BOSTON — Here's a pop quiz following the recent Kyoto climate conference: Have scientists shown that human activity is causing global warming? If you believe the media hype surrounding the conference, you might well answer "yes." But climate scientists themselves aren't so sure.
They can show that Earth's average surface temperature has risen 0.5 degrees C during this century. They can show that the atmosphere's concentration of man-made carbon dioxide (CO2) and other heat-trapping "greenhouse" gases has risen sharply at the same time. But when it comes to putting those two facts together, how a scientist interprets the data "depends very much on your prior belief," says Klaus Hasselmann, director of the Max-Planck-Institute for Meteorology in Hamburg, Germany.
Dr. Hasselmann says that he is "convinced we're having a global impact" on climate. Yet when it comes to uncovering mankind's dirty fingerprints in the data, scientists are stymied. "We simply don't have sufficiently reliable estimates of the natural variability" to tell what, if anything, is unnatural in the trend that has made the 20th century the warmest century of the past 500 years, Hasselmann explains.
That may give comfort to critics of the Kyoto agreement to limit CO2 emissions. But climatologists like Hasselmann warn that the scientists' ignorance is no excuse to delay action to curb greenhouse-gas pollution. They believe they have sound theoretical reasons to warn that such pollution can drive undesirable global warming. They are working hard to reduce the uncertainties in their estimates of what greenhouse-gas pollution has already done to our climate.
It's a tough job. Hasselmann and several dozen colleagues gave a progress report on their efforts during the recent meeting of the American Geophysical Union in San Francisco. Their problems lie both with the available climate data and with the computer-based climate simulations that complement those data.
Good meteorological instrument records go back only about 150 years. Scientists must use what they call proxies to get a hold on such earlier climate factors as temperature and precipitation.
Tree rings reflect temperature and moisture in their annual growth. Pollen in lake sediments indicates ancient plant populations and, by inference, the climates in which they grew. Annual layers in glacial ice record past temperatures and precipitation. And subsurface rock temperatures are an "archive of temperature change at the surface," according to geologist Henry Pollack at the University of Michigan in Ann Arbor.
Professor Pollack reviewed data from 300 underground boreholes taken by researchers in Australia, Europe, North America, and South Africa. His analysis suggests that Earth's average surface temperature has warmed by about 1 degree C over the past 500 years. Half the increase has come since 1900. Pollack called this result "fully consistent with the conclusion ... that human activity is a significant driving force behind global warming." But it does not prove the point.
The trouble with proxy data is that they don't do much to define the natural year-to-year climate variability. This makes it tough to discern whether or when any human impact cuts in. This means scientists have to look to computer simulations to try to estimate the natural variability, says Tim Barnett of Scripps Institution of Oceanography in La Jolla, Calif. He adds that those simulations are themselves fraught with uncertainties.
The bottom line, Dr. Barnett says, is that present scientific knowledge can't resolve the cause and effect of 20th-century climate. Yet, it is "more compatible" with the conclusion that both natural and man-made forces are at work than it is with the notion that 20th-century warming is benignly natural or, conversely, is largely mankind's fault. That means politicians and special interests will continue to make what they will of scientific uncertainties. The rest of us should take their claims with the proverbial grain of salt.