| Stream flow simulation, influenced by snow melting processes, has been studied for the past decades. Studies have shown that snow is variable in shorter time scales (daily and hourly) and rapid snowmelt can be a cause of flooding. The current hydrological models use a temperature driven snow melting algorithm. This efficient although simplistic algorithm does not accurately represent the state of the snowpack which could lead to large errors in the calculations of the snow melting amount and consequent streamflow simulation. Accurate estimates of snowpack properties are essential in order to have precise prediction of rapid snowmelts and consequent streamflow events. Assimilated distributed meteorological data is used to better capture the spatial variability of the meteorological parameters. Using the hydrological response unit (HRU) approach, the energy and mass balance snowpack model (Snow Thermal Model) is spatially distributed within a watershed. These results are incorporated into a hydrological model (Soil and Water Assessment Tool) to improve discharge simulation. The new model (named SWAT-SNT) was tested in the West Branch Delaware River. The results obtained with SWAT-SNT showed better agreement to the observed streamflow at the USGS station than the standard SWAT. SWAT-SNT RMSE was 6.8 m3/sec lower, and R2, NSE, and KGE were 14%, 21%, and 17% higher respectively. |
