The Pause in 2017 - Part 1 : An Introduction

I've been meaning to write a piece looking at the old pause and where it is now - based on a few posts on WUWT in the past month. But it was in danger of becoming an epic, so I've decided to turn it into an ongoing series. This will inevitably mean much of it will be out of date by the time I actually complete some of the posts - but what else is new? In this first post I'm just going to state my views on the pause.

Being Skeptical

In case it's not clear from any of my posts last year on the subject, I'm a pause skeptic. By that I genuinely mean I'm skeptical; I don't deny the possibility of a pause, I would just like to see some clear evidence for its existence.

The problem so far is that all the claims I've seen for the pause are based on short term trends, carefully selected (aka cherry-picked) to show the longest possible static period, or the slowest possible warming period. The problem here is that, as temperature records fluctuate from year to year, you will see short term fluctuations in the trend, purely by chance. The usual approach to claims such as this, is to look for statistically significant evidence; that is, you want evidence that is unlikely to have arisen purely by chance.

All the claims of a pause I've seen either ignore the question of whether their evidence is significant, or abuse the notion of significance by turning it on its head and claiming the lack of statistically significant warming over a period is confirmation of a pause. That is they claim absence of evidence is evidence of absence.

Even if statistically evidence can be found for a particular pause period, the evidence is corrupted by the cherry-picking. The probability that a given period will produce a small of negative trend might be low - and hence statistically significant, but if you know that the given period was selected because it showed little or not warming, then the probability of such a period existing goes up considerably. An example of this problem is to consider tossing a coin. The probability that you if you toss a coin 7 times and it comes up heads each time is 1 / 128 (around 0.8%). That is statistically significant. But if you toss a coin 100 times, the likelihood of finding a sequence of 7 heads increases considerably, as there are now 94 possible sequences of seven coin tosses, so the probability of a sequence of at least 7 heads is now over 50%.

There are other problems that have to be taken in to consideration with any such statistical calculation. See Global temperature evolution: recent trends and some pitfalls for more and better explanations.

Just What is the Pause?

The other major problem with talk of any pause is that the term the pause has become an umbrella phrase for many different concepts. Sometimes it means a period of no warming, or negative warming, sometimes for a period of slower warming, sometimes for a period of warming that is less than some forecasts, and sometimes for a period where the warming wasn't statistically significant. There are a number of problems with using conflicting definitions for the same phenomenon.

Firstly, it's simply confusing the language. If you say there has been a pause in warming, anyone would assume this meant warming has stopped - there has been no warming. But with these multiple definitions it could mean warming was continuing, possibly faster than before the pause.

Secondly, it makes any statistical evidence even more suspect. All the points made above about the dangers of selection a specific period count double when you can change your definition to reflect the data you have.

Thirdly, it allows a bait and switch argument by those using the pause as an argument against global warming. A scientist might right a study accepting a weak definition of a pause, but that agreement is used as if it was evidence for a much stronger claim.

Through the Looking Glass

The thing that amuses me most about the so-called skeptics talking about the pause is how their own arguments against global warming get inverted. People who react in affected outrage if they are called deniers will happily talk about pause deniers. People who insist arguments based on a consensus of scientists ("as science isn't a democracy"), will point to a flurry of scientific papers apparently acknowledging the hiatus, as proof that the pause is real.