Integrated knowledge- and data-driven approaches for groundwater potential zoning using GIS and multi-criteria decision making techniques on hard-rock terrain of Ahar catchment, Rajasthan, India

Abstract

This study utilizes for the first time integrated knowledge-driven and data-driven methods for groundwater potential zoning in the hard-rock terrain of Ahar River catchment, Rajasthan, India by employing remote sensing, geographical information system, multi-criteria decision making (MCDM), and multiple linear regression (MLR) techniques. Thematic maps of the eleven hydrological/hydrogeological factors, i.e. geomorphology, soil, topographic elevation, slope, drainage density, proximity to surface waterbodies, pre- and post-monsoon groundwater depths, net recharge, transmissivity, and land use/land cover, influencing the groundwater occurrence were used. The themes and their features were assigned suitable weights, which were normalized by the MCDM technique. Finally, the knowledge-driven groundwater potential map, generated by weighted linear combination, revealed that the good, moderate and poor groundwater potential zones are spread over 90.94 km2 (26%), 135 km2 (39%) and 122.36 km2 (35%), respectively. Furthermore, the data-driven precise groundwater potential index (GPI) map was computed by MLR technique. The results of both the knowledge- and data-driven approaches were validated from the well yields of 18 sites and were found to be comparable to each other. Moreover, exogenous and endogenous factors affecting the good, moderate and poor groundwater potential were identified by applying principal component analysis. The results of the study are useful to water managers and decision makers for locating appropriate positions of new productive wells in the study area. The novel approach and findings of this study may also be used for developing policies for sustainable utilization of the groundwater resources in other hard-rock regions of the world.