Assessing the recharge of a coastal aquifer using physical observations, tritium, groundwater chemistry and modelling.

Assessing recharge is critical to understanding groundwater and preventing pollution. Here, we investigate recharge in an Australian coastal aquifer using a combination of physical, modelling and geochemical techniques. We assess whether recharge may occur through a pervasive layer of floodplain muds that was initially hypothesized to be impermeable. At least 59% of the precipitation volume could be accounted for in the shallow aquifer using the water table fluctuation method during four significant recharge events. Precipitation events <20mm did not produce detectable aquifer recharge. The highest recharge rates were estimated in the area underneath the floodplain clay layer rather than in the sandy area. A steady-state chloride method implied recharge rates of at least 200mm/year (>14% of annual precipitation). Tritium dating revealed long term net vertical recharge rates ranging from 27 to 114mm/year (average 58mm/year) which were interpreted as minimum net long term recharge. Borehole experiments revealed more permeable conditions and heterogeneous infiltration rates when the floodplain soils were dry. Wet conditions apparently expand floodplain clays, closing macropores and cracks that act as conduits for groundwater recharge. Modelled groundwater flow paths were consistent with tritium dating and provided independent evidence that the clay layer does not prevent local recharge. Overall, all lines of evidence demonstrated that the coastal floodplain muds do not prevent the infiltration of rainwater into the underlying sand aquifer, and that local recharge across the muds was widespread. Therefore, assuming fine-grained floodplain soils prevent recharge and protect underlying aquifers from pollution may not be reasonable.