uni-bremen.de
In present-day climatology, data assimilation has become an important tool for quantifying the state of the climate system based on observations. However, this method is not readily available for paleodata due to their sparse spatial distribution and the uncertainties associated with proxy reconstructions, which are frequently not well constrained.
Several groups have recently started to work on the assimilation of paleodata into state-of-the-art climate models. The majority of work is directed towards the state of the large-scale ocean circulation during the last glacial maximum (LGM), since the corresponding proxy-data coverage is relatively good. The goal is to find the best estimate of the LGM ocean circulation that is dynamically consistent (within prior error estimates) with model and proxy data.
Several data-assimilation techniques exist, including variational techniques, sequential filtering and statistical methods. The advantages and disadvantages of the various methods in a paleoclimatic context are currently not clear. Particular questions exist with respect to the implementation for sparsely distributed proxy data.
The workshop aims to bring together scientist involved in data-assimilation projects using proxy data, to discuss the technical issues associated with the various methods. By bringing the various groups together at an early stage of the research, we intend to create important synergies with respect to technical implementation issues.