Abstract

Large climate changes have been observed in West Antarctica for several decades, including mass loss from glaciers due to intrusions of warm oceanic mass into the ice shelf cavities. The satellite era is too short to understand the natural and forced drivers of these changes, in particular with the strong internal variability that prevails there. To extend historical changes beyond the satellite period, we need to rely on indirect climate records from natural archives that record past environmental conditions, namely paleoclimate records. In polar regions, ice cores provide the most common records. However, these records only provide local information for specific variables. In order to provide a spatially complete multi-field reconstruction, paleoclimate data assimilation methods have been emerging over the past years. Based on a probabilistic approach, data assimilation optimally combines local information from paleoclimate records with the physics of an Earth System Model. It thus ensures that the resulting reconstruction is dynamically consistent with the climate physics given by the Earth System Model. In this talk, we will see how relevant paleoclimate data assimilation is to provide historical changes beyond the satellite era, and improve our knowledge about the dynamical relationships. In particular, a case study on the contribution of tropical variability to 20th century atmospheric changes in the Amundsen Sea sector will be presented, along with future directions regarding historical ice shelf melting in West Antarctica.