Personal tools

You are here: Home / Science and society / Science for all / Just published! / Winds amplified abrupt millennial-scale climate changes during ice age

Winds amplified abrupt millennial-scale climate changes during ice age

How to explain the abrupt climate oscillations of the late ice age? Modelling shows that, among other components of the coupled ocean-ice-atmosphere system, the wind pattern played a major role through its action on the southward drift of sea ice.


Studying past climatic events helps understanding the interactions between atmosphere, ocean and continents. The Earth already experienced abrupt climate charges, and those who occurred during the last ice age are of high interest to climatologists. These so-called Dansgaard–Oeschger (DO) events caused a rapid warming (up to 16°C within a few decades in Greenland) followed by a progressive cooling; between 30,000 and 60,000 years before present, they occurred every 1,000 to 2,000 years.


Climate evolution in Greenland for the last 100,000 years as shown by the isotope ratio δ18O, a good indicator of air temperature. DO events (numbers) are absent from the most recent interglacial period (Holocene).

It is generally accepted that the Atlantic Meridional Overturning Circulation (MOC) is somehow involved. This large-scale circulation is caused by seawater density differences and by winds; it can be described as a surface northward flow (Gulf Stream and north Atlantic drift) and a returning deep southward flow. Its variations cannot explain alone the large temperature variations observed. In fact their effect on high latitude climate is strongly amplified by sea-ice, and particularly by the feed-back effect of its reflecting power (albedo). If the ocean transports more heat to the North in the Atlantic Ocean, arctic sea ice melts. Surface water in the Arctic then absorbs more solar radiations and warms up, thereby amplifying the initial ice melting.

While the ice-albedo feedback is well understood, much less is known about the influence of atmospheric feedbacks through which winds could alter sea ice distribution and thus influence the response of air temperature to a change in the strength of the Atlantic MOC.

This study consisted in a simulation of the ocean-ice-atmosphere coupled system to improve the understanding of the influence atmospheric circulation changes could have had on its stability and on the abrupt millennial variability observed in glacial climates. The scientists compared the results of numerical simulations where winds were allowed to evolve interaction with the climate and where they followed a constant seasonal cycle independent of the climatic and oceanic conditions. They used an ocean-ice-atmosphere coupled model whose complexity was intermediate between those of very simple conceptual models and of climate prediction models.

The model geometry is that of an idealized ocean basin whose dimensions are similar to those of the Atlantic. A channel is periodically opened in southern subpolar latitudes to represent the Antarctic Circumpolar Current. Space is divided by a 2° horizontal grid and by a vertical grid of 19 layers 50 to 450 meters thick. The idealization of both geometry and physics proved necessary to cope with the long time-scales of the simulations. Using a general circulation coupled model of the kind used for 21st century climate predictions would indeed have required computer resources much beyond what presently exists. This approach is also justified by the fact that parameter sensitivity experiments are necessary to establish causal relations between different physical processes in a complex coupled system.


Simplified model geometry. Arrows show the main features of the circulation generated in the basin

The main experiment consisted in cooling the climatic system over a very long period (100,000 years), with or without considering the interaction between wind and climate; other simulations were performed to refine the analysis. A first result is that, at intermediate climates, MOC looses sits stability and is affected by very strong millennial-scale oscillations whose structure in space and time is similar to observed DO events. This agrees with analysis of paleoclimate archives from Greenland ice cores that have suggested that abrupt millennial-scale climate transitions during the last ice age developed only for cooling rates associated with intermediate ice-sheet sizes. The model also shows that the interaction between wind and climate increases the amplitude of the oscillations and shifts them towards colder climates.

MOC intensity (in Sverdrups, 1 Sv = 106 m3 s-1) during the simulated progressive cooling, with (red) or without (blue) considering the interaction between wind and climate

What mechanisms are responsible for the larger amplitude of abrupt millennial-scale variability? Whereas the ice-albedo positive feedback is simple, the processes involving atmosphere, ocean and sea ice interact in a highly complex way within a chain of cause-to-effect relations which are generally difficult to establish. In the present case however, the authors could identify the causal relations at play in the ocean-ice-atmosphere coupled system. The effect of interaction between wind and climate combines with the ice-albedo feedback to increase the amplitude of the oscillations; a new positive feedback loop emerges within the coupled system. This is through its effect on sea ice export that the interaction between wind and climate is able to amplify the atmospheric response to a change in ocean circulation.

Mechanisms of positive feedback between atmosphere (beige), ice (green) and ocean (mauve), and involving either albedo (left) or the interaction between wind and climate (right). A parameter may influence another one positively (+) or negatively(-)


The second effect of the interaction between wind and climate is to shift the domain of existence of abrupt millennial variability towards colder climates. To understand this, it is necessary to consider not the variability of the coupled system but its stability. Atlantic MOC stability stems from the northward oceanic heat transport: if the circulation weakens, the north-south temperature gradient increases and drives it back on track. Weakening of this regulator causes the instability of the circulation to occur and (under certain conditions) abrupt millennial-scale climate oscillations to emerge. The numerical simulations show that or the interaction between wind and climate strengthens this negative feedback between temperature and circulation: the oscillation window thus shifts towards colder climates because of the higher stability of circulation facing climate "coolness".

This study shows that the interaction between wind and climate can significantly alter the properties of abrupt millennial-scale climate variability as observed during the last glacial period in Greenland ice cores. The robustness of this result will have to be evaluated in coupled models of increasing complexity and of course checked against the future reconstructions of glacial climates.


The paper

O. Arzel et M. H. England, 2012. Wind-stress feedback amplification of abrupt millennial-scale climate changes. Climate Dynamics publié en ligne, doi 10.1007/s00382-012-1288-1
See the first page

The authors

This work was conducted in collaboration by researchers of the Laboratoire de physique des océans (LPO) of IUEM and of the Climate Change Research Centre (University of New South Wales, Sydney, Australia).


The journal

Issued by the international editor Springer, Climate Dynamics covers all aspects of the dynamics of the global climate system. The journal publishes paleoclimatic, diagnostic, analytical and numerical modeling research on the structure and behavior of the atmosphere, oceans, cryosphere, biomass and land surface as interacting components of the dynamics of global climate.



Authors : browse IUEM staff directory
Communication and outreach service:

Back to publications list

Picture of the month