Application of fuzzy C-means clustering and fuzzy EDAS to assess groundwater irrigation suitability and prioritization for agricultural development in a complex hydrogeological basin

• Published in: Environmental science and pollution research international 2023-04, Vol.30 (20), p.57529-57557
• Main Authors: Halder, Sudipa, Bhattacharya, Shuvoshri, Roy, Malabika Biswas, Roy, Pankaj Kumar
• Format: Article
• Language: English
• Subjects: Agricultural development; Agricultural Irrigation; Agriculture; Aquatic Pollution; Aquifer systems; Aquifers; Atmospheric Protection/Air Quality Control/Air Pollution; Calcium; Calcium ions; Cluster analysis; Clustering; Coefficients; Contamination; Earth and Environmental Science; Ecotoxicology; Electrical conductivity; Electrical resistivity; Entropy; Environment; Environmental Chemistry; Environmental Health; Environmental Monitoring; Environmental science; Fluorides; Fuzzy logic; Geographic information systems; Geology; Groundwater; Groundwater - analysis; Groundwater irrigation; Groundwater quality; Hydrogeochemistry; Hydrogeology; India; Ion exchange; Irrigation; Magnesium; Parameters; Permeability; Research Article; River basins; Sodium; Sodium - analysis; Sodium bicarbonate; Sodium carbonate; Soft computing; Spatial distribution; Vector quantization; Waste Water Technology; Water analysis; Water Management; Water Pollutants, Chemical - analysis; Water Pollution Control; Water Quality; Water sampling; Water Supply
• ISSN: 1614-7499; 0944-1344; 1614-7499
• DOI: 10.1007/s11356-023-26394-7

The current research is focused on detecting a river basin suitable for agriculture and priority for management using a new clustering tool of groundwater quality with fuzzy logic technique in R and Geographical Information System. A new fuzzy clustering-soft computing technique has been executed to determine the different hydrochemical zones considering 13 essential parameters such as electrical conductivity, hardness, chloride, sodium adsorption ratio, residual sodium carbonate, soluble sodium percent, magnesium hazard, permeability index, potential salinity, residual sodium bicarbonate, Kelly’s ratio, synthetic harmful coefficient, and exchangeable sodium percentage. The derived fuzzy C-mean clustering (FCM) outperformed other available hard computing techniques like hierarchical clustering, K-means clustering, and agglomerative clustering. It divided the sampling sites into 2 clustering groups (FCM I and FCM II) which has been validated using fuzzy silhouette index (0.85), the partition coefficient (0.76), the partial entropy (0.68), and the modified partition coefficient (0.52). The hydrogeochemical analysis confirmed that the rock-water interaction, chemical weathering, and ion exchange process are predominant in the aquifer system of the study area. According to the correlation plots, the studied groundwater samples largely evolved from Ca 2 + - HCO 3 - , mixed Ca 2 + - Mg 2 + - Cl - types, and HCO 3 - - Ca 2 + types. The spatial distribution map and the hydrochemical analysis also gives a clear depiction of the fluoride (> 1.0 mg/l) and high iron (> 0.3 mg/l) contamination in groundwater quality, making it unsuitable for both drinking and irrigation. A fuzzy EDAS priority map has been prepared based on all the irrigation suitability parameters which concludes that the groundwater at the upstream and downstream section of the basin requires the most attention. Based on the highest priority for management, five zones have been delineated: very high (5.98%), high (22.31%), medium (16.39%), low (32.30%), and very low (23.02). The findings of this study will be beneficial to planners and policymakers as they can develop schemes to solve similar problems across the country.