A GIS-based linear regression modeling approach to assess the impact of geologic rock types on groundwater recharge and its hydrological implication

Abstract

In this study, the efficacy of the linear regression modeling technique in the company of recharge rate-induced geophysical parameters was implemented through geographical information system to establish the essentiality of geology factor in recharge rate estimate and prediction. The study empirically assessed the in situ and the spatiotemporal variability in the recharge rate in response to the multi-faceted geologic settings in the Perak province, Malaysia. The used data sets were 2D resistivity data, climate data and the ancillary data (surface geologic rock types; [quaternary (QUA); Devonian (DEV); Silurian (SIL); igneous(ING)]. The acquired 2D data were interpreted for delineating vadose zone units having varying geoelectrical parameters consisting of resistivity (ρ) and its thickness (D) (depth to aquifer top). The combined GIS and linear regression techniques were applied to relate the vadose zone geoelectrical parameters with the spatially modeled rainfall–recharge estimate. The obtained results were processed in R statistical software environment to develop GIS linear regression (GLR)-based recharge models for the area underlain geologic rock types. The developed GIS linear regression (GLR)-based recharge models were validated using the models’ parameter significance evaluation and models’ forecasting accuracy assessment approaches. The results of the applied GLR-based recharge models at each 2D location were processed to produce groundwater recharge maps for the rock types. Applying the concept of GIS-based geoelectrical parameters recharge estimate and rainfall–recharge rate–subsurface media estimate approaches, the output of the GLR-based recharge models was quantitatively evaluated to give expected recharge estimates of 242 mm/yr, 248.mm/yr, 268 mm/yr and 224 mm/y and total recharge water values of 377.9 × 106 m3/year, 319 × 106 m3/year, 161.6 × 106 m3/year, 227.3 × 106 m3/year, for the area underlain rock types, namely QUA, DEV, SIL and ING, respectively. Based on these results, the impact of geologic rocks’ factor on the area recharge rate estimates is established. Thus, the consideration of the geology factor is very important in accurate estimation and prediction of the groundwater recharge rate. The developed GLR-based recharge models and the produced groundwater recharge maps are viable decision tools for efficient management of groundwater resources in the study area and any area of similar geologic settings.