UNDERSTANDING THE HYDROLOGICAL BEHAVIOR OF GADARWARA WATERSHED USING SWAT MODEL

Abstract

For the effective planning and sustainable management of land and water resource in a river basin, proper assessment of water resource is required. The assessment of water resource can be done effectively with help of appropriate well calibrated hydrological model. The present study is undertaken with an aim to test the performance of SWAT model Shakkar River located at located at Gadarwara Watershed at Narsinghpur District in Madhya Pradesh state India. Shakker River is the left bank tributary of Narniada River. Total drainage area is 2183.24 km2. For model application, the catchment area is divided into 53 sub-watersheds. The whole catchment area is classified into 6 land use categories. Available hydrological data (i.e. loon 1995-2002) was split into two groups for calibrating and validating parameter of the model. The model is calibrated at Gadarwara gauging site on monthly basis data of stream flow and sediment concentration. The model was auto-calibrated and -validated using SWAT cup SUFl-2 software. Model performance was analyzed based on visual comparison of observed and computed stream flow and sediment concentration as well as on statistical measures based on Moriasi et al (2007). Comparative analysis of result is done on monthly time scale analysis reveals that during calibration period, the Nash Sutcliffe efficiency for stream flow and sediment are 0.81 and 0.71, respectively. For the validation period NSE for stream flow and sediment are 0.65 and 0.71, respectively. Gadarwara gauging site show good agreement between observed and model computed hydrograph. The detailed study was carried out to understand sensitivity of model parameters in term of model output such as water yield, surface runoff and baseflow to gain depth understanding of the role of different model parameter affecting water yield are CN2, SOL_K and USLE_P. On the basis of sensitivity analysis of model parameters, the ranking of most sensitive parameter from highest sensitive to relative lesser sensitiveness on stream flow and sediment concentration is CN2, SOL_K, USLE_P, GW_DELAY, ALPHA_BF, GWQMN, SLSUE3BSN, CH_ERODMO and USLE_K so on. Barren and range land from 16 sub-basins from watershed were found the most sediment loss contributing areas in the watershed. Sub-basins 47, 51,42 and 38 were found the most active areas in the watershed due to high slope and hilly area. The 21% of barren and range land were contributing the 58-60 % sediment loss loon watershed. Overall, based on visual and statistical evaluation, it can be concluded that the SWAT model can be used with confidence to simulate stream flow and sediment concentration response in Gadarwara basin on monthly basis.