A binary 'hot Jupiter' object would be a very weird thing indeed. Would they be tide locked to each other, or would both be tide locked to the host star ?

If the double planets were not mutually tide locked to each other, the dissipation and heating would be large. So too if they are not tide locked to the host star.

I'm thinking as fun to study as such a system might be, there aren't going to be any real examples . . .

Hi! About such strange binnary planet, i remember the excellent story wrote by Robert L. Forward, "The flight of Dragonfly", where there is a planet named "Rocheworld", about the name of a matematician Edouard Roche who calculated the possibily and stability of a double planet.

Another science download completed, with information stored since early July.

Link

Given the reported unknowns being discovered about new classes of stellar variability, I may have to revise my earlier confidant predictions about coming up with early estimated numbers of planets being seen based on statistics. Apparently, it might be harder to do that.

Just noticed these powerpoint slides on the Kepler site:

http://www.kepler.arc.nasa.gov/ed/ppt/Boru...resentation.ppt

These are from a talk the PI gave at the IAU meeting last month. Most interesting (for me) was a dozen or so light curves from various stars illustrating things the Kepler team haven't managed to explain yet.

--Greg

Very interesting. I especially like the middle and bottom figures in page 11. I doubt a binary system can cause these light curves, most likely a tertiary system (two close binaries with an out of plane third star).

One thing I've been wondering is how much Kepler can tell about the shapes of stars. Since so many stars spin really fast (often ~90% of breakup velocity) there are a lot of oblate stars out there. Now the Von Zeipel effect means that the equatorial regions of oblate stars are darker than the polar regions (due to a lower temperature) -- this on its own doesn't help us though because a spinning oblate star doesn't result in any apparent change in brightness from the point of view of Kepler. However, I have been wondering if there is a star analogue of Haumea -- which is a Kuiper belt object that spins so fast that hydrostatic equilibrium results in a more cigar shaped/rugby ball object (prolate isn't correct because that implies a different rotational axis) rather than a oblate spheroid). I presume that this occurs at even higher spin rates, just before breakup velocity. If stars like this exist, then different regions of the star's equator would be at different temperatures -- this means that if the spin axis is suitably aligned it should be possible to determine the stars spin rate and shape based on the light curve. This could be backed up by further investigations using optical interferometry.

Anyway... Just a random thought! If this does occur, it must be in the lightcurve data somewhere.

There's an interesting Van Karman lecture on using optical interferometry to measure star shapes
http://www.jpl.nasa.gov/events/lectures/mar06.cfm
I have no idea why the main Von Karman page doesn't link to the pre-2008 lectures any more -- I even sent a note to the webmaster, sigh!

"Many candidates have been found and are being examined with RV observations to check for false positives and to measure masses."

Even though they stated elsewhere in the document that most of these would be false positives, it is still welcome news.

All nine figures look at least a LITTLE weird. Offering my non-professional opinion (in hopes of teasing a real professional opinion from someone) :-)

Starting with the top one on page 9:

1) Looks like a textbook curve for an eclipsing binary star EXCEPT where is the secondary eclipse and why are the two curves so different? (Is it possible that the second dip actually IS the secondary eclipse and that we're seeing a single period of two almost-equal eclipsing stars)?

2) I'm trying to figure out how "non-aligned spin axes of hot, fast-rotating stars" explains this one.

3) Definitely not a full period. Wonder if this is actually variability in the star itself? Perhaps some sort of flare?

Looking at page 10:

4) Looks like a text-book illustration of a limb-darkened eclipsing binary star.

5) I loved the two-day transit. When they say "dwarf" of course they mean "dwarf star" (maybe brown dwarf) not "dwarf planet". Transit time is about what we'd expect for Saturn, but the depth is 6x bigger than for Jupiter. Since Jupiter is about as big as a planet can be (apparently super Jupiters are thought to have about the same diameter -- they're just denser) this really seems to need at least a brown dwarf -- if I'm not totally off base. :-)

6) What could possibly cause a sawtooth curve? Some kind of weird variable star?

Page 11

7) Look how wide the relative flux is for this one. Hope that's not hardware or software trouble with the sensor.

8) Like Sirivan, I don't think this could be just a single planet. However, I'm thinking "variable star" again -- not planet at all.

9) This looks to me a lot like the graph of two or three superimposed sinusoids. Again, looks like some complex kind of variable star, not a planet or binary star.

Page 12

10) I liked their illustration for this one, but it might be easier to believe it's a simple binary where the brighter star is brighter on one side than the other and rotates in a 1:2 resonance with the dimmer star.

Incredibly, this is still all just from 9.7 days of callibration data! They should have 10x that much by now, plus some time to double-check with ground-based telescopes. With any luck, they've got another paper or two in the pipeline by now.

--Greg

Thanks, Greg, that was really helpful.

BTW on 7) there - it's actually really low flux. if you look closely there's a "x 10^-3" that somehow ended up on top of the graph.

Also, 6) reminded me of a Cepheid variable, but admittedly I know next to nothing about all this and am probably wrong.

Well, well ...

Kepler could find habitable exomoons

Good catch. For some reason I had it in my mind that Cepheid's light curves were sinusoidal, but they're clearly not.

http://www.astrosociety.org/education/publ...realastro3.html

So, yeah, #6 looks like a classical Cepheid variable.

Good catch on the scale for #7. When we compare it to #3 (the only other one with a 10^-3) I guess it shows a randomly noisy star, but at a level that couldn't be seen without Kepler.

Then perhaps #8 and #9 are similar to "Delta Scuti" variable stars, since those can have multiple superimposed periods with periods of hours.

http://users.skynet.be/bho/researchprojects.htm

Here's the actual paper:

http://xxx.lanl.gov/PS_cache/arxiv/pdf/0907/0907.3909v2.pdf

He models both mass and orbital distance of the moons, down to a mass of 0.2 Earth (about 2x Mars). The basic idea would be to spot giant planets in earth-like orbits, then add them to Kepler's short list of frequently sampled targets (he needs one sample per minute during the transit, and Kepler only does a few hundred that frequently.) He depends on the fact that a moon will affect both the timing of the start of the transit and the duration of the transit. He needs to watch it for four years (so eating into the extended mission time) but he achieves the same detection confidence that Kepler does for Earth-like planets in general.

I wonder if he could reduce the data load merely by upping the sampling rate during the expected transit period.

--Greg

I was looking at the 100X HAT-P-7B light curve . . .

I've got to sleep on this and the above paper first, more later, thanx

I took some time to play with the math for computing what a light curve ought to look like, assuming round, uniformly-bright stars and round, completely dark planets. Some of the results surprised me, so I thought I'd share them. I made some graphs, which turned into tiny GIF files, so I hope it's okay to post them inline.

This first one is for a Jupiter-sized planet, radius 1/10 that of the star.

Click to view attachment

The x-axis is time, where -1 is start of transit for a right-across-the-equator transit and +1 is end of transit. The y-axis is the brightness, where 1 is normal. Obviously these curves go from 1 to 0.99. "Center" is a "perfect" transit. "Half" is a transit that crosses half-way between the equator and the pole. "Graze" is a transit that is only full for an instant. (That is, second and third contact are at the same time.) And "Miss" means the center of the planet just misses the edge of the star, but because the planet isn't a point, it does block some light.

The biggest surprise to me was that there's not much difference between the center transit and the half-way transit. You actually have to get pretty close to "graze" before you see much curve at the bottom.

The next one is Earth-sized, radius 1/100 that of the star.

Click to view attachment

The effect I noticed with Jupiter is much more pronounced. On reflection, it's obvious that a small planet is unlikely to have a "miss" type of transit. It has to be lined up just right.

Then, just for grins, I modeled a "planet" that's star-sized; 75% the size of its sun.

Click to view attachment

Now it's hard to get a flat-bottomed light curve because it's hard to get a "full" transit. The "planet" is much more likely to pass partly in front of its star, but not right in front.

So, in light of that, when I look at the HAT-P-7B light curve, the first thing that strikes me is that it looks like a grazing transit. However, HAT-P-7 is bigger than Sol, and HAT-P-7B is a bit bigger than Jupiter, but not by as much, so it's probably unlikely this really is a grazing transit. The limb-darkening effect (where the edge of the star is darker-looking than the middle) might explain it, or maybe I do need to treat the planet as not-entirely-dark.

Anyway, I had fun with this, and I'll happily share the equations with anyone who wants to play too.

--Greg

I would be quite interested in those equations, if you don't mind.

Sure. Let me know if you can't download this -- it's my first try at using skydrive.

http://cid-6f25d46a94c04426.skydrive.live....ht%20Curves.pdf

This is a bit rough, since I just threw it together on Saturday, but I'm pretty confident it's correct. If you see something wrong, don't hesitate to let me know.

The derivations don't require anything beyond 10th-grade algebra and trigonometry. There is a faster way to get there using multi-variable calculus (and you have to do it that way to get limb-darkening) but I didn't put that into this document.

--Greg

Thank-you so much for that, works great

A minor mission update has been posted, as usual the link is here.

Basicly everything is routine right now, and they are making some headway in understanding the safing events.

Perhaps the bigger question is whether stars exist which are significantly brighter on one side than the other -- for whatever reason.

--Greg

I thought maybe Kepler data would lend itself to studying surface phenomena like flares and star spots, but was surprised to see that the BIG thing is going to be using the data to study star interiors!

The Kepler team will use seismic techniques to probe the cores of a large number of stars.
"The quality of the Kepler data and the large number of stars observed are expected to lead to a huge step
forward in understanding of stellar evolution. During the first nine months in space Kepler will survey more
than 5000 stars for oscillations. Based on those measurements around 1100 stars will be followed for
detailed studies throughout the full mission. The accuracy with which Kepler will be able to measure stellar
oscillations is so high that the science team expects to watch directly the change in stars as they age."


There is an international group formed specifically to do these studies, the Kepler Asteroseismic Science Consortium.

One of the members, the Danish Asteroseismology Centre has an informative website with lots of pretty graphics which gives some basic background.

Another Mission Manager update:

http://www.kepler.arc.nasa.gov/about/manager.html

The second quarterly roll went well. Also, some of the data are now publicly available!

http://archive.stsci.edu/kepler/

Be interesting to see what's in there.

--Greg

An SQL query with Availability_Flag = 2 returns 165831 results.


I guess that means the science team is interested in ~166k targets, at least so far.

Yeah, it looks like they won't make the first quarter's raw data available for another twelve months. Then they'll release more of it annually. Oh well. At least they DO plan to make it all available eventually.

--Greg

That would be every star currently being watched for transits. All the giant stars they can identify are already off the list. The number will fall as more and more dwarf stars are eliminated as being too active to be good candidates. In a few months time it should be down to about 100K.

What was the reason for excluding giant stars? Are they all too variable? Or simply too bright to be able to measure the effect of a transit of even a giant planet?

--Greg

To tell the truth, I'm not really sure. I just know that they are being excluded. The fact that they are much more luminous does make it harder to detect planets, I'm certain. Other factors might be that the transit time will be much longer, allowing stellar variablity to play a larger noise factor. Also, they are concentrating on looking for solar system analogs with possible earth like worlds. Any that avoided being swallowed and still existed would be fried. Anything made habitable would have an orbital period too long for confirmation.

I saw some indications on internet search engines that guest observers were inquiring about the possibility of seaching for transits around sub-giants and giants, so there might be some kind of side program going on for these. But the web pages for those were pulled.

Holder of the Two Leashes when approximately report about discovery of the first planets? On conferences DPS 2009 the new discovery Kepler are expected?

As I understand it, all else held constant, brighter transits are easier to detect due to a higher S/N.
Giant stars have much larger angular surface areas, so when the planet transits across, the planet blocks a significantly lower fraction of light, producing much fainter transit signals.

Yes, you're right. Perhaps I didn't state that as well as we could. The surface brightness doesn't matter, the planet will be blocking the same fraction of light no matter what. Giants are more luminous due to a much larger surface as well as a much greater total energy output. The fraction of blocked light for any given planet size is less for larger stars. And the signal will be stronger depending on the apparent magnitude, as seen by Kepler and not how bright the star actually is. The fact that there are so many giants showing up and being ignored is simply due to the fact you can see them from further away.



Sorry, all I know is from public sources, so it's what I'm told and what I can infer.
They have stated that by the end of the year they will be ready to announce a lot of short period planets, those that have orbits measured in days.

If you want insider information, you're going to have to track down ustrax. Either that or maybe crack the code on one of his frickin anagrams .

If I'm not mistaken, Kepler's focus on finding Earth-sized planets seems to also lean a bit towards the serendipitous possibility of finding worlds around suitable stars in "Goldilocks zones", which would be one reason to exclude giant stars.

Another would be that smaller worlds probably aren't likely to be found around giants. Massive luminous stars often die too young to form mature planetary systems (or blow all the excess gas & dust away before one has the chance to form), and old red giants would tend to swallow up inner planets. At the other extreme, red dwarfs are thought to have extensive starspotting which would complicate detection confirmations considerably.

All in all, it's probably more statistically productive to focus on middle main-sequence dwarfs given Kepler's strategy & capabilities.

Since brightness of a star goes up as the fourth power of mass, it occurs to me that another problem presents itself: if a giant star did have a planet orbiting in the Goldilocks zone, that planet would have a period of decades or even centuries. There's simply no hope of seeing repeated transits.

--Greg

Indeed. Another thing to consider is that even if this planet is in the habitable zone, it hasn't been for long, and won't be much longer. Habitable Earth-like planets around giant stars seem unlikely.

The official objectives for Kepler aren't limited just to potentially habitable planets:

I personally think it would be quite interesting to learn what sort of planets, if any, orbit giant stars. However, it does seem that Kepler wouldn't be likely to find them, even if it were trying to -- for all the reasons we've already given.

--Greg

Several planets have been discovered orbiting giant stars through radial velocity. So far, most of them have been gas giant planets, probably the result of the biased nature of the radial velocity method.

Not that I can tell. The most massive star listed in the Extrasolar Planets Encyclopedia has a mass 4.5 times that of the Sun. Giant stars start at 10x the mass of the Sun.

http://www.exoplanet.eu/catalog-all.php?&a...=8&more=yes

I'd be interested to learn otherwise, though. What's your source?

EDIT: Nevermind; it's radius, not mass, and the Encyclopedia gives 21 like that -- none smaller than 2.6 times the mass of Jupiter.

--Greg

Ah, that sounds right. I couldn't recall at the time I made my post if all of them were gas planets, or if there was a Neptune somewhere in there, thus the use of "most".

During yesterday's Belgian Association for Astronomy's yearly meeting, some astronomers pointed out that the Kepler team will have to wait for the installation of new high precision spectrograph at Herschel Telescope at La Palma as HARPS in La Silla cannot reach Kepler's Field of View (Cygnus)...

A new Mission Manager update at Kepler:

http://www.kepler.arc.nasa.gov/about/manager.html

They're getting a handle on those earlier safing events -- something to do with a low-voltage power supply, but at least not a processor defect. No word on how they're planning to deal with it though.

--Greg

http://www.universetoday.com/2009/11/02/no...ler-until-2011/

I interpreted those reports to mean that data acquisition is fine... they just need new algorithms to massage the data, and writing the software will take a couple years. No?

The article mentioned uploading the new algorithms to Kepler. That would seem to me that the data processing change will occur on the spacecraft (assuming the article got it right). No telling what you can do with the data already on the ground, at least from this news item.

Sort of. I read it as...

The data acquisition is ok but not as good as hoped. The data is noisier than hoped. The noise can be removed with software algorithms. The software will take a couple years to develop.

It's a major bummer but it's also completely understandable. I really wish we could build another with all the kinks worked out and send it up next month, but that's just not the way it works.

Kepler mission uses double differential photometry?

I'm pretty sure so.

I.e. already collected for 6 months photometry is not possible to fix, since the processing is on board a space ship?

http://blogs.discovery.com/space_disco/200...until-2011.html

It would help a lot if we knew what they mean by 84 "channels". Are the data transmitted back at 84 different frequencies? If so, how are the results divided among them? Obviously not by position on the CCD.

This is probably written up somewhere, but my first shot at finding it came up dry.

--Greg

This article describes the channels in detail (in Russian)
http://infox.ru/science/universe/2009/11/0...netsearch.phtml

TheoA on the exoplanet forum said that that

There are 24 CCD's - each with two 'havles. Each half with two channels

http://kepler.nasa.gov/gif_files/Channel_numbers.jpg

Kepler does have a page on "differential photometry" but that doesn't really address the problem:

http://kepler.nasa.gov/sci/basis/diffphot.html

If anything, it implies problems with a channel should be limited to the stars on that channel. It definitely doesn't say "all the stars" -- it just says "nearby stars". Of course, it doesn't actually talk about "channels" at all . . .

--Greg