A notorious example occurred in 1908 when an asteroid in this size range is believed to have exploded above the uninhabited Tunguska region of Siberia, leveling trees for some 800 square miles (2,000 square kilometers) around. Astronomers have for a decade or so said so-called Tunguska events probably occur about once every hundred years, leading some to speculate that we're about due for another.

Harris stressed that impacts are random events; the idea that another could be "about due" is incorrect.

"The fact that one occurred only a century ago makes it no less (or more) likely that one will happen tomorrow instead of 1,000 years from now," he told SPACE.com. "My lower estimate of collision frequency makes it less likely on any given day (year, decade, century) that an impact will occur, but does not allow one to make any statement as to when specifically the next one might occur."

Harris, along with several NASA scientists and many asteroid hunters, worries far more about larger Near Earth Objects (NEOs); those bigger than 1 kilometer (0.62 miles) that could cause global destruction and even threaten civilization.

About 1,100 large NEOs are thought to exist. More than 600 have been found by programs stemming from a NASA effort to locate 90 percent of them by 2008.

None are known to be on a collision course with Earth.

Smaller asteroids are harder to detect, however. Now and then, one passes relatively close to the planet and some are spotted only after such flybys. A few vocal scientists have long called for stepped-up funding to find these smaller NEOs, because they are more likely to strike and are a more immediate threat.

NASA's line has been to find the big ones first, then consider going after the smaller ones. Only recently, however, have discussions on how to do this become serious. New telescopes would be needed, and telescopes generally take a few years to design and build.

Coincidentally on Tuesday, the University of Hawaii announced a $3.4 million Air Force grant to design an array of telescopes that would find faint objects, including smaller asteroids. The planned observatory, called the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS), is expected to begin operation in 2006.

Rolf Kudritzki, who leads the university's Institute for Astronomy, said Pan-STARRS would help complete the search for large NEOs "and will extend the search to much smaller objects."

Alan Harris said that because smaller, Tunguska-sized impacts are regarded by most asteroid experts as "a very minor component of the overall risk," this overall risk is not changed much, in practical terms, by his new calculations.

There are smaller object out there, too.

Asteroids smaller than 165 feet (50 meters), which might in fact hit the Earth about once per century, would not be expected to cause significant damage "beyond a few broken windows and such," Harris said, "though I'm sure such an event would be spectacular from under it."

The smallest rocks, car-sized or smaller, routinely enter the atmosphere but generally vaporize or break apart before reaching the ground.

Harris' new computer-generated estimate was made possible because about 30 asteroids in the Tunguska size range have now been discovered by MIT's Lincoln Near Earth Asteroid Research (LINEAR) survey, which is designed primarily to find large NEOs.

Harris' estimate is uncertain by a factor of about three.

More Asteroid News | Astronotes