Relative importance of conceptual and computational errors when delineating saltwater intrusion from resistivity inverse models in heterogeneous coastal aquifers

Abstract

Highlights:-Inverted resistivity data are commonly used to delineate saltwater intrusion (SWI).-Errors associated to inversion and petrophysical transformation are assessed.-A coupled modelling approach was applied to four synthetic heterogeneous aquifers.-Inversion and salinity recovery errors are equivalent when neglecting heterogeneity.-Spatial quantification of the errors is key for geophysics-aided SWI management.

Imaging the Structure and the Saltwater Intrusion Extent of the Luy River Coastal Aquifer (Binh Thuan, Vietnam) Using Electrical Resistivity Tomography

“…On the other hand, geological conditions, such as the presence of confining layers combined with paleohydrological conditions, can also yield complex saltwater distributions related to ancient seawater trapped in sediments [5]. In addition, geological heterogeneity also is identified as one of the most important factors affecting the spatiotemporal dynamics of saltwater in coastal aquifers [1,6]. It not only affects the evolution of SI through creating preferential flow paths but also makes the conceptualization of SI more difficult by preventing the generalization of the processes.…”

“…Geophysical techniques offer a cost-effective alternative, by providing indirect information on the aquifer properties at a high spatial resolution [7]. Geophysical data can be used as a characterization tool but can also be more systematically integrated into the development of groundwater conceptual models and the calibration of numerical models [6]. Electrical and electromagnetic methods are particularly well suited for investigating SI because of their sensitivity to the bulk electrical conductivity of the ground, which is directly related to the salinity of pore water [8].…”

“…Quantitative interpretation is even more challenging. In a recent contribution, González-Quiros and Compte [6] studied the ability to delineate salinity boundaries from inverted ERT data. They identified three main sources of errors limiting the direct interpretation of ERT.…”

“…This makes it virtually impossible to quantify the salinity in absence of colocated data, i.e., ERT data collected at the location of existing boreholes and wells to validate petrophysical relationships. Second, the difficulty is further enhanced by the regularization used for the ERT inversion, often leading to over smoothing [6] and the loss of resolution of surface ERT at depth [14,15,21], which affects the petrophysical relationship [22,23]. It often leads to discrepancies when comparing inverted results with borehole data [14,24,25].…”

“…It often leads to discrepancies when comparing inverted results with borehole data [14,24,25]. The salinity is often overestimated in the low-salinity zone and underestimated in the high-salinity zone [6]. This effect can be reduced but not eliminated by integrating crossborehole [25] or surface-to-hole measurements [14] or by choosing more appropriate regularization [21].…”