18 ·8 ·6 ·4 -2 8 4 8 18 12 14 .. rt •II ... ... ·I ·Z . l II ll II TEMPERA lURE <DECREES Cl .... ~ IDfWIUi£ ~D \0 (D -..! OQ. .. f-'· Ill >-'rt \Ot'i 00 f-'• or:t JULY SEPTEMBER .. !::: rt T£Hf'UIA lURE LAPSE RAT£ lEMPERAIURE U.PSt: RATE l-'f-'o HEICIIT IH METERS HEIGHT IN METERS \00 3000 3008 00:;::1 -. 2500 • • ..., rl' • 2500 D Ill o I -> • • \0:;::1 00() ~ • , .. 2a08 N::r' • • •• • 0 t'i Ill J!il!B 0 "UDg •• • 1500 OQ !IIIII '116' . ... (D IBWI • • 10Ba •• § Q. 500 500 •• D •• • • ... • • •• B • ·10 -8 -6 -4 -2 II 12 14 -10 -s -6 -4 -z B 2 6 10 12 14 T~PERA TLIRE WE CREES D TEMPERA lURE WECREES Cl -, -~ >rJ JUNE AUGUST ..... OQ c::: llli:Hl IK IOEIERS 1£1Qtl IK IIEI£115 11 (1) ~ 3111111 00 •a'lo ~ D D llg Pi ...... <I \0(1) 0'\11 OOrt 2llOII w ..... D D n ...... jl) \0 I-' (1'\ 151111 15111 \0 11 .. (1) I-' ...... Ill \0 rt IIIII IIIli ....... ..... 0 < .. (1) DOD p ...... [ \0 00 0 ..... .. llo ..... 1-'rt u ll l2 u ... l5 u l7 ll ll 1.1 u ll l2 u 1!.4 l5 l6 L7 ... u I. i \0'-<l 00 II£Uilil: lliiiDITl, 11£111/.1. Rlllllii IIJIJOJIY, ll:CJMAL ...... llo ..... (I) ...... rt \0 11 JULY SE~TD·IBER 00 ..... N C" c::: l£lllil IK I(1{RS rt t(JQ,T IN II[IEI!S ..... lllU 0 l:S .. Zoill D rn a a'll til-lbo o a ' p ()' r:r' 0 a 1'1 Ill OQ (1) ISilll lb l'iilll • D ~ p. lllilll 11100 Do DO >rJ Ill ..... 11 dl aD 5011 a a C" N ~ w :>;' (I) ... ll lZ Ll ... ? g, 5 l6 L7 u u 1.0 u e. 1 i. 2 e.! 1,4 ·-~ u 1!.7 1.8 0.9 1.1 RllAfii'E IIJIIDIIY. ll:CIMAI. RElATIVE JUIJOIII. OEmi.L r - - """' ' r - .... from Watana reservoir (RM 184.4), the model predicts the water temperatures at any specific location downstream to Sunshine Station (RM 83.8). This network (Figure 9) is easily modified for simulations with flows and water temperatures at Devil Canyon. To obtain starting temperatures at Watana for validation and calibration simulations, we defined a 40-mi reach from the USGS Cantwell gage (RM 223.5) downstream to Watana. River mile distances are based on interpolations of maps in the Susitna River Mile Index (R&M 1981). Tributaries between Watana and Sunshine Station were included in the Susitna stream network. The flow and thermal contributions of smaller tributaries not explicitly included were estimated, and either incorporated into a nearby tributary or were linearly distributed to the neighboring mainstem reach. A more detailed description of these hydrologic approximations appears in the section "Water Balance Accounting for the Susitna Basin." STREAM STRUCTURE Segmenting the system network into reaches with similar physical characteristics (Figure 10) provided the physical model of the system. Reach selection was based predominately on orientation and local topo- graphy with consideration of significant slope change, width change, and elevation drop. Topographic Shading Topographic shading may significantly affect Susitna River tempera- tures especially in the winter months. The orientation of the reach and the elevation of surrounding canyon walls limits the amount of sunshine the stream surface receives. As previously mentioned, SNTEMP was modified to accept changes in stream shading for each month. The variable which defines the amount of shading is the sunrise/sunset altitude angle (a ). We chose a representative midchannel point in the s reach to compute this angle. A compass rose was centered on this point and terrain elevation versus distance transects collected from 30 degrees east of North to due South to 30 degrees west of North at 15 degree increments. Maximum terrain altitude angles were determined from 24 Figure 9. r I I TSUSENA 2!2l8.6 ~ DEVIL 174.3 .... PORTAGE 175.5 - INDIAN 159.6 -WHISKERS 113. 2 -CHULITNA 116. 5 -i f""' TRAPPER 115. 4 - - Stream network from Watana to Sunshine. WATANA DAM SITE 184.4 181. 3 176. f2l 161. 3 148. 8 138.6 U211. 4 98.6 97.2 91. 2 SUNSHINE RM 83.8 H FOG 195.9 CHINCHEE 163. 3 TALKEETNA 11212. 121 H refers to tributary head~ waters as defined in the stream network J refers to tributary junction with the mainstem Numbers refer to River Mile as interpolated from R&M River Mile Index (1981). 25 1 ~M I Rovermole lli.O <!) Reach _ ..... , ...... 0 r I - - each of these transects and then transferred to solar altitude versus bearing angle plots (Siefert 1981). We computed the average sunrise/sunset altitude angles for each month from these plots (Appendix A) • Stream Widths The quantity of radiative energy entering or leaving the stream is a function of the stream surface area. An estimate of the stream width is necessary for surface area determination. Mainstem wetted widths used in SNTEMP from the Talkeetna River confluence to Watana were determined from the R&M cross sections and REC-2 simulations (R&M 1982d). The stage-discharge relationships developed by ADF&G (1983) were not available when our width analysis was being performed. However, since the stage discrepancies noted between the R&M simulations and ADF&G observations would not result in significant width differences, we do not propose to modify the width functions at this time. Water surface widths simulated by R&M were measured from the cross section diagrams (R&M 1982d) and plotted as a function of flow (Appendix B). We calculated width/flow functions from these plots. Other methods were used to estimate top width for other mainstem reaches and tributaries. USGS (1980, 1981) observations at Cantwell, Chulitna, and Talkeetna provided some stream width and flow data. Width data at the Chulitna and Talkeetna gages were available for several flows. Several width measurements within a narrow range of flows provided a constant width estimate for the Susitna River between Cantwell gage and the Watana dam site. The width of the reach below the Chulitna junction to Sunshine Station was determined from transects collected by R&M (Coffin 1983). This width was also assumed constant with flow. Field personnel estimated widths of the tributaries (Sauntner 1983; Schoch 1983; Quane 1983) which were assumed constant with flow. Figure 11 presents width/flow functions in tabular form with graphic presentations in Appendix B. The plots present data points connected by line segments and the computed function. 27 """ Figure 11. Tabular values of width function parameters. ! r Stream Reach II Start End a b I (mile) (mile) ii r 184.5 179.5 98.26 0.1577 ,I Susitna 1 Susitna 2 ' 179.5 175.5 105.40 0.1708 r I Susitna 3 175.5 166.0 98.13 0.1820 u r Susitna 4 166.0 163.0 189.96 0. 077 4 " Susitna 5 163 .o 146.5 144.88 0.1005 !""" Susitna 6 146.5 142.5 98.15 0.1845 Susitna 7 142.5 124.0 13.16 0.4078 r Susitna 8 124.0 115.0 33.95 0.3117 -Susitna 9 115.0 99.5 29.77 0.3390 Susitna 10 99.5 83.8 1256 Tsusena 1 208.6 181.3 80 Fog 1 195.9 176.0 50 Devil 1 174.3 161.3 35 Chinchee* 1 163.3 154.6 25 Portage 1 175.5 148.8 60 F"' Indian 1 159.6 138.6 50 Whiskers 1 113.2 101.4 20 [""" I Chulitna 1 116.5 98.6 60.70 0.2086 Talkeetna 1 102.0 !""" 97.2 97.92 0.1761 Trapper 1 115.4 91.2 18 Values for "a" and "b" in the function width (feet) b If "b" = a . flow (cfs) • is undefined, "a" represents a constant width (feet). *A synthetic stream representing the combined Chinook and ~hechako tributaries. 28 r J r I r" I l r i I .... r ! - r Hydraulic Retardance SNTEMP does not require stream velocity estimates to predict average daily downstream water temperatures (see nDescription of the Stream Temperature Model"). On the other handJ daily minimum and maximum temperature predictions do require estimates of stream velocities. If daily maximum and minimum temperature estimates are desired later, it will become necessary to obtain the Manning's n values to compute stream velocities. Tributary Assumptions Except for the tributaries simulated Chulitna and by SNTEMP Talkeetna rivers, all Susitna are essentially self-starting. Simulation of these tributaries starts from their estimated headwaters where a constant headwater temperature of 0 C is assumed. Since the headwater flow is assumed to be zero, this seasonally constant initial water temperature is not critical (the heat content of zero mass would be zero, exclusive of the temperature assigned). Flow is added to these tributaries based on the flow balance schemes discussed in the section "Water Balance Accounting for the Susitna Basin." Predicted tributary temperatures are highly sensitive to the temperature assumed for distri- buted flow. Techniques for estimating these temperatures will be discussed in the section "Temperatures of Distributed Flow." Tributary widths were based on field estimates and lengths were measured from topographic maps with an opisometer. Each tributary in the model is assumed to be a single stream. For branched tributaries we estimated a sub-basin area-weighted average length. Tributary reaches were defined based on 300 m elevation drops. HYDROLOGY Flows As described in the section "Water Balance Accounting for the Susitna Basin," we investigated three types of flow balancing techniques for supplying flow estimates to the temperature model. These techniques are used both with historical flows and for gaming with reservoir releases. 29 r ~~ r I' ' !I"" II I r ,... I .... r I r Stream Temperatures Observed Temperatures. SNTEMP uses observed water temperature data of two types--initial water temperatures necessary for starting the model and validation/calibration water temperatures. Only three initial temperatures are required for the Watana (or Cantwell) to Sunshine Station simulations. These are Susitna River at Watana (or Cantwell), Chulitna River at the USGS Gage, and Talkeetna River at the USGS gage. The remaining observed water temperatures are essential in determining how well the mainstream and tributary temperatures are being simulated and in servirLg as a calibration target. Most of the validation/calibration temperature data for this study are being collected by ADF&G (1981, 1983); USGS (USGS 1980, 1981; Bigelow 1983) collected the three initial water temperatures. Unfortunately, most of these initial temperatures are unusable as a result of incomplete records or discrete sampling. Usable data are defined as those data which are complete for the month or, if not complete, synunetric around the middle of the month. Data which cluster evenly around the middle of the month should result in an unbiased measure of the monthly mean. Figure 12 presents the available data, and Figure 13 presents usable data collected for the June to September periods of 1980 through 1982. Data collected by USGS at Gold Creek were not used in this study since it had been observed that the temperature recorder was in the plume of the Gold Creek tributary (Trihey 1983) and thus not representative of mainstem flow. USGS recently relocated this temperature recorder, and future data provided by USGS and ADF&G should allow adjustment of the historical USGS data to be representative of the mainstream temperatures. Synthetic Temperatures. USGS Cantwell gage on the Susitna River (RM 223. 7). Stx·eam temperature data were recorded at the Cantwell gage during the 1980 and 1982 June through September periods. To verify downstream temperature predictions with stream temperatures observed by ADF&G (1981) and R&M (1982b), we estimated water temperatures at Cantwell for 1981. SNTEMP incorporates a regression technique for data filling, but, as discussed previously, more data than are available are necessary for this technique to produce physically reasonable results. 30 r-Figure 12. Monthly stream temperatures, available data June to Sept. 1980, 1981, 1982. i MainstemiTributary Number of Days I River Mile River name I description 1980 1981 . 1982 J J A s J J A s J J A s 10.1/0.5 Alexander Cr. 25 31 31 30 10.1 Susitna above Alexander Cr. 25 31 31 25.8 Susitna A., Su Station . 30 31 31 30 10 r 28.012.0 Yentna R. 26 31 31 14 I 28.0/4.0 Yentna A. 23 31 31 27 29.5 Susitna R. above Yentna R. 20 31 31 30 32.3 Susitna R. above Yentna R. 25 31 31 12 ,_ 40.6/1.2 Deshka A. 21 31 31 30 49.8/4.9 ••Deception Cr. near Willow 5 8 8 49.8111.6 **Willow Cr. near Willow 5 18 22 50.5/1.0 Little Willow Cr. 7 31 31 30 50.5 Susitna R. above Uttle Willow Cr. 7 31 31 24 61.2 Susitna R. above Kashwitna R. 2 27 -77.210.0 Montana Creek 19 24 1 n.5 Susitna R. above Montana Cr. 19 3 2 30 83.8 Susitna A., east shore-Parks Hwy. 20 14 30 83.9 Susitna R., west shore-Parks Hwy. 23 9 10 30 97.0 Susitna A.-LRX 1 17 97.215.0 **Talkeetna R. near Talkeetna 1 -- 97.0/1.0 Talkeetna R. 10 31 31 30 97.211.5 Talkeetna R. 17 31 30 98.5118.0 **Chulitna R. near Talkeetna 1 -27 30 3 20 98.6/0.5 Chulitna R. 11 17 20 98.6/0.6 Chulitna R. 17 10 25 -103.0 Susitna R.-TKA fish wheel 11 10 19 22 7 28 31 25 113.0 Susitna A.-LRX 18 25 31 30 120.7 Susitna A.-Curry 25 31 30 126.0 Susitna A.-Slough SA 4 31 30 126.1 Susitna R.-LRX 29 22 31 30 129.2 Susitna A.-Slough 9 4 31 24 130.8 Susitna A.-LAX 35 23 4 17 1"" 131.3 Susitna A. above 4th of July Cr. 15 31 30 26 136.5 ••s_IJsitna A. near Gold Cr. 30 31 31 30 8 25 29 12 30 136.8/0.0 Gold Creek 7 3 -138.6/1.0 Indian R. 23 31 4 28 138.6/0.1 Indian R. 10 25 14 138.7 Susitna A. above Indian R. 11 29 16 140.0 Susitna A.-Slough 19 5 13 140.1 Susitna R.-LRX 53 23 142.0 Susitna A.-Slough 21 4 29 4 31 30 148.8 Susitna R. above Portage Cr. 13 31 29 148.8/0.1 Portage Cr. 13 26 28 181.3/0.0 Tsusena Cr. 12 7 31 30 184.4 *Susitna A. at Watana dam site 30 -31 30 194.1/0.0 Watana Cr. 11 31 15 16 206.8/0.0 Kosina Cr 4 31 17 12 223.7 ··susitna A. near Cantwell ---27 31 31 22 231.3/0.0 Goose Creek 31 31 30 233.4/0.0 Oshetna Creek 31 31 30 *A&M gages -• *USGS gagt:S ' All others are ADF&G gages. 31 Figure 13. MOnthly stream temperatures, usable data June to r Sept. 1980, 1981, 1982. I; Mainstem/Tributary Number of Days River Mile River name I description 1980 1981 1982 J J A ·s J J A s J J A s ,~"""' 10.1/0.5 Alexander Cr. 18 31 31 26 10.1 Susitna above Alexander Cr. 18 31 27 25.8 Susitna A., Su Station 30 31 31 30 28.0/2.0 Yentna R. 20 31 31 28.0/4.0 Yentna A. 14 31 31 24 29.5 Susitna R. above Yentna R. 10 31 31 30 32.3 Susitna A. above Yentna A. 18 31 29 6 .... 40.6/1.2 Deshka A. 10 31 31 30 49.8/4.9 **Deception Cr. near Willow 2 49.8/11.6 •*Willow Cr. near Willow 13 4 .... 50.5/1.0 Little Willow Cr. 31 31 28 50.5 Susitna A. above Little Willow Cr. 31 31 10 61.2 Susitna R. above Kashwitna A. 22 -77.210.0 Montana Creek 6 17 77.5 Susitna R. above Montana Cr. 8 30 83.8 Susitna R., east shore-Parks Hwy. 8 30 83.9 Susitna A., west shore-Parks Hwy. 14 30 .... 97.0 Susitna A.-LAX 1 14 97.215.0 .. Talkeetna A. near Talkeetna ---- 97.0/1.0 Talkeetna R. -31 31 30 !""' 97.2/1.5 Talkeetna A. 14 31 30 I 98.5/18.0 ••chulitna R. near Talkeetna - - --24 30 10 I 98.6/0.5 Chulitna A. 3 12 98.6/0.6 Chulitna R. 14 18 !""' 103.0 Susitna R.-TKA fish wheel 17 13 21 31 16 113.0 Susitna A.-LAX 18 17 31 30 120.7 Susitna A.-Curry 17 31 30 !""" 126.0 Susitna A.-Slough 8A 29 30 126.1 Susitna R.-LRX 29 13 31 30 129.2 Susitna A.-Slough 9 31 20 130.8 Susitna A.-LAX 35 6 r 131.3 Susitna R. above 4th of July Cr. 31 26 22 136.5 • *Susitna R. above Gold Cr. 30 31 31 30 24 24 30 136.810.0 Gold Creek -138.6/1.0 Indian R. 16 31 138.6/0.1 Indian R. 17 8 138.7 Susitna A. above Indian R. 21 10 r 140.0 Susitna A.-Slough 19 140.1 Susitna A.-LAX 53 23 142.0 Susitna A.-Slough 21 28 31 30 148.8 Susitna R. above Portage Cr. 31 28 -148.8/0.1 Portage Cr. 15 25 181.3/0.0 Tsusena Cr. 31 31 30 184.4 ·susitna R. at Watana dam site 30 -31 30 """ 194.110.0 Watana Cr. 31 6 206.810.0 Kosina Cr 31 3 223.7 .. Susltna R. near Cantwell ----24 31 31 15 231.310.0 Goose Creek 31 31 30 233.410.0 Oshetna Creek 15 31 24 "R&M gages .... ··usGs gages All others are AOF&G gages. 32 r r - - To fill this missing year, we simplified, but retained, the logic of the SNTEMP regression technique. SNTEMP uses what may be termed a "physical process" regression model for data filling and smoothing. The regression model is based on a simplified version of the heat transport model used to predict downstream water temperatures. These models employ an equilibrium stream temperature assumption calculated equilibrium temperature (T ) e represents the where value the the stream is asymptotically approaching. The standard regression model of SNTEMP uses the calculated T and the rate of approach to T as independent variables. e e For the Susitna River application, this model was simplified to use only the equilibrium temperature (Figure 14). USGS gage data collected on the Chulitna and Talkeetna rivers. Only three usable water temperatures were available for the Talkeetna and Chulitna rivers during the June to September periods. These temperatures were recorded on the Chulitna River during June, July, and September of 1982 (Bigelow 1983). Because of the limited data at these ... stations, regressions similar to those used for the Cantwell gage were of little value (Figures 15 and 16). However, the values predicted by these regressions were used to fill in the missing data and to smooth those observ1ed data points with only one observation per month. Where available, ADF&G temperature data were used to adjust the temperatures at the gages so that simulated temperatures matched the observed data at the ADF&G s:ites. Figures 15 and 16 list the values assumed by the model, but the reader should note the low confidence associated with these values. Temperatures of Distributed Flow. Flow accretions from groundwater or surface inflow are included in the network as continuous additions to the stream flow, referred to as distributed flows. This is the primary mechanism for simulating Susitna tributary flows. Water temperature predictions for smaller tributaries depend on the water temperatures assigned to tributary distributed flows. Thus, the accuracy of temperatures assigned to distributed flow is critical to the simulation. Contribution from surface or groundwater flows have not been quantified 33 Figure 14. Temperature regression for Susitna River at Cantwell gage. r i' ,(: IF" I 1 """' i - F- 1 - -I ! ,... Month June 81 July 81 August 81 September 1""' 34 81 15 10 5 Regression 95 Percent· Regression Prediction Confidence Intervals {C) (C) 8.82 ± 1.53 8.96 ± 1.54 8.20 ± 1.51 5.30 ± 1. 74 5 lfZl 15 20 25 EQUILIBRIUM TDfPERATURE I"" I""' I - I""' .... - r r l r I r ' r Figure 15. Temperature regression for Chulitna River at USGS gage. Regression 95 Percent Regression Confidence Observed (C)/ Prediction Intervals Month Sample Size (C) (C) June 81 7.4/1. 6.68 ± 2.85 July 81 -/0 6.90 ± 2.89 August 81 7.2/1 6.64 ± 2.84 September 81 -/0 4.95 ± 3.83 June 82 7.3/24 6.53 ± 2.84 July 82 5.7/31 7.01 ± 2.92 August 82 -/0 7.01 ± 2.92 September 82 4.6/10 5.22 ± 3.56 *Temperature at gage was adjusted so downstream simulation matched data collected by ADF&G. 15 5 0 0 5 11Zl 15 20 25 EQUILIBRIUM TEMPERATURE Value Used (C) 6.68 7 .10* 6.64 5.25* 5.45* 5.7 7.01 4.6 35 .... I r r- 1 .... .... ..... Figure 16. Temperature regression for Talkeetna River at USGS gage. Regression 95 Percent Regression Confidence Observed (C)/ Prediction Intervals Month Sample Size (C) (C) June 81 10.2/1. 8.37 ± 5.85 July 81 9.0/1 9.28 ± 5.96 August 81 9.7/1 8.14 ± 5.84 September 81 1.5/1 2.86 ± 7.38 June 82 -/0 8.54 ± 5.86 July 82 -/0 9.67 ± 6.04 August 82 -/0 9.52 ± 6.01 September 82 -/0 2.74 ± 7.45 *Temperature at gage was adjusted so downstream simulation matched data collected by ADF&G. 36 ~ ~ 15 ~ 1121 w p.., ~ E-o A ~ 5 ~------~--------~~----~------~------~ w t;/1 ~ 0 0 0 5 10 15 20 25 EQUILIBRIUM TEMPERATURE Value Used (C) 8.37 8.60* 8.40* 5.70* 7.00* 9.67 9.20* 5.50* r -I - r - in Susitna tributaries; therefore, they must be estimated. Presently, two techniques can be used to estimate these temperatures. The first is to assume groundwater inflow at a constant temperature for all time periods and all locations. G. Nelson (1983) of the USGS suggested a value of 3 C as representative of a wide range of conditions encountered by that organization in adjacent drainages. This assumption does not allow for 1) seasonal ground temperature variation, 2) ground tempera- ture variation with site elevation, or 3) the possibility of surface runoff. Rather than assuming a constant temperature for distributed flows, an alternative technique is to vary temperature by location and depth. AEIDC modified the ground temperature function presented by Williams and Gold (1976): where: T (x,t) = T +(~T /2) cos [(2~t/t ) -xl~/at ] exp(-xl~/at ) g g g 0 0 0 T g AT g t t 0 X a = = = = = = average annual ground surface temperature (C) annual range of ground surface temperature variation (C) time from occurrence of peak temperature (days) time for one cycle of temperature variation (365 days) depth (m) thermal diffusivity (m 2 /day) = thermal conductivity/volumetric heat capacity This formula can be used to predict ground temperatures at variable depths and times if the average annual ground surface temperature (T ) g and annual range of ground surface temperature variation (~T ) are g known. The annual range of ground temperature can be assumed to be the same as the annual range of air temperature variation (Williams and Gold 1976) which is 28.2 C at Talkeetna. Data presented in Williams and Gold (1976) . indicates that the average annual ground temperature is approximately 1 to 7 C warmer than the average annual air temperature in regions with persistent snow cover. If, for notational purposes, we 37 (10) r I I r - - - designate this 1 to 7 C offset by Toff and define A the formula becomes: = 21r/t , B 0 Tg(x,t) = Tair + Toff + 14.1 cos (At-Bx) exp(-Bx) = l1r/at , 0 (11) Air and ground temperature data collected at Gulkana, Alaska (Aitken 1964b) and Big Delta, Alaska (Aitken 1964a) suggest that this offs.et temperature is in the range of 4. 3 to 4. 9 C. For purposes of further discussion in this paper a value of 4. 6 C will be assumed, although in the SNTEMP implementation of this ground temperature model Toff will be used as a calibration variable. The mean annual air temperature of an arbitrary location at elevation Z can be computed from the mean annual air temperature at Talkeetna (0. 3 C) using the lapse rate equations discussed in the modifications section: (12) where: Tz = air temperature at elevation Z (C) To = observed air temperature at elevation zo (C) zo = elevation of site where air temperature is known (Z = 0 105 m for Talkeetna) z = elevation of site where air temperature is desired (m) y = air temperature lapse rate (C/m) By substituting the air temperature lapse rate expression for air temperature at elevation Z, the ground temperature formula can be rewritten as: T (x,t,Z) = 4.9-y(Z-105) + 14.1 cos(At-Bx) exp(-Bx) g (13) If a value is assumed for the · thermal diffusivity, the only undefined variable for any location and time period is the depth of the ground temperature. There are two depths of interest which correspond 38 - ,.,. - to two separate forms of heat flux--conduction to and from the streambed and mass transfer of heat (distributed flow). Streambed conduction is a function of the depth at which the ground temperature variation is essentially zero for the simulation time period. Given an estimate of a., a depth can be computed where daily temperature fluctuations are essentially zero. Williams and Gold (1976) give an a.-value for wet sand of 0.01 cm 2 /sec. This value is also used to represent the thermal diffusivity of sand, gravel, cobbles, and boulders in the Susitna slough hydrogeology study (Acres 1983b). Using this value, daily temperature fluctuations penetrate to a depth of approximately 0.8 m. Substituting 0.01 cm 2 /sec for a. and 0.8 m for depth, the above formula reduces to: T (t,Z) = 4.9 -y(Z-105) + 10.3 cos(At-0.316) g (14) The distributed flow heat flux is a function of the average depth from which the water flows. Rather than assume a value, this depth has been retained as a variable for calibration purposes. This ground temperature model must be considered provisional as the assumptions made cannot be tested or validated without further data collection. Temperature at depth data at several locations within the Susitna Basin would be required for validation of this model and improving estimates of assumed values. AEIDC believes this model currently provides the best available approximation of the physical conditions existing in the Susitna Basin and will be applied without validation until better estimates of existing conditions are obtained. METEOROLOGY Selection of Meteorologic Data The SNTEMP model is designed for climatic data input from only one representative meteorologic data station per stream network. The only long-term meteorologic data station within the Susitna Basin is the U.S. National Weather Service station located in Talkeetna. This station has summarized monthly data (air temperature, wind speed, relative humidity, and percent cloud cover)--the data required by SNTEMP--for the period 1968 to 1982. In addition, unreduced data are available from 1950 to 1968 on computer tape from the National Climatic Data Center. This 39 r r I - - period of record allows stream temperature simulations under extreme and normal meteorologic conditions once these data are· adjusted to better represent Susitna Basin conditions. We used meteorologic data collected specifically for the Susitna study (R&M 1980, 1982a, 1982b, 1982c, 1982e, 1982f) to validate this meteorologic data adjustment and SNTEMP solar model predictions. Ground Reflectivity and Atmospheric Dust The stream temperature model predicts solar radiation based on site latitude, period of the year, cloud cover, ground reflectivity, and atmospheric dust. AEIDC determined monthly ground reflectivity values for the Susitna Basin using the percent area groundcover vegetation types presented in McKendrick (1982) and Bredthauer and Drage (1982). The remaining component necessary to predict solar radiation is an estimate of atmospheric dust. Dust was estimated by calibrating monthly average predicted solar radiation to observed values using the published solar radiation and percent possible sunshine data collected at the Palmer Agricultural Experiment Station (Matanuska Station as recorded in Wise 1979). Figure 17 presents these coefficient values. Meteorologic Predictions Conditions observed at Talkeetna are not necessarily representative of the entire basin. SNTEMP adjusts most of the recorded variables to better represent the local conditions within the basin. For example, the predicted solar radiation considers local topographic shading. The following discussion compares Susitna Basin meteorologic predictions with data collected by R&M. As was previously discussed, SNTEMP has been modified to accept monthly air temperature/elevation and humidity/elevation functions. The air temperatures and humidities predicted by these equations using observed data are compared to the data collected by R&M (Figures 18 and 19). From these plots it appears that the humidity lapse model is a poor predictor of basinwide conditions; however, we retained it in SNTEMP for three reasons: 1) Talkeetna humidity data are based on wet/dry bulb measurements which are inherently more accurate than ceramic plate recorders (Wise 1983); 2) balloon-carried radiosondes are 40 --] .. 1 J OUST COEFFICIENT GROUND REFLECTIVITY J -··- 1.0 "· 5 -1 J ~--~~ l " --1 ] . ---] ---~----) -~--1 --·1 1 ----~ l ---I COEFFICIENT VALUE - .r-··- r .r--·~ / / . V . ""-·'\ I \ / . I .___ ~---------7· ~ -~ ---·~ \ ............_ ----. --......---....__ . ·--.--·-·----\ / / ./ - _I _I I I I I .I I JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC MONTHS ] ...... ..... . f~ C"t 0 § 8 c:: p. (I) ~ti Ill Cll Hl >I-' OQCil t-t n 1-'• C"t n 1-'· c:: < ..... 1-'· C"I'C"t C::'<l ti Ill Ill 1-'1:1 p. t<:l .a~ ~ g 1-'• (I) m-g. r:s Cll C"t ti 1-'· cnn C"t Ill p. C"t = 1-'-(1) 0 C"t r:s .. n 0 >-diD Ill Hl 1-'Hl a .... Cll n 11 I-'• • Cll :::1 C"t (I) .. .... """" I I ..... I , .... !""" ! - - - !""" Figure 18. Susitna Basin observed air temperatures vs. temperatures predicted from Talkeetna data. JOOE 1980. 1981, 1982 PREDICTED TEMP. OEG C SHERMAN 14 (l 12 IJ£VIL C/o.NYD/1 A Iii VAT ANA 8 • 5 I<OSIHA 2 B 11 2 4 6 B 11! 12 lliSERVED TEIIP. OEG C JULY 19Ba 1981. 1982 PREDICTED TEMP. llEGC SllERl~.'Jl 14 D D 12 OEV!l CJ.NYCN A Ul ~ VAHJII. 8 ! r • t 6 1- KCSIW. f 4 • 2 z a 2 6 a ~~ 12 OBSERVE:! m?. !:'EG C 4? 14 A H Figure 18. (Continued) """ AUCUST 19Bil. 1981. 1982 PREDICTED TEH?. QEG C SHE.~!o!.•.li 14 f""' 0 12 DEVIL CJ.NYON ll. A 10 ~HUlA 8 ~ !""" 6 KOSINA • 2 - z !""' ~ 2 6 8 :a 12 14 OOSERVED TEf.!P, DEGC -SE?TO!BER 1982. 1981. 1982 PREDICTED TEIIP. DEG C SHEltii~.N 14 0 12 !IV IL C/lllYCN A 10 I""' ~m:-:A 8 I""' ~ 6 KCS!NA -I • """ 2 !""' 0 • " 2 4 6 9 1~ 12 14 . OBSERVED iEH?. :EG C -43 ,- Figure 19. Susitna observed humidities vs. humidities predicted from Talkeetna data. Jtn'r ·~ IllS~. 1981, 1982 PREDICTED RH. DECIMAL II""' u SHE~\~Mt ~. 9 D ll.B cr:'ii~ Cf.NYOlJ 11.7 ll u ~ABNA il.S + a.4 KOSIW. 0.3 "·2 !il. I Z.lil u 1.1 0.2 a3 Ill. 4 !il.S 11.6 ll.7 l.l.e lUI u ,.... 09SS~IED RH. DEC;~~-l JULY 1983. 1981. 1982 PRE;JJCiEO RH. OECIH~.L 1.11 SH'"..Rfllll 11.9 0 3.8 rr.Vll Cf~JYCIJ 1!.7 .. r"" a.6 'jATf.NA as F"' • a.4 K!JSlNA !""' B.3 9 lil.2 """' Ill u I""", :l.:il <t.l 2.2 ~.3 :il. 4 ll.S ~.5 B. 7 z.a a.9 l.Z ~SSERV8J RH. rECitli.L -44 Figure 19. (Continued) .... f.UGUST )ga3, 1981. 1982 :--PREDICtED RH. OECI!VJ. 1.11 9l~N a.9 i Q 11.8 D ~ DmL CANYON • A At 3.7 A 11.5 r' VAT.IiiA as + r 0.4 i(OSl/i}. &.3 • 0.2 ll.l lil.0 ll.ll 0.1 9.2 3.3 0. ~ 0.5 11.6 ta.7 z.a 11.9 l.ll OBSERVED R~. OE.Cl)!f.L -SE?!EMa"..R 19BD. 1981. 1982 PREDiCtED Rll. OEClH~L 1.0 -S/!£RHf.N 11.9 0 0 3.8 .- DEY !L CUI'I:lH 17 + A ..... 11.5 ~mN. ll.S + :.u i(0SJilA ..... I ~.J • ;1.2 _, I ll.l a.o ~.a ll.l <1.2 3.3 :J.I, li.S u 0.7 ll.B 0. 9 1.11 ,... ::asF.R'I:D PH, ~EC:~~.L 45 - r i" i ,- calibrated at the time of release and resultant data are the means of twice-daily observations; and 3) erratic behavior (e.g., daily 0 to 100 percent oscillations) was noted in several of the R&M humidity recordings. The wind speeds at Talkeetna are not currently adjusted in any way to better represent winds within other parts of the Susitna River basin. Wind speeds recorded at Talkeetna were compared to wind speeds recorded by R&M at various locations within the basin (Figure 20). It would be relatively simple to incorporate a linear adjustment equation to translocate observed Talkeetna wind speed data to locations which would be better represented by the observed R&M data. However, the wind speed data collected by R&M does not necessarily represent the wind speeds which occurred directly above the water surface and are responsible for the rates of convective and evaporative heat flux. Since it appears to be impractical to collect wind speed data within the canyons below the existing meteorological data sites (Bredthauer 1983), the wind speed data collected at Talkeetna will be used as representative of average basin winds. Figure 21 compares observed solar radiations to predicted solar radiation. The simulated data are a reasonable approximation of the field measurements. VALIDATION The purpose of model validation is to locate systematic prediction errors. Systematic errors result when observed or assumed data for a particular study do not represent actual conditions. Since the stream temperature model has been verified with previous applications (Theurer and Voos 1982; Theurer et al. 1983) and, since some adjustments have been made to SNTEMP to account for conditions particular to the Susitna application, it is assumed that any remaining systematic errors are the result of nonrepresentative input data. An initial validation run of the Susitna-modified SNTEMP demonstrated a tendency to underpredict the upper tributary temperatures (Figure 22). Since most of the data defining these tributaries are assumed or estimated values, much uncertainty exists in the definition of each tributary. Several poorly defined variables which might be 46 Figure 20. Average monthly wind speeds' (M/S), 1980, 1981, 1982. ~ AVE. WIND SPEED - - I""' r - r- - sr--------------------------------------- WATANA DEVIL CANYON TALKEETNA KOSINA ---_... ' 0 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC AVE. WIND SPEED 5 --1 I WATANA I I 4 I I DEVIL CANYON /i 3 ' I TALKEETNA ::c-....~ /"-.'-.......' ·-·"" . 1 .""-I ! -·-·· 2 ·-.--~ I · KOSINA -.....__ _,) .....__ / .....__ ----·---1 JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC sr=A~VE~·~WI~ND~S~P~E=ED=--------------------- WATANA 4 DEVIL CANYON TALKEETNA 2 SHERMAN ........ __ - KOSINA 1 -/ "-., ......... .., ................ ..... ,...-... ----~--------- o~----~~--~--~~--~--~~--~~ JAN FEB MAR APR HAY JUN JUL AUG SEP OCT NOV DEC 0 co 0'\ ...... ...... co 0'\ ...... N co 0'\ ...... 47 -l +"- 00 to:j .... OQ PREDICTED SOLAR (J/M2/SEC> ~ (1) 259 N ...... . WATANA 1-0 11 (1) p. D ~· n 1'1' (1) p. 200 < DEVIL CANYON (I) . 0 0' 1:. (I) (1) ~ (1) p. (I) SHERMAN 0 150 1-' Ill 11 + "d 11 (1) p. ~· n 1'1' .... 0 ::s (I) .. 100 ...... 1.0 00 0 .. ...... \0 00 ...... .. ...... \0 50 00 N . 50 10~ 150 2~~ 250 OBSERVED SOLAR (J/M2/SEC> l 1 l-%j ..... PREDICTED TEMP <Cl ~ 12 ltl N N . ~ ..... 0"' ~ 19 rt ~ rt ltl .eJ ltl ti cu B rt ~ ltl Cll w C"l 0 OQ 6 ti g ::s ~ cu rt ltl ti I-'• 4 ::s Hl ...... ~ cu Cll Cll ~ ltl p. 2 . .j::-2 4 6 8 1~ 12 1.0 OBSERVED TEMP <C> r - - ,.... I. l l r ! !"""' - ..... adjusted to improve model predictions are 1) stream flow, 2) initial stream temperature, 3) stream length, 4) stream width, and An effort has been made to adjust 5) distributed flow temperatures. other variables to better represent prevailing conditions (e.g., air temperature, relative humidity, and topographic shading). Of the five poorly defined variables, most improvement could be gained from focusing on temperatures of distributed flows. determination was based on the following logic. This 1. · Without the benefit of continuous tributary flow gaging, substantially 2. 3. 4. present stream flow estimates cannot be enhanced. With the subsequent necessary assumption of zero flow at the tributary headwaters, initial tributary temperatures have no influence on the predictions. Tributary lengths were measured from maps. Stream widths are based on field estimates and initial tests with SNTEMP demonstrated that this variable was not sensitive enough to remove the existing predictive bias. Rather than arbitrarily modifying the constant 3 C estimate of groundwater temperature, the ground temperature model previously described was employed to generate physical process-based temperature estimates. This model introduced three variables which must be estimated--the average annual air/ground temperature offset (Toff), thermal diffusivity (a.), and depth of inflow (x). AEIDC is currently seeking techniques and data for estimating values of these variables. Until solid estimates can be obtained, these variables will be adjusted to calibrate to observed water temperature data. While the tributary temperatures have a relatively small influence on the natural mainstem temperatures, this influence could increase in importance during construction or operation of the dams. Future temperature simulations will provide an indication of this effect and perhaps suggest alternate modeling techniques or point out the need for more tributary temperature data • 50 r i I r I ' .... r - CALIBRATION Tributary temperature predictions were improved by adjusting the three groundwater temperature parameters (Figure 23). The resulting values were: Toff = 1.0 C and a = 0.01 cm~/sec for the entire basin, Z = 0. 4 m for Kosina Creek, Z = 0. 7 m for Watana Creek, and Z = 2. 0 m for the mainstem and remaining tributaries. necessary to validate these values. Further analysis is The goodness of fit was determined by using the following statistics: where: s = lt(o. -~)2/(n-1) 1. (15) (16) (17) t5i = difference between ith predicted and observed temperatures, c Ti .th published temperature, c = 1. Ti th predicted by SNTEMP, c = i temperature ~ = mean difference, c n = number of observed temperatures S = standard error estimate, C These statistics can be combined with Z values to define prediction confidence intervals. For example, 90 percent of the predicted values fall within ~±(1.645)8 of the observed values. Postcalibration statistics for the tributaries indicate that predicted values are on the average 0.28 C (~) higher than the published values, and 90 percent of the predicted values can be expected to fall between 2.10 above and -1.54 C below the published water temperature (S = 1.11 C, n = 12). The model fit could be improved with additional adjustment of Toff' Z, and a. However, it was decided that additional calibration be postponed 51 Figure 23. Tributary temperatures; postcalibration, including distributed flow temperature model. -N ---s -~ --I " ----' CQ ..... I '\. I (Q ..... -.... ..., ...... (..) ....... ~ ·-C) I:!:! (..) ...... ~ N N e5ll CQ co .... N -I""' 52 I""" 1--I -u ....... ~ ffi 1-~ ffi c.n CQ C) - .... l r- i""" .... - I""" i - """" ! until research is completed to define reasonable physical limits of these parameters. Once the tributary predictions had been improved, the entire mainstem/tributary system essentially was calibrated, and no additional parameter adjustments were attempted. Statistics for the mainstem are ~ = -0.05 C, S = 0.90 C, n = 28. Figure 24 presents these statistics as computed for each month. Figure 24. Temperature model calibration statistics for tributary predictions. n ~ (C) S (C) June 1981 2 -2.08 1. 39 July 1981 1 0.41 August 1981 6 0.59 0.42 September 1981 6 -0.002 0.74 June 1982 2 0.08 0.44 July 1982 3 -0.89 0.28 August 1982 4 0.19 0.80 September 1982 4 0.14 0.20 Average 28 -0.05 0.90 The statistics for June 1981 indicate a poor fit. This is understandable since the three required initial water temperatures (Cantwell, Chulitna, and Talkeetna) were synthesized with linear regression models. This is the only month which had all three initial temperatures synthesized. A more reasonable estimate of the simulation performance for the mainstem is obtained by eliminating this month from the computations: ~ = 0.10 C, S = 0.66 C, n = 26. percent confidence interval is 0.10 ± 1.09 C. The corresponding 90 Appendix C provides 53 r r - - - r .... longitudinal temperature predictions for the 1981 and 1982, June through September periods. RESULTS AND DISCUSSION The Susitna River temperature model has been validated and calibrated for the months of June through September 1981 and 1982. We estimate that mainstem temperature predictions will be within 1.09 to -0. 99 C of actual values, and upper tributary temperature predictions will be within 2.10 to -1.54 C of actual values (90 percent confidence intervals). This estimate assumes that the statistics computed from simulations using two years of historical data will apply to project conditions and there is no way of knowing if this assumption is valid. Nevertheless, these statistics are a measure of the model's performance given the best possible conditions and the available input data. Tributary and mainstem temperature data from the 1983 field season are expected to improve estimates of the model's accuracy and precision. Additional analysis of distributed flow and temperature regimes and tributary flow regimes will be required if the model's predictive capabilities are to be improved, especially with respect to the upper basin tributaries. We used a ground temperature model to estimate the temperature of distributed flow. This model has not been validated with data from within the Susitna Basin. If the parameter values defining the model can be measured, or at least assigned physically relevant constraints, the model can be applied with confidence to simulations of the proposed project. 54 r r r - ..... I""' I ! .... - - FUTURE APPLICATIONS AND ENHANCEMENTS AEIDC will continue the Susitna flow and stream temperature analysis by the following steps. 1. Normal and extreme flow regimes within the basin will be defined by statistical analysis of the pre-and postproject 32-year flow records. 2. 3. 4. 5 • 6. 7. 8. Using statistical analysis, AEIDC will determine the location where postproject flows are significantly different from natural flows. This will identify the area facing possible hydrologic/hydraulic impacts. Combinations of hydrology and meteorology which produce normal and extreme stream temperature changes will be determined from simulations using recorded meteorologic and hydrologic data. Ranges of expected flows and temperatures resulting from the filling and operational phases of the project will be used as input to the temperature model for simulating downstream effects. These simulations will use normal and extreme basin hydrology and meteorology. Results of these simulations will be analyzed and a zone of predictable impacts identified. This zone will be partially defined by estimates of the model's performance statistics. Weekly or daily prediction capabilities will be pursued if the need is indicated by analysis of the monthly simulations. Results of the 1983 field season will be incorporated into the model and new model performance statistics calculated. Techniques will be developed for improving the distributed flow temperature model. 9. Fall and winter conditions will be used for water temperature simulations to provide estimates of tbe most upstream limit of ice cover. If the stream temperature model reliably predicts the recorded limits, the model will be applied to proposed project conditions. Ice observations by R&M will be used for validation of these simulations. 55 - -I ' - - r 56 .... I""" ' - -i r - BIBLIOGRAPHY Acres American, Inc. 1983a. Application for license for major project, Susitna Hydroelectric Project, before the Federal Energy Regulatory Commission. Vol. SA. Exhibit E, Chaps. 1 and 2 (figures). Alaska Power Authority. Susitna Hydroelectric Project. 1 vol. 1983b. Slough hydrogeology report. Draft Report. Alaska Power Authority. Susitna Hydroelectric Project. 27 pp. 1982. Susitna Hydroelectric Project. Feasibility report. Vol. 4. Appendix A. Hydrologic studies. Final draft. Prepared for the Alaska Power Authority. 1 vol. Aitken, G.W. Alaska. Hanover, 1964a. Ground temperature observations, Big Delta, Cold Regions Research and Engineering lab. , U.S. Army, NH. CRREL Technical Report 104. 15 pp. 1964b. Ground temperature observations, Gulkana, Alaska. Cold NH. Regions Research and Engineering Lab • , U. S. Army, Hanover, CRREL Technical Report 106. 13 pp. Alaska Dept. of Fish & Game. 1983. 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. 1981. report. Studies. Aquatic habitat and instream flow project. Final draft Anchorage, AK. Alaska Power Authority. SuHydro Aquatic Report for Acres American, Inc. 2 vols. in 3. Alaska, University, Arctic Environmental Information and Data Center. 1983. Methodological approach to quantitative impact assessment for the proposed Susitna hydroelectric project. Alaska Power Authority. Susitna Hydro Aquatic Studies. Anchorage, AK. Report for Harza/Ebasco Susitna Joint Venture. 71 pp. Bigelow, B.B. 1983. Letter, March 8. U.S. Geological Survey, Anchorage, AK. 7 pp. Bredthauer, S. 1983. Personal communication. Telephone conversation. R&M Consultants, Inc., Anchorage, AK. Bredthauer, S., and B. Drage. 1982. River morphology. R & M Consultants Inc., Anchorage, AK. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. Brett, J.R. 1971. Energetic responses of salmon to temperature. A study of some thermal relations in the freshwater ecology of sockeye salmon (Oncorhynchus nerka). American Zoologist. 11:99-113. 57 - .... ..... .... r - Cherry, D.S., K.L. Dickson, and J. Cairns. 1975. Temperatures selected and avoided by fish at various acclimation temperatures. Journal of the Fisheries Research Board of Canada. 32:485-491 • Coutant, c. C. 1970. Thermal resistance of adult coho (Oncorhynchus Kisutch) and jack chinook (0. tshawytscha) salmon, and adult steelhead trout (Salmo gairdneri) from the Columbia River. Battelle, Pacific Northwest Labs., Richland, WA. USEAC R&D Dept., BNWL-1508. Kilday, G.D. 1974. Mean monthly and annual precipitation in Alaska. National Weather Service, U.S. National Oceanic and Atmospheric Administration, Anchorage, AK. NOAA technical memorandum • NWSAR-10. 1 vol. McKendrick, J. 1982. Plant ecology studies. Final Report. Agri.cultural Experiment Station, University of Alaska, Palmer, AK. Alaska Power Authority. Susitna Hydroelectric Project. 124 pp. Nelson, G. 1983. Personal communication. Telephone conversation. U.S. Geological Survey, Anchorage, AK. Quane, T. 1983. Personal communication. Interview. Alaska Dept. of Fish & Game, SuHydro, Anchorage, AK. R&M Consultants, Inc. 1982a. Field data collection and processing, Supplement 1. Task 3-Hydrology. Alaska Power Authority. Susitna Hydroelectric Project. Prepared for Acres American, Inc. 215 pp. 58 1982b. Field data collection and processing. 3-Hydrology. Alaska Power Authority. Susitna Project. Prepared for Acres American, Inc. 1 vol. Vol. 3. Task Hydroelectric 1982c. Field data index. Anchorage, Authority. Susitna Hydroelectric Project. American, Inc. 1 vol. AK. Alaska Report for Power Acres 1982d. Hydraulic and ice studies. Anchorage, AK. Alaska Power Authority. Susitna Hydroelectric Project. Report for Acres American, Inc. 1 vol. 1982e. Processed climatic data. Vol. 6. Devil Canyon Station. Anchorage, AK. Alaska Power Authority. Project. Report for Acres American, Inc. Susitna Hydroelectric 1 vol. 1982f. Processed climatic data. Vol. 5. Watana Station. Anchorage, AK. Alaska Power Authority. Project. Report for Acres American, Inc. Susitna Hydroelectric 1 vol. 1981. Attachment Authority. can, Inc. Susitna River mile index: Mouth to Susitna Glacier. D to Hydrographic Surveys. Anchorage, AK. Alaska Power Susitna Hydroelectric Project. Report for Acres Ameri- 1 vol. ,.... ' - i ! ..... .... I i R&M Consultants, Inc. 1980. Field data index. Anchorage, AK. Alaska Power Authority. Susitna Hydroelect~ic Project. Report for Acres American, Inc. 1 vol. Reiser, D.W., and T.C. Bjornn. 1979. Habitat requirements of salmonids. No. 1. in U.S. Forest Service. Influence of forest and rangeland management on anadromous fish habitat in western North America. Anadromous Fish Habitat Program. General Technical Report PNW-96. 54 pp. Sauntner, J. 1983. Personal communication. Interview. Alaska Dept. of Fish & Game, SuHydro, Anchorage, AK. Schoch, C. 1983. Personal communication. Telephone conversation. R&M Consultants, Inc., Anchorage, AK. Siefert, R.D. 1981. A solar design manual for Alaska. Water Resources, University of Alaska, Fairbanks, AK. 6. 163 pp. Institute of Bulletin No. Theurer, F., and K. Voos. 1982. An instream water temperature model of the upper Colorado River basin. Unpublished. Paper for presen- tation at the International Symposium on Hydrometeorology, Denver, co. 6 pp. Theurer, F., K. Voos, and W. Miller. 1983. Instream water temperature model. Draft report. Instream Flow and Aquatic Systems Group, U.S. Fish and Wildlife Service, Fort Collins, CO. Instream Flow Information Paper No. 16. 263 pp. Trihey, E.W. 1983. Personal communication. consultant, Anchorage, AK • Interview. Private U.S. Geological Survey. 1981. Water resources data for Alaska. Anchorage, AK. Water-Data Report AK-81-1. 395 pp. 1980. Water resources data for Alaska • Water-Data Report AK-80-1. 373 pp. Anchorage, AK. U.S. National Weather Service. 1980. Climatological data national summary, Vol. 30, No. 9. Washington, DC. 1970. Climatological data national summary, Vol. 20, No. 8. Washington, DC. 1969. Climatological data national summary, Vol. 19, No. 7. Washington, DC. 1968. Climatological data national summary, Vol. 18, No. 6. Washington, DC. U.S. Soil Conservation Service. 1982. Precipitation and water yield. Alaska Rivers Cooperative Study, Susitna River Basin, Matanuska-Susitna Borough, Alaska. 4 pp. 59 i" I I ,_ ' - r Williams~ G.P., and L.W. Gold. 1976. Ground temperatures. CBD180 in National Research Council Canada. Canadian Building Digests 151-200. Ottawa, Canada, 1979. Wise, J.L. 1983. Personal communication. Interview. Arctic Environmental Information and Data Center, University of Alaska, Anchorage~ AK. 1979. Alaska solar radiation analysis. Arctic Environmental Information and Data Center, University of Alaska, Anchorage, AK. 27 pp. 1977. Mean annual precipitation in inches. Arctic Environ- mental Information and Data Center, University of Alaska~ Anchorage, AK. 1 map (scale 1:2,500,000). World Meteorological Organization. 1982. Monthly climatic data for the world, Vol. 35, No. 1. National Climatic Center, Asheville, NC. 1981. Monthly climatic data for the world, Vol. 34, No. 1. National Climatic Center, Asheville, NC. 60 - """ I - - I""' I APPENDIX A TOPOGRAPHIC SHADING - r - .... I TOPOGRAPHIC SHADING These plots present the solar shading characteristics of the Susitna reaches (refer to Figure 10). Mainstem reaches 9 and 10 and the Talkeetna and Trapper tributaries were estimated to be unshaded for all months. Fog Creek was assigned the same shading characteristics as reach 1. The synthetic tributary (Cheechin) was assigned the same characteristics as reach 4. The continuous curves represent the path of the sun for each month. The hatched area represents the potential shading of the surrounding terrain. A-1 1 ] Reach 1 RM 179.5·184.5 64NL - - .. - - - ,,, nc on ~ -r-~ 10 ~ ~ ..... 21 lui 21 t:t r-r--., prn - / I ~ Sa I pm ~ y I -+ ~·· 21 Au,, 21 ~ +-- "{.. ~ I Flpm a. \ vt I - / I I I ~ \/ I I ~/ 1a v \ /'\ \ I j I ,r-......., I I~ pm ~ M•r 21 s •• 21 I I ~ :~ l)( I I T'----, I ~ -A I a. \ \ ~ I pm I I I / I """ I ?} A lX I ~ I ~ ~ ~ ~ I ~ I ~ ~ ~~ \ \ I 1:1 I I \ I \ ~ ~ ~ -~ ~ ~ ~ I N \ I "!:_ !'-. ~ p"" 4 .. ny \ I I ~ \ V\ \ ~ ~ ~ ~ ~~ ~ ~ ~ ~ I " ~ '"' r& it' I d ~ ~ 'v \ ~~'\ ~ I ~ \ ~ I ~ ~ ~· I ~ ~ "'~ ~ '1 oo· '13o· "~liD' '1D5'H 715'B :;~~s~ Na~ '15' g15' 30' 45' so· 75· M '105· 120· '135· '15 a· ANGLES 00 EAST SOUTH WEST Sun palh diagram lor 64 •N lalilude. so· 70' Ul w 60' .J [!] 2 50• ct 40' w 0 J 30' ... -~ ~· <! 'lO' ] l 1 Reach 2 RM 175.5-179.5 64NL - - - - - ,:unc on ~ -~ 10 ~ M•v21 111121 ~ It--pm - 9P n/ w \ ' ~ pm v K I I Aor 21 Ao. 21 r+--I "{. ~ \ ...-\ ~ flpm a. I vt I - / \ I ' I ~ , ......... t ' -......! ,,. v I V", I I ' I I' I I~ prn 10 l)( ~ M•r 2L s •• ~· rr---I I ~ ~ I I r--I - \ A h I So I I pm I \ \ II I I I X V( V( I /\ I I I I ~ I ~ ""' 15 \ I ' I I I \ I r.h ~· O<l...l.!_ ' ~" ~ \ ~ I - \ ;; I V-1 r--.-, ' N 4"":'/ \ ~ r--... I ~'-... !!P~. ~ ~ ~ ~ ~ ~ ~ ~ No.2JI I ........... ~ I "" ~ ~ ~ ~ ~ 3. J.an 21 , I ~ ~ ..... t-d I ~~ ~I ~'h\0 ~ . 1 150' 1315• 1QO' 1015"H 715'8 ;;~;.6~ N3;;; ,5. B 15' 30' 45' 60' 715· H 105· 120· 1:.1:5· 15 o· ANGLES 00 EAST SOUTH WEST Sun path diagram for 64"N latitude. ao· 70' Ul w ..J SO' (.!) 2 50' ct 40' w 0 J :.10' 1--~ ~· ct "KJ• l ~--~ 1 I ] ~~~ 1 l ~-~·-~·~1 -l c) l -~"C~ .. , ] 1 1 ) l Reach 3 RM 166.0*175.5 64NL - - - - - .~t nc lon I vr ~ 10 ~ ~ Mav21 '"' 21 ~ r-r--. pm -n/ w I ~ So I pm ~ v I I Aor 21 ~u~ 21 M--i "-,( ~ I !--\ ~ ~pm a. \ Vr I - / I I I I f'-,..,..... I I I r--..../ ,., v \ ~, . \ I I I I'"' I N pm !r:: vt oA•• 21 s •• 21 I I It( ~ l)( I ~ r-~ I I -a. \ I ~ I pm \ I I L<5 l" "\." ~ I I I rY, v lX I l2 I ~ ~ t&~ ~ ~ ~ I "' } ~n \ es I I I 'v \ ~ ~ - Q.,'o/ \ /; ~ ~ ~ I N Nt--.~ \ ~ b<Y I V\ \ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1'-I ~ ~ ~ ~ h ~ \ ~" ~~ I I I t.Ll liCK" .~ I I oo· 1315' -uzo· 105' Q 7!5' B ~~ ;.5; N3;; 16' B 15' 30' Q5• 60' 75· H 105· 120· 135· 15 o· ANGLES 90 1 EAST SOUTH WEST Sun path diagram lor 64 "N latitude. ao· 70' Ul w ..J 60' (!) 2 so· q QO• w 0 J 30• 1--~ ~·~ 10• j --~-J 0 --1 Reach 4 RM 163.0·166.0 64NL ~--~---4----+----+----~--~--~~--;----+----+----r--~r---;---~----+----T----r---~r--~----,ao· b---~--~P---~----~--~----~---+----f----+----+----+----+----+----+----+----1----+----1----+----1Ba ~--~-----~--~----~--~----~---+----+----+----+----+----+----+----+----+----1---~----1---~----~70· (/) w 60' ..J ~~---+---+--~--~--~--+---~~r--;---+---+---r--,_--~--+---r-~r--;--~ ~ 2 ex ~--+---+---~--~---+~~~~~~+--+4---~---+~-4-.r-~~~--,_---r---+--~---.r---140' w 0 J ~---~--~--~~~~~~r-~~~~~--~--~--~~~~--~k-+--r~~~=-~---r---r---130' ~ ~ ~--4~--1~~~~~~~~~~~+~~~~~~~~-~ BEARING ANGLES EAST SOUTH WEST Sun palh diagram lor 64 "N latitude. I Reach 5 RM 146.5-163.0 B4NL J U1 w ..J 60" Cl 2 ~~---+--~--~--~~~~--~~~~~~~r-~---+---r--~--+---r-~---T--~50· ~ ~~~-L--l---L--l~~~~~~~~~~~~~~~~~~--~--+-~---+--~~·w 0 J ~~~~~--~~~~~--~~~~~~~~~~~~~~~~~~~~~~~+---+---+-~30' ~ ~ ~~~~~~~~~~~-+--~~~~~~K~~~~~~~n~~~~~~~~~~--~~r-~~· ~ BEARING ANGLES EAST SOUTH WEST Sun palh diagram lor 64 "N lalilude. J ..•. ) Reach 6 RM 142.5·146.5 B4NL - ~--~--~--+---+---~--4---~---+---+---4---4--~r---r---r---+---;---~--~---r---iao• - ~--~--+---+---4---~--~--~---+---+---4~~--~r---r---+---+---;---~---r---r---i~ - EAST SOUTH WEST Sun palh diagram lor 64 "N lalilude. ~, 1 .. --~ . '" .. I --, l ') ---~~ 1 J c---~--1 ., ] -~. Reach 7 RM 124.0·142.5 B4NL - - - - - .un_c:> pn ~ Vi M• 21 Jut 21 =t: ~ 10 ~ pm - ~ I I ~ Bo I pm v [:\(' I I Anr 21 Ar , 21 ~ I "i-~ I ~ ~ ~pm a. I v-r I - / \ I I I ~ ,_...... l -......! ,,. v w /"\ \ I i I ,-........, I I~ prn ~ H ~,~.,71 S•D1J , I I ~ ~prn \ I .-rj-... I -A t--....... I Sr I I I I \ ,J I I I ~ v; V( I V\ I \ I I ~ I ~ I'~ -tl I I I I I I \ I rrh 11 Oo I ~I I I I I 'y 2: ~ 1-~ ~ ~ \ I I ~"'" ~ I 'i ~ I',. !!P~r-~"'"/ \ I L. ~ I ~ \ V\ j v I v ~ ~1-0 ~) ~ ~ ~ ~ ~ '"' ~ I ~ 3nm ):_ ~ 1 1 Opn ~ \ ~~ ~~ J ~ ;, I ,-· I .V'\.1"\. ~ .X ~' I'\. I'\: oo· 131:5· '10lO' 1oe;·H 71:5'B ~o~~5~ Na;; 15· M 15· 30' 45' 60' 75· M '105· 120· '13~· '15 o· ANGLES 00 , EAST SOUTH WEST Sun path diagram lor 64 •N latitude. -] eo· 70• Ul w ..J 60" (!) 2 50' <l 40' w D J 30' 1--~ ~· q: 'X:l' .. l Reach 8 RM 115.0·124.0 64NL - - - - ' - ,,. nc !on ~ ~ M•• 21 '"''' ~ t:r 10 ~ rr--pm - tn/ w ' ~ . an l i prn ~ v I H-A'"t 21 N--I "! ~ I Aor 21 I ~ ppm B• \ V( I - / I I I I ~ 1/ I I .,.....,, 7D v \ V', \ l I I ''"""' I N pm lr;: l)( l--t .~ .. 21 S•n >t I I I ~ ~ \ I rr--I - s. \ ~ ~ I pm I I I I \ II I I b; [;;' l)( ' V\ I \ I I ~ . I 1' ~ ~~ ts I I ' I I \ I I I frb 21 Ort >t I I \ v--' I{ \/ ): f'---r--N \ \ I I I N r--. ~--"'"""'!'..~ ~ \ I / ...< ["11, I ~~ ~ ~ ~ l7 I v ~ ~ ~ ~ "-' ~ 7 'tt;is / Jan2l ~ ~ ~ 9prr ~ l v-~ ~'§l 1 I l'....~ I ~""' No. ""!. ..... '\ 1'\..f~ 1 oa· 130' "UlO' 105'~ 70'B ~~.;s; Na~ 15' hl ~s· 30' 45' sa· 75· H 105· 120· 13:5· 15 0' ANGLES 00 EAST SOUTH WEST Sun path diagram lor 64 •N lalilude. eo· 70' rn w ..J 60" (!) 2 50' c:t 40' w 0 :J 30' .... ~ oo· c:t -.a· Chulitna River B4NL - - - - - .unc ion ~ Vi ... 21 IUflJ ~ ~ 10 ~ r-r--pm -'n/ w I ~ Bo I prn ~ [\/ I l---1-An• 21 t 2l t+--I ~ N I ~ Flprn a. \ ~ I - / I I I f'.... 1/ I \ t--..1 -, .. v I V', \ I I I ''""" I N pm ~~ LY vt M~tr2 s.~ 21 I I I rr\ ~~m I ~ I - s. " ~ ~ I I I I I I II I I ry V\ V( I V\ I \ I I ~ I ~ }-I I I I 1:! I I I ~~ \ \ \ ~I r.b 21 O<t 'I ' ~ y \ )\ I I ~ I I N 4am/ I I _/ ,......., I I )'.._ £!£!::. --~ I V\ ): [71 v '-..., I " ! 1"-J'~ I / I ~ 'v )ar\21 Hn~ )J ~: I ~ ~ I I ~ "-" I ~ bl..-I I ~ ~~. ~~ ~~ \O« 21 t\J"\.J~ ~~ ~~ ~~~ "~ ·'~" N. PTI I :'!o.i-~"'f., r-,..~~ :l.._~"\._"'\."'1_"\ ., 1:50' 130' 1420' 1oe·-g 7~· B ;;~ ;:; N3~ 1s· B 15' :Jc• 45' sa· 75' g 105' 120' 135• 15 c· ANGLES DC EAST SOUTH WEST Sun palh diagram lor 64 "N latitude. eo· 70' Ul w .J 60' (!) 2 50• ct 40' w 0 J 30' t--l-.J. ~· c:t '10' 1 1 Devil Creek . - -J Upper ·tsssst Lower .177771 i B4NL -----1 Ul w ..J ~ 2 ct ~--~--+---+---~--~~~~~~-+--~--~---4~~~~~~~--+---~--~--~---r--~40·W 0 J ~---~--+---+-~~~~~~~~~r+--4+--~---4~~~--~~~~~~~=-~-7~--~---430" ~ -~ ..J ~~~~rr~~~~~~~~~~+-~~---~--4~~1)~~~~~~~~~~~~~~·~ BEARING ANGLES EAST SOUTH WEST Sun path diagram for 64 •N latitude. Indian River ] Upper bS\SS) Lower V/7//1 . -~~ ---I 64NL - ~--+---~--~r---+---~--~r---+---~---4----+---~---4----+---~---1----+---~---1~---t---490' - ~--~---~----~--~---+----~--~---+----~--4----+----~--4----+--~~--;----+-----r-~4----i70" - Suo patti diagram lor 64 "N latitude. 1 ~ -----, Portage Creek B4NL ~--~--~----+----+----~--~--~~--4----+~--~--~--~~--4---~~--+----+----~---r--_, __ __,ao. ~----+---~----~--~---4----+----+----~---r--~~--4----+----+---~----~--~---1----+----1BCr ~---+----+---~----~----~---+----4-----:~--~----+----4----~----~---+----4-----~---+----+---~----~70· ({J w ..J 60" (!) 2 ct ~--+---~---+--~~--~~~~~~~--~~--4---~+--4~~~~~--~---1----~--+---~---440" w D J ~---~--~--~~~~~--~~~~~-+~--~--~--++--~~+--r~~~~+---+---+---430· r ~ +--t!----t·!--~+--H-f'-1:--f--P~"i----t--+-4 20· c:t BEARING ANGLES EAST SOUTH WEST Sun path diagram lor 64 •N latitude. 1 1 . -1 , .. -, l Tsusena Creek 64NL ~--~~-+----~--4----+----~--4-~-+--~~--4----+--~~--+---~---1,---+---~---1----t---~ao· ~--~-----~--~----~--~----+----+----+----+----+----+----+----+----+----4----~---4----4----4----~~ ~--~--~----~---+----~--~----~--~---4----+----+----~--~---4----4---~----+---~--~r---;?O· UJ w ..J ~--+---4---~---+---4--~~--+---4---~---+---4--~~--~--+---4----+---+---1----r---;s~ ~ 2 ct ~--+---4---~---+---4~~~-7+-~4--4~---+---4~~~~~=-~~~----~--+---4---~---;40-w 0 J ~---~--~--~~~~~--+-~+-~+--++---+---~--+4--~*-+--+~~~=-~--~--~~30· ~ ~ ~--~~--~~~-L~~-+--~~4+--~~---~--~~--~~~4~--~~~~~r+~-~~--~~~-~ BEARING ANGLES EAST SOUTH WEST Sun palh diagram for 64 •N latitude. 1 - - .... ..... ,... r i APPENDIX B WIDTH/FLOW FUNCTIONS I""" ! r I""' I .... WIDTH/FLOW FUNCTIONS These graphs represent the relationship of wetted river width to flow on a log/log scale. The solid lines connect HEC-2 predicted widths for the six different flows used in the R&M (1982d) simulations. The numbers associated with these solid lines are R&M cross-section identi- fiers. Several R&M cross-sections were used for each reach as defined for the SNTEMP network (refer to Figure 10). For more readable plots, several plots are presented for a single reach when necessary. The dashed line presents the flow/width function used in the SNTEMP simu- lations. B-1 b:j I N 1 1E4 STREN4 VlDTH IN FEET .I-... f- - 5 - - 2 - 1£3 -~ --- 1- 5 1- f- 2 - lE2 ~ I I 2 E3 I 3 REACH 1 CROSS SECJlOOS 106.3 -11 ~ GRAPH 1 OF 2 I -- 1 . ----11 fZI ' lf219 ;;;.. - COMPUTED WiOr:--~-------- ~ 11ZJ6. 3 n- liZJB H'l7 I I I . I I I I I I I I I I I I I I -1 I I 4 5 6 7 8 9 2 3 5 6 1 8 9 1E4 1E5 FLOV IN CUBIC FEET/SEUW I:P I w 1E4 . J STREAM VIDTH IN FEET 5 2 5 2 2 3 1E3 . ·-~] . .) REACH 1 CROSS SECTIONS 111 -114 • GRAPH 2 (If 2 - -COMPUTED WIDTH 4 5 6 7 8 9 1E4 FLOV IN WJIC FEET/SECOND 111 --~-----::::====::::: --113 112 2 3 4 5 6 7 9 9 IF.S ] REACH 2 CROSS SECTIONS 115 !'" 117 I GRAPH 1 OF 1 STREAM WIDTH lN FEET 1£4 5 2 115 1~ 4-------------------------------r-------------------~---------t 5 2 2 4 5 6 7 9 9 2 3 4 5 6 7 9 9 1~ 1£4 1E5 FLOV IN rune FEET/SECOND td I \Jl 1E4 1E3 1E2 STREAM WIDTH IN FEET ~ r- ~ ·r 5 r - - 2 r- i---- 5 - ~ ~ 2 r- I 2 1E3 REACH 3 CROSS SECTJ(IIS 119 -129 I GRAPH 1 Of 1 118 ... 12121 ~ 119 ----- COMPUTED WIDTH / I I I I I I I I I I I I I I I 3 4 5 6 7 9 9 2 3 4 5 6 7 9 9 1E4 1F.5 FLOV IN CliJIC FEETISECOOD t:d I 0\ 1E4 1£3 1E2 STREAM WIDTH IN FEET 1- 1- '- - 5 - - - 2 - -- r 1- 5 1- r- '- 2 - I 2 1£3 1 ) REACH 4 CROSS SECTIONS 120 -121 I GRAPH 1 (f 1 12(21 --~--___. __.... ----COMPUTED WIDTH 121 - I I I I I I I I I I I J J I I 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1E4 1£5 FLOV IN WJIC FEET/SEcm:J -----] -'1 J REACH 5 CROSS SECTIONS 60-64 ' GRAPH 1 Cf 2 1E4 STREAM VIDTH IN FEET ..... f-- f- 5 f- f- 2 1- 1E3 f- f- - 6121 5 r- L.. ------------sr 62 .... COMPUTED WIDTH 61 63 2 f- 1E2 I I I I I I I I I I I I I I I I 2 3 4 5 6 7 8 9 2 3 4 5 6 7 B 9 IE3 1£4 IES FLOV IN OJIUC FEET/SECOND b:l I OJ 1£4 J£3 1E2 5 2 5 2 j STREAM WIDTH IN fEET --- r 1- r ... ~ 1-- ,... i"' '- - - - I 2 1£3 ] -1 ..... 1 -1 -l 1 -·1 REACH 5 CROSS SECTIONS 65-68 1 GRAPtl 2 Cf 2 65 ~L--::------/ COMPUTED WIDTH .......... 68 I I I I I f I f r I f f I I I 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1£4 J£5 flOW IN r.tiBIC fEET/SECml ~-~ . 1 REACH 6 · CROSS SECil ONS 57 -59 1 GRAPH 1 Of 1 1E4 STREAM VIDTH IN fEE~ f- r- r- f- 5 - - - 2 - 1E3 59 f--.-::::: 59 '---/~-------·· 57 --COMPUTED WIDTH 5 - f- 2 f- 1E2 I I I I I r 1 r I I I I I I I I 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1E3 1E4 lES FLOV IN CUBIC FEET/SECOND tJj I 1-' 0 REACH 7 CROSS SECTlONS 28 -32 r GRAPH 1 lJ 6 STREAM VIDTH IN FEET l~ ~----------------------------------~--------------------------------------------~ 5 3121 2 28 32 ---l£3 31 29 5 ------COMPUTED WIDTH 2 1£2 ~------- 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1E3 1E4 1E5 FLOV IN tlJBJC FEETISECGIO -] b:l I 1-' 1-' 1 1 c --, --1 REACH 7 CROSS SECTIONS 33 -'!Is GRAPH 2 Cf 6 STREAM VIDTH IN FEET 1£4 ,-------------------------------~----------------------------~ 2 5 2 2 1£3 33 36 37 ------3j_- COMPUTED WIDfii .!-------35 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1£4 FLOV IN CUBIC FEET/SECWl lES b;j I t-' N STREAM VIDTtl IN FEET JE~ 5 2 REACH 7 CROSS SECTIOliS 38 -42 I GRAPH 3 lf 6 41 1---------------~~-42 :38 JB -~·--------------------------~-----------~~~~~~~----- 5 2 2 3 1E3 ---COMPUTED WIDTH 4 5 6. 7 8 9 JE4 FLOV IN UJBIC fEET/SECOND 2 3 4 5 6 7 8 9 lES 1 _-1 C:l I ...... w -l 1 --~ J REACH 7 CROSS SECTiotiS 43 -47 I GRAPH 4 Of 6 STREAM VJDTH IN FEET lU ~------------------------------~------------------------------ 5 2 1£3 2 2 1£3 ---COMPUTED w16fH . 3 5 6 1 B 9 1E4 FLOV m CUBIC FEET/SECOND 47 4!--,... --------45 2 3 4 5 6 7 8 9 1£5 -·) J -1 REACH 7 CROSS SECTIONS 48 -52 a GRAPH 5 (f 6 STREAM WJDTJI IN FEET 1£4 5 2 ------52 48 ..--- 1~ ~------------------------------~==~------~~----~~~---~~4g---- 5 2 2 3 1~ 51 ~-~~~~---------~50 ..------L---------------~ COMPUTED WIDTH 4 5 6 7 8 g 2 3 4 5 6 7 8 9 1E4 FLOV IN CUBIC FEETISEIDID 1ES l REACH 7 CROSS SECTIOOS 53 -56 1 GRAPH 6 OF 6 STRU..14 VJDTtt IN FEET 1E4 ..,..-.---------------r-----------------, 5 ~ 2 2 2 3 1£3 ------ko::=~~=------53 COMPUTED WIDTH 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1E4 FlOV IN CUBIC FEET/SEmiJ 1E5 ~ .. 1 REACH 8 CROSS SECTIONS 19 -23 1 GRAPH 1 (f 2 STREAM VIOTtl IN FEET 1£4 ,-------------------------------r-----------------------------~ 5 2 3 4 5 6 1 8 9 2 3 4 5 6 7 8 9 1£3 1£4 lES FLOV IN CliJIC FEETISECOOD txl I --...J 1E4 STREAM WIDTH JN fEET 5 2 5 2 2 3 1E3 REACH 8 CROSS SECTIOOS 24 -l/, GRAPH 2 (f 2 25 ----COMPUTED WIDTH 27 t----~--~==--~~=24 4 5 6 7 8 9 2 3 5 6 7 8 9 1E4 FLOV IN CI.IHC FEET/SECCW . 1 lES ~·· .. 1 REACH 9 CROSS SECTilllS 3 - 7 z GRAPH 1 {f 3 STREAM VIOnt IN fEET 1E4 ~------------------------------~----------------------------~ 5 3 4 5 2 6 7 --1E3 ----5 -COMPUTED W[DTH 2 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1E3 1E4 1£5 FLOV IN UIBIC FEET/SEC£Nl --J -J . J REACH 9 CROSS SECTIONS 8 -12 1 GRAPH 2 lf 3 STREAM VIDUI IN FEET 1£4 - f- 1- !- 5 :- '- - 2 f- --JE3 -----t0 ----/---.-8 -11 ----v COMPUTED WIDTH 12 -5 I !- 9 r- 2 - 1E2 I I . I I I I 1 I I I I I J I I I 2 3 4 5 6 7 B 9 2 3 4 5 6 7 8 9 JE3 1E4 JES FlOV IN llJIJIC FEET/SEcmJ -. J 0::1 I N 0 1 _-·· -~~ ---] STREAM WIDTH IN FEET J£4 - r- 1- r- 5 r- - r- 2 - 1£3 1- r- 1- - 5 - r- 1- 2 r- 1E2 I I 2 3 JE3 REACH 9 CROSS SECTIONS 13 -18 1 GRAPH 3 OF 3 14 16 .--:::::::::.. 1 7 ~~15 ~ -:...-----/13 ~ ::;:;;;-: ::::;::::;-- ~------~ 18 COMPUTED WIDTH ~!----- I I I I I I I I I I I I I I 4 5 6 7 8 9 2 3 4 5 6 7 8 9 1£4 JES FLOV IN OJBIC FEET/SECOND -' .... I -I ,_ I - APPENDIX C LONGITUDINAL TEMPERATURE PROFILES JUNE TO SEPTEMBER 1981-1982 - - .... ' - -' - - - LONGITUDINAL TEMPERATURE PROFILES JUNE TO SEPTEMBER 1981-1982 These graphs represent both the predicted and observed temperatures for the June, July, August, and September period of 1981 and 1982. The observed data points are shown with 95 percent confidence intervals. These confidence intervals are measures of the monthly variations in the usable historical data for the Susitna Basin (Figure 13). Predicted temperatures are from the postcalibration simulations with SNTEMP. C-1 . -) ('") I N 20 15 10 5 0 .. l Predicted Longitudinal Temp. Profiles June 81 (957. confidence intervals) T emperoture (C) I I ----------_ ___..__........ ---- ~--~ -------~~--~--~~ lL---- 190 180 170 160 150 140 130 120 110 9~ . 80 River Mila ("') I . w 20 15 10 5 Temperature (C) 190 180 170 Predicted Longitudinal Temp. Profiles July 81 (95% confidence intervals) 160 150 140 130 120 110 90 80 River Mile (') I ~ 20 15 10 5 J ... 1 --1 . J . ] Predicted Longitudinal Temp. Profiles August 81 (95% confidence intervals) Temperature (C) r~--~-+}--Irl'f~---:rnl~~ 190 180 170 160 150 140 130 120 110 100 90 80 River Mila (") I ln 20 15 10 5 J ...... J .1 .. -J Predicted Longitudinal Temp. Profiles Saptambor 81 (95% confidence intorvols) T omporatura (() I 1~----~I~ri~I----~~~~ 190 180 170 160 150 140 130 120 110 90 80 River Mi Ia 1 -1 -1 .. -] . ) .l Predicted Longitudinal Temp. Profiles June 82 (95% confidence intervals) Temperature (C) 20 15 10 5 190 180 170 160 150 140 130 120 11~ 80 River Mile (") I -...I 20 15 10 5 T emparature (C) 190 180 170 Predicted Longitudinal Temp. Profiles July 82 (95t confidence intervals) I I I l-- 160 !50 140 130 120 1H1 100 90 80 River Mile ("') I CXl 20 15 10 5 .... J. Predicted Longitudinal Temp. Profiles August 82 (95% confidonco intervals) Temperature (C) 190 ---~------_JI __ _JI-rr-;--I 180 170 160 150 140 130 120 ' 110 River Mile ] lL--- 100 90 80 -1 (') I 1.0 ,, 20 15 10 5 _,, l ~r'"" ''] J ,-, T ompe~ctu~e {C) 190 180 170 '1 r--I ~-----' 1 --· -, ----, ,,,_,, 1 ., ·=---J --1 l Predicted Longitudinal Temp. Profiles Septombe~ 82 (95% confidence intervals) I I I I 160 150 140 130 120 110 90 80 Rive~ Mile