Assessing Groundwater Quality and Land Use, Land Cover Changes (Case Study: Gharasu Basin, Golestan Province)

In the present study, a spatial model has been developed, which couples artificial neural networks (ANN) with a geographical information system and remote sensing to evaluate the impact of land-use and land-cover change on groundwater quality. Input parameters for running ANN model were land use, distance from roads, built-up areas and rivers, population density and cultivated land, underground water table for the years 1992, 2002 and 2008, and soil and geology. The output parameter was the observed groundwater quality properties in the sampled wells that included TDS, SO4-2, NO3-, and Cl-. For NO3-, the model was run with data from the years 2002 and 2008 and for TDS, SO4-2 and Cl-, the model was run with data from the years 1992 and 2008. The results indicate that the density of cultivated land and geological structures of the region have a large impact on the quality of underground water supplies. Also, the effect of built-up areas, population density and distance from rivers demonstrate increases through time. Spatial distribution maps of various pollution parameters were prepared that demarcate the geographic distribution of water pollutants in a comprehensive manner and help in suggesting groundwater pollution control and remedial measures in a holistic way.