On Super-Earths and Alpha Centauri

by Paul Gilster on November 13, 2012

The discovery of Centauri B b, a small planet with a mass similar to Earth, continues to percolate in the news even if the initial buzz of discovery has worn off. Science News gives the new world a look in a recent article, noting the fact that with an orbital period of 3.236 days, this is not a place even remotely likely for life. Surface temperatures in the range of 1200 degrees Celsius are formidable obstacles, but of course the good news is the potential for other planets around Centauri B and, indeed, around its larger companion.

Centauri A may well host interesting worlds, but it’s a tough study because it’s given to the kind of stellar activity that can more readily mask a planetary signature than the quieter Centauri B. Even so, we can imagine the possibility of two planetary systems in close proximity, a scenario that would surely propel any technological civilization around one to investigate the other. We don’t have the driver for spaceflight in our system that an Earth-like world around Centauri B might have, a second habitable planet breathtakingly close around another star.

If we’ve ruled out planets larger than Neptune around any of the three Alpha Centauri stars, that leaves the door open for the small worlds that could be the most interesting if one or more turned up in the habitable zone, and it’s worth noting that on this score, Proxima Centauri is still in the game. But right now the incredibly tricky detection of Centauri B b needs confirmation, which could be delivered by Debra Fischer (Yale University). You’ll recall that Fischer has been working at Cerro Tololo (CTIO) in Chile to develop the high-resolution spectrometer known as CHIRON, commissioned in March of 2011, as part of her team’s search for rocky Alpha Centauri planets.

Leaving Copernicus Behind

Centauri B b might also be confirmed through detection of a transit, the chances being estimated in the region of 10% and perhaps, according to Greg Laughlin (UC-Santa Cruz) as high as 25%. While the necessary work continues, let’s move beyond the Alpha Centauri stars for a moment to talk about Laughlin’s latest work with Eugene Chiang. Exoplanet hunters have learned through experience to question assumptions, the most obvious of which is that our Solar System is in some sense ‘normal.’ Or as Laughlin writes on systemic, “There is an intriguing, seemingly anti-Copernican disconnect between the solar system and the extrasolar planets.”

Image: Exoplanet hunter Greg Laughlin, whose latest work re-examines how super-Earths form close to their stars. Credit: UC-Santa Cruz.

Maybe the reason exoplanets so often surprise us is that we base our thinking on our own Solar System, and the minimum-mass Solar nebula from which it grew, considering this a template. The rest of the galaxy may have other ideas. Consider that close-in super-Earths are common. Planets like these, showing up in abundance in Kepler data and Doppler velocity surveys, are a challenge to explain. Laughlin and Chiang say that more than half, if not nearly all Sun-like stars have planets with radii between 2 and 5 times that of Earth and orbital periods of less than 100 days. The researchers write:

Super-Earths are not anomalous; they are the rule that our Solar System breaks. In a sense, the burden of explaining planetary system architectures rests more heavily on the Solar System than on the rest of the Galaxy’s planet population at large.

The problem, then, is that our Solar System has no planets inside Mercury’s 88-day orbit. Is it possible our Solar System did not undergo the same kind of formation history that may be the dominant mode in the galaxy? To explore this, the researchers look at migration issues, for it is commonly thought that short-period planets formed several AU out from their stars and then migrated to their present location. But disk migration is poorly understood, and while it may be necessary to explain hot Jupiters, Laughlin and Chiang say it may not be the mechanism to explain the majority of planetary systems with super-Earths in inner orbits.

The alternative: Forget orbital migration and consider the possibility that super-Earths form right where they are, in circumstellar disks that extend inward from 0.5 AU. The researchers go to work on constructing a new template, the Minimum Mass Extrasolar Nebula (MMEN), which allows them to explore how such planets could form near their star:

Our order-of-magnitude sketches in this regard are promising. In-situ formation at small stellocentric distances has all the advantages that in-situ formation at large stellocentric distances does not: large surface densities, short dynamical times, and the deep gravity well of a parent star that keeps its planetary progeny in place.

The basic properties of planets forming at their current orbital distance are made clear:

In-situ formation with no large-scale migration generates short-period planets with a lot of rock and metal and very little water. The accretion of nebular gas onto protoplanetary cores of metal produces H/He-rich atmospheres of possibly subsolar metallicity that expand planets to their observed radii. Retainment of primordial gas envelopes against photoevaporation leads to planets that can be similar in bulk density to Uranus and Neptune while being markedly different in composition. Close-in planets are not water worlds.

Laughlin and Chiang believe there should be observational consequences to such predictions, making the theory falsifiable. Hot young stars should lack close-in super-Earths because they would be too hot for planetesimal-building dust to survive. Brown dwarfs and M dwarfs should have close-in super-Earths and Earths orbiting them. Close-in planets grown from close circumstellar disks should also have orbital planes aligned with the equatorial planes of their host stars. I’ll send you to the paper to go through the entire list of predictions, all of which should allow these ideas to be probed, but I do want to mention one last prediction having a bearing on Centauri B b. For if Laughlin and Chiang are right, then binary systems offer a good test.

After all, close binary systems should make planetary migration extremely difficult. The close companion would disrupt planet formation at large distances from the star. Planets orbiting Centauri B inside the 0.5 AU boundary would be incompatible with migration, and of course, we now have such a planet, along with the likelihood of finding more. Here I drop back to the Science News article, which quotes Laughlin on Centauri B: “I think that the odds that there’s an interesting planet, a truly interesting planet in the system, are very high, given that this one is there.” And if he’s right, that interesting, potentially habitable world may serve as further evidence for the theory that such planets formed right where they are found.

The paper is Chiang and Laughlin, “The Minimum-Mass Extrasolar Nebula: In-Situ Formation of Close-In Super-Earths,” submitted to Monthly Notices of the Royal Astronomical Society (preprint).

tzf_img_post

{ 27 comments }

Mike November 13, 2012 at 12:53

This to my mind has to be the most astonishing results so far from Kepler and the ground based RV searches. That many if not most planetary systems consist of super Earths and or mini Neptunes crammed into close in orbits. This also requires a new approach to explaining how what maybe the majority of planets form. Further observations will certainly confirm or refute this assertion. Much fun for the planet formation theorists.

I’d read the ARXIV paper yesterday and it really gave me pause for thought.
The big question that will need to be addressed by further observations is do these type of systems commonly have planets orbiting further out, in the habitable zone of the parent star? At least in the case of G and K dwarfs orbits much longer then 100 days, rather 200 to 400 days are needed for a HZ location. Do these crowded inner systems usually have outer planets? Inquiring minds what to know!

Thomas Mazanec November 13, 2012 at 13:35

Brown dwarfs and M dwarfs should have close-in super-Earths and Earths orbiting them.
In view of your recent column on the long lived red dwarf stars, this is quite interesting.

andy November 13, 2012 at 14:34

Might as well post this one here as well that I posted in the HD 40307g thread: Wu and Lithwick (arXiv 2012) come to a similar conclusion about the nature of small super-Earths (predominantly rocky, small water-mass fraction and hydrogen-rich atmospheres either from outgassing or accretion) based on the trends observed in the Kepler planets.

I am still fairly pessimistic about the prospects for habitable worlds at Alpha Centauri: we can still face the prospect that the entire habitable zone was accretion-hostile and no planets formed there, i.e. the scenario detailed by Thébault et al. (2009) – see also the Centauri Dreams article about this.

Hopefully it won’t be too long before we know for sure.

Wojciech November 13, 2012 at 18:37

Not all are so pessimistic about AC Bb, Karl Schroeder, a hard SF author, imagined both pessimistic and optimistic scenario for the possible shape of the planet, and came to conclusion that in both cases it is possible to either colonize it or help in colonization.
http://www.kschroeder.com/weblog/archive/2012/10/17/colonizing-alpha-centauri-the-least-and-most-we

Phil November 13, 2012 at 18:47

What if it turns out that most rocky-ish planets in the Hz are outer planets of tightly packed systems, like the still to be confirmed HD 40307g?
What if most rocky-ish planets in the Hz turn out to be moons of gas giants?
Or what if….oh never mind :)

One thing I do find interesting is that various teams seem to be deliberately turning their attention to systems that are RV quiet (so, no pesky gas giants). Whereas once just a detection of a hot Jupiter was a major event, now sensitivity is getting good enough to deliberately target stars that previously ‘had nothing to say to us’. What about all those pole-on stars out there that wont have big Keplarian RV variations? Tau ceti is one these I believe. Maybe Prox cen is as well (no reason to think its spin axis is aligned with A and B)? This is where space based astrometry is going to have to finally pick up the ball….Exciting times….

P

xcalibur November 14, 2012 at 2:05

Great work!

However, it’s too early to make blanket assertions about the layout of all solar systems. Observation bias is still too much of a limitation on the relevant data. Naturally, planets closer to their star will be easier to find than those further out. Not that I think distant super-earths are common, but it’s important to fill in the gaps with observation. Once we have more complete solar system models, including the outliers, we’ll have a stronger foundation for understanding solar system formation.

Still, there’s nothing wrong with preliminary/partial study and speculation while data comes in.

Enzo November 14, 2012 at 2:29

@Thomas Mazanec

That’s what I thought too : “A compact systems around a G star, if it stays at the same distance, should be ok with a lower mass star”.
However, looking at Kepler data, compact systems around smaller star are even more compact, resulting in inhabitable planets again :-(

kzb November 14, 2012 at 8:40

Seems a bit odd to me, that these close-in super Earths can retain hydrogen and helium, yet at the same time have no water?

jkittle November 14, 2012 at 11:22

with the announcement of another rogue planet, we are reminded that the closest target for an extra solar system probe or mission may not be a star system, but could be a rogue planet. The earlier discussion of detection of tidally heated moons is interesting in this connection , but unfortunately I do not think that moons would survive the stellar system ejection process, if a “planet ” forms in the accretion disk of a protostar. However, if a rogue planet were to directly accrete from a molecular cloud and not be initial bound to a star, then moons, even close- in ones many be stable .
Serious consideration needs to be given to better and more sensitive detection of any nearby “rogue” planets.. some form of low temperature survey / high proper motion or maybe some other novel method for identifying candidates fo the JWST to zoom in on.

stephen November 14, 2012 at 12:25

Heavier elements would be more abundant (as compared to carbon and the other lighter elements), wouldn’t they? So maybe life forms based on heavier elements might be a possibility. They might be micro-organisms in isolated pockets around the planet–if it’s tidally locked. If it’s not tidally locked, they might be floating in the atmosphere.

andy November 14, 2012 at 14:47

However, it’s too early to make blanket assertions about the layout of all solar systems. Observation bias is still too much of a limitation on the relevant data. Naturally, planets closer to their star will be easier to find than those further out.

If you find a very high incidence of the detectable planets (which are not found in the solar system), you can still conclude that the solar system is in the minority even if you can’t detect the inner solar system analogues. That is indeed what seems to have happened: the incidence of close-in super-Earths seems to be >50% (this seems to arise from both the HARPS sample and the Kepler sample), and we do not have any hot super-Earths in our solar system.

So yes, it does look like we can say that our solar system is in the minority. Perhaps inner system analogues may even turn out to be rarer than Jupiter-analogues. If so it is quite ironic since the modern era of planet detection basically began with the detection of a fairly decent analogue of our inner system.

So the question is: what led our inner solar system to end up looking like the planetary system around a millisecond pulsar?

Ronald November 14, 2012 at 20:37

Such a compact system may still possess a habitable earthlike planet, if there is a small terrestrial planet in the HZ, beyond the super-earths and Neptunes in the innermost system.
Until now, data, especially Kepler data, are mainly known for the inner systems up to about 0.5 AU. The anxious waiting is now for more data to come in.
Do planet sizes in such compact systems drop toward the outside?
So far, I would not count on that, at least not for most systems.

Furthermore, it would be interesting and relevant to be able to relate the difference between our type of system and this compact type of system to certain differences in elemental abundance.

R Kelley November 14, 2012 at 20:59

In a 2011 article, authors Brad Hansen and Norm Murray set forth the results of a study in which they found that the observed HARPS frequency of Hot Neptunes and Super Earths could be duplicated with a simple model of in-situ formation but only if there was 30-100 earth masses of rocky material within 1 AU of the star. The authors go on to note that this amount of disc material is in excess of traditional estimates of the amount of material within 1 AU but could be achieved if there was significant radial migration from the outer part of the disc and assembly after the migration. The case of Alpha Centauri B argues against there being any migration at all. How this affects potential planet formation in the habitable zone (either positively or negatively) is very unclear. Case 1: Assume there was migration – In this case, the passage of 30-100 Earth masses through the habitable zone into a very tight solar orbit would not necessarily have disrupted planet formation in the habitable zone. In fact, it may have added essential volatiles. Clearly, the passage of Jupiter to within 1.5 AUs of the Sun did not prevent the formation of Mars (although it lessened the size of Mars). Would a much smaller mass of about 50 earth masses have not permitted a world like Earth to still develop after the disc was eliminated? Of course, the alternative is that the Sun essentially serves as a giant sinkhole in which all planetary material within 1 AU of the star falls victim. This would be a very unfortunate result. Case 2: Assume there was no migration – In this case, it seems that the MMEN is gigantic compared to the MMSN. In this case, one must ask why did the Sun end up with such a much smaller reservoir of material than most stars? Is it possible that photoevaporation from the Sun dispersed the material before it could form into planets? In any event, again it would not seem that the giant amount of material in a very tight solar orbit would lead to disruption of a forming Earth in the habitable zone.
Hopefully, we will know the answers to these questions someday.

Ronald November 15, 2012 at 6:59

Stephen: the abundance of an element does not make it any more or less suitable for life. Take our own earth and life, for example. Carbon is relatively uncommon in comparison with silicon. And yet life is based on carbon, because of its unique properties to form very complex molecules. Silicon also has that to a certain degree, but pales in comparison with carbon.

I am pretty convinced that this wil be found to be a universal rule: life will always be based on carbon, and water. Oxygen is also very handy, but maybe not always absolutely necessary.

That makes this recent work by Laughlin and Chiang extra sobering: water may be very scarce in those inner compact systems, while carbon may be very common.

When I saw this paper by Laughlin and Chiang about in-situ compact systems of super-earths and Neptunes, with the MMEN graph, I got the impression I was looking at a landmark piece of work, a scientific truth.
The paper by Wu and Lithwick, quoted by andy above, on the two distinct classes of medium-sized planets (very rocky super-earths and vary gaseous Neptunes) adds to this important reality.
Though I sincerely wonder whether there is not a tiny touch of wishful thinking, from which even the greatest scientists are not entirely immune, in the idea that there is a ‘very high’ chance of there being a ‘truly interesting planet’ in the HZ of Alpha Centauri B, also given Thébault´s modelling work quoted by andy above.

stephen November 15, 2012 at 11:53

Ronald: Thanks, but…there have been speculations on other possibilities. Wikipedia has an article on hypothetical types of biochemistry.
http://en.wikipedia.org/wiki/Hypothetical_types_of_biochemistry
Also, David Darling has many articles on his site.
http://www.daviddarling.info/encyclopedia/A/astrobiology_entries.html

Not that any of this is all that likely…
Yes, there is some wishful/hopeful thinking here, most of them might be microorganisms, and I was thinking in terms of sf story ideas…
It’s fun to speculate.

Zen Blade November 15, 2012 at 12:29

Quick question or two and a comment or two as well.

Comment: Ronald, “I am pretty convinced that this wil be found to be a universal rule: life will always be based on carbon, and water. Oxygen is also very handy, but maybe not always absolutely necessary.” If you mean O2, I would agree with you. After all, the first creepings of life on Earth formed in an O2-poor atmosphere: http://en.wikipedia.org/wiki/File:Sauerstoffgehalt-1000mj2.png
But water does have oxygen, and if an organism can figure out a way to take advantage of even a small amount of O2 production [such as O2 production from photosynthesis or something], then there’s a chance an organism will evolve to take advantage of the O2 as aerobic metabolism is far more efficient than anaerobic.
So, basically I’d argue that you don’t need O2 to get things going (and actually you probably want to avoid it), but once life kicks off… O2 is great.

Question: Isn’t a premature to argue that our solar system is distinctly different in terms of planet formation/presence? I could be completely wrong here, but I guess my point would be that if we can detect thing “X” 100x better than we can detect thing “Y” and we see that thing “X” occurs 100x more frequently than thing “Y”… don’t we have to say that we don’t yet have enough data? Anyone? Thoughts? Explanations why we can conclude that we are really that different.

Question 2: Are we looking only at single-star systems or binary, etc…? Isn’t our planetary system already in the minority because we aren’t a multiple-star system? And are most of the systems being looked at single-star or multiple-star?

Cheers,
-dennis

FrankH November 15, 2012 at 16:11

@Ronald – at http://www.exoplanets.org/plots you can plot Fe/H vs # planets and semi-major axis; metallicity doesn’t seem to have an effect on the number of planets or their orbits. It does seem to have a small effect on the masses of the planets (slight skew towards more massive planets). I agree that the Laughlin and Chiang paper has a sense of “truth” to it; it may have an impact similar to the Nice model.

The oddity of our solar system may be due in part to the size of the cluster it was formed in and the Sun’s position in it as well as who the neighbors were. It would be interesting to see the types of planetary systems found in open clusters, especially if the types change towards the periphery or nearer the more massive members.

Ronald November 16, 2012 at 10:23

FrankH: yes, I know that very valuable site. But I find that graph a bit crude. I mean, I would rather like to see a graph of largest planet and/or total planetary mass per (inner) system against metallicity.
And there are indications that Fe as a parameter alone is also too crude, other elemental abundances may play an important role as well.

It is already well established that there is a (very) strong correlation between metallicity and the occurrence of giant planets: nearly all giant planets occur around stars with minimum (Fe) metallicity of -0.15 or higher, i.e. almost none below this limit, at least among solar type stars.

Other trends, such as heavier (refractory) elements versus lighter (volatile) elements, are being investigated.

andy November 16, 2012 at 15:57

And there are indications that Fe as a parameter alone is also too crude, other elemental abundances may play an important role as well.

Indeed, there has recently been some focus on the alpha-elements: magnesium seems to be a pretty good indicator for planets. See for example Adibekyan et al. (2012) “Overabundance of alpha-elements in exoplanet host stars” and Adibekyan et al. (2012) “Exploring the alpha-enhancement of metal-poor planet-hosting stars. The Kepler and HARPS samples“.

Question: Isn’t a premature to argue that our solar system is distinctly different in terms of planet formation/presence? I could be completely wrong here, but I guess my point would be that if we can detect thing “X” 100x better than we can detect thing “Y” and we see that thing “X” occurs 100x more frequently than thing “Y”… don’t we have to say that we don’t yet have enough data?

No. Suppose for example we only had the observational capability to detect, say, hot Jupiters, and we found that they occurred around 90% of stars (this is not actually the case, this is just an example). We’d then still be able to say that our solar system is in the minority despite not being able to detect solar system analogues at all.

What’s happened with the super-Earth systems is that we are finding that over half of solar-type stars have them. So despite the fact that (inner) solar system analogues are still out of reach of the observations, we’re still able to say that we’re in the minority, because all the fractions have to add up to 1.

at http://www.exoplanets.org/plots you can plot Fe/H vs # planets and semi-major axis; metallicity doesn’t seem to have an effect on the number of planets or their orbits.

Of course, that graph is biased by only showing the stars which have detected (and hence detectible) planets around them…

xcalibur November 16, 2012 at 21:11

@andy, what you’re saying is true. It’s a very safe bet at this point that our solar system is unusual. It’s interesting, because our G class star doesn’t seem particularly unique.

So the question is: what led our inner solar system to end up looking like the planetary system around a millisecond pulsar?

A compelling question, which will have to wait until we advance further in solar system formation theory. It’s likely that there were particular factors early on in the stellar life-cycle that skewed things, but there’s only speculation at this point.

Eniac November 17, 2012 at 12:57

No. Suppose for example we only had the observational capability to detect, say, hot Jupiters, and we found that they occurred around 90% of stars (this is not actually the case, this is just an example).

So, what is the actual percentage here? Is it 10%? 1%?

What’s happened with the super-Earth systems is that we are finding that over half of solar-type stars have them.

This is very interesting information. If true, you could indeed say that the solar system can no longer be the most common “type” of system, if type means with/without super Earths, with/without hot Jupiters, etc. It could still be the second most common among a half dozen such types, which leaves a lot of room for “ordinaryness”.

andy November 17, 2012 at 13:56

So, what is the actual percentage here? Is it 10%? 1%?

HJ systems seem to be at around the 1% level around solar-type stars (Wright et al. (2012)). They aren’t even common for giant planets, the majority of which seem to fall into the “eccentric Jupiters” category (periods over ~100 days, a large range of eccentricities). The eccentric Jupiters are the right hand population of giant planets in this figure from the Systemic blog.

spaceman November 19, 2012 at 6:32

It seems as though a false dichotomy existed for quite some time in the astronomical community regarding the subject of planets around other stars. One school of thought maintained that planets would be very rare since their creation relied on debris left over from the collision of passing stars. The other school of thought posited that if planets exist around other stars then they are likely to resemble the way the planets are configured in our solar system with gas giants further out and smaller terrestrial type planets closer to the parent star. However, as it turns out the most common type of system is one with medium sized worlds closer in.

One question I have is this: in the systems in which there exist close in super-earths which formed in-situ might the smaller earth sized worlds exist further out near or in the HZ? Of course, in the migration scenario any earth sized orbs in the habitable zone would be knocked out of the system into the interstellar void.

So, the apparent rarity of planetary systems resembling our own (now well established by the Kepler mission and radial velocity surveys) combined with the intrinsically low probability of inanimate matter changing into living matter provide, in my opinion, one of the most likely, viable solutions to the Fermi paradox.

Ronald November 19, 2012 at 10:48

spaceman: “One question I have is this: in the systems in which there exist close in super-earths which formed in-situ might the smaller earth sized worlds exist further out near or in the HZ?”

That is precisely the million dollar question! As fas as I know, this is not well established yet at present, because both with Kepler (transit method) and with HARPS (RV method) the innermost planets are usually the easiest and first to be discovered (with RV there is also the technological lower limit to the mass and orbital radius combination). Time will tell.

It would be interesting to know what theoretical models already exist for this.
Catanzarite and Shao, 2011, The Occurrence Rate of Earth Analog Planets Orbiting Sunlike Stars, estimate eta Earth between 1 and 3%, based on extrapolation of early 2011 Kepler data.
The ‘scaled’ HZ is taken from 0.95 – 1.37 AU in our solar system (according to Kasting 1993) and an earthlike planet is defined as from 0.8 – 2 Re. The lower limit for R (corresponding to a mass of 0.5 Me) may be a bit too pessimistic, but the higher limit already includes a lot of super-earths. If we wish to exclude the real super-earths (beyond about 3 Me) because of dense and different atmospheres and put the upper limit at a more earthlike 1.5 Me, then the fraction of eta Earth becomes even lower, probably closer to 1 than 3%.
The main reason for this rather pessimistic view is that “Interestingly, we find that the density of super-Earth and Neptune planets decreases toward longer periods. This differs markedly with the findings of previous RV surveys of Saturns and Jupiters”.

So, terrestrial planets in these compact systems, even when corrected for observational bias, seem to drop of in abundance toward wider orbits.
This does not bode very for these compact systems, by far the most common type. But we are anxiously looking forward to a newer release of Kepler data.

Eniac November 20, 2012 at 0:18

spaceman:

Of course, in the migration scenario any earth sized orbs in the habitable zone would be knocked out of the system into the interstellar void.

Why could there not be other planets migrating inwards into the habitable zone trailing behind the ones we see, resurrecting the possibility of habitable planets even in the migratory scenario?

Eniac November 20, 2012 at 0:24

In my view, the presence of so many super-Earth’s close to the star where we can see them is a good sign. Likely, they are there because there are just more planets in the system, so there should be lots of them further out where we can’t see them, as well. A much more massive protoplanetary disk, perhaps? Perhaps the sun just has very few planets compared to more “normal” systems.

Ronald November 21, 2012 at 5:28

Eniac: I hope you are right in both your above comments, but a hot or warm inward migrating (spiraling) giant planet is likely to sweep the primordial dust disc empty, not leaving much building material for other planets.

And with regard to more of those super-earths further out, the extrapolations by Catanzarite and Shao, that I quoted above, are not too optimistic:
“Interestingly, we find that the density of super-Earth and Neptune planets decreases toward longer periods”.
But maybe you are right and it means that toward wider orbits the planets become smaller, about earth-sized, that would be nice. Let’s hope so, time will tell.
When is the next batch of Kepler data going to be released?

Comments on this entry are closed.

{ 1 trackback }