Global warming underestimated?

Are the official estimates of a 1.4° to 5.8°C (2.5° to 10.4°F) increase in global mean surface temperatures by the year 2100 significantly in error? That was the conclusion of MIT professor Dr. Peter H. Stone, in a lecture I attended last week at the annual meeting of the American Meteorological Society. Dr. Stone's results were also published January 13, 2006 in Geophysical Research Letters. The "official word" in the science of climate change comes from the United Nations-sponsored Intergovernmental Panel on Climate Change (IPCC), a collaborative effort between over 2,000 scientists from over 100 countries, including many of the top climate researchers in the U.S. The IPCC publishes an extensive assessment of the state of the science every six years. The most recent report, issued in 2001, predicted the 1.4° to 5.8°C increase. If Dr. Stone is right, the next IPCC assessment, due out in 2007, will have to revise that estimate upwards.

Dr. Stone started his talk by posting this quote from the Executive Summary of the 2001 IPCC model evaluation chapter: "Confidence in the ability of models to project future climates is increased by the ability of several models to reproduce the warming trend in 20th century surface air temperature when driven by radiative forcing due to increasing greenhouse gases and sulphate aerosols." (The term "forcing" in climate research refers to any process, natural or human-caused, that "forces" the climate to respond in a significant way.) The IPCC report supported their statement by comparing climate simulations of the observed 20th century climate that used just natural processes ("forcings" such as volcanic eruptions and natural changes in the sun's brightness) with simulations done including human-caused "forcings" (greenhouse-effect gases added since pre-industrial times, plus aerosol particle pollution). Dr. Stone presented Figure 4 (below), a modified version of a figure from the 2001 IPCC report. The figure shows a typical 20th century climate simulation by one of the major climate models used for the IPCC assessment--the UK Hadley Center model. The results look good. The model is able to reproduce the observed climate of the 20th century. In addition, the simulation shows that one cannot explain the observed 20th century global warming of 0.6°C without including human-caused (anthropogenic) climate forcings.



Dr. Stone argued that the IPCC's confidence in the ability of models such as the UK Hadley Center Model to predict future climate was invalid, and that the good agreement between the observed climate and model prediction seen in the figure above could have been coincidence. He outlined several ways that compensating errors in two or more areas of model uncertainty could have produced a climate simulation that matched the observed 20th century record.

Major uncertainties in climate change computer models include:

1) Climate Sensitivity (how much global surface temperature changes when CO2 is doubled)
2) Rate at which the oceans take up heat
3) Strength of forcing by aerosol particles
4) Natural variability

For example, if a model has a Climate Sensitivity that is too great (the model predicts too much warming for a given increase in CO2), and improperly assumes too much cooling will occur due to pollution from aerosol particles, the two errors will cancel each other out and lead to a realistic-looking simulation. The Climate Sensitivities of the 11 key models used to generate the 2001 IPCC results varied by about a factor of 2.5, from 1.5°C to 4.5°C. Similarly, the amount of heat taken up by oceans varied by about a factor of 2.5 in the models. Additional uncertainties exist in the models' treatment of aerosols and natural variability.

Rather than dismiss the climate models as being too filled with uncertainty to be useful for performing climate simulations, Dr. Stone maintained that one can do an intelligent uncertainty analysis by varying two of the major uncertainties in a model simultaneously, and study the resulting model predictions. He described his group's research to evaluate the uncertainties in 11 of the key models used to formulate the 2001 IPCC climate report. The study was done using data from the Coupled Model Intercomparison Project (CMIP), an international research program begun in 1995. The talk then became quite technical, with several plots showing Probability Distribution Functions on parameter-space diagrams. It was at this point I bemusedly watched the audience member next to me who hadn't had enough cappuccino that morning repeat the classic pecking bird "doze-droop-jerk-I'm awake!" pattern. Meteorology talks aren't always filled with captivating displays of 3-D Category 5 hurricanes! There's a lot of hard science needed to understand the concepts.

Finally, Dr. Stone finished his uncertainty analysis, and he presented some rather startling conclusions:

1) Models have been over-estimating the rate of mixing of heat into the deep ocean.

2) This implies that their projections of surface warming for the 21st century are too low.

The guy next to me jerked fully awake now, and the audience got noticeably more attentive. "And this worries me," Dr, Stone continued. "It worries me enough that we've made many extensive tests of our methodology that try to make sure that there are no flaws. I would be delighted if anybody here could come up with a test that we might look at to see if we've done anything wrong." The audience, filled with several hundred people, including many of the world's foremost climate experts, was silent. No one could come up with a reason to dispute Dr. Stone's gloomy conclusion.

So how much in error are the climate models? Dr. Stone didn't give a number in his talk, and when I asked him about this later he said he had only a rough preliminary idea of what this error might be. His research team is currently analyzing their results to see how much additional warming we can expect. When they publish some specific error estimates, I'll be sure to post a follow-up blog on the subject.

Professor Stone's talk can be heard on-line for free. To do so, you must install the free WebEx player for IE or Netscape. Note: this will not work for other browsers, such as Firefox! The talk is about 40 minutes long, and includes figures. Alternatively, you can read the paper on the subject that he co-authored along with C.E. Forest and A.P. Sokolov of MIT's Joint Program on the Science and Policy of Global Change:

Forest, C.E., P.H. Stone, and A.P. Sokolov, "Estimated PDFs of climate system properties including natural and anthropogenic forcings", Geophysical Research Letters, 33, L01705, doi:10.1029/2005GL023977, 2006.

A free abstract of the paper is available from the agu.org website. A full version costs $9 for non-subscribers.

A note on my global warming blogs
In an issue as complex, contentious, and important as global warming, it is impossible for anyone to present an unbiased and fair treatment of the subject. My bias will be towards presenting new scientific findings published in peer-reviewed journals, as well as calling attention to the political aspects of the debate when it appears that one side or the other is attempting to twist or hide the truth. While thus far I have only focused on the NASA/Dr. James Hansen affair, I also have criticism of those claim that Hurricane Katrina was significantly enhanced by global warming. Although it is possible that global warming did contribute significantly to Katrina's intensity, the current best hurricane science supports only a 1-2 mph enhancement in Katrina's winds by global warming. I have a blog on this topic I plan to post next week, highlighting recent questionable statements by the editor of Science magazine on the matter.

For those of you following the NASA/Dr. James Hansen affair, see this morning's New York Times article, where George C. Deutsch, the young NASA press aide who resigned on Tuesday amid claims that he had tried to keep Dr. James Hansen from speaking publicly about global warming, defends himself publicly.

A note on media bias on the global warming issue
I'm of the opinion that articles in the New York Times on global warming tend to be biased in favor of dramatizing the problem and calling for action. Articles in the Wall Street Journal, Washington Times, and Newsweek magazine generally have the opposite bias. Time magazine seems pretty neutral, and CNN.com may have a pro-action bias. I'm not sure about the USA Today, Washington Post, or other sources. One of my favorite sources of global warming info (but a little too technical for many readers) is from realclimate.org, which is maintained by some of the top climate scientists in the field. They have serious disagreements with the Wall Street Journal.

Jeff Masters