Showcasing a global-scale, 3D coastal groundwater salinity model

Being extensively available and of high quality, groundwater is the primary source of freshwater in coastal regions globally. However, due to anthropogenic and natural drivers, groundwater salinisation is a growing threat to this resource’s long- and short-term viability. The causes and timescales of aquifer salinisation are complex and difficult to quantify, information essential for suitably timed mitigation strategies. One way to inform these strategies and develop storylines of future freshwater (un)availability is through 3D groundwater salinity modelling. These models can predict current groundwater distributions and quantitatively assess the impacts of a projected increase in groundwater extraction rates and sea-level rise. Until recently, detailed 3D models on this scale have been largely unattainable due to computational burdens and a shortage of in-situ data. Fortunately, recent developments in code parallelization, reproducible modelling techniques, and access to high-performance computing (e.g., via parallel SEAWAT) have made this feasible. Machine learning and data mining developments have also allowed an unprecedented opportunity to constrain and calibrate those models. With this in mind, we present our progress towards global 3D salinity modelling by showcasing a regional-scale model in the Mediterranean Sea area. This test case uses newly developed, automated geological and salinity interpolation methods to create initial conditions while implemented in a parallelized version of SEAWAT. The modelling outcomes highlight the potential of supra-regional scale modelling in the context of global (planetary) processes and localised anthropogenic effects.