ALBERT

Description: Oxygen and hydrogen isotope ratios in polar ice typically show variations over a large range of timescales. Since the isotope ratios are interpreted as a proxy for atmospheric temperatures, their variations can provide essential information about the natural climate variability and cycles. Nowadays high-resolution isotope samplings corresponding to depth intervals below or around the local accumulation of snow per year are routinely performed, and observed variations in the isotopic composition at a given site have frequently been interpreted as the reflection of the seasonal cycle in temperature and also to indicate multi-year quasi-periodic climatic cycles. However, studies from strongly different accumulation conditions in East Antarctica reported similar isotopic variability and comparable apparent cycles in isotope profiles with typical wavelengths of around 20 cm, which is inconsistent with a climatically driven origin. Here we show, based on spectral analysis, that these features do not correspond to truly or quasi-periodic cycles. In addition, the typical wavelengths increase with depth for most East Antarctic sites, which is inconsistent with the effect of burial and compression on a climatic cyclic signal. We explain these results by isotopic diffusion acting on a noise-dominated isotope signal. The firn diffusion length is rather stable across the Antarctic Plateau, leading to similar power spectral densities of the isotopic variations, and increases with depth in the near-surface firn. Since the first moments of the spectral density govern the characteristic spacing of the extrema of a time series – a fundamental relationship known as Rice’s law – the similar isotope spectra in turn imply similar average distances between the isotopic minima and maxima that get larger with increasing depth. Our results bear important implications for the interpretation of isotope records in terms of cyclical climate variability. They underline that simply counting isotopic extrema is not sufficient to detect periodicities, instead robust spectral analyses have to be applied in order to differentiate between true climate cycles and the apparent cycles created in the diffusion process. This has consequences for the dating of ice-core records, which is often based on or underpinned by counting isotopic maxima, but also for the detection and interpretation of quasi-periodic climate phenomena on longer timescales. Finally, the general implications of our findings are not restricted to ice cores but likely also apply to other paleo-climate archives, as other smoothing processes, e.g. the bioturbational smoothing of proxy records from marine sediments, might lead to similar apparent cycles.

Description: The pollution input in polar ice sheets in Greenland and Antarctica is of atmospheric aeolian origin, just as all natural non-ice impurities as well. They thus provide potential information on the evolution of the atmospheric share of pollutants in the ocean. Aerosols found in ice are transported with atmospheric circulation and wind patterns and are deposited e.g. with precipitating snow. The impurity content in this so-called meteoric ice is relatively low compared to many other natural materials such as rocks (ppb to ppm range). The reason is that most aerosols in the atmosphere have been removed by fall-out or precipitation during transport from the impurities’ sources to the remote ice sheet. Non-ice constituents in polar ice cores have been studied in the last decades mainly for reconstructions of past atmospheric aerosol concentrations, with respect to questions conceding the global climate change. The fastest and easiest analytical way is chemical analysis of the melted water from ice cores. However despite the tiny concentrations, the interactions with and effects of impurities in the solid ice influence the physical properties of the material as a whole: e.g. electric as well as dielectric response and, in particular, mechanical behaviour thus “softness” of the material seems to be strongly controlled by impurities. Smaller concentrations of impurities (up to a few ‰) do soften the material as a whole, while larger concentrations of particles harden it, depending on the type of impurities of course. The underlying processes are partly hypothesised for decades, but not yet proven or understood satisfactorily as the quest for ppb to ppm concentrations in solid matrix material is a search for a “needle in a haystack”. To improve the data basis regarding the in-situ form of incorporation and spatial distribution of impurities in ice we used micro-cryo-Raman spectroscopy to identify the location, phase and composition of micrometer-sized inclusions in natural ice samples (NEEM ice core from Greenland and EPICA-DML ice core from Antarctica). The combination of Raman results with ice-microsctructure measurements and complementary impurity data provided by the standard analytical methods (IC, CFA, and DEP) allows for a more interdisciplinary approach interconnecting ice core chemistry and ice core physics. While the samples originating from interglacial times were dominated by sulfate salts—mainly gypsum, sodium sulfate (possibly thenardite) and iron–potassium sulfate (likely jarosite)—the glacial ice contained high numbers of mineral dust particles—in particular quartz, mica, feldspar, anatase, hematite and carbonaceous particles (black carbon). We cannot confirm cumulation of impurities in the grain boundary network as reported by other studies, neither micro-particles being dragged by migrating grain boundaries nor in form of liquid veins in triple junctions. We argue that mixing of impurities on the millimeter scale and chemical reactions are facilitated by the deforming ice matrix. Refs.: doi: 10.5194/tc-11-1075-2017 doi: 10.3389/feart.2019.00020 https://www.humboldt-foundation.de/web/trilateral-jagfos-2019.html http://www.nasonline.org/programs/kavli-frontiers-of-science/past-symposia/2019-jagfos.html Invited poster.