Assessment of saltwater intrusion in southern coastal Broward County, Florida

Of the counties in southeastern Florida, Broward County has experienced some of the most severe effects of saltwater intrusion into the surficial Biscayne aquifer because, before 1950, most public water-supply well fields in the county were constructed near the principal early population centers located less than 5 miles from the Atlantic Ocean. The construction of major regional drainage canals in the early 20th century caused a lowering of the water table and a gradual inland movement of the saltwater front toward the well fields. The U.S. Geological Survey began field investigations of saltwater intrusion in the Biscayne aquifer of southeastern Broward County in 1939. As part of the present study, the positions of the saltwater front in 1945, 1969, and 1993 were estimated using chloride concentrations of water samples collected between 1939 and 1994 from various monitoring and exploratory wells. The data indicate that, between 1945 and 1993, the saltwater front has moved as much as 0.5 mile inland in parts of the study area. The position and movement of the saltwater front were simulated numerically to help determine which of the various hydrologic factors and water-management features characterizing the coastal subsurface environment and its alteration by man are of significance in increasing or decreasing the degree of saltwater intrusion. Two representational methods were applied by the selection and use of appropriate model codes. The SHARP code simulates the position of the saltwater front as a sharp interface, which implies that no transition zone (a zone in which a gradational change between freshwater and saltwater occurs) separates freshwater and saltwater. The Subsurface Waste Injection Program (SWIP) code simulates a two-fluid, variable-density system using a convective-diffusion approach that includes a representation of the transition zone that occurs between the freshwater and saltwater bodies. The models were applied to: (1) approximately replicate predevelopment and current positions of the interface in the study area; and (2) study the relative importance of various factors affecting the interface position. The model analyses assumed a conceptual model of uniform easterly flow in the aquifer toward points of offshore discharge to tidewater. Measurements of water-table altitude and the depth to the interface in the study area exhibit an interrelation that differes substantially from the classical Ghyben-Herzberg relation. However, both model codes simulated water-table altitudes and interface positions that were generally consistent with the Ghyben-Herzberg relation but differed substantially from observed data. The simulate interface positions were inland of the known positions, and simulate water-table altitudes were higher than measured ones. The SHARP and SWIP simulations were in general agreement with each other when a low value of longitudinal dispersivity was specified in the SWIP simulation and also for higher values of longitudinal dispersivity when modified dispersion algorithms were used in SWIP that greatly reduced the simulated degree of vertical dispersion. Sensitivity analyses performed using the SHARP code indicated simulation results to be relatively insensitive to a substantial change in the specified slope of the base of the aquifer and moderately sensitive to a 150-percent change in net atmospheric recharge to the aquifer (rainfall minus evapotranspiration). Representing well-field pumping by the City of hallandale had only a minor, localized influence on the simulated regional interface position. Using various cross-sectional grid designs in applications of the SWIP code, near convergence of all lines of equal concentrations in the transition zone was achieved within a simulation time of 10 years. The simulated equilibrium interface location was sensitive to substantial spatial variations in the specified hydraulic conductivity values, but was relatively insensitive to seasonal varying