Home Abstract Introduction Hydrogeology of S. Florida Ground-Water Movement - Movement Based on Natural Isotopes   >  Carbon Isotopes   -  Uranium Isotopes - Flow Patterns - Effects of Rising Sea Level - Upwelling Ground Water Subsurface Storage Summary and Conclusions References PDF Version

Ground-Water Movement in the Floridan Aquifer System in Southern Florida: Ground-Water Movement Based on Natural Isotopes as Tracers

Carbon Isotopes

The radiocarbon dating technique has been an important and accepted research tool in archeology and geology since its inception in 1946 by Libby (1955). However, its use in hydrogeology has been dubious because of the uncertainties in comparing the carbon-14 in dissolved carbon species in ground water with respect to that in the water when it was last in contact with the atmospheric reservoir of carbon-14. An understanding of the involved chemical processes and the reservoir through which the ground water moves is essential to the interpretations and corrections that would apply to the measured carbon-14 in the sample.

Carbon-14 measurements by the U.S. Geological Survey were by liquid scintillation counting of benzene which was synthesized from the carbonate in a 30-gallon (gal) sample of ground water (Pearson and Bodden, 1975; Thatcher and others, 1977); however, measurements by the Tritium Laboratory, Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, were by gas proportional counting of carbon dioxide from a 55-gal water sample (Stuiver and Ostlund, 1980). By convention, the measurements were compared with standard National Bureau of Standards oxalic acid to determine the relative activity (R) for age-dating. The carbon-13/carbon-12 stable isotope ratio (¹³C) was determined by mass spectrometry. Also by convention, the measured relative carbon-14 activity (R) was normalized to a ¹³C value of -25 (per mil), the value for wood with respect to the standard Pee Dee Belemnita (PDB), to compensate for isotopic fractionation. The generally accepted equation for isotopic fractionation normalization is

D14C = d14C-2 (13C + 25)

[ 1+ d14C  ]   (3) 1,000

where the normalized activity, D¹⁴C, equals the determined activity, d¹⁴C, corrected for isotopic fractionation, assuming that a carbon-13/carbon-12 change from the accepted value for wood (¹³C =25 with respect to PDB) induces a change in the carbon-14/carbon-12 ratio of the sample. A 10 change in ¹³C value, therefore, results in a 20 change in the carbon-14 millesimal difference (d¹⁴C).

The percent of modern carbon (PMC), corrected for isotopic fractionation normalization or the percentile difference, was calculated using the following equation:

PMC = 100+

D14C 10

±s

D14C 10

(4)

where s D¹⁴C/10 is the standard error. The age of the sample is expressed as apparent age owing to the uncertainty about past production of carbon-14 and the origin of the carbon in the water. The apparent age is based on the Libby half-life of 5,568 yr and was calculated using the following equation:

Apparent age (before 1950) = -8033 ln (PMCx10^-2)   (5)

The measured relative carbon-14 activities, R (that is, uncorrected for isotopic fractionation normalization), ranged from about 0.04 to 65.9 percent, and the normalized activities, PMC, ranged from about 0.04 to 62.9 percent. The experimental error for the analyses by the U.S. Geological Survey ranged from 0.2 to 1.5 percent, and that for the Tritium Laboratory, University of Miami, ranged from 0.1 to 0.3 percent. The corrections for isotopic fractionation normalization had relatively insignificant effects on the activities of the samples. No corrections for dilution or chemical reaction during transit were applied; therefore, the activities do not represent absolute age. Ages calculated according to equation 5 are maximum estimates. Corrections for reactions led to younger age estimates. However, insufficient data are available to correct PMC for reaction effects. Generally, the samples of brackish ground water from the Upper Floridan aquifer had lowest carbon-14 activities.

The PMC (carbon-14 activity corrected for isotopic fractionation normalization) of samples of brackish ground water from the Upper Floridan aquifer at sites 5, 9, 10, 12, and 14 (table 4) ranged from about 0.04 (a value that is near the limit of the dating technique) to 16.7 percent. The stable isotope ratios (¹³C) ranged from +4.3 to -3.9 (table 4).

Variations in the PMC with depth are indicated. At site 10, location of the Alligator Alley test well, the PMC of samples between 895 and 1,618 ft ranged from 4.5 to 16.7 percent and the carbon-13 stable isotope ratios (¹³C) ranged from -1.6 to -2.6 . The well was cased to a depth of 895 ft (225 ft below the estimated top of the Floridan aquifer system), and the open borehole was drilled to a depth of 2,811 ft into the limestone and dolostone of the Lower Floridan aquifer. Samples from various depths in the open hole were obtained chiefly with packers. A monitor tube 2 inches in diameter with perforations from 811 to 816 ft provided samples of ground water from near the top of the Upper Floridan aquifer. The chief water-bearing zone in the Upper Floridan aquifer at the well site is at about 1,030 ft, at the unconformable contact of the Ocala Limestone with the overlying Suwannee Limestone. Samples of ground water, presumably from the Suwannee-Ocala contact in the Upper Floridan aquifer, generally had higher carbon-14 activities (except the seawater-like samples from the Lower Floridan aquifer), thereby suggesting that they are relatively younger in age.

At site 5 (well STL-255), the PMC for samples of brackish ground water from depths of about 898 to 1,663 ft in the Upper Floridan aquifer were virtually zero (0.18 and 0.04 percent), and the respective carbon-13 stable isotope ratios were -3.3 and 4.3 . The deeper sample (1,610 to 1,663 ft), which had a chloride concentration of only 530 mg/L, contained an unusually high concentration of carbon-13. The high carbon-13 concentration in the deeper sample may indicate dissolution of dolomite (Hanshaw and Back, 1971, p. 147) and, therefore, greater dilution of carbon-14. The carbon-14 activity of the deeper sample would, therefore, be lower chiefly because of the dilution. The presence of brackish ground water at that depth is further evidence of the variability of water chemistry in the aquifer system.

Figure 11. Carbon-14 activities and carbon-13 stable isotope ratios in samples of ground water from the Upper Floridan aquifer, peninsular Florida. [larger version]

The distribution of carbon-14 and carbon-13 in the brackish ground water of the Upper Floridan aquifer is apparently very complex and probably is more affected by dissolution-precipitation reactions than is the distribution in the Lower Floridan aquifer (saltwater). The distribution of carbon-14 activity (corrected for isotopic fractionation) and carbon-13 stable isotope ratios (¹³C) in the upper part (from about 900 to 1,200 ft deep) in southern Florida is shown in figure 11; also shown is the distribution in central Florida, from Hanshaw and others (1965) and Plummer (1977). These isotope values are plotted on a predevelopment potentiometric surface map of the Upper Floridan aquifer to show the relation with the Upper Floridan flow system. The ¹³C values generally increase southward, from -11.4 at Polk City (site A) to -2.4 at sites 9 and 10 in southern Florida. The carbon-14 activity decreases generally southward from the recharge area near the Polk City high (site A) in central Florida (activity of 33.4 percent) until the activity becomes very small (virtually that of dead carbon-14) at site 5 east of Lake Okeechobee. Southward from Lake Okeechobee, however, the activity increases to about 6.5 percent at site 14 near Miami.

The cause of the apparent reversal in carbon-14 activity is unclear, but at least two hypotheses are possible. The first hypothesis requires the inland flow of seawater in the lower part of the Floridan aquifer system, as proposed by Kohout (1965) and supported by the isotope data in this report. This hypothesis is supported by the evidence that carbon-14 activity in water in the Upper Floridan aquifer decreases southward from central Florida. The reduction of carbon-14 activity is due to radioactive decay during the residential time and to dilution by inorganic carbon from limestone and dolomite in transit from central Florida toward the area of Lake Okeechobee. Beyond Lake Okeechobee, the southward-flowing water is blended with younger seawater from the Lower Floridan aquifer, and ground water in the Upper Floridan aquifer becomes brackish. As the ground water continues flowing generally southward toward the coastal areas, it becomes increasingly salty and contains greater amounts of carbon-14 as more seawater is contributed from the Lower Floridan aquifer. Thus, the carbon-14 activities in ground water decrease from central Florida to the area around Lake Okeechobee and then increase again southward from the lake. This hypothesis also is supported by the vertical distribution of carbon-14 in the Alligator Alley test well.

The second hypothesis assumes that horizontal and vertical permeability in the system is highly variable, and that velocity differences are the chief cause of the apparent reversal in carbon-14 activity. If the permeability is assumed to be greatest along a flow line roughly from Polk City in central Florida to site 14 in southern Florida, then the time of travel along that line is assumed to be the least. Rates of ground-water flow in central Florida estimated by Hanshaw and others (1965) and modified by Plummer (1977) ranged from about 6 to 32 feet per year (ft/yr); therefore, the transit time from the central Florida recharge area at Polk City (33.4 percent of modern carbon) to the Alligator Alley test well (site 10) in southern Florida (4.5 percent of modern carbon), a distance of about 150 mi, could range from 132,000 to 24,750 yr. On the basis of the measured activity at Polk City and a rate of 32 ft/yr (the highest rate), the carbon-14 activity at site 10 (the Alligator Alley test well) in southern Florida would be about 1.5 percent of modern carbon, compared with the measured value of 4.5 percent. Also, the analysis does not consider the chemical reactions that would have occurred in transit and reduced the concentration of carbon-14 by dilution to less than 1.5 percent. The measured carbon-14 activity at site 10 (4.5 percent) is probably lower than the actual activity as a result of dilution while in transit. The results of this sample analysis suggest that either the velocity greatly exceeds 32 ft/yr or significant amounts of carbon-14 enter the flow system downgradient from Polk City and closer to site 10. Possible sources of closer freshwater recharge are the sinkhole lakes about 80 mi north-northwest of site 10 in southern Highlands County (Kohout and Meyer, 1959). By comparison, the first hypothesis seems to be the most plausible because the second hypothesis needs two constrained assumptions. However, more data are required to test these hypotheses.

The distribution of ¹³C values suggests that dissolved carbon-13 in the ground water of the Upper Floridan aquifer increases (values become more positive) southward from the central recharge area to near Lake Okeechobee, where the values approximate the value for the native rock (dolostone). Hanshaw and Back (1971) reported values of -3.1 for dolomite and -0.4 for calcite in samples of Tertiary carbonate rocks from a well near the western shore of Lake Okeechobee. The relatively small variation in ¹³C values in southern Florida suggests that locally the water is saturated with respect to calcite and dolomite and that in southern Florida, the net effect of dissolution-precipitation reactions on the already small carbon-14 activity probably is insignificant. The southward increase in ¹³C (carbon-13 enrichment) from central Florida to Lake Okeechobee would result in substantial dilution of the dissolved carbon-14 by inorganic carbon from the limestone and dolostone. The net effect of the additional carbon would result in relatively less carbon-14 activity, and the age of the water samples would appear to be greater.

Attempts to adjust the measured carbon-14 activities in the samples from southern Florida, using the chemical dilution and isotopic dilution methods suggested by Mook (1980, p. 58), yielded adjusted ages that seem to be inconsistent with the time-of-travel estimates. No attempts were made to use mass-balance and mass transfer calculation schemes, as suggested by Plummer (1977) and Wigley and others (1978), to adjust the measured carbon-14 activity or ¹³C values because existing data are insufficient to accurately describe the hydrochemistry of the flow system.

The carbon-14 activities, presented by percent of modern carbon (PMC), of samples of salty (salinity comparable to seawater) ground water from the Lower Floridan aquifer ranged from less than 0.69 to 62.9 percent (table 4). Of the eight samples, only four are considered representative of the native ground water. Apparently, four samples were contaminated with salty drilling fluid. At site 10, for example, three samples from the Lower Floridan aquifer had activities of about 29.0, 7.3, and 37.1 percent. The sample of October 19, 1981, which had an unusually high chloride concentration (22,700 mg/L) and the least carbon-14, probably contained residual, artificially produced, salty drilling fluid. The artificial drilling fluid was chiefly artesian water from the Upper Floridan aquifer and sodium chloride.

Likewise, three samples collected at site 9 yielded activities of about 0.69, 1.12, and 62.9 percent. The samples of October 21, 1981, and December 16, 1982, which had low carbon-14 activities, were from an isolated zone in a deep monitor well (G-2331) that apparently had not been completely flushed of artificially produced drilling fluid. However, the sample of April 28, 1983, which had the highest carbon-14 activity, was pumped at a high rate (10,000 gal/min) from a nearby municipal wastewater disposal well (G-2333) and is, therefore, considered representative of the native fluid (a conclusion that is supported by the concentrations of uranium isotopes).

The normalized carbon-14 activities (PMC) and the carbon-13/carbon-12 stable isotope ratios (¹³C) for salty ground water from the Lower Floridan aquifer at sites 5, 9, 10, and 14 (table 4) are shown with the estimated values for modern seawater in figure 12. The carbon-14 activity in the salty ground water generally decreased radially inland from site 9 at Fort Lauderdale, whereas the stable isotope ratio (¹³C) decreased generally southward from site 5. The apparent ages of the four samples are shown in figure 13 with the differences in age relative to that at site 9 (Fort Lauderdale). Because the deep ground-water flow system is virtually closed to atmospheric carbon, except in the Straits of Florida where exchange with modern seawater is inferred, the relative age difference probably indicates depletion of carbon-14 primarily by radioactive decay (aging) rather than dilution of the carbon-14 content in the water by dissolution-exchange processes. The apparent southward depletion of carbon-13 is probably related to reactions with dolostone, but the relative effect on the activities is insignificant.

The apparent age of the saltwater in the Lower Floridan aquifer increases radially inland from site 9, suggesting that seawater enters the aquifer chiefly through submerged karst in the Straits of Florida, east of Fort Lauderdale, and then travels inland through highly permeable dolostones (chiefly the Boulder Zone). The rate of movement between sites 9 and 10 is about 44.5 mi in about 4,300 yr, or about 54.6 ft/yr, an average rate that greatly exceeds Hanshaw and others' (1965) estimated range in average rates (7 to 39 ft/yr) for the Upper Floridan aquifer in central Florida. Most of this difference is probably due to the extremely high transmissivity of the Lower Floridan aquifer (Boulder Zone).

The rate of ground-water movement between the subsea outcrop (karst of the Miami Terrace) in the Straits of Florida and site 9 at Fort Lauderdale is about 10.5 mi in about 3,200 yr, or about 17.3 ft/yr; the rate between the subsea outcrop and site 10 is about 55.0 mi in about 7,500 yr, or about 38.7 ft/yr. The relative increase in rates of movement with distance inland from the source suggests that inland movement of seawater through the highly permeable zones of the Lower Floridan aquifer could be related in part to the rapid rise of sea level during the Holocene transgression. The effects of changing sea level on ground-water movement and hydraulic gradients in the Lower Floridan aquifer are discussed in a later section of this report.