<%BANNER%>

Using Radon-222 as a tracer of mixing between surface and ground water in the Santa Fe River Sink/Rise System

University of Florida Institutional Repository

PAGE 1

USING RADON-222 AS A TRACER OF MIXING BETWEEN SURFACE AND GROUND WATER IN THE SANTA FE RIVER SINK/RISE SYSTEM By LAUREN A. SMITH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Lauren A. Smith

PAGE 3

iii ACKNOWLEDGMENTS I wish to express my appreciation to my advisor, Dr. Jonathan Martin, for his constant patience and enthusiasm. I wish to express my gratitude to my committee members, Dr. Elizabeth Screaton and Dr. Philip Neuhoff, both of whom are wonderful mentors and friends. I wish to thank all of the employees at O’Leno State Park for their cooperation. A special thanks goes to Dr. Jaye Cable, for her assistance; to Dr. Jason Curtis, for his support and never-ending supply of liquid nitrogen; and to Kevin Hartl, for construction of a beautiful radon-extraction line. My thanks go to those who supported me through their friendship and love, especially my family. Last but certainly not least, I would like to give the greatest thanks to Brooke Sprouse and Jennifer Martin: without their support, patience, and laughter in the field, the ticks and snakes would have gotten to me.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS..................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES...........................................................................................................vii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION.......................................................................................................1 Study Area....................................................................................................................4 Location and Climate............................................................................................4 Physiography........................................................................................................6 Geology and Hydrostratigraphy............................................................................6 Background 222Rn Studies............................................................................................8 Background O’Leno Studies........................................................................................9 2 METHODS................................................................................................................14 Field Methods............................................................................................................14 Sample Bottle Construction................................................................................14 Water Sampling..................................................................................................15 Ground Water Sampling.....................................................................................17 Lab Methods..............................................................................................................18 Absorbance.................................................................................................................21 3 RESULTS..................................................................................................................22 Precipitation and River Stage.....................................................................................22 Depth Profiles............................................................................................................22 Excess 222Rn, Precipitation, and River Stage.............................................................28 Sample Periods I to VI........................................................................................28 Well Samples......................................................................................................30 Mixing Model............................................................................................................31 Decay Equation..........................................................................................................32 Sink vs. Rise.......................................................................................................32 Sink vs. Karst Windows......................................................................................34

PAGE 5

v Gas Exchange Equation.............................................................................................36 Color Absorbance vs. Activity...................................................................................36 4 DISCUSSION............................................................................................................40 Precipitation and River Stage.....................................................................................40 Depth Profiles............................................................................................................40 Excess 222Rn...............................................................................................................41 Radioactive Decay Along the Flow Path...................................................................46 Distribution of 222Rn in the Groundwater..................................................................48 5 CONCLUSIONS........................................................................................................51 APPENDIX MEASURED AND CALCULATED 222RN, TOTAL 226RA, TRAVEL TIME, AND EVASION FOR ALL SITES...............................................................53 REFERENCES..................................................................................................................59 BIOGRAPHICAL SKETCH............................................................................................63

PAGE 6

vi LIST OF TABLES Table page 1-1 Geologic and hydrogeologic units found in the Santa Fe River Basin....................10 2-1 Dates, river stages, and sample sites of each sample event.....................................16 2-2 Drilling records of six wells drilled during 2003 ...................................................17 2-3 Average percent efficiency and standard deviation for each Lucas cell.................21 3-1 Average, standard deviation, and coefficient of variation for each sample time....30 3-2 Relationship between measured 222Rn-activity and travel time and calculated 222Rn-activity and travel time using the decay equation..........................................34

PAGE 7

vii LIST OF FIGURES Figure page 1-1 Map of the Santa Fe River basin, shown in shaded area...........................................5 1-2 Site map of the Santa Fe River Sink/Rise system in O’Leno State Park...................7 2-1 Schematic of the sample bottles used for collection of water for 222Rn analysis....15 2-2 A simplified schematic of the Radon Extraction Line............................................19 2-3 A simplified schematic of the radon counting system modified from Operations Manual 1012899A...................................................................................................20 3-1 The precipitation record for O'Leno State Park from June 2001 to July 2003, which encompasses the entire study period.............................................................24 3-2 The Santa Fe River Stage recorded from O’Leno State Park from mid April 2001 till the end of June 2003.................................................................................25 3-3 Vertial depth profiles at the River Rise...................................................................26 3-4 Vertical depth profiles at Vinzants Landing............................................................27 3-5 Radon-222 activities of samples from karst windows, the River Sink, and the River Rise versus distance along the flow path for sample dates of A. 5/8/02 and 9/12/02, B. 1/30/03, and C. 2/19/03, 2/26/03, and 3/5/03......................................29 3-6 Radon-222 activities of wells 1, 2, 3, 4, 6, and 7....................................................31 3-7 Percent X (%X) versus Stage (m) for each site in the study area............................32 3-8 The gain or loss in activity downstream from the River Sink.................................35 3-9 Radon-222 activities of the River Sink at various sample periods vs. Absorbance..............................................................................................................38 3-10 Radon-222 activities of samples taken at Sweetwater Lake, Jim Sink, Ogden Pond, the River Sink, and Hawg Sink on 3/5/03.....................................................38 3-11 Absorbance values taken during the study..............................................................39

PAGE 8

viii 4-1 A conceptual model of interactions between the Floridan Aquifer and water in conduits during the sample periods.........................................................................43 4-2 Evasion at the River Rise increases as travel time increases...................................44

PAGE 9

ix Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science USING RADON-222 AS A TRACER OF MIXING BETWEEN SURFACE AND GROUND WATER IN THE SANTA FE RIVER SINK/RISE SYSTEM By Lauren A. Smith May 2004 Chair: Dr. Jonathan B. Martin Major Department: Geological Sciences Karst aquifers (which provide 25% of the world’s population with potable water) are characterized by links between surface and ground water, making them susceptible to contamination. Linkage between groundwater and surface water occurs through the unconfined Floridan Aquifer in north-central Florida. An example of surfacegroundwater linkage occurs when the Santa Fe River flows into a 32-m deep sinkhole, the River Sink, and resurges approximately 8 km down gradient at the River Rise. This system provides opportunities to observe mixing between surface and ground water within a karst aquifer. Mixing can be observed using natural chemical and isotopic tracers, one of which is the isotope 222Rn (half-life = 3.82 days). Radon-222 is formed by the alpha decay of 226Ra, an alkaline earth element common in carbonate and clay minerals. Surface waters have low 222Rn activities (~20 dpm/L) due to atmospheric evasion and ground water have high 222Rn activities (>200 dpm/L). The difference in

PAGE 10

x activities of these two end members should make 222Rn a good tracer for surface-ground water mixing in karst areas. Activities of 222Rn were measured to depths of 8 and 10 m at two sinkholes (Vinzants Landing and River Rise) and indicated that 222Rn is heterogeneous in the sinkholes. Water samples were collected from the River Sink, River Rise, and karst windows between the Sink and Rise from May 2002 to April 2003. These samples encompass drought, base flow, and flood stages. As the river stage rises to flood conditions (~11.51 masl) the excess 222Rn activities decrease from approximately 100 dpm/L to less than 25 dpm/L, indicating that no matrix water enters the conduit; and that the conduit is filled with 222Rn-poor surface water from the Santa Fe River. During low stage (9.25 masl), when past studies have found groundwater dominates the system, the water activities are also low ( 25 dpm/L) possibly caused by evasion and radioactive decay of 222Rn as the water slowly flows through the Sink/Rise System. At times of base flow conditions for the river (10.70 m stage), activities are ~100 dpm/L, indicating some ground water and surface water mixing. The fraction of surface and groundwater in the system could not be determined through a two end-member mixing model because evasion and radioactive decay complicate the simple mixing model. 222Rn may be a good tracer of mixing between surface and ground water; however, further studies need to be conducted to fully understand the impacts of decay, evasion, and lithology over the 222Rn activities in the Sink/Rise system.

PAGE 11

1 CHAPTER 1 INTRODUCTION Karst aquifers provide 25% of the world’s population with potable water (Ford and Williams, 1989). These aquifers are typically characterized by direct links between the surface and ground water systems through sinkholes, swallets, and highly permeable rocks, which make karst groundwater resources susceptible to surface contamination (Ford and Williams, 1989). For example, tannins, pesticides, and nitrates from agricultural and animal byproducts have been shown to move rapidly into and through karst groundwater systems, contaminating water resources (Katz, 1992; Boyer and Pasquerell, 1996; Kincaid, 1998). The loss of undersaturated surface water with respect to calcite to the aquifer may also increase dissolution of carbonate rocks (Drever, 1988). Tracking contaminants can be difficult because the heterogeneous permeability of karstic rocks complicates modeling of fluid flow and contaminant transport. Complications are due to the exchange of water among three types of porosity: intergranular matrix porosity, fracture porosity, and conduits (White, 1999). Many studies of karst aquifers have been conducted, but most focus on aspects of either conduit flow or matrix flow. Highly altered karst aquifers, like those found in Paleozoic midcontinent carbonates are dominated by conduit and fracture flow because recrystallization has led to dense, low permeability matrix rocks. Less altered karst aquifers may have matrix rocks with high permeability, creating flow characteristics that differ from karstic aquifers formed in dense and altered carbonates with low permeability.

PAGE 12

2 An important example of a less altered karst system dominated by both conduit and matrix flow is the Floridan Aquifer system (Bush and Johnston, 1988), a regionally important aquifer because it is the main source of potable water in north central Florida. The Floridan Aquifer represents a class of karst aquifers with high intergranular porosity and permeability (Budd and Vacher, 2002), which may allow a significant fraction of the subsurface flow through matrix rocks (Martin and Screaton, 2001). Typical of all karstic aquifers, the Floridan Aquifer has direct links to surface water through sinkholes, swallets, and highly permeable rocks. Bush and Johnston (1988) suggested that on a regional scale, it could be assumed that the conduit and matrix systems in the Floridan Aquifer of the southeastern United States could be treated as homogenous or as an equivalent porous medium. At a local scale, even in porous karst aquifers such as the Floridan Aquifer, conduit flow predominates and the equivalent porous medium approach is invalid (Bush and Johnston, 1988; Padilla et al, 1994; Halihan et al, 1998). Prevention and remediation of contamination in these aquifers require an understanding of the mixing among surface and ground water and controls of flow paths in the subsurface. Many techniques have been used for studies of flow paths, mixing, and flow velocities through karst aquifers. Common techniques include injection of artificial dyes (e.g., rhodamine and fluorescein), studies of physical characteristics of the water, and measurement of chemical and isotopic compositions of the water and dissolved constituents (e.g., Smart, 1988; Ryan and Meiman, 1996; Martin and Dean, 1999). One dissolved radioactive component with good potential for studying the mixing between ground and surface water is the isotope 222Rn (half-life=3.82 days). Radon is a product of the natural radioactive decay series of 238U and a direct product of alpha disintegration of

PAGE 13

3 226Ra (Bertin and Bourg, 1994). As an alkaline earth element, radium is found in carbonate and clay minerals; and thus large amounts of radon are generated in ground water. However, because of atmospheric evasion (loss of radon gas to the atmosphere), radon activity should be low in surface waters. In addition, radon is chemically conservative (i.e., no source other than 226Ra and no sinks other than decay); and is easily measured at low activities (e.g., 0.02 pCi/L (0.044 dpm/L)) through alpha-counting techniques. Therefore, activities of 222Rn in karst water should reflect the relative amounts of water from the surface with low 222Rn activity and water that has been stored in matrix porosity and thus gained a high 222Rn activity. Rogers (1958) was the first to use 222Rn to study ground water and surface water relationships. He demonstrated that 222Rn activities in springs (ground water) were much higher than in surface water and that the springs were a source of 222Rn in the streams surrounding his study area. A study by Ellins et al. (1990) of Puerto Rican streams in an upland karst region confirmed Rogers’ (1958) work. Ellins et al. (1990) described low activities of 222Rn in the streams (18 to 200 dpm/L) as a function of the atmospheric evasion and the high levels of 222Rn (up to 693 dpm/L) in related springs due to the input of enriched groundwater. Bertin and Bourg (1994) were able to use the difference in 222Rn activities in ground and surface water to trace the seeping of Lot River water in France into an alluvial aquifer. Though Bertin and Bourg (1994) did not work in a karstic terrain, their study suggests that 222Rn can be used as a tracer of infiltrating river water and can be used in the presence of mixing within the aquifer. This study has two principle objectives: To evaluate the potential for using 222Rn in a karst environment for tracing groundwater and surface water interactions

PAGE 14

4 To define the interaction between surface water and ground water in a karst system with high matrix permeability using 222Rn. To satisfy these objectives, several questions concerning 222Rn activities in a karstic terrain were addressed: Does 222Rn trace surface water and ground water mixing in a karst region? Will 222Rn trace mixing of the water from the matrix and conduit? How can the interactions between surface and ground water be quantified? Will lithology prove to be a main control over 222Rn activities? Study Area Location and Climate The Santa Fe River Basin (Figure 1-1) covers an area of 3584.5 km2 and is a tributary basin of the Suwannee River (Hunn & Slack, 1983). The Santa Fe River originates in Altho and Santa Fe lakes in north-central Florida (Hunn & Slack, 1983) and flows westward approximately 50 km until it reaches a 32 m deep sinkhole at O’Leno State Park. River water that enters the sinkhole mixes with groundwater from the Floridan Aquifer and travels through areas of mapped and unmapped conduits. This water resurfaces at the River Rise (~8 km down gradient of the Sink) and continues as the lower Santa Fe River (Hisert, 1994; Dean, 1999) (Figure 1-2). Located in a semi-tropical climate, average air temperature in the Santa Fe River Basin is 20 C and average groundwater temperature is around 22 C (Hisert, 1994; Florida Geological Survey, 1992). Precipitation averages 123 cm/yr with most occurring between June and September. Most summer rainfall occurs during local thunderstorms and seasonal tropical storms. As a result, precipitation can be heterogeneous over small areas. Winter rainfall is from extratropical storms, which causes most flooding.

PAGE 15

5 Figure 1-1. Map of the Santa Fe River basin, shown in shaded area. The location of O’Leno State Park is shown by the square and is seen in detail in Figure 1-2. Modified from Hunn, J. D., Slack, L. J., 1983. Water resources of the Santa Fe River Basin, Florida. U. S. Geological Survey. O’Leno State Par k

PAGE 16

6 Physiography Three dominant physiographic divisions in the region include the Northern or Proximal zone, the Central or Mid-peninsular zone and the Southern or Distal zone (White, 1970). O’Leno State Park is located on the border between the Northern and Central physiographic zones within the Western Valley, a 140 mile long lowland, and the High Springs Gap, an opening within the Western Valley. This portion of Florida is a well-drained area of high recharge, and variably developed karst. The High Springs Gap includes the Cody Scarp, which represents the erosional edge of the Hawthorn Group. This scarp is the most prominent topographic feature in peninsular Florida, yet it is a subtle feature (with ~25m per 10km) (Puri and Vernon, 1964; White, 1970). The Hawthorn Group is the confining unit for the Floridan Aquifer, and thus karst features are common where the Hawthorn Group is absent. Most streams, including the Santa Fe River, either disappear or become losing streams to the aquifer systems below as they cross the scarp. Geology and Hydrostratigraphy Three aquifer systems comprise the hydrostratigraphy of the study area and include the Surficial Aquifer system, Intermediate Aquifer system, and the Floridan Aquifer system. The connection between the stratigraphy and hydrostratigraphy is shown in Table 1-1. The Surficial Aquifer system provides small-yield domestic and agricultural water supplies (Scott, 1992) and is composed of undifferentiated Plio-Pleistocene sediments containing sinkhole fill, fine to medium sands, and layers of clay and silt (Hunn and Slack, 1983; Scott, 1992). In O’Leno State Park, the Surficial Aquifer system is less than 2 meters thick to absent (Hunn and Slack, 1983).

PAGE 17

7 Figure 1-2. Site map of the Santa Fe River Sink/Rise system in O’Leno State Park. The dashed line represents locations of caves mapped by cave divers (Old Bellamy Exploration Team, unpublished report). The solid line represents roads and the pentagons represent wells drilled in 2003. Modified from Hisert (1994); Dean (1999). Cody Scarp

PAGE 18

8 The Intermediate Aquifer is contained within the Hawthorn Group (Bush and Johnston, 1988). The Hawthorn Group in Northern Florida consists of interbedded phosphatic carbonates and siliciclastics with a trend of increasing siliciclastics in the younger sediments. These sediments have low permeability and form an effective aquiclude, the intermediate confining unit, which confines the Floridan Aquifer where present (Bush and Johnston, 1988). In areas where the intermediate confining unit is absent such as the western portion of the Santa Fe River Basin, the Surficial Aquifer directly overlies carbonates of the Floridan Aquifer (Scott, 1992). The lack of the confining unit limits lakes and wetlands in the surface (Scott, 1992). Karst features, such as sinkholes are common through both the Surficial and Intermediate Aquifer systems and provide direct recharge to the Floridan Aquifer System. The Floridan Aquifer System is composed of several hundreds of meters of limestone and dolostone and is the main source of water in northern Florida and parts of Georgia, South Carolina and Alabama (Hunn and Slack, 1983; Stingfield and LeGrand, 1966). The stratigraphic units comprising the Floridan Aquifer system from lateEocene to Oligocene include the Oldsmar Limestone, Avon Park Limestone, Ocala Limestone, and Suwannee Limestone. The Ocala Limestone is the uppermost stratigraphic unit in O’Leno State Park (Hisert, 1994). Background 222Rn Studies Two surface and groundwater mixing studies using 222Rn were conducted previously in portions of the Santa Fe River Basin. Ellins et al. (1992) used background 222Rn activities of the Santa Fe River (~10 dpm/L) and spring 222Rn activities in the lower Santa Fe River Basin (~1000 dpm/L) to observe a ratio of ground to river water 222Rn activities of 100:1. Hisert (1994) sampled 222Rn in the Santa Fe River along

PAGE 19

9 Hollingsworth Bluff, located southwest of O’Leno State Park. In samples collected at the riverbed, the depth halfway between the riverbed and surface, and the surface of the river, Hisert (1994) found that 222Rn activities decreased with depth and were dependent on groundwater influxes. Hisert (1994) also measured 222Rn activities within the O’Leno State Park. At an average discharge of 42 m3/s, 222Rn activities in water from selected karst windows in the northern section of the park (River Sink and Jim Sink) are identical to atmospheric background levels (< 50 dpm/L) (Hisert, 1994). At these discharges, activities in water from the karst windows in the southern section of the park (Two Hole, Sweetwater Lake, and River Rise) were close to ground water levels (~450 dpm/L). Hisert interpreted this increase in activity to result from an influx of groundwater into the southern karst windows. It is unclear as to whether the Hawthorn Group or the carbonates in the Floridan Aquifer is the major source of the 222Rn in the study. Smoak et al. (2000) found well water from the Floridan Aquifer near Tampa, FL to have 226Ra activities for the Floridan Aquifer to range from 5.8 to 6.6 dpm/L. These 226Ra activities are similar to those seen at wells in this study, which correspond to 222Rn activities of approximately 1000 dpm/L. Wherett (1992) found average soil gas 222Rn activities of the Hawthorn Group to be approximately 2000 dpm/L. Although it is unclear how average soil gas converts to the liquid gas used in this study, it can be assumed that the numbers are within the same order of magnitude. Background O’Leno Studies Skirvin (1962) conducted the first study of water flow through the Sink/Rise system of the Santa Fe River. Based on the tannic acid content of the water, Skirvin

PAGE 20

10Table 1-1. Geologic and hydrogeologic units found in the Santa Fe River Basin. Modified from Hunn and Slack (1983), Scott (19 92), Hisert (1994), and Dean (1999). Series Stratigraphic Unit Hydrogeologic Unit Lithologic Description Thickness (m) Holocene Undifferentiated Surficial Aquifer Sinkhole fill, fluvial terraces, and 0-25 Pleistocene Sediments thin surficial sand Pleistocene Alachua Formation Intermediate aquifer/ Reddish-white sands, with clays, 0-30 to Miocene Upper confining unit sandy clays, and phosphate pebbles Middle to Hawthorn Group Phosphatic clayey sand-sandy clay Lower Miocene with varying amounts of Fullers Earth and carbonate Oligocene Suwannee Limestone Floridan Very pale yellow, moderately indurated, 0-100 Aquifer porous, fossil-rich calcarenite Eocene Ocala Limestone Very permeable limestone, 275-300 Avon Park dolomitic limestone, and dolomite Limestone Oldsmar Limestone Paleocene Cedar Keys FormationSubFloridan confining unit Limestone, some evaporites and clay ?

PAGE 21

11 (1962) suggested that most water entering the River Sink discharged at the River Rise. Skirvin (1962) also observed that water discharged from the Rise and intermediate karst windows when water was blocked from the River Sink by a temporary dam. Skirvin found that more water discharged from the Rise than flowed into the Sink in May and November 1961. The overall discharge gain in May and November were 5.30 m3/s and 5.08 m3/s, respectively. These observations indicate that the system gains water from sources other than the River Sink. Hunn and Slack (1983) suggested that the water quality of the basin at this time was more influenced by natural factors, such as limestone dissolution, than by anthropogenic factors. Hunn and Slack (1983) describe the sinking eastern rivers as having lower concentrations of iron, calcium, magnesium, and bicarbonate than the water within the Floridan Aquifer. The rising western rivers have concentrations of iron, calcium, magnesium, and bicarbonate similar to the Floridan Aquifer. These chemical compositions indicate that the rivers contain water from the Floridan Aquifer, which is similar to the results reported by Skirvin (1962). Hisert (1994) provided the first in-depth look into the complex underground system of O’Leno State Park by using sulfur hexafluoride (SF6), temperature, and 18O to trace groundwater flow from the River Sink to the River Rise. The SF6 tracer experiment, conducted in July 1991, required two injections to link flow from the Sink to the Rise. The first injection, performed at a discharge of 31.3 m3/s, connected the River Sink to Sweetwater Lake through multiple karst windows, but SF6 from this injection was not recovered at the River Rise. A second injection into Sweetwater Lake was detected at the River Rise (Fig. 1-2). The discharge rate during this second injection was not reported,

PAGE 22

12 but a third injection was made at Jim Sink when River discharge was 5.8 m3/s into the River Sink, which connected Jim Sink to the River Rise via Sweetwater Lake. Even though a direct connection was never made between the River Sink and Rise, Hisert concluded that the karst windows in the Santa Fe River Sink/Rise system are hydraulically connected. Hisert was also able to use background temperatures of the river (15 C) and groundwater (22 C) to conclude that the 2.5 C increase in temperature found within some of the windows, including the Rise, indicated a 37.5% influx of groundwater. This conclusion was supported by variations in oxygen ratios of the water samples. The karst windows have a slightly enriched 18O signature compared to the River Rise reflecting an influx of 18O-enriched groundwater. Dean (1999) used water chemistry to determine the extent of mixing between the river and ground water, and he used temperature as a tracer of flow rate from the River Sink to the River Rise. Using natural variations in temperature, Dean (1999) found that subsurface travel time of the water from the River Sink to the River Rise varies from approximately 12 hours to nearly 8 days, depending on river stage. By using temperature as a tracer and discharge measurements, Ginn (2002) and Screaton et al. (2004) found a subsurface conduit flow at rates of 3.11 km/day, which are similar to the subsurface flow rate found in Hisert’s thesis. Both Ginn (2002) and Dean (1999) agree with Hisert’s conclusion that the sinks in O’Leno State Park are all hydraulically well connected. Chloride (Cl-) concentrations from Dean (1999) increased between the River Sink and Sweetwater Lake by an average of 35.8% during low flow conditions. This increase suggests an addition of Clrich water, probably groundwater. During high flow

PAGE 23

13 conditions the Clconcentrations between the Sink and Sweetwater Lake change by an average of 4.3%, which suggests very little addition of ground water.

PAGE 24

14 CHAPTER 2 METHODS The objectives of this project were met through a series of physical and chemical measurements including daily precipitation, river stage records, and measurements of isotopic compositions in water samples collected from various sites in the region of the Santa Fe River Sink/Rise system. Park staff collected daily precipitation and river stage measurements for the O’Leno State Park station on the Santa Fe River. New chemical analyses used in this thesis include measurement of 222Rn activities in all collected water samples and measurement of absorbance in selected samples. Field Methods Sample Bottle Construction Sample bottles were constructed from emptied, cleaned 2.5 L glass acid bottles, which were wrapped in duct tape to prevent breakage (Figure 2-1). A two-holed rubber stopper along with two lengths of 1/40-inch diameter copper pipes, one 2-inch long piece and the other 12-inches long were placed in the mouth of each bottle. The base of the 12-inch pipe was attached to an Aqua-Tech air diffuser. The rubber stopper was secured with thin wire and attached to the bottle with silicon to make a gas tight seal. Rubber tubing with male/female connectors was attached to the ends of the two copper pipes that extend from the rubber stopper. Clamps were attached to each rubber tube to control airflow. Each bottle was evacuated in the lab for approximately 5 min. before water sampling. The vacuum allows the water samples to be sucked into the bottle and prevents atmospheric contamination of the sample.

PAGE 25

15 Gas tight silicon seal Connector Clamp 2” copper tubing 12” copper tubing Air bubbler Duct Tape 2.5 L acid bottle Rubber Hose Figure 2-1. Schematic of the sample bottles used for collection of water for 222Rn analysis. Water Sampling Water was sampled at 11 sites along the river (Vinzants Landing, River Sink, Ogden Pond, Big Sink, Parener’s Sink, Jim Sink, Jug Lake, Hawg, Two Hole, Sweetwater Lake, and the River Rise) and at 5 wells (Wells 1, 2, 3, 4, 6, and 7) located within the park (Figure 1-2). Samples were collected between May 2002 and May 2003 at stages ranging from 9.78 to >11.51 masl (Table 2-1). One objective of this study was to develop simple and robust techniques for sampling and observing 222Rn activities in karst systems. Consequently, sampling techniques evolved as the study proceeded. Two types of samples were collected initially and include grab samples and vertical profile samples. Grab samples were taken approximately one foot from shore of the various sinks by submerging the tube with the male connector and loosening the clamp. Vertical profile samples were taken at the River

PAGE 26

16Table 2-1. Dates, river stages, and sample sites of each sample event. Sample Periods Sampling Dates River Stage (m)Grab or Peristaltic SamplesSample sites I 5/8/02 9.78 Grab Samples All 11 sites 6/6/02 9.61 Grab Samples Vertical profile of River Rise (1, 2, 4, 6, 8, 10 m depths) 7/3/02 9.56 Grab Samples Vertical profile of River Rise (0.3, 1, 2, 4, 6, 10 m) Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m) 8/29/02 10.12 Grab Samples Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m) II 9/13/02 10.16 Grab Samples All 11 sites III 1/31/03 10.53 Peristaltic Samples All 11 sites IV 2/19/03 11.49 Peristaltic Samples River Sink, Ogden, Jim, Hawg, Sweetwater, River Rise V 2/27/03 11.51 Peristaltic Samples River Sink, Ogden, Hawg, Sweetwater, River Rise VI 3/5/03 >11.51 Peristaltic Samples River Sink, Ogden, Jim, Hawg, Sweetwater, River Rise 4/30/03 10.56 Peristaltic Samples Wells 5/7/03 10.47 Peristaltic Samples Vertical profile of River Rise (0.3, 1, 2, 4, 6, 10 m) Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m)

PAGE 27

17 Rise and Vinzants Landing to depths of 10 and 8 meters, respectively. These samples were collected using a rubber raft and a rope to secure the raft position. Once in position, a tube was lowered to various depths, and water was siphoned to the tip of the tube. After the tube was filled with water, it was attached to the evacuated bottle, which was allowed to fill with sample water. Subsequent samples were collected from shore using a peristaltic pump (Geotech Geopump 2). This sample system included a tube that was extended up to 20 meters from shore. One end was connected to the pump and the other was weighted and screened with a woven mesh of polypropylene with a mesh size of 210 m. The tube was set to a 2-meter depth and attached to a long PVC pipe with floats on either side and pushed from shore. The tubing was purged with 2 liters of water before being connected to the sample bottle. The peristaltic pump was also used for the vertical profiles. A weighted rubber tube was lowered from the side of the boat at various depths (0.3, 1, 2, 4, 6, 8, and 10 m) and then connected to the pump, which like the other samples, was allowed to purge for two minutes before each sample was taken. Ground Water Sampling Ground water samples were taken from 6 wells (Well 1, 2, 3, 4, 6, and 7) that are located within the Park. Completed depth, screened interval, and depth to bedrock are given in Table 2-2. Table 2-2. Drilling records of six wells drilled during 2003 Wells Completed Depth (ft) Screened Interval (ft) Depth to Bedrock (ft) 1 75 75-55 56 2 100 100-80 20 3 93 93-73 10 4 97 97-77 15 6 102 102-82 16 7 98 98-78 18

PAGE 28

18 At each well a Redi-flow 2” variable performance submersible pump was used to pump water directly into the base of an 8.4-liter bucket, which was allowed to overflow ensuring no atmospheric interference. The wells were purged with at least three well volumes before 222Rn samples were collected. The samples were collected by submerging the tubing to the base of the bucket and allowing the vacuum to fill the bottle. Lab Methods Radon was extracted from the water samples within 24 hours from sample time to insure a minimal loss of 222Rn to decay. Bottles were attached to the extraction line (Figure 2-2), clamps were loosened, and ultra pure (99.999992 %) helium was allowed to bubble into the sample through the l2-inch copper pipe (Figure 2-1). The helium sparged gases from the water, and all gases flowed from the bottle through the shorter copper pipe. Gases flowed to a U-shaped collection tube, which is filled with copper filings and submerged in liquid nitrogen (Figure 2-2). Prior to entering the U tube, water and carbon dioxide are removed from the carrier gas by Hammond drierite and Thomas ascarite. As the gases flow through the U tube, the 222Rn freezes to the surfaces of the copper filings and other remaining gases (e.g. He and O2) escape. After one hour, the liquid nitrogen is removed, the bottles are disconnected, and the U tube is warmed. Counting cells (Lucas cells) are attached to the extraction line and helium pushes the trapped 222Rn into the cells (Figure 2-2). The cells are then allowed to sit for a minimum of three hours for the alpha particle from the 222Rn decay to allow an ingrowth of daughters (Operations Manual, 1985) (Figure 2-3) and then are placed on the alpha counter. After 222Rn has been extracted from a sample, it is allowed to sit for a minimum of 7 days to allow regrowth of

PAGE 29

19 Figure 2-2. A simplified schematic of the Radon Extraction Line. Each number below corresponds with the number in the diagram. (J. Cable, 2003 personal communication) 1. By-pass valve purges line between samples or will be used when evacuating Lucas cells. 2. Valve controls flow into the sample bottle. Flow is regulated by the Flow meter (400 ml/min is recommended). 3. Valve controls flow out of sample bottle. Gas from bottle headspace is forced out of bottle and carried through the line. Flow passes through Drierite to remove water vapor and Ascarite to remove carbon dioxide. These gases must be removed to eliminate their freezing inside the LN2 trap and taking up surface area reserved for radon molecules. 4. Valve controls flow into LN2 cold trap. This valve is open during the sample processing and is closed after the collection period (about 60 minutes) is complete. Time series experiments using standards have shown that after 50 minutes 90% of radon in water is collected on the trap. After 60 min, 99% of radon is collected. 5. Valve controls flow into the Lucas cell. During sample processing, Lucas cell is left off of the line and helium is allowed to exit the cell port. Valve 5 is open. Radon and helium are forced into the cell after the trap has been heated and the cell is evacuated. 6. Valve controls flow to vacuum pump. Left closed during sample processing. It is opened only when the trap is closed (4 & 5 valves) so that you can evacuate the Lucas cell.

PAGE 30

20 Figure 2-3. A simplified schematic of the radon counting system modified from Operations Manual 1012899A. The voltage travels to the PM tube and gives off pulses of light with each detected particle. The pulse of light is then sent to the counter. 222Rn from the 226Ra dissolved in the water. The ingrown 222Rn is then extracted and measured following the techniques for initial 222Rn analysis, which provides a measurement of 222Rn activity in the sample. Once all samples have been run, the bottles are emptied to determine water volume. The error associated with radon analysis are those associated with sample counting, cell background counting, volume, line efficiency, and operator error. Efficiency of the extraction line and counter were calculated by running 222Rn standards. The primary standard used was NIST 268.2 Bq/g (16,092 dpm/g). The standard was diluted into 8 standards in bottles identical to those used for the measured 222Rn samples and were extracted for 222Rn using the same process as with the measured samples. Each Lucas cell was used three times. Table 2-3 shows that the average percent efficiency of the

PAGE 31

21 Table 2-3. Average percent efficiency and standard deviation for each Lucas cell. extraction line, lucas cells, and alpha counter range from 41 – 80 % and standard deviation ranges from 0.01 to 0.54. The precision of the procedure was determined by running duplicates. Duplicates were taken at Vinzants Landing at a depth of one and six meters, at the River Rise at a depth of one and six meters, and at Well 2. The percent difference of the duplicates taken at Vinzants Landing (1m) is 11% and is 64% at Vinzants Landing (6 m). The percent difference of the duplicates taken at the River Rise 1 m and 6 m are 73% and 18%, respectively. The percent difference of the duplicates of Well 2 is 29%. Absorbance A spectrophotometer measures the color absorbance quantitatively within the visible spectrum (360-375 nanometers). The instrument used in this study was the Milton Roy Spectronic 401 spectrophotometer. To get an absorbance value a wavelength was set on the spectrophotometer that will be maximally absorbed by river water. In this case the wavelength is 375 nanometers. Water with a higher amount of color due to tannins; such as river water, will have a higher absorbance value than water with little color (groundwater). It would be expected that the samples composed of groundwater will have a higher 222Rn activity and should have the lowest absorbance. Therefore, absorbency should be a good check for the 222Rn activities of this study. Counter Red Green Blue Yellow Cells I II III I II III I II III I II III Average % efficiency 45 41 39 56 44 48 72 70 53 65 80 57 Standard Deviation 0.08 0.29 0.120.20.01 0.110.040.160.250.39 0.03 0.54

PAGE 32

22 CHAPTER 3 RESULTS Precipitation and River Stage The precipitation and river stage records for the study period are shown in Figures 3-1 and 3-2. A total of 120.45 cm of rainfall fell at the O’Leno station during the study period (April 1, 2002 – May 7, 2003) with 57% of the total (120.45 cm) occurring between September 2002 and April 2003, or 62% of the study interval. River stage is controlled by precipitation and evapotranspiration, but river stage varies seasonally primarily due to evapotranspiration, which ranges from approximately 5 cm/month during the winter to approximately 14 cm/month during the summer (Gordon, 1998). The river stage between April 1, 2002 and May 7, 2003 ranged from a minimum of 9.48 masl on August 1, 2002 to a maximum of 14.43 masl on March 13, 2003 (Figure 3-2). Between the stages of 9.48 and approximately 10.50 masl, the river is completely captured by Vinzants Landing, a swallet upstream of the River Sink. Once stage reaches approximately 12 masl, the river overflows its banks at the south/southwestern portion of the Sink and continues overland flow in a south/southwestern direction within the Park. At stage of approximately 14.3 masl the river overflows its banks at the eastern portion of the Sink and begins to connect to all the other intermediate karst windows and the River Rise via overland flow (Dean, 1999). Depth Profiles Profiles of 222Rn activity with depth were measured at Vinzants Landing and the River Rise. River Rise profiles are shown on Figure 3-3 A and B and Vinzants Landing

PAGE 33

23 profiles are shown on Figures 3-4 A and B. Figure 3-3 A has two profiles, one taken on June 5, 2002 and the other on July 2, 2002. On these dates all river water was captured by the sinkhole at Vinzants Landing. Both profiles have low 222Rn activities, but show distinctly different trends. The July 2nd profile decreases with depth from a maximum of ~13 dpm/L to a minimum of ~2 dpm/L, whereas the June 5th profile is variable at the surface but increases slightly with depth below the sample taken at 4 m below the surface. Figure 3-4 B shows the first two profiles along with a third profile taken on May 7, 2003, when the river water was being captured by the River Sink at a stage of 10.43 masl. There is no pattern with depth; activities vary from a minimum of 25 dpm/L to a maximum of 186.7 dpm/L. This profile also shows no similarity with the first two profiles. The first two Vinzants profiles were taken during the summer of 2002 (July 2, 2002 and August 29, 2002) when all river water was captured by the sinkhole at Vinzants Landing (Figure 3-4 A). The profile taken on July 2nd shows the highest activity (32.86 dpm/L) at a depth of 0.3 m with a minimum at a depth of 4.0 m. The August 29th profile has the highest activity (30.37 dpm/L) at a depth of 0.3 m, but the activity changes with depth less than on July 2nd. Figure 3-4 B. shows the first two profiles as well as a third profile taken on May 7, 2003. The May 7th profile has the highest activity (393 dpm/L) at depths of 0.3 meters and passes through a minimum between 2 and 6 meters below the water surface.

PAGE 34

24 Santa Fe River Stage9 10 11 12 13 14 15 4/19/20017/28/200111/5/20012/13/20025/24/20029/1/200212/10/20023/20/20036/28/2003 DateRiver Stage (m) 5/8/02 9/12/02 1/30/03 2/19/03 2/27/03 3/5/03 6/5/02 7/2/02 8/29/02 5/7/03 Depth Profile Sample Times Individual Site Sample Times I II III IV V VI Figure 3-1. The precipitation record for O'Leno State Park from June 2001 to July 2003, which encompasses the entire study peri od. The red arrows indicate when individual sinks were sampled and the green arrows indicate when the depth profile sampling was performed. The roman numerals indicate the sample periods (Table 2-1).

PAGE 35

25 O'Leno State Park Precipitation0 1 2 3 4 5 6 7 8 9 4/19/20017/28/200111/5/20012/13/20025/24/20029/1/200212/10/20023/20/20036/28/2003 DatePrecipitation (cm) Depth Profile Sample Times Individual Site Sample TimesIIIVV 5/8/02 9/12/02 2/19/03 1/30/03 2/27/03 3/5/03 8/29/02 7/2/02 5/7/03 6/5/02I III VI Figure 3-2. The Santa Fe River Stage recorded from O’Leno State Park from mid April 2001 till the end of June 2003. The red ar rows indicate when individual sites were sampled and the green arrows indicate when the depth profile samples were performed. The roman numerals indicate the sample periods (Table 2-1).

PAGE 36

26 02468101214 0 2 4 6 8 10 12 6/5/02 7/02/02 Excess 222Rn Activity (dpm/L) Depth (m) 050100150200 0 2 4 6 8 10 12 6/5/02 7/02/02 5/7/03 48.33 48.33 4.60 4.60Excess 222Rn Activity (dpm/L) Depth (m) A Excess 222Rn Activity (dpm/L) Figure 3-3. Vertical depth profiles at the River Rise. A. Depth profiles of the River Rise taken when the river was captured by the sinkhole at Vinzants Landing. B. The first two profiles as seen in graph A and the profile taken in May 2003 when the river was flowing to the River Sink. The values at 1 m and 6 m depths are averages of duplicate samples. Note the scale change between profile A and B. B

PAGE 37

27 0 5 10 15 20 25 30 35 0 2 4 6 8 8/29/02 7/2/02 Excess 222 Rn Activity (dpm/L) Depth (m) 050100150200250300350400 0 2 4 6 8 7/2/02 8/29/02 5/7/03 28.77 28.77 103.82 103.82Excess 222Rn Activity (dpm/L) Depth (m)A B Figure 3-4. Vertical depth profiles at Vinzants Landing. A. The depth profiles of Vinzants Landing when it captured the river B. The first two profiles as seen in graph A versus the profile taken in May 2003 when the river was flowing to the River Sink. The values at 1 m and 6 m depths are averages. Note the scale change between profile A and B.

PAGE 38

28 This profile also has no pattern with depth and shows no similarity with the first two profiles Excess 222Rn, Precipitation, and River Stage Sample Periods I to VI Excess 222Rn (dpm/L) records for the first two sample periods (May 8, 2002 and September 12, 2002) are shown in Figure 3-5 A, when river stage was 9.78 and 10.16 masl, respectively. Little to no precipitation had occurred in this area directly preceding the sampling events and all activities for both dates are less than 50 dpm/L. River stage was sufficiently low at these times that all river water flowed into Vinzants Landing and none into the River Sink. Activities measured for sample period III (January 30, 2003) are shown in Figure 3-5 B. Approximately 40.21 cm of rain fell between September 12, 2002 and January 30, 2003, causing the river stage to rise to 10.53 masl (Figure 3-1, 3-2). Most of the 222Rn activities sampled on January 30, 2003 are higher than the first two sample periods, with the highest activity of 295.0 dpm/L at Hawg Sink. Ogden Pond, Jug Sink, and Two Hole have activities similar to the first two sample periods. By sample period IV (February 19, 2003) the river stage reached 11.49 masl and all activities returned to values less than or equal to 50 dpm/L (Figure 3-5 C). This trend of decreasing 222Rn activities with increasing stage continued to sampling period five (February 27, 2003) where, at a stage of 11.51 masl, all locations, with an exception of Hawg Sink have activities less than 25 dpm/L. At that location, the 222Rn activity was 97.5 dpm/L but this value also represents a decrease in activity of 33 % from its peak activity of 295 dpm/L on January 30, 2003 (Figure 3-5 C). By March 5, 2003, sample period six, the river water had risen over the lower of two staff gages and overflowed its

PAGE 39

29 0 25 50 75 100 125 150 02000400060008000 5/8/02 9/12/02Excess 222Rn (dpm/L)Distance (m) 0 25 50 75 100 125 150 02000400060008000 1/30/03Excess 222Rn (dpm/L)Distance (m) 0 25 50 75 100 125 150 02000400060008000 2/19/03 2/26/03 3/5/03Excess 222Rn (dpm/L)Distance (m) Figure 3-5. 222Rn activities of samples from karst windows, the River Sink, and the River Rise versus distance along the flow path for sample dates of A. 5/8/02 and 9/12/02, B. 1/30/03, and C. 2/19/03, 2/26/03, and 3/5/03. Big A Sink Ogden Parener’s Jim Jug Hawg Two Hole Sweetwater Rise B Sink Ogden Parener’s Jim Jug Hawg (287.7) Two Hole Sweetwater Rise Big Sink Jim Hawg Sweetwater Rise Ogden

PAGE 40

30 banks. At this time, most sample locations have activities less than or equal to 25 dpm/L (Figure 3-5 C). The relationship between the average, standard deviation, and coefficient of variation for samples collected during sample periods 1-6 are shown in Table 3-1. The averages shown on Table 3-1 do not include Hawg Sink or Ogden Lake for 2/19/03. At that time Ogden Lake had an activity of 0.09 dpm/L, which is likely in error. The different activities at Hawg Sink suggest that it may not be on a similar flow path to the other karst windows. Table 3-1. Average, standard deviation, and coefficient of variation for each sample time. Sample TimeNumber of samplesAverage (dpm/L)Standard DevationCoefficient of Variation 1922.58.70.4 21118.38.40.5 3960.435.10.6 4526.219.50.7 5314.69.80.7 6511.35.30.5 Well Samples During the course of this study seven wells were drilled in locations throughout the field area (Figure 1-2). Six of the wells were sampled on April 30, 2003 and 222Rn activities are shown on Figure 3-6. Wells 1, 2, and 3 have 222Rn activities ranging from 1172 dpm/L to 732 dpm/L and are either located in the northern portion of the park (Wells 1 & 2) or up gradient (northeast) of the mapped conduit (Well 3). Wells 4, 6, and 7 have activities less than wells 1, 2, and 3 and range from 432 to 152 dpm/L. Wells 4, 6, and 7 are located in the southern portion of the study area. Well 3 is also located in the southern portion of the study area, but it is located up gradient from the other southern wells and the mapped conduit.

PAGE 41

31 0 200 400 600 800 1000 1200 1400 012345678Excess 222Rn (dpm/L)Wells Well 1 Well 2 (avg) Well 3 Well 4 Well 6 Well 7 Figure 3-6. 222Rn activities of wells 1, 2, 3, 4, 6, and 7. A duplicate was taken at well 2. The activity plotted is the average of the two samples taken with a range of 213 dpm/L. Mixing Model Mixing between the Santa Fe River and the Floridan Aquifer can be estimated by modeling the quantities of each water type in the samples. Assuming there is only two end member mixing with distinct 222Rn signatures, the amount of river water in the samples (%X) can be calculated by a mixing model (Kincaid, 1998), described by %X = ((Rs – Raq)/(Rriv – Raq)) 100 (1) where Rs represents the activity of 222 exRn measured in each sample, Raq is the activity of 222 exRn in the groundwater, which is determined by the average well activities of wells 1, 2, and 3, and Rriv is the amount of 222 exRn in the river. The River Sink has large changes in activity due to the variable input of ground and surface water in the upper Santa Fe River. Therefore, the 222 exRn activities for the River Sink during each sample period were used for Rriv. This model has been used to calculate percentages of river water in North and/or East of Conduit South and West of Conduit

PAGE 42

32 samples taken from the River Sink, Rise, and the intermediate karst windows. Figure 3-7 shows the fraction of river water for each site with respect to river stage according to values calculated using equation (1). Sample periods I and II have approximately 100% river water in each sample. Sample period III shows more variability; the X% ranges from 60 to 100%. Sample period IV, V, and VI have approximately 100% river water in each sample. Because the River Sink 222Rn activities were used as the Rriv variable all %X values for the River Sink are 100% and are not reported in Figure 3-7. 65 70 75 80 85 90 95 100 105 9.51010.51111.51212.5 Ogden Big Pareners Jim Jug Hawg Two Hole Sweetwater Rise 9.51010.51111.51212.5 % X Stage (m) Figure 3-7. %X versus Stage (m) for each site in the study area. Lower the %X represents higher the groundwater content in the samples. Decay Equation Sink vs. Rise The amount of decay that takes place over a period of time can be determined in the basic radioactive decay equation described by N = Noe – t (2)

PAGE 43

33 N is the activity of parent atoms that remain after a certain travel time (t), No is the original activity of radioactive parent atom, and is the decay constant, which is 0.263 day-1 for 222Rn. Rearranging the equation allows calculation of travel time from calculated and measured 222Rn activities. t = ln(N) – ln( N o ) (3) Table 3-2 shows how travel time, decay, and the measured activities of the samples taken at the River Sink and Rise interact. Martin (2003) was able to calculate velocities and travel time by correlating temperature peaks as water flowed through the Sink/Rise system. Travel time (t) was based on these travel times and velocities given in Martin (2003) for 2/19/03, 2/23/03, and 3/3/03, which are compared to sample periods 2/19/03, 2/26/03, and 3/5/03 of this study. No water was entering the River Sink during the first two sample periods (5/8/02 and 9/12/02); therefore, it is possible that there was no flow from the Sink to the Rise. Because of this, travel times cannot be calculated for these two dates. Martin (2003) did not have velocities calculated for a date close to sample period III (1/30/03) therefore the travel time here is from Dean (1999). Calculated activity for the River Rise (222 exRn Rise calc) was found by using equation (2), assuming the initial 222Rn activity (No) is the activity at the River Sink. Calculated travel time (T222 exRn) is the travel time found by using equation (3), where No is the 222Rn activity of the River Sink, and N is the 222Rn of the River Rise. The measured activity of the Rise sample collected January 30, 2003 is higher than the calculated activity. For example, the measured activity of the water at the River Rise was 53.50 dpm/L, but the calculated value was 17.0 dpm/L. The last three sample

PAGE 44

34 Table 3-2. Relationship between measured 222Rn-activity and travel time and calculated 222Rn-activity and travel time using the decay equation. Date Stage (m) 222 exRn Sink1 (dpm/L) 222 exRn Rise2 (dpm/L) 222 exRn Rise calc3 (dpm/L) Tt4 (days) T 222 exRn5 (days) 5/8/02 9.78 22.55 10.6 ------2.9 9/12/02 10.16 18.32 8.54 ------2.93 1/30/03 10.53 48.6 53.5 17.00 4 ---2/19/03 11.49 20.32 4.45 13.70 1.5 5.78 2/26/03 11.51 24.25 4.62 16.34 1.5 6.31 3/5/03 >11.51 17.94 4.54 13.8 1 5.25 1 & 2 Measured 222Rn activities of the River Sink and Rise, respectively. 3 N calculated by the equation 2 using 1 as No and 4 at travel time. 4 Travel time from Martin (2003) based on river stage during the sample times. 5 Travel time calculated by equation 3, using 1 and 2 as N0 and N, respectively. periods (February 19, 2003, February 26, 2003, and March 5, 2003) have a different relationship than the first three sample periods between the measured and calculated 222 exRn activities for the River Rise. The measured activities are lower than the calculated values. For example, the measured activity of the Rise on February 19, 2003 was 4.45 dpm/L and the calculated activity was 13.7 dpm/L. Travel times for each sample period were also calculated using the decay equation. However, the calculated travel times do not correspond with the travel times from Dean (1999) and Martin (2003) and are not used in any other calculations or discussion. Sink vs. Karst Windows A 222Rn value for each site was calculated by using equation (1) and the value of 222Rn activity at the River Sink. These values are reported in Appendix Table A-1, which gives a comparison of measured and calculated 222Rn activity, 226Ra activity, and travel time between the River Sink and all other sites. Martin (2003) only calculated travel times for the River Sink, Parener’s, Two Hole, Sweetwater, and the Rise. Travel times

PAGE 45

35 for the other sites (Ogden, Big, Jim, Jug, and Hawg) were calculated from velocities reported by Martin (2003) Radon-222 activity increased from the River Sink to most karst windows on May 8, 2002, September 12, 2002, and January 30, 2003 and decreased from the River Sink to most karst windows on February 19, 2003, February 26, 2003, and March 5, 2003. Locations where this trend is not followed include Ogden Pond, Parener’s Sink and Hawg Sink. Activity decreased between the River Sink and Ogden Pond for all sample periods. Parener’s Sink shows a increase for all period. Hawg Sink has higher 222Rn activity for all sample times, except February 19, 2003 (Figure 3-8). -40 -20 0 20 40 60 80 100 010002000300040005000600070008000 5/8/02; 9.78 m 9/12/02, 10.16 m 1/30/03, 10.53 m 2/19/03, 11.49 m 2/26/03, 11.51 m 3/5/03, > 11.51 mGain (dpm/L)Distance (m)Loss (dpm/L)Ogden Big Paraners Jim Jug Hawg Two Hole Sweetwater Rise Figure 3-8. The gain or loss in activity downstream from the River Sink. If there is a gain in activity then the measured activity at the site is higher than the measured activity at the Sink. If there is a loss in activity the measured activity at the site is lower than the measured activity at the Sink. Activity at Hawg Sink for January 30,2003 sample time is 295 dpm/L.

PAGE 46

36 Gas Exchange Equation To determine the amount of 222Rn that is lost to the atmosphere by evasion, Elsinger and Moore (1983) and Ellins et al (1990) used the gas exchange equation. Cd = Cu e-[D/(zhv)] x (4) Cu and Cd are the 222Rn activities up and downstream. D is the molecular diffusivity of 222Rn, which is 1.2 x10-9 m2/s at 23C and h, v, and x are the average stream depth, velocity and distance between sample locations, respectively. The variable z is the thickness of stagnant film layer at the surface of the stream. Elsinger and Moore (1983) found the thickness of the stagnant film of the Pee Dee River in South Carolina to be between 19 and 48 m. Within this range, the thickness of the stagnant film layer does not significantly effect the calculations (< 1%) therefore the average of this (33.5 m) is used for the current study. However, if the thickness of the stagnant film decreases to lower than 19 m, the calculations are effected. Even though this equation is not for karst environments, it was used to establish a first order estimate of the amount of evasion occurring in the Sink/Rise system. This equation assumes that both stream and the air above it constitute two well mixed reservoirs with uniform vertical activities separated by a stagnant film of water (Ellins, 1990). The amount of exchange is mainly governed by flow generated turbulence (Ellins, 1990); however, the thicker the film the slower the rate of transfer. Color Absorbance vs. Activity Figure 3-9 and 3-10 suggest that the initial hypothesis of absorbance may not be correct. Figure 3-9 shows activity vs. absorbance values for the River Sink in samples collected on 1/30/03, 2/19/03, and 3/5/03. Contrary to expectations, the sample with the

PAGE 47

37 lowest absorbance has the highest activity. In Figure 3-10, the sample with the highest activity has the lowest absorbance; however, the sample with the second highest activity also has the highest absorbance. The relationship between absorbance and stage can be seen in Figures 3-11A, B, and C. Figure 3-11 A. gives averaged absorbance values for each site versus stage and Figures 3-11 B and 3-11 C. plot Ogden and Hawg sinks versus stage. R2 values for the regression curves on Figure 3-11 A, B, and C are 0.888, 0.854, and 0.9589, respectively. All graphs show an increase in absorbance with the increase of stage.

PAGE 48

38 Sink 0 0.1 0.2 0.3 0.4 0.5 0.6 0204060 Activity (dpm/L)Absorbence 2/19/03 3/5/03 1/30/03 Figure 3-9. 222Rn activities of the River Sink at various sample periods vs. Absorbance. The sample with the highest activity has the lowest absorbance. 3/05/03 0.4 0.42 0.44 0.46 0.48 0.5 0510152025 Activity (dpm/L)Absorbence Hawg Sink Ogden Jim Sweetwater Figure 3-10. 222Rn activities of samples taken at Sweetwater Lake, Jim Sink, Ogden Pond, the River Sink, and Hawg Sink on 3/5/03.

PAGE 49

39 River Stage vs Absorbencey = 0.119x 0.9755 R2 = 0.8884 0 0.1 0.2 0.3 0.4 0.5 0.6 1010.51111.51212.5 Stage (ft)Absorbe n 1/30/03 2/19/03 2/26/03 3/5/03 Ogdeny = 0.128x 1.0834 R2 = 0.854 0 0.1 0.2 0.3 0.4 0.5 910111213 Stage (m)Absorbe n 1/30/03 2/19/03 2/26/03 3/5/03 Hawg y = 0.2467x 2.5052 R2 = 0.9589 0 0.1 0.2 0.3 0.4 0.5 910111213 Stage (m)Absorbenc e 1/30/03 2/19/03 2/26/03 3/5/03 A B C Figure 3-11. Absorbance values taken during the study. A. Average absorbance for all the samples collected. B. Absorbance values for Ogden Pond. C. Absorbance values for Hawg Sink. Each graph covers samples collected on 1/30/03, 2/19/03, 2/26/03, and 3/5/03.

PAGE 50

40 CHAPTER 4 DISCUSSION Precipitation and River Stage River stage varies little during the summer months due to low amounts of rainfall and high evapotranspiration rates (~ 14 cm/month) (e.g. Gordon, 1998). However, low evapotranspiration rates (~ 5 cm/month) during the fall and winter months cause higher stage fluctuations than the summer, particularly during associated times of increased precipitation and passage of cold fronts (e.g. Gordon, 1998). During the winter months of this study, river stage increased rapidly with rainfall: between March 3-7, 2003, 2.27 cm of rainfall caused an increase of stage by 1.24 meters (Figure 3-1and 3-2). In all of June 2002, 11.07 cm of rain fell; however, evapotranspiration caused river stage to decrease by 0.03 m during this time. Depth Profiles Assuming that surface water loses 222Rn to the atmosphere and sources of 222Rn are from the Hawthorn group and solid material in the Floridan Aquifer, there should be an increase of 222Rn with depth in the sinkholes. However, depth profiles in this study indicate that the water has heterogeneous 222Rn activities with depth. All profiles taken from the River Rise and Vinzants Landing reflect heterogeneous activities with depth that may reflect mixing between river water and groundwater (Figures 3-3 and 3-4). The main difference between the profiles shown in Figures 3-3 and 3-4 is that Vinzants Landing has its highest activities at the surface. The lithology of the sinkholes should not have an impact over the activity; however, this is unknown. It is

PAGE 51

41 possible that the high activities seen at both sites are due to higher amounts of sediment mixing with the waters. Chung (1973) and Berelson et al (1982) indicate that the increase in 222Rn activity with depth (which should be seen in depth profiles) is due to fluxes of sediment caused by turbidity currents within the water column. 226Ra is bonded with the sediment and the decay to 222Rn causes the increase in activity. Both of these studies are conducted in coastal environments in the Santa Barbara and San Nicholas basins, California. Even though the environments differ from the current study, they provide an analog for what could occur at Vinzants Landing and the River Rise. Excess 222Rn Three capturing processes control the 222Rn activity in the karst windows of the Santa Fe River: mixing of low activity surface water and 222Rn rich ground water, atmospheric evasion, and radioactive decay. All three processes depend on the flow through the Sink/Rise System. During the first two sample periods, all water in the Santa Fe River was captured by the sinkhole at Vinzants Landing. The river stages for these two dates are 9.78 masl and 10.16 masl, respectively. Martin and Screaton (2002) suggest that during low flow conditions, the conduit will have a lower head than the matrix and will act as a drain for the surrounding matrix porosity (Figure 4-1 A). With no water flowing into the River Sink, groundwater should be the main influence over the 222Rn activities measured from the samples taken at these times and even during 5/8/02 and 9/12/02 water seemed to be resurging at the River Rise (i.e., the River Rise was not dry). Therefore the conduit should have water with high 222Rn activity that may approach values found at the wells (~200 to ~1200 dpm/L). Flow rates are slow at low stage conditions (~10.00 masl), with travel times from the River Sink to the Rise of approximately 8 days or longer (Dean, 1999; Martin &

PAGE 52

42 Dean, 1999). However, if there is no flow entering the River Sink, then travel times are indeterminable. Slow to no flow in the Sink/Rise system, combined with 222Rn’s short half-life (3.84 days), indicates that radioactive decay will reduce the activities. This also allows longer time for evasion of 222Rn to the atmosphere. Figure 4-2 indicates that more evasion does occur during slow flow, but not enough to account for the low activity. Rainfall prior to sample period I and II is minimal therefore there is no dilution caused by surface water recharge. Therefore, it is possible that the diffuse flow from the matrix to the conduit is slow enough that significant decay occurs. This along with other factors such as radioactive decay within the conduit and evasion probably account for the low activities during these sample periods. Prior to sample period III, 18.8 cm of rain fell and caused the river stage to rise to 10.53 masl, which is considered average river stage based on the hydrograph of river stage for the study period (Figure 3-2). Some 222Rn activities are higher than seen during the first two sample periods. Travel time from the Sink to the Rise is approximately four days on the basis of the stage vs. travel time relationship in Dean (1999). More rapid travel time may prevent most evasion and decay allowing 222Rn activities to stay elevated. For example, calculated 222Rn activity (N from equation 2) for Ogden Pond (Appendix Table A-1) during sample period III is 48.3 dpm/L, but activity from the River Sink (No) is 48.6 dpm/L. Also, results from the gas exchange equation (Appendix Table A-3) show that at Ogden Pond only 0.17 dpm/L was lost to evasion. Both of these results indicate that the quicker travel time creates less loss of activity due to radioactive decay and evasion. Another possible explanation for the elevated 222Rn activities is that an increase in diffuse flow from the rain has created a more rapid influx of 222Rn rich water

PAGE 53

43 River Water Mixing Mixing Matrix Water Santa Fe River Water Sample Periods IV, V, and VI C Decay over time Evasion Matrix Water Sample Periods I and II A Mixing Matrix Water Santa Fe River Water Sample Period III B Figure 4-1. A conceptual model of interactions between the Floridan Aquifer and water in conduits during the sample periods. A. No water is entering the River Sink; therefore, the matrix water enters the conduits. Since there is little flow through the Sink/Rise system at this time, the water in the conduits decays. B. River water is now flowing to the River Sink and water from the matrix is entering the conduit and mixing with the river water. There is flow in the system and the decay is minimal. C. The river is flooding and the abundant amount of river water in the conduit is now moving to the matrix.

PAGE 54

44 Evasion vs. Travel Timey = 0.4965x + 0.1155 R2 = 0.9679 0.00 1.00 2.00 3.00 4.00 5.00 0.002.004.006.008.0010.00 Travel Time (days)Evasion (dpm/L) 9.78 10.16 10.53 11.49 11.51 >11.51 Figure 4-2. Evasion at the River Rise increases as travel time increases. The numbers by the data points indicate river stage at the various sample times. (ground water) to flow into the conduit and mix with the incoming river water (Figure 41 b). A similar flow of groundwater to conduit was observed using Clconcentrations in the water. At stages of 10.45 to 10.70 masl, Dean (1999) saw an increase in Clby 24.7% to 43.2 % from the River Sink to Sweetwater Lake, which he interpreted to indicate that ground water flowed to the conduits. Compared to the first two sample periods, the current study reports an increase in 222Rn during this stage of greater than or equal to 80, this is probably due to the activity being lower than expected during the first two periods. An increase in diffuse recharge should cause an increase in flow from the matrix to the conduit; therefore, most sites should show a higher activity than the first two sample periods. However, Ogden Pond, Jug Lake, and Two Hole Sink have activities similar to the first two sample periods. It is unclear as to why Ogden Pond, Jug Lake, and Two Hole Sink differ from the other samples. Of all the karst windows, Hawg Sink has the highest 222Rn activity at 295.0 dpm/L. A study conducted by Sprouse (2004, in process) reports calcium concentrations of approximately 60 mg/L and an alkalinity of ~140 mg/L for

PAGE 55

45 Hawg Sink in January 2003. Compared to the chemistry of the other sinks, these concentrations are high and are possible indicators of ground water input. Along with the high 222Rn activity measured at Hawg Sink, this may indicate that it is not on the same flow path as the other sinks and could represent an unmapped groundwater source. During sample periods IV, V, and VI, the river stage was 11.49, 11.51, and >11.51 masl, respectively. During sample period VI, elevation of the Santa Fe River was greater than the staff gage located at the River Sink preventing stage measurements and an exact stage value. At this time, it was observed that the river overflowed the banks on the southwestern portion of the Sink and began to flow over land. The travel times from the Sink to the Rise are 1.3 days for 2/19/03 and 2/26/03 and 1 day for 3/5/03 (Martin, 2003). These rapid travel times prevent a large loss of activity due to radioactive decay and evasion. For example, the calculated activity for the River Rise is ~ 4 dpm/L less than the River Sink (Table 3-2) and the gas exchange equation indicates that only 0.5 dpm/L was lost to evasion. A possible cause for the observed low activities is dilution by rainwater. Two days prior to sampling on February 19th, 7.57 cm of rain fell, 1.06 cm of rain fell the week before sampling on February 26th, and 3.4 cm fell the week prior to March 5th. It takes 15 days for secular equilibrium between 226Ra and 222Rn to be reached. If the surrounding sediments have been previously flushed with water, then a pulse of water within 15 days of the first pulse, that flushed the system, will have less 222Rn activity. However, depending on the timing of the water pulses, it is still possible that rainwater would gain 222Rn activity as it moves through the sediment column and a dilution effect from the rainwater and groundwater mixing in the matrix would not be seen. It is most

PAGE 56

46 likely that the low activities seen at the various sinks during the sample periods are due to dilute 222Rn-activity water entering the conduit at the River Sink. Based on Clconcentrations, Dean (1999) reported that at approximately 11.9 masl most of the water at the River Rise originated from the River Sink, suggesting that there is little loss of water from matrix porosity to the conduit during high flow. Martin (2003) supported this conclusion and suggested that heads are higher in the conduit then in the matrix when water is leaving the conduit and flowing to the matrix. Therefore, the river water in the conduit is flowing out into the matrix (Martin and Screaton, 2002) (Figure 4-1 c). Color absorbance values also indicate that the water with in the karst windows is consistent with an origination at the River Sink. At most of the karst windows, there is an increase in absorbance with an increase in stage (Figures 3-11, 12,and 13), which suggests that higher flow rates flush the ground water from the conduit. Radioactive Decay Along the Flow Path At times of low flow, groundwater input and radioactive decay should be the main controls over the 222Rn activities of the karst windows. The low 222Rn activities in samples collected at the River Rise on May 8, 2002 and September 12, 2002 (Figure 3-4 A) indicate that radioactive decay and evasion control the water’s activity. Samples collected January 30, 2003 have higher activities and a quick travel; therefore the effects of decay and evasion may not be as strong. Radon-222 activities calculated from equation (2) are not corrected for atmospheric evasion. However, results of equation (4) indicate that approximately 4.5 dpm/L and 3.6 dpm/L are lost by evasion for May 8th and September 12th, respectively and only 2.5 dpm/L is lost for January 30th (Appendix Table A-3). For each of these dates, the calculated 222Rn activities for the River Rise from equation (2) are approximately twice

PAGE 57

47 the measured activities, which indicate that the measured activities are higher than if decay was the only influence over the water 222Rn activity (Table 3-2). Therefore, there is some source of 222Rn to the water (i.e., ground water). Samples taken from the River Rise on February 19, 2003, February 27, 2003, and March 5, 2003 all have measured 222Rn activities that are lower than calculated activities, (Table 3-2, and Figure 3-8). Travel times during these sample periods are shorter (1-1.5 days) than the half-life of 222Rn (3.84 days) suggesting that a lesser amount of decay will occur. The calculated activities for the River Rise are not much lower than the activities at the River Sink, indicating that not much decay has occurred. For example, the calculated activities at the River Rise for 2/19/03, 2/26/03, and 3/5/03 are all approximately 5 dpm/L less than the activities at the River Sink. The amount of evasion for 2/19/03, 2/26/03, and 3/5/03 was calculated to be 0.75, 0.71, and 0.48 dpm/L, respectively and the amounts of evasion during 5/8/02, 9/12/02, and 1/30/03 are 4.45, 3.62, and 2.45 dpm/L, respectively. According to the Gas Exchange equation, this indicates the quicker travel times also prevent most evasion. Figure 4-2 indicates that the amount of evasion decreases with increasing travel time. The low 222Rn-activity river water is not mixing with the high 222Rn-activity groundwater within the conduit but moving out into the matrix. Other processes thus appear to control the loss of 222Rn from the conduit. If the loss of radon from the River Sink to the River Rise is not due to evasion what is causing this? It is possible that the water in the conduit takes a longer time to travel through the system than the temperature tracking indicates. Alternatively, it is possible that the loss in activity is from mixing with a low 222Rn activity water source. Old

PAGE 58

48 Bellamy Cave Exploration Team have mapped a main conduit which enters the system from the east (Old Bellamy Cave Exploration Team, unpublished report); however, there are no known surface water sources for the eastern system, suggests that it does not supply dilute water. With the large amount of precipitation (> 20 cm) that occurred during the later sample periods it is possible that the loss of activity is caused by a influx of low activity surface water that travels to the conduit. With a few exceptions (Ogden Pond, Parener’s Sink, and Hawg Sink) the other sample sites used in this study agree with the 222Rn activities at the Rise. Ogden Sink has calculated activities higher than the measured activities during all sample periods but May 8, 2002. This could be caused either by less mixing with radon-rich ground water, more degassing, or a combination of both. Parener’s Sink shows a decrease in activity on 9/12/02; however, it is so slight (-0.5 dpm/L) that it is smaller than analytical error of the measurement. Hawg Sink shows an increase in activity from the River Sink during every sample period but February 19, 2003 indicating a different source of groundwater not available to the other karst windows. Distribution of 222Rn in the Groundwater Regional groundwater flow is from the northeast to the southwest (Miller, 1997); therefore the wells on the eastern side of the conduit should sample groundwater that is not influenced by water from the conduits. Martin (2003) indicates that local water flow is towards the conduit during base flow conditions. Samples from the wells were collected at a stage of 11.05 masl. Water collected from wells 1, 2, and 3 have a higher activity than wells 4, 6, and 7. Well 1 and 3 are located in the eastern portion of the study area and the measured activity of 1173 and 1134 dpm/L, respectively, suggests that they are influenced by the groundwater. Even though Well 2 is on the western side of the

PAGE 59

49 conduit, the high activity (733 dpm/L), which is measured, indicates that it is also influence by the matrix water. Wells 1 and 2 are located in the northern portion of the park, in which they are in closer proximity to the Hawthorn Group, which is a source of 222Rn. Wells 4, 6, and 7 are in the southwestern portion of the park, just west of the River Rise. The low activities measured at these wells (152, 432, and 294 dpm/L, respectively) indicate that may be influenced by the mixing between the river and ground water within the matrix. It is uncertain as to whether the Hawthorn provides more 222Rn activity to the area then the limestone within the Floridan Aquifer. Crandall (1996) found that surficial aquifer water influenced by the Hawthorn Group had 222Rn activities of 1200 dpm/L. Smoak et al. (2000) found 226Ra activities for the Floridan to be approximately 6 dpm/L. Wells 1, 2, and 3 have similar 226Ra activities, which indicates that the 222Rn activities of the water measured in Smoack et al (2000) might also be relatively high. Well 1, 2, and 3 activities are also similar to those found in Crandall (1996). Therefore, at this point the 222Rn activities provided by the Hawthorn Group and the 222Rn activities provided by the Floridan Aquifer limestone are indistinguishable. Well activities were used as an end member (Raq) in the mixing equation. Kincaid (1998) was able to use this model successfully and found the amount of river water to be between 50 and 90 %. The %X, in this study, range from approximately 60 to > 100. Figure 3-7 suggests that during sample periods I, II, IV, V, and VI primarily river water is in the conduit and that during sample period III water in the karst windows is influenced by ground water. However, previously discussed results suggest that this is not the case. Water chemistry data from Dean (1999) indicated that water within the karst windows

PAGE 60

50 was influenced by ground water. The 222Rn activities during sample periods I and II are low, which is the result of radioactive decay and atmospheric evasion, not an influx of river water. Therefore, in a future study of the Sink/Rise system more end members need to be defined and quantified than what was provided by equation (1). Such end members include influx of radon from the surrounding sediment and the influx of radon from the Hawthorn Group. It is also important to use a variety of tracers, each with distinct geochemical behavior, such as 18O, Cl-, SO4 -2, as well as 222Rn, to better constrain surface and ground water mixing.

PAGE 61

51 CHAPTER 5 CONCLUSIONS Karst aquifers are important hydrologic systems and provide potable water to most of our world. Therefore, it is necessary to understand the hydrologic characteristics of these aquifers. One characteristic is the abundance of sinking streams, which allow for mixing between the surface and ground water. These streams have potential to carry pollutants that, once mixed with the potable groundwater, could spread quickly and harm thousands of individuals. Natural tracers that record mixing of surface and ground water would be valuable to developing an understanding of mechanisms and quantities of mixing. Results from this study suggest that 222Rn activity varies with stage. During low stage, the low 222Rn activities suggest a loss of activity to decay and evasion. In the future, to determine the exact amount of evasion that is occurring the stagnant film thickness needs to be corrected for the Santa Fe Sink/Rise system. During base flow, the high 222Rn activities suggest mixing between the river and ground waters within the conduit. During high stage, low activities indicate there is little to no mixing between the river and ground waters; river water is flowing from the conduit to the matrix. The results also indicate that Hawg Sink may not be on the main flow path. Furthermore, results indicate that a two end-member-mixing model is not adequate to quantify the amounts of water involved in the mixing. In the future, when modeling the quantities of water involved in mixing, other end members, such as, decay, evasion, and sediment input need to be defined. The effects of lithology on the 222Rn activities

PAGE 62

52 cannot be determined at this time. This is something that should be resolved in the future. Other future work involves more depth profiles at other locations, as well as, the River Rise and Vinzants Landing to fully understand vertical mixing with depth. Radon222 may prove to be a useful tracer of mixing between surface and ground water; however, other tracers, such as Cland 18O should be used to support the results from the measured 222Rn activities and further analysis of the complex karst system, the Santa Fe River Sink/Rise system needs to continue.

PAGE 63

53 APPENDIX MEASURED AND CALCULATED 222RN, TOTAL 226RA, TRAVEL TIME, AND EVASION FOR ALL SITES Table A-1. Travel time, measured and calculated 222Rn, and total 226Ra Location Date Stage (m) Travel Time (days) ^ 222 exRn calc (dpm/L) ** 222 exRn (dpm/L) ** 226Ra (dpm/L) River Sink 5/8/02 9.78--38.21 -0.25 9/12/02 10.16--14.21 0.00 01/31/03 10.53--12.64 1.34 02/19/03 11.49--14.28 -2.49 2/26/03 11.51--0.09 -1.45 3/5/03 <11.51--14.8 1.62 Ogden 5/8/02 9.78>0.24 >24.17 22.55 2.02 9/12/02 10.160.24 17.73 18.32 1.59 01/31/03 10.530.126 48.3 48.6 1.64 02/19/03 11.490.025 20.18 20.32 -2.45 2/26/03 11.510.025 24.08 24.25 1.14 3/5/03 <11.51< 0.025 <17.81 17.94 0.26 Big 5/8/02 9.78> 1.16 >16.62 18.43 1.25 9/12/02 10.161.16 13.5 27.09 1.15 01/31/03 10.530.61 41.41 92.10 1.41 02/19/03 11.490.119 ---2/26/03 11.510.119 ---3/5/03 <11.51<0.119 ---Parener’s 5/8/02 9.78>1.45 >15.42 23.37 1.17 9/12/02 10.161.45 12.53 12.01 1.32 01/31/03 10.530.756 39.8 96.96 -0.44 02/19/03 11.490.15 ---2/26/03 11.510.15 ---3/5/03 <11.51<0.15 ---Jim 5/8/02 9.78> 1.71 >14.38 31.20 -0.39 9/12/02 10.161.71 11.68 24.48 0 01/31/03 10.530.89 38.61 84.79 1.64 02/19/03 11.490.25 21.8 38.93 0.53 2/26/03 11.510.25 ---3/5/03 <11.51<0.25 <16.73 11.36 0.51

PAGE 64

54 Table A-1. Continued Location Date Stage (m) Travel Time (days) ^ 222 exRn calc (dpm/L) ** 222 exRn (dpm/L) ** 226Ra (dpm/L) Jug 5/8/02 9.78> 2.99 >10.23 25.18 2.41 9/12/02 10.162.99 8.31 32.05 1.64 01/31/03 10.531.56 32.25 29.71 1.11 02/19/03 11.490.43 ---2/26/03 11.510.43 ---3/5/03 <11.51<0.43 ---Hawg 5/8/02 9.78> 3.62 >8.55 17.82 -0.18 9/12/02 10.163.62 6.94 11.51 1.28 01/31/03 10.531.89 29.48 287.66 376.24 02/19/03 11.490.52 17.67 14.62 -2.58 2/26/03 11.510.52 21.08 97.50 -0.84 3/5/03 <11.51<0.52 <15.60 21.29 0.78 Two Hole 5/8/02 9.78> 3.67 >8.63 11.45 1.16 9/12/02 10.163.67 7.01 16.93 2.51 01/31/03 10.531.91 29.5 20.74 0.95 02/19/03 11.490.53 ---2/26/03 11.510.53 ---3/5/03 <11.51<0.53 ---Sweetwater 5/8/02 9.78> 4.82 >6.33 21.37 0.13 9/12/02 10.164.82 5.14 25.10 0 01/31/03 10.532.52 25.11 104.37 1.37 02/19/03 11.490.7 16.97 52.94 -1.87 2/26/03 11.510.7 20.26 14.83 1.44 3/5/03 <11.51<0.70 15 8.08 0.38 River Rise 5/8/02 9.78> 8 days >2.76 10.6 -0.11 9/12/02 10.168 days 2.24 8.537 1.24 01/31/03 10.534 17.01 53.5 -0.07 02/19/03 11.491.3 14.46 4.45 -0.08 2/26/03 11.511.3 17.26 4.62 -0.08 3/5/03 <11.511 12.78 4.54 0.36 Travel time (days) is based on Dean (1999) and Martin (2003) ** Measured activities of 222Rn and 226Ra ^ Activities calculated using the decay equation

PAGE 65

55 Table A-2. Excess 222Rn and total 226Ra for Depth Profiles and Wells Sites Sample Dates 222 exRn (dpm/L) 226Ra (dpm/L) Rise 1m 6/5/02 3.69 1.56 Rise 2m 6/5/02 7.12 2.24 Rise 4m 6/5/02 1.91 1.17 Rise 6m 6/5/02 3.04 2.47 Rise 8m 6/5/02 3.36 0.68 Rise 10m 6/5/02 4.77 1.86 Vinzants 0.3048m 7/2/02 32.86 1.16 Vinzants 1m 7/2/02 25.25 1.10 Vinzants 2m 7/2/02 20.51 0.76 Vinzants 4m 7/2/02 13.78 0.81 Vinzants 6m 7/2/02 30.22 -2.18 Vinzants 8m 7/2/02 24.04 1.52 Rise 0.3048m 7/2/02 13.23 1.30 Rise 1m 7/2/02 7.14 1.24 Rise 2m 7/2/02 3.63 2.43 Rise 4m 7/2/02 3.02 2.05 Rise 6m 7/2/02 2.78 0.75 Rise 10m 7/2/02 2.39 0.88 Vinzants 0.3048m 8/29/02 30.37 0.51 Vinzants 1m 8/29/02 16.94 2.19 Vinzants 2m 8/29/02 19.69 0.80 Vinzants 4m 8/29/02 19.23 0.73 Vinzants 6m 8/29/02 21.19 1.60 Vinzants 8m 8/29/02 20.88 0.79 Vinzants 0.3048m 5/7/03 392.93 0.24 Vinzants 1m (1) 5/7/03 275.38 0.30 Vinzants 1m (2) 5/7/03 247.76 0.65 Vinzants 2m 5/7/03 36.78 0.92 Vinzants 4m 5/7/03 129.25 0.66 Vinzants 6m (1) 5/7/03 110.52 1.15 Vinzants 6m (2) 5/7/03 213.93 0.70 Vinzants 8m 5/7/03 321.54 0.17

PAGE 66

56 Table A-2. Continued Sites Sample Dates 222 exRn (dpm/L) 226Ra (dpm/L) Rise 0.3048m 5/7/03 31.55 0.22 Rise 1m (1) 5/7/03 123.60 0.49 Rise 1m (2) 5/7/03 8.79 0.98 Rise 2m 5/7/03 149.79 0.41 Rise 4m 5/7/03 186.74 0.14 Rise 6m (1) 5/7/03 27.76 0.13 Rise 6m (2) 5/7/03 23.24 0.78 Rise 8m 5/7/03 61.23 1.32 Rise 10m 5/7/03 110.25 -0.34 Well 1 4/30/03 1051.82 -7.08 Well 2 (1) 4/30/03 762 4.76 Well 2 (2) 4/30/03 625.87 1.94 Well 3 4/30/03 1120.69 0.77 Well 4 4/30/03 131.96 -0.36 Well 6 4/30/03 431.61 5.78 Well 7 4/30/03 293.61 1.49

PAGE 67

57 Table A-3. Gas Exchange Equation Site Date Cd (dpm/L) Amount of Evasion (dpm/L) Ogden Pond 5/8/02 22.40 0.15 9/12/02 18.19 0.13 1/30/03 48.43 0.17 2/19/03 20.31 0.01 2/26/03 24.23 0.02 3/5/03 17.93 0.01 Parener’s Sink5/8/02 21.82 0.73 9/12/02 17.72 0.60 1/30/03 47.77 0.83 2/19/03 20.25 0.07 2/26/03 24.17 0.08 3/5/03 17.88 0.06 Big Sink 5/8/02 21.64 0.91 9/12/02 17.58 0.74 1/30/03 47.56 1.04 2/19/03 20.23 0.09 2/26/03 24.15 0.10 3/5/03 17.86 0.08 Jim Sink 5/8/02 21.48 1.07 9/12/02 17.45 0.87 1/30/03 47.38 1.22 2/19/03 20.18 0.14 2/26/03 24.08 0.17 3/5/03 17.81 0.13 Jug Lake 5/8/02 20.71 1.84 9/12/02 16.82 1.50 1/30/03 46.48 2.12 2/19/03 20.07 0.25 2/26/03 23.95 0.30 3/5/03 17.72 0.22 Hawg Sink 5/8/02 20.34 2.21 9/12/02 16.52 1.80 1/30/03 46.05 2.55 2/19/03 20.02 0.30 2/26/03 23.89 0.36 3/5/03 17.67 0.27

PAGE 68

58 Table A-3. Continued Site Date Cd (dpm/L) Amount of Evasion (dpm/L) Two Hole Sink 5/8/02 20.31 2.24 9/12/02 16.50 1.82 1/30/03 46.02 2.58 2/19/03 20.02 0.30 2/26/03 23.89 0.36 3/5/03 17.67 0.27 Sweetwater Lake 5/8/02 19.65 2.90 9/12/02 15.97 2.35 1/30/03 45.23 3.37 2/19/03 19.92 0.40 2/26/03 23.77 0.48 3/5/03 17.59 0.35 River Rise 5/8/03 18.10 4.45 9/12/02 14.70 3.62 1/30/03 20.10 2.45 2/19/03 19.58 0.74 2/26/03 23.36 0.89 3/5/03 17.44 0.50

PAGE 69

59 REFERENCES Berelson, W. M., Hammond, D. E., and Fuller, C, 1982. Radon-222 as a tracer for mixing in the water column and benthic exchange in the southern California borderland. Earth and Planetary Science Letters 61: 41-54. Bertin, C., Bourg, A. C. M., 1994. Radon-222 and Chloride as natural tracers of the infiltration of river water into an alluvial aquifer in which there is significant river/groundwater mixing. Environmental Science Technology 28: 794-798. Boyer, D. G., Pasquarell, G. C., 1996. Agricultural land use effects on nitrate concentrations in a mature karst aquifer. Water Res. Bull. 32: 565-573. Budd, D. A., Vacher, H. L., 2002. Facies control on matrix permeability in the upper Floridan Aquifer, West-Central Florida: Implications to diffuse flow. In Hydrogeology and Biology of Post-Paleozoic Carbonate Aquifers, Karst Waters Institute Special Publication 7, ed. J. B. Martin, C. M. Wicks, and I. D. Sasowsky, 14-24. Charles Town, West Virginia: Karst Waters Institute. Bush, P. W., Johnston, R. H., 1988. Ground-water hydraulics, regional flow, and groundwater development of the Floridan Aquifer system in Florida and in parts of Georgia, South Carolina, and Alabama. U. S. Geological Survey: 80. Chung, Y., 1973. Excess radon in the Santa Barbara Basin. Earth and Planetary Science Letters 17: 319-323. Corbett, R. D., Burnett, W. C., Cable, P. H., Clark, S. B., 1997. Radon tracing of the groundwater input into Par Pond, Savannah River Site. Journal of Hydrology, 203: 209-227. Crandall, C. A., 1996. Shallow ground-water quality in selected agricultural areas of south-central Georgia. U. S. Geological Survey Water Resources Investigative Report # 96-4083. Dean, R. W., 1999. Surface and ground water mixing in a karst aquifer: An example from the Floridan Aquifer. M. S. Thesis, Gainesville, Florida, University of Florida, 74 p. Drever, J. I., 1988. The Geochemistry of Natural Waters, 2nd edition. Englewood Cliffs, NJ, Prentice Hall, 437 p.

PAGE 70

60 Ellins, K. K., Hisert, R. A., Kincaid, T. R., 1992. Hydrogeology of the western Santa Fe River Basin. Field Trip Guide, South Eastern Geological Society Spring Meeting, Gainesville, FL. Ellins, K. K., Roman-Mas, A., Lee, R., 1990. Using 222Rn to examine groundwater/surface discharge interaction in the Rio Grande De Manati, Puerto Rico. Journal of Hydrology 115: 319-341. Elsinger, R. J., Moore, W. S., 1983. Gas exchange in the Pee Dee River based on 222Rn evasion. Geophysical Research Letters 10:443-446. Ford, D. C., Williams, P. W., 1989. Karst Geomorphology and Hydrology. Winchester, MA, Unwin Hyman, Ltd. Ginn, B., 2002. Using Temperature and Water Elevation measurements to model conduit properties in karst Aquifers; An Example from the Santa Fe Sink-Rise System, Florida, senior undergraduate thesis, University of Florida, 23 p. Gordon, S. L., 1998. Surface and groundwater mixing in an unconfined karst aquifer, Ichetucknee River ground water basin, Florida: Ph.D. Dissertation, University of Florida. Halihan, T., Wicks, C. M., Engein, J. F., 1998. Physical response of a karst drainage basin to flood pulses: Example of the Devil's Icebox cave system (Missouri, USA). Journal of Hydrology. 204:24-36. Hamada, H., 2000. Estimation of groundwater flow rate using the decay of 222Rn in a well. Journal of Environmental Radioactivity 47: 1-13. Hamada, H., Komae, T., 1998. Analysis of recharge by paddy field irrigation using 222Rn concentration in groundwater as an indicator. Journal of Hydrology 205: 92-100. Hisert, R. A., 1994. A multiple tracer approach to determine the ground and surface water relationships in the western Santa Fe River, Columbia County, Florida, Gainesville, FL, Ph.D. Dissertation, University of Florida: 213 p. Hunn, J. D., Slack, L. J., 1983. Water resources of the Santa Fe River Basin, Florida. U. S. Geological Survey. Hussain, N., Church, T. M., Kim, G., 1999. Use of 222Rn and 226Ra to trace groundwater discharge into the Chesapeake Bay. Marine Chemistry 65: 127-134. Katz, B. G., 1992. Hydrochemistry of the Upper Floridan Aquifer, Florida. USGS Water Resources Investigation Report 91-4196, 37 p. Kincaid, T. R., 1998, River water intrusion to the unconfined Floridan Aquifer, Environmental and Engineering Geoscience IV: 361-374.

PAGE 71

61 Martin, J. B., Dean, R. W.,1999. Temperature as a natural tracer of short residence times for ground water in karst aquifers. Karst Modeling. A. N. Palmer, M. V. Palmer and I. D. Sasowsky, Karst Waters Institute Special Publication #5: 236-242. Martin, J. B., Screaton, E. J., 2002. Exchange of matrix and conduit water with examples from the Floridan Aquifer, U.S. Geological Survey Karst Interest Group proceedings, St. Petersburg, Florida. Water Resources Investigations Report 014011, 38-44. Martin, J. M., 2003. Quantification of the matrix hydraulic conductivity in the Santa Fe River Sink/Rise system with implications on the exchange of water between the matrix and conduit, Gainesville, FL. M. S. Thesis, University of Florida: 100p. Miller, J. A., 1997. Hydrogeology of Florida: In: The Geology of Florida, eds. A. F. Randazzo and D. S. Jones, Gainesville, Florida: University Press of Florida, p. 6988. Padilla, A., Pulido-Bosch, A., Mangin, A., 1994. Relative importance of base flow and quick flow from hydrographs of karst springs. Ground Water 32: 267-277. Puri, H. S., Vernon, R. O., 1964. Summary of the Geology of Florida and a guidebook to the classic exposures, Florida Geological Survey Special Publication #5. Rogers, A. S., 1958. Physical behavior and geological control of radon in mountain streams, Geological Bulletin No. 1052-E, U. S. Geological Survey. Ryan, M., Meiman, J., 1996. An examination of short-term variations in water quality at a karst spring in Kentucky. Ground Water 34, 1: 23-30. Scott, T. M., 1992. A Geological Overview of Florida, Florida Geological Survey. Skirvin, R. T., 1962. The underground course of the Santa Fe River near High Springs, Florida, Gainesville, FL. M. S. Thesis, University of Florida, 62 p. Smart, C. C., 1988. Artificial tracer techniques for the determination of the structure of conduit aquifers. Groundwater 26: 445-453. Smoak, J. M., Brenner, M., Allen, M. S., Shelskie, C. L., Leeper, D. A., 2000. Biological Accumulation of 226Ra in groundwateraugmented Florida Lake. Limnology and Oceanography 45, 3: 710-715. Stringfield, V. T., LeGrand, H. E., 1966. Hydrology of limestone terrains in the coastal plains of the southeastern United States: Geological Society of America Special Paper 93, 43 p. Tennelec, Operations Manual, 1985, Dual Radon Counting System, Oak Ridge, Tennessee.

PAGE 72

62 Wherett, Judith L., 1992. Geologic characterization of soil gas radon concentrations in Florida and comparison with soil permeability parameters, Gainesville, FL. M. S. Thesis, University of Florida: 87p. White, W. A., 1970. The geomorphology of the Florida Peninsula, Geological Bulletin #51. White, W. B., 1999. Conceptual models for karstic aquifers. Karst Modeling. A. N. Palmer, M. V. Palmer and I. d. Sasowsky, Karst Waters Institute Special Publication #5: 11-16.

PAGE 73

63 BIOGRAPHICAL SKETCH Lauren A. Smith was born to Roger D. Smith and Martha A. Smith on August 30, 1977, in Ft. Hood, Texas. Shortly after her birth, her family moved to Dublin, Georgia, and has lived there ever since. After graduating from Dublin High School in 1995, Lauren attended Furman University in Greenville, South Carolina, where she graduated, in 2000, with a B. S. in earth and environmental science. Later that year, she moved to Florida to begin graduate school at the University of Florida. She worked as a teaching assistant for undergraduate classes and as a research assistant. She is currently seeking employment in the environmental industry.


Permanent Link: http://ufdc.ufl.edu/UFE0003260/00001

Material Information

Title: Using Radon-222 as a tracer of mixing between surface and ground water in the Santa Fe River Sink/Rise System
Physical Description: x, 63 p. ; ill.
Language: English
Creator: Smith, Lauren A. ( Dissertant )
Martin, Jonathan B. ( Thesis advisor )
Screaton, Elizabeth ( Reviewer )
Neuhoff, Philip ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Geological Sciences thesis, M.S
Dissertations, Academic -- UF -- Geological Sciences

Notes

Abstract: Karst aquifers (which provide 25% of the world's population with potable water) are characterized by links between surface and ground water, making them susceptible to contamination. Linkage between groundwater and surface water occurs through the unconfined Floridan Aquifer in north-central Florida. An example of surface- groundwater linkage occurs when the Santa Fe River flows into a 32-m deep sinkhole, the River Sink, and resurges approximately 8 km down gradient at the River Rise. This system provides opportunities to observe mixing between surface and ground water within a karst aquifer. Mixing can be observed using natural chemical and isotopic tracers, one of which is the isotope ²²²Rn (half-life = 3.82 days). Radon-222 is formed by the alpha decay of ²²⁶Ra, an alkaline earth element common in carbonate and clay minerals. Surface waters have low ²²²Rn activities (approx. 20 dpm/L) due to atmospheric evasion and ground water have high ²²²Rn activities ( > 200 dpm/L). The difference in activities of these two end members should make ²²²Rn a good tracer for surface-ground water mixing in karst areas. Activities of ²²²Rn were measured to depths of 8 and 10 m at two sinkholes (Vinzants Landing and River Rise) and indicated that ²²²Rn is heterogeneous in the sinkholes. Water samples were collected from the River Sink, River Rise, and karst windows between the Sink and Rise from May 2002 to April 2003. These samples encompass drought, base flow, and flood stages. As the river stage rises to flood conditions (approx. 11.51 masl) the excess ²²²Rn activities decrease from approximately approx. 100 dpm/L to less than 25 dpm/L, indicating that no matrix water enters the conduit; and that the conduit is filled with ²²²Rn-poor surface water from the Santa Fe River. During low stage (9.25 masl), when past studies have found groundwater dominates the system, the water activities are also low (less than or equal to 25 dpm/L) possibly caused by evasion and radioactive decay of ²²²Rn as the water slowly flows through the Sink/Rise System. At times of base flow conditions for the river (10.70 m stage), activities are approx. 100 dpm/L, indicating some ground water and surface water mixing. The fraction of surface and groundwater in the system could not be determined through a two end-member mixing model because evasion and radioactive decay complicate the simple mixing model. ²²²Rn may be a good tracer of mixing between surface and ground water; however, further studies need to be conducted to fully understand the impacts of decay, evasion, and lithology over the
Subject: ²²²Rn, ground, hydrogeology, karst, Radon222, river, surface, water
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 73 pages.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.
Thesis: Thesis (M.S.)--University of Florida, 2004.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0003260:00001

Permanent Link: http://ufdc.ufl.edu/UFE0003260/00001

Material Information

Title: Using Radon-222 as a tracer of mixing between surface and ground water in the Santa Fe River Sink/Rise System
Physical Description: x, 63 p. ; ill.
Language: English
Creator: Smith, Lauren A. ( Dissertant )
Martin, Jonathan B. ( Thesis advisor )
Screaton, Elizabeth ( Reviewer )
Neuhoff, Philip ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Geological Sciences thesis, M.S
Dissertations, Academic -- UF -- Geological Sciences

Notes

Abstract: Karst aquifers (which provide 25% of the world's population with potable water) are characterized by links between surface and ground water, making them susceptible to contamination. Linkage between groundwater and surface water occurs through the unconfined Floridan Aquifer in north-central Florida. An example of surface- groundwater linkage occurs when the Santa Fe River flows into a 32-m deep sinkhole, the River Sink, and resurges approximately 8 km down gradient at the River Rise. This system provides opportunities to observe mixing between surface and ground water within a karst aquifer. Mixing can be observed using natural chemical and isotopic tracers, one of which is the isotope ²²²Rn (half-life = 3.82 days). Radon-222 is formed by the alpha decay of ²²⁶Ra, an alkaline earth element common in carbonate and clay minerals. Surface waters have low ²²²Rn activities (approx. 20 dpm/L) due to atmospheric evasion and ground water have high ²²²Rn activities ( > 200 dpm/L). The difference in activities of these two end members should make ²²²Rn a good tracer for surface-ground water mixing in karst areas. Activities of ²²²Rn were measured to depths of 8 and 10 m at two sinkholes (Vinzants Landing and River Rise) and indicated that ²²²Rn is heterogeneous in the sinkholes. Water samples were collected from the River Sink, River Rise, and karst windows between the Sink and Rise from May 2002 to April 2003. These samples encompass drought, base flow, and flood stages. As the river stage rises to flood conditions (approx. 11.51 masl) the excess ²²²Rn activities decrease from approximately approx. 100 dpm/L to less than 25 dpm/L, indicating that no matrix water enters the conduit; and that the conduit is filled with ²²²Rn-poor surface water from the Santa Fe River. During low stage (9.25 masl), when past studies have found groundwater dominates the system, the water activities are also low (less than or equal to 25 dpm/L) possibly caused by evasion and radioactive decay of ²²²Rn as the water slowly flows through the Sink/Rise System. At times of base flow conditions for the river (10.70 m stage), activities are approx. 100 dpm/L, indicating some ground water and surface water mixing. The fraction of surface and groundwater in the system could not be determined through a two end-member mixing model because evasion and radioactive decay complicate the simple mixing model. ²²²Rn may be a good tracer of mixing between surface and ground water; however, further studies need to be conducted to fully understand the impacts of decay, evasion, and lithology over the
Subject: ²²²Rn, ground, hydrogeology, karst, Radon222, river, surface, water
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 73 pages.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Original Version: Text (Electronic thesis) in PDF format.
Thesis: Thesis (M.S.)--University of Florida, 2004.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0003260:00001


This item has the following downloads:


Full Text












USING RADON-222 AS A TRACER OF MIXING BETWEEN SURFACE AND
GROUND WATER IN THE SANTA FE RIVER SINK/RISE SYSTEM















By

LAUREN A. SMITH


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Lauren A. Smith
















ACKNOWLEDGMENTS

I wish to express my appreciation to my advisor, Dr. Jonathan Martin, for his

constant patience and enthusiasm. I wish to express my gratitude to my committee

members, Dr. Elizabeth Screaton and Dr. Philip Neuhoff, both of whom are wonderful

mentors and friends. I wish to thank all of the employees at O'Leno State Park for their

cooperation. A special thanks goes to Dr. Jaye Cable, for her assistance; to Dr. Jason

Curtis, for his support and never-ending supply of liquid nitrogen; and to Kevin Hartl, for

construction of a beautiful radon-extraction line. My thanks go to those who supported

me through their friendship and love, especially my family. Last but certainly not least, I

would like to give the greatest thanks to Brooke Sprouse and Jennifer Martin: without

their support, patience, and laughter in the field, the ticks and snakes would have gotten

to me.
















TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ....................................... ......... ............................................iii

LIST OF TABLES ............ ................ ............... ............. vi

LIST OF FIGURES ........... .... .................................... ............vii

A B S T R A C T ................................ix.............................

CHAPTER

1 IN TR O D U C T IO N ............ ............................................. ........ .. ........... .. 1

Study A rea................................................ ........ 4
Location and Clim ate......................................... .......... ...... .............. 4
Physiography ..................................................... .............. 6
Geology and Hydrostratigraphy.................. ........ .......................... 6
B background 222R n Studies....... ......................................................... .............. 8
B background O 'L eno Studies ......... ................................ ..................... .............. 9

2 M E T H O D S ............................................................................ 14

Field M ethods ........................................ .............. .............. 14
Sam ple B ottle Construction .......................................................... .......... .... 14
W after Sam pling ..................................................... .............. 15
G round W ater Sam pling ......................................................... .............. 17
L ab M methods ........................................ 18
A b sorbance .................................................................................................... ....... 2 1

3 R E S U L T S ............................................................................. 2 2

Precipitation and River Stage............................. .............. 22
Depth Profiles ....................................... .............................22
Excess 222Rn, Precipitation, and River Stage ........................................................... 28
Sample Periods I to VI............................................................ 28
W ell Sam ples ................................................................... ... ........ 30
M ixing M odel ........................................ 3 1
D ecay E qu action ................................................................... ......................... 32
S in k v s. R ise ........................................................................................... 3 2
Sink vs. K arst W indow s..................................................... ......................... 34









Gas Exchange Equation .. ...... ................................................... .............. 36
Color Absorbance vs. Activity ......................................................... .............. 36

4 D ISCU SSION ................................................. .. ...... ...... .. ........ .. 40

Precipitation and River Stage..................... ............................... ........................... 40
D epth P profiles ........................................ 40
Excess 222Rn ............... ........ ..... .................. 41
Radioactive Decay Along the Flow Path ........................ ..................... 46
Distribution of 222Rn in the Groundwater .................. ............. ......... ............ 48

5 CON CLU SION S .................. ..................... ........ .............. .......... .. 51

APPENDIX MEASURED AND CALCULATED 222RN, TOTAL 226RA, TRAVEL
TIME, AND EVASION FOR ALL SITES .......................................................... 53

R E F E R E N C E S ................................................................. .. ..... ......... .. .. ....... ..... 59

BIOGRAPHICAL SKETCH .. ............................................................ .............. 63
















LIST OF TABLES


Table pge

1-1 Geologic and hydrogeologic units found in the Santa Fe River Basin.................. 10

2-1 Dates, river stages, and sample sites of each sample event................................. 16

2-2 Drilling records of six wells drilled during 2003. ............................................. 17

2-3 Average percent efficiency and standard deviation for each Lucas cell. ................ 21

3-1 Average, standard deviation, and coefficient of variation for each sample time .... 30

3-2 Relationship between measured 222Rn-activity and travel time and calculated
222Rn-activity and travel time using the decay equation ................................... 34
















LIST OF FIGURES


Figure pge

1-1 Map of the Santa Fe River basin, shown in shaded area ................................ 5

1-2 Site map of the Santa Fe River Sink/Rise system in O'Leno State Park................ 7

2-1 Schematic of the sample bottles used for collection of water for 222Rn analysis .... 15

2-2 A simplified schematic of the Radon Extraction Line .................................. 19

2-3 A simplified schematic of the radon counting system modified from Operations
M annual 1012899A ................................................... ..... .. .......... 20

3-1 The precipitation record for O'Leno State Park from June 2001 to July 2003,
which encompasses the entire study period....................................................... 24

3-2 The Santa Fe River Stage recorded from O'Leno State Park from mid April
2001 till the end of June 2003 ................................................................. .. 25

3-3 Vertial depth profiles at the River Rise ......................................... ..... ......... 26

3-4 Vertical depth profiles at Vinzants Landing......................................... 27

3-5 Radon-222 activities of samples from karst windows, the River Sink, and the
River Rise versus distance along the flow path for sample dates of A. 5/8/02 and
9/12/02, B. 1/30/03, and C. 2/19/03, 2/26/03, and 3/5/03 .................................... 29

3-6 Radon-222 activities of wells 1, 2, 3, 4, 6, and 7 ....... .... .................................... 31

3-7 Percent X (%X) versus Stage (m) for each site in the study area........................ 32

3-8 The gain or loss in activity downstream from the River Sink............................. 35

3-9 Radon-222 activities of the River Sink at various sample periods vs.
A bsorbance ............................................................... ... ..... ......... 38

3-10 Radon-222 activities of samples taken at Sweetwater Lake, Jim Sink, Ogden
Pond, the River Sink, and Hawg Sink on 3/5/03 ......................................... ..... 38

3-11 Absorbance values taken during the study .............. ............................. ....... ....... 39









4-1 A conceptual model of interactions between the Floridan Aquifer and water in
conduits during the sample periods .............. ...... ........................... ........... 43

4-2 Evasion at the River Rise increases as travel time increases........................... 44














Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

USING RADON-222 AS A TRACER OF MIXING BETWEEN SURFACE AND
GROUND WATER IN THE SANTA FE RIVER SINK/RISE SYSTEM

By

Lauren A. Smith

May 2004

Chair: Dr. Jonathan B. Martin
Major Department: Geological Sciences

Karst aquifers (which provide 25% of the world's population with potable water)

are characterized by links between surface and ground water, making them susceptible to

contamination. Linkage between groundwater and surface water occurs through the

unconfined Floridan Aquifer in north-central Florida. An example of surface-

groundwater linkage occurs when the Santa Fe River flows into a 32-m deep sinkhole,

the River Sink, and resurges approximately 8 km down gradient at the River Rise. This

system provides opportunities to observe mixing between surface and ground water

within a karst aquifer. Mixing can be observed using natural chemical and isotopic

tracers, one of which is the isotope 222Rn (half-life = 3.82 days). Radon-222 is formed by

the alpha decay of 226Ra, an alkaline earth element common in carbonate and clay

minerals. Surface waters have low 222Rn activities (-20 dpm/L) due to atmospheric

evasion and ground water have high 222Rn activities (>200 dpm/L). The difference in









activities of these two end members should make 222Rn a good tracer for surface-ground

water mixing in karst areas.

Activities of 222Rn were measured to depths of 8 and 10 m at two sinkholes

(Vinzants Landing and River Rise) and indicated that 222Rn is heterogeneous in the

sinkholes. Water samples were collected from the River Sink, River Rise, and karst

windows between the Sink and Rise from May 2002 to April 2003. These samples

encompass drought, base flow, and flood stages. As the river stage rises to flood

conditions (-11.51 masl) the excess 222Rn activities decrease from approximately 100

dpm/L to less than 25 dpm/L, indicating that no matrix water enters the conduit; and that

the conduit is filled with 222Rn-poor surface water from the Santa Fe River. During low

stage (9.25 masl), when past studies have found groundwater dominates the system, the

water activities are also low (< 25 dpm/L) possibly caused by evasion and radioactive

decay of 222Rn as the water slowly flows through the Sink/Rise System. At times of base

flow conditions for the river (10.70 m stage), activities are -100 dpm/L, indicating some

ground water and surface water mixing. The fraction of surface and groundwater in the

system could not be determined through a two end-member mixing model because

evasion and radioactive decay complicate the simple mixing model. 222Rn may be a good

tracer of mixing between surface and ground water; however, further studies need to be

conducted to fully understand the impacts of decay, evasion, and lithology over the 222Rn

activities in the Sink/Rise system.














CHAPTER 1
INTRODUCTION

Karst aquifers provide 25% of the world's population with potable water (Ford and

Williams, 1989). These aquifers are typically characterized by direct links between the

surface and ground water systems through sinkholes, swallets, and highly permeable

rocks, which make karst groundwater resources susceptible to surface contamination

(Ford and Williams, 1989). For example, tannins, pesticides, and nitrates from

agricultural and animal byproducts have been shown to move rapidly into and through

karst groundwater systems, contaminating water resources (Katz, 1992; Boyer and

Pasquerell, 1996; Kincaid, 1998). The loss of undersaturated surface water with respect

to calcite to the aquifer may also increase dissolution of carbonate rocks (Drever, 1988).

Tracking contaminants can be difficult because the heterogeneous permeability of

karstic rocks complicates modeling of fluid flow and contaminant transport.

Complications are due to the exchange of water among three types of porosity:

intergranular matrix porosity, fracture porosity, and conduits (White, 1999). Many

studies of karst aquifers have been conducted, but most focus on aspects of either conduit

flow or matrix flow. Highly altered karst aquifers, like those found in Paleozoic mid-

continent carbonates are dominated by conduit and fracture flow because recrystallization

has led to dense, low permeability matrix rocks. Less altered karst aquifers may have

matrix rocks with high permeability, creating flow characteristics that differ from karstic

aquifers formed in dense and altered carbonates with low permeability.









An important example of a less altered karst system dominated by both conduit and

matrix flow is the Floridan Aquifer system (Bush and Johnston, 1988), a regionally

important aquifer because it is the main source of potable water in north central Florida.

The Floridan Aquifer represents a class of karst aquifers with high intergranular porosity

and permeability (Budd and Vacher, 2002), which may allow a significant fraction of the

subsurface flow through matrix rocks (Martin and Screaton, 2001). Typical of all karstic

aquifers, the Floridan Aquifer has direct links to surface water through sinkholes,

swallets, and highly permeable rocks. Bush and Johnston (1988) suggested that on a

regional scale, it could be assumed that the conduit and matrix systems in the Floridan

Aquifer of the southeastern United States could be treated as homogenous or as an

equivalent porous medium. At a local scale, even in porous karst aquifers such as the

Floridan Aquifer, conduit flow predominates and the equivalent porous medium approach

is invalid (Bush and Johnston, 1988; Padilla et al, 1994; Halihan et al, 1998). Prevention

and remediation of contamination in these aquifers require an understanding of the

mixing among surface and ground water and controls of flow paths in the subsurface.

Many techniques have been used for studies of flow paths, mixing, and flow

velocities through karst aquifers. Common techniques include injection of artificial dyes

(e.g., rhodamine and fluorescein), studies of physical characteristics of the water, and

measurement of chemical and isotopic compositions of the water and dissolved

constituents (e.g., Smart, 1988; Ryan and Meiman, 1996; Martin and Dean, 1999). One

dissolved radioactive component with good potential for studying the mixing between

ground and surface water is the isotope 222Rn (half-life=3.82 days). Radon is a product of

the natural radioactive decay series of 238U and a direct product of alpha disintegration of









226Ra (Bertin and Bourg, 1994). As an alkaline earth element, radium is found in

carbonate and clay minerals; and thus large amounts of radon are generated in ground

water. However, because of atmospheric evasion (loss of radon gas to the atmosphere),

radon activity should be low in surface waters. In addition, radon is chemically

conservative (i.e., no source other than 226Ra and no sinks other than decay); and is easily

measured at low activities (e.g., 0.02 pCi/L (0.044 dpm/L)) through alpha-counting

techniques. Therefore, activities of 222Rn in karst water should reflect the relative

amounts of water from the surface with low 222Rn activity and water that has been stored

in matrix porosity and thus gained a high 222Rn activity.

Rogers (1958) was the first to use 222Rn to study ground water and surface water

relationships. He demonstrated that 222Rn activities in springs (ground water) were much

higher than in surface water and that the springs were a source of 222Rn in the streams

surrounding his study area. A study by Ellins et al. (1990) of Puerto Rican streams in an

upland karst region confirmed Rogers' (1958) work. Ellins et al. (1990) described low

activities of 222Rn in the streams (18 to 200 dpm/L) as a function of the atmospheric

evasion and the high levels of 222Rn (up to 693 dpm/L) in related springs due to the input

of enriched groundwater. Bertin and Bourg (1994) were able to use the difference in

222Rn activities in ground and surface water to trace the seeping of Lot River water in

France into an alluvial aquifer. Though Bertin and Bourg (1994) did not work in a

karstic terrain, their study suggests that 222Rn can be used as a tracer of infiltrating river

water and can be used in the presence of mixing within the aquifer.

This study has two principle objectives:

* To evaluate the potential for using 222Rn in a karst environment for tracing
groundwater and surface water interactions









* To define the interaction between surface water and ground water in a karst system
with high matrix permeability using 222Rn.

To satisfy these objectives, several questions concerning 222Rn activities in a karstic

terrain were addressed:

* Does 222Rn trace surface water and ground water mixing in a karst region?
* Will 222Rn trace mixing of the water from the matrix and conduit?
* How can the interactions between surface and ground water be quantified?
* Will lithology prove to be a main control over 222Rn activities?

Study Area

Location and Climate

The Santa Fe River Basin (Figure 1-1) covers an area of 3584.5 km2 and is a

tributary basin of the Suwannee River (Hunn & Slack, 1983). The Santa Fe River

originates in Altho and Santa Fe lakes in north-central Florida (Hunn & Slack, 1983) and

flows westward approximately 50 km until it reaches a 32 m deep sinkhole at O'Leno

State Park. River water that enters the sinkhole mixes with groundwater from the

Floridan Aquifer and travels through areas of mapped and unmapped conduits. This

water resurfaces at the River Rise (-8 km down gradient of the Sink) and continues as the

lower Santa Fe River (Hisert, 1994; Dean, 1999) (Figure 1-2).

Located in a semi-tropical climate, average air temperature in the Santa Fe River

Basin is 200C and average groundwater temperature is around 220C (Hisert, 1994;

Florida Geological Survey, 1992). Precipitation averages 123 cm/yr with most occurring

between June and September. Most summer rainfall occurs during local thunderstorms

and seasonal tropical storms. As a result, precipitation can be heterogeneous over small

areas. Winter rainfall is from extratropical storms, which causes most flooding.











































Figure 1-1. Map of the Santa Fe River basin, shown in shaded area. The location of
O'Leno State Park is shown by the square and is seen in detail in Figure 1-2.
Modified from Hunn, J. D., Slack, L. J., 1983. Water resources of the Santa Fe
River Basin, Florida. U. S. Geological Survey.









Physiography

Three dominant physiographic divisions in the region include the Northern or

Proximal zone, the Central or Mid-peninsular zone and the Southern or Distal zone

(White, 1970). O'Leno State Park is located on the border between the Northern and

Central physiographic zones within the Western Valley, a 140 mile long lowland, and the

High Springs Gap, an opening within the Western Valley. This portion of Florida is a

well-drained area of high recharge, and variably developed karst. The High Springs Gap

includes the Cody Scarp, which represents the erosional edge of the Hawthorn Group.

This scarp is the most prominent topographic feature in peninsular Florida, yet it is a

subtle feature (with ~25m per 10km) (Puri and Vernon, 1964; White, 1970). The

Hawthorn Group is the confining unit for the Floridan Aquifer, and thus karst features are

common where the Hawthorn Group is absent. Most streams, including the Santa Fe

River, either disappear or become losing streams to the aquifer systems below as they

cross the scarp.

Geology and Hydrostratigraphy

Three aquifer systems comprise the hydrostratigraphy of the study area and include

the Surficial Aquifer system, Intermediate Aquifer system, and the Floridan Aquifer

system. The connection between the stratigraphy and hydrostratigraphy is shown in

Table 1-1. The Surficial Aquifer system provides small-yield domestic and agricultural

water supplies (Scott, 1992) and is composed of undifferentiated Plio-Pleistocene

sediments containing sinkhole fill, fine to medium sands, and layers of clay and silt

(Hunn and Slack, 1983; Scott, 1992). In O'Leno State Park, the Surficial Aquifer system

is less than 2 meters thick to absent (Hunn and Slack, 1983).



















































Figure 1-2. Site map of the Santa Fe River Sink/Rise system in O'Leno State Park. The
dashed line represents locations of caves mapped by cave divers (Old Bellamy
Exploration Team, unpublished report). The solid line represents roads and
the pentagons represent wells drilled in 2003. Modified from Hisert (1994);
Dean (1999).









The Intermediate Aquifer is contained within the Hawthorn Group (Bush and

Johnston, 1988). The Hawthorn Group in Northern Florida consists of interbedded

phosphatic carbonates and siliciclastics with a trend of increasing siliciclastics in the

younger sediments. These sediments have low permeability and form an effective

aquiclude, the intermediate confining unit, which confines the Floridan Aquifer where

present (Bush and Johnston, 1988). In areas where the intermediate confining unit is

absent such as the western portion of the Santa Fe River Basin, the Surficial Aquifer

directly overlies carbonates of the Floridan Aquifer (Scott, 1992). The lack of the

confining unit limits lakes and wetlands in the surface (Scott, 1992). Karst features, such

as sinkholes are common through both the Surficial and Intermediate Aquifer systems

and provide direct recharge to the Floridan Aquifer System.

The Floridan Aquifer System is composed of several hundreds of meters of

limestone and dolostone and is the main source of water in northern Florida and parts of

Georgia, South Carolina and Alabama (Hunn and Slack, 1983; Stingfield and LeGrand,

1966). The stratigraphic units comprising the Floridan Aquifer system from late- Eocene

to Oligocene include the Oldsmar Limestone, Avon Park Limestone, Ocala Limestone,

and Suwannee Limestone. The Ocala Limestone is the uppermost stratigraphic unit in

O'Leno State Park (Hisert, 1994).

Background 222Rn Studies

Two surface and groundwater mixing studies using 222Rn were conducted

previously in portions of the Santa Fe River Basin. Ellins et al. (1992) used background

222Rn activities of the Santa Fe River (-10 dpm/L) and spring 222Rn activities in the lower

Santa Fe River Basin (-1000 dpm/L) to observe a ratio of ground to river water 222Rn

activities of 100:1. Hisert (1994) sampled 222Rn in the Santa Fe River along









Hollingsworth Bluff, located southwest of O'Leno State Park. In samples collected at the

riverbed, the depth halfway between the riverbed and surface, and the surface of the river,

Hisert (1994) found that 222Rn activities decreased with depth and were dependent on

groundwater influxes.

Hisert (1994) also measured 222Rn activities within the O'Leno State Park. At an

average discharge of 42 m3/s, 222Rn activities in water from selected karst windows in the

northern section of the park (River Sink and Jim Sink) are identical to atmospheric

background levels (< 50 dpm/L) (Hisert, 1994). At these discharges, activities in water

from the karst windows in the southern section of the park (Two Hole, Sweetwater Lake,

and River Rise) were close to ground water levels (-450 dpm/L). Hisert interpreted this

increase in activity to result from an influx of groundwater into the southern karst

windows.

It is unclear as to whether the Hawthorn Group or the carbonates in the Floridan

Aquifer is the major source of the 222Rn in the study. Smoak et al. (2000) found well

water from the Floridan Aquifer near Tampa, FL to have 226Ra activities for the Floridan

Aquifer to range from 5.8 to 6.6 dpm/L. These 226Ra activities are similar to those seen at

wells in this study, which correspond to 222Rn activities of approximately 1000 dpm/L.

Wherett (1992) found average soil gas 222Rn activities of the Hawthorn Group to be

approximately 2000 dpm/L. Although it is unclear how average soil gas converts to the

liquid gas used in this study, it can be assumed that the numbers are within the same

order of magnitude.

Background O'Leno Studies

Skirvin (1962) conducted the first study of water flow through the Sink/Rise

system of the Santa Fe River. Based on the tannic acid content of the water, Skirvin












Table 1-1. Geologic and hydrogeologic units found in the Santa Fe River Basin. Modified from Hunn and Slack (1983), Scott (1992),
Hisert (1994), and Dean (1999).
Series Stratigraphic Unit Hydrogeologic Unit Lithologic Description Thickness (m)

Holocene Undifferentiated Surficial Aquifer Sinkhole fill, fluvial terraces, and 0-25
Pleistocene Sediments thin surficial sand

Pleistocene Alachua Formation Intermediate aquifer/ Reddish-white sands, with clays, 0-30
to Miocene Upper confining unit sandy clays, and phosphate pebbles

Middle to Hawthorn Group Phosphatic clayey sand-sandy clay
Lower Miocene with varying amounts of Fullers
Earth
and carbonate

Oligocene Suwannee Limestone Floridan Very pale yellow, moderately 0-100
indurated,
Aquifer porous, fossil-rich calcarenite

Eocene Ocala Limestone Very permeable limestone, 275-300
Avon Park dolomitic limestone, and dolomite
Limestone
Oldsmar Limestone

Paleocene Cedar Keys Formation Sub- Floridan Limestone, some evaporites and ?
confining unit clay









(1962) suggested that most water entering the River Sink discharged at the River Rise.

Skirvin (1962) also observed that water discharged from the Rise and intermediate karst

windows when water was blocked from the River Sink by a temporary dam. Skirvin

found that more water discharged from the Rise than flowed into the Sink in May and

November 1961. The overall discharge gain in May and November were 5.30 m3/s and

5.08 m3/s, respectively. These observations indicate that the system gains water from

sources other than the River Sink.

Hunn and Slack (1983) suggested that the water quality of the basin at this time

was more influenced by natural factors, such as limestone dissolution, than by

anthropogenic factors. Hunn and Slack (1983) describe the sinking eastern rivers as

having lower concentrations of iron, calcium, magnesium, and bicarbonate than the water

within the Floridan Aquifer. The rising western rivers have concentrations of iron,

calcium, magnesium, and bicarbonate similar to the Floridan Aquifer. These chemical

compositions indicate that the rivers contain water from the Floridan Aquifer, which is

similar to the results reported by Skirvin (1962).

Hisert (1994) provided the first in-depth look into the complex underground system

of O'Leno State Park by using sulfur hexafluoride (SF6), temperature, and 8180 to trace

groundwater flow from the River Sink to the River Rise. The SF6 tracer experiment,

conducted in July 1991, required two injections to link flow from the Sink to the Rise.

The first injection, performed at a discharge of 31.3 m3/s, connected the River Sink to

Sweetwater Lake through multiple karst windows, but SF6 from this injection was not

recovered at the River Rise. A second injection into Sweetwater Lake was detected at the

River Rise (Fig. 1-2). The discharge rate during this second injection was not reported,









but a third injection was made at Jim Sink when River discharge was 5.8 m3/s into the

River Sink, which connected Jim Sink to the River Rise via Sweetwater Lake. Even

though a direct connection was never made between the River Sink and Rise, Hisert

concluded that the karst windows in the Santa Fe River Sink/Rise system are

hydraulically connected.

Hisert was also able to use background temperatures of the river (150C) and

groundwater (22C) to conclude that the 2.5C increase in temperature found within

some of the windows, including the Rise, indicated a 37.5% influx of groundwater. This

conclusion was supported by variations in oxygen ratios of the water samples. The karst

windows have a slightly enriched 8180 signature compared to the River Rise reflecting an

influx of 6180-enriched groundwater.

Dean (1999) used water chemistry to determine the extent of mixing between the

river and ground water, and he used temperature as a tracer of flow rate from the River

Sink to the River Rise. Using natural variations in temperature, Dean (1999) found that

subsurface travel time of the water from the River Sink to the River Rise varies from

approximately 12 hours to nearly 8 days, depending on river stage. By using temperature

as a tracer and discharge measurements, Ginn (2002) and Screaton et al. (2004) found a

subsurface conduit flow at rates of 3.11 km/day, which are similar to the subsurface flow

rate found in Hisert's thesis. Both Ginn (2002) and Dean (1999) agree with Hisert's

conclusion that the sinks in O'Leno State Park are all hydraulically well connected.

Chloride (Cl-) concentrations from Dean (1999) increased between the River Sink

and Sweetwater Lake by an average of 35.8% during low flow conditions. This increase

suggests an addition of Cl- rich water, probably groundwater. During high flow






13


conditions the C1 concentrations between the Sink and Sweetwater Lake change by an

average of 4.3%, which suggests very little addition of ground water.














CHAPTER 2
METHODS

The objectives of this project were met through a series of physical and chemical

measurements including daily precipitation, river stage records, and measurements of

isotopic compositions in water samples collected from various sites in the region of the

Santa Fe River Sink/Rise system. Park staff collected daily precipitation and river stage

measurements for the O'Leno State Park station on the Santa Fe River. New chemical

analyses used in this thesis include measurement of 222Rn activities in all collected water

samples and measurement of absorbance in selected samples.

Field Methods

Sample Bottle Construction

Sample bottles were constructed from emptied, cleaned 2.5 L glass acid bottles,

which were wrapped in duct tape to prevent breakage (Figure 2-1). A two-holed rubber

stopper along with two lengths of 1/40-inch diameter copper pipes, one 2-inch long piece

and the other 12-inches long were placed in the mouth of each bottle. The base of the

12-inch pipe was attached to an Aqua-Tech air diffuser. The rubber stopper was secured

with thin wire and attached to the bottle with silicon to make a gas tight seal. Rubber

tubing with male/female connectors was attached to the ends of the two copper pipes that

extend from the rubber stopper. Clamps were attached to each rubber tube to control

airflow. Each bottle was evacuated in the lab for approximately 5 min. before water

sampling. The vacuum allows the water samples to be sucked into the bottle and

prevents atmospheric contamination of the sample.













[ -Clamp
Gas tight. Clamp
silicon seal 4 Connector
2" copper tubing
12" copper tubing




Duct Tape
2.5 L acid bottle
Air bubbler

Figure 2-1. Schematic of the sample bottles used for collection of water for 222Rn
analysis.

Water Sampling

Water was sampled at 11 sites along the river (Vinzants Landing, River Sink,

Ogden Pond, Big Sink, Parener's Sink, Jim Sink, Jug Lake, Hawg, Two Hole,

Sweetwater Lake, and the River Rise) and at 5 wells (Wells 1, 2, 3, 4, 6, and 7) located

within the park (Figure 1-2). Samples were collected between May 2002 and May 2003

at stages ranging from 9.78 to >11.51 masl (Table 2-1).

One objective of this study was to develop simple and robust techniques for

sampling and observing 222Rn activities in karst systems. Consequently, sampling

techniques evolved as the study proceeded. Two types of samples were collected initially

and include grab samples and vertical profile samples. Grab samples were taken

approximately one foot from shore of the various sinks by submerging the tube with the

male connector and loosening the clamp. Vertical profile samples were taken at the River












Table 2-1. Dates, river stages, and sample sites of each sample event.
Sample Periods Sampling Dates River Stage (m) Grab or Peristaltic Samples
I 5/8/02 9.78 Grab Samples
6/6/02 9.61 Grab Samples


9.56

10.12
10.16
10.53
11.49
11.51
>11.51
10.56
10.47


Grab Samples

Grab Samples
Grab Samples
Peristaltic Samples
Peristaltic Samples
Peristaltic Samples
Peristaltic Samples
Peristaltic Samples
Peristaltic Samples


Sample sites
All 11 sites
Vertical profile of River Rise
(1, 2, 4, 6, 8, 10 m depths)
Vertical profile of River Rise (0.3, 1, 2, 4, 6, 10 m)
Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m)
Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m)
All 11 sites
All 11 sites
River Sink, Ogden, Jim, Hawg, Sweetwater, River Rise
River Sink, Ogden, Hawg, Sweetwater, River Rise
River Sink, Ogden, Jim, Hawg, Sweetwater, River Rise
Wells
Vertical profile of River Rise (0.3, 1, 2, 4, 6, 10 m)
Vertical profile of Vinzants Landing (0.3, 1, 2, 4, 6, 8 m)


7/3/02

8/29/02
9/13/02
1/31/03
2/19/03
2/27/03
3/5/03
4/30/03
5/7/03









Rise and Vinzants Landing to depths of 10 and 8 meters, respectively. These samples

were collected using a rubber raft and a rope to secure the raft position. Once in position,

a tube was lowered to various depths, and water was siphoned to the tip of the tube.

After the tube was filled with water, it was attached to the evacuated bottle, which was

allowed to fill with sample water.

Subsequent samples were collected from shore using a peristaltic pump (Geotech

Geopump 2). This sample system included a tube that was extended up to 20 meters

from shore. One end was connected to the pump and the other was weighted and

screened with a woven mesh of polypropylene with a mesh size of 210 Pm. The tube

was set to a 2-meter depth and attached to a long PVC pipe with floats on either side and

pushed from shore. The tubing was purged with 2 liters of water before being connected

to the sample bottle. The peristaltic pump was also used for the vertical profiles. A

weighted rubber tube was lowered from the side of the boat at various depths (0.3, 1, 2, 4,

6, 8, and 10 m) and then connected to the pump, which like the other samples, was

allowed to purge for two minutes before each sample was taken.

Ground Water Sampling

Ground water samples were taken from 6 wells (Well 1, 2, 3, 4, 6, and 7) that are

located within the Park. Completed depth, screened interval, and depth to bedrock are

given in Table 2-2.

Table 2-2. Drilling records of six wells drilled during 2003.
Wells Completed Depth (ft) Screened Interval (ft) Depth to Bedrock (ft)
1 75 75-55 56
2 100 100-80 20
3 93 93-73 10
4 97 97-77 15
6 102 102-82 16
7 98 98-78 18









At each well a Redi-flow 2" variable performance submersible pump was used to

pump water directly into the base of an 8.4-liter bucket, which was allowed to overflow

ensuring no atmospheric interference. The wells were purged with at least three well

volumes before 222Rn samples were collected. The samples were collected by

submerging the tubing to the base of the bucket and allowing the vacuum to fill the

bottle.

Lab Methods

Radon was extracted from the water samples within 24 hours from sample time to

insure a minimal loss of 222Rn to decay. Bottles were attached to the extraction line

(Figure 2-2), clamps were loosened, and ultra pure (99.999992 %) helium was allowed to

bubble into the sample through the 12-inch copper pipe (Figure 2-1). The helium sparged

gases from the water, and all gases flowed from the bottle through the shorter copper

pipe. Gases flowed to a U-shaped collection tube, which is filled with copper filings and

submerged in liquid nitrogen (Figure 2-2). Prior to entering the U tube, water and carbon

dioxide are removed from the carrier gas by Hammond drierite and Thomas ascarite. As

the gases flow through the U tube, the 222Rn freezes to the surfaces of the copper filings

and other remaining gases (e.g. He and 02) escape. After one hour, the liquid nitrogen is

removed, the bottles are disconnected, and the U tube is warmed. Counting cells (Lucas

cells) are attached to the extraction line and helium pushes the trapped 222Rn into the cells

(Figure 2-2). The cells are then allowed to sit for a minimum of three hours for the alpha

particle from the 222Rn decay to allow an ingrowth of daughters (Operations Manual,

1985) (Figure 2-3) and then are placed on the alpha counter. After 222Rn has been

extracted from a sample, it is allowed to sit for a minimum of 7 days to allow regrowth of
















S#2 #3~ Drierite Vacuum/Pressure
T 6. o Gauge
HeliUm Inflow ; CGa
SO Ascarite Vau

01 0 -n

Bubbles
Sample Helium
Bottle LN2 Exhaust
Flask
Cold Lucas
Trap Cell


Figure 2-2. A simplified schematic of the Radon Extraction Line. Each number below
corresponds with the number in the diagram. (J. Cable, 2003 personal
communication)

1. By-pass valve purges line between samples or will be used when evacuating Lucas
cells.
2. Valve controls flow into the sample bottle. Flow is regulated by the Flow meter
(400 ml/min is recommended).
3. Valve controls flow out of sample bottle. Gas from bottle headspace is forced out
of bottle and carried through the line. Flow passes through Drierite to remove
water vapor and Ascarite to remove carbon dioxide. These gases must be removed
to eliminate their freezing inside the LN2 trap and taking up surface area reserved
for radon molecules.
4. Valve controls flow into LN2 cold trap. This valve is open during the sample
processing and is closed after the collection period (about 60 minutes) is complete.
Time series experiments using standards have shown that after 50 minutes 90% of
radon in water is collected on the trap. After 60 min, 99% of radon is collected.
5. Valve controls flow into the Lucas cell. During sample processing, Lucas cell is left
off of the line and helium is allowed to exit the cell port. Valve 5 is open. Radon
and helium are forced into the cell after the trap has been heated and the cell is
evacuated.
6. Valve controls flow to vacuum pump. Left closed during sample processing. It is
opened only when the trap is closed (4 & 5 valves) so that you can evacuate the
Lucas cell.




























Figure 2-3. A simplified schematic of the radon counting system modified from
Operations Manual 1012899A. The voltage travels to the PM tube and gives
off pulses of light with each detected ac particle. The pulse of light is then
sent to the counter.

222Rn from the 226Ra dissolved in the water. The ingrown 222Rn is then extracted and

measured following the techniques for initial 222Rn analysis, which provides a

measurement of 222Rn activity in the sample. Once all samples have been run, the bottles

are emptied to determine water volume.

The error associated with radon analysis are those associated with sample counting,

cell background counting, volume, line efficiency, and operator error. Efficiency of the

extraction line and counter were calculated by running 222Rn standards. The primary

standard used was NIST 268.2 Bq/g (16,092 dpm/g). The standard was diluted into 8

standards in bottles identical to those used for the measured 222Rn samples and were

extracted for 222Rn using the same process as with the measured samples. Each Lucas

cell was used three times. Table 2-3 shows that the average percent efficiency of the









Table 2-3. Average percent efficiency and standard deviation for each Lucas cell.

Counter Red Green Blue Yellow
Cells I II III I II III I II III I II III
Average % 45 41 39 56 44 48 72 70 53 65 80 57
efficiency
Standard 0.08 0.29 0.12 0.2 0.01 0.11 0.04 0.16 0.25 0.39 0.03 0.54
Deviation


extraction line, lucas cells, and alpha counter range from 41 80 % and standard

deviation ranges from 0.01 to 0.54. The precision of the procedure was determined by

running duplicates. Duplicates were taken at Vinzants Landing at a depth of one and six

meters, at the River Rise at a depth of one and six meters, and at Well 2. The percent

difference of the duplicates taken at Vinzants Landing (1m) is 11% and is 64% at

Vinzants Landing (6 m). The percent difference of the duplicates taken at the River Rise

1 m and 6 m are 73% and 18%, respectively. The percent difference of the duplicates of

Well 2 is 29%.

Absorbance

A spectrophotometer measures the color absorbance quantitatively within the

visible spectrum (360-375 nanometers). The instrument used in this study was the Milton

Roy Spectronic 401 spectrophotometer. To get an absorbance value a wavelength was

set on the spectrophotometer that will be maximally absorbed by river water. In this case

the wavelength is 375 nanometers. Water with a higher amount of color due to tannins;

such as river water, will have a higher absorbance value than water with little color

(groundwater). It would be expected that the samples composed of groundwater will have

a higher 222Rn activity and should have the lowest absorbance. Therefore, absorbency

should be a good check for the 222Rn activities of this study.














CHAPTER 3
RESULTS

Precipitation and River Stage

The precipitation and river stage records for the study period are shown in Figures

3-1 and 3-2. A total of 120.45 cm of rainfall fell at the O'Leno station during the study

period (April 1, 2002 May 7, 2003) with 57% of the total (120.45 cm) occurring

between September 2002 and April 2003, or 62% of the study interval.

River stage is controlled by precipitation and evapotranspiration, but river stage

varies seasonally primarily due to evapotranspiration, which ranges from approximately 5

cm/month during the winter to approximately 14 cm/month during the summer (Gordon,

1998). The river stage between April 1, 2002 and May 7, 2003 ranged from a minimum

of 9.48 masl on August 1, 2002 to a maximum of 14.43 masl on March 13, 2003 (Figure

3-2). Between the stages of 9.48 and approximately 10.50 masl, the river is completely

captured by Vinzants Landing, a swallet upstream of the River Sink. Once stage reaches

approximately 12 masl, the river overflows its banks at the south/southwestern portion of

the Sink and continues overland flow in a south/southwestern direction within the Park.

At stage of approximately 14.3 masl the river overflows its banks at the eastern portion of

the Sink and begins to connect to all the other intermediate karst windows and the River

Rise via overland flow (Dean, 1999).

Depth Profiles

Profiles of 222Rn activity with depth were measured at Vinzants Landing and the

River Rise. River Rise profiles are shown on Figure 3-3 A and B and Vinzants Landing









profiles are shown on Figures 3-4 A and B. Figure 3-3 A has two profiles, one taken on

June 5, 2002 and the other on July 2, 2002. On these dates all river water was captured

by the sinkhole at Vinzants Landing. Both profiles have low 222Rn activities, but show

distinctly different trends. The July 2nd profile decreases with depth from a maximum of

-13 dpm/L to a minimum of-2 dpm/L, whereas the June 5th profile is variable at the

surface but increases slightly with depth below the sample taken at 4 m below the

surface. Figure 3-4 B shows the first two profiles along with a third profile taken on May

7, 2003, when the river water was being captured by the River Sink at a stage of 10.43

masl. There is no pattern with depth; activities vary from a minimum of 25 dpm/L to a

maximum of 186.7 dpm/L. This profile also shows no similarity with the first two

profiles.

The first two Vinzants profiles were taken during the summer of 2002 (July 2, 2002

and August 29, 2002) when all river water was captured by the sinkhole at Vinzants

Landing (Figure 3-4 A). The profile taken on July 2nd shows the highest activity (32.86

dpm/L) at a depth of 0.3 m with a minimum at a depth of 4.0 m. The August 29th profile

has the highest activity (30.37 dpm/L) at a depth of 0.3 m, but the activity changes with

depth less than on July 2nd. Figure 3-4 B. shows the first two profiles as well as a third

profile taken on May 7, 2003. The May 7th profile has the highest activity (393 dpm/L) at

depths of 0.3 meters and passes through a minimum between 2 and 6 meters below the

water surface.
















Santa Fe River Stage


VI
V 3/5/03
2/27/03


7/28/2001 11/5/2001 2/13/2002 5/24/2002 9/1/2002 12/10/2002 3/20/2003 6/28/2003


Individual Site Sample Times


Depth Profile Sample Times


Figure 3-1. The precipitation record for O'Leno State Park from June 2001 to July 2003, which encompasses the entire study period.
The red arrows indicate when individual sinks were sampled and the green arrows indicate when the depth profile sampling
was performed. The roman numerals indicate the sample periods (Table 2-1).


9 -/
4/19/2001
















O'Leno State Park Precipitation


II IV
9/12/T2 2/19/03


7/28/2001 11/5/2001 2/13/2002 5/24/2002 9/1/2002 12/1012002 3/20/2003 6/28/2003


SIndividual Site Sample Times


Depth Profile Sample Times


Figure 3-2. The Santa Fe River Stage recorded from O'Leno State Park from mid April 2001 till the end of June 2003. The red arrows
indicate when individual sites were sampled and the green arrows indicate when the depth profile samples were performed.
The roman numerals indicate the sample periods (Table 2-1).


V
2/27/03
VI
13/5/03


0 -
4/19/2001
















A
Excess 222Rn Activity (dpm/L)
0 2 4 6 8 10 12


B
222
Excess 2Rn Activity (dpm/L)
50 100 150


0



2



4



6



8



0


Depth (m)


Depth (m)









1F



1:





Figure 3-3


I I I I IL Z
G 6/5/02 675/02
S6/5/02
7/02/02
-- 7/02/02 -o 5/7/03

Vertical depth profiles at the River Rise. A. Depth profiles of the River Rise taken when the river was captured by the
sinkhole at Vinzants Landing. B. The first two profiles as seen in graph A and the profile taken in May 2003 when the
river was flowing to the River Sink. The values at 1 m and 6 m depths are averages of duplicate samples. Note the scale
change between profile A and B.


-
- - -
















2 A
Excess Rn Activity (dpm/L)


0 5 10 15 20 25 30 35


Depth (m)


222
Excess Rn Activity (dpm/L)

0 50 100 150 200 250 300 350 400


I I










\\
i \ i

\\
- ------------- -------


--3 ------------- 10 .82--


- --........' 103.82


i I
/-
28.77-0-28.77
























I I-
- -- ------,- ---- --- ---- ---,-------












103.82----------------------------





--. -^--------


-e- 7/2/02
e 8/29/02
-B-- 8/29/02

7/2/02 -1 5/7/03

Figure 3-4. Vertical depth profiles at Vinzants Landing. A. The depth profiles of Vinzants Landing when it captured the river. B.
The first two profiles as seen in graph A versus the profile taken in May 2003 when the river was flowing to the River
Sink. The values at 1 m and 6 m depths are averages. Note the scale change between profile A and B.


Depth (m)


2 ---*-----






4





6





8









This profile also has no pattern with depth and shows no similarity with the first two

profiles

Excess 222Rn, Precipitation, and River Stage

Sample Periods I to VI

Excess 222Rn (dpm/L) records for the first two sample periods (May 8, 2002 and

September 12, 2002) are shown in Figure 3-5 A, when river stage was 9.78 and 10.16

masl, respectively. Little to no precipitation had occurred in this area directly preceding

the sampling events and all activities for both dates are less than 50 dpm/L. River stage

was sufficiently low at these times that all river water flowed into Vinzants Landing and

none into the River Sink.

Activities measured for sample period III (January 30, 2003) are shown in Figure

3-5 B. Approximately 40.21 cm of rain fell between September 12, 2002 and January 30,

2003, causing the river stage to rise to 10.53 masl (Figure 3-1, 3-2). Most of the 222Rn

activities sampled on January 30, 2003 are higher than the first two sample periods, with

the highest activity of 295.0 dpm/L at Hawg Sink. Ogden Pond, Jug Sink, and Two Hole

have activities similar to the first two sample periods.

By sample period IV (February 19, 2003) the river stage reached 11.49 masl and all

activities returned to values less than or equal to 50 dpm/L (Figure 3-5 C). This trend of

decreasing 222Rn activities with increasing stage continued to sampling period five

(February 27, 2003) where, at a stage of 11.51 masl, all locations, with an exception of

Hawg Sink have activities less than 25 dpm/L. At that location, the 222Rn activity was

97.5 dpm/L but this value also represents a decrease in activity of 33 % from its peak

activity of 295 dpm/L on January 30, 2003 (Figure 3-5 C). By March 5, 2003, sample

period six, the river water had risen over the lower of two staff gages and overflowed its


















































































Figure 3-5. 222Rn activities of samples from karst windows, the River Sink, and the River
Rise versus distance along the flow path for sample dates of A. 5/8/02 and
9/12/02, B. 1/30/03, and C. 2/19/03, 2/26/03, and 3/5/03.


0 5/8/02 -- 9/12/02
150

125 -
150 ~ ~ ~ ~ --------- ---------------------------



75 ------------- ------------- ------- A
SParener's Hawg
50 A---
s o ... -Og i--.... -- ------- -- T 6w H 0 ole- -
5 /^ Jim Jug l
25 --- ... Rise .

0o Big Sweetwater
0 2000 4000 6000 8000
Distance (m)


150
1250 ---------------- ---------- ------------------ --------------f-~

Sweetwater
E Parener's
a 100 ------
Big
c i h Jim l
S 75 -. ........ / ........... ...- .- ........ --- B





Ogden Jug Two Hole
o i------------------i-
0
0 2000 4000 6000 8000
Distance (m)


150

125
-j
E 100
-o

Qr 75

S 50
o

25

0


-
0 2000 4000 6000 8000
Distance (m)










banks. At this time, most sample locations have activities less than or equal to 25 dpm/L

(Figure 3-5 C). The relationship between the average, standard deviation, and coefficient

of variation for samples collected during sample periods 1-6 are shown in Table 3-1. The

averages shown on Table 3-1 do not include Hawg Sink or Ogden Lake for 2/19/03. At

that time Ogden Lake had an activity of 0.09 dpm/L, which is likely in error. The

different activities at Hawg Sink suggest that it may not be on a similar flow path to the

other karst windows.

Table 3-1. Average, standard deviation, and coefficient of variation for each sample
time.
Sample Time Number of samples Average (dpm/L) Standard Devation Coefficient of Variation
1 9 22.5 8.7 0.4
2 11 18.3 8.4 0.5
3 9 60.4 35.1 0.6
4 5 26.2 19.5 0.7
5 3 14.6 9.8 0.7
6 5 11.3 5.3 0.5

Well Samples

During the course of this study seven wells were drilled in locations throughout the

field area (Figure 1-2). Six of the wells were sampled on April 30, 2003 and 222Rn

activities are shown on Figure 3-6. Wells 1, 2, and 3 have 222Rn activities ranging from

1172 dpm/L to 732 dpm/L and are either located in the northern portion of the park

(Wells 1 & 2) or up gradient (northeast) of the mapped conduit (Well 3). Wells 4, 6, and

7 have activities less than wells 1, 2, and 3 and range from 432 to 152 dpm/L. Wells 4, 6,

and 7 are located in the southern portion of the study area. Well 3 is also located in the

southern portion of the study area, but it is located up gradient from the other southern

wells and the mapped conduit.










North and/or East of
Conduit
1400

1200
S Wll 1 Well 3
1000
E
80 South and West of
*-' 800 -
S0 Conduit
600 \ Well 2 (avg)



200


0 1 2 3 4 5 6 7 8
Wells
Figure 3-6. 222Rn activities of wells 1, 2, 3, 4, 6, and 7. A duplicate was taken at well 2.
The activity plotted is the average of the two samples taken with a range of
213 dpm/L.

Mixing Model

Mixing between the Santa Fe River and the Floridan Aquifer can be estimated by

modeling the quantities of each water type in the samples. Assuming there is only two

end member mixing with distinct 222Rn signatures, the amount of river water in the

samples (%X) can be calculated by a mixing model (Kincaid, 1998), described by

%X = ((Rs Raq)/(Rriv Raq)) 100 (1)


where Rs represents the activity of 222exRn measured in each sample, Raq is the activity of

222exRn in the groundwater, which is determined by the average well activities of wells 1,

2, and 3, and Rriv is the amount of 222exRn in the river. The River Sink has large changes

in activity due to the variable input of ground and surface water in the upper Santa Fe

River. Therefore, the 222exRn activities for the River Sink during each sample period

were used for Rriv. This model has been used to calculate percentages of river water in










samples taken from the River Sink, Rise, and the intermediate karst windows. Figure 3-7

shows the fraction of river water for each site with respect to river stage according to

values calculated using equation (1). Sample periods I and II have approximately 100%

river water in each sample. Sample period III shows more variability; the X% ranges

from 60 to 100%. Sample period IV, V, and VI have approximately 100% river water in

each sample. Because the River Sink 222Rn activities were used as the Rriv variable all

%X values for the River Sink are 100% and are not reported in Figure 3-7.




9.5 10 10.5 11 11.5 12 12.5
105


-100 Ogden
95 ---------------..............-----...... -" .. ------------ ................ -Og den
\- Big
90 ............... -o Pareners
%X x-- Jim
85 Jug
\ /i : Hawg
80 Two Hole
5 / -- Sweetwater
-- Rise
70

65
9.5 10 10.5 11 11.5 12 12.5
Stage (m)

Figure 3-7. %X versus Stage (m) for each site in the study area. Lower the %X
represents higher the groundwater content in the samples.

Decay Equation

Sink vs. Rise

The amount of decay that takes place over a period of time can be determined in the

basic radioactive decay equation described by

N =Noe Xt (2)









N is the activity of parent atoms that remain after a certain travel time (t), No is the

original activity of radioactive parent atom, and ) is the decay constant, which is 0.263

day'1 for 222Rn. Rearranging the equation allows calculation of travel time from

calculated and measured 222Rn activities.

t = ln(N) In( No) (3)
-h
Table 3-2 shows how travel time, decay, and the measured activities of the

samples taken at the River Sink and Rise interact. Martin (2003) was able to calculate

velocities and travel time by correlating temperature peaks as water flowed through the

Sink/Rise system. Travel time (t) was based on these travel times and velocities given in

Martin (2003) for 2/19/03, 2/23/03, and 3/3/03, which are compared to sample periods

2/19/03, 2/26/03, and 3/5/03 of this study. No water was entering the River Sink during

the first two sample periods (5/8/02 and 9/12/02); therefore, it is possible that there was

no flow from the Sink to the Rise. Because of this, travel times cannot be calculated for

these two dates. Martin (2003) did not have velocities calculated for a date close to

sample period III (1/30/03) therefore the travel time here is from Dean (1999).

Calculated activity for the River Rise (222exRn Rise calc) was found by using equation (2),

assuming the initial 222Rn activity (No) is the activity at the River Sink. Calculated travel

time (T222exRn) is the travel time found by using equation (3), where No is the 222Rn

activity of the River Sink, and N is the 222Rn of the River Rise.

The measured activity of the Rise sample collected January 30, 2003 is higher than

the calculated activity. For example, the measured activity of the water at the River Rise

was 53.50 dpm/L, but the calculated value was 17.0 dpm/L. The last three sample









Table 3-2. Relationship between measured 222Rn-activity and travel time and calculated
222Rn-activity and travel time using the decay equation.
Date Stage (m) exRn Sink exRn Rise2 ,exRn Rise calc Tt4 T 2exRn
(dpm/L) (dpm/L) (dpm/L) (days) (days)
5/8/02 9.78 22.55 10.6 ---- ---- 2.9
9/12/02 10.16 18.32 8.54 ---- ---- 2.93
1/30/03 10.53 48.6 53.5 17.00 4 -
2/19/03 11.49 20.32 4.45 13.70 1.5 5.78
2/26/03 11.51 24.25 4.62 16.34 1.5 6.31
3/5/03 >11.51 17.94 4.54 13.8 1 5.25
1 & 2 Measured 222Rn activities of the River Sink and Rise, respectively.
3 N calculated by the equation 2 using 1 as No and 4 at travel time.
4 Travel time from Martin (2003) based on river stage during the sample times.
5 Travel time calculated by equation 3, using 1 and 2 as No and N, respectively.


periods (February 19, 2003, February 26, 2003, and March 5, 2003) have a different

relationship than the first three sample periods between the measured and calculated

222exRn activities for the River Rise. The measured activities are lower than the

calculated values. For example, the measured activity of the Rise on February 19, 2003

was 4.45 dpm/L and the calculated activity was 13.7 dpm/L. Travel times for each

sample period were also calculated using the decay equation. However, the calculated

travel times do not correspond with the travel times from Dean (1999) and Martin (2003)

and are not used in any other calculations or discussion.

Sink vs. Karst Windows

A 222Rn value for each site was calculated by using equation (1) and the value of

222Rn activity at the River Sink. These values are reported in Appendix Table A-i, which

gives a comparison of measured and calculated 222Rn activity, 226Ra activity, and travel

time between the River Sink and all other sites. Martin (2003) only calculated travel

times for the River Sink, Parener's, Two Hole, Sweetwater, and the Rise. Travel times











for the other sites (Ogden, Big, Jim, Jug, and Hawg) were calculated from velocities


reported by Martin (2003)


Radon-222 activity increased from the River Sink to most karst windows on May 8,


2002, September 12, 2002, and January 30, 2003 and decreased from the River Sink to


most karst windows on February 19, 2003, February 26, 2003, and March 5, 2003.


Locations where this trend is not followed include Ogden Pond, Parener's Sink and Hawg


Sink. Activity decreased between the River Sink and Ogden Pond for all sample periods.


Parener's Sink shows a increase for all period. Hawg Sink has higher 222Rn activity for


all sample times, except February 19, 2003 (Figure 3-8).


S-awgl Sweetwater
80 --- --
Si 5/8/02, 9 78 m
Paraners /
60 ----I / ---- ---...----- ------ -. 9/12/02, 10 16 m
E Big Jim
Ris -- 1/30/03, 10 53 m
40---- -------
40 / Jug I -y --x--2/19/03, 11 49 m
/ ./
S gden / + 2/26/03, 11 51 m
20 1 \ ^ ^ ^ ^
S6' 3/5/03, > 11 51 m
0 --

oTwo Hole
o -20 i I i I
-40
0 1000 2000 3000 4000 5000 6000 7000 8000
Distance (m)
Figure 3-8. The gain or loss in activity downstream from the River Sink. If there is a
gain in activity then the measured activity at the site is higher than the
measured activity at the Sink. If there is a loss in activity the measured
activity at the site is lower than the measured activity at the Sink. Activity at
Hawg Sink for January 30,2003 sample time is 295 dpm/L.









Gas Exchange Equation

To determine the amount of 222Rn that is lost to the atmosphere by evasion,

Elsinger and Moore (1983) and Ellins et al (1990) used the gas exchange equation.



Cd = Cu e-[D/(zhv)] x (4)
C" and Cd are the 222Rn activities up and downstream. D is the molecular diffusivity of

222Rn, which is 1.2 x10-9 m2/s at 23 C and h, v, and x are the average stream depth,

velocity and distance between sample locations, respectively. The variable z is the

thickness of stagnant film layer at the surface of the stream. Elsinger and Moore (1983)

found the thickness of the stagnant film of the Pee Dee River in South Carolina to be

between 19 and 48 pm. Within this range, the thickness of the stagnant film layer does

not significantly effect the calculations (< 1%) therefore the average of this (33.5 pm) is

used for the current study. However, if the thickness of the stagnant film decreases to

lower than 19 am, the calculations are effected. Even though this equation is not for

karst environments, it was used to establish a first order estimate of the amount of

evasion occurring in the Sink/Rise system. This equation assumes that both stream and

the air above it constitute two well mixed reservoirs with uniform vertical activities

separated by a stagnant film of water (Ellins, 1990). The amount of exchange is mainly

governed by flow generated turbulence (Ellins, 1990); however, the thicker the film the

slower the rate of transfer.

Color Absorbance vs. Activity

Figure 3-9 and 3-10 suggest that the initial hypothesis of absorbance may not be

correct. Figure 3-9 shows activity vs. absorbance values for the River Sink in samples

collected on 1/30/03, 2/19/03, and 3/5/03. Contrary to expectations, the sample with the









lowest absorbance has the highest activity. In Figure 3-10, the sample with the highest

activity has the lowest absorbance; however, the sample with the second highest activity

also has the highest absorbance. The relationship between absorbance and stage can be

seen in Figures 3-11A, B, and C. Figure 3-11 A. gives averaged absorbance values for

each site versus stage and Figures 3-11 B and 3-11 C. plot Ogden and Hawg sinks versus

stage. R2 values for the regression curves on Figure 3-11 A, B, and C are 0.888, 0.854,

and 0.9589, respectively. All graphs show an increase in absorbance with the increase of

stage.










Sink

0.6
3 0.5
0.4
0.3
0
a 0.2
< 0.1
0
0 20 40 60
Activity (dpm/L)


Figure 3-9. 222Rn activities of the River Sink at various sample periods vs. Absorbance.
The sample with the highest activity has the lowest absorbance.


3/05/03

0.5
8 0.48
W 0.46
o 0.44
0.42
0.4
0 5 10 15 20 25
Activity (dpm/L)


Figure 3-10. 222Rn activities of samples taken at Sweetwater Lake, Jim Sink, Ogden
Pond, the River Sink, and Hawg Sink on 3/5/03.











River Stage vs Absorbence


10 10.5
y = 0.11 9x 0.9755
R2 = 0.8884


y= 0.128x 1.0834 8-3e4
Ogdel
R2 = 0.854

0.5
o 0.4
0.3
o
0.2
< 0.1
0
9 10


11 11.5 12
Stage (ft)


11 12 13


Stage (m)

y = 0.2467x- 2.5052 Hawg
R2 = 0.9589

0.5
0.4
0.3
0 0.2
0.1 C
0
9 10 11 12 13
Stage (m)


Figure 3-11. Absorbance values taken during the study. A. Average absorbance for all
the samples collected. B. Absorbance values for Ogden Pond. C.
Absorbance values for Hawg Sink. Each graph covers samples collected on
1/30/03, 2/19/03, 2/26/03, and 3/5/03.


12.5














CHAPTER 4
DISCUSSION

Precipitation and River Stage

River stage varies little during the summer months due to low amounts of rainfall

and high evapotranspiration rates (- 14 cm/month) (e.g. Gordon, 1998). However, low

evapotranspiration rates (- 5 cm/month) during the fall and winter months cause higher

stage fluctuations than the summer, particularly during associated times of increased

precipitation and passage of cold fronts (e.g. Gordon, 1998). During the winter months of

this study, river stage increased rapidly with rainfall: between March 3-7, 2003, 2.27 cm

of rainfall caused an increase of stage by 1.24 meters (Figure 3-land 3-2). In all of June

2002, 11.07 cm of rain fell; however, evapotranspiration caused river stage to decrease by

0.03 m during this time.

Depth Profiles

Assuming that surface water loses 222Rn to the atmosphere and sources of 222Rn are

from the Hawthorn group and solid material in the Floridan Aquifer, there should be an

increase of 222Rn with depth in the sinkholes. However, depth profiles in this study

indicate that the water has heterogeneous 222Rn activities with depth.

All profiles taken from the River Rise and Vinzants Landing reflect heterogeneous

activities with depth that may reflect mixing between river water and groundwater

(Figures 3-3 and 3-4). The main difference between the profiles shown in Figures 3-3

and 3-4 is that Vinzants Landing has its highest activities at the surface. The lithology of

the sinkholes should not have an impact over the activity; however, this is unknown. It is









possible that the high activities seen at both sites are due to higher amounts of sediment

mixing with the waters. Chung (1973) and Berelson et al (1982) indicate that the increase

in 222Rn activity with depth (which should be seen in depth profiles) is due to fluxes of

sediment caused by turbidity currents within the water column. 226Ra is bonded with the

sediment and the decay to 222Rn causes the increase in activity. Both of these studies are

conducted in coastal environments in the Santa Barbara and San Nicholas basins,

California. Even though the environments differ from the current study, they provide an

analog for what could occur at Vinzants Landing and the River Rise.

Excess 222Rn

Three capturing processes control the 222Rn activity in the karst windows of the

Santa Fe River: mixing of low activity surface water and 222Rn rich ground water,

atmospheric evasion, and radioactive decay. All three processes depend on the flow

through the Sink/Rise System. During the first two sample periods, all water in the Santa

Fe River was captured by the sinkhole at Vinzants Landing. The river stages for these

two dates are 9.78 masl and 10.16 masl, respectively. Martin and Screaton (2002)

suggest that during low flow conditions, the conduit will have a lower head than the

matrix and will act as a drain for the surrounding matrix porosity (Figure 4-1 A). With

no water flowing into the River Sink, groundwater should be the main influence over the

222Rn activities measured from the samples taken at these times and even during 5/8/02

and 9/12/02 water seemed to be resurging at the River Rise (i.e., the River Rise was not

dry). Therefore the conduit should have water with high 222Rn activity that may approach

values found at the wells (-200 to -1200 dpm/L).

Flow rates are slow at low stage conditions (-10.00 masl), with travel times from

the River Sink to the Rise of approximately 8 days or longer (Dean, 1999; Martin &









Dean, 1999). However, if there is no flow entering the River Sink, then travel times are

indeterminable. Slow to no flow in the Sink/Rise system, combined with 222Rn's short

half-life (3.84 days), indicates that radioactive decay will reduce the activities. This also

allows longer time for evasion of 222Rn to the atmosphere. Figure 4-2 indicates that more

evasion does occur during slow flow, but not enough to account for the low activity.

Rainfall prior to sample period I and II is minimal therefore there is no dilution caused by

surface water recharge. Therefore, it is possible that the diffuse flow from the matrix to

the conduit is slow enough that significant decay occurs. This along with other factors

such as radioactive decay within the conduit and evasion probably account for the low

activities during these sample periods.

Prior to sample period III, 18.8 cm of rain fell and caused the river stage to rise to

10.53 masl, which is considered average river stage based on the hydrograph of river

stage for the study period (Figure 3-2). Some 222Rn activities are higher than seen during

the first two sample periods. Travel time from the Sink to the Rise is approximately four

days on the basis of the stage vs. travel time relationship in Dean (1999). More rapid

travel time may prevent most evasion and decay allowing 222Rn activities to stay

elevated. For example, calculated 222Rn activity (N from equation 2) for Ogden Pond

(Appendix Table A-i) during sample period III is 48.3 dpm/L, but activity from the River

Sink (No) is 48.6 dpm/L. Also, results from the gas exchange equation (Appendix Table

A-3) show that at Ogden Pond only 0.17 dpm/L was lost to evasion. Both of these results

indicate that the quicker travel time creates less loss of activity due to radioactive decay

and evasion. Another possible explanation for the elevated 222Rn activities is that an

increase in diffuse flow from the rain has created a more rapid influx of 222Rn rich water







43


Evasion


Sample Periods I and II


Santa Fe River Water



Matrix Water

Mixing


Sample Period III


Santa Fe River Water


Sample Periods IV, V, and VI


Figure 4-1. A conceptual model of interactions between the Floridan Aquifer and water
in conduits during the sample periods. A. No water is entering the River Sink;
therefore, the matrix water enters the conduits. Since there is little flow
through the Sink/Rise system at this time, the water in the conduits decays. B.
River water is now flowing to the River Sink and water from the matrix is
entering the conduit and mixing with the river water. There is flow in the
system and the decay is minimal. C. The river is flooding and the abundant
amount of river water in the conduit is now moving to the matrix.











Evasion vs. Travel Time


5.00
S4.00
1 3.00
0 2.00
> 1.00
w
n nn


9.78
-in 10.16



11.51

>11 .51


0.00 2.00 4.00 6.00 8.00 10.00
y= 0.4965x+ 0.1155
R2 = 0.9679 Travel Time (days)

Figure 4-2. Evasion at the River Rise increases as travel time increases. The numbers by
the data points indicate river stage at the various sample times.

(ground water) to flow into the conduit and mix with the incoming river water (Figure 4-

1 b). A similar flow of groundwater to conduit was observed using C1 concentrations in

the water. At stages of 10.45 to 10.70 masl, Dean (1999) saw an increase in Cf by 24.7%

to 43.2 % from the River Sink to Sweetwater Lake, which he interpreted to indicate that

ground water flowed to the conduits. Compared to the first two sample periods, the

current study reports an increase in 222Rn during this stage of greater than or equal to 80,

this is probably due to the activity being lower than expected during the first two periods.

An increase in diffuse recharge should cause an increase in flow from the matrix to

the conduit; therefore, most sites should show a higher activity than the first two sample

periods. However, Ogden Pond, Jug Lake, and Two Hole Sink have activities similar to

the first two sample periods. It is unclear as to why Ogden Pond, Jug Lake, and Two Hole

Sink differ from the other samples. Of all the karst windows, Hawg Sink has the highest

222Rn activity at 295.0 dpm/L. A study conducted by Sprouse (2004, in process) reports

calcium concentrations of approximately 60 mg/L and an alkalinity of -140 mg/L for









Hawg Sink in January 2003. Compared to the chemistry of the other sinks, these

concentrations are high and are possible indicators of ground water input. Along with the

high 222Rn activity measured at Hawg Sink, this may indicate that it is not on the same

flow path as the other sinks and could represent an unmapped groundwater source.

During sample periods IV, V, and VI, the river stage was 11.49, 11.51, and >11.51

masl, respectively. During sample period VI, elevation of the Santa Fe River was greater

than the staff gage located at the River Sink preventing stage measurements and an exact

stage value. At this time, it was observed that the river overflowed the banks on the

southwestern portion of the Sink and began to flow over land. The travel times from the

Sink to the Rise are 1.3 days for 2/19/03 and 2/26/03 and 1 day for 3/5/03 (Martin, 2003).

These rapid travel times prevent a large loss of activity due to radioactive decay and

evasion. For example, the calculated activity for the River Rise is 4 dpm/L less than

the River Sink (Table 3-2) and the gas exchange equation indicates that only 0.5 dpm/L

was lost to evasion.

A possible cause for the observed low activities is dilution by rainwater. Two days

prior to sampling on February 19th, 7.57 cm of rain fell, 1.06 cm of rain fell the week

before sampling on February 26th, and 3.4 cm fell the week prior to March 5th. It takes

15 days for secular equilibrium between 226Ra and 222Rn to be reached. If the

surrounding sediments have been previously flushed with water, then a pulse of water

within 15 days of the first pulse, that flushed the system, will have less 222Rn activity.

However, depending on the timing of the water pulses, it is still possible that rainwater

would gain 222Rn activity as it moves through the sediment column and a dilution effect

from the rainwater and groundwater mixing in the matrix would not be seen. It is most









likely that the low activities seen at the various sinks during the sample periods are due to

dilute 222Rn-activity water entering the conduit at the River Sink. Based on C1

concentrations, Dean (1999) reported that at approximately 11.9 masl most of the water

at the River Rise originated from the River Sink, suggesting that there is little loss of

water from matrix porosity to the conduit during high flow. Martin (2003) supported this

conclusion and suggested that heads are higher in the conduit then in the matrix when

water is leaving the conduit and flowing to the matrix. Therefore, the river water in the

conduit is flowing out into the matrix (Martin and Screaton, 2002) (Figure 4-1 c). Color

absorbance values also indicate that the water with in the karst windows is consistent

with an origination at the River Sink. At most of the karst windows, there is an increase

in absorbance with an increase in stage (Figures 3-11, 12,and 13), which suggests that

higher flow rates flush the ground water from the conduit.

Radioactive Decay Along the Flow Path

At times of low flow, groundwater input and radioactive decay should be the main

controls over the 222Rn activities of the karst windows. The low 222Rn activities in

samples collected at the River Rise on May 8, 2002 and September 12, 2002 (Figure 3-4

A) indicate that radioactive decay and evasion control the water's activity. Samples

collected January 30, 2003 have higher activities and a quick travel; therefore the effects

of decay and evasion may not be as strong.

Radon-222 activities calculated from equation (2) are not corrected for atmospheric

evasion. However, results of equation (4) indicate that approximately 4.5 dpm/L and 3.6

dpm/L are lost by evasion for May 8th and September 12th, respectively and only 2.5

dpm/L is lost for January 30th (Appendix Table A-3). For each of these dates, the

calculated 222Rn activities for the River Rise from equation (2) are approximately twice









the measured activities, which indicate that the measured activities are higher than if

decay was the only influence over the water 222Rn activity (Table 3-2). Therefore, there

is some source of 222Rn to the water (i.e., ground water).

Samples taken from the River Rise on February 19, 2003, February 27, 2003, and

March 5, 2003 all have measured 222Rn activities that are lower than calculated activities,

(Table 3-2, and Figure 3-8). Travel times during these sample periods are shorter (1-1.5

days) than the half-life of 222Rn (3.84 days) suggesting that a lesser amount of decay will

occur. The calculated activities for the River Rise are not much lower than the activities

at the River Sink, indicating that not much decay has occurred. For example, the

calculated activities at the River Rise for 2/19/03, 2/26/03, and 3/5/03 are all

approximately 5 dpm/L less than the activities at the River Sink. The amount of evasion

for 2/19/03, 2/26/03, and 3/5/03 was calculated to be 0.75, 0.71, and 0.48 dpm/L,

respectively and the amounts of evasion during 5/8/02, 9/12/02, and 1/30/03 are 4.45,

3.62, and 2.45 dpm/L, respectively. According to the Gas Exchange equation, this

indicates the quicker travel times also prevent most evasion. Figure 4-2 indicates that the

amount of evasion decreases with increasing travel time. The low 222Rn-activity river

water is not mixing with the high 222Rn-activity groundwater within the conduit but

moving out into the matrix. Other processes thus appear to control the loss of 222Rn from

the conduit.

If the loss of radon from the River Sink to the River Rise is not due to evasion what

is causing this? It is possible that the water in the conduit takes a longer time to travel

through the system than the temperature tracking indicates. Alternatively, it is possible

that the loss in activity is from mixing with a low 222Rn activity water source. Old









Bellamy Cave Exploration Team have mapped a main conduit which enters the system

from the east (Old Bellamy Cave Exploration Team, unpublished report); however, there

are no known surface water sources for the eastern system, suggests that it does not

supply dilute water. With the large amount of precipitation (> 20 cm) that occurred

during the later sample periods it is possible that the loss of activity is caused by a influx

of low activity surface water that travels to the conduit.

With a few exceptions (Ogden Pond, Parener's Sink, and Hawg Sink) the other

sample sites used in this study agree with the 222Rn activities at the Rise. Ogden Sink has

calculated activities higher than the measured activities during all sample periods but

May 8, 2002. This could be caused either by less mixing with radon-rich ground water,

more degassing, or a combination of both. Parener's Sink shows a decrease in activity on

9/12/02; however, it is so slight (-0.5 dpm/L) that it is smaller than analytical error of the

measurement. Hawg Sink shows an increase in activity from the River Sink during every

sample period but February 19, 2003 indicating a different source of groundwater not

available to the other karst windows.

Distribution of 222Rn in the Groundwater

Regional groundwater flow is from the northeast to the southwest (Miller, 1997);

therefore the wells on the eastern side of the conduit should sample groundwater that is

not influenced by water from the conduits. Martin (2003) indicates that local water flow

is towards the conduit during base flow conditions. Samples from the wells were

collected at a stage of 11.05 masl. Water collected from wells 1, 2, and 3 have a higher

activity than wells 4, 6, and 7. Well 1 and 3 are located in the eastern portion of the study

area and the measured activity of 1173 and 1134 dpm/L, respectively, suggests that they

are influenced by the groundwater. Even though Well 2 is on the western side of the









conduit, the high activity (733 dpm/L), which is measured, indicates that it is also

influence by the matrix water. Wells 1 and 2 are located in the northern portion of the

park, in which they are in closer proximity to the Hawthorn Group, which is a source of

222Rn. Wells 4, 6, and 7 are in the southwestern portion of the park, just west of the River

Rise. The low activities measured at these wells (152, 432, and 294 dpm/L, respectively)

indicate that may be influenced by the mixing between the river and ground water within

the matrix.

It is uncertain as to whether the Hawthorn provides more 222Rn activity to the area

then the limestone within the Floridan Aquifer. Crandall (1996) found that surficial

aquifer water influenced by the Hawthorn Group had 222Rn activities of> 1200 dpm/L.

Smoak et al. (2000) found 226Ra activities for the Floridan to be approximately 6 dpm/L.

Wells 1, 2, and 3 have similar 226Ra activities, which indicates that the 222Rn activities of

the water measured in Smoack et al (2000) might also be relatively high. Well 1, 2, and 3

activities are also similar to those found in Crandall (1996). Therefore, at this point the

222Rn activities provided by the Hawthorn Group and the 222Rn activities provided by the

Floridan Aquifer limestone are indistinguishable.

Well activities were used as an end member (Raq) in the mixing equation. Kincaid

(1998) was able to use this model successfully and found the amount of river water to be

between 50 and 90 %. The %X, in this study, range from approximately 60 to > 100.

Figure 3-7 suggests that during sample periods I, II, IV, V, and VI primarily river water is

in the conduit and that during sample period III water in the karst windows is influenced

by ground water. However, previously discussed results suggest that this is not the case.

Water chemistry data from Dean (1999) indicated that water within the karst windows









was influenced by ground water. The 222Rn activities during sample periods I and II are

low, which is the result of radioactive decay and atmospheric evasion, not an influx of

river water. Therefore, in a future study of the Sink/Rise system more end members need

to be defined and quantified than what was provided by equation (1). Such end members

include influx of radon from the surrounding sediment and the influx of radon from the

Hawthorn Group. It is also important to use a variety of tracers, each with distinct

geochemical behavior, such as 180, C-, S04-2, as well as 2Rn, to better constrain

surface and ground water mixing.














CHAPTER 5
CONCLUSIONS

Karst aquifers are important hydrologic systems and provide potable water to most

of our world. Therefore, it is necessary to understand the hydrologic characteristics of

these aquifers. One characteristic is the abundance of sinking streams, which allow for

mixing between the surface and ground water. These streams have potential to carry

pollutants that, once mixed with the potable groundwater, could spread quickly and harm

thousands of individuals. Natural tracers that record mixing of surface and ground water

would be valuable to developing an understanding of mechanisms and quantities of

mixing.

Results from this study suggest that 222Rn activity varies with stage. During low

stage, the low 222Rn activities suggest a loss of activity to decay and evasion. In the

future, to determine the exact amount of evasion that is occurring the stagnant film

thickness needs to be corrected for the Santa Fe Sink/Rise system. During base flow, the

high 222Rn activities suggest mixing between the river and ground waters within the

conduit. During high stage, low activities indicate there is little to no mixing between the

river and ground waters; river water is flowing from the conduit to the matrix. The

results also indicate that Hawg Sink may not be on the main flow path.

Furthermore, results indicate that a two end-member-mixing model is not adequate

to quantify the amounts of water involved in the mixing. In the future, when modeling

the quantities of water involved in mixing, other end members, such as, decay, evasion,

and sediment input need to be defined. The effects of lithology on the 222Rn activities






52


cannot be determined at this time. This is something that should be resolved in the

future. Other future work involves more depth profiles at other locations, as well as, the

River Rise and Vinzants Landing to fully understand vertical mixing with depth. Radon-

222 may prove to be a useful tracer of mixing between surface and ground water;

however, other tracers, such as CF and 6180 should be used to support the results from

the measured 222Rn activities and further analysis of the complex karst system, the Santa

Fe River Sink/Rise system needs to continue.















APPENDIX
MEASURED AND CALCULATED 222RN, TOTAL 226RA, TRAVEL TIME, AND
EVASION FOR ALL SITES

Table A-1. Travel time, measured and calculated 222Rn, and total 226Ra
Location Date Stage Travel Time 222exRn calc ** 222exRn ** 22Ra
(m) (days) (dpm/L) (dpm/L) (dpm/L)
River Sink 5/8/02 9.78 -- -- 38.21 -0.25


9/12/02
01/31/03
02/19/03
2/26/03
3/5/03


Ogden 5/8/02
9/12/02
01/31/03
02/19/03
2/26/03
3/5/03


Big


5/8/02
9/12/02
01/31/03
02/19/03
2/26/03
3/5/03


Parener's 5/8/02
9/12/02
01/31/03
02/19/03
2/26/03
3/5/03


Jim


5/8/02
9/12/02
01/31/03
02/19/03
2/26/03
3/5/03


10.16
10.53
11.49
11.51
<11.51

9.78
10.16
10.53
11.49
11.51
<11.51

9.78
10.16
10.53
11.49
11.51
<11.51

9.78
10.16
10.53
11.49
11.51
<11.51

9.78
10.16
10.53
11.49
11.51
<11.51


14.21
12.64
14.28
0.09
14.8


>0.24
0.24
0.126
0.025
0.025
< 0.025

> 1.16
1.16
0.61
0.119
0.119
<0.119

>1.45
1.45
0.756
0.15
0.15
<0.15

> 1.71
1.71
0.89
0.25
0.25
<0.25


>24.17
17.73
48.3
20.18
24.08
<17.81

>16.62
13.5
41.41





>15.42
12.53
39.8





>14.38
11.68
38.61
21.8

<16.73


22.55
18.32
48.6
20.32
24.25
17.94

18.43
27.09
92.10





23.37
12.01
96.96





31.20
24.48
84.79
38.93

11.36


0.00
1.34
-2.49
-1.45
1.62


2.02
1.59
1.64
-2.45
1.14
0.26

1.25
1.15
1.41





1.17
1.32
-0.44





-0.39
0
1.64
0.53

0.51









Table A-1. Continued
Location Date Stage Travel Time A 2exRn calc ** exRn ** 2Ra
(m) (days) (dpm/L) (dpm/L) (dpm/L)
Jug 5/8/02 9.78 > 2.99 >10.23 25.18 2.41
9/12/02 10.16 2.99 8.31 32.05 1.64
01/31/03 10.53 1.56 32.25 29.71 1.11
02/19/03 11.49 0.43 -- -- --
2/26/03 11.51 0.43
3/5/03 <11.51 <0.43 -- -- --

Hawg 5/8/02 9.78 > 3.62 >8.55 17.82 -0.18
9/12/02 10.16 3.62 6.94 11.51 1.28
01/31/03 10.53 1.89 29.48 287.66 376.24
02/19/03 11.49 0.52 17.67 14.62 -2.58
2/26/03 11.51 0.52 21.08 97.50 -0.84
3/5/03 <11.51 <0.52 <15.60 21.29 0.78

Two Hole 5/8/02 9.78 > 3.67 >8.63 11.45 1.16
9/12/02 10.16 3.67 7.01 16.93 2.51
01/31/03 10.53 1.91 29.5 20.74 0.95
02/19/03 11.49 0.53 -- -- --
2/26/03 11.51 0.53
3/5/03 <11.51 <0.53 -- -- --

Sweetwater 5/8/02 9.78 > 4.82 >6.33 21.37 0.13
9/12/02 10.16 4.82 5.14 25.10 0
01/31/03 10.53 2.52 25.11 104.37 1.37
02/19/03 11.49 0.7 16.97 52.94 -1.87
2/26/03 11.51 0.7 20.26 14.83 1.44
3/5/03 <11.51 <0.70 15 8.08 0.38

River Rise 5/8/02 9.78 > 8 days >2.76 10.6 -0.11
9/12/02 10.16 8 days 2.24 8.537 1.24
01/31/03 10.53 4 17.01 53.5 -0.07
02/19/03 11.49 1.3 14.46 4.45 -0.08
2/26/03 11.51 1.3 17.26 4.62 -0.08
3/5/03 <11.51 1 12.78 4.54 0.36
* Travel time (days) is based on Dean (1999) and Martin (2003)
** Measured activities of 222Rn and 226Ra
^ Activities calculated using the decay equation









Table A-2. Excess 222Rn and total 226Ra for Depth Profiles and Wells
Sites Sample 222Rn 226Ra
Dates (dpm/L) (dpm/L)
Rise Im 6/5/02 3.69 1.56
Rise 2m 6/5/02 7.12 2.24
Rise 4m 6/5/02 1.91 1.17
Rise 6m 6/5/02 3.04 2.47
Rise 8m 6/5/02 3.36 0.68
Rise 10m 6/5/02 4.77 1.86

Vinzants 0.3048m 7/2/02 32.86 1.16
Vinzants Im 7/2/02 25.25 1.10
Vinzants 2m 7/2/02 20.51 0.76
Vinzants 4m 7/2/02 13.78 0.81
Vinzants 6m 7/2/02 30.22 -2.18
Vinzants 8m 7/2/02 24.04 1.52

Rise 0.3048m 7/2/02 13.23 1.30
Rise Im 7/2/02 7.14 1.24
Rise 2m 7/2/02 3.63 2.43
Rise 4m 7/2/02 3.02 2.05
Rise 6m 7/2/02 2.78 0.75
Rise 10m 7/2/02 2.39 0.88

Vinzants 0.3048m 8/29/02 30.37 0.51
Vinzants Im 8/29/02 16.94 2.19
Vinzants 2m 8/29/02 19.69 0.80
Vinzants 4m 8/29/02 19.23 0.73
Vinzants 6m 8/29/02 21.19 1.60
Vinzants 8m 8/29/02 20.88 0.79

Vinzants 0.3048m 5/7/03 392.93 0.24
Vinzants Im (1) 5/7/03 275.38 0.30
Vinzants Im (2) 5/7/03 247.76 0.65
Vinzants 2m 5/7/03 36.78 0.92
Vinzants 4m 5/7/03 129.25 0.66
Vinzants 6m (1) 5/7/03 110.52 1.15
Vinzants 6m (2) 5/7/03 213.93 0.70
Vinzants 8m 5/7/03 321.54 0.17









Table A-2. Continued
Sites Sample 222,eRn 226Ra
Dates (dpm/L) (dpm/L)
Rise 0.3048m 5/7/03 31.55 0.22
Rise Im (1) 5/7/03 123.60 0.49
Rise Im (2) 5/7/03 8.79 0.98
Rise 2m 5/7/03 149.79 0.41
Rise 4m 5/7/03 186.74 0.14
Rise 6m (1) 5/7/03 27.76 0.13
Rise 6m (2) 5/7/03 23.24 0.78
Rise 8m 5/7/03 61.23 1.32
Rise 10m 5/7/03 110.25 -0.34

Well 1 4/30/03 1051.82 -7.08
Well 2 (1) 4/30/03 762 4.76
Well 2 (2) 4/30/03 625.87 1.94
Well 3 4/30/03 1120.69 0.77
Well 4 4/30/03 131.96 -0.36
Well 6 4/30/03 431.61 5.78
Well 7 4/30/03 293.61 1.49









Table A-3. Gas Exchange Equation
Site Date Cd (dpm/L)


Ogden Pond


Parener's Sink







Big Sink







Jim Sink







Jug Lake







Hawg Sink


5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03


22.40
18.19
48.43
20.31
24.23
17.93

21.82
17.72
47.77
20.25
24.17
17.88

21.64
17.58
47.56
20.23
24.15
17.86

21.48
17.45
47.38
20.18
24.08
17.81

20.71
16.82
46.48
20.07
23.95
17.72

20.34
16.52
46.05
20.02
23.89
17.67


Amount of Evasion
(dpm/L)
0.15
0.13
0.17
0.01
0.02
0.01


0.73
0.60
0.83
0.07
0.08
0.06

0.91
0.74
1.04
0.09
0.10
0.08

1.07
0.87
1.22
0.14
0.17
0.13

1.84
1.50
2.12
0.25
0.30
0.22

2.21
1.80
2.55
0.30
0.36
0.27









Table A-3. Continued
Site Date Cd (dpm/L) Amount of Evasion
(dpm/L)


Two Hole Sink







Sweetwater Lake







River Rise


5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/02
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03

5/8/03
9/12/02
1/30/03
2/19/03
2/26/03
3/5/03


20.31
16.50
46.02
20.02
23.89
17.67

19.65
15.97
45.23
19.92
23.77
17.59

18.10
14.70
20.10
19.58
23.36
17.44


2.24
1.82
2.58
0.30
0.36
0.27

2.90
2.35
3.37
0.40
0.48
0.35

4.45
3.62
2.45
0.74
0.89
0.50
















REFERENCES


Berelson, W. M., Hammond, D. E., and Fuller, C, 1982. Radon-222 as a tracer for mixing
in the water column and benthic exchange in the southern California borderland.
Earth and Planetary Science Letters 61: 41-54.

Bertin, C., Bourg, A. C. M., 1994. Radon-222 and Chloride as natural tracers of the
infiltration of river water into an alluvial aquifer in which there is significant
river/groundwater mixing. Environmental Science Technology 28: 794-798.

Boyer, D. G., Pasquarell, G. C., 1996. Agricultural land use effects on nitrate
concentrations in a mature karst aquifer. Water Res. Bull. 32: 565-573.

Budd, D. A., Vacher, H. L., 2002. Facies control on matrix permeability in the upper
Floridan Aquifer, West-Central Florida: Implications to diffuse flow. In
Hydrogeology and Biology of Post-Paleozoic Carbonate Aquifers, Karst Waters
Institute Special Publication 7, ed. J. B. Martin, C. M. Wicks, and I. D. Sasowsky,
14-24. Charles Town, West Virginia: Karst Waters Institute.

Bush, P. W., Johnston, R. H., 1988. Ground-water hydraulics, regional flow, and ground-
water development of the Floridan Aquifer system in Florida and in parts of
Georgia, South Carolina, and Alabama. U. S. Geological Survey: 80.

Chung, Y., 1973. Excess radon in the Santa Barbara Basin. Earth and Planetary Science
Letters 17: 319-323.

Corbett, R. D., Burnett, W. C., Cable, P. H., Clark, S. B., 1997. Radon tracing of the
groundwater input into Par Pond, Savannah River Site. Journal of Hydrology, 203:
209-227.

Crandall, C. A., 1996. Shallow ground-water quality in selected agricultural areas of
south-central Georgia. U. S. Geological Survey Water Resources Investigative
Report # 96-4083.

Dean, R. W., 1999. Surface and ground water mixing in a karst aquifer: An example from
the Floridan Aquifer. M. S. Thesis, Gainesville, Florida, University of Florida, 74
p.

Drever, J. I., 1988. The Geochemistry of Natural Waters, 2nd edition. Englewood Cliffs,
NJ, Prentice Hall, 437 p.









Ellins, K. K., Hisert, R. A., Kincaid, T. R., 1992. Hydrogeology of the western Santa Fe
River Basin. Field Trip Guide, South Eastern Geological Society Spring Meeting,
Gainesville, FL.

Ellins, K. K., Roman-Mas, A., Lee, R., 1990. Using 222Rn to examine
groundwater/surface discharge interaction in the Rio Grande De Manati, Puerto
Rico. Journal of Hydrology 115: 319-341.

Elsinger, R. J., Moore, W. S., 1983. Gas exchange in the Pee Dee River based on 222Rn
evasion. Geophysical Research Letters 10:443-446.

Ford, D. C., Williams, P. W., 1989. Karst Geomorphology and Hydrology. Winchester,
MA, Unwin Hyman, Ltd.

Ginn, B., 2002. Using Temperature and Water Elevation measurements to model conduit
properties in karst Aquifers; An Example from the Santa Fe Sink-Rise System,
Florida, senior undergraduate thesis, University of Florida, 23 p.

Gordon, S. L., 1998. Surface and groundwater mixing in an unconfined karst aquifer,
Ichetucknee River ground water basin, Florida: Ph.D. Dissertation, University of
Florida.

Halihan, T., Wicks, C. M., Engein, J. F., 1998. Physical response of a karst drainage
basin to flood pulses: Example of the Devil's Icebox cave system (Missouri, USA).
Journal of Hydrology. 204:24-36.

Hamada, H., 2000. Estimation of groundwater flow rate using the decay of 222Rn in a
well. Journal of Environmental Radioactivity 47: 1-13.

Hamada, H., Komae, T., 1998. Analysis of recharge by paddy field irrigation using 222Rn
concentration in groundwater as an indicator. Journal of Hydrology 205: 92-100.

Hisert, R. A., 1994. A multiple tracer approach to determine the ground and surface water
relationships in the western Santa Fe River, Columbia County, Florida, Gainesville,
FL, Ph.D. Dissertation, University of Florida: 213 p.

Hunn, J. D., Slack, L. J., 1983. Water resources of the Santa Fe River Basin, Florida. U.
S. Geological Survey.

Hussain, N., Church, T. M., Kim, G., 1999. Use of 222Rn and 226Ra to trace groundwater
discharge into the Chesapeake Bay. Marine Chemistry 65: 127-134.

Katz, B. G., 1992. Hydrochemistry of the Upper Floridan Aquifer, Florida. USGS Water
Resources Investigation Report 91-4196, 37 p.

Kincaid, T. R., 1998, River water intrusion to the unconfined Floridan Aquifer,
Environmental and Engineering Geoscience IV: 361-374.









Martin, J. B., Dean, R. W.,1999. Temperature as a natural tracer of short residence times
for ground water in karst aquifers. Karst Modeling. A. N. Palmer, M. V. Palmer
and I. D. Sasowsky, Karst Waters Institute Special Publication #5: 236-242.

Martin, J. B., Screaton, E. J., 2002. Exchange of matrix and conduit water with examples
from the Floridan Aquifer, U.S. Geological Survey Karst Interest Group
proceedings, St. Petersburg, Florida. Water Resources Investigations Report 01-
4011, 38-44.

Martin, J. M., 2003. Quantification of the matrix hydraulic conductivity in the Santa Fe
River Sink/Rise system with implications on the exchange of water between the
matrix and conduit, Gainesville, FL. M. S. Thesis, University of Florida: 100p.

Miller, J. A., 1997. Hydrogeology of Florida: In: The Geology of Florida, eds. A. F.
Randazzo and D. S. Jones, Gainesville, Florida: University Press of Florida, p. 69-
88.

Padilla, A., Pulido-Bosch, A., Mangin, A., 1994. Relative importance of base flow and
quick flow from hydrographs of karst springs. Ground Water 32: 267-277.

Puri, H. S., Vernon, R. O., 1964. Summary of the Geology of Florida and a guidebook to
the classic exposures, Florida Geological Survey Special Publication #5.

Rogers, A. S., 1958. Physical behavior and geological control of radon in mountain
streams, Geological Bulletin No. 1052-E, U. S. Geological Survey.

Ryan, M., Meiman, J., 1996. An examination of short-term variations in water quality at
a karst spring in Kentucky. Ground Water 34, 1: 23-30.

Scott, T. M., 1992. A Geological Overview of Florida, Florida Geological Survey.

Skirvin, R. T., 1962. The underground course of the Santa Fe River near High Springs,
Florida, Gainesville, FL. M. S. Thesis, University of Florida, 62 p.

Smart, C. C., 1988. Artificial tracer techniques for the determination of the structure of
conduit aquifers. Groundwater 26: 445-453.

Smoak, J. M., Brenner, M., Allen, M. S., Shelskie, C. L., Leeper, D. A., 2000. Biological
Accumulation of 226Ra in groundwater- augmented Florida Lake. Limnology and
Oceanography 45, 3: 710-715.

Stringfield, V. T., LeGrand, H. E., 1966. Hydrology of limestone terrains in the coastal
plains of the southeastern United States: Geological Society of America Special
Paper 93, 43 p.

Tennelec, Operations Manual, 1985, Dual Radon Counting System, Oak Ridge,
Tennessee.






62


Wherett, Judith L., 1992. Geologic characterization of soil gas radon concentrations in
Florida and comparison with soil permeability parameters, Gainesville, FL. M. S.
Thesis, University of Florida: 87p.

White, W. A., 1970. The geomorphology of the Florida Peninsula, Geological Bulletin
#51.

White, W. B., 1999. Conceptual models for karstic aquifers. Karst Modeling. A. N.
Palmer, M. V. Palmer and I. d. Sasowsky, Karst Waters Institute Special
Publication #5: 11-16.















BIOGRAPHICAL SKETCH

Lauren A. Smith was born to Roger D. Smith and Martha A. Smith on August 30,

1977, in Ft. Hood, Texas. Shortly after her birth, her family moved to Dublin, Georgia,

and has lived there ever since. After graduating from Dublin High School in 1995,

Lauren attended Furman University in Greenville, South Carolina, where she graduated,

in 2000, with a B. S. in earth and environmental science. Later that year, she moved to

Florida to begin graduate school at the University of Florida. She worked as a teaching

assistant for undergraduate classes and as a research assistant. She is currently seeking

employment in the environmental industry.