Comments 1 to 9:

1. Lionel A at 03:15 AM on 15 October, 2012
   I see that David Rose has another dreadful piece of creative writing in the Mail this weekend. I am trying to track down the Ben Weller of that 'temperature graph' to discover what it is based upon as I smell a Monckton Manoeuvre here.
   0 0
2. KR at 04:20 AM on 15 October, 2012
   Lionel A - Well, to start with, the graph in that article is mislabeled. Using either the SkS Trend Calculator (1997.7) or WoodForTrees (1997.6, zero based decimal year) it's clear that the graph data from the Mail starts in September 1997, excluding cooler temperatures earlier in that year - while the Mail axis is (mis)labeled as if it were the start of 1997.
   
   And that displayed data has a rising trend, albeit small and rather noisy - the autocorrelation corrected (i.e., conservative in uncertainty) data shows:
   
   Trend: 0.071 ±0.166 °C/decade (2σ), a 2σ range of -0.095 to 0.237.
   
   While an upward trend, that's not enough data to exclude either the null hypothesis or the ~0.17 °C/decade long term trend given the variations. And whoever compiled that graph failed to show the statistics. Hmmm...
   
   So - this is an argument from insufficient, statistically insignificant data at the 2σ level. If you include enough data to exclude either 0.0 or 0.17 °C/decade (which using the Trend Calculator on HadCRUT4 is two more years of data to ~1995.6) that data indicates that the zero warming null hypothesis is false with a trend of: 0.142 ±0.139 °C/decade (2σ). So when looking at sufficient data to statistically call either warming or no warming? Warming.
   
   This is yet another 'skeptic' claim from insufficient data, with cherry-picked intervals, and absolutely no statistics - to put it bluntly, bullpucky.
   0 0
3. KR at 04:28 AM on 15 October, 2012
   Lionel A - I should also note another deceptive bit of graphing in that piece from the Mail; the labeling of single timepoints at the start and end of the graph with equal values. That is ridiculous given the variations in the data. I could choose points (from that graph) in mid-2000 and mid-2006 and show a difference of 0.7C, or late 1997 to mid-2007 and show a difference of -0.6C. None of these choices is statistically justifiable.
   
   On the other hand, looking at the trend of the entire set of data, even over this statistically insignificant period, you see a trend of 0.071 °C/decade. The very data the Mail graphed disproves their assertion - it is warming.
   0 0
4. John Russell at 04:50 AM on 15 October, 2012
   I'd like to point out that the UK Met office has almost immediately issued a thorough rebuttal of Rose's article, which clearly reveals how cynical has been his deception of the Daily Mail's readers.
   0 0
5. Phil at 04:53 AM on 15 October, 2012
   KR @2;
   
   Your calculated trend of 0.071+/-0.166 C/decade is higher than the historical one stated (or at least implied) by Rose himself in the article; 0.75/13 = 0.057 C/decade. So Rose not only presents evidence for continued warming but also that the warming is accelerating - shame about his headline !
   0 0
6. JosHagelaars at 06:39 AM on 15 October, 2012
   @KR#2
   
   The HadCrut4 data in the SkS Trend Calculator as well as in the WoodForTrees link does not go any further than December 2010.
   Using the HadCrut4 4.1.1.0 monthly data from September 1997 to August 2012 I get (using the SkS Trend Calculator uncertainty calculation):
   +0.033 ± 0.145 °C/decade
   
   The range is clearly cherry-picked, it starts with the extraordinary 1998 El Nino. The range ends with the La Nina years 2008 and 2011. In this short period the influence of ENSO and the sun on the global temperature is clearly negative and one would expect the trend to be lower than the long term average. The trend for the last 30 years for the new HadCrut4 is +0.165 ± 0.051 °C/decade, so I would say your conclusion still stands when the data are extended to august 2012.
   
   It is absolute nonsense to state that 'Global Warming has stopped' based on such a short time-frame. When you just start 1 year later with the HadCrut4 data, from September 1998 to August 2012, the trend more than doubles:
   +0.078 ± 0.147 °C/decade
   Using the same 15 years interval from May 1992 to April 2007 HadCrut4 gives a trend of +0.287 ± 0.151 °C/decade. About 5 years ago the Mail On Sunday would probably have concluded that global warming was getting totally out of control. They should have a look at the SkS Escalator.
   0 0
7. dana1981 at 06:53 AM on 15 October, 2012
   It's also of course important to remember that surface temperatures are a very small fraction of overall global warming, which Nuccitelli et al. (2012) shows has not slowed.
   0 0
8. Bob Loblaw at 08:51 AM on 15 October, 2012
   As for making claims using short time periods, we can only follow the sage advice of Roger Pielke himself (I think it's from Junior, not Senior), who thinks that even thirty years is not enough:
   
   

   "Thirty years is not an appropriate length of time for a climate analysis, much less finding causal factors like climate change," says Roger Pielke, a professor of environmental studies at the University of Colorado.
   

   
   
   (as quoted in USA Today)
   0 0
9. John Russell at 21:55 PM on 15 October, 2012
   Further to the Met Office's rebuttal (see #4 above) I'd just like to point out that Judith Curry has a problem with David Rose's piece too.
   0 0

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.