To project overview

The global anatomy of abrupt climate changes

Annual Review Article 2020

The ice age climate in the North Atlantic area was characterised by repeated abrupt climate changes. Although we will most likely not experience equivalent climate changes in the current interglacial period, it is important to understand the underlying physical mechanisms that may again come into play in our future climate.

By Sune Olander Rasmussen, associate professor, PhD, the Niels Bohr Institute, Copenhagen University

The ChronoClimate project examines the interaction between the oceans, the atmosphere and the cryosphere (particularly the ice caps and sea ice) in order to map how climate changes propagate geographically and between the components of the climate system. 

Ice cores from Greenland show around 30 abrupt climate changes – known as Dansgaard-Oeschger events – during the last ice age. 

Time and again, the climate of the North Atlantic area changed very abruptly in the course of a few decades, presumably as a result of changes in energy transport by ocean currents and the atmosphere from low to high northerly latitudes. 

These changes had a significant impact on Greenland, with changes to atmospheric circulation patterns and rises in temperature of 8-17°C, often within a human lifespan or less. This was not an example of global warming, but rather of a redistribution of energy between north and south, and between ocean and atmosphere.

Using a combination of ice-core data and climate models, we in the ChronoClimate project were the first to identify the mechanisms behind Antarctica’s delayed and opposite reaction

Greenland calls, Antarctica responds

Warming Antarctica’s ice-cold heart

Climate changes can have vastly different regional manifestations although they have a common cause. Variations in the strength of the North Atlantic ocean currents affect Greenland, for example, rapidly and significantly by changing the temperature and sea-ice cover in the surrounding ocean, while other regions that lack the same direct contact with the affected ocean react more slowly. 

Antarctica is particularly difficult to impact, as the continent is effectively cut off from the rest of the world by a belt of winds and ocean currents circling around it, known as the Antarctic Circumpolar Current (ACC). 

Until now, it has been a mystery how the rise in temperature propagates across the ACC and heats Antarctica, as ice-core data show is the case. As the North Atlantic ocean currents weaken, and temperatures fall rapidly in Greenland, our new computer models show how the heat builds up in the interior of the oceans.

Thanks to ice cores from both Greenland and Antarctica, carefully synchronised with the aid of volcanic layers and other traces that, under favourable conditions, leave synchronous impressions on both ice caps, we know that when Greenland experiences abrupt climate changes, the first signs of changes in the atmospheric circulation around Antarctica occur simultaneously. By ”simultaneously” we mean within approximately 50 years. 

A huge amount of work to reduce the timing uncertainties lies behind this conclusion. However, it takes 100-200 years for the change in temperature, which is extremely abrupt in the North Atlantic area, to manifest as a change in Antarctic temperature. 

And when they do come, the changes are much more gradual. When Greenland moves from a cold to a warm phase, Antarctica responds with a gradual cooling a century or two later and, vice versa, Antarctica starts to warm up 100-200 years after Greenland has returned to the cold glacial period state. 

Using a combination of ice-core data and climate models, we in the ChronoClimate project were the first to identify the mechanisms behind Antarctica’s delayed and opposite reaction.

Sune Oleander Rasmussen from the Niels Bohr Institute, University of Copenhagen shows a Greenlandic ice core. Photo: Ola Jakub Joensen

Timing is crucial

One of the premises for the new results is the long, painstaking dating work carried out over several decades by researchers at the former Centre for Ice and Climate – now part of the section for the Physics of Ice, Climate and Earth at the Niels Bohr Institute. 

Highly accurate dating and synchronisation of different data series is an indispensable ingredient when it comes to reconstructing how past climate changes propagated from north to south and from atmosphere to ocean. 

The iconic main building ’main dome’ built for the ice core project NEEM in North West Greenland, where it housed up to 35 researchers and camp staff during the summers of 2007-2011.
Photo: Dorthe Dahl-Jensen

Danish researchers have led the way here, right back from when Willi Dansgaard took the first tentative steps towards ice-core-based climate research in Copenhagen more than 50 years ago. 

Another element is comprehensive computer models capable of reproducing climate fluctuations that are similar to the Dansgaard-Oescgher events of the past. This makes it possible to delve into the details and, for example, calculate the energy transport in the oceans and how changes in the sea-ice extent change the heat transfer between ocean and atmosphere.

Highly accurate dating and synchronisation of different data series is an indispensable ingredient when it comes to reconstructing how past climate changes propagated from north to south and from atmosphere to ocean

Fig. 1 The map of model results shows how the North Atlantic cools by 4°C or more and the surface air temperatures above the Southern Ocean both north and south of the ACC rise.

From the big picture to the fine details

Having gained a better understanding of the communication at global scale between the climates of Greenland and Antarctica, we are now focusing on the fine details and have thrown ourselves into a large, new dataset that shows the year-onyear development of abrupt climate changes in a wealth of data series from two parallel Greenland ice cores. 

Although the 30 or so Dansgaard-Oeschger events are similar, differences emerge in the detail, and we hope to be able to coax more information about the underlying mechanisms out of the dataset. 

At the same time, we are working to refine the dating of the most recent ice core so that, going forward, we continue to have the best possible basis for interpreting data from the ice cores and are able to compare this with data from other components of the climate system. 

This is the broad approach the ChronoClimate project is taking to gain a better understanding of the climate’s fundamental mechanisms. 

Fig. 2 This figure shows the ocean temperature along a cross section between the Atlantic Ocean 200-300 years after the abrupt cooling in Greenland. As expected, we can see that the ACC effectively blocks the heat in the South Atlantic from propagating further southwards through the ocean.

The aim is twofold: to achieve a better understanding of the climate in terms of the basic science and to operationalise this knowledge to improve the climate models, which are one of the most important tools we have to prepare us for both abrupt and gradual climate changes in the future. 

We can use the model to analyse the physical mechanisms that unfold and hence study how the warming propagates across the ACC zone. 

The key proves to be that the sea-ice extent changes, and that the resulting change in the area covered by sea ice changes both how much energy is reflected from the sun and the energy exchange between ocean and atmosphere. 

The full story can be read in the article: J. Pedro et al., Beyond the bipolar seesaw: Toward a process understanding of interhemispheric coupling, Quaternary Science Reviews 192, pp 27-46, 2018.