Using of hybrid fuzzy models to predict spatiotemporal groundwater quality parameters

Spatiotemporal groundwater quality simulation is very important for management of water resources and environmental activities. The present study integrated a number of hybrid methods such as Adaptive Neuro Fuzzy Inference System (ANFIS) with Genetic Algorithm (GA), and ANFIS with Partial Swarm Optimization (PSO) to simulate three groundwater quality parameters in Kerman plain (including Chloride concentration, Electrical Conductivity (EC), and PH). This research investigated the abilities of hybrid techniques as well, to predict groundwater quality. Considering the complexity of different aquifer materials and difficulty of collecting desirable samples, as it is both time- and cost-consuming, a number of hybrid models have been developed, presuming various combinations of monthly variables of rainfall and groundwater level and three different quality parameters. The results suggested that both models indicated an acceptable efficiency in the spatiotemporal simulation of groundwater quality. The study also revealed that groundwater level fluctuations across the aquifer as well as rainfall contribute as two important factors in predicting groundwater quality.