A new hybrid framework of site selection for groundwater recharge

Since incorrect site selection has sometimes led to the failure of artificial recharge projects, it is necessary to increase the effectiveness of such projects and minimize their failure by employing new techniques. Therefore, the present research used a combination of decision-making models, numerical groundwater modeling and clustering technique to determine suitable sites for implementation of an artificial recharge project. This hybrid approach was employed for the Yasouj aquifer located in southwestern Iran. In the first stage, by employing an AHP decision-making model, hydraulic conductivity, specific yield, slope, land use, depth to groundwater, and aquifer thickness were selected from 21 criteria used in previous research. The selected criteria were then entered as input into the classical k-means clustering model. Using the output, aquifer was divided into seven different regions or clusters. These clusters were then matched with the land use map, and some of the abandoned land areas were selected as the final option for implementing the artificial recharge project. Finally, the MODFLOW code in the GMS software was used to simulate the groundwater level and cluster the sites selected, with regards to increase in groundwater level. Results indicated that the most significant increases in groundwater level (43 and 27 cm) were those of Clusters 2 and 6 in the northern and western parts of the aquifer, respectively. Therefore, this approach can be used in other similar aquifers in arid and semi-arid regions to select the best sites for artificial recharge and to prevent loss of floodwaters. (c) 2021 China University of Geosciences (Beijing) and Peking University. Production and hosting by Elsevier B.V.