a.m moradi - graduated student
a.b dariane - associate professor
m. teshnehlab

River flow forecasting can reduce excessive water deficiencies flood damages and enhance proper management of agricultural water supplies however development of an appropriate river flow forcasting model is a complex process despite numrous applications of multilayer perceptron neural networks MLP in river flow forcasting there is not any application using the cerebellar model articulation controller CMAC thus in this research two algorithms namely MLP with error backpropagation and CMAC are developed and applied for forcasting monthly inflows to dez reservoir located in southwestern iran.

monthly forecasting , neural networks,CMAC,reservoir,and Dez