In this study, we analyzed core samples (10.2 cm in diameter, 139 m cumulative length) obtained from five core holes. Three test core holes were drilled under the direction of the Miami-Dade County Water and Sewer Department in 1998 (test core holes adjacent to production wells S-3168, S-3169, and S-3170), and the U.S. Geological Survey (USGS) supervised drilling of two continuously cored test wells (observation well G-3772 and injection well G-3773) during 2002 (Figs. 1C and 1D). The distance of a transect including these five wells is ~1.7 km, with well-to-well distances of separation ranging from 0.07 to 0.83 km. The five cores were slabbed and visually analyzed using a 10x magnification hand lens and binocular microscope. Standard transmitted-light petrography aided examination of 111 thin sections of core samples from the G-3772, G-3773, and S-3169 wells (Figs. 1C and 1D). Core and thin-section analyses determined lithofacies, vertical patterns of lithofacies, sedimentary structures, cycle boundaries, and assessed lateral correlation or variability of these features. Lithofacies were defined by allochem types, fabric, sedimentary structures, bedding type, and diagenetic features, using a combination of classification schemes and terminology from Dunham (1962), Embry and Klovan (1971), and Lucia (1999). Relationships between lithofacies and petrophysical properties (porosity and permeability) were assessed using methods of Lucia (1995, 1999).

A Mount Sopris OBI-40 digital optical logging tool produced continuous digital images of the borehole walls in the G-3772 and G-3773 core holes. These images were used to appraise the distribution of highly porous stratal intervals not recovered in the core record and to image the borehole cyclostratigraphy and associated pore network. Four air-rotary wells (G-3257A, G-3258A, G-3816, and G-3817), drilled under supervision of the USGS, were optically logged for this study, and continuous images of their borehole walls were compared to the cyclostratigraphy and pore systems observed in the cored wells (Fig. 1D). Observation wells G-3257A and G-3258A were drilled in 1983, and borehole injection wells G-3816 and G-3817 were completed in 2003. Percent vuggy porosity observed in digital borehole images in the G-3772 and G-3773 wells was calculated employing a technique described in Cunningham et al. (2004a). Cyclostratigraphic correlations were corroborated by comparison to larger-scale, inclusive hydrogeologic and cyclostratigraphic relations currently being established for the entire Lake Belt area (Cunningham et al., 2004b, 2004c, 2006).

Taxonomy of mollusks and foraminifera from selected lithofacies was determined to assist in interpretation of paleoenvironments. Mollusks from 13 core samples obtained from the S-3168 and S-3170 test core holes (Fig. 1C) were prepared and identified. Preservation of most of the mollusks present in the samples was as either molds or casts. Core samples were initially examined under a binocular microscope to observe diagnostic characteristics of the molluskan taxonomy and to compare with published documents (e.g., Mansfield, 1939; Olsson and Harbison, 1953; DuBar, 1958; Olsson and Petit, 1964; Abbott, 1974; Portell et al., 1992). Where appropriate, clay squeezes or latex casts were made of the molluskan molds to help in identification. Many molluskan species present in the Pleistocene units of south Florida are extant or represented by close relatives, so interpretation of paleoenvironmental settings is based on living fauna. Publications by Perry and Schwengel (1955), Warmke and Abbott (1961), Abbott (1974), Andrews (1977), and Brewster-Wingard et al. (2001) support our paleoenvironmental interpretations. Information contained in Bock et al. (1971) and Poag (1981) aided our benthic foraminiferal identifications at the genus level for the 111 thin sections. Data described in Bock et al. (1971), Rose and Lidz (1977), Poag (1981), and Lidz and Rose (1989) assisted our interpretation of paleoenvironmental conditions based on foraminiferal taxonomy.

Data combined from borehole fluid-conductivity and fluidtemperature logs, caliper logs, digital borehole image logs, and borehole flow-meter measurements delineated intervals with inflow or outflow from boreholes. Considered together, these logs provided information on formation permeability (Keys, 1990; Paillet, 2004). Fluid-conductivity logs have been used to assess changes in concentration of dissolved solids in the borehole fluid column (Keys, 1990). Asharp change in borehole fluid conductivity or fluid temperature or both can identify borehole intervals showing inflow or outflow (Keys, 1990). Caliper and digital borehole image data used in combination with boreholefluid logs can identify potential high-permeability flow zones. A suite of caliper, digital borehole image, borehole fluid-conductivity, and fluid-temperature logs were obtained for the G-3772 observation well and G-3773 injection well. In the G-3773 well, a Century Geophysical Corporation electromagnetic flow meter was used in static (stationary measurements) and trolling modes (measurements while the tool moves up or down the well bore) to measure vertical borehole groundwater flow under stressed and ambient conditions. A 10.2-cm-diameter suction-lift centrifugal pump withdrawing at an average rate of 454 L/min was used in the G-3773 well to produce stressed conditions during a single-well flow-meter test. Ambient measurements were collected at the G-3773 well during unstressed conditions.

The injection well G-3773, observation well G-3772, and a well-field production well (S-3164) were used during a forced-gradient, convergent tracer test using a fluorescent dye and deuterium in April 2003 (Fig. 1D). The G-3816, G-3817, G-3772, and S-3164 wells were used for additional tracer tests, using bromide, dissolved gas, deuterium, and colloidal particles, conducted in February and March 2004 (Fig. 1D). This study uses only results from the 2003 test; however, Renken et al. (2005) discuss some results of the 2004 tests.