Using GIS-based order weight average (OWA) methods to predict suitable locations for the artificial recharge of groundwater

This study aims to determine suitable locations for artificial recharge of groundwater (ARG) using the GIS-based analytic hierarchy process (AHP) and order weight average (OWA). To determine the weights of the different parameters, the AHP method of pair-comparison was used after preparing a fuzzy map for each attribute. After that, using the OWA-AHP method for different levels of confidence (different values), the weighting process was used for each parameter to produce land suitability maps of varied risks. In addition, the adaptive network-based fuzzy inference system (ANFIS) method was used to predict land suitability classes using input parameters. Then, using the best subset regression method, the most important effective ARG parameters were identified. Fuzzy-AHP results show that 27% of the area has good and very good conditions for ARG. Under low-level risk and no trade-off, the combined OWA-AHP method shows that the more area is in the very low class (80%) while in case of higher level of risk and average trade-off, the highest values are in the very low class (27%). The results of the ANFIS method indicate that both fuzzy c-means (FCM) and sub-clustering methods can be used to predict appropriate places for ARG. The results of the best subset regression method show that slope, lithology, land use, and altitude with the lowest C-p values (5.2) are effective parameters to determine the suitability of ARG locations.

Automated summary

This study utilized GIS to determine suitable locations for ARG, assigning weights to different parameters using AHP and OWA methods. ANFIS method was used to predict land suitability classes, while best subset regression identified key ARG parameters.