As the administration had hoped, the meteorologists used the occasion to opine about climate change—but what many of them said wasn’t quite what Al Gore had in mind. “There’s still a significant segment of the scientific community that’s not sold on this,” Harvey Leonard, then the weatherman at WHDH in Boston, told The Washington Post. Others loudly refused to attend the summit, including all but one of the weathercasters in the Oklahoma City market. “I’m not smart enough to know [if the earth is warming], and I don’t think any person on the planet is,” KOKH meteorologist Tim Ross told the Daily Oklahoman. The following month, twenty TV weather personalities added their names to the Leipzig Declaration, a petition opposing the global warming theory.

It was only a blip on the radar, but it presaged the broader rejection of climate science that would come a decade later. The question was, why? No doubt, some of the blame belonged to the White House. In positioning themselves as advocates for not only a policy position but also a scientific one, Clinton and Gore had conflated the political question of what to do about climate change—one that was, and remains, deeply partisan in the U.S.—with the apolitical question of whether it was happening. This put the weathermen in a tricky spot—embracing what was, even then, the majority position in the scientific community would make them look like shills for the administration. “Since the White House is behind it, it’s political,” Leonard told the Post. “I’m not a lap dog,” Gary England of KWTV in Oklahoma City—now a prominent climate skeptic—told the Daily Oklahoman. “I think Al Gore’s motives were pretty good—he saw early on the potential that these people had,” Kris Wilson says. “But he was probably the wrong spokesman. As journalists, we’re taught to be skeptical, right? We’re taught that if your mother says she loves you, get a second source.”

But the disagreement, then as now, also came down to the weathercasters themselves, and what they knew—or believed they knew. Meteorology has a deceptively close relationship with climatology: both disciplines study the same general subject, the behavior of the atmosphere, but they ask very different questions about it. Meteorologists live in the short term, the day-to-day forecast. It’s an incredibly hard thing to predict accurately, even with the best models and data; tiny discrepancies matter enormously, and can pile up quickly into giant errors. Given this level of uncertainty in their own work, meteorologist looking at long-range climate questions are predisposed to see a system doomed to terminal unpredictability. But in fact, the basic question of whether rising greenhouse gas emissions will lead to climate change hinges on mostly simple, and predictable, matters of physics. The short-term variations that throw the weathercasters’ forecasts out of whack barely register at all.

This is the one explanation that everyone who has mulled the question seems to agree on—and indeed, when I spoke with meteorologists who were skeptical of or uncertain about the scientific consensus, it was the one thing they all brought up. “Meteorologists know our models,” Brian Neudorff, a meteorologist at WROC in Rochester, New York, told me. “There’s a lot of error and bias. We’ll use five different models and come back with five different things. So when we hear that climatological models are saying this, how accurate are they?”

But that hardly explains why so many meteorologists have disregarded the mountain of evidence of global warming that has already occurred—or why, in the case of the hard-line skeptics, they are so fixated on proving a few data sets’ worth of tree-ring and ice core measurements wrong. “I think a lot of people have theories,” Robert Henson says, “but nobody knows for sure.”

Charles Homans is an editor at Foreign Policy.