On the resilience of freshwater lenses on small islands: evidence of the long-term effects of groundwater extraction on Bonriki Island, Kiribati. Field data and modeling results highlight the strong control of precipitation variability and pumping on the temporal dynamics of the freshwater lens. 2014) used numerical simulations to infer the effect of variable rainfall on the thickness of freshwater lenses for the atoll islands of the Maldives.
A comprehensive, calibrated model of Home Island, part of the Cocos (Keeling) Islands, was produced by Ghassemi et al. Here we present a study of the freshwater lens on the island of Bonriki, South Tarawa, Kiribati. The black solid rectangle in the main figure outlines the location of the map shown in Figure 2.
The blue lines in the western region of the island indicate the location of the inland lakes. Tarawa Atoll is located in the hot and humid tropical zone that covers most of the area. There is a marked demarcation between the hydraulic properties of the Holocene and Pleistocene units of Bonriki Island (Falkland and Woodroffe, 2004).
The chronology of the studies and the annual volumes of groundwater extracted are summarized in Figure 4.
Numerical model
Three-dimensional conceptual block diagram of the Bonriki groundwater system, modified from Bosserelle et al. Map showing the outline of the finite difference network and the distribution of recharge areas across Bonriki Island. A salinity water head equal to mean sea level was assigned to all cells with the specified head.
The TDS concentration of the supplement was made dependent on the charging rate to account for its effect. In this case, the difference with Bonriki Island is mainly due to a higher bottom zone thickness in the Cocos (Keeling) Islands model (500 mm versus the 300 mm assumed for our study), with our WATBAL. A moderate response of water levels to rainfall was observed in this zone compared to the other parts of the Lens Reserve (Sinclair et al., 2015).
The infiltration galleries are implemented in the model by assigning abstraction rates to the cells in the top layer of the model grid that coincide with the location of the galleries. Three additional scenarios were considered in which the scaled abstraction rates were 75%, 50%, and 25% of the abstraction rates used in the calibration simulation (i.e., as listed in Table 1). The initial heads and concentrations for the transient model were determined based on the heads and concentrations after the final time step of the model with the updated calibration.
The concentrations of the multilevel observation wells were compared with the model-calculated concentrations in the cells corresponding to the location and depth of the measuring point. Comparison of calculated and measured EC values of the pumped water at the mainstem during January 1987 to June 2014. An improved interpretation of the lens evolution is possible by plotting ΔVf versus Vabs,tot, as given in Figure 11.
At the onset of the pumping period (i.e. 1987), ΔVf shows an increasing trend, which is due to the lens adapting to the introduction of abstraction. The color of the data points reflects the value of ΔV'f (positive values indicate lens growth). In a similar way, QR and Qcoast are correlated at the one-month temporal resolution of the model (Figure 13).
The average ratio of the annual rate of volumetric abstraction to the annual rate of volumetric replenishment during the period was 36%. However, the observed trends call for a re-evaluation of freshwater lens management strategies.
Conclusions
As this study has shown, the response times of the lens, even in permeable, small-scale lens systems such as Bonriki Island, mean that monitoring must be sustained over many decades to fully capture the abstraction impacts. While the salinity of the water extracted from the galleries appears to be more responsive to droughts in recent years, there are currently no immediate problems with the salinity of the water supply. Furthermore, given the high water consumption of the deep-rooted palm trees, vegetation management can be adopted that takes into account the trade-off between the value of the trees for nutrition and environmental functioning, and their negative impact on the lens through withdrawals from evapotranspiration.
The results clearly show that under natural conditions, the volume of fresh water is subject to significant changes, which are caused by the intensity of rainfall. This showed that abstraction causes sustained lens shrinkage over the 27.5 years of pumping considered in this study. The rate of decline of freshwater volume showed no decrease with time, meaning that freshwater volume has not yet reached a new equilibrium with respect to pumping stresses.
The efforts to monitor groundwater and withdrawal water salinities over several decades have proven essential in assessing the state of the freshwater resource. But as thinning of the lens continues and shallow groundwater salinities become more variable over time, the relationship between R12 and extraction water salinity is no longer fixed. Caution is therefore needed in relying on such indicators to determine the health of the freshwater reserve.
Due to its size and rainfall, the Bonriki freshwater lens is relatively resilient compared to other islands, and thus this is especially true for smaller islands with less recharge. Further work needs to focus on management options for groundwater extraction on Bonriki Island and a reassessment of the appropriate sustainable yield. We thank the Australian Government and the European Union for funding, the staff of PUB and MPWU, Kiribati, for their help with fieldwork and data input, Anthony Falkland and Ian White for technical expertise provided during the projects, the Water and Sanitation Program of Geoscience, Energy and Maritime Division at SPC for support, and our colleagues who worked with us.
Kiribati – Tarawa Water Resources Pre-Design Study, Australian Government Department of Housing and Construction, Canberra, Australia. Climate and Abstraction Impacts in Atoll Environments (CAIA): Sustainable Management of the Bonriki Water Reserve, Tarawa, Kiribati. From days to decades: numerical modeling of freshwater lens response to climate change on small low-lying islands, Hydrology and Earth System Sciences doi: 10.5194/hess.
The El Niño-Southern Oscillation determines the salinity of the freshwater lens beneath a coral atoll in the Pacific Ocean. Climate change, recharge and salinity of freshwater lenses of a coral atoll in the Pacific Ocean, A new focus on groundwater-seawater interactions.
