Interaction of sea ice sediments and surface sea water in the Arctic Ocean: Evidence from excess 210Pb



[1] We measured the activities of 210Pb, 226Ra, 238U and 137Cs in a suite of ice-rafted sediments (IRS) from the Arctic Ocean in an attempt to assess the interaction of sea ice sediments and surface water. The concentrations of these nuclides were compared to those of the benthic sediments in the coastal and shelf regions of the Arctic Ocean, which are believed to be the major source region for the IRS. The concentration factors (CF = activity of a nuclide in IRS/average activity in benthic sediments) are ∼1 and 4-92 for 137Cs and 210Pb, respectively. The CF values for 137Cs are comparable to the values that can be obtained from the previously published data while we report the first set of high CF values of 210Pb. A major portion of 210Pb in some IRS samples is likely derived from surface waters and thus, the concentrations of 210Pb combined with another particle-reactive radionuclide (such as 7Be, 234Th) in IRS might provide information on the residence time and transit time of sea ice-laden sediments.

1. Introduction

[2] The extent, composition, thickness, and drift patterns of sea ice in the Arctic can influence the exchange of heat and gases between the ocean and the atmosphere, thus affecting the global energy balance and oceanic circulation [e.g., Aagaard et al., 1985; Nürnberg et al., 1994]. A major portion of the Arctic Ocean is covered by either perennial or seasonal sea ice. Large amounts of sediments have been reported to be entrained into the ice cover through several processes, including incorporation of suspended sedimentary particles into frazil-ice crystals, uplift of sediments by anchor ice, discharge of river-borne sediments into sea ice, and atmospheric deposition of dust particles [Darby et al., 1974; Reimnitz et al., 1987; Eicken et al., 1997]. A major portion of the sediments are presumably incorporated in the coastal and shelf areas of the East Siberian Arctic region and subsequently transported to other regions in the Arctic [Reimnitz et al., 1987]. This is consistent with the common occurrence of continental shelf foraminifera in the deep Arctic basins, which was attributed to incorporation of benthic sediments into the sea ice in the shallow shelf areas and transport by ice drift [Wollenburg, 1993]. The movement of sediments incorporated in the sea ice and glacial icebergs is controlled by the surface currents of the Arctic Ocean and is dominated by the clockwise Beaufort Gyre in the Amerasian Basin and east flowing Transpolar Drift in the Eurasian Basin [e.g., Pfirman et al., 1995; Eicken et al., 1997; Tucker et al., 1999]. In the limited studies conducted on IRS so far, no exact source areas for sea ice sediments have been located due to the complexity of the drift regime in the Arctic Ocean [e.g., Nürnberg et al., 1994].

[3] The presence of sediments in sea ice affects the radiation balance of the Arctic Ocean. For example, Tucker et al. [1999] found an albedo of 0.48 for sea ice with sediments compared to 0.83 for fresh snow. A decreased albedo leads to melting of ice surrounding the sediment, forming pits (cryoconite holes) that ultimately could lead to the concentrated accumulation of sediments at the surface [Pfirman et al., 1995]. Drifting sediment-laden sea ice also may play a key role in the long-range redistribution of contaminants in the Arctic [e.g., Pfirman et al., 1995; Cooper et al., 1998] as well as affect the particle flux to the deep sea [Hebbeln and Wefer, 1991]. Further, pollutants derived from the Atlantic and Pacific Ocean surface water currents and atmospheric deposition, mainly through Artic haze during the winter-spring seasons, have led to increased pollution in sediments and the water column. The sea ice and IRS could potentially serve as atmospheric collectors where the surface ocean is ice-covered. The sediments in sea ice could also pick up additional particle-reactive contaminants during contact with surface waters.

[4] Particle-reactive radionuclides that are delivered to the ocean surface at a known rate, such as 210Pb (and 7Be), could be used to investigate the accumulation of contaminants from surface waters in sea ice, as well as yield information on the time scale of interaction between IRS and surface water. A comparison of the concentrations of less-particle-reactive and particle-reactive nuclides provides information on the importance of scavenging of particle-reactive nuclides at the sea ice-water interface. We have measured the activities of excess 210Pb (210Pbxs = 210Pbtotal226Ra) from a suite of ice-rafted detritus from the Arctic Ocean along with 226Ra, 238U and 137Cs. Here we report the first set of very high enrichment factors for 210Pb and discuss the factors and processes that could cause the one- to two orders of magnitude higher activities of excess 210Pb compared to those in source sediments in the coastal and continental shelves of the Arctic Ocean.

2. Materials and Methods

[5] One set of sea-ice rafted detritus (1B, 2A, 3, 4A-B, and 5) was collected from the Canada Basin during summer 2000 and the second set was collected during 2001 from the Eurasian Basin (IRS-1, 7, 9, and 10), from Svalbard to the North Pole (Figure 1). Sea ice samples (10–50 L, yielding 10–40 g of dry sediments in Canada Basin and 4–50 g in the Eurasian Basin) were collected and thawed inside the ship. The sediments were separated by decanting the supernatant, after a few days of sediment settling and brought to the laboratory for further analysis. The activities of excess 210Pb, 226Ra, 238U and 137Cs in IRS from the Canada Basin were measured using gamma-ray spectrometry while 210Pb from the Eurasian Basin were measured using alpha spectrometry [Baskaran and Naidu, 1995]. The activities of 210Pbxs and 137Cs were decay corrected to the date of collection.

Figure 1.

Sample location map in the Arctic Ocean.

3. Results and Discussion

[6] The activities of 226Ra, 137Cs and 238U in IRS (Table 1) are comparable to the values reported for the benthic sediments in the coastal and continental shelf regions of the Arctic region [Baskaran and Naidu, 1995; Baskaran et al., 1996, 2000; Meese et al., 1997; Landa et al., 1998; Cooper et al., 1998]. In contrast, the activities of excess 210Pb in IRS are about 1–2 orders of magnitude higher, up to 175 dpm/g. These high values cannot be derived from surface water suspended particulate matter (SPM), which have been found to range from 6.9 to 45.7 dpm/g (mean = 25 dpm/g, 3 data points from 5–10 m in surface waters off Barrow, Alaska; unpublished data). Also, most of the SPM are of biogenic origin while the IRS are of primarily benthic origin. Roberts et al. [1997] reported 330 dpm/g 210Pbxs in one particulate matter in sea ice in the Northeast Polynya. Masque et al. [2002] reported 7Be and 210Pb data for a suite of IRS from the western Fram Strait and suggested that the source of 7Be and 210Pb was atmospheric input to the sea ice.

Table 1. Location and Concentrations of 210Pb, *226Ra and 137Cs in Sea Ice Sediments From the Arctic Ocean
SampleLatitude, NLongitude, E/W210Pbtotal, dpm/g238U, dpm/g226Ra,a dpm/g137Cs,b dpm/g210Pbxs,b dpm/g
  • a

    226Ra concentration in IRS-1, 7, 9 and 10 was assumed to be the average of the remaining 6 samples.

  • b

    210Pbxs = 210Pbtotal226Ra; the error from 222Rn loss [Imboden and Stiller, 1982] is expected to be very small, as 210Pbtotal226Ra; 210Pbxs activities were decay corrected to the time of collection; numbers in parenthesis denote concentration factor, CF. CF = concentration/mean concentration in surficial sediments in Ob, Yenisey Rivers, Kara and Laptev Sea; the mean 210Pbxs and 137Cs activities in 134 samples was found to be 1.90 (range: below detection limit (BD) to 6.56 dpm/g) and 0.75 dpm/g (range: BD to 4.28 dpm/g), respectively [Baskaran et al., 1996, 2000]); NM; Not measured.

IRS-185°32′40°04′E32.4 ± 1.5NM(1.95 ± 0.15)NM32.9 ± 1.6 (17.3)
IRS-788°44.58′1°37.73′W16.8 ± 2.6NM(1.95 ± 0.15)NM16.0 ± 2.8 (8.4)
IRS-988°35.97′0°43.73′E102.2 ± 6.4NM(1.95 ± 0.15)NM108.3 ± 6.9 (57.0)
IRS-1088°25.3′6°37.8′W26.5 ± 1.4NM(1.95 ± 0.15)NM26.5 ± 1.5 (13.9)
1B73°15′153°35.5′W20.9 ± 0.22.88 ± 0.132.04 ± 0.090.34 ± 0.04 (0.47)21.0 ± 0.2 (11.1)
2A73°15′153°35.5′W30.7 ± 0.32.50 ± 0.661.80 ± 0.52BD32.2 ± 0.7 (16.9)
373°15′153°35.5′W20.4 ± 0.23.40 ± 0.192.17 ± 0.130.21 ± 0.38 (0.28)20.3 ± 0.3 (10.7)
4A72°40.7′154°33.5′W159.0 ± 1.83.68 ± 0.332.00 ± 0.190.75 ± 0.08 (1.0)175.0 ± 2.0 (92.1)
4B72°40.7′154°33.5′W114.3 ± 1.12.19 ± 0.161.76 ± 0.130.76 ± 0.02 (1.0)125.5 ± 1.2 (66.1)
572°40.7′154°33.5′W8.60 ± 0.091.93 ± 0.151.95 ± 0.140.16 ± 0.08 (0.23)7.4 ± 0.2 (3.9)

[7] The activities of 137Cs and 210Pb in IRS can be compared to possible source sediments in the coastal/shelf areas of the Arctic Ocean. There are no significant variations on the surficial activities of 210Pb or 137Cs from different regions of the Arctic Ocean (such as coastal Alaskan waters and Russian marginal seas [Weiss and Naidu, 1986; Baskaran and Naidu, 1995; Baskaran et al., 1996, 2000]). For comparison, the extensive database (for 210Pb and 137Cs) for the surficial sediments in the Ob and Yenisey Rivers, the Kara Sea, and the continental shelf sediments from the Pechora Sea, that have 210Pbxs activities ranging up to 6.56 dpm/g (mean = 1.90 dpm/g; n = 134 [Baskaran et al., 1996, 2000]) were utilized. The 137Cs activities varied from below detection limit to 4.28 dpm/g (mean = 0.75 dpm/g). Using these mean values, the concentration factors (CF = activity in IRS/mean activity in bottom sediments) are ∼1 for 137Cs (Table 1), similar to the values calculated from the published data reported for the Arctic Ocean [Cooper et al., 1998; Landa et al., 1998; Meese et al., 1997]. Cesium is not strongly particle-reactive in the marine environment (Kd (= activity (dpm/g) in the solid/activity in water (dpm/cm3)) ∼500 cm3/g, IAEA, 1985) and hence the adsorption of 137Cs from seawater on to IRS is expected to be minimal.

[8] The CF values for 210Pb vary between 4 and 92, and there are no published data for comparison. As discussed above, benthic sediments in the shelf and suspended sediments do not have such high concentrations of 210Pbxs. This is true even if finer sediments with potentially greater adsorptive surface areas were selectively incorporated, since studies have shown that there is no significant correlation between the bulk radioactivity and grain size of the IRS [Cooper et al., 1998]. Other sources include direct atmospheric deposition onto sea ice followed by incorporation into sea ice-bound sediments during melting/refreezing cycles; or scavenging by IRS from the underlying water during contact over an extended period immediately after entrainment or by exposure during subsequent reworking of the ice. Of these sources, as discussed before, the benthic sediments do not have sufficiently high 210Pb to account for the highest IRS values reported here. A comparison can also be made with Pu, another particle-reactive nuclide measured in IRS. The distribution coefficient of Pu (both in the oxidized and reduced form) is comparable to Pb [Sholkovitz, 1983; International Atomic Energy Agency (IAEA), 1985] and hence the concentration factors for Pu and 210Pb should be comparable. Earlier studies have shown that there is no significant enrichment of Pu in IRS [Landa et al., 1998; Cooper et al., 1998]. However, while 210Pb is constantly supplied through the atmosphere, 239,240Pu is a transient tracer with virtually no atmospheric depositional flux since early 1970s. Since Pu is efficiently removed by particle scavenging, most of the Pu in the surface waters derived from the atmospheric fallout in early 1970s has been quantitatively removed.

[9] An estimate on the amount of 210Pbxs in IRS contributed by direct atmospheric deposition can be made by assuming that all 210Pb deposited onto the ice is eventually collected in melt water and scavenged onto IRS. The IRS concentration in dirty sea ice is highly variable, and values ranging between 8 and >250 g m−2 have been reported for the Arctic [Nürnberg et al., 1994; Eicken et al., 1997; Tucker et al., 1999]. Assuming a typical IRS concentration of 100 g m−2, a residence time of IRS in seasonal ice cover of 3 yrs [Nürnberg et al., 1994], atmospheric depositional flux of 210Pb of 0.06 dpm cm−2 y−1 [Weiss and Naidu, 1986; Baskaran and Naidu, 1995; Huh et al., 1997], then a value for the sediments of ∼18 dpm/g is obtained, under the assumption that all the atmospherically-delivered 210Pb is eventually collected onto IRS. This is significantly lower than the three highest values reported in Table 1. Further, it is unlikely that the processes involved are completely efficient, so that only a fraction of 210Pb is likely to be collected by IRS. However, due to the extremely heterogeneous distribution of sediments and the possible range of the area from which 210Pb might be scavenged, it is not known how great a range of IRS concentrations can be generated, although reaching values as high as those measured here may be unlikely.

[10] In most oceans, the highest 210Pb activity in the water column is found in the surface waters due to atmospheric deposition. Surface waters are an obvious source of 210Pb. Direct scavenging may occur during a prolonged period after entrainment while IRS are at the ice/water interface or at subsequent times when collisions or melting move sediment within the ice back in contact with the water. It may also be possible that ice around sediments becomes porous through processes such as melting due to solar heating of the sediments at the surface, and ice reworking may submerge these regions and locate them where seawater may pass through. When sea ice melts as a result of solar heating, sediments can be moved by snowmelt and can concentrate in patches. The transfer of 210Pb from snow or sea ice on to IRS must involve water and hence during this process additional 210Pb could be picked-up by IRS from water. Overall, each sediment sample may have experienced an involved and different history. The amount of water that must have been scavenged to generate the highest 210Pb activity (175 dpm/g) can be obtained using the average concentration of dissolved 210Pb of 5.7 dpm/100L (particulate 210Pb is ∼10% of this) in the upper 5 m from two stations where 1B and 4A (Table 1) were collected (M. Baskaran, unpublished data). Assuming the 210Pb was quantitatively scavenged, each gram of the IRS was in contact with ∼3 m3, corresponding to a Kd of 3 × 106 cm3/g. This Kd value is comparable to the values reported for 210Pb in marine setting [IAEA, 1985; Baskaran and Santschi, 2002]. If some of the atmospherically delivered 210Pb is added to IRS through melting/freezing of sea ice, then, each gram of the IRS need to be in contact with <3 m3 water.

[11] From the discussion above, it appears that a major portion of excess 210Pb is derived from surface waters by effective sorption onto ice-rafted sediments. Sea ice could undergo considerable recycling during alternating melting/refreezing cycles, leading to varying amounts of contact time with the seawater, and thus affecting the amount of 210Pb sorbed from seawater. The variations in the activities of 210Pb clearly indicate that there are differences in the life history of IRS and their extent of interaction with seawater. Recent supplemental data indicate that there are significant differences in Sr- and Nd-isotopic compositions, and so different IRS source regions [Andersson et al., 2004]. In addition, combining 210Pbxs activities with those of other short-lived particle-reactive radionuclides, such as 7Be and/or 210Po, could be utilized to determine the residence/transit time of IRS [Masque et al., 2002].

[12] The implication of our observation is that other atmospherically-delivered contaminants such as Hg and other heavy metals and PCBs to the seasonally covered Arctic Ocean surface could significantly concentrate on IRS and be redistributed to farther distances from the source region. Some recent studies suggest the importance of IRS in redistributing particle-reactive nuclides in the upper water column. Recently, Baskaran et al. [2003] proposed that particulate matter trapped in porous sea-ice could serve as a sieve, by picking-up additional particle-reactive radionuclides when surface water freely flows through the porous particle-laden sea ice. In addition, the activity ratios of 230Th/232Th in particulate matter from the upper 5–10 m in the deep Canada Basin were significantly higher than the crustal average and these were attributed to recycling of sedimentary particles by ice rafting, from ice resulting in additional removal of 230Th from the water column [Trimble et al., 2004].

4. Conclusions

[13] Combining all the recent evidences, it appears that the scavenging of particle-reactive nuclides is a dynamic and ongoing process, and the time-scale of interaction between sea ice-laden sediments and surface water at the sea-ice interface can be investigated using a suite of particle-reactive radionuclides (such as 234Th, 7Be). Our results indicate that the enrichment factor for 210Pb in IRS can be very high and the source of this 210Pb appears to be mainly derived from the interaction of IRS with upper surface waters. We propose that the sea ice and the sediments therein can interact with surface waters, leading to high enrichments of particle-reactive contaminants, both organic and inorganic species.


[14] The work was supported in part by a grant from the National Science Foundation (NSF-OPP-9996337). The IRS sampling in 2001 was supported by The Swedish Polar Research Secretariat (AO-01). We thank the crews and captains of the POLAR STAR (AWS-2000) and ODEN for their assistance with the sample collection. Constructive reviews from Don Porcelli, Per Andersson, and two anonymous reviewers are deeply appreciated.