A Model for Fast Estimation of Seawater Intrusion in Coastal Aquifers: Simulating Flow and Land Subsidence

Abstract

We introduce a novel groundwater flow model designed for the rapid estimation of seawater intrusion (SWI) in coastal aquifers. Drawing on Girinskii's potential theory (1947) for 2D aquifer flow, the model extends Strack's (1976) adaptation to coastal aquifers under transient state conditions. Key features include applicability to heterogeneous unconfined aquifer systems under time-dependent pumping and spatially distributed groundwater recharge. Assumptions include a sharp seawater-freshwater interface, prevailing horizontal flow, and neglect of flow rates within the saltwater wedge. Built upon a finite-element 2D saturated groundwater simulator, the model derives a solution for the potential using a non-linear formulation of the classic flow equation and can accommodate prescribed-head or prescribed-flux boundary conditions at inland boundaries, as well as time-varying head conditions at the shore boundaries, which is crucial for addressing long-term sea-level rise (SLR). Noteworthy modifications enable the model to estimate land subsidence through a 1D vertical consolidation model directly in terms of potential, providing a holistic view of aquifer dynamics. The model is successfully tested against analytically based solutions and can facilitate swift calculations of aquifer vulnerability indicators, such as SWI toe location, increase in dissolved salt mass, and apparent land subsidence due to compounded effects of SLR, groundwater pumping, and long-term variable recharge conditions. Our presentation will showcase the model's features, preliminary results focusing on the effects of aquifer heterogeneities on the intensity of SWI and land subsidence patterns, model strengths and limitations. Emphasis will be placed on the model computational efficiency, enabling rapid estimation of crucial SWI indicators. Furthermore, the discussion will outline future steps, highlighting the need to balance the simplicity of working hypotheses with the computational demands of more intricate variable density models in assessing complex coastal aquifer dynamics.