Vol.:(0123456789) 1 3 Environmental Earth Sciences (2020) 79:405 https://doi.org/10.1007/s12665-020-09148-3 ORIGINAL ARTICLE Comparison of stochastic and traditional water quality indices for mapping groundwater quality zones Khalid Mahmood 1  · Rida Batool 2 Received: 10 March 2020 / Accepted: 21 August 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020 Abstract This study has compared the use of standard water quality overlay techniques with that of Getis-Ord Gi* statistical techniques for visualizing the spatial distribution of water quality parameters in an unconfned aquifer beneath the city of Lahore in Pakistan. The non-uniformly distributed groundwater sampling points were uniformly transformed to a grid of evenly dis- tributed values to apply Getis-Ord Gi* statistics. An optimized neighborhood distance value of 700 m was determined for the Getis-Ord Gi* assessment, showing that the aquifer is regionally continuous and there are no barriers to lateral groundwater fow. This statistical approach was initially applied to individual parameters and was found to better defne hotspots than that of the conventional method. Similarly, the use of Getis-Ord Gi* values improved the assessment of hotspots of water quality index (WQI) values than conventional overlay techniques. Keywords Groundwater quality · Geographic information system · Spatial interpolation · Getis-ord Gi* · Hotspot analysis · Water quality indices Introduction Groundwater is a key resource which, under suitable hydro- geological conditions, is important for water supply and sus- tainability of certain ecosystems. However, the provision of safe sources of drinking water to rapidly increasing popula- tions is becoming increasingly difcult which are further severely afected by pollution caused by land-use activities (Caizhi et al. 2016; Mahmood and Batool 2019). The efects of climate change, over-exploitation of groundwater to meet needs of rapidly increasing populations, urban sprawl with predominantly impervious surfaces and modern living stand- ards are all factors that are limiting the availability of sus- tainable groundwater for use (Abbas et al. 2015). Consequently, there has been an increased focus by pol- icy makers on conserving groundwater quality. To get an insight into the quality of underlying aquifer, samples from boreholes provide the basic data needed (Jang et al. 2017). Increasingly, these feld-based investigations are being sup- plemented by the use of geostatistical and geospatial algo- rithms to determine how groundwater interacts with the environment. Although interpretation of groundwater quality of a region is important, it is equally important to determine the spatial distribution and infuence of groundwater qual- ity variations in an aquifer (Mahmood et al. 2016). This is typically carried out using geographic information systems (GIS) which provide many tools to analyze the spatial trends caused by environmental variables in a region. GIS systems can also be used to relate land use and natural geological fea- tures to changes in groundwater quality in an aquifer (Alqadi et al. 2014). As only a limited number of samples are usually col- lected during a groundwater investigation, it is, generally, not possible to take into account the spatial interrelations of each measured variable. Therefore, specifc operations and algorithms are needed to convert discrete data values to accurate continuous surfaces for precise assessment of groundwater conditions. In the past, diferent statistical methods have been devised to attain better estimations for in-depth analysis of groundwater quality (Mahmood et al. 2014). Measures that are currently employed in this con- text include the use of spatial interpolations to generate a * Khalid Mahmood khalid.m270@yahoo.com; khalid.spsc@pu.edu.pk 1 Remote Sensing, GIS and Climatic Research Lab (National Center for GIS and Space Applications), Department of Space Science, University of the Punjab, Lahore, Pakistan 2 Department of Space Science, University of the Punjab, Lahore, Pakistan