NASA's Kepler mission, designed to find extrasolar planets transiting across the faces of their stars, is off to a great start, project scientists announced Monday morning (Jan. 4, 2010) at the convention of the American Astronomical Society (AAS) in Washington, DC. Attending their presentation was S&T editor in chief Robert Naeye. He rushed me back his voluminous notes.

Kepler in space

An artist's impression of the Kepler spacecraft. It is basically a wide-field, 95-megapixel photometer fed by a 36-inch (0.95-cm) telescope.

NASA / Ames Research Center

Kepler science team leader William Borucki announced that Kepler has already logged more than 100 transiting-planet candidates, but said most of these will probably prove to be false alarms once followups are done by other methods. The team did release detailed information on five fully confirmed Kepler exoplanets that were discovered in just the first 43 days of data-taking last May and June, following the mission's launch last March. In these cases, radial-velocity measurements made by ground-based observatories showed the transiting objects to have planet-like masses, based on their stars' slight wobbles due to the planets' gravitational influence.

Four of the five new worlds are puffed-up hot Jupiters (with 1.3 to 1.5 times Jupiter's diameter) orbiting very close to their stars. One of them, named Kepler 7b, is among the lowest-density planets yet discovered, with a measured diameter and mass that yield an average density of just 0.17 grams per cubic centimeter (compared to Saturn’s 0.69 and Earth’s 5.52). That's the density of styrofoam. The result is “something theoreticians will be delighted to look at in terms of the structure of this planet,” Borucki said.

The other planet is a hot Neptune with about our own Neptune’s density — even though it orbits so close to its star that its surface layer should be roasted to 1,900°C. All five, in fact, should have outer-layer temperatures hotter than lava, and two should be hotter than molten iron.

Such hot worlds in fast, close orbits are the likeliest to transit their stars from our viewpoint, and are the most readily detectable if they do, and are the most quickly confirmable as well. Longer-period planets orbiting farther out, in a star's habitable zone for life where climates are more reasonable, with require a few years to find and confirm.

Lots more planets are coming — probably hundreds by the time the mission is scheduled to end three years from now. The host stars in all five cases announced today are somewhat larger and brighter than our Sun, with 1.4 to 2.0 times the Sun’s diameter. This is because these stars were chosen for early followups based on showing many narrow spectral lines, good for radial-velocity measurements. Such large stars are not very common; this bodes well for greater numbers of planets to be found around smaller dwarf stars, which are much more abundant.

In the first 43 days of data-taking, Kepler found about 175 transit candidates. Fifty of these were scrutinized to find the five confirmed planets that were announced. Some 125 candidates remain, and that's just from the first six weeks of data. Tidal waves of subsequent data are already in hand.

Kepler's celestial targets

The fields of view covered by Kepler's CCD chips sample a rich area of the Milky Way. More detailed map.

NASA / Carter Roberts

Kepler is watching, nearly continuously, a selection of about 156,000 stars from 9th to 15th magnitude (out of 4.6 million in its field of view) in a patch of sky covering 100 square degrees a little north of the plane of the Milky Way between Vega and Deneb, as shown at right. The satellite will keep watching these stars nearly 24/7 for at least 3½ years, in order to catch at least three transits of any luckily aligned planets that are in wide, Earth-like, 1-year orbits.

Any such planet has only a 1-in-200 chance of being in an orbit that's oriented just right to cross a Sun-sized host star as seen from our viewpoint. That's one reason why Kepler is watching so many stars.

Another is statistics. Kepler is intended not just to identify a few individual exo-Earths. It was designed, Borucki stressed, to watch enough stars to give a firm statistical reading on the abundance — or rarity — of terrestrial-size planets generally, throughout the galaxy and the universe.

Kepler's primary mirror in cleanroom

The honeycombed blank for Kepler's primary mirror, 1.4 meters wide. The mirror yields an effective aperture of 0.95 meter (36 inches) in the telescope's wide-field design. Even with such a large aperture, Kepler generally takes exposures 30 minutes long to build up good photon statistics.

Ball Aerospace

Kepler is routinely achieving 1-part-in-40,000 brightness precision (0.000025 magnitude) for measurements of 12th-magnitude stars. That is good enough to find transits of worlds as small as Earth, as planned.

Variable Stars, Oscillating Stars, Rotating Stars. . .

Some other news from the team's press conference and just-released papers:

• Kepler’s measurements are so precise that most “false positives,” such as an eclipsing binary star blended with the image of another star, can probably be weeded out upfront, without tedious and expensive radial-velocity measurements from the ground. This is very good news. Eclipsing binaries are the main source of false “transits.”

Transiting exoplanet HAT-P-7b

Early observations from Kepler revealed telltale dips as the known exoplanet HAT-P-7b passed behind (center) and in front of (left and right) its host star. Note the overall light curve's gradual rise and fall throughout the planet's 2.2-day orbit.

W. Borucki & others / Science

• For hot Jupiters — large bodies dazzlingly lit by their stars — Kepler can even see the "secondary eclipse": the planet's light being blocked by the star when the planet passes behind it, as shown at right. The depth of the secondary eclipse, combined with the planet's known diameter from the primary transit, can help tell the planet's albedo (reflectivity; how light or dark it is). Details of the phase effect, or how the reflected light changes as the planet waxes and wanes from crescent to full and back again, should tell more about its atmosphere. Moreover, the light curve also indicates how the shape of the star itself is slightly distorted by a close giant planet's tidal effect. (More on the secondary eclipse above.)

• For a few stars, Kepler has measured slight surface oscillations like those on the Sun. These arise from low-frequency sound (pressure) waves resonating through a star’s body. These resonant oscillations reveal a star’s diameter, mass, and state of evolution with very high precision — refining, in turn, the diameter, mass, orbit, and age of any planet it may have.

For instance, a Kepler oscillation study has refined the diameter of the star HAT-P-7 from an uncertainty of 10% to 1%, which in turn has refined the density of its previously known planet from 50% to 5% uncertainty. Ronald Gilliland, a stellar-oscillations specialist at the Space Telescope Science Institute, said at a press conference later in the day, “The data coming back from Kepler for stellar seismology is just spectacular. We expect it to reinvigorate and revolutionize our field.”

• Thousands of new variable stars are turning up, and Kepler’s extremely high-precision, near-continuous light curves offer rich material for new study. For instance, a third of the stars most similar to our Sun turn out to have tiny, short-term variabilities greater than the Sun’s. But not by much. And in time, we may be able to see the equivalent of our own star's sunspot cycle.

• A star’s placement on the detector’s array of pixels yields its position to better than a thousandth of a pixel-width, or 4 milliarcseconds. This is that way that many eclipsing binary stars are being weeded out, by their proper-motion wobbles. Such precision will also provide the best-yet parallaxes (distances) for most of the faint stars in Kepler’s catalog. Good distances are needed to help to characterize any planets the stars are seen to have.

• Slight, periodic variations can also reveal a star’s rotation rate, due to temporary starspots rotating in and out of view. In this way Kepler scientists expect to better calibrate the relation between a star’s rotation rate and its age and mass, the most widely useful way to assign ages to individual stars everywhere.

• Two strange, very hot objects that Kepler found orbiting white, type-A stars appear to be too hot to be planets but too low-mass to be stars. These may be an odd class of "not-really-white-dwarfs" — in which a normal white dwarf has shed all but about 0.15 solar mass to its companion star, and has thus enlarged beyond normal white-dwarf size due to its lower mass, gravity, and degree of compression.

Only the Beginning

So far, the Kepler team members have been holding most of the satellite's data close to the vest as they work on planet confirmations themselves. This morning Borucki said that Kepler data will be released publicly on a regular basis starting in June 2010, around the one-year anniversary of the start of its regular science observations.

At the afternoon press conference, independent commentator Caty Pilachowski (Indiana University) summed up astronomers' excitement. The new finds, she said, "represent the first of a new breed"; Kepler can, for the first time, compile an unbiased sample of transiting exoplanets free of many selection effects, to "get a true picture of what planets inhabit close-in zones and eventually farther out. What we’ll learn from Kepler five years from now will be astounding — what percentage are rocky planets, steam planets, or Jupiter-like." So, she urged the reporters present, please be patient.

One important piece of Kepler news, she said, is the finding that Sun-like stars are generally quiescent, showing only very small brightness variations. That's good news for astrobiology, and it's also good news for us. It tells us that solar-type stars spend most of their time in quiescent states, so we ourselves are in a good, safe place, not just enjoying a lucky quiet time next to a star that may act up.

"The precision is so exquisite, it makes future of stellar astronomy very exciting," she said. "What we can learn from slow rumbles of red giants is very exciting; we will understand insides of these stars, it will tell us about detailed evolution of stars going from tip of the main sequence to red giants. The statistics on the fraction of binary stars will lead to new science results . . . Kepler will give us all of that data. It will change everything about how we do astronomy."

See NASA's press release on today's announcements.

View the slides from the press conference.

Here are the past year's research papers from and about Kepler on the astronomy preprint archive,

The five new worlds will bring the total in the Extrasolar Planets Encyclopaedia up to 420, including 69 that have been seen to transit.


Image of Frank J. Cernik

Frank J. Cernik

January 9, 2010 at 3:56 am

I find it to be absolutely mind boggling how much information can be culled from just staring at a patch of sky non-stop.

The precision with which the light curves of so many stars can be tracked, the data kept discrete for each one, is just amazing! The atmosphere of Earth certainly limits what can be detected from the ground. It's just incredible the kind of precision that can be achieved by placing the detectors above the atmosphere.

It also says a lot about how computers have revolutionized recording and organizing data, as well as how CCD's have gone so far beyond what film cameras and, of course, the human eye are capable of.

Kepler shows, even at this early stage (43 days), just how little we really know about the Universe, and just hints at how much there still is to learn about it! If Kepler is typical of the Space observatories' previous track records of actual performed mission time compared to planned minimum time, we will be finding things in that small patch of sky that we can hardly imagine now.

Undoubtedly, as time goes on, Kepler will give indications of possible Earth-type planetary objects.

Maybe even signs of advanced,intelligent life out there. Could a "styrofoam ball" density planet possibly be a "cloud" of artifacts orbiting a close-in small planet? a hollow constructed object such as a Dyson-type sphere? or even a "simple" circular collector array? Maybe it is too close to it's star to be habitable, but just maybe a close in "stellar" energy collector could be sending power to another, habitable world farther out from it's parent star. We need to keep an open mind about the possibilities.

Those light curves could tell us a lot...

It's something to think about, isn't it?

Clear Skies,


You must be logged in to post a comment.

Image of Rod


January 9, 2010 at 8:51 am

Concerning the crop of 422 exoplanets currently reported the average mass is 3.07 Jupiters, the average radius is 1.194 Jupiters and the average host star mass is 1.1 solar. 27% orbit <= 0.1 AU from the host star. We are not seeing evidence that the Milky Way is teeming with habitable earths and extraterrestrial life in these observations.

You must be logged in to post a comment.

Image of Matt


January 10, 2010 at 4:52 am

> We are not seeing evidence that the Milky Way
> is teeming with habitable earths and
> extraterrestrial life in these observations.

Not yet, anyway. But there's no big mystery why we're finding Jupiter-sized planets first: they're the easiest to find. If we were looking at our own solar system at the distances involved, we would have detected Jupiter and maybe Saturn, possibly even Uranus and Neptune, but we almost certainly would not have detected the inner rocky planets.

The abundance of hot Jupiters out there suggests that many or perhaps even most stellar systems evolve differently from how ours has. On the other hand, if only 1 in 100 systems has rocky worlds in the habitable zone, that still leaves many billions of candidates in the Milky Way. That doesn't mean that we will find life, or that we won't. The search has just barely started in earnest.

You must be logged in to post a comment.

Image of Rod


January 11, 2010 at 5:26 pm

Matt, good observation about hot Jupiters and the formation of our solar system, these are a problem when we consider that about 27-30% of the exoplanets discovered are hot Jupiters. Concerning the 422 exoplanets in the library now, 33 show masses < 0.05 Jupiter or less than Neptune. This group has an average mass of 0.024 Jupiter (close to 8 earth masses), average semi-major axis is 0.282 AU (closer than Mercury is to our Sun) and average host star mass of 0.63 solar so many of the host stars are red dwarfs. I remain skeptical that the Milky Way is teeming with habitable earths and extraterrestrial life which according to Darwinian evolution, evolved from non-living matter vs. created by God.

You must be logged in to post a comment.

Image of Marc Dubbeldam

Marc Dubbeldam

January 19, 2010 at 5:36 am

Rod, the statistics you present mean absolutely NOTHING as the sample on which they are based is fundamentally biased by the techniques currently used for detecting exoplanets. Radial velocity measurements and earth-based photometry are simply not sensitive enough to detect low-mass planets in ~ 1 AU orbits. These will only show up after several years of observations with space-based instruments like Kepler. Only then will we have a statistically more relevant sample (but by no means complete or perfect, as Kepler is still biased towards finding hot Jupiters!) which can tell us whether our solar system is the norm or an anomaly. My money is on the former.

Whether any of these new-found planets are inhabited by (intelligent) life is another matter. I personally do not believe in aliens; I sometimes struggle to convince myself that there is "intelligent" life on Earth 😉

You must be logged in to post a comment.

You must be logged in to post a comment.