Performance assessment of multicriteria decision making techniques and artificial neural network to predict future groundwater potential zones

Identifying Groundwater Potential Zones (GWPZs) is essential for sustainable water security amid anthropogenic pressures and climate variability. This study develops an integrated geospatial framework to project GWPZ spatiotemporal dynamics in Gazipur District, Bangladesh, using three MCDM techniques: Analytic Hierarchy Process (AHP), Frequency Ratio (FR), and Artificial Neural Network (ANN). Eleven hydro-environmental factors were incorporated, with rainfall and LULC projected as dynamic variables for 2030, 2040, and 2050 using CNRM-CM6–1 (CMIP6) under SSP1–2.6, SSP2–4.5, and SSP5–8.5 scenarios. Model validation via pixel-wise correlation and ROC-AUC ranked FR highest (AUC = 0.80), followed by AHP (0.77) and ANN (0.71). All models consistently projected contraction of moderate-to-high potential zones and expansion of low-potential areas. Under SSP2–4.5, moderate zones decreased from 1053.78 km 2 (2030) to 955.26 km 2 (2050), while low-potential zones increased from 321.73 km 2 to 440.96 km 2 . These results indicate rising groundwater stress under future climates, highlighting the urgency of adaptive, resilient governance. The framework provides a robust, scalable decision-support tool for prioritizing interventions, managed aquifer recharge, and policy formulation in data-scarce, hydro-climatically vulnerable regions. • The study delineates future groundwater potential zones (GWPZ) using AHP, FR, and ANN, considering eleven factors, with LULC and rainfall as dynamic variables projected for 2030, 2040, and 2050 under CMIP6 SSP scenarios. • The study classifies GWPZs into five categories, revealing a decline in “Moderate” and “High” zones, especially under SSP2–4.5 and SSP5–8.5, with a shift toward lower potential areas. • Performance assessment shows the FR model (80% AUC) outperforms AHP (77%) and ANN (71%). • The findings support sustainable groundwater management, aiding agriculture, economic growth, climate resilience, and policy-driven resource planning.