Evaluating the Effectiveness of the Integrated Model of Precipitation Time Clustering and Climate Forecasting in Estimating Monthly Flow

Rainfall is one of the variables that have the most impact on runoff. In this study, the inflow of Karun 4, Dez, Karun 3, Rudbar, Sardasht and Zayandeh-Roud dams was modeled using the NAM rainfall-runoff model and past meteorological and hydrological data. In order to determine the amount of rainfall input to the model of North America were received. Then, to select the best precipitation forecasting model, precipitation temporal clustering method and artificial neural network with radial basis function were used. After determining the rainfall and predicting the inflow to the dam to check the performance of the rainfall model selection, the predicted monthly inflow was compared with the actual monthly inflow. The results showed that more than 60 percent of the time in Karun 4 and Sardasht dams, with the cooperation of the best climate model, the rainfall values are close to the observed values. The normalized mean error rate, normalized root mean sum of squares and Nash Sutcliffe in Karun Dam 4 are calculated as -0.064, 0.046 and 0.649, respectively. The results of this research show the importance of rainfall time clustering method to choose the best rainfall climate model to improve irrigation results.