After successfully employing a new technique for detecting planets around distant stars, a group of astronomers announced Jan. 24 that they have identified the smallest planet yet detected orbiting a normal star.

Its orbit is farther from its host star than Earth is from the Sun, which is significant because most known extrasolar planets reside inside the equivalent of Mercury's orbit.

The planet is estimated to be about 5.5 times as massive as Earth and thought to be rocky. It orbits a red dwarf star about 28,000 light-years away. Red dwarfs are about one-fifth as massive as the Sun and up to 50 times fainter. But they are among the most common stars in the universe.

So the finding suggests rocky worlds may be common.

"The team has discovered the most Earth-like planet yet," said Michael Turner, assistant director for the mathematical and physical sciences directorate at the National Science Foundation, which supported the work. The discovery is detailed in the Jan. 26 issue of the journal Nature.

Prior to this discovery, the smallest extrasolar planet found around a normal star was about 7.5 Earth masses. Earth-sized planets have been detected, but only around dying neutron stars.

The newfound planet, named OGLE-2005-BLG-390Lb, is probably too cold to support life as we know it, astronomers said. With a surface temperature of minus 220 degrees Celsius, it is nearly as frigid as Pluto.

It was discovered using a technique called gravitational microlensing, where scientists search for light from a distant star that is bent and magnified by the gravitational field of a foreground star. The presence of a planet around the foreground star causes light from the distant star to become momentarily brighter.

Astronomers hailed the discovery as the first of a new class of small, rocky worlds located at far-out distances from their stars.

The planet and star are separated by about 2.5 astronomical units (AU). One AU is equal to the distance between the Earth and the Sun. Until now, no small planet had been found farther than 0.15 AU from its parent star.

The finding means planet hunters are one step closer to detecting their holy grail: a habitable Earth-like planet that can sustain liquid water and support life.

"We may predict with reasonable probability that microlensing will discover planets with masses like that of Earth at a similar distance from their stars and with comparable surface temperature," said study co-author Bohdan Paczynski from Princeton University.

Of the more than 150 planets that have been discovered so far, most were found using the Doppler technique, in which astronomers look for wobbles in a star caused by the gravitational pull of a planet. This method has uncovered dozens of huge worlds but cannot spot small planets that are far from their stars.

Microlensing can detect small planets, but it is 50 times more likely to find a gas-giant planet like Jupiter. "Microlensing should have discovered dozens of Jupiters by now if they were as common as these five-Earth-mass planets," said study co-author David Bennett.

That suggests most of our galaxy's planets are small and rocky.

This prediction agrees with the standard model for solar system formation, known as the "core accretion" model. It goes like this: Dust around newborn stars forms clumps that stick together and eventually become asteroids, comets and planet precursors. In this scheme, relatively few planets successfully become gas giants, and they are outnumbered by small, rocky worlds.

"It's incredible to think that we went from 10 years ago having no planet to now having over 100 gas giants and even starting to find the first terrestrial planets," said Alan Boss, a theorist at the Carnegie Institution of Washington who did not participate in the discovery. "That's just an amazing leap."

Since star alignments are unique events, a microlensing experiment can never be repeated. Todd Henry, an astronomer at Georgia State University who was not involved in the study, said the discovery was an "intriguing result from this particular technique, but unfortunately you can't follow it up."

Many astronomers view the lack of repeatability as an acceptable trade-off, however, because thousands of star systems can be screened in a relatively short period of time compared to other techniques.

"You can't learn a whole lot about the details of individual systems ... but it's a wonderful alternative for learning about what the mass distribution of extrasolar planets might be and the frequency at which they occur," said David Latham, an astronomer at the Harvard-Smithsonian Center for Astrophysics who was not part of the study.

In a telephone interview, Jean-Phillipe Beaulieu, a co-author in the study, said that while the observations can never be repeated, the discovery was simultaneously verified by different telescopes around the world.

The microlensing event was detected July 11 by telescopes in the OGLE (Optical Gravitational Lensing Experiment) project. Planet hunters are very protective of their data and cooperation between different teams is rare. But astronomers around the globe were alerted so the event could be detected by multiple telescopes.

"The only way to realize the full scientific benefit of our observations is to share the data with our competition," said Paczynski, an OGLE co-founder.

Overall, the study involved 73 researchers from 32 institutions worldwide.

"The fact that they've got a whole bunch of folks using multiple telescopes all observing the same event and calibrating themselves self-consistently makes the data look very sound," Boss said. "I think it's a pretty solid detection."