Quasi Two-Dimensional Modeling of Flow Hydraulics and Bed Load Transport in Zaremrood River

Determining the amount of sediment carried by rivers is important in several ways. This parameter is effective in design of dimensions and geometric characteristics of flow regulation and diversion structures, reservoir dams as well as pumping stations. In this study, the calculation of flow discharge and bed load of Zaremrood river located in Mazandaran province has been investigated using Shiono and Knight quasi-two-dimensional model. This model is based on the Navier-Stokes continuity and momentum equations and has been simplified by depth averaged concept. For this purpose, using the finite element method, this model was solved numerically and the lateral velocity distribution was calibrated at the Garmrood hydrometric station. Comparison of obtained results by Shiono and Knight model in different flow discharges against measured data indicates the high accuracy of the model for lateral velocity distribution. Then, by using the computed lateral velocity distribution, the distribution of bed load across the river was simulated. The results showed that among the 17 empirical bed load equations selected in this study, the Duboy formula (1879) has the best accuracy in both one and quasi-two-dimensional modeling cases. In 1D modeling case, this formula with standard deviation of the discrepancy ratio of 0.34 percent had better agreement with the measured bed load in comparison to the Frijlink (1952) and Meyer-Peter and Mueller (1948) equations with the standard deviation of 3.46 and 7.32 percent, respectively. In 2D modeling bed load transport, the root mean square error (RMSE) was obtained 7.45, 98.8 and 172.9 for three equations of Duboy, Frijlink and Meyer-Peter and Mueller, respectively which indicates that only Duboy formula has an acceptable accuracy while Frijlink and Meyer-Peter and Mueller equations have large errors. The results also showed that the bed load transport in quasi-two-dimensional model using Duboy equation is more accurate than one-dimensional case.