The whole thing is on the official NASA TV youtube channel http://www.youtube.com/watch?v=qRN7fNkZ-IQ

I'm extremely excited. But there's no way to detect the composition of the atmosphere, right?

Depends on the planet. If the star is sufficiently bright then you might get a high enough SNR to say something about the atmosphere (i.e. like HD 209458 b and HD 189733 b, but these planets are exceptional cases). It won't be done with Kepler though, all it has is a photometer.

That came up in the press conference, and the answer is that Kepler can't do that, but some of the future missions being proposed would be able to.

By the way, in case anyone missed it, this is what a Jovian planet looks like when you have four transits to work with. Had this been Earth and the Sun, the transit (the big drop) would have been about 2/3 the depth of the occultation (the little drop). So, yes, Earth-like planets should be detectable, but it's going to be close.

Something else interesting from the conference and some of the links was that the light curves of the variable stars are often unlike anything in the literature. Apparently the atmospheric noise has been hiding significant behavior. This complicates finding planets, of course, since their models for spotting variable stars have to be reworked.

--Greg

ustrax's couple writeups were extremely interesting.

I thought it was cool that the warm Spitzer was mentioned as an observatory likely to follow up on new discoveries. I love it when old missions come in handy.

I have some questions, if anyone can answer them.

1. Kepler's mission is only 3.5 years, barely enough time to confirm it's own initial discoveries. It sounds likely that the mission can be extended, so what is the expected longevity of the mission (assuming funding is not the problem)?

2. From the mission website: "Expected Results:

From transits of terrestrial planets in one year orbits:

About 50 planets if most are the same size as Earth (R~1.0 Re) and none larger,
About 185 planets if most have a size of R~1.3 Re,
About 640 planets if most have a size of R~2.2 Re,
About 12% with two or more planets per system. "

-- would these expected numbers scale linearly with mission extensions (i.e. would another 4 years of observing double these numbers)? Or is Kepler's field of view fixed to one region, so the sample set is difficult to change, and if this is the case, would most new planets found be longer orbital periods?

3. "Stellar evolution models are used to estimate the mass, radius and metalicity of the parent star"
-- how reliable are these models? Is there any way to directly determine these values, or does it require an instrument like the Terrestrial Planet Finder?

4. For gas giants found within a habital zone of the star, would it be possible to search for large moons with either Kepler or astrometry of the gas giant using ground-based instruments?

Four years or so I think. Fuel is the problem. Kepler's gyroscopes must be de-saturated every so often. Eventually the fuel will run out, the gyroscopes will become saturated, and the entire spacecraft will lose its ability to point.

Yes, it's fixed. As such, the useful data return will decline with age.

I would imagine very reliable.

We had that discussion earlier in the thread I think. If I remember, yes it's possible but very difficult.

That is not a free article. Anyone with access care to summarize (not plagiarize) the new and interesting parts, if any?

It's mostly the same things reported in the press conference yesterday. There is a good summary in Sky & Telescope's website
http://www.skyandtelescope.com/community/s...g/52657352.html

Kepler is using a defocused star image, right? Is it a defocused mirror? What if instead of a defocused mirror, they put in a lens with severe chromatic abberation. They'll still get defocused images, since the colors would be spread around the whole star, but a circular integral around the star could yield some useful information. Dumb idea, or should I call a patent lawyer?

IIRC, Kepler has a spherical mirror, which I think is much easier to make than parabolic mirrors needed to fix the spherical aberration. I guess there is not much point is going through the extra work of making it parabolic, just to defocus it with a secondary lens.

Metalicity is measured directly spectroscopically. Mass can be measured directly if the star is binary (or tertiaty...). Of course, for exoplanets the main focus of interest is on single stars (as least partly due to the old discussion about possibility of habitable planets orbiting binary stars) and this is one of the reasons to use stellar evolution models.

There was some discussion of exoplanet moons with help from the transit timing method in this post (#158):

http://www.unmannedspaceflight.com/index.p...mp;#entry139639

And I think I read of the possibility of "amateur" transit timings even contributing to the search. Was this in Sky and Telescope or somewhere? There are some websites mentioning this as well.

We can get masses for members of binary stars. The standard relations between mass and luminosity come from members of widely separated binaries, where the stars are too small compared to the orbits to have affected one another's evolution (yet). Here is a typical set of mass-luminosity and mass-radius relations for main-sequence stars. Eyeballing that scatter, it looks like 20% in luminosity if mass is known or 10% in mass if luminosity is known (since it's a steep function). Radius looks a bit worse; that often has to come from blackbody laws and the effective temperature and luminosity; which come from spectroscopy and from photometry plus parallax. That can be improved for stars not too distant; I saw a result from the CHARA interferometer in which they resolved the disk of one of the stars with a transiting planet, reducing its uncertainty in radius. (They are a long way from getting an interferometric signal from the dark planetary disk, alas).

As has been posted already, we get metallicity from spectroscopy, calibrated to the Sun. (Kind of odd that the best-fitting spectroscopic oxygen abundance there is not the best-fitting one for helioseismology. A lot of other things may shift a bit when that gets sorted out).

I'm quite curious to learn more about the light curves of some of the other variable stars, not necessarily curves which look like they may be planet transits. It sounds like they may have unique and exciting data. I wonder where and when these other data will be reported.
Wouldn't it be something if it turns out that Kepler's Earth sized planets are a minor part of what its data yields?
steve