The Week That Was
November 16-22, 1998

With Candace Crandall recovering from the Buenos Aires political climate extravaganza, I thought you might be interested in science. Herewith my comments on two papers in the current literature that are likely to attract a great deal of attention by both promoters and skeptics of global warming. The authors include such icons of the GW-faith as James Hansen, Tom Wigley, and Ben Santer. Jim, an astrophysicist, has always been an independent thinker; Tom comes up with some surprisingly sensible views from time to time; and Ben is still a "martyr of the movement," for having altered Chapter 8 of the IPCC Report.. Not surprisingly, the two papers don't exactly jibe.

The paper by Hansen (and collaborators) is being hailed by global warming opponents as proof that climate predictions are meaningless. Supposedly, Jim Hansen, one of the earliest and strongest proponents of global warming, has recanted. Although a plain reading of his abstract might suggest so, this is not fair to Hansen. It is certainly not his intention to renounce greenhouse warming, but his paper sure doesn't give it much support.

The paper by Wigley, Smith, and Santer, due to be released to the press by Science magazine on November 27, is an attempt to shore up the ambiguous conclusion of the IPCC Report that "there is a discernible human influence on global climate." Their analysis is a dubious exercise in massaging the temperature data of both hemispheres (1880-1995) and contains a fundamental flaw - I think.

1. Climate Forcings are Uncertain -- So Where does this Leave Us?

Ref: J.E. Hansen, M. Sato, A. Lacis, R. Ruedy, I. Tegen, and E. Matthews, "Climate Forcings in the Industrial Age" Proc.Nat. Acad. Sci. USA 95, 12753-12758 (1998)

Hansen is well known as the scientist who was "99 percent confident" in 1988 that evidence for a catastrophic global warming was at hand. His current paper has been widely interpreted as stating that climate predictions for the next century are meaningless. To quote from the abstract: "The forcings that drive long-term climate change are not known with an accuracy sufficient to define future climate change."

Yet a careful study of his paper leads to a somewhat different conclusion. What he claims is that the uncertainties of forcing are greater than the uncertainties of "climate sensitivity" (defined as temperature rise for doubled CO2). But he then waves his arms and tries to reinforce the IPCC mantra that temperature data and climate models are "broadly consistent."

His discussion of aerosol forcing is more sophisticated than that of other modelers (most of whom consider only the direct scattering effect of manmade sulfates). But he arrives at a climate sensitivity of about 3 deg C, without much discussion about the treatment of clouds or upper-tropospheric water vapor, based on indirect paleoclimate evidence. (He assigns the 5 deg rise since the last ice age to a forcing of 7.5 W/m2.) He then adjusts his aerosol scenarios in a fairly arbitrary way to give the observed temperatures since 1940 (taking no account of "fingerprints" or even of NH/SH differences).

There are other surprising features of his treatment. Ozone depletion, which plays such a large role in climate effects in the Hadley model (of the British Met Office) and in Hansen's earlier publications, is barely mentioned here. Nor do we learn that solar effects, being cyclic, must average out to zero in the long run. Aerosol effects, too, cannot build up too much because of their short residence time in the atmosphere. Thus by choosing a high value of climate sensitivity -- arbitrarily, I think -- Hansen virtually guarantees a large future temperature rise.

What to make of this paper? On the one hand, it says that climate change cannot be predicted. But on the other hand, it provides enough free parameters (through complicated and uncertain aerosol forcings) to fit the standard GCMs to the observed temperature record. (Notice that Hansen is really in conflict with the IPCC Report, which claims a similar result but with entirely different assumptions.)

2. New Analysis of Global Temperature Claims Human Fingerprint

Ref: T.M.L. Wigley, R.L. Smith, B.D. Santer. "Anthropogenic Influence on the Autocorrelation Structure of Hemispheric-Mean Temperatures" Science 282, 1676-1679 (Nov. 27, 1998)

Yet another attempt to demonstrate a "discernible human influence" by massaging the same old tired data set. This time round, Wigley et al (two of them lead authors of IPCC Chapter 8 that provided meager and disputed evidence for such an "influence") have tried lagged auto-correlations (and cross-correlations between hemispheres) of the temperature data from 1880 to 1995. These correlations are shown to differ from control-runs of two coupled ocean/atmosphere GCMs that are run "unforced" (i.e. with no increase in GHG forcing) and that are thought to encompass all of the natural variations of climate.

This difference between observed and modeled correlations is supposed to be a "fingerprint" of human influence. That's about it.

The problems with this labored approach (trying to prove a preconceived hypothesis) are manifold. Ignore for a moment the poor quality of the data before 1945, esp. in the SH, and the idea that existing GCMs can simulate all important natural climate fluctuations. The real problem is that there was an unusually large and sustained warming between 1880 and 1940, undoubtedly a natural recovery from the preceding Little Ice Age, as seen in all climate records.* This 60-year warming period is not likely to be representative of the so-called natural variations shown in the GCM runs.

Thus the WSS conclusion may simply be an artifact of their analysis. An obvious check would be to divide the data set and analysis into two periods: 1880 - 1940, when the human contribution to atmospheric GHG level was minor; and 1940 - 1995, when most of CO2, CFCs, etc was released.

Perhaps the referee who reviewed the paper did make this obvious suggestion -- although the paper seems to have gone through rapidly: submitted Sept 16 and accepted Nov 2.

This issue of TWTW was prepared by Fred Singer. Candace Crandall and her insouciant news commentary will be back next week…

* There is no agreement about the post-1940 temperature data: a decided cooling between 1940 and 1975 (according to publications by Jim Hansen, Tom Karl, and others) -- while the IPCC showed little change. And finally, there is also a conflict (between 1979 and 1995) between the independent balloon data and satellite data (showing no warming) and the surface data (showing a substantial warming, likely due to contamination from urban heat island effect). But that's another story. For more information, read Hot Talk, Cold Science.

Go to the Week That Was Index