geochemical characterization of dust from the Talos Dome ice core

Polar ice cores contain many proxies, of which mineral dust is a key one in understanding past climate variability. In fact, dust and climate have a strong influence on each other. Dust has both direct and indirect effects on climate by interacting with solar radiation and influencing cloud formation processes, while climate itself can strongly affect production, transport and deposition of dust. For example, it is well known that dust fluxes respond to the transition from glacial to interglacial regimes. Deposition of dust on the Antarctic continent is controlled by a number of climatic and environmental factors, and a reliable reconstruction of the dust record is essential to understand how these factors have changed in the past. But evidence has shown that, at certain depths in the ice, the dust record may be subjected to some degree of alteration. This work was conceived with the main objective of enhancing our comprehension of deep ice processes through the use of an array of different techniques applied to the Talos Dome ice core. In particular, we aim at studying the chemical and physical anomalies present in the deeper part of the dust record and confirm the existence of post-depositional processes which may alter the climatic signal embedded in deep ice. Moreover, we wish to observe how different elements partition between soluble and insoluble phase, at different depths of the ice core and link the geochemical patterns of the considered elements to the main climatic oscillations covered in the Talos Dome ice core. The Talos Dome ice core, drilled from a peripheral dome in Eastern Antarctica, is 1620m long and covers more than 250k years of climate history. We prepared samples from the entire length of the ice core, with a focus on depths lower than 1450m, which have not yet been dated. In this work, the published dust record has been integrated in its deepest part, by analyzing 125 samples through Coulter Counter. The dust concentration in the Talos Dome ice core exhibits the well known correlation with the oxygen isotopic ratio; levels are low during interglacials and rise during the last glacial stage. South America, and Patagonia in particular, has been recognized as the major dust supplier for Eastern Antarctica during glacial stages, while during interglacials other sources become more relevant. When the hemispheric dust influx from remote sources is dampened, local sources become more important to the total dust budget at Talos Dome. We calculated two indexes, Fine Particle Percentage (FPP) and Coarse Local Particle Percentage (CLPP), to assess the relative contribution of fine and coarse particles to the total dust record. The deeper section of the ice core displayed some anomalies: FPP drops to very low values, while CLPP shows a significant increase. Moreover, we found that modal values for the volume size distribution shift from 2µm to higher values at deeper depths. We interpret this as a sign of dust aggregation, which was confirmed by our observations at SEM. We found evident weathering features in grains belonging to samples from the deep section. Elemental maps of individual grains show that precipitates present in the deep part are typically composed by Fe, S and K, which is compatible with the chemical formula of jarosite, while the “residual” cores of the grains mostly consist in Si or a mixture of Si and Al. These results are compatible with the acidic weathering of basalt in a closed system. Jarosite is also known to act as a cement during weathering, explaining the difficulty we encountered in breaking the aggregates when we performed tests with an ultrasonic bath. We further investigated the elemental composition of dust in the Talos Dome ice core with INAA and ICP-SFMS measurements, covering 36 different elements with the former and 38 with the latter; 22 elements overlapped between the two techniques, allowing comparisons between the two. Our principal aim was to study the fractionation of elements between soluble and insoluble phase and how this may change at different climatic stages. We consider INAA results to reflect the composition of the insoluble fraction of dust, while filtered ICP-SFMS samples should reflect the soluble fraction, and untreated ICP-SFMS samples are considered as a benchmark for the total dissolvable concentration. We also used enrichment factors and correlation matrices to assess the crustal or non crustal origin of the considered elements. The high correlations and low enrichment factors found among insoluble elements confirm a prevalent crustal source for mineral dust. Other minor contribution to mineral dust can derive from volcanic eruptions and marine emissions, in the form of sea salt sprays and biogenic emissions, as testified by our records of mercury and selenium. The concentration of some elements, namely As, Sr, Ni and Ca, is the probable result of a mixture of different sources, both terrestrial and marine, making the signal harder to read. Rare Earth elements were measured both with INAA and ICP-SFMS. They present a typical crustal behavior, with crustal-like concentrations and low solubility. No evident differences were observed along the core, confirming that these set of elements is among the most suited to track crustal material, regardless of weathering. In our solubility study, the majority of elements exhibits a minimum in solubility during the last glacial maximum. This is the result of higher fluxes of mineral dust which were transported to the Antarctic continent from remote sources and is consistent with the crustal origin we documented for the better part of the elements considered. Many elements also showed a maximum in solubility in shallower samples, belonging to the Holocene. Despite this, no clear trends connected to depth were found, except for iron. The decrease of iron's solubility with depth further supports the chemical weathering of the deeper sections of the ice core. When comparing INAA and ICP-SFMS data, the soluble fraction typically remains constant while the total and insoluble fraction vary. Concentrations are higher during glacial stages and lower during interglacials, showing a good correlation with total dust content. Deep samples always display an intermediate behavior, which could point to a disturbance in the paleoclimatic signal of this section, leading to a homogenization of the concentration from different climatic periods. Five elements (Ag, Ni, Ta, Li, Na) were found to be completely soluble throughout the entire ice core, and three more (Rb, Cs, Bi) have shown strong solubility tendencies. We suggest this may be due to a significant bond of these elements with insoluble particles smaller than 0.2µm, which would not be retained by the filter thus altering our soluble fraction. Because these elements do not display the typical glacial/interglacial pattern of aeolian dust, we can also hypothesize a predominant local origin for these elements, and in particular the result of aeolian remobilization of local volcanic material.