Sensitivity Analysis of Total Sediment Load Transport Equations in Rivers

In order to estimate the sediment transport discharge in rivers, empirical equations commonly are used. Variables of these equations are flow and sediment properties, such as flow depth and velocity, sediment grain size, or other properties. Estimation of sediment transport by using these equations is based on the measurement of the physical properties. In engineering applications, measuring errors of these properties affect the accuracy of sediment fluxes. The present study quantifies error propagation from the input properties and investigates its effect on the sediment transport calculations. This analysis determines the sensitivity, strengths and weaknesses of four total load equations, including: Engelund and Hansen, Shen and Hung, Molinas and Wu, and Yang and Lim. Due to non-linearity of most of the sediment transport equations, a Monte Calro numerical method is used to compare these equations. Results show that among the input physical properties, flow velocity, sediment grain size, and roughness coefficient mostly affect the sediment fluxes, respectively. Variation of flow depth has the least effect on sediment transport estimations. Results also show that among four investigated equations, Molinas-Wu and Shen-Hung equations are the least and most sensitive to the errors in the input physical data, respectively. In most cases the deference between the Shen-Hung and Molinas-Wu equations is very high and about several hundred percent. Hence, it is recommended not to use the Shen-Hung equation for sediment flux estimation or in hydrodynamic models, except if the physical properties are measured with precision.