publications > paper > PP 1403-G > ground-water movement > movement based on natural isotopes > uranium isotopes
Hydrogeology, Ground-Water Movement, and Subsurface Storage in the Floridan Aquifer System in Southern FloridaBy Frederick W. Meyer
Ground-Water Movement in the Floridan Aquifer System in Southern Florida: Ground-Water Movement Based on Natural Isotopes as TracersUranium IsotopesDissolved uranium (U) isotopes have been used with varying degrees of success to relate samples of ground water to aquifers and to deduce ground-water flow patterns. Dissolved uranium in ground water from the Floridan aquifer system ranges widely, from several micrograms per liter (µg/L) to a trace. Generally, uranium concentrations are relatively high in areas where the aquifer is unconfined (oxidizing environment) and relatively low in areas where it is confined (reducing environment). Also, concentrations are relatively high in areas where the ground water represents a mixture of seawater, and anomalously high concentrations of dissolved uranium have been found in dolomitized parts of the aquifer system. Uranium concentrations in the ground water of the karst in west-central Florida (oxidizing environment) are generally one or two orders of magnitude higher than concentrations in the ground water of central and southern Florida (reducing environment). Concentrations are relatively high in oxidizing environments (unconfined aquifer) because the uranyl ion (U02+2) forms strong bicarbonate, tricarbonate, and phosphate complexes, and concentrations in reducing environments (confined aquifer) are relatively low because the complex uranium ions are either precipitated or adsorbed. The relative abundance of 234U (uranium-234) to 238U (uranium-238) in the ground water is attributed to the selected accumulation of 234U in the liquid phase by either selective leaching of 234U or by direct-recoil transfer of 234Th (thorium-234), which is a 234U precursor (Cowart, 1978, p. 713). The relative deficiency of 234U to 238U in water is attributed to solution of rocks that were already deficient in 234U, a result of the above-mentioned mechanisms.
The uranium concentrations and AR values for 62 samples of ground water from the Floridan aquifer system are presented with the corresponding worldwide average values for seawater (Ku and others, 1974) in table 5. Eight of the samples represent salty, seawater-like ground water from the Lower Floridan aquifer (chiefly the Boulder Zone), and 54 samples represent ground water ranging in salinity from that of freshwater to that of seawater from the Upper Floridan aquifer. The data are shown graphically in figure 14 by geographic location and hydrogeologic environment (that is, an unconfined aquifer generally represents an oxidizing environment and a confined aquifer generally represents a reducing environment). The uranium concentrations for the 62 samples ranged from about 7.97 µg/L in sample 3605 (table 5) to 0.003 µg/L in sample 525, and the AR values ranged from 3.9 in sample 13 to about 0.46 in sample 3773. For the 62 samples, the average AR value is 1.40 and the average uranium concentration is 0.653 µg/L. The geographic-hydrologic distribution of the data (fig. 14) suggests that (1) a general relation exists between samples of seawater-like ground water from the confined Lower Floridan aquifer (Boulder Zone) in southeastern Florida, samples of fresh ground water from the unconfined Upper Floridan aquifer in west-central Florida, and samples of brackish ground water from the confined Upper Floridan aquifer in distal southeastern Florida, and (2) a general relation exists between samples of fresh ground water from the confined Upper Floridan aquifer in central Florida and samples of brackish ground water from the confined Upper Floridan aquifer in southern Florida. The first relation suggests that the shared characteristics of the samples are indicative of their close proximity to sources of recharge (that is, direct infiltration of freshwater in the karst of west-central Florida and direct infiltration of seawater in the Upper and Lower Floridan aquifers along the southeastern coast of Florida). The second relation suggests that the shared characteristics of samples from central and southern Florida result from transit of the water from the distant recharge areas (that is, the uranium concentration decreases with increased distance downgradient, but the AR increases because of the alpha-recoil mechanism).
Upper Floridan Aquifer Uranium isotope analyses of 54 samples of ground water from the Upper Floridan aquifer (table 5, fig. 14) were used to deduce ground-water movement from the recharge areas in central Florida and the discharge areas in central and southern Florida. The salinities of the samples ranged from that of freshwater to that of seawater. Ten of the samples were from large freshwater springs in central Florida, and 44 were from deep wells. Of the 44 samples from deep wells, 4 were from the outcrop area (oxidizing environment) for the Upper Floridan aquifer in western and central Florida. Of the 10 springs sampled, 7 were from the outcrop area (oxidizing environment) and 3 were from downgradient areas where the aquifer is confined (reducing environment). Osmond and others (1974) used isotope dilution techniques to estimate ground-water contributions to Rainbow Springs (sample 91) and Silver Springs (sample 99) based on areal variations in uranium concentration and AR values. The distribution of dissolved uranium and AR values by region and hydrogeologic environment is summarized in table 6. In west-central Florida, the area where the aquifer is unconfined, the uranium concentrations in 11 samples of ground water ranged from 0.10 to 1.91 µg/L, averaging 0.75 µg/L; the AR values ranged from 0.72 to 1.47, averaging 0.98. In east-central and south-central Florida, generally the area south and east of the outcrop area to the northern half of Lake Okeechobee, the uranium concentrations in 24 samples of ground water ranged from 0.004 to 0.41 µg/L, averaging 0.080 µg/L; the AR values ranged from 0.94 to 3.9, averaging 1.83. In southern Florida, generally the area from the southern half of Lake Okeechobee to distal southeastern Florida, the uranium concentrations in seven samples ranged from 0.003 to 0.072 µg/L, averaging 0.048 µg/L; the AR values ranged from 0.66 to 1.94, averaging 1.24. In distal southeastern Florida (Dade and Monroe Counties), the uranium concentrations in 12 samples ranged from 0.034 to 4.70 µg/L, averaging 0.568 µg/L; the AR values ranged from 0.46 to 2.12, averaging 1.11 (table 6).
The general trend indicated by the average values of dissolved uranium concentration is decreasing concentration from the outcrop area (unconfined aquifer, oxidizing environment) southward to central and southern Florida (confined aquifer, reducing environment), then increasing concentration near the coastline in distal southeastern Florida. The general trend in average AR values is increasing AR values (234U increasing relative to 238U) from the outcrop area southward into central Florida (confined aquifer, reducing environment), then decreasing values from central Florida into distal southeastern Florida.
The areal distribution of AR values (fig. 16) indicates a general trend (with some exceptions) of increasing AR values eastward and southward from the outcrop area in west-central Florida. The values substantially increase as flow conditions change from unconfined (oxidizing) to confined (reducing). Of the 11 samples in the outcrop area (oxidizing environment), 9 have AR values ranging from 0.7 to 1.0 (zone A) and 2 (samples 95 and 151) have values that are slightly greater than 1.0 (zone B). Seven of the nine samples in zone A are from large freshwater springs that rapidly drain relatively large areas of the karst, and two (samples 96 and 106) are from wells tapping deeper parts of the Upper Floridan aquifer. The range in AR values for the springs suggests that groundwater contributions to the springs are chiefly from zone A (AR<1), but that some ground water is contributed from zone B (AR>1), or that the ground water has been in transit long enough to be affected by selective leaching, or the alpha-recoil mechanism. For example, contributions of ground water from the northern part of the outcrop area (samples 95 and 151) apparently have significant effects on the flow from Rainbow Springs (sample 91) and Silver Springs (sample 99). Ground water in the major part of the peninsula is confined, and AR values generally range from about 1.1 to 2.0 (zone B), with locally higher (zone C) and lower (zone D) values. Eastward and downgradient from the outcrop area in zone B are three freshwater springs (samples 1120, 1123, and 1124) with AR values ranging from 1.19 to 1.46, which suggest slow transit of the ground water from the distant outcrop area. Southward and downgradient from the outcrop area, the AR values vary widely, but values along the southward-plunging nose of the Polk City potentiometric surface high are relatively lower as a result of local recharge from sinkhole lakes on the high sand ridge, which forms the major axis of the present-day peninsula to a point about 25 mi northwest of Lake Okeechobee. Coincident with the sand ridge is the ground-water divide that extends to the southern terminus of the peninsula and separates flow to the east and to the west. In central Florida, there is subtle but reasonable evidence that AR values increase downgradient (east and west) from this ground-water divide. West and east of the divide in central Florida, AR values range from about 2.1 to 3.0 (zone C). The highest values chiefly are in the Peace River valley, west of the divide, extending southward about 70 mi from the Polk City high. Samples in this area were obtained from wells that are progressively deeper downdip and southward from the outcrop area. The highest AR value (3.9) was determined for a sample of water (sample 13) from a 1,103-ft-deep municipal supply well cased to 404 ft for the city of Wauchula. At Zolfo Springs, about 5 mi south of Wauchula, a sample of water (sample 699) from a 1,002-ft-deep municipal supply well cased to 350 ft yielded an AR value of 2.92. At Arcadia, about 22 mi south of Wauchula, a sample of water (sample 703) from a 1,410-ft-deep supply well cased to 900 ft yielded an AR value of 1.61. AR values for the three samples decrease southward (downdip), and the lowest value (1.61) is from the well that is both cased and drilled the deepest. The variations in AR values in the Peace River valley probably are due to vertical variations in water-bearing characteristics (permeability and porosity) and to variations in well construction. The relation between depth and AR value implies that the higher AR values in the upper Peace River valley could be associated with wells that receive significant contributions of ground water from the top of the aquifer. The higher AR values could be related to contributions of water from overlying confining beds (hence, longer transit times) by downward leakage during stress from pumping. A similar explanation can probably apply to the high AR value (2.75) in sample 727 in the easterly flow regimen. In addition to the above, high AR values were determined for samples from a deep monitor well at McKay Creek (sample 654) near St. Petersburg and a deep supply well on Big Pine Island (sample 448) near Fort Myers. The McKay Creek sample was similar in composition to seawater, and the high AR value (2.06) would be related to geochemical mechanisms during inland transit of seawater (coastal seawater intrusion). The Big Pine Island sample was brackish water (the salinity equivalent to about 4 percent of seawater), and the AR value (2.98) could be related to long-term seaward flow of freshwater from the principal recharge areas in central Florida. AR values ranging from 0.46 to about 1.0 (zone D) were determined for 11 samples from three areas in central and southern Florida where the aquifer is confined. As previously stated, values of about 1.0 or less are mostly found in unconfined oxidizing environments; therefore, the occurrence of values of less than unity in the confined reducing environment is considered to be anomalous. Samples 481 and 708 near the Peace River estuary have identical AR values (0.94) although the sources are vastly different. Sample 481 is from a 1,640-ft-deep flowing well cased to 648 ft producing saltwater (chloride concentration of about 18,400 mg/L) at 96.0 °F. Sample 708 is from a flowing 917-ft-deep well cased to 595 ft producing brackish water (chloride concentration of about 380 mg/L) at 80.6 °F. The similarity in AR values led Osmond and Cowart (1977, p. 145) to conclude that both samples are related to local upwelling of deep, warm saltwater; however, the salinity and temperature data for sample 708 do not support that conclusion. Sample 481 probably represents a blend of saltwater (seawater-like salinity) from the bottom of the well with fresher water from higher water-bearing zones. The AR value is lower than that for modern seawater (about 1.14) and much lower than that for seawater that had traveled significantly far (for comparison see section of report concerning AR values in the Lower Floridan aquifer). The anomalously high temperature of sample 481 (96.0 °F) is undoubtedly related to geothermal heating. The temperature of the saltwater in the underlying Boulder Zone of the Lower Floridan aquifer at the well site (sample 481) is about 110 °F (see fig. 7). Cowart and others (1978, p. 169) hypothesized that the low AR value for sample 457 south of Lake Okeechobee may be a relic of the late Pleistocene Epoch, when oxidized waters (AR1.0) were transported farther downgradient through discrete solution features (zones of high transmissivity) during low stands of the sea. Samples 524 and 3772 may have similar origins. The anomalously low AR values in distal southeastern Florida (samples 455, 512, 505, 2183, 3773, and 3774) may have origins similar to that of sample 457. However, an alternate explanation was proposed by Cowart and others (1978), one that involves dissolution of previously precipitated uranium that was originally deficient in 234U. The close proximity of the wells in distal southern Florida to the ocean suggests that the low AR values there are related to geochemical reactions that involve seawater. According to J.B. Cowart (Florida State University, written commun., 1984), the mechanisms hypothesized to be responsible for uranium isotope disequilibrium all assume (on the basis of extensive radiochemical data) that 234U is likely to be more mobile (and, thus, end up in the liquid phase) than 238U. The reason for the increased mobility of 234U is associated with the radiogenic origin of the isotope. The formation of 234U results from the radioactive decay of 238U so that (1)234U is likely to be situated in a lattice-damaged site (the damage resulting from its formation from 234Th and 234Pa (palladium-234)), or (2) the 234U (or its precursors 234Th and 234Pa) is directly propelled into the liquid phase surrounding the solid phase by direct alpha recoil. A 234U atom in a lattice-damaged site is less tightly bound to the solid and is, therefore, more likely to be leached from the solid than is the 238U situated in an undamaged site. There is no known mechanism by which elemental 238U is more mobile than elemental 234U. The fact that there are waters in a few places that have a deficiency of 234U relative to 238U can be explained using the previously discussed isotope fractionation mechanisms. If the liquid phase is enriched in 238U, the source of the uranium (the solid) must be depleted in 234U relative to 238U. If the environmental geochemical conditions change such that all of the uranium in the solid phase (or all of the uranium in part of the solid phase) is mobilized, the uranium entrained will be deficient in 234U. Usually, such profound changes in the geochemical environment are associated with near-surface locations (water-table changes and erosion)-hence the "oxidized unconfined" location of most samples in which the AR is less than 1. AR's of less than 1 in deep samples pose a different problem. One possibility is that the water was transported rapidly to a deep and distant location by way of conduits so that the water retains aspects of its "oxidized unconfined" origin. The other possibility is that changes of a rather profound type have occurred in the deep environment. Cowart and others (1978) suggested that the AR's of less than 1 in the Upper Floridan aquifer of southeastern Florida (samples 455, 457, 505, 512, and 524) resulted from rapid movement of water from the recharge area to south Florida. Since that publication, other samples from the Upper Floridan aquifer in the area have been analyzed by Cowart (samples 2183 and 3771 through 3774). Of the samples more recently collected, two pairs representing vertical sets of water from the Upper Floridan aquifer have been analyzed (samples 3771 and 3772 from the Alligator Alley test well and samples 3773 and 3774 from Fort Lauderdale's Port Everglades wastewater treatment plant monitor wells). Of the four samples, three have AR's of less than 1. The Alligator Alley test well samples have AR's of 1.73±0.15 and 0.92±0.13 for depths of 811 to 816ft and 895 to 1,150 ft, respectively. The Fort Lauderdale samples have AR's of 0.46±0.03 and 0.78±0.02 for depths of 1,021 to 1,072 ft and 1,466 to 1,562 ft, respectively. The uranium concentrations for the Fort Lauderdale samples are 0.305±0.014 and 4.70±0.30 µg/L, respectively. The magnitude of the disparity between the samples in the Fort Lauderdale well strongly suggests that the source of the <1-AR uranium is relatively nearby. Furthermore, the dilution or mixing of the higher concentration sample (sample 3774) with any water having an AR of 0.46 or more to produce the uranium concentration and isotope ratio of sample 3773 is not possible. It is not known whether the other samples in southern Florida having AR's of less than 1 result from relict circulation or from changes in environment; it is possible they represent some combination. Interestingly, the samples from the Upper Floridan aquifer in southern Florida that have AR's of less than 1 and the highest uranium concentrations (samples 455, 505, 512, 3773, and 3774) are located near the coast. A scenario consistent with the results is one in which uranium has been precipitated at a geochemical barrier, possibly a redox barrier, the source of the dissolved uranium being seawater. With changes in sea level and hydraulic head, the precipitated uranium was isolated in a reducing environment, wherein by the process of alpha recoil it became depleted in 234U. With the advent of geochemical-redox changes, the precipitated uranium became mobile (thus, the relatively high concentration), and it was, of course, deficient in 234U (AR of less than 1). Such a scenario is unlikely for the more inland samples (samples 481, 708, 457, and 3772) that have AR's of less than 1. The explanation for the apparent AR anomalies are, therefore, many and diverse and could involve several processes. Dolomitization, for example, may be important in the selective precipitation of elemental uranium. An investigation of high gamma-ray emissions from dolostone in the Lower Floridan aquifer at Stuart (fig. 2, site 6) in 1974 by borehole spectrometry indicated that uranium was the probable source (W.S. Keyes, U.S. Geological Survey, written commun., 1974). The low AR anomaly in distal southeastern Florida could be linked to the mixing of magnesium-rich seawater with brackish bicarbonate-type ground water (Hanshaw and Back, 1971) to yield dolomite that perhaps has excess 234U. Lowered temperatures (less than 77 °F) along the southeastern coast may enhance the dolomitization process. In other areas, upwelling of warm, magnesium-rich, relatively young seawater from the Boulder Zone may also produce uranium-rich dolomite and AR anomalies. Also, some of the AR anomalies may be due to experimental error since the uranium concentration in the brackish artesian water of southern Florida is generally very low. The trends in dissolved uranium concentration and AR values suggest that flow originated from the outcrop area (figs. 15, 16); however, predevelopment heads in the Upper Floridan aquifer, as indicated by the potentiometric surface map, suggest that flow in the outcrop area is chiefly westward toward the Gulf of Mexico. This discrepancy could be related to changes in flow patterns as a result of sea-level fluctuations, and perhaps climatic changes, during the Holocene transgression. Plummer (1977, p. 811) estimated that the ground water at Polk City (the highest heads) -the northernmost source of flow to central and southern Florida-is about 3,200 to 8,000 yr old (water samples contain about 67 percent of modern carbon) and that the average southward velocity was about 32 ft/yr. Therefore, the present-day concentration gradients are perhaps, to a large degree, a relic of antecedent flow patterns when the principal recharge area was perhaps 20 mi north of the Polk City potentiometric surface high. To date, the uranium isotope data for the Upper Floridan aquifer are meager and present a complex picture of the regional flow system. Data from discrete water-bearing zones in the Upper Floridan aquifer are necessary prerequisites for correct interpretation, and further investigation is needed into the possible causes of the apparent anomalies. Lower Floridan Aquifer Uranium isotope analyses of eight samples of salty ground water from the Lower Floridan aquifer (chiefly the Boulder Zone) in southeastern Florida were used to evaluate the potential for inland circulation of seawater from the Straits of Florida (tables 4, 5; fig. 14). A sample from a deep monitor well at site 9 (well G-2331) was not included because the sample was contaminated with drilling fluid. The uranium concentration in the eight samples ranged from 0.45 µg/L in sample 520 to 7.97 µg/L in sample 3605, averaging 2.90 µg/L. The AR values ranged from 1.14 to 1.50, averaging 1.24. The uranium concentrations in six samples are less than that of worldwide seawater (U=3.30±0.14 µg/L). The anomalously high uranium concentration (7.97 µg/L) in sample 3605 is unexplained. The areal distribution of uranium (fig. 17) shows that concentrations generally decrease radially inland from site 9 in the Fort Lauderdale area, where the concentration compares best to that of seawater. The inland concentration gradient suggests that uranium is either precipitated or adsorbed during transit from a source in the Straits of Florida. A similar flow was indicated by the radiocarbon data. The AR values for the samples from the Lower Floridan aquifer generally increase radially inland from site 9 in the Fort Lauderdale area, where the AR value is identical to that in seawater (fig. 17). The inverse gradient suggests that 234U increases relative to 238U, probably by the alpha-recoil phenomenon during transit inland of seawater from a source area in the Straits of Florida, east of Fort Lauderdale. Therefore, the increase in AR values is an indication of time in transit through the Boulder Zone, and the AR data correspond with both the uranium concentration data and the radiocarbon data.
Previous: < Carbon Isotopes | Next: Flow Patterns > |
U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/papers/pp1403g/uranium.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 06 January, 2005 @ 02:05 PM(TJE)