Automatic calibration of SWAT hydrological model by SUFI-2 algorithm

Abstract

In recent development, scientists have shown increasing interest in applying geographic information systems (GIS) and satellite-based remote sensing (RS) techniques to the hydrological modelling. This study incorporates satellite-based RS and GIS techniques to estimate the spatial variation of the hydrological parameters, which are used as input to the model. The hydrological model is calibrated to resemble the hydrological cycle of the watershed. The model SWAT is calibrated by SUFI-2 algorithm performing sensitivity analysis and then validation is done successfully. Four sensitive parameters namely CN2, ALPHA_BF, GW_DELAY and GWQMN are chosen to calibrate the model for surface runoff over sub-basin number 9. After the rigorous permutations in the algorithm, the respective values of sensitive parameters are obtained as 75, 0.7, 240 and 1.00. The NSE for the run is 0.90. The watershed of Western India is taken to demonstrate the work.