Call for applications for a thesis : Coherent dating of deep polar ice cores and implications for climatic mechanisms

Supervisors :
Frédéric Parrenin (DR2 CNRS, IGE, Grenoble), frederic.parrenin univ-grenoble-alpes.fr , 04.76.82.42.65.
Amaëlle Landais (DR2 CNRS, LSCE, Gif sur Yvette), amaelle.landais lsce.ipsl.fr, 01.69.08.46.72

Keywords :
glaciology, paleoclimates, dating, isotopes, greenhouse gases, bayesian methods

Subject :
Deep ice cores in Antarctica and Greenland make it possible to reconstruct key climatic parameters of the last hundreds of thousands of years, such as the temperature and precipitation over the polar ice caps, the impurity content of the atmosphere and the atmospheric composition (in particular the content of greenhouse gases such as CO2, CH4 and N2O). In order to accurately determine causal relationships in the functioning of the climate system, a precise chronology of events recorded in the ice cores must be constructed. This is a complex task because there are a multitude of dating methods, all of which have different advantages and fields of application. The counting of annual layers provides an accurate assessment of the duration of events but it is only applicable to recent periods and to sites with high snow accumulation rates. Moreover the absolute ages provided by this method quickly become imprecise. The synchronization of various records to variations in astronomical parameters of the Earth (which are calculated very precisely by the equations of celestial mechanics) has the advantage of being applicable to almost all periods but its accuracy is no better than a few thousand years. Synchronization with other dated paleoenvironmental records, such as volcanic lava and cave speleothems, provides precise but disparate control points for absolute ages. Modeling of snow accumulation, ice densification (with air trapping) and ice flow provides continuous dating of ice cores for both trapped air bubbles and the ice matrix, but this dating is imprecise when considering ancient periods. Finally, the synchronization of the different ice cores using common markers makes it possible to study the phase relationships between parameters measured in different cores but does not make it possible to determine absolute ages. Thus, current ice core dating, such as the AICC2012 dating (Bazin et al., CP, 2013 ; Veres et al., CP, 2013) is obtained by combining different sources of chronological information through a probabilistic model (Lemieux-Dudon et al., QSR, 2010 ; Parrenin et al., GMD, 2015).

This thesis will consist in :
1) Improving the different chronological information available on the different boreholes. A modeling of the sedimentation process for major ice cores will be updated. At LSCE, new elemental (O2/N2 ratio) and isotopic (d15N of N2 and d18O of O2) composition data in air trapped in the ice recently obtained will provide orbital constraints, notably for the Dome C deep core in East Antarctica over the last 800,000 years, as well as constraints on the relative chronology of ice and air over different boreholes.
2) Combining the different chronological information existing on the different ice cores using the paleochrono probabilistic model (Parrenin et al., GMD, 2015). New functionalities of this model could be explored. Compared to the current dating AICC2012, a maximum of new drillings will be included (notably the American WAIS drilling and the Japanese Dome F drilling).
3) Studying the resulting phase shifts between key records on glacial - interglacial transitions : phase relationship between CO2 and Antarctic temperature increases, between temperature changes in Greenland and Antarctica, and between variations in astronomical parameters and climate variations.

This thesis involves extensive collaborations with the international ice core community to provide a new federative reference dating. The work will be carried out under the supervision of Amaëlle Landais at LSCE and Frédéric Parrenin at IGE (Grenoble) and thus involves travel between the two sites. Finally, this thesis is part of the European project ERC ICORDA with strong links with the H2020 BE-OIC project.

Contact : frederic.parrenin univ-grenoble-alpes.fr