Simulation of surface runoff using GIS-based model of Limburg soil erosion model (LISEM)

Simulation of runoff resulted from any significant rainfall events can help to estimate the severity of floods. To protect watersheds, application of a model which is able to predict runoff at certain times and places is essential. Therefore, this study aimed to simulate runoff generation in upstream catchments of the Shahrake Behzisti as well as assessing the effect of implementation of vegetation management scenarios, using three relatively intense rainfall events occurred in recent years. The model of physically based Limburg soil erosion model (LISEM) within PCRaster GIS and programming language was applied in this study. In order to run the LISEM model, four basic raster maps of digital elevation model, landuse, soil texture and roads (impervious areas) were prepared. Digital elevation model map of the catchment was extracted from the Ultracam image of 1 m resolution within the ArcGIS environment. During this study, only two events on 2/11/2014 and 16/9/2015 led to runoff generation. Thus, the first event was used for calibration and the second one for validation of the model. Statistical comparison of the simulated and observed hydrographs of the rain events on 2/11/2014 and 16/9/2015 showed the correlation coefficients of 0.80 and 0.76 and Nash-Sutcliffe criteria of 0.8 and 0.5 respectively. The results of the simulation showed that the model was able to predict total runoff, peak discharge, runoff initiation time and time to peak discharge with relatively good performance. The results of the application of the LISEM model showed that runoff volume and peak discharge significantly decreased by implementation of vegetation management scenarios.