On Sept. 11 the Federal Aviation Administration grounded all commercial aircraft in the United States after the terrorist attacks. For three days the skies were nearly as clear and quiet as they'd been since before the Wright Brothers skidded aloft.

It was a difference not lost on David Travis. "I realized this was an opportunity to study what the U.S. used to be like before the aviation age began," says the University of Wisconsin atmospheric scientist.

Specifically, he realized he had an unprecedented chance to answer a question that has vexed him and others for decades: Do the wispy tracings left behind by jets - known formally as contrails - somehow alter the environment?

Travis' study, published today in the British journal Nature, offers the first direct evidence that contrails do have a slight but measurable effect. When jets are flying, the days are a little cooler and the nights a little warmer.

On a typical travel day, scientists have counted more than 14,000 flights crisscrossing the skies. In the long term, the skies are expected to become even more crowded, making contrails a growing concern for some atmospheric scientists.

Short for "condensation trail," contrails are essentially tiny ice balls, formed when the water vapor in the jet exhaust meets the frosty air of the upper atmosphere.

"Within seconds there's a flash freeze," Travis says.

Contrails form most frequently in the spring and fall when the air above 25,000 feet - the typical cruising altitude for airplanes - is cold and moist. Typically they dissipate in an hour, but if conditions are just so, they can linger for six or more. Although they are thickest in the Northeast, Midwest and Pacific Northwest - where jet traffic is heaviest - contrails show up nearly everywhere: On satellite photos, they have been spotted over the Sahara and the South Pole.

Before atmospheric scientists came onto the scene, the military was the first to take an interest in contrails.

"Contrails were considered a hell of a problem," says Stanley Changnon, an atmospheric scientist at the University of Illinois at Urbana-Champaign. "With or without radar, the enemy could say, 'There goes a bunch of bombers.'"

After World War II, the U.S. Air Force embarked on an unsuccessful effort to eliminate contrails. During the Cold War, pilots of U-2 spy planes installed rear-view mirrors. If they spotted exhaust trailing them as they crossed into the Soviet air space, they wheeled around rather than risk detection.

Atmospheric scientists got involved in the 1970s, when they began wondering whether contrails were connected to an increase in high-altitude cloud cover in the Midwest. Early research done by Changnon and others suggested there was a link.

But until Sept. 11 there was no way to conduct a proper experiment, no way to measure the atmosphere when no planes were flying.

"Here's a lifetime opportunity to really have a control case," Changnon says.

To find an answer, Travis used ground-level temperature readings collected by more than 4,000 weather stations across the continental United States. The scientists measured the diurnal temperature range - the difference between the daytime high and the nighttime low temperatures.

That's because contrails were thought to have a dual effect on the atmosphere: Like the wispy high-altitude cirrus clouds, they deflect incoming radiation from the sun, cooling the planet. Like low-hanging cumulus and stratus clouds, contrails also trap infrared radiation emanating from the Earth.

When the scientists compared average temperature data from 1971 to 2000 with the measurements recorded during the three-day grounding period, they found an average difference of slightly more than 1 degree Celsius.

In other words, without contrails in the sky to deflect or trap radiation, the nights were, on average, slightly cooler and the days slightly warmer than usual.

The effect was even more striking in regions where air traffic is heavy. In the Northeast, for example, the temperature difference between day and night could be as much as 3 degrees Celsius, he says.

Although it's possible that other weather effects are contributing to the temperature changes, Travis and others say it's hard to believe it's a coincidence that when planes stopped flying there could be such a drastic -in scientific terms anyway - change in temperature. "It's pretty solid proof the contrail effect is real," Changnon says.

Travis wasn't the only researcher to jump at the chance to study contrails during the three-day lull in air traffic. Although no commercial jets were flying during that period, military jets were in the air. And that gave Patrick Minnis, a scientist at the National Aeronautics and Space Administration's Langley Research Center in Hampton, Va., an idea.

Using a series of photos snapped by National Oceanic and Atmospheric Administration satellites Sept. 12, he traced the growth of a half-dozen contrails made by jets flying at high altitudes over an area from Kentucky to Pennsylvania.

(One of the contrails he saw streaked from Nebraska to Washington, prompting Minnis to tell a New York Times reporter that it was probably Air Force One. White House officials apparently were not happy that atmospheric scientists and contrails could compromise security. Now some scientists are having trouble finding the satellite photos in any database. When asked about it now, Minnis says tactfully, "I have no comment.")

In a matter of hours, the photos showed, the contrails change from pencil-thin etches in the sky to hazy clouds covering more than 20,000 square kilometers. "If anybody had been on the ground, they wouldn't have known it was a contrail," Minnis says. The contrails, he says, essentially became plain old clouds.

Using a computer model, Minnis and his collaborators calculated that contrails from a typical day's air traffic passing through the same region would cover 100,000 square kilometers.

Of course, scientists have yet to prove whether the contrail-smeared skies are harming the environment. Some scientists have argued that it's upsetting the radiation balance of the planet.

Although a change of a single degree throughout the United States may not do much harm, Travis and others say a few degrees in a region such as the Midwest might affect crops or other things sensitive to the weather.

Of course, Changnon says, the temperature change caused by contrails could also be viewed in a positive light: By making the days cooler, it means less need for expensive air conditioning.