Wairau River-Wairau Aquifer Interaction

Abstract

This report presents initial findings from an investigation into understanding the nature of transient recharge from the Wairau River to the Wairau Aquifer. This investigation has been a year-long collaboration between Marlborough District Council (MDC), the Water & Earth System Science Competence Cluster (WESS) at University of Tübingen in Germany, and Lincoln Agritech. The intention of this phase of the Wairau Aquifer recharge project is twofold: To review our conceptual understanding of the river and aquifer, and identify knowledge gaps; To develop a numerical model to quantify the river-aquifer exchange based on that conceptual understanding .

Historical flow gauging data shows that approximately 7.5 m³/s is recharged from the river to the Wairau Aquifer between Rock Ferry and Wratts Road at flows less than 20 m³/s. Temporary flow sites have been established at Rock Ferry, SH6 and Wratts Road, and a good flow-rating curve has been established at Rock Ferry. The difference between flows at Rock Ferry and SH1 indicates that aquifer recharge increases as the river flow increases. The preliminary evidence suggests that 15 m³/s or more may be lost as recharge pulses during high flow events, although further work is required to correct flow for time lags in the river which will improve these estimates.

One of the key findings of this report is that the river appears to be perched above the Wairau Aquifer across its main recharge reach. This means that there is a vertical hydraulic gradient between the river and aquifer where most of the recharge occurs, which theoretically simplifies the calculation for estimating transient recharge rates.

Another key finding is that the hydraulic nature of the aquifer is more complex than our previous simple unconfined aquifer conceptual model. Groundwater monitoring records and aquifer test data indicate that the aquifer is highly stratified. This stratification explains the observation that the aquifer may be perched. The reason for this is that a strong vertical to horizontal anisotropy in permeability enables groundwater to potentially drain faster than it can be recharged by the river. There is also the possibility of distinct upper and lower aquifer horizons, although this needs to be explored further.

The implication of aquifer stratification for estimating river recharge is that the groundwater monitoring sites are representative of quite localised conditions. Therefore, for the prediction of recharge rates, groundwater data can be used to constrain hydraulic parameters within our numerical model, but it is preferable to calculate recharge rates based on river observations rather than changes in groundwater levels at a particular site. Furthermore, if the is indeed river perched over its main recharge reach, as the available evidence suggests, the prediction of transient recharge rates will be largely determined by river geometry, and the relationship between stage and wetted river bed perimeter.

A steady state numerical model has been built and calibrated to accord with our conceptual understanding. A combination of groundwater level observations and surface water fluxes were used for calibration targets. The best-fit "compromise" solution resulted in data fits that are well within the measurement uncertainty ranges. One of the key findings of the model was that the river was required to be perched to enable calibration, which supports field evidence outlined in this report.

Further work for 2015 will involve a more intensive field program to improve our understanding of the relationship between river flows and groundwater levels immediately adjacent to and beneath the river. Transient calibration of the numerical model will also begin, and we intend to use the final calibrated model to estimate transient aquifer recharge rates.