Determination of the Best Model for Estimation of Suspended Sediment by using Statistical Error Criteria (Case study: Some Sub-Watersheds of Kashaf Roud)

The phenomena of erosion, sediment transport and sedimentations have tremendously destructive effects on environment and hydraulics structures. In general, sediment transportation depends on river discharges, but the proposed equations inherited with large errors. To evaluate the suspended sediment loads and an optimized model on them, in this research, data were collected from some sub-watersheds of Kashaf Roud basin (including Moshang, Golmakan, Dowlat Abad-Khoram Dareh, Saraseeyab-Shandiz, Golestan-Jaghargh, and Hesar-Dehbar). To decrease the error terms of equation model of sediment, we have used the combined methods in regression analysis based on temporal classification of discharge data and their sediment in a period of 30 years. Different base time models were examined using 12 statistical criteria, to select an optimal estimation of suspended sediment loads. The results showed the best equations are hydrological, vegetation and climatological based models, respectively. So, these models had the lowest error, selected as the optimal models for estimation of suspended sediment loads. But annual-based model (without any classifications on discharge and sediments data), was among the most insufficient models. In general, the error rates in sediment estimation for dry months with water shortages were less significant than the wet months with high water supplies. In these months, the effectiveness of suspended sediment loads of discharges were more pronounced to compare to other sediments transferring agents. Based on the results of this paper, it is recommended to evaluate more watersheds to reach to a precise model for estimation of sediments rating curves.