Investigation the Ability of Artificial Neural Network in Simulation of Rainfall-Runoff Process under the Climate Change Conditions (Case Study: Pashakola Babol Dam Basin)

River flow forecasting plays an important role in planning, management and operation of water resources. To achieve this goal and according to the phenomenon of global warming, it is necessary to simulate the daily time series of rainfall and runoff for future periods. Therefore, it is important to survey the detection of climate change event and its impact on precipitation and runoff in the basin. In the first step of this research, using Mann-Kendall trend statistical test, climate change event in the Pashakola Babol basin in Mazandaran province is confirmed. The results of the survey on 36 years daily mean temperature data, reflect the increasing trend in average temperature in the basin. In the second step, the LARS-WG model under general circulation models HadCM3 and A2 scenario is used to generate the daily rainfall time series in the future period. In order to rainfall data, minimum temperature, maximum temperature and sunshine hours is entered in the model for 12-year base period (2004-2015) and Then daily rainfall time series in the basin is predicted for the next 10-year period (2016-2025). In the third step, the artificial neural network model is used to simulate the process of rainfall - runoff in the climate change condition and to generate the daily runoff time series in the future period. Finally, in order to enhance the capability of the artificial neural network model in predicting the daily runoff, besides the predicted daily precipitation data, the rainfall and runoff data one day before as effective factors on the current day runoff is also entered in the model and using rainfall and runoff neurons. Correlation coefficient was obtained equal to 0.8. This correlation coefficient is significant at 1% and show the ability of model to simulate rainfall–runoff process.