Optimizing coefficients of Sediment rating curve equation using Genetic Algorithm (Case study: Ghazaghli and Bagh abbasi stations)

The proper estimation of sediment yield of rivers is important for planning and managing water resources. Various methods have been developed to determine the relationship between discharge and sediment concentration. Sediment rating curve is one of the most common methods for estimating the suspended sediment yield in rivers, which is always associated with a large error. In order to improve estimation of sediment yield using sediment rating curve, the equation coefficients can be optimized using artificial intelligence methods. The purpose of this research is to use a genetic algorithm to optimize the Sediment rating equation coefficients for Gorganroud river (Ghazaghli station) and Fariman river (Bagh Abbasi station). With this aim in mind, discharge and suspended sediment concentration data have been acquired for the stations. Sediment rating curve equation was calculated for each station with training data. Then an optimal coefficient of equation was achieved using genetic algorithm model defined in MATLAB 2017 software. The study results showed that the genetic algorithm model for Ghazaghli and Bagh Abbasi stations had a better performance than the sediment rating curve with the Nash-Sutcliff coefficient of 0.5 and 0.72 and coefficient of determination of 0.5 and 0.89, respectively. Also, the genetic algorithm for Bagh Abbasi Station with the limited samples has better accuracy than the sediment rating curve method. The study results indicate a high performance of the genetic algorithm in optimizing the coefficients of Sediment rating curve equation, especially in low data stations.