Intercomparison of marine tracer and catchment-based submarine groundwater discharge estimates for a karst aquifer

Abstract

Submarine groundwater discharge (SGD) is a major source of solutes to the ocean. Consequently, the accuracy of estimates of the magnitude and seasonal variability of SGD is critical for the management of coastal marine ecosystems. However, very few intercomparison studies between different conceptual approaches have been made to assess the reliability and consistency of SGD flow estimates while simultaneously gauging the applicability of underlying models and budget assumptions in real-world scenarios. To date, intercomparison efforts have focused on homogeneous aquifers, and less so on highly heterogeneous aquifers such as karst systems. Here we compare the results of tracer-based (salt mass balance, tidal prism and radon mass balances) and catchment-based methods (water balance, semi-distributed karst network model) to assess SGD at a data-rich coastal location. All approaches tested here reveal a positive relationship between SGD fluxes and groundwater level measured close to the preferential groundwater flow pathways in the phreatic karst aquifer (e.g. water-saturated conduits or fracture network). We use this finding to derive site-specific relationships between groundwater level and SGD flow rates. We derive these relationships from least square fitting of Darcian and Non-Darcian models and regression analysis. The relationships validate the findings derived independently from hydraulic/hydrogeological modelling of the complex karst networks. By combining various estimates made with distinct assumptions, this method produces robust estimates of the SGD seasonal variability, essential to develop optimal management strategies of SGD resources in coastal karstic regions. Where direct measurements of SGD are not possible, monitoring of coastal groundwater level close to preferential phreatic groundwater pathways can be thus used as a useful indicator to compare past groundwater flow measurements and derive site-specific relationships, particularly in settings where large flow rate changes are present such as karst aquifers.