Comparison of Autocorrelated Model and Artificial Neural Network in Estimating River Flows, a Case Study; Kaka Reza River, Lorestan Province

Accurate estimation of flow in streams can have a significant role in water resources management. In this study, estimation of yield of Kakarza River in Lorestan Province was assessed by using aoutocoorelated model then the results compared with intelligent methods such as neural networks. River discharge parameters on a monthly time scale during the period (1361- 1386) are used as input data in the selected model. Correlation coefficient Criteria, root mean square error and coefficient of Nash Sutcliff was used to evaluate and compare the performance of the models. The results show that both two mentioned models were able considerable accuracy to estimate river discharge, but in terms of accuracy, the ANN model with the highest correlation (0.856), minimum root mean square error (0.049 m3/s) and Nash-Sutcliffe efficiency 0.677) were put in the verification phase. The results show that the artificial neural network model has high ability to estimate of minimum and maximum values of river discharge.