You are here:
HYDROLOGIC AND STREAM TEMPERATURE MODELING FOR ANADROMOUS FISH HABITAT RESTORATION IN A WILDLAND WATERSHED
Citation:
Chen, Y. D., S C. McCutcheon, W. L. Nutter, AND R F. Carousel. HYDROLOGIC AND STREAM TEMPERATURE MODELING FOR ANADROMOUS FISH HABITAT RESTORATION IN A WILDLAND WATERSHED. Presented at American Water Resources Association Annual Summer Specialty Conference on Wildland Hydrology, Bozeman, MT, June 30-July 2, 1999.
Description:
Reduction or removal of streamside vegetation by logging and grazing can alter stream temperatures by reducing riparian shading. In the Pacific Northwest of the United States and other parts of the world, elevated stream temperatures in summer are a major fish habitat degradation problem that affects coldwater species such as salmon and trout. For example, the lethal temperature for Chinook Salmon is approximately 26oC, and sublethal effects on juveniles can occur at significantly lower temperatures. Projects to restore riparian forest cover are often intended to reestablish shading and reduce stream temperatures to levels that can support coldwater communities. To provide guidance for riparian vegetation restoration activities, comprehensive and dynamic information about stream temperature regimes can be cost-effectively generated by watershed-scale, continuous stream temperature modeling. The Hydrologic Simulation Program - FORTRAN (HSPF), a major watershed modeling tool developed and supported by US Environmental Protection Agency (USEPA) and US Geological Survey (USGS), together with its supporting data management programs and expert system software for model calibration, form a comprehensive watershed hydrology and water quality modeling system that may be used to conduct hydrologic/hydraulic and stream temperature simulations.
Solar radiation is the primary source of energy for stream heating. To simulate stream temperature dynamics, hillslope topographic shade angles and geometric dimensions of riparian vegetation buffers (if any) must be used for estimating the amount of solar energy that actually reaches the stream surface. Because of the limitation of modeling technology and the lack of basin-wide riparian shading characteristics, watershed-scale and continuous-based simulation of stream temperature in forest basins was not possible in the past. Traditional field observations using solar pathfinder and other equipment can only estimate a few averaged parameters of riparian shading characteristics as input to reach-scale stream temperature models. With the advances in remote sensing and GIS technologies, extensive data representing the detailed characteristics of stream channel, hillslope topography, and riparian vegetation buffers can be effectively created and processed. In order to use the GIS-derived riparian shading characteristics for watershed-scale stream temperature modeling, a computer model called SHADE was developed for computing the shading dynamics and thus the effective solar energy for stream heating. The SHADE-generated radiation data are used by HSPF to simulate hourly stream temperatures that can be used to target critical stream reaches for fish habitat restoration and to evaluate the effectiveness of replanting riparian vegetation. The SHADE-HSPF modeling system was calibrated, and then validated for the Upper Grande Ronde (UGR) watershed in northeast Oregon, USA.
The module section HTRCH in HSPF is a one-dimensional code for simulating water temperature of each reach of a stream network. In HTRCH, the energy budget analysis technique is employed to determine the net heat exchange for simulating stream temperature dynamics. However, HTRCH did not originally have the capability to provide a realistic estimate of the amount of incoming solar radiation that actually enters stream water, due to the lack of adequate algorithms for vegetation and topographic shading computations. Therefore, SHADE was developed for dynamically estimating the contribution of riparian vegetation buffers and topography to stream surface shade. SHADE, together with, HSPF and its four supporting programs (ANNIE, METCMP, SWSTAT, and HSPEXP) form a watershed hydrology and stream temperature modeling system. In the SHADE-HSPF modeling system, GIS-derived topographic and vegetation shading characteristics are used by SHADE for computing the width of shade which is compared against the stream surface width (TWID, simulated by the hydraulic module section HYDR in HSPF) for estimating the percentage of shaded stream surface at each time interval. By doing so, SHADE computes the amount of effective solar radiation for stream heating, as indicated by the input of unadjusted radiation and the output of adjusted radiation for stream temperature simulation by HTRCH in HSPF. SHADE also generates solar radiation factor (SRF, the ratio of radiation effective for stream heating divided by the incoming radiation before any reduction by shading) values that can be used to characterize the dynamic shading conditions.
The SHADE-HSPF modeling system was applied to simulate the watershed hydrology and stream temperatures in the UGR watershed. Simulation results confirmed the accuracy and robustness of the modeling system. To identify the possible causes for reducing the high summer stream temperatures, the impacts of hydroclimatic shifts and hypothetical riparian vegetation buffers were evaluated. Simulations demonstrated that natural weather cycles of ?10% or ?20% in air temperature, solar radiation, and precipitation can not sufficiently moderate the stream temperature regimes to insure the survival and reproduction of salmon. Therefore, riparian vegetation is the only critical factor that can be managed to alleviate significantly the lethal and sub-lethal stream tempera-tures. Stream temperature forecasts for restored riparian buffers demonstrate that 44 out of 51 reaches in the watershed could achieve a no-effect standard that includes a maximum summer temperature of 16oC and average 7-day maximum temperature of 14.5oC. Downstream reaches (the 5th order) on the mainstem Grande Ronde River are too wide to be sufficiently shaded by restored buffers to meet this standard. The creation of thermal refugia or other management practices, inaddition to riparian restoration, may be required if studies of the threatened salmon species show that the lower mainstem is the critical habitat.
Simulated maximum values of stream temperature, on which the riparian restoration forecasts are based, are accurate to 2.6 to 3.0oC. Hourly simulations have approximately the same accuracy and precision. The phase, diurnal fluctuations, and day-to-day trends in stream temperature simulations are very good, confirming the validity of shading computations and the observed air temperatures. Occasionally, the model conservatively oversimulated, especially in localized areas where cool ground-water in-flow may dominate. The difference in the spatial resolution between the reach-averaged simulations and the point measurements of stream temperature, together with other data uncertainties such as the limited precision and accuracy of riparian shading characteristics and the lack of extensive channel morphological data, caused some systematic simulation errors for over a third of the 27 calibration sites (11 sites in 1991 and 12 sites in 1992).
The application study demonstrated the validity and usefulness of the SHADE-HSPF modeling system. With the SHADE-generated solar radiation data, HSPF can accurately simulate reach-averaged stream temperatures at the watershed scale by accounting for the riparian shading characteristics and the thermal impacts of basin-wide land cover on the runoff temperatures. Compared to the 8 to 10oC violations of the temperature standards under the present riparian vegetation conditions in the Upper Grande Ronde, the model accuracy of 2.8oC is more than adequate to assess riparian restoration and management scenarios. To apply the new SHADE-HSPF modeling system for other forest watersheds, model runs for different kinds of specific management objectives can be made easily in the future.