Using Artificial Neural Network (ANN) for Estimating Rainfall Relationship with River Pollution

Nature, and the experimental site, and flexible neural networks, to cause the problems are such categories have predicted that such attitude is all the structure and is used for nonlinear behavior and ungovernable well. In recent years due to universal climate change, precipitation reduction is happened. The Rivers are the most important domestic water resources and agricultural consumptions and are affected by the rainfalls reduction. Amount of salts in the rivers is increasing and threat life of organism. Therefore expectancy of amount of salts in the river considering to rainfalls reduction is necessary. Three years information of Lenj station is used at Zayandehrood River in this research. The monthly precipitation considered as the input and the river pollutants as output. Artificial nervous network is BP, it used tow ways for enter information input information once without variation and raw. Once between 0 and 1 transformation then enter to network. In this modeling, correlation coefficient 0.999, to compare tow stage, information shows that second way (0 and 1) has low error.