Comparison of 10-year trends from two estimates of global mean surface air temperature. Notice how variable the 10 year trends are in both data sets. The most recent red analysis from NOAA shows no evidence of a hiatus, but the real story is that 10 years is just too short a time to make reliable ‘trend’ estimates. For comparison, look at the 30-year trends on the same data at the end of this story. By the way: The two data sets would normally overlap each other but have been offset to allow you to see the data more clearly. Calculations by, and thanks to, Ed Hawkins. (CLICK FOR A LARGER GRAPH!)
Friends: Sorry for the reduced frequency of blog postings. Work has been busy and in my spare time I have been indulging myself by reading books and recording music. These activities feel great, but I am still constantly itching to write about … well, everything!
Anyway: to the business at hand…
Some of you may have seen articles around the web stemming from an article in Science describing a new estimate of the trend in global temperatures. The new work has increased our estimate of the warming trend in recent decades, invalidating the use of the term ‘hiatus’ to describe the recent trend. And in a small sense – I actually helped!
The new estimates are based on re-consideration of existing data on the air temperature above the ocean and land surfaces.
For the Land Surface Air Temperature (the temperature record derived from conventional thermometers in Stevenson screens), the most significant effect has been the use of a new data bank of surface temperature records, developed under the auspices of the International Surface Temperature Initiative (ISTI).
ISTI was born – or at least conceived – at a meeting at the Met Office in 2010, and since its start I have sat on its steering committee! My role on the committee has been minor – essentially keeping open channels between the metrological and meteorological communities. But I was there!
It has taken 5 years of work – by other people – to reach the point at which a ‘data bank’ has been released, and then still further work to allow the formation of a ‘data product’.
The data bank can be downloaded for free and if anyone disagrees with the conclusion then they can devise a method of analysis and test it for themselves.
And the point is…
… no serious enquirer can reasonably doubt that the Earth is warming – the evidence is overwhelming.
Skeptics at first rubbished the estimates of global surface air temperature and said they were – in essence – bogus.
When the Berkeley Earth estimate showed essentially the same trend as the other estimates, skeptics changed their tune
Now they believed the data in detail and said ‘look: the warming has stopped!’
And at first sight it might seem that the new study refutes that claim. This brings us to the key point.
The new estimate differs only minutely from the old estimate. The real significance of the new analysis lies in the fact that this minute change was enough to remove the trend. This shows how statistically fragile the trend was.
Now the new estimate is likely to be better than the old estimate since its includes more station records from new locations. But ’10-year trends’ in this data are just to ‘noisy’ to draw any conclusions.
For comparison compare the 30-year trends in the graph below with the 10-year trends shown at the head of this article.
Comparison of 30-year trends from two estimates of global mean surface air temperature. Notice how similar the 30-year trends are in both data sets. Calculations by, and thanks to, Ed Hawkins
It is interesting to note that the ‘noise’ on the data is not statistical noise. It occurs because there are many ‘cycles’ around the globe with periods of ‘a few’ years, and one must average over many of these cycles before one get reasonable trend estimates. You may also like to check out articles by:
- Victor Venema who runs the Variability Variability blog. He has a nice article on this subject.
- Ed Hawkins – the author of the graphs.
===================================== *With apologies to Spike Milligan
Protons for Breakfast 
July 9, 2015 at 1:03 pm |
We’ve missed your posts. Glad you’re ok, sir
July 9, 2015 at 2:49 pm |
Does removing the trend undermine the physics of the Berkeley Earth overview?
Click to access examining-the-pause.pdf
Or were they just wasting their time, since the hiatus no longer “exists”?I think we will soon find out given that the AMO is looking very cool.
July 19, 2015 at 10:12 pm |
Stephen
I think the point is that a 10 year trend is too noisy to deduce anything. I think we need to be patient: this is the future and it is very hard to predict.
This ‘hiatus’ is only an issue because in most countries the electoral timescale is 4 years or even less. And Climate Change is too slow to provide news stories that at the necessary rate to inject momentum into the debate. Some people are deliberately seeking to undermine the climate science by any ,means they can because they dislike the potential solutions to the carbon problem. It would be more honest to just state that, but they can’t bring themselves to do that. I don’t know why.