A cooling pause

A mathematician may say anything he pleases, but a physicist must be at least partially sane.

- Gibbs

Before taking a climate holiday, let's pause to summarize the months of blog discussion on climate and climate change.

A snapshot of the theory. The major thrust of theoretical discussion here has been to spell out how rising infrared (IR) opacity from IR-active gases such as carbon dioxide (CO2), methane (CH4), and water vapor (H2O) might change Earth's atmosphere with regards to temperature and water distribution. The focus has been on pinning down qualitative trends, rather than worry about specific numbers that are very difficult to predict.

The theory of Earth's climate (from the point of view of just the atmosphere) requires three physical processes: radiation and radiative transport; air turbulence and heat convection; and water phase transformations. The first is well understood conceptually and quantitatively and can be subjected to controlled approximations of different types, including - but not limited to - numerical (computer) models.

The second and third are poorly understood and are approximated by uncontrolled methods. Converting these into numerical methods calculated on computers in no way mitigates these deficiencies. It just means calculating wrong answers more reliably and quickly.

Two extra-atmospheric effects will have significant impact on the CO2 concentration, absorption by the oceans (understood somewhat) and by plants (poorly understood).

The "global warming theorem" and why it fails. If we confine an atmospheric model to radiative transport alone, we can prove a theorem: increasing IR opacity must lead to a steeper temperature lapse rate. Holding cloud cover constant, we can also conclude that the temperature overall must increase at every altitude, with the increase being largest at the surface and declining until it reaches zero at the top of the atmosphere. The "global warming" craze has no other scientific basis than this. Computer climate models get this part - and only this part - right.

Better than the misleading metaphor of the greenhouse is that of a closed attic. It gets hot in an attic (hotter than the air outside) because the attic has no way of exporting heat other than to radiate in the IR. But suppose we open the attic windows, loosen or take off the roof, and hose it down with water?

Then the theorem fails: increased temperature is no longer a "must." The atmosphere, like the attic, has two other ways of sending heat flow on its way upward, turbulence and convection, and water phase transformations. It's not difficult to prove that convection must be enhanced in such a situation. The hard part is pinning down by how much.

The case of water is even more difficult. But the amount of water below the upper atmosphere is fixed. The rate of evaporation from the surface will increase. The water has nowhere else to go, so precipitation and therefore cloudiness must increase. Again, the hard question is, by how much.

Both of these atmospheric responses (convection/turbulence and water phase changes) will blunt the original tendency of the temperature and its lapse rate to increase.

Empirical snapshot. The observational situation is less clearcut, but not favorable to the hypothesis of significant human-induced warming.

The spatio-temporal patterns of the last few decades do not match what we should expect in the IR-opacity-dominant case. There's no systematic narrowing of daily or seasonal temperature differences. And while the Arctic might have warmed somewhat in the last three decades, the Antarctic has cooled, and there's no narrowing of equatorial-polar temperature differences. There's been no long-term fall-off of tropical cyclones. The most striking spatio-temporal fact is the lack of any systematic steepening of the temperature vertical lapse rate, the most fundamental signature of enhanced IR opacity.

When considering global temperature trends, use extreme caution: spatial averages of a temperature field are not legitimate, and climate needs at least a humidity variable as well for a complete description. Nonetheless, thermometers, satellites, and biogeo-proxies are valuable for indicating temperature trends, if spread out enough in space and time and provided we take care to compare proxies, not crunch them all together into a fictitious average. Also reject the misuse of composite time-series extrapolation methods like the "hockey stick" and similar mistakes; use the year-by-year measured temperatures, especially for the last 15 years, after the most recent warming period ended. This issue is thoroughly explored by ClimateAudit. The striking result is a fatal double blow to the anthropogenic "global warming" hypothesis.

In the short term, over the last century, there is no simple trend of upward temperatures. Instead, there is a pattern of up and down, with the strongest warming occurring before major CO2 emissions started after World War Two. The most recent trend, of the last 15 years, is neutral-to-cooling. There's no obvious human "thumbprint."

In the long term, over the last 400 years, there is a clear warming trend (with smaller short-term wiggles), with glacial retreat and rising sea levels. But this trend has been going on too long to be attributed to human activity. Again: no obvious human thumbprint.

The false politics of "consensus." To cut to the heart of the controversy requires facing the reality that the IPCC is a small minority of scientists. The active group issuing the hysterical summary reports under the aegis of the UN is about 50, with a staff of a couple hundred. They are apparently driven by faulty preconceived dogma and have mainly succeeded in manipulating the name of much a larger group of scientists - the couple thousand associated in some way with the IPCC's scientific working groups or perhaps all climate scientists - to make it sound as if the latter back the "global warming" craze. As the IPCC has gotten more aggressive in its claims, so more and more climate scientists have publicly rebelled: the Revolt of the Captive Ventriloquists' Dummies, to use an earlier metaphor. But by getting the ear of funding agencies and the editorial boards of scientific journals, the IPCC has done disastrous damage to the geosciences, climate science, and meteorology. Twenty years have now been wasted on wrong questions and dead-ends. Distorted funding and sociological incentives have enticed younger scientists to ride the bandwagon for careerist reasons.

Add to this the persistent gap between the caution and skepticism of the science reports, both from the IPCC and the US federal government, and the hysterical conclusions of the summaries for non-scientists and policymakers. A recent, striking example occurred in connection with the non-increase of the vertical lapse rate.

The IPCC is also responsible for foisting on scientists a false ideal of "consensus." Science doesn't work by consensus. When something is really solid, scientists virtually all assent to it because of its evident truth. When something is murky, controversial, or faulty, consensus is missing for a good reason: no one really knows. Good scientists swarm all over such challenging questions. If preconceived dogma is not enforced, and everybody works hard and honestly, progress results.

Hiking map for the final stages of the climate journey. After climate vacation, the journey will take us over two final high ridges of scientific thought with essential vistas on how much we don't understand about climate.

The first ridge to cross is the double approximation embodied in climate computer models: the controlled approximation of their discretized spacetime grid, and the uncontrolled approximations inherent in their treatment of convection, turbulence, and water phase transformations. We've looked at the latter somewhat already, but they are used in climate models in a peculiar way worth a brief look.

The second ridge is the thrilling intellectual achievement often labeled "chaos." We'll look hard at what it tells us about climate and, indeed, whether "climate" means anything. In the context of computer models, chaos is usually filtered out by design, by missing very-long-term behavior. But no attempt to predict climate that ignores chaos is credible.

As we descend, we'll make a final scientific basecamp at the question of climate cycles, which overlaps with chaos in connection to very-long-term climate change. Climate cycles are real but not sufficiently studied or understood. They range as blocks of time over the tens-to-hundreds of thousands of years (ice ages) to decades-centuries-millennia (solar luminosity cycles) to years (atmosphere-ocean oscillations), with the latter two essential to understanding climate change in recent history.

The journey ends back in the everyday with three final questions. The science comes first, starting with the climate cycle question: where should climate science be, and what went wrong in the last generation? How should society outside the science be supporting it, and what should society be looking for? And, given the apparently mild risk of human-induced climate change in the next couple centuries, what countermeasures can and should we take - and what should we avoid?

POSTSCRIPT: Here's a roundup of good climate blogs and books. McIntyre's ClimateAudit should win some kind of prize. Besides needing no sleep and his sly sense of humor, its author went more or less single-handedly mano-a-mano with the "hockey stick" nonsense and won. And he's not even American. The IPCC should have known better: Canadians know all about hockey. As Mark Steyn quipped, he's another immigrant doing a dirty job Americans won't do. And keep WorldClimateReport, GlobalWarming and CO2Science in mind too.

Of recent books, those of Lomborg, Leroux & Comby, Soon, and Essex & McKitrick are the most worth your time.*

A fine thumbnail sketch of the "global warming" craze is this talk by S. Fred Singer of the University of Virginia and George Mason University. He has a new book out too, but I have not yet read it.
---
* Mark this fact: not one of these gentlemen is American-born.