WASHINGTON — It began with an anonymous Internet posting, and a link to a wonky set of e-mails and files. Stolen, apparently, from a research center in Britain, the files showed the leaders of climate-change science discussing flaws in their own data, and seemingly scheming to muzzle their critics.
Now it has mushroomed into what is being called "Climate-gate," a scandal that has done what many slide shows and public-service ads could not: focus public attention on the science of a warming planet.
Except now, much of that attention is focused on the science's flaws. Leaked just before international climate talks begin in Copenhagen — the culmination of years of work by scientists to raise alarms about greenhouse-gas emissions — the e-mails have cast those scientists in a political light and given new energy to others who think the issue of climate change is all overblown.
The e-mails don't say that: They don't provide proof that human-caused climate change is a lie or a swindle.
But they do raise hard questions. In an effort to control what the public hears, did prominent scientists who link climate change to human behavior try to squelch a back-and-forth that is central to the scientific method? Is the science of global warming messier than they have admitted?
The stolen electronic files include more than 1,000 e-mails and 3,000 documents, all taken from servers at the Climatic Research Unit in Britain.
Phil Jones, the unit's director, wrote a colleague that he would "hide" a problem with data from Siberian tree rings with more accurate local air temperature measurements. In another message, Jones talks about keeping research he disagrees with out of a U.N. report, "even if we have to redefine what the peer-review literature is!"
Since then, Jones has stepped down temporarily. And Penn State is exploring whether the e-mails, some of which were written by one of its professors, Michael Mann, warrant an investigation. Mann said he is confident that neither he nor any of the other researchers whose e-mails were pirated "did anything improper."
But recent debate — some scientists say the Earth hasn't warmed as predicted over the past 10 years — show that climate science is still science, with researchers drawing different lessons from the same data. The problem is that it plays out before an audience that won't wait for certainty.
Politicians say, " 'We need to reduce the uncertainty,' and I think that's contributed to a certain mind-set where (climate scientists) try to reduce the uncertainty" when they talk about their research, said Judith Curry, chair of the school of Earth and atmospheric sciences at Georgia Tech. "I'm a little bit worried about that political pressure," she said.
But the climate establishment — including the U.S. government's top scientists on the subject — say that nothing in the e-mails disproves their bedrock ideas. Carbon dioxide and other greenhouse gases are still gathering in the atmosphere and trapping more of the sun's heat, and the consequences of that will still be dire in the long run, they say.
"Our collective understanding of how the Earth is warming … rests on a wealth of scientific information that is very diverse and comes from multiple sources and multiple groups," said Jane Lubchenco, who heads the National Oceanic and Atmospheric Administration. "Regardless of what happened in one place, it doesn't undermine the totality of what we know."
For a few, however, the stolen files were confirmation that the climate establishment was trying to keep them out of the debate.
These include the familiar kind of climate skeptics, those who think that the climate isn't changing or that it isn't a crisis. But they also include a handful of researchers who think climate change is happening, but — for various reasons — are skeptical that mainstream science fully understands the phenomenon.
"To me, it's unambiguous … humans are altering the climate system," said Roger Pielke Sr., a research scientist at the University of Colorado. "It's just that, it's much more than CO2."
Pielke said his research shows that, in addition to carbon dioxide and other factors, Earth's warming is affected by how people alter the land. When a forest becomes a farm, or a farm becomes a suburb, that changes the amount of heat and moisture coming off the ground, he said.
But Pielke said he has seen some papers rejected and has felt so marginalized that he quit a U.S. panel summing up climate change a few years ago. One of the stolen e-mails seems to confirm the idea that he was being excluded: In 2005, Jones wrote to colleagues about some of Pielke's complaints, "Maybe you'll be able to ignore them?"
"These individuals, who are very sincere in their beliefs, have presumed that that gives them permission to exclude viewpoints that are different from their own," Pielke said.
Mainstream climate scientists say they have kept an open mind but have rejected papers that lack proper evidence. In Pielke's case, "the literature doesn't show" his ideas about the importance of land use are correct, said Tom Karl, head of the NOAA's National Climatic Data Center.
Top climate scientists say that in recent years most of the new, worthy research has only made the threat of climate change seem bigger and faster.
But the current debate over what's happening to global temperatures shows the noisy, confusing disagreement of scientists trying to make nature make sense.
These are the facts: After an increase in 1998, the world has been historically warm, but its average temperatures have not climbed steadily. Does that mean climate change has stopped?
Many mainstream scientists say no: This is just a tic of nature, as cycles of currents in the Pacific Ocean and a decrease in heat coming off the sun have temporarily dampened warming. Some researchers, though, have said the models — and, by extension, the human researchers who built them — could be missing something about how the climate works. That point was made in one stolen e-mail, in which climate researcher Kevin wrote it was a "travesty" that models could not explain why the Earth hadn't warmed more.