Global footprints of dansgaard-oeschger oscillations in a GCM
Kenji Izumi, Edward Armstrong, Paul J. Valdes
Abstract
The mechanisms driving the Dansgaard-Oeschger (DO) cycles remain uncertain, but promising hypotheses could explain it, and a comprehensive assessment of its impact is needed. We have identified DO-like millennial-scale climate variability in a glacial simulation based on the salt oscillator hypothesis with the HadCM3B-M2.1 coupled atmosphere-ocean-vegetation model. The simulated DO-like warming shows temperature changes of 7.1 ± 2.5 °C in Greenland, comparable to DO9 and DO6, the smallest temperature changes of the observed DO events. This study focuses on the impacts of DO-like events on climate and vegetation beyond the North Atlantic. We find that the simulated warming in the Northern Hemisphere extratropics during Greenland interstadials agree with available proxy estimates. The simulated tropical hydroclimatic responses during the interstadials, such as northward propagation of the Intertropical Convergence Zone, strengthening of some Northern Hemisphere summer monsoons, and weakening of some Southern Hemisphere monsoons, are also consistent with proxy estimates. Moreover, simulated vegetation cover in the Northern Hemisphere increases during interstadials relative to stadials, while no large-scale dominant vegetation changes are consistent with observed biome changes associated with DO6. Our simulation based on the salt oscillator hypothesis can account for many observed features of DO events. However, it cannot simulate several observed climate responses in the equatorial Indian and western Pacific Oceans, including the East Asian and South African monsoon. The causes of those mismatches need to be further explored, whether they are hypothesis-dependent or model-dependent.