Spatiotemporal Geostatistical Techniques Approach for Evaluation of Groundwater Quality and Its Suitability for Potable Uses in Delhi City, India

The present study is focussed on the Delhi region of the northern parts of India; this region is a part of the Indo-Gangetic plain mainly covered by Quaternary alluviums and Delhi Supergroup rocks of the Pre-Cambrian age. A lack of information is present in this region regarding spatiotemporal changes in groundwater quality. The study area is highly susceptible to groundwater deterioration in quality and quantity aspects. Six hydro-geochemical variables have been considered for the study area. These hydro-geochemical variables are acidity (pH), magnesium (Mg+2), sulphate (SO4−2), calcium (Ca+2), chloride (Cl−), and fluoride (F−). The study aims to analyse the long-term spatial–temporal changes in groundwater quality from 2013 to 2017. The modeling study involves a twofold objective. First, it estimates and evaluates spatial and temporal variations in groundwater quality using the ordinary kriging methods for each year (2013–2017). A theoretical semi-variogram was fitted to an experimental semi-variogram enabling estimation of nugget, continuity, sill, and range of influence. Subsequently, kriging was carried out using the semi-variogram parameters that generated prediction maps with uncertainty for six geochemical variables. The point kriging cross-validation (PKCV) results revealed that the kriged mean error lies between 0.001 and 0.002 m, the ratio of estimated variance to the kriged variance range lies between 1.05 and 0.95 and the R2 values are significant through t-test of significance on R which is optimum, acceptable, and supports the unbiasedness hypothesis of kriging. Second, the study applies a pixel-based Mann–Kendall trend analysis on groundwater quality kriged raster surfaces. The trend, significant or not, is determined by the Mann–Kendall test, while Sen's slope estimator determines the slope magnitude of the trend. Results revealed a significantly (at 90% confidence interval) high-rate decreasing and increasing rate of geochemical variables in the study area from 2013 to 2017. The kriged results of each geochemical variable have been compared with the Bureau of Indian Standards (BIS), and it is found that there is a contamination of geochemical variables above the desirable level in the study area. The results clarified that kriging with pixel-based trend modelling is a valuable tool for identifying the critical regions of groundwater contamination zones that need more attention for sustainable groundwater treatment.