The Space Reviewin association with SpaceNews

Kepler illustration
Kepler, scheduled for launch Friday night, will look for Earth-sized worlds around other stars, the next step in the search for life beyond our solar system. (credit: NASA)

In search of other Earths

If all goes as currently planned, on Friday night a Delta 2 rocket will lift off from Cape Canaveral and, a little over an hour later, inject NASA’s Kepler spacecraft into an Earth-trailing heliocentric orbit. Kepler is the latest in NASA’s long-running Discovery program of low-cost planetary science missions—although “low-cost” is a relative term: Kepler’s total cost is about $600 million, and budget overruns at one point jeopardized the mission—that have ventured throughout the inner solar system, from Mercury to Mars, as well as comets and asteroids.

Kepler, though, has a more distant scientific destination. From its perch in orbit around the Sun, the spacecraft will peer out at a particularly dense star field in the Orion spiral arm of the Milky Way. Its purpose: to detect minute variations in the brightness of some of the stars in that field caused by planets as small as the Earth passing in front of them. It’s a relatively simple mission, but one that promises to usher in a new era in the study of extrasolar planets as well as the search for life beyond Earth.

Looking for Earth-like exoplanets

Since the wave of extrasolar planet, or exoplanet, discoveries began in the mid-1990s, over 300 such worlds have been discovered. During that time astronomers have assembled an impressive menagerie of worlds from “hot Jupiters”, gas giants that orbit extremely close to their parent stars; to extremely massive planets that approach the limits of what’s considered a brown dwarf; to planets in highly elliptical orbits. There are also an increasing number of multi-planet solar systems, such as 55 Cancri, which has at least five planets.

The concept behind Kepler is simple: look for stars that dim when a planet passes in front of them. “The technique is understandable to a kindergartner,” quipped Marcy.

What’s missing from the exoplanet catalogs today is anything with the basic characteristics—mass and orbit—of the Earth. This absence is not alarming given that most exoplanets have been found with what’s known as the “radial velocity” technique, where astronomers look for periodic Doppler shifts in the spectra of stars caused by the gravitational tugs of orbiting planets. The heavier the planet and the closer it is to its star, the greater the Doppler shift it can produce, so it’s little surprise that the exoplanet census includes so many hot Jupiters and supermassive planets that, by the standards of our own solar system, are so exotic.

Bigger telescopes and more sensitive instruments can enable astronomers to discover smaller exoplanets. Astronomers have to date discovered a handful of so-called “super-Earth” planets with masses ten times that of the Earth or less. One example is HD 7924 b, a recently discovered planet weighing in at about nine Earth masses. “We hope that with the Doppler technique we hope to find planets well under nine Earth masses, maybe even well under three or four Earth masses, by improving our Doppler precision and getting more observations,” said Geoff Marcy, a University of California Berkeley astronomer and one of the leading exoplanet discoverers, during a talk at the American Association for the Advancement of Science (AAAS) annual conference in Chicago last month.

Getting below that, though, might be too much of a challenge even for the biggest terrestrial telescopes and most advanced spectrometers. Enter Kepler. The spacecraft eschews the radial velocity technique in favor of looking for transits, when a planet passes through the line of sight between its star and the observer. When a transit takes place the star’s brightness dims by a tiny fraction; repeated, periodic transits reveal not just the planet’s orbit and mass but its radius as well (based on the fraction of starlight blocked by the planet during transits) and thus its density. “The technique is understandable to a kindergartner,” quipped Marcy, who is a co-investigator on the mission.

While the technique may be simple, its application to looking for Earth-sized exoplanets is more difficult. Terrestrial telescopes have detected dozens of exoplanets with transits, and a French spacecraft, CoRoT (a convoluted acronym for “Convection Rotation and planetary Transits”), has detected several, including a super-Earth announced a month ago. However, the transit technique requires the planet to be almost perfectly aligned with the Earth: perhaps only one percent of solar systems would qualify, given the random distribution of orientations. It also requires a sensitive detector: an Earth-sized planet transiting a Sun-like star would cause the star’s brightness to drop by less than one one-hundredth of one percent. And such transits would last only a matter of hours, taking place once a year in an Earth-like orbit.

All these factors drive the development of a dedicated space-based instrument that can continuously monitor a large number of stars, looking for a handful of transits that might be missed by other telescopes. That is exactly the mission Kepler is designed to do. The spacecraft is essentially a flying photometer: an instrument designed to measure changes in stellar brightness as small as 20 parts per million. Starlight is captured by the spacecraft’s 1.4-meter primary mirror and focused on a 95-megapixel CCD array. When pointed at a particularly dense star field, like the region of the Orion spiral arm selected by project scientists, Kepler can monitor over 100,000 stars simultaneously.

Because multiple transit detections are needed to confirm an exoplanet discovery, it will take essentially the entire 3.5-year mission to detect planets in Earth-like orbits. (Closer in planets with shorter periods will be found more quickly; Kepler should be able to find any planets in Mercury-like orbits in its first year of observations.) “In about three or four years from now there will be a press conference at NASA Headquarters,” predicted Alan Boss, a Carnegie Institution of Washington astronomer who studies extrasolar planets and recently published a book on the topic (see “Review: The Crowded Universe”, The Space Review, February 16, 2009), where Kepler project officials “will stand up and tell us just how frequently Earths occur.”

Life, intelligent and otherwise

While scientists and the public wait for that eventual press conference, there’s no shortage of speculation today about the frequency of extrasolar Earths. Boss believes that such planets are common. “We’re on the verge of finding out how frequently habitable planets occur in the universe,” Boss said during a press conference at the AAAS conference. “I think we’re going to find out that number is very close to one,” referring to the fraction of Sun-like stars having Earth-like planets.

“We’re on the verge of finding out how frequently habitable planets occur in the universe,” Boss said. “I think we’re going to find out that number is very close to one,” referring to the fraction of Sun-like stars having Earth-like planets.

Boss based that optimistic prediction on a projection of existing observations of larger super-Earths. “The radial velocity studies imply that something like one third of solar-type stars have objects in the super-Earth mass range, say five to ten Earth masses, on relatively short-period orbits,” he said. “So I would claim that if you have that many ‘oddball’ planets, they’re just the tip of the iceberg of a much larger population of lower-mass planets. It’s that much harder to build a ten-Earth-mass object than it is a one-Earth-mass object.”

Boss said if that line of thinking is correct, the average number of Earth-like planets around Sun-like stars is close to one. Multiply that by the number of G-type stars in the galaxy—about 100 billion, Boss said—and by the number of galaxies in the universe—also about 100 billion—and you end up with a staggering number of potential Earth-like worlds in the universe: on the order of 1022, or ten billion trillion. “Give or take a few,” he added.

“I think we would be absolutely astonished if Kepler and CoRoT didn’t find any Earth-like planets, because we’re basically finding them already,” Boss said, referring to the discovery of super-Earths in relatively close orbits. “I think what we’re going to find is that the number of Earths is quite large.”

Part of this interest in detecting Earth-like worlds is to better understand the distribution of exoplanet sizes. In his AAAS talk Marcy showed a histogram of exoplanets organized by mass. “The real holy grail for the next decade is to fill in this diagram,” he said.

However, the interest in such planets, particularly among the general public, has little to do with filling in gaps in charts. “It’s not just Earth-like planets that many members of the public and even Congress would like us to detect, it’s habitable Earth-like planets,” Marcy said.

Detecting habitable Earth-like planets will have to wait for a new generation of spacebased telescopes like NASA’s Space Interferometry Mission (SIM) and Terrestrial Planet Finder (TPF) spacecraft and ESA’s Darwin mission. Those missions have been in the planning stages for years and are unlikely to launch for years to come, in part because of their high costs and budget constraints. Nonetheless, Boss believes that these missions will benefit from Kepler and CoRoT, since the number of potential targets for these follow-on missions is “going to tell us how to build the next telescopes that are going to examine these planets.”

And, arguably, the real interest is not in habitable worlds, but inhabited worlds, which in turn raises the question of the prospects for intelligent life. Marcy, using more conservative estimates for the number of stars with planetary systems, estimated that there are 30 billion planetary systems in our galaxy. “The most pessimistic answer that most people will agree to is about one in a million,” he said, referring to the probability of a habitable world supporting intelligent life. “Then you do the math, and you see that a millionth of 30 billion leaves you with not just thousands of civilizations, but many of them millions of years more evolved compared to us.”

“I think we would be absolutely astonished if Kepler and CoRoT didn’t find any Earth-like planets, because we’re basically finding them already,” Boss said.

This then raises what’s know as the Fermi Paradox: if intelligent civilizations are that common in the universe, why haven’t we seen any evidence of them? Marcy raised several possibilities why that “pessimistic” one-in-a-million estimate might be “grossly wrong”. The Earth, he noted, is about 0.03 percent water: a little bit less—say, 0.01 percent—and the planet would be as dry as Mars; a little bit more—0.05 percent—and the planet would be a water world that could support life but not a technological civilization. There’s also no guarantee that evolution would always result in the development in intelligent species. And, he added, if technological civilizations do develop, they may not last long enough to overlap with one another.

All that, Marcy freely admitted, was speculation. However, the scientific foundation that speculation is built upon is getting firmer, with Kepler set to help scientists take the next step and understand just how common Earth-like planets may be. Another scientist at the AAAS conference noted that the field of exoplanet research was evolving from “butterfly collecting” to studying the nature of these planets. While that may be true, there’s still at least one more round of collecting to do, this time for the most elusive—and prized—specimens.