Effect of Soil Map Scale on the Results of SWAT Model in Monthly Runoff Simulation of Tuyserkan Watershed

Soil erosion models are useful tools to predict runoff, sediment and soil erosion in watersheds. Although, the SWAT model is used for evaluating runoff discharge and the long-term effects of management operation on water, sediment and agricultural chemical yields in the large watersheds, in this research, its efficiency was investigated in monthly runoff simulation for Tuyserkan watershed (with an area of about 3800 ha) and also the effect of soil map scale on the simulation results was studied. In order to study the effect of soil map scale, SWAT model was run with three different map scales (1:40000, 1:250000 and 1:1000000), in which all input models (such as land use map, digital elevation map, number of simulations and effective primary parameters in simulation results) were same, except soil map scale. The SWAT model calibration and uncertainty analysis for each scale in a period of 1998 to 2008 were taken using SUFI-2 program. The values of statistical indices (R2, Nash-Sutcliffe coefficient, p-factor and r-factor) for each of the three map scales were virtually the same. It seems that the key factor in determining the impact of soil map scale on the SWAT results, is not scale number of the studied map, but soils diversity plays an important role on the results of the model.