Impact of Saharan dust events on radionuclide levels in Monaco air and in the water column of the northwest Mediterranean Sea

Impact of Saharan dust events on radionuclide levels in Monaco air and in the water column of the northwest Mediterranean Sea Mai Khanh Pham a , * , Elena Chamizo b , Jos e Luis Mas Balbuena c , Juan-Carlos Miquel a , Jacobo Martín d , Iolanda Osvath a , Pavel P. Povinec e a IAEA-Environment Laboratories, Monte Carlo 98000, Monaco b Centro Nacional de Aceleradores, Universidad de Sevilla, Isla de la Cartuja, 41092 Sevilla, Spain c Departamento de Física Aplicada I, Universidad de Sevilla, 41012 Sevilla, Spain d CADIC-CONICET, Bernardo Houssay 200, Ushuaia 9410, Argentina e Faculty of Mathematics, Physics and Informatics, Comenius University, SK-84248 Bratislava, Slovakia article info Article history: Received 10 November 2015 Received in revised form 7 April 2016 Accepted 13 April 2016 Available online 23 April 2016 Keywords: Air monitoring Radionuclides Saharan dust Northwest Mediterranean DYFAMED Sediment trap abstract Characterization of atmospheric aerosols collected in Monaco (2004e2008) and in sediment traps at 200 m and 1000 m water depths at the DYFAMED (Dynamics of Atmospheric Fluxes in the Mediterranean Sea) station (2004) was carried out to improve our understanding of the impact of Saharan dust on ground-level air and on the water column. Activity concentrations of natural ( 210 Pb, 210 Po, uranium and radium isotopes) and anthropogenic ( 137 Cs, 239 Pu, 240 Pu, and 239þ240 Pu) radionuclides and their isotopic ratios confirmed a Saharan impact on the investigated samples. In association with a large particulate matter deposition event in Monaco on 20 February 2004, the 137 Cs (~40 Bq kg 1 ) and 239þ240 Pu (~1 Bq kg 1 ) activities were almost a factor of two higher than other Saharan deposition dust events. This single-day particle flux represented 72% of the annual atmospheric deposition in Monaco. The annual deposition of Saharan dust on the sea was 232e407 mBq m 2 for 137 Cs and 6.8e9.8 mBq m 2 for 239þ240 Pu and contributed significantly (28e37% for 137 Cs and 34e45% for 239þ240 Pu) to the total annual atmospheric input to the northwest Mediterranean Sea. The 137 Cs/ 239þ240 Pu activity ratios in dust samples collected during different Saharan dust events confirmed their global fallout origin or mixing with local re-suspended soil particles. In the sediment trap samples the 137 Cs activity varied by a factor of two, while the 239þ240 Pu activity was constant, confirming the different behaviors of Cs (dissolved) and Pu (particle reactive) in the water column. The 137 Cs and 239þ240 Pu activities of sinking particles during the period of the highest mass flux collected in 20 February 2004 at the 200 m and 1000 m water depths represented about 10% and 15%, respectively, of annual deposition from Saharan dust events. 1. Introduction Radionuclides found in ground-level air are of natural or anthropogenic origin. Sources of natural radioactivity include cosmogenic radionuclides such as 7 Be and 22 Na and radiogenic radionuclides such as 210 Pb that emanate (via 222 Rn) from the earth's crust to the atmosphere. Anthropogenic radionuclides including 137 Cs, isotopes of plutonium have been introduced to the atmosphere via nuclear weapons tests and releases from nuclear reprocessing facilities and accidents such as Chernobyl and Fukushima (e.g. Livingston and Povinec, 2000, 2002). Radionuclides in the atmosphere rapidly attach on submicron-sized aerosols, and their variability in ground-level air is driven by the behavior of aerosols (Masson et al., 2009; Sýkora et al., 2012; Povinec et al., 2012; Hirose and Povinec, 2015). Atmospheric radionuclides are deposited from the air onto the land and sea surface by wet and dry deposition. In this way, the terrestrial and marine environments are labelled by natural and anthropogenic radionuclides that can be used as tracers of environmental processes. Radionuclide activities and ratios in the environment can vary due to changes in radioisotope production, such as releases from nuclear installation or radon emanating from soil. Variability in radionuclides can also arise from processes such as soil resuspension or biomass burning (Amiro et al., 1996; *Corresponding author. E-mail address: [email protected] (M.K. Pham). http://dx.doi.org/10.1016/j.jenvrad.2016.04.014 0265-931X/ Wotawa et al., 2006; Povinec et al., 2012; Hirose and Povinec, 2015). Saharan dusts events, which can be observed in the atmosphere of southern Europe, are associated with the re-suspension source term. This transport of dust particles from northern Africa to Europe can have a substantial impact on the atmospheric processes in southern Europe. It has also been identified as an important pathway for particle delivery into surface seawater of the northwest (NW) Mediterranean (Moulin et al., 1997; Lee et al., 2002, 2003; Pham et al., 2003, 2005; Ternon et al., 2010). Similar dust events originating in Mongolian and Chinese deserts have also been reported in Japan (e.g. Igarashi et al., 2005). Out of the nine significant Saharan dust events registered between 1998 and 2013 at the Monaco air monitoring station of the International Atomic Energy Agency, two major deposition events were detected on 23e24 November 2002 and 20e21 February 2004, when significant amounts of red-colored particles were collected (Pham et al., 2013). The impact of the exceptional Saharan dust event in February 2004 was also studied by French colleagues using aerosol samples collected in southern France (Masson et al., 2010; Menut et al., 2009). Atmospheric aerosol radioactivity has been continuously monitored and studied in the Principality of Monaco by International Atomic Energy Agency - Environment Laboratories (IAEA-EL) since the 1980s to understand sources and variations of natural ( 7 Be, 40 K, 210 Po, 210 Pb, radium and uranium isotopes) and anthropogenic ( 137 Cs, 239 Pu, 240 Pu, and 239þ240 Pu) radionuclides in the atmosphere, including impact studies of the Chernobyl and Fukushima accidents on the atmospheric and marine environments (Then et al., 1980; Ballestra et al., 1987; Lee et al., 2002; Pham et al., 2011, 2012, 2013). In our previous studies we focused on variations of radionuclides in the atmosphere, on their dry and wet depositions, and on their inputs to the Mediterranean Sea (Lee et al., 2001, 2003; Pham et al., 2011, 2012, 2013). Other studies were devoted to investigations of Saharan dust particles in ground-level air (Lee et al., 2002; Pham et al., 2003, 2005, 2013). The aim of the present work was to study the impact of Saharan dust events between 2004 and 2008 on atmospheric radioactivity and the transport of dust particles in seawater. The main focus is on the Saharan dust event observed in Monaco on 20 February 2004, and its impact on natural and anthropogenic radionuclides in ground-level air, and the water column of the NW Mediterranean Sea using sediment trap samples at the DYFAMED (Dynamics of Atmospheric Fluxes in the Mediterranean Sea) station off the Monaco coast (Fig. 1). The characterization of collected particles was done using X-ray fluorescence, gamma-spectrometry, alphaspectrometry, ICP-MS (Inductively Coupled Plasma - Mass Spectrometry), and AMS (Accelerator Mass Spectrometry) techniques, which represent the state of the art technologies applied in environmental radioactivity research. Data on concentrations and isotope ratios of natural ( 40 K, 210 Pb, 210 Po, 226 Ra, 238 Ra, 234 U, 235 U and 238 U) and anthropogenic ( 137 Cs, 239 Pu, 240 Pu and 239þ240 Pu) radionuclides as well as major and trace elements are presented and discussed. This is the first time that such an extensive set of natural and anthropogenic radionuclides emanating from a Saharan dust event has been studied simultaneously in the atmosphere and marine environment using sediment traps. 2. Materials and methods 2.1. Collection of Saharan dust samples in Monaco The collection of Saharan dust was carried out on the roof of the IAEA-EL premises in Monaco (43  45 0 N, 07  25 0 E) (Fig. 2) using a 22m 2 stainless steel funnel-type collector located about 15 m above the ground (Fig. 2). At the bottom of the collector a rectangular container made of high density polypropylene was placed to collect the dry and wet deposition (Pham et al., 2013). The dust samples were collected by rinsing the surface of the collector with acidified distilled water. Finally, the collected liquid with the dust particles was slowly evaporated to dryness and placed in polyethylene containers for analysis by gamma spectroscopy. Several Saharan dust deposition events were recorded in Monaco between 2002 and 2008. The most two important events were detected during 23e24 November 2002 and 20e21 February 2004 (Pham et al., 2005, 2013). In this study we will mainly focus, however, on the latter event which happened in the night of 20 February 2004 (coded as SD1, Table 1a), when a significant amount of red-colored particles (mass flux of 19,800 mg m 2 ) was collected. Smaller dust events, which were observed on 1 May 2004, 29 July 2005, and 26 May 2008 (named as SD2, SD3 and SD4, respectively) are also investigated in the present work (Table 1a). 2.2. Sediment trap samples collected in the NW Mediterranean The input of the 20 February 2004 Saharan dust event to the water column of the Mediterranean Sea was studied using sediment trap samples collected at 200 m and 1000 m at the DYFAMED station located at 43  25 0 N and 07  52 0 E in the Ligurian Sea (NW Mediterranean), about 52 km off the Monaco coast (Fig. 1). Details of the deployment of the DYFAMED sediment traps has been described in previous work (Martín et al., 2009; Miquel et al., 2011). The collection of the sediment trap samples started on 21 December 2003 and ended on 23 May 2004, with a 14 day collection periods (Table 1b). A total of eight sediment trap samples were collected at a water depth of 200 m, and 11 samples were collected at a depth of 1000 m. The sample coded as ST 9e12 was a combination of four samples (i.e. ST9, ST10, ST11 and ST12) due to the scarcity of material over the sampling periods. The Saharan dust event of 20 February 2004 manifested as red particles in the samples collected at the 200 m, mainly during the deployment time interval of 15e29 February (ST6, mass flux of 1228 mg m 2 d 1 ), with a smaller contribution (853 mg m 2 d 1 ) up to 14 March (ST7). In the 1000 m trap samples the peak-particle flux (893 mg m 2 d 1 ) was found between 29 February and 14 March (ST7), with smaller fluxes in the preceding interval, and the two subsequent intervals. 2.3. Analytical methods Activity concentrations of gamma-emitters ( 40 K, 137 Cs, 210 Pb, 226 Ra and 228 Ra) were determined by gamma-ray spectrometry. Well-type HPGe detectors of 150% and 200% relative efficiency (compared to 7.6 cm in diameter and 7.6 cm high NaI(Tl) detector), operating in the IAEA-EL's underground laboratory with very low background, were used for the analysis of collected samples. The samples were counted for one to two weeks. Further details of the low-background gamma-ray detection system and calibration procedures can be found in Povinec et al., 2004, 2005. The uncertainties (below 10% at 1 sigma) include the counting statistics, detector efficiency calibration, and a background correction. 210 Pb was measured non-destructively (4 g in the well-type HPGe detector) in the whole sample (or a subsample for a big event such as SD1) by gamma-spectrometry. 210 Po was determined by alpha-spectrometry on a sub-sample (around 300 mg each) after radiochemical separation using an Ortec counting system. The calculated activities were corrected for the radioactive decay to the mid-collection period. Typical propagated uncertainties at 1 sigma were around 10%. Uranium isotopes ( 234 U, 235 U, and 238 U) were determined by ICP-MS at the CITIUS laboratory (Departamento de Física Aplicada I, Universidad de Sevilla, Spain) after a total decomposition of the 3 samples. Plutonium isotopes ( 239 Pu and 240 Pu) were determined by AMS at the CAN laboratory (Seville, Spain) after applying radiochemical procedures described in Chamizo et al. (2015). The elemental composition of particles was determined using X- ray fluorescence spectrometry (Spectro X-Lab 2000). Details of the X-ray fluorescence technique can be found in Martín et al. (2009) and Pham et al. (2005). 2.4. Quality assurance of the analytical methods In order to achieve high quality in the reported radionuclide results, regular analyses of reference materials were carried out during the sample measurements. A certified reference material provided by the IAEA (IAEA-385, Irish Sea sediment) (Pham et al., 2008), and a standard reference material (SRM 4353, soil) developed by NIST (National Institute of Standards and Technology, Gaithersburg, USA) were analyzed by gamma- and alphaspectrometry, ICP-MS and AMS. To check for possible contamination problems, blank samples were also analyzed during ICP-MS and AMS measurements. A Saharan soil sample (SS) (26  02 0 34.2 00 N and 14  21 0 17.6 00 E) from nearby the Gerboise site, where French nuclear tests were carried out in the 1960s (Menut et al., 2009; Masson et al., 2010) was also analyzed for radionuclides and trace elements. The source of the Saharan dust collected at the Monaco station may not come from the exact place where the Saharan soil (SS) was collected. The SS sample was sieved, and the <63 m m fraction was used for comparison with the deposit samples collected in Monaco. 3. Results and discussion 3.1. Elemental composition The major element composition of the Saharan dust deposits collected in Monaco during the four studied events, in the Saharan soil (SS) and in the sediment trap samples collected at the DYFAMED station, were similar consisting mainly of Al (5e7%), Si (up to 20%), Ca (up to 12%) and Fe (4%). The concentrations of Al, Si, Ca and Fe were comparable with those observed previously in Saharan dust samples collected in Monaco (Pham et al., 2005), and with those observed in marine particles collected in the Southeast Mediterranean (Herut et al., 2001). However, their concentrations were higher than in the dust originating from the European continent (Guieu et al., 1997; Pham et al., 2003). Trace metal concentrations were similar between samples, except for Cu, Zn, Cr, Sn and Ni, whose concentrations were higher in the Saharan dust deposits compared to in the sediment trap samples and the SS. It has been previously demonstrated that Zn decreases notably from the atmosphere through the water column to the sea floor (Martín et al., 2009, and references therein). A Fig. 1. Map of the Northwestern Mediterranean basin with two studied stations: ( : Monaco principality (43450N, 07250E) and : DYFAMED site (43250N, 07520E). The comparison with Manosque station (southern France: 43500N, 5470E, Masson et al., 2010) is also shown. Note is eastern Magreb region where 80% of Saharan dust event, 21e22 February 2004, arrived to southern France (Menut et al., 2009). 4 sediment core collected from the sea floor at the DYFAMED site revealed that Pb, Zn and Cu were enriched over background levels in the upper centimeters of the sediment column (Martín et al., 2009). The concentration of Zr was exceptionally high in the SS sample (7900 m g/g) compared to the dust and sediment trap samples (maxima at 200 m g/g). Conversely, S, Br, Sr and I were higher in the sediment trap samples; the highest S concentration (up to 3770 m g/ g) was found in particles collected between 28 March 2004 and 23 May 2004 at 200 m depth, and between 9 May 2004 and 23 May 2004 at 1000 m depth. It is well known that particles collected at the coastal sites in the NW Mediterranean are mixtures of urbandominated (European) and crust-dominated (Saharan) components (Chester et al., 1997), and that concentrations of some trace elements (Fe, Mn, Cr, Ni, Cu, Zn, Cd and Pb) in particles coming from Sahara Desert are lower than the ones of European origin (Guieu et al., 1997; Pham et al., 2003). 3.2. Radionuclide compositions 3.2.1. Natural radionuclides Activities of 210 Pb, 210 Po, 226 Ra, 228 Ra, 234 U, 235 U, 238 U and 40 Kare listed in Tables 2a and b. Activity ratios for some of the radionuclides are given in Table 4. 3.2.1.1. Saharan soil and Saharan dust deposition. The activity of uranium isotopes ( 234 U, 235 U and 238 U) was similar in SD1, SD2, and SD3 samples and slightly higher in SD4 (Table 2b). The reported activities for SD1 are comparable with results obtained at the Manosque station (southern France: 43  50 0 N, 5  47 0 E; Fig.1) one or two days later (Masson et al., 2010). They are also comparable with previous results from the Monaco station (Pham et al., 2005). The activities of 226 Ra, 228 Ra (Table 2a) in the Saharan dust samples were very similar, but much lower than the soil sample (SS). The 210 Pb and 210 Po levels determined by gamma and alpha spectrometry, respectively for SD and SS samples were in good agreement (Table 2a). Over the sampling period, their values in the Saharan dust deposit samples varied four-fold (Table 2a). 210 Pb is produced by radioactive decay from its progenitor, 222 Rn, a noble gas in the 238 U decay chain, and emanates primarily from land surface. The 210 Po/ 210 Pb activity ratio for SD1, SD2 and SD3 as well as for SS was1(Table 4), almost two times higher than ratios found in a previous study of Saharan dust deposits at the Monaco station (Pham et al., 2005). The 210 Po activity at the time of radiochemical separation from 210 Pb was the same as of 210 Pb, thus confirming an equilibrium between 210 Pb and 210 Pb at the sampling date. Lower concentrations of 210 Po and 210 Pb in the SD1 sample compared to the other SD samples (Table 2a) could be explained by a direct (unmixed) transfer from the Saharan Desert or in the case of SD2 and SD3 by partial mixing with local re-suspended dust particles. The concentrations of 210 Pb, 210 Po and uranium isotopes in the SS sample were lower than in the SD samples, while the 40 K activity was similar. The activity ratios of 235 U/ 238 U were ~0.046, confirming that the collected particles were of natural origin (crust-dominated from Saharan desert), as demonstrated in previous studies (Pham et al., 2003, 2005). The activity ratios of 234 U/ 238 U were ~1, indicated a radioactive equilibrium between 238 U and its daughter 234 U. However, a further look at other 238 U daughter product, 226 Ra and 210 Pb calls into question this equilibrium as 226 Ra and 210 Pb were enriched relative to the 238 U in all the Saharan dust deposition samples. In the Saharan soil sample (which may not be representative of the Saharan dust deposits collected in Monaco), the 235 U/ 238 U and 234 U/ 238 U activity ratios were respectively about 0.046 and 1, confirming that they were of natural origin (terrestrial crust). Similar to the Monaco dust samples, 226 Ra and 210 Pb were enriched relative to 238 U. Uranium and radium have different behaviors in the Fig. 2. Stainless steel collector on the roof of IAEA-EL premises in Monaco for collection of dry and wet deposits (Note in the lower part shown the important residue of dust deposition SD1 collected on the 20 February 2004). Table 1a Saharan dust deposits collected at Monaco station (43  45 0 N, 07  25 0 E). Sample code Sampling date start Sampling date end Mass flux (mg m 2 d 1 ) SD1 2004-2-20 2004-02-21 19,800 SD2 2004-05-01 2004-05-02 2675 SD3 2005-07-29 2005-07-31 1275 SD4 2008-05-26 2008-05-27 1375 5 environment: uranium is soluble (U 4þ form being oxidized to uranyl anion UO 2 2þ ), while radium is insoluble (existing as RaO) (Cochran, 1992). This may explain the depletion of uranium (compared to radium) in the soil and dust samples. The 210 Po/ 210 Pb activity ratio of Saharan soil was about 1.14 (Table 4), confirming that 210 Pb and 210 Pb were in equilibrium at the sampling date. 3.2.1.2. Sediment trap samples. The 226 Ra levels in sediment trap samples were higher during the three Saharan dust events (samples # ST6, ST7 and ST8) at both depths compared to the atmospheric level (Table 2a). The concentration of uranium isotopes ( 234 U, 235 U and 238 U) varied little between depth and sampling date (Table 2b). The 226 Ra/ 238 U activity ratios of sediment trap samples were similar at both depths and higher than the dust deposition samples (Table 4). This is likely due to solubility of uranium and insolubility of radium in seawater. The levels of 210 Pb and 210 Po (Table 2a) varied in time at both depths, with lower activities in samples ST6, ST7 and ST8 compared to samples collected before the SD1 event (for instance ST4 sampled at 200 m, and ST5 sampled at 1000 m). The 210 Po/ 210 Pb activity ratios (Table 4) in samples ST3 to ST9-12 at 200 m varied from 0.69 to 2.9. The 210 Po activity at the time of radiochemical separation was higher than that for 210 Pb for all samples at 200 m (except for the sample ST3), suggesting that there was already some 210 Po in the Saharan dust deposit before it was transported into the water column. The 226 Ra and 210 Pb showed enrichment compared to 238 U and 234 U. Radon progeny likely adsorbed onto atmospheric Table 1b Depth, date and mass flux of the mooring deployments during the time-series DYFAMED 39 (DYF 39) survey (43  25 0 N, 07  52 0 E). Note the sediment trap samples in this series were principality influenced from the SD1 event. Depth (m) Sample code Sampling date start Sampling date end Mass flux (mg m 2 d 1 ) 200 ST2 2003-12-21 2004-01-04 584.8 200 ST3 2004-01-04 2004-01-18 175.8 200 ST4 2004-01-18 2004-02-01 130.9 200 ST5 2004-02-01 2004-02-15 41.7 200 ST6 2004-02-15 2004-02-29 1228 200 ST7 2004-02-29 2004-03-14 852.7 200 ST8 2004-03-14 2004-03-28 234.1 200 ST9-12 2004-03-28 2004-05-23 209.1 1000 ST2 2003-12-21 2004-01-04 156.3 1000 ST3 2004-01-04 2004-01-18 140.6 1000 ST4 2004-01-18 2004-02-01 85.3 1000 ST5 2004-02-01 2004-02-15 114.9 1000 ST6 2004-02-15 2004-02-29 723.7 1000 ST7 2004-02-29 2004-03-14 893.4 1000 ST8 2004-03-14 2004-03-28 558.7 1000 ST9 2004-03-28 2004-04-11 564.9 1000 ST10 2004-04-11 2004-04-25 173.0 1000 ST11 2004-04-25 2004-05-09 243.8 1000 ST12 2004-05-09 2004-05-23 248.1 Table 2a Radionuclide activities (in Bq kg 1 ) determined by gamma-spectrometry and alpha-spectrometry (for 210 Po). Sample 137 Cs Unc. 226 Ra Unc. 228 Ra Unc 40 K Unc. 210 Pb Unc. 210 Po Unc. SD1 40.7 2.1 38.7 2.4 42.2 1.6 633 15 980 10 1050 20 SD2 37.9 1.9 31.1 1.6 34.6 0.5 574 12 2450 6 2370 40 SD3 38.1 2.2 26.4 2.4 42.4 2.3 680 19 2940 30 2930 50 SD4 23.2 1.2 32.2 1.7 44.5 0.5 531 11 4032 6 n.d. SS 23.0 1.3 117.7 6.1 171.4 1.9 497 11 188 5 215 4 200 m ST2 12.2 0.9 39.0 3.3 38.8 2.6 581 18 323 20 1030 30 ST3 10.2 1.9 82.7 4.8 34.4 1.7 489 13 1510 60 1040 30 ST4 10.6 1.0 95.1 5.2 41.7 1.8 465 12 534 16 1660 50 ST6 22.9 1.3 59.5 3.9 32.0 2.3 483 14 298 12 740 30 ST7 25.4 1.4 51.0 3.6 33.4 2.2 545 15 302 12 780 30 ST8 23.3 1.3 50.9 3.3 34.0 1.8 494 13 287 10 820 30 ST9-12 14.2 0.9 83.2 4.6 24.0 1.7 383 11 347 11 830 30 1000 m ST2 12.7 0.9 61.3 3.5 34.6 1.8 553 14 372 11 ST3 10.4 4.8 86.0 4.6 35.3 1.4 498 12 1760 200 ST4 10.4 0.9 95.4 5.4 37.8 2.2 508 13 509 16 ST5 8.5 2.0 108.6 5.8 32.3 1.4 477 12 1420 70 ST6 22.2 1.6 61.6 3.7 31.0 1.9 538 14 714 20 ST7 24.1 1.7 60.4 4.1 34.9 2.4 559 17 781 22 ST8 19.4 1.2 62.4 3.6 31.4 1.8 527 14 405 12 ST9 16.3 1.5 96.5 5.4 27.8 1.7 480 13 928 26 ST10 14.5 1.5 82.4 4.7 28.3 1.9 467 13 1140 30 ST11 12.1 3.2 85.6 4.5 27.3 1.2 471 11 1130 80 ST12 9.7 0.8 102.0 5.3 23.6 1.2 340 8 366 11 n.d. not determined. 6 aerosol particles, leading to 210 Pb enrichment relative to the soil and the degree of enrichment increased with travel time through the atmosphere. Both polonium and lead have affinity for particles; however, lead preferentially adsorbs to inorganic particles, whereas polonium has a high affinity for biological particles via biological uptake. Therefore particles traveling in the water column incorporated more Po than Pb from the seawater. Similarly, the other elements with a high affinity for organic matter (e.g. Ca) were also enriched in sediment trap samples when compared to Saharan dust and Saharan soil. 40 K levels varied little between samples, as potassium dissolves readily in the water column. The 228 Ra (the decay product in the 232 Th decay chain) concentrations were comparable with 226 Ra levels. 3.2.2. Anthropogenic radionuclides The concentrations of 137 Cs, 239 Pu, 240 Pu and 239þ240 Pu in Saharan dust deposit samples (SD), Saharan soil (SS) sample as well as in sediment trap samples (ST) are shown in Tables 2a and 3. The atom ( 240 Pu/ 239 Pu) ratio and the activity ( 239þ240 Pu/ 137 Cs, 137 Cs/ 239þ240 Pu) ratios are shown in Table 4. 3.2.2.1. Saharan dust deposits and Saharan soil. The activity concentration of 137 Cs in Saharan dust deposits was rather stable from one event to another with higher concentrations (from about 38 to 40 Bq kg 1 ) in the first three events SD1, SD2 and SD3 compared to the last event SD4 (23 Bq kg 1 )(Table 2a), which was closest to Saharan soil (Table 2a). The 137 Cs deposition flux from the Saharan dust event SD1 was 0.79 Bq m 2 . This deposition event, which lasted just one day, corresponds to 72% of the cumulative 137 Cs deposition during 2004 at the Monaco station (1.1 Bq m 2 y 1 ; Pham et al., 2013). It is comparable to the 137 Cs activity (38.7 Bq kg 1 ) and 75% cumulative 137 Cs deposition flux found by Masson et al. (2010) from the same event collected one or two days later at the Manosque station. The 137 Cs activity in the Saharan dust deposits was of the same order of magnitude as the activity of local re-suspended particles collected during dry seasons (Pham et al., 2013). The main sources of 137 Cs variations in the atmosphere may be summarized as follows: (i) Local re-suspension of 137 Cs from soil (Pham et al., 2011; Sýkora et al., 2012); (ii) 137 Cs input from Saharan dust events (Masson et al., 2010; Pham et al., 2005, 2011); (iii) Contributions from field/forest fires burning biota containing 137 Cs of the Chernobyl and global fallout origin (Wotawa et al., 2006; Masson et al., 2009; Hirose and Povinec, 2015). Menut et al. (2009) showed that in the case of the 21e22 February 2004 Saharan dust event, the main part (80%) of the dust found in southern France was from the eastern Maghreb region. Only 0.7% of the African emissions were from the Gerboise site, which could explain 1e5% to the peak-day value observed at southern France station. The 239þ240 Pu levels in Saharan dust deposits ranged from 0.98 Bq kg 1 to 0.68 Bq kg 1 significantly higher than in the Saharan soil (0.096 Bq kg 1 )(Table 3). The plutonium value obtained for SD1 event was also similar to the activity measured in the samples collected at Manosque station (Masson et al., 2010). The 240 Pu/ 239 Pu atom ratio was similar in all dust samples (~0.20) (Table 4) and within uncertainty the same as the Saharan soil (0.17 ±0.02). These values are similar to the global fallout ratio of 0.186 ±0.009 reported by Kelley et al. (1999). The 239þ240 Pu/ 137 Cs activity ratios in dust samples also varied little (Table 4) and were similar to the global fallout value (0.024) reported by Ikeuchi et al. (1999). These results are in agreement with results obtained from other Saharan dust events previously observed at the Monaco station (Pham et al., 2005), and also those measured at the Manosque station (Masson et al., 2010). We can therefore conclude that the origin of anthropogenic radionuclides in Saharan dust particles is primarily global fallout. The 137 Cs/ 239þ240 Pu activity ratios in SD1, SD2 and SD4 also support this hypothesis (Menut at al., 2009, and references therein), however the dust collected on 29 July 2005 (SD3) may have been influenced by another source than global fallout. The higher 137 Cs/ 239þ240 Pu ratio of SD3 may be due to the contribution of locally re-suspended 137 Cs from soil. This hypothesis is supported by the presence of grey/reddish colored particles, compared to the red or reddish-colored particles found in the SD1, SD2 and SD4 samples. The 137 Cs/ 239þ240 Pu activity ratio in the Saharan soil was much higher (239 ±19), reflecting the 137 Cs dominance in the Saharan region. 3.2.2.2. Sediment trap samples. In sediment trap time-series, cesium and plutonium isotopes exhibited different behaviors. The activity of 137 Cs was double in samples ST6-ST8 relative to samples collected before and after the deposition event. These activities were however only half the activity of 137 Cs in the dust sample from Table 2b Activity concentration of uranium isotopes (in Bq kg 1 ) determined by ICP-MS. Sample 234 U Unc. 235 U Unc. 238 U Unc. SD1 28.0 0.6 1.17 0.01 25.2 0.2 SD2 28.0 0.5 1.17 0.01 25.4 0.2 SD3 28.6 2.1 1.20 0.01 25.9 0.2 SD4 36.8 0.7 1.51 0.01 32.6 0.3 SS 16.3 0.6 0.71 0.01 15.4 0.1 200 m ST3 19.1 1.1 1.02 0.01 22.0 0.2 ST6 21.4 0.5 0.95 0.01 20.5 0.1 ST7 22.7 0.4 0.98 0.01 21.4 0.2 ST8 25.7 0.5 1.02 0.01 22.2 0.2 ST9-12 20.5 0.3 0.84 0.01 18.4 0.1 1000 m ST3 19.4 0.8 1.03 0.01 22.3 0.2 ST6 23.3 0.5 0.98 0.01 21.2 0.1 ST7 18.5 0.9 0.91 0.01 19.9 0.2 ST8 22.7 0.5 1.00 0.01 21.5 0.2 ST9 23.2 0.4 1.01 0.01 21.8 0.2 Table 3 Activity concentrations of plutonium isotopes (in Bq kg 1 ) determined by AMS. Sample 239 Pu Unc. 240 Pu Unc. 239þ240 Pu Unc. SD1 0.57 0.01 0.41 0.01 0.98 0.02 SD2 0.54 0.01 0.39 0.01 0.93 0.02 SD3 0.44 0.01 0.32 0.01 0.76 0.02 SD4 0.37 0.01 0.31 0.01 0.68 0.01 SS 0.059 0.003 0.037 0.004 0.096 0.01 200 m ST3 0.61 0.01 0.41 0.01 1.02 0.02 ST6 0.64 0.01 0.46 0.01 1.09 0.02 ST7 0.66 0.01 0.47 0.01 1.13 0.02 ST8 0.68 0.02 0.47 0.01 1.16 0.02 ST9-12 1.16 0.02 0.86 0.02 2.02 0.03 1000 m ST3 0.68 0.01 0.49 0.01 1.17 0.02 ST6 0.34 0.01 0.24 0.01 0.58 0.01 ST7 0.64 0.01 0.46 0.01 1.10 0.02 ST8 0.66 0.01 0.49 0.01 1.15 0.02 ST9 0.65 0.01 0.48 0.01 1.14 0.02 7 SD1. It is well known that cesium is soluble in water (Pham, 1997, and references therein); therefore its levels in particulate matter should be low. The activity of 239 Pu, 240 Pu, 239þ240 Pu were similar for sediment trap samples ST6, ST7 and ST8 which were collected during the Saharan dust event and sample ST3 collected before the event (Table 3). However, in sample ST9-12 collected after the event (between 28 March 2004 and 23 May 2004 at 200 m), the plutonium concentration was two times higher. This sample is unique in that it is a combination of four very low-flux sediment trap samples. In the 1000 m sediment trap sample, ST6 (sampled between 15 February 2004 and 29 February 2004), which was the first sample of sinking particles from the Saharan dust event at this depth, the plutonium activity was two times lower than other samples. There was little variation in the 240 Pu/ 239 Pu atom ratios at either depth across the entire sampling period (Table 4). The 137 Cs/ 239þ240 Pu activity ratio of ST6, ST7 and ST8 samples were two times higher than for other ST samples. Still, these ratios were half those measured in the atmospheric particles (SD1). With the exception of sample ST9-12 at 200 m and ST6 at 1000 m, the 239þ240 Pu activity was around 1.1 Bq kg 1 . This behavior of plutonium isotopes reflects its particle-reactivity compared to the high solubility of cesium in seawater. Taking into account the 137 Cs and 239þ240 Pu activities in dust particles and the reported fluxes of marine particles (around 1mgcm 2 yr 1 in the western Mediterranean (Guieu et al., 1997), the annual deposition from Saharan dust was estimated to 232e407 mBq m 2 for 137 Cs and to 6.8e9.8 mBq m 2 for 239þ240 Pu. These represent significant contributions (28e37% for 137 Cs and 34e45% for 239þ240 Pu) to the total annual atmospheric input of radionuclides to the NW Mediterranean Sea. The 137 Cs and 239þ240 Pu activities during ST6 (period of the highest mass flux at 200 m) and ST7 (period of the highest mass flux at 1000 m) represented about 10% of 137 Cs and about 15% of 239þ240 Pu annual Saharan dust fluxes at each depth. It was worth noting that the total flux of 137 Cs in sediment traps at both depths in 2004 was 3e5 times greater than the annual deposition reported in the previous studies (Lee et al., 2003; Pham et al., 2005). Although cesium has been diluted through the water column, its downward transport appears to be very efficient. 4. Conclusions To understand the impact of Saharan dust on the air and sea of the northwest Mediterranean region, radioisotopes in aerosol and marine particles were characterized from 2004 to 2008. Four specific dust events were captured including a particularly large event in February 2004. The main observations were: (i) The activity concentrations of natural radionuclides ( 210 Pb, 210 Po, uranium and radium isotopes) and their activity ratios, as well as the elemental composition of collected particles, confirmed a Saharan-dust source fingerprint. In a huge particulate matter deposition event captured at the Monaco station on 20 February 2004, the 137 Cs (about 40 Bq kg 1 ) and 239þ240 Pu (about 1 Bq kg 1 ) activities were almost a factor of two higher compared to the other Saharan deposition events. This single-day particle flux represented 72% of annual atmospheric deposition in Monaco. (ii) The annual deposition of the Saharan dust on the sea was estimated to be 232e407 mBq m 2 for 137 Cs and 6.8e9.8 mBq m 2 for 239þ240 Pu, accounting for a significant proportion of the total annual atmospheric input to the NW Mediterranean Sea (28e37% for 137 Cs and 34e45% for 239þ240 Pu). (iii) The 137 Cs/ 239þ240 Pu activity ratios of different Saharan dust events ranged from 31 to 40 with one exceptional value of 50, showing, a global fallout origin for the former cases, or their mixing with local re-suspended soil particles for the latter case. (iv) While the 137 Cs activity varied by a factor of two in marine particles, the 239þ240 Pu activity remained unchanged at the time scale studied at both depths, thus reflecting the different behaviors of Cs (soluble in seawater) and Pu (particle reactive) in the water column. The 137 Cs and 239þ240 Pu activities of the sinking particles at the peak mass fluxes at 200 m and 1000 m represented about 10% of 137 Cs, and about 15% of 239þ240 Pu total annual deposition fluxes from Saharan dust deposits at each depth. Acknowledgment The authors would like to thank Drs. I. Levy and L. Liong Wee Kwong for assistance during XRF and alpha-spectrometry analysis Table 4 Activity and atom ratios for natural and anthropogenic radionuclides. Sample 240 Pu/ 239 Pu Unc. 239þ240 Pu/ 137 Cs Unc. 137 Cs/ 239þ240 Pu Unc. 226 Ra/ 238 U Unc. 210 Po/ 210 Pb Unc. SD1 0.20 0.01 0.024 0.001 41.4 2.3 1.54 0.10 1.08 0.02 SD2 0.20 0.01 0.024 0.001 40.9 2.2 1.22 0.06 0.97 0.02 SD3 0.20 0.01 0.020 0.001 50.3 3.0 1.02 0.09 1.00 0.02 SD4 0.23 0.01 0.029 0.002 34.1 1.9 0.99 0.05 SS 0.17 0.02 0.004 0.001 239 19 7.64 0.40 1.14 0.04 200 m ST3 0.18 0.01 0.100 0.019 10.0 1.9 3.76 0.22 0.69 0.02 ST6 0.20 0.01 0.048 0.003 21.0 1.3 2.90 0.19 2.48 0.14 ST7 0.20 0.01 0.044 0.003 22.6 1.3 2.38 0.17 2.58 0.14 ST8 0.19 0.01 0.050 0.003 20.1 1.2 2.29 0.15 2.86 0.14 ST9-12 0.20 0.01 0.142 0.010 7.0 0.5 4.52 0.25 2.39 0.11 1000 m ST3 0.20 0.01 0.113 0.052 8.8 4.1 3.86 0.21 ST6 0.20 0.01 0.026 0.002 38.1 2.9 2.91 0.18 ST7 0.20 0.01 0.046 0.003 21.9 1.6 3.04 0.21 ST8 0.20 0.01 0.059 0.004 16.9 1.1 2.90 0.17 ST9 0.20 0.01 0.070 0.006 14.3 1.3 4.43 0.25 8 of investigated samples. Special thanks to Dr. S. Morris for a final check and edit for language. Special thanks to the anonymous reviewers. PPP acknowledges a support provided by the EU Research and Development Operational Program funded by the ERDF (project No. 26240220004). The IAEA is grateful to the Government of the Principality of Monaco for the support provided to its Environment Laboratories. References Amiro, B.D., Sheppard, S.C., Johnston, F.L., Evenden, W.G., Harris, D.R., 1996. Burning radionuclide question: what happens to iodine, caesium and chlorine in biomass fires? Sci. Total Environ. 187, 93e103. Ballestra, S., Holm, E., Walton, A., Whitehead, N.E., 1987. Fallout deposition at Monaco following the chernobyl accident. J. Environ. Radioact. 5, 391e400. Chamizo, E., L opez-Lora, M., Villa, M., Casacuberta, N., L opez-Guti errez, J.-M., Pham, M.K., 2015. Analysis of 236 U and plutonium isotopes, 239,240 Pu, on the 1 MV AMS systems at the CNA, as a potential tool in oceanography. Nucl. Instrum. Meth. Phys. Res. B. http://dx.doi.org/10.1016/j.nimb.2015.02.066. Chester, R., Nimmo, M., Corcoran, P.A., 1997. Rain water-aerosol trace metal relationships at Cap Ferrat: a coastal site in the western Mediterranean. Mar. Chem. 58, 293e298. Cochran, J.K., 1992. In: Ivanovich, M., Harmon, R.S. (Eds.), Uranium-series Disequilibrium: Application to Earth, Marine and Environmental Sciences. Clarendon, Oxford, pp. 334e395. Guieu, C., Chester, R., Nimmo, M., Martin, J.-M., Gueroni, S., Nicolas, E., Mateu, J., Keyse, S., 1997. Atmospheric input of dissolved and particulate metals to the north-western Mediterranean. Deep-Sea Res. II 44, 655e674. http://dx.doi.org/ 10.1016/S0967-0645(97)88508-6. Herut, B., Nimmo, M., Medway, A., Chester, R., Krom, M.D., 2001. Dry atmospheric inputs of trace metals at the Mediterranean coast of Israel (SE Mediterranean): sources and fluxes. Atmos. Environ. 35, 803e807. Hirose, K., Povinec, P.P., 2015. Sources of plutonium in the atmosphere and stratosphere- troposphere mixing. Sci. Rep. 5, 15707. http://dx.doi.org/10.1038/ srep15707. Igarashi, Y., Aoyama, M., Hirose, K., Povinec, P., Yabuki, S., 2005. What anthropogenic radionuclides ( 90 Sr and 137 Cs) in atmospheric deposition, surface soils and aeolian dusts suggest for dust transport over Japan. Water Air Soil Pollut. 5, 51e69. Ikeuchi, Y., Amano, H., Aoyama, M., Berezhnov, V.I., Chaykovskaya, E., Chumichev, V.B., Chung, C.S., Gastaud, J., Hirose, K., Hong, G.H., Kim, C.K., Kim, S.H., Miyao, T., Morimoto, T., Nikitin, A., Oda, K., Pettersson, H.B., Povinec, P.P., Tkalin, A., Togawa, O., Veletova, N.K., 1999. Anthropogenic radionuclides in seawater of the far eastern seas. Sci. Total Environ. 237/238, 203e212. Kelley, J.M., Bond, L.A., Beasley, T.M., 1999. Global distribution of Pu isotopes and 237 Np. Sci. Total Environ. 237 (238), 483e500. Lee, S.-H., Povinec, P.P., Pham, M.K., 2001. Atmospheric input of radionuclides to the northwestern Mediterranean. Rapp. Comm. Int. Mer. M edit. 36, 142. Lee, S.-H., Pham, M.K., Povinec, P.P., 2002. Radionuclides variations in the air over Monaco. J. Radioanal. Nucl. Chem. 254, 445e453. Lee, S.-H., LaRosa, J., Levy-Palomo, I., Oregioni, B., Pham, M.K., Povinec, P.P., Wyse, E., 2003. Recent inputs and budgets of 90 Sr, 137 Cs, 239þ240 Pu and 241 Am in the northwest Mediterranean Sea. Deep-Sea Res. II 50, 2817e2834. Livingston, H.D., Povinec, P.P., 2000. Anthropogenic marine radioactivity. Ocean Coast. Manag. 43, 689e712. Livingston, H.D., Povinec, P.P., 2002. A millennium perspective on the contribution of global fallout radionuclides to ocean science. Health Phys. 82, 656e668. Martín, J., Sanchez-Cabeza, J.A., Eriksson, M., Levy, I., Miquel, J.-C., 2009. Recent accumulation of trace metals in sediments at the DYFAMED site (northwestern Mediterranean Sea). Mar. Pollut. Bull. http://dx.doi.org/10.1016/ j.marpolbul.2009.03.013. Masson, O., Piga, D., Le Roux, G., Mary, J., de Vismes, A., Gurriaran, R., Renaud, P.H., Saey, L., Paulat, P., 2009. Recent trends and explanation for airborne 137 Cs activity level increases in France. Radioprotection 44, 327e332. Masson, O., Piga, D., Gurriaran, R., D'Amico, D., 2010. Impact of an exceptional Saharan dust outbreak in France: PM10 and artificial radionuclides concentrations in air and in dust deposit. Atmos. Environ. 44, 2478e2486. Menut, L., Masson, O., and Bessagnet, B., 2009. Contribution of Saharan dust on the radionuclide aerosol activity levels in Europe? The 21e22 February 2004 case study. J. Geophys. Reas. 114, D16202, doi:10.1029/2009JD011767. Miquel, J.-C., Martín, J., Gasser, B., Rodriguez-y-Baena, A., Toubal, T., Fowler, S.W., 2011. Dynamics of particle flux and carbon export in the northwestern Mediterranean Sea: a two decade time-series study at the DYFAMED site. Prog. Oceanogr. 91, 461e481. Moulin, C., Lambert, C.E., Dulac, F., Dayan, U., 1997. Control of atmospheric export of dust from north Africa by the north Atlatic oscillation. Nature 387, 691e694. Pham, M.K., 1997. Comportement des m etaux traces rejet e en milieu fluvial: Application  a la Vienne (Civaux, France), p. 229. Th ese de Doctorat de l'Universit e Marie-Curie Paris VI. Pham, M.K., Povinec, P.P., Lee, S.-H., Oregioni, B., 2003. Atmospheric transport of particles from the north Africa to Monaco. Bull. Bur. Natl. M etrol. 2, 143e146. Pham, M.K., La Rosa, J.J., Lee, S.-H., Oregioni, B., Povinec, P.P., 2005. Deposition of Saharan dust in Monaco rain 2001-2002: radionuclides and elementary composition. Phys. Scr. T118, 14e17. Pham, M.K., Sanchez-Cabeza, J.A., Povinec, P.P., Andor, K., Arnold, D., Benmansour, M., Bikit, I., Carvalho, F.P., Dimitrova, K., Edrev, Z.H., Engeler, C., Fouche, F.J., Garcia-Orellana, J., Gasc o, C., Gastaud, J., Gudelis, A., Hancock, G., Holm, E., Legarda, F., Ik€ aheimonen, T.K., Ilchmann, C., Jenkinson, A.V., Kanisch, G., Kis-Benedek, G., Kleinschmidt, R., Koukouliou, V., Kuhar, B., La Rosa, J., Lee, S.-H., LePetit, G., Levy-Palomo, I., Liong Wee Kwong, L., Llaurado, M., Maringer, F.J., Meyer, M., Michalik, B., Michel, H., Nies, H., Nour, S., Oh, J.-S., Oregioni, B., Palomares, J., Pantelic, G., Pfitzner, J., Pilvio, R., Puskeiler, L., Satake, H., Schikowski, J., Vitorovic, G., Woodhead, D., Wyse, E., 2008. A new certified reference material for radionuclides in Irish sea sediment (IAEA-385). Appl. Radiat. Isot. 66, 1711e1717. Pham, M.K., Nies, H., Betti, M., Povinec, P.P., 2011. Temporal changes of 7 Be, 137 Cs and 210 Pb activity concentration in surface air at Monaco and their correlation with meteorological parameters. J. Environ. Radioact. 102, 1045e1054. Pham, M.K., Eriksson, M., Levy, I., Nies, H., Osvath, I., Betti, M., 2012. Detection of Fukushima Daiichi nuclear power plants accident radioactive traces in Monaco. J. Environ. Radioact. 114, 131e137. Pham, M.K., Povinec, P.P., Nies, H., Betti, M., 2013. Dry and wet deposition of 7 Be, 210 Pb and 137 Cs in Monaco air during 1998e2010: seasonal variations of deposition fluxes. J. Environ. Radioact. 120, 45e57. Povinec, P.P., Comanducci, J.F., Levy-Palomo, I., 2004. IAEA-MEL's underground counting laboratory in Monaco ebackground characteristics of HPGe detectors with anticosmic shielding. Appl. Rad. Isot. 61, 85e93. Povinec, P.P., Commanducci, J.-F., Levy-Palomo, I., 2005. MEL-IAEA's underground counting laboratory (CAVE) for the analysis of radionuclides in the environment at very low-levels. J. Radioanal. Nucl. Chem. 263, 441e445. Povinec, P.P., Holý, K., Chudý, M.,  Sivo, A., Sýkora, I., Je skovský, M., Richt arikov a, M., 2012. Long-term variations of 14C and 137Cs in the Bratislava air - implications of different atmospheric transport processes. J. Environ. Radioact. 108, 33e40. Sýkora, I., Povinec, P.P., Brest akov a, L., Florek, M., Holý, K., Masarik, J., 2012. Resuspension processes control variations of 137Cs activity concentrations in the ground-level air. J. Radioanal. Nucl. Chem. 293, 595e599. Ternon, E., Guieu, C., Loye-Pilot, M.-D., Leblond, N., Bosc, E., Gasser, B., Miquel, J.-C., Martin, J., 2010. The impact of Saharan dust on the particulate export in the water column of the north western Mediterranean Sea. Biogeosciences 7, 809e826. Then, M., Ballestra, S., Yamato, A., Fukai, R., 1980. Delivery of transuranic elements by rain to the Mediterranean Sea. Geochim. Cosmochim. Acta 44, 1091e1097. Wotawa, G., De Geer, L.-E., Becker, A., D'Amours, R., Jean, M., Servranckx, R., Ungar, K., 2006. Inter- and intra-continental transport of radioactive cesium released by boreal forest fires. Geophys. Res. Lett. 33, L12806. http://dx.doi.org/ 10.1029/2006GL026206. 9