Stockwell on CSIRO Drought Report

David Stockwell has posted up an analysis of the CSIRO Drought Report, using the data grudgingly made public by CSIRO after public pressure. Key claims of the CSIRO report do not pass obvious statistical test for “significance”. Please visit David at his blog.

25 Comments

  1. Robert Wood
    Posted Aug 5, 2008 at 6:12 PM | Permalink

    I read this article and my head spun, as it so often does on this site.

    But one question, which has been nagging me for a long time, is: “Has no one made a basic physical analysis of a gas mixture surrounding a spherical body?” My goodness, I can even tell which horse will win the race, given a spherical horse.

    This would be a basic thermodynamic stability model, where P,V and T, and R are all varying, under the force of a constant gravity. And once we have that basic model, then we can perturb it with humidity, insolation, convection and albedo (clouds). Why has this not been done, or am I missing something?

  2. DeWitt Payne
    Posted Aug 5, 2008 at 11:17 PM | Permalink

    It has. They’re called GCM’s or General Circulation Models or variants of that. If you really want to get lost, I direct you to the Exponential Growth thread part 1, part 2 and part 3 which discusses whether these models actually work or can ever work even with unlimited computing power.

  3. fred
    Posted Aug 6, 2008 at 12:57 AM | Permalink

    Its a very sobering read. Stockwell has not done anything very sophisticated to blow this thing up. What seems to be happening, here and elsewhere, is that Climate Scientists are engaging with topics which require statistical treatment, but they do not themselves have the statistical expertise, and are unwilling to co-opt it from professional statisticians. Consequently they make what are, in statistical terms, fairly basic mistakes, and misrepresent both the trends of their data and its significance. One understands them not having the expertise themselves, but to refuse to bring statisticians into the team, and for this to be a general phenomenon, not just a few individuals, reflects really badly on the whole discipline. The reluctance, or refusal, to publish data, which allows independent examination of the work and reproduction of results, makes it even worse.

    It is hard to see how anyone thinks this is helping the cause. In the end it just promotes a general scepticism about Climate Scientists which lowers the credibility of both the people involved and the AGW hypothesis.

  4. Ian Castles
    Posted Aug 6, 2008 at 1:54 AM | Permalink

    Australia’s Prime Minister Kevin Rudd announced the dismal findings of the CSIRO/Bureau of Meteorology (BoM) report exactly one month ago. He said that the ‘historical assumption’ that exceptional or extreme drought [and] extremely high temperatures’ occurred once every 20 years had ‘now been revised down to between every one and two years.’ The Minister for Agriculture followed up some hours later with the statement that some of the predictions of drought in the new report ‘read more like a disaster novel than a scientific report’ – a claim that predictably made the headlines.

    David Stockwell has now found that all 13 of the GCMs used in the CSIRO/BoM study fail standard internal validation tests for regional droughted area in Australia over the last century, and that there’s no credible basis for the claims of increasing frequency of Exceptional Circumstances declarations made in the report. It seems that there was no external scrutiny of the report and that the authors did not even consult the statisticians in CSIRO’s own Mathematical and Information Services Division. Perhaps the reason that the CSIRO/BoM study doesn’t read like a scientific report is that it isn’t.

    But I can’t agree with Fred that reports such as this promote a general scepticism about climate scientists, or that they don’t help the cause. What cause? Since the report was released the Minister for Agriculture has told farmers in several States what ‘the best available science’ means for their regions, and has detailed the additional spending that the Government has decided upon in the light of the findings in the report. By all means go to David’s site and take up his invitation to review the CSIRO/BoM findings and his re-analysis of the data on which they are based – but the validity or otherwise of those findings has little relevance at the political level.

  5. Vincent Guerrini Jr
    Posted Aug 6, 2008 at 1:57 AM | Permalink

    I think this why the current government is thinking of closing down/downgrading the CSIRO

  6. Andrey Levin
    Posted Aug 6, 2008 at 2:18 AM | Permalink

    Re#1,2 Robert Wood, DeWitt Paine:

    This would be a basic thermodynamic stability model, where P,V and T, and R are all varying, under the force of a constant gravity.

    Me think that adiabatic model of atmosphere is more to the task than GCMs:

    http://freepdfhosting.com/d8ba65962a.pdf

  7. D. Patterson
    Posted Aug 6, 2008 at 2:26 AM | Permalink

    5 Vincent Guerrini Jr says:
    August 6th, 2008 at 1:57 am
    I think this why the current government is thinking of closing down/downgrading the CSIRO

    Yes, but close down and downgrade which parts of CSIRO, those which supported the theme of portending climate catastrophes or those which could threaten such themes of climate change catastrophes?

  8. Philip_B
    Posted Aug 6, 2008 at 6:12 AM | Permalink

    I searched Google news using ‘stockwell drought’ and got zero hits and therein lies the problem. The media is so invested in AGW that what would be a big story on almost any other topic gets ignored.

    I’ll email my MP directing him to Stockwell’s analysis. If politicians bring this up in parliment, then the media will find it hard to continue ignoring Stockwell’s work.

  9. Jonathan Schafer
    Posted Aug 6, 2008 at 7:00 AM | Permalink

    #8

    Searching for the terms enclosed in quotes returns a link to David’s blog page. But searching without enclosing in quotes returns plenty of hits. When you enclose in quotes, it’s going to search for those two terms right next to each other, which may not be how the text is laid out in any partiular web page.

  10. Philip_B
    Posted Aug 6, 2008 at 8:40 AM | Permalink

    I was searching google news (no quotes). 2 hits returned both unconnected to David Stockwell. I.e. no media reporting of this.

  11. Morry Weiskop
    Posted Aug 6, 2008 at 9:12 AM | Permalink

    Fred says “Climate Scientists are engaging with topics which require statistical treatment, but they do not themselves have the statistical expertise, and are unwilling to co-opt it from professional statisticians”

    Probably because of the fear that professional statisticians might tell them that their models are worthless.

  12. Michael Jankowski
    Posted Aug 6, 2008 at 9:23 AM | Permalink

    In the early days of RC, someone over there was talking about all the “unprecedented” drought conditions in Australia due to “global warming” according to news reports – and getting worse. I went to the BoM website, and they had both color maps and raw data to show there wasn’t anything unusual at all going-on relative to the rest of the 20th century. The poster then decided to get more specific and say the drought was especially bad in his/her neck of the woods, so I used more localized BoM data to show that it wasn’t very unusual at all in their specific region. It was just a matter of either looking at maps or downloading small text files and putting them in Excel to make graphs (no rigorous statistical analysis). But the poster wanted to believe what the news was reporting, rather than what the data showed.

  13. Mark T.
    Posted Aug 6, 2008 at 10:07 AM | Permalink

    I’ll email my MP directing him to Stockwell’s analysis. If politicians bring this up in parliment, then the media will find it hard to continue ignoring Stockwell’s work.

    As I always suggest in similar cases, send a letter, not an email. 100 times (very scientific number) more likely to get read, and appreciated.

    Mark

  14. Posted Aug 6, 2008 at 12:00 PM | Permalink

    But the poster wanted to believe what the news was reporting, rather than what the data showed.

    This is something that has been bothering me for quite a while, people want bad news. In fact, if you give them news that should make them happy, like GW or drought isn’t as bad as they thought, they are disappointed. Some people actually get angry. Why wouldn’t they breathe a sigh of relief hearing news like that?

    Isaac Crawford

  15. jcspe
    Posted Aug 6, 2008 at 1:50 PM | Permalink

    Steve,

    My apologies if I step over the line on your ban on policy discussions here. It is not really my intent. I just wanted to get some considered responses about when/where engineering should enter the AGW arena.

    I took Steve’s advice and visited David’s blog. In the comments (http://landshape.org/enm/agw-where-the-evidence/#comment-168414) I came across Neal J. King’s statement:

    But in the course of what Kuhn calls “normal science”, there is always “bad news” (inconsistencies) and “good news” (resolutions of inconsistencies): a theory that always and only has good news ceases to be an area of research (it becomes part of engineering), and an area that only has bad news becomes a construction zone for a new theory. I would say that climate science is far from being part of engineering, but has plenty of solidity as a real science.

    He followed that a few posts later and wrote:

    There are always discrepancies in ongoing scientific research, otherwise the area is designated as engineering.

    Anyway, it got me to thinking about Steve’s request for an engineering-level exposition of AGW resulting in x-degrees of warming in y-years. I began to wonder if the reason he can’t get such an exposition is that AGW has jumped directly from scientific research to policy implementation without ever having an engineering phase. If that is true, is AGW alone in that regard?

    If anyone can address the premise and answer that question without crossing the policy discussion line, please do. If not, let’s respect Steve’s blog home. Maybe we can have the discussion somewhere else.

    Steve: People make decisions all the time, not just in government, but in business, often based on judgment rather than engineering quality information. I’m sure that other precedents exist, perhaps in drug testing. But analogies don’t always help. In this issue, arguably the key “science” is settled; the problem is estimating parameters, which is the sort of thing that engineers do – with the added feature that they report carefully and comprehensively and not in little sound bites for Nature and Science. I don’t preclude the possibility that matters may be worse than people think. When people say to me, if the Hockey Stick is wrong, then the situation is worse than we think. My answer – well, then people other than me should be dissecting the HS and, by the way, why would you support the withholding and obstruction of efforts to examine the HS?

  16. trevor
    Posted Aug 6, 2008 at 2:06 PM | Permalink

    Andrew Bolt has covered the story at his blog here: http://blogs.news.com.au/heraldsun/andrewbolt/index.php/heraldsun/comments/dud_studies_behind_rudds_freakish_claims/

    But still no hits for – stockwell csiro – in Google News.

  17. jae
    Posted Aug 6, 2008 at 2:09 PM | Permalink

    jcspe: You can bring this matter up at the Message Board, if you want to hear from the Peanut Gallery. It is ironic that there are several fairly in-depth analyses by physicists that show no significant effects of CO2, but evidently none that show the opposite.

  18. Posted Aug 6, 2008 at 2:39 PM | Permalink

    I don’t think its fair to single out CSIRO. You need to identify the enemy — IMO bias and pseudoscience. There are targets for review everywhere. The public face of science has shifted from atom splitters to GHG accounting.

  19. Joe Solters
    Posted Aug 6, 2008 at 2:47 PM | Permalink

    Re: 15 When should ‘engineering’ enter the AGW arena? Quick answer. Assuming ‘engineering’ means proving a theory by using actual engineering concepts, real data, and thorough checks, then this stage must be completed before any climate theories are used to support significant public policy recommendations. In a few years, engineers, statisticians et cet will start closing the technical loop on climate science, and then , perhaps public policy issues can be addressed. Clearly, not today.

  20. Sam Urbinto
    Posted Aug 6, 2008 at 4:13 PM | Permalink

    When making decisions based upon incomplete information, the best information available (or actually, the least worst information available) should be used to make intelligent decsions based upon some sort of logical risk/reward cost/benefit analysis; do I spend 20 billion to sequester carbon dioxide, or do I spend it on shorelines, famine, disease, etc. How much more study of the issues is needed? Should I focus R&D on this, or that? Or not do R&D but use resources on solving issue X or Y? This isn’t about anything other than checking the information behind the conclusions so as to make these types of decisions; which is difficult to do when the information isn’t available.

  21. Ian Castles
    Posted Aug 6, 2008 at 5:24 PM | Permalink

    On 7 July the Australian Minister for Agriculture announced that ‘What we’ve decided today, in the light of the information we received yesterday [in the CSIRO/BoM report] is to more than triple the $15 million commitment to research and development in agriculture.’ Of course the Government might have decided to triple spending on R&D in agriculture anyway: the case for doing so doesn’t stand or fall on the CSIRO/BoM findings. But if David Stockwell is right that there’s no credible basis for those findings, the expenditure may well be misdirected.

  22. Jonathan Schafer
    Posted Aug 6, 2008 at 8:41 PM | Permalink

    #8, 16

    Yep, still no hits on Stockwell drought at Google or at Yahoo news either. But regular Google and Yahoo search turn up plenty of references. I suspect most people dont’ search just at news.google.com. In fact, I rarely go there for news, since their indexing engine for the web picks up changes so quickly. Most of the news ends up in regular google, if not all.

  23. Posted Aug 6, 2008 at 9:23 PM | Permalink

    Ian

    the expenditure may well be misdirected

    Particularly as its for pointless carbon equivalent reductions.

    The $46.2 million fund will deal with reducing greenhouse pollution, better soil management and adapting to a changing climate. Under reducing greenhouse pollution, the money provides an opportunity to look at ways – whether it be through nitrous oxide with cropping or methane gases in livestock production – of reducing an emissions profile in agriculture, without reducing production.

  24. kuhnkat
    Posted Aug 6, 2008 at 10:30 PM | Permalink

    [no policy please]

  25. Ian Castles
    Posted Aug 21, 2008 at 10:10 PM | Permalink

    On 5 August Steve McI opened this thread with a link to David Stockwell’s analysis of the CSIRO Drought Report, and expressed the view that key claims in that report ‘do not pass obvious statistical tests for “significance”‘. Steve urged CA readers to ‘Please visit David at his blog.’

    I don’t want to get in the road of the spirited discussion of a later analysis by David Stockwell that’s taking place on the ‘Koutsoyiannis et al 2008: On the credibility of climate predictions’ thread. But CA readers should know that, nearly three weeks after the publication of David’s analysis of the Drought Report, there’s been no response to his paper from any of the 11 authors of that Report – and very little from anyone else.

    There was some discussion at ‘Open Mind’ which was triggered by a post in which Lazar said that “This is the first time I am actually angry about … denialists pestering scientists, producing disinformation and setting themselves [up] as auditors in order to sell that disinformation.” After accusing Stockwell of selling “deliberately false information … with the intention of influencing the policies or opinions of those who receive it”, I hope Lazar felt a twinge of embarrassment when David prefaced his reply on his own blog with the comment “Thanks to Lazar for taking the time to provide the following feedback at Open Mind.”

    I urge CA experts who have the time to do so to study David Stockwell’s critique of the Drought Report and post their reactions at ‘Niche Modeling’. Meanwhile I’ll post below the text of an email that I sent to 23 Australian statisticians on 7 August. The letter is, I think, self-explanatory, but I do want to stress the importance of this matter. The Australian Minister for Agriculture has made several speeches telling farmers of the dire prospects facing their regions unless urgent action is taken to reduce GHG emissions. If these statements are based on findings that in the view of expert statisticians don’t pass obvious tests for ‘significance’, the farmers are entitled to know this.

    [Letter of 7 August from Ian Castles to Selected Australian Statisticians begins]
    I am writing to you as Accredited Statisticians who are recognised by the Statistical Society of Australia Inc (SSAI) as having achieved an acceptable level of professional competence in the understanding and application of statistical methods, and who are listed on the SSAI’s website as experts who wish their contact details to be made public. Your email addresses indicate that you are affiliated with the ANU (10), the CSIRO (9) or the ABS (4).

    Last Monday, Dr. David Stockwell, author of Niche Modeling and host of the Niche Modeling website, posted an evaluation of the validity of the modeling in the recently-published CSIRO/Bureau of Meteorology ‘Drought Exceptional Circumstances Report’ (DECR). This report is available at http://www.bom.gov.au/climate/droughtec/ , and Dr. Stockwell’s analysis is published at http://landshape.org/enm/effects-of-global-warming/#more-422 . He has posted his R programming and links to the data sets and invites others to verify the accuracy of his code and to carry out their own analyses of the data.

    The conclusion of Dr. Stockwell’s analysis is that all 13 of the models used in the DECR fail standard internal validation tests for regional droughted area in Australia over the last century. If that finding is accepted, there is no credible basis for the claims of increasing frequency of Exceptional Circumstances declarations made in the report.

    At the time of writing (Noon on 7 August, which is more than 60 hours after the posting of Stockwell’s evaluation and invitation) there has been NO response to that invitation. Nor, to the best of my knowledge, has there been any reference to this matter in the mainstream media – even in newspapers that reported the original findings under headlines such as ‘Climate change study like disaster novel …’ (The Age, Melbourne) or ‘Climate change report like a disaster novel, says Australian Minister’ (The Guardian, London). There have been several reports on blogs: Andrew Bolt of the Herald-Sun (Melbourne) made a posting ‘Dud studies behind Rudd’s freakish claims’ and attracted over 150 responses within 24 hours, and there have been posts on Climate Audit (‘Stockwell on CSIRO drought report’) and on the website of trade and public policy analyst Peter Gallagher (‘CSIRO’s drought model “fails”’).

    I know that you are all busy people, but I hope that at least some you can find the time to examine David Stockwell’s analysis and to post your assessment at Niche Modeling.

    The Government has already made important decisions on the basis of the CSIRO/BoM report: for example, the Minister for Agriculture, Tony Burke, announced on 7 July that ‘What we’ve decided today, in the light of the information we received yesterday, is to more than triple the $15 million commitment to research and development in agriculture.’

    Of course the Government might have decided to triple spending on R&D in agriculture anyway – the case for doing so doesn’t stand or fall on the CSIRO/BoM findings. But it is still important to know who is right: David Stockwell or Australia’s leading climate science research institutions. In speeches around the country (e.g., the Tasmanian Farmers and Graziers Association on 17 July and the NSW Farmers and Graziers Association on 31 July) the Minister for Agriculture is ‘selling’ the CSIRO/BoM findings, including the dire predictions for individual regions, as ‘the best science available’.

    I have pasted below, for your personal information, an article that I’m expecting to be published tomorrow in the e-journal ‘Online Opinion’. I understand that it is to carry the apt title ‘Scientists, Politicians and Public Policy.’ For statisticians (including official statisticians as well as those in academia, CSIRO and the private sector) it should be seen as a serious matter that the Government has accepted without question a report that does not meet ‘an acceptable level of professional competence in the understanding and application of statistical methods.’ Please note the reference in my last paragraph to the Accredited Statisticians list maintained by the SSAI.
    [Letter of 7 August from Ian Castles to Selected Australian Statisticians ends]

Follow

Get every new post delivered to your Inbox.

Join 3,249 other followers

%d bloggers like this: