The best Leonid show is yet to come?

The streaks of light came fast and furious. Some raced across the sky in nearly parallel tracks, leaving behind hazy trails. A few seemed to dive into the moon.

 If last month's Leonid meteor shower proved disappointing in the United States, it took Europe and the Middle East by storm. And if the predictions of two astronomers continue to hold true, Earth will be in for a really big show in 2001 and another in 2002.

 At the shower's peak, on Nov. 17, some observers saw between 3,000 and 5,000 shooting stars, or meteors, in a single hour. Activity reached a crescendo at 9:05 p.m. EST-just 3 minutes earlier than predicted by David J. Asher of Armagh Observatory in Northern Ireland and Rob McNaught of the Australian National University in Weston.

 Scenes from the 1999 Leonid shower: Meteor's fireball and its fading light seen for more than 20 minutes over the Italian Alps. (CLorenzo Comolli)

 This is the first accurate prediction of a meteor storm, says Brian G. Marsden of the Smithsonian Astrophysical Observatory in Cambridge, Mass.

 The Leonid meteor shower happens every November, when Earth passes through a stream of dusty debris, or meteoroids, expelled by Comet 55P/Tempel-Tuttle. Dust grains slam into Earth's atmosphere and burn, creating the streaks of light known as meteors. About every 33 years, when the comet passes near, Earth encounters a large amount of debris, resulting in a heavy shower or storm.

 Exactly which years the Leonid dust particles will generate a storm has been difficult to predict. That's because astronomers hadn't realized that the debris stream is composed of distinct, narrow strands of dust, each expelled by the comet during a different passage by the sun, notes Asher. It's a matter of hit or miss: If Earth plows through the center of a dense strand, a storm will occur.

 By simulating the motion of strands in the solar system, Asher and McNaught conclude that the dust strand Earth traveled through last month was shed by the comet in 1899. Although that's the same material the planet traveled through during the spectacular storm of 1966, last month's event wasn't as dazzling because Earth crossed the strand's edge rather than its center, Asher says.

 Donald K. Yeomans of NASA's Jet Propulsion Laboratory (JPL) in Pasadena, Calif., says he agrees with the pair's explanation for the 1999 event. "I do take their future predictions more seriously now," he adds.

 Next year, McNaught and Asher calculate, Earth will pass for the first time through the edge of a band of dust cast off by the comet in 1866, yielding a puny shower. In 2001, however, Earth will plow sequentially through no less than three trails-debris expelled in 1767, 1699, and 1866-and the light show should prove more stunning than last month's. In 2002, when Earth again encounters material from 1866, as well as from 1933, the Leonids should also put on a great show, McNaught and Asher say.

 Their findings may shed light on a puzzling feature seen just hours after the Leonid shower reached its 1999 peak. Observers saw flashes of light near the moon, as if meteoroids had crashed on its surface. Researchers reported the phenomenon in a Nov. 26 circular of the International Astronomical Union.

 The brilliance of these flashes requires that the meteoroids have as much mass as a bowling ball-a rare but not extraordinary occurrence, estimates Alan W. Harris of JPL. Moreover, Asher and McNaught calculate that the moon intercepted a denser part of the 1899 stream than Earth did and thus encountered a greater number of large meteoroids at the time the flashes occurred. However, cautions Paul R. Weissman of JPL, the flashes could merely have been sunlight glinting off satellites or space debris.