Characterization of Paleochannel and Floodplain Aquifers Using Vertical Electrical Sounding: A Case Study from the Western Part of Bengal Basin

The subsurface architecture of paleochannel and floodplain deposits, as well as their hydrogeological significance, remains insufficiently characterized in the Ganges upper delta region. The present study evaluates the hydrogeological implications for groundwater resource assessment and aquifer vulnerability in Chakla, North 24 Parganas, West Bengal. Descriptions and discrimination of different subsurface regimes are provided based on electrical resistivity. Vertical Electrical Sounding (VES) surveys are conducted at 51 sites, distributed across paleochannel (np = 31) and floodplain (nf = 20) geomorphic settings. Six-layer resistivity models are developed for each site through inversion analysis. Regime-specific hydrogeological properties are quantified through non-parametric statistical testing (Wilcoxon rank-sum and Kruskal-Wallis) on the modeled VES data. Furthermore, longitudinal conductance and transverse resistance, as obtained from the Dar-Zarrouk parameter analysis, are explored. VES inferences are found to be in well accordance with the borehole lithology from six cores. Paleochannel aquifers and floodplain aquitards exhibit significantly different resistivity distributions due to different grain sizes and saturation. Paleochannel sites reveal higher median resistivity (49.5 Ω·m) and coarser grain sizes, indicating high-capacity aquifers with enhanced investigation depth. On the other hand, floodplain sites are characterized by lower resistivity (19.2 Ω·m), finer grain sizes, and low-permeability confining layers. The findings of the present study support targeted groundwater exploration in paleochannel zones and aquifer protection in floodplain areas, providing key insights for water supply assessment and contaminant vulnerability.Keywords: Vertical Electrical Sounding (VES), Paleochannel aquifer, Electrical resistivity, Dar-Zarrouk parameters, Hypothesis testing, Non-parametric statistics