Towards a high-quality interannual to millennial ice-core based temperature reconstruction for East-Antarctica

Abstract

Instrumental observations of climate cover only the past ~200 years. Our knowledge about historical natural climate variability beyond this time is exclusively based on climate proxies, stored within natural archives. The polar ice sheets of Greenland and Antarctica have formed by continuous accumulation of snow over > hundred-thousand of years, recording local climate conditions such as the air temperature by means of the isotopic composition of the water molecules composing the snow and ice. However, in regions with very little snow accumulation, multiple processes hamper the interpretation of isotopic data in terms of climate variations and temperature reconstructions from ice-cores from the East Antarctic Plateau are still highly uncertain on decadal to millennial timescales. One approach to overcome the local spatial noise in isotope records was to stack several independent vertical profiles of a specific site. In order to extend this approach into space and time, innovative developments are required to enable (1) the extensive spatial sampling of the snowpack in the field, (2) the temporal analysis providing high-quality isotope records of an increasing amount of profiles from Antarctic snow- and ice-cores. As a part of the AWI Strategyfund project COMB-i, this dissertation addresses the technical development of tools to meet both requirements and to be applicable to the East Antarctic snowpack and ice column: 1) Spatially, I developed a novel sampling technique for a fast and precise sampling of the snow, allowing to investigate the spatial variability of the isotopic composition of the snowpack. We applied this new tool in a study to regionally investigate the relationship between local topography, accumulation rate and isotopic composition. We identify a new mechanism leading to a non-climatic variability in isotopic records retrieved from ice cores, which may impact a large area of the East Antarctic Plateau. Such a finding will improve the interpretation of existing ice-core records, affect the choice of future ice-core drilling site, and contribute to develop statistical models for non-climatic variations of water isotope record. 2) Temporally, I enabled the analysis of isotopic composition with high-precision of snowpack profiles using the Continuous Flow Analysis (CFA) technique. This will allow high-pace measurement of snow-cores, and hence increase statistics of temporal analyzes in order to ultimately strengthen the understanding of the isotopic signal formation. The CFA setup was further applied to generate high-quality paired records of stable water isotopes and chemical impurities along the top 70-meters of an ice-core from East Antarctica. The data produced will foster the newly developed method within the COMB-i project to utilize impurity records to improve the temperature reconstruction from East Antarctic ice-cores. Overall, through technical innovations and their application, this thesis contributes to push further the limit towards high-quality temperature records on the interannual to millennial time scales in East Antarctica.