intdots.gif (546 bytes)
white1.GIF (834 bytes)

Proving the Negative

Who says you can't "prove a negative"? In a recent issue of the refereed scientific journal Geophysical Research Letters (GRL) several WCR colleagues published a paper that gets pretty close to showing that the climate models being used as the basis for gloom-and-doom projections are simply wrong over the balance of the lower atmosphere.

We obtained this result months ago but have been impatiently sitting on it so as not to jeopardize publication in GRL, which, like most scientific journals, publishes only previously unreleased results. Truth be told, an oversight by a third party led to some of our results appearing on the Internet for a couple of days while our submission was in review for the journal Science. They rejected it, and rightly so, given their policy that a submission must "remain a privileged document and...not be released to the press or the public before publication."

(We find it somewhat ironic that shortly after our paper was rejected on these grounds, that same publication ran a piece describing some of the contents in the draft version of the yet-to-be-released IPCC Third Assessment Report, even though each and every page of the IPCC document was marked "Do not cite. Do not quote." Go figure.)

Still, we understood. So we expanded the study and sent it to GRL.

Our paper is a bullet through the heart of the global warming scare, which requires that the computer models used as the excuse for the United Nations' climate treaty match reality.

The well-known problem we examine stems from satellite and weather-balloon data for the balance of the lower atmosphere that appear to show very little warming during their period of concurrency, which is the 21-plus years (since Jan. 1, 1979). But the computer models all indicate that there should have been a dramatic warming. Literally billions of our tax dollars have now gone to try to explain away that discrepancy.

Finally, last March, newspapers around the world trumpeted—some on the front page—that new research by federal climatologist Ben Santer had reconciled the difference. Along with several co-authors, he argued in Science that computer models could account for the lack of warming after all, mainly because of the cooling influence of the 1991 eruption of the Philippine volcano Mt. Pinatubo.

Two things troubled us: Pinatubo wasn't the only big volcano in recent decades (El Chichón caused a cooling in the early 1980s of about half of the magnitude of Pinatubo), and the paper's data ended precisely at a very hot point—the mega–El Niño of 1998. (To those who think history repeats itself here—you recall correctly that these same guys tried using a different, highly fortuitous dataset a few years ago to prove that the models were OK. See sidebar).

The recent Santer Science paper argued that including Mt. Pinatubo's cooling effects made the difference between GCM-predicted temperatures and those measured by the satellites only 0.045ºC per decade, which they found to be statistically insignificant and which therefore led them to conclude that they had reconciled observed and modeled temperatures.

But when we add in all of the volcanic action (including the cooling from El Chichón) and allow for the fact that El Niños are pretty much random occurrences (in other words, a huge one happened to occur in 1998, which happened to produce this happy result), the difference between the models and the observed temperatures works out to a whopping 0.162ºC per decade, or 360 percent of the amount Santer and colleagues published in Science (Figure 1).

Figure 1.  The difference in temperature trends between the satellite observations of global temperature from 1979 to 1998 and the climate model output incorporating the Mt. Pinatubo eruption's interruption was reported by Santer and colleagues as only 0.045°C per decade.  But when the effect of ending their study during a strong El Niño is factored in, the temperature difference increases to 0.081°C per decade.  And when the other volcanic eruptions that occurred during this time are considered, the difference further increases to 0.162°C per decade—a value 360 percent larger than the one reported by the Santer team in Science.

Interestingly, this is almost exactly the difference in warming between surface temperatures and those of the rest of the lower atmosphere—proving, as we have maintained in these pages for more than half a decade now, that recent warming is confined to the very lowest layers of the atmosphere and, further research confirms, that it is largely confined to shallow, coldest air masses of winter that no one likes anyway.

In other words, the gloom-and-doom models don't work, which effectively leaves us with no scientifically based projection of 21st-century climate at all, except the projection that results from observed data and the laws of physics that dictate that human-induced warming should be relatively constant, rather than increasing in an alarming exponentiality.

That leaves us about 0.65ºC of warming to "worry" about for the next 50 years. That's the only conclusion we can take from the recent GRL paper. The models are wrong and nature has told us the answer.

Will we ever have a climate model that works? We think so, and we think we know how to "make" that happen. As NASA scientist James Hansen recently did, simply adjust the warming radiation down in the models to make them consistent with reality.

But that's called throwing in the towel, sending the champagne to World Climate Report, and finding something else to do for a living. Not very likely.


Michaels, P.J., and P.C. Knappenberger, 2000, Natural Signals in the MSU Lower Tropospheric Temperature. Geophysical Research Letters, 27, 2905–2908.

Santer, B.D., et al., 2000, Interpreting differential temperature trends at the surface and in the lower troposphere, Science, 287, 2000.


The Way Science Works

Longtime devotees of World Climate Report know of at least one other instance where data selection highly influenced the conclusion that the computer models were correctly simulating global warming. It was in the July 1996 paper by Santer et al. in which they compared lower atmospheric temperatures from 1963 through 1988 and found a statistically strong match. We examined their result in light of the complete record that was available (1957–1995 at the time of publication) and showed (Figure 1) that the main region of strong warming in fact showed no warming when all the data were used.

Figure 1.  In their 1996 paper, Santer and colleagues used only data from 1963 to 1988 (filled circles)—although data were available from 1958 through 1995.  A more complete record provides a clearer picture.

Thomas Kuhn, the late, great historian of science, wrote in his classic The Structure of Scientific Revolutions that such actions are in fact the norm in science when a "paradigm," or overarching logical framework, is assaulted by inconvenient data.

Maintaining the paradigm, he wrote, is the work of "normal science." In 1962, Kuhn wrote:

Closely examined, whether historically or in the contemporary laboratory, that enterprise [normal science] seems an attempt to force nature into the preformed and relatively inflexible box that the paradigm supplies.


In science…only the anticipated and usual are experienced even under circumstances where the anomaly is later to be observed. Further acquaintance, however, does result in an awareness of something wrong or does relate the effect to something that has gone wrong before.

What this means for climatology: The reigning paradigm is that computer models can simulate the behavior of the atmosphere. When data appear that show that they can't, the scientists' natural response is to ignore reality or to convolute the facts in a way that props up the paradigm. Thus the current tendency to either selectively cite data or to ignore inconveniences is, sadly, the real way that science works—until the entire house of cards implodes, which is what the recent Geophysical Research Letters paper might have accomplished.


Michaels, P.J., and P.C. Knappenberger, 1996, Human effect on global climate? Nature, 384, 533–23.

Santer, B.D., et al., 1996, A search for human influences on the thermal structure of the atmosphere, Nature, 382, 36–45.