The study area selected for the present study is Upper Krishna river catchment having an area of ~15000 sq-km whose average annual rainfall is 1300 mm. Spatial variation of the rainfall in the basin is considerably high which varies between 500 and 6200 mm. Land-use land-cover map of scale 1:250000 obtained from National Remote Sensing Centre, India indicates that forested and agriculture lands dominate most of the catchment. Considering the availability of the observed hydro-meteorological data, data of 2002-2005 and 2006-2007 were used for calibration and validation respectively. The rainfall data used for the present analysis is IMD gridded daily rainfall whose spatial resolution is 0.25 degree. Only one discharge site at Kurundwad is available in the selected basin. Hence, the same obtained from Central Water Commission (CWC), India used for calibration and validation. Similarly, SRTM 30m DEM was used to extract topography related information using HEC GeoHMS (a geospatial extension of HEC-HMS). Due to non-availability of high-resolution soil maps, FAO Soil map of scale 1: 5000000 was used to generate soil related parameters.
For representing the various hydrological processes of the study catchment, considering the availability of data, Soil Moisture Accounting (SMA) method is used as a loss model for estimation of excess rainfall. The Clark unit hydrograph method is used for conversion of runoff volume (excess rainfall) to discharge hydrograph at outlet/ discharge station. Base flow is modeled using exponential recession method and finally, all the modeled stream flow components were routed using Muskingum routing method.
Considering the hydrometerological characteristics of the study area, sensitivity analysis of one parameter at a time has been performed to identify the sensitive parameters during both wet as well as dry seasons. Thus obtained results indicate the parameters such as soil storage, soil percolation, maximum infiltration rate, percentage impervious area, and tension soil storage in the five most sensitive parameters in both the seasons however, their sensitivity levels are different. Similarly, season specific model calibration and validation model has been carried out for both the seasons and evaluated using the measures of Nash-Sutcliff Efficiency, Coefficient of Determination, Percentage Error in simulated volume and peak flow.
Comparision of the season specific simulations and the simulations without considering the seasonal effects indicates that the season specific simulations are far better during both calibration and validation runs. Hence, it can be concluded that the HEC-HMS modeling software tool can be used for continuous hydrological modeling over a large tropical catchments of India such as Upper Krishna Basin, however with season based calibration for better results.