Sediment transport prediction in rivers using quasi-two dimensional model

Accurate prediction of flow and sediment discharges, as basic information, is important for many river engineering projects. Flow and sediment rating curves are the main tools for computation of these two parameters, respectively. In spite of great importance of flow discharge in river hydraulic computations, no specific approach or mathematical model has yet been presented and still the field measurement of lateral velocity and then its lateral integration are considered as the main solution. Field measurement of velocity across the river is a time consuming and costly work and during the river flooding, is very dangerous (5). Opposite to the flow discharge, plenty of relationships and mathematical models have been developed by the researchers for computing the suspended and total sediment load in rivers (4, 8, 26). These models have often complex theory and need large input parameters. Due to specific conditions of our countries’ rivers and their lack of hydraulic and sediment transport data with good quality, it’s necessary to propose suitable solutions. One of the key elements in this regards is use of quasi two-dimensional mathematical models (9, 10, 11, 28). Shiono and Knight (1991) by developing a quasi-two-dimensional model for lateral velocity distribution, presented a new method with suitable accuracy for computation of flow discharge in flooded rivers (24). In this study, in order to mathematical models being applied for simultaneous computation of river flow hydraulics and sediment transport, a quasi-two dimensional mathematical model has been selected. This model has been calibrated and then verified in Karoun river at Molasani hydrometric station. For simulation of sediment transport in this river, three well known sediment transport equations including Engelund-Hansen (1967), Ackers-White (1973) and Yang (1979) have been used. Few input data needed is the main advantage of this model against the complicated two and three dimensional models. The results showed that calculated stage-discharge curves in calibration and validation phases have mean errors of 2.4 and 4.0 percent, respectively. Such accuracy is noticeable for a large river, such as Karoun river. Also, it’s revealed that among the selected empirical equations for total sediment transport capacity, Yang formula gives good predictions in low flow discharges (less than 800 m3/s). On the other hands, Engelund-Hansen and Ackers-White formulas give better predictions of the sediment discharge for high flow discharges (more than 1000 m3/s). Accordance to the high importance of sediment transport in flood condition, it seems that application of two sediment transport equations of Engelund-Hansen and Ackers-White give more reasonable results for Molasani hydrometric station and hence are recommended.