Clouds, which reflect sunlight, ought to cool the earth. But they can also hold in warmth. That second effect swamps the first. According to Minnis' calculations, increased cloud cover since the 1970s ought to have led to a warming of .36 to .54F per decade. The actual warming during this period falls within that range, at just under .5. That may not sound like much, but when only 9F separate our current temperature from the last Ice Age, it's clear that a little warming makes a big difference. "This study," says Minnis, "demonstrates that contrails should be included in climate-change scenarios."
The only way to prove the point is to keep the jets on the tarmac and see what happens. That's exactly what occurred in 2001, between Sept. 11 and 14, when U.S. air travel was shut down following the terrorist attacks. During that period, the swing between daytime highs and nighttime lows sometimes measured more than twice as much as usual, perhaps owing to a reduction in cirrus clouds that allowed collected solar heat to radiate away. New and larger passenger planes might exacerbate the problem, but it is the frequency of flights that matters most. One way to tackle warming would be to have planes fly roughly 25% lower — altitudes less conducive to cirrus-cloud formation. But there's a catch: gas consumption would go up if planes were forced to plow through thicker air.