Comparison of Automatic Extraction of Sediment Delivery of Watershed and Traditional Method in Geographic Information System (Case Study: Yekechenar Watershed –Golestan Province)

Determining the sediment delivery rate and watershed delineation is among the first stage in environmental research, especially in water erosion estimation. The determining the accurate boundary of watershed is important for the hydrological and morphologic characteristics of watersheds. The study aims to present an automatic extraction model of watershed delineation and calculate sediment delivery rate. The traditional method of determining the watershed delineation, and subsequently, calculating sediment delivery rate is performed manually that using a topographic map, the boundary of the watershed is determined and main flow length is calculated. However, nowadays due to the advancement of digital analytical spatial-based methods in GIS software, automatic delineation of watershed is feasible. For this purpose, the errors of digital elevation model are removed. After calculating the flow direction, and flow accumulation, watershed boundaries can be determined having pourpoints. Then, sediment delivery rate is calculated by local height of watershed, its perimeter, and flow length. In this research, the digital elevation models of ASTER-DEM and SRTM-DEM are utilized to design watershed delineation model and to evaluate the overall accuracy and the correspondence of them, so the border of watershed is delineated using Google map, which is used as a ground truth. The findings of this study show the automatic extraction of watershed boundary and calculation of sediment delivery rate do not have significant differences with traditional mothed. So that, the overall accuracy and Kappa index of the watershed boundary based on ASTER-DEM are 93 percent and 0.92, respectively, and their values for the watershed boundary based on SRTM-DEM are 94.3 percent and 0.94, respectively. The correlation coefficient (r2) of calculated sediment delivery rate based on SRTM-DEM is 0.98 and its value based on ASTER-DEM is 0.95.