The role of high-latitude oceans in transient climate change

Abstract

In this thesis we explore the role of the large-scale ocean circulation in the North Atlantic and the Southern Ocean (SO) in setting the regional and globally averaged sea surface temperature (SST) response to atmospheric forcing. We focus on the impact of anthropogenic greenhouse gases (AGHGs) and the Antarctic ozone hole and use output from general circulation models (GCMs) to estimate the corresponding climate response functions (CRFs). We show that the strength and the vertical extent of the time-mean Atlantic Meridional Overturning Circulation (AMOC) set the effective heat capacity of the World Ocean and affect the global CRF to greenhouse gas (GHG) forcing. A large fraction of the anomalous surface heat uptake induced by GHGs takes place over the North Atlantic. However, the SO also plays a significant role in removing excess heat from the atmosphere. Compared to the rest of the World Ocean, the SO warms at a much slower rate under GHG forcing. In this region the background Meridional Overturning Circulation (MOC) upwells unmodified deep water masses to the surface where they take up atmospheric heat. The modified water masses are then advected northward and subducted in the mid-latitudes. This geographical imprint of the MOC is reflected in the regional CRFs to GHGs, as seen in idealized numerical experiments with GCMs. However, GHGs are not the only major source of anthropogenic forcing on the SO. Stratospheric ozone depletion around Antarctica gives rise to an atmospheric pattern similar to the positive phase of the Southern Annular Mode (SAM): a strengthening and a southward shift of the westerlies. This poleward intensification of the winds changes the ocean circulation and gives rise to an SST response. We examine the SO CRF to a SAM pattern that arises either in the form of natural variability in unforced control experiments or as a result of imposed ozone perturbations. We analyze the SO SST response to SAM on multiple timescales and across an ensemble of GCMs from the Climate Modeling Intercomparison Project phase 5 (CMIP5). We show that the corresponding SO CRF is governed by the anomalous wind-driven MOC redistributing the background heat reservoir. The intermodel diversity in the fast and slow SST responses to SAM is partly explained by differences in the climatological thermal stratification across the ensemble of GCMs. Furthermore, we demonstrate that the sea ice response to SAM in models is very well correlated with the geographic pattern of the SST anomalies. Finally, we convolve our estimated CRFs with timeseries of historical forcing to recover the SO SST trends in numerical simulations and in observations. We contrast the multidecadal SO cooling trends against the SST warming rate in the Northern Hemisphere high latitudes. Our results imply that the recent cooling in the SO may be explained by the Antarctic ozone hole projecting on a positive SAM trend. We furthermore attempt to understand why CMIP5 models have been unable to reproduce the observed negative SST trends in the SO and instead predict regional warming. Many GCM simulations underestimate the historical SAM evolution. Another subset of CMIP5 models have biases in their climatological SO stratification, which affects their SO CRFs to SAM. The successful application of the CRF framework in the context of observed and simulated SST trends validates the results of our analysis. We are thus able to interpret the CRFs as inherent characteristics of the climate system and elucidate the importance of the high latitude oceans in transient climate change.