Estimation of the monthly Runoff from precipitation and flow measurement networks using Artificial Neural Network

Abstract

The Ourika river originated in the High Atlas generate the average water resources of 157 Mm3/year, Since there is no dam in this river, the water supplies are regulated by traditional channels to irrigate 19 855 ha, hence the importance of developing runoff-rainfall model for water estimation in this context. In the presented study Artificial Neural Network is applied to forecast the monthly runoff in outlet of basin. This study uses runoff from two stations and monthly precipitation data recorded at measurement network composed of 5 stations located in Ourika basin during 15 years from 2000 to 2015. For evaluate the performance of model in the phases training and validation the appropriate statistical methods were used.