Modeling of rainfall-runoff relation with artificial neural network methods: Kurukavak basin casestudy

Flow estimation in a base, the help of the obtained hydrometeorologicdata, is important in terms of water resources projecting studies. In this study,based modelling applications are performed with flow estimation which isobtained by the help of antecedent hydrometeorologic data. In the study, the flowestimations are made by previously measured rainfall, evaporation and antecedentflow data obtained in Kurukavak Basin, Pazaryeri, Bilecik. In the study, theartificial neural network methods of the feed forward back propagation method,the generalized regression neural network and the radial based artificial neuralnetwork method are used. The calculations are performed in Matlab 6,5programme. The new flow data are obtained from a computer programme inmatlab written about rainfall, flow and evaporation data. Many simulations aredone for each different artificial neural network architecture to get the best resultsand the real results are obtained in a short time.Keywords: Artificial neural network, feed forward back propagation method,generalized regression neural network, radial basis function neuralnetwork