Nice to know that, according to the latest issue of Science—the official newsmagazine of our profession's Washington lobby, the American Association for the Advancement of Science— climate science has so far avoided the fraudulent behavior that has cropped up in other "hot" areas where science intersects policy.

At least that's the position of Science editor Donald Kennedy, who used to be president of Stanford University. The reason for his seemingly downward mobility has to do with the Feds getting mad about National Science Foundation funds (i.e., our tax dollars) being spent on Stanford's yacht and his house. So he has some experience in this touchy area. Welcome to Washington: give the appearance of bilking the government from afar, resign rather than get canned, and welcome home.

Alas, though, experience may not be a sufficient teacher. In an April 18 editorial, "Research Fraud and Public Policy," Kennedy lamented the loss of reproducible data in two social studies projects on gun control and then opined that at least other politically hot sciences, among them climatology, weren't subject to such shenanigans. He wrote:

Other sciences engage political passions just as hot as those in the gun control controversy. For example, in analyzing the relationship between atmospheric chemistry and the only climate we have, we need solid scientific work.... So far, thankfully, passion has not overtaken truth in the actual doing of science, even in these hotly contested areas.

Oh, yeah? We consider passion to have over- taken reason when scientists either don't share their data or selectively use bits of it to paint a passionate, if politically correct picture. Here are three examples that might blow a bit of the bloom off Kennedy's science rose:

1995: IPCC Scientists Withheld Data from Reviewers. In 1995, WCR Chief Editor Patrick Michaels was a reviewer of the Second Assessment Report of the United Nations' Intergovernmental Panel on Climate Change (IPCC). One of the featured climate models was produced by the United Kingdom Meteorological Office (UKMO). A look at its output for early in the 21st century suggested that it was predicting much more warming than had been observed in recent decades in the high latitudes of the Northern Hemisphere—a critical area for climate change since substantial warming in this region reduces snowcover, which accelerates warming further.

An email was sent to the UKMO requesting intermediate output to see if, in fact, it was gravely overestimating current high-latitude temperatures. On May 11, 1995, head modeler John Mitchell replied that he "felt it inappropriate to send [Michaels] gridpoint data at this time." In response, Michaels reminded Mitchell that "Science is a cooperative effort in which information should be freely shared" and that the data requested was critical to a proper review of the IPCC report. (Mitchell, incidentally, was a senior author of the IPCC chapter on climate models.)

After failing there, Michaels sent five separate inquiries to the IPCC. Yet the model output was never forthcoming.

Years later, perhaps we know why. Another Met Office scientist, Myles Allen, adjusted their model with observed data and the resultant projection of warming for the next 50 years dropped to precisely the value WCR has always proffered: a mere 0.75°C.

Anyone who wants to read more on this should look at Michaels' testimony to the Subcommittee on Energy and the Environment of the U.S. House of Representatives, November 16, 1995.

1993: Ozone Scare Based on "Forgotten" Data. Remember the big splash 10 years ago, when two Canadian scientists, J.B. Kerr and C.T. McElroy, reported that the intensity of burning solar radiation had "increased by 35 percent per year in winter"? Their study included only five years of data, ending in March 1993—oops, make that four. Winter 1991–1992 (which showed no particularly high radiation readings) was not analyzed "because the instrument was out of service." Wrongo. In fact, the data were mislabeled in their Science paper. When they were included, the maximum increase in radiation dropped by 50 percent. What's worse, all of the "significant" trend was induced by four days in March 1993 at the end of their record. That just happened to be the same time as the "storm of the century," the deepest extratropical low pressure ever measured in eastern North America. It is well known that that phenomenon would disturb the ozone profile.

Every scientist knows to be careful about extreme values at the beginning or the end of an analysis, because such values often create "significant" trends that in fact are not real, once a longer data stream is analyzed. But, in their "passion," both the authors and the reviewers in, yes, Science missed that point. Nor did they notice that there was actually a lot of data in 1991–1992.

1996: Selective Use of Climate Data for Political Effect. It was no accident that Nature published the biggest bombshell paper on climate change ever written just a week before the United Nations was to meet in Geneva to lay the groundwork for the atrocious Kyoto Protocol.

The paper, by Benjamin Santer and just about every other big name in government climate science, demonstrated a rather striking match between observed trends in temperature from 5,000 to 30,000 feet and those predicted by climate models that also projected dramatic warming for this century. News of their article made the front page of every newspaper worth reading and appeared on every major news program.

The problem is, their match was obtained using only data from between 1963 and 1988, yet data from a longer period of record existed. Why did they select only the 1963–1988 period? See Figure 1. We don't accept any answer here but a clear attempt to paint a particular picture.

Figure 1. Temperature history from a crucial portion of the atmosphere, 1958–1995. For some reason, in the important paper "demonstrating" climate change, Ben Santer and colleagues only used the portion of the data circled here.

As a final thought, we draw your attention to the pull-quote from Kennedy's editorial in the April 18 Science. Though he was referring to gun data, which constituted "social" science, we find it useful to substitute just one word: "Sound [climate] science, not cooked data, is what we need" (bold in original). Thus amended, we couldn't agree more.

References:

Allen, M.R. et al., 2000. Quantifying the uncertainty in forecasts of anthropogenic climate change, Nature, 407,617–620.

Kennedy, 2003. Research fraud and public policy, Science, 300, 393.

Kerr, J.B., and C.T. McElroy, 1993, Evidence for large upward trends of ultraviolet-B radiation linked to ozone depletion, Science, 262, 1032–1034.

Michaels, P.J., F. Singer and P.C. Knappenberger, 1994. Analyzing ultraviolet-B radiation: Is there a trend, Science, 264, 1341–1342.

Michaels, P.J., and Knappenberger, P.C., 1996. Human effect on global climate? Nature, 384, 522–523.

Michaels, P.J., 1995. Testimony to the Subcommittee on Energy and the Environment, U.S. House of Representatives, November 16.

Santer, B.D., et al, 1996. A search for human influence on the thermal structure of the atmosphere. Nature, 382, 39–46.