Despite nearly 2 decades of concern, experts still don't have a good handle on exactly how quickly tropical forests are disappearing. In the 9 August issue of Science, however, researchers describe an effort to fill that data void: one of the first studies to assess humid tropical forest with satellite data rather than on-the-ground measurements and guesswork.
The source of most estimates of global forest loss has for years been the United Nations' Food and Agriculture Organization (FAO)--reports that are notoriously inaccurate (Science, 23 March 2001, p. 2294). Aiming for more reliable results, Frédéric Achard and colleagues at the European Commission's Joint Research Centre in Ispra, Italy, sampled remote-sensing data for humid tropical forests. The researchers first identified deforestation "hot spots" using early 1990s low-resolution satellite maps and by consulting with experts. Then they selected 100 sampling sites, statistically weighting them so more fell in hot spots. Using high-resolution images of these 100 patches, they calculated how much forest had been lost between 1990 and 1997.
Extrapolating, the team found that the world lost an average of 5.8 million hectares of humid tropical forest each year--an area twice the size of Maryland--give or take 1.4 million hectares. The study's net total for deforestation (including regrowth) is 23% less than the FAO estimate for the same time period. The highest percentage deforestation rates were in Southeast Asia, followed by Africa and South America.
But some researchers say the study's approach is flawed. "I dispute that they got the right hot spots," says David Skole, a remote-sensing expert at Michigan State University, East Lansing. By relying on maps from the early 1990s, he says, the study likely missed areas where deforestation began later. Skole is a co-author on another new satellite forest study, currently under review, that uses a different approach. Although that study also finds that the FAO estimates are too high for the 1990s, it gets different results for each continent.
While these two studies aren't likely to be the last word, they've already got climate change experts rethinking an important number: how much carbon dioxide land plants are absorbing. "It's a very big deal," because predictions of global warming rely on that number, says ecologist David Schimel of the National Center for Atmospheric Research in Boulder, Colorado.