drought forecast using neural network in uremia lake basin located in kurdistan province

In this research, to predict the probability of drought in uremia lake basin located in kurdistan province was used gmdh neural network method genetic algorithm-based as an instrument with high capability in complex nonlinear dynamic systems modeling. in this study, the effects of input parameters on drought were investigated. in order to ensure the accuracy of the proposed method, the accuracy of the used models were examined in comparison with experimental data using correlation coefficients and root mean square error.the data obtained by the model were compared with experimental data. the results showed a very good compliance between data obtained from the model with experimental data. as a result, it was determined that the resulting gmdh neural networks model is reliable to determine the extent of drought and it’s a good alternative to determine the drought in various parts of the country.