Author/Authors :
Kiyani Majd ، Mahnaz Department of Rangeland and Watershed Management - Faculty of Water and Soil - University of Zabol , Nohtani ، Mohammad Department of Rangeland and Watershed Management - Faculty of Water and Soil - University of Zabol , Dahmardeh Ghaleno ، Mohammad Reza Department of Rangeland and Watershed Management - Faculty of Water and Soil - University of Zabol , Sheikh ، Zahra Department of Combat Desertification - Desert Studies Faculty - Semnan University
Abstract :
The digital elevation model (DEM) map plays a crucial role in the SWAT model, significantly impacting its output results. This study examines the influence of various accuracies of Digital Elevation Models (DEMs) with resolutions of 12.5, 30, 50, 90, 450, and 1000 meters on runoff simulation using the SWAT model for the Lar watershed. The model was simulated over a 30-year period (1988-2017), with the initial three years allocated for model warm-up, the subsequent period (1988-2005) for calibration, and the period from 2006 to 2017 for validation. Model accuracy was assessed using Nash-Sutcliffe efficiency (NS), coefficient of determination (R²), R-factor, and P-factor. The runoff simulation results reveal that variables such as area, sub-basin, Hydrologic Response Unit (HRU), watershed slope, and average channel slope are particularly affected by DEM accuracy. With increasing DEM accuracy, the length of the main channels decreases, and lower-order channels are eliminated, leading to a reduction in the calculated runoff depth. During validation and calibration, DEM resolutions of 12.5 and 1000 meters exhibited lower accuracy compared to resolutions of 30, 50, 90, and 450 meters, suggesting that extremely high or low spatial data accuracy does not enhance simulation accuracy. Additionally, minimal differences were observed among the results for DEM resolutions of 30, 50, 90, and 450 meters, which can be attributed to adjustments in calibrated variable values and the application of the SUFI-2 method across different DEM accuracies. Finally, based on the results of this research, the DEMs with spatial resolutions of 30, 12.5, 50, 90, 450, and 1000 meters demonstrated the best performance for simulating monthly discharge in the Lar Watershed. Additionally, based on the value of the objective function (NS) during the calibration and validation stages, it can be concluded that the SWAT model can simulate the monthly discharge of the Lar Watershed with acceptable accuracy. Therefore, it can be said that the SWAT hydrological model can be used in arid regions to implement management scenarios quickly and at a low cost for decision-making.
Keywords :
SWAT , CUP , DEM , SUFI , 2 method , Nash , Sutcliffe efficiency , SWAT hydrological model
