the analyses of geomorphological artificial neural network in estimating of runoff in jajrood sub basin

estimation of pure runoff is a house virtually is complex and different methods of calculation have been proposed. modern methods of solving problems in river engineering and water homes and assess the flow method is used, is that the artificial network pattern of human brain neural network training process, while implementation of the internal relationships between the data and discover for other situations will generalize. the main objective of this study to estimate the runoff through the analysis of relations rainfall-runoff by geomorphology and quantitative data and using geomorphologic artificial neural networks technique in amameh sub basin in jajrood basin. this study based on geomorphologic stracture and ed hydrologic network, a geomorphologic system with three-layer neural network with the number of intermediate nodes or the number of paths situations hydrology network in order to estimate direct runoff was established. the weights of system input connections within the network structure model determined with using geomorphologic variables. results from the above network model with information derived from direct observations to indicate the efficiency was compared. assessment results, indicate a very good function (r2=0.97) geomorphologic network model to determine the response of hydrological basin is studied. means, this model shows the preference of them than the other common procedure methods.