Suspended sediment load prediction based on river discharge and genetic programming method

The correct prediction of effective factors in water resource projects is one of the most important problems of water recourse engineers. Suspended sediment volume carried by rivers is one of these important factors due to its negative issues in water quality، reservoirs capacity and river morphology. In fact deriving a proper method for sediment volume estimation can be one of the most important problems in erosion and sedimentation process. Although During recent decades، some black box models based on artificial neural networks (ANN)، have been developed to overcome this problem and those accuracy privilege to empirical relations such as sediment rating curves have been shown، But these type of models are implicit that can not be simply used by other investigators. Therefore it is still necessary to develop an explicit model for the discharge–sediment relationship. It is aimed in this study، to develop an explicit model based on genetic programming (GP). Explicit models obtained using the GP are compared with artificial neural network technique in suspended sediment load estimation. The daily stream flow and suspended sediment data from one station on Lighvan River in Orumieh lake basin are used as a case study. The results indicate that the proposed GP method performs quite well compared to artificial neural network models and is quite practical for use.