Climate Feedback

Human prints on the poles

Gillett.JPGHumans are at fault for warming at both the Earth’s poles: so say unique findings published in Nature Geoscience today.

With stories of dwindling sea ice and collapsing ice shelves already saturating the media, this at first may hardly sound like news. But in fact, researchers had never formally pinned the Arctic’s rapid warming on humans, because limited data were swamped by great natural variability. Down south, where warming on the Antarctic Peninsula contrasts with cooling in some other regions of the continent, the IPCC’s 2007 assessment report concluded: “Anthropogenic influence has been detected in every continent except Antarctica (which has insufficient observational coverage to make an assessment)”.

In the new study, Nathan Gillett of the University of East Anglia and colleagues found a way to squeeze clear results from those sparse data. Their method was based on state-of-the-art models of the polar climates that either incorporated anthropogenic as well as natural influences on variability, or included natural factors only. (Human influences include greenhouse gases that cause warming and a cooling effect from depletion of stratospheric ozone; natural ones are solar variation and volcanic eruptions.) This type of study, pioneered by Peter Stott and co-authors in 2000, has greatly boosted the IPCC’s confidence that humans are causing climate change globally. The new Gillett et al. study – co-authored by Stott – gave the technique an important tweak, say Andrew Monaghan and David Bromwich in an accompanying News and Views article, by focusing on model results for only those places with observational temperature records:

By this restriction, the group is able to perform an ‘apples with apples’ comparison of model simulations and polar near-surface temperature records during the twentieth century. Their analysis implies that the models can simulate trends better than previous studies had suggested.

In the records since 1900 put together by Gillett et al., the average temperature across monitoring stations has risen at both poles. And the models match these trends only when they factor in all influences, including human hands (producing the line labelled ALL in the above figure).

“We detected the human fingerprint in both the Arctic and Antarctic regions,” says Gillett. The familiar litany of impacts – balding Arctic summer sea ice, ecological and human displacements, sea level rise – are likely down to us too, the authors say. Stott adds that the results make the poles’ future all the bleaker, since “the human component isn’t going to let up anytime soon.”

meltwater.jpgOne caveat from Monaghan and Bromwich: the average trends still come from partial data, and not all of Antarctica, in particular, is heating up.

Because the authors focus only on the grid points for which there are observations, caution must be used in extrapolating the results to the entire polar regions. This is especially true for much of the vast interior of Antarctica, for which there are few observations, and for which other studies suggest that no statistically significant temperature changes exist. Indeed, the distribution of observational records along the coastal margins, and especially on the rapidly warming Antarctic Peninsula, may be biased towards a limited region of Antarctica for which warming has been more pronounced.

So the human-caused warming could be finer-grained, occurring in many of the regions studied in this paper while other places in Antarctica stay stable.

Anna Barnett

Images: Top, Figure 1 from Gillett et al. (with labels added); bottom, meltwater ponds on sea ice off Coburg Island, Canada / Sandy Briggs.


  1. Report this comment

    Bob Tisdale said:

    Anna Barnett: You list only solar and volcanic eruptions as natural influences included in the “state-of-the-art models” used by the authors of the study. Did their GCMs also include ENSO, AO, AMO?

  2. Report this comment

    Bob Tisdale said:

    Hmm. My question from 10/31 never made it past the filter. Let’s try again.

    Were solar variations and volcanic aerosols the only natural variables considered by this study?

  3. Report this comment

    Anna Barnett said:

    Bob: I asked Nathan Gillett for details on what was included. Here’s his reply:

    “Volcanic aersols and solar irradiance were the only natural influences external to the climate system that we considered in our study (these are the main ones). However, we also compared the observations against long control simulations in order to test whether the warming might be explained by random variations internal to the climate system. The blog post asks about ENSO, the AO and the AMO – these are all modes of variability internal to the climate system, which are simulated by the climate models we use. So the answer to the question is that the GCMs did include these modes, and we did test against this kind of internal climate variability.”

  4. Report this comment

    Bob Tisdale said:

    Anna: Thanks to you and Nathan Gillett for following up on my question and sorry about duplicating it above. The reason for my curiosity is that the instrument temperature record appears to indicate the impact of the 97/98 El Nino event has been underestimated in the study. The following is a graph of global, Northern Hemisphere, and Arctic (land and sea) surface temperature anomalies from December 1978 to April 2008. The source is UAH MSU. Note how the Arctic diverges from the global and Northern Hemisphere data after the 97/98 El Nino. Arctic temperatures then remain elevated due to the influence of the subsequent El Nino events that occurred in 2002/03, 2004/05, and 2006/07.

    To assure that the effect is not solely a function of the UAH MSU data, I created two additional graphs using long-term and short-term ERSST.v3 Sea Surface Temperature data. They are graphs of global, Northern Hemisphere, and Arctic SST. The long-term data covers the time period of January 1880 to April 2008 and the short-term data covers January 1978 to April 2008.

Comments are closed.