Ths small satellite "FAME", scheduled for 2004, will spend five years measuring the positions of millions of stars with an accuracy of only 14 billionths of a degreee -- enough to detect planets only twice Jupiter's mass around several hundred stars within 75 light-years.

NASA hopes to launch the much bigger "Space Interferometry Mission" in 2006, which will survey several hundred nearby stars with an accuracy 50 times better -- good enough to detct Earth from a distance of 9 light-years! (The greater sensitivity of these satellites will also enable us to detect giant planets much farther from their parent stars, allowing us to gauge just how unusual our own Solar System really is.) And astrometry -- unlike Doppler velocity measurements -- will allow us to determine the "flatness" of the star's looping path as seen by us, and so measure the tilt of the planet's orbit toward us and thus its exact mass.

Another promising near-future technique is "transit observations", in which one simply measures the very slight decrease in the star's apparent brightness as a planet repeatedly crosses in front of it -- which about one-half percent of Earth-sized planets of other stars should do for us. This technique was successfully used for the first time this year to confirm an extrasolar planet (one of the hot Jupiters) -- and it allows determination of the planet's diameter.

Laurence Doyle of the SETI Institute, in a Conference poster, reported the results of his long-term transit survey of the binary red dwarf star pair CM Draconis, looking for a rocky terrestrial planet within its Habitable Zone.

(Since the stars are so dim, such a planet would have a period of only a few weeks.) He could find no such planet down to a diameter of 31,000 km: "We believe this is the first survey of another stellar system for possible habitable planets to be performed."

But transit observations, again, can be made far more sensitively from space.

The French Corot satellite, to be launched in 2004, will make very sensitive light-fluctuation measurements of 6000 stars, enabling it to look for large transiting planets.

But Corot is a small satellite with different scientific goals, which is not really designed for this.

Ames Research Center astronomer William Borucki, in a Conference poster, described his "Kepler" satellite, which is small and cheap enough that he has already proposed it for NASA's low-cost Discovery program.

(It didn't make the list of finalists, but it got good enough appraisals that he is proposing an improved version this year.) Kpeler -- which can be launched on a Delta -- is designed (unlike Corot, FAME and SIM) to focus entirely on planet-hunting.

It would spend four years observing a single field of 100,000 stars, checking all of the simultaneously for transiting planets, with half of the stars being F, G and K-types that might have habitable Earth-type planets.

It would be able to firmly detect 150-200 Earth-sized planets in these stars' Habitable Zones, and over 400 larger terrestrial-type planets (as well as over 1500 giant planets near their stars, which it could also directly detect by measuring their reflected light).

And thus it might single-handedly give us our first really good estimate of how many alien planetary systems may contain planets that might support complex life.