CHARACTERIZATION OF TIDAL VARIATIONS IN THE MIXING ZONE OF A LATE-PLEISTOCENE CARBONATE AQUIFER ON SAN SALVADOR ISLAND, BAHAMAS

A one-month monitoring study (January, 2008) provided measurements of water level, pH, specific conductance (SpC), dissolved oxygen (DO), and temperature (T) from one of thirteen abandoned production wells at the Line Hole well field on the northwest coastline of San Salvador Island, Bahamas. Our primary interest is the diurnal and longer fluctuation of this Late-Pleistocene carbonate aquifer to external mechanisms such as tides and rainfall. In particular, we seek to better characterize the nature of the mixing zone at the base of this fresh-water lens in an aquifer historically impacted by salt water intrusion from overutilization.

We installed one YSI piezometer (Level Scout) and two YSI datasondes (600XLM) in the well; one datasonde at a depth of seven meters below top of casing (~5.5 m below the water table, position one) and the piezometer and other datasonde at a depth of ten meters below top of casing (~8.5 m below the water table, position two). These deployment depths reflect field observations of SpC that signify halocline positions in the well. All instruments were programmed to collect data every ten minutes.

Semi-diurnal, mixed-magnitude fluctuations of water level were observed that are in sync with, but of lesser amplitude than, the calculated predictions for high and low tide. Specifically, we observe fluctuations with a minimum range of ~10 cm and a maximum range of ~60 cm that align with the neap and spring tides, respectively. Values of SpC and T at positions one (SpC_av = 1,310 μS/cm ± 224 μS/cm, T_av= 26.8 °C ± 0.049 C°) and two (SpC_av = 10,810 μS/cm ± 965 μS/cm, T_av= 26.1 °C ± 0.033 C°) track changes in water levels. Values of pH vary little from the mean values of 7.10 ± 0.03 and 7.37 ± 0.02 at positions one and two, respectively. Values of DO span a broad range at both positions (DO₁ = 4.6 mg/L ± 3.5 mg/L, DO₂ = 2.6 mg/L ± 1.2 mg/L) with minor changes in sync with the tides. No excursions in any dataset align with observed precipitation events.