Bishop Hill NYT "almost always" exaggerated
Jun 12, 2013 Climate: Surface Justin Gillis, the New York Times' deep green correspondent writes about the plateau in surface temperatures today. If truth is to be told he starts pretty well, explaining scientists's bemusement and confusion over exactly what is behind it.
...given how much is riding on the scientific forecast, the practitioners of climate science would like to understand exactly what is going on. They admit that they do not, even though some potential mechanisms of the slowdown have been suggested. The situation highlights important gaps in our knowledge of the climate system, some of which cannot be closed until we get better measurements from high in space and from deep in the ocean.
Then, however, he goes on to say this:
As you might imagine, those dismissive of climate-change concerns have made much of this warming plateau. They typically argue that “global warming stopped 15 years ago” or some similar statement, and then assert that this disproves the whole notion that greenhouse gases are causing warming.
Rarely do they mention that most of the warmest years in the historical record have occurred recently. Moreover, their claim depends on careful selection of the starting and ending points. The starting point is almost always 1998, a particularly warm year because of a strong El Niño weather pattern.
Is that last point true? We sceptics "almost always" pick 1998 as the start point? I know Lucia doesn't. I certainly don't. David Whitehouse didn't in his report on the plateau. A brief perusal of the results of a search for "no warming for 15 years" turned up nothing at all.
Now don't get me wrong. I'm sure there are people out there who have started at 1998. But my impression is that these are in a small minority.
I wonder if Justin Gillis can back his claim up?
Reader Comments (75)
No comments allowed. A good indicator for a lightweight, poorly researched or partisan story.
Jun 12, 2013 at 8:49 AM | Manfred
---------------------------------------------------
I wish the NYT would stop sending their lightweight comments to me by email after I have requested that they take me off their mailing list.
Okay. So you removed my link. Probably a good thin, anyway.
I have never seen an answer to the question which I keep posing to the warmists/alarmists:
'When was the climate perfect..?'
Here are my reactions to the Gillis article:
Whatever the actual effect of CO2 and co-amissions — I don’t think they are zero, but they are likely much less than the IPCC says, more like what Pat Michaels and all the new articles about climate sensitivity say — Gillis’ piece is propaganda. Here is why:
In the second paragraph, Gillis says:
“True, the basic theory that predicts a warming of the planet in response to human emissions does not suggest that warming should be smooth and continuous. To the contrary, in a climate system still dominated by natural variability, there is every reason to think the warming will proceed in fits and starts.”
Two problems with this:
1. That isn’t what Gillis and the people this article represents were saying a couple of years ago, they were denying there was any significant flattening of temperatures, while they denigrated the people who pointed out the diversion between model and reality; and
2. The models say this shouldn’t be happening, contrary to Gillis’ new spin about natural variability.
So this is more propaganda: deny anything is wrong with the models (who is the “Denier”?} until it gets too obvious that there is an issue, relative to reality. Then and only then do you admit to the obvious, fail to apologize to those you denegrated or acknowledge that they were right, and say that, OK, the models and reality aren’t exactly in sync, but you wouldn’t really expect them to be in sync. The opposite of what you said a few years ago.
These are standard tools of a politician, not of a reporter (especially a science reporter), or of a scientist.
Deeper into the article, Gillis says:
“So the real question is where all that heat is going, if not to warm the surface. And a prime suspect is the deep ocean. Our measurements there are not good enough to confirm it absolutely, but a growing body of research suggests this may be an important part of the answer.”
Note how this sentence silently slides by another, very likely possibility: that because we don’t understand how clouds work at the microphysics level, it is possible that more of the heat that Gillis asserts must be building up somewhere might actually be escaping to space. In other words, he still implicitly assumes the models are right in every important way, but doesn’t say this explicitly, or his readers might actually think for themselves and recall the issue of modelers not understanding cloud microphysics.
Does Journalism school now teach propaganda as a vital part of the curriculum?
I can certainly understand why the NY Times wouldn’t allow comments on this article. Their intelligent but easily herded followers might be exposed to ideas that might open their eyes!
sherlock1 says-
"'When was the climate perfect..?'"
Hansen's 1988 paper clearly concludes that the climate is perfect right now.
Steven Goddard at RealScience recently pointed out that Hansen's 1988 scenario C assumed anthro CO2 emissions zeroed out after 2000, resulting in a temperature plateau that Hansen considered 'safe'. Since measured temperatures are at or below scenario C, Goddard then correctly concluded that the current global temperatures are Hansen-safe, and therefore canonically safe.
QED.
http://stevengoddard.wordpress.com/2013/03/28/hansen-climate-is-deadly-and-safe-at-the-same-time/
PS- Note that Hansen's 1988 temperature prediction for scenario C plateaus 3-4 years AC*. More missing heat, this time the centennial heat allegedly in the pipeline.
*AC = After Carbon
Roy Spencer, former NASA scientist, naturally starts in 1979, at the beginning of the satellite era. The graph in his piece has been widely distributed.
http://www.drroyspencer.com/2013/06/still-epic-fail-73-climate-models-vs-measurements-running-5-year-means/
A reprise of my post on Gillis' article on WUWT:
Whatever the actual effect of CO2 and co-amissions — I don’t think they are zero, but they are likely much less than the IPCC says, more like what Pat Michaels and all the new articles about climate sensitivity say — Gillis’ piece is propaganda. Here is why:
In the second paragraph, Gillis says:
“True, the basic theory that predicts a warming of the planet in response to human emissions does not suggest that warming should be smooth and continuous. To the contrary, in a climate system still dominated by natural variability, there is every reason to think the warming will proceed in fits and starts.”
Two problems with this:
1. That isn’t what Gillis and the people this article represents were saying a couple of years ago, they were denying there was any significant flattening of temperatures, while they denigrated the people who pointed out the diversion between model and reality; and
2. The models say this shouldn’t be happening, contrary to Gillis’ new spin about natural variability.
So this is more propaganda: deny anything is wrong with the models (who is the “Denier”?} until it gets too obvious that there is an issue, relative to reality. Then and only then do you admit to the obvious, fail to apologize to those you denegrated or acknowledge that they were right, and say that, OK, the models and reality aren’t exactly in sync, but you wouldn’t really expect them to be in sync. The opposite of what you said a few years ago.
These are standard tools of a politician, not of a reporter (especially a science reporter), or of a scientist.
Deeper into the article, Gillis says:
“So the real question is where all that heat is going, if not to warm the surface. And a prime suspect is the deep ocean. Our measurements there are not good enough to confirm it absolutely, but a growing body of research suggests this may be an important part of the answer.”
Note how this sentence silently slides by another, very likely possibility: that because we don’t understand how clouds work at the microphysics level, it is possible that more of the heat that Gillis asserts must be building up somewhere might actually be escaping to space. In other words, he still implicitly assumes the models are right in every important way, but doesn’t say this explicitly, or his readers might actually think for themselves and recall the issue of modelers not understanding cloud microphysics.
Does Journalism school now teach propaganda as a vital part of the curriculum?
I can certainly understand why the NY Times wouldn’t allow comments on this article. Their intelligent but easily herded followers might be exposed to ideas that might open their eyes!
Justin Gillis should just sit down quietly for an hour and watch this http://www.youtube.com/watch?v=2ROw_cDKwc0&feature=player_embedded
-followed by: - The Feynman lecture on Seeking new Laws: Feynman http://research.microsoft.com/apps/tools/tuva/#data=4%7C4dbfe549-e795-47a0-bda2-9597fe5bb344%7C%7Cwhich Murray Salby cites at the end of his talk.
This applies to Zed too, if its attention span is up to that.
There's a nice demolition of the NYT's alarmism here:
Reassurance for the climate change faithful: the end remains nigh
I like 2003 as the inflection point. Why start with 1998, the year with the third highest anomaly?
GISS 2010 0.66C
2005 0.66C
1998 0.61C
Hadrut4 2010 0.54C
2005 0.53C
1998 0.52C
Quite right Entropic Man, best to start with now and work backwards. That way you'll see how long it's been since there was any statistically significant warming.
Er, oh...
SteveW
If you want to know if two years are significantly different look at the 95% confidence limits. If they do not overlap, there is a <5% chance that they are similar.
http://data.giss.nasa.gov/gistemp/graphs_v3/Fig.A2.gif
Note the 95% confidence limits icons in green. The 5-year means for the last decade are all significanly higher than any records before the mid-1990s. They are also not significantly higher than later years.
I wouldn't get too excited about that. The variability and 95% confidence limits of the data since 1970 mean that it takes 15-20 years for the long term trend to emerge significantly from the short term noise. The Met Office, GISS or any other competent scientist will agree that there has been no significant warming since the late 1990s. They will also know why.
Those pushing the current campaign in Parliament are either ignorant of this, or deliberately promoting disinformation.
EM - "They will also know why."
Please can you support this with a reference?
Here's a sceptic who thinks there's been no warming since 1998.
http://blogs.telegraph.co.uk/news/jamesdelingpole/100095506/there-has-been-no-global-warming-since-1998/
not banned yet
Look at my 8.03 pm post. No reference is necessary. This is elementary statistics.
You may remember that in February 2010 Phil Jones said that there had been no significant warming between 1995 and 2009. He did not say there had been no warming. He said it was close to, but not quite, significant to 95% confidence.
Given another year of data he was able to say in June 2011 that the warming since 1995 was significant, having crossd that 95% confidence threshold.
Note the timescale. It needed 16 years of data to generate a significant change at latter 1990s/2000s rates of warming. At the current rates, it may take a couple of decades to say with any confidence whether this is a genuine change in slope or just short term noise obscuring a continuing warming trend. Anyone claiming anything else is going way beyond the data.
Ah - I thought you meant that they would know why against the rising CO2 levels, contrary to their theory of CO2 induced warming, temperatures are static.
As far as elementary statistics go, I'm afraid I find your posts confusing. You say:
"At the current rates, it may take a couple of decades to say with any confidence whether this is a genuine change in slope or just short term noise obscuring a continuing warming trend. "
Surely the science of statistics should allow you to formulate a more specific hypothesis than this along with its rejection criteria?
The blog said "A brief perusal of the results of a search for "no warming for 15 years" turned up nothing at all." What sort of search did he do? Try Google. Here are a few searches:
"no warming for 15 years" - 43,300 hits
"no global warming for 15 years" - 49,800 hits
"no warming since 1998" - 98,900 hits
I didn't have to try very hard. Now I know these will not all be relevant quotes, but all the same, the search hardly "turns up nothing".
"Surely the science of statistics should allow you to formulate a more specific hypothesis than this along with its rejection criteria?"
Jun 13, 2013 at 11:11 PM | not banned yet
Given the sample sizes, sample variances and the timescale of the climate data; no.
I suggest you research mean, variance, sample size, standard deviation, 95% confidence limits, significance.Once you understand these terms and how they interact, you might better understand why there is a threshold amount of data below which the significance of any differences between means cannot usefully be determined.
"Given the sample sizes, sample variances and the timescale of the climate data; no"
Tell us then guru - how will we know with any confidence whether this is a genuine change in slope or just short term noise obscuring a continuing warming trend.?
not banned yet
I presume you didn't look up the terms I suggested, or you would understand enough not to ask the question.
Very quickly, before I go to bed, a simple example
Are men taller than women?
Assuming that there is a real difference in average height, the ability to show that the difference is significant depends on the properties of the samples. All else being equal this is how each factor affects the calculation of significance.
Increased difference between the means increases significance.
Increased sample sive increases significance.
Increased variation within samples (expressed as variance, standard deviation or 95% confidence limits) decreases significance.
The difference between the means, and the variation within each sample, should become more or less constant once you get past small sample sizes. Significance improves as sample size gets larger, so more data gives you more confidence that observed differences are real.
Now onsider trend data like climate change, with a constant rate of change, the difference between the mean temperaures for individual years increases with the time between the two samples. Comparing two years close together the difference between the means will below and the significance of the difference is low.
As you compare more widely spaced years the difference between their means is greater and the significance increases. After a sufficient number of years the difference becomes significant.
Interesting example EM but I'm afraid I don't immediately see how it relates to the problem at hand.
Please can you flesh it out as a worked example and highlight how to interpret it with respect to time series of temperature data? Thanks.
Bump for EM - still looking forward to that example calc. and commentary.
Before I can tutor you in statistics I need a tutorial myself.
Using this word processing software how do I produce tables that do not collapse leftward? How do I write formulae and generate mathmatical notation and symbols?
How are you getting on with mastering the IT? Surely it's succumbed to your powers by now?
Got the wp cracked yet EM?