Assessment of groundwater quality for agricultural purposes in Qazvin Province, northwestern Iran: A fuzzy inference and indicator Kriging approach

Abstract

This study addresses the challenge of assessing groundwater quality for agriculture in Qazvin Province, northwestern Iran. In this region, over-extraction has led to significant degradation of groundwater resources. Traditional assessment methods often overlook uncertainties and spatial variability in groundwater quality. To address this, our study aimed to integrate fuzzy inference and geostatistical methods to assess groundwater quality under uncertain conditions. The research was conducted in two stages. First, a fuzzy inference system classified six key water quality parameters: Electrical Conductivity (EC), Sodium Adsorption Ratio (SAR), Residual Sodium Carbonate (RSC), Total Hardness (TH), sodium ion concentration (Na⁺), and chloride concentration (Cl⁻), into three categories: "desirable," "acceptable," and "unacceptable," using 54 fuzzy rules. In the second stage, we applied ordinary Kriging and indicator Kriging to spatially interpolate these classifications and produce probabilistic maps of groundwater quality risk across the study area. In ordinary Kriging, the average Root Mean Square Error (RMSE) values for EC, SAR, RSC, TH, Na⁺, and Cl⁻ were 0.94 dS/m, 1.54, 1.42 meq/L, 3.13 mg/L, 3.03 mg/L, and 2.62 mg/L, respectively, indicating reliable assessments of groundwater quality parameters. Results also suggest that by 2023, areas classified as "unacceptable" increased by 142.0% since 2009, with an additional 25.2% of the region facing a 40 to 80% probability of further degradation. These findings highlight important trends in groundwater quality, assisting local authorities in prioritizing areas for preventive interventions. This supports sustainable agricultural practices and aligns with the United Nations Sustainable Development Goal 6 for water resource management.