Two sites were picked to conduct a dye-tracer groundwater flow experiment. One site is located on the Florida Bay side of Key Largo near the Westin Inn (old Sheraton) Resort and the other directly across the island of Key Largo on the Atlantic Ocean side near Rock Harbor. Each site has a well cluster associated with it: one central site and 8 satellite sites located 30m from the center. Each well site, center and satellite, contained two wells. The deep well (A) was 40ft below the seafloor (fbsf) and the shallow well (B) was 20fbsf.

During drilling, cores of the Pleistocene limestone were collected and subsequently described and archived. After drilling to 40ft was complete, a 1-inch-diameter PVC pipe with 5-ft-long well screen at the base was inserted into the borehole. The well was finished off with sand and cement. A shallow well (20ft) was inserted into the same borehole and a cap of hydraulic cement placed around the PVC pipe and annulus of the borehole to prevent surface water from entering the borehole. After a day or two, the wells were purged using a 5 gallon per minute (gpm) pump to develop the well.

Dye Injection-

Two different fluorescent dyes were used for this study: Rhodamine and Fluorescein. The red Rhodamine dye was injected into the deep 40ft well and the yellow-green Fluorescein dye injected into the shallow 20ft well. A very concentrated solution for each was concocted such that a positive and distinct signal would be detected in the satellite wells. Five-hundred grams (500g) of Fluroescein dye (powder form) was mixed with 75 gallons of overlying seawater and injected into the shallow central well with a small impeller pump. Three liters of a 20% Rhodamine solution was mixed in 75 gallons of surface water and similarly pumped into the deep central well. Prior to mixing the dyes, all wells were sampled to obtain background fluorescent measurements.

Water sampling-

The central wells were not sampled throughout the study because they contained/retained high concentrations of dye; most likely a result of some of the solution binding to the sand or on the PVC screens. The satellite wells were sampled regularly using a small impeller pump. Salinity and temperature were measured in the field and samples collected from the wells were placed in dark amber polypropylene bottles. Fluroescein is highly photo reactive and will degrade in the presence of light, hence the collection in the dark bottles. After all samples were collected from both BSWC and OSWC, bottles were taken back to the USGS, St. Petersburg office and run on a filter fluorometer (Turner Design). The fluorometer was set up to run Fluorescein first and then filters within the light detection chamber were switched out and then samples run for Rhodamine. Fluorescein and Rhodamine fluoresce at different wavelengths, therefore different filters and lamps were used. Values, in ppb, from the fluorometer were directly read off the instrument and recorded.

Seepage Meter Construction-

For the seepage meter experiment in south Florida (Florida Bay, Florida Keys) a new design of seepage meters was developed. A fiberglass mold (~70cm x 70cm) was constructed from an off-the-shelf household skylight. Approximately 60 seepage domes were constructed and installed at various locations throughout the study region. Because the study regions underlying Pleistocene limestone is very heterogeneous several domes were installed in certain locations to combat hydrogeologic variations which control flow in and out of the subsurface. The fiberglass seepage meter provided a rigid frame which was then mounted to the bare rock seafloor throughout Florida Bay and along the Keys. Portland cement was mixed and used to seal the edges of the fiberglass domes. A hole was previously cut in the dome top to be used for connecting a plastic bag: the preferred bag was a turkey basting bag because of its construction material.

Seepage Meter Flux Measurements-

After domes were in place and the cement had cured, basting bags were placed on domes. These bags were left on the dome for a 24hr period and subsequently removed and the volume in the bag measured with a calibrated graduated cylinder. Knowing the volume over the 24 hour period and the area of the seepage meter a flux (L/m²/d) could be calculated. These measurements were recorded and placed in spreadsheets to determine variations by location and other environmental factors (i.e., rain, tide, and wave).