NASA's Kepler Spots Its First Rocky Exoplanet 97
coondoggie writes "NASA today said its star-gazing satellite Kepler has identified its first rocky planet orbiting a sun similar to our own — 560 light years from our solar system. While not in an area of space considered habitable, the rocky planet known as Kepler-10b is never-the-less significant because it showcases the ability of Kepler to find and track such small exoplanetary movements. 'Kepler's ultra-precise photometer measures the tiny decrease in a star's brightness that occurs when a planet crosses in front of it. The size of the planet can be derived from these periodic dips in brightness. The distance between the planet and the star is calculated by measuring the time between successive dips as the planet orbits the star. Kepler is the first NASA mission capable of finding Earth-size planets in or near the habitable zone, the region in a planetary system where liquid water can exist on the planet's surface. However, since it orbits once every 0.84 days, Kepler-10b is more than 20 times closer to its star than Mercury is to our sun and not in the habitable zone.'"
20x closer to sun than Mercury (Score:5, Informative)
That's less than 2 million miles, or .05 AU from the sun.
Quite toasty.
Re:Plane (Score:5, Informative)
Yes. Mostly. For this (transit photometry) method.
There are several methods of finding an extrasolar planet.
Briefly:
1) Pulsar variations: If a planet orbits a pulsar, the pulsar's timing will vary in a manner that can be detected by us, and we can use 3-D trig to figure out relevant parameters such as mass and radial distance.
2) Doppler shift of a star's emission lines: If a planet orbits a solar-type star, we can use the star's doppler shift of certain spectra to determine the various parameters of the body (or bodies) orbiting the star.
3) Gravitational microlensing: If two stars align just right to create a microlensing effect, the star further from us will show up as several images or as an Einstein ring, and its brightness will be amplified. If there's a planet orbiting the star that's closer to us, those mirror images or the ring will change with time, and they will be a bit brighter than without the planet.
4) Astrometry (measurements of the variation of a star's position relative to the "plane of the sky"): If there's a massive planet with an eccentric orbit, the star will orbit a barycenter that's outside of its mass, causing the star to move relative to the background.
5) Direct imaging: with certain techniques for processing stellar imagery, we can detect whether or not there's a planet reflecting some of that star's light to us.
6) Transit photometry: observing the star's brightness decrease as the planet eclipses the star. This works best for planets with a perfect orbital alignment with us, but we can still detect and work out minimum values for the relevant parameters.
7) Radio flux: Certain jovian-type planets can emit radio fluxes that differ significantly from most stars. These fluxes can be difficult, though not impossible, to detect from the interstellar noise.
There are more methods...
Re:Is there a doctor (of astronomy) in the house? (Score:5, Informative)
The Roche limit is defined as:
d = R ( 2 rhoM/rhom) ^ (1/3).
d is the orbital distance.
R is the primary (star in this case) radius.
rhoM is the primary's density.
rhom is the satellite's density.
If rhom > 2 rhoM, d is inside the radius of the primary.
The star in question is similar to ours, so I'll use our sun's density: 1.4 g/cm^3
The planet's density is 8.8 g/cm^3.
Therefore, the roche limit is within the star's radius and the planet will not be ripped apart.
This presumes a nearly circular orbit, which is good enough for this case.
Re:Wonder what the discovery curve is expected to (Score:4, Informative)
And that is only to find a single transit. Then add another year to get the orbit, probably another year at least to confirm.
Well, probably yes, assuming they're looking for yearly (like Earth's) orbits. Makes a bit of sense, but an Earth-like planet might be closer or further away from its host star, and be perfectly OK for liquid water, life, all that (depending on the host star's energy output). Probably not very different from a year though, it rather depends on the sizes (mass and orbital radius) involved.
As for the confirmation, it might not get that long; since the dip might be a starspot or a different agent, a Doppler effect study (or astrometry, in the future) might confirm or dismiss it because, to some extent, different methods of detection can be used on the same source for confirmation. Though, on the Kepler mission, I think the confirmation is 'included' and the timeframe is set for 3 years.
To me it seems that it is going to be a very slow start (apart from these totally hotrock type planets with insanely quick orbit) but then the taps will be turned on and they will start finding exponentially more and more?
Hopefully yes. For the moment the methods of detection are biased- each of them is capable of locating specific groups of planets based on two parameters, those parameters being the planetary mass and its distance from its host star- there's also gravitational lensing that can 'see' better, but its a one-timer.
Encouragingly enough, if one plots the findings so far (mass vs orbital distance) it is not hard to imagine that the so far covered areas will start to expand. My point being that, before the Kepler mission, 'hot Jupiters' kept being the majority of bodies discovered, because they are the only ones we had the means detecting- Kepler has been watching the same patch of space, and it should see more than 'hot Jupiters' (provided they're out there and we are going around this the right way).
Re:So to find earth... (Score:4, Informative)
So to find a truly earthlike planet, won't they have to focus on a single star for more than a year in order to detect the planet passing the star more than once?
Yep. And for Jupiter-like planet we'd need to be watching it for hundreds, if not thousands of years if we were to use this method.
What if the planet's orbit never aligns to eclipse the sun?
Then we would never detect it via this method.
What if there are two or three planets in very similar orbits?
It depends on how well they are aligned. Even if they're perfectly aligned, we're liable to see the first one before the second or third one as it passes in front of the star. If they are even slightly out of phase, they will eventually be in an orbit in which we see all three distinctly. In any case, the radius and shape of the occlusion in front of the star is determined by the shape of the light intensity vs. time graph. Circular disks have a very specific light occlusion shape, while abberant occlusions have different shapes.