Simulation of Carbonated Fault Zones Hydrodynamics and Transport Considering Parametric and Predictive Uncertainty

Hydrodynamic understanding of karstic aquifers is a real challenge due to the complexity of their internal structures. However, their societal significance lies in the substantial quantity of groundwater resources they embody. Among these complexities, faults partially control the organization of flows in these systems. The nature of this control can either facilitate rapid flow transfer or act as a barrier, impacting both groundwater quality and quantity. Understanding the behavior of fault zone features is crucial for efficiently management of karstic aquifer resources. However, there is a lack of studies that estimate and simulate flow within fault zones. In this study we estimate the hydraulic properties of the fault zone within carbonate karstic aquifers for flow and transport forecasting purposes based on cross-hole pumping tests simulation and inversion. The flow and transport are modeled using MODFLOW6 and MODPATH7. The inverse modeling approach is based on the Gauss-Levenberg-Marquardt Algorithm (GLMA) and the Iterative Ensemble Smoother (IES) integrated into the PEST++ code. Initially, we applied and validated the approach on a synthetic fault zone and subsequently on a real case studies within karstic carbonate aquifers of interest (Lez aquifer, Montpellier (France)). The inverse modeling approach has proven efficient in exploring hydrodynamic properties and then obtained both flow and transport forecasts with a satisfactory level of uncertainty. These works contribute to a better understanding of the hydrodynamic aspects of fault zones in carbonate environments through an innovative approach specific to its application. This study offers a reproducible method to understand and quantify hydrodynamics in aquifer in general and carbonated aquifer fault zones in particular. This improvement enhances the management strategies for groundwater resources in carbonated aquifers.