Monday, July 14, 2008

Climate of discontent

Enthusiasts for action on climate change may have put the cart so far before the horse that there are no longer any reins and other horse tack connecting them. Worldwide solutions to projected problems carry enormous potential costs, tens to hundreds of trillions of dollars, or years to decades of the total economic output of the United States. Before there can be any sustained commitment approaching such scales, there will have to be general agreement and near total certainty as to benefits and costs. Today there is only limited agreement and great uncertainty. Areas of agreement are recent trends of rising temperatures, rising sea levels and rising concentrations of trace gases in the atmosphere. Uncertainties include distinguishing causes from effects, assessing relative strengths of causes and predicting future effects from human efforts directed at causes.

Key knowledge and projections of climate change are coming from a few hundred people worldwide, who rarely if ever expose details of their work candidly to public view. Most reports rest on hackneyed assumptions, stated or not, that have never been conclusively proven. Only insiders at the private clubs get meaningful access. However, mavericks among them have explained that no model achieves much accuracy in predicting many decades of observations based solely on the profile of the Earth and on physics, chemistry and scalable, validated engineering. Models are tweaked with arbitrary adjustments, parameters determined by statistical inference rather than from physical principles, so that they get the right answers to match observations[1]. Who has the crystal ball to tweak them correctly for future years? An article evaluating several thousand models showed that only 15 years into the future, global temperature projections varied up to nearly 15 degrees Farenheit[2].

That situation does not reflect some kind of "junk science" [3] but rather an endeavor of great complexity, comparable perhaps to understanding the causes of and cures for cancer, and similar in that about 50 years of science has produced partial knowledge and limited success, leaving vast territories yet to be covered. Atmosphere and ocean circulations are dynamic, chaotic and dissipative, with many interacting factors and much potential for novel and unpredictable behavior. After 50 years of science, we are able to predict weather fairly accurately for about two days, but at about two weeks all our supercomputing and automatic monitoring does not improve much over what we can predict from typical weather for a time of year.

Climate, of course, is weather, averaged over years, decades and centuries. On a historic time scale, let alone a geologic scale, our records are spotty and shallow. We have a handful of temperature series extending back a few hundred years, some measurements of atmospheric gases for about a century and a half, and direct if controversial measurements of solar output since 1979. The past thirty years have seen major improvements in the density and sophistication of measurements, but such an amount of time is at best marginal as a baseline for estimating climate. Solar output, a critical factor in climate, can be estimated from longer records of sunspot activity, but the correlations are noisy and imprecise.

Historical records tell about sizable shifts in climate before the Industrial Age, such as the Medieval Warm Period, a shift in annual temperatures up to 2 degrees Farenheit found in Europe from about 800 to 1,300 CE. We have little to explain some such events except unmeasured changes in solar output. Modern measurements of the geologic record over the last several million years tell about more dramatic episodes of warming and cooling, shifting annual temperatures by more than 20 degrees Farenheit, with some substantial changes taking place in less than 100 years. Persistence of some patterns over centuries and millenia, recurrences at intervals around 40,000 and 100,000 years, and abrupt spikes and dips make explanations appear unlikely other than orbital patterns and solar output changes, stimulating changes in atmosphere and ocean circulation. In some warming episodes, atmospheric carbon dioxide concentrations lag rather than lead temperatures, thus appearing as effects rather than causes.

Many current climate models fail to consider records older than about 100 years, but useful records extend to three times that age. Among other things, they indicate warming trends since the early 1700s, while large effects of the Industrial Age on atmospheric gases emerge in the mid-1800s. Adjustment factors of current climate models may be tuning them to correlations between temperatures and atmospheric gases that would not be found in records over longer periods. Atmospheric emissions from power plants, motor vehicles and agriculture exploded since the 1940s, yet although it has been shown that the average exchange time for carbon dioxide between atmosphere and ocean is about ten years, records of temperatures and sea levels do not show closely correlated accelerations of most trends.

Experts at climate modeling seem increasingly focused on inside pursuits, such as cross-comparisons and increased areal and time densities. However, development of the science shows greatest improvements from understanding new factors, such as cloud cover, aerosols and particulates. Trustworthy knowledge is more likely to emerge from thorough understanding of systematics and interactions than from refinements of calculations. Unless we can obtain robust, accurate models grounded on principles of physics, chemistry and engineering, few people will trust the models with their lives and fortunes. Critical issues are whether, how much and how rapidly climate change could now be affected by human efforts and what levels of efforts may be required. If it were to turn out that rises in sea levels had become practically unstoppable and irreversible, then we might better spend resources reinforcing or relocating coastal communities rather than rushing to build photovoltaic panels.

[1] A. Arakawa and W.H. Schubert, Interaction of a cumulus cloud ensemble with the large-scale environment, part 1, Journal of the Atmospheric Sciences 31(4):674-701, 1974. A.H. Van Tuyl, Physical initialization with the Arakawa-Schubert scheme in the Navy's operational global forecast model, Meteorology and Atmospheric Physics 60(1):47-55, 1996. T.M. Wagner and H-F. Graf, A spectral convection parameterisation with a dynamical Arakawa-Schubert quasi-equilibrium closure, Geophysical Research Abstracts 10, EGU2008-A-12443, 2008.

[2] D.A. Stainforth, et al., Uncertainty in predictions of the climate response to rising levels of greenhouse gases, Nature 433(1):403-406, January, 2005.

[3] Steven Milloy, ed., Martin Durkin, dir., “The Great Global Warming Swindle,” British television film, March 8, 2007, revised edition available as DVD. Lars Mortensen, dir., "Doomsday Called Off," Danish television film, 2004, first shown in Canada, November 27, 2005. Hilary Lawson, dir., "The Greenhouse Conspiracy," British television film, August 12, 1990.

No comments: