Analyzing Effects of Two Different Land Use Datasets on Hydrological Simulations by Using SWAT Model

Abstract

Land use/cover (LULC) dataset can be addressed as one of the significant factors governing watershed hydrology. Thus, there is a necessity for using appropriate LULC data especially while working on the hydrology of sensitive watersheds from which drinking water is provided. The aim of this study is to analyze the effect of using two different LULC datasets on the hydrological components of a basin in Turkey. Omerli Basin that covers one of the important drinking water reservoirs of Istanbul with a drainage area of 621 km2 is selected as the study area. The Soil and Water Assessment Tool (SWAT), a physically-based hydrological and public domain model, is used as the hydrological model to evaluate the water budget of the basin. SWAT model is calibrated with an automatic calibration and uncertainty analysis program (SWAT-CUP) by using SUFI-2 algorithm. Model setup is conducted for two different LULC datasets which are CORINE obtained for year 2006, and LULC data produced from 2006 dated Landsat satellite image. This pioneer study evaluates and compares the impact of CORINE and Landsat LULC datasets used in hydrological modeling by using SWAT model at sub-basin level. It revealed that different LULC datasets yielded quite similar results in terms of surface water runoff and actual evapotranspiration in the Omerli Basin; however, different spatial distribution was observed especially in urbanized sub-basins. Despite the coarser land-use classification in Landsat data, it enabled the detection of different LULC classification with higher spatial resolution, and thus, different model outputs were achieved especially in the urbanized sub-basins.