Understanding climate
for the benefit of society

Nour-Eddine Omrani writes about his recent study of oscillations in the North Atlantic. Photo: Ellen Viste

Variations in the North Atlantic influence global temperatures

Variations on various time scales can lead to apparent intensifications or pauses in global warming. Nour-Eddine Omrani writes about his recent study on the mechanisms behind oscillations in the North Atlantic.


Written by Nour-Eddine Omrani, researcher at the Bjerknes Centre for Climate Research and the Geophysical Institute at the University of Bergen

On top of the long-term anthropogenic climate change, the Northern Hemisphere climate has experienced various coherent wintertime multidecadal climate variability in stratosphere, troposphere, ocean, Arctic sea ice and surface temperature.

Depending on the sign of its change, the multidecadal variability can add to or counteract the long-term climate change for a limited time period, leading to an apparent intensification or hiatus – pause – in the overall long-term climate trends.

Such hiatus and intensifications can last over a few decades and were registered in the historical record of the wintertime global temperature and North Atlantic Arctic sea-ice in the 1950s–1970s and 1980s–2000s respectively. For the near future climate projection, the multidecadal variability can therefore be very crucial.

Several efforts have been made to understand different mechanisms of North Atlantic variability separately. But a general framework linking all the mechanisms is still missing. In a new article we have proposed a general physical framework that explain the coherent multidecadal fluctuations of the stratosphere, troposphere and ocean, including Arctic sea ice.

Observations show influence of North Atlantic Oscillation and climate change
Temperatures increase with climate change, but are at the same time affected by natural oscillations in the climate system. The example demonstrates this for observed and predicted North Atlantic sea surface temperatures. The final temperature is a result of climate change and the phase of the oscillation, together with variations on shorter time scales. Ill.: Nour-Eddine Omrani / Ellen Viste

Coupled stratosphere / troposphere / Atlantic Oscillation

Since the observational record is short, and the overall distribution of the oceanic data is missing, we must refer to model simulations for a deeper understanding of the long-term climate fluctuations. In our study we used a millennial (more than 1000 years) model simulation forced by anthropogenic – such as greenhouse gases and manmade aerosol emissions – and natural forcings – such as solar forcing and volcanos – as well as appropriate statistical and dynamical analysis.

Our results show that the coherent multidecadal changes in different climate components in the Northern Hemisphere can be understood as a damped quasiperiodic oscillation that couple the stratosphere, troposphere and Atlantic Ocean including Arctic sea ice.

The concept of quasiperiodic oscillations has been widely used to describe and understand quasiperiodic climate variability such as ENSO dynamics. It requires positive feedback that intensifies the magnitude of the oscillation and subsequent negative feedback that weakens it and reverse the phase of the oscillation.

In our proposed oscillation, we have showed that the positive feedback responsible for the multidecadal intensification of the North Atlantic warming and the resulting North Atlantic Arctic sea ice melting, is maintained mainly by the strengthening of atmospheric wind, described by an index called North Atlantic Oscillation (NAO) index.

A signal propagates from the stratosphere into the troposphere during the wintertime and intensifies the deep Atlantic meridional overturning circulation and wind-driven North Atlantic surface circulation. The net results strengthen oceanic heat transport from the warm tropical into the cold-high latitude Atlantic areas. This causes an intensification of the large-scale Atlantic warming and Arctic sea ice melting.

The large-scale Atlantic warming acts to weaken the westerly wind forcing. This weakening of the westerlies acts as subsequent negative feedback that weaken the ocean circulation and reverse the warming and ice melting into the opposite phase, that is cooling and southward sea ice expansions.

The weakening of the westerlies in response to North Atlantic warming and the response to the Atlantic circulation to the NAO has been demonstrated in previous studies using well designed model experiments.

Importance for the near future climate

Because of its quasi-periodicity, the coupled multidecadal oscillations has high potential of predictability. We showed that a simple multi-regression model can reproduce the observed multidecadal variability quite well, if the coupling between the ocean, atmosphere and sea ice is considered.

However, the pronounced long-term anthropogenic climate cannot be predicted from such a coupled oscillation. The overall near future decadal-to-multidecadal climate projection depends also on the long-term anthropogenic climate change, which is very pronounced for the surface temperature and Arctic sea-ice.

If we assumed that the long-term anthropogenic trends will not change, the evolution of the present and projected near future climate resemble strongly the condition seen in the 1950s–1970s. Similar to this period, the present and projected climate are associated with a weakening of the westerlies. According to the mechanism proposed, this should weaken the ocean circulation and lead to north Atlantic cooling and Arctic sea-ice extension, which can dampen the effect of anthropogenic long-term trends and leads to hiatus in the wintertime surface temperature and North Atlantic Arctic sea ice.

Present hiatus can just offer time to work out solutions for the next accelerated warming

Compared to the 1950s–1970s, the climate change makes the present and projected climate much warmer. The global cooling associated with multidecadal oscillation can dampen temporarily the observed anthropogenic warming and sea-ice melting, but it cannot bring the climatic conditions of 1950s–1970s back.

In particular, the cooling phase of the multidecadal fluctuation can just retard temporarily the subsequent acceleration of the global warming and sea-ice melting associated with the warm phase of the Atlantic multidecadal oscillation.

The global warming is in turn making the effect of multidecadal warming and sea ice melting much extreme and the effect of cooling and sea-ice extension less intense, in such way that the net extreme cooling and sea ice extension seen 1950s–1970s cannot not be reached anymore. In particular, the next multidecadal warming will start from much higher and lead to unpreceded warming, sea ice meeting and associated extremes.

The present and predicted hiatus can just offer time for the decision makers to work out technical, political and economic solutions to the next accelerated global warming and sea-ice melting.


Omrani, NE., Keenlyside, N., Matthes, K., Boljka, L., Zanchettin, D., Jungclaus, J.H., Lubis, S.SH.: Coupled stratosphere-troposphere-Atlantic multidecadal oscillation and its importance for near-future climate projection. npj Clim Atmos Sci 5, 59 (2022). https://doi.org/10.1038/s41612-022-00275-1