Warming Is Unequivocal

by Bill Chameides | February 23rd, 2010
posted by Erica Rowell (Editor)

Permalink | 7 comments

Is the Wall Street Journal making like an ostrich and sticking its head in melting permafrost?

If you haven’t already, check out the editorial page from yesterday’s Wall Street Journal. On it you’ll find a spirited, one might say angry piece by L. Gordon Crovitz entitled “Climate Change and Open Science: In the Internet age, transparency is the foundation of trust.” The piece riffs off a BBC interview with Phil Jones, the embattled director of the Climate Research Unit at the University of East Anglia, to reach its inevitable conclusion that “equivocation has replaced ‘unequivocal’” in the climate science world.

That word unequivocal in this context carries symbolic meaning, having appeared in the latest assessment of the Intergovernmental Panel on Climate Change: “warming of the climate system is unequivocal.”

The word was included in the report after much discussion and debate and, as I understand it, was retained over the objections of some politicians at the insistence of the scientists. Now the climate skeptics, including apparently the Wall Street Journal’s editorial board, would like to use the climategate incident to take the ‘un’ out of unequivocal.

Is the WSJ Editorial Page Getting Yellower?

As noted in yesterday’s TheGreenGrok, the media coverage of climategate has not been exemplary of journalism at its best. Now outlets like the WSJ are using the occasion of the Jones interview to pile on — using quotations taken out of context to press their attack on climate scientists. RealClimate has a nice piece on how the Daily Mail inaccurately spun the Jones interview to undermine the science.

The WSJ’s editorial follows a similar path. Here is one example.

The piece states: “Phil Jones … acknowledged to the BBC that there hasn’t been statistically significant warming since 1995.” OK, that is what he said, sort of. Why sort of? Because the WSJ conveniently neglected to include the context.

Actually, Jones said that there was a warming trend in global temperatures since 1995 — at a rate of 0.12 degrees Centigrade per decade (or 0.2 degrees Fahrenheit per decade). But because of the short time period, the rate was not quite statistically significant at the 95 percent confidence level. It was “quite close to the confidence level” but not quite there.

Newspaper Decontextualizes Quotes, Misleading Readers

Simply put, a 95 percent confidence level means that there is a 95 percent probability that the actual temperature trend was positive and a five percent probability it was not. The 95 percent confidence interval is referred to by statisticians as “2-sigma” because it covers two standard deviations from the mean.

Scientists often choose the 2-sigma or 95 percent confidence level, instead of the 1-sigma or 68 percent confidence level, to be conservative in their conclusions. Note: climate scientists are being conservative in their pronouncements about global warming, not the other way around.

Here’s the point, Mr. Crovitz: saying that there was no statistically significant trend is not the same as saying the temperature trend between 1995 and today was positive but not significant at the 95 percent confidence level — in fact it was statistically significant at a slightly lower confidence interval. Doing so can be particularly misleading when reporting to a public that is not aware that scientists commonly use a 95 percent confidence level to establish statistical significance. It is especially misleading since the lack of statistical significance in the trend at the 95 percent confidence level was related to the shortness of the time period over which the trend was calculated.

I am left with three possible inferences from the Crovitz piece:

  1. He never actually read the transcript from Jones’s interview and just cribbed from the Daily Mail. Tsk tsk.
  2. He does not understand statistics, in which case what is he doing writing about science?
  3. He has intentionally misled his readership.

I suppose there is a finite probability that I have got it wrong, and there is another explanation. My confidence level is only 80 percent so I guess I’m being a bit unconservative in this instance.

It Is Unequivocal — Just Look at Glaciers, Sea Ice, Permafrost, Earlier Springs …

So much of the arguments about global temperature trends focus on how to interpret temperature records from myriad weather stations. Indeed, a lot of the climategate controversy surrounding Jones was about a paper he wrote in 1990 trying to quantify the influence of urban heat islands based on weather stations in China.

The debate over the temperature record is important for establishing the magnitude of the warming but is unnecessary to establish that the globe is warming. The globe integrates the temperature signals from all those individual stations and provides very obvious, large-scale signs of climate change. What are those signs? How about melting glaciers? How about shrinking sea ice? How about earlier arrival of spring? How about melting permafrost?

And speaking of melting permafrost: this just out from the journal Permafrost and Periglacial Processes. Authors Simon Thibault and Serge Payette of Laval University in Quebec report on a study of permafrost extent near the James Bay area of Quebec.

Using a combination of aerial and ground surveys and historical aerial photographs, the authors concluded that permafrost in the region had retreated northward by about 130 kilometers (80 miles) over the past 50 years. They note that the changes they found are similar to the findings of numerous other permafrost studies, such as here and here. (See related article and photographs here, here, here, and here.)

Don’t believe it? Congratulations, you’ve joined the ranks of the skeptical ostriches with their heads buried deep in the melting permafrost.

filed under: climate change, faculty, global warming, science
and: , , , , , , , , , , , , , , ,


All comments are moderated and limited to 275 words. Your e-mail address is never displayed. Read our Comment Guidelines for more details.

  1. Ken Towe
    Mar 3, 2010

    Ken: what is statistically significant in a scientific sense and in a public policy sense are quite different. And Phil Jones, in comparing 1910-1940 and 1975-1998, was comparing warming RATES not temperatures. Some ecosystems may be “repeating themselves” but others are assuredly not, e.g., melting permafrost. I suspect that had the Jones rates been significant in a scientific sense it would have “vindicated” the “settled” view that today is different from the recent past. Again, in 1975 the “normal” was 15°C for the northern hemisphere and the late 30s were plus 0.6 = 15.6°C. Not so today. The Jones et al. NH “normal” is 14.6°C and the late 30s are now shown as only ~15.1. What happened? According to Jones et al. the entire period from 1850 to 1940 was BELOW normal. But in 1975 this same period was essentially “normal”. Please explain. Then we can analyze the meaning of rates and the melting arctic today and in the 1920s…

  2. Thomas Backhaus
    Feb 27, 2010

    Maybe one should consider the actual meaning of any confidence belt, and that is, that there is a 95% chance that the trend is actually positive, GIVEN THE AMOUNT OF AVAILABLE DATA. Meaning that “the warming was not significant on a 95% level” is fundamentally different to “there is no warming”. There can very well be a trend, only that we are currently (!) not able to prove that it is there. And maybe it’s not the best option to simply wait until it becomes significant… Sadly enough, I can only agree to the general statements in your post.

  3. Hank Roberts
    Feb 28, 2010

    Cherrypicked timespan, and Jones should’ve called him on picking an interval short enough to barely fail the 95% test and suggested using at least 30 years. Bogus claim, bad interview response, and s a result, creation of a lovely cherry being reposted many places like this “Frontiers of Freedom” ( politics blog. Do better, please.

  4. Ken Towe
    Feb 25, 2010

    Most statisticians agree to accept a certain level of confidence before starting their experiments, not after. The majority of scientific studies use the 95% confidence level as the de facto choice. Thus, a result is either significant or it is not. It might be “not quite significant” or “almost significant”, but from a strictly statistical point of view it remains non-significant… or vice versa. Jones said this: “Achieving statistical significance in scientific terms is much more likely for longer periods, and much less likely for shorter periods.” It is true that small samples are more likely to be misleading, generating false positives… or negatives. One could easily have said “achieving statistical non-significance…”. If climate scientists were to be really conservative they might have chosen the 99% confidence level. To say that they are conservative because they didn’t use 68% might be just a bit disingenuous? Yes, we can be “100% confident that the climate has warmed” and that “warming is unequivocal” but can we be as confident that this recent warming is unequivocally different or statistically significant from earlier similar warming periods… e.g., the 1930s? Phil Jones said: “As for the two periods 1910-40 and 1975-1998 the warming rates are not statistically significantly different.” The real question is a bit more than whether it is unequivocally warmer today, on average, than it was 30 years ago. One example: One conclusion: “We are presently experiencing a warming trend, not only in the northern North Atlantic but globally. Several of the ecosystem changes observed during the 1920s and 1930s in the northern North Atlantic appear to be repeating themselves.”

    • Bill Chameides
      Mar 2, 2010

      Ken: what is statistically significant in a scientific sense and in a public policy sense are quite different. And Phil Jones, in comparing 1910-1940 and 1975-1998, was comparing warming RATES not temperatures. Some ecosystems may be “repeating themselves” but others are assuredly not, e.g., melting permafrost.

  5. Greg Jones
    Feb 24, 2010

    Great article. I am looking for existing established companies with an edge in environmntal economics. Until we can show that eco friendly manufacturing and products are more cost effective then the alternative the simple law of Value vs Price will doom environmental efforts. Taxes and policies that are not universial will also fail. Manufacturing will simply move to China or some other 3rd world company when tax or policies are not in effect. I believe there is an oil company that has found a way to recicle tires and produce oil at a cheeper cost then what comes out of the ground (may have been natural gas). If this is true it is prime example of environmental economics at its best. Waste dumps producing natural gas to heat and supply energy is anoter example. Is there an easy place I can go to get information on successful environmental economics? Thank you for your time (please forgive any spelling errors)

    • Bill Chameides
      Feb 25, 2010

      Greg, Check out Chad Holliday et al’s book entitled “Walking the Talk: The Business Case for Sustainable Development” and this book by Daniel Esty and Andrew Winston: “Green to Gold: How Smart Companies Use Environmental Strategy to Innovate, Create Value, and Build Competitive Advantage.”

©2015 Nicholas School of the Environment at Duke University | Box 90328 | Durham, NC 27708
how to contact us > | login to the site > | site disclaimers >

footer nav stuff