A Novel Methodology for the Stochastic Integration of Geophysical and Hydrogeological Data in Geologically Consistent Models

Abstract

Groundwater resources represent 99% of the total liquid fresh water on Earth and provides approximately 50% of the total volume of water withdrawn for domestic use in the world (United Nations Educational Scientific and Cultural Organization, 2022). However, because the resource is at depth below the ground surface, the interest of the general public and policy makers tends to wane due to its invisible aspect (United Nations Educational Scientific and Cultural Organization, 2022). In a changing climate, being able to understand, predict and model underground flows and resources is crucial. These questions can only be properly addressed by local to regional groundwater modeling.Another issue owing to the underground position of groundwater is the difficulty in collecting data. Data are usually sparse and often represent only a tiny fraction of the total volume of the aquifer. The data can therefore miss most of the spatial heterogeneity. Conjectures based on such groundwater models could lead to inadequate decisions. A recent paper by Adams et al. (2022) reviews the different existing remote sensing techniques (including geophysical techniques) and possible future approaches that could help fill this gap. They argue that satellite or airborne remote sensing could be key methods for understanding groundwater dynamics at various scales, which would have been impossible with classical spare in situ observations. One method mentioned in their paper is the electromagnetic induction techniques, and especially the Time Domain Electromagnetic (TEM). The TEM method is relatively inexpensive, fast and has a depth of investigation ranging from a few meters to a few hundred meters (Christiansen et al., 2009;Fitterman, 1987). Due to its inductive principle, the Electromagnetic Magnetic (EM) method will be mainly sensitive to underground electrical resistivity. Consequently, it will also be sensitive