Advocates of "geoengineering"--proposed approaches to cool the planet to offset the effects of greenhouse gases--point to major volcanic eruptions as evidence that the techniques could work. Very large eruptions, like the one at Mount Pinatube in 1991, actually did cool the planet by injecting sulfates into the upper reaches of the atmosphere, where they circulated for over a year, shading the earth. Such advocates have been hoping that another major eruption would come along that could be studied for clues on how best go about artificially cooling the planet.

The eruptions of Iceland's Eyjafjallajökull volcano, while impressive and large enough to shut down air traffic in Europe, don't fit the bill. Here's what Alan Robock, a professor of environmental studies at Rutgers University, who has studied the impact of volcanoes on climate, told me:

So far the volcano has only put out less than 0.004 Mt of SO2, compared to the 20 Mt that the 1991 Mt. Pinatubo eruption injected into the stratosphere. Furthermore, the Icelandic emission was only into the lower atmosphere, where the lifetime is on the order of one week, as opposed to a couple years in the stratosphere. So too little, and staying in the atmosphere for too short a time. So far.

To make an impact globally, the volcano would have to erupt much more violently (or perhaps trigger nearby volcanoes to erupt violently) so that the larger amounts of sulfates would reach the stratosphere and stay in place, he says. To follow the size of eruptions at the volcano, click here.

The prospect of geoengineering, the intentional large-scale manipulation of the climate to offset the effects of global warming, continues to draw the attention of top scientists and policymakers. The prestigious journal Science is publishing two papers on the subject in its next issue, and next week the Science Committee of the House of Representatives will hold its second hearing on the subject.

According to some climate models, climates could change very quickly as greenhouse gas emissions rise, such as when thawing permafrost releases methane (a powerful heat-trapping gas), or loss of ice causes the earth to absorb more heat. Such outcomes might not be very likely, but even the chance that they could happen has scientists worried, as they could lead to widespread severe drought and famine and rapid sea level rise, among other things. Even the most ambitious efforts to curb greenhouse emissions might not be enough to prevent disastrous climate change, some say. And ambitious efforts to cut emissions seem ever more unlikely, after the limited success of the recent summit in Copenhagen, and opposition to a climate change bill in the Senate.

In one of the Science articles, researchers say testing of geoengineering strategies may prove necessary, but before testing can occur, they say an international agreement should be reached. Such testing could cause changes in weather patterns that could hurt some countries, for example, and an agreement needs to be in place to establish standard testing procedures and methods for dealing with international disputes arising from such testing.

Such an agreement might be even more important if the argument of another Science paper proves to be correct. In that paper the researchers argue that the only way to effectively test geoengineering is through full scale implementation of a particular geoengineering scheme--and such a large deployment could have big, unforeseen consequences.

In the past couple of weeks I've been looking into a comment from one of the people who read my recent feature on geoengineering, "The Geoengineering Gambit." The comment had to do with the first paragraph of the feature, which says that glaciers in the Himalayas are in danger of disappearing in just 25 years, threatening the water supply for people in India, China and other Asian countries. That statement was based on a report by the Intergovernmental Panel for Climate Change (IPCC), which states:

Glaciers in the Himalaya are receding faster than in any other part of the world (see Table 10.9) and, if the present rate continues, the likelihood of them disappearing by the year 2035 and perhaps sooner is very high if the Earth keeps warming at the current rate.

A reader pointed out that in recent months there has been speculation that the number 2035 was in error, and that the original source estimated the glaciers would disappear by 2350. I've since contacted the IPCC and read the relevant sources. It's still unclear whether the original date was indeed 2350, although there is, I believe, a strong case that it was, and that the IPCC was therefore off by about 300 years.

What is clear is that the IPCC got this date, not from peer-reviewed literature, but from an article in the magazine New Scientist. The IPCC report does not cite the New Scientist in the report, so there was no way of knowing that the organization, in this case, decided not to rely on peer-reviewed literature. Today the IPCC issued an official statement saying the information was from "poorly substantiated sources" and that "the clear and well-established standards of evidence, required by the IPCC procedures, were not applied properly."

In other words, there is no solid source for the statement that the glaciers will disappear in 25 years, and I regret that this statement was included in the feature. We'll print a correction in the next issue, but I also wanted to call reader's attention to this error here.

The main point of the feature isn't affected by this mistake. Scientists are concerned enough about rapid climate change to recommend the study of geoengineering methods for rapidly cooling down the earth, and the disappearance of the Himalayan glaciers was only one of the reasons, and indeed not even the main reason cited by the people I talked to. The potential for drought, species extinction, ocean acidification and rising sea levels were their top concerns.

A story about the IPCC error, and how it occurred, and what it means for the credibility of that organization, will appear tomorrow on our website.