The atoll's increasing vulnerability to ENSO-driven droughts and their water-related impairments prompted the assessment of the electrical resistivity tomography (ERT) and ground penetrating radar (GPR) to generate subsurface models that inform/guide drought response actions and investments with confidence. The ABEM Terrameter resistivity kit and the Mala Easy Locator Pro WideRange GPR (dual frequency, 160 and 670 MHz) were tested on Nukulau Island, Fiji, and on South Tarawa, Kiribati. The test was conducted under high rainfall conditions with the identification of the water table and the estimation of freshwater lens thickness being the main objectives.
On Nukulau island, high resolution ERT data was generated with the Res2DInv software with a distinctive lateral and vertical variation in resistivity exhibiting an estimated freshwater lens thickness of up to 7 m atop basal saline water. Clear distinction between unsaturated and saturated zones (water table) was easily made. Contrasts between saturated materials at depth based on salinity levels and the possible presence of undulating and porous limestone was easily noticed – calibration of these models was made through known depth and salinity records from similar atoll settings.
The respective GPR survey was completed faster and the data processing, using the RadExplorer software, required signal improvement techniques, such as background removals, and assigning appropriate velocity values to subsurface materials at different depths and salinity composition, converting arrival time models to depth models. The model showed similar freshwater lens shape with a lens thickness of 3.5 m, 50 % less than the ERT model estimation.
The second GPR traverse was undertaken in Bonriki in South Tarawa, Kiribati. Two monitoring bores along the profile provided freshwater lens thickness of up to 18 m before decreasing to 7 m towards the end. GPR results, again, showed a reduced lens thickness of up to 6 m. GPR is limited in confidently defining the lens, possibly attributed to the limited signal strength and penetration of the two frequencies as opposed to the detailed depth-varying ERT. The availability of monitoring bores to provide known depth and salinity measurements are critical to both methods to allow for calibration. Further, the identification of the water table can be difficult during the wet season and in moderate to high vegetation areas due to the increased moisture content in the unsaturated zone coupled with the influence of the capillary fringes causing reflections, and thus, increasing uncertainties of around 0.2 m as tested in similar conditions elsewhere.