What’s the purpose of auditing surface temperature data?
Like many, I try and follow the latest in the global warming discussion. I’m not a climate scientist…I don’t pretend to be. When it comes to the general conclusions regarding the measurements and attribution of global warming, I accept the IPCC’s statements (more or less). Do we know everything about how the climate system works? Of course not. But, we have to try and pull all the available information together and make reasonable interpretations. That being said, I don’t necessarily and wholeheartedly agree with all the policy recommendations. I’m not sure what the best course of action is…but the policy discussion is the discussion to be having now.
Once in a while I head over to Climate Audit just to see what they are up to. Essentially it is a group of statisticians, engineers, and computer scientists…from what I can gather. If you’ve never been there, it is difficult to get a good overview of what they are doing. The FAQ is rather narrow and requires some digging into the history of why the site was started. I once asked if they could produce a high-level ‘executive summary’ and was quickly scolded and chased away. And it seems someone else in recent days inquired about a similar thing…and they told that person to get lost. So, it seems the blog is more of a forum for those involved in the data-mining and statistical analyses…it’s not for the outside reader. It’s not meant for general consumption. That’s fine, I have no problem with that. Maybe they should put that in the FAQ.
In the last few weeks, these ‘auditors’ have been painstakingly sifting through surface temperature data used by NASA GISS to calculate and report longer-term trends. Firstly, I’m not a statistician and can’t evaluate what they are doing in any technical way and, secondly, I do not have the time to try and figure it out. So, as an outsider, I try and get an idea from the post and the associated comments. A good portion of the comments are very celebratory in tone, as if they have torn down the very foundation of our understanding of global temperature trends. Indeed, they may be finding actual errors, and they may be rightly confused about NASA’s methods for adjusting surface temperature data. Most of the general comments regarding what they think of climate scientists in general, and Hansen and Mann of NASA specifically, have a tone similar to this:
I’m beginning to think that one of the main divides is that climate scientists have no idea what an engineering study looks like. They are only familiar with little articles in journals (or literature reviews.) This thing cries out for an engineering study.
I really hope that they work on communicating what it is they are doing, and why it is so important to a general audience. If their work has the implications they all think it does, it behooves them to communicate it effectively. For all their criticism of the RealClimate blog (NASA’s team), which is quite harsh, RealClimate at least has a very thorough index and ‘start here’ page communicating what they are doing. Actual audits typically create a short (1-2 page) summary of their findings…should CA do the same?
So, to sum it up…I’m not against what they are doing in principal. I say go for it. More scrutiny is always good in science. But, is this really for the benefit of our understanding of climate science?
Oh, and by the way…remember they are now claiming that the surface temperature data is, at best flawed and, at worst, completely unreliable as an indicator of global trends. Below is a plot (from Open Mind) showing the relationship of surface temperature data and satellite data. I suppose they’ll have to audit the satellite data now too.
This is being discussed in much more detail here.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Clastic Detritus 
For further insight, I recommend:
1) Allan Brandt’s “The Cigarette Century” pp306-307 regarding “sound science” in which the “tobacco industry responded by criticizing the EPA for inadequate and poorly analyzed data.”
2) Chris Mooney’s “The Republican War on Science”: look up Data Quality ACt and how it has been often abused to raise the bar so high for inconvenient results as to keep them from being used.
3) You might want to ask yourself: Steve McIntyre is Canadian, so one might expect that the first priority might be to audit Canadian data, especially since:
a) Canada is (slightly) larger than the US (total, including Alaska, which probably would make more sense if aggregated with Canada).
b) Consensus results expect it to warm faster further North.
For both of these reasons, if there is global warming, Canada (& the Arctic, and Greenland, and Russia) would contribute substantially more to the warming than the US, which is not expected to warm so much.
Re your questions about the purpose of Climateaudit
a) firstly about the primary purpose of the blog seems to be to bring a knowledge of mathematics and statistics to the issue. This is knowledge that is perceived to be sorely lacking in the climate science community. The fist issue addressed by Climateaudit was the infamous Mann hockey stick curve. Climateaudit showed that the derivation of this curve was based on the use of incorrect mathematics with the use of proxies that did not contain any valid temperature information. The whole area of paleo-temperature extraction was shown to be rife with ignorance of the appropriate mathematics and the use of unsound proxies
b) secondly the contributors of Climateaudit are not restricted to statisticians and engineers. There are quite a number of self-identified economists, physicists and geologists who contribute. Economists contribute because there is a large body of mathematics that has been developed in econometrics to discern the same sort of trends that climate scientists are trying to discover. Climate scientists instead of taking advantage of this knowledge and have insisted on attempting to develop their own techniques. Climateaudit has shown this to be unsuccessful due to the lack of mathematical skill in the community and the refusal to take advantage of knowledge outside of the their own small numbers.
c) Thirdly, the suggestion for an engineering document refers to the inadequacy of the current IPCC AR documents. These documents rely on the per-reviewed literature of the climate science community. It has been demonstrated that his community is small and lacking in the necessary mathematical skills. Thus peer review in this community is relying on those lacking in necessary skills reviewing the work of each other.
The purpose of auditing surface temperature measurements is to make sure hat they do not suffer from the same inadequate mathematics and mistakes that have affects the hockey stick and other climate science work. The recent discovery of a simple programming blunder that rendered the US surface temperature record incorrect from 2000 to 2007 illustrates this point. This was an inexcusable blunder that was made worse by the fact that it went undetected for seven years and had to be discovered by an outside party.
This does not inspire confidence in the climate science community’s skills and the effectiveness of their peer review process. To trust this community to develop knowledge on which to base policies to manage the world economy would be a foolish and incorrect thing to do.
“The purpose of auditing surface temperature measurements is to make sure that they do not suffer from the same inadequate mathematics and mistakes that have affects the hockey stick and other climate science work.”
That’s all fine and good…but, the site should (1) make the purpose very clear up front; make this blindingly obvious to the general interested reader, (2) put up a list of the papers and studies that are currently being audited; for example, what does ‘other climate science work’ include, (3) put together and outline of all the papers where CA has demonstrated inadequate mathematics, and (4) comment on the significance of said inadequacy.
TAG says: “The whole area of paleo-temperature extraction was shown to be rife with ignorance of the appropriate mathematics and the use of unsound proxies”
Which proxies are sound?