Sensitivity Analysis of Conditioning Factors In Groundwater Potential Prediction (Case Study: Ghorve-Dehgolan Plain)

Groundwater is known as one of the most valuable natural resources which play an important role in providing water supply in arid areas. In this study, multi-criteria decision-making (MCDA) methods were used for sensitivity analysis of the effective parameters in groundwater potential prediction based on the lithology, annual rainfall, drainage density, lineament density and land slope parameters in Ghorveh-Dehgolan plain as a case study. The normalized weights of the parameters were determined based on the paired comparisons of AHP approach. After preparing the groundwater potential prediction map, validation of the model was conducted by using the pumping tests data of 20 wells in the study area. Sensitivity analysis of the effective parameters in groundwater potential prediction was measured using the elimination method and finally accuracy of the maps was evaluated. The results of the sensitivity analysis showed that all parameters have a positive impact on the accuracy of the groundwater potential prediction while the best accuracy was obtained by combination of all layers (85%), and the worst accuracy resulted when the lithology was eliminated from the decision layers (60%). Therefore, the lithology parameter through effects on physical properties such as density of lineaments and faults, porosity and permeability has the highest impacts on the accuracy of groundwater potential prediction. The results also revealed that rainfall and slope variables, in despite of their little changes in a plain, are effective in accuracy of groundwater potential prediction.