Regional scaling of groundwater recharge

This project is aimed at improving groundwater recharge estimates in a shallow coastal aquifer as part of a wider investigation of sustainable yield. Current techniques for measuring recharge provide estimates at various scales, both temporal and spatial. These scales may be different from those required for yield calculations. It is therefore necessary to provide methods that can both aggregate and disaggregate recharge measurements. Yield calculations do not adequately account for uncertainty in recharge estimates. The aggregation and disaggregation techniques provided in this thesis show how to estimate the uncertainty in recharge and provide indications of the data requirements of the methods. Finally it is shown that yield calculations require accurate determination of mass balance. For aquifers with shallow water tables, this means that it is necessary to determine plant water use and its impact on recharge.

Small spatial scale estimates of recharge have been made using the water table fluctuation method and the chloride mass balance method. A time series model has been developed for the water table fluctuation method of estimating recharge. The water table fluctuation method measures gross recharge, or the sum of all rainfall reaching the water table. The model uses a variable specific yield based upon the soil moisture retention curve, filters out the influence of the Lisse effect on the water table and is capable of detecting recharge even if the water table does not rise. It is shown that high temporal resolution water level data and a variable specific yield are essential for the accurate estimation of recharge.

This study has shown that considerable amounts of data are required to minimize uncertainty in the chloride mass balance method. It is shown that a single observation of the chloride concentration of the groundwater can deviate as much as 50% from the average. If observations are averaged over 100 days the error reduces to 20% and after 10,000 days the error is expected to be less than 10%. The chloride mass balance measures long term net recharge, or the gross recharge remaining after accounting for evapotranspiration. A temporal disaggregation method based on a water balance is presented for determining the short term variability of recharge .

In the aquifer under investigation, the average annual recharge volume is between a half and a third of the available storage. This means that the temporal variability of the recharge is extremely important in determining the sustainable yield of the aquifer. Annual recharge estimates were found to be well correlated with the annual rainfall, but the annual evapotranspiration sourced from the groundwater was found to be inversely related to the annual rainfall. In a drought period there is reduced recharge and increased discharge by phreatophytes. A simple model has been used to show how these observations can be incorporated into a probabilistic determination of the sustainable yield of the aquifer.