Spatial optimization of rainwater harvesting infrastructure using comparative GIS-MCDM approaches in a semi-arid basin

Abstract Water scarcity in semi-arid regions is intensifying under climate variability and growing pressure on water resources, necessitating the development of efficient and spatially optimized Rainwater Harvesting (RWH) systems. This study presents a GIS-based decision-support framework to optimize RWH infrastructure siting in the Dinarta basin, by integrating hydrological analysis with Multi-Criteria Decision Making. Five methods, Weighted Overlay, SAW, TOPSIS, ELECTRE, and VIKOR, were applied to spatial criteria derived from remote sensing and terrain analysis. Model performance was evaluated through statistical and hydrological validation. ROC analysis against existing RWH infrastructure locations showed that VIKOR achieved the highest spatial discrimination capacity (AUC = 0.822), followed by Weighted Overlay (AUC = 0.816). Independent hydrological validation against SCS-CN-derived runoff using the Quality Sum framework further confirmed the physical consistency of the predicted priority zones. The final VIKOR map indicates that about 10.5% (42.7 Km 2 ) of the basin falls within high to very high suitability classes, concentrated in southern and central lowlands where specific conditions enhance storage feasibility and recharge potential. Within the infrastructure-feasible corridors identified by the VIKOR model, SCS-CN modelling indicates annual surface-runoff depths reaching a maximum of 671.7 mm in the most productive zones within the GIS layer. Technique-specific mapping suggests ponds, percolation tanks, and check dams are most suitable in these priority corridors; while terracing and gully plugs are preferable in steeper uplands for runoff and sediment control. By coupling statistical discrimination with independent hydrological validation, the framework introduces a dual-validation approach that confirms predicted suitability zones correspond to genuine runoff-contributing areas, offering a more reliable and physically grounded basis for prioritizing RWH investments.