[M-GI30] Near Surface Investigation and Modeling for Groundwater Resources Assessment and Conservation

[MGI30-02] Applicability of Dupuit-Ghijben-Herzberg (DGH) analysis to quantifying the volume of freshwater lenses in extremely permeable aquifers

*Satoshi Tajima¹, Jiaqi Liu¹, Tomochika Tokunaga¹ (1.The University of Tokyo)

Keywords:Freshwater lens, Seawater intrusion, Coral island, Limestone, Numerical simulation, FEFLOW

Freshwater lenses are vulnerable water resources in small, low-lying coral islands. An analytical model based on the Dupuit-Ghijben-Herzberg approximation (DGH analysis) delineates the geometry of freshwater lenses simply and effectively, and hence, it has been widely used for practical purposes. DGH analysis assumes a steady-state and ignores vertical groundwater flow and dispersion processes. When the hydraulic conductivity (K) of an aquifer is extremely high, these simplifications may not work properly. Few studies, however, have investigated the freshwater lenses formed in extremely permeable aquifers in detail, and the applicability of DGH analysis remains unclear. This research aims to clarify whether DGH analysis is applicable to quantifying freshwater volume in an aquifer having an extremely high K. A case study in Tarama Island, Japan, was performed by comparing DGH analyses, numerical simulations, and in-situ salinity measurements. It was found that when K was 1.1×10^-3 m s^-1 or higher, the freshwater volumes estimated by the DGH analyses were larger than those by the numerical simulations, and this discrepancy increased with the increase of K. In particular, when K was 2.0×10^-2 m s^-1, which is the estimated value for the aquifer in the Tarama Island, the freshwater volume estimated by the DGH analyses was around 14 times larger than that by the numerical simulation. It should be noted that the in-situ measurements were much closer to the numerical simulations compared with the DGH analyses. Moreover, the numerical simulations showed that saline pore water exists in the unsaturated zone near the coastline. The large discrepancy between the DGH analyses and the numerical simulations may be attributed to the fact that DGH analysis ignores the dispersion processes. Therefore, DGH analysis should be applied with caution to assessing freshwater lenses in extremely permeable aquifers.