the florida state university college of arts …nsgl.gso.uri.edu/flsgp/flsgpx05005.pdf · the...
TRANSCRIPT
THE FLORIDA STATE UNIVERSITY
COLLEGE OF ARTS AND SCIENCES
MULTIPLE ISOTOPIC TRACERS FOR STUDY OF COASTAL
HYDROLOGICAL PROCESSES
By
HENRIETA DULAIOVA
A Dissertation submitted to the Department of Oceanography
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
Degree Awarded: Summer Semester, 2005
ii
The members of the Committee approve the dissertation of Henrieta Dulaiova defended on June 28, 2005.
____________________________ William C. Burnett
Professor Directing Dissertation
____________________________ Joseph F. Donoghue
Outside Committee Member
____________________________ Jeffrey P. Chanton Committee Member ____________________________ William M. Landing Committee Member ____________________________ Willard S. Moore Committee Member ____________________________ Joel E. Kostka Committee Member
Approved: __________________________________________ Nancy H. Marcus, Chair, Department of Oceanography The Office of Graduate Studies has verified and approved the above named committee
members.
iii
ACKNOWLEDGEMENTS
I would like to thank my major professor Bill Burnett for his support throughout
my studies. He is not only an excellent scientist but also a great educator who always has
time and patience to advise on scholarly and research problems. I thank him for giving
me the opportunity to work on great projects and introducing me to the international
“SGD community.”
During the development of the Mn-fiber method described in Chapter 2, Christina
Stringer analyzed companion samples by the Rn-emanation technique. Billy Moore
helped with the sample collection and measurement techniques, the analysis of the Sicily
samples, and provided valuable comments. I also thank the staff of the FSU
Oceanography Machine Shop for their assistance. Financial support came from the
Biological and Chemical Oceanography Program, Office of Naval Research (Grant #
N00014-00-0175).
The co-authors of Chapter 3 were Bill Burnett, Derek Lane-Smith and Rick
Peterson. I also wish to thank personnel from the Florida State University Marine
Laboratory for their assistance during this study. Christina Stringer helped with the
fieldwork and radon emanation analyses. Scientific support for this research was
provided by grants from NOAA’s Cooperative Institute for Coastal and Estuarine
Environmental Technology, CICEET (02-606), and the Biological & Chemical
Oceanography Program of the Office of Naval Research (N00014-00-0175).
The SGD study in West Neck Bay was done in co-operation with Jeff Chanton,
Billy Moore, Henry Bokuniewicz, Matt Charette, and Edward Sholkovitz. I thank the
organizers of the experiment and the teams from Stony Brook University, Cornell
Cooperative Extension, and the Suffolk County Department of Health for the logistical
and personnel support during the experiment. Christina Stringer and Rick Peterson
helped with fieldwork and sample analysis. The SGD intercomparison experiment was
iv
partially funded by SCOR, LOICZ, UNESCO (IOC and IHP), CICEET (Grant# 1368-
810-41) and ONR (Grant# 1368-769-27).
The Chao Phraya River project was done in co-operation with Makoto Taniguchi
from the Research Institute for Humanity and Nature, Japan and Gullaya Wattayakorn
and Pramot Sojisuporn from the Chulalongkorn University, Bangkok. I thank the crew of
the R/V Chula Vijai for their helpful assistance during the river-estuary cruises. Supitcha
Chanyotha and others from the Nuclear Technology Department, Engineering Faculty,
Chulalongkorn University provided some of the well sampling and radon analyses. I also
acknowledge the support from the SARCS, the Southeast Asia Regional Committee for
START (SyStem for Analysis, Research, and Training) and the National Science
Foundation (Award No. OCE-0350514).
I thank Lee Edmiston, Lauren Levi, Jennifer Wanat and others from the
Apalachicola National Estuarine Research Reserve for providing excellent logistical
support and helpful assistance with the field work in Apalachicola Bay. Brent McKee
and Dan Duncan from the Department of Earth and Environmental Sciences, Tulane
University kindly provided the continuous-flow centrifuge and helped with the suspended
particulate matter sampling. Behzad Mortazavi provided helpful comments on the
residence time estimates. I acknowledge support from the NOAA National Estuarine
Research Reserve System’s Graduate Research Fellowship (Award #
NA03NOS4200055). Scientific support for this research was also provided by a grant
from NOAA's Cooperative Institute for Coastal and Estuarine Environmental
Technology, CICEET (02-606).
Finally, I would like to thank the members of my dissertation committee Jeff
Chanton, Bill Landing, Joel Kostka, Billy Moore, and Joseph Donoghue for their
guidance and comments on my dissertation.
v
TABLE OF CONTENTS
List of Tables ................................................................................................................... viii List of Figures .................................................................................................................... xi Abstract ............................................................................................................................ xvi
Chapter 1. INTRODUCTION TO THE DISSERTATION.................................................1 Geochemical tracers and groundwater discharge ........................................1 Research objectives......................................................................................3 Organization of the dissertation ...................................................................3 Chapter 2. AN EFFICIENT METHOD FOR GAMMA SPECTROMETRIC
DETERMINATION OF 226,228Ra VIA Mn FIBERS...............................................6
Introduction..................................................................................................6 Materials and procedures .............................................................................7 Fiber preparation and Ra preconcentration......................................7 Short-lived radium isotopes measurement.......................................8 Long-lived radium isotopes counted by gamma-spectrometry........8 Calibration......................................................................................11 Assessment and discussion ........................................................................12 Comments and recommendations ..............................................................15
Chapter 3. A MULTI-DETECTOR CONTINUOUS MONITOR FOR ASSESSMENT
OF 222Rn IN THE COASTAL OCEAN.................................................................16
Introduction................................................................................................16 Experimental ..............................................................................................17 Single detector system ...................................................................17 Multi-detector system ....................................................................20 Results and discussion ...............................................................................21 Laboratory tests..............................................................................21 Field tests .......................................................................................22 Conclusions................................................................................................24 Chapter 4. ASSESSMENT OF GROUNDWATER DISCHARGES INTO WEST NECK
BAY, NEW YORK VIA NATURAL TRACERS ................................................26 Introduction................................................................................................26
vi
Study site and methods ..............................................................................28 Results........................................................................................................30 Discussion..................................................................................................34 Calculation of SGD using radium isotopes....................................34 Offshore flux of 222Rn....................................................................36 Calculation of SGD from continuous radon model .......................37 Residence time of water in West Neck Bay...................................41 Summary ....................................................................................................43 Chapter 5. RADON AND RADIUM ISOTOPES IN THE CHAO PHRAYA RIVER
AND ESTUARY ...................................................................................................45 Introduction................................................................................................45 Study site....................................................................................................46 Methods..........................................................................................48 Sampling techniques ......................................................................48 Results........................................................................................................51 Horizontal surface and vertical salinity profiles ............................51 Radon surveying ............................................................................52 Dissolved radium isotopes .............................................................53 Radium on suspended particulates and bottom sediments.............59 Discussion..................................................................................................61 Radium isotopic trends ..................................................................61 Estuarine radium ages ....................................................................65 Radon atmospheric evasion ...........................................................68 Estimate of SGD based on a mixing model ...................................71 Evidence of SGD based on 226Ra balance......................................76 Conclusions and applications of the tracer results.....................................79 Chapter 6. EVALUATION OF THE FLUSHING RATES OF APALACHICOLA BAY,
FLORIDA VIA NATURAL GEOCHEMICAL TRACERS.................................80 Introduction and study area........................................................................80 General approach .......................................................................................81 Sampling and measurement .......................................................................84 Sampling plan ................................................................................84
Dissolved radium isotope ratios measured in river and bay waters .............................................................................................85
Radium isotopes measured on suspended sediments and sediments........................................................................................87
Radioanalytical measurement techniques ......................................87 Radon and radium isotope survey on the periphery of
Apalachicola Bay ...........................................................................89 Results........................................................................................................91 Initial tracer survey ........................................................................91 Environmental parameters and radium isotopes ............................91
vii
Discussion................................................................................................100 Radium isotopes in Apalachicola River.......................................100 Radium isotopic ratios in the bay.................................................101 Apparent radium ages ..................................................................104 Conclusions..............................................................................................106 CONCLUSION................................................................................................................111 APPENDIX A..................................................................................................................112 APPENDIX B ..................................................................................................................115 REFERENCES ................................................................................................................121 BIOGRAPHICAL SKETCH ...........................................................................................128
viii
LIST OF TABLES
2.1. Sample characteristics and results of the IAEA’s intercomparison study of radium and uranium in natural waters. The IAEA values of natural samples (N) represent consensus values [n = 57 (IAEA-424), n = 49 (425), and n = 50 (427)] calculated as robust mean and robust standard deviation (Muller, 2000). The gamma spectrometric results for 226Ra represent mean values from 214Pb (295, 352 keV) and 214Bi (609 keV) peaks and the 228Ra results are averages of the two 228Ac (338, 911 keV) photopeaks. The listed uncertainties are based on the standard deviation of these averages. ...................................................................13
4.1. Ranges of activities (dpm.100L-1) of excess 222Rn, 223Ra, excess 224Ra, 226Ra,
228Ra, methane (nM), and salinity measured in wells, piezometers, seepage meters and seawater within 50 meters from the shore from the West Neck Bay study site. “Ex.” refers to excess activities unsupported by radioactive parents. .......31
4.2. Activities of 222Rn measured in wells adjacent to the study site and pore water
estimates based on sediment equilibration techniques. Standard deviations with asterisks indicate duplicates, otherwise S.D. estimated via counting statistics. Radon values from well cluster S-1, thought to be influenced by heterogeneity in the aquifer, were excluded from the average..........................................................33
4.3. Values of advection rates calculated by: (1) the continuous radon model using a
mixing term estimated by inspection of “net fluxes”; (2) the radon model using a radium derived mixing term; and (3) calculated solely by the distribution of radium isotopes. For the radon balance approach the table shows the range in specific discharge that occurred between 17 May 2002 and 23 May 2002 and the average for the entire period. The same values are also expressed as a shoreline flux of groundwater per meter shoreline per day. Also shown are SGD values measured by the WHOI dye-dilution seepage meter positioned 10 meters seaward of mean tide (inshore) and 20 meters seaward of mean tide (offshore). ...................................................................................................................44
5.1. Dissolved radium isotopes and 222Rn activities measured in wells on land near
the coastline of Gulf of Thailand. Known depths and salinities are also indicated. The “nd” indicates that we do not have depth information about the particular well. The “x” indicates pore water values obtained from surface sediments in the coastal zone, either collected from benthic chambers (S-1, NC, C-1) or determined by sediment equilibration (pw1, pw2). The benthic chambers and some of the wells (SR, HH) were located in Sri Racha in the
ix
eastern Gulf of Thailand and Hua Hin in the western coastline of the gulf (Fig. 5.1).. ............................................................................................................................54
5.2. Ra isotope activities in the Chao Phraya River and estuary. ......................................56 5.3. Ra isotopes on the towed fibers. .................................................................................58 5.4. Radium activity ratios measured in zero-salinity water at station B25 located -
33 km upstream in Chao Phraya River. The particles were collected on a 0.45 µm cartridge-type filter, and the river water values are the average of the B25 samples collected in duplicate. Uncertainties represent ± 1 σ based on counting statistics.......................................................................................................................60
5.5. Gamma-spectrometric results of sediments from the eastern shore (Sri Racha)
of the upper Gulf of Thailand. Uncertainties represent ± 1 σ based on counting statistics.......................................................................................................................61
5.6. The Ra isotope regeneration time on particles to 99 % of the desorbable
activity. .......................................................................................................................64 6.1. Sampling dates, Apalachicola River discharge calculated as an average of the
flow three days before and during our sampling, wind speed and direction measured at NOAA CO-OPS station #8728690 located in Apalachicola, Florida. The listed wind speed and direction includes measurements two days before and during our sampling. The tidal stage was measured at station #8728690. The indicated salinities were measured nearby the three main outlets of Apalachicola Bay: at station 6 by Indian Pass, at station 8 by West Pass and station 10 at Sike’s Cut as indicated on Figure 6.1......................................93
6.2. Salinities measured during the buoy deployments in the surface and near
bottom waters at each station and the radium isotopes measured on the deployed fibers. Uncertainties represent ± 1 σ based on counting statistics .............94
6.3. Radium isotopes measured in Apalachicola River. Uncertainties represent ± 1
σ based on counting statistics .....................................................................................98 6.4. Radium isotopes measured in bottom and suspended sediments in Apalachicola
River. Uncertainties represent ± 1 σ based on counting statistics .............................99 6.5. Comparison of 224Ra/223Ra ratios measured in bay water and corresponding
prefilters. Uncertainties represent ± 1 σ based on counting statistics........................99 6.6. 224Ra/223Ra activity ratios and the corresponding apparent radium ages at the
different stations sampled during the four fiber deployments. Uncertainties for the 224Ra/223Ra activity ratios represent ± 1 σ based on counting statistics. The
x
average error of the apparent radium ages calculated based on the 224Ra/223Ra AR uncertainties is estimated to be ± 1-1.5 days. ....................................................108
A.1. Radium isotopes and salinities measured in the West Neck bay study site and
the transect leading from the study site to Gardiners Bay. .......................................113 A2. Radon-222 measured in the West Neck bay study site and the transect leading
from the study site to Gardiners Bay. The two different sample types were counted by the radon-emanation technique (Sample I) and using the RAD-7 H2O (Sample II). .......................................................................................................114
xi
LIST OF FIGURES
2.1. The stainless-steel tool-wrap sheet is cut into a circle and a crucible is formed on a hand press. The Mn-fiber is packed into the crucible and ashed at 550 oC. The crucible is then folded, hydraulically pressed, and sealed with silicone sealant to create a ~3 mm thick counting vessel...........................................................9
2.2. Count rate data for a 226Ra standard compared to theoretical ingrowth. (A)
Mn-fiber was spiked with 226Ra, ashed, folded and measured at certain time intervals. After 37 days from preparation the counting vessel was sealed with silicone sealant. This resulted in a 5 % increase in the 222Rn daughter activities most likely because of 222Rn loss before sealing. (B) The Mn-fiber was prepared in the same manner as above but sealed when prepared. Its activity followed the theoretical ingrowth line precisely, demonstrating that there was no Rn escape. ..............................................................................................................11
2.3. Results of the 2003 IAEA intercomparison study of radium in waters. (A)
Comparison of the IAEA reference values to our results of 226Ra measurements by gamma-spectrometry using the average of 295, 352, and 609 keV photopeaks, and the radon emanation technique. (B) Comparison of the IAEA reference values to our results of 228Ra measurements using the average of the 228Ac photopeaks at 338 and 911 keV. Note the scale break on the activity axis of both plots. The IAEA values represent consensus values calculated as robust value of the average and robust standard deviation. The robust analysis is based on using the median instead of the mean to assure better stability against outliers (Muller, 2000). Uncertainties shown for our analyses represent ± 1 σ based on counting statistics. .......................................................................................14
3.1. Diagrammatic view of the experimental setup for a single RAD-7 exchanger
continuous radon-in-water monitor. The “Drystik” is a Nafion drying tube that helps preserve the Drierite desiccant. .........................................................................18
3.2. Count rate in the 218Po channel of the RAD-7 when the water entering the
exchanger is switched from low-radon to high-radon water and then back again. This experiment was performed with an exchanger using a 60o nozzle and with water flow rate of 5.0 L/min. .......................................................................20
3.3. A simplified sketch of an equilibrium three-stage radon measurement system.
Lines shown represent the water pumped through the exchanger by a submersible pump (dashed) and a closed air loop (solid) that flows through three RAD-7 radon analyzers in parallel. ...................................................................21
xii
3.4. Results from a 3-stage system starting in tap water (~13,300 Bq/m3) and
cycling to radium-free (~350 Bq/m3) deionized water as indicated by the dashed lines. The 3 detection systems were 7 minutes out of phase and each had a counting interval of 21 minutes. The experiment was performed with a 60o nozzle and a water flow rate of 1.5 L/min............................................................24
3.5. Distribution of 222Rn as a function of distance offshore the FSU Marine
Laboratory on March 28, 2003. The measurements were collected within a few hour period using the automated radon system described in this paper. The open circles represent the earlier measurements made at a lower tide and the closed circles were taken near the end of the survey when the tide was higher. .........................................................................................................................25
4.1. Map of Shelter Island, New York, and the study site in West Neck Bay. The
symbols refer to station locations of water samples for geochemical tracers. The squares indicate the radium and circles represent the radon sampling points with the corresponding sample numbers..........................................................28
4.2. Activities of radium isotopes (dpm.100L-1) for samples collected in the West
Neck Bay study site and on the transect leading from the study site to Gardiners Bay. ............................................................................................................32
4.3. Continuous 222Rn measurements (dpm.L-1), methane (CH4), and water depth
(tidal) records from the study site in West Neck Bay. The methane concentrations (nM) have been divided by 10 to fit onto the same scale. Radon and methane concentrations tend to be the highest shortly after the lowest tide........34
4.4. (A) Natural logarithm of 223Ra concentration over distance on transect from
study site to Gardiners Bay that is used to calculate the mixing coefficient in West Neck Bay. (B) Radon-222 activities (dpm.m-3) along the same transect. Both plots have regression lines shown for the West Neck Bay part of transect. The dashed lines represent the 95 % confidence intervals around the regression....................................................................................................................37
4.5. Calculated net 222Rn fluxes based on the change in inventories per unit time
after corrections for tidal effects and atmospheric evasion. The mixing losses estimated via the maximum negative Rn fluxes (dashed line) and the 223Ra derived mixing loss (solid line) is indicated on the figure. ........................................38
4.6. Fluid advection rates assessed from the radon model using the Ra-derived
mixing loss. The advection rates were calculated by division of the total radon flux by our best estimate of the radon concentration in the advecting fluids (173 ± 17 dpm.L-1; see Table 4.2). The gray interval around the advection rate is the total combined uncertainty based on the errors of the analytical measurements as well as the estimated uncertainties of the atmospheric flux
xiii
and mixing calculations and represent ± 1σ. ..............................................................40 4.7. Plot showing net 222Rn fluxes, water level, and seepage rates measured by a
dye-dilution seepage meter developed at Woods Hole Oceanographic Institution (Sholkovitz et al., 2003). Note that the maximum 222Rn fluxes and the highest measured SGD tend to occur at low tides, while the main radon losses and lowest SGD occur at high tides. ................................................................41
5.1. (A) Map of Thailand depicting the upper Gulf of Thailand; (B) Chao-Phraya
River with the January 2004 radon data; (C) July 2004 trajectories. Radon (222Rn) activities measured in the near-surface waters are plotted on the transects where larger circles represent higher activities. Towns of Hua Hin and Sri Racha indicated on panel (A) were our two additional study sites. ...............47
5.2. (A) Horizontal salinity profile in near surface water in the river and estuary.
Zero km is at the river mouth, negative distances are upriver, positive are offshore. Differences in salinity at the same location during the same survey were due to tidal effects. (B) Salinity and temperature vertical profile at the farthest offshore station (+42 km) in July showing stratification. The January profile showed the water column well mixed at approximately the same location. .....................................................................................................................52
5.3. Activity of 222Rn and its parent 226Ra as a function of distance in the river and
Gulf of Thailand in January and July, 2004. ..............................................................53 5.4. Radium isotope distribution in surface water samples plotted against salinity in
the Chao Phraya River and estuary in January and July, 2004...................................55 5.5. Activity ratios 224Ra/223Ra plotted against 224Ra/226Ra and 222Rn for January
and July. The three groups represent samples collected in the I: River, II: offshore, III: intertidal zone........................................................................................62
5.6. Apparent radium ages of water in the Chao-Phraya estuary calculated using
224Ra/223Ra and 224Ra/228Ra activity ratios for January and July, 2004. .....................67 5.7. Radon and 224Ra isotopes plotted together (A) on different scales to compare
their trend against distance, and (B) plotted on the same scale against the apparent radium ages in the estuary, where 0 age is at 6 km. Distance d on figure (A) represents the part of the transect that was used for the radon loss calculations. ................................................................................................................70
5.8. Radon losses from the water column by atmospheric evasion calculated using a
theoretical approach based on measured Rn in air and water concentrations, wind speed and water temperature (crosses); and calculated from the comparison of 222Rn and 224Ra activities along a transect (triangles). .......................71
xiv
5.9. Estimates contribution of river, seawater and groundwater discharge to the water column plotted against distance. The fractions were calculated using a 3-end-member mixing model based on the respective concentrations of 222Rn and 224Ra/223Ra activity ratios. The bottom panels show the horizontal salinity profiles during the two cruises....................................................................................75
6.1. Map of Apalachicola Bay, Florida showing the Apalachicola River and the
four passes to the Gulf of Mexico. Our sampling stations are marked by diamonds. The groundwater study site at the St. George Island State Park is also indicated. ............................................................................................................81
6.2. Monthly historical discharge of Apalachicola River measured near Sumatra,
Florida since 1980. The gauge (USGS 02359170) is about 33 km upstream from the river mouth. The circles on the expanded part of the plot indicate sampling time and discharge during our seasonal study. ...........................................84
6.3. Design of moored buoys for deployment of Mn-fibers. .............................................86 6.4. Radon and radium isotope ratio levels measured along the coastline of
Apalachicola Bay during July16 - 17, 2003 (A) Radon concentrations (dpm m-3) measured by a Rn-surveying system; (B) 224Ra/223Ra activity ratios measured
on towed Mn-fibers, where one towed fiber is represented by 8 points. The darker colors and larger circles represent higher activities (Rn) and activity ratios (Ra). ............................................................................................................................90
6.5. Ratios of 224Ra/223Ra measured in selected river water and sediment samples.
Uncertainties represent ± 1 σ based on counting statistics. ......................................101 6.6. Ratios of 224Ra/223Ra measured in surface and near bottom waters in
Apalachicola Bay and River during the four sampling periods. Uncertainties represent ± 1 σ based on counting statistics..............................................................102
6.7. Ratios of 224Ra/228Ra measured in surface and near bottom waters in
Apalachicola Bay and River during the four sampling periods. Uncertainties represent ± 1 σ based on counting statistics..............................................................103
6.8. Calculated water ages estimated for surface waters during our four sampling
periods in August 2003, March and August 2004, and January 2005. The indicated river discharge is an average before and during the sampling. .................109
6.9. Wind speed and directions measured at NOAA CO-OPS station #8728690
located in Apalachicola, Florida. The lines indicate the time of our sampling. .....110 B.1. Submarine groundwater discharge rates measured at the St. George Island
study site. (A) Dry period; (B) Groundwater advection measured at the same site during several rain events (indicated by bars). ..................................................118
xv
B.2. Rain and well water level measured at the St. George Island study site during
the summer of 2001. Tropical storm Barry and our sampling period right after the storm passed are indicated on the figure. The SGD results are shown in Fig. B.3. ....................................................................................................................119
B.3. Submarine groundwater discharge measured on the St. George Island study
site in the summer of 2001 right after tropical storm Barry passed over the area (A) Groundwater advection rates assessed based on continuous radon measurements; (B) SGD measured by a continuous automated seepage meter. Both instruments were deployed at the same place at the same time. ......................120
xvi
ABSTRACT
This study focused on the combined use of radon and radium isotopes as tracers
of near-shore geophysical processes including submarine groundwater discharge (SGD),
water exchange rates, and atmospheric evasion. Methods were developed for easier
measurement of long-lived radium isotopes in natural waters and for continuous radon
surveying over larger areas in the coastal zone. These tracer techniques were used to
study the mentioned processes at study sites in Shelter Island Sound (New York), the
Gulf of Thailand, and Apalachicola Bay, Florida.
Groundwater fluxes calculated for Shelter Island using isotopic techniques
produced results consistent with those measured directly via seepage meters.
Groundwater discharge in the Chao Phraya Estuary (Thailand) was shown to be in the
range of 2 to 20 m3 s-1, small compared to river discharge but much higher than seepage
rates measured in nearby locations.
An experimental assessment of 222Rn evasion to the atmosphere was performed
using radon and 224Ra profiles in the Chao-Phraya Estuary in Thailand. The different
trends in radium and radon isotopes measured in the estuary provided an estimate of
atmospheric exchange that agreed with a theoretical approach.
Short-lived radium isotopes were applied in a seasonal study of water residence
time in Apalachicola Bay, Florida. The water transport within different sectors of the bay
was evaluated as a result of advective (estuarine) and dispersive physical mixing. The
distribution of the radium isotope ratios was used to determine apparent radium ages of
the water within the bay. The results clearly showed how the water-residence time in the
bay changes seasonally and that the winds and tides influence the water circulation in the
bay. The radium tracer approach gave turnover times of 6 to 12 days in Apalachicola
Bay during the studied periods.
1
CHAPTER 1
INTRODUCTION TO THE DISSERTATION
Geochemical Tracers and Groundwater Discharge
Geochemical tracers can provide important information on hydrological processes
and mixing rates in coastal areas. This work describes the use of radon and radium
isotopes to study submarine groundwater discharge (SGD) and exchange rates of near-
shore and estuarine waters.
SGD is the direct flow of groundwater into the coastal ocean. Like surface water,
groundwater flows down gradient and SGD occurs wherever a coastal aquifer is
connected to the sea. Groundwater is delivered to the ocean either by slow seepage
mainly in near-shore areas or by discrete spring flow (Cable et al., 1996a, Burnett et al.,
2001a). Some confined aquifers extend for considerable distances from shore. If the
groundwater is under sufficient pressure a relatively large volume of water may be
discharged through springs and fractures in the coastal zone, even out to the shelf break
(Swarzenski, 2001).
SGD consists of various mixtures of fresh groundwater and seawater. There are
several oceanic processes that cause seawater to enter coastal aquifers and then flow back
as re-circulated seawater. The circulation can be driven by the hydraulic gradient on the
land and various oceanic forces. Fresh and saltwater components often mix, especially
near the coast, resulting in brackish water discharge.
SGD is a widespread coastal feature. Regional parameters that may affect the
occurrence and magnitude of SGD are climate, geology, topography, soil and sediment
type and their hydraulic conductivity, hydraulic head of the underlying aquifer, tidal
2
range and other oceanic forces (Bokuniewicz, 2001; Burnett et al. 2003a). There is also
indication that substantial groundwater inputs may occur in river-dominated margins
(Moore, 1997; Krest, 1999; Moore and Krest, 2004).
When one measures submarine groundwater discharge, the traditional concept of
groundwater as terrestrial fresh water cannot be applied. If we view water in the pores of
sediments as groundwater, then SGD is any flow of groundwater through a sea floor that
can be fresh, brackish or salty, and it is defined regardless of its origin or driving force
(Burnett et al. 2003a). Submarine groundwater recharge (SGR) is a flow opposite to
SGD and occurs in marine environments due to tides, waves, currents, groundwater and
sea level fluctuations, and density differences. The net flow across a certain area of sea
floor is then the difference between SGD and SGR in the area and it can result as
outflow, inflow or zero net flow.
The basic approaches for assessments of submarine groundwater discharge
include hydrologic modeling, direct physical measurement using seepage meters, and
tracer techniques. The principle of natural tracer techniques is that groundwater is
enriched in certain elements that occur in much lower concentrations in seawater.
Natural geochemical tracers have been used successfully for assessment of SGD
in a number of studies. Investigators used radium isotopes as tracers estimating the
quantity and effects of SGD and coastal residence times (Moore, 1996; Moore, 2000a,b;
Krest and Harvey, 2003; Charette et al., 2001; Kelly and Moran, 2002). Radon has also
been shown to be an excellent tracer for work performed in the Gulf of Mexico (Cable et
al., 1996a; Burnett et al., 2002; Burnett and Dulaiova, 2003) and in Florida Bay (Chanton
et al., 2003; Corbett et al., 1999, 2000). Radon and radium isotopes are radioactive
elements from the uranium and thorium decay series, which are relatively easy to
measure, and they belong to the group of radioactive conservative tracers.
Radium isotopes and radon have been successfully used as two separate tracer
techniques to assess SGD in various coastal settings. The research described in this
dissertation shows that the combined application of radon and radium isotopes provides
an even more powerful tool to evaluate the magnitude and dynamics of submarine
groundwater discharge.
3
Research Objectives
The main goal of this research was to apply radon and the radium isotopes
simultaneously to test the following hypotheses:
a) Usually, the most important losses of radon in a coastal setting are by mixing
with lower concentration waters offshore. I hypothesized that use of short-lived radium
isotopes with complementary radon measurements can be applied to determine the
horizontal eddy dispersion coefficient and calculate the mixing losses of radon from the
coastal zone. This would constrain an important loss term in the radon mass balance
model.
b) Another loss term of radon from the water column is its evasion to the
atmosphere. I hypothesize that simultaneous 222Rn and 224Ra measurements from a
common main source can be used to determine atmospheric loss of radon from seawater.
These estimates could then be compared to losses calculated with standard gas exchange
equations.
c) Radium isotopes and their ratios may be used to determine the water flushing
rate in shallow estuaries.
Organization of the Dissertation
Geochemical tracer techniques require reliable and efficient sampling and
measurement methods. Radium concentrations are typically at such low levels in natural
waters, that their measurement requires preconcentration from very large water samples.
Radium from large volumes (20-150 liters) of seawater or groundwater can be collected
by passing the water through a cartridge loaded with MnO2-impregnated acrylic fiber
(Moore et al., 1995; Moore, 1996, 2000a, b; Turekian et al., 1996; Charette et al., 2001;
Kelly and Moran, 2002; Burnett et al., 2002; Burnett and Dulaiova, 2003). Chapter 2
presents a new approach for converting “Mn-fiber” into a form suitable for measurement
of long-lived 226Ra and 228Ra by gamma-spectrometry. The method is simple to
implement as it does not require any wet chemistry and it is appropriate for radium
4
measurements in natural waters over a wide range of salinity and radium activity making
it ideal for coastal studies.
The usefulness of radon as a tracer of groundwater discharge has been limited by
the time consuming nature of collecting individual samples and traditional analysis
schemes (Mathieu, 1988; Stringer and Burnett, 2004). Burnett et al. (2001b) developed a
“continuous” radon monitor that provides high-resolution measurements of the radon
concentration at one location over time. While this was a significant improvement, there
was a further need to measure radon concentrations over larger coastal areas in shorter
times. Chapter 3 demonstrates an automated multi-detector system that can be used in a
continuous survey basis to map radon activities. This system produces continuous radon
results together with GPS navigation, water depth, temperature, and conductivity
measurements. It also allows for large spatial coverage in much shorter times than the
traditional methods of collecting discreet samples.
Chapter 4 reports on submarine groundwater discharge rates measured in West
Neck Bay, New York. West Neck Bay is a shallow, enclosed embayment located on
Shelter Island in the eastern half of Peconic Bay, New York in the northeast United
States. There is significant groundwater seepage into the bay from fresh water-table
aquifer on Shelter Island. The aquifer has high hydraulic heterogeneity, and the
hydraulic gradient as well as SGD vary along the seepage face. In West Neck Bay we
applied the radon mass balance model (Burnett and Dulaiova, 2003) in combination with
coastal mixing rates estimated via short-lived radium isotopes (Moore, 2000a). This way
the radium isotopes helped to constrain the uncertainties in the radon balance and also
provided an independent estimate of submarine groundwater discharge.
Identification of SGD in river dominated margins is challenging as the
geochemical signals may be the combination of the river and groundwater inputs. The
major source of radon to most estuaries is likely via groundwater discharge. Substantial
amounts of radium are transported to estuaries by rivers and particles transported by
rivers as well as via saline or brackish groundwater discharge. Chapter 5 presents a
study to evaluate groundwater discharge in the estuary of the Chao Phraya River in
Thailand using radon and radium isotopes. The Chao Phraya River is the largest river
flowing into the Gulf of Thailand and represents about half of the total river flow into the
5
gulf. Since there is a significant difference in the river discharge in the dry and wet
seasons, we measured the geochemical tracers during both periods. Higher radon levels
and different radium isotope signatures indicated that there is a significant amount of
groundwater seeping into the estuary. A mixing model using the river, groundwater, and
offshore radon and radium end-member concentrations was constructed to decipher the
groundwater flow contribution. The study revealed that the groundwater discharge can
be as high as 10 % of the river discharge. In addition, the onshore-to-offshore
distribution of 222Rn and 224Ra was used to make an independent estimate of radon loss to
the atmosphere. These results agreed with traditional gas exchange calculations.
Moore (2000b) used radium isotopes to determine radium ages of continental
shelf waters and Moore and Krest (2004) applied the same approach to estimate the
apparent water age in the Mississippi River plume. We hypothesize that the same radium
method can be used to determine water ages in a small, shallow embayment. Chapter 6
reports on the results of the water residence time calculations using radium isotope ages
in Apalachicola Bay, Florida. The bay is a shallow bar-built estuary, with a stratified
water column. The major fresh water and radium source to the bay is the Apalachicola
River. We performed a radon and radium survey on the coastline of the bay and
concluded that because of the small magnitude of SGD Ra inputs by groundwater
discharge to this bay are insignificant compared to the river input. We selected several
stations across the bay and collected surface water samples in four different seasons. The
radium ages were calculated based on the short-lived radium isotopic ratios measured in
the stations. The apparent radium ages were used to create age contour plots of
Apalachicola Bay for the four different sampling periods. The method can be easily
applied to other embayments where the river is a the major source of radium isotopes.
6
CHAPTER 2
AN EFFICIENT METHOD FOR GAMMA SPECTROMETRIC
DETERMINATION OF 226,228Ra VIA Mn FIBERS
Introduction
There are four naturally occurring radium isotopes - 223Ra: T1/2=11.4 days, 224Ra:
3.6 d, 226Ra: 1600 years, and 228Ra: 5.8 y. Measurement of these isotopes in natural
waters is of great interest in geochemistry as well as in radioprotection. For example,
geochemists use isotopic ratios for calculations of various geochemical processes, like
tracing submarine groundwater discharge or calculation of offshore and inshore water
mixing coefficients (Moore et al., 1995; Moore, 1996, 2000a, b; Turekian et al., 1996;
Charette et al., 2001; Kelly and Moran, 2002; Burnett et al., 2002; Burnett and Dulaiova,
2003).
Radium isotopes are typically at such low levels in natural waters, especially in
seawater, that their measurement requires preconcentration from very large samples.
Moore (1976) developed a method where radium is collected from 200 – 1000 liters of
seawater by passing the water through MnO2-impregnated acrylic fiber. At near neutral
pH and under controlled flow-rate, the fiber quantitatively adsorbs Ra, Pb, Th, Ac and
other elements. The fiber is subsequently processed for gamma-spectrometry by leaching
the MnO2 off the fiber with mineral acids and co-precipitating radium as BaSO4 (Moore,
1984); or ashing the fiber and packing the ash into a vial for gamma counting (Charette et
al., 2001). Baskaran et al. (1992) used polypropylene fiber impregnated by MnO2
because of its lower 228Ra blank activity and smaller residue after ashing. We describe
here a method using MnO2-impregnated acrylic fiber for preconcentration of radium and
7
a modified approach for gamma-spectrometry for the measurement of 226Ra and 228Ra
using custom-made stainless steel crucible – counting vessels.
Materials and Procedures
Fiber preparation and Ra preconcentration
The MnO2-impregnated acrylic fiber is prepared by immersing the raw acrylic
fiber for about 20 minutes in saturated KMnO4 solution heated to 75oC. When the fiber
turns jet black, it is removed from the bath and rinsed thoroughly (Moore, 1976). Ra-free
water, prepared by passing de-ionized water through previously prepared Mn fiber is
used throughout the procedure. After a final rinse in Ra-free water, the fiber is stored
damp in plastic bags until use. For applications, approximately 150 cm3 (~25 grams dry
weight) of fiber is packed into a cylindrical cartridge.
We have been using the fiber to measure radium isotope concentrations in
groundwater and coastal seawater. The sample volumes usually range between 20 to 150
liters depending on the expected activities. The water is either slowly (1-2 L/min) passed
through the fiber without collection at a controlled flow-rate, or it is collected in
containers and processed at a more convenient time. The flow-rate of water passing
through the fiber has to be below 2 L/min to achieve quantitative radium adsorption
(Moore, 1976; Moore et al., 1995). Kim et al. (2001) have checked for complete radium
uptake by pumping 500-liters of groundwater through two columns connected in series.
No detectable radium was found on the downstream column, while significant radium
was on the initial column confirming quantitative adsorption.
In certain cases groundwater can be reducing, containing hydrogen sulfide or
other reducing agents. In such cases the water can reduce and thus dissolve the
manganese on the fiber, causing less than quantitative radium recoveries. In such cases
we pump the water into an open container and the sample is degassed and oxidized
before processing.
In a different approach, one could pass water samples through Mn-fiber using a
high flow-rate to yield only the ratios of 223Ra, 224Ra and 228Ra to 226Ra, and then the 223Ra, 224Ra and 228Ra is quantified by a separate measurement of 226Ra using standard
radon-emanation or some other convenient method. This later approach is preferred if
8
the sampling time needs to be as short as possible. In a similar manner, if one desires to
measure just the activity ratios of radium isotopes, the fiber could be either immersed in-
situ in the water to passively collect radium or towed through the water for times
depending on the expected activities.
After exposing natural waters to the Mn-fiber by whatever means, the fiber is
flushed with Ra-free de-ionized water. This rinse is important to wash out any
particulates and sea salts that can interfere with radon emanation during the measurement
(Moore, 2000a; Sun and Torgersen, 1998).
Short-lived radium isotopes measurement
After radium is extracted and the Mn-fiber is rinsed, the moisture of the fiber is
adjusted to have a water-to-fiber weight ratio in a range from 0.7 to 2.5 (Kim et al., 2001;
Sun and Torgersen, 1998). This adjustment is easily performed by either hand squeezing
or drying via compressed air. The short-lived isotopes 223Ra (T1/2=11.4 days) and 224Ra
(T1/2=3.6 days) are then measured by a delayed coincidence counter system developed by
Moore and Arnold (1996). The partially dried fiber is placed in a helium-circulation
system in which the short-lived radon daughters of 223Ra and 224Ra, 219Rn and 220Rn, are
swept into a scintillation detector and a delayed coincidence circuit discriminates the
alpha decays of the different radium daughters by the timing of the alpha-decay events.
The system is calibrated using 232Th and 227Ac standards that are known to have their
daughters in radioactive equilibrium and are adsorbed onto an MnO2-coated fiber.
Alternatively, one could measure 224Ra on the fiber by a method developed by Kim et al.
(2001) that uses a commercially available radon-in-air monitor (Rad-7, Durridge Co.,
Inc.) to count 220Rn released from the fiber.
Long-lived radium isotopes counted by gamma-spectrometry
When the short-lived radium measurements are completed, the fiber is processed
for the measurement of long-lived 226Ra and 228Ra by gamma-spectrometry. The sample
is packed into a custom-made crucible (Fig. 2.1) that is fashioned from stainless-steel
sheets in our laboratory. The crucible also serves as a counting vessel to eliminate any
transfer steps. We needed to choose a material for the “crucible-counting vessel” that has
9
low gamma-ray attenuation, is easy to form, and withstands high temperatures for
prolonged periods. We selected a stainless steel tool-wrap sheet Type 321 (MSC
Industrial Supply CO., Inc, Jacksonville, Florida), that has high titanium content,
withstands temperatures up to 1000oC, and has a thickness of only 0.05 mm. To test its
gamma-ray absorption at the energies of interest for this study we used a high activity
mixed gamma source that we measured both directly positioned on the detector and than
again after placing the stainless steel sheet between the source and the detector.
Repetitive measurements showed that while there was a 15 % attenuation of gamma rays
at 46 keV, the absorption was only 2 % at 186 keV and less than 1 % at the higher
energies (295, 338, 352, 609 and 911 keV) that are most important for our application.
We also measured the surface area of our germanium detector (19 cm2) and we prepared
a die to press the crucibles at that exact diameter achieving the best possible measurement
geometry.
Figure 2.1. The stainless-steel tool-wrap sheet is cut into a circle and a crucible is formed on a hand press. The Mn-fiber is packed into the crucible and ashed at 550 oC. The crucible is then folded, hydraulically pressed, and sealed with silicone sealant to create a ~3 mm thick counting vessel.
10
We use these crucibles to ash the Mn-fiber at 550oC for 6 hours. During the
ashing procedure the sample mass is reduced by approximately 90 %. The stainless steel
sheet has the advantage over materials like aluminum that it does not get brittle during
this heating process. We allow a few hours cooling period after the ashing otherwise the
steel would loose its elasticity and crack easily. The crucibles are than folded to enclose
the ash and pressed using 1 kg/cm2 pressure with a hand press, creating an approximately
3 mm thick wafer (Fig. 2.1). This approach has the advantage that the sample is not
transferred after ashing to a different measurement container. The sheet is folded in a
way that there is no loss of the ash containing the radium and there is apparently little or
no radon escape from the pressed ‘wafer.’ Prevention of radon escape is important
because 222Rn (T1/2=3.8 days) is the direct daughter of 226Ra and its short-lived daughters
(214Bi, 214Pb) are used for 226Ra determination.
We tested radon escape by preparing a fiber containing a known amount of 226Ra
and we measured its activity from the time of preparation in short time intervals. Without
radon losses the measurements should follow a theoretical ingrowth line (Fig. 2.2A).
Thirty-seven days after the preparation of the standard we sealed the folded side of the
vessel with silicon sealant. Shortly after that the measured activity increased by 5 %
implying that 5 % of the radon had escaped from the vessel most likely because of 222Rn
loss before sealing. In order to ensure no radon escape, we now seal all wafers with the
caulking material. The activity ingrowth of another standard prepared in this manner
shows that the ingrowth follows the theoretical line (Fig. 2.2B).
Radium-228 is determined by counting its daughter 228Ac (T1/2=6.1 hours) that
has photopeaks at energies 338 and 911 keV. Radium-226 is determined either by its
direct photopeak at 186 keV that should in this case be clear from interferences from 235U
(U is not extracted on Mn fiber from oxidizing waters) or by counting its granddaughters 214Bi at 609 keV and 214Pb that have photopeaks at energies 295 and 351 keV. If one
wants to measure 226Ra via the 186 keV peak the sample can be counted right away.
Otherwise, approximately 3 weeks are needed to reach equilibrium between 226Ra and its
granddaughters. Counting 214Bi and 214Pb has the advantage over the 186 keV peak, in
that they have lower backgrounds and significantly higher gamma yields, both of which
result in improved sensitivities and lower measurement errors.
11
Time (days)
0 50 100 150 200
Peak
(cpm
)
0
1
2
3
4
Theoretical Activity (B
q)
0
2
4
6
8
295 keV352 keV609 keVTheoretical
wafer sealed on 37th dayA
Time (days)
0 20 40 60 80 1000
10
20
30
40
0
20
40
60
80B
Figure 2.2. Count rate data for a 226Ra standard compared to theoretical ingrowth. (A)
Mn-fiber was spiked with 226Ra, ashed, folded and measured at certain time intervals. After 37 days from preparation the counting vessel was sealed with silicone sealant. This resulted in a 5 % increase in the 222Rn daughter activities most likely because of 222Rn loss before sealing. (B) The Mn-fiber was prepared in the same manner as above but sealed when prepared. Its activity followed the theoretical ingrowth line precisely, demonstrating that there was no Rn escape.
Calibration
Blanks were prepared from the same amount of fiber as used for the sample
collection, and these blanks were ashed and packed in the same way as described for the
samples. Blanks were measured to obtain a background count rate value that was
subtracted from every measurement.
The counting efficiencies of our gamma-spectrometer were determined using
NIST-traceable 226Ra and 228Ra standard solutions. As a part of efficiency calibration we
had to be sure that the fiber extracts radium quantitatively and that radon does not escape
from the counting vessel. In order to make these evaluations, we first packed standard
sized Mn-fibers into stainless-steel crucibles and spiked with 133Ba (15 Bq), 226Ra (38 Bq)
and 228Ra (2.7 Bq). Pipetting directly onto the Mn-fiber ensured that 100 % of the
activity stayed on the fiber. The fiber was then dried and ashed and the crucible pressed
into a counting wafer. A second set of two standards was prepared by passing 2 liters of
pH=7 spiked water (133Ba, 226Ra, 228Ra) through the fibers. After passing the water
through the fibers we collected the effluent from two standards to see how efficiently the
12
fibers extracted the radioisotopes. We did this by co-precipitating radium and barium
isotopes from the effluent with BaSO4. This precipitate was filtered, air-dried and then
measured by gamma-spectrometry to evaluate whether any Ba or Ra remained. We did
not find any statistically significant Ba or Ra activity in the precipitate from either
standard. According to the gamma-spectrometric measurements these standards resulted
in a chemical recovery for radium at least 97 %. We thus concluded that the radium and
barium isotopes are quantitatively extracted from the water by the Mn-fiber using these
techniques.
Assessment and Discussion
Blank measurements of Mn-fibers prepared in the same manner as the samples
showed values indistinguishable from the normal detector backgrounds for the peaks of
interest. We did not encounter any difficulties with high 228Ra blanks from the acrylic
fiber as had been reported by Baskaran et al. (1992).
We validated our method by participating in a proficiency test organized by the
Analytical Quality Control Services (AQCS) of the International Atomic Energy Agency
(IAEA), called the “Interlaboratory Study on Determination of Radium and Uranium
Radionuclides in Water.” We received 6 acidified water samples (3 natural and 3
synthetic) from the IAEA in January 2003 (Table 2.1). For analysis, sample volumes
were adjusted to 5 liters with radium-free de-ionized water, which raised their pH to
approximately 5. The samples were first measured for 226Ra by standard Rn-emanation
technique (Stringer and Burnett, 2004). Since Rn-emanation is a non-destructive method
we could process the same samples for 226Ra and 228Ra determination using the MnO2-
fiber method. We adjusted the acidity of the water samples to near neutral pH (6.5-8) and
then passed each sample through an MnO2-fiber column. The fibers were ashed and
sealed as described earlier, and we measured them after at least a 21-day holding period
on a gamma-spectrometry system.
We calculated 226Ra results using an average calculated from the photopeaks of 214Pb (295 and 352 keV) and 214Bi (609 keV) and 228Ra using the mean results from the 228Ac photopeaks at 338 and 911 keV (Fig. 2.3). The IAEA reference values of the
13
natural samples had large standard deviations and were not appropriate for validation of
our method. However, the synthetic sample reference values had standard deviations
lower than 2 % and the agreement between our results and the IAEA reference values
were very good. Our mean 226Ra value relative to the IAEA reference value has a ratio of
1.05±0.24 and for 228Ra is 0.95±0.09. This comparison indicates that our results are not
biased by a systematic error and there is a good agreement between the reference values
and our results both with low and high radium activities. The comparison of 226Ra results
measured using the Mn-fiber method and by Rn-emanation (Fig. 2.3A) also shows an
excellent agreement.
Table 2.1. Sample characteristics and results of the IAEA’s intercomparison study of radium and uranium in natural waters. The IAEA values of natural samples (N) represent consensus values [n = 57 (IAEA-424), n = 49 (425), and n = 50 (427)] calculated as robust mean and robust standard deviation (Muller, 2000). The gamma spectrometric results for 226Ra represent mean values from 214Pb (295, 352 keV) and 214Bi (609 keV) peaks and the 228Ra results are averages of the two 228Ac (338, 911 keV) photopeaks. The listed uncertainties are based on the standard deviation of these averages.
Cl- Volume IAEA reference value Gamma Spectrometric
Result 226Ra 228Ra 226Ra 228Ra
Sample Code IAEA
(mg/L) * (kg) (Bq/kg) (Bq/kg) (Bq/kg) (Bq/kg)
421 760 S 0.999 0.055±0.001 0.124±0.002 0.05±0.01 0.12±0.05
422 760 S 0.486 0.568±0.002 0.84 ±0.02 0.52±0.07 0.65±0.12
424 930 N 2.015 0.054±0.026 0.1 ±0.1 0.081±0.009 0.062±0.056
425 1600 N 1.093 0.3 ±0.1 0.5 ±0.2 0.3±0.1 0.5±0.1
427 72300 N 0.594 65 ±17 36 ±4 71±2 31.6±0.4
431 52200 S 0.487 24.9 ±0.4 4.14 ±0.07 30.7±0.5 4.14±0.10 *S-Synthetic Sample; N-Natural Sample
Radium-226 was also measured by the 186 keV peak as well as the radon
daughter peaks in order to demonstrate the improved sensitivity using the 214Pb + 214Bi
peaks. Our results show that the measurement uncertainty from the counting statistics for
the 186 keV peak is 3 to 5 times larger than that for the 214Pb and 214Bi peaks for
14
comparable counting periods. The peak at 186 keV resulted in 1σ uncertainties that
averaged 40 %, while the others had uncertainties that averaged 10 %. However, the 186
keV peak can be used if one wishes to measure the samples immediately after ashing, and
the counting uncertainties can be reduced by longer acquisition times.
Sample
IAEA 421 422 424 425 427 431
226 R
a ac
tivity
(Bq/
kg)
0.0
0.5
1.0
20.0
40.0
60.0
80.0IAEA Ref.Mn-fiberRn-eman
Sample
IAEA 421 422 424 425 427 431
228 R
a ac
tivity
(Bq/
kg)
0.0
0.5
1.0
10.0
20.0
30.0
40.0IAEA Ref.Mn-fiber
A B
Figure 2.3. Results of the 2003 IAEA intercomparison study of radium in waters. (A) Comparison of the IAEA reference values to our results of 226Ra measurements by gamma-spectrometry using the average of 295, 352, and 609 keV photopeaks, and the radon emanation technique. (B) Comparison of the IAEA reference values to our results of 228Ra measurements using the average of the 228Ac photopeaks at 338 and 911 keV. Note the scale break on the activity axis of both plots. The IAEA values represent consensus values calculated as robust value of the average and robust standard deviation. The robust analysis is based on using the median instead of the mean to assure better stability against outliers (Muller, 2000). Uncertainties shown for our analyses represent ± 1 σ based on counting statistics.
We used the MnO2-fiber method to measure radium activities in coastal ocean
waters off the Florida State University Marine Laboratory (FSUML) and in Sicily, Italy
during recent experiments. At each location we collected measured amount of seawater
(50-150 liters) and passed it through MnO2-fibers. After counting the short-lived radium
isotopes we processed the samples for gamma-spectrometric measurement as described in
this paper. In order to compare our results to a reference method we also collected
parallel samples at both sites that were analyzed by another method. At the FSUML site
15
we collected 5-liter grab samples for 226Ra measurement by standard Rn-emanation
technique (Mathieu et al., 1988). In the Sicily study, W. S. Moore (University of South
Carolina) collected parallel Mn-fiber samples that he analyzed for 226Ra and 228Ra by acid
leaching the fibers and co-precipitating Ra with BaSO4 before gamma-spectrometry
(Moore, 1984). The average ratio of the method described here to the reference methods
is 1.01±0.3.
We have shown that this technique provides good results, requires minimal
operator time, and it is simple to implement. It is appropriate for radium measurements
in natural waters over a wide range in salinity and radium activity.
Comments and Recommendations
While radium collection on Mn-fiber is well known within the oceanographic
community it has also proven useful in analysis of radium isotopes in spring, lake and
river waters. We have also used the method successfully for radium analysis in public
water supplies.
16
CHAPTER 3
A MULTI-DETECTOR CONTINUOUS MONITOR FOR ASSESSMENT
OF 222Rn IN THE COASTAL OCEAN
Introduction
There are many reasons why one might want to make continuous measurements
of 222Rn in natural waters. Geophysical applications include groundwater monitoring
associated with earthquake and volcanic eruption prediction, air-sea gas exchange
processes, horizontal and vertical eddy diffusion in the ocean, and assessment of
submarine groundwater discharge. In addition, continuous monitoring of the radon
content of municipal water supplies and of water supplied to bottling plants and other
industries would be attractive for radiation protection purposes. In such cases continuous
monitoring may be preferable to the conventional, standard process of taking a sample
and sending it for later analysis because radon activities in natural waters may change on
many time scales.
Our interest in radon concerns its use as a tracer of groundwater inputs to the
coastal ocean. Groundwater is an important source of nutrients and other dissolved
constituents to the coastal ocean in some areas (Burnett et al., 2003). Estimates of the
location and magnitude of groundwater flow are scarce because measurements cannot be
performed easily and sites of discharge are not always obvious – groundwater is an
“invisible pathway” between land and sea (Moore, 1996; Moore, 1999; Zekster, 1996).
Groundwater discharge estimates via radon tracing is an indirect method of
assessment. Due to the huge radon concentration difference between groundwater and
surface water (often 1000-fold or higher 222Rn levels in groundwater), dilution is not as
17
important as other potential tracers. In addition, radon is completely inert so
biogeochemical reactions do not need to be considered. By using a modeling approach in
conjunction with radon analyses in the receiving waters, one can estimate the total flux of
groundwater into a region (Cable et al., 1996b; Burnett and Dulaiova, 2003). A practical
application of such an approach would be the identification of seepage sites and
assessment of fluxes of a contaminant that has a groundwater pathway into a coastal
zone.
In spite of the fact that we have made significant progress in our ability to assess
groundwater fluxes in one area using a radon tracing approach, we are hampered in
making regional-scale assessments by the time-consuming logistical requirements of
collecting and analyzing samples in the conventional manner (Burnett et al., 2002). We
recently developed a “continuous” radon monitor that provides high-resolution
measurements of the radon concentration at one location over time (Burnett et al.,
2001b). While this has been a significant improvement, we still do not have the ability to
map radon concentrations very effectively because the method relies on counting the
radioactive daughters of 222Rn and this typically requires thirty minutes to two hours per
analysis at the activities encountered in most coastal waters. We describe here a new
version of this system that we specifically designed for surveying that uses a high-flow
air-water exchanger and multiple detectors in order to increase analysis throughput.
Experimental
Single Detector System
Our single-detector system (Fig. 3.1) analyses 222Rn from a constant stream of
water passing through an air-water exchanger that distributes radon from the running
water to a closed air loop. The air stream is fed to a commercial radon-in-air monitor
(RAD-7, Durridge Co.) that determines the activity of 222Rn by collection and
measurement of the α-emitting daughters, 214Po and 218Po. Since the distribution of radon
at equilibrium between the air and water phases is governed by a well-known temperature
dependence, the radon activity in the water is easily calculated if one also measures the
water temperature.
18
The RAD-7 uses a high electric field with a silicon semiconductor detector at
ground potential to attract the positively-charged polonium daughters, 218Po+ (T1/2 = 3.10
min; alpha energy = 6.00 MeV) and 214Po+ (T1/2 = 164 µs; 7.67 MeV) which are then
counted as a measure of the radon activity in air. For faster analyses, the 218Po is
preferred, as it will reach radioactive equilibrium with 222Rn in only about 15 minutes
(214Po requires about 3 hours for equilibrium because of the intermediate 214Pb and 214Bi
daughters).
The air-water exchanger is simply a plastic cylinder that has water entering
continuously via a nozzle that aspirates the water into fine droplets so that radon is
emanated into a stream of air that is circulated through the exchanger, a drying system,
then to the RAD-7 for measurement, and then returned to the exchanger to begin another
loop. After some time, the radon activity in the air reaches equilibrium with the radon in
the water, the ratio at equilibrium being determined by the water temperature (Weigel,
1978).
Figure 3.1. Diagrammatic view of the experimental setup for a single RAD-7 exchanger continuous radon-in-water monitor. The “Drystik” is a Nafion drying tube that helps preserve the Drierite desiccant.
19
In order to make quicker measurements, it is important to achieve the air-water
equilibrium as quickly as possible. Faster flow rates of the water phase should aid the
equilibrium process. We tested a high-flow exchanger system by running tap water from
our laboratory using a nozzle (WL-4) that has a large enough opening to allow flow rates
of up to about 9 L/min. The tap water in our building is always moderately high in 222Rn
(~6,000-14,000 Bq/m3; not unusual for a groundwater source) although the levels are not
constant over long periods because the city utility mixes different water sources. The
results of our tests showed that there is a substantial decrease in the time required for
equilibration between the lowest flow (2.5 L/min) tested and the intermediate rate (5.5
L/min) while the improvement is only marginal when the flow rate is raised to the
maximum level possible (9 L/min) on our experimental set up. It appears that under
conditions of flow exceeding about 5 L/min, equilibrium is obtained in 20-25 minutes,
just slightly longer than the theoretical value for radioactive equilibrium.
In order to determine the time lag that we can expect when waters of variable
radon activity are encountered, we designed an experiment that would evaluate the
response when the equilibrium system is subjected to drastic changes in the activity of
radon in the water flowing through the exchanger. We set up a reservoir of radium-free
deionized water that had a radon activity of ~350 Bq/m3 during the period of the test and
alternated the inflow to the exchanger between this relatively low radon water and the
much higher radon tap water. When the system was switched from the low to high radon
source the response is fast, reaching equilibrium in approximately 20 minutes (Fig. 3.2),
close to the results from the earlier tests. When switched back to the low radon water, the
response is initially rapid with a drop of about 85% from the high equilibrium activity in
30 minutes, but then the rate slows with a tail extending out to about 85 minutes. We
think that the system is more sluggish on the change to lower activity because the excess
radon in the air from the high-radon water must first re-dissolve in the water phase in
order to be eliminated from the system. This is apparently a slower process than
releasing radon from the water to the air phase that is assisted by spraying the water
through a nozzle.
20
Time (min)
0 25 50 75 100 125 150 175 200 225 250 275 300
218 P
o (c
ps)
0
1
2
3
4
5
6
7
45 min 95 min
180 min
low Rn low Rnhigh Rn
65 min
Figure 3.2. Count rate in the 218Po channel of the RAD-7 when the water entering the
exchanger is switched from low-radon to high-radon water and then back again. This experiment was performed with an exchanger using a 60o nozzle and with water flow rate of 5.0 L/min.
Multi-Detector System
We have now developed a 3-stage approach that uses one high-flow exchanger
with a drying system connected in parallel to 3 radon detection systems (Fig. 3.3). The
advantage of this approach is that we can triple the time resolution with little additional
effort. Since the air-water equilibrium is based on concentration, there should be no
reduction in the activity of radon in the air being circulated, in spite of the much larger
volume of air circulating through the system. Thus, there would be no reduction in the
sensitivity of each of the 3 radon monitors compared to a single detector system. Since
approximately three times the air is being run through the system, it would be desirable to
increase the flow rate of the water in order to minimize the time required to reach
equilibrium. Another relevant factor is that the flow rate of the air coming back to the
exchanger will be effectively tripled as three air flows (each driven by the internal pump
21
in the RAD-7 at about 1 L/min) are joined together at a four-way valve and then sent
back to the single exchanger. This increased flow should assist in stripping radon from
the water phase in the exchanger. Thus, there are two effects in expanding the
equilibrium system from one detector to three: (1) an increase in the air volume by
approximately a factor of three; and (2) an increase in the air flow rate through the
exchanger, again by roughly a factor of about three. These processes affect the timing of
the approach to equilibrium in opposite directions and thus may nearly counterbalance
each other.
air
RAD-7 RAD-7 RAD-7
exchanger
water
air
drier
Drystik
Figure 3.3. A simplified sketch of an equilibrium three-stage radon measurement system. Lines shown represent the water pumped through the exchanger by a submersible pump (dashed) and a closed air loop (solid) that flows through three RAD-7 radon analyzers in parallel.
Results and Discussion
Laboratory tests
The response time of the system depends upon the half life of 218Po, the volume of
the air loop, the speed of transfer of radon from the water to the air (which likely depends
on the size of the water droplets and the efficiency of the aeration), the flow rate of the
re-circulating air, the volume of water in the exchanger, and the flow rate of water
through the exchanger (Burnett et al., 2001b; Lane-Smith and Shefsky, 1999). The half-
life of 218Po, 3.1 min., dictates an ultimate theoretical limit, for the 95% response time, of
about 15 minutes, assuming everything else was instantaneous. The solubility coefficient
22
of 222Rn at 20oC is about 0.25, meaning that there is about four times more radon in the
air phase than the aqueous phase at equilibrium. Thus, at least four times more water
must flow through the system to deliver all the radon that is required. Again, that is
assuming everything is working at maximum efficiency, which is unlikely.
We first tested this 3-stage set-up in the laboratory by connecting it to our
building tap water. A completely independent single detection system was also deployed
in a neighboring laboratory (same water supply) to ensure that the results were
compatible. All three detectors reached equilibrium quickly (~20 minutes) and gave
statistically identical results to each other, as did the control system in the other
laboratory. In addition, the system displayed no significant difference in activity when
the flow rate was lowered from 4.5 L/min to 2.0 L/min. All detectors also showed a
similar and rapid (~20 min) return to baseline when switched from tap water to air.
In order to maximize the data while in a survey mode, we would run the three
systems out of phase, i.e., the timing between each unit would be set so data outputs
would be equally spaced. To test an out-of-phase operation, we programmed each RAD-
7 to run a 21-minute cycle and started each counting interval seven minutes apart. This
results in a new reading every 7 minutes. The system was started in tap water (~13,300
Bq/m3 on the day this experiment was performed), switched to radium-free deionized
water (~ 350 Bq/m3), and then the cycle was repeated. The results show a very good
coherence between the results from the three detectors (Fig. 3.4). The approach to
equilibrium was slower and the apparent activity in the low-radon deionized water
appears somewhat too high because the low-flow (1.5 L/min.) pump used that day
resulted in more sluggish response times.
Field tests
We performed a field test of the system on March 28, 2003 off the Florida State
University Marine Laboratory (FSUML) on the northeast Gulf of Mexico. We positioned
a submersible bilge pump with a capacity of 230 L/min (3700 gph) over the side of a 10-
meter pontoon boat that provided the exchanger approximately 5 L/min flow of seawater
from about 1.5 meters below the surface. We programmed the three RAD-7 monitors for
30-minute cycles, 10 minutes out of phase, so a new reading would be obtained every 10
23
minutes. Before beginning the survey, we stayed at the dock with the system running
long enough to obtain air-water and radioactive equilibriums. We then left the dock,
driving at a slow speed (approximately 2-3 knots) to increase the spatial resolution of our
data and continued straight out to sea for a distance of about 5 km and then returned with
the system running continuously. We also collected occasional grab samples using a
peristaltic pump with a tube whose opening was positioned adjacent to the pump sending
water to the exchanger. These grab samples were analyzed using conventional radon
emanation techniques (Mathieu et al., 1988) over the next several days for both 222Rn and 226Ra and total 222Rn was calculated after making appropriate decay corrections.
While it is not possible to make actual side-by-side comparisons because of the
nature of how we are operating this system, we will show that the trends in the data from
both the automated system and grab samples are essentially the same. Grab samples
represent an “instantaneous” look at a limited volume of water (5 liters in this case) while
the automated analyses are based on an integration of the radon concentrations over a
discrete distance (up to several hundred meters in this case). More direct comparisons of
our automated system deployed in a stationary mode to traditional grab sampling and
analysis have been presented before and showed very good agreement (Burnett et al.,
2001b; Lambert and Burnett, 2003). The results of this field test (Fig. 3.5) look very
encouraging. The location “uncertainties” arise because the boat is moving during data
collection – the reading thus represents an integrated view of the activity over a finite
distance (depending upon speed), rather than discrete sampling points. In addition to the
observed spatial variation, there is typically a change in radon activity over time
depending upon the tidal stage at this location (Burnett et al., 2002). The results collected
earlier in the day going in a seaward direction (open circles) were from a lower tide
period when the 222Rn is generally higher while the later results obtained going back
towards land (closed circles) were collected during a rising tide when 222Rn tends to be
lower (Burnett et al., 2002). The grab samples (closed triangles) have virtually no
uncertainty in the sample location (except for navigation) and somewhat less
measurement uncertainty but it is clear that the automated system produces much more
data with basically the same pattern and with much less effort. Based on our experience
with the system thus far, we are able to measure radon concentrations with precisions
24
varying from ± 5-15 %, depending upon the radon concentrations. A precision of ± 20 %
is usually sufficient for the use of radon as geochemical tracer.
Time (min)
0 50 100 150 200 250 300 350
Wat
er 22
2 Rn
(Bq/
m3 )
0
2000
4000
6000
8000
10000
12000
14000
Det # 1Det # 2Det # 3
Tap Water DIWater
Tap Water DI Water
Figure 3.4. Results from a 3-stage system starting in tap water (~13,300 Bq/m3) and
cycling to radium-free (~350 Bq/m3) deionized water as indicated by the dashed lines. The 3 detection systems were 7 minutes out of phase and each had a counting interval of 21 minutes. The experiment was performed with a 60o nozzle and a water flow rate of 1.5 L/min.
Conclusions
We find these results very encouraging in terms of having a system that can
analyze radon on the fly with results essentially complete when data are downloaded
from the RAD-7 monitors after arriving back at the dock. Future enhancements under
consideration include integrating the continuous monitors into a navigational GIS-based
system so results can be more effectively displayed and tied to more exact locations. We
also are considering expanding the system with a larger pump and exchanger and up to
six detectors for work further offshore where radon concentrations are generally lower
and longer integration times would be necessary.
25
Distance (m)
0 1000 2000 3000 4000 5000
222 R
n (B
q/m
3 )
0
20
40
60
80
100
120
seawardlandwardRn emanation
Figure 3.5. Distribution of 222Rn as a function of distance offshore the FSU Marine Laboratory on March 28, 2003. The measurements were collected within a few hour period using the automated radon system described in this paper. The open circles represent the earlier measurements made at a lower tide and the closed circles were taken near the end of the survey when the tide was higher.
26
CHAPTER 4
ASSESSMENT OF GROUNDWATER DISCHARGES INTO WEST
NECK BAY, NEW YORK VIA NATURAL TRACERS
Introduction
It is now recognized that in certain regions the direct discharge of groundwater
into the coastal oceans can be significant. Whether in the form of fresh groundwater or
recirculated seawater, submarine groundwater discharge (SGD) complements river
inputs, as a quantitatively smaller, but still important source of dissolved solutes into the
coastal zone. Submarine groundwater discharge is therefore a concern of both
hydrologists and oceanographers because of its influence on the water balance on land
and biogeochemical inputs into the ocean. Still, the assessment of groundwater discharge
rates and associated chemical mass flux remains difficult due to a high degree of
uncertainty in the methodologies (Moore, 1999; Burnett et al. 2001a; Taniguchi et al.,
2002).
The basic approaches for quantitative assessments of groundwater discharge
include hydrologic modeling, direct physical measurement using seepage meters, and
tracer techniques. A series of systematic comparisons of assessment methods has been
performed over the past few years under sponsorship by the Scientific Committee on
Oceanic Research (SCOR), the Land-Ocean Interactions in the Coastal Zone (LOICZ)
project, UNESCO’s Intergovernmental Oceanographic Commission (IOC) and
International Hydrologic Project (IHP), and the International Atomic Energy Agency
(IAEA). These systematic comparisons are conducted at a series of sites selected as
coastal “prototypes” representative of important types of coastline selected by the LOICZ
27
typology group (Bokuniewicz, 2001). In addition to comparing differing approaches of
tracers for the evaluation of SGD, an additional goal of the project is to determine which
tracers/approaches are best applicable in certain coastal situations or at certain types of
coastlines to offer guidance to future researchers. We report here on the geochemical
tracer results of one of these experiments held on Shelter Island, New York, May 17-22,
2002. This site represented a glacial till setting. The techniques applied by various
investigators included manual, heat-based, ultrasonic, and dye-dilution based automated
seepage meters deployed side by side at the study site; geoelectric measurements of the
area; and the use of naturally occurring geochemical tracers. Estimates of groundwater
discharge in the area had previously been made by water balance calculations and
hydrogeologic modeling (DiLorenzo and Ram, 1991, Schubert, 1998).
Geochemical tracers have been used successfully for assessment of SGD in a
number of other studies. For example, Moore used radium isotopes as a tracer in a series
of papers reporting on the quantity and effects of SGD off the coast of the southeastern
U.S. (Moore, 1996; Moore, 2000a,b). Others used radium isotopes to study SGD and
coastal residence times (Krest and Harvey, 2003; Charette et al., 2001; Kelly and Moran,
2002). Radon has also been shown to be an excellent tracer for work performed in the
Gulf of Mexico (Cable et al., 1996a; Burnett et al., 2002; Burnett and Dulaiova, 2003)
and in Florida Bay (Chanton et al., 2003; Corbett et al., 1999, 2000). Radon and radium
isotopes are radioactive elements from the uranium and thorium natural decay series,
which are relatively easy to measure, and they are radioactive conservative type tracers.
Methane has been employed as a tracer of groundwater inputs into near-shore
waters along the coast of the northeastern Gulf of Mexico (Bugna et al., 1996 and Cable
et al., 1996b), Florida Bay (Corbett at al., 1999), and Korea (Kim and Hwang, 2002).
Although methane is not a conservative tracer it has proven to be useful where its
concentration in groundwater highly exceeds methane inventories in the water column.
One of the advantages of geochemical tracers over seepage meters is that the
coastal water column integrates the tracers and smoothes small-scale variations in
discharge. In this study we show that while both radium isotopes and radon have been
shown to be effective tracers of SGD, they are even more powerful when applied together
in the same system.
28
Study Site and Methods
West Neck Bay (WNB) is a shallow, enclosed embayment located on Shelter
Island in the eastern half of the Peconic Bay, New York. The bottom sediment in the bay
is mostly sand with some muddy deposits over an unconsolidated aquifer. There is
significant groundwater seepage into the bay from fresh water-table aquifer on Shelter
Island. Paulsen et al. (2001) measured high SGD rates ranging between 60 and 240 cm
day-1 occurring along this section of shoreline. The upper glacial aquifer is recharged
solely by precipitation. The aquifer has high hydraulic heterogeneity, and the hydraulic
gradient as well as SGD vary along the seepage face. The upper glacial aquifer is
underlain by a base clay at ~ 27 m below sea level. Groundwater in the deeper (100 m)
aquifer on the island is mostly saline (Schubert, 1998). Our study site was located in the
northeastern part of WNB (Fig. 4.1).
Figure 4.1. Map of Shelter Island, New York, and the study site in West Neck Bay. The symbols refer to station locations of water samples for geochemical tracers. The squares indicate the radium and circles represent the radon sampling points with the corresponding sample numbers.
29
During the intercomparison experiment, we collected water samples in the WNB
vicinity and analyzed them for radon, radium isotopes, and methane. For the radium
isotope analysis, water samples of large volume (20-100 liters) were collected from wells,
piezometers, seepage meters, and ambient seawater. We measured the salinity of each
sample and than passed the water through MnO2-coated acrylic fiber that retains radium,
thorium, and actinium (Moore, 1976). The fibers were counted for 223Ra, 224Ra, and 228Th on a delayed coincidence counter system (Moore and Arnold, 1996). Ra-226 and 228Ra were measured by leaching and barium sulfate co-precipitation followed by
gamma-spectrometry at the University of South Carolina (Moore, 2000a). Smaller
volume samples (0.25 – 5 L) were collected for 222Rn measurements. Radon activities in
grab samples were determined using a radon-in-air monitor (RAD-7, Durridge Co., Inc.)
with an adaptor for water samples (RAD-H2O) and by a standard radon emanation
technique with a newly-designed plastic bottle (Stringer and Burnett, 2004). Methane
samples in groundwaters were collected into 50-mL glass bottles and analyzed by
headspace equilibration technique and flame ionization gas chromatography. Methane
analyses of coastal seawater at the site were collected in a semi-continuous manner by
equilibration in an air-water exchanger.
Six sediment samples were collected from several places in the seepage-meter
measurement area. About 150-gram sediment samples were used to assess radon pore
water concentrations by a sediment equilibration technique (Corbett et al., 1998).
Water samples for radium, radon, and methane were also collected on an
approximately 20-km long transect leading from the study site at WNB to the south and
than east to open waters in Gardiners Bay (Fig. 4.1). These samples were collected from
the top 1-meter of the water column using a submersible pump.
In order to assess temporal variations in groundwater seepage rates, we also made
continuous methane and radon measurements in the water column at the WNB study site.
We deployed a continuous radon monitor (Burnett et al., 2001b) on a boat anchored ~25
meters from the shoreline. This system consists of a submersible pump bringing a steady
stream of water from a depth of 0.7 m to an air-water exchanger where radon is degassed
and delivered to a commercial radon-in-air analyzer. The instrument made integrated
measurements of radon concentration in the seawater every 2 hours over a 7-day period.
30
On a nearby dock we installed a water level meter (to monitor the tides), a radon-in-air
monitor, and weather station, which made continuous measurements of wind speed, air
and water temperature. By continuously monitoring the water depth at the same location,
we can calculate 222Rn inventories as well as concentrations over the study period. This
assumes that the water column is well mixed which is likely the case in this very shallow
(0.5-1.5 m) system.
Methane was measured continuously with an equilibration sampler similar to that
of radon. Bay water was pumped into a 3-liter chamber and sprayed into the headspace
at a rate of 1 to 1.5 L/min. The headspace was flushed with nitrogen at a rate of 20 mL
per minute. Prior experiments had shown that equilibrium between methane in air and
water in this system was reached after an initial period of 45 minutes. Headspace
methane samples were taken every 15 minutes during selected periods of the experiment
and were analyzed by flame ionization gas chromatography.
Results
A summary of results from water samples analyzed for radium isotopes, Rn,
methane, and salinity are shown in Table 4.1. (Detailed results are shown in Appendix
Tables A.1 and A.2.) Radium isotope activities were lower in wells and piezometers
where the salinity of the samples was 0‰. This was expected because radium isotopes in
freshwater environments are predominantly bound to particles. The radium
concentrations increased markedly in water collected from the seepage meters, because as
fresh groundwater and saltwater mix, radium is released from solid phases into solution.
The radium concentration in these samples also depended on the time of sample
collection because SGD at this site tends to be the highest at low tide. The ranges of the
measured activities are displayed in Table 4.1, where the higher values are from samples
collected at low tides and the lower values correspond to samples collected at high tides.
The variation of radium isotope activities in surface waters as a function of distance
offshore on the 20-km long transect leading from the WNB study site shows that all
radium isotopes have high activities near the coast and decreasing activity with increasing
distance (Fig. 4.2). Seawater samples collected from the seepage meters (Table 4.1) and
31
Tabl
e 4.
1. R
ange
s of a
ctiv
ities
(dpm
.100
L-1) o
f exc
ess 22
2 Rn,
223 R
a, e
xces
s 224 R
a, 22
6 Ra,
228 R
a, m
etha
ne (n
M),
and
salin
ity (p
pt)
mea
sure
d in
wel
ls, p
iezo
met
ers,
seep
age
met
ers a
nd se
awat
er w
ithin
50
met
ers f
rom
the
shor
e fr
om th
e W
est N
eck
Bay
stud
y si
te.
“Ex.
” re
fers
to e
xces
s act
iviti
es u
nsup
porte
d by
radi
oact
ive
pare
nts.
E
x. 22
2 Rn
223 R
a E
x. 22
4 Ra
226 R
a 22
8 Ra
Met
hane
Sa
linity
dpm
.100
L-1
dpm
.100
L-1
dpm
.100
L-1
dpm
.100
L-1
dpm
.100
L-1
nM
ppt
Wel
ls
15,0
00-3
6,00
0 <
1 1-
20
6-34
12
-20
4-30
0-
0.1
Piez
omet
ers
15,0
00-2
5,00
0 <
1 5-
7 7-
15
28
14-3
5 0-
0.1
Seep
age
met
ers
300-
3,00
0 1-
24
20-2
80
6-37
27
-170
8-
33
25.3
-28.
2 Se
awat
er
100-
1,20
0 4-
6 22
-44
13-1
8 60
-90
11-3
7 26
.2-2
9.3
32
within 50 meters from the shore were distinctly enriched in radium isotopes compared to
offshore samples. The near-shore waters are highest in the thorium-series nuclides 228Ra
and 224Ra with the uranium-series isotopes 226Ra and 223Ra being much lower. This is
likely a reflection of the predominance of Th over U in the aquifer sediments and the
different regeneration times of radium isotopes from their Th parents.
Distance (km)0 5 10 15 20
Ra
activ
ity (d
pm.1
00L-1
)
0
20
40
60
80
100Ra-223 Ra-224 Ra-226 Ra-228
Figure 4.2. Activities of radium isotopes (dpm.100L-1) for samples collected in the West Neck Bay study site and on the transect leading from the study site to Gardiners Bay.
Radon activity (Table 4.1) was highest in wells and piezometers where the water
was in contact with material containing its parent 226Ra. Ra-226 is mainly attached to
particles and decays to 222Rn that tends to escape because of alpha recoil processes and its
gaseous state into the surrounding water. Radon activities in wells and piezometers were
up to two orders of magnitude higher than in the coastal seawater. On the transect
leading from the study site offshore, radon activities were highest near the coast and
showed a decreasing pattern with distance similar to the 224Ra pattern. Results of radon
33
activities in wells and from the sediment equilibration experiments are shown in Table
4.2.
Methane concentrations within groundwater were not always enriched over bay
water concentrations thus ruling out the use of methane as an SGD tracer in this
environment. This is most likely due to the low organic content of the glacial till aquifer
which resulted in low methane production in the aquifer.
The continuous 222Rn and methane records together with the observed water
levels are shown in Figure 4.3. The methane record is not continuous throughout the
entire study period because it required manual sampling (the radon is completely
automated). The continuous radon record clearly changes with a 12-hour periodicity,
apparently reflecting tidal modulation of the SGD.
Table 4.2. Activities of 222Rn measured in wells adjacent to the study site and pore water estimates based on sediment equilibration techniques. Standard deviations with asterisks indicate duplicates, otherwise S.D. estimated via counting statistics and represent ±1σ. Radon values from well cluster S-1, thought to be influenced by heterogeneity in the aquifer, were excluded from the average.
Sample
Location Sediment Porosity
Bulk Density
222Rn
S.D. g.cm-3 dpm.100L-1
Beach seep Beach at low tide 16,400 3,200 Piezometer 10’ 19,200 2,800*
Well S-1 S-1 B 36,300 6,500*
S-1 C 36,700 6,400 S-1 D 34,800 1,800*
Well S-2 15,800 2,300* Tap water Pridwin Hotel 20,100 5,200*
Piezometers 0.26 1.74 18,300 3,000*
Sediment Stony Brook 2 0.37 2.05 15,400 2,100* equilibration FSU 1 0.41 1.93 17,700 5,400* experiment K 2 0.45 2.02 16,600 1,800
K 1 0.52 1.66 15,100 1,400 Small Krupaseep 0.45 1.78 17,900 1,700
AVERAGE ( EXCLUDING S-1 ) 17,300 1,700
34
Time/Date 2002
17-May 19-May 21-May 23-May
222 R
n (d
pm.L
-1) &
CH
4 (10
-1 n
M)
0
2
4
6
8
10
Water Level (cm
)
0
40
80
120
Rn (dpm.L-1) Water Level (cm) CH4 (10-1 nM)
Figure 4.3. Continuous 222Rn measurements (dpm.L-1), methane (CH4), and water depth (tidal) records from the study site in West Neck Bay. The methane concentrations (nM) have been divided by 10 to fit onto the same scale. Radon and methane concentrations tend to be the highest shortly after the lowest tide.
Discussion
Calculation of SGD using radium isotopes
The results from radium isotopes and radon analyses indicate that all these tracers
have potential to be good SGD tracers, and their combination allows us to evaluate the
magnitude and dynamics of local and regional groundwater flow at the same time. The
spatial distribution of radium isotopes on the transect leading from the WNB study site
going offshore to Gardiners Bay (Fig. 4.2) allows us to quantify material flux from the
coastal area to offshore. All four radium isotope activities display an apparent break in
slope at a distance of about 4 km from the beginning of the transect. This distance
corresponds to the mouth of the WNB inlet. There were no samples collected between
~4-12 km from the study site. Samples beyond 12 km were collected in Shelter Island
Sound and Gardiners Bay. The break in slope could thus be a result of different mixing
35
patterns in the sound than WNB. Since we are interested mainly in the processes in the
nearshore area, we will restrict the use of these results to the first 4 km in WNB.
A previous study of West Neck Sound reported a water residence time of about
11.7 days (DiLorenzo and Ram, 1991). On this time scale, the radioactive decay of 226Ra
(half-life: 1600 y) and 228Ra (5.7 y) can be neglected. From our results it appears that
both long-lived radium isotopes have a linear trend with distance on the 4-km long part of
the transect that may indicate that their distribution is controlled more by diffusive
mixing than advection. Advection would cause negative or positive curvature of the
activity of the long-lived radium isotopes depending on its direction offshore or onshore.
In systems controlled by eddy diffusion an eddy diffusion coefficient (Kh) can be
calculated applying a principle developed by Moore (2000a) using the distributions of
short-lived 223Ra (half-life: 11.4 d) or 224Ra (3.6 d). The patterns of these short-lived
radium isotopes with distance offshore will depend on two processes, radioactive decay
and mixing. The decay rates are known and the mixing rates can thus be estimated based
on the slope of the natural logarithm of activity as a function of distance. The
distributions of both short-lived isotopes reflect mixing on a several-day long time scale.
Because of its longer half-life the 223Ra profile is preferable since short-term disturbances
would be smoothed out (Fig. 4.2). The plot of ln 223Ra versus distance (Fig. 4.4A) has a
slope of -0.289 ± 0.029 km-1 from which we calculate a mixing coefficient of 8.6 ± 1.2 m2
s-1 based on Eq. 4.1:
hKslope 223λ
= . (4.1)
The flux of 226Ra from the study site can then be calculated as the product of the
concentration gradient of 226Ra and the mixing coefficient derived from 223Ra. For a
linear 226Ra concentration gradient of -0.84 ± 0.23 dpm m-3 km-1 calculated from the 4-
km long part of the transect and the assumption that the tracer is transported in a 2.6 m
deep layer (average depth in WNB), the offshore 226Ra flux is (1.6 ± 0.5)x106 dpm km-1 d-
1. Assuming that this flux is in steady state, 226Ra must be balanced by an input in the
coastal zone – most likely from SGD. We assume that groundwater discharge is the sole
source of radium at our study site in West Neck Bay. From previous measurements
(Paulsen et al., 2001) we know that SGD is high in this area and that compared to SGD,
36
sediment resuspension and surface runoff are likely negligible sources of radium.
According to the ultrasonic seepage meter measurement results at the field site, it appears
that the groundwater discharge is most intensive in a 50-meter wide section along the
shoreline. Using the measured 226Ra activities in seepage meters (average = 220 ± 130
dpm m-3, n = 18, Table 4.1) we converted the radium flux to a water flux by dividing by
this concentration. This results in a groundwater seepage flux of 7 ± 5 m3 m-1 d-1 i.e., 7
m3 of groundwater flowing into the sea per unit meter of shoreline per day in the study
area. Assuming a 50-m wide seepage face, this flow translates into an upward velocity
flux of 0.15 m d-1. We can get a conservative (lower limit) estimate of SGD using the
maximum measured 226Ra activity (370 ± 16 dpm m-3 from a seepage meter), which
produces an apparent seepage flow of 4 ± 1 m3 m-1 d-1 or an upward flux of 0.09 m d-1.
Offshore flux of 222Rn
The concurrent measurements of radon and radium isotopes allow us to use the
mixing coefficient derived from 223Ra to calculate the offshore flux of radon from our
study site. Quantification of the loss of 222Rn from nearshore waters via mixing is an
important component of the radon mass balance. The linear radon activity gradient along
the study transect (Fig. 4.4B) was –450 ± 120 dpm m-3 km-1. The negative trend is due to
losses by mixing, radioactive decay, and atmospheric evasion. Both radioactive decay
and atmospheric loss are more significant for the farther points of the transect because the
age (time since radon enters water from near-shore seepage) at the more distant points is
likely at least several days. We calculate the radon flux offshore from the WNB study
area by multiplying the linear gradient in the 222Rn concentration along the transect (-450
dpm m-3 km-1) by the mixing coefficient derived from analysis of the 223Ra gradient (8.6
m2 s-1) and the average depth of WNB (2.6 m). This calculation results in a total 222Rn
offshore flux of (3.6 ± 1.0) x 104 dpm m-1 h-1. In order to convert this seaward flux to a
flux of 222Rn from the nearshore seepage face on the seabed, we divide the offshore flux
by the estimated width of the seepage face (50 m). The resulting 222Rn flux is equal to
730 ± 260 dpm m-2 hr-1. This independent estimate of the loss of radon via mixing will
assist us to constrain the mass balance of 222Rn in our continuous radon model approach
for assessing SGD (see following section).
37
Distance (km)0 5 10 15 20
ln 22
3 Ra
0.0
0.5
1.0
1.5
2.0
ln223Ra slope = -0.290±0.029R2=0.93
Distance (km)
0 5 10 15 20
222 R
n (d
pm.m
-3)
0
500
1000
1500
2000
2500
3000222Rn slope = -450±120 dpm.m-3.km-1
R2=0.79
A B
Figure 4.4. (A) Natural logarithm of 223Ra concentration over distance on transect from study site to Gardiners Bay that is used to calculate the mixing coefficient in West Neck Bay. (B) Radon-222 activities (dpm.m-3) along the same transect. Both plots have regression lines shown for the West Neck Bay part of transect. The dashed lines represent the 95 % confidence intervals around the regression.
Calculation of SGD from continuous radon model
The continuous radon monitor provided high-resolution data on radon
concentration in the water column at one location over time (Fig. 4.3). We used this
record to quantify rates of groundwater discharge by calculating radon fluxes using a
mass balance approach (Burnett and Dulaiova, 2003). Using the continuous water level
information and 226Ra concentrations from spot measurements in the water column, we
calculated the unsupported 222Rn inventories for each 2-hour long measurement interval.
The inventories were then corrected for tidal changes as described in Burnett and
Dulaiova (2003). Radon losses by atmospheric evasion were calculated for each
measurement interval. The total radon gas flux across air-water interface depends on the
molecular diffusion produced by the concentration gradient across this interface and
turbulent transfer governed primarily by wind speed. We used the gas exchange
equations presented by Macintyre et al. (1995) that calculate the gas exchange across sea-
air interface using the radon concentration gradient, temperature and wind speed. In this
manner we were able to calculate the net radon fluxes (Fig. 4.5) that represent the
observed fluxes of 222Rn into the water column with all necessary corrections except
losses via mixing with lower radon activity waters offshore. From the net radon fluxes it
38
is apparent that most of the radon enters the water column during low tide. The flux is
always highest at low tide and lowest at high tides. We assume that the apparent negative
fluxes observed in the diagram are due to mixing processes between coastal and offshore
waters with lower radon concentration. We estimated these mixing losses for different
periods based on the maximum absolute values of the observed negative fluxes. We
assume that the maximum negative net fluxes are conservative estimate of the mixing
loss as greater losses could be masked by concurrently higher inputs. Based on this
assumption, the mixing losses would range between 500 and 1150 dpm m-2 hr-1 with an
average of 670 dpm m-2 hr-1 over the 7-day measurement period. The short-dashed line
in Figure 4.5 shows this approach.
Time/Date 2002
18-May 20-May 22-May 24-May
222 R
n Fl
ux (d
pm.m
-2.h
r-1)
-1500
-1000
-500
0
500
1000
1500
2000
Water level (cm
)
0
20
40
60
80
100
120
140
est. mixing loss Ra calc. mixing loss
Figure 4.5. Calculated net 222Rn fluxes based on the change in inventories per unit time after corrections for tidal effects and atmospheric evasion. The mixing losses estimated via the maximum negative Rn fluxes (dashed line) and the 223Ra derived mixing loss (solid line) is indicated on the figure.
This estimated mixing loss is in a very good agreement with the independent
estimate of mixing losses based on the eddy diffusion coefficient derived from 223Ra as
described in section 4.2. That flux equals 730±260 dpm m-2 hr-1 (Fig. 4.5, solid line).
The dynamic changes in mixing are not reflected in the radium derived estimate but it
39
represents a radon flux integrated over at least several days before and during the
measurements.
In the radon mass-balance approach for assessing SGD, the estimated mixing
losses are added to the net fluxes resulting in total radon flux. Dividing the total radon
fluxes by the 222Rn activity of the advecting fluids results in estimated water fluxes. The
presumed 222Rn activity in seepage water was estimated by measuring the radon activity
in piezometers, wells, and seepage meters as well as sediment equilibration experiments
(Corbett et al., 1998) where one measures the total amount of radon that a sediment can
produce into a unit amount of pore water. In general these different approaches resulted
in very uniform radon activities with an average of 173 ± 17 dpm L-1 (n=10). The well
cluster S-1 was an exception with a much higher radon activity of 359 ± 10 dpm L-1. We
excluded the results from well S-1 from the average because we believe that these higher
radon activities are not representative of this system.
We describe the SGD calculations using both mixing-loss scenarios. Submarine
groundwater discharge calculated based on mixing losses estimated from the apparent
negative net radon fluxes cycles between 0 and 34 cm d-1 with an average and standard
deviation of 11 ± 7 cm d-1. This flow is equivalent to 0-17 m3 m-1 d-1 of groundwater flux
per meter shoreline per day if we assume a seepage face of 50 m. Very similar results
were estimated using radon fluxes calculated from the radium isotope mixing approach,
when the resulting SGD ranges between -5 and 32 cm d-1 with an average of 12±7 cm d-1.
This flow is equivalent to -2 to16 m3 m-1 d-1. The apparent negative advection rates are
artifacts resulting from greater net radon losses than the average integrated value
occurring over short periods. The calculated uncertainty of the individual SGD results is
about 40 %.
The SGD results assessed from the radon model using the radium-derived mixing
are shown on Figure 4.6 where the gray area represents the estimated ±1 σ uncertainty
limits. The discharge clearly fluctuates with an apparent semidiurnal period of 12 hours.
The groundwater, and the radon that it carries, are responding to lower hydrostatic
pressure at low tides, causing increased seepage and higher radon fluxes. A comparison
between our measured net radon flux and the seepage rate measured by a dye-dilution
seepage meter (Sholkovitz et al., 2003) shows that all of the peaks in radon flux and
40
groundwater discharge occur at low tides (Fig. 4.7). This supports the theory that with a
presumably constant hydraulic gradient in the freshwater aquifer a decrease in hydrostatic
pressure due to low tide results in increased seepage. The same dynamic seepage pattern
was confirmed by several other automated seepage meters deployed at this site (Paulsen
et al., 2001).
Time/Date 2002
17-May 19-May 21-May 23-May
Adv
ectio
n R
ate
(cm
.d-1
)
-10
0
10
20
30
40
50
Figure 4.6. Fluid advection rates assessed from the radon model using the Ra-derived mixing loss. The advection rates were calculated by division of the total radon flux by our best estimate of the radon concentration in the advecting fluids (173 ± 17 dpm.L-1; see Table 4.2). The gray interval around the advection rate is the total combined uncertainty based on the errors of the analytical measurements as well as the estimated uncertainties of the atmospheric flux and mixing calculations and represent ±1σ.
A similar model could not be set up for the methane fluxes because as Table 4.1
shows, there was no significant difference between the water column and groundwater
methane concentrations. Overall, one must conclude that methane is not a good tracer of
groundwater discharge in this glacial till setting. This may be because there was not
sufficient organic matter within the aquifer matrix to produce a SGD-methane signal. In
other settings when CH4 has been a successful tracer (Bugna et al., 1996; Cable et al.,
41
1996b; Corbett et al., 1999 and Kim and Hwang, 2002) there have been significantly
greater CH4 concentrations in groundwater relative to surface water due to greater
concentration of organic materials in the aquifer matrix.
Time/Date 2002
18-May 19-May 20-May 21-May
Net
222 R
n flu
x (d
pm.m
-2.h
r-1)
-2000
-1000
0
1000
2000S
eepage meter (cm
3.cm-2.day
-1)
0
10
20
30
40
50
Wat
er le
vel (
cm)
0
50
100
150
Rn FluxWHOI SeepageTides
Figure 4.7. Plot showing net 222Rn fluxes, water level, and seepage rates measured by a dye-dilution seepage meter developed at Woods Hole Oceanographic Institution (Sholkovitz et al., 2003). Note that the maximum 222Rn fluxes and the highest measured SGD tend to occur at low tides, while the main radon losses and lowest SGD occur at high tides.
Residence time of water in West Neck Bay
We assume that the major input of radium isotopes in WNB occurs along the
coastline in the form of SGD rather than from sediment resuspension or diffusion from
the bottom of the bay. After the water mass carrying radium looses contact with its
source, in this case SGD occurring in a ~50 meter wide zone along the coast, the activity
of the short-lived radium isotopes should decrease due to radioactive decay and dilution
(mixing) with offshore waters. This principle is used in a model developed by Moore
(2000b) that gives a measure of the time elapsed since the water became enriched in
42
radium isotopes. A ratio of radium isotopes is used to calculate the water age when a
short-lived radium isotope is normalized to one of the long-lived isotopes to compensate
for dilution. Radium-223 has a longer half-life (T1/2=11.4 days) compared to 224Ra
(T1/2=3.6 days) therefore it provides a better tool for age calculation of water that is
presumably more than 10 days old at the bay inlet. In our case, we elected to use the
ratio of 223Ra to 228Ra (T1/2=5.7 years). The distribution of both isotopes is consistent
with near-shore additions by SGD and dilution with constant activity water from Shelter
Island Sound (Fig. 4.2). The activities of 223Ra and 228Ra in Shelter Island Sound are not
zero and thus these “background” activities should be subtracted from the observed
activities to determine the “excess” 223Ra and 228Ra (ex223Ra and ex228Ra) supplied from
SGD. In the model the 223Ra/228Ra ratio observed on a transect leading from the coast out
of WNB is compared to an initial 223Ra/228Ra ratio that we measured in seepage water.
The half-life of 228Ra is long with respect to the mixing time and its decay may be
neglected. We may thus estimate a radium age from Eq. 4.2:
.* *228
223
228
223223 t
iobs
eRaexRaex
RaexRaex λ−
=
(4.2)
Based on several water samples collected right along the coast at low tide and
from seepage meters we calculated an initial ex223Ra/ex228Ra ratio of 0.212 ± 0.039
(n=6). After solving equation (4.2) for time (t) we calculated that the “radium age” of the
water exiting WNB is 13 ± 4 days. This age is in good agreement with the estimated
flushing time of 11.7 days calculated by DiLorenzo and Ram (1991) by a time-dependent
box-model called STEAM.
The solution of equation (4.2) using ex224Ra rather than ex223Ra and an
ex224Ra/ex228Ra ratio of 2.3 ± 0.6 (n=6) results in a “radium age” of 7 ± 2 days. The
difference between the two calculations is likely due to the shorter half-life of 224Ra. If
the residence time is actually closer to 13 days, than >90 % of the ex224Ra will have
decayed during that time. We stress that these are apparent ages that do not consider the
effects of mixing of waters having different ages (Moore 2000b).
43
Summary
The combination of radon and radium isotopes proved to be an excellent tool for
revealing the magnitude and temporal variation of SGD in our study area in West Neck
Bay. Methane could not be used to make SGD estimates in the glacial till environment of
West Neck Bay because of its low concentration in groundwater. Radium isotopes
provided information about material transport in the embayment. Using 223Ra we
calculated a horizontal eddy diffusivity coefficient of 8.6 m2 s-1. We used this coefficient
to make an independent estimate of the mixing loss of Rn for our continuous radon
model. The resulting Rn flux by mixing was 730 dpm m-2 hr-1. This is in a good
agreement with our conservative estimate (500-1100 dpm m-2 hr-1) that is based on
inspection of the net radon fluxes in the continuous radon mass-balance model.
The resulting SGD rates from the different approaches are shown in Table 4.3.
The advection rates shown include Rn estimates using the 223Ra-mixing coefficient and
mixing by inspection and the concentration gradient of 226Ra. The table shows both the
groundwater advection rate (specific discharge, cm day-1) and flux per unit width of
shoreline (m-3 m-1 day-1). All approaches produced results that were overlapping and in
the range of seepage meter results. The continuous radon mass-balance approach using
the mixing losses from inspection of Rn fluxes produced an estimate of 6 m3 m-1 day-1
(average of 0-17 m3 m-1 day-1) and gave very valuable information about the temporal
variation of SGD. The calculated SGD flux from the radon model using the Ra-derived
mixing was also 6 m3 m-1 day-1 (average of 0-16 m3 m-1 day-1) and the flux calculated
using Ra isotopes alone was 4-7 m3 m-1 day-1. These results are in agreement with the
fluxes measured by various types of seepage meters deployed at the site. Estimates of the
shoreline flux based on manual, dye-dilution, and continuous-heat type seepage meters
were 2-6 m3 m-1 day-1. On the other hand, the ultrasonic seepage meter deployed at
different locations from 0 to 50 meters offshore measured a total integrated seepage of 18
m3 m-1 day-1. This higher value may be attributed to the influence of a pier that ran
perpendicular to the shoreline. The pilings of the pier had apparently pierced a shallow
aquitard allowing local discharge of groundwater. The ultrasonic meter was deployed at
locations near the pier and thus may have a high bias. We should also mention that the
44
tracers, measured in the water column, integrate a larger area than the seepage meters. In
addition, conductivity measurements showed that a significant portion of nearshore SGD
measured by the seepage meters was fresh water while the radiotracers integrating a
larger area than seepage meters reflected total flow comprised of both fresh and saline
water.
Despite some uncertainty about the best integrated seepage values at the site, the
continuous measurement techniques (Rn-model and seepage meters) all agree that there
is a reproducible pattern of higher SGD flux during low tides. Both the concentrations of
radon and the net radon fluxes (Figs. 4.3, 4.5 and 4.7) showed that the highest radon input
to the water column is at low tides. The same pattern was confirmed by automated
seepage meters deployed in the same area (Paulsen et al., 2001; Sholkovitz et al., 2003).
We also estimated the water residence time in WNB using radium isotopes. We
used an age equation proposed by Moore (2000b) and although not all the assumptions
required for the model were confirmed, we believe that the calculations produced
reasonable results. Our estimate of the residence time based on 223Ra was within the
uncertainty of an earlier estimate based on a tidal wedge model.
Table 4.3: Values of advection rates calculated by: (1) the continuous radon model
using a mixing term estimated by inspection of “net fluxes”; (2) the radon model using a radium derived mixing term; and (3) calculated solely by the distribution of radium isotopes. For the radon balance approach the table shows the range in specific discharge that occurred between 17 May 2002 and 23 May 2002 and the average for the entire period. The same values are also expressed as a shoreline flux of groundwater per meter shoreline per day. Also shown are SGD values measured by the WHOI dye-dilution seepage meter positioned 10 meters seaward of mean tide (inshore) and 20 meters seaward of mean tide (offshore).
Method Specific Discharge Shoreline Flux
cm.day-1 m3.m-1.day-1
Radon Inspection of Rn fluxes
Range: 0-34 Average: 11±8
0-17 6±4
Radon 223Ra mixing model
Range: 0-32 Average: 12±7
0-16 6±3
Ra Isotopes 15 7 WHOI meter - inshore WHOI meter - offshore
Range: 2-37 Range: 3-12
45
CHAPTER 5
RADON AND RADIUM ISOTOPES IN THE CHAO PHRAYA RIVER
AND ESTUARY
Introduction
Groundwater discharge is recognized as a significant, but poorly quantified source
of nutrients and dissolved species to coastal waters and estuaries. It may also be a
pathway for land derived contaminants of anthropogenic origin. One of the geographic
areas where groundwater transport to the marine environment has been scarcely studied,
but could be particularly important is in Southeast Asia. The area has a wet climate and
large regions are covered by limestone terrains that are particularly conducive to
underground flow. An understanding of submarine groundwater discharge (SGD) is
needed especially because many countries in the region are undergoing rapid agricultural
and industrial development. Such activities may have a significant impact on surface and
groundwater chemistry with potential influence on the water quality and ecology of the
coastal zone.
In recent years, progress has been made in the number of SGD investigations on
different types of coastal zones. However, the importance of groundwater inputs in river-
dominated margins is still an open question. Identification of SGD in a region with large
riverine inputs is challenging as the geochemical signals may be similar. Moore (1997),
Krest et al. (1999), and Moore and Krest (2004) have identified groundwater discharge in
the Mississippi, Atchafalaya and Ganges-Brahmaputra estuaries using radium isotopes.
In this study we focused our investigation on the Chao Phraya River that flows
into the upper Gulf of Thailand (Fig. 5.1). We combined the application of radon and
46
radium isotopes to estimate groundwater discharge and estuarine mixing in the Chao-
Phraya Estuary. An earlier investigation of barium, 226Ra and 228Ra distribution in this
estuary by Nozaki et al. (2001) indicated that there could be substantial groundwater
discharge in the area. Unfortunately, those investigators were unable to estimate the
magnitude of SGD because they had no way of estimating the residence time of waters in
the estuary - Gulf of Thailand system. In our study we measured short-lived radium
isotopes to calculate the estuary-Gulf exchange rates, and combined this information with
radon and long-lived radium isotope distributions to estimate groundwater discharge in
the estuary.
Study Site
Thailand has a tropical climate with wet and dry seasons influenced by the
southwest monsoon and by tropical cyclonic storms from the South China Sea (Stansfield
and Garrett, 1977). These monsoon patterns strongly affect the flows in the Chao Phraya
(“River of Kings”) and its tributaries. The average annual rainfall is about 900 to 1400
mm/year. The Chao Phraya is the largest river in central Thailand (Fig. 5.1). It empties
into the Gulf of Thailand and represents approximately half of the gulf’s river input. The
Chao Phraya has a large influence on industrial and agricultural development in central
Thailand. There has been extensive land use development over the last 40 years in the
river basin that included irrigation projects, engineering projects for salinity control, and
navigation. The river flows through Bangkok and a heavily industrialized region near the
estuary.
The diurnal tide in the river mouth has an amplitude of about 1 meter. Because of
the strong seasonality, we designed our sampling program for both dry (January) and wet
(July) field programs. During our dry season sampling in January 2004 at low river
discharge, we witnessed a day of unusually high water level in the river apparently
caused by an abnormally high tidal surge. The water flooded the roads stopping traffic
and when drained away it flushed the debris from the streets into the river. Such events
probably influence both the surface and the groundwater composition and quality in the
river basin aquifers.
47
.
Fi
gure
5.1
: (A
) Map
of T
haila
nd d
epic
ting
the
uppe
r Gul
f of T
haila
nd; (
B) C
hao-
Phra
ya R
iver
with
the
Janu
ary
2004
rado
n da
ta;
(C) J
uly
2004
traj
ecto
ries.
Rad
on (22
2 Rn)
act
iviti
es m
easu
red
in th
e ne
ar-s
urfa
ce w
ater
s are
plo
tted
on th
e tra
nsec
ts w
here
larg
er
circ
les r
epre
sent
hig
her a
ctiv
ities
. To
wns
of H
ua H
in a
nd S
ri R
acha
indi
cate
d on
pan
el (A
) wer
e ou
r add
ition
al st
udy
site
s. H
uaH
inSr
iRac
haB
angk
ok
48
Methods
The most widely used methods of assessment of SGD are direct measurement
techniques (seepage meters), hydrological modeling and the application of geochemical
tracers (Burnett et al., 2001a; Taniguchi et al., 2002). Radon and radium isotopes are the
most frequently used geochemical tracers. Rn-222 is a naturally occurring radioactive
element with a half-life of 3.8 days. It is a non-reactive noble gas and its only losses
from the water column are due to its radioactive decay and evasion to the atmosphere.
Because groundwater is in contact with radon emanating aquifer material, 222Rn
concentrations in groundwater are about two to three orders of magnitude higher than
most surface waters. This makes radon very useful in identifying areas of groundwater
input into lakes, rivers and the coastal ocean, and can be used to estimate the dynamics
and magnitude of groundwater discharge (Cable et al., 1996a,b; Corbett et al., 1998,
1999, 2000; Burnett et al., 2002; Burnett and Dulaiova, 2003; Chanton et al., 2003).
Naturally occurring radium isotopes are also used as tracers of SGD (Moore,
1996; Moore, 2000a,b; Krest and Harvey, 2003; Charette et al., 2001; Kelly and Moran,
2002; Rutgers van der Loeff et al., 2003) and are widely applied to study estuarine and
coastal mixing rates. The distribution and flux of radium isotopes in estuaries have been
investigated by many investigators including Li at al. (1977), Moore and Todd (1993),
Plater et al. (1995), Moore et al. (1995), Moore et al. (1996), Moore (1997), Krest et al.
(1999), Nozaki et al. (2001), and Moore and Krest (2004). Particles transported by rivers
are one of the major sources of the four naturally occurring radium isotopes in estuaries: 224Ra (half life = 3.6 days), 223Ra (11.4 days), 228Ra (5.7 years) and 226Ra (1600 years).
As river water mixes with sea water, the water’s ionic strength increases and the radium
isotopes are released from particles by ion exchange reactions. This results in the well
documented non-conservative distribution of radium versus salinity with a maximum
concentration occurring in the mixing zone (Li et al. 1977).
Sampling techniques
The sample collection and radon surveying was done on two cruises aboard the
R/V Chula-Vijai of Chulalongkorn University (Bangkok, Thailand). The dry season
sampling took place in January, 2004 at low river discharge (47 m3s-1), and the wet
49
season sampling was done in July, 2004 during a river discharge of 430 m3s-1. These
discharge figures are the sum of two gauging stations on different branches of the river
located at the Chao Phraya and Rama VI Dams. The location of these gauges is about
270 and 190 km upriver to prevent tidal effects and the discharge estimates are likely
somewhat low as additional surface and groundwater inputs to the lower reaches of the
river are ignored. In both seasons we sampled along transects leading from the river
mouth to a distance of about 48 km upriver (a low bridge prevented the ship from going
farther) and approximately 42 km from the river mouth into the upper Gulf of Thailand
(Fig. 5.1). The surveys were performed at relatively slow speed (5-7 km/hr) to ensure
good spatial resolution for our automated radon measurements.
We used a multi-detector continuous radon measurement system for radon
mapping. The automated radon system analyses 222Rn from a constant stream of water
pumped from about 1 m depth. The water passes through an air-water exchanger that
distributes radon from the running flow of water to a closed air loop. The air stream is
fed to a series of commercial radon-in-air monitors (Durridge Co., Inc. RAD-7). We
used three detectors that were arranged in parallel and they determined the concentration
of 222Rn by electrostatic collection and measurement of the α-emitting daughter, 218Po.
Since the distribution of radon at equilibrium between the air and water phases is
governed by a well-known temperature dependence, the radon concentration in the water
is easily calculated. We used a protocol that provided a new reading of the 222Rn
concentration in the surface waters every 10 minutes. Our radon mapping system also
incorporates integrated GPS navigation, depth sounding, salinity (conductivity), and
temperature measurements (Burnett et al., 2001b; Dulaiova et al., 2005). We set up
another RAD-7 detector on the boat to measure radon concentration in air along the
transect during the cruise. We received wind speed data for the period of our study from
a meteorological station of the Thai Navy in Hua Hin (westerm side of upper Gulf of
Thailand).
We also measured the content and isotopic composition of dissolved radium in the
river and gulf. We used MnO2-impregnated fibers to concentrate radium from large
volume water samples (~100 liters) collected along the same track as our radon survey.
These fibers quantitatively adsorb radium from seawater as long as the flow rate of water
50
passing through the fiber is maintained at a rate of less than 2 L/min (Moore, 1976).
Discrete water samples from the river and offshore were collected by quickly pumping
into large (>100 L) plastic barrels and then processing measured volumes of water by
slowly pumping the water through cartridges filled with ~150 cm3 of Mn fiber. We
connected two cartridges in series and put raw acrylic fiber in the upstream cartridge to
filter out the bulk of the suspended particles. The Mn-coated fiber was in the
downstream cartridge. Volumetric measurements were made using a calibrated
mechanical volume totalizer.
We also collected radium on Mn-fibers by towing fibers for 30-60 minutes in a
mesh bag behind the ship. Such samples cannot quantify the activities of Ra isotopes
because there is no volume measurement but are useful for assessing radium activity
ratios (AR). After either collection method, the excess water was squeezed out of the
fiber by hand and the fibers were sent via express courier back to our laboratories at
Florida State University (FSU), where they were washed with radium-free water to wash
the salt out of the fibers before radium measurement. We also collected the raw acrylic
fiber from the upstream cartridge containing particulate material from selected samples
and the more heavily loaded samples were also analyzed for radium activity ratios. The 223Ra and 224Ra measurements were made within a few days of sampling using a specially
designed delayed coincidence system (Moore and Arnold, 1996). The long-lived 226Ra
and 228Ra were measured at a later date by gamma spectrometry after ashing the fiber in
custom-made stainless steel crucibles that are later pressed into wafers. The samples are
aged for about 3 weeks and then counted on an intrinsic germanium detector using the 228Ac peaks at 338 and 911 keV for 228Ra and the 214Pb and 214Bi peaks at 259, 352 and
609 keV for 226Ra (Dulaiova and Burnett, 2004).
We measured CTD profiles and collected surface water samples for nutrient
analysis at each radium station. The nutrient results have been presented elsewhere
(Wattayakorn et al., 2004).
Groundwater wells from three locations were sampled for conductivity, radon, and
radium isotope concentrations. The radon was measured using a RAD-H2O system
which is basically a portable 250 mL Rn-emanation bottle attached to the RAD-7 radon-
51
in-air detector. The radium isotopes were measured from 20-L water samples using the
Mn-fiber technique.
We also collected several surface sediment samples in the eastern part of the Gulf
of Thailand that were dried and analyzed by gamma-spectrometry.
River flux measurements were made at a downstream station following a method
developed by Kjerfve (1979) during the same periods as our isotopic measurements in
January and July, 2004. The main purpose of these measurements was to estimate the
actual net river flow at a station near the mouth of the river. Hourly measurements of
water current (speed and direction) together with temperature, salinity and total
suspended solids were carried out for 49 hours at two sampling stations in a cross section
of the Chao Phraya River 18 km upstream of the river mouth. The stations were off the
left and right banks of the river at water depths of 8 and 5 meters respectively. One
observation that will have bearing on our results was that the water level at the tide gauge
on January 23, 2004 at high tide was 2.55 m above mean sea level which is 0.9 m above
the predicted tide. The water levels measured in July were close to the predicted ranges.
Results
Horizontal surface and vertical salinity profiles
The difference in fresh water flow during the two seasons had a dramatic effect on
the distribution of surface salinity in the river (Fig. 5.2). Salt intrusion was observed
more than 40 km upstream during the January sampling. In July, fresh water was found
throughout the entire river and the salt wedge was pushed seaward within a few
kilometers of the river mouth. At the farthest station offshore (42 km) in January we
measured a salinity of 31 throughout the water column, while in July the surface waters
in the same area were 27 and the bottom water 30 ppt. The gulf water was stratified in
July with an 8-meter thick mixed layer on the surface while there was little stratification
in January (Fig. 5.2).
52
Distance (km)
-60 -40 -20 0 20 40 60
Sal
inity
(ppt
)
0
5
10
15
20
25
30
35
JanuaryJuly
Salinity Temp
24 26 28 30 32
Dep
th (m
)
0
5
10
15
20
(A) (B)
Figure 5.2. (A) Horizontal salinity profile in near surface water in the river and estuary.
Zero km is at the river mouth, negative distances are upriver, positive are offshore. Differences in salinity at the same location during the same survey were due to tidal effects. (B) Salinity and temperature vertical profile at the farthest offshore station (+42 km) in July showing stratification. The January profile showed the water column well mixed at approximately the same location.
Radon surveying
Variations in freshwater discharge also made an impact on the concentrations of
radon detected in the river. The concentrations in the river were roughly 30% lower
during the high flow period in July compared to January (~4,000 compared to ~6,000
dpm m-3) (Fig. 5.3). However, the spatial pattern of the concentrations remained much
the same in the two seasons, although there was an apparent double peak in the river
during January. In both surveys, the highest concentrations of 222Rn occurred at or just
beyond the mouth of the river, and the drop off from the high values near the river’s
mouth out to sea was rapid, systematic, and nearly identical. This reduction in radon
concentration in the offshore direction is due to a combination of mixing with low-
activity waters offshore, radioactive decay, and emanation of radon into the atmosphere.
The radon activity decreases to the same level in both seasons within 10 kilometers of the
river mouth (25 ppt in January and 7 ppt in July).
July, 2004
53
Distance (km)
-40 -20 0 20 40
222 R
n &
226 R
a (d
pm.m
-3)
0
1000
2000
3000
4000
5000
6000
7000
Rn JanuaryRn July226Ra January226Ra July
Figure 5.3. Activity of 222Rn and its parent 226Ra as a function of distance in the river and
Gulf of Thailand in January and July, 2004.
The radon activity in a nearby groundwater well with zero salinity measured by
an air-water exchanger was 376,000 ± 70,000 dpm m-3 (n=6) and 372,000 ± 48,000 dpm
m-3 (n=1) measured by a RAD-H2O. We measured a radon concentration of 930,000 ±
64,000 dpm m-3 (n=4) also by RAD-H2O in a nearby well that had 1.2 ppt salinity.
Another well in the same location and depth (70 m below surface) had 594,000± 53,000
dpm m-3 (n=2) of radon. Other wells sampled in various locations in the eastern and
western Gulf of Thailand ranged from 100,000-2,000,000 dpm.m-3 (Table 5.1).
The atmospheric radon concentrations measured on the research vessel varied
between 0 and 400 dpm m-3 with the average of 180±90 dpm m-3.
Dissolved radium isotopes
The spatial distribution of dissolved radium isotopes showed a variation with river
discharge (Fig. 5.4, Tables 5.2, 5.3). Unlike radon, the distribution of radium isotopes is
expected to be strongly influenced by salinity because the salt content governs ion
exchange reactions that result in release of radium adsorbed onto particle surfaces (Li et
54
al., 1977; Moore et al., 1995). So when radium isotopes are plotted against salinity, the
concentrations are more comparable at the various salt levels rather than geographic
position. Radium isotopes were also measured in several wells in the region (Table 5.1).
Table 5.1. Dissolved radium isotopes and 222Rn activities measured in wells on land near the coastline of Gulf of Thailand. Known depths and salinities are also indicated. The “nd” indicates that we do not have depth information about the particular well. The “x” indicates pore water values obtained from surface sediments in the coastal zone, either collected from benthic chambers (S-1, NC, C-1) or determined by sediment equilibration (pw1, pw2). The benthic chambers and some of the wells (SR, HH) were located in Sri Racha in the eastern Gulf of Thailand and Hua Hin in the western coastline of the gulf (Fig. 5.1).
Well Depth Salinity 222Rn 223Ra 224Ra 226Ra 228Ra
m ppt dpm m-3 dpm m-3 dpm m-3 dpm m-3 dpm m-3 River bank
W 1 80 0 376,000 5.6 1,560 1,050 1,780 W 2 72 1.2 930,000 58 783 13,510 988 W 3 68 594,000
Sri Racha
SR 1 nd 0.4 255,000 10 820 380 460 SR 2 nd 0.6 251,000 130 2,260 760 1,480
SR S-1 x 31.1 40 490 440 380 SR NC x 30.4 240,000 1,220 32,520 1,100 3,410 SR C-1 x 31 270 8,270 1,920 2,880 SR pw1 x 885,000 SR pw2 x 440,000
Hua Hin
HH 1 100 0 2,013,000 130 5,610 750 5,190 HH 2 20 0 353,000 HH 4 5 0 100,000 120 1,200 2,120 6,260
HH S-1 x 320 8,680 390 2,690
Although the spatial distributions showed seasonal variations, the dissolved
radium isotope distributions depending on the salinity are similar in January and July.
Except for several samples that seem anomalously low in 224Ra activity in January, the 224Ra, 223Ra and 226Ra activities increased from a low activity at zero salinity to a peak
activity between about 10-17 ppt, then decreased as salinity increased further. However,
we never reached completely marine conditions or South China Sea salinities (32-33 ppt)
55
and the short-lived radium activities did not decrease to supported thorium levels as one
would expect for an offshore water mass. We found that the 223Ra and 226Ra peak
activities were higher in January (low flow) than July (high flow), a pattern that has been
observed before for other large rivers (Moore and Krest, 2004). The long-lived 228Ra, on
the other hand, reaches maximum activity in January between 25 and 31 ppt that is at
much higher salinities than the other radium isotopes. We never observed a decrease in
the 228Ra activity in July - perhaps because our transect did not extend far enough out into
the gulf. Nozaki et al. (2001) had found very similar trends of 226Ra and 228Ra versus
salinity in the Chao Phraya Estuary in their July and November 1996 samplings. These
investigators measured 226Ra activities ranging from 170 to 260 dpm m-3 in the river and
estuary respectively, and 228Ra ranging from 200 to 1100 dpm m-3. Our values for 226Ra
ranged from 100 to 330 dpm m-3, and 228Ra from 200 to 1400 dpm m-3. 22
3 Ra
(dpm
.m-3
)
0
20
40
60
80
224 R
a (d
pm.m
-3)
0
200
400
600
800
1000
1200 JanuaryJuly228Th January228Th July
Salinity
0 5 10 15 20 25 30 35
226 R
a (d
pm.m
-3)
0
100
200
300
400
500
Salinity
0 5 10 15 20 25 30 35
228 R
a (d
pm.m
-3)
0
200
400
600
800
1000
1200
1400
1600
1800
Figure 5.4. Radium isotope distribution in surface water samples plotted against salinity
in the Chao Phraya River and estuary in January and July, 2004.
56
Table 5.2. Ra isotope activities in the Chao Phraya River and estuary
Sample Salinity Distance 223Ra 224Ra 226Ra 228Ra ppt km dpm m -3 dpm m-3 dpm m-3 dpm m-3
January B 1 16.2 -17.4 63 400 310 950 B 2 13.1 -22.8 57 530 270 770 B 3 7.4 -30.2 34 680 330 600 B 4 4.4 -40.0 21 610 310 610 B 5 2.0 -46.8 15 340 160 310 B 6 0.4 -44.7 8 200 190 240 B 7 0.9 -42.1 9 190 140 200 B 8 8.6 -29.4 54 790 250 680 B 9 8.6 -29.4 27 720 370 960 B 10 11.5 -23.2 19 390 120 130 B 11 14.7 -17.2 48 290 350 930 B 12 19.4 -6.9 50 330 270 880 B 13 22.2 -0.2 50 470 270 1160 B 14 27.3 2.7 44 700 180 950 B 15 29.4 9.7 33 310 120 990 B 16 30.2 17.7 19 250 210 1240 B 17 30.2 25.9 12 140 100 790 B 18 30.9 42.2 12 100 140 910 B 19 29.0 8.2 41 580 220 1230 B 20 30.1 12.1 27 230 140 980 B 21 30.3 16.1 22 190 180 1070 B 22 30.9 19.4 20 240 120 1190 B 23 31.2 23.5 11 120 140 800 July B 24 0.1 -47.0 5 170 120 210 B 25 0.1 -33.1 5 200 110 100
B25dupl 5 210 90 120 B 26 0.2 -3.5 7 270 160 170 B 27 0.8 0.2 13 290 190 210 B 28 8.0 6.0 36 710 190 650 B 29 3.5 -2.4 27 510 220 360 B 30 8.5 4.2 34 690 250 690 B 31 19.6 9.3 27 460 120 700 B 32 25.4 17.0 31 260 180 1450 B 33 26.4 24.2 22 210 260 1390 B 34 27.0 31.6 28 190 180 1290 B 35 27.7 35.3 25 180 250 1380 B 36 28.0 42.6 16 60 210 1340 B 37 14.9 7.4 37 890 310 1100 B 39 2.5 -2.9 22 420 160 430 B 40 12.0 2.4 33 700 240 830
57
Table 5.2 – continued. Sample Salinity Distance 223Ra 224Ra 226Ra 228Ra
ppt km dpm m -3 dpm m-3 dpm m-3 dpm m-3
B 41 6.0 2.8 25 580 210 650 B 42 17.2 6.7 41 770 180 800 B 43 21.4 10.3 28 530 200 900 B 44 27.2 24.2 35 220 200 1420 B 45 25.1 13.7 38 550 220 1420
58
Table 5.3. Ra isotopes on the towed fibers
Sample Salinity Distance 223Ra 224Ra 226Ra 228Ra ppt km dpm/fiber dpm/fiber dpm/fiber dpm/fiber
January CV 1 14.0 -18.8 3.2 43 39 112 CV 2 12.9 -22.5 2.2 33 19 58 CV 3 9.6 -25.8 1.4 21 11 14 CV 4 8.0 -29.1 1.2 27 13 13 CV 5 6.4 -33.1 2.3 33 23 30 CV 6 5.6 -36.7 0.8 21 12 15 CV 7 3.8 -40.5 1.1 13 20 12 CV 8 2.4 -43.9 0.5 9 11 13 CV 9 1.7 -47.5 0.4 5 nd nd CV 10 0.7 -42.8 1.1 13 22 16 CV 11 1.6 -39.9 1.5 23 13 21 CV 12 3.5 -35.9 0.6 19 6 6 CV 13 6.2 -31.5 1.0 23 6 18 CV 14 7.6 -26.6 1.5 29 11 20 CV 15 10.8 -21.5 2.0 47 21 52 CV 16 14.3 -14.7 3.4 16 33 103 CV 17 17.1 -7.0 5.2 69 17 55 CV 18 19.1 -0.7 5.2 81 21 86 CV 19 24.4 2.4 5.0 74 17 110 CV 20 26.9 10.8 1.9 22 9 73 CV 21 27.5 18.6 2.8 28 15 138 CV 22 27.6 27.7 6.4 34 19 187 CV 23 28.3 38.5 1.9 21 18 152 CV 24 27.2 10.1 2.7 37 21 145 CV 25 27.9 14.0 4.8 28 18 106 CV 26 27.9 30.5 2.3 34 39 251 July
CV 27 0.1 -46.5 0.3 6 8 8 CV 28 0.1 -48.8 0.6 7 6 7 CV 29 0.1 -32.8 0.1 7 2 6 CV 30 0.1 -21.9 0.5 7 3 3 CV 31 0.1 -10.5 0.4 9 3 3 CV 32 0.2 -3.5 0.6 9 10 5 CV 33 0.3 1.9 0.9 32 16 26 CV 34 5.1 5.8 2.4 47 25 51 CV 35 2.4 0.3 1.2 23 16 16 CV 36 3.5 -0.7 0.8 20 17 25 CV 37 10.7 6.2 0.6 11 10 16 CV 38 18.7 10.6 0.9 9 12 21 CV 39 22.8 16.7 1.8 13 17 58 CV 40 23.6 23.6 2.5 13 27 106
59
Table 5.3 – continued.
Sample Salinity Distance 223Ra 224Ra 226Ra 228Ra ppt km dpm/fiber dpm/fiber dpm/fiber dpm/fiber
CV 41 24.2 31.6 1.4 11 21 96 CV 42 25.0 38.1 2.5 10 35 167 CV 43 24.8 38.5 6.3 29 55 331 CV 44 23.7 25.7 4.3 26 40 196 CV 45 18.9 11.9 2.0 56 15 41 CV 46 5.6 3.2 3.9 89 47 79 CV 47 4.6 0.5 2.2 41 24 36 CV 48 10.7 5.2 1.9 34 20 50 CV 49 17.9 9.4 4.4 96 43 131 CV 50 22.5 13.3 3.2 34 21 118 CV 51 24.6 20.8 0.9 9 13 56 CV 52 24.4 19.9 3.1 46 46 243 CV 53 20.3 13.4 4.5 47 34 121 CV 54 7.0 6.3 2.3 55 26 62 CV 55 0.5 -0.3 0.4 18 18 30
nd = not determined
Radium on suspended particulates and bottom sediments
Total suspended sediment (TSS) concentrations were measured at 0.2 m depth at
the fixed station located 18 km upstream from the river mouth. The amount of TSS
varied between 2 and 72 g m-3 (n=49) in January and 20 to 80 g m-3 (n=49) in July. The
average TSS during the periods of our isotopic measurements was 27 g m-3 in January
and 39 g m-3 in July. At station B25 (33 km upstream from the river mouth) we
processed duplicate water samples and we also preserved the prefilter from the cartridges.
The radium activity ratios measured on the suspended particulates from this station (zero
salinity) collected by the raw acrylic fiber are provided in Table 5.4. These ratios have
high uncertainties due to the low particulate mass collected on the raw fibers. The 234Th
(238U)/ 235U activity ratio measured by gamma-spectrometry on these particles is ~22,
nearly identical to the natural abundance ratio of 21.7 expected from the 238U/235U natural
decay chains. The 226Ra / 223Ra activity ratio is also very close to this number both on the
particles and in the river (AR=18 ± 5). The other isotopic activity ratios on the
particulates and dissolved in the water agree within the estimated uncertainties. One
exception is the 224Ra / 228Ra AR which is much higher in the dissolved state in the river
60
water than on the particulates. We were not able to calculate the specific radium isotope
activities on these samples because the prefilters were not pre-weighed, the radon
emanation coefficients from these particles are unknown, and there was evidence of
particulate losses from the fibers during sample handling. However, we were able to
estimate the particulate radium isotope activities from one river water sample that we
passed through a 0.45 µm filter. During the July sampling at the same station (B25), we
passed 11.5 liters of water through a cartridge-type filter and measured the radium
activity on the particulates that it collected. Based on the water volume and assuming
that the average TSS value (39 g m-3) was applicable to this sample, we estimated the 226Ra specific activity on the suspended sediments at about 5 dpm g-1.
The results of gamma-spectrometric analysis of sediments collected in the eastern
part of Gulf of Thailand are listed in Table 5.5. At this location the sediments are
constantly exposed to salty sea water with high ionic strength (32 ppt) and thus should
have lost their desorbable radium activity. The highest 226Ra and 228Ra concentrations
were measured in the muddy sediments.
Table 5.4. Radium activity ratios measured in zero-salinity water at station B25 located -
33 km upstream in Chao Phraya River. The particles were collected on a 0.45 µm cartridge-type filter, and the river water values are the average of the B25 samples collected in duplicate. Uncertainties represent ± 1 σ based on counting statistics.
Particles River 226Ra/228Ra 0.85 ± 0.11 0.87 ± 0.21 224Ra/228Ra 0.67 ± 0.28 1.78 ± 0.31 224Ra/223Ra 27 ± 17 37 ± 4 226Ra/223Ra 18 ± 9 18 ± 5 234Th/235U 22 ± 5 nd
nd = not determined
61
Table 5.5. Gamma-spectrometric results of sediments from the eastern shore (Sri Racha) of the upper Gulf of Thailand. Uncertainties represent ± 1 σ based on counting statistics.
Sediment 226Ra 228Ra 234Th 235U
dpm.g-1 dpm.g-1 dpm.g-1 dpm.g-1
Silty 5.4±0.6 4.3±0.7 4.7±1.0 0.4±0.1
Sand/Silt 3.5±0.2 3.4±0.4 5.9±1.1 0.9±0.2
Sandy 2.0±0.1 1.7±0.1 3.2±0.4 0.2±0.1
Discussion
Radium isotopic trends
There was an unusual distribution of 224Ra during the January sampling expressed
as an apparent double peak at salinities 10 and 27 ppt with a low point at 15 ppt (15 km
upstream). One would expect a similar trend of 224Ra versus salinity in both sampling
periods as observed for the other radium isotopes. However, in January the downstream
river flux measurements indicated a much higher apparent discharge (281 m3 s-1) relative
to the gauged values upstream (47 m3 s-1). The estimated net river flux in July (324 m3 s-
1) at that time was much closer to the gauged values (430 m3 s-1). We interpret the
apparent high flux values in January as being due to a tidal surge of gulf water entering
the lower reaches of the river. Those results indicated that there was a higher than usual
water level (water levels about 0.9 m above predicted tidal stage) and current velocity on
January 23, 2004 (our sampling dates were January 23 to 25, 2004). This surge may
have been caused by a high pressure front passing through the South China Sea and
pushing seawater into the gulf and consequently into the river. We think that the low 224Ra in between the two peaks is a reflection of “old” gulf water brought into the river
during this event. The short duration of this surge would explain why only the radium
isotope with the shortest half-life was affected. The surge could also be a reason why the
radon distribution in January shows a similar double peak as 224Ra and the area of
maximum radon activity is more landward than during our July sampling when no such
surge occurred. The similarity in the 222Rn and 224Ra horizontal profiles can be explained
by the fact that they likely have similar sources and the half-life of 222Rn (3.8 d) is very
close to that of 224Ra (3.6 d).
62
.
224Ra/223Ra
01020304050
224 R
a/22
6 Ra
01
23
45
224Ra/223Ra
01020304050
222 R
n
01
23
45
67
I
IIII
I
I
I
II
III
II
III
III
III
Janu
ary
July
Janu
ary
July
Fi
gure
5.5
. Act
ivity
ratio
s 224 R
a/22
3 Ra
plot
ted
agai
nst 22
4 Ra/
226 R
a an
d 22
2 Rn
for J
anua
ry a
nd Ju
ly.
The
thre
e gr
oups
repr
esen
t sam
ples
co
llect
ed in
the
I: R
iver
, II:
offs
hore
, III:
inte
rtida
l zon
e
63
The distribution of 228Ra plotted against salinity in both seasons revealed a
different pattern than the other radium isotopes (Fig. 5.4). In January, the 228Ra
distribution showed a peak at about 27-28 ppt, and in July its activity increased
throughout the sampled salinity range (ended at 27 ppt). The water depth at the stations
where the 228Ra maxima occurred was around 16-20 meters. We did not collect samples
in the deeper water layer so we cannot say whether the 228Ra comes from the bottom
waters or by horizontal mixing. In January, the water column throughout the sampling
interval was well mixed but in July there was a distinct stratification in the upper Gulf
with a 3 ppt salinity gradient at a depth of 8 meters. The onset of the higher 228Ra activity
in the surface water corresponds to the end of the boat channel in the intertidal zone of
the estuary where a break occurs in the water depth. This is also an area where we
observed an abrupt change in water clarity in July, 2004.
Due to the non-conservative distribution of radium isotopes against salinity, the
activity ratios of the various radium isotopes may provide better information about
possible input sources than the absolute activities. Plots of 224Ra / 223Ra against 224Ra / 226Ra and 224Ra / 223Ra versus 222Rn indicate that there are at least three distinctive water
masses with different isotopic signatures (Fig. 5.5). After examining the location of the
samples in these groups it is clear that the respective end-members are: (I) river water
with high 224Ra / 223Ra ratios and high radon concentrations; (II) Gulf of Thailand water
with very low 224Ra / 223Ra and very low radon activities; and (III) samples collected
between +0 and +10 km on the transect with intermediate 224Ra / 223Ra, low 224Ra / 226Ra
activity ratios and the highest radon concentrations.
The 224Ra / 223Ra ratios in group (I) are lower in January (average 27) than in July
(average 37). If the particles in the river originate in the same drainage area in both
seasons we would expect these ratios to be similar. The reason why we see lower river 224Ra / 223Ra ratios in January might be that we were not able to collect samples in zero
salinity river water that month. In calculations in the following sections we will assume
that the fresh water endmember for the 224Ra / 223Ra AR is the same in both seasons (37).
Group (II) has the highest 228Ra signature. There is a possibility that where the estuary
gets deeper 228Ra rich water from a more distant source mixes with the river plume by
lateral mixing. As evident from Figures 5.4 and 5.5 this water is deficient in the short-
64
lived isotopes relative to 228Ra, implying that the water mass is at least several days old
and the short-lived radium isotopes have mostly decayed.
Radon concentrations are the highest in group (III), which represents samples
from the intertidal zone called the Bangkok Bar (0 to +10 km on our transect). In this
area the water depth is 5-10 meters in the boat channel and only 1 - 2 meters on either
side in the intertidal zone. Our trajectory led through waters generally about 3 meters
deep. High radon concentrations in this area can be supported by diffusion from
sediments, sediment resuspension, groundwater discharge, or a mixture of these
processes.
Radium isotopes, whether from sediment resuspension or by groundwater
discharge in the intertidal zone might come from the shoal at 1-7 km created by particle
deposition. Such particles were at some point exposed to salty water and presumably lost
desorbable radium. Newly deposited sediments are thus depleted in radium. However,
as the sediments age the radium isotopes are regenerated by radioactive ingrowth from
their thorium parents. The isotopes 224Ra and 223Ra are regenerated on the timescale of
days, 228Ra in few years, while 226Ra ingrowth is very slow, requiring thousands of years
for significant ingrowth. Therefore any radium added to the system from benthic sources
in this area would be relatively high in short-lived isotopes, with some 228Ra but not a
significant source of 226Ra. The parent-daughter pairs of radium isotopes and their
regeneration time to 99 % equilibrium with their respective parents on particles is
indicated in Table 5.6. These calculations assume that Th isotopes are also adsorbed onto
the particle surfaces.
Table 5.6. The Ra isotope regeneration time on particles to 99 % of the desorbable
activity.
Parent - daughter Regeneration time 228Th - 224Ra 22 days 227Th - 223Ra 73 days 230Th - 226Ra 8000 years 232Th - 228Ra 37 years
65
Estuarine radium ages
The radium isotopic activity ratios in the estuary may give a measure of the time
elapsed since the water first became enriched in radium isotopes assuming no further
additions occur (Moore, 2000b). When radium isotopes desorb from the particles in the
mixing zone they stop being supported by their parent radionuclides. Assuming that the
radium enters into the coastal waters with a constant isotopic composition at least over a
period comparable to the effective mean life of the 224Ra / 223Ra activity ratio (7.8 days)
one can estimate “radium ages” of the water. Apparent radium ages may also be
calculated using the short-lived radium isotope normalized to a long-lived isotope to
compensate for dilution. Based on the 224Ra / 223Ra activity ratio, one may use the
following equation (Moore, 2000b):
t
t
iobs ee
RaRa
RaRa
223
224
223
224
223
224
λ
λ
−
−
=
, (5.1)
where [224Ra/223Ra]obs and [224Ra/223Ra]i represent the observed and initial activity ratios
of the radium source water for each data set, and λ224 and λ223 are the decay constants of 224Ra and 223Ra, respectively. For the age calculations we used the initial 224Ra / 223Ra
ratio of 37 for both sampling periods as explained earlier.
Solving for t, equation (5.1) becomes:
223224223
224
223
224
1*lnage Raλλ −
=
obs
i
RaRa
RaRa
, (5.2)
The same approach can be applied for the ratio of 224Ra and long-lived isotope 228Ra. Since the thorium-series isotopes (224Ra and 228Ra) are much more abundant in the
coastal waters off Thailand than the uranium-series isotopes (223Ra and 226Ra), estimating
radium ages based on the 224Ra / 228Ra activity ratios could be more precise, at least in
terms of the analytical uncertainties. However, as described earlier, we suspect that 228Ra
may have another source that contributes to what we observed along the transect that
would distort the 224Ra / 228Ra relationship. We have calculated ages based on these
ratios with an initial 224Ra / 228Ra AR of 2.0 (based on analyses of zero salinity river
66
water). Ra-228 has a half life of 5.7 years that is long with respect to the half-life of 224Ra as well as the mixing time of nearshore waters, so we may neglect decay of 228Ra:
t
iobs
eRaRa
RaRa
224228
224
228
224λ−
=
, (5.3)
or, in terms of Ra age:
224228
224
228
224
1*lnage Raλ
obs
i
RaRa
RaRa
= . (5.4)
The calculated apparent radium ages were plotted against distance to examine
possible relationships between position offshore and transport time (Fig. 5.6A, B). The
trends and calculated ages are similar for both isotope ratios and during both samplings,
but the age versus distance trends show more inflections in January than the smooth
trends in July. There is a clear linear relationship between radium ages and distance
offshore for both isotope pairs in July. The lack of a clear pattern in January is likely
related to the tidal surge event that was hypothesized based on the unusual tidal range, the
river flux data, and our observation of lower 224Ra values at about 15 km upstream. This
is consistent with the idea that the tidal surge brought “older” water up into the lower
reaches of the river. The anomalously “old” ages in January between ~ 10-20 km upriver
are likely a reflection of this event.
If we assume that the river-derived radium had the same initial 224Ra / 223Ra AR
for both seasons, the ages in January start to increase at ~40 km upriver. This could be
related to the lower river discharge in January and the fact that salt intrusion was
observed upstream in the river causing mixing of “old” sea water with river water. As a
result, there is a much more gradual increase in ages in the river and going offshore for
the January dataset than in July. The radium apparent ages increase from 0 to 15 days
over 80 kilometers in January but only 40 kilometers in July.
There are other factors that might influence the radium ages. In our calculation we
did not allow for any additional inputs of radium either in the intertidal area or offshore.
For example, in both seasons the higher 228Ra isotope concentration in the salinity range
of 25-27 ppt, which may have a separate source, will make the apparent radium ages in
67
that section of the transect appear older. This is likely the reason why the 224Ra / 228Ra
AR tends to give slightly greater ages compared to the ages calculated from the 224Ra / 223Ra AR.
Overall, we can conclude based on this analysis that the apparent ages offshore
are similar in both seasons and the residence time of the water in the estuary out to ~40
km offshore is approximately two weeks.
224 R
a/22
3 Ra
appa
rent
age
(day
s)
0
5
10
15
20
Distance (km)
-60 -40 -20 0 20 40 60
224 R
a/22
8 Ra
appa
rent
age
(day
s)
0
5
10
15
20
Distance (km)
-60 -40 -20 0 20 40 60
River mouth
January July
January July
(A)
(B)Tidal surgeinfluence
Figure 5.6. Apparent radium ages of water in the Chao-Phraya estuary calculated using
224Ra/223Ra and 224Ra/228Ra activity ratios for January and July, 2004.
68
Radon atmospheric evasion
The measurements of 222Rn and 224Ra made in this study provide us with a unique
opportunity to evaluate the exchange of radon between the sea surface and the
atmosphere. This exchange may be an important component of the radon mass balance
approach to estimate SGD. Radon-222 is a chemically inert noble gas with a half life
(T1/2) of 3.8 days. Its losses from shallow coastal waters occur by radioactive decay,
mixing with lower concentration offshore water masses, and degassing to the atmosphere.
Radon losses to the atmosphere are governed by molecular diffusion produced by the
concentration gradient across the air-water interface and turbulent transfer, which is
dependent on physical processes, primarily governed by wind speed. The flux (Fatm) of
radon across the air-water interface can be calculated from a formula presented by
Macintyre et al. (1995):
( )atmwatm CCkF α−= (5.5)
where Cw and Catm are the radon concentrations in surface water and air, respectively; α is
Ostwald’s solubility coefficient; and k is the gas transfer coefficient. This coefficient is a
function of kinematic viscosity, molecular diffusion, and turbulence, which is dependent
on wind speed. The evaluation of k in Macintyre’s approach is based on an empirical
relationship observed in lake studies and its application to coastal seawater environments
should be verified. Earlier studies by Elsinger and Moore (1983) indicated that the
existing models of gas transfer rates across water-air interface might not be appropriate
for all environments.
We can make an independent estimate of the calculated atmospheric radon flux
based on the dataset from the Chao Phraya estuary. Ra-224 is a radioactive alkaline earth
element with a very similar half-life (T1/2 = 3.6 days) as 222Rn; it is a radioactive
conservative tracer in seawater and stays dissolved until it decays or it is transported
away by mixing. So its losses from the water column occur in the same manner as radon,
except for atmospheric evasion. Simultaneous 222Rn and 224Ra measurements away from
a common source of the two nuclides can thus be used to determine atmospheric loss of 222Rn from seawater.
As mentioned earlier, the trends during the January sampling seem to have been
influenced by an unusual tidal surge. We will thus use the radon and radium transects
69
measured in the Chao Phraya Estuary in July for the atmospheric radon loss calculations
as both 222Rn and 224Ra activity maxima occur at the same place (+ 6 km, Fig. 5.7A).
Since 222Rn and 224Ra do not decay exactly at the same rate, we have corrected the 224Ra
values further by multiplying by the λ224/λ222 ratio of the decay constants (this makes a
difference of about 1% per day). The isotope distributions are then plotted against the
corresponding radium age in the estuary, and fitted by an exponential regression (Fig.
5.7B), resulting in the following equations and R2 values:
Radon =( 363 ± 17)*e-(0.368 ± 0.030)t R2= 0.80, (5.6)
Radium = (76 ± 5)*e-(0.186 ± 0.033)t R2= 0.88, (5.7)
where t is the apparent radium age. Radon has a steeper profile with a more negative
slope. We normalized the two equations to the same y intercept by multiplying both
sides of the radium equation by 363/76. The slope of each equation at different points
along the transect (dpm m-3 day-1) is equivalent to the rate of loss of the isotope from the
water column over time and may be calculated as the first derivative. The difference in
the slopes of the two curves at corresponding points should then equal radon loss by
atmospheric evasion (dpm m-3 day-1). These losses should vary as functions of radon
concentration, wind speed, changes in water temperature or any other factors effecting
radon evasion at the various sampling sites. The difference of the slopes at each point is
than multiplied by the depth of the surface mixed layer (8 meters) to derive a radon flux
from the water column (dpm m-2 day-1). The result is an exponential function with the
apparent radium age as a variable. Because we know the relationship between the
distance and the radium age we can plot atmospheric evasion also against distance in the
estuary (Fig. 5.8). This provides a way to compare our estimates of atmospheric evasion
based on measured 222Rn and 224Ra trends as described above to calculated losses of
radon based on the theoretical gas-exchange formulations.
Radon atmospheric losses were calculated based on the equations in Macintyre et
al. (1995) using the same radon in water concentration dataset as used for the previous
estimation. These losses were computed using the measured average wind speed (4 m s-
1), the measured Rn in water and Rn in air concentrations and the respective water
temperatures along the transect between +6 and +20 kilometers on July 17th. The results
(Fig. 5.8) show that the two methods of atmospheric evasion provide estimates that are
70
Dis
tanc
e (k
m)
-40
-20
020
40
222Rn (dpm.100L
-1)
0
100
200
300
400
500
224Ra (dpm.100L-1)02040608010
0
222 R
n22
4 Ra
App
aren
t wat
er a
ge
02
46
810
1214
0
100
200
300
400
500
224 R
a =
76*e
-0.1
86x
222 R
n =
363*
e-0.3
68x
(A)
(B)
d
Fi
gure
5.7
. Rad
on a
nd 22
4 Ra
isot
opes
plo
tted
toge
ther
(A) o
n di
ffer
ent s
cale
s to
com
pare
thei
r tre
nd a
gain
st d
ista
nce,
and
(B) p
lotte
d on
the
sam
e sc
ale
agai
nst t
he a
ppar
ent r
adiu
m a
ges i
n th
e es
tuar
y, w
here
0 a
ge is
at 6
km
. Dis
tanc
e d
on fi
gure
(A) r
epre
sent
s the
pa
rt of
the
trans
ect t
hat w
as u
sed
for t
he ra
don
loss
cal
cula
tions
.
71
within 0-30% of each other at fluxes greater than about ~25 dpm m-2 day-1. The approach
becomes imprecise where radon and radium concentrations get close to their supported
values offshore.
The much higher fluxes near the mouth of the river are a reflection of the higher 222Rn concentrations in the water in this area. The ratio of the radon atmospheric flux to
the radon inventory represents the fraction of radon lost from the water column per time.
In the estuary, this fraction is approximately 50 % of the radon inventory per day, and
drops to about 2 % at a distance of 20 km offshore.
Distance (km)
0 10 20 30 40
222 R
n ev
asio
n (d
pm.m
-2.h
r-1)
0
50
100
150
200
250
Theoretical calc.
Rn vs. 224Ra estimate
Figure 5.8. Radon losses from the water column by atmospheric evasion calculated using
a theoretical approach based on measured Rn in air and water concentrations, wind speed and water temperature (crosses); and calculated from the comparison of 222Rn and 224Ra activities along a transect (triangles).
Estimate of SGD based on a mixing model
The radon concentration in the river measured during our surveys (3,000-6,000
dpm m-3) was much higher than can be supported by its parent 226Ra that ranged from
100 to 350 dpm m-3 as dissolved and ~200 dpm m-3 in suspended particles. The
maximum total radon activity supported by its parent in the river water column is
therefore about 550 dpm m-3. Additional radon sources may include diffusion from
sediments, resuspension, and groundwater inputs. The amount of radon diffusing from
72
the bottom sediments can be estimated from an experimentally defined relationship
between 226Ra content of sediments and the corresponding measured 222Rn flux by
diffusion (Burnett et al., 2003b). That empirical relationship is based on experimental
data from several different environments (both marine and fresh), where
Flux (dpm m-2 day-1) = 495 x 226Ra conc. + 18.2. (5.8)
From gamma-spectrometric analysis of suspended particles in the river (0
salinity), we estimate that the bottom sediment 226Ra activity in the river is ~5 dpm g-1.
We take this as a maximum activity which may actually be lower in much of the river
because most sediments would have been exposed to saline water at low river discharge
during the dry season. As an estimate of 226Ra activity in the shallow intertidal zone we
used the average activity measured in sediments collected in the coastal area of Gulf of
Thailand (3.5 dpm g-1). This activity may be lower than the activity in the river bottom
because it is certain that these sediments were exposed to saline water and must have lost
any desorbable 226Ra. Radon diffusion calculated from equation (5.8) from river bottom
sediment (5 dpm g-1) would be about 2500 dpm m-2 day-1 and the diffusion from
sediments in the intertidal zone is estimated at 1,800 dpm m-2 day-1. Assuming a steady-
state system, the radon flux can also be expressed as Flux = Rn Inventory x λRn, from
which the radon inventory supported by diffusion equals 2,500 dpm m-2 day-1/0.1824 day-
1 = 13,600 dpm m-2 in the river and analogously 9,600 dpm m-2 in the intertidal zone.
These would be maximum estimates as we have not considered concurrent losses by
atmospheric evasion. We convert the inventory into concentration by dividing by the
water depth which is 8 meters on average in the river and only 3 meters in the intertidal
zone. The corresponding maximum 222Rn concentrations that can be supported by
diffusion from the sediments are thus 1,700 dpm m-3 in the river and 3,000 dpm m-3 in the
intertidal zone. Radon concentration supported by 226Ra in the water column and
diffusion from sediments totals to a maximum of 2,000 dpm m-3 and 3,400 dpm m-3 in the
river and in the intertidal zone, respectively. We observed activities as high as 5,000-
6,000 dpm m-3.
We suspected that the excess radon concentrations near the mouth of the Chao
Phraya could be a reflection of enhanced groundwater seepage in this area. It is very
likely that groundwater in this topographically very low region has a significant saline
73
component, i. e., recirculated Gulf of Thailand water. Any fresh groundwater flow would
be directed down gradient along topographic contours and thus would tend to be
concentrated towards the central axis of the river valley. The highest radon
concentrations in our surveys were consistently in the central axis of the river at or just
beyond its mouth.
Using the information from Figure 5.5 that identifies the groups influenced by the
different sources of radon and radium isotopes we are able to estimate the fraction of
groundwater discharge into the river and estuary. The radon concentration and 224Ra / 223Ra AR signature in the estuary can be combined into a 3-endmember mixing model:
fO +fR + fGD = 1.00 (5.9) 222RnOfO + 222RnRfR + 222RnGDfGD = 222RnM (5.10)
(224Ra/223Ra)OfO + (224Ra/223Ra)RfR + (224Ra/223Ra)GDfGD = (224Ra/223Ra)M (5.11)
where
f is the fraction of ocean (O), river (R), or groundwater discharge (GD) end-members; 222RnO and (224Ra/223Ra)O represent radon activity and the radium AR in the ocean end-
member; 222RnR and (224Ra/223Ra)R are the radon activity and radium AR in the river end-member; 222RnGD and (224Ra/223Ra)GD are identified as the radon activity and radium AR in the
groundwater end-member; 222RnM and (224Ra/223Ra)M represent the radon activity and radium AR measured in
individual samples. For purposes of this analysis, we have neglected radon losses by
atmospheric evasion from the water column.
The calculated supported 222Rn activity in the freshwater end-member of the river
is 2,000 dpm m-3 (3,400 in the intertidal zone) and the river values for the 224Ra/223Ra
ARs are 37 both in January and July. The ocean end-member radon equals 80 dpm m-3
(South China Sea 226Ra activity; Nozaki et al. (2001)), and the ocean 224Ra/223Ra AR
equals 4 for both seasons. The estimated groundwater 224Ra/223Ra AR equals 19 for both
seasons.
A good estimate for the groundwater end-member radon concentration is more
difficult. Although groundwater radon concentrations could be estimated from values
measured in the surrounding wells (Table 5.1), wells sampled closest to the river had very
74
different radon activities ranging from 370,000 to 930,000 dpm m-3 (measured in 3 wells,
n=11). Using a radon activity of 370,000 dpm m-3 the calculated groundwater discharge
equals 3.6 % of the river flow in January (1.7 m3 s-1) and 1.2 % in July (5 m3 s-1). We
feel that there is a high probability that any groundwater discharge consist largely of re-
circulated Gulf of Thailand water with some input from a shallower source than the
aquifer we sampled. If the effective radon concentration was lower, the groundwater
discharge values would be proportionally higher. For example, a radon activity of
100,000 dpm m-3 (measured in one of the shallow wells on land close to the coast in the
western portion of the upper Gulf of Thailand) would make groundwater discharge equal
~ 14.5 % of the river flow (6.8 m3 s-1) in January and 4.4 % (19 m3 s-1) in July. Such low
radon levels in the shallow aquifer are caused by a shorter residence time of the water and
shorter exposure to soils and sediments. While unconfirmed, these estimates show that
groundwater discharge in the river mouth may be significant. Consider, for example, that
the estimated July groundwater flow is equal to about 40 % of the total river flow in
January.
In relative terms, the groundwater discharge represents a higher fraction of the
river flow in January although the total flux is higher in the wet season. According to
salinities measured in a well sampled near the river mouth (1.2 ppt in well W 2) we
believe that groundwater consists of fresh plus recirculated seawater. Seawater enters the
aquifer especially at high tide and during tidal surge events as we experienced in January.
Tidal inputs therefore shorten the water residence time and increase the salinity in the
aquifer. Groundwater discharge in July is likely fresher because the aquifers are more
saturated with fresh water from high rainfall.
The 3-end-member mixing calculations could also be performed using a
combination of salinity and one of the tracers. However, the combination of the radon
and radium isotope AR suits the calculations better because radon can identify total
groundwater inputs while the radium isotope AR helps identify different radium sources
and responds more to saline flows. In addition, the groundwater discharge in this area is
likely dominated by saline inflows. Figure 5.9 shows the composition of the water
column as estimated fractions of river, groundwater and seawater calculated using the
100,000 dpm m-3 estimate as the groundwater radon concentration. The mixing curves
75
Fraction
0.0
0.2
0.4
0.6
0.8
1.0
Riv
er
Sea
Gro
undw
ater
Dis
char
ge
Janu
ary
Dis
tanc
e (k
m)
-40
-20
020
40
Salinity (ppt)
0102030
Dis
tanc
e (k
m)
-40
-20
020
40
July
Sea
Riv
erG
roun
dwat
er D
isch
arge
Fi
gure
5.9
: Est
imat
es c
ontri
butio
n of
rive
r, se
awat
er a
nd g
roun
dwat
er d
isch
arge
to th
e w
ater
col
umn
plot
ted
agai
nst d
ista
nce.
The
fr
actio
ns w
ere
calc
ulat
ed u
sing
a 3
-end
-mem
ber m
ixin
g m
odel
bas
ed o
n th
e re
spec
tive
conc
entra
tions
of 22
2 Rn
and
224 R
a/22
3 Ra
activ
it y ra
tios.
The
bot
tom
pan
els s
how
the
horiz
onta
l sal
inity
pro
files
dur
ing
the
two
crui
ses.
76
show that there is groundwater discharge occurring in the river that appears to peak near
the river mouth. No groundwater discharge is apparent in distances farther than +10 km
offshore.
There is a good agreement between the tracer-derived water fractions and the
salinity profile which corresponds to the fraction of seawater over distance (plotted under
the mixing curves for both seasons on Figure 5.9). The use of the isotopic tracer profiles
have several advantages over salinity values. Note, for example, that the tidal surge in
January with high seawater component at ~10-20 km upriver is evident from the tracer-
mixing curve, but is not apparent from the January salinity profile. If we wanted to use
salinity in similar mixing calculations we would have to either assume that the
groundwater was completely fresh or would have to know the fraction of re-circulated
seawater in the discharging groundwaters.
Evidence of SGD based on 226Ra balance
In order to compare our radium fluxes to Nozaki’s (2001) estimate, we calculated
the flux of 226Ra from the river during both seasons. The maximum measured 226Ra
concentrations in the Chao Phraya estuary were 370 and 308 dpm m-3 in our January and
July samplings, respectively. These maxima occur in both seasons at mid salinities. We
calculated the total 226Ra flux that leaves the estuary using an estuarine mixing model for
non-conservative elements (Officer, 1979). From the 226Ra distribution versus salinity
(Fig. 5.4) we estimated the effective river concentration as the y-intercept of the slope of
the best fit curve plotted through the upper end of the measured radium values. The
effective 226Ra concentrations estimated in this way were 420 dpm m-3 and 308 dpm m-3
in January and July, respectively. Using the river discharge of 47 and 430 m3 s-1, the
estimated total 226Ra flux from the Chao Phraya Estuary to the Gulf of Thailand is
1.71x109 and 1.14x1010 dpm day-1 in January and July, respectively. The January flux is
almost one order of magnitude lower than in July because of the lower river discharge in
January. Our July 226Ra flux is slightly lower than that estimated by Nozaki et al. (2001)
who determined 1.78x1010 dpm day-1 for a river discharge of 400 m3 s-1.
We used the measured 226Ra concentrations to estimate groundwater discharge in
the estuary. We used this isotope because of its long half-life (1600 y) there is no need to
77
make decay corrections. We will construct a 226Ra balance calculation based on the July
data to avoid the complications caused by the tidal surge in January. We selected a pie-
shaped area with an arc of 140o delineating the estuary from 2 to 24 km offshore, which
we divided into 3 sections that allowed us to more precisely estimate the water depths
and salinities in each section. This geometry is felt to most accurately portray the shape
of the river plume based on examination of aerial photographs (www.noaa.gov). The
depths were estimated from a nautical map (MapSource BlueChart Pacific v5, Garmin).
We calculated the volume of water by multiplying the area by the water depth in each
compartment and then summing the three portions (1.97x109 m3). Based on the 226Ra
activity of samples collected on our transect from 2 to 24 km (average 212 dpm m-3, n=
11) we then estimated the total 226Ra present in this volume (212 dpm m-3 x 1.97x109 m3
= 4.17x1011 dpm). We assume that this dissolved radium is derived from the river,
desorption of radium from particles, groundwater discharge, and mixing with offshore
seawater. Offshore water in the South China Sea at 33 ppt contains about 79 dpm m-3 226Ra (Nozaki et al., 2001). We used this value and the measured salinities to correct for
the 226Ra activity in the volume that is supplied by offshore seawater (1.1x1011 dpm).
The difference in the total 226Ra and 226Ra activity supplied by offshore seawater is
3.03x1011 dpm, which must be supplied by the river, desorption from particles from the
river and groundwater discharge. From Figure 5.6 we determined that the radium age at
2 km is 4.3 days and at 24 km is 14.9 days. Therefore the residence time of water in the
volume between these two points is 10.6 days and at steady-state the estuarine radium
flux (FRaEst) has to equal 3.03x1011 dpm / 10.6 days or 2.86x1010 dpm day-1. This flux
can be expressed as:
GGPRRRRaEst CQCQCQF ⋅+⋅+⋅= , (5.12)
where QR is the river discharge which was 430 m3 s-1 in July;
CR is the dissolved 226Ra activity in the river (120 dpm m-3);
CP is the 226Ra activity that desorbs from the particles transported by the river as they
encounter saline water;
QG is the groundwater discharge;
CG is 226Ra activity in the groundwater, which we estimate at 13,500 dpm m-3 based on
the measurement in the well with salinity of 1.2 ppt.
78
The equation can be rearranged to solve for groundwater discharge QG:
( )G
PRRRRaEstG C
CQCQFQ ⋅+⋅−= . (5.13)
From the particles collected on the cartridge filter at station B25 we determined
that the 226Ra released from river particles is ~200 dpm m-3. Since we did not measure
the desorbable fraction of this radium, we will assume that 100 % of the radium desorbs
to the water. This will result in a conservative estimate of groundwater discharge as this
is subtracted from the estuarine radium flux. Using 200 dpm m-3 for the desorbable 226Ra
activity, the calculated groundwater discharge calculated from equation (5.13) equals
14.4 m3 s-1. A lower desorbable fraction can be estimated from the difference between
the 226Ra concentration on the suspended particles at zero salinity (5 dpm g-1) and on
particles that have been in contact with seawater and lost their desorbable radium (3.5
dpm g-1), which is ~1.5 dpm g-1 or 60 dpm m-3 (~30 % of the total concentration on
particles). The solution of equation (5.13) with this CP results in groundwater discharge
of 18.9 m3 s-1.
While the assumption of a 100% desorbable 226Ra is probably not realistic, it
gives a conservative lower limit of the estimate of the magnitude of the groundwater
discharge. The estimated groundwater flux also depends heavily on the applied
groundwater radium concentration. In this calculation we assumed that the discharging
groundwater is saline and has a relatively high 226Ra concentration. Use of lower
activities (Table 5.1) would make the groundwater flux even higher. Our calculated
range of groundwater discharge of 14 m3 s-1 to 19 m3 s-1 is consistent with our best
estimate of 19 m3 s-1 for the July data based on the mixing model in the previous section.
Although the estimated groundwater discharge represents only a small fraction
(~4 %) of the river discharge, it does represent a significant amount of groundwater
seepage. Using an estimated width of the river mouth of ~8 km, 19 m3 s-1 translates to a
unit shoreline seepage flux of over 200 m3 m-1 day-1. We measured SGD in the Gulf of
Thailand on two sites in Sri Racha and Hua Hin as indicated on Figure 5.1 (A)
(Wattayakorn et al., 2004). SGD in the Chao Phraya estuary represents a large flow
compared to these other areas around the Gulf of Thailand (6 - 22 m3 m-1 day-1;
79
Wattayakorn et al., 2004), and Florida (20 -30 m3 m-1 day-1; Burnett et al., 2002; Burnett
and Dulaiova, 2003).
Conclusions and Applications of the Tracer Results
Natural geochemical tracers proved to be very useful in assessing groundwater
discharge into the Chao Phraya River and in estimating the exchange dynamics of the
estuary. The radon and radium transects suggested that there are at least three different
sources of these tracers into the river and estuary. We calculated groundwater discharge
via a tracer mixing model using 222Rn and the 224Ra/223Ra activity ratios. Our estimates
are 6.8 and 19 m3 s-1 for January and July, respectively. Another independent estimate of
groundwater discharge in July using a 226Ra balance approach resulted in values between
14 and 19 m3 s-1 depending upon the estimated desorbable 226Ra from particles
transported by the river. The groundwater inputs therefore represent at least 5% of the
river flow in January and 1% of the river flow in July, 2004. Using less conservative
assumptions these numbers could be as high as 10 and 5 % for January and July
sampling, respectively. The unit shoreline fluxes in this river-dominated area (~200 m3
m-1 day-1 in July) are much higher than other areas in the Gulf of Thailand and elsewhere.
We estimated that the apparent radium age at about 40 km offshore is 15 days
from 224Ra/223Ra AR. This is very similar to 224Ra/223Ra apparent radium ages measured
in the Mississippi plume, where Moore and Krest (2004) estimated the apparent age of 14
days for samples 50 km offshore from the river mouth.
The Chao Phraya River is a very contaminated river with extremely high nutrient
concentrations from industrial and domestic activities in Bangkok. This geochemical
tracer study has shown that groundwater fluxes contribute significantly to the isotopic
balance of the estuary. Presumably, other constituents in the estuarine waters are also
significantly effected by groundwater inputs.
80
CHAPTER 6
EVALUATION OF THE FLUSHING RATES OF APALACHICOLA
BAY, FLORIDA VIA NATURAL GEOCHEMICAL TRACERS
Introduction and Study Area
Estuarine residence time and salinity variations directly influence the ecological
conditions and production rates of estuaries. Circulation models that are based on
monitoring of river discharge, tides, weather conditions, and salinity distribution have
traditionally been used to evaluate the residence time of water within estuaries. In our
study, we applied natural radium isotopes to trace the direction and speed of water
movement in a shallow estuary in northwestern Florida. Radium isotopes have
previously been applied as tracers to determine transport rates in large estuaries, bays,
and sounds (Moore, 1984; Krest et al., 1999; Torgensen et al., 1996; Turekian et al.,
1996; Kelly and Moran, 2002).
We used Apalachicola Bay, Florida as our study site. The bay is ideal for this
application because of the availability of a 3-D circulation model for comparison (Huang
et al., 2002a, b) and the excellent logistics afforded by the Apalachicola National
Estuarine Research Reserve (ANERR). This bay is a sub-tropical, shallow estuary in the
northeastern Gulf of Mexico (GOM), surrounded by a chain of barrier islands. The main
source of fresh water to the bay is from the Apalachicola River and it exchanges water
with the GOM at four inlets: St. George Sound, Sike’s Cut, West Pass and Indian Pass
(Fig. 6.1). River discharge and exchange through these inlets are important factors for
determining the flushing rate and the salinity variations in the bay.
81
Figure 6.1. Map of Apalachicola Bay, Florida showing the Apalachicola River and the
four passes to the Gulf of Mexico. Our sampling stations are marked by diamonds. The groundwater study site at the St. George Island State Park is also indicated.
Currents in the bay are primarily tidal (Huang et al., 2002b), but are strongly
affected by wind direction and speed, and the river flow (Huang et al., 2002a; Dawson,
1955; Niu, et al., 1998). Huang et al. (2002b) developed a tidal circulation model of the
estuary, and Mortazavi et al. (2000, 2001) calculated the water export from the estuary to
the Gulf of Mexico with a 3-dimensional numerical model based on freshwater inflow,
tidal stage, temperature, salinity, and wind-stress forcing. During Mortazavi’s two-year
study in 1994 and 1995 they found that the water from St. George Sound (east of the bay)
accounted for all the seawater input from the GOM, except in June, 1995 when Indian
Pass, on the western side, contributed some seawater. Their results indicated that 69% of
the water outflow from the estuary generally occurred through West Pass. However, in
April and June, 1995, St. George Sound was the major outlet for the bay water. The
water residence time within the estuary during their study period in 1994 and 1995 varied
from 2 to 12 days, while the river discharge ranged from 300 to 2750 m3 s-1.
General Approach
We based our tracer study on a method pioneered by Moore (2000b) who used
natural radium isotopes to derive “ages” of continental shelf waters and large river
82
plumes (Moore and Krest, 2004; Moore and Todd, 1993). The source of 223Ra, 224Ra, 226Ra and 228Ra isotopes to coastal waters may include river discharge of dissolved
radium, diffusion from sediments, discharge of salty groundwater, and desorption from
suspended particles in the case of estuaries. As particles from the river encounter
saltwater, radium isotopes are desorbed from particles due to the increase of ionic
strength of the surrounding waters. Unlike radium isotopes, the thorium parents stay
attached to these particles because of thorium’s highly particle-reactive nature. The
dissolved radium isotopes are therefore no longer supported by the parent nuclides and
decay according to their own decay constants.
The half-lives of the four naturally occurring radium isotopes are 224Ra = 3.6
days, 223Ra = 11.4 days, 228Ra = 5.7 years, and 226Ra = 1600 years. Apparent radium ages
may be calculated using a ratio of the two short-lived radium isotopes or one of the short-
lived isotopes to the long-lived one. The approach assumes that radium is added from a
source with a fixed isotopic composition and that the change in this ratio occurs solely by
radioactive decay. In small estuaries like Apalachicola Bay where the expected range of
the flushing rate is 2 to 12 days (Mortazavi et al., 2000) the ratio of 224Ra/223Ra is the
most useful, because both isotopes have half-lives on the order of days. As the water
mass moves away from its source and ages, 224Ra decays faster than 223Ra. Assuming
that the river supplies the estuary with a constant radium isotopic composition at least
over a period comparable to the effective mean life of the 224Ra/223Ra activity ratio (7.8
days), one can estimate apparent radium ages of the water (Moore, 2000b):
t
t
iobs ee
RaRa
RaRa
223
224
223
224
223
224
λ
λ
−
−
=
, (6.1)
where [224Ra/223Ra]obs and [224Ra/223Ra]i represent the observed and initial activity ratios
of the radium, and λ224 and λ223 are the decay constants of 224Ra and 223Ra, respectively.
The equation can be rearranged to solve for t which represents the “radium age” i.e., the
amount of time since radium was added to the water:
83
223224223
224
223
224
1*ln t λλ −
=
obs
i
RaRa
RaRa
. (6.2)
The basic assumptions for the age model are: (1) there is a single major source of 224Ra and 223Ra into the estuary; (2) the source supplies a constant 224Ra / 223Ra activity
ratio; (3) the losses of radium after leaving the source are only by dilution by waters with
no excess 223Ra or 224Ra (which does not effect the ratio) and radioactive decay; and (4)
the Gulf of Mexico water contains negligible amounts of excess 224Ra and 223Ra. The
excess 224Ra refers to radium activities unsupported by its dissolved thorium parent 228Th
that occurs in very small amounts in the water. All 224Ra / 223Ra activity ratios hereafter
refer to 224Ra in excess to its thorium parent.
In the case of shallow, bar-built estuaries, like Apalachicola Bay, there are several
reasons why assumptions (1) and (2) have to be verified. The average depth of
Apalachicola Bay is only about 2 meters. Inputs of newly regenerated radium from
sediments via resuspension or diffusion could alter the radium isotopic ratios and make
the apparent radium ages appear younger. The bottom sediments in the river delta and
most of the bay consist of medium to fine sand, silt and clay. About 80 % of the bay area
consists of soft, unvegetated sediments and only 20 % is covered by oyster reefs and
submerged aquatic vegetation (Livingston, 1983). The muddy sediments in this shallow
bay may be disturbed by wind events allowing additional radium input into the water by
sediment resuspension.
Apalachicola River drains about a 50,000 km2 land area. The historical river
discharge records investigated back to 1978 show an annual long-term average of
710±200 m3 s-1 making this the river with the highest discharge in Florida (data from
USGS, Figure 6.2). The river discharge is usually highest from January to April, and
lowest in the summer months. The discharge over the past 27 years ranged from 180 to
2750 m3 s-1. Would the isotopic composition of radium remain constant in spite of such
large variations in discharge?
Another source of radium into the bay could be groundwater discharge that may
occur along the coastline of the land and barrier islands. To locate areas influenced by
84
submarine groundwater discharge and estimate its magnitude we made a survey along the
entire coastline during which we monitored the radon concentration and radium isotope
ratios in the water. High radon and radium isotope ratios would indicate areas where
groundwater may be entering the bay floor. We also investigated groundwater discharge
in more detail in a study site located on the bay side of St. George Island (Fig. 6.1) using
automated seepage meters (Taniguchi and Fukuo, 1993) and time-series radon
experiments (Burnett et al., 2001b; Burnett and Dulaiova, 2003).
Year
1980 1985 1990 1995 2000 2005
Apal
achi
cola
Riv
er D
isch
arge
(m3 s
-1)
0
1000
2000
3000
4000
Jul Oct Jan Apr Jul Oct Janm
3 s-1
0
500
1000
1500
2004 2005
Figure 6.2. Monthly historical discharge of Apalachicola River measured near Sumatra, Florida since 1980. The gauge (USGS 02359170) is about 33 km upstream from the river mouth. The circles on the expanded part of the plot indicate sampling time and discharge during our seasonal study.
Sampling and Measurement
Sampling plan
In order to calculate the apparent radium ages in the bay we arranged sampling at
17 stations in Apalachicola Bay as indicated on Figure 6.1. Because of large salinity and
water compositional changes over short periods at many of these stations due to tidal
circulation, we felt that grab samples would not be a good representation of the average
85
state of the bay. Instead, we wished to obtain radium activity ratios (AR) that would
integrate a signal over a period of at least one tidal cycle.
To confirm whether inputs from bottom sediments are important, we monitored
the radium isotopes in bottom water layers as well as surface waters during our sampling.
We also sampled zero salinity river water in a time series to determine if the river
supplies a constant 224Ra / 223Ra activity ratio.
Dissolved radium isotope ratios measured in river and bay waters
We sampled radium isotopes in the bay on four occasions, July 30 - August 4,
2003; March 3 - 4, 2004; August 30 - 31, 2004; and January 25 – 26, 2005. We
distributed passive radium collectors called “Mn fibers” (Moore, 1976) at the same 17
locations within the estuary during each sampling trip. Two stations were in the river,
one at the river mouth, one in East Bay, and 13 were positioned systematically across
Apalachicola Bay (Fig. 6.1). Each station consisted of a float attached by a line to a
cinder block that served as an anchor. We attached a mesh bag filled with ~ 200 ml of
MnO2-coated acrylic fiber to the line about 0.3 m from the float (Fig. 6.3) so when it was
deployed it collected radium from 0.3 m below the water surface. We also attached
several of these passive collectors 0.3 m above the bottom of the block to sample the
deeper layers. In August 2003 and March 2004 we sampled the bottom waters only at
selected stations, while in August 2004 and January 2005 we attached bags to the bottom
at every station. These passive collectors were deployed for 4 days in August 2003 and
only one day during the other seasons. At each deployment we measured the water
salinity at the top and bottom of the station with a YSI 85 conductivity probe.
During the deployment the mesh-bags with fibers accumulated large amounts of
suspended particles. After we gathered the fibers from the stations, the accumulated
particles were washed out by rinsing the fibers in radium-free water. The fibers were
then immediately prepared for the measurement of the short-lived radium isotopes.
In March, 2004 during our fiber deployment, we also collected grab samples at
stations 3 (river mouth) and 12 (central bay). We passed measured volumes of water
from these two stations through a column filled with Mn-fiber and used raw acrylic fiber
upstream from the Mn-fiber as a prefilter.
86
Figure 6.3. Design of moored buoys for deployment of Mn-fibers.
To confirm our assumption that the river 224Ra / 223Ra AR is constant over at least
a period of several days, we made a time series of measurements in the river between
August 24 and 31, 2004. We selected a dock 7 km upstream where the saltwater
intrusion does not extend in the river and thus no significant radium desorption should
occur from suspended particles. We attached a mesh bag filled with Mn-fiber to a line
that was tied to the dock and immersed it in the river for 2 days. We repeatedly sampled
4 times over an 8-day period, with the last two bags being deployed in duplicate.
We collected a river water sample to investigate the difference between filtered
and unfiltered aliquots, since the fibers deployed at the stations in the river and bay will
collect radium from unfiltered waters. In September 2003 we collected ~200 liters of
river water into two barrels. We pumped about 40 l of each sample through a 0.45 µm
filter cartridge and we passed the rest through an Mn-fiber column without filtration.
Because of logistic reasons we did not sample outside the bay area so we do not
have information about the radium isotopes outside of the inlets of the bay. According to
the measurements of Moore and Krest (2004) in surface water 80 km offshore of the
Mississippi River and Moore (2003) 28 km offshore in Apalachee Bay, short-lived
radium isotope activities in the Gulf of Mexico are equal to their radioactive parents and
excess radium activities are negligible.
87
Radium isotopes measured on suspended sediments and sediments
In August 2004, we collected water samples in three different locations in the
river. At each location we pumped about 160 liters of water into a barrel and also
collected about 140 liters into several 20-liter collapsible plastic containers. A one liter
aliquot of the river water from each station was set aside to determine the concentration
of suspended particulate matter by filtering the water through pre-weighed filters. The
water from the barrels was pumped through a Mn-fiber column with raw acrylic fiber
used as prefilter. We saved both the Mn-fibers and the prefilters with the collected
particles for radium analysis.
The plastic containers were moved to the ANERR laboratory in Eastpoint where
we passed the water through a continuous-flow centrifuge system (Contifuge Stratos,
manufactured by Kendro Laboratory Products) to separate river particles from the water.
We used a relative centrifugal force (g-force) of 20,000 g and passed the water
continuously through the system at a flow-rate of 0.4 l min-1. Under these conditions the
centrifuge separates and retains particles larger than 2 µm. We collected particles from
all three samples and saved them wet for later processing.
We also collected sediment from the river bottom at two stations. We took one
grab sample of sandy sediment at 8 km upstream from the river mouth and we collected
one silty sample at 2.6 km upstream. These bottom sediments were dried at 60oC,
homogenized and packed into 100 ml containers that were counted on a germanium
detector for determination of radium isotopes. We did not collected sediments in the bay
because these sediments are isolated from the surface water due to the water stratification
in the bay.
Radioanalytical measurement techniques
The short-lived radium isotopes 224Ra and 223Ra collected on fibers were
measured on a delayed coincidence counting system. The Mn-fibers were placed in a
measurement column and partially dried with a stream of air to adjust the moisture of the
fiber, to achieve optimal radon emanation during counting. During the measurement
helium is circulated through the column to sweep the radon daughters of 224Ra and 223Ra
to a scintillation cell. The cell counts alpha-decays of 219Rn and 220Rn and their short-
88
lived polonium daughters (215Po and 216Po). The events are registered by a
photomultiplier tube and the counts are sorted by a delayed coincidence circuit, based on
the decay constants of the polonium daughters (Moore and Arnold, 1996).
The fibers deployed for four days in August 2003 collected very high radium
concentrations and had to be allowed to decay for several days before the first
measurement could be done. All other deployments (~1 day each) were counted very
shortly after collection. After the first counting, the fibers from all deployments were
measured on the counting system again within 7 to 10 days when the initial high 224Ra
activity partially decayed and we were able to obtain a more precise 223Ra value. The
fibers were recounted again after about one month for 228Th collected on the fiber. The 228Th was subtracted from the measured 224Ra to estimate the excess 224Ra activity.
The radium sorption capacity of the Mn-fibers might have reached maximum
during the 4-day deployment causing an uncertainty in the effective deployment period.
This would not be important for the 1-day deployments. We assumed a linear radium
uptake over the time of deployment in the bay for all four samplings.
Assuming a linear radium uptake model the measured 223Ra and 224Ra were decay
corrected to the time of sampling. In the case of the passive collection we applied a
combined decay correction term to consider differences in decay rates while the deployed
Mn-fiber collected radium from the water:
22241224 11 2224
224224 ttmi et
eAA λλ
λ−− −
××= , (6.3)
22231223 11 2223
223223 ttmi et
eAA λλ
λ−− −
××= , (6.4)
where,
A224i and A223i are the 224Ra and 223Ra activities at the time of deployment;
A224m and A223m are the 224Ra and 223Ra activities at the time of measurement;
λ224 and λ223 are the 224Ra and 223Ra decay constants;
t1 is the time between the end of sampling and the mid-point of measurement; and
t2 is the duration of deployment of the Mn fiber in the water.
The long-lived 226Ra and 228Ra on the fibers were measured at a later date by
gamma-spectrometry after ashing the fibers in custom-made stainless steel crucibles. The
89
crucibles were than closed, sealed, and aged for about 3 weeks to allow the 222Rn
daughters 214Pb and 214Bi grow to approach secular equilibrium with 226Ra. The samples
were then counted on an intrinsic germanium detector (Dulaiova and Burnett, 2004).
The raw acrylic fibers that served as prefilters were also kept and we measured
them for 223Ra and 224Ra on the delayed coincidence counter. The prefilters were packed
into the measurement columns and measured in the same manner as the Mn-fibers. We
re-measured the prefilters several times over a one-month period to see if the 224Ra and 223Ra were supported. The prefilters were then ashed, packed, and counted for long-lived
radium isotopes in the same manner as the Mn-fibers.
The three samples of river suspended particles collected using the continuous
centrifuge were kept wet and we packed them into a measurement flask that consisted of
an Erlenmayer flask, a rubber stopper with two copper tubes, and Tygon tubing attached
to the copper tubes. About 2 grams of wet sediments were weighed into each of the
flasks. A Tygon tubing reaching 1 cm from the bottom of the flask was attached to one
of the copper tubes on the bottom of the rubber stopper. We then sealed the flask with the
stopper, and attached Tygon tubing to the top of the two pieces of copper tubes. These
two tubes were connected to the delayed coincidence counter and allowed the helium to
circulate through the flask and sweep the radon isotopes into the scintillation cell. We
did not calibrate the system, so the obtained results served only as approximate values to
obtain the short-lived radium AR. After the measurement we re-wet the sediments and
counted them repeatedly several times over a one-month period to evaluate the amount of 224Ra and 223Ra in excess of their thorium parents. The sediments were then dried and
packed for gamma-spectrometry into small plastic vials and were measured using a well-
type germanium detector.
Radon and radium isotope survey on the periphery of Apalachicola Bay
We performed a spatial radon and towed Mn-fiber survey on the periphery of
Apalachicola Bay over two days in July, 2003. The survey was done aboard the R/V C-
Hawk of Apalachicola NERR at a relatively slow speed (5-7 km/hr) to ensure good
spatial resolution for our automated radon measurements. We used a multi-detector
continuous radon measurement system for radon mapping (Burnett et al., 2001b;
90
Dulaiova et al., 2005). The radon results were matched to the GPS locations and a radon
map was created along the shoreline of Apalachicola Bay.
Figure 6.4. Radon and radium isotope ratio levels measured along the coastline of
Apalachicola Bay during July16 - 17, 2003 (A) Radon concentrations (dpm m-3) measured by a Rn-surveying system; (B) 224Ra/223Ra activity ratios measured on towed Mn-fibers, where one towed fiber is represented by 8 points. The darker colors and larger circles represent higher activities (Rn) and activity ratios (Ra).
B
A
91
Results
Initial tracer survey
The results of the radon survey reveal that the highest radon activity (14,000 dpm
m-3) occurs near the river mouth (Fig. 6.4A). It is possible that there is enhanced
groundwater seepage either immediately in this area or farther upstream. The survey
showed very low radon concentration (2,000 dpm m-3) in all the other parts of the bay.
The radium isotope ratios measured on the towed fibers also showed low AR without any
higher values along the coastline except near the river mouth. The 224Ra / 223Ra ARs
ranged from 4 in St. George Sound, ~6 along the coastlines in the east part of the Bay,
and up to 9 along the coastline in the western part of the bay. The ratios in the river
mouth area and in East Bay were 16 and 17, respectively (Fig. 6.4B). Overall, based on
this survey and some seepage meter studies on the barrier island (Schneider and Kruse,
2005 and our unpublished results, see Appendix) it appears that groundwater inputs of Ra
into Apalachicola Bay are very minor compared to the river flux. However, higher radon
activities and 224Ra / 223Ra ARs during the radon survey revealed a possible groundwater
source near the river mouth. SGD in this area would influence the apparent radium ages
only if its 224Ra / 223Ra AR was different from that in the river. Otherwise the
groundwater input around the river mouth can be regarded as part of the river input.
Environmental Parameters and radium isotopes
River discharge values measured at a gauging station near Sumatra, Fl at about 30
km from the river mouth (USGS 02359170) were obtained from a website maintained by
the USGS
http://nwis.waterdata.usgs.gov/fl/nwis/dv?format=html&period=730&site_no=02359170.
The river discharge during our four main samplings ranged from 338 to1016 m3 s-1 (Fig.
6.2). Wind and water level data (Table 6.1) were gathered from a station (#8728690)
operated under NOAA’s Center for Operational Oceanographic Products and Services
(CO-OPS) in Apalachicola, Fl (www.tidesonline.nos.noaa.gov/geographic.html).
The salinities measured at our 17 stations ranged from 0 ppt in the river to 31 ppt
in East Bay, with higher salinities in the bottom water layers. The water column was
stratified at most stations in the bay (Table 6.2). The activities of radium isotopes
92
collected on the fibers (dpm/fiber) at the 17 stations during the four deployments are
listed in Table 6.2. The collectors were deployed for over four days in August 2003 and
some of the fibers collected more than 1000 dpm/fiber 224Ra. The fiber collectors in the
other seasons were deployed only for 24 hours and collected significantly less activities
(224Ra ranged from ~10 to a maximum of 800 dpm/fiber). The radium values were used
to calculate the 224Ra / 223Ra ARs at each station. The uncertainty of each ratio was
calculated from the individual measurement uncertainties of 224Ra and 223Ra via error
propagation (1 σ) and represent approximately 10-15 % of the reported values. The river 224Ra / 223Ra AR calculated as the average of stations 1, 2, and 3 were 28±6 in August,
2003; 22±6 in March, 2004; 25±6 in August, 2004 (we did not include station 3 because
it had a salinity of 2 ppt at that time); and 25±9 in January, 2005 .
The results (in dpm/fiber) of the time series measurements from the upstream
river dock are listed in Table 6.3 (sample series RB). Samples RB-3 and 4, and RB-5 and
6 are duplicates. The 224Ra / 223Ra ARs on these fibers agree within the calculated
uncertainties and their average value is 23±5, essentially the same as the average values
for the river mouth stations. Radium isotope activities from known water volumes (dpm
m-3) are listed in Table 6.3 as samples series RS and RA. The average radium isotope
concentrations in the unfiltered river samples are: 223Ra =2.2 ± 1.2 dpm m-3, 224Ra =54 ±
15 dpm m-3, 226Ra =12 ± 7 dpm m-3, and 228Ra =31 ± 13 dpm m-3. The average 224Ra / 223Ra AR on the unfiltered samples is 27 ± 7. The amount of suspended sediment
material in the river in August, 2004 based on three measurements from zero salinity
water was 12.1 ± 0.8 mg L-1. The results of the prefilter and suspended sediment
analyses are shown in Table 6.4. Due to uncertainty in the sample mass and emanation
efficiency during the delayed coincidence counting, the results are not given in absolute
concentration but as dpm/sample. The uncertainties in the sample mass and emanation
efficiency did not influence the measured radium activity ratios. The average 224Ra / 223Ra AR in the suspended sediments is 23 ± 5, the same value as the average from the
time-series experiment that should represent dissolved radium. Table 6.5 shows the
absolute radium isotope activities determined in Apalachicola River and Bay water
(stations 3 and 12) and on the prefilters with the particles in March, 2004.
93
Tabl
e 6.
1: S
ampl
ing
date
s, A
pala
chic
ola
Riv
er d
isch
arge
cal
cula
ted
as a
n av
erag
e of
the
flow
thre
e da
ys b
efor
e an
d du
ring
our
sam
plin
g, w
ind
spee
d an
d di
rect
ion
mea
sure
d at
NO
AA
CO
-OPS
stat
ion
#872
8690
loca
ted
in A
pala
chic
ola,
Flo
rida.
The
list
ed
win
d sp
eed
and
dire
ctio
n in
clud
es m
easu
rem
ents
two
days
bef
ore
and
durin
g ou
r sam
plin
g. T
he ti
dal s
tage
was
mea
sure
d at
st
atio
n #8
7286
90.
The
indi
cate
d sa
liniti
es w
ere
mea
sure
d ne
arby
the
thre
e m
ain
outle
ts o
f Apa
lach
icol
a B
ay: a
t sta
tion
6 by
In
dian
Pas
s, at
stat
ion
8 by
Wes
t Pas
s and
stat
ion
10 a
t Sik
e’s C
ut a
s ind
icat
ed o
n Fi
g. 6
.1.
Fi
ber
depl
oym
ent
Dat
e*
Riv
er
Dis
char
ge
W
ind
Spee
d D
omin
ant w
ind
dire
ctio
n T
idal
ran
ge
max
/ m
in
Salin
ity a
t st
atio
ns 6
/ 8
/ 10
m3 s-1
m
s-1
m
pp
t 1
8/1/
2003
10
16
0 –
6 di
urna
l 0.
54 /
-0.0
1 6.
4 / 1
6 / 7
.3
2 3/
3/20
04
966
2 –
6 E
0.42
/ -0
.07
3.9
/ 16.
8 / 3
0.3
3 8/
30/2
004
338
0 –
5 di
urna
l 0.
62 /
-0.0
2 18
.6 /
24 /
19.8
4
1/25
/200
5 80
9 2
- 11
W
0.13
/ -0
.30
27 /
14.9
/ 9.
8
10
/9/2
003
400
8/
16/2
004
438
*sam
plin
gs o
n 10
/9/2
004
and
8/16
/200
4 w
ere
from
Apa
lach
icol
a R
iver
94
Table 6.2. Salinities measured during the buoy deployments in the surface and near bottom waters at each station and the radium isotopes measured on the deployed fibers. Uncertainties represent ± 1 σ based on counting statistics.
July 30 - August 4, 2003
Sample*
Salinity surface
Salinity bottom
223Ra
ex. 224Ra
226Ra
228Ra ppt ppt dpm/fiber dpm/fiber dpm/fiber dpm/fiber
AB 1 0.0 0.0 7 ± 1 192 ± 7 219 ± 5 198 ± 12 AB 2 0.1 0.1 4 ± 1 102 ± 5 139 ± 10 130 ± 12 AB 3 0.1 6 ± 1 199 ± 25 254 ± 12 246 ± 16 AB 4 0.0 8 ± 1 62 ± 4 55 ± 5 73 ± 2 AB 5 3.0 21.0 19 ± 2 318 ± 13 297 ± 5 382 ± 30 AB 6 13.0 20.0 63 ± 4 573 ± 21 447 ± 19 777 ± 24
AB 6B 21 ± 2 306 ± 14 154 ± 2 289 ± 1 AB 7 7.0 22.0 66 ± 6 728 ± 22 545 ± 5 904 ± 5
AB 7B 17 ± 3 240 ± 9 108 ± 4 211 ± 1 AB 8 12.0 28.0 93 ± 8 1071 ± 37 909 ± 22 1622 ± 123 AB 9 13.0 21.0 99 ± 6 1043 ± 42 894 ± 48 1480 ± 104
AB 9B 18 ± 2 196 ± 8 113 ± 7 175 ± 26 AB 10 12.0 24.0 78 ± 5 852 ± 29 703 ± 7 1110 ± 52
AB 10B 18 ± 3 171 ± 8 147 ± 11 174 ± 5 AB 11 4.1 22.0 29 ± 3 425 ± 17 351 ± 9 515 ± 11 AB 12 1.5 17.7 16 ± 2 323 ± 12 301 ± 9 358 ± 48
AB 12B 18 ± 2 272 ± 15 176 ± 4 254 ± 9 AB 13 2.7 19.3 34 ± 3 464 ± 21 458 ± 10 623 ± 32 AB 14 12.0 13.0 nd nd nd nd AB 15 15.0 21.0 65 ± 5 871 ± 37 778 ± 21 1194 ± 52
AB 15B 27 ± 3 220 ± 13 187 ± 15 195 ± 24 AB 16 20.0 29.0 102 ± 7 979 ± 38 1013 ± 10 1451 ± 107 AB 17 20.0 29.0 158 ± 10 1207 ± 48 1277 ± 4 1612 ± 31
AB 17B 20 ± 3 236 ± 15 nd nd *B=near bottom measurement nd means the values were not determined
95
Table 6.2 – continued. March 3 - 4, 2004
Sample
Salinity surface
Salinity bottom
223Ra
ex. 224Ra
226Ra
228Ra ppt ppt dpm/fiber dpm/fiber dpm/fiber dpm/fiber
AB 1 0 0 1.7 ± 0.3 30 ± 11 31 ± 1 37 ± 4 AB 2 0.1 0.1 0.9 ± 0.1 16 ± 6 14 ± 2 12 ± 2 AB 3 0 0.4 ± 0.4 16 ± 6 18 ± 3 12 ± 3 AB 4 0 0.5 ± 0.1 7 ± 2 10 ± 2 7 ± 1 AB 5 0.9 1.1 1.4 ± 0.8 32 ± 11 43 ± 11 40 ± 4 AB 6 3.9 3.9 1.2 ± 0.6 21 ± 8 20 ± 4 15 ± 2
AB 6B 0.8 ± 0.1 13 ± 5 17 ± 5 16 ± 5 AB 7 5.1 6.8 1.6 ± 0.6 28 ± 10 44 ± 7 39 ± 4
AB 7B 0.6 ± 0.2 6 ± 2 15 ± 9 6 ± 1 AB 8 16.8 23.1 11 ± 2 107 ± 39 135 ± 13 129 ± 14 AB 9 14.2 17.2 5 ± 1 34 ± 13 52 ± 6 47 ± 5
AB 10 30.3 30.4 5 ± 2 52 ± 19 75 ± 2 82 ± 10 AB 10B 1.5 ± 0.3 10 ± 4 24 ± 14 14 ± 3 AB 11 6.2 6.3 8 ± 2 72 ± 28 83 ± 12 86 ± 12 AB 12 4.1 4.2 ± 0.7 85 ± 30 105 ± 16 100 ± 3
AB 12B 1.4 ± 1.0 26 ± 10 34 ± 10 28 ± 6 AB 13 9.6 25.3 5 ± 2 52 ± 22 74 ± 5 70 ± 2 AB 14 24.6 24.6 19 ± 7 129 ± 52 181 ± 18 178 ± 6
AB 14B 4 ± 3 44 ± 20 80 ± 13 70 ± 5 AB 15 27.3 27.5 5 ± 1 35 ± 15 nd nd
AB 15B 3 ± 1 13 ± 4 nd nd AB 16 27.1 27.1 8 ± 1 36 ± 15 nd nd AB 17 28.3 28.4 6 ± 2 40 ± 17 nd nd
*B=near bottom measurement nd means the values were not determined
96
Table 6.2 – continued. August 30 - 31, 2004 Sample
Salinity surface
Salinity bottom
223Ra
ex. 224Ra
226Ra
228Ra ppt ppt dpm/fiber dpm/fiber dpm/fiber dpm/fiber
AB 1 0.4 1 2 ± 1 37 ± 18 89 ± 6 70 ± 10 AB 1B 3 ± 2 44 ± 15 50 ± 2 59 ± 1 AB 2 0.7 1.5 2 ± 1 62 ± 26 131 ± 6 127 ± 2
AB 2B 4 ± 1 36 ± 12 46 ± 5 57 ± 4 AB 3 2 2 ± 1 27 ± 8 39 ± 3 44 ± 3 AB 4 6.5 6.5 6 ± 2 62 ± 19 55 ± 9 108 ± 5 AB 5 12.2 21.3 7 ± 2 61 ± 12 79 ± 4 163 ± 1
AB 5B 3 ± 2 31 ± 10 26 ± 2 66 ± 6 AB 6 18.6 21.3 12 ± 2 86 ± 34 88 ± 10 215 ± 19
AB 6B 6 ± 1 48 ± 18 32 ± 4 88 ± 7 AB 7 12.6 24.1 9 ± 1 50 ± 15 70 ± 9 138 ± 3
AB 7B 2 ± 1 10 ± 4 11 ± 1 20 ± 1 AB 8 24.1 28.4 4 ± 1 33 ± 12 38 ± 6 63 ± 1
AB 8B 6 ± 1 33 ± 12 32 ± 3 47 ± 1 AB 9 27.7 27.7 9 ± 1 46 ± 17 41 ± 3 90 ± 2
AB 9B 3 ± 1 17 ± 7 13 ± 1 29 ± 1 AB 10 19.8 24.1 19 ± 2 96 ± 35 138 ± 7 293 ± 25
AB 10B 2 ± 1 13 ± 3 17 ± 3 26 ± 2 AB 11 17.5 24.5 23 ± 2 149 ± 49 195 ± 5 397 ± 20
AB 11B 6 ± 1 44 ± 17 34 ± 3 78 ± 1 AB 12 7.2 19.1 11 ± 4 118 ± 50 135 ± 2 230 ± 10
AB 12B 10 ± 3 82 ± 32 65 ± 4 143 ± 4 AB 13 nd nd nd nd nd nd AB 14 20.1 20.1 34 ± 19 501 ± 249 503 ± 4 885 ± 6
AB 14B 20 ± 8 202 ± 83 255 ± 23 463 ± 10 AB 15 25.9 26.9 25 ± 3 171 ± 71 239 ± 7 393 ± 15
AB 15B 9 ± 1 45 ± 20 68 ± 5 98 ± 10 AB 16 29.5 30.5 20 ± 1 88 ± 28 250 ± 12 359 ± 4
AB 16B 12 ± 3 58 ± 22 152 ± 13 206 ± 2 AB 17 nd nd nd nd nd nd
*B=near bottom measurement nd means the values were not determined
97
Table 6.2 – continued. January 25 - 26, 2005 Sample
Salinity surface
Salinity bottom
223Ra
ex. 224Ra
226Ra
228Ra ppt ppt dpm/fiber dpm/fiber dpm/fiber dpm/fiber
AB 1 0 0 0.6 ± 0.1 14 ± 1 18 ± 4 16 ± 1 AB 1B 0.9 ± 0.2 18 ± 1 17 ± 3 15 ± 1 AB 2 0 0 0.3 ± 0.1 11 ± 1 4 ± 1 4 ± 1
AB 2B 0.2 ± 0.1 9 ± 1 4 ± 1 4 ± 1 AB 3 0 0.6 ± 0.1 10 ± 1 13 ± 2 11 ± 1 AB 4 2.7 3 ± 1 42 ± 3 28 ± 1 27 ± 3 AB 5 14.1 22.6 3 ± 1 41 ± 3 35 ± 2 38 ± 5
AB 5B 4 ± 1 45 ± 3 40 ± 3 41 ± 4 AB 6 27 28.1 5 ± 1 56 ± 3 63 ± 2 61 ± 16
AB 6B 0.8 ± 0.2 11 ± 1 9 ± 1 8 ± 4 AB 7 12.6 24.1 5 ± 1 59 ± 4 62 ± 4 63 ± 7
AB 7B 0.9 ± 0.2 9 ± 1 6 ± 3 6 ± 1 AB 8 14.9 28.9 4 ± 1 45 ± 3 67 ± 5 67 ± 1
AB 8B 0.9 ± 0.2 8 ± 1 11 ± 1 6 ± 1 AB 9 13.1 14.4 5 ± 1 65 ± 4 60 ± 5 84 ± 10
AB 9B 0.8 ± 0.1 11 ± 1 5 ± 1 5 ± 1 AB 10 9.8 32.7 6.2 ± 1 73 ± 3 82 ± 6 83 ± 6
AB 10B 0.9 ± 0.1 6 ± 1 12 ± 1 11 ± 1 AB 11 9.1 21.7 4 ± 1 52 ± 2 46 ± 4 57 ± 3
AB 11B 1.3 ± 0.2 21 ± 1 16 ± 3 17 ± 1 AB 12 21.3 17.6 4 ± 1 51 ± 3 53 ± 2 51 ± 3
AB 12B 2 ± 1 34 ± 2 20 ± 4 16 ± 2 AB 13 3.3 20.7 2 ± 1 45 ± 3 32 ± 2 26 ± 1
AB 13B 1.7 ± 0.2 19 ± 1 17 ± 2 12 ± 1 AB 14 7.1 9.4 14 ± 1 171 ± 15 172 ± 2 188 ± 3
AB 14B 1.9 ± 0.2 19 ± 1 24 ± 3 17 ± 2 AB 15 15.3 18.7 4 ± 1 46 ± 4 65 ± 7 62 ± 2
AB 15B 0.6 ± 0.1 7 ± 1 10 ± 3 11 ± 1 AB 16 19.1 22.4 13 ± 1 113 ± 6 197 ± 11 210 ± 29
AB 16B 1.7 ± 0.2 9 ± 1 18 ± 1 20 ± 1 AB 17 nd nd nd nd nd nd
*B=near bottom measurement nd means the values were not determined.
98
Tabl
e 6.
3: R
adiu
m is
otop
es m
easu
red
in A
pala
chic
ola
Riv
er.
Unc
erta
intie
s rep
rese
nt ±
1 σ
bas
ed o
n co
untin
g st
atis
tics.
Dat
e
223 R
a 22
4 Ra
228 T
h ex
224 R
a 22
6 Ra
228 R
a 22
4 Ra/
223 R
a Sa
mpl
e sa
mpl
ed
dpm
m-3
dp
m m
-3
dpm
m-3
dp
m m
-3
dpm
m-3
dp
m m
-3
R
S-1
filte
red
10/9
/03
1.8
± 0.
8 47
± 3
9
± 1
37 ±
3
21 ±
10
RS-
1
3.2
± 0.
8 91
± 4
16
± 1
75
± 4
24
± 6
RS-
2 fil
tere
d 10
/9/0
3 2.
6 ±
0.8
54 ±
3
3 ±
1 52
± 3
20
± 6
R
S-2
3.
9 ±
0.7
80 ±
4
18 ±
1
62 ±
4
16 ±
3
R
A-1
8/
16/0
4 1.
7 ±
0.5
58 ±
4
7 ±
1 51
± 4
13
± 3
16
± 5
30
± 9
R
A-2
8/
17/0
4 1.
3 ±
0.5
44 ±
3
3 ±
1 42
± 3
5
± 2
37
± 5
31
± 1
2 R
A-3
8/
17/0
4 1.
2 ±
1 43
± 3
3
± 1
40 ±
3
19 ±
5
40 ±
7
34 ±
18
Ave
rage
unf
ilter
ed
2.
2 ±
1.2
54 ±
15
12 ±
7
31 ±
13
27 ±
7
Ave
rage
filte
red
2.
2 ±
0.6
45 ±
11
21 ±
6
dp
m/fi
ber
dpm
/fibe
r dp
m/fi
ber
dpm
/fibe
r dp
m/fi
ber
dpm
/fibe
r
RB
-1
8/24
/04
0.24
± 0
.07
7.2
± 0.
3 0.
9 ±
0.1
6.3
± 0.
4 10
± 2
10
± 2
27
± 8
R
B-2
8/
26/0
4 0.
10 ±
0.0
5 1.
9 ±
0.3
0.3
± 0.
1 1.
6 ±
0.2
16 ±
8
RB
-3
8/28
/04
0.14
± 0
.08
4.5
± 0.
4 0.
7 ±
0.2
3.8
± 0.
4 1
± 1
1.
8 ±
1.8
27
± 1
4 R
B-4
8/
28/0
4 0.
18 ±
0.0
5 3.
4 ±
0.3
0.4
± 0.
1 3.
0 ±
0.3
17 ±
6
RB
-5
8/30
/04
0.13
± 0
.05
4.1
± 0.
3 0.
8 ±
0.1
3.3
± 0.
4 0.
7 ±
0.1
1.
0 ±
0.1
26
± 1
0 R
B-6
8/
30/0
4 0.
06 ±
0.0
3 2.
0 ±
0.2
0.5
± 0.
1 1.
5 ±
0.2
24 ±
11
Ave
rage
23 ±
5
fil
tere
d –
sam
ples
filte
red
by 0
.45
mic
ron
filte
r car
tridg
e, o
ther
wis
e on
ly a
raw
acr
ylic
fibe
r pre
filte
r was
use
d
99
Tabl
e 6.
4: R
adiu
m is
otop
es m
easu
red
in b
otto
m a
nd su
spen
ded
sedi
men
ts in
Apa
lach
icol
a R
iver
. U
ncer
tain
ties r
epre
sent
± 1
σ b
ased
on
cou
ntin
g st
atis
tics.
Sa
mpl
e D
ate
sam
pled
22
3 Ra
224 R
a 22
6 Ra
228 R
a 22
4 Ra/
223 R
a
dp
m g
-1
dpm
g-1
Sedi
men
t - sa
nd
10/9
/03
0.39
±0.0
3 0.
49±0
.17
Se
dim
ent -
mud
10
/9/0
3
1.
3±0.
1 2.
2±0.
1
dpm
/sam
ple
dpm
/sam
ple
dpm
/sam
ple
dpm
/sam
ple
Su
sp. p
artic
les
8/16
/04
0.09
± 0
.08
1.3
± 0.
2
14
± 1
2 Su
sp. p
artic
les
8/17
/04
0.41
± 0
.24
9 ±
3 5.
0 ±
0.3
5.1
± 0.
5 22
± 1
5 Su
sp. p
artic
les
8/17
/04
0.21
± 0
.13
5 ±
1 5.
1 ±
0.5
4.7
± 1.
3 26
± 1
6
dp
m/s
ampl
e dp
m/s
ampl
e
RS-
2 Pr
efilt
er
10/9
/03
0.8
± 0.
5 13
± 2
RA
-1 P
refil
ter
8/16
/04
0.44
± 0
.20
8.6
± 1
3 ±
2
5 ±
2 20
± 5
R
A-2
Pre
filte
r 8/
17/0
4 0.
42 ±
0.1
9 13
± 2
30
± 1
4 R
A-3
Pre
filte
r 8/
17/0
4 0.
48 ±
0.1
3 11
± 2
4
± 1
4 ±
1
23 ±
6
Tabl
e 6.
5: C
ompa
rison
of 22
4 Ra/
223 R
a ra
tios m
easu
red
in b
ay w
ater
and
cor
resp
ondi
ng p
refil
ters
. Unc
erta
intie
s rep
rese
nt ±
1 σ
bas
ed
on c
ount
ing
stat
istic
s.
D
ate
sam
pled
22
3 Ra
224 R
a 22
8 Th
ex22
4 Ra
226 R
a 22
8 Ra
224 R
a/22
3 Ra
Sam
ple
Salin
ity
dpm
m-3
dp
m m
-3
dpm
m-3
dp
m m
-3
dpm
m-3
dp
m m
-3
AB
3
3/4/
04
1.8
± 0.
8 47
± 3
9
± 1
36 ±
3
46 ±
9
76 ±
32
21 ±
10
AB
3 P
refil
ter
0 pp
t 3.
2 ±
0.8
91 ±
4
16 ±
1
75 ±
4
41 ±
20
69 ±
17
24 ±
6
A
B 1
2 3/
4/04
2.
6 ±
0.8
54 ±
3
3 ±
1 51
± 3
17
2 ±
6 16
2 ±
36
20 ±
6
AB
12
Pref
ilter
4.
1 pp
t 3.
9 ±
0.7
80 ±
4
18 ±
1
62 ±
4
26 ±
3
55 ±
46
16 ±
3
100
Discussion
Radium isotopes in Apalachicola River
From the river water radium analysis (measured volumes) we estimated that the
Apalachicola River transported 880 dpm s-1 dissolved 223Ra and 17,800 dpm s-1 dissolved 224Ra into Apalachicola Bay in September, 2003; 1,740 dpm s-1 223Ra and 34,800 dpm s-1 224Ra in March, 2004 and 613 dpm s-1 223Ra and 19,400 dpm s-1 224Ra in August, 2004.
These amounts do not account for radium released from the particles transported by the
river. Based on the comparison of grab samples collected at stations 3 (0 ppt) and 12 (4
ppt) in March 2004, desorption from particles adds at least 40 % more dissolved radium
to the water. The desorbed radium has the same AR as the river. While the radium
inputs change by season according to the river discharge, based on our three river
samplings the absolute radium concentration in the river seems to be in a close range
(Tables 6.3 and 6.4). For our age model it is the activity ratio of the radium isotopes that
needs to be constant over at least several days before and during the radium sampling in
the bay. The time series radium collection in the river on August 24-31, 2004 showed
that the radium isotopic ratio did not change significantly over 8 days (Table 6.3). While
the measurement uncertainties are rather high, the determined 224Ra / 223Ra AR (average
= 23 ± 5) compare very well with the river 224Ra / 223Ra ARs measured a few days later
as the average of stations 1 and 2 measured during our station deployments on August
30-31, 2004 (AR = 25 ± 6).
Moreover, all samples collected in the river end-member during our two year
study had very comparable 224Ra / 223Ra activity ratios with an overall average of 24 ± 5
(n = 25). Figure 6.5 is a plot of the short-lived radium ratios of all samples collected in
the river, where the RS series are filtered and unfiltered samples from October 9, 2003;
the RA series are water and suspended particles collected on August 17, 2004; and RB
are bags deployed from a dock during the time series measurement in August 24-31,
2004.
For comparison, Figure 6.5 also contains the 224Ra / 223Ra activity ratios measured
on prefilters and in sediment samples. The sediment activity ratios tend to be lower and
have higher uncertainties which are due to low radium activity and low 223Ra count rates
on the delayed coincidence counting system. Based on these results we can assume that
101
the dissolved and particulate radium of the river represents a constant 224Ra / 223Ra
activity ratio source to the Apalachicola Bay over a time scale of several days.
RS-1fRS-1
RS-2fRS-2
RA-1RA-2
RA-3RB-1
RB-2RB-3
RB-4RB-5
RB-6sand
mud
224 R
a/22
3 Ra
0
10
20
30
40
50
60
waterprefiltersediment
Figure 6.5. Ratios of 224Ra/223Ra measured in selected river water and sediment samples.
Uncertainties represent ± 1 σ based on counting statistics.
Radium isotopic ratios in the bay
The sampling design of the radium collection on Mn-fibers deployed in a mesh
bag appeared to work well. The fibers collected high enough radium activities and the
samples were easy to measure on both the delayed coincidence counter and by gamma-
spectrometry. The fibers deployed near the surface collected significantly more (one
order of magnitude higher) radium activity than those deployed near the bottom. This is
likely due to stronger currents at the surface compared to near the bottom. Thus more
water circulated through the surface fibers, resulting in larger amounts of radium being
collected on these fibers.
In order to assess radium inputs from the bay sediments into the water column we
compared the radium isotopic ratios in the bottom and surface water. Figure 6.6 shows
102
the surface and bottom 224Ra / 223Ra activity ratios plotted by stations. If there is new
radium input from the sediments, we would expect the ratios to be higher in the bottom
waters because 224Ra is regenerated in the sediments faster than 223Ra due to its shorter
half-life. Figure 6.6 shows that the surface ratios are higher (more influenced from the
river) or the same as the bottom ratios in almost all cases. Even if the radium activity
ratios were higher in some bottom waters (indicating more input of 224Ra from
sediments), the water in the bay is stratified with limited mixing. We thus assume that
radium input from the bottom sediments to the surface waters is negligible.
August, 2003
1 2 3 4 5 6 7 8 9 1011121314151617
224 R
a/22
3 Ra
0
10
20
30
40
50
SurfaceBottom
March, 2004
1 2 3 4 5 6 7 8 9 1011121314151617
August, 2004
Station #
1 2 3 4 5 6 7 8 9 1011121314151617
224 R
a/22
3 Ra
0
10
20
30
40
50
January, 2005
Station #
1 2 3 4 5 6 7 8 9 1011121314151617
Figure 6.6. Ratios of 224Ra/223Ra measured in surface and near bottom waters in
Apalachicola Bay and River during the four sampling periods. Uncertainties represent ± 1 σ based on counting statistics.
103
August, 2003
1 2 3 4 5 6 7 8 9 1011121314151617
224 R
a/22
8 Ra
0.0
0.5
1.0
1.5
2.0
SurfaceBottom
March, 2004
1 2 3 4 5 6 7 8 9 1011121314151617
August, 2004
Station #
1 2 3 4 5 6 7 8 9 1011121314151617
224 R
a/22
8 Ra
0.0
0.5
1.0
1.5
2.0
January, 2005
Station #
1 2 3 4 5 6 7 8 9 1011121314151617
Figure 6.7. Ratios of 224Ra/228Ra measured in surface and near bottom waters in
Apalachicola Bay and River during the four sampling periods. Uncertainties represent ± 1 σ based on counting statistics.
It is interesting that the plots of AR of 224Ra to a long-lived 228Ra (5.7 years) in
the surface and bottom waters (Fig. 6.7) reveal a different pattern than 224Ra / 223Ra. At
most stations there are higher 224Ra / 228Ra ratios in the bottom than in surface waters.
The bottom waters are generally salty or brackish all over the bay so the radium isotopes
are constantly desorbed from the surface of the sediments. While these additions are
small, their higher AR ratios (up to 1.5) are a result of faster regeneration of 224Ra than 228Ra from the sediments.
104
The results from four repeated measurements of the suspended particles from the
river on the coincidence counting system over one month show that most of the 224Ra and 223Ra present on the particles is supported by their thorium parents. The particles
transported by the river loose their desorbable radium when they encounter saline water.
But radium is continually regenerated on the suspended particles by radioactive decay of
thorium. This newly regenerated radium desorbs to the saline bay water. Because after
desorption the 224Ra is regenerated on the particles faster than 223Ra, the additional newly
desorbed 224Ra / 223Ra activity ratio would be higher than it would be at equilibrium. The
influence of the regenerated radium from the suspended particles on the total radium
composition in the water will depend on the amount of suspended particles in the water
and how large a fraction of the regenerated radium is desorbable from the particles. We
neglect these radium inputs because we believe that they are insignificant compared to
the amount of radium transported to the bay by Apalachicola River.
Apparent radium ages
The apparent radium age of the water at each station was calculated using
equation (6.2) based on the 224Ra / 223Ra activity ratios measured in the surface water
samples. Table 6.6 lists the initial [224Ra/223Ra]i AR applied for each sampling, the
observed [224Ra/223Ra]obs AR of the surface water at each station, and the calculated
apparent radium ages. The age results are also presented as contour plots in Figure 6.8.
The ages represent a situation averaged over four days in August, 2004 and one day
during the other periods. The average error of the apparent radium ages calculated based
on the 224Ra / 223Ra activity ratio uncertainties is estimated to be ±1-1.5 days. The
contours plotted in East Bay and the southern part of St. George Sound are only rough
approximations, we had a limited number of stations in these areas. Another possible
source of uncertainty in the estimated radium ages is the excess 224Ra and 223Ra present in
the coastal waters in the Gulf of Mexico outside of Apalachicola Bay. Measurements in
coastal waters in nearby areas outside of Apalachicola Bay indicate that the coastal 224Ra
/ 223Ra AR can be as high as 4-5 (Burnett and Dulaiova, 2003, Moore, 2003). When
GOM water enters the bay it brings in excess 224Ra and 223Ra influencing the bay water 224Ra / 223Ra AR. This occurs mainly in the inlet areas. The excess radium affects
105
especially ages calculated in St. George Island Sound which is the major inlet of the
GOM water to the bay. Due to the GOM excess 224Ra / 223Ra AR the apparent radium
ages in St. George Sound may appear 3-4 days younger than the real water age.
Younger radium ages imply that the water leaves the bay faster. According to the
observed river discharge we would expect to see the youngest bay-water ages in August,
2003 when the discharge was higher (1016 m3 s-1), and the oldest ages during the August,
2004 sampling (338 m3 s-1). The contour plots (Fig. 6.8) confirm this with much younger
ages in August 2003, and they also reveal different circulation pattern. In August, 2003
the river plume is more symmetrical than in August, 2004. The apparent radium ages at
West Pass and St. Vincent Sound, the two major water outlets, are >10 days in August,
2004 and 7 - 8 days in August, 2003.
Besides river discharge, wind patterns and tides will influence water movement in
the bay. If we compare August, 2003 and March, 2004, two seasons with similar river
discharge, we still see distinct differences in the age contours (Fig. 6.8). In March, right
before our deployment, there was a steady wind blowing over one day from the east at 2-
6 m s-1 (Fig. 6.9). The age contours during this period reflect the influence of this wind.
The water is forced from St. George Sound towards the western part of the bay where the
water exits through West Pass and Indian Pass. The apparent radium ages in the eastern
part of the bay are over 12 days, while in St. Vincent Sound the water is only 2 days old.
During the deployments the tidal ranges measured in the bay were generally 0.6
m. However, on January 24, 2005 one day before our sampling, the low tide was -0.304
meters below the mean lower low water (MLLW) and the high tide only +0.13 m above
the MLLW, which is 0.3 m below the predicted tide levels. We surmise that such low
water levels in the bay were caused by the combined effect of the intense west winds
prevailing at that time (Fig. 6.9) and the tides. While collecting fibers from our stations
the next day it was low tide and we noticed a very strong eastward current in the bay.
The apparent tracer ages also showed that the river plume moved eastward. The oldest
age within the area of our investigation was 7 days in the eastern and 6 days in the
western part of the bay.
These circulation patterns are not unusual in Apalachicola Bay. Huang et al.
(2002a) examined the effects of surface wind on the salinity distribution in the bay. They
106
found that steady winds can induce a very large volume flux. Mortazavi et al. (2000,
2001) also observed a change in the bay-water circulation due to the strong influence of
winds and tides when instead of West Pass, St. George Sound may became the main
water outlet from the bay.
The calculated apparent radium ages represent the rate at which river water moves
through the bay. This flux refers to the flushing rate with respect to the freshwater
(radium input) only. We estimated the flushing rates as the age of the oldest river water
present in the bay, excluding the area of St. George Sound. These estimated flushing
rates were 8, 10, 12 and 6 days for August, 2003, March and August, 2004 and January,
2005, respectively. These rates fall into the same range as the turnover times estimated
by the circulation model reported in Mortazavi et al. (2000). In their model the authors
estimated the residence time as a combination of fresh and salt water inputs and outputs
into the bay. We plotted their turnover time data against river discharge at that time and
derived a dependency curve. We neglected wind and tide effects so the comparison is
only assessing one of the important variables. Based on this relationship and our river
discharge values we estimated that their model would predict residence times of about 7,
7, 12 and 8 days for the same four periods that we investigated.
We also calculated a mean radium age of the water in the bay. From the contour
plots on Figure 6.8 we estimated the area represented by each age section. For example
we assumed that the area of the bay that is between contours 2 and 3 days represents the
fraction of the water that is 2.5 days old. We then calculated the weighted average of the
ages according to the extent of the total area each age represents. These estimated mean
ages were 5.5 days in August 2003, 7.6 days in March 2004, 9 days in August 2004, and
5.1 days in January 2005. This approach would be more accurate if we applied volume
instead of area, but because of the shallow nature of this bay the differences in depths are
probably small.
Conclusions
We applied a radium tracer technique to estimate apparent radium ages with
respect to fresh water in Apalachicola Bay. Our results confirm the findings of previous
107
studies that the water circulation in the bay highly depends on river discharge, but is also
influenced by prevailing winds and tidal patterns. However, the goal of our investigation
was not to discern the relationship between flushing rate and these parameters, but to
apply a simple technique that does not require discharge, wind, tide, and water current
measurements. Our approach would be useful, therefore, even in remote locations where
this type of information is not available.
Our results indicate that the Apalachicola Bay flushing times range from 6 to 12
days. We found that even above-average river discharge does not necessary result in
faster flushing rates. These radium ages can be used to quantify transport processes of
dissolved substances in the bay. Should soluble pollutants enter the bay via the river, the
radium tracer approach would show how such contaminants would be dispersed and how
fast they are leaving the bay. The radium tracer approach thus provides a tool for
environmental managers to evaluate pollution dispersion in the estuary.
We conclude that the radium age technique is very well suited for residence time
calculations in river dominated shallow estuaries where the river is the only or major
radium source to the system. In estuaries with additional radium sources like
groundwater discharge or sediment resuspension one has to be aware of the possible
interferences from radium inputs by these other sources. In such cases the apparent
radium age calculations would have to assume radium inputs from more than one end-
member.
108
Tabl
e 6.
6. 22
4 Ra/
223 R
a ac
tivity
ratio
s and
the
corr
espo
ndin
g ap
pare
nt ra
dium
age
s at t
he d
iffer
ent s
tatio
ns sa
mpl
ed d
urin
g th
e fo
ur fi
ber
depl
oym
ents
. U
ncer
tain
ties f
or th
e 22
4 Ra/
223 R
a ac
tivity
ratio
s rep
rese
nt ±
1 σ
bas
ed o
n co
untin
g st
atis
tics.
The
ave
rage
err
or o
f th
e ap
pare
nt ra
dium
age
s cal
cula
ted
base
d on
the
224 R
a/22
3 Ra
AR
unc
erta
intie
s is e
stim
ated
to b
e ±1
-1.5
day
s.
A
ug, 0
3 R
a M
ar, 0
4 R
a A
ug, 0
4 R
a Ja
n, 0
5 R
a
ex.22
4 Ra/
223 R
a ag
e ex
.224 R
a/22
3 Ra
age
ex.22
4 Ra/
223 R
aag
e ex
.224 R
a/22
3 Ra
age
days
days
days
da
ys
Initi
al
AR
28
± 6
22 ±
6
25
± 6
25 ±
9
AB
1
27 ±
6
0 18
± 7
0
21 ±
16
0 24
± 5
0
AB
2
23 ±
6
0 18
± 7
0
29 ±
22
0 34
± 1
0 0
AB
3
34 ±
8
0 29
± 2
0 0
12 ±
7
5.4
17 ±
3
3.1
AB
4
8 ±
1 10
.2
15 ±
5
2.7
10 ±
4
7.3
14 ±
2
4.7
AB
5
17 ±
3
3.9
22 ±
15
-0.1
9
± 3
8.2
12 ±
2
5.6
AB
6
9 ±
2 8.
9 17
± 1
0 1.
9 7
± 3
9.6
10 ±
1
6.9
AB
7
11 ±
2
7.3
13 ±
8
3.8
6 ±
2 11
.5
13 ±
1
5.3
AB
8
11 ±
3
7.1
7 ±
4 8.
4 8
± 3
9.2
10 ±
1
6.9
AB
9
11 ±
2
7.7
5 ±
3 10
.8
5 ±
2 12
.0
13 ±
2
4.9
AB
10
11 ±
2
7.4
7 ±
4 8.
8 5
± 2
12.5
12
± 1
5.
8 A
B 1
1 15
± 3
5.
1 7
± 4
9.1
7 ±
2 10
.4
13 ±
1
5.3
AB
12
20 ±
4
2.8
15 ±
8
2.8
11 ±
6
6.6
13 ±
2
4.8
AB
13
14 ±
3
5.6
8 ±
5 7.
9 nd
21 ±
3
1.3
AB
14
nd
5
± 3
11.6
15
± 1
1 4.
1 13
± 1
5.
3 A
B 1
5 13
± 3
5.
8 5
± 3
11.8
7
± 3
10.2
11
± 2
6.
4 A
B 1
6 10
± 3
8.
4 3
± 2
14.9
4
± 1
13.5
9
± 1
8.1
AB
17
8 ±
1 10
.2
5 ±
3 11
.9
nd
nd
nd
= n
ot d
eter
min
ed
109
Figu
re 6
.8. C
alcu
late
d w
ater
age
s est
imat
ed fo
r sur
face
wat
ers d
urin
g ou
r fou
r sam
plin
g pe
riods
in A
ugus
t 200
3, M
arch
and
Aug
ust
2004
, and
Janu
ary
2005
. Th
e in
dica
ted
river
dis
char
ge is
an
aver
age
befo
re a
nd d
urin
g th
e sa
mpl
ing.
1016
m3 s-1
99
6 m
3 s-1
338
m3 s-1
80
9 m
3 s-1
110
Fi
gure
6.9
. Win
d sp
eed
and
dire
ctio
ns m
easu
red
at N
OA
A C
O-O
PS st
atio
n #8
7286
90 lo
cate
d in
Apa
lach
icol
a, F
lorid
a. T
he li
nes
indi
cate
the
time
of o
ur sa
mpl
ing.
111
CONCLUSIONS
The overall purpose of the project was to combine the use of radon and radium
isotopes as tracers of coastal hydrological processes. The approach was successfully
applied to assess submarine groundwater discharge (SGD) in various coastal settings
(discharge from sandy and glacial till aquifers). Radon is an excellent tracer of SGD
because of its large enrichment in groundwaters compared to surface waters. The source
of radium isotopes to the coastal zone is discharge of salty groundwater providing
somewhat different information than radon, which is enriched in any groundwater. The
advantage of the combined use of these tracers is that we were able to: (1) derive mixing
rates from the short-lived radium isotopes to make an independent estimate of radon loss
due to mixing; (2) estimate the radon losses due to atmospheric evasion by comparing 222Rn and 224Ra offshore profiles; (3) determine the apparent radium ages of the water in
the coastal zone and estuaries as a measure of coastal residence time.
Radon and radium isotopes have been shown to be effective tracers of coastal
hydrological processes and based on our work we can conclude that they are even more
powerful when they are applied together in the same system.
112
APPENDIX A
TABLES
113
Table A.1. Radium isotopes and salinities measured in the West Neck bay study site and the transect leading from the study site to Gardiners Bay.
Sample Salinity 223Ra 224Ra 226Ra 228Ra
ppt dpm m-3 dpm m-3 dpm m-3 dpm m-3 R 4 - 20.8 29.3 13 70 135 521 R 5 - 16.8 29.2 24 182 144 593 R 6 - 12.4 29.1 13 159 134 579 R 7 - 3.1 29.1 21 184 155 668
R 13 - 3.1 29.1 23 121 146 595 R 14 - 2.2 28.9 36 248 149 738 R 15 - 0.6 28.1 53 394 158 745 R 16 - 0.1 28.2 61 405 168 911
dock 55 441 184 861 dock 58 295 153 696 dock 48 388 159 823 dock 38 221 141 644 dock 27.5 54 280 201 907 dock 27.5 52 357 191 899 dock 28.2 36 308 139 681 dock 4 53 124 34 dock 39 319 150 742 dock 28 47 363 158 767 dock 27.51 37 295 124 629 dock 26.15 42 431 146 802
SM - M 22 183 109 415 SM - M 11 377 106 460 SM - M 15 233 56 275 SM - M 15 246 58 267 SM - Kb 28.15 65 596 201 928 SM - Kc 27.9 89 348 199 1024 SM - Kc 26.15 143 1569 163 1015 SM - Ka 25.32 237 2805 236 1696 SM - Ka 208 2615 211 1506 PM - 40 0.1 8 70 155 286 PM - 10 0 5 58 77 275 Well 1 0 7 44 106 185 Well 2 0 5 78 48 117
*R - 20.8 = Radium samples collected on a transect with the distance of a sampling point indicated in km dock = samples collected along the dock on the study site SM = samples collected from seepage meters: M - manual, Ka - Krupa large, Kb - Krupa small, Kc - Krupa baby PM = piezometers 10' and 40' Well = samples collected from wells S-1 and S-2
114
Table A.2. Radon-222 measured in the West Neck bay study site and the transect leading from the study site to Gardiners Bay. The two different sample types were counted by the radon-emanation technique (Sample I) and using the RAD-7 H2O (Sample II).
Sample I 226Ra 222Rn Sample II 222Rn Methane dpm m-3 dpm m-3 dpm m-3 nM
T 1 - 19.2 218 183 SM - Kb 7,000 T 2 - 10.8 122 926 SM - Kb 22,000 T 3 - 8.1 217 458 SM - Kb 30,000 T 4 - 5.6 24 551 BW 187,000 33 T 5 - 2.9 87 843 BW 142,000 8 T 6 - 2.4 100 1,188 PM - 10 226,000 17 T 7 - 0.9 163 1,147 PM - 10 202,000 19 T 8 - 0.6 107 1,927 PM - 10 164,000 14 T 9 - 0.3 209 1,882 PM - 10 175,000 16
T 10 - 0.1 98 902 PM - 10 204,000 T 10b - 0.1 98 1,931 PM - 40 91,000 32
dock 167 1,080 PM - 40 35 dock 67 992 Well 1 339,686 30 dock 202 3,385 Well 1 317,000 6 dock 106 2,429 Well 1 409,000 10 dock 274 12,096 Well 1 367,000 4 dock 109 5,978 Well 1 336,000 19 dock 146 4,975 Well 1 361,000 4 dock 209 4,954 Well 2 194,952 6 dock 334 1,471 Well 2 160,000 dock 93 4,590 Well 2 155,000 dock 160 5,180 Tap water 198,000 dock 200 1,300 Tap water 204,000 dock 230 3,480
SM - Kc 370 8,657 SM - Kb 212 2,788 SM - Kb 250 2,876 Well 1 260 202,320 Well 1 330 367,220
T - 19.2 = radon samples collected on a transect with the distance of a sampling point indicated in km dock = samples collected along the dock on the West Neck Bay study site SM = samples collected from seepage meters: Kb - Krupa small, Kc - Krupa baby PM = piezometers 10' and 40' Well 1, 2 = samples collected from wells S-1 and S-2 BW = fresh water flowing on the beach at low tide Tap water = collected in the hotel
115
APPENDIX B
ASSESSMENT OF SUBMARINE GROUNDWATER DISCHARGE ON ST.
GEORGE ISLAND
We deployed an automated seepage meter off the St. George Island State Park in
January and February, 2001. The seepage meter consists of a benthic chamber made
from the top of a 55-gallon oil drum with a vent on the top. Fluid advection through the
sediments displaces the water in the chamber that then exits through the vent. This flow
is determined using an automatic device that measures flow via heat transport (Taniguchi
and Fukuo, 1993). We also monitored rain and water level in the bay at the same time.
The meter was deployed for 6 days each time (Fig. B.1A, B).
At the same site in St. George Island State Park, there are several wells on the
land with depths reaching 1 to 6 meters. We sampled these wells to determine the radon
and radium isotope concentrations in groundwater. After purging the wells we collected
about 40 liters of water that we passed through Mn-fibers and measured for the short-
lived radium isotopes. We collected about 10 ml of water into vials containing 10 ml of
mineral oil and measured the 222Rn by liquid scintillation counting.
At the same study site we measured the radium isotope ratios on a transect
perpendicular to the coast leading 3 kilometers to the bay. We collected about 40 liters of
seawater at 6 stations and passed through Mn-fiber columns.
We deployed the automated seepage meter at the same site again in August, 2001
after tropical storm Barry passed over the area (Fig. B.2). This time we also deployed a
continuous radon monitor (Burnett et al., 2001b) on a float set up about 50 meters from
the shore. Both instruments were operating continuously for 5 days. The radon monitor
made repeated integrated 2-hour measurements of radon concentration in the seawater at
this station to assess temporal variations of 222Rn that may be related to groundwater
116
seepage (Fig. B.3A, B). The advection rates were estimated from radon inventories
following the methodology described in Burnett and Dulaiova (2003).
The radium isotope ratios measured in freshwater samples collected from shallow
wells on St. George Island State Park were very low. The 224Ra / 223Ra AR in the wells is
5 ± 2 (n=14). The 223Ra in the wells ranged from 43 to 470 dpm m-3 (n=7) and 224Ra
were 160 to 2550 dpm m-3 (n=7). The average radon concentration in the same wells is
55,000±30,000 dpm m-3 (n=12). The highest tracer signals were measured in a 6-meter
deep well with a salinity of 6 ppt located 20 meters from the shore. In this well the
measured 224Ra / 223Ra AR is 30 and the 222Rn activity equaled 530,000 dpm m-3.
The radon concentrations measured in bay waters at this site were relatively low,
500 to 2,000 dpm m-3. The radium isotope ratios measured on the transect perpendicular
to the coast of the St. George Island State Park did not show any enrichment. There were
no higher radium ARs close to the shore.
The seepage meter results from January and February, 2001 indicate very low
submarine groundwater discharge. The baseflow ranges from 0.005 to 0.04 m day-1 and
only increased to 0.1 m day-1 during a short interval as a result of rainfall on the island
(Fig. B.1A, B). The seepage rates measured right after tropical storm Barry passed
through the area on August 6, 2001, were higher but persisted only a short time (Fig. B.2
and B.3A, B). Both the automated seepage meter and calculations based on the radon
inventories indicated groundwater flows of up to 0.25 m day-1.
From our radon survey map it seems that SGD is uniformly low along the
coastline, therefore we can scale up our measurement results obtained at the study site as
representative fluxes of the whole 53 km long coastline. If we assume that the seepage
occurs at a 100 m wide seepage face (Schneider and Kruse, 2005) and that the
groundwater discharge rate is similar along the 53-km long coastline, we can estimate the
total SGD that occurs from the barrier islands. If we multiply the SGD advection rate by
the length of the barrier islands and the width of the seepage face, the resulting submarine
groundwater discharge is 1.84 m3 s-1. This flow represents less than 1 % of the river
flow. The groundwater flux spikes to ~15 m3 s-1 after intense rain events (still only max 3
% of river flow), but according to our measurements these spikes last only several hours.
117
The magnitude of 224Ra and 223Ra input by submarine groundwater discharge was
calculated using the radium activities measured in a shallow well located 3 meters from
the shore. The water salinity in this well changes according to the tides and groundwater
level. The 223Ra in the well ranged from 43 to 470 dpm m-3 (n=7, median 86 dpm m-3)
and 224Ra activities were 160 to 2550 dpm m-3 (n=7, median 470 dpm m-3). Lower
radium activities were detected after rain events when the groundwater levels were higher
because the residence time of the groundwater is shorter, the rain dilutes the groundwater
radium concentrations and the fresh water changes the distribution coefficient of radium.
The 223Ra flux by SGD under base-flow conditions is thus 158 dpm s-1 and the 224Ra flux
is 870 dpm s-1. At SGD flows of 0.25 m d-1 the radium inputs may increase to 600 dpm s-
1 and 2000 dpm s-1 of 223Ra and 224Ra, respectively. We compared these SGD inputs to
our estimated the river fluxes. We multiplied the measured river 223Ra and 224Ra
concentrations (Tables 6.2 and 6.4) with the corresponding river discharge values and
received 223Ra fluxes of 1,500 - 2,500 dpm s-1 and 224Ra fluxes up to 30,000 to 50,000
dpm s-1. These fluxes are calculated using only the dissolved radium activity measured at
0 salinity river water. We don’t have an exact estimate of how much of the radium is
desorbable from the particles at increased salinities. Our best estimate is based on the
activities measured from grab samples at stations 3 (0 ppt) and 12 (4 ppt) in March 2004
according to which we saw a 40 % increase in the radium activities between the two
stations. The radium concentration at station 12 is already diluted by mixing so we can
state that there is at least 40 % increase in the radium activity due its desorption from
particles transported by the river. The radium inputs by SGD from St. George Island
therefore represent less than 1-2 % and only in sporadic cases up to 10 % of the river
radium fluxes. These higher fluxes however, last only for few hours.
In general, radium inputs by groundwater are higher if the SGD is brackish or
salty. It is because the recharging saltwater causes desorption of radium from the solids
in the aquifer and transport it to the bay upon discharge. Since according to our
measurements the magnitude of SGD on St. George Island is so sensitive to rainstorms,
the spikes in groundwater advection rates are probably due to fresh groundwater
discharge. Rain events would dilute the radium concentration in the groundwater and the
radium flux after rain is perhaps much lower than the above estimated values.
118
.
Date
2/10/01 2/11/01 2/12/01 2/13/01 2/14/01 2/15/01 2/16/01
Adv
ectio
n R
ate
(cm
/day
)
0
2
4
6
8
10
12
A
.
Date
1/26/01 1/27/01 1/28/01 1/29/01 1/30/01 1/31/01 2/1/01 2/2/01 2/3/01
Adv
ectio
n ra
te (c
m/d
ay)
0
2
4
6
8
10
12
Rain (cm
)
0.0
0.2
0.4
0.6
0.8
1.0
Advection rateRain
Figure B.1. Submarine groundwater discharge rates measured at the St. George Island
study site. (A) Dry period; (B) Groundwater advection measured at the same site during several rain events (indicated by bars).
B
119
Date/2001
Jul-23 Jul-30 Aug-06 Aug-13 Aug-20 Aug-27 Sep-03
Rai
n (m
m)
0.0
0.5
1.0
1.5
2.0
2.5
3.0R
elative well w
ater level (m)
0.2
0.4
0.6
0.8
1.0
1.2
SGD measurement
Tropical Storm Barry
Figure B.2. Rain and well water level measured at the St. George Island study site during
the summer of 2001. Tropical storm Barry and our sampling period right after the storm passed are indicated on the figure. The SGD results are shown in Figure B.3.
120
Date
8/7/01 8/8/01 8/9/01 8/10/01 8/11/01 8/12/01 8/13/01
Adv
ectio
n R
ate
(cm
/day
)
0
5
10
15
20
25
30
Water level (m
)
0.0
0.2
0.4
0.6
0.8
1.0
8/7/01 8/8/01 8/9/01 8/10/01 8/11/01 8/12/01 8/13/01
Adv
ectio
n R
ate
(cm
/day
)
0
5
10
15
20
25
30
Water level (m
)
0.0
0.2
0.4
0.6
0.8
1.0Advection Rate Water level
Figure B.3. Submarine groundwater discharge measured on the St. George Island study
site in the summer of 2001 right after tropical storm Barry passed over the area (A) Groundwater advection rates assessed based on continuous radon measurements; (B) SGD measured by a continuous automated seepage meter. Both instruments were deployed at the same place at the same time.
B
A
121
REFERENCES
Baskaran, M., D. J. Murphy, P. H. Santschi, J. C. Orr, and D. R. Schink, 1992. A method
for rapid in situ extraction and laboratory determination of Th, Pb, and Ra isotopes from large volumes of seawater. Deep-Sea Research I. 40, 849-865.
Bokuniewicz, H., 2001. Toward a coastal ground-water typology. Journal of Sea
Research 46, 99-108. Buddemeier, R. W. (Ed.) 1996. Groundwater discharge in the coastal zone: Proceedings
of an International Symposium. LOICZ Reports and Studies No. 8, LOICZ, Texel, The Netherlands, p. 179.
Bugna, G. C., J. P. Chanton, J. E. Cable, W. C. Burnett, and P. H. Cable, 1996. The
importance of groundwater discharge to the methane budgets of nearshore and continental shelf waters of the northeastern Gulf of Mexico. Geochimica et Cosmochimica Acta 60 (23), 4735-4746.
Burnett, W. C., J. E. Cable, D. R. Corbett, 2000. Radon Tracing of Submarine
Groundwater Discharge in Coastal Environments. IAEA Advisory group meeting on “Nuclear and Isotopic Techniques in the Management of Coastal Zones and Submarine Groundwater Discharge”, Vienna, Austria.
Burnett, W. C., M. Taniguchi, and J. Oberdorfer, 2001a. Measurement and significance
of the direct discharge of groundwater into the coastal zone. Journal of Sea Research 46, 109-116.
Burnett, W. C., G. Kim, and D. Lane-Smith, 2001b. A continuous radon monitor for
assessment of radon in coastal ocean waters. Journal of Radioanalytical and Nuclear Chemistry 249, 167-172.
Burnett, W.C., J. P. Chanton, J. Christoff, E. Kontar, M. Lambert, W. S. Moore, D.
O’Rourke, C. Smith, L. Smith, and M. Taniguchi, 2002. Assessing methodologies for measuring groundwater discharge to the ocean. EOS 83, 117-123.
Burnett, W. C., H. Bokuniewicz, M. Huettel, W. S. Moore, and M. Taniguchi, 2003a.
Groundwater and pore water inputs to the coastal zone. Biogeochemistry 66, 3-33.
122
Burnett, W. C., J. E. Cable, and D. R. Corbett, 2003b. Radon tracing of submarine groundwater discharge in coastal environments. pp. 25-42. In: Land and Marine Hydrogeology. M. Taniguchi, K. Wang and T. Gamo (eds.), Elsevier. 200 p.
Burnett, W. C. and H. Dulaiova, 2003. Estimating the dynamics of groundwater input
into the coastal zone via continuous radon-222 measurements. J. of Environmental Radioactivity 69, 21-35.
Cable, J. E., W. C. Burnett, J. P. Chanton, and G. Weatherly, 1996a. Modeling
groundwater flow into the ocean based on 222Rn. Earth and Planetary Science Letters 144, 591-604.
Cable, J.E. G. C. Bugna, W. C. Burnett, and J. P. Chanton, 1996b. Application of 222Rn
and CH4 for assessment of groundwater discharge to the coastal ocean. Limnology & Oceanography 41, 1347-1353.
Chanton, J.P., W. C. Burnett, H. Dulaiova, D. R. Corbett, and M. Taniguchi, 2003.
Seepage rate variability in Florida Bay driven by Atlantic tidal height. Biogeochemistry 66, 187-202.
Charette, M. A., K. O. Buesseler, and J. E. Andrews, 2001. Utility of radium isotopes for
evaluating the input and transport of groundwater-derived nitrogen to a Cape Cod estuary. Limnology & Oceanography 46 (2), 465-470.
Church, T. M., 1996. An underground route for the water cycle. Nature 380, 579-580. Corbett, D. R., W. C. Burnett, P. H. Cable, and S. B. Clark, 1998. A multiple approach to
the determination of radon fluxes from sediments. Journal of Radioanalytical and Nuclear Chemistry 236, 247-252.
Corbett, D.R., J. P. Chanton, W. C. Burnett, K Dillon, C. Rutcowski and J. Fourqurean.
1999. Patterns of groundwater discharge into Florida Bay. Limnology & Oceanography 44, 1045-1055.
Corbett, D. R., K. Dillon, W. C. Burnett, and J. P. Chanton, 2000. Estimating the
groundwater contribution into Florida Bay via natural tracers, 222Rn and CH4. Limnology & Oceanography 45, 1546-1557.
Dawson, C. E. 1955. A contribution to the hydrography of Apalachicola Bay, Florida.
Publication of Texas Institute of Marine Science 4, 14-35. DiLorenzo, J. L., and R. V. Ram, 1991. Flushing-time estimates for West Neck Harbor:
a small tidal embayment of the Peconic Bays, New York. Report - Najarian Associates, L.P. Eatontown, New Jersey. 44p.
123
Drogue, C., 1996. Groundwater discharge and freshwater-saline water exchange in karstic coastal zones. In: Buddemeir, R. W. (Ed.) Groundwater discharge in the coastal zone: Proceedings of an International Symposium. LOICZ Reports and Studies No. 8, LOICZ, Texel, The Netherlands, p. 179.
Dulaiova, H. and W.C. Burnett, 2004. An efficient method for gamma spectrometric
determination of 226,228Ra via Mn fibers. Limnology & Oceanography: Methods 2, 256-261.
Dulaiova, H., R. Peterson, W. C. Burnett, and D. Lane-Smith, 2005. A multidetector continuous monitor for assessment of 222Rn in the coastal ocean. Journal of Radioanalytical and Nuclear Chemistry 263(25), 361-365.
Elsinger, R. J., and W. S. Moore, 1983. Gas exchange in the Pee Dee River based on
222Rn evasion. Geophysical Research Letters 10(6), 443-446. Hornberger, G. M., J. P. Raffensperger, P. L. Wiberg and K. N. Eshleman, 1998.
Elements of physical hydrology. The Hopkins University Press, USA, p. 7 Huang, W., W. K. Jones, and T. S. Wu, 2002a. Modelling wind effects on subtidal
salinity in Apalachicola Bay, Florida. Estuarine, Coastal and Shelf Science 55, 33-46.
Huang, W., H. Sun, S. Nnaji, W. K. Jones, 2002b. Tidal Hydrodynamics in a multiple-
inlet estuary: Apalachicola Bay, Florida. Journal of Coastal Research 18(4), 674-684.
Kelly, R. P. and S. B. Moran, 2002. Seasonal changes in groundwater input to a well-
mixed estuary estimated using radium isotopes and implications for coastal nutrient budgets. Limnology & Oceanography 47, 1796-1807.
Kim, G., W. C. Burnett, H. Dulaiova, P. W. Swarzenski, and W. S. Moore, 2001.
Measurement of 224Ra and 226Ra activities in natural waters using a radon-in-air monitor. Environmental Science Technology 35, 4680-4683.
Kim, G. and D.-W. Hwang, 2002. Tidal pumping of groundwater into the coastal ocean
revealed from submarine 222Rn and CH4 monitoring. Geophysical Research Letters 29 (14), 23-27.
Kjerfve, B., 1979. Measurement and analysis of water current, temperature, salinity and
density. pp. 186-216. In: Hydrography and sedimentation in Estuaries. K.R. Dyer (ed.), Cambridge University Press. 230 pp.
Kohout, F. A. 1996. Submarine springs: a neglected phenomenon of coastal hydrology.
Hydrology 26, 391-413.
124
Krest, J. M., W. S. Moore, and Rama, 1999. 226Ra and 228Ra in the mixing zones of the Mississippi and Atchafalaya Rivers: indicators of groundwater input. Marine Chemistry 64, 129-152.
Krest, J. M., and J. W. Harvey, 2003. Using natural distributions of short-lived radium
isotopes to quantify groundwater discharge and recharge. Limnology & Oceanography 48 (1), 290-298.
Lambert, M. J., and W.C. Burnett, 2003. Submarine groundwater discharge at a Florida
coastal site based on continuous radon measurements. Biogeochemistry 66, 55-73. Lane-Smith, D. and S. Shefsky, 1999. Proceedings of the American Association of Radon
Scientists and Technologists, Las Vegas, Nevada, November 7-10, 1999. Lee, D. R., 1977. A device for measuring seepage flux in lakes and estuaries. Limnology
& Oceanography 22, 140-147. Li, Y.-H., G. Mathieu, P. Biscaye, H. J. Simpson, 1977. The flux of 226Ra from estuarine
and continental shelf sediments. Earth and Planetary Science Letters 37, 237-241.
Livingston, R.J. 1983. Resource atlas of the Apalachicola estuary. Florida Sea Grant
Publ. 64 pp. Macintyre, S., R. Wanninkhof, and J. P. Chanton, 1995. Trace gas exchange across the
air-sea interface in freshwater and coastal marine environments. In: Matson, P. A. and R. C. Harris (Eds.), Biogenic trace gases: measuring emissions from soil and water. Blackwell Science Ltd, pp. 52-97.
Mathieu, G., P. Biscayne, R. Lupton, and D. Hammond, 1988. System for measurements
of 222Rn at low levels in natural waters. Health Physics 55, 989-992. Moore, W. S., 1976. Sampling 228Ra in the deep ocean. Deep Sea Research and
Oceanographic Abstracts 23, 647-651. Moore, W. S., 1984. Radium isotope measurements using germanium detectors. Nuclear
Instruments and Methods in Physics Research 223, 407-411. Moore, W. S., and J. F. Todd, 1993. Radium isotopes in the Orinoco and Eastern
Caribbean Sea. Journal of Geophysical Research 98(C2), 2233-2244. Moore, W. S., H. Astwood, and C. Lindstrom, 1995. Radium isotopes in coastal waters
on the Amazon shelf. Geochimica et Cosmochimica Acta 59, 4285-4298. Moore, W. S., 1996. Large groundwater inputs to coastal waters revealed by 226Ra
enrichments. Nature 380, 612-614.
125
Moore, W. S., D. J. DeMaster, J. M. Smoak, B. A. McKee and P. W. Swarzenski, 1996. Radionuclide tracers of sediment-water interactions on the Amazon shelf. Continental Shelf Research 16(5-6), 645-665.
Moore, W. S., and R. Arnold, 1996. Measurement of 223Ra and 224Ra in coastal waters
using delayed coincidence counter. Journal of Geophysical Research 101 (C1), 1321-1329.
Moore, W. S., 1997. High fluxes of radium and barium from the mouth of the Ganges-
Brahmaputra River during low river discharge suggest large groundwater source. Earth and Planetary Science Letters 150, 141-150.
Moore, W. S., 1999. The subterranean estuary: a reaction zone of ground water and sea
water. Marine Chemistry 65, 111-125. Moore, W. S., 2000a. Determining coastal mixing rates using radium isotopes.
Continental Shelf Research 20, 1993-2007. Moore, W. S., 2000b. Ages of continental shelf waters determined from 223Ra and 224Ra.
Journal of Geophysical Research 105 (9), 22,117-22,122. Moore, W. S, 2003. Sources and fluxes of submarine groundwater discharge delineated
by radium isotopes. Biogeochemistry 66, 75-93. Moore, W. S., and J. Krest, 2004. Distribution of 223Ra and 224Ra in the plumes of the
Mississippi and Atchafalaya Rivers and the Gulf of Mexico. Marine Chemistry 86, 105-119.
Mortazavi, B., R.L. Iverson, W. Huang, F.G. Lewis, J.M. Caffrey, 2000. Nitrogen budget
of Apalachicola Bay, a bar-built estuary in the northeastern Gulf of Mexico. Mar. Ecol. Prog. Ser. 195, 1-14.
Mortazavi, B., R. L. Iverson, and W. Huang, 2001. Dissolved organic nitrogen and
nitrate in Apalachicola Bay, Florida: spatial distributions and monthly budgets. Marine Ecology Progress Series 214, 79-97.
Muller, J. W., 2000. Possible advantages of a robust evaluation of comparisons. J.
Research National Institute of Standards and Technologies 105, 551-555. Niu, X.-F., H.L. Edmiston, and G.O. Bailey, 1998. Time series models for salinity and
other environmental factors in the Apalachicola estuarine system. Estuarine Coastal and Shelf Science 46 (4), 549-563.
Nozaki, Y., Y. Yamamoto, T. Manaka, H.Amakawa, A. Snidvongs, 2001. Dissolved
barium and radium isotopes in the Chao Phraya River estuarine mixing zone in Thailand. Continental Shelf Research 21, 1435-1448.
126
Officer, C. B. 1979. Discussion of the behavior of non-conservative dissolved constituents in estuaries. Estuarine and Coastal Marine Science 9, 91-94.
O`Rourke, R. Paulsen, C. Smith, L. Smith, and M. Taniguchi, 2002. Assessing
methodologies for measuring groundwater discharge to the ocean. EOS, 83, 117-123.
Paulsen, J. R., C. F. Smith, D. O’Rourke, and T.-F. Wong, 2001. Development and
evaluation of an ultrasonic ground water seepage meter. Ground Water 39 (6), 904-911.
Plater, A. J., M. Ivanovich, R. E. Dudgale, 1995. 226Ra contents and 228Ra/226Ra activity
ratios of the Fenland rivers and The Wash, eastern England: spatial and seasonal trends. Chemical Geology (Isotope Geoscience Section) 119, 275-292.
Rosenau, J., G. Faulkner, C. Hendry and R. Hull, 1977. Springs of Florida. US
Geological Survey Bulletin 31, 438-453. Rutgers van der Loeff, M., S. Kuhne, M. Wahsner, H. Holtzen, M. Frank, B. Ekwurzel,
M. Mensch, V. Rachold, 2003. 228Ra and 226Ra in the Kara and Laptev seas. Continental Shelf Research, 23, 113-124.
Schneider, J.A. and S.E. Kruse, Assessing natural and anthropogenic impacts on
freshwater lens morphology on small barrier islands: Dog Island and St. George Island, Florida, in press, Hydrogeology Journal.
Schubert, C. E., 1998. Areas contributing ground water to the Peconic estuary, and
ground-water budgets for the North and South forks and Shelter Island, Eastern Suffolk County, New York. U.S. Geological Survey, Water resources investigations report 97-4136.
Shinn, E. A., C. D. Reich, and T. D. Hickey, 2002. Seepage Meters and Bernoulli’s
Revenge. Estuaries 25 (1), 126-132. Sholkovitz, E., C. Herbold, and M. Charette, 2003. An automated dye-dilution based
seepage meter for time-series measurement of submarine groundwater discharge. Limnology & Oceanography: Methods 1, 16-28.
Soren, J., 1978. Hydrogeologic Conditions in the Town of Shelter Island, Suffolk
County, Long Island, New York. U. S. Geological Survey Water-Resources Investigations Report 77, 22.
Stansfield, K. and C. Garrett, 1977. Implications of the salt and heat budgets of the Gulf
of Thailand. Journal of Marine Research 55, 935-963.
127
Stringer, C. E., and W. C. Burnett, 2004. Sample bottle design improvements for radon emanation analysis of natural waters. Health Physics 87, 642-646.
Stringfield, T., 1966. Artesian water in the Tertiary limestone in the southeastern States.
US Geological Survey Prof. Paper 517, 226. Sun, Y. and T. Torgersen, 1998. The effects of water content and Mn-fiber surface
conditions on 224Ra measurements by 220Rn emanation. Marine Chemistry 62, 299-306.
Swarzenski, P. W., C. D. Reich, R. M. Spechler, J. L. Kindinger and W. S. Moore, 2001.
Using multiple geochemical tracers to characterize the hydrogeology of the submarine spring off Crescent Beach, Florida. Chemical Geology 179 (1-4), 187-202.
Taniguchi, M., and Y. Fukuo, 1993. Continuous measurements of ground-water seepage
using an automated seepage meter. Ground Water 31, 675-679. Taniguchi, M., W. C. Burnett, J. E. Cable, and J. V. Turner, 2002. Investigation of
submarine groundwater discharge. Hydrological Processes 16, 2115-2129. Torgersen, T., K. K. Turekian, V. C. Turekian, N. Tanaka, E. C. DeAngelo, and J.
O’Donnell, 1996. 224Ra distribution in surface and deep water of Long Island Sound: sources and horizontal transport rates. Continental Shelf Research 16 (12), 1545-1559.
Turekian, K. K., N. Tanaka, V. C. Turekian, T. Torgersen, and E. C. Deangelo, 1996.
Transfer rates of dissolved tracers through estuaries based on 228Ra: a study of Long Island Sound. Continental Shelf Research 16, 863-873.
Zekster, I. S. 1996. Groundwater Discharge in the Coastal Zone, R.W. Buddemeier (ed.)
122. Visser, W. A. (ed.), 1980. Geologic Nomenclature. Gorinchen: Royal Geological and
Mining Society of Netherlands, 540. Wattayakorn, G., W. C. Burnett, M. Taniguchi, P. Sojisuporn, S. Rungsupa, 2004.
Contribution of carbon and nutrient species into SE Asian waters via submarine groundwater discharge. The Southeast Asia Regional Committee for START (SARCS), Carbon pilot research project, Final report (http://www.sarcs.org/new/initiated/Carbon/Carbon%20Pilot%20Project%20Final%20Report/Thailand%20Final%20Report.pdf), October, 2004, pp. 35.
Weigel, F. 1978. Chemiker Zeitung 102, 287.
128
BIOGRAPHICAL SKETCH
Henrieta DULAIOVA
P E R S O N A L H I S T O R Y BORN: 29 OCTOBER, 1973 ZELIEZOVCE, CZECHOSLOVAKIA E D U C A T I O N FLORIDA STATE UNIVERSITY, DEP. OF OCEANOGRAPHY TALLAHASSEE, FLORIDA, USAPh.D. Candidate in Chemical Oceanography 2000 - 2005
Dissertation Thesis: Multiple Isotopic Tracers for Study of Coastal HydrologicProcesses
CZECH TECHNICAL UNIVERSITY PRAGUE, CZECH REPUBLICFACULTY OF NUCLEAR SCIENCES AND PHYSICAL ENGINEERING 1992 - 1997Master of Sciences in Nuclear Chemistry
Thesis: Determination of Uranium by Gamma-Spectrometry MASARYK INSTITUTE OF ADVANCED STUDIES PRAGUE, CZECH REPUBLICCertificate in Pedagogical Studies 1994 - 1996
Complementary Pedagogical Education E M P L O Y M E N T H I S T O R Y FLORIDA STATE UNIVERSITY, DEPARTMENT OF OCEANOGRAPHY TALLAHASSEE, FLORIDA, USA Graduate Research Assistant 2000 - 2005 NATIONAL RADIATION PROTECTION INSTITUTE PRAGUE, CZECH REPUBLICSpecialist - Radiochemistry Group 1997 - 2000
Research focus: Determination of Actinides in Biological and Environmental Samples A W A R D S
WHOI Postdoctoral Fellowship Award recipient NOAA NERR Graduate Research Fellowship Award recipient AGU Student Travel Grant and IUGG Travel Grant recipient FSU Department of Oceanography Outstanding Graduate Student for the 2002-2003
Academic Year
129
P U B L I C A T I O N S
Dulaiova, H., W. C. Burnett, G. Wattayakorn, and P. Sojisuporn. Radon and RadiumIsotopes in the Chao Phraya River and Estuary, Forthcoming
Dulaiova, H. and W. C. Burnett. Seasonal Factors Affecting Radon and Radium
Isotopes in Coastal Waters, Forthcoming
Dulaiova, H. and W. C. Burnett. Evaluation of Flushing Rates of Apalachicola Bay,Florida via Natural Geochemical Tracers, Forthcoming
Dulaiova, H., W. C. Burnett, J. P. Chanton, W. S. Moore. H. J. Bokuniewicz, M. A.
Charette, and E. Sholkovitz. Assessment of Submarine Groundwater Discharges intoWest Neck Bay, New York, via Natural Tracers, Continental Shelf Researchsubmitted.
Burnett, W.C. and H. Dulaiova, 2005. Radon as a Tracer of Submarine Groundwater
Discharge into a Boat Basin in Donnalucata, Sicily, Continental Shelf Researchsubmitted.
Burnett, W. C., H. Dulaiova, C. Stringer, and R. Peterson, 2005. Submarine
Groundwater Discharge: Its Measurement and Influence on the Coastal Zone, Journal of Coastal Research Special Issue 39, in press.
Dulaiova, H., R. Peterson, W. C. Burnett, and D.-L. Smith, 2005. A Multi-Detector
Continuous Monitor for Assessment of 222Rn in the Coastal Ocean., Journal of Radioanalytical and Nuclear Chemistry 263 (2), 361-365
Swarzenski, P., W. C. Burnett, C. Reich, H. Dulaiova, R. Peterson, and J. Meunier,
2004. Novel Geophysical and Geochemical Techniques Used to Study SubmarineGroundwater Discharge in Biscayne Bay, Florida, USGS Fact Sheet 2004-3117
Dulaiova, H. and W. C. Burnett, 2004. An Efficient Method for Gamma Spectrometric
Determination of Radium-226, 228 via Manganese Fibers, Limnology & Oceanography: Methods 2, 256-261
Burnett, W. C., and H. Dulaiova, 2003. Estimating the Dynamics of Groundwater
Input into the Coastal Zone via Continuous Radon-222 Measurements, Journal of Environmental Radioactivity 69 (1-2), 21-35
Chanton, J. P., W. C. Burnett, H. Dulaiova, D. R. Corbett, and M. Taniguchi, 2003.
Seepage Rate Variability in Florida Bay Driven by Atlantic Tidal Height,Biogeochemistry 66, 187-202
Lane-Smith, D. R., W. C. Burnett, and H. Dulaiova, 2002. Continuous Radon-222
Measurements in the Coastal Zone, Sea Technology October 2002, 37-45
130
Kim, G., W. C. Burnett, H. Dulaiova, P. W. Swarzenski, and W. S. Moore 2001.
Measurement of 224Ra and 226Ra Activities in Natural Waters Using a Radon-in-Air Monitor, Environmental Science & Technology 35, 4680-4683
Dulaiova, H., G. Kim, W. C. Burnett and E. P. Horwitz, 2001. Separation and
Analysis of Am and Pu from Large Soil and Sediment Samples, Radioactivity & Radiochemistry 12 (3), 4-15
C O N F E R E N C E P R E S E N T A T I O N S 50TH ANNUAL CONFERENCE ON BIOASSAY, ANALYTICAL AND ENVIRONMENTAL RADIOCHEMISTRY CINCINNATI, OH, USA November 2004
H. Dulaiova, W. C. Burnett, and N. Dimova: High Concentrations of Radon andThoron Discovered in Public Water Supplies
49TH RADIOBIOASSAY & RADIOCHEMICAL MEASUREMENTS CONFERENCE JACKSON, WY, USA October 2003
H. Dulaiova and W. C. Burnett: Determination of Radium Isotopes in Natural Watersusing MnO2-Coated Fiber and Gamma Spectrometry
SIXTH INTERNATIONAL SYMPOSIUM & EXHIBITION ON ENVIRONMENTAL CONTAMINATION IN CENTRAL & EASTERN EUROPE AND THE COMMONWEALTH OF INDEPENDENT STATES PRAGUE, CZECH REPUBLIC September 2003
H. Dulaiova and W. C. Burnett: Evaluation of Water Residence Times in Embaymentsvia Natural Radium Isotopes
Member of the organizing committee and the student poster session jury 2003 IUGG GENERAL ASSEMBLY SAPPORO, JAPAN July 2003
H. Dulaiova, W. C. Burnett, W. S. Moore and H. Bokuniewicz: GroundwaterDischarge and Mixing Estimates in West Neck Bay, Shelter Island, NY via IsotopicApproaches
AGU Student Travel Grant and IUGG Travel Grant recipient 48TH RADIOBIOASSAY & RADIOCHEMICAL MEASUREMENTS CONFERENCE KNOXVILLE, TN, USA November 2002
H. Dulaiova, D. Lane-Smith, R. Peterson and W. C. Burnett: Improved ContinuousRadon Measurements for Coastal Waters
14TH RADIOCHEMICAL CONFERENCE MARIANSKE LAZNE, CZECH REPUBLIC April 2002
H. Dulaiova, W. C. Burnett: Improved Measurement of Radium Isotopes at LowConcentrations in Natural Waters
131
47TH ANNUAL RADIOCHEMICAL MEASUREMENTS CONFERENCE HONOLULU, HI, USA November 2001
H. Dulaiova, G. Kim, W. C. Burnett: Measurement of 224Ra and 226Ra Activities in Natural Waters Using a Radon-in-Air Monitor
EICHROM TECHNOLOGIES, INC. NORTH AMERICAN USERS` GROUP WORKSHOP KNOXVILLE, TN, USA May 2001
H. Dulaiova, G. Kim, W. C. Burnett, E. P. Horwitz: Analysis of Actinide Elementsfrom Large Samples
IRPA CONGRESS ON RADIATION PROTECTION IN CENTRAL EUROPE
BUDAPEST, HUNGARY, August 1999 Poster presentation: H. Dulaiova, V. Beckova, I. Bucina: Determination of Americium
by Extraction Chromatography in Urine Samples 13TH NATIONAL SEMINAR ABOUT SEPARATION CHEMISTRY
LAZNE BOHDANEC, CZECH REPUBLIC, June 1999 H. Dulaiova: Review of Methods for the Determination of 241Am in Excretion
Analysis F I E L D T R I P S A N D C R U I S E S MANILA BAY PHILIPPINES, ASIAAPN Project January 2005
Groundwater Discharge as an Important Land-Sea Pathway in Southeast Asia GULF OF THAILAND AND CHAO-PHRAYA ESTUARY THAILAND, ASIASTART Project January & July 2004
Contribution of Carbon and Nutrient Species into SE Asian Waters via SubmarineGroundwater Discharge
PENNSGROVE AND FLORENCE TOWNSHIPS, NJ & ESCAMBIA COUNTY, FL NEW JERSEY AND FLORIDA , USA
Investigation of natural radioactivity in public water supplies April & May 2004 BISCAYNE BAY AND LOXAHATCHEE RIVER FLORIDA, USAUSGS Project June 2004
Using Novel Geophysical and Geochemical Techniques to Study SubmarineGroundwater Discharge
SICILY ITALY, EUROPEIAEA, CRP Project April 2002
Nuclear and Isotopic Techniques for the Characterization of Submarine GroundwaterDischarge in Coastal Zones
132
WEST NECK BAY SHELTER ISLAND, NY, USAIOC, IHP, SCOR, LOICZ Project May 2002
Assessment and Management Implications of Submarine Groundwater Discharge Intothe Coastal Zone
APALACHICOLA BAY FLORIDA, USANOAA NERR Graduate Research Fellowship 2003-2005
Evaluation of Flushing Rates of Estuaries and Embayments via Natural GeochemicalTracers
GULF OF MEXICO: APALACHEE BAY, SARASOTA BAY& FLORIDA KEYS FLORIDA, USA
Geochemical Tracer Studies 2000-2004 FSU MARINE LABORATORY, TURKEY POINT FLORIDA, USASCOR, LOICZ Project August 2000
Assessing Methodologies for Measuring Groundwater Discharge to the Ocean I N T E R N S H I P S BUDAPEST TECHNICAL UNIVERSITY BUDAPEST, HUNGARYInstitute of Nuclear Techniques January 1999
Analyses of actinides in water samples from the nuclear power plant at Paks MEDITERRANEAN INTERNATIONAL UNIVERSITY NARBONNE, FRANCESummer School August 1998
Environmental protection and chemistry TECHNICAL UNIVERSITY DELFT DELFT, NETHERLANDSInterfaculty Reactor Institute April 1997
Instrumental Neutron Activation Analysis (BISNIS) NUCLEAR POWER PLANT TEMELIN TEMELIN, CZECH REPUBLICDepartment of Analytical Chemistry July 1996
NPP reactor water analysis (AAS, HPLC) P R O F E S S I O N A L S O C I E T Y M E M B E R S H I P S
American Geophysical Union