We saw yesterday how a newly refined radial velocity technique allowed researchers to identify five planet candidates around the nearby star Tau Ceti. The latter has long held fascination for the exoplanet minded because it’s a G-class star not all that different from the Sun, and one of the planets around it — if confirmed — appears to be in its habitable zone. But smaller stars remain much in the news as well, as witness Gl 667C, a red dwarf (M-class) star in a triple system that also contains two closely spaced K-class stars with a semimajor axis of 1.82 AU.
M-class stars offer a lot to planet hunters, as new work using the HARPS spectrograph at La Silla is making clear. For one thing, a planet of a given size induces more radial velocity variation around a low-mass star than around a larger one, making the planet easier to spot. For another, red dwarfs are dimmer than G and K-class stars, with a habitable zone much closer to the star. Here again we get a larger radial velocity wobble than we would find with a larger star.
Philip Gregory (University of British Columbia) has performed a re-analysis of the HARPS data on Gl 667C that is getting play in the press because it identifies not one but three planets in the habitable zone. The star has about a third of the mass of the Sun, so according to Gregory’s figures, a habitable zone planet there produces seven times the radial velocity signature that a similar planet around a G-class star would generate. In 2011 Gl 667C was already known to be orbited by at least one planet, Gl 667C b, with a 7.2 day orbit, and there was evidence for other worlds. Later work confirmed the planet Gl 667C c in a 28-day orbit in the habitable zone.
All of this, says Gregory in the paper on his new work, has driven a serious effort to improve the statistical tools used to analyze radial velocity data, and we saw yesterday how Mikko Tuomi (University of Hertfordshire) and team put their own methods to work on Tau Ceti. Using his own algorithms, Gregory has already been able to confirm the existence of a second planet around 47 Ursae Majoris, but Gl 667C will draw more attention because of the dramatic placement of the planet candidates here. In this case, the astronomer included stellar jitter and orbital drift factors to re-analyze the HARPS data using Bayesian methods to refine the probability estimates.
The result: The detected signals include the already established planets in 7.2 and 28.1-day orbits, but also show possible planets in 30.8 (d), 38.8 (e), 53.2 and 91.3-day orbits (f). Gregory discounts the 53.2-day signal because it seems to be the result of surface activity on the star. All these candidates are more massive than Earth, but e is only 2.4 Earth masses. The signals at 30.8 and 38.8 days, if confirmed, would join Gl 667C c as planets in the habitable zone. Gregory finds that the 91.3 day orbit would take that planet inside the outermost edge of the habitable zone, although its eccentric orbit would keep it outside the HZ for the majority of time.
Image: The darker area is the orbital region that remains continuously habitable during at least 5 Gyr as a function of the stellar mass (Selsis et al. 2007). The light grey region gives the theoretical inner (runaway greenhouse) and outer limits with 50% cloudiness, with H2O and CO2 clouds, respectively. The dotted boundaries correspond to the extreme theoretical limits, found with a 100% cloud cover. The dashed line indicates the distance at which a 1 M? planet on a circular orbit becomes tidally locked in less than 1 Gyr. Credit: Philip Gregory.
The scientist is careful to note that new simulations will be needed to determine which of the planetary signals are consistent with a stable planetary system. Look particularly at the closeness of the 28.1 and 30.8 day orbits, where the semi-major axis differs by a mere 0.007 AU. This sets up a closest approach, as Gregory notes, every 323 days, doubtless a fascinating astronomical spectacle from the surfaces of these possible worlds. And the author cites the Kepler mission’s own findings of systems with close planetary separations, including Kepler 36 b and c (0.014 AU), Kepler 42 b and d (0.0038 AU), and KOI 55b and c (0.0016 AU).
We may have to start getting used to solar systems with close planetary separations, unlike the relatively spacious inner system we see around the Sun. I’m reminded of something Steve Vogt (UC-Santa Cruz) said in the news release on the Tau Ceti story: “We are now beginning to understand that Nature seems to overwhelmingly prefer systems that have multiple planets with orbits of less than one hundred days. This is quite unlike our own solar system where there is nothing with an orbit inside that of Mercury. So our solar system is, in some sense, a bit of a freak and not the most typical kind of system that Nature cooks up.” Vogt was not speaking of M-dwarfs, of course, but the statement has no better illustration than Gl 667C’s possible planets.
The paper is Gregory, “Additional Keplerian Signals in the HARPS data for Gliese 667C from a Bayesian Re-analysis” (preprint).
The more solar systems found the more the words “ours is freak”, “ours is strange” spoken. The more we want to throw “rare solar system”, the more it bite us in the ass.
Paradox much?
Has Gregory done a long-term stability analysis on the newly configured system(s)? I believe that there will be no problem with the two closely spaced planets, BUT; the eccentricity of planet “F” may prohibit the existance of ANY hypothetical planet “g” in a 55 day orbit, thus confirming his suspicion that THAT “signal is indeed a false positive.
Sorry! I meant 53.2 instead of 55.
I am not so sure of this. Our observational methods overwhelmingly prefer such systems. As for Nature, I just don’t believe we know, yet.
No, and he mentions that such a stability analysis needs to be done. Presumably if these planets do exist the eccentricities have to be very close to zero: this is compatible with the error bars but the central values for the eccentricity lead to crossing orbits for planets c and “d”.
You have to wonder what any intelligent species that evolved there would make of another planet going past every ~320 days large enough to present a visible disc…
With closely-spaced planets now being verified, perhaps we will find a “Nemesis” system soon, where two planets share the same orbit 180 degrees apart, or even Trojan systems with planets 60 degrees from a more massive one. Orbital mechanics specialists: would any of these situations be stable for long periods?
As far as our system being a freak, we need to keep in mind the bais these surveys have for large, in tight planets. Also we shouldn’t necessarily expect M class stars to have similar setups to G class stars.
I’d question whether this system is stable. For a very small habitable zone, having three planets would mean they’re all awfully close to each other.
Here’s another possibility to throw into the mix: Could there exist stable (and habitable) planetary systems in which a planet orbits an M class (red dwarf) star (or a brown dwarf) *outside* the red dwarf’s (or the brown dwarf’s) habitable zone, but is habitable because the red dwarf (or brown dwarf) orbits a hotter (K or G) class star fairly closely? In such a case, the combined light and heat from both stars would make the planet habitable. In order to have a stable orbit in such a circumstance, the planet would have to orbit quite close to a very small and cool red dwarf (or a brown dwarf) so that the larger and brighter star’s gravity wouldn’t disrupt the planet’s orbit.
@James
The problem with that scenario is not stability, but the amount of energy the planet can receive itself. The system that you imagined is like Alpha Centauri system where a planet orbiting outside HZ Centauri B and get some light frrm Centauri A. But as we can see, half of year, the planet will in the opposite orbit of A and will get no additional light from it. Which means half of year, the life on the planet has to be suspended.
@ Areceibo, not so if the M star’s planet’s orbit lies totally within the bright star’s HZ, which is quite possible. Extreme summer vs winter but ‘habitable’.
Well yes there are biases, but that does not mean the derivations are hopelessly inaccurate. In fact the studies that are coming to the conclusion that our solar system represents a fairly rare mode of planetary formation do take into account the detection biases. Taking those into account, for solar-type stars it looks like the hot/warm super-Earth systems are somewhere around 50% or more, while systems with Jupiter analogues are at maybe 10% of systems (and a Jupiter analogue does not mean the system is a solar system analogue, e.g. Upsilon Andromedae hosts a Jupiter-analogue planet “e”, but has a very different inner system to our own). Hot Jupiters are down at maybe 1% so we are perhaps not the rarest type of system out there…
And in any case if the M-dwarf systems do exhibit substantial differences from the ones around G-dwarfs, we are more likely to be an outlier by virtue of having a relatively rare star that is more massive than the average.
It is difficult to imagine how planetary systems with orbital separations as little as 0.007, 0.014, 0.0038, and 0.0016 AU can have long-term stability.
Ok, most of these may be M dwarfs, but Kepler-36 is a G-star.
Arecibo wrote:
[The problem with that scenario is not stability, but the amount of energy the planet can receive itself. The system that you imagined is like Alpha Centauri system where a planet orbiting outside HZ Centauri B and get some light frrm Centauri A. But as we can see, half of year, the planet will in the opposite orbit of A and will get no additional light from it. Which means half of year, the life on the planet has to be suspended.]
The stellar/planetary system I had in mind would be much more compact than the Alpha Centauri system (more like a Jovian planet with a habitable moon, whose primary orbits within the habitable zone of its star). A close-in, tidally-locked (rotation-wise) planet around a minimum-size red dwarf (or a brown dwarf) could have a year that would be just one Earth day or so long (or even less). If the red dwarf or brown dwarf orbited close to its more massive and brighter stellar companion (actually, both stars would be orbiting around a common barycenter between them, like Jupiter and the Sun), the planet could constantly receive sufficient light and heat from the brighter star, with a “night” (when the red dwarf or brown dwarf was between the planet and the brighter star) just a few hours long. Also:
Because of the constantly-changing phase angles between the three bodies, even the hemisphere of the planet that always faced toward the cooler star would receive light and heat from the brighter star as the two stars orbited about the system’s barycenter. Such (habitable) systems probably wouldn’t be common, but they could be another way that some planets manage to be abodes of life.
Just exactly what counts as a “solar system analog” anyways? It would seem at the very least it would have to have an earth analog in the habitable zone and a Jupiter analog.
Seeing as we have been observing barely long enough to detect a single orbit of a true Jupiter analog, and a true earth analog is barely within our detection capability at all, it would seem that the requirements for a “solar system analog” are stringent enough that they would have to be rare even if planetary distributions in the galaxy fell under no pattern at all and were entirely random from star to star.
At least the outer Solar system isn’t something very unusual, we have HR 8799 example which looks very similar (though scaled up, and we don’t know it’s inner structure and it may be different). My guess is that such systems with gas giants on wide orbits are common among more massive stars with less dense disks, formed if 1) protostar’s radiation had blown volatiles and fine dust out of the inner system before material could coalesce into superearths and hot neptunes and 2) there’s no catastrophic planetary migration from outer system. If the Sun is unusually small for such kind of systems (providing necessary main sequence lifetime), and life is unlikely to arise in densely packed ones for whatever reason (say, inevitable 100km+ deep oceans if volatiles aren’t blown beyond the snow line), the rare Earth hypothesis gets some support…
About “M-dwarf in the G-star’s HZ” binaries, the imagination can go extreme here, to the world with one side a habitable earthlike surface with seas and forests, and another a lava ocean (the atmosphere is thin with many small convective cells on the dayside and without significant heat transfer to the earthlike side, though there are some librations and lava tsunamis from the asteroid impacts and geologic activity on the dayside) – and still firmly in the realm of possibility – what a treasure for sci-fi writers! :-)
Amphiox wrote: Just exactly what counts as a “solar system analog” anyways?
Hmmm… based on how many discussion we do, i believe “solar analog” need 4 criterions:
1. No planet so close in to it’s parent star. Based on mercury as our model, no planet should be orbiting inner than orbit that receive 10x earth’s sunlight/solar constant. Maybe we can “tolerate” to orbit that receive 15 solar constant.
2. No planet bigger than super-earth in habitable zone. Let say the upper limit is 8 earth mass. Neptune type that up to 15 earth mass is a no.
3. Big planet in the frost line zone.
4. The distance between planet is not close. Let say that a neighbor planet shouldn’t be be as big as earth’s moon seen from earth.
Asteroid belt is a wild card, though.
I’ve finally done a quick analysis of that graph, and am astonished at how much must be assumed in the planetary biogeochemical cycles implied in its modelling.
The very outer limit of the HZ of a 0.3 solar mass star can be read off as 0.22 to 0.23 on the said graph. Lets put it at 0.225AU. Now the bolometric luminosity of such a star should be 0.017 of our sun,, meaning that the sub solar point of a tidally locked planet there should receive fully a third more energy than the tropics on Earth, yet the model for this graph finds even that sweltering heat insufficient!!!