The mechanisms of transient groundwater recharge are often not well understood or quantified, but can be explored through numerical simulations. There are numerous ways to simulate recharge for groundwater models, with varying levels of complexity. In this study, we compare simulations of groundwater recharge in the floodplains of the lower River Murray, where recharge from rainfall and flooding is a critical part of the water balance. The complexity of conceptualising the timing and amount of groundwater recharge triggered this research.
The selected codes for comparison are: RCH with EVT in MODFLOW, UZF in MODFLOW, HYDRUS, and LEACHM. Each is used to simulate a dynamic scenario of twenty years of rainfall, evapotranspiration, changes in river level, and flooding within a thin unsaturated zone of 0.5-3 m. Loxton weather station data was used to simulate seasonal changes from 1992 to 2011. The soil parameters of the Coonambidgal and Monoman Formation are derived from field measurements and pedotransfer functions.
The RCH-EVT and UZF methods generated similar volumes of recharge, but the difference in timing of recharge between two approaches can be significant, with two to five weeks of lag-time between applied water and net recharge. In some cases, the wetting front from rainfall overtook and mixed with a flood. The higher recharge chiefly happens during the cold seasons, from March to August, due to lower evapotranspiration rates. The lags in infiltration depend on the thickness of the unsaturated zone. To understand the salt movement in the vadose zone and magnitude of recharge, the LEACHM and HYDRUS model were developed as benchmarks, which investigated the salinity and water stress on vegetation health as well. A long-term simulation revealed that HYDRUS delivered a significant recharge, 42% above UZF, and ET removed by LEACHM was 61% larger than UZF. It is instructive to compare results for a significant climate event, as observed during the 2002 drought where the groundwater recharge was hampered.
This research contributes to understanding hydrological features, surface water-groundwater interaction, within the MDB floodplain in SA. The developed method in this research can be employed in models within the study area, which can be used in simulating solute transport. This research contributes to modelling tools for managing ecosystem health in the MDB floodplain.