Assessment of Hydrologic Impacts of Spatial Discretization in Koshi River Basin by GIS and HEC-HMS Semi - Distributed Model

Most hydrological modeling are complex process to recognize due to immense spatial and temporal variability of catchment characteristics, vegetation distinctiveness, precipitation data, snow pack and number of concerned variables in modeling the physical processes. The hydrological models are useful for spatial and temporal extrapolation of hydrological data. These models help in admitting the conceptualization of governing hydrologic methods working in catchments. Hydrological models are mainly classified as lumped, semi-distributed and distributed models. The model describing catchment in terms of average spatial and temporal quantities is considered lumped model while describing spatially variable system are termed "distributed models". Hydrological models that fall between lumped and distributed is known as "semi distributed models". Semi-distributed hydrological models generally have advantages of short calculation time, low calibration needs and high model efficiency. The sensitivity of this model in simulating the rainfall runoff, as a role of sub basin descriptions, was investigated for the Koshi river basin by using digital elevation model, climate data, land use and soil type of Koshi watershed. The results showed that increasing the number of subdivision from 7 to 35, decreases hydrograph peak discharge, average stream flow, total runoff volume but increasing efficiency of model with increasing sub basin divisions. However this trend becomes less significant when number of subdivisions exceeds twenty-five subdivisions.