Climate science

The origin of regional Arctic warming

Observational data and modelling show that the rapid warming of the northeastern Canada and Greenland sector of the Arctic over the past three decades has been strongly driven by cooling in the tropical Pacific Ocean. See Letter p.209

Over the past 30 years, Earth has become a warmer place. One of the most striking examples of surface-temperature warming is the polar regions in the Northern Hemisphere (Fig. 1). The greater warming in the Arctic1, compared with the global mean, is associated with a reduction in sea ice2 and dynamical and radiative feedbacks3, and is widely attributed to anthropogenic climate change. But the fact that the warming is not spatially uniform raises the question of whether natural climate variability has a role in driving it and causing regional climate change. On page 209 of this issue, Ding et al.4 show that the most prominent Arctic warming has occurred in northeastern Canada and Greenland, and that cooling in the tropical Pacific Ocean forced half of the warming in these two regions. The findings indicate that a substantial part of regional Arctic climate change is therefore a result of natural climate variability.

Figure 1: Trend in annual mean surface temperature.

The graphic shows the observed change per decade of annual mean surface and near-surface temperature for the period 1979–2012, based on the ERA-interim climate data set. The most marked warming has occurred in northeastern Canada, Greenland and north Siberia. (Adapted from Extended Data Fig. 1 of ref. 4.)

Much of the interannual and decadal variability in atmospheric climate can be described by the evolution of the leading modes of climate variability, such as the North Atlantic Oscillation (NAO). The NAO consists of variations in the difference of sea-level atmospheric pressure between the Icelandic low-pressure system (Icelandic low) and the Azores high-pressure system (Azores high), and is most pronounced during boreal winter. In the positive phase of the NAO, there is a considerable difference in pressure between these two systems, with both the Icelandic low and the Azores high being intensified. In the negative phase, the two pressure zones are weakened and the difference between them is less.

The NAO is linked to changes in the intensity and location of the North Atlantic jet stream and storm track, and to large-scale temperature and precipitation variations over Europe, Greenland and North America. It is an intrinsic atmospheric phenomenon, but fluctuations in sea surface temperature (SST) can also affect it. In fact, it can be influenced by both variations in local North Atlantic SST and remote SST in the tropics5. Changes in tropical SSTs lead to changes in convection throughout the lowest portion of the atmosphere (the troposphere) at low latitudes, which in turn excite large-scale atmospheric waves called Rossby waves. These waves can propagate to mid- and high latitudes and affect the NAO.

In their study, Ding et al. demonstrate that Rossby waves and the NAO are involved in regional Arctic warming. Their finding that the Arctic warming in northeastern Canada and Greenland since 1979 is strongly driven by cooling in the tropical Pacific is supported by observational data indicating that warming in these two regions is not limited to the surface but also extends to the upper troposphere. The authors argue that it is unlikely that decadal temperature changes in the upper Arctic troposphere are locally forced by variations in surface temperature. They suggest instead that warming at the surface and in the troposphere are the result of atmospheric-circulation changes in the high troposphere, and that these changes are remotely forced. Specifically, Ding et al. show that the recent warming in northeastern Canada and Greenland is associated with a negative NAO phase driven by Rossby-wave activity caused by SST cooling in the tropical Pacific.

These results are confirmed by modelling experiments. The authors demonstrate that an atmospheric general circulation model forced by the observed tropical SST can simulate the connection between tropical Pacific SST cooling and regional Arctic tropospheric warming. However, they also show that coupled ocean–atmosphere climate models used in the fifth assessment report of the Intergovernmental Panel on Climate Change fail to reproduce the observed regional pattern of Arctic warming. Two plausible reasons for this failure are worth mentioning. First, the recent cooling in the tropical Pacific can probably be attributed to intrinsic variability of the climate system6, because it is not simulated by coupled climate simulations that incorporate observed changes in the concentration of greenhouse gases and aerosols. Second, despite continued improvements to climate models, it is still a challenge to simulate the influence of remote climatic phenomena correctly.

By linking cooling in the tropical Pacific with trends in atmospheric circulation and regional Arctic warming, Ding and colleagues highlight the complexity of processes involved in regional climate change. Even remote climatic fluctuations can have a substantial impact. Improving the representation of such teleconnections in climate models should therefore remain a high priority for climate scientists. The authors also note the importance of natural internal climate variability for present and near-future regional Arctic climate. But as greenhouse-gas concentrations are likely to increase in the future, it is only a question of time before external forcing dominates regional Arctic warming.


  1. 1

    Graversen, R. G., Mauritsen, T., Tjernström, M., Källén, E. & Svensson, G. Nature 451, 53–56 (2008).

    Google Scholar 

  2. 2

    Screen, J. A. & Simmonds, I. Nature 464, 1334–1337 (2010).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Shindell, D. & Faluvegi, G. Nature Geosci. 2, 294–300 (2009).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Ding, Q. et al. Nature 509, 209–212 (2014).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Hoerling, M. P., Hurrell, J. W. & Xu, T. Science 292, 90–92 (2001).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Kosaka, Y. & Xie, S.-P. Nature 501, 403–407 (2013).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Jürgen Bader.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bader, J. The origin of regional Arctic warming. Nature 509, 167–168 (2014).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.