A comprehensive assessment of spatial interpolation methods for the groundwater quality evaluation of Lahore, Punjab, Pakistan

Spatial interpolation is commonly used to generate water quality surfaces, but different spatial interpolation methods yield different surfaces from the same data. The water quality map produced using one model of spatial interpolation method may be significantly different from the map produced using another model of the same spatial interpolation method. The purpose of this study was to evaluate the performance of different spatial interpolation methods to depict the water quality of Lahore correctly. The water samples (n = 73) were collected from tube wells and tested for physicochemical parameters (pH, turbidity, hardness, total dissolved solids, alkalinity, calcium, and chlorides). The data exploration was performed using SPSS software. The inter-comparison of different powers of inverse distance weighting (IDW) and different functions of radial basis functions (RBF) was completed using geostatistical analyst extension in ArcGIS 10.3. Moreover, these deterministic interpolation methods (IDW and RBF) were compared with geostatistical interpolation methods (ordinary kriging and ordinary co-kriging) based on cross-validation statistics; root means square error (RMSE). The analysis showed that ordinary co-kriging performed much better than ordinary kriging, RBF, and IDW, for water quality assessment of Lahore. Hence, ordinary co-kriging with appropriate auxiliary variable and the best-fitted semi-variogram model was used to generate the spatial distribution map for each water quality parameter. The water quality index (WQI) was computed using the tested physicochemical parameters, and the results showed that 98% of the tube wells were providing ‘excellent’ to ‘good’ water quality in Lahore city. However, there were few areas of City and Anarkali subdivisions where it indicated poor to very poor water quality. The procedure used in this study is valuable for the water management authorities to better understand and monitor the groundwater quality.