date: Thu, 23 Oct 2008 18:09:59 +0200
from: Stefan Rahmstorf <rahmstorf@ozean-klima.de>
subject: Re: global temperatures
to: P.Jones@uea.ac.uk

<x-flowed>
Dear Phil,

enviable, to be in Iceland. How are they dealing with their bankruptcy?

I take it that you basically agree with the interview text. I have 
adapted it below, taking some of what you wrote into account. I have cut 
the bit about the differences between Hadley and GISS, I think that is 
getting too complex. The change you mentioned that happened in 1990 
could not explain why the 1998-2007 trend is so different between GISS 
and Hadley.

My trend calculations you can trust, I computed them myself from 
http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/annual
In fact, I just double-checked and recalculated them. Looks like the 
2007 value was updated since I last used it, it is 0.01 C lower than 
before. This is why the 9-year trend has changed from 0.22 to 0.21. The 
10-year trend is unaffected.


Thus the little interview now reads:

KlimaLounge: The German journalist Ulli Kulke has repeatedly claimed 
that global warming has stopped and global temperatures "show no trend 
in this decade". You are one of the leading experts on global climate 
data. Has the global warming trend stopped, or at least slowed down?

Jones: No. What happens is that some people take the trend starting in 
1998, which was an extremely warm year, well above the long-term trend 
line. This exceptional warmth was due to an El Nio event in the Pacific 
ocean. If you take the trend 1998-2007 in our data, it is + 0.09 C per 
decade. But if you take the trend 1999-2007, it is + 0.21 C per decade. 
For comparison: the long-term global warming trend for the past 50 years 
is 0.13 C per decade. Thus, if you start in 1998 you get a 
below-average trend, if you start in 1999 you get an above-average 
trend. That simply is a result of natural year-to-year variability, 
which is always superimposed on the long-term warming trend. There is no 
indication for a change in the trend. You explained this very nicely in 
your Realclimate article earlier this year.

KlimaLounge: But Kulke claims that his statements are based on you 
saying that there has been no statistically significant warming since 
1998, and that therefore we can relax our mitigation efforts.

Jones: That is a serious misinterpretation of what I said. Indeed the 
trend since 1998 is not statistically significant - simply because of 
the natural variability just mentioned, the time span of 10 years is too 
short to reliably determine a trend. The uncertainty on a 10-year trend 
is +/- 0.2 C per decade (2-sigma). Therefore, one certainly cannot 
claim that the warming trend has slowed down on this basis. This is 
nonsense and just confuses the signal of global warming with the 
superimposed noise of natural variability. Such claims are 
scientifically simply incorrect.

KlimaLounge: Thanks for this clarification. Let's hope that Kulke will 
stop making such false claims in future. How long a period do you need 
to obtain a reliable trend?

Jones: If you take 20-year trends the result is getting more robust, the 
uncertainty here is only +/- 0.07 C per decade. The most recent 20-year 
trend, 1988-2007, is 0.20 C per decade. With this more robust measure, 
the most recent trends are the highest. Compare this for example with 
the trend centered ten years earlier, 1978-1997: this is only 0.11 C 
per decade. Thus, on those time scales where one can make scientifically 
sensible statements, the global warming trend has accelerated and 
certainly not slowed down.

If you can live with this, I'll make a German version. Enjoy the hot 
springs.

Cheers, Stefan

-- 
Stefan Rahmstorf
www.ozean-klima.de
www.realclimate.org

</x-flowed>
