HR: 16:55h
AN: H54B-03    [Abstracts]
TI: Evaluation of Snowmelt-Dominated Forest Hydrology Simulations in a Mountainous Watershed with the Swat Model
AU: * Ahl, R S
EM: rob.ahl@umontana.edu
AF: The Univeristy of Montana, Department of Forest Management, College of Forestry and Conservation, 32 Campus Dr. , Missoula, MT 59812, United States
AU: Zuuring, H R
EM: hans.zuuring@cfc.umt.edu
AF: The Univeristy of Montana, Department of Forest Management, College of Forestry and Conservation, 32 Campus Dr. , Missoula, MT 59812, United States
AU: Woods, S W
EM: scott.woods@cfc.umt.edu
AF: The University of Montana, Department of Ecosystem and Conservation Sciences, College of Forestry and Conservation, 32 Campus Dr. , Missoula, MT 59812, United States
AB: The Soil and Water Assessment Tool (SWAT) has a long track record of successful application in lowland and temperate environments, but little is known about the model's performance in forested mountain watersheds where hydrologic patterns are dominated by seasonal cycles of snow accumulation and melt. In this study, the ability of SWAT to simulate annual, monthly, daily, and seasonal streamflow in a snow-dominated upland watershed that represents conditions commonly found in high elevation environments in the Rocky Mountains of North America was evaluated. The model was calibrated with 4 years of continuous daily climate data collected within the research watershed, and corresponding streamflow records measured at the defined outlet. Hydrologic flow predictions were assessed graphically and with relative error (RE), mean paired deviation (DV), and Nash- Sutcliffe model efficiency (NS) statistics. The calibrated model was validated with an independent dataset spanning an additional 4 years, using both traditional performance criteria obtained over the calibration and validation time periods, and objective regression-based methods. After calibration, model performance was very good, with a relative simulation period error of 2%, mean deviations of 36% and 31%, and Nash-Sutcliffe efficiencies of 0.90 and 0.86 for monthly and daily streamflow predictions, respectively. Model predictions were validated over annual, monthly, and daily time steps using both traditional and objective procedures, with RE 4%, DV 43 and 32, and NS 0.90 and 0.76. Seasonally, SWAT performed well during the snowmelt-induced runoff periods, but could not be validated for baseflow simulation. Assessment of key factors indicated that adjustment of snow process parameters contributed most significantly to model calibration. Other important parameters were surface runoff lag, groundwater, soil, and curve number parameters, in decreasing order of influence. Ultimately, model results indicate that when calibrated SWAT can predict annual, monthly, and daily hydrologic processes in forested mountain watersheds with efficiency levels that are similar to those obtained in other regions where it has been applied.
DE: 1846 Model calibration (3333)
DE: 1847 Modeling
DE: 1860 Streamflow
DE: 1863 Snow and ice (0736, 0738, 0776, 1827)
DE: 1879 Watershed
SC: Hydrology [H]
MN: 2007 Joint Assembly