Sky & Telescope | Kelly Beatty | 2011 July 27
From all accounts, NASA's Kepler spacecraft has been an unabashed success since its launch 2½ years ago. In February, mission scientists announced that they'd found an astounding 1,235 candidate planets in just the first four months' observations. This cache brims with multiple-planet systems: 115 doubles, 45 triples, and 10 with at least four.An artist's impression of the Kepler spacecraft, which is essentially
a wide-field photometer fed by a 38-inch (0.95-cm) telescope.
(Credit: NASA / Ames Research Center)
Moreover, beyond its planet-hunting prowess, Kepler has returned paradigm-changing data on the nature of stars themselves, heralding what one researcher calls a "golden age for stellar physics." An estimated 500 astronomers are poring through the spacecraft's ultraprecise light curves, churning out roughly one new research paper per week on everything from a detailed census of stellar sizes and masses to using stars' subtle, rhythmic pulsations to probe their interiors.
But project managers have become quietly concerned that Kepler's top priority — finding true analogs of Earth circling Sunlike stars in temperate, life-friendly orbits — can't be achieved by the time the spacecraft completes its planned 3½-year-long mission just 16 months from now.
Kepler has been staring at a patch of sky about 10° across near the Cygnus-Lyra border, continually recording the brightness of more than 145,000 stars. It's designed to detect the tiny drop in a star's light as a planet passes in front of it. Spotting big planets in this way is a snap for the orbiting telescope's 36-inch (95-cm) aperture and supersensitive detector array.However, it's much more challenging to pick out the intermittent blips from Earth-size bodies, which create brightness drops of less than 0.01%. "We've always believed that there are more small planets than big ones," comments Sara Seager, an MIT exoplanet specialist. Yet among Kepler's 1,235 announced candidates, only 68 of them (fewer than 6%) are Earth-size — a tally far short of the anticipated windfall.In February 2011, Kepler scientists announced that the spacecraft
had identified more than 1,200 candidate planets. But the fraction
of Earth-size worlds was far below theorists' expectations.
(Credit: NASA / Kepler / Wendy Stenzel)
So are other Earths truly rare, or is something amiss with the spacecraft? The answer likely neither of these. Instead, it appears that (recasting Shakespeare), "The fault lies not in ourselves, but in our stars."
The first stirrings of a problem appeared early in the mission. "We weren't getting anywhere near the precision we should be," explains principal investigator William Borucki (NASA-Ames Research Center). Yet an end-to-end check showed that the spacecraft itself was behaving as expected: its pointing was rock solid (10 times better than the Hubble Space Telescope's stability) and electronic noise very low.
It soon became apparent that the problem lay with the targeted stars themselves. Kepler scientists had assumed that stars like the Sun would behave, well, just as the Sun does. Up-and-down churning of gas in the solar photosphere causes the Sun's brightness to vary by about 10 parts per million (10 ppm) over time scales of a few hours.But brightness fluctuations from the stars being watched by Kepler are greater than that — averaging about 20 ppm for 12th-magnitude dwarfs. This added noise makes their light curves messier and identifying transits by small bodies all the more difficult. "They're much more variable than the Sun," Borucki admits. "It was a big surprise to us."Kepler scientists expected solar-type stars to fluctuate in brightness by only
about 10 ppm, as the Sun does. But as this plot of 12th-magnitude stars
shows, they were surprised to find that the fluctuations are much greater.
(Credit: R. Gilliland et al / Astrophysical Journal)
The team estimates that a planet like ours, passing in front of a star like ours, should yield a brightness drop of about 85 ppm. "That's going to be a fairly subtle signature in the data," explains Ron Gilliland (Space Telescope Science Institute). Stars with lower masses than the Sun are much more common, but they're also smaller and dimmer. Unless they happen to be relatively close by, less of their light reaches Kepler's detectors and noisy detections become an even bigger problem.
The main reason Kepler has identified any Earth-size candidates is that a planet in a tight orbit will cross its star's face many times in just a few months. So, eventually, the recorded number of transits grows and a distinct repetitive pattern emerges even from small planets in the noisier-than-expected brightness measurements.But these close-in worlds are far too hot to be habitable. Kepler's raison d'être is to identify true analogs of Earth — small planets orbiting far enough from their stars to maintain nurturing temperatures. Mission scientists have known all along that these prized candidates would take the most time to spot, both because they cross in front of their host stars only about once per year (lasting at most 13 hours for an Earth-and-Sun clone) and because their much larger orbits have only a 1-in-200 chance of appearing exactly edge-on from our perspective.If stars were no "noisier" than the Sun, then Kepler should take less than
6 months to identify an Earth-size world circling a solar-type star in a
10-day-long orbit — and less than 4 years to spot one in a 300-day-long
orbit. But stars are noisier than the Sun, so more time is needed to make
such detections. (Credit: D. Koch et al / SPIE)
That's why Kepler is staring at so many stars, and that's why the basic mission is 3½ years long. Over that stretch a candidate planet circling roughly 100 million miles from its star (as we do) will complete three transits — the minimum to be considered statistically robust. But throw in the added difficulty of noisy stars, and the chance of finding true Earth analogs around Sunlike stars plummets.
More Time Needed
There is a way around this unexpected setback: the longer Kepler can stare, the better its odds of identifying small worlds. According to a comprehensive analysis just concluded by Gilliland and 16 colleagues, the spacecraft would need to amass at least 7 or 8 years of observations — double the planned mission length — to identify all the Earths passing in front of solar-type stars in the spacecraft's field of view.
Ordinarily, getting NASA managers to extend Kepler operations for a few more years would be nearly automatic. After all, even though the total mission cost is roughly $600 million, it would take no more than $17 million annually to keep it going. The spacecraft is healthy, with enough consumables aboard to last through most of this decade. But these are turbulent times for the space agency. Substantial cost and schedule overruns by the James Webb Space Telescope threaten to gobble up an ever-greater fraction of NASA's space-science budget. Worse, within the past month Congressional budgeteers have proposed not only to cancel JWST outright but to slash more than $1.9 billion from the Obama administration's NASA budget for fiscal year 2012.
Borucki remains hopeful that an extension will be approved. "I think this is the greatest mission that NASA has ever flown," he exults. "I can't imagine any other mission being more important." Borucki's bravado is understandable, given that he's championed Kepler and its cause for more than 25 years. Still, plenty of others agree with him.
"We certainly need an extended mission," says Geoff Marcy (University of California, Berkeley). "Kepler offers the clear ability to answer a profound question for humanity: are Earth-size, rocky planets common or rare in the Milky Way galaxy?"
Jon Morse, director of NASA's astrophysics division, understands all this but doesn't have "a crystal ball that can tell us what the outcome is going to be." For now, despite all the Congressional posturing, he's assuming that Kepler will get the $19 million requested for the upcoming fiscal year. That's enough for the project to collect observations through November 2012 and to analyze them for a year thereafter. But Morse acknowledges that greater forces are in play. "One of the hardest things we do is to plan our next year's activities in uncertain times."
To get Kepler's mission extended, first Morse and his staff must affirm that it's worth doing (a decision is due within the next two weeks). Then next spring a formal request will go to a scientific review committee, where, he says, "missions that project the highest science return in the future have priority."
Since Kepler can't complete its most basic objective without an extension, it figures to be high on the review panel's approval list. But if NASA money doesn't materialize, some in the exoplanet community feel private funding should be found. "Kepler is the political cornerstone of the exoplanet search," says Seager. "There's no other option."
Meanwhile, the next batch of planet candidates won't be announced until next June. Until then, the spacecraft's discoveries will keep researchers hopping. "Remarkably, planets of near-Earth-size are emerging," says Marcy, "and we are working day and night to confirm and validate them as bona-fide planets."
