Application of Adaptive Neuro-Fuzzy Inference System to Estimate the Groundwater Quality Index in Telkaif, Iraq

In this study, an innovative application of the ANFIS artificial neural network is presented to predict the water quality index based on physicochemical parameters for groundwater in the Telkaif District, Iraq. Samples of water were collected from 16 wells located in the Telkaif district for nine months, from January to September 2024. Measured parameters include total dissolved solids (TDS), magnesium, calcium, sodium, sulfate, and dissolved oxygen. The spatial distribution pattern of all measured parameters is prepared and indicated that most of the parameters are high in the central region of the study area. The groundwater quality index is estimated using ANFIS, and the results show that one well is excellent, 3 wells are good, one is marginal, 2 are moderate, and 9 are poor. The validity of the ANFIS model is confirmed using the Canadian model, which shows (very good R and R2) values, while the RMSE value is (good) (0.97, 0.93, and 8%) respectively. The ANFIS model can be relied upon to assess the quality of drinking water due to its time-saving, ease of application and accuracy of results