Simulation of Runoff Using SWAT Model at Annual Scale (Case Study: Neyshabour Watershed)

Runoff is a main component of water balance in the watershed scale which its accurate estimation helps to better hydrological simulation of watershed. In this study, the SWAT (Soil Water Assessment Tool) model was used to simulate runoff in Neyshabour watershed with area of 9500 km2 for the 2007 to 2010 period. To this means, SWAT model was calibrated based on measured runoff data at three hydrometric stations (Andarab, Kharve Majmoo and Hossein Abad) during 2000-2007. Calibration, uncertainty and sensitivity analysis were done using SUFI-2 algorithm. According to the results, parameters such as fraction of transmission losses, moisture condition curve number and factors related to snowmelt showed the greatest sensitivity to run off estimation. The model calibration in all three stations; Andarab, Kharve Majmoo and Hossein Abad, was evaluated as appropriate by referring the values of 0.84, 0.77 and 0.92 for the Nash-Sutcliff efficiency coefficient, respectively, although underestimations for the peak amounts of hydrographs were observed relatively. The values of the mentioned coefficient were obtained as 0.92, 0.66 and 0.71 for the validation period at the stations of Andarab, Kharve Majmoo and Hossein Abad, respectively. Also root mean square error values varied between 0.46 to 0.57 m3/s and 0.06 to 0.19 m3/s in the same three hydrometric stations for the calibration and validation periods, respectively. The results showed that considering the crop management concepts significantly improved runoff estimation.