Groundwater Quality Indexing Optimization Based on Geographical Information System in Lahore-Pakistan

Abstract

The study has been intended to find best possible solution for developing continuous spatial distribution of groundwater quality, being used for drinking in the mega city of Lahore, Pakistan. A variety of interpolation methods have been compared to map spatial trends in different indicators of groundwater quality in the most realistic and effective way. For this purpose, concentration of nine groundwater parameters (Alkalinity, Ca, Cl, Conductivity, Hardness, Mg, pH, TDS and Turbidity) for 348 drinking water production wells have been used with their spatial reference. Ultimate accuracy check over different interpolation methods is applied using contour analysis, surface analysis, semivariogram, error plots, average standard error and root mean square error (RMSE). Kriging family, particularly simple kriging, is found to be most suitable for predicting most of the groundwater quality surfaces with the existing distribution of the production wells. The only exception found is for predicting Cl surface that has been best developed using redial basis function. RMSEs of the beast predicted surfaces for Alkalinity, Ca, Cl, Conductivity, Hardness, Mg, pH, TDS and Turbidity are 69.84, 19.04, 19.66, 212.10, 73.65, 10.21, 0.15, 131.70 and 0.86 respectively.