Climate instabilities during the last interstadial period - modelling and analyzing the sensitivity of the bipolar seesaw climate mechanisms along a global transect of sites

Project leader


Funding source

Swedish Research Council - Vetenskapsrådet (VR)

Project Details

Start date: 01/01/2016
End date: 31/12/2019
Funding: 4900000 SEK


The end of the Last Ice Age, the ‘Last Termination’, was characterized by distinct climate shifts, which are documented in a variety of paleoclimate archives and proxies. With its markedly different boundary conditions (low sea levels, massive continental ice sheets) as compared to today, this time period represents a major challenge for Climate or Earth System Models, but provides an excellent opportunity to increase our understanding of non-linear climate feedbacks and of climate system sensitivity to internal vs. external forcing. A key role in explaining Last Termination climate variability is attributed to episodes of high freshwater input from melting N Hemisphere ice sheets impeding N Atlantic deep water formation, thus reducing northward heat transport. This triggered two major cold intervals in the N Hemisphere and corresponding warm intervals in the S Hemisphere and also impacted climate on a global scale. While the effect of this bipolar seesaw mechanism is prominent during longer time intervals, the underlying processes, which led to short-term climate instabilities, are still not known. To study these climate instabilities, we combine high-resolution proxy data sets with transient climate model simulations using the Community Earth System Model (CESM 1.0.5). The model has been implemented and tested in a parallel project with realistic sea levels, topographies and ice sheet extent/height. Our focus is on the time interval 14,700-12,900 years ago, a period with several short climate shifts in the N Hemisphere. These are well documented as cold and dry phases in three Swedish lake sediment records. Two lake sediment sequences located just north of the Subtropical Front in the central S Atlantic; four S American sites, situated in the centre of the southern westerlies belt; a high-resolution dated lake sediment record from S Thailand in the equatorial Indian Ocean; and one sequence from the Southern Indian Ocean, in the centre of the westerlies, also show distinct hydroclimatic variability during this time span. These eleven sites form a transect encompassing the N Atlantic region with its melting ice sheets, the S Atlantic with its zonal circulation and bipolar seesaw response, and the Indian Ocean with its variable monsoon climate and its zonal circulation in the south. The establishment of high-resolution radiocarbon chronologies for sites with insufficient age control and the use of Bayesian age modelling will allow detailed comparisons of the climate signals recorded in all sites and detecting possible leads/lags in response to short-term climate shifts. Our transient model focuses on the processes that ensue at transitions from cold to warm or warm to cold climate states, on the role, amount and location of melt water to initiate a climate shift, and on the global propagation of the climatic signals. Our working hypothesis is that northern melt water pulses also triggered short climate oscillations in the north by decreasing surface water salinity, which hampered the thermohaline circulation. As a result, sea ice expanded and pushed the Polar Front south and as an effect the Intertropical Convergence Zone (ITCZ) moved towards the warmer hemisphere. This led to rapid shifts of the S Hemisphere circulation belts (Subtropical Front, southern Westerlies, southern Polar Front) and impacted regions in the S Atlantic through changes in precipitation, temperature and wind strength. We hypothesize that the climate signal rapidly spread south, but that the response could have been delayed in the Indian Ocean region due to the complicated patterns of the ITCZ. Climate simulations will allow testing this hypothesis and will further our understanding of climate system sensitivity to short-lived climate oscillations, especially in a world whose climate was strongly influenced by changes in the cryosphere, a fact even valid for today’s world. The paleo-data vs. climate model simulation approach will thus be of great importance for understanding processes in a world with rapidly melting ice sheets.

Last updated on 2017-28-07 at 08:36