Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.

Settings

Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup

Settings


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Support

Twitter Facebook YouTube Mastodon MeWe

RSS Posts RSS Comments Email Subscribe


Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...



Username
Password
New? Register here
Forgot your password?

Latest Posts

Archives

Baked Curry: The BEST Way to Hide the Incline

Posted on 1 November 2011 by dana1981

Sadly, the so-called climate "skeptics" continue to find new ways to spin the Berkeley Earth Surface Temperature (BEST) results.  First they came up with bogus excuses why the BEST results are not valid.  Then they tried to change the subject, doubling-down on other climate myths.  Now Judith Curry - a member of the BEST team no less (though her involvement in the project has been relatively minimal) - has claimed that the BEST team has tried to "hide the decline" in recent temperatures.

"This is “hide the decline” stuff. Our data show the pause, just as the other sets of data do. Muller is hiding the decline."

Curry's comments were in response to a BBC radio interview with the leader of the BEST team, Richard Muller, who said:

"We see no evidence of it [global warming] having slowed down"

So is Muller right that BEST shows no evidence of global warming slowing down, or is Curry correct in accusing her colleagues of hiding the decline in temperatures?

Misguided Curry

Firstly, it's worth noting that the BEST team addressed the myth that global warming stopped in recent years in their FAQ:

"decadal fluctuations are too large to allow us to make decisive conclusions about long term trends based on close examination of periods as short as 13 to 15 years."

In short, Curry's comments are contradicted by actual statistical analysis done by other members of the BEST team.  As SkS has discussed at length with Dr. Pielke Sr., over short timeframes on the order of a decade, there is too much noise in the data to draw any definitive conclusions about changes in the long-term trend. 

On his blog, tamino does the statistical analysis of the BEST data and finds that because the timeframe in question is so short, the uncertainty is too large to say for certain that the short-term trend in question is any different than the long-term trend.  Right off the bat, it's clear that Dr. Muller was correct to say there is no definitive evidence that global warming has slowed down.

Hiding the Incline

The Daily Mail article containing the Curry interview includes a graph of BEST data originating from serial misinformation source GWPF, with a cherrypicked starting point of January 2001, through the final BEST data point in May 2010 (a period shy of a decade).  Figure 1 highlights the magnitude of the cherrypick by comparing the full BEST record to the fraction of the record included in the article.

BEST cherrypick

Figure 1: Entire BEST record vs. the data examined in the Judith Curry Daily Mail article

Eagle-eyed readers may notice a problem towards the end of the record, as tamino did: the April 2010 BEST anomaly is -1.04°C, which represents a sharp cooling of 1.9°C from the previous month, and is followed by a 2.1°C warming the following month (Figure 2).

BEST with outliers

Figure 2: BEST record since January 2001, with the April 2010 anomaly highlighted in red

Was there really such a large temperature drop and rise between March and May 2010?  It doesn't show up in any other surface temperature record.  When we examine the BEST data, the problem is immediately apparent.  The uncertainty levels in April and May 2010 are 2.8°C and 2.9°C, respectively.  Going back to January 2001, the next-largest uncertainty level is 0.21°C, and the average uncertainty is less than 0.1°C.  Tamino plots the monthly data hockey stick-like uncertainties (Figure 3).

BEST uncertainty

Figure 3: BEST monthly uncertainties since 2001, with a huge spike in April and May 2010

So what happened with the April and May 2010 data?  While the March 2010 anomaly was based on 14,488 stations, April and May were based on only 47 stations, all in the Antarctic (h/t Nick Stokes).  In other words, April and May 2010 should be excluded from BEST data analysis because they are incomplete, their uncertainties are just too large, and April 2010 is quite obviously an anomalous outlier.  Frankly they should not have been published in their current state.

Figure 4 shows how the short-term trend changes when we exclude those two unreliable data points.

hiding the incline

Figure 4: BEST data and linear trend since January 2001 including and excluding April and May 2010

The BEST linear trend increases from 0.03°C per decade when including the faulty data points, to 0.14°C per decade when they are excluded.  It's also important to remember that according to NOAA, which is the dataset most similar to BEST, 2010 was the second-hottest year on record over land (behind 2007), and the hottest globally (effectively tied with 2005).  Ironically, the analysis the "skeptics" are using to argue that global warming has stopped ends in a record hot year for global surface temperatures.

In short, the problem is not that Muller is hiding the decline, the problem is that Curry is hiding the incline.

Enough Cherrypicking Already

As we have recently discussed, although we can't say for certain statistically, it's likely that the global surface temperature warming trend has slowed over the past decade, because virtually all short-term temperature impacts have been in the cooling direction over that timeframe.  Climate "skeptics" desperately want us to believe that the trend has slowed because global warming has magically disappeared, but that's simply not the case.  In fact, as Santer et al. (2011) showed,

"Because of the pronounced effect of interannual noise on decadal trends, a multi-model ensemble of anthropogenically-forced simulations displays many 10-year periods with little warming. A single decade of observational TLT data is therefore inadequate for identifying a slowly evolving anthropogenic warming signal.  Our results show that temperature records of at least 17 years in length are required for identifying human effects on global-mean tropospheric temperature."

Although the Santer et al. analysis applied to the temperature of the lower atmosphere, the same argument applies to surface temperatures.  Figure 5 shows the BEST trend from March 1993 to March 2010 (the most recent 17-year period available in the data, excluding the two final unreliable points).

BEST 17 year

Figure 5: BEST most recent 17 years of data with linear trend

Over the most recent 17-year period, the BEST trend is 0.36°C per decade*, clearly showing the anthropogenic warming trend over that period. 

Examining the causes of decadal variability is both interesting as useful, but exploiting decadal variability to try and incorrectly argue that global warming has magically stopped is neither.  And of course those who argue that global warming has magically stopped conveniently ignore the continued increase in ocean heat content (Figure 6).

warming

Figure 6: Total Global Heat Content from Church et al. 2011

Note that Curry has agreed that tamino's analysis (replicated in this post) is correct and useful, but

"my statement to Rose was about the plot with the 10 year running mean ending in 2006 being misleading. It is misleading."

As it so happens, most of the BEST graphs include 12-month running means (i.e. see their research papers).  Moreover, how plotting a 10-year running mean is automatically "misleading" is a mystery (unexplained by Curry), and Curry's accusations of deception remain unsupported, unwarranted, and unwise.

Accusing other scientists (especially one's colleagues) of deceit for doing proper statistical analysis is simply unconscionable, and why Curry would accuse others of "hiding the decline" while herself hiding the incline is a mystery.  On  her blog, Curry complains that the article misrepresented her to some degree (while standing behind some of her worst comments, and acknowledging that the quotes attributed to her in the article are correct), but frankly when dealing with fake skeptics like those at The Daily Mail, most of us know that misrepresentation is the norm. 

We conclude by offering Dr. Curry the same advice we recently offered Dr. Pielke: DNFTD (Do Not Feed The Delayers).  In short, if an interviewer tries to "tease out" from you a quote about "hiding the decline," or some other denialist myth, don't let them.

===================

* There is probably a small contribution to this trend from the 1991 Mount Pinatubo eruption. The 15-year trend is 0.30°C per decade, 16-year is 0.28°C per decade, and 18-year is 0.41°C per decade.

0 0

Printable Version  |  Link to this page

Comments

1  2  3  Next

Comments 1 to 50 out of 101:

  1. Being a bit pedantic (giving Curry as much of the beneit of the doubt as possible), Muller is not absolutely correct in saying "We see no evidence of it [global warming] having slowed down". The measured trend from 2001-present is lower (0.13642 per decade) than the trend since, say 1993 (0.357871 per decade), which is evidence that the rate of warming had fallen. However it is not statistically significant evidence - it doesn't reach the minimum standard scientists generally regard as being sufficient to proceed with an hypothesis. So Muller's comment is perhaps an over-statement, but Curry's comment is silly; she needs to demonstrate that the decline really exists before she can complain that the BEST team are hiding it.
    0 0
  2. Dikran: First we had Pielke Sr. and now Curry making the same error. Do you suspect a trend? Here's one: Suppose one student from all four grade levels in high school walk into my classroom. The 9th grader was shorter than the 10th grader who was shorter than the 11th grader. However, the 12 grader was shorter than the 11th grader. My conclusion: teenage growth stops in 11th grade - alert the media!
    0 0
  3. 2, muoncounter,
    Do you suspect a trend?
    I object. Your erroneous-global-cooling-trend-statements trend is not statistically significant. Really, how can you identify such a trend with just two data points? Surely you jest, sir.
    0 0
  4. Dikran - fair enough, evidence but not statistically significant (or less than weak) evidence. Really the issue is that, as often happens, the short-term noise is temporarily slowing the long-term trend (in a couple of years the converse will likely happen). The so-called "skeptics" are screaming that this proves global warming has stopped. The scientists who are trying to "bridge the gap" between real science and hysterical "skeptics" (a few of whose names have been mentioned here) don't seem to know how to react to this short-term change in trend. Dr. Pielke has said we should lower the standards of our analysis and essentially admit that global warming has slowed/paused/stopped/whatever. Dr. Curry is frankly all over the place. Dr. Muller started out saying there's no evidence the warming has slowed, then backtracked and apparently said we're in the midst of a "pause". The real failure amongst these scientists is in communicating that short-term pauses are expected, commonly occur, don't tell us anything about the long-term warming trend, and the current one is no surprise given that so many short-term effects have been in the cooling direction in recent years.
    0 0
  5. Sphaerica @3, "Really, how can you identify such a trend with just two data points?" Good point ;) Good one. But we do have a fairly large sample size. Watts, Singer, Curry, Pielke, Delingpole, Eschenbach, McLean, Monckton and many others I suspect claim that the warming has stopped, some of those even claim that we are cooling or have entered a long-term cooling trend. Where is that neat graph that you showed elsewhere? I do not wish to steal your thunder :)
    0 0
  6. I'm considering writing a post that centers around the graph Sphaerica created showing global warming "pauses" every 8 years or so (if you don't mind me stealing your idea, Sphaerica). It looks like "global warming has magically stopped" is becoming the new favorite "skeptic" argument, and it's argument they can always make with some creative cherrypicking of dates, as Sphaerica showed. And the likes of Pielke and Curry and Spencer are not helping matters, effectively encouraging the propagation of this myth.
    0 0
  7. Re: Sphaerica @ 3, dana @ 4. Quite to the contrary. If you want to do a linear fit, I highly recommend that you stop collecting data after you have two points. Anything more just adds to the confusion. The trick is making sure that you have only the two data points you want to show the trend you want. [If anyone wants to disagree with me, keep in mind that I disagree with me, before you start arguing.] I actually read a paper once that did a linear regression on about 100 points, and found that the relationship was not statistically significant. They then duplicated the data set - i.e., they added a second copy of the same data to make 200 points - and then claimed that the regression was now significant, so they had shown that the relationship was real and the only reason for the lack of significance in the original data was that they didn't have enough data. They then proceeded under the assumption that they had come up with a proper result. I have no idea how it got past peer review.
    0 0
  8. Why don't the BEST team have the data from 2010 like everyone else? Could they be "hiding the incline" ;)? Really though, how can they be the "BEST" data set when they are so far out of data?
    0 0
  9. michael @8 - I'm not sure why the BEST analysis effectively ends a year and a half ago. I found that odd as well.
    0 0
  10. The 'let's have fun with BEST graphs' started here. It's a hoot that this comes so quickly on the heels of the Pielke statements as summarized by Albatross and Dikran. In law, if you don't have the facts, you argue the law. In this newly evolving pseudo-skeptic 'science,' if you don't like the signal, you argue the noise.
    0 0
  11. 9, dana1981, My guess is that they just haven't finished yet. Curry's site says that Muller said they rushed to publication before November so that it would be eligible for inclusion in AR5. That makes sense, and if that's the case, and they started from the oldest date and worked forward, the obvious inference is that by publication they'd only gotten halfway through 2010, and are still working on the rest. But... the smart move would have been to completely leave out April/May, not include it with only 47 data points processed.
    0 0
  12. I like muoncounter's new law at 10, and I thought I'd add one: if you don't have any noise, make some.
    0 0
  13. To show how easily the human senses can be fooled, J S Bach used to have compositions where the pitch seemed to ever rise. For more see here http://en.wikipedia.org/wiki/Shepard_tone I think the opposite is happening here with the wished for 'decline' being seen as obvious by our highly evolved pattern recognition systems working overtime in some people. As muoncounter pointed out at 9.22 am the BEST graphs can be potrayed as an ever rising set of 'steps' with the noise hiding the 'risers'. Bert
    0 0
  14. To build on muoncounter's excellent post @10: The Merchants of Doubt (Watts, Curry, Pielke etc.) are behaving just like defense lawyers do and using the same tactics/tricks that they use on jurors: caste doubt, confuse, obfuscate, and make ad hominem attacks. I'd suggest one small change to "muoncounter's law": "If you don't have the science, and you do not like the signal, then argue the noise"
    0 0
  15. Sphaerica @11 - that sounds like a likely explanation. I guess if they convince the IPCC not to rely on HadCRUT as the main surface temperature record in the AR5, that might actually be a useful result.
    0 0
    Moderator Response: [John Hartz] AR5=IPCC's Fifth Annual Assessment report.
  16. I'm not convinced about that November cut-off date mentioned by Sphaerica, supposedly mentioned by Muller. On the IPCC website is the following : WGI AR5 literature cut-off for submitted papers, 31 July 2012
    0 0
  17. To add some light entertainment to this important conversation. http://www.youtube.com/watch?v=TrURLJ6Vlsg But I don’t think the Watts, Singer, Curry crowd are in denial – it’s a lot more complicated than that.
    0 0
  18. Enough cherrypicking already `nough said!!!!
    0 0
  19. 16, JMurphy, I didn't think I could stomach the comments on Curry's blog, but looking it now, they saw the same thing and quickly cried "foul" on Muller's November deadline excuse. The plot thickens. Why am I not surprised?
    0 0
  20. Some opponents of BEST results are arguing that about inaccuracies of thermometer records, and the supposed unreliability of temperature proxy data. Sorry to drift off topic, but I've found a few threads here at SkS that are very helpful but ask if anyone here can direct me to any articles which deal with those two specific questions?
    0 0
    Response:

    [dana1981] Temp record, proxies

  21. Sphaerica#19: "quickly cried "foul"" So, releasing info to the press (or online) prior to publication in peer-reviewed journals is a foul? This rule did not apply to Mme Curry's blog release of 'The Uncertainty Monster,' a full 3 months before it appeared in press. I guess that's how they roll in faux-skeptic world: the rules are different for the other guys.
    0 0
  22. Wow that one month looks like about the coldest month since 1956. Just like McClean predicted. Except he predicted that it would be next year. More seriously I find it ironic that the more sensible sounding opponents of the climate consensus do focus on two major talking points: - that the temperature record should be more open and transparent - that warming has slowed/stopped in the last decade, and that this pause is significant and reflects a reduction in Co2 warming, and not a temporary cooling caused by some other factor. And that HADCRUT shows less warming than GISS over the last decade, and is also the series with the biggest issues around transparency. And now BEST was created by skeptics specifically to address their concerns around openness and transparency shows a greater warming rate than GISS over the last decade - as long as that cool value in Apr 2010 is excluded,.
    0 0
  23. Michael Hauber, Well in fairness to that month... it is only including stations from Antarctica haha
    0 0
  24. Seems a bit silly to just include the temperatures from Antarctica for that month. Why not wait until other data come in? People will be very confused by that blip.
    0 0
  25. The fouls just keep on coming: In this comment at RC, it appears that Fred the-patron-saint-of-denial Singer committed the very same foul. Singer has violated both Nature’s strictures against science by press conference, and the very principle he pretends to defend, by giving it to Watts to publicize Of course, didn't Dr. Roy also play the science-by-press-conference game? I suspect - but have no proof of it - that Dessler was under pressure to get this paper published to blunt the negative impact our work has had on the IPCC's efforts.
    0 0
  26. Commenting on Dikran Marsupial @1 (who knows all this), he raises an interesting point. I prefer to use a definition of evidence by which: x is evidence of y if and only if the probability of y given that x is true > the probability of y given that x is false. By this definition, the slope of the BEST temperature data from January 2001 to May 2010 is certainly evidence that global temperatures will not continue to rise in the near to medium future. I think it is important to recognize that if we are to be guided by the whole body of evidence, but it is even more important to recognize the weight of the recent temperature record as evidence. The importance of recognizing this is illustrated by the paradox of the raven. The paradox of the raven, pointed out by Carl Hempel, is one of many paradoxes of induction. Hempel pointed out that "All ravens are black" is logically equivalent to "All things which are not black are not ravens". It follows that discovering something which is both not black, and not a raven is evidence that all ravens are black. As paradoxical as it seems, that result is sound. Finding a green apple does indeed make it more likely that you will never discover a raven which is not black - by an imperceptible amount. This can even be shown with Bayes theorem, so it cannot be denied without logical inconsistency. But while the green apple is evidence that all ravens are black, it is not significant evidence. Scientists, who only have a limited time on Earth to discover a great many things, try to work efficiently. Consequently, if a scientist was trying to test the hypothesis, "all ravens are black", she would go around examining ravens, not apples. In like manner, when examining temperature series for evidence of future behavior, they concentrate on those series long enough to have a narrow error range. By convention, they focus only on those results which lie within the 95% confidence interval, a standard temperature trends with less than 10 years data do not achieve. Not only are there issues of significance involved; there is also the issue of using all the available data. The temperature series from January, 2001 to May 2010 is not the only evidence that bears on the issue of future temperature rises. The series from December 2000 to May 2010 is also relevant. So is the series from November 2000 to May 2010, and so on. If we were to take all possible trends from the BEST data terminating in May 2010, and weight them according to statistical significance to determine what the balance of evidence shows, it would show a rising trend. In fact, that is the point of Tamino's post. The slope of the temperature data between January 2001 and May 2010 isn't significant evidence of anything much, because of the very wide error bars on the regression. But that does not cause the slope on the data from January 1975 to May 2010 to stop being evidence of future behavior of the temperature series. On the contrary, it remains evidence, and highly significant evidence. Consequently we can categorically state that the balance of scientific evidence indicates that global warming is continuing unabated. This does not mean that there is no evidence, which taken in isolation, and no matter how weak, does not indicate the opposite. But it does mean the great bulk of the evidence strongly favours that hypothesis. It is not clear that this long discussion really adds anything to what Dikran said. However, given the experience of Phil Jones on "statistical significant warming", I think it necessary to be absolutely clear when you say things like "There is evidence that global warming has slowed down", or that somebody is incorrect in denying that. In fact, the response to Muller's claim shows clearly that so called "climate skeptics" will roll rough shod over any subtleties to extract the message they want to hear, regardless of what the data shows. Given that, I fully understand Muller's shorthand expression. What he should have said is that, "There is no significant evidence that global warming has stopped", and "The balance of evidence shows global warming to be continuing unabated". But for a popular audience, that is correctly summed up with the claim that, "There is no evidence global warming has stopped" - to which, were the faux-skeptics not trying to manufacture faux-controversy, nobody would have demurred.
    0 0
  27. Stevo - go here (taxonomy) and look at articles around "temperature record is unreliable" right at the top. If you dont find the answers you want there, how about some more exact claims (and their source) and we could look at it.
    0 0
  28. scaddenp @27, I used the "temperature record is unreliable" thread yesterday. I suspect I might be being "trolled" but have passed on your request for exact claims and specific sources and thank you very much for your kind offer to assist. An article I was refered to was at WUWT entitled 'the metrology of thermometers'. Of course it read to me as sounding like "we reserve the right to ignore any instrument readings we dislike because they were probably inaccurate".
    0 0
  29. Wow, look at that decline. Maybe that McLean fellow was right after all...
    0 0
  30. Note that there are several sub-heading articles on "temperature record in unreliable" covering proxy records. Thermometer records have issues with screens, time of day, recording interval, height etc. For the GHCN stations are homogenised by procedures which are linked at GISTEMP site. A tougher call is why would it show a warming bias? However, this is best discussed on the appropriate thread. The point of BEST is that the best effort by skeptics have yet to show reason to reject GISTEMP. If warming is just an instrumentation problem, then explain the satellite MSU readings, glacier retreat and sealevel rise which broadly agree.
    0 0
  31. scaddenp @30 Thankyou for that. I'll refer any further questions to the appropriate thread.
    0 0
  32. NewYorkJ @29, yup, his retrodiction that in April 2010, the globe would shrink so that it would only include Antarctica was right on the money.
    0 0
  33. Tom C#26: "the point of Tamino's post." I find that tamino is making a much stronger point than you've highlighted. In the text surrounding this graph, he speaks directly to Dr. C:
    "The red dashed line shows the trend rate from 1975. Note that not one single start year gives an estimate which contradicts that rate. That is evidence — damn strong evidence in fact — that the underlying trend rate has not changed since 1975. Hey Judith: there’s a REAL scientific basis for you."
    --emphasis in original

    I think that's the key point: The uncertainties in the trends calculated from the more recent data values are so large that we cannot say with confidence that any such change has occurred. In other words, we have no evidence of that change; on the contrary, we have evidence that it has not changed. This is still a signal processing problem. Until we can separate the signal (the trend) from the noise (short term variations), we should not be fooled into thinking that noise is signal. That was exactly what Dr. Pielke insisted we should be doing: to ignore the obvious that the warming of the lower troposphere has halted, when averaged over the time period 1998 (or 2002) to the present, ignores the obvious signal in the data. That's only true if you like the noise more than you like the signal.

    0 0
  34. I had posted about this over on the no warming since 1998 thread but this specific argument seems to be taking place here. I posted a question there (124) that while it got a number of responses, it at least seems to me that they missed the basic point. According to the daily mail article, Muller made the following statement before Curry made any public statement. When asked about recent data showing global warming has stopped, Muller answered, "There is no evidence of it [global warming] having slowed down." The first part of this is timing. At least from what I understand, Muller made his comment first, before Curry, so Curry's comment is a response to Muller. Is that accepted as true? Second, Muller said there is no evidence of global warming having slowed down, is it accepted that he meant that statement in relation to recent climate data (i.e. roughly the last 12 years)? When Curry responded to Muller, is it accepted that when she says, ‘There is no scientific basis for saying that warming hasn’t stopped’ that she's talking about roughly the last 12 years? So Muller spoke first and said: there is no statistically significant scientific evidence looking at data from the last 12 years proving global has slowed down. Curry responded saying: there is no statistically significant scientific evidence looking at data from the last 12 years proving global warming hasn't stopped. 1. From what I can tell, those two statements are accurate restatements of what they said. 2. Those two statements are logically identical. 3. Both are technically true. 4. As Tamino and many others point out, both are meaningless and useless for intelligent discussion. Whatever people think about Curry or other things she may have said, Muller spoke first and made the first meaningless, misleading statement. Curry called him on it, argued that the exact same data he cited could be used to say the exact opposite of what he said and that she was embarrassed he would put out such a misleading statement. Her position was simply, don't over-speak, we just put out the most accurate, reliable data set to date, let it speak for itself. No matter what you think about the bigger argument, Muller made a technically true but meanignless statement in talking to the Daily Mail reporter. He could have simply said 10-12 years is too small of an amount of time to know anything about what's happening with global warming, he could have talked about decadal fluctuations, but he started the "there is no evidence" ball rolling. It seems seriously inconsistent to hammer on Curry but not put the blame on Muller who started it.
    0 0
  35. muoncounter @33, with respect, I believe I was making the same logical point that Tamino made with his entire post, not just one graph. Using Tamino's preferred ARMA(1,1) model, the uncertainties of successive trends to March 2010 in BEST data are as follows: You can see from that that if we only had the data from 1993 onwards, we could not say with certainty that the underlying trend was positive, let alone that it was close to 0.27 degrees C per decade. Given that, we would still be in the position to say that "not one single start year gives an estimate which contradicts" a rate of 0.27 degrees C per decade. But we would also be in a position to say that "not one single start year gives an estimate which contradicts" a rate of -0.01 degrees C per decade. As it happens, we are in a position to say something far stronger than that, ie, that the underlying trend is very likely (greater than 0.95 probability) to be close to 0.27 degrees per decade. We are in that position because the trend from 1992 to May 2010 is also evidence regarding that underlying trend. And so on for the years that go back. (We can also say that for reasons of physics, which are more important that mere statistical reasons, but that is beside the point here.) Perhaps it would be more accurate to state Tamino's point this way: 1) We have strong evidence from medium term trends (greater than 17 years) that the temperature trend is close to 0.27 degrees C; 2) No short term trends provide evidence significant enough to bring suggest the evidence in (1) is misleading about the future temperature trend. I believe you are focusing on (2), but logically, without (1), (2) is largely irrelevant. (I do not exclude the possibility that we are merely talking at cross purposes.)
    0 0
  36. Rickoxo @34, taking your questions in order: 1) Yes, Muller made his statement first. 2) No, Muller was (probably) talking about the whole body of evidence. The trend from 1975 to current is just as much evidence about what will happen in the future as the trend from 2000 to present. Muller chose not to ignore that evidence, and to make his statement based on the whole of the evidence before him. 3) No. Curry explicitly only talks about the period from January 2001 to May 2010, a period of just under 10 years. It is important to her case that she uses just under 10 years, rather than 10 years or 12 years. Had she used either of the longer periods, a clear (but not statistically significant) positive trend would have been present. Finally even if we confine ourselves to consideration of just the last 12 years data, we can only conclude that global warming has stopped by ignoring the physics. We know from evidence in the last 12 years that early in that period we had frequent El Nino's, whereas since 2008 ENSO has tended to La Nina's. We know that El Nino's result in warmer years, and La Nina's result in cooler years. We know that solar activity has reduced over the last 12 years to levels not seen since 1910, so much so that some solar physicists are predicting a new maunder minimum. Therefore we know that without an significant warming factor temperatures would have declined drastically over the last 12 years, instead of remaining fairly constant. That they have not is clear evidence of a countervailing warming force that will restore warming once the ENSO cycle switches to neutral or El Nino's. In essence, Curry has taken a line with troughs and peaks, and drawn a line from a peak to a trough and insisted that that is the trend. We don't need to know statistics to know that that is garbage.
    0 0
  37. Hi Rick @34, I'm sorry, but you seem to be under the impression that repeating claims makes them true. It doesn't. The primary subject of this thread is Judith Curry, not Dr. Muller. Additionally, the only people posting the alleged comments made by Dr. Muller are some "skeptic"/denier sites, David Rose and you here. So I have to question the validity of the alleged statement by Muller and would like to see a transcript of the radio interview before accepting what he is alleged to have said. The fact remains that that Tamino offered a detailed statistical analysis to make his points. In stark contrast, Curry has not offered any scientific or statistical evidence to support her assertion that: “There has been a lag/slowdown/whatever you want to call it in the rate of temperature increase since 1998.” I do agree with you that Curry's claim that "There is no scientific basis for saying that warming hasn’t stopped.", is "meaningless and useless for intelligent discussion." Someone has already explained the logical fallacy of your claim that Muller's and Curry's claims are logically identical here, you might also wish to read this post by Tom Curtis. "Curry called him on it, argued that the exact same data he cited could be used to say the exact opposite of what he said" Not true. Curry has "called" Muller other things too, at least she suggested that he tried to hide the decline. A serious allegation that is not supported by the data and proper statistical analysis. As has been shown here and at Tamino's that claim is not true. If you have not yet read Tamino's post (it seems that you have not, at least not in detail) then I suggest that you do so. You can find it here. "she was embarrassed he would put out such a misleading statement." Judith Curry has a great deal many other very real reasons for which she should be embarrassed, and they pertain to her own musings and actions. "Her position was simply, don't over-speak, we just put out the most accurate, reliable data set to date, let it speak for itself." That is not true either. Her position is to caste doubt and inflate uncertainty, and she is not letting the data speak for itself, she is making unsubstantiated and demonstrably false claims about the data. As Tamino noted though, "It looks like the “uncertainty monster” decided to turn around and bite her on the ass.". That statement by Tamino really does sum things up nicely about Judith Curry on this one. Have you tried to make these same arguments to Tamino? I'd be interested to see what he has to say about your claims. "It seems seriously inconsistent to hammer on Curry but not put the blame on Muller who started it. " This is now beginning to sound like a schoolyard tiff-- "but he started it first!". Except the basis for that claim is only traceable back to highly questionable sources-- I would not believe anything that David Rose writes Rick. That we are having to "debate" this BS is beyond me, and why you feel compelled to try and distract from Dr. Curry's despicable and unprofessional behavior is also beyond me. Or is this your attempt to fabricate debate and controversy and/or derail the thread Rick?
    0 0
  38. You guys are really < -snip ->. how can you look at this graph and say that warming has not stopped.
    0 0
    Moderator Response:

    [DB] Please note that posting comments here at SkS is a privilege, not a right.  This privilege can and will be rescinded if the posting individual continues to treat adherence to the Comments Policy as optional, rather than the mandatory condition of participating in this online forum.

    Moderating this site is a tiresome chore, particularly when commentators repeatedly submit offensive or off-topic posts that demonstrate that they have not bothered to read the OP of the thread that they are commenting on. We really appreciate people's cooperation in abiding by the Comments Policy, which is largely responsible for the quality of this site.
     
    Finally, please understand that moderation policies are not open for discussion.  If you find yourself incapable of abiding by these common set of rules that everyone else observes, then a change of venues is in the offing.

    Please take the time to review the policy and ensure future comments are in full compliance with it.  Thanks for your understanding and compliance in this matter.

  39. garythompson @38 Simple, by looking at the long term data and reconciling it with observations such as ENSO and solar cycle. How can you say warming has stopped when the Arctic ice cap has melted to its smallest extent in recorded history twice in the last five years? How can sea level rises in Tuvalu and Kiribati be explained if warming has stopped. Please read the article and comments in full.
    0 0
  40. garythompson @ 38- why do you comment on a post you don't even appear to have read? This is about the BEST global temperature data. Note Figure 4 and comment @ 1 by Dikran Marsupial. Simply put: short-term trends aren't very informative of long-term trends. What happens when the surface oceans switch to a warm phase (El Nino), or the dimming effect of pollution aerosols is overwhelmed by greenhouse gases? That's why a mechanistic understanding is important too, and why SkS stresses the importance of looking at the whole picture.
    0 0
  41. Tom, Ok, I get your take on #2, but near the end of the article, the Daily Mail wrote: However, he [Muller] admitted it was true that the BEST data suggested that world temperatures have not risen for about 13 years. But in his view, this might not be ‘statistically significant’, although, he added, it was equally possible that it was – a statement which left other scientists mystified. This is why it seems like his original statement was about this same period, that's why Curry picked and after Curry responded, he was challenged on what he had said earlier. Please don't hear me saying I know what the Daily Mail wrote was true. There seems to be some serious push back from both Curry and Muller that the various articles didn't faithfully report what they said. But no one has challenged Muller's original statement and his response fits with other things he said. So, it seems like the best evidence shows him talking about the same rough period Curry responded to. As for Curry not being technically right, what is inaccurate about what she said? There's no way to think she didn't mean statistically significant, it's the whole point she's trying to make and it's the exact point Tamino makes critiquing her. Data from last week doesn't prove or disprove global warming, it's stupid to ask if it does. Data from the last ten years doesn't prove or disprove global warming, it's too small of a period of time. But Muller made the first comment and Curry simply pointed out it was a stupid statement, as well as a few other stupid things he had said (now there's no cause for skepticism when there used to be ...) One last thing Tom, I don't know why you're attributing the graph the Daily Mail printed to Curry. She never mentions it on her blog, she makes comments on her blog that make it sound like she understood that the last two data points in the data set were problematic and she never presents that graph anywhere. I was just doing some checking and one of the articles attributes the second graph showing no temperature rise to "a report to be published today by the Global Warming Policy Foundation". So when you say Curry picked out that special under ten year window, she didn't pick that cause it's not her graph. Last, regarding your comment at the end, Curry specifically says the data set project has nothing to do with physics, El Nino's, La Nina's or anything else other than providing the most accurate and accessible record of land temperature data. She wrote, "Our study addressed only one area of the concerns: was the temperature rise on land improperly affected by the four key biases (station quality, homogenization, urban heat island, and station selection)?" To say she's wrong because she isn't referencing something that is entirely outside the scope of the BEST project isn't fair.
    0 0
  42. Rob Painting @40, short term trends are known to uninformative, but short term trends from start points carefully selected with a predetermined result in mind, from a carefully selected data set when other equivalent data sets do not show the same thing, and when the data set selected is known to be biased in favour of the view of the person selecting the data are even less informative. Consequently Thompson's carefully selected period of a CRU graph tells us far more about him than future temperature trends.
    0 0
  43. Rickoxo @41: 1) I am uncertain what Muller "admitted" about the last 13 years of BEST data, because the claim is made by David Rose who has a reputation for gross distortions and inaccuracy. Nor does Rose give us the complete quote, but only partial quotes as part of his narrative, a classic method of spinning quotes out of context to give a false impression. 2) Assuming Muller "admitted" that "world temperatures have not risen for about 13 years" (Rose's words), then he admitted a falsehood, which would strongly suggest he is not even familiar with his own data, which I find highly unlikely. Below is a detail of the last few years of the BEST temperature index taken from the methods paper (PDF): The BEST data is given in black, with HadCRU (blue), GISS (green) and NOAA (red) also shown. The grey vertical line is 2000. The inner yellow box shows the data plotted by the GWPF, while the outer yellow box shows the last 13 years. Very clearly, had the plot by the GWPF included data from 1997 (13 years), it would have shown a significant positive trend. Indeed, a termination at 2001 is the longest interval they could have included and still retained near zero trend. This gives the lie to David Whitehouse's claim that,
    "Incidently you could extend the graph back a few years before 2001 and it doesn't make much difference because the 'super el nino' of 1998 and the two subsequent cooler years of 1999 and 2000 do not show up as dramatically in the Best land data as they do in HadCrut3."
    More importantly it shows that claims that "world temperatures have not risen for about 13 years" (David Rose) are simply false, whether Muller concurred or not. 3) Muller's purported claim that "this might not be ‘statistically significant’, although, he added, it was equally possible that it was" (David Rose) are hardly mystifying. If that is indeed what Muller said, it merely means he has not calculated the statistical significance of the trend for what ever interval of data Rose asked him about. If he has not made that calculation, or seen it done, his claims about global warming having stopped for 10 years are not made based on the full available evidence. But that is irrelevant to discussion here, because Tamino has made the calculation, and made the results publicly available so we know that Muller's claim (about the 10 years) was true regardless of whether he himself was justified in making it. These possibilities do not paint Muller in an attractive light, but his character flaws are nothing new in the climate change debate. Regardless of his known flaws, given the source of the truncated quotes, I would be very loath to assume Muller had not made the relevant analysis to justify his comments without verification from his own mouth. 4) Muller's statement about the 10 years was made on October 21 to the BBC. The purported statements about the 13 year period where made sometime on October 29th, or later. Therefore those later comments provide no context from which to understand his earlier comments. Any inference from the later comment that in the earlier comment he was speaking about the 10 years data to the exclusion of any other relevant information is unwarranted. 5) Curry's statement as quoted by the Mail (and explicitly not disavowed by her on her blog) was:
    "There is no scientific basis for saying that warming hasn’t stopped"
    As copiously shown by Tamino, there is a scientific basis for saying the warming hasn't stopped. That scientific basis is that: a) The period 2001-2011 is a sub-period of an interval with a statistically significant positive trend of approx 0.27 degrees C per decade; and b) The period 2001-2011 is not a sub-period of any interval with a statistically significant zero or negative trend. (a) provides a sound scientific reason to believe the underlying trend in that period is positive, and near 0.27 degrees C per decade. (b) shows that the sound scientific reason in (a) is not defeated by any counter-evidence. The two combined give us a very sound scientific reason for saying the warming has not stopped. In fact, the case is stronger when known physics is included, but that is beside the point. Based purely on the statistics, Curry is not wrong. She is egregiously in error. Sufficiently so that we must either doubt her competence or honesty to explain the error. 6) (And finally), you are corect that Curry did not prepare the graph, and I apologize for my error. However, Curry was shown the graph, and does not disavow it. Further, her comments where made in reference to the graph. Consequently my points made above still stand.
    0 0
  44. An excellent and timely piece. I think it's a good thing that Curry et al are nailing their tattered colours so firmly to the mast of warming having stopped. It's become increasingly obvious to everyone that it hasn't; and this conviction is likely to deepen with time. Curry and her ilk can only cherry pick these figures for so long before there are no cherries left on the tree - and no fig leaves either...
    0 0
  45. You know, I read Curry's statement, "This is 'hide the decline' stuff", and immediately thought; 'what, you mean people grossly misrepresenting the data?' She then proceeded to grossly misrepresent the data. So... I agree with Judith Curry. This is 'hide the decline stuff'. The same sort of nonsense as the claims that 'hide the decline' referred to a decrease in global temperatures, rather than the divergence problem in recent tree ring proxies. That Curry is still promoting that old falsehood, and now 'shoring it up' with new misrepresentations places her towards the bottom of the 'Scientific' Denier Credibility Scale. Somewhere around Spencer's level I'd say (i.e. higher than Singer / lower than Pielke).
    0 0
  46. Tom said:
    a) The period 2001-2011 is a sub-period of an interval with a statistically significant positive trend of approx 0.27 degrees C per decade; and b) The period 2001-2011 is not a sub-period of any interval with a statistically significant zero or negative trend.
    Wow. Logical reasoning is hard. Much harder than spotting the flaws in other people's reasoning. But you nailed it here; that's one for the quote file.
    0 0
  47. 38, garythompson,
    ...how can you look at this graph and say that warming has not stopped.
    You are so right! Take a look at this graph. Warming hasn't merely stopped, but cooling has started. In fact, cooling started back in 1973 and the globe has cooled since. The fact is that all trends are negative since 1973, without pause, and therefore the inescapable conclusion is that there has never been any warming. [Hint: A "skeptic" can pick any point in time in the past 30 years and prove that the globe is cooling. What does that tell you?] Albatross: Sorry. I missed your request for this back at comment #5.
    0 0
  48. Tom C #35: If I may repeat your points with a minor correction, "1) We have strong evidence from medium term trends (greater than 17 years) that the temperature trend is close to 0.27 degrees C; 2) No short term trends provide evidence significant enough to bring suggest the evidence in (1) is misleading about the future temperature trend." I agree with your statements 1 (where the stated trend is in deg C per decade) and 2; a most cogent and compelling summary. The implication of (2) is clear: it is nonsensical to go on about 'there is evidence for a change in trend' or 'there is no evidence that the trend hasn't changed.' In the first case, because it is incorrect; in the latter case, there is no evidence that a lot of things have not happened; what kind of science deals with that? Unfortunately, those who are in denial must find something to deny, so they will go on with this nonsense. That leaves them to practice their chosen field: pseudo-skeptic climastrology, which now includes the study of the lack of evidence of that which hasn't happened.
    0 0
  49. Tom Curtis wrote: "The period 2001-2011 is not a sub-period of any interval with a statistically significant zero or negative trend." Presumably you intended this to refer just to the duration of the BEST record, but without that caveat being explicitly stated there are technically very large intervals (e.g. the past 50 million years) for which it would not be true. Here's a thought... when was the last statistically significant zero or negative trend within the confines of the BEST record? I'm guessing that it was decades ago.
    0 0
  50. muoncounter it may not be a trend, but it is significant! ;o) dana1981 I am in complete agreement. Even giving Prof. Curry the benefit of the doubt, her comments do her little credit. It doesn't overly surprise me that scientists don't do a good job of communicating the science, the really susprising thing is that they have not actually performed the required statistical test of their hypothesis. Bob Loblaw ;o) I once read a microbiology paper where a "quadratic response surface" was fit to two points with one independent variable! Fortunatly the software wisely decided that the coefficient of the squared term was zero (which was faithfully written down in the equation for the model). Tom Curtis Nice comment. Perhaps the Bayes factor, which gives the ratio of the support from the data for two hypothesis (taking the uncertainties into account), would be a better approach. If the Bayes factor is less than three the evidence is generally considered "barely worth mentioning", which I suspect would sum up the 2001-present trend. Gary Thompson Simple, we have a decent grasp of basic statistics principles and know that the eyecrometer is not a reliable instrument in detecting relatively small magnitude signals in the presence of noise.
    0 0

1  2  3  Next

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



The Consensus Project Website

THE ESCALATOR

(free to republish)


© Copyright 2024 John Cook
Home | Translations | About Us | Privacy | Contact Us