They no have high-precision spectrographs to determine the masses of the planets. Perhaps only for detecting spectrally double stars, planetary false candidates. But it takes relatively little observational time.

Probably all telescopes with high-precision spectrographs in the northern hemisphere involvement for the project Kepler.

I understand that Kepler data are being made public after a certain proprietary period. The public can then look for things like variable stars that the science teams aren't focusing on. I wondering if the science teams will have time to sift for all the planets in the data, or if there would be any waiting to be found by the public (e.g. UMSFers)?

The place I think they might miss some would be planets orbiting variable stars and/or multiple stars. They seem to be focusing only on quiet, well-behaved stars.

--Greg

Bill Borucki's talk at the Exoplanets Rising conference this morning, "Status of Kepler Results," is already online here. I'm only about 5 minutes into the talk, but it appears to be mostly (if not entirely) a summary of what was already presented in January. Although someone besides me might enjoy watching it, too.

So was anything new announced there? I'm guessing not, since I'm not seeing anything on Emily's blog, nor the Kepler page, nor Spaceflight now. Sometimes they leak things in talks, but I'm too lazy to listen to a whole talk without at least some hint there's something special. ;-)

--Greg

Nothing new in terms of "announcements." He did present data on an apparent sixth planet as an example of how careful they have to be in follow-on observations. (In this case, there was a star in the background that was impossible to separate out by the Kepler instrument.) He said they are close to announcing this planet, though. He threw out a lot of teasers about how many "interesting signals" they are seeing, but of course didn't give direct answers to questions about unpublished data. If you're familiar with the data presented at the January AAS, then you might just want to fast forward the first 30 minutes of the presentation and only watch the Q&A session during the last 15 minutes. Some of the questioners were actually discussing unpublished work (in press, I think) about the anomaly of putative planets being hotter than their star, and Borucki commented that they're seeing a lot more than two of those. I'm assuming that the next exoplanet discovery announcements will come at the 216th AAS Meeting in Miami the last week of May. The timing and venue both seem right: another six weeks or so of follow on observations, plus a month to write up papers and get them safely in the publication pipeline before disclosing the data.

The oddball hot companions are looking more like white dwarfs. This paper evaluated the secondary masses from Kepler data in a more direct way than the discovery paper, and found them to be more reasonable for white dwarfs (especially if the binaries had a history of mass transfer). I was blown away by their technique - the Kepler light curves are so precise that they can measure changes in stellar radial velocities from Doppler shifts as reflected solely in the arrival rate of photons, to 1 km/s accuracy.

Alien Planet Hunter Develops a Blind Spot

hmmm I assumed this was old news.

It would take years to find an earth orbiting one of these stars. Wouldn't we get into mission extension before we hit the motherlode?

@Sunspot Yeah, I don't see anything new on the topic, nor does the New Scientist article appear to link to anything new.

@brellis The original plan was four years, of which we have three to go. Mission extension could add two more years, but it's not essential. Essentially, the plan is that when they think they've seen two transits, they predict a third. If it shows up on schedule, that call it a planet. For an Earth-type planet in an Earth-type orbit, that means you need from three to four years, depending on when you observed the first transit during the first year.

Anything in a 100-day orbit or less ought to be at about the 3-transit point by now. I'll be surprised if there aren't any Earth-sized planets in close orbits -- especially around smaller stars. Once they're comfortable with the data they're getting and the validity of their analysis, I expect they'll deliver a huge dump of planets in that range. If this next report still limits itself to planets large enough to be verified by ground-based telescopes, that'll suggest to me that they're still not quiet confident in something.

--Greg

An interesting thing to ponder is that if any Earth-sized planets exist in 1 AU orbits around the sun-like target stars, they have already probably been detected now. Just a matter of getting two more transits and RV followup.

In his talk earlier this week, Borucki understandably dodged a big question (Do you have any candidates that you know to be multi-planet systems?) with "we see all sorts of interesting signals." He also spoke of being able to do a very rough estimate of orbital period using transit times. Reading between the lines, I suspect they're actually feeling very good that they've hit paydirt already with interesting systems--including those with multiple planets, some of them at cool temperature, orbiting a star. But they are going to err on the side of caution. I would, too. The last thing anyone needs is to lose credibility with a headline grabbing announcement that turns out to be just a false positive. (Remember cold fusion?)

Still, if I were going to guess at the discovery announcement timeline, I'd predict (1) many more hot jupiters and neptunes, similar to the first five, later this spring; (2) hot super-earths and temperate gas giants by January 2011; (3) super-earths in habitable zones before the end of 2011; (4) earths in habitable zones before the end of the primary mission.

Not that I take my own speculation all that seriously. Just having fun with it.

Another Mission Manager Update:

http://www.kepler.arc.nasa.gov/news/mmu/in...s&NewsID=36

Not too exciting. Another roll. Another download. No new safemode event since February 5, but a software update is planned for April 21-23 to mitigate any future ones.

--Greg

Of course a number of Kepler science teams focus on variables (as I do myself) in the Kepler Asteroseismic Consortium. For non-public data, even us non-exo guys had to sign a non-disclosure agreement, to not to tell anyone if we somehow manage to find planets on our own, with light-time method or whatever except of the PI of the exo programme. And they even had some sort of Transit Filter that messed up certain properties of the data but they dumped that recently. But after the data is made public, you can do whatever you want with it. But I think they have the manpower to go through all the lightcurves and have their designed pipelines to find eclipse-signals...

Announcing and publishing new exoplanets take a long time apparently for such missions. I have heard a talk about COROT planets recently, it took them almost a year to collect enough RV data for COROT-7b (the super-Earth), and of course all other candidates had to wait, not that it wasn't worth it. They just have published Corot-9b but the speaker have already seen numbers up to COROT-15b (not officially though, of course)...

Wondering if anyone would care to brainstorm a bit...

Let's say that at the end of the Kepler extended mission, there have been 10 planets of 1-1.5 x earth mass discovered in potentially habitable zones around stars in the sample. What kind of follow up would you propose if:

(1) You were the PI of a grant for a few hundred million dollars?

(2) Bill Gates (or better yet, a space agency) asked you to administer a $5-10 billion fund for further researching these planets?

(1) Build and launch the biggest telescopic spectrometer possible to study those planets.
(2) Build and launch the biggest telescopic spectrometer possible to study those planets.

Seconding what centswroth says. If anything even remotely habitable (at least more than what has been found so far) is discovered, follow-ups to confirm it should be the highest priority.

Would an oxygen-rich atmosphere even be detectable at interstellar distances? I seem to recall a couple of years back some probes did measurements of Earth from a distance (I think Galileo or Cassini or some other craft that needed gravity assists?) and 'discovered' life on Earth this way.
Someone refresh my memory? Thanks!

Re: centsworth_II

Sounds good. But can we find some money in the budget for Kepler++? After the spectrometer has worked all targets identified by Kepler, I'd like for it to have some fresh targets.

Maybe if TPF or Darwin come back from the dead? It would be a shame not to have another dedicated planet-hunting mission after Kepler....

Well PLATO has a good chance now to fly around 2017. But we really need a huge telescope, like VLT-sized at least, with a very powerful spectrograph, dedicated entirely to exoplanets. And then you don't have to fight always with all the other proposals when applying for big telescopes. I guess the new extension of the propretiary period on Kepler candidates comes down directly to the lack of followups and RV measurements.

Not at these interstellar distances. Spitzer and HST have been used to measure the atmospheric compositions of a few planets, namely HD 209458 b and HD 189733 b, but they are very bright stars. The Kepler stars are several magnitudes dimmer.

It isn't so much distance that matters, but host star visible magnitude, because that contributes to the Signal-to-noise ratio needed to observe such small spectral changes when the planet transits.

400 "objects of interest" plus 328 more "candidates"? News story here.

ADMIN: fixed broken link

Broken link. (duplicate "http://")
Try this link.

Here's the original article, from nature.com. Telescope team may be allowed to sit on Exoplanet data. It has a lot more information in it than the softpedia article did. It makes it clear it's about Kepler wanting to change the policy to release the raw data even later than the current plan.

This was informative:


That sounds reasonable, however



There's a lot more in the article itself, of course, including the fact that Borucki thinks about 50% of the 328 current exoplanet candidates are probably false positives. Wish it said when we might expect another list of confirmed planets though.

--Greg

Borucki is giving a talk at Stanford in a few days: "Status of the Kepler Mission and Early Discoveries." (Tuesday, April 27, 2010. 4:15-5:45 PM). It's free and open to the public. Of course there won't be any new revelations in that kind of venue, but folks in the area might be interested in attending and plying him with questions, just for fun. The event website, such as it is, can be found here.

I'll be there for that one.

Also this one:

http://www.seti.org/csc/lectures
http://archive.seti.org/pdfs/csc-May-10.pdf

Great! Please pass on your observations and impressions from the talks. Thanks.

http://www.nasa.gov/mission_pages/kepler/n...year_anniv.html


Where see distribution the size and the period of these 328 candidates to Kepler planet on 7 months photometry?

on arXiv today An Independent Analysis of Kepler-4b through Kepler-8b

Also new on arXiv: the discovery of a nova-like cataclysmic variable star in the Kepler data (here). Stare unblinkingly at a field of stars and you'll see all kinds of interesting things...

I haven't finished the article yet, but from the abstract it seems like Kipping & Bakos have done a lot of the interesting and important follow-up of the data. Of particular interest is their constraints on resonant perturbers, exomoons and exotrojans... negative results are less sexy but provide interesting information as well.

As for Kepler's new data-withholding policy, I agree with Greg that it is understandable but may be wrongheaded, and it is interesting that ESA is moving in the opposite direction for PLATO, as ESA's specialist for both PLATO and COROT, Malcolm Fridlund, points out in the Nature article. "'You get a larger community and you get a bigger workforce for free,' he says. 'It's clear that the more people you get involved, the more support you get.'" Something for NASA to bear in mind when deciding whether to withhold KEPLER data, perhaps.

No big news.

- Kepler has 10 years of "expendables." I thought the number was 5 years.
- Funding for continuation will be competed for in 2012. Borucki is "confident" they'll receive funding for additional years of observations.
- "200 to 300" candidate Earth-like planets around Sol-like stars. Don't yet know the size of the parent stars, so "candidate" uses a liberal application.
- Less than one year of observations on the ground to date. (Or perhaps I misunderstood -- perhaps processed less than one year of observations.)
- Ground-based observations occur only during the northern hemisphere's summer months.
- Any new announcements won't come until January 2011 at the earliest during the AAS conference in Seattle, for the simple reason that there is nothing to announce until then.
- He seemed a little peeved at Nature for implying the team is sitting on data.
- The team will propose a follow-on mission for spectroscopy in a couple years.

I dunno... 200 to 300 candidate Earth-sized planets sounds like pretty big news to me. Even if 90% of them are duds, that's a lot of Earth-sized planets (especially after just one year of observations!). Plus, it's good to know they've got 10 years worth of consumables.

Say - does anyone have any ideas for a catchy word to use in place of "Earth-sized planets"? Earthoids? Terroids? No, wait - "oid" is Greek, so gaioids?

Thanks!

10 years is huge. We should be able to see any earth-sized or larger planet up to nearly 2 AU from each star. That will give us extremely valuable insights into these systems.

How I do wish that there was at least one large scope located far enough north that the star-field would be circumpolar. Too late now, but wouldn't there be some location in the northern USA with good enough climate and elevation and dark enough skies for a serious research scope? Wouldn't it have been great if construction of a dedicated follow-on instrument would have been part of the original Kepler proposal? Maybe it would be coming online right about now. Nah--with all the red tape, probably not.

In scientific literature, they're called "Earths."


HARPS-N was constructed partly for this purpose.

http://sci.esa.int/science-e/www/object/in...fobjectid=46020

Sadly, this probably won't work, as "geoid" is already formed from these roots - "ge" is an alternate form of "gaia".




Yeah, I'll say -- that's pretty awesome. I wonder how many are in near-Earthlike orbits? (Or, well, the equivalent from a smaller star.)

I'd guess that by "candidate" they mean they've seen at least one transit. They'll need at least two to estimate the period, and three to confirm.

Kind of sad to have to wait until January for more data, but maybe it'll be really exciting data by then. :-)

--Greg

Correct.

Moids. As in M class. We'd have:

Moid_radius
Moid_mass
Moid_{orbital period}
Moid_temperature
Moid_atmosphere

Within the context of this thread, Moid could mean just the radius.

I think with even a single transit, one could get a rough back-of-the envelope estimate of the period, right? If you assume a near-circular orbital geometry and have a decent idea of the diameter of the star, you should be able at least to estimate the period well enough to place the exoplanet reasonably well within its planetary system. You could tell whether it was probably a "hot" planet, a potential habitable zone planet, or a "cold" planet that you caught making a rare transit. And a single transit curve could also give you a decent idea of the planet's size, again if you make some reasonable assumptions. You could at least tell whether you were looking at a likely Jupiter, Neptune, or super-Earth.

Of course the three-transit gold standard, with radial velocity confirmation if possible, would still be required before any announcement. But if I'm thinking right, then of their 200-300 "candidates," they probably already have a short list of potential headline-grabbers.

Are my understanding and reasoning correct here? Thoughts?

Transit duration is not just a function of the planet's orbital velocity, but also the impact parameter of the orbit. From a single transit alone, it is not possible to distinguish a short period planet in a perfect, i=90° transit from a long period planet in a grazing transit. The only way around this would be to measure the steepness of the transit (a function of the impact parameter only), but Kepler does not sample the light curve with a high enough cadence to measure this in a single transit. The pretty light curves are arrived at by folding the data over itself so that the curve is much more visible.

So no, unless you have an amazing cadence, a single transit is insufficient for a back-of-the-envelope estimate. Two transits will be needed for an estimate of the orbital period, and a third for consistency/confirmation.

But yes you can pretty much know the planet's size from a single transit, assuming you had a radius for the star.

I've attached the light curve of Kepler-6 b. Note that each transit only has a handful of data points. The pretty light curve is arrived at by folding the data and stacking multiple transits on top of each other.

Ah, thanks. That clears things up a lot. So I guess if they have a "short list" by now, it's probably very incomplete, since two transits are required. They've seen everything with an orbital period of 4-5 months transit at least twice. Some planets with periods as long as 9-10 months could have been observed twice, if Kepler was lucky enough to have them transit during the first month of data gathering...

Borucki said "we don't yet know the size of the parent stars." They won't know until sometime during the summer.

Having said that, I bet the data processing room at Ames is a very exciting place to be once per month from here on out. I can picture them drooling over the data, saying "we'll have to check out that one!" only to renege the next month when something even more juicy comes down.

It might have been useful if Kepler had been designed to detect the beginning of a transit and increase the sample rate. By buffering data at a higher sample rate for say, half an hour, you could detect the transit with a high statistical certainty but not lose the beginning of the transit.

On the other hand, once the period is known it's pretty easy to train a large aperture telescope on the star and get photometric readings with a much better SNR and temporal resolution. I don't think Kepler gets a great SNR for dim targets, especially at high temporal resolutions, so there may not have been any significant benefit from doing this.

There may be a better way to compute it, but the way *I* figure it, here's what you need to compute the velocity of a planet from transit data:

R Radius of the star
r Radius of the planet
T Duration of the entire transit
t Duration of egress (that is, the amount of time between third and fourth contacts)

You could do it with first and second contacts too, but that requires the instrument to be able to predict the planet before it has seen it!

Then squared velocity will be 4Rr/t(T-t) or, given T>>t just 4Rr/Tt

(If you use diameters instead of radii, you can even drop the factor of 4)

Then if you know the mass of the star, and assume a circular orbit, you've got the period.

For the Earth and Sun, figuring diameters of 8,000 and 1,000,000 miles, respectively, with a fifteen-hour (equatorial) transit and a seven-minute egress gives me about 18.5 miles/sec, which is correct.

Note that this is very sensitive to the egress duration. Kepler's Long Cadence, of 30 minutes, can probably estimate T with less than 10% error, but even Short Cadence (1 minute) is going to be worse than that for t, which means even if Kepler could switch to SC once a transit was in progress, it couldn't give good enough estimates for t to make other than rough estimates.

Add to that the fact that, except for giant planets, Kepler can't actually tell with any confidence that a transit is actually occurring at all.

Of course, for giant planets we only see once, it would be nice to have at least a rough idea of the period.

--Greg

With as many false positives as there is going to be, if Kepler switches over to this high-cadence mode, it's going to be doing so a lot, far more than needed. Not sure why you would really need this. I'm not sure why one would make the telescope more complicated so as to determine the period in one transit when follow-up from the ground could be done.

This is my thinking as well. Of course the problem is that if Kepler detects only one transit during its (however long) mission, we'll have no clue when to look for the second, third transits. Not sure who is going to put forth the resources into such a long photometric campaign.


Also considering how many stars Kepler is looking at, I wonder if "quick! Switch to high-cadence mode!" would even be feasible. Maybe short period planet hosts could be neglected in favor of switching to a high-cadence mode for stars that have not yet been observed to have transits after some period of time.

Another Mission Manager update.

http://www.kepler.arc.nasa.gov/news/mmu/in...s&NewsID=39

Some interestinig tidbits:

Kepler’s list of interesting candidates has grown to well over 200 . . . the constellation Cygnus has returned to the night sky of the northern hemisphere, so the Science Team is actively involved in ground-based follow-up observations . . . After the completion of these observations over the next several months . . . announcement of the new discoveries will be made this winter.

--Greg

I'm resigned to no new Kepler discovery announcements until January 2011. However, there may still be some interesting discussion and perhaps independent analysis of prior results at a few major conferences this summer. Here are some to keep an eye on:

• COSPAR Meeting: "Extrasolar Planets: Recent Results from Space Missions and Future Prospects" July 18-25 (overview)
  
• 2010 Sagan Exoplanet Summer Workshop: "Stars as Homes for Habitable Planetary Systems" July 26-30, 2010 (agenda)
  
• Observatoire de Haute-Provence: "Detection and Dynamics of Transiting Exoplanets" August 23-27 (program)

I went to the SETI lecture today where Jon Jenkins from NASA Ames was presenting Kepler light curve analysis.

- Majority of the lecture was about how binary stars and noise (spacecraft heating) affect the light curves.
- Today is the one-year anniversary of the beginning of the science phase of the mission.
- "Over 300" candidates at this time.
- The newest algorithm to detect planet signatures is operational. It has found additional planets in previously mined data.
- The team will release light curve data for all but 300 stars in 44 days.
- The team will have to be "choosy" about which 300 stars to redact. Implying light curves will be released which may contain planet signatures (perhaps false positives).
- From just the calibration data from April of 2009, the team had 200 candidates by which they could choose to confirm Kepler's performance. I think from those 200 they chose the five that ultimately gave them Kepler 4, 5, 6, 7, and 8b.
- The team has isolated the instrument that caused some of the safe mode events. The expected safe mode frequency as a result of the instrument is 18 months. The instrument is sensitive to radiation.
- Headquarters is pushing the team to have a press conference with results in October, then another one in November.

Thanks for the report.


Sweet! I hope it happens. Waiting until January seemed so conservative. Surely by the end of this ground observation season, there will be something well into the publication pipeline--another batch of short period hot jupiters, if nothing else...