GIS−Based Hydrologic Modelling in Taleghan Watershed

Modelling hydrological processes accurately has usually occurred as a challenge, especially in mountainous watershed due to complex governing process and lack of hydrometry data sets. Therefore, precise estimation of water resource potential has been usually difficult, especially for large watersheds. The purpose of this study is assessing the capability of the SWAT models for stream flow simulation in Taleghan watershed under different soil types and land use classes. To this end, water and meteorological data, land use, soil and digital elevation model layer were prepared in accordance with the standards and were used for parameter sensitivity analysis, model calibration and validation. A sensitivity analysis was performed; using LH_OAT method to evaluate stream flow variations in response to changes in model input hydrologic parameters. The curve number, soil available capacity, snow melt base temperature and evaporation compensation factor were found to be the most sensitive parameters. Model calibration for the period 1997 through 2002, and validation was performed for 2003 through 2005. The model performance was evaluated by visual comparison of time series plots and well-established statistical methods, namely Nash-Sutcliffe efficiency (NS), coefficient of correlation (), percent bias (PBIAS), and the T-test for student. The results of monthly calibration NS, and PBIAS values for runoff at the watershed outlet were respectively 0.8, 0.93 and -14.8; and for validation, these statistics were 0.84, 0.88 and -2. T-test results also showed that there was no significant difference between the observed values and simulated ones. In general, the results of the study have exposed the ability of SWAT model for flow forecasting in Taleghan watershed.