The data from Kepler’s first 136 days of operation could not be more interesting. As discussed in yesterday’s news conference, we now have fully 1235 exoplanet candidates from Kepler’s transit observations, and it’s worth quoting principal investigator William Borucki (NASA Ames) on the significance of the results thus far:
“We went from zero to 68 Earth-sized planet candidates and zero to 54 candidates in the habitable zone – a region where liquid water could exist on a planet’s surface. Some candidates could even have moons with liquid water. Five of the planetary candidates are both near Earth-size and orbit in the habitable zone of their parent stars.”
Statistical analysis by the Kepler team shows that between 80 and 90 percent of these candidates are likely to be real planets. Remember that the spacecraft is staring at 156,453 stars in a patch covering 1/400th of the sky, located in the constellations Cygnus and Lyra. What it’s giving us is a statistical sample of stars in a particular part of the Milky Way, one we can use to extrapolate planetary populations throughout the galaxy. In a helpful post, Franck Marchis (UC-Berkeley) lays out the properties of these planets classified by size:
- 68 Earth-size exoplanets with a radius (Rp) of less than 1.25 Earth radius (Re)
- 288 super-Earth size exoplanets with 1.25 x Re < Rp ? 2.0 x Re
- 662 Neptune-size exoplanets with 2.0 x Re < Rp ? 6.0 x Re
- 165 Jupiter-size exoplanets with 6.0 x Re < Rp ? 15 x Re
- 19 very-large-size with 15.0 x Re < Rp ? 22 x Re
The numbers are useful as well as deeply exciting, because they suggest how vast the exoplanet population must be on the galactic scale. Kepler is seeing only the small fraction of planets whose orbital alignment as seen from Earth causes them to transit their hosts stars. Roger Hunter, Kepler project manager, sums up what many had long suspected but lacked evidence for until now: “It’s looking like the galaxy may be littered with many planets.”
Image: Kepler’s planet candidates by size. Credit: NASA/Wendy Stenzel.
Where is Earth 2?
One of the questioners at yesterday’s news conference asked Douglas Hudgins, a Kepler program scientist, whether buried within the latest data release the project’s ‘holy grail’ might be lurking; i.e., a planet in the habitable zone of a star like the Sun. We do see planets in the habitable zone of certain stars from Kepler’s work, but Hudgins had to point out that the true ‘grail’ Kepler was after was a planet similar enough to Earth that it would be found in a roughly one-year orbit around a star like ours, and squarely in the habitable zone. We haven’t yet had enough time for such an observation to be made, since it would take more than one transit to reveal such a recurring event.
So right now we’re talking about planets in the habitable zone of stars that are cooler and smaller than the Sun, places where the star in the sky would be a lot redder than the Sun as seen from Earth. The next data releases will get more and more intriguing on that score as we home in on true Earth-analogs. At that point Kepler’s statistics really will be giving us a glimpse of the likelihood of the kind of planets we could live on and how common they are in the galaxy. As the overall catalog grows, it will become more accurate and should reveal not just exoplanets orbiting at larger distances from their host stars but very possibly moons around some of them.
We’re also getting a better read on just how skewed our early exoplanet results were toward large planets, particularly the so-called ‘hot Jupiters.’ That was an inevitable result of picking off planets that were the easiest to detect, but Kepler is now showing us that stars in our galaxy are more likely to be orbited by smaller worlds. Let me quote Marchis on this:
…we can now say that stars in our Milky Way galaxy are more likely to host small exoplanets since 75% of the Kepler catalog exoplanets are smaller than Neptune, with a peak of exoplanets only 2-3 times larger than Earth.
Using model predictions which take into account the probability of having the correct geometry to detect these exoplanets, the Kepler team extrapolated that 6% of the stars in our Milky way have Earth- and super-Earth size exoplanets, 17% of them have Neptune-size candidates and only 4% of them have Jupiter-size exoplanets.
Image: Kepler’s planet candidates as of Feb. 1, 2011. Credit: NASA/Wendy Stenzel.
Why so comparatively few multi-planet systems (about 170 thus far)? Considering the short period of time Kepler has been in operation, we’re working with fairly short orbital periods. That’s one thing that makes Kepler-11, discussed here yesterday, so unusual. Given that we’re talking about orbits close to the parent star, it’s astounding to find a star with five planets crammed inside the orbital distance of Mercury, and a sixth at a distance between Mercury and Venus around our Sun. The five inner planets have orbital periods varying between a scant 10 to 47 days around this G-class star, and from all the indications we have, the system is dynamically stable.
Bringing the Hunt Back to Earth
Where would we have gone next in our exoplanet hunt if money were no object? The Space Interferometry Mission immediately comes to mind, because while Kepler could give us a statistical look at a particular patch of sky, SIM would have targeted nearby stars, giving us detections via interferometry of Earth-like worlds within thirty light years. Without SIM or, for that matter, Terrestrial Planet Finder, we’re finding other ways of locating such worlds from Earth’s surface. Ground-based surveys like MEarth and new technologies like laser frequency combs may help us fill out our target lists for the upcoming James Webb Space Telescope.
Lee Billings looks at these technologies in a fine new article in Nature. Take Steven Vogt and colleagues, who have built the Automated Planet Finder (APF), a robotic telescope paired with a powerful spectrometer at Lick Observatory in California. Like MEarth, APF targets short period planets around nearby stars, and that includes the brightest M-dwarfs in the sky. One way to save money while still hunting exoplanets is to maximize observing time. Billings quotes Vogt:
“The coin of the realm is observing nights,” he says. “It’s not new technology; it’s not laser combs or some newfangled near-infrared spectrometers that can take advantage of M-dwarfs. Take $50 million, which is chump change in the NASA regime, build a 6–8-metre telescope with enough light-gathering power to reach a large fraction of the nearest M-dwarfs, put a nice spectrometer on it and dedicate it to this work every single night of the year. You’d have these planets pouring out of the sky.”
And maybe there’s a cheaper way in space, too, as revealed in Sara Seager’s ExoplanetSat program. Seager (MIT) has the notion of entire fleets of tiny CubeSats, dozens at a time scanning individual stars, each satellite with its own particular target. Planets around nearby Sun-like stars should be detectable if they transit, and Billings says we should see a functional prototype as early as 2012, with subsequent satellites launching for as little as $250K apiece. So there are ways to proceed beyond Kepler, and they’ll surely be in full gear even as the Kepler data continue to arrive and the planets we discover begin more and more to resemble our own.
For more, see Ford et al., “Transit Timing Observations from Kepler: I. Statistical Analysis of the First Four Months” (preprint) and Borucki et al., “Characteristics of planetary candidates observed by Kepler, II: Analysis of the first four months of data” (preprint).
Quoting Steven Vogt: “Take $50 million, which is chump change in the NASA regime, build a 6–8-metre telescope with enough light-gathering power to reach a large fraction of the nearest M-dwarfs, put a nice spectrometer on it and dedicate it to this work every single night of the year. You’d have these planets pouring out of the sky.”
So quantity rules all? Or planets per dollar? :-/ Machine-harvesting the lowest hanging fruit will hardly advance science…
Take $1500 million, put the “S” back in “NASA” and build space-based missions like the Terrestrial Planet Finder and Darwin, observe G-dwarfs, and then we’ll find “Earth 2” (or if not, then we’ll ~know~ that this is no nearby “Earth 2.”)
“…the Kepler team extrapolated that 6% of the stars in our Milky way have Earth- and super-Earth size exoplanets, 17% of them have Neptune-size candidates and only 4% of them have Jupiter-size exoplanets.”
That means that only 17-27% of stars have planets at all. My question is why are >= 3/4 of all stars in the neighborhood apparently without planets based on these statistics?
Its means water planets (small Neptunes) are the most common in the galaxy. Just not Earth-like water planets. Then again, the Earth doesn’t really have that much water compared to a planet like Neptune.
I complain a bit now and again how Kepler doesn’t help us find exoplanet targets in our immediate vicinity.
Let me emphasize that the beauty of the Kepler data is that it demonstrates absolutely unequivocally that planets are literally common as dirt, and for those of us who grew up back when “planets around other stars” were generally assumed, but still purely theoretical, this is huge.
It’s also huge because it helps us at last refine with real statistical certainty the f-sub-p term of the Drake Equation.
Hi folks,
This is interesting because the number of planets around Neptune-size seems to be larger than smaller planets around the size of Earth. Do models predict this? I thought the trend was in favor of smaller and smaller planets being more and more common, but this latest set of data seems to suggest otherwise. Or, is there a bias in favor of finding these slightly larger than Earth planets? Afterall, in our own solar system, small objects exist in far larger numbers than the larger objects.
you should post the slide that shows witch of these planets are in the Goldilocks zone ( page 14 ?)
Mostly Neptune sized worlds and one Jupiter + sized world, sadly I remember a paper by a Ms Penelope( ?) from the south west institute that claims Jupiter mass planets can not produce exomoons the size of earth :(
But hopefully Kepler ( or you folks?)can find these worlds
If low-mass planets do frequently turn out to be hydrogen-rich, the usual calculations of habitability which assume rocky planets go out of the window. I recall some calculations on the possibility of liquid water oceans in ice giant planets that were published in 2006: it would seem that these results are going to take on new relevance for the Kepler habitable zone candidates.
Alex Tolley, have a look at this analysis:
http://nextbigfuture.com/2011/02/kepler-telescope-related-what-is.html#more
http://kepler.nasa.gov/files/mws/FebDataRelease_revised_020211.pdf
a 106 page paper!
page 22 table 6
look at planet 326.01 it has less then 1 earth radii.
search self luminous in this paper, several false positives may be in fact planets.
there is one binary brown dwarf
binary stars with planets might mean a lot of the planets are really larger then the data suggests,so does this mean a lot of the Neptune’s and below in the Goldie locks zone might be Jupiter’s?
off course there should not be any Jupiter’s in a binary right?
this paper states that the data base is so huge that the Kepler team will concentrate on small planets, so you exomoon hunters will have to do the guest observer hunt
so now we have a list of super earths and Neptune’s in many a Goldilocks zones, is it possible that a glancing planetary collision might produce an earth mass exomoon? but, would it not be a water world?
spaceman.
in the paper above there are clues that stars of different luminosity’s or sizes affect the ratio of planet sizes but the Kepler team cautions that the data has yet 2 more years of gathering to tell.
figure 12
People may not have appreciated the answer to the question “where is Earth 2?” since it wasn’t very highly emphasized in the post. The answer is: *if* “Earth 2” exists around one of the observed stars then Kepler may yet discover it but there hasn’t been enough time yet. Firm evidence of discovery requires 3 transits, which will take a minimum of 2 years and as much as 3 years, and Kepler simply has not been in operation long enough.
So hold your horses, there will invariably be more discoveries coming as Kepler continues operations, it’s still too early to detect Earth 2’s quite yet.
A couple of comments,
1. The extrapolation of systems with planets is on the low side. I am not sure how they extrapolated to ~17% (to ~25%), but this is almost surely the lower limit/bound.
2. Andrew W, I disagree with that link. The blogger’s numbers with regard to how he reaches,
“The most significant “take-away” from this is that it is all but certain that virtually every star has planets., and not just “one or two”, but many – on the order of our own system.”
just does not make sense. I agree that the number of systems with planets will be larger, but I don’t fully agree with his “arse pull”.
There is a good point to be made about detection frequency based on how far out the planet is from the star, but I really don’t get how his numbers were generated. I would trust the extrapolations from the paper much much more than the extrapolations on a blog posting.
Granted, once the Kepler data are expanded, the number may approach ~50+ %, but the blogger’s different “basis of detection” does not mean much to me.
3. More exciting stuff to come in 2012. :)
My impression of the Kepler results is that, yes, there are lots of planets. But they don’t seem to be very Earth-like. Every time you look for an Earth, you get a Neptune.
Can someone tell me the significance of the three different colors in the diagram of Kepler candidates above?
Spaceman.
I do not think the Kepler results are any more indicative of the size distribution of planets than the doppler data is. First, off Kepler detects a much higher percentage of very close planets than ones farther out. Secondly, the minimum size that can be detected also falls of with distance. So, there are strong sample selection effects. The results do however point to there being a paucity of small planets with less than 4 day orbits, but this I attribute to the fact that small planets evaporate away over stella lifetimes.
I think that once we find the mass of these close in Earth-sized bodies , we will find they are very dense, and what they are, are the cores of much larger planets that are in the process of disappearing. It’s a bit like with black holes; the smaller they get, the faster they evaporate. So, you will get very few small bodies this close in.
I would bet that the size of planets follows a power law curve, with smaller being more frequent. They only thing that we are not certain of is the size of the exponent.
Zen Blade: “I would trust the extrapolations from the paper much much more than the extrapolations on a blog posting.”
Marchis: “…we can now say that stars in our Milky Way galaxy are more likely to host small exoplanets since 75% of the Kepler catalog exoplanets are smaller than Neptune, with a peak of exoplanets only 2-3 times larger than Earth.
Using model predictions which take into account the probability of having the correct geometry to detect these exoplanets, the Kepler team extrapolated that 6% of the stars in our Milky way have Earth- and super-Earth size exoplanets, 17% of them have Neptune-size candidates and only 4% of them have Jupiter-size exoplanets.”
I’m not sure Marchis has interpreted the analysis correctly, I think the numbers he sites only refer to candidate planets within 0.5 AU of their host star, ie only those that could possibly have been found in the first 4 months of Kepler observations.
The actual paragraph he’s referring to is at the bottom of page 32 of the paper.
I think I can answer my own question of February 3, 2011 at 22:48.
The three colors indicate the stellar classification of the host star, with relatively few exceptions only F,G,K, stars are included in the Kepler study (the actual colors used probably make this obvious to an astronomer).
The ATA has done a first pass search on 50 of the 54 KOIs in habitable zones:
http://shrewdraven.org/content/seti-searching-habitable-zone-kois
Do es this mean that neptune size planets/water worlds around red drawfs (in the habitable zone) are likely to “turn” into earth type planets over time?
The notion that due to the flares that smaller stars are likely to send out in their youth, much of the atmosphere of the neptunes will get whittled away. So as the smaller stars start to settle down, the neptunes will have a thinner atmosphere and less water, compared to what they started with and so conditions for life would be better?
I for one don’t find the new Kepler data release very promising for the total numbers of earth-sized planets in habitable zones. They should have ALREADY found most of the “good” planets around K & M dwarfs but the number released is only 5, a very small fraction, less than 0.2 %, of even all
the M dwarfs in their field of view (yes, that may increase by a factor of order 100x given the detection probability). Since G stars are much less common than K & M dwarfs, the total numbers are not likely to increase by more than, at most, a factor of 2 or 3 by the end of the survey. Extrapolating this to the galaxy as a whole gives, optimistically, perhaps a billion earth sized planets in habitable zones (this number has the detection probability due to transits factored in). Earth sized moons around Jupiter sized planets in habitable zones (IF such moons exist) are unlikely to increase this number very much.
Notably, NO moons were mentioned among the objects found in habitable zones, even though the orbital periods are short enough that, if they were common, they would have already shown up in the data (earth or mars sized moons, that is) . Using this number of “good” planets, it’s exceedingly easy to get very low numbers of technological civilizatons in the galaxy by plugging “reasonable” guesses into the Drake Equation. In fact, it’s very hard to get a large number for N even if you’re very optimistic about the unknown factors in the Drake Equation. Feb. 2, 2011 may well mark the first big step towards resolving the Fermi Paradox, but not in the direction many of us had hoped for.
I think the Fermi paradox is based on the assumption that ETI, once they start spreading from star to star, will not stop. One such occurrence can fill the entire galaxy with ETI in a relatively short time. The density of planets is therefore not really that relevant for the Fermi paradoxon, unless it would make the difference between N>=1 and N=0.
Eniac
“I think the Fermi paradox is based on the assumption that ETI, once they start spreading from star to star, will not stop.”
In part:
Some ETI may not have technological civilizations, some ETI may not venture to other stars for various reasons, etc.
But, if there many ETI, it becomes far likelier that at least one/some of them are interested in star travel.
Also, if there are many ETI, star travel on their part may not even be necessary in order for us to ‘hear’ them or ‘see’ their engineering accomplishments.
So, where are they?
Determinig the number of ETI civilizations in our galaxy is low/very low would explain why there is no sign of their presence among the stars – no signals, astroengineering, etc. It would explain why our galaxy appears to be untouched by intelligence, a wilderness formed only by natural forces. It would explain the Fermi paradox.
@Eniac I’m not sure I quite agree with your defintion of the Fermi Paradox . My interpretation is that Fermi was asking if ETIs are common, then where are they all? I don’t think he was assuming that N=1 would populate the galaxy given long enough lifetimes (in fact, most physicists don’t seem to believe that at all, but the opposite) but rather that if N is large, then we should have heard or seen some ETI by now. The point I was trying to make, though I may have made it badly, is that the Kepler data seems to indicate that N may actually be very small (unless L is very large, on the order of millions or tens of millions of years), so we should EXPECT to have not detected (by radio, laser, handshake, astro-engineering, whatever) any.
Even a value of ten million years for L can, plausibly, only give N on the order of 100, which is damn small, in my opinion.
I may well be wrong on my interpretation of Fermi’s Paradox above. Perhaps it needs to be reformulated in terms of SETI (communication based) rather than just physical contract = in the same solar system etc…
If Kepler had been pointed across interstellar space at our solar system, after 4 months of observations, the sun would not have a single planet listed as a candidate. Mercury is too small for Kepler to detect and Venus wouldn’t make the three transits required for confirmation until there had been at least 470 days of observations, and even then Venus would only have a 1 in 1250 chance of passing between the sun and the distant Kepler.
So if all 156,000 star systems under observation were models of our solar system, Kepler, thus far, would have failed to have confirmation of a single planet!
Given that, the detection of even one planetary system in 150 stars thats been achieved is remarkable.
It leads me to wonder if the main answer to Fermi’s paradox is that in most star systems there’s just too much happening, too much interplanetary junk in too many exotic orbits, there might be a billion “habitable” planets out there but they’re almost all getting blasted too often for complex life to develop.
If that’s the case, there are rich pickings out there for the colonists of future space arks.
Hello. I read very often this site and I think articles and comments are very interesting. This time around I think I want to say something very shocking: looking for life in the universe “as we know it”, looking for Earth 2 or Earth 3 or Earth 1000, is on the same level as stating the Earth is flat and evreything else rotate around it, including the Sun of course.
I’m totally astonished about this obsession of finding Earth X around the universe. If there is one thing that’s costant throughout the entire universe, from quarks to galaxies, is the VARIETY. There is nothing identical to another thing in this universe and this principle is valid for a biosphere like Earth too, of course.
Some months ago the first alien life form was discovered: an alien from our very same planet, not a distant one. I’m talking about the arsenic bacteria.
Now in the universe is the “life as we know it” the only one possible: no it is not… we’ve got the very evidence on our planet.
Planets like Venus can be one of the most life-friendly place in the Solar System, at some altitude (50 kms, if I rember well), Mars too, but even Moons around gas giants like Jupiter or Saturn can be colonized in principle (the fact we haven’t the technology now doesn’t mean another civilizations doesn’t have it too). Each one of these potentially colonizable places aren’t even remotely similiar to each other!
Not to mention the possibility of immense space stations in which milions of individuals can live into.
But there is something even more shocking in all of this “as we know it” obsession: the water is essential for life. Yeah sure… what about life “as we DO NOT know it”?
And what about life “as we know it” but it’s different from us? Let’s take the machines for example: they’re getting better and better in a very short time, considering the time scales of the universe, and it’s reasonable to assume they will become sentient sooner than later.
It’s reasonable to assume they may have become sentient already in some other place of the universe, so they should be pratically in evrey place where there is energy. So what is all this obsession about rocky planets in the so called habitable (for us..) zone?
I’m still not clear on the probabilities of planets transiting their star such that Kepler can detect them. When the Kepler mission first started I saw articles that said the chances would be between 5% and 1%.
Andrew W’s probabilities are far lower than that. If his numbers are correct the vast majority of stars must have solar systems given how many planets Kepler has already found around just 156,000 stars.
I’ve been assuming the math by the guy at thenextbigfuture was correct, after having a think about the geometry (and I no mathematician) I think he might be in error. My very poor maths suggests about a 1% chance of a transit by a Venus orbiting a sun being detectable. Earth 0.6% Jupiter 0.114%.
great blog and comments
they gave us nice planet info but how many are around binary star systems? do we have a tatooine yet? ;)
Its strange – or is this just me. I can see obvious reason for complaining that Eniac UNDERSTATES the magnitude of the Fermi paradox, but everyone in this forum is wining that he overstates it. Do all the rest of you work on SETI?
In reality the Drake only examines the Emergence of fresh civilisations on their home planets. It assumes that these never spread. You multiply several factors together to find N = the number of CURRENT, independently emerged, civilisations, that posses such grand outwardly-looking ambitions that the are prepared to send out unsolicited messages to all corners of the galaxy in the (forlorn) hope of reply.
Perhaps many of you can now see the problem. How many years could they carry on this expensive procedure before the cumulative cost is more than sending out a starship. Now since the starship would give its expected return in centuries, but the most optimistic (pre-Fermi paradox) estimates for unsolicited signalling as given by the Drake equation are a few millennia, we have a starting point for our problem. Obviously most such communicators will spread if they survive long enough. The first such civilising race that has a spreading rate that is faster than its self-destruction rate will inevitably and rapidly colonise the galaxy.
But this little assessment is just the grand underestimate used by Eniac that equates N and galactic colonisation chances. Actually N is just PROBABLE number of EXTANT civilisations with the above properties, that persist on their home planet. Now comes the nasty bit, L = the expected longevity of such a civilisation. A typical value for L used here is 10,000 years – about 25 times longer than human technological civilisation has lasted, so if we are to expect that there is one such entity with which we can communicate today, then their must have been about a million such civilisations spread out over the 10 billion year history (forgetting the first few billion years were the metalicity was too low) of our galaxy. We only need THIS figure to be 1 for Eniac to have his problem, not N. He should have stated “unless it would make the difference between N>=0.000001 and N=0″ not “unless it would make the difference between N>=1 and N=0″
Shame on you Eniac for such wanton understatement.
Andrew W, Drakend – great points from you both. What I’d like to add is that we have to remember all the time how limited we are by the current state of the world’s engagement in space exploration matters. I mean, the government-sponsored, cold war-infused exploration all but died out already and at the same time governments all around the world got stronger at the expense of robust, innovative, daring private sector. So sadly there’s no Heinlein-like future on the horison. Our civilization inevitably turns inwards just before our eyes and as of today an honest look at the state of affairs has to affirm that the only exploration we’re going to enjoy in the near future is the one provided by the Eve Online computer game.
So, taking it back to Drake, for all we know now, the always assumed interstellar transportation may well be beyond any intelligent life form’s reach, because we so far even in the slightest haven’t proved it to be otherwise. Trying to go further ‘mentally’ before going further ‘physically’ is just it – a speculation and fantasy. And frankly with years I grew more and more tired of it. We have to step beyond Earth, just like Europeans had to step beyond Europe to shove civilization to the next stage of evolution. Otherwise we are just running in circles, sadly.
Following up on the above discussion with regard to planet abundance, and in particular Alex Tolley and Andrew W, I am also surprised by the statement by Kepler team member Marchis that:
” the Kepler team extrapolated that 6% of the stars in our Milky way have Earth- and super-Earth size exoplanets, 17% of them have Neptune-size candidates and only 4% of them have Jupiter-size exoplanets”.
With due respect for the Kepler team I wonder whether this is not a slight error in interpetation or at least quotation.
Some facts and assumptions:
– Nearly 1000 stars so far have been shown to possess at least one planet, out of a total sample of 156,000 stars (mainly F,G,K).
– 86% of all these (1000) stars have only one planet detected so far.
– The most common detected planet size classes are super-earth and sub-giant (Neptune class).
– Of these detected planets 68 are near-earth size, 54 are in the HZ, 5 are overlapping: both near-earth size and in the HZ.
– The detection chance of a planet near-earth size and larger in the ‘Kepler realm’ so far is almost 1% (also the cited nextbigfuture estimates are in this realm for the ‘Kepler distance’ so far).
– The observation period of 136 days corresponds to the innermost system up to about 0.5 AU around each observed star, equivalent to just beyond Mercury’s orbit in our own solar system.
The above would imply that:
– At least 2/3 of all stars have at least one planet in its innermost system (up to 0.5 AU, about Mercury orbit).
– About 5% of all stars have a near-earth size planet in their innermost system (within 0.5 AU), about 4% have a planet in their HZ, about 0.4% have a near-earth size planet in their HZ.
– The fact that 86% of stars have only one planet within 0.5 AU corresponds very well with our own solar system.
– The main striking difference between the majority of detected planetary systems and our own solar system is that the innermost planet is not a very small terrestrial planet but a much larger (super-earth, Neptune class) planet.
– As noted by others, the smallest category of planets (e.g. Mercury) cannot reliably be detected by Kepler (?), so that the true tally of planets is probably even somewhat higher.
– With regard to HZ planets, only HZ within 0.5 AU has so far been considered, implying late K (>= K3) stars (plus some M dwarfs). The HZ of most of the true solar type stars within the sample has not even been reliably searched yet.
From the magnitude and temperature data in the above-cited Kepler document (partic. the graphs on page 10 and the table on pages 22/23), I roughly deducted that about half of all stars in the total sample have their HZ within the range now surveyed ( up to 0.5 AU), implying that when the Kepler survey is complete and has covered the HZ of all stars in the sample, my guesstimate is that about 8 (6 – 10) % of all stars will appear to have a planet in the HZ, and about 1 % will appear to have a near-earth size planet in the HZ.
Still there is planets that kepler can’t detect, and probably can be made of rock , like mercury or moon size planets,cold earth-size made of rock than orbit far way from the star (a microlesing space mission would great to detect such planets and ever free floating planets). Other thing the small earth-size planets it’s hard to be detect,the kepler team still refine the data from the noise.
Still could have a lot of earth-size planets on this new data undetected yet, rock planets than someday the man can walk on.
I’ve done some more maths, and by my reckoning if Kepler was studying our solar system these are the odds of it detecting a single transit of each of the innermost 5 planets, first column is for 43 days of operation, second column for the time necessary for an entire orbit of the target planet (26 years for Saturn):
Mercury 1:274* 1:134
Venus 1:1235 1:236
Earth 1:2852 1:336
Jupiter 1:176,260 1:1750
Saturn 1:800,000 1:3200
*but Mercury is probably too small for Kepler to detect anyway so that’s more likely ZERO chances of detection.
Goat guy at the next big future, whose odds I was using earlier, has made the assumption that the observations of each star aren’t continuous, that Kepler only studies a selection at any one time, my interpretation is that Kepler is focused on each and every star under observation continuously.
Hi All,
I’m an amateur astronomer with a passion for telling and writing stories based on facts and sensible conjecture. Can anyone help me with a few planks of hard science on this:
Just suppose there were a planet with approximate Rp of 0.7 to 1.3 orbiting Centauri ‘B’ in the goldilocks zone….
What might the time period of it’s orbit be?
Would any part of the planet ever be in ‘night’ since Centauri ‘A’ is or would there be a number of ‘days’ with light and dusk, alternating with days of light and darkness depending on the positions of the two stars.
Lastly, would the presence of Centauri ‘A’ have a gravitational impact on a planet orbiting ‘B’? If the planet had liquid water would this cause immense tidal movements? Would there be more volcanic activity?
Any further contributions or musings on a hypothetical planet around Centauri ‘B’ would be most welcome!
Thanks
Andy
Andy C, one place you might want to start is P. Wiegert & M. Holman, (1997) The Stability of Planets in the Alpha Centauri System, Astron. J, 113, 1445-1450, which lays out the possibilities. You can have a look here:
http://www.astro.uwo.ca/~wiegert/papers/1997AJ.113.1445.pdf
It’s certainly been a good week for the planet hunt. We take a really close look at just one tiny patch of sky and we find this jewel box after only 6 months. Plainly the galaxy is littered with nameless worlds of every kind. And many kinds we haven’t imagined yet. In every direction, a wilderness of wonders .
Andrew W February 4, 2011 at 15:28
“If Kepler had been pointed across interstellar space at our solar system, after 4 months of observations, the sun would not have a single planet listed as a candidate. Mercury is too small for Kepler to detect and Venus wouldn’t make the three transits required for confirmation until there had been at least 470 days of observations, and even then Venus would only have a 1 in 1250 chance of passing between the sun and the distant Kepler.”
I am just now getting the chance to see the press conference,Andrew’s observations are correct,
http://www.youtube.com/watch?v=zZHSptpDoLQ
only one earth size “candidate” world has been found in a Goldilocks orbit of a 150 days, and this is around a K star.
the full three years of data are needed to come to a conclusion on the question of the ages.
the, “we will not find earth size moons around exo Jupiters paper”
http://www.nature.com/nature/journal/v441/n7095/abs/nature04860.html
I wonder if there is a Alpha Centauri analog in Kepler’s field of view?
I can not understand why it is so often repeated here than Jupiter sized planets can not have Mars sized moons. I realise that there have been some recent papers purporting to have demonstrated that the formation of giant planets does not allow regular moons of this size to develop, but to me these just seem to be remodeling much older work that came to the same conclusions anyway.
Surely other recent work showing that large captured moons, such as Triton, might not be as rare as we previously thought, is much more relevant here.
The number of rock planets still could be much common that we can imagine,the kepler data is just the tip of the iceberg.
if our solar system was in kepler field in 136 days data,Mercury wouldn’t be detectable by kepler because it’s too small, the still there is not enough data to confirm Venus,because with this Venus would transit the sun only once.
then if all 156,000 stars in the kepler field was like our solar system,the telescope wouldn’t detect not ever one planet!
So I think that there much more solid planet out there,than neptune or gas planet like what they found in the kepler 11 system,still kepler can have rock planets undetectable yet,mercury-size planets and below,small planets in long orbit periods far from the star,undetectable
I think than the galaxy its full of rock planets,its just hard to be detect,and kepler just show to us, the easy detectable new class of gas planet,this not meaning the rock planets are uncommon,this show than there is a big variety of planets,and there is enough room for rock, habitable planets, in the galaxy
Another thing that I must point out: Remember than kepler search for planets around stars most F ,G, K class and few M Class (~3000) and 75% of the star in the galaxy are M class,then the kepler haven’t a good statistic about rock planet around the habitable zone of this stars,and recent studies show than this stars it’s more likely form small rock planets than gas giants
steven rappolee point out:
“the, “we will not find earth size moons around exo Jupiters paper”
http://www.nature.com/nature/journal/v441/n7095/abs/nature04860.html”
like hot jupiter was impossible in theory before 1995 and then the scientist found 51 Pegasus and the first hot jupiter transit HD209458 in 2000, eart -size exomoons are impossible by a theoretical model of a science paper, but this not mean than this earth-size exomoons doesn’t exist.
the Kepler scientists still work on the data, exomoons are hard to detect,by transit or TTV,this earth-size exomoons can be right now on the last kepler science download.
it’s only 136 days,still years of data to come yet, the nature always show us surprises than we can’t ever imagine before.
Hi Everyone
After having read much of the paper, I think that the following quoted figures for planet frequency, “Kepler team extrapolated that 6% of the stars in our Milky way have Earth- and super-Earth size exoplanets, 17% of them have Neptune-size candidates and only 4% of them have Jupiter-size exoplanets”, are most likely lower limits and will not be the final word as to what will be the “true” planet frequency as determined by Kepler observations. Now, although some of the candidates will eventually be rejected as false positives, the Kepler team notes:
“… the inherent frequency would be higher than shown in Table 7 and associated figures. Furthermore, throughout the mission we will continue
to make improvements to the data analysis pipeline. Thus, the capability of the system to recognize small candidates will continue to improve, and more candidates will be discovered. The latter is expected to occur in mid-year when the capability to stitch together quarters of observations becomes operational.” (Borucki et al 2011)
Also, earlier in the paper, and with respect to small candidates the team notes:
” All panels in Figure 3 show a scarcity of candidates with radius Rp smaller than 1 R!. The paucity of small candidates at even the shortest orbital periods may be due in part to a decrease in the actual distribution, but part if not all of it is due to incompleteness for the smaller signals, coupled
with analysis of only a portion of the eventually expected Kepler data, and higher than expected noise levels (Borucki et al 2011).”
In other words, the dearth of candidates below 1 Earth-radius in the three figures on page 12 will probably start to be filled in with additional candidates as the pipeline gets better and more transits are observed; thus, the actual percentage of stars with Earth and sub-Earth-sized planets which stands at ~6.0% is almost sure to rise in future data releases thereby bringing the overall planet frequency to a larger percentage. My guess as the to the eventual over all planet frequency after the mission is finished: fp>50%. Does this assessment sound about right?
Another tidbit that I found interesting is that the overall planet frequency as it is measured now by Kepler does not seem to decrease with decreasing stellar luminosity, as is nicely shown in the lower left of Figure 12 on page 28.
Lastly, what a wonderful accomplishment all of this is, as I can remember growing up in the 1980s– a time when the existence of planets around other stars was still mere speculation!
Thanks to Paul for the reply on Centauri ‘B’ planets. Andy
Regarding exomoons, the article Daniel pointed out is also downloadable outside Nature’s paywall:
http://www.fas.harvard.edu/~planets/journalclub/nature04860.pdf
I note that contributors to these pages seem very good at extrapolating to find preliminary conclusions from incomplete data sets. Perhaps then one of you will know how significance Kepler’s non-detection of exomoons is. To me such calculations must be complicated by the following factor.
The closer a planet orbits to its star the more quickly its rotation slows towards tidal lock. Moons that orbit slower than their planet rotates, spiral in towards their planet at an accelerating rate, until they reach the Roche limit.
How significant is this effect, and thus how significant is Kepler’s failure to detect exomoons so far? Can we say anything about the prospects for Earth sized moons at > 0.5 AU as yet?
It would seem that a planet orbiting one of the stars in a binary system would have a “winter”, where both suns are in the same direction and a bright two-sun-day is followed by dark night, and a “summer”, where it is never dark and the two suns take turns in the sky. In between, you would have a short night, followed by a long one-sun-two-sun-other-sun day. The “other” sun would be weaker, probably considerably. Come to think of it, it would actually be very similar to the sun and the moon on Earth, except that the month would be as long as the year.
In the middle of “winter” there would be a chance for a sun-sun eclipse, which could be quite interesting if the other sun is bright and not too distant.
Does this finally put to rest the notion that planets are rare? What other factors in Drake’s equation need reworking?
Oops, above I meant “moons that orbit faster than their planet rotates” not slower.
I note that there is still much comment on reworking of decades old news that giant planets limit their own moon formation to a mass fraction of 1:10,000. I’m still mystified as to why no one seems interested in other mechanisms of aquiring moons such as
http://www.nature.com/nature/journal/v441/n7090/full/nature04792.html
So, perhaps Earth-sized moons are not likely to form around gas giants–even around the so-called “super-Jupiters”, but what about the possibility of gas giants capturing a large Earth-sized protoplanet that formed elsewhere in the planetary system? What is the minimum mass for a habitable planet (I have heard estimates ranging from 0.3 Me to 0.8 Me)?