2011-06-13 04:40:31Phil Jones - Warming Since 1995 is now Statistically Significant
dana1981
Dana Nuccitelli
dana1981@yahoo...
69.230.102.37

Post on the Jones comment that warming since 1995 is now statistically significant.

Phil Jones - Warming Since 1995 is now Statistically Significant

2011-06-13 05:09:28
grypo

gryposaurus@gmail...
173.69.56.151

been totally consistent and accurate

I'd lose the 'totally' there and write

been consistent and accurate throughout

or something

 

****

 

Somewhat confusing is when HadCrut was statisically significant after removing noise effects.  At one point you say 2001 (tamino) and then at the end you say 2000 (not attributed, but I assume tamino). 

 

****

The reason we describe this question as "loaded" is that 1995 is the closest starting point for which the answer to this question is "yes".

I think this should be accentuated.  Something like:

Why choose 1995?  Well, that is the closest year for which the answer to this "loaded" question is "yes"

*****

This also might be a good time to have a statisitian do a post on confidence levels and what they mean in research.

2011-06-13 05:09:52
grypo

gryposaurus@gmail...
173.69.56.151

thumbs

2011-06-13 05:31:52
Albatross
Julian Brimelow
stomatalaperture@gmail...
199.126.232.206

Great Dana...I'll have a look when I get a chance.  IMMHO, we really ought to stick it to Lindzen here...he was advising Watts how to manipulate and cherry-pick the data to arrive at the desired result...no warming.  So this email clearly demonstrates a conscious effort to mislead people.  I would go so far as to so for someone of Lindzen's stature, it amounts to scientific misconduct.

2011-06-13 06:19:41
Alex C

coultera@umich...
67.149.101.148

- "Several of the questions were gathered from "climate sceptics", including the second, which has been widely re-published and misrepresented:"

It's not the question per se that has been re-published an misrepresented, it's Jones' answer.  I think "Several of the questions were gathered from "climate skeptics," and Jones' response to the second one has been widely re-published and misrepresented." is a better way to phrase that sentence.

- "...and that the media subsequently grossly distorted Jones' response." - a bit redundant from a previous paragraph, Fox's "gross" misrepresentation.  "Grossly" isn't really needed in this case.

- "...was picked up and posed as a loaded question to Phil Jones in the BBC interview." - It didn't pose as a loaded question, it was a loaded question that posed as honest.  "...up and passed on in..." might be better.

- "This story seems to clearly distinguish the media outlets whose goals are to accurately inform their readers from those whose goals are to misinform their readers."

...which were the ones to accurately inform?  I think a different concluding statement is needed, as in order to say "clearly distinguish" you need different outlets to distinguish with.  So far the ones you have named have all misrepresented his statements and have equivocated.

2011-06-13 07:31:39
Rob Painting
Rob
paintingskeri@vodafone.co...
118.93.24.85

I wonder if it's worth adding the the 95% level is a figure reached by consensus? (a subtle dig) Also, I notice Robin from Carbon Brief's comment about this topic (which I guess prompted this post), unimportant given we're talking about climatic trends, but has anyone crunched the numbers?

 

 

2011-06-13 07:32:47
Rob Painting
Rob
paintingskeri@vodafone.co...
118.93.24.85

Aww, yeah, and the thumb.

2011-06-13 09:34:03Two things I'd like to see in this treatment
John Cook

john@skepticalscience...
121.222.9.229
The problem with this climate myth is it's so technical and shrouded in technobabble (statistically significant warming) that people don't understand it and all they hear is "no warming since 1995". Which plays into the deniers hands as they well know. The fact that the warming switched to significant this year is almost irrelevant. The key point to communicate is this:

Statistical significant warming means 95% chance of warming. But from 1995 to 2009, there was a 93% chance of warming. This whole argument is taking 93% chance of warming and giving people the impression of NO warming. This is the key narrative - turn the denier's strength into a weakness by exposing how they use it to mislead people.

Also important is the GISS record DID show statistically significant warming because it covers the entire globe. But I wouldn't say it like that. I'd talk about it in concrete terms the public can understand. HADCRUT doesn't cover the entire globe and missing out the sharply warming Arctic, it underestimates the trend. GISS covers the entire globe and consequently, over 1995 to 2009, it shows a 9X% chance of warming. The European reanalysis also covers the entire globe and shows a 9X% chance of warming.

Don't get lost in the technobabble, otherwise the deniers win (eg - don't wrestle with pigs in the mud, you both get dirty and the pigs like it). To clearly rebut this myth, you need a narrative (deniers try to turn 93% chance of warming into no warming).

2011-06-13 10:21:51
Albatross
Julian Brimelow
stomatalaperture@gmail...
199.126.232.206

John,

 

I do not agree with the phrasing that you are using....this is all about extracting a signal from the noise.  It has been warming, that is not under dispute, what is under dispute is if the time window selecte dis long enough to establish at a very high livel of confidence that the warming is not arising by chance.  IMHO, the point that needs to be made is that to obtain a robust signal in a relatively noisey time series such as the SAT, we need at least 20 years of data...EOS, and Lindzen knows that damn well.

Anyhow, my wife is giving me the evil eye here, gotta run.

2011-06-13 10:44:29
dana1981
Dana Nuccitelli
dana1981@yahoo...
69.230.102.37

Okay, I made a number of changes to incorporate the suggested revisions.

2011-06-13 11:16:17Alby's explanation
John Cook

john@skepticalscience...
121.222.9.229

There's nothing Alby says that I disagree with (especially the wife's evil eye bit, I'm quite familiar with that experience while climate blogging also). But there is also the key issue that we have to explain this issue in a simple, compelling manner - that tells a story as well as explains the abstract concepts. If it's possible to explain the science and tell a story (Rob Honeycutt is a master at this), then you tick all the boxes.

BTW, Dana, I would strongly recommend someone who is better at statistics than me (probably most people on this forum) confirm the HadCRUT trend from 1995 to 2009 is significant at a 93% confidence level. Someone will inevitably ask us to substantiate this.

2011-06-13 11:44:42
dana1981
Dana Nuccitelli
dana1981@yahoo...
69.230.102.37

...and better at statistics than me.  Alby, if you get time, can you verify the 93% confidence level?

2011-06-13 14:11:19
Albatross
Julian Brimelow
stomatalaperture@gmail...
199.126.232.206

Dana and John,

Sorry do not have time to do a hypothesis test on the HadCRUT data, besides I'd be surprised if someone challenges that (and I have eyes glaring my way again and some deadlines...sigh).  Maybe KR could help out?

More importantly it is critical to show what the trend was for 1995-209 versus 1995-2010.  Emphasize the positive sign and degree of warming (i.e., X.XX C/decade), that positive sign is key-- it has been warming, the warming did not stop or go away.  This is all simply another great example of prestidigitation by the 'skeptics' and deniers....I'm really now beginning to place Lindzen squarely in the 'denial' category, AGW is more than accepting the greenhouse effect.

This might be a good post to have a lay person version, but I realize that doing so is very easy for me to suggest.  As an alternative and/or short cut one could ling to the David Black piece which is for the most part in layman's terms.

Sorry for not being able to pull my weight on this one.  I have still to give the post a thorough and critical read, I'll do my best to tackle that tomorrow. 

2011-06-13 15:11:02Got some numbers
Albatross
Julian Brimelow
stomatalaperture@gmail...
199.126.232.206

Oh nevermind, you got me curious ;) 

Of course the values calc. by Jones were correct.  I used MINITAB and I used the data from here.

I get a p-value of 0.072 (so stat sig at 92.8% level of confidence for 1995-2009, rate of warming +0.108 C/decade.  Warming trend for 1995-2010 is stat sig at 95.8% level of confidence (p-val of 0.042), rate of warming +0.109 C/decade.

 

Regression Analysis: HadCRUT3 (1995-2009) versus Year

The regression equation is
HadCRUT3 (1995-2009) = 0.293 + 0.0108 Year


Predictor      Coef   SE Coef     T      P
Constant    0.29342   0.05038  5.82  0.000
Year       0.010839  0.005541  1.96  0.072


S = 0.0927165   R-Sq = 22.7%   R-Sq(adj) = 16.8%


Analysis of Variance

Source          DF        SS        MS     F      P
Regression       1  0.032897  0.032897  3.83  0.072
Residual Error  13  0.111753  0.008596
Total           14  0.144650


Unusual Observations

              HadCRUT3
Obs  Year  (1995-2009)     Fit  SE Fit  Residual  St Resid
  2   2.0       0.1390  0.3151  0.0410   -0.1761     -2.12R
  4   4.0       0.5290  0.3368  0.0326    0.1922      2.21R

R denotes an observation with a large standardized residual.

 

Regression Analysis: HadCRUT3 (1995-2010) versus Year

The regression equation is
HadCRUT3 (1995-2010) = 0.293 + 0.0109 Year


Predictor      Coef   SE Coef     T      P
Constant    0.29328   0.04685  6.26  0.000
Year       0.010865  0.004845  2.24  0.042


S = 0.0893442   R-Sq = 26.4%   R-Sq(adj) = 21.2%


Analysis of Variance

Source          DF        SS        MS     F      P
Regression       1  0.040134  0.040134  5.03  0.042
Residual Error  14  0.111754  0.007982
Total           15  0.151888


Unusual Observations

              HadCRUT3
Obs  Year  (1995-2010)     Fit  SE Fit  Residual  St Resid
  2   2.0       0.1390  0.3150  0.0386   -0.1760     -2.18R
  4   4.0       0.5290  0.3367  0.0312    0.1923      2.30R

R denotes an observation with a large standardized residual.

2011-06-13 18:31:33Random comments
James Wight

jameswight@southernphone.com...
121.79.14.24

“distortethd” should be “distorted”

“it originated from Motl and/or Lindzen” – it could also have originated from Watts.