There's a new research paper out that will probably set the blogosphere ablaze in coming days, especially the fake skeptic part of it. Why? Because it espouses a theory that explains the slowdown in the rate of temperature rise of the past decade and a half (depending on your starting point).
According to the paper by Cowtan and Way, called Coverage bias in the HadCRUT4 temperature series and its impact on recent temperature trends, the slowdown is actually smaller, but it's not apparent because temperatures aren't recorded properly in a couple of areas around the globe, the most important being, of course, the Arctic.
This doesn't really come as a surprise, as we have seen the massive temperature spikes since 2007, especially during freeze-up when the Arctic Ocean releases all the heat accumulated during the melting season to be able to freeze up. Apparently there are a lot differences in the way these temperatures are measured or inferred by the various global temperature data sets.
You can read more about this in a blog post by Stefan Rahmstorf on RealClimate (where I took the graph on the upper right from), but I'd also like to draw attention to a blog post from last April that also looks at some discrepancies between the GISS LOTI and NCEP/NCAR data sets (Cowtan and Way focus on HadCRUT4). The analysis is done by - who else - Chris Reynolds on his Dosbat blog. Here's an excerpt and a graph:
As with the previous post on Arctic Amplification it is apparent that summer warming is negligible, this is due to energy going into melting ice, not raising temperatures. However the latent heat of energy melting ice, and ocean warming, is released in September to November, with later warming possibly due to thinner ice and ice growth.
These temperature anomalies will only be picked up by a handful of meteorological stations on islands in the Arctic, therefore will not be being strongly represented in GISS LOTI. It seems to me that rather than concluding that NCEP/NCAR is wrongly showing warming of a pattern shown by other reanalysis datasets, and backed up by mechanistic understanding of the changes in sea ice, it is more likely that GISS LOTI is missing a significant element of warming that could be as large as an extra 0.5degC in the annual average.
Read all of it here.
So it could be that the 'hiatus' in global temperature rise really isn't one. What will happen when the Sun becomes more active again, ENSO isn't dominated by La Niña, China and India start filtering aerosols out of their emissions? The prospect worries me, to be frank, as I don't think temperatures in the Arctic will drop any time soon, despite the cold summer we just witnessed.
Here's a very nice graph produced by Jos Hagelaars from the Our Changing Climate blog that shows that the slowdown (or 'hiatus' or 'pause') in global atmospheric temperature increase hasn't disappeared, it has just become smaller because of Cowtan and Way's inclusion of the Arctic and other data-less regions (update: due to a calculation error the graph has been updated):
The fact that the upward trend isn't going up as fast in the period since the 70's still probably has to do with ENSO being dominated by La Niñas, an inactive Sun, Asian aerosols, etc. There's a good chance that even if all these things don't flip back and keep dampening atmospheric temps, the rate of temperature increase will gradually go back to what it was. And the Arctic will probably continue to get warmer much faster than the rest of the globe.
Commenter Dominik Lenné also has an interesting back-of-the-spreadsheet analysis on his blog showing how some of the 'missing' heat (I don't miss it at all) might have gone into melting Arctic sea ice. Rob Dekker, get over there and have a look (if you have the time): Global Warming 'hiatus': it might be the ice