Assimilating Continental Mean Temperatures to Reconstruct the Climate of the Late Pre-industrial Period

Abstract:

Climate reconstructions for the pre-instrumental period are either based on climate proxy data or on numerical simulations. However, both approaches are associated with substantial uncertainties. In principle, the best state estimates can be expected by employing data assimilation (DA) techniques, which systematically combine the empirical information from proxy data with the representation of the processes that govern the climate system given by climate models. Here, an ensemble-based DA method is performed to reconstruct the climate of the period 1750-1850 AD, and the performance is evaluated on large and small spatial scales. We use a low-resolution version of the Max Planck Institute for Meteorology's MPI-ESM model and assimilate the PAGES 2K continental mean temperature reconstructions for the Northern Hemisphere (NH). The ensembles are generated sequentially for sub-periods based on the analysis of previous sub-periods.

The assimilation has good skill for large-scale temperatures, but there is no agreement between the DA analysis and proxy-based reconstructions for small-scale temperature variability within Europe or with reconstructions for the North Atlantic Oscillation (NAO) index. To explain the lack of added value in the DA, a Maximum Covariance Analysis (MCA) of links between NH temperature and sea level pressure (SLP) is performed on a control run with MPI-ESM. The analysis shows that the Northern Annular Mode (NAM) is the pattern that is most closely linked to the NH continental temperatures. This link is reproduced in the assimilation run for winter, spring and annual means, providing potential for constraining the NAM/NAO phase and in turn, for regional temperature variability in the DA. It is shown that the lack of actual skill for these aspects is likely due to the fact that either the link might be too weak, as the continental mean temperatures are not the best predictors for the NAM (they are reconstructed by many proxies in locations not linked to the NAM), or that the PAGES 2K temperatures might be different from reality, as they include noise. The NAM phase in the DA analysis might therefore be substantially different from reality, leading to unrealistic representation of small-scale temperature variability.