ARTICLE IN PRESS

Applied Radiation and Isotopes 66 (2008) 1599– 1603

Contents lists available at ScienceDirect

Applied Radiation and Isotopes

0969-80

doi:10.1

� Tel.:

E-m

journal homepage: www.elsevier.com/locate/apradiso

Monitoring natural and artificial radioactivity enhancement in the AegeanSea using floating measuring systems

C. Tsabaris �

Hellenic Centre for Marine Research, P.O. Box 712, Anavyssos, Gr-19013, Greece

a r t i c l e i n f o

Keywords:

Underwater radioactivity222Rn137Cs

Rainfall

Oceanographic buoys

43/$ - see front matter & 2008 IAEA. Publish

016/j.apradiso.2008.01.020

+30 22910 76410; fax: +30 22910 76323.

ail address: tsabaris@ath.hcmr.gr

a b s t r a c t

In the present work, the enhancement of radioactivity due to rainfall in the Aegean Sea using floating

measuring systems was observed and quantified. The data were acquired with a NaI underwater

detection system, which was installed on a floating measuring system at a depth of 3 m. The results of

natural and artificial radioactivity are discussed taking into account the rainfall intensity and wind

direction. The activity concentration of 214Bi increased up to (9917102) Bq/m3 after strong rainfall in the

North Aegean Sea in winter (humid period) with east wind direction. On other hand, the maximum

activity concentration reached the level of (110710) Bq/m3 in summer (dry period) during south winds.

& 2008 IAEA. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Monitoring of the seas is a major concern for the marinescientific community and the need for good and calibratedmeasurements at sea is obvious. In the marine environment,natural radionuclides mainly result from the weathering andrecycling of terrestrial mineral and rocks, where their distribu-tions depend on their physical, chemical and geological properties(Satyajit et al., 2000). The artificial radioactivity is extremely lowcompared to the concentration of natural radionuclides, so that,neither poses a risk to marine flora nor a health hazard to thepopulation consuming seafood. However, radioactivity monitor-ing of the marine environment based on a continuously operatingnetwork could offer very important information on the concen-trations of specific radionuclides (natural and anthropogenic) forother applications (like rainfall, submarine groundwater dischargeand submarine faults) (Wedekind et al., 1999; Tsabaris and Ballas,2005), improving the methodologies for quantifying marineradioactivity with the in situ gamma-ray spectrometry technique.

137Cs is of special interest among the artificial radionuclides,because it is a long-lived radionuclide (30.05 years); it is used as aradionuclide tracer in seawater and constitutes the artificialradionuclide of greatest radiological significance in the marineenvironment (Papucci and Delfanti, 1999; Volpe et al., 2002;Delfanti et al., 2004). It persists in the environment from falloutfrom Chernobyl, weapon tests and nuclear power and processingfacility discharge; it is transported over long distances by water

ed by Elsevier Ltd. All rights reser

currents and contributes towards radioactive contamination ofthe marine food chain.

222Rn (half life 3.823 d) is a noble gas, and its progeny (214Pband 214Bi) are found in aerosol particles in accumulation-mode.Rainfall brings atmospheric 222Rn progeny to the marine environ-ment by its scavenging effect (within and below the cloud).Rainfall increases temporarily the radon progeny activity concen-trations from several percent to several tens of percent of intensitycompared to dry conditions with no rain (Nishikawa et al., 1995).However, the variation of radon progeny activity is not constantmainly due to rainfall intensity, rainfall type and humidity(Yoshioka, 1992). It has been measured (laboratory conditions)that the activity concentration of radon progeny in rainwater inFinland amounts up to 105 Bq/L (Paatero, 2000).

The continuous monitoring of 222Rn progeny, which aretransported by rainfall to the seawater, is necessary to assessproperly the cause of any variation in the marine environmentalgamma-ray dose rate. For instance, 222Rn progeny variation is aserious problem for the monitoring of any release in the air ofradionuclide gas from a nuclear facility. In addition, the rainfallconcentration ratio of 214Bi to 214Pb can give significant informa-tion on the type of cloud (Takeyasu et al., 2006), where the rainfallis produced.

The quantitative information on activity concentrations usingon line networks is very scarce because a lot of preparation isneeded before the deployment of the system, and carefuladjustments are needed for power consumption, stability condi-tions and data transmission. Using newly developed methods forthe application and data analysis of the system, the present worktries to overcome the above problems. Many applications havebeen done in the past with similar systems (Aakenes, 1995a;Povinec et al., 1996; Wedekind et al., 1999; Van Put et al., 2004).

ved.

ARTICLE IN PRESS

C. Tsabaris / Applied Radiation and Isotopes 66 (2008) 1599–16031600

The problem of the quantitative estimation of the detectedactivities as well as the reduction of the lower level of detectiondemands many improvements, although a lot of progress has beenmade by developing software for simulating underwater NaIdetection systems (Vojtyla, 2001; Vlachos and Tsabaris, 2005;Ntziou, 2004).

The continuous gamma ray monitoring in the Aegean Sea hasbeen carried out to demonstrate the enhancement of natural andanthropogenic radioactivity due to rainfall. The measurementshave been performed by an underwater NaI(Tl) system RADAM(Aakenes, 1995b) with high efficiency (100%) and low powerconsumption (2 W). The measured data were provided by thePOSEIDON network (Soukissian et al., 1999) and analyzed forseparating the artificial from the natural part of radioactivity. Theresults of the artificial and natural radioactivity due to rainout inthe Aegean Sea are correlated with the season, the rain intensity(from model prediction) and the wind direction.

s

10-1

100

214 B

i & 20

8 Tl &

137 C

s

40K

214 P

b

2. Material and methods

The setup and application of the stationary monitoringnetwork POSEIDON (Soukissian et al., 1999) has been incontinuous mode in the Aegean Sea since 1998. The POSEIDONnetwork comprises now nine (9) stations and the measured dataare transmitted every 3 or 6 h to the operational center (HCMR).The network consists of stationary gamma ray underwaterspectrometers operating at open sea with satellite data transmis-sion for long term monitoring. A description of the floatingmeasuring system together with the underwater detector hasbeen published recently elsewhere (Tsabaris and Ballas, 2005).The study area was focused mainly at the North Aegean Sea,named Athos (A), at the North-East, named Lesvos (L) and closeto Athens, named Saronikos Gulf (S) (see Fig. 1). The LAT andLON coordinates are 39 57.932, 24 43.230 for Athos (A), 39 09.410,25 48.515 for Lesvos (L) and 37 51.151, 23 42.370 for SaronikosGulf (S).

The radioactivity detectors were installed for 80% of themonitoring time in the location (A), due to the strong events ofwater mass transfer in the specific region. The marine efficiencycalibration was performed at the National Technical University ofAthens in a tank of 2 m diameter and 2 m height by diluting threereference radioactive sources: 137Cs, 40K and 99mTc (Tsabaris et al.,2005). These values were used to estimate the efficiency curve inthe whole energy interval (0–2000 keV) in order to quantify theactivity concentration (in Bq/m3) of the detected photopeaks.

20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.0035.00

36.00

37.00

38.00

39.00

40.00

41.00

A

L

S

Fig. 1. The location of the floating measuring systems at the Aegean Sea.

3. Results and discussion

The data were sorted for a period of one day and stored forfurther analysis. During the periods of 1999–2005 a lot of datatransmitted to operational center of HCMR for long periods (morethan six months per year). The detector drifts were eliminated bysaving the energy spectrum every 6 h and performing the energycalibration using two energy peaks which are always present:energy threshold in seawater (55 keV when the salinity value is38 psu) and natural radiation of 40K (1461 keV). The data analysisfor the net area and other useful parameters was performed withthe software package ‘‘SPECTRG’’ (Kalfas, 1991). The deconvolutionprocedure of the measured spectra for resolving the energygamma ray peaks, has been published elsewhere (Tsabaris andBallas, 2005).

The NaI underwater detector was placed in a monitoringfloating buoy at 3 m depth under the sea level. The measurementsof special interest concern data with rainfall, since rainwaterenhances the artificial and natural radioactivity level. Typicalspectra are shown in Fig. 2. The dashed line represents data beforerainfall and the solid line the spectrum immediately after arainfall event. The acquired spectrum without rainfall (dashedline) is defined as a baseline spectrum (background) for measur-ing radioactivity in the sea at 3 m depth, where the cosmicradiation is attenuated, and it does not interfere in the measure-ment. The background spectrum can be subtracted from theforeground spectrum (spectrum after rainfall), in order to resolvethe fallout contributions from any other potential radioactivityenhancement in the region. This spectrum is salinity dependentand is stable for each season. A better approach of the background

0

cp

10-4

10-3

10-2

214 B

i

500 1000 1500 2000Eγ [keV]

Fig. 2. The in situ spectrum of the NaI underwater detection system. The dashed

line depicts background data and the solid line the foreground spectrum due to

rainfall.

ARTICLE IN PRESS

Table 3The measured data for the artificial and natural radioactivity detected by the NaI

system (November 2002) at North Aegean Sea (Athos)

Radionuclide Energy (keV) Net area (counts) Volumetric activity (Bq/m3)

Natural radioactivity214Pb 351 3600 13807130214Bi 609 4143 9917102214Bi 1120 1780 926794214Bi 1764 2660 94979840K 1461 17 078 12 1007610

Artificial radioactivity137Cs 662 550 2172

The integration time of the measurement was 24 h and the averaged rain intensity

for one day was 19 mm.

C. Tsabaris / Applied Radiation and Isotopes 66 (2008) 1599–1603 1601

spectrum is deduced by averaging spectra of the previous dayswithout rainfall.

A typical spectrum of this subtraction is shown in Fig. 3 wherethe radon progeny 214Pb and 214Bi are clearly seen. Information ofthe analyzed gamma energy photopeaks is given in Tables 1 and 2.On the base of the analysis of the final spectrum, the activityconcentrations are given by taking into account the count rate, theintensity and the effective volume of each photopeak. The marinedetection efficiency values for all energies are given also in theprevious work (Tsabaris et al., 2005), and the infinite volume iscalculated in a spherical geometry using as the infinitive radiusthe estimated values from literature (Kedhi, 2004). Using theseparameters the units of the results are deduced in Bq/m3. Someextreme observations of radioactivity variations are given inTables 3 and 4 for dates on which strong rainfall occurred. Theuncertainties are given within 1s. The radioactivity level varieddue to the following radionuclides: 40K, 214Bi, 214Pb and 137Cs. Theanalysis result at 609 keV (214Bi) is corrected by deconvolutionmethods (Kalfas, 1991), since the natural radionuclide 208Tl at583 keV (86%) interferes with the analysis of the peak. In addition,the results from the three gamma rays of 214Bi give similar activityconcentrations within statistical uncertainties (see Table 3). Thedifference of the activity concentrations between the progeny of

105

104

103

102

101

0 1000 2000 3000Eγ (keV)

CO

UN

TS

Pb-214Bi-214

Bi-214Bi-214

nov7_var.spe

Fig. 3. The subtracted spectrum of 7th November 2002 (location A) from the

background spectrum.

Table 1The energies and the intensities of the most intense gamma rays of the 222Rn

progeny 214Bi and 214Pb

Radionuclide Energy (keV) Intensity (%)

214Bi 609 46

1120 16

1764 15

214Pb 242 7.5

295 19

352 37

Table 2The energies and the intensities of the gamma rays of 137Cs and 40K

Radionuclide Energy (keV) Intensity (%)

137Cs 662 8640K 1461 11

222Rn (214Pb and 214Bi) is due to the decay half-life difference andto the solubility factors in the seawater. The activity concentrationof 40K is in good agreement with the salinity sensor values anddecreases during rainfall due to the dilution of rainwater withseawater. The data from the three stations (as drawn in Fig. 1) areaccumulated and analyzed for the specific radionuclides.

The measured data were correlated with wind direction andrain intensity, which are provided from the data base of thePOSEIDON network. The rainfalls were evident experimentallyfrom the salinity value decrease (38.8–35.5 psu) and from theincrease of the gross gamma ray activity of the seawater. Inaddition, the rainfall intensity was estimated by the forecastingmodel (Papadopoulos et al., 2002).

The highest activity enhancement of 214Bi (up to 991 Bq/m3) wasobserved on 7th November 2002 (rain intensity was 19 mm/6 h)during east winds. On 13th of August 2001 the activity concentra-tion of 214Bi (the rainfall intensity was 13 mm/6 h) reacheda maximum value up to 110 Bq/m3 during south winds. Theactivity enhancement of the 222Rn progeny at the other twostations (S and L) was also monitored and quantified. For instance,on 15th February 2000 at station (L) and on 5th of December atstation (S) a similar 222Rn progeny enhancement was observed(see Table 4) due to east winds. The floating measuring systemsrecorded data for both stations (S and L) during humid periods(winter season). The enhancement of 222Rn in winter during windswith east direction was 8–9 times higher compared to winds withsouth direction. This dependence on the wind direction is alsoconfirmed from analysis for other dates with rainfall. The 222Rnenhancement at north Aegean Sea (A) was also affected by theseason as expected from the radon emanation process whichdepends strongly on the water content of the soil (Markkanen andArvela, 1992). This was easily observed from measured values insummer time (see Table 4 at Athos station 13/8/01 and at Lesvosstation 7/5/00).

The highest activity enhancement of 137Cs (up 31 Bq/m3) wasobserved on 13th of August 2001 with winds of north (north-east)direction. This is the maximum value during the period1998–2005 and amounts to an increase of 5–6 times comparedwith reference values (4–6 Bq/m3) in the North Aegean Sea(Papucci and Delfanti, 1999). At 7th of November 2002 theactivity concentration of 137Cs exhibits an enhancement up to21 Bq/m3 (see Table 4), which is 35% lower compared withmeasured data with north wind direction (31 Bq/m3—13/8/01).The measurement on 13th of August 2001 was tested performingan intercalibration by sampling seawater (40 L) after rainfall at thesame position. The samples were analyzed for 137Cs using the AMPmethod and HPGe detector. The chemical yield of the method was92% and it was estimated using as tracer 134Cs. The intercalibratedvalue at Athos location at 13th of August 2001 was 1972 Bq/m3.

ARTICLE IN PRESS

Table 4The activity concentrations measured together with the rain intensity predicted, wind and current direction

Date of rainfall Volumetric activity

(Bq/m3)

Rain intensity

(mm/6 h)

Wind

direction (1)

Wind type Location

137Cs 27/11/00 1872 21 180 South A214Bi 27/11/00 110710214Pb 27/11/00 15071240K 27/11/00 12 0457120

137Cs 13/08/01 3173 13 0 until 50 North, north-east A214Bi 13/08/01 140712214Pb 13/08/01 28072540K 13/08/01 12 8007122

137Cs 07/11/02 2172 19 90 until 160 East, east-south A214Bi 07/11/02 9917102214Pb 07/11/02 1380713040K 07/11/02 12 1007610

137Cs 20/10/03 1772 14 90 until 160 East, east-south A214Bi 20/10/03 630758214Pb 20/10/03 88078140K 20/10/03 12 1807610

137Cs 5/12/99 Below LLD 8 0 until 50 North, north-east S214Bi 5/12/99 717759214Pb 5/12/99 98777640K 5/12/00 12 2307120

137Cs 15/2/00 Below LLD 10 90 East L214Bi 15/2/00 642755214Pb 15/2/00 91177840K 15/2/00 12 8507125

137Cs 27/5/00 2973 15 10 North L214Bi 27/5/00 141712214Pb 27/5/00 26972440K 27/5/00 12 9877128

The averaged values correspond to an integration time of 24 h (LLD is the lower limit of detection of the system).

C. Tsabaris / Applied Radiation and Isotopes 66 (2008) 1599–16031602

The uncertainty concerns statistics due to the net areas of thethree analyzed photopeaks (661, 605 and 796 keV). At the othertwo monitoring stations (S and L) the activity concentration of137Cs was below the lower limit of detection of the system, excepton 27th of May 2000 at Lesvos location, where strong north windsappeared with rainfall.

The variation of the levels of 222Rn progeny is attributed to thewind direction and due to the type of the rainfall as observedelsewhere (Yoshioka, 1992). At location (A) the level of the radonprogeny is enhanced after rainfall due to a continental air masstransfer from Lemnos Island (or easterly) and from the northernpart of Greece (or northernmost). On the other hand, south windsat the same location do not favor radon progeny enhancement dueto maritime air mass transfer. In addition, the 222Rn progenyin seawater were strongly enhanced when a cold front passed on5th of December 1999 at station (S). Similar phenomena wereobserved recently elsewhere (Takeyasu et al., 2006).

The variation of the 137Cs radioactivity level is attributed to theseasonal variation and to the wind direction. This variation isfavoured from rainfalls during summer and after north and north-east winds for similar rain intensities. It was very difficult tocorrelate precisely the enhancement of 137Cs volumetric activitieswith rain intensities during dry periods (summer) due to the shorttime of rainfall in summer. During summer the maritime masstransfer from the Black Sea favours the 137Cs enhancement. Inaddition, the enhancement due to the direction of the wind can beinterpreted from the air mass transfer from the north and/ornorth-east, where terrestrial mass activity and air activity of 137Csis higher compared to the south Mediterranean region (from theChernobyl accident).

4. Conclusions

The variation of gamma ray activity has been observed withthe continuous monitoring network POSEIDON by acquisitionwith a NaI-system at open sea. During rainfall periods a lot ofmeasured data were produced concerning the qualitative andquantitative information about the source of gamma radiation.According to the performed analysis significant enhancement ofthe counting rate was observed after rainfall depending mainly onthe season and the wind direction.

The 137Cs activity concentration enhancement is favored withnorth (north-east) direction winds and during dry periods. Thenatural radioactivity level as observed from 214Bi and 214Pb, isfavored during East winds (North Aegean Sea) from the con-tinental air mass transfer and during humid periods. At allstations at the Aegean Sea the seasonal variation of 222Rn progenyactivity concentration measured in seawater during winter periodis higher than in summer periods. This seasonal variation dependson the weather conditions at the Aegean Sea for the studiedperiod.

The use of underwater NaI systems in seawater is a significanttool for monitoring studies. The continuous monitoring of theradioactivity variation relative to oceanographic parameters canimprove our understanding of the mechanisms governing therainout of other natural nuclides. In future, such data togetherwith a rain gauge will offer a better understanding of the activityconcentration enhancement of 137Cs and 222Rn progeny at theopen sea. Apart from using such a system for radioprotectionstudies and early warning applications, other applications such as222Rn progeny continuous monitoring will be very beneficial for

ARTICLE IN PRESS

C. Tsabaris / Applied Radiation and Isotopes 66 (2008) 1599–1603 1603

research and long term monitoring in submarine groundwaterdischarge, volcanic areas and on submarine faults.

In the near future, a newly developed system will be used forsuch applications with improved electronics and housing materi-al, for resolving clearly the energy photopeaks and increasing thestability of the detection system in cases of long term operation(more than a year) in the marine environment.

Acknowledgment

The author would like to thank the POSEIDON group for thecontinuous support for the maintenance of the monitoringnetwork systems.

References

Aakenes, U.R., 1995a. Radioactivity monitored from moored oceanographic buoys.Chem. Ecol. 10, 61–69.

Aakenes, U.R., 1995b. RADAM-expandable gamma monitoring units for militaryand civil use. In: International Conference on Environmental radioactivity inthe Artic, Oslo, p. 315.

Delfanti, R., Tsabaris, C., Papucci, C., Kaberi, H., Lorenzelli, R., Zervakis, V.,Tangherlini M., Georgopoulos, D., 2004. Re-distribution of 137Cs Chernobylsignal in the Aegean Sea. In: Proceedings of the International Conference onIsotopes in Environmental Studies—Aquatic Forum, Monaco, p. 7.

Kalfas, C., 1991. SPECTRG—software package for analysis of gamma-ray spectra.Personal Communication.

Kedhi, M., 2004. Evaluation of the infinite radius of a radioactive waterenvironment. In: Proceedings of the International Conference on Isotopes inEnvironmental Studies—Aquatic Forum, Monaco, p. 484.

Markkanen, M., Arvela, H., 1992. Radon emanation from soils. Radiat. Prot. Dosim.45, 269–272.

Nishikawa, T., Tamagawa, Y., Aoki, M., Okabe, S., 1995. Analysis of the time variationof environmental gamma radiation due to the precipitation. Appl. Radiat. Isot.46, 603–604.

Ntziou, I., 2004. Simulation of a NaI underwater detection system in the marineenvironment. Master Thesis, National Technical University of Athens.

Paatero, J., 2000. Wet deposition of radon-222 progeny in northern Finlandmeasured with an automatic precipitation gamma analyzer. Radiat. Prot.Dosim. 87, 273–280.

Papadopoulos, A., Kallos, G., Katsafados, P., Nickovic, S., 2002. The Poseidonweather forecasting system: an overview. Global Atmos. Ocean Syst. 8,219–237.

Papucci, C., Delfanti, R., 1999. 137Caesium distribution in the eastern MediterraneanSea: recent changes and future trends. Sci. Total Environ. 237, 67–75.

Povinec, P.P., Osvath, I., Baxter, M.S., 1996. Underwater gamma-spectrometry withHPGe and NaI(Tl) detectors. Appl. Radiat. Isot. 47, 1127–1133.

Van Put, P., Debauche, A., De Lellis, C., Adma, V., 2004. Performance level of anautonomous system of continuous monitoring of radioactivity in seawater.J. Environ. Radioact. 72, 177–186.

Satyajit, G., Alam, M.N., Islam, M.N., 2000. Concentrations of 222Rn, 226Ra and228Ra in surface sea water of the Bay of Bengal. J. Environ. Radioact. 47,291–300.

Soukissian, T.H., Chronis, G.T., Nittis, K., 1999. POSEIDON: operational marinemonitoring system for Greek Seas. Sea Technol. 40, 31–37.

Takeyasu, M., Iida, T., Tsujimoto, T., Yamasaki, K., Ogawa, Y., 2006. Concentrationsand their ratio of 222Rn decay products in rainwater measured by gamma-rayspectrometry using a low background Ge detector. J. Environ. Radioact. 88,74–89.

Tsabaris, C., Ballas, D., 2005. On-line gamma ray spectrometry at open sea. Appl.Radiat. Isot. 62, 83–89.

Tsabaris, C., Vlachos, D.S., Papadopoulos, C.T., Vlastou, R., Kalfas, C.A., 2005. Set upand application of an underwater g-ray spectrometer for radioactivitymeasurements. Mediterr. Mar. Sci. 6, 35–40.

Vlachos, D.S., Tsabaris, C., 2005. Response function calculation of an underwatergamma ray NaI (Tl) spectrometer. Nucl. Instrum. Methods A 539, 414–420.

Vojtyla, P., 2001. Calibration of monitors used for surveillance of radioactivity ineffluent water from CERN’s accelerator installations. Appl. Radiat. Isot. 55,81–88.

Volpe, A.M., Bandgong, B.B., Esser, B.K., Nianchini, M., 2002. Radiocesium in NorthSan Franscisco Bay and Baja California coastal surface waters. J. Environ.Radioact. 60, 365–380.

Wedekind, Ch., Schilling, G., Gruttmuller, M., Becker, K., 1999. Gamma-radiationmonitoring network at sea. Appl. Radiat. Isot. 50, 733–741.

Yoshioka, K., 1992. The seasonal variation of rainout activity of short-lived radondaughters. Radiat. Prot. Dosim. 45, 395–398.