Intelligent forecasting of rainfall and temperature of Shiraz city using neural networks

Most parts of southern Iran have frequently experienced drought and flooding events. The occurrence of these natural disasters is common in Fars province which supplies about 25% of the national wheat product. Shiraz (capital city of the province) is studied as a representative of the province. Estimating rainfall and temperature can help in agricultural water management, protection from water shortages, and flood damage, thus having significant economic impacts. The prediction is, however, a complicated procedure and conventional mathematical methods are not able to easily capture such a relationship. To overcome the problem, neural network-based models were used for forecasting temperature and rainfall in Shiraz (Iran). Various simulation results based on the real data are presented. The results suggest that the applied methodology is suitable and more practical than the previous approaches for the prediction of rainfall and temperature. The developed model is able to predict rainfall and temperature one season ahead with reasonable error.