Evaluation of Climate Change Impacts on Hydrological Characteristics of Watershed (Case study: Aji- Chai Watershed)

 Studying climate change impact on hydrological variables such as runoff has attracted more attention over recent years mainly due to the imposed imbalance in climate system by greenhouse gas emission. Hence, initially in this research, the trend and abrupt changes of hydro – climatic variables were studied using data mining approaches in 18 stations over 40 years (1972-2011) in Aji-Chai watershed located in northwestern Iran and secondly, the outputs of climatic variables including temperature and rainfall of HadCM3 under A2 and B2 scenarios downscaled by SDSM in order to studying the climate change impact on runoff. Finally, five data mining and intelligent models including ANFIS, SVM, MLP, M5, M5 Rules were used to rainfall – runoff modeling. The results showed that the hydro – climatic variables have been changed over the last four decades which make it necessary to study climate change and consider its impact on the runoff in the study area. The results showed that the maximum temperature will increases 2.04 and 2.40 degrees centigrade until 2010 to 2050 and 3.87 and 5.11 degrees centigrade until 2051 to 2091 under A2 and B2 scenarios, respectively.  Also the minimum temperatures will increases 1.63 and 1.90 degrees centigrade until 2010 to 2050 and 3.76 and 3.00 degrees centigrade until 2051 to 2091 under A2 and B2 scenarios, respectively. The results of future rainfall prediction showed no specific trend while in some seasons showed increasing and in other cases showed decreasing trend. The results of the model performance evaluation determined the M5 and MLP models the lowest errors. In general, the results of rainfall – runoff modeling indicated inherit uncertainty of the models and the runoff will increase or decrease over the seasons.