Given that I’ve written a number of posts about the so-called “pause”, I thought I would mention a recent paper by Lewandowsky, Risbey & Oreskes called On the definition and identifiability of the alleged “hiatus” in global warming. I don’t want to say too much about it as there are various articles about it already.All I really wanted to do is mention the figure on the right. The top panel is the linear trend for a particular vantage year (end year looking back in time), and the number of years included. The colour is the actual trend, and the dots are those combinations of vantage years and years included for which the trend is statistically significantly different from 0 (no trend). The bottom panel is essentially just illustrating this showing which vantage years and years included have -values less than 0.05 (the standard test of statistical significance). What it shows is that if you include less than 17 years, trends are typically not statistically significant, while if you include 17 years or more, they are almost always significant – for the time period considered, at least.
So, this figure essentially shows what’s been pretty obvious to many; if you define the “pause” as the trend being statistically consistent with 0 (no trend) then if you consider periods that are too short, you will almost always find a “pause”. Given the natural variability in the data, you need to consider periods of – typically – more than 17 years if you want to determine if there is a non-zero trend, or not. This is essentially what Richard Telford pointed out when he illustrated that Ross McKitrick’s paper was essentially describing a recipe for a hiatus.
So, I think the recent Lewandowsky, Risbey & Oreskes paper is simply illustrating nicely what has been pretty obvious to many people for a while now. This doesn’t mean that the recent period isn’t interesting from a model/observation perspective and that interesting and unexpected things didn’t happen. It’s just clear that there hasn’t really been any kind of “pause”.