Collaborative Research: An Inverse Model Study of Abrupt Climate Change

Woods Hole Oceanographic Institution, Woods Hole MA

Investigators

Abstract

Under this award the PIs will use data assimilation techniques developed in the framework of optimal control theory and Bayesian stochastic inversion, to combine the Greenland paleotemperature record and numerical models of the ocean-atmosphere system in the Atlantic (a relatively simple box model and the more complex Stocker/Wright zonally-averaged model). The PIs hope to estimate the temporal evolutions of the Atlantic MOC, Fh, and surface freshwater forcing during the marine isotope stage 3 (24.1-59.0 ka BP) that are consistent with both the Greenland record and the physics embedded in the models. The estimated evolutions, in concert with paleoceanography records from Atlantic sediments, will be used to assess the adequacy of the theory to explain rapid climate change. Paleoclimate records, in particular those from the North Atlantic area, document a series of warm phases during the last glaciation, commonly referred to as Dansgaard-Oeschger interstadials (DOI). The onsets of DOIs recorded in Greenland ice corresponded to extremely large temperature jumps (7-16oC) that occurred within a span of a few decades. The discovery of DOIs demonstrates that it is dynamically possible for the climate system to make an abrupt transition to a new climate state. Understanding the origin of DOIs is an important part of assessing the possibility for future 'surprise' transitions. A leading theory of abrupt change invokes the susceptibility of the meridional overturning circulation (MOC) and zonally- and depth-integrated heat flux (Fh) in the Atlantic, to glacial meltwater discharges from northern hemisphere ice sheets.