The scientific community fights for years over one Hubble telescope - and some shady agency has two?
They can afford to "give them away" now. Probably because they have something much better now?
Am I the only one who thinks there is something simply "wrong" with all this? (And yes, I find it good those things are *now*, better: *finally*, used for science)
No, some shady agency does not have two. They have two surplus and obsolete (for their purposes) telescopes that were never launched. The NRO had many, many more than two such satellites in actual operation, and now we are being told those are no longer cutting edge, so they definitely have something much better.
There is a saying in astronomy that you cannot compare ground-observatory project costs to space-observatory project costs (every grad student ever has pointed out "for the cost of HST, imagine the huge telescope we could have built on the ground!" only to be rebuked with "space dollars are not the same as ground dollars). Similarly, military dollars are not the same as space dollars are not the same as ground dollars. Otherwise, one could naively say "for just the cost of one F22, we could have paid for XYZ science program by now."
Actually, HST has a 96 minute orbital period around the Earth, so it cannot stare continuously. But it can coadd several exposures over several orbits.
The real advantage of HST is that it is diffraction limited even on very faint objects (adaptive optics requires bright guide stars; even laser-guide-star adaptive optics needs a relatively bright natural star for the first-order correction) and the background light is much lower.
As I understand it... the bulk of the large moons in our solarsystem are made of low density materials essentially they're mini-frozen Jupiters in themselves or water ice balls like Enceladus. Wouldn't the formation of an Earthlike planet be precluded so close to a Jovian mass... Something with a substantial rocky core like Earth's forming at the same distance as a Jovian would have become a Jovian itself and most likely would have merged with the parent body? The Inner planets as I understand it were Jovians that had the bulk of thier gas envelopes blown clear during our Sun's T'Tauri phase of super strong solar winds.
The major moons in our solar system have densities between those of the terrestrial worlds and those of the giants. I certainly would not consider them to be mini-Jupiters.
Planet formation isn't as simple as our old ideas of "if it forms here, it will be a gas giant, if not, it will be terrestrial". Clearly there are a lot more details that we don't know yet.
Would it be easier to detect the existence of large (small planet-sized) moons around a gas giant than earth-sized planets around a star? Would not the perturbation of the gas giant be easier to detect because the mass ratios are closer (large moon to gas giant vs earth-sized planet to star)?
If so, detecting the gas giant in the habitable zone and then looking for evidence of large moons or companion bodies could allow detection of candidates for life.
I assume this would entail detailed, direct observation of the gas giant, but I would imagine that will happen sooner than detailed, direct observation of an earth-sized body.
Not with this method, since we are seeing the reflex motion of the star, which wouldn't be very different at all if the thing orbiting it were a planet by itself or a planet with a moon.
However, for the transit method of planet formation that the Kepler mission is doing, we see the planet move in front of its star, periodically blocking the light. If the timing isn't perfectly periodic, that may mean its arriving too soon sometime, and too late others. This could be due to a moon orbiting the planet. This is a fairly large effect for Kepler.
With current telescopes what's the distance limit that we can use astrometry with any hope of accuracy (how many parsecs out can this technique be used)? In a similar vein are you using a single viewing session with one (or a single set) of telescopes or are you making multiple observations at different points in the year to create a virtual optical array? Does this increase the astrometric measurements in a meaningful way?
With current telescopes? Well, the telescopes we used to do this work have been decommissioned (and bulldozed!), so I guess they're not current anymore. HST and some large ground-based telescopes are doing astrometry about 5x less precise than our program. The Europeans are building an array in Chile that should be able to do comparable precisions as our study (though over larger and more versatile set of target stars). Let us assume by "current" we mean something operating with precision similar to our program (35 micro-arcseconds).
A Jupiter around a Sunlike star at 10 parsecs would cause the star's position to vary by 1000 microarcseconds. An SNR of about 5-6 is needed to make a detection. Depending on the number of measurements made, this means the smallest signal detectable is about 200 micro-arcseconds, so this technique could work to around 50 parsecs (160 light-years). For truly "current" telescopes, HST or ground-based AO might work to 10 parsecs.
NASA/JPL has demonstrated technology for the Space Interferometry Mission (SIM) using the same method as our study, and shown that 1 micro-arcsecond astrometry is possible (again, on a much larger and far more versatile set of target stars). This could find Earthlike planets in the habitable zones of Sunlike stars to 10-20 parsecs. However, NASA canceled this program 2 weeks ago.
We used an array of 3 telescopes working together as an interferometer, creating a system with resolving power equal to that of a 100 meter telescope. We observed our target stars several nights per week, every week, for several years. To see the signal, repeated measurements are necessary---we are looking for motion of the stars with time, as the planet slowly sweeps out its orbit. The number of measurements on a given star (50-100 over the years) helps improve the measurement noise some, but not a lot.
Do you have an impression that double / multiple star systems were up to this point neglected in searches of planets? While "gravitational instability" model perhaps even suggests they are at least comparably likely to form planets?
It's actually a bit harder to find planets in binary systems using most of the current techniques. For example, the Doppler method is much more challenging when there are two spectra present to disentangle. When making measurements of extremely high, cutting-edge precision, and such complications can drastically reduce that precision. Similar, the transit technique is more challenging because there is more light present and because those studies usually require Doppler follow-up anyways to avoid false-positive signals (which have other, non-planetary, origins).
So binaries were not neglected due to choice or because astronomers dismissed them as being uninteresting for exoplanet studies, but rather because the measurement methods could not work sufficiently well for them.
We developed our new astrometry technique specifically because other methods were having problems with binaries, and because we recognized that binaries were an interesting laboratory to investigate planet formation (a null result also would have been enlightening on this topic).
(question is how much we would care; those radiation numbers certainly look awfully high)
Regarding magnetic field: due to interactions in the field of Jupiter, its radio emissions can apparently "outshine" the Sun; occasionally... Maybe somebody who's strongly into radio astronomy, and available to you, can comment if our radio telescopes have the potential of resolving such source?
We don't yet have radio telescopes powerful enough to detect the radio emission of a planet like Jupiter around another star. There has been some recent speculation that if the radio brightness increases as the planet is brought closer to its star, and if that scaling is optimistically strong, then maybe some next generation radio telescopes could do so for the closest-in giant planets (i.e. the hot Jupiters), though this is planet not one of those.
I've always thought that a binary system would create eddies in the dust and that mass caught in the eddies would coalesce quite quickly (incidentally, becoming a mass big enough to draw in more mass at an increasingly faster rate).
I haven't seen any of the planetary creation models, so how much do they consider this kind of eddying and coalescence? I would think the eddying would be greater in a binary or trinary system than a single star system.
Yes, an alternative model of planet formation called "gravitational instability" or "gravitational collapse" predicts that planets form in this way. That method is predicted to form planets very rapidly, and while there is not universal agreement on the subject, it seems likely that this is enhanced in binary systems. In the study, we discuss that this alternative model is one way to solve the problem of how this planet (and similar other ones) formed. This finding offers significant support to that alternative theory.
Note that we are not claiming that the "core accretion" model does not happen in nature. Rather, it cannot be the only method by which planets form.
My understanding is that the moons of Jupiter are not human-habitable with any current technology on account of fierce charged particle radiation from the strong magnetic fields. Do I have this right, or does this only apply to Io, which is in one of the radiation belts?
With sufficient shielding, this could probably be overcome. However, there are other reasons those moons might not be comfortable places for humans: wrong temperature, no liquid water on the surface, no atmospheres, etc. On this new planet, it doesn't seem likely we'll be measuring its magnetic field any time soon, so it's a bit early to speculate more.
Any questions? I'll try to answer responses to this post.
How can so much about the planet be observed without knowing which star the planet orbits? I'd think that information would be critical before any of the other information could be inferred.
The planet was discovered by measuring variations in the separation of the two stars. Their separation changes very slowly as the stars orbit each other, and on top of that motion, we found a very small wobble in their separation that repeats every ~3 years. That 3 year effect is the reaction of one of the stars to the planet orbiting it. Since we are measuring the relative separations of the stars, there is no way to know which one is wobbling. For the science content, it turns out not to matter nearly as much as one might think.
I am an author of the paper in which this discovery was reported. You can find a copy of the paper here.
While the planet probably is near the habitable zone, this isn't the first time a giant planet has been found in the habitable zone of a star, and while it could have moons, there isn't any reason to speculate more about this planet than any of the others.
However, this planet is important for two other reasons: 1. It was the first planet discovered using a technique called "astrometry", which is measuring the positions of stars in the sky, as the move up/down and left/right in reaction to a planet orbiting it. This technique has the potential to find earthlike planets in the habitable zones of nearby stars. 2. It is found in a binary system and the second star is close enough that its gravity would have impacted planet formation. The leading theory of planet formation, called "core accretion", requires millions of year for planets to form, as dust in a disk around the star collides together and clings electrostatically (similar to the way dustballs collect on a hardwood floor). Eventually the dustballs grow large enough to be considered rocks, those collide and grow bigger, etc. But the second star's gravity would cause the dust to be swept out of the system in just thousands of years, far too little time for core accretion to occur. Thus, we need a different mechanism to explain planet formation in this system. This isn't the only such binary, but it this study does offer more controlled statistics of how frequently such binaries host planets, and these facts combined show that some had to form in the binary itself---the chances of a binary interacting with another star (that originally hosted the planet), leading to an exchange where the binary picks up the star, are much too small to explain the high rate observed.
Also, here is another press story covering the discovery (by the way, stars have multiple names---don't be confused that this article calls it "HR 7162" and the other one refers to "HD 176051"---they really are the same system). The third figure on the right hand panel is particularly useful.
Any questions? I'll try to answer responses to this post.
Eric, is this you? If so, you actually know me in real life. A couple of years ago I was in the office 2 doors away from you.
Anyways, yes, a moon might tidally lock with the planet, preventing tidal locking with the star, but I wouldn't be surprised if simulations showed that those same tides would disrupt the planet-moon pair, causing that subsystem to be unstable. The stability of large moons in such a configuration is worth checking. Want to work on a new paper?
Still, even without tidal locking, there is good likelihood of habitable zones on the planet, but the tidal locking would be nice to help guarantee it.
It wasn't Butler, but rather Vogt who overstepped and said life was 100% likely. That was certainly overstepping things---a lot.
These are actually some rather complex atmospheric questions, but it seems likely that winds on the planet would help mix the temperatures all over, making them more moderate. But it is possible the dark side would have more liquidification and freezing of parts of the atmosphere. On Earth we call this rain and snow...not necessarily bad things.
It would be fascinating to study this planet's weather patterns to compare to the Earth's, from a scientific point of view. But it seems likely, no matter the patterns, that some stable point exists where life could thrive.
The comparison to Mercury seems to be based on the planet's proximity to its star. The star is much colder that the Sun, so a closer-in planet like Mercury would not be nearly as hot as Mercury finds itself.
In terms of size of the planet, this one is much more like the Earth. Mercury is really very small in comparison, and does not have much gravity to retain any atmosphere even if it were located where Earth is. So here the comparison to Mercury really doesn't work well.
It isn't clear that life only happened on Earth once. In particular, there are life communities around thermal vents in the deep ocean that very well may be the result of a completely different spark of life. For the rest of the Earth, once life gets going, it seems to become rather dominant, not leaving much opportunity for a second genesis, though how would we know if it happened?
In chemistry labs, we find that if the right basic elements are collected and put in early-Earth conditions, some of the complex molecules of life are assembled quite naturally. This is encouraging, at least.
But, 60/40 is just a complete guess, you are correct.
I'm not sure which comment of his this refers to, but almost certainly it is (b). The planet in question does not transit its star as view from Earth, which would have been an unlikely geometry, but if it had happened would have allowed very detailed follow-up study of the planet.
There is some controversy here. GJ 581 doesn't seem to be to dramatically variable. But others are. The lead of SETI wrote a recent paper claiming M dwarfs are not so active as to prevent life or even advanced life. However, this was in response to papers claiming the opposite. It's uncertain, but it seems GJ 581 is stable enough for long enough periods that life can evolve. Even our Sun isn't super stable, yet life exists. Thus ice ages, the Maunder Minimum and Mini-Ice-Age, and the like.
The spectrum of the star wouldn't necessarily tell us about the composition of planets. Some planet-star spectrum correlations have been seen as far as whether stars have planets, but these have not necessarily been tied to causation, and certainly not to composition of the planets. We would certainly need to calibrate any such tracer first, anyways.
The composition-age relationship for stars that you mention has more to do with the generation of stars. Stars today are made out of the waste products from the exploded material from previous stars. That material is enriched by the nuclear processes from those previous stars, meaning they start with more heavy elements. The current generation includes stars today and those from at least as long ago as 10 billion years. Beyond that you start to get to the beginnings of the universe and earlier generations of stars. So no big changes are really expected here, and the phenomenon you cite isn't currently believed to be planet-related, but rather just evolution-of-the-universe related, a very different topic.
I don't think anything about the spectra of the star could identify water at this level of precision. Planets are a billion times fainter than their stars. The spectra had signal-to-noise ratios of order 300:1, which is impressive enough, but nowhere close to enough to see features of the planet. (If Bill Gates, the man of $60 billion, woke up tomorrow with $60x300 = $18,000 to his name, he might need to be put on suicide watch. That is the level of change we are talking about.)
I answered this above, but probably after you posted this. Just for completion my answer is as follows. The RV-of-the-star itself data didn't imply the tidal locking, but rather extrapolations based on gravitational interactions, as below:
I believe they determined it as follows:
The planet is close to its star.
The planet has a fairly well known size.
The gravitational force on the near vs. far side can be calculated based on the planet-star distance and the planet size.
Guessing the planet is mostly rock (a very safe guess based on lots of planetary science information), we can guess how much frictional energy is lost in that differential stretching.
Based on the elements observed in the star, we can estimate the age as billions of years old.
The frictional forces would slow down the planet rotation much faster than billions of years (I forget the exact value, but less than 1 billion years; if you really want me to spend a few hours doing the calculation for a better estimate, let me know, but it wouldn't really matter). Thus, by now, it would be tidally locked.
The key is that the planet is closer to its star than the Earth. For example, Mercury (which isn't even as close to the Sun as GJ581g is to its star) is in a 3:2 tidal lock between its orbit and rotation. The full 1:1 lock is expected for closer planets. This is the case for the Earth's Moon, which is why we always see the same side of the Moon. This tidal locking is extremely well established with the Earth's Moon.
The gravitational force on the near vs. far side can be calculated based on the planet-star distance and the planet size.
Guessing the planet is mostly rock (a very safe guess based on lots of planetary science information), we can guess how much frictional energy is lost in that differential stretching.
Based on the elements observed in the star, we can estimate the age as billions of years old.
The frictional forces would slow down the planet rotation much faster than billions of years. Thus, by now, it would be tidally locked.
The key is that the planet is closer to its star than the Earth. For example, Mercury (which isn't even as close to the Sun as GJ581g is to its star) is in a 3:2 tidal lock between its orbit and rotation. The full 1:1 lock is expected for closer planets. This is the case for the Earth's Moon, which is why we always see the same side of the Moon. This tidal locking is extremely well established with the Earth's Moon.
Certainly life as we know it has evolved to day-night cycles. Life here would be different. Raccoons (night-animals) would be as confused as deer (day-animals). But there isn't reason to believe they couldn't have evolved differently.
As far as the narrow bands of tropics, this actually helps us determine that there are temperate zones. I posted the following above, but after your post, I just don't want to retype: "However, the great thing about this planet is that there is almost certainly a "too-hot" part, and a "too-cold" part, for humans, due to the tidal locking that you point out. However, somewhere between there, there must be a "just-right" part. This helps confirm that there is a habitable zone on the star."
The gravitational dynamics are rather well studied, for orbital stability. This is a rather robust part of the study (which, as someone interested in many-body dynamics, a very complex subject, is always surprising to me).
There might be some bizarre weather patterns, but there will be a region of what would be, to us humans, a comfortable region. This strongly suggests a nice region for life as we know it.
Could life exist as-we-do-not-know-it in a different extreme environment? Maybe. But a simpler jump is to life-as-we-do-know-it being elsewhere, since we have evidence such life does exist here, so that is why finding a human-suitable environment is so promising.
The weather might not be fun, that's for sure. But ask people in Alaska and the Mojabe---life exists nonetheless. It might be fun (or not) to be a weatherman there.
Honestly, that conclusion was a bit premature. The other coauthors (including my coworkers) avoided speculating on this point.
His conclusion was based on the idea that where liquid water can be present, so far we have always found life to within out ability to identify it. Thus, since there seems some high probability that liquid water *could* exist on this planet (though no evidence thereof, yet---it just seems likely due to the temperature and because water is such a simple and universally common molecule), and where we've found water we've found life (even in circumstances that would be considered unpleasant), he jumped to saying life was likely.
I personally think that it is premature to speculate on life in this system, since so little evidence is available. If pushed to make a call by Vegas, I'd have to say life was more likely than not on this planet, but my line would not be near 100%. Probably closer to 60/40.
Though a big fan of sci-fi (I would have to be as someone who studied astronomy), I'm afraid I'm not familiar with this one.
However, the great thing about this planet is that there is almost certainly a "too-hot" part, and a "too-cold" part, for humans, due to the tidal locking that you point out. However, somewhere between there, there must be a "just-right" part. This helps confirm that there is a habitable zone on the star.
I actually work quite closely with 2 of the authors of the paper that reports these results. Any questions? I'll try to respond to posts between now and 2 October.
Slashdot Mirror
The scientific community fights for years over one Hubble telescope - and some shady agency has two?
They can afford to "give them away" now. Probably because they have something much better now?
Am I the only one who thinks there is something simply "wrong" with all this? (And yes, I find it good those things are *now*, better: *finally*, used for science)
No, some shady agency does not have two. They have two surplus and obsolete (for their purposes) telescopes that were never launched. The NRO had many, many more than two such satellites in actual operation, and now we are being told those are no longer cutting edge, so they definitely have something much better.
There is a saying in astronomy that you cannot compare ground-observatory project costs to space-observatory project costs (every grad student ever has pointed out "for the cost of HST, imagine the huge telescope we could have built on the ground!" only to be rebuked with "space dollars are not the same as ground dollars). Similarly, military dollars are not the same as space dollars are not the same as ground dollars. Otherwise, one could naively say "for just the cost of one F22, we could have paid for XYZ science program by now."
Actually, HST has a 96 minute orbital period around the Earth, so it cannot stare continuously. But it can coadd several exposures over several orbits.
The real advantage of HST is that it is diffraction limited even on very faint objects (adaptive optics requires bright guide stars; even laser-guide-star adaptive optics needs a relatively bright natural star for the first-order correction) and the background light is much lower.
That(just(about(sums(it(up(.))))))
As I understand it... the bulk of the large moons in our solarsystem are made of low density materials essentially they're mini-frozen Jupiters in themselves or water ice balls like Enceladus. Wouldn't the formation of an Earthlike planet be precluded so close to a Jovian mass... Something with a substantial rocky core like Earth's forming at the same distance as a Jovian would have become a Jovian itself and most likely would have merged with the parent body? The Inner planets as I understand it were Jovians that had the bulk of thier gas envelopes blown clear during our Sun's T'Tauri phase of super strong solar winds.
The major moons in our solar system have densities between those of the terrestrial worlds and those of the giants. I certainly would not consider them to be mini-Jupiters.
Planet formation isn't as simple as our old ideas of "if it forms here, it will be a gas giant, if not, it will be terrestrial". Clearly there are a lot more details that we don't know yet.
Would it be easier to detect the existence of large (small planet-sized) moons around a gas giant than earth-sized planets around a star? Would not the perturbation of the gas giant be easier to detect because the mass ratios are closer (large moon to gas giant vs earth-sized planet to star)?
If so, detecting the gas giant in the habitable zone and then looking for evidence of large moons or companion bodies could allow detection of candidates for life.
I assume this would entail detailed, direct observation of the gas giant, but I would imagine that will happen sooner than detailed, direct observation of an earth-sized body.
Not with this method, since we are seeing the reflex motion of the star, which wouldn't be very different at all if the thing orbiting it were a planet by itself or a planet with a moon.
However, for the transit method of planet formation that the Kepler mission is doing, we see the planet move in front of its star, periodically blocking the light. If the timing isn't perfectly periodic, that may mean its arriving too soon sometime, and too late others. This could be due to a moon orbiting the planet. This is a fairly large effect for Kepler.
With current telescopes what's the distance limit that we can use astrometry with any hope of accuracy (how many parsecs out can this technique be used)? In a similar vein are you using a single viewing session with one (or a single set) of telescopes or are you making multiple observations at different points in the year to create a virtual optical array? Does this increase the astrometric measurements in a meaningful way?
With current telescopes? Well, the telescopes we used to do this work have been decommissioned (and bulldozed!), so I guess they're not current anymore. HST and some large ground-based telescopes are doing astrometry about 5x less precise than our program. The Europeans are building an array in Chile that should be able to do comparable precisions as our study (though over larger and more versatile set of target stars). Let us assume by "current" we mean something operating with precision similar to our program (35 micro-arcseconds).
A Jupiter around a Sunlike star at 10 parsecs would cause the star's position to vary by 1000 microarcseconds. An SNR of about 5-6 is needed to make a detection. Depending on the number of measurements made, this means the smallest signal detectable is about 200 micro-arcseconds, so this technique could work to around 50 parsecs (160 light-years). For truly "current" telescopes, HST or ground-based AO might work to 10 parsecs.
NASA/JPL has demonstrated technology for the Space Interferometry Mission (SIM) using the same method as our study, and shown that 1 micro-arcsecond astrometry is possible (again, on a much larger and far more versatile set of target stars). This could find Earthlike planets in the habitable zones of Sunlike stars to 10-20 parsecs. However, NASA canceled this program 2 weeks ago.
We used an array of 3 telescopes working together as an interferometer, creating a system with resolving power equal to that of a 100 meter telescope. We observed our target stars several nights per week, every week, for several years. To see the signal, repeated measurements are necessary---we are looking for motion of the stars with time, as the planet slowly sweeps out its orbit. The number of measurements on a given star (50-100 over the years) helps improve the measurement noise some, but not a lot.
Do you have an impression that double / multiple star systems were up to this point neglected in searches of planets? While "gravitational instability" model perhaps even suggests they are at least comparably likely to form planets?
It's actually a bit harder to find planets in binary systems using most of the current techniques. For example, the Doppler method is much more challenging when there are two spectra present to disentangle. When making measurements of extremely high, cutting-edge precision, and such complications can drastically reduce that precision. Similar, the transit technique is more challenging because there is more light present and because those studies usually require Doppler follow-up anyways to avoid false-positive signals (which have other, non-planetary, origins).
So binaries were not neglected due to choice or because astronomers dismissed them as being uninteresting for exoplanet studies, but rather because the measurement methods could not work sufficiently well for them.
We developed our new astrometry technique specifically because other methods were having problems with binaries, and because we recognized that binaries were an interesting laboratory to investigate planet formation (a null result also would have been enlightening on this topic).
(question is how much we would care; those radiation numbers certainly look awfully high)
Regarding magnetic field: due to interactions in the field of Jupiter, its radio emissions can apparently "outshine" the Sun; occasionally...
Maybe somebody who's strongly into radio astronomy, and available to you, can comment if our radio telescopes have the potential of resolving such source?
We don't yet have radio telescopes powerful enough to detect the radio emission of a planet like Jupiter around another star. There has been some recent speculation that if the radio brightness increases as the planet is brought closer to its star, and if that scaling is optimistically strong, then maybe some next generation radio telescopes could do so for the closest-in giant planets (i.e. the hot Jupiters), though this is planet not one of those.
I've always thought that a binary system would create eddies in the dust and that mass caught in the eddies would coalesce quite quickly (incidentally, becoming a mass big enough to draw in more mass at an increasingly faster rate).
I haven't seen any of the planetary creation models, so how much do they consider this kind of eddying and coalescence? I would think the eddying would be greater in a binary or trinary system than a single star system.
Yes, an alternative model of planet formation called "gravitational instability" or "gravitational collapse" predicts that planets form in this way. That method is predicted to form planets very rapidly, and while there is not universal agreement on the subject, it seems likely that this is enhanced in binary systems. In the study, we discuss that this alternative model is one way to solve the problem of how this planet (and similar other ones) formed. This finding offers significant support to that alternative theory.
Note that we are not claiming that the "core accretion" model does not happen in nature. Rather, it cannot be the only method by which planets form.
My understanding is that the moons of Jupiter are not human-habitable with any current technology on account of fierce charged particle radiation from the strong magnetic fields. Do I have this right, or does this only apply to Io, which is in one of the radiation belts?
With sufficient shielding, this could probably be overcome. However, there are other reasons those moons might not be comfortable places for humans: wrong temperature, no liquid water on the surface, no atmospheres, etc. On this new planet, it doesn't seem likely we'll be measuring its magnetic field any time soon, so it's a bit early to speculate more.
Any questions? I'll try to answer responses to this post.
How can so much about the planet be observed without knowing which star the planet orbits? I'd think that information would be critical before any of the other information could be inferred.
The planet was discovered by measuring variations in the separation of the two stars. Their separation changes very slowly as the stars orbit each other, and on top of that motion, we found a very small wobble in their separation that repeats every ~3 years. That 3 year effect is the reaction of one of the stars to the planet orbiting it. Since we are measuring the relative separations of the stars, there is no way to know which one is wobbling. For the science content, it turns out not to matter nearly as much as one might think.
I am an author of the paper in which this discovery was reported. You can find a copy of the paper here.
While the planet probably is near the habitable zone, this isn't the first time a giant planet has been found in the habitable zone of a star, and while it could have moons, there isn't any reason to speculate more about this planet than any of the others.
However, this planet is important for two other reasons:
1. It was the first planet discovered using a technique called "astrometry", which is measuring the positions of stars in the sky, as the move up/down and left/right in reaction to a planet orbiting it. This technique has the potential to find earthlike planets in the habitable zones of nearby stars.
2. It is found in a binary system and the second star is close enough that its gravity would have impacted planet formation. The leading theory of planet formation, called "core accretion", requires millions of year for planets to form, as dust in a disk around the star collides together and clings electrostatically (similar to the way dustballs collect on a hardwood floor). Eventually the dustballs grow large enough to be considered rocks, those collide and grow bigger, etc. But the second star's gravity would cause the dust to be swept out of the system in just thousands of years, far too little time for core accretion to occur. Thus, we need a different mechanism to explain planet formation in this system. This isn't the only such binary, but it this study does offer more controlled statistics of how frequently such binaries host planets, and these facts combined show that some had to form in the binary itself---the chances of a binary interacting with another star (that originally hosted the planet), leading to an exchange where the binary picks up the star, are much too small to explain the high rate observed.
Also, here is another press story covering the discovery (by the way, stars have multiple names---don't be confused that this article calls it "HR 7162" and the other one refers to "HD 176051"---they really are the same system). The third figure on the right hand panel is particularly useful.
Any questions? I'll try to answer responses to this post.
Eric, is this you? If so, you actually know me in real life. A couple of years ago I was in the office 2 doors away from you.
Anyways, yes, a moon might tidally lock with the planet, preventing tidal locking with the star, but I wouldn't be surprised if simulations showed that those same tides would disrupt the planet-moon pair, causing that subsystem to be unstable. The stability of large moons in such a configuration is worth checking. Want to work on a new paper?
Still, even without tidal locking, there is good likelihood of habitable zones on the planet, but the tidal locking would be nice to help guarantee it.
It wasn't Butler, but rather Vogt who overstepped and said life was 100% likely. That was certainly overstepping things---a lot.
I'm right here. I promise I do exist. Really.
These are actually some rather complex atmospheric questions, but it seems likely that winds on the planet would help mix the temperatures all over, making them more moderate. But it is possible the dark side would have more liquidification and freezing of parts of the atmosphere. On Earth we call this rain and snow...not necessarily bad things.
It would be fascinating to study this planet's weather patterns to compare to the Earth's, from a scientific point of view. But it seems likely, no matter the patterns, that some stable point exists where life could thrive.
The comparison to Mercury seems to be based on the planet's proximity to its star. The star is much colder that the Sun, so a closer-in planet like Mercury would not be nearly as hot as Mercury finds itself.
In terms of size of the planet, this one is much more like the Earth. Mercury is really very small in comparison, and does not have much gravity to retain any atmosphere even if it were located where Earth is. So here the comparison to Mercury really doesn't work well.
It isn't clear that life only happened on Earth once. In particular, there are life communities around thermal vents in the deep ocean that very well may be the result of a completely different spark of life. For the rest of the Earth, once life gets going, it seems to become rather dominant, not leaving much opportunity for a second genesis, though how would we know if it happened?
In chemistry labs, we find that if the right basic elements are collected and put in early-Earth conditions, some of the complex molecules of life are assembled quite naturally. This is encouraging, at least.
But, 60/40 is just a complete guess, you are correct.
I'm not sure which comment of his this refers to, but almost certainly it is (b). The planet in question does not transit its star as view from Earth, which would have been an unlikely geometry, but if it had happened would have allowed very detailed follow-up study of the planet.
Good point!
There is some controversy here. GJ 581 doesn't seem to be to dramatically variable. But others are. The lead of SETI wrote a recent paper claiming M dwarfs are not so active as to prevent life or even advanced life. However, this was in response to papers claiming the opposite. It's uncertain, but it seems GJ 581 is stable enough for long enough periods that life can evolve. Even our Sun isn't super stable, yet life exists. Thus ice ages, the Maunder Minimum and Mini-Ice-Age, and the like.
The spectrum of the star wouldn't necessarily tell us about the composition of planets. Some planet-star spectrum correlations have been seen as far as whether stars have planets, but these have not necessarily been tied to causation, and certainly not to composition of the planets. We would certainly need to calibrate any such tracer first, anyways.
The composition-age relationship for stars that you mention has more to do with the generation of stars. Stars today are made out of the waste products from the exploded material from previous stars. That material is enriched by the nuclear processes from those previous stars, meaning they start with more heavy elements. The current generation includes stars today and those from at least as long ago as 10 billion years. Beyond that you start to get to the beginnings of the universe and earlier generations of stars. So no big changes are really expected here, and the phenomenon you cite isn't currently believed to be planet-related, but rather just evolution-of-the-universe related, a very different topic.
I don't think anything about the spectra of the star could identify water at this level of precision. Planets are a billion times fainter than their stars. The spectra had signal-to-noise ratios of order 300:1, which is impressive enough, but nowhere close to enough to see features of the planet. (If Bill Gates, the man of $60 billion, woke up tomorrow with $60x300 = $18,000 to his name, he might need to be put on suicide watch. That is the level of change we are talking about.)
I answered this above, but probably after you posted this. Just for completion my answer is as follows. The RV-of-the-star itself data didn't imply the tidal locking, but rather extrapolations based on gravitational interactions, as below:
I believe they determined it as follows:
The planet is close to its star.
The planet has a fairly well known size.
The gravitational force on the near vs. far side can be calculated based on the planet-star distance and the planet size.
Guessing the planet is mostly rock (a very safe guess based on lots of planetary science information), we can guess how much frictional energy is lost in that differential stretching.
Based on the elements observed in the star, we can estimate the age as billions of years old.
The frictional forces would slow down the planet rotation much faster than billions of years (I forget the exact value, but less than 1 billion years; if you really want me to spend a few hours doing the calculation for a better estimate, let me know, but it wouldn't really matter). Thus, by now, it would be tidally locked.
The key is that the planet is closer to its star than the Earth. For example, Mercury (which isn't even as close to the Sun as GJ581g is to its star) is in a 3:2 tidal lock between its orbit and rotation. The full 1:1 lock is expected for closer planets. This is the case for the Earth's Moon, which is why we always see the same side of the Moon. This tidal locking is extremely well established with the Earth's Moon.
I believe they determined it as follows:
The planet is close to its star.
The planet has a fairly well known size.
The gravitational force on the near vs. far side can be calculated based on the planet-star distance and the planet size.
Guessing the planet is mostly rock (a very safe guess based on lots of planetary science information), we can guess how much frictional energy is lost in that differential stretching.
Based on the elements observed in the star, we can estimate the age as billions of years old.
The frictional forces would slow down the planet rotation much faster than billions of years. Thus, by now, it would be tidally locked.
The key is that the planet is closer to its star than the Earth. For example, Mercury (which isn't even as close to the Sun as GJ581g is to its star) is in a 3:2 tidal lock between its orbit and rotation. The full 1:1 lock is expected for closer planets. This is the case for the Earth's Moon, which is why we always see the same side of the Moon. This tidal locking is extremely well established with the Earth's Moon.
Certainly life as we know it has evolved to day-night cycles. Life here would be different. Raccoons (night-animals) would be as confused as deer (day-animals). But there isn't reason to believe they couldn't have evolved differently.
As far as the narrow bands of tropics, this actually helps us determine that there are temperate zones. I posted the following above, but after your post, I just don't want to retype:
"However, the great thing about this planet is that there is almost certainly a "too-hot" part, and a "too-cold" part, for humans, due to the tidal locking that you point out. However, somewhere between there, there must be a "just-right" part. This helps confirm that there is a habitable zone on the star."
The gravitational dynamics are rather well studied, for orbital stability. This is a rather robust part of the study (which, as someone interested in many-body dynamics, a very complex subject, is always surprising to me).
There might be some bizarre weather patterns, but there will be a region of what would be, to us humans, a comfortable region. This strongly suggests a nice region for life as we know it.
Could life exist as-we-do-not-know-it in a different extreme environment? Maybe. But a simpler jump is to life-as-we-do-know-it being elsewhere, since we have evidence such life does exist here, so that is why finding a human-suitable environment is so promising.
The weather might not be fun, that's for sure. But ask people in Alaska and the Mojabe---life exists nonetheless. It might be fun (or not) to be a weatherman there.
Honestly, that conclusion was a bit premature. The other coauthors (including my coworkers) avoided speculating on this point.
His conclusion was based on the idea that where liquid water can be present, so far we have always found life to within out ability to identify it. Thus, since there seems some high probability that liquid water *could* exist on this planet (though no evidence thereof, yet---it just seems likely due to the temperature and because water is such a simple and universally common molecule), and where we've found water we've found life (even in circumstances that would be considered unpleasant), he jumped to saying life was likely.
I personally think that it is premature to speculate on life in this system, since so little evidence is available. If pushed to make a call by Vegas, I'd have to say life was more likely than not on this planet, but my line would not be near 100%. Probably closer to 60/40.
Though a big fan of sci-fi (I would have to be as someone who studied astronomy), I'm afraid I'm not familiar with this one.
However, the great thing about this planet is that there is almost certainly a "too-hot" part, and a "too-cold" part, for humans, due to the tidal locking that you point out. However, somewhere between there, there must be a "just-right" part. This helps confirm that there is a habitable zone on the star.
I actually work quite closely with 2 of the authors of the paper that reports these results. Any questions? I'll try to respond to posts between now and 2 October.