On March 6 NASA’s Kepler space telescope embarked on a four-year mission to discover Earth-like planets in the Milky Way. Like its predecessor, the French-led COROT mission launched in December 2006, Kepler will monitor a selection of stars for temporary decreases in brightness. One dip could mean anything, probably just a blip in the star’s energy output; a second dip would still signify relatively little; a third dip, occurring after the same time interval as that between the first and second, would seem highly provocative; and a fourth dip after an identical interval would almost certainly mean that a planet is on an orbit that carries it directly between the star and us. Every time the planet passed, or transited, across the face of its star, it would block some of the starlight. A world roughly the size of ours diminishes its star’s light by about one part in 10,000 [see “Searching for Shadows of Other Earths,” by Laurance R. Doyle, Hans-Jörg Deeg and Timothy M. Brown; Scientific American, September 2000].

Earlier this year COROT found a planet with about twice Earth’s diameter, orbiting so close to its parent star that each revolution takes only 20 hours. Kepler, with a mirror three and a half times wider than COROT’s, should find dozens or hundreds of Earths orbiting at more comfortable distances from their star. Most current searches, which look for the slight gravitational tug that a planet exerts on its parent star, could not detect such comparatively small worlds. The trade-off is that the planets’ orbits must be aligned with our line of sight, and the laws of probability suggest that only about one in 100 will be so lucky. Nevertheless, Kepler will be able to create a statistically valid sample of Earth’s galactic cousins.

But if this triumph occurs, astronomers will find themselves bereft of the information that they would most dearly like to obtain: What conditions exist on these planets? Are they suitable for life? When a gas-giant planet transits its star, astronomers can analyze its atmosphere by measuring the amount of dimming produced at different wavelengths. But planets the size of Earth are far too small for this technique to work. So the search strategy employed by COROT and Kepler can find Earths but cannot tell us much about them. They cannot discern any of the signs of life, such as the distinctive colors of chlorophyll or its alien equivalents [see “The Color of Plants on Other Worlds,” by Nancy Y. Kiang; Scientific American, April 2008]. Even the Space Interferometry Mission (SIM), loosely planned for launch in 2015, will say little about the Earths it discovers.

The instruments capable of assessing habitability lie still further in the future, primarily because they are so expensive. NASA’s Terrestrial Planet Finder (TPF) and the European Space Agency’s Darwin mission could take spectroscopic measurements of planets’ surfaces and atmospheres, but neither has yet proceeded beyond the design study phase. Even if the agencies pool their resources, the mission could cost about $2 billion and take nearly a decade to build. For now, the best hopes for gleaning more information about planets are the James Webb Space Telescope (JWST), scheduled for launch in 2013, and the next generation of ground-based telescopes [see “Giant Telescopes of the Future,” by Roberto Gilmozzi; Scientific American, May 2006].

Although they were not specifically designed for planet analysis, these telescopes will be equipped with coronagraphic instruments designed to block starlight, allowing researchers to see any small bodies hiding in the glare. These instruments could produce images of young gas-giant planets, if they exist, around some of the nearest stars. They might also be able to piece together spectroscopic information about tightly orbiting objects.