Trend changing assessment magnesium of using neural networks and GIS

Shafts are one of the most important supplies of potable water, agriculture and industry.
The quality of the water is the most important factor between these qualitative factors. Therefore the investigation and prediction of pollution is one of the managers and programmer’s purposes.
In this research, Artificial neural network and geospatial information system were used to determine the pollution of these materials; magnesium in east villages of Golestan province during these years; 87, 88, 89, 90.
In this model the artificial neural network by perspetron structure with hidden layers and several neurons were evaluated. In the present time, the underground water pollutions are increasing because of the chemical activities. Therefore it is require to investigate the vulherable regions of the area in order ro prevent underground water pollution.
There are also different kinds of maps in this research such as: topography, geology, situation of shafts, … that were used in Arc GIS software to determine this underground water pollution and the suitable weight for the slightly parameter was specified in Matlab and the output achieved so that we finally could acquired the pollution maps by putting this output in Arc GIS.
By observing these pollution maps and the available data during these mentiond years, we concluded that the rate of pollution was low and this could not be dangerous, because these polluted areas are vacant.