For stability and river basin management and preventing soil and sediment losses in river basins, as well as conducting catchment conservation plans, it is necessary to estimate suspended sediment content in rivers. In the present study, in order to select the best method of suspended sediment estimation at Barzoo, Baba Aman, Ghatlesh, Darband, Aghmazar, Tabarok Abad hydrometric stations located in Atrak river basin, corresponding data of flow and sediment discharges in the available statistical periods , were collected and analyzed. By creating a relationship between corresponding water and sediment discharges based on five linear, combined linear (multiline), linear with applied correction FAO, mean load within discharge classes and schematic method (tangent to maximum concentration line), we have selected an appropriate model which has the best predictability potential, based on the statistical indices. Therefore the index of root mean square error (RMSE), coefficient of determination (R2), mean estimated to observed data ratio (r), coefficient of variation (CV),ratio of the root mean square error to mean estimated data (GSD), the relation coefficient between estimated and observed sedimentation (R), mean absolute error (MEA), and mean bias error (MBE), were used. Results showed that the model of mean load within discharge classes at Tabarrok Abad station, linear model at Aghmazar and Barzoo stations, graphical models (maximum concentration) at Baba Aman and Ghatlesh stations, hybrid models (multiline) at Darband stations have the lowest mean square error and the best predictability in estimating sediment content among the tested models at the studied stations.

Keywords