Isotope tracing of submarine groundwater discharge offshore Ubatuba, Brazil: results of the IAEA-UNESCO SGD project

Results of groundwater and seawater analyses for radioactive ((3)H, (222)Rn, (223)Ra, (224)Ra, (226)Ra, and (228)Ra) and stable (D and (18)O) isotopes are presented together with in situ spatial mapping and time series (222)Rn measurements in seawater, direct seepage measurements using manual and automated seepage meters, pore water investigations using different tracers and piezometric techniques, and geoelectric surveys probing the coast. This study represents first time that such a new complex arsenal of radioactive and non-radioactive tracer techniques and geophysical methods have been used for simultaneous submarine groundwater discharge (SGD) investigations. Large fluctuations of SGD fluxes were observed at sites situated only a few meters apart (from 0cmd(-1) to 360cmd(-1); the unit represents cm(3)/cm(2)/day), as well as during a few hours (from 0cmd(-1) to 110cmd(-1)), strongly depending on the tidal fluctuations. The average SGD flux estimated from continuous (222)Rn measurements is 17+/-10cmd(-1). Integrated coastal SGD flux estimated for the Ubatuba coast using radium isotopes is about 7x10(3)m(3)d(-1) per km of the coast. The isotopic composition (deltaD and delta(18)O) of submarine waters was characterised by significant variability and heavy isotope enrichment, indicating that the contribution of groundwater in submarine waters varied from a small percentage to 20%. However, this contribution with increasing offshore distance became negligible. Automated seepage meters and time series measurements of (222)Rn activity concentration showed a negative correlation between the SGD rates and tidal stage. This is likely caused by sea level changes as tidal effects induce variations of hydraulic gradients. The geoelectric probing and piezometric measurements contributed to better understanding of the spatial distribution of different water masses present along the coast. The radium isotope data showed scattered distributions with offshore distance, which imply that seawater in a complex coast with many small bays and islands was influenced by local currents and groundwater/seawater mixing. This has also been confirmed by a relatively short residence time of 1-2 weeks for water within 25km offshore, as obtained by short-lived radium isotopes. The irregular distribution of SGD seen at Ubatuba is a characteristic of fractured rock aquifers, fed by coastal groundwater and recirculated seawater with small admixtures of groundwater, which is of potential environmental concern and has implications on the management of freshwater resources in the region.     

Isotope tracing of submarine groundwater discharge offshore Ubatuba, Brazil: results of the IAEA–UNESCO SGD project

Results of groundwater and seawater analyses for radioactive (³H, ²²²Rn, ²²³Ra, ²²⁴Ra, ²²⁶Ra, and ²²⁸Ra) and stable (D and ¹⁸O) isotopes are presented together with in situ spatial mapping and time series ²²²Rn measurements in seawater, direct seepage measurements using manual and automated seepage meters, pore water investigations using different tracers and piezometric techniques, and geoelectric surveys probing the coast. This study represents first time that such a new complex arsenal of radioactive and non-radioactive tracer techniques and geophysical methods have been used for simultaneous submarine groundwater discharge (SGD) investigations. Large fluctuations of SGD fluxes were observed at sites situated only a few meters apart (from 0 cm d⁻¹ to 360 cm d⁻¹; the unit represents cm³/cm²/day), as well as during a few hours (from 0 cm d⁻¹ to 110 cm d⁻¹), strongly depending on the tidal fluctuations. The average SGD flux estimated from continuous ²²²Rn measurements is 17 ± 10 cm d⁻¹. Integrated coastal SGD flux estimated for the Ubatuba coast using radium isotopes is about 7 × 10³ m³ d⁻¹ per km of the coast. The isotopic composition (δD and δ¹⁸O) of submarine waters was characterised by significant variability and heavy isotope enrichment, indicating that the contribution of groundwater in submarine waters varied from a small percentage to 20%. However, this contribution with increasing offshore distance became negligible. Automated seepage meters and time series measurements of ²²²Rn activity concentration showed a negative correlation between the SGD rates and tidal stage. This is likely caused by sea level changes as tidal effects induce variations of hydraulic gradients. The geoelectric probing and piezometric measurements contributed to better understanding of the spatial distribution of different water masses present along the coast. The radium isotope data showed scattered distributions with offshore distance, which imply that seawater in a complex coast with many small bays and islands was influenced by local currents and groundwater/seawater mixing. This has also been confirmed by a relatively short residence time of 1–2 weeks for water within 25 km offshore, as obtained by short-lived radium isotopes. The irregular distribution of SGD seen at Ubatuba is a characteristic of fractured rock aquifers, fed by coastal groundwater and recirculated seawater with small admixtures of groundwater, which is of potential environmental concern and has implications on the management of freshwater resources in the region.     

Reconnaissance of submarine groundwater discharge at Ubatuba coast,Brazil, using 222Rn as a natural tracer

Submarine groundwater discharge (SGD), which includes fresh groundwater and recycled seawater, has been recognized as a widespread phenomenon that can provide important chemical elements to the ocean. Several studies have demonstrated that SGD may approach or even exceed freshwater sources in supplying nutrients to coastal zones. This work reports preliminary results of a study carried out in a series of small embayments of Ubatuba, S?o Paulo State, Brazil, covering latitudes between 23 degrees 26'S and 23 degrees 46'S and longitudes between 45 degrees 02'W and 45 degrees 11'W. The main aims of this research were to set up an analytical method to assess 222Rn and 226Ra activities in seawater samples and to apply the excess 222Rn inventories obtained to estimate SGD. Measurements made during the summer of 2001 included 222Rn and 226Ra in seawater, 226Ra in sediment, seawater and sediment physical properties, nutrients and seepage rates. A continuous 222Rn monitor was also used to determine in situ collection of data to study short-term changes at one location. All methods indicated significant inflow of subsurface fluids at rates in excess of several cm per day.     

An efficient and simple method for measuring (226)Ra using the scintillation cell in a delayed coincidence counting system (RaDeCC)

A delayed coincidence counter (RaDeCC), developed to determine ultra-low levels of (223)Ra (half life = 11.1 days) and (224)Ra (half life = 3.6 days) in seawater, was adapted to measure (226)Ra (half life = 1622 years). After pre-concentration of Ra from seawater onto MnO(2)-coated fiber we show in this study that the (226)Ra activity can be determined using the RaDeCC's ability to record alpha decay of its daughters as total counts. For sufficient ingrowth of (222)Rn, the Mn-fiber is hermetically sealed in a column for a few days. Then, the ingrown (222)Rn is circulated through the RaDeCC air-loop system followed by shutting down of the pump and closure of the scintillation cell for equilibration. Counting may be completed within a few hours for seawater samples. Sample measurements with this method agreed well with data obtained using gamma-ray spectrometry. This proves that a set of Ra isotopes ((223)Ra, (224)Ra, and (226)Ra), commonly used for geophysical studies such as mixing rates of different water masses and submarine groundwater discharge, can be efficiently and rapidly measured using the RaDeCC.     

Inferring coastal processes from regional-scale mapping of Radon and salinity: examples from the Great Barrier Reef, Australia

The radon isotope ²²²Rn and salinity in coastal surface water were mapped on regional scales, to improve the understanding of coastal processes and their spatial variability. Radon was measured with a surface-towed, continuously recording multi-detector setup on a moving vessel. Numerous processes and locations of land-ocean interaction along the Central Great Barrier Reef coastline were identified and interpreted based on the data collected. These included riverine fluxes, terrestrially-derived fresh submarine groundwater discharge (SGD) and the tidal pumping of seawater through mangrove forests. Based on variations in the relationship of the tracers radon and salinity, some aspects of regional freshwater inputs to the coastal zone and to estuaries could be assessed. Concurrent mapping of radon and salinity allowed an efficient qualitative assessment of land-ocean interaction on various spatial and temporal scales, indicating that such surveys on coastal scales can be a useful tool to obtain an overview of SGD locations and processes.     

Estimating the dynamics of groundwater input into the coastal zone via continuous radon-222 measurements

Submarine groundwater discharge (SGD) into the coastal zone has received increased attention in the last few years as it is now recognized that this process represents an important pathway for material transport. Assessing these material fluxes is difficult, as there is no simple means to gauge the water flux. To meet this challenge, we have explored the use of a continuous radon monitor to measure radon concentrations in coastal zone waters over time periods from hours to days. Changes in the radon inventories over time can be converted to fluxes after one makes allowances for tidal effects, losses to the atmosphere, and mixing with offshore waters. If one assumes that advective flow of radon-enriched groundwater (pore waters) represent the main input of 222Rn in the coastal zone, the calculated radon fluxes may be converted to water fluxes by dividing by the estimated or measured 222Rn pore water activity. We have also used short-lived radium isotopes (223Ra and 224Ra) to assess mixing between near-shore and offshore waters in the manner pioneered by. During an experiment in the coastal Gulf of Mexico, we showed that the mixing loss derived from the 223Ra gradient agreed very favorably to the estimated range based on the calculated radon fluxes. This allowed an independent constraint on the mixing loss of radon-an important parameter in the mass balance approach. Groundwater discharge was also estimated independently by the radium isotopic approach and was within a factor of two of that determined by the continuous radon measurements and an automated seepage meter deployed at the same site.     

Occurrence of 222Rn, 226Ra, 228Ra and U in groundwater in Fujian Province, China

222Rn, 226Ra, 228Ra and U were determined in a total of 552 groundwater samples collected throughout Fujian Province of China. The geometric mean concentrations of 222Rn, 226Ra, 228Ra and total U in the groundwater were 147.8 kBq m-3, 12.7 Bq m-3, 30.2 Bq m-3 and 0.54 microgram kg-1, respectively. High groundwater 222Rn was explained by the predominantly granitic rock aquifers in Fujian. A lifetime risk of 1.7 x 10(-3) was estimated for the ingestion of groundwater 222Rn. High ratios of 228Ra to 226Ra contents (geometric mean of 2.4) and their disproportion suggest that 228Ra should also be measured in the assessment of population doses from drinking water in the regions of high rock or soil 232Th. No significant correlation between the 222Rn concentrations in groundwater and air was found.     

A nationwide survey of outdoor radon concentration in Japan

Nationwide outdoor radon (222Rn) concentrations in Japan were measured to survey the environmental outdoor 222Rn level and to estimate the effective dose to the general public from 222Rn and its progeny. The 222Rn concentration was measured with a passive-type radon monitor. The 222Rn monitors were installed at about 700 points throughout Japan from 1997 to 1999. The annual mean 222Rn concentration in Japan was estimated from four quarters measurements of 47 prefectures in Japan. Nationwide outdoor mean 222Rn concentration was 6.1 Bq m(-3). This was about 40% of the indoor 222Rn concentration in Japan. The 222Rn concentration in Japan ranged from 3.3 Bq m(-3) in the Okinawa region to 9.8 Bq m(-3) in the Chugoku region, reflecting geological characteristics. Seasonal variation of outdoor 222Rn concentration was also found to be lowest in July to September, and highest in October to December. From the results of this 222Rn survey and previous indoor 222Rn survey program, the effective dose to the general public from 222Rn and its progeny was estimated to be 0.45 mSv y(-1).     

Present status of 222Rn in groundwater in Extremadura

Radon-222 was measured in groundwater sources of Extremadura (Spain), analyzing 350 samples from private and public springs, wells, and spas by liquid scintillation counting (LSC) and gamma spectrometry. The (222)Rn activity concentrations ranged from 0.24 to 1168BqL(-1). The statistical analysis showed a log-normal distribution with a mean of (111+/-7)BqL(-1) and a median of (36+/-3)BqL(-1). A hydrogeological study revealed correlations between the activity concentration and the aquifer material's characteristics. A map of (222)Rn in groundwater was elaborated and compared with the natural gamma radiation map for this region. About 35% of the samples showed (222)Rn activity concentrations above the Euratom recommended limit of 100BqL(-1). Three uranium series radionuclides ((238)U, (234)U, and (226)Ra) were also assayed by alpha-particle spectrometry, estimating the annual effective dose due to the presence of these natural radionuclides in drinking water.     

Development and calibration of a portable radon sampling system for groundwater Rn activity concentration measurements

The assembling of a system for field sampling and activity concentration measurement of radon dissolved in groundwater is described. Special attention is given in presenting the calibration procedure to obtain the radon activity concentration in groundwater from the raw counting rate registered in a portable scintillation detector and in establishing the precision of the activity concentration measurements. A field procedure was established and the system tested during one year of monthly observations of ²²²Rn activity concentration in groundwater drawn from two wells drilled on metamorphic rocks exposed at Eastern São Paulo State, Brazil. The observed mean ²²²Rn activity concentrations are 374 Bq/dm³ in one well and about 1275 Bq/dm³ in the other one. In both wells the ²²²Rn activity concentrations showed a seasonal variation similar to variations previously reported in the literature for the same region.     

Time-series sampling of Ra and Ra at the inlet to Great South Bay (New York): a strategy for characterizing the dominant terms in the Ra budget of the bay

Ra isotopes are a powerful tool for quantifying the flux of submarine groundwater discharge (SGD) into the sea. Previous studies of ²²³Ra and ²²⁴Ra mass balances in coastal embayments have shown that the Ra balance is dominated by supply via SGD, exchange with the open ocean and radioactive decay. The current study shows that a single time series over a tidal cycle at the principal inlet to Great South Bay (NY, US) is sufficient to determine the net flux of Ra across the inlet, and also can be used to estimate the decay of short-lived Ra in the bay. Estimates of the net Ra flux obtained from a single tidal time-series by using three different approaches agree with those determined from a more time-consuming survey of Ra within the bay, and may represent a first step of estimating SGD in bays and coastal lagoons.     

Investigation of residence time and groundwater flux in Venice Lagoon: comparing radium isotope and hydrodynamical models

The four naturally-occurring isotopes of radium were coupled with a previously evaluated hydrodynamic model to determine the apparent age of surface waters and to quantify submarine groundwater discharge (SGD) into the Venice Lagoon, Italy.Mean apparent age of water in the Venice Lagoon was calculated using the ratio of ²²⁴Ra to ²²⁸Ra determined from 30 monitoring stations and a mean pore water endmember. Average apparent age was calculated to be 6.0 d using Ra ratios. This calculated age was very similar to average residence time calculated for the same period using a hydrodynamic model (5.8 d).A mass balance of Ra was accomplished by quantifying each of the sources and sinks of Ra in the lagoon, with the unknown variable being attributed to SGD. Total SGD were calculated to be 4.1 ± 1.5, 3.8 ± 0.7, 3.0 ± 1.3, and 3.5 ± 1.0 × 10¹⁰ L d⁻¹ for ^{223,224,226, 228}Ra, respectively, which are an order of magnitude larger than total mean fluvial discharge into the Venice Lagoon (3.1 × 10⁹ L d⁻¹). The SGD as a source of nutrients in the Venice Lagoon is also discussed and, though significant to the nutrient budget, is likely to be less important as the dominant control on SGD is recirculated seawater rather than freshwater.     

222Rn and 220Rn concentrations in soil gas of Karkonosze-Izera Block (Sudetes, Poland)

Soil gas 222Rn and 220Rn concentrations were measured at 18 locations in the Karkonosze-Izera Block area in southwestern Poland. Measurements were carried out in surface air and at sampling depths of 10, 40 and 80 cm. Surface air 222Rn concentrations ranged from 4 to 2160 Bq m(-3) and 220Rn ranged from 4 to 228 Bq m(-3). The concentrations for 10 and 40 cm varied from 142 Bq m(-3) to 801 kBq m(-3) and 102 Bq m(-3) to 64 kBq m(-3) for 222Rn and 220Rn, respectively. At 80 cm 222Rn concentrations ranged from 94 Bq m(-3) to >1 MBq m(-3). The 220Rn concentrations at 80 cm varied from 45 Bq m(-3) to 48 kBq m(-3). The concentration versus depth profiles for 222Rn differed for soils developed on fault zones, uranium deposits or both. Atmospheric air temperature and soil gas 222Rn and 220Rn were negatively correlated. At sampling sites with steep slopes, 220Rn concentrations decreased with depth.     

Radon concentration in soil gas around local disjunctive tectonic zones in the Krakow area

The purpose of this study was to investigate radon in the vicinity of geologic fault zones within the Krakow region of Poland, and to determine the influence of such formations on enhanced radon concentrations in soil. Radon ((222)Rn and (220)Rn) concentration measurements in soil gas (using ionization chamber AlphaGUARD PQ2000 PRO and diffusion chambers with CR-39 detectors), as well as radioactive natural isotopes of radium, thorium and potassium in soil samples (using gamma ray spectrometry with NaI(Tl) and HPGe detectors), were performed. Site selection was based on a geological map of Krakow. Geophysical methods (ground penetrating radar and shallow acoustic seismic) were applied to recognize the geological structure of the area and to locate the predicted courses of faults. Elevated levels of radon and thoron in soil gas were found in the study area when compared with those observed in an earlier survey covering Krakow agglomeration. For (222)Rn, the arithmetic mean of registered concentration values was 39 kBq/m(3) (median: 35.5 kBq/m(3)). For (220)Rn, the arithmetic mean was 10.8 kBq/m(3) and median 11.8 kBq/m(3).     

Measurements of indoor radon concentrations in the Santiago de Compostela area

Galician soils are among those with the highest (222)Rn exhalation rates in Spain. A year-round study of the indoor (222)Rn concentration in buildings in the Santiago de Compostela area (Galicia, Northwest of Spain) was performed. The study is based on systematic samplings with active charcoal canisters, following a modified EPA 520/5-87-005 protocol. These measurements were complemented by others obtained using etched track dosimeters. Each data set follows a log-normal distribution, with a geometric mean of (253+/-3)Bqm(-3) for charcoal canisters and (285+/-2.5)Bqm(-3) for etched track detectors. After correcting for the different measuring conditions, the mean value of both methods differed by only 2%. A careful analysis of the seasonal dependence of our measurements did not reveal any significant seasonal variations in the (222)Rn concentration. Parallel to these measurements, different meteorological parameters were recorded, which revealed a direct correlation between the indoor radon concentration and the outdoor temperature derivative with respect to time.     

Coupled radon, methane and nitrate sensors for large-scale assessment of groundwater discharge and non-point source pollution to coastal waters

We constructed a survey system of radon/methane/nitrate/salinity to find sites of submarine groundwater discharge (SGD) and groundwater nitrate input. We deployed the system in Waquoit Bay and Boston Harbor, MA where we derived SGD rates using a mass balance of radon with methane serving as a fine resolution qualitative indicator of groundwater. In Waquoit Bay we identified several locations of enhanced groundwater discharge, out of which two (Childs and Quashnet Rivers) were studied in more detail. The Childs River was characterized by high nitrate input via groundwater discharge, while the Quashnet River SGD was notable but not a significant source of nitrate. Our radon survey of Boston Harbor revealed several sites with significant SGD, out of these Inner Harbor and parts of Dorchester Bay and Quincy Bay had groundwater fluxes accompanied by significant water column nitrogen concentrations. The survey system has proven effective in revealing areas of SGD and non-point source pollution.     

A survey of indoor workplace radon concentration in Japan

Radon ((222)Rn) concentration was measured at indoor workplaces in Japan to estimate effective dose to the public from (222)Rn and its progeny. Measurements were made from 2000 to 2003 at 705 sites in four categories of office, factory, school and hospital. Passive type Rn monitors equipped with two sheets of polycarbonate thin films for measuring radon concentrations were installed at observation sites and replaced every 3 months to observe seasonal variations in (222)Rn concentrations. The range of annual mean (222)Rn concentrations for all sites was 1.4-182 Bq m(-3), with the arithmetic mean and standard deviation were 20.8 and 19.5 Bq m(-3). Annual mean (222)Rn concentration observed at office, factory, school and hospital were 22.6, 10.1, 28.4 and 19.8 Bq m(-3), respectively. Seasonal variations in (222)Rn concentrations at offices, schools and hospitals were similar to those found in dwellings, and variations in factories were similar to those found in outdoor environments. (222)Rn concentration observed in every quarter period was found to decrease as follows: school>office>hospital>factory. The average effective dose to the public due to (222)Rn was estimated to be 0.41 mSv y(-1) weighted by the working population. Considering the (222)Rn exposure in indoor workplaces, effective dose to the general public is estimated to be in the range from approximately 0.42 to 0.52 mSv y(-1).     

A case study of radon-222 transport from continental North-East Asia to the Japanese islands in winter by numerical analysis

A case study of the regional transport ( approximately 3000 km) of radon-222 ((222)Rn) from continental North-East Asia to the Japanese islands was performed by numerical analysis using five separate source areas (South, Middle and North China, Russia and Korea), while a seasonal northwest wind blew over the Japan Sea. The results for three periods (Term I: 16-18, Term II: 22-25 and Term III: 27-28 in December 1990) were compared with concentrations measured at the Kanazawa site (near the coast of the Japan Sea facing the seasonal wind) and the Nagoya site (overland and downwind on the shores of the Pacific Ocean). Most of the (222)Rn at the Kanazawa site was calculated to come from North China and Korea in Term I, Middle China, North China, and Korea in Term II, and Russia and Korea in Term III. The considerable differences in the origins of (222)Rn emanated from the continent were estimated between Terms I, II and III, even though the similar northwest wind was dominant over the Japan Sea. A contour line analysis indicated movement of (222)Rn emanated from Middle China in a northerly direction first and then a southeasterly direction, resulting from low pressure. The results suggest that the low-pressure systems play an important role in the transport of (222)Rn in North-East Asia.     

Natural radioactivity in Brazilian groundwater

More than 220 groundwater samples were analyzed for 228Ra, 226Ra, 222Rn, 210Pb, U(nat), Th(nat), pH, conductivity, fluoride and some additional elements determined by ICP-MS. Since samples from several Brazilian states were taken, involving areas with quite different geologies, no general trend was observed relating the chemical composition and the natural radionuclide content. On the other hand, 210Pb strongly depends on the water content of its progenitor, 222Rn. The values obtained during the present work were compared with those reported by Hainberger et al. [Hainberger, P.L., de Oliveira Paiva, I.R., Salles Andrade, H.A., Zundel, G., Cullen, T.L., 1974. Radioactivity in Brazilian mineral waters. Radiation Data and Reports, 483-488.], when more than 270 groundwater samples were analyzed, mainly, for 226Ra. Based on the results of both works, it was possible to build a database including the results of both works, generating a set with the radium content of circa 350 groundwater sources. It was demonstrated that 228Ra, 226Ra, 222Rn, 210Pb and U(nat) content in Brazilian groundwater follows a lognormal distribution and the obtained geometric mean were 0.045, 0.014, 57.7, 0.040 BqL(-1) and 1.2 microgL(-1), respectively.     

Contribution of (222)Rn-bearing water to the occupational exposure in thermal baths

The present study investigates the short- and long-term effects of radon ((222)Rn) released from water on the progeny exposure in a thermal spa. For the purposes of this work, the Polichnitos spa was used as a case study. The bathroom was supplied with water containing 110-210 kBq m(-3) of (222)Rn. The (222)Rn concentration in air and the short-lived (222)Rn progenies in attached and unattached form were monitored into the bathroom and the surrounding premises. The equilibrium factor (F-factor) and the unattached fraction were estimated. The results of this study show that water flow during bath filling is by far the dominant mechanism by which (222)Rn is released in the air of the bathroom. The progeny exposure was correlated linearly with the (222)Rn concentration in the entering water. The annual effective dose received by a worker was found to be below the lower limit value of 3 mSv recommended by ICRP 65. The dose limit was exceeded only for water containing more than 300 kBq m(-3).