During the summer work seasons on the Antarctic island of Berkner, Bory et al. (2010) excavated samples from a slab of snow 3 meters long and 1 meter wide, taking 14 successive layers from the first fresh-snow deposits downwards, each sample about 6.5–10 cm thick from the slab. The 14 samples represented about two years worth of dust deposition and precipitation from the summer of 2001 to early December, 2003. Cut with a stainless steel saw and protected against particulate contamination with plastic sheets and no downwind activity, each layer was melted down into 60 liter samples of meltwater, which Bory et al. analyzed for radioisotope ratios of strontium (Sr) and neodymium (Nd), dust particulate size, and oxygen 18 isotope ratios. The research team then used the isotope signatures and particulate size to determine the source from which the dust deposits arrived, comparing their samples to those from the East Antarctic Plateau (EAP) in both glacial and interglacial eras. The concentration and isotopic make-up of the Berkner island samples varied seasonally as well as in regards to the EAP samples, meaning that the Berkner deposits came from other places than the EAP’s, and that the source of dust may shift from summer to winter as the amount of dust and the wind currents in various land masses fluctuate. Such results suggest that the air masses traveling to the Antarctic are more specified than previously thought, and perhaps from more diverse sources.–Elise Wanger
Bory, A., Wolff, E., Mulvaney, R., Jagoutz, E., Wegner, A., Ruth, U., Elderfield, H., 2010. Multiple sources supply eolian dust to the Atlantic sector of coastal Antarctica: Evidence from recent snow layers at the top of Berkner Island ice sheet. Earth and Planetary Science Letters 291, 138–148.
Bory et al. filtered each 60 l sample from the snow-pit slabs to separate impurities, leaving 12 ml of larger particulate materials to be analyzed for grain size and composition. The rest of the sample was measured for the oxygen 18 isotope ratio. Oxygen naturally comes in three isotopes: oxygen-18, 17 and 16, with oxygen-16 being the most prevalent. Since oxygen-18 has two more neutrons, making it significantly heavier, it falls more readily in precipitation and therefore is first to fall as water vapor cools and condenses into precipitation. By the time water vapor reaches the Antarctic, more oxygen-16 is left. Therefore a higher concentration of oxygen-16 correlates to colder, wetter seasons. Bory et al. used this to date the sample layers of snow. The samples were excavated in 2003 from the surface layer downwards, so the oxygen isotope ratios could be fitted with the record of surface temperatures from that given starting date. These analyses were also compared with samples taken down the entire snow pit wall (1 m deep) in 2 cm intervals, which allowed Bory et al. to check their fitted dates for coherency and provide a higher resolution. The resulting time-scale provided periods as discrete as three weeks to as much as three months. The layers of snow in the samples were visible from various storms due to different grains sizes and colorations, and these horizons were fairly straight and blatant, indicating that each statigraphic layer would be accurately compatible with the fitted time periods across the 3 meter width.
Next Bory et al. analyzed the concentrations and isotope ratios of Sr and Nd. Strontium and neodymium are both earth metals present in both the crust and the mantle. The crust is the geologic layer on the surface of the earth in the form of land and oceanic sediment and makes up about 1% of the earth’s thickness. The mantle is the layer below the crust and composed of a high pressure, high temperature liquid. The concentration and isotope ratios of Sr and Nd are unique to each geologic land mass depending on when the land was formed and how fast (and therefore the temperature) the rock cooled on the surface. Strontium can bond in the same ways calcium can, and Nd can bond in the same manner as potassium (K). Therefore the amount of non-soluble Sr and Nd is also contingent upon the concentration of Ca and K, respectively, even while still below the crust. As the mantle material cools, different minerals crystallize at different temperatures, and therefore each rock-forming event carries different concentrations of elements like Ca and K at different levels. And since different continental land masses formed at different times, the radioactive decay of 87Sr (with a half-life 4.88×1010 years) and 144Nd (with a half-life of 2.29×1015 years) can also act as a geological fingerprint, describing with land mass the deposit originated. For convenience in graphing and calculations, Bory et al. measured the 143Nd/144Nd ratio in respect to the deviation from the standard 143Nd/144Nd ratio of chondrites, which are essentially meteorites impervious to geologic melting and chemical weathering (making them a default reference point for universal original starting ratios and chemical compositions).
In order to account for the possibility of Sr from oceanic distributions (the Berkner island site is right along the coast, and high amounts of Sr exist in sea-salt), the dry residue of the filtered particles (about 12 mL) were analyzed for saltwater content by finding the concentration of sodium (Na+) and chloride (Cl–) ions (which are the components of salt) through ion chromatography. These ions would be almost entirely sea-water derived. If the Sr/Cl and Sr/Na ratios were comparable to those of seawater (4.15×10–4 and 4.33×10–4), this would further indicate a direct seawater influence. Possible Sr from soluble carbonates like ikaite—a crystalline mineral that only exists in conditions too cold for calcite—were also a concern, which could have entered the sample layers with sea-water contributions. The researchers checked for this possibility by measuring the Ca2+ cation content, since carbonates such as ikaite contain significant calcium contents. Also, as with seawater, a predictable Sr/Ca ratio from ikaite would be expected, about 0.01. Of course, any calcium form could just as likely be dust-derived, even in the form of a carbonate, so to assume the calcium-based Sr is completely from salt-water sources provides a “maximum correction” (Bory et al., 2010; 141). In other words, if anything, compensating for Sr from calcium by multiplying the 87Sr/86Sr ratio to that of seawater (0.709) will underestimate the ratio to something lower than the reality.
Assuming that the dust in the sample contains the standard 25 parts per million (ppm) of neodymium (Nd), the concentration of dust in the snow samples could be determined by measuring the total Nd in the sample, deriving the assumed mass of dust from that value, and dividing the mass of dust by the mass of the entire sample. Bory et al. compared this value to the mass of dust particles measured using a laser-sensing particle detector, which yielded the same results within standard error, and also checked the results against previous studies. All the different methods indicated extremely low values of 1 to 6 ppb (parts per ten thousand) of dust. The concentrations varied seasonally, as did the 87Sr/86Sr and 143Nd/144Nd ratios, although to assume consistent seasonal fluctuations from a two-year diagnosis would be quite imprudent, especially since they were inconsistent even within the short temporal range of data. The fall and winter of 2002 and 2003 had the lowest concentrations of dust, while the following springs and summers had the highest. The summer of 2001 (from the earliest samples) defied this pattern with relatively low concentrations. Correspondingly, the two spring/summers with the highest concentrations also had less negative 143Nd/144Nd deviations from the chondritic value than average— meaning that the 143Nd/144Nd ratio was closer to the chondritic value, and thus lower and more radiogenic, than usual—while the summer of 2001 was anomalously quite negative. The 87Sr/86Sr ratio of these seasons was lower—meaning less radiogenic—which makes sense since a higher 143Nd/144Nd ratio (ie. less radiogenic) typically corresponds with a less radiogenic Sr ratio. The Sr ratio between summers, however, do not correspond in the same range of values and the assumption that the fluctuations are seasonal (as opposed to completely random) is questionable. As with Nd, the Sr ratio in the summer of 2001 defies the trend, in the case of the 87Sr/86Sr ratio by being abnormally high and therefore quite radioactive. Bory et al. suggest the patterns from 2001 could be related to the peculiar atmospheric circulation patterns over the Weddell Sea during that season, although such a suggestion is purely conjecture.
Either way, the purpose of the elemental analyses in the samples was primarily to determine where the dust derives and if the sources change seasonally, as well as if these sources differ from those in other regions of Antarctica, especially the East Antarctic Plateau (EAP). Samples from the glacial period of the EAP (during the last ice age) show significant isotopic differences: glacial EAP samples are less radiogenic in regards to Sr and more radiogenic for Nd. Interglacial samples from the EAP—during our era—are more radiogenic in Sr than the glacial EAP but still less than the Berkner samples, and less radiogenic in Nd that the glacial EAP but still more radiogenic on average than Berkner samples. In other words, the interglacial EAP samples are intermediate between the two, and in some cases even overlap the Berkner isotope ratios. Berkner and the EAP could therefore plausibly receive some dust from the same continental origins, also Bory et al. also propose that various mixing could create incidental similarities in ratios too, even without sharing sources. No one has yet determined the exact location of dust sources in Antarctica in any region, except for the glacial EAP that has a “young” radioisotope signature distinct of Patagonia and the Puna-Altiplano region in Southern South America. Past and present wind currents also validate Patagonia as the probable source of glacial EAP dust. Interglacial EAP also has the characteristic isotope signature of Patagonia and the Puna-Altiplano area. For both the glacial and interglacial EAP samples, other possibilities have yet to be confirmed or rejected, and areas such as around the Magellan Strait, the Andes, or Argentina could all possibly contribute, but only Patagonia and the Puna-Altiplano region have been definitively confirmed.
Still, the EAP samples all fall in the range of the Southern South American candidates, unlike the Berkner isotope ratios. Some of the negative deviations from the chodritic 143Nd/144Nd ratios exceed any range in Southern South America, and is less cohesive with the Patagonian signature. The Puna-Altiplano region matches the Berkner region better, and has the ratios to be the sole provenance of the 2002 and 2003 spring/summer seasons. But for the other periods an older land mass is necessary to explain the low Nd radioactivity. Bory et al. propose Australia for an older ratio, but the lower latitude of the land makes the long-distance dust travel less feasible than from Patagonia, and no data supports the transport of Australian dust to Antarctica. Whether dust could travel such as distance on the given wind patterns is undetermined, and even if it could the particulate size would be much smaller than that observed in the samples. Furthermore, the Berkner isotope ratios still fall far outside even the most Australian-favored models of a realistic mixing with Patagonian dust. Another older-crust like isotopic ratio in the southern hemisphere is southern Africa. The Kalahari desert and Coastal Namibia fit with the ratios observed in Brekner, but once again the complications of travel hinder the plausibility of the dust deriving from these sources. Although these particulate sources cannot be discounted, the best option given the grain sizes of the sediments and the dominant wind patterns is for the other source beyond southern South America to be the snow-free regions of EAP. At the edges of the East Antarctic ice caps, between the southern ocean and the Berkner Islands, the land is geologically pre-Cambrian: old enough to have the low radiogenic ratios of Nd, with scattered 87Sr/86Sr ratios. And wind patterns are conducive to having Patagonian and East Antarctic dust mixing and landing on Berkner island together, granted such a possibility is still purely hypothetical. And since such dust would be traveling from low altitudes in the direction of Berkner, it would completely evade the EAP samples, reinforcing the differentiation observed in isotope analysis.
If climate change is as highly contingent on atmospheric circulation as many scientists suspect, understanding global wind circulations could provide insight into creating more accurate future climate models. Our current wind patterns are what keeps the Mediterranean climate temperate, the Sahara hot and dry, and the eastern United States snowy and rainy. As they change, so does global weather. Looking at where dust derives in Antarctica provides further insight into these wind currents and where they travel. If we can compare the current dust sources to those of the last glacial period, we can even understand the differences between our contemporary currents now and those of the ice ages. The Berkner isotope analyses add one more piece to the puzzle of where particles are traveling, and what weather may be traveling with them.