Rainfall-Runoff modelling using Support Vector Regression and Artificial Neural Network Models ( Case study: SafaRoud Dam Watershed)

Rainfall-runoff modeling is an important and complex aspect in most water resource management and planning projects. In this study, Perespetron multi-layered artificial neural network (MLP), Radial basis function Neural Network (RBF), and support vector machine regression with linear kernel functions (SVR linear) were used to develop some models in SPSS to simulate Rainfall-runoff process in subarea of Safaroud dam, located in Halil Rood watershed. To do so, hydrometric data of Hanjan station and rainfall data of Hanjan, Rabor, Cheshme Aroos, and Meidan stations, located in the studied area, were used. 70% of the data were used as training data and 30% were used as test data. After calculating the partial correlation coefficients of the rainfall and discharge, six different patterns were used to model the daily rainfall of Hanjan station. In the best pattern of the test level, for SVR Linear 5 model, Mean Absolute error (MAE), Root Mean Squared Error (RMSE) and correlation coefficient (r) were equal to 0.032, 0.229, and 0.967, respectively. The results proved the efficient performance of MLP and SVR Linear in Rainfall-runoff modeling in the studied area.