Upon the advice of Roger Pielke Jr., who in a recent post at Prometheus praises the appearance of two new blogs, I checked out William M. Briggs, Statistician. Although the most recent post there, "Is climatology a pseudoscience?" begins with an intriguing premise, it eventually deteriorates into a sad self-parody by invoking the venerable Bob Parks' Seven Warning Signs of Bogus Science.
Here's how Briggs starts out:
...suppose, if you are able, that significant man-made climate change is false; further, that it cannot happen, and that all changes to the climate system are due to external forcings, such as those caused by changes in solar output. Just suppose all this is true for the sake of argument.
Now put yourself in the place of a climatologist, one of the many hundreds, in fact, who was involved with the IPCC and so shared in that great validator, the Nobel Peace Prize*. You have spent a career devoted to showing that mankind, through various forms of naughtiness, has significantly influenced the climate, and has caused temperatures to grow out of control. Your team, at a major university, has built and contributed to various global climate models. Graduate students have worked on these models. Team members have traveled the world and lectured on their results. Many, many papers were written about their output, and so forth.
But something has gone wrong. The actual temperature, predicted to go up and up, has not cooperated and has instead stayed the same and even has gone down. What do to?
This is a not-so-subtle reference to recent claims that global warming has stopped. It hasn't, of course. NASA says 2007 was the second-warmest on record, and as even Pielke has shown, all the major temperature records, including that of the RSS group and the Christy-Spencer data from the University of Alabama, show it hasn't. But still, interesting thought experiment. What if in a few years, that does turn out to be the case?
Well, nothing. Speculation is waste of time. Running sophisticated climate models on supercomputers is better use of same.
Anyway, then Briggs trots out Parks' Seven Signs. He only specifically mentions one (No. 3), but by introducing such a highly regarded concept, he implies (there he goes again, implying rather than asserting) that climatology is somehow less than respectable. So let's take Parks' list and see if any of the items actually apply.
1. The discoverer pitches the claim directly to the media.
I can't think of a reputable climatologist who does this. Only so-called denialists (whom I prefer to call pseudoskeptics) do this. Jim Hansen of NASA does write non-peer-reviewed summaries and letters (about which I will have more to say in a future post), but they aren't attempts to take science to the public by skipping peer review -- he does plenty of that.
2. The discoverer says that a powerful establishment is trying to suppress his or her work.
Not applicable in the slightest, as it's the pseudoskeptics who are swimming against the current of the anthropogenic global warming (AGW) consensus.
3. The scientific effect involved is always at the very limit of detection.
This is Brigg's issue. It is certainly debatable. But the IPCC has, with each new report, increased the degree of confidence its authors and reviewers have in AGW. After the politicians have had at the IPCC documents, what was language equivalent of 95 percent certainly ends up still a strong 90 percent.
4. Evidence for a discovery is anecdotal.
Ummm, no. One glacier is an anecdote. When 98 percent of the world's glaciers are in retreat, that's a trend. One warm year is an anecdote. When 11 of the past 12 years are the warmest on record, that's a trend. And when the warming trend begins with the onset of widespread fossil-fuel use and continues, with AGW-consistent interruptions, for 150 years, that's not anecdotal data.
5. The discoverer says a belief is credible because it has endured for centuries.
On the contrary. While the scientific principles involved date back more than a century, the AGW notion just 30 years of age.
6. The discoverer has worked in isolation.
I wouldn't call the 2500 IPCC's working conditions, "isolation," would you?
7. The discoverer must propose new laws of nature to explain an observation.
There are those who like to propose new mechanisms by which gamma-ray bursters millions of light years distant are responsible for changes in Earth temperatures, but the mainstream climatology community has no such need to change the laws of physics. See No. 5. and look up Svante Arrhenius.
So much for Brigg's attempt to cast aspersions on climatology, and so much for Pielke's blog judgment. Too bad.
- Log in to post comments
Hey James,
Actually, I was not trying to impugn climatology as it is today. In fact, I even start out by answering the question about pseudoscience in the negative. The seven signs do not hold now.
No, the entire post asks how can we recognize when and if certain climatologists lapse into pseudoscience *given* that man-made significant warming is false. I try to be careful and remind readers, about half-way through, that this assumption holds for the sake of argument.
Only if it is false would we have the potential for certain scientists who are so devoted to a beloved theory that they could fall pray to Park's 7 signs. (I am of course also tweaking Park, whom I love, but has become, as all old men do, crotchety.)
Now, if man-made significant warming is true, then of course, the potential for pseudoscience is negligible. We of course do not know for certain whether significant warming is true or false. Many say 90% true, or even higher. My belief is more in the 30% range, based on my work on forecast verification and the statistics of uncertainty in observations; meaning, naturally, that I think there is some chance that my conditional will obtain, etc. etc.
Also, I wouldn't actually say my post was "self-parody." I have a post like that, where I talk about a magic trick I use with my statistics students.
Oh, at my web page you can go the "Resume" tab for my bona fides.
Briggs
I am amazed, over and over again, that people with no training in atmospheric science, must less in climatology, presume to decide that they can figure out climate better than the scientists who have devoted years, and even decades to the study. Aren't they afraid of looking like fools? But I repeat myself.
Small point - when I take the time to add a comment to your blog which corrects an error of yours that flipped the meaning of a sentence 180 degrees, it might be friendly to see a "Fixed it ... thanks" comment from you rather than just have my append and the error vanish into the ether ...
Hey Mark,
Always dangerous to assume...
For ease, I thought I'd give you some bona fides. Like I said before, more complete details can be had on my resume.
I am a statistician: Ph.D. Cornell. My topic was forecast evaluation. My masters is in Atmospheric Science, also Cornell, with a topic on how to use climate forecasts. My Bachelors was Meteorology, and I actually spent a year as a weather forecaster for the National Weather Service.
I am on the American Meteorological Society's Probability & Statistics Committee; I am also an Associate Editor of Monthly Review; I regularly review (peer review) articles for all the major weather and climate journals.
I have a paper about to appear in the Journal of Climate on hurricanes (tropical storms), showing they have probably not increased in number or intensity. I just gave a talk at the Annual Meeting of the AMS, where all the bigwigs go, on an expanded version of this model: talk got a good reception.
I have published many theoretical and practical papers on how to best look at forecasts: generally, if forecasts do poorly, it is right to question their underlying models; if they do well, then, while you can still question the underlying models, it gives more evidence that they underlying models are adequate physical explanations. Climate models, at this date, anyway, are not yet making very skillful forecasts. I believe that someday (soon) they will, however.
Well, just thought I'd help out those who might wonder about my qualifications.
How about yours?
Briggs
It is disreputable to make arguments on behalf of a person who does not make those arguments, for the purpose of disproving them all.
just gotta throw it out there... back in the 70's Global Cooling was the trend touted by the NY times and other similar publications
http://denisdutton.com/newsweek_coolingworld.pdf
therefore, if without looking into the future, this politically motivated science shreiked and cried that the world was going to freeze over, and the same type of science backed it up... and it WAS TRUTH, then why aren't we talking about it now?
instead it was proven wrong and everyone jumped on the "global warming" bandwagon... which, as it seems is based on the same type of science... most of which is politically motivated... reasons to blame bush for katrina or excuses to expand government and whatever else GW is being blamed for...
but i ask this... if climatology touted global cooling, and now touts global warming...
doesn't that make it a false prophet? if it was wrong once, how can it reclaim validity?
it trys to explain it away by saying "global warming causes global cooling" this is the equivalent of "proposing new laws of nature to explain the discovery"
The worst part about the GW debate these days is that it is very emotional. Everyone you talk to seem to have staked their claim or bet on their team so to speak. So whenever you attempt to carry a reasonable debate, emotions immediately fly high, ad-homs are tossed about, and people huddle yet closer to their likeminded tribe members, chanting whatever mantra they have.
I think alot of the blame for this type of discourse stems from the celebrities and celebrutants that with not much more than backing by money and fame try to push their issue du jour, whatever makes them get a couple of headlines.
Actually, looking at this blog's other posts, it is clear that the author enjoys taking part in emotional debates.
James- Interesting interpretation ... I thought Briggs' post was a interesting riff on the philosophy of science. Especially after he answers his question on whether climatology is a pseudoscience with "The short answer, I will disappoint many of you by saying, is no." ;-)
Hi William. I have a PhD in atmospheric science from Georgia Tech. But I am not a climatologist and for that reason would not even consider trying to debunk the studies and conclusions of scores to hundreds of climatologists. Statistics is a wonderful area. I would not dream of trying to debunk a study by a statistician. But I would point out that climate is not weather. Climate models forecast in a different sense from weather forecasts. As to whether climate models do a good job of forecasting, do you say that because the effects of AGW seem actually to be worse than the more conservative forecasts? And do you think that all of the trends identified by analysis of existing data do not show an effect of AGW?
And lastly, do you consider yourself a climatologist?
Caligula - you are touting an urban myth. And even if the NYT had a story, that is generally not considered to be a peer-reviewed publication, under even the loosest definition of that term.
I didn't miss the opening answer to the question as answered by William. I just thought there was a significant disingenous element in the post, one that ventured into silliness by invoking Park's seven signs.
I acknowledge Williams' creds, but am not impressed by the approach he takes to the subject.
It is possible, of course, that I completely misread Williams' intent...
To Scott: thanks for catching the typo. I generally like to thank volunteer proofreaders by email, though, as I prefer to keep my comments area clean of administrative items. I make an exception in this case because I clearly ruffled your feathers. Sorry about that. Please continue to alert me to typos, especially those that constitute semantic reversals, by email. Thanks
No training or expertise here, just a curious nature. Definitely not willing to speak to the true topic at hand. However, I must point out to Mark P. that Caligula is not touting an urban myth. I have had a curious nature for many decades, and clearly recall the "coming ice age", in peer reviewed publications no less. I am not, however, old enough to recall first hand, and must rely on history books to relay the point, that it was, I believe the British Royal Society that espoused in the 1890s they knew all there was to know about physics, and merely had to just keep filling out the accepted mathematical formulae. Which provides a great anecdote for both A) good science can and must change tack when demanded (e.g. ice-age to warming), and subsequently B) consensus means nothing, only the data.
DPR, the history of climatology has some interesting twists and turns, including some earlier debate and apparent resolution about the effects of greenhouse gases. However, I would appreciate it if anyone could point me to a recent (last 30 years) peer-reviewed article that predicts a coming ice age. In fact, I would like to see a reference to any peer-reviewed article in the last 50 years that predicts an ice age on any basis other than a reference to periodic ice ages - in other words, any reference based on a predictive climate model.
To Mark P.
No, I do not consider myself a climatologist. And you're right that climate forecasts are not the same as weather forecasts. They are, however, in the same genre. That is to say, they both produce, in multi-dimensions, projections of various observable parameters. What statistics does is to measure the "distance" between the projections (in all the dimensions) and the observations in order to say which forecasts have skill and which do not.
Statistically, then, climate and weather forecasts are the same. Particularly easy to deal with are global mean temperature forecasts: these, generally, are a single number (both in forecast and observation; though the later is measured with some error, perhaps slight, but of enough magnitude to deal with, though it is shockingly nearly always ignored).
The statistics of forecast evaluation are, then, in some sense, "blind" to the source of projection and observation, and so we can bring the mathematical techniques of evaluation to bear regardless how the forecast was produced. If a model says "global mean temperature (defined as such and such) will increase in five years by three degrees C" then we can look at the global mean in five years and see if the increase was three or not. If not, then this might be cause enough to say the model underlying that forecast is suspect.
Right now, the verification of GCMs is largely unknown. Some work has been done; enough to show that predictions made are too confident (the error bounds are too tight and should be wider, say by a factor of 2).
So I argue, somewhat indirectly and teasingly in the post quoted, that a more thorough understanding of forecast bias and error should be understood. With your experience with similar dynamic models, I think you might agree.
Now, I have never argued that mankind has had no effect of any aspect of his environment, temperature included (see my post "What is the environment"): I instead argue that mankind *must* have an effect. the only real question is: how much? I do not think that all "trends" (particularly those touted about tropical storms; see my upcoming J. Climate paper on this) have been shown to have significant man-made signal. But I again emphasize that I do not think this is impossible: it is just not sufficiently proven.
Your first comment is the most worrying: just because "scores" of people say something, it, I think you will agree, does not mean that they are correct. Anybody can criticize, and each argument should be taken on its merit. The world of science has been far too compartmentalized: "Oh, you're not a statistician, eh? Then you can't comment on Briggs's critique of climate forecast goodness." I say, Oh yes you can.
Briggs
I don't think predictions about of the strength and number of hurricanes can be assessed particularly well at this time. Neither the strength nor the number could have been measured with any accuracy at all before satellite data, and even after that not so well. It is impossible to know either beyond a relatively short time ago. Even given how well we can do now, there are obvious problems in assessing strength without actually in-situ measurement. Also, the number of hurricanes is small and thus I suspect it will be hard to detect small changes. I leave that to the statisticians like you. However, temperature can and has been measured quite well, at least at some spatial resolution, for a long time, and vicariously for much longer. Since not many thoughtful people have asserted a real increase in either strength or number as real evidence of AGW, I think that is probably not the most profitable place to look. Temperatures are probably a better place to look, but even then a small signal is hidden in relatively large variation. Again, a good place for statistical analysis.
As to who can criticize, of course anyone can criticize. And I certainly hope you and other statisticians add your voices to areas like the statistical analysis of climate data. And yes, the validity of a model must be assessed by how well it predicts. There is no other real criterion. Since climate predictions are of necessity long-term, it might be a little early to think we can see results at a strong level.
I think it is presumptuous, if not arrogant, to think that someone untrained in climate modeling can simply jump into the debate and point out all the errors in climate modeling, like certain people (not necessarily you, but some who ought to know better), on the presumption that all the climatologists in the world simply overlooked a significant factor. For example, natural variation, solar activity, water vapor and all the rest of the denialist litany. And yes, simply because scores of people say something does not mean it is correct. But when scores of climatologists say essentially the same thing, the safe betting position is that they are probably onto something.
William,
What is the citation to the work which has shown that GCM error bounds should be widened by ~2x?
Question: why is Greenland called Greenland?
Extra credit: is Mars warming and, if so, why?
Further extra credit: Why does Antarctica refuse to fall into line?
Extra, extra credit: care to earn a quick $150K? The prize "will be awarded to the first person to prove, in a scientific manner, that humans are causing harmful global warming."
Doug. The bet of $150K on the "proof" that GW is anthropogenic will stay in the pocket of the bettor. Right now the consensus of the careful investigators is enough for rational action. If we take steps to increase efficiency, reduce waste and approach sustainability nothing is lost. My (multi) great grandchildren will appreciate our start now.
I can't wait around for the certainty that the "proof" requires before taking rational action.
Given that the noise in climate and weather are completely different how can anyone say they are statistically alike? If nothing else they differentiate themselves by persistence.
Global warming which may be largely caused by human activity is as political as it is a scientific matter.
Politicians of both sides in the democracies have adopted the AGW orthodoxy, intimidated by the supposed scientific 'consensus', the supposed authority of the IPCC and spurred on by hysterical media reports, but the public have rarely been given the opportunity to hear both sides of the matter put fairly.
Governments in Europe, North America and Australia have already introduced petty regulations and imposed taxes affecting the way people live but will have negligible effect on CO2 emissions and yet the science has not been settled.
James Hrynyshyn is pleased to refer to so called 'denialists' (an appallingly pejorative term given the historic allusion) as "pseudoskeptics" which is an improvement at least.
I would ask the following:
Would it be possible, in James' opinion, for a respectable climatologist to be genuinely skeptical of the hypothesis that CO2 is the main driver of current global warming, if so, are there any and who are they?
Mark,
Your questions and comments are so excellent that I will take them up in a day or two on my own site. The question "Who is allowed to criticize?" deserves a lot of thought. For instance, what gives non-climatologists, such as mere writers like Mr Hrynyshyn, the right to praise and support climatologists? (I hope, given that some tend to get emotional, that that last comment, while demanding an answer, is a joke.)
Mr Rabett,
Actually, all I claim is that climate and weather forecasts are the same, or should be treated the same way, statistically. And of course, it is impossible not to define climate without the use of statistical methods.
Briggs
Suppose a bunch of scientists proposed an experiment. They say that this experiment will take decades to a century or longer to perform and some results will take additional decades to see. The results could range anywhere from beneficial in certain areas through inconsequential all the way to catastrophic in certain areas. The number affected could be as high as millions, and the results could include virtually the entire surface of the Earth. The experiment will certainly include great economic benefits for a great number of people and will completely change virtually everyone's standard of living. The scientists can't say exactly what the rest of the results will be with great certainty, although many of them think the results will tend towards the catastrophic. If the results are catastrophic, the long-term cost will certainly be very high; potentially millions of people could be displaced, food supplies could be threatened and a number of species could be threatened with extinction. Once this stage of the experiment is reached, it will take about as long to reverse as it took to reach this stage, if the results can be reversed. Additional scientists, many not included in the proposal, think the results will not be catastrophic and the benefits will greatly outweigh any risk. Would you recommend that this experiment proceed? Would you participate?
but i ask this... if climatology touted global cooling, and now touts global warming...
doesn't that make it a false prophet? if it was wrong once, how can it reclaim validity?
Physicists once believed that the atom was absolutely indivisible. Then it turned out that it wasn't. Does that make physics a "false prophet"? How can it reclaim validity?
Have you ever made a mistake of any kind at any time in your life? If the answer is "yes", why should we listen to anything you say?
You see, science isn't revealed truth. It's a process of discovery. And the key to that is to acknowledge when you're wrong, and change your ideas accordingly. That's what makes the whole thing work. Far from being a weakness, it is science's greatest strength.
"Falsus in uno, falsus in omnibus" is completely stupid, all-or-nothing, black-or-white thinking. Most people manage to grow out of it.
Well, I can't answer for James, but as far as I am aware the answers would be "yes - if that was what the evidence showed", "no", and "see previous answer".
The argument over what was and wasn't said in the 70s is a long one.
Yes, there were published papers in the 70s that discussed global cooling - naturally enough, since there was a decades-long trend for cooling in place at that time, which we now believe to be the result of aerosol pollution - which we were also creating a lot of, and of opposite sign to the warming coming from CO2. A particular paper by Rasool and Schneider spectacularly over-estimated the impact of aerosols.
There was also a growing acceptance of Milankovitch's work, which carried the unmistakable conclusion that the orbital forcing had reached its peak and we should start sliding towards a new ice age. The primary paper making that claim included the time frame for that ice age (20,000 years) and an explicit statement about the possibility that "human interference has already altered the environment so much that the climatic pattern of the near future will follow a different path"...and CO2 was one such alteration specifically called out.
And there were plenty of papers published on global warming in the 70s. In fact, there were more papers on global warming in the 70s than there were on global cooling, by a ratio of 6:1. Citations to papers on global warming exceeded cites to those on global cooling by over 7:1. While the NY Times may have pitched a huge article on cooling, the developing consensus at the time was for warming, a consensus which has grown stronger with time as models, techniques, and evidence improve. So, yes, there is an urban legend aspect to the "global cooling of the 70s".
See this paper from last week's AMS meeting for more detail:
http://ams.confex.com/ams/88Annual/techprogram/paper_131047.htm
quote: Well, I can't answer for James, but as far as I am aware the answers would be "yes - if that was what the evidence showed", "no", and "see previous answer" unquote. Posted by: Dunc | January 30, 2008 8:17 AM
I hadn't realized that anthropogenic global warming had passed from hypothesis to immutable fact.
When did this 'eureka' moment occur?
Hey James, nice post! I just checked, and Roger Pielke has updated his IPCC projection/temperature measurement comparison for the corrected RSS data, and so might you!
Mark,
I finally finished, as promised, the article on "Who is allowed to criticize." It's at my site.
Thanks again for the idea.
Briggs
A small point made in the spirit of keeping the system honest -- or at least directionally so. When someone says that global warming has not stopped and offers as support the factoid that 2007 was the 'second warmest on record' according to NASA, this is a fine example of a 'cherry picktoid.' The writer disregards the fact that NASA alone finds this so; all other major keepers of the thermometers do not. Worse yet, any statement of the sort "200x was the n'th warmest" is by its nature meaningless; it signals either lack of knowledge or bias -- or both. It's only the trends that count, not individual years. And quoting regressed slopes is not enough either because any regression comes with a standard deviation of its slope. In this case it is a factoid that all major global average temperature records INCLUDE ZERO in their statistical envelope (conservatively defined by 1 standard deviation) over the past six years. Therefore, it is valid to say that global warming has in fact stopped over the past six years, whatever this may signify.
Galigula: "just gotta throw it out there... back in the 70's Global Cooling was the trend touted by the NY times and other similar publications"
I'd just like to point out that the NYT is not exactly the gold standard for climate science. Offhand, I wonder if there isn't a single field of science which wouldn't be embarrassed if we were to use NYT articles to judge it by.
Pelkor: "the worst part about the GW debate these days is that it is very emotional. Everyone you talk to seem to have staked their claim or bet on their team so to speak. So whenever you attempt to carry a reasonable debate, emotions immediately fly high, ad-homs are tossed about, and people huddle yet closer to their likeminded tribe members, chanting whatever mantra they have."
One side it doing science, the other isn't. The blame for emotion is pretty clear.
"DPR, the history of climatology has some interesting twists and turns, including some earlier debate and apparent resolution about the effects of greenhouse gases. However, I would appreciate it if anyone could point me to a recent (last 30 years) peer-reviewed article that predicts a coming ice age. In fact, I would like to see a reference to any peer-reviewed article in the last 50 years that predicts an ice age on any basis other than a reference to periodic ice ages - in other words, any reference based on a predictive climate model."
Posted by: Mark P
Mark, you should make this a standard challenge.
I'd also like to point out that Briggs' argument is basically arguing that the guys who have been right for decades now should not be believed because there is a non-zero chance that things have changed to make them wrong. Without offering good evidence that in fact things have changed.
It's the argument of a fraud - 'let's not dwell on *who* h as been right and *who* has been wrong, let us consider a hypothetical'.
For future reference, a few years back William Connolley put a lot of work into organizing and analyzing all of the scientific "global cooling" material. It's an excellent resource with all sorts of links for anyone who wants all the details.
He doesn't have much of the popular press stuff, but frankly nobody has ever been able to produce much more than a smattering of such (meaning that memories of a "panic" are wishful thinking). Most of this material was based on conflating the very minor short-term then-cooling trend with the then-new research that proved the Milankovitch-glaciation link.
This RealClimate article is a nice summary.
Also, note that the Rasool and Schneider paper mentioned above did indeed overestimate the potential cooling impact of aerosols (by 2X IIRC) but did so in the context of a 4X increase in aerosol levels that was postulated to be unphysical; i.e. the paper made no real-world prediction.