Probabilistic delineation of well capture zones in a multi-aquifer system under geological and parametric uncertainty

In the Venetian-Friulian plain, in northeast Italy, groundwater is the primary source of drinking water. In light of the concerning effects of climate change, invasive pollution, and evolving population needs, developing adequate strategies for managing the regional water resource is of utmost importance. Any such strategy must take into account the role of groundwater and rely on a comprehensive understanding of its availability, vulnerability, and potential evolution. However, the complexity of the hydrogeological context in the area contributes to the uncertainty that is naturally inherent in groundwater systems modeling, which can affect the reliability of model predictions. To address the issue of uncertainties, probabilistic approaches are increasingly recommended. In this study, we adopted a probabilistic framework with the objective of delineating well capture zones and time-related capture zones for two major wellfields located in the Venetian-Friulian Plain. The wells under investigation extract groundwater from deep confined aquifers of a multilayered aquifer system and represent major sources of drinking water in the area; hence, ensuring their protection is crucial. The adopted framework relied on groundwater flow and transport modeling and addressed both geological and parametric uncertainty. Geological uncertainty was taken into account by relying on a multiple geological modeling approach based on categorical geostatistical simulations, while parametric uncertainty was incorporated by sampling model parameters from prescribed distributions. The groundwater flow model also included a simplified surface water module to represent river–aquifer interactions, thus rendering a basic representation of coupled surface–subsurface processes. A global sensitivity analysis allowed us to identify the most influential parameters, which were then calibrated, while the less influential parameters were fixed. Moreover, it allowed us to quantify the influence of geological uncertainty. Post-calibration Monte Carlo simulations yielded ensemble statistics of hydraulic heads, river flow rates, river–aquifer exchanges, and transport-related variables, including probabilistic capture zones and time-related capture zones. The results of this research show that uncertainty originating from the spatial configuration of the hydrogeological facies (i.e., geological uncertainty) has a significant influence on model predictions, especially when dealing with the transport problem. Furthermore, we demonstrate that surface water bodies, such as rivers and springs, play a primary role in the behavior of the investigated groundwater body. These control the main groundwater flow directions and, consequently, the shape and extent of the probabilistic capture zones, even for the relatively deep confined aquifers considered. Finally, a comparison of the model results under diverse hydroclimatic settings revealed that persistently different climatic conditions could substantially alter the principal groundwater flow directions, potentially leading to a complete reconfiguration of the capture zones. These findings highlight the importance of explicitly considering geological uncertainty, surface water–groundwater interactions, and climatic variability in the sustainable management and protection of groundwater resources.