Why is it so difficult to detect planets around Alpha Centauri? Proxima Centauri is one thing; we’ve found interesting worlds there, though this small, dim star has been a tough target, examined through decades of steadily improving equipment. But Centauri A and B, the G-class and K-class central binary here, have proven impenetrable. Given that we’ve found over 4500 planets around other stars, why the problem here?
Proximity turns out to be a challenge in itself. Centauri A and B are in an orbit around a common barycenter, angled such that the light from one will contaminate the search around the other. It’s a 79-year orbit, with the distance between A and B varying from 35.6 AU to 11.2. You can think of them as, at their furthest, separated by the Sun’s distance from Pluto (roughly), and at their closest, by about the distance to Saturn.
The good news is that we have a window from 2022 to 2035 in which, even as our observing tools continue to improve, the parameters of that orbit as seen from Earth will separate Centauri A and B enough to allow astronomers to overcome light contamination. I think we can be quite optimistic about what we’ll find within the decade, assuming there are indeed planets here. I suspect we will find planets around each, but whether we find something in the habitable zone is anyone’s guess.
Image: This is Figure 1 from today’s paper. Caption: (a) Trajectories of ?-Cen A (red) and B (blue) around their barycenter (cross). The two stars are positioned at their approximate present-day separation. The Hill spheres (dashed circles) and HZs (nested green circles) of A and B are drawn to scale at periapsis. (b) The apparent trajectory of B centered on A, with indications of their apparent separation on the sky over the period from CE 2020 to 2050. The part of trajectory in yellow indicates the coming observational window (CE 2022–2035) when the apparent separation between A and B is larger than 6 and the search for planets around A or B can be conducted without suffering significant contamination from the respective companion star. Credit: Wang et al.
If we don’t yet have a planet detection around the binary Centauri stars, we continue to explore the possibilities even as the search continues. Thus a new paper from Haiyang Wang (ETH Zurich), who along with colleagues at the university has been modeling the kind of rocky planet in the habitable zone that we hope to find there. The idea is to create the benchmarks that predict what this world should look like.
The numerical modeling involved examines the composition of the hypothetical world, drawing on what we do know, based on spectroscopic measurements, of the chemical composition of Centauri A and B. Here there is a great deal of information to work with, especially on so-called refractory elements, the iron, magnesium and silicon that go into rock formation. Centauri A and B are among the Gaia “benchmark stars” for which stellar properties have been carefully calibrated, and up to 22 elements have been found in high-quality spectra, so we know a lot about their chemical makeup.
But a key issue remains. While rocky planets are known to have rock and metal chemical compositions similar to that of their host stars, there is no necessary correspondence when it comes to the readily vaporized volatile elements. The authors suggest that this is because the process of planetary formation and evolution quickly does away with key telltale volatiles.
The researchers thus develop their own ‘devolatilization model’ to project the possible composition of a supposed habitable zone planet around Centauri A and B, linking stellar composition with both volatile and refractory elements. The model grew out of Wang’s work with Charley Lineweaver and Trevor Ireland at the Australian National University in Canberra, and it continues at Wang’s current venue at ETH. This is fundamentally new ground that extends our notions of exoplanet composition.
Wang and team call their imagined world ‘a-Cen-Earth,’ delving into its internal structure, mineralogy and atmospheric composition, all factors in evolution and habitability. The findings reveal a planet that is geochemically similar to Earth, with a silicate mantle, although carbon-bearing species like graphite and diamond are enhanced. Water storage in the interior is roughly the same as Earth, but the deduced world has a somewhat larger iron core mixed with a possible lack of plate tectonics. Indeed, “…the planet may be in a Venus-like stagnant-lid regime, with sluggish mantle convection and planetary resurfacing, over most of its geological history.”
As to the atmosphere of the hypothetical world that grows out of Wang’s model, its early era shows an envelope rich in carbon dioxide, methane and water, which harks back to the Earth’s atmosphere in the Archean era, between 4 and 2.5 billion years ago. That gives life a promising start if we assume abiogenesis occurring in a similar environment.
Image: ? Centauri A (left) and ? Centauri B viewed by the Hubble Space Telescope. At a distance of 4.3 light-?years, the ? Centauri group (which includes also the red dwarf ? Centauri C) is the nearest star system to Earth. Credit: ESA/Hubble & NASA.
How far can we take a model like this? We may soon have data to measure it against, but it’s worth remembering what the paper’s authors point out. After noting that planets around the “Sun-like” Centauri A and B cannot be extrapolated from the already known planets around the red dwarf Proxima Centauri, they go on to say:
Second, although ? Cen A and B are “Sun-like” stars, their metallicities are ?72% higher than the solar metallicity (Figure 3). How this difference would affect the condensation/evaporation process, and thus the devolatilization scale, is the subject of ongoing work (Wang et al. 2020b).
That’s a big caveat and a useful pointer to the needed clarification that further work on the matter should bring – metallicity is obviously significant. The paper adds:
Third, we ignore any potential effect of the “binarity” of the stars on their surrounding planetary bulk chemistry during planet formation, even though we highlight that, dynamically, the planetary orbits in the HZ around either companion are stable. Finally, we have yet to explore a larger parameter space, e.g., in mass and radius, but have only benchmarked our analysis with an Earth-sized planet, which would otherwise have an impact on the interior modeling…
So we’re in early days with planet modeling using these methods, which are being examined and extended through the team’s collaborations at Switzerland’s National Centre of Competence in Research PlanetS. Note too that the authors do not inject any catastrophic impact into their model of the sort that could affect both a planet’s mantle and/or its atmosphere, with dramatic consequences for the outcome. We know from the Earth’s experience in the Late Heavy Bombardment that this can be a factor.
With all this in mind, it’s fascinating to see the lines of observation and theory converging on the Alpha Centauri binary pair. Finding a habitable zone planet around Proxima Centauri was exhilarating. How much more so to go beyond the many imponderables of red dwarf planet habitability to two stars much more like our Sun, each of which might have a planet in its habitable zone? The Alpha Centauri triple system may turn out to be a bonanza, showing us both red dwarf and Sun-like planetary outcomes in a single system that just happens to be the closest to us.
The paper is Wang et al,, “A Model Earth-sized Planet in the Habitable Zone of ? Centauri A/B,” The Astrophysical Journal Vol. 927, No. 2 (10 March 2022). Abstract/Full Text. Preprint also available.
Given how close these stars are, what is the expected interaction of the nebular disk during planet formations? Does it create chaotic interactions disturbing planet formation, possibly enough to disrupt planet formation entirely? Are outer planets inevitably absent, or is it possible some sort of resonance or orbit swapping is possible?
[I am sure I have seen/read some modeling about planet formation in close binaries, but perhaps someone with the requisite knowledge can answer those questions?]
Will the JWST be able to take advatage of the seperation in the next 13 years?
What is used to create tremendous pressure, the diamond anvil, here on earth could cause a diamond crystal inner layer that might create a very unusual core, maybe ringing like a bell. Super earths with diamonds and carbon graphene aerogel spewing from their volcanoes. Graphane carpets with weird electrical interplays from Alpha Centauri’s solar storms. Maybe we can catch a reflection off those beautiful diamond mountains or it could be black as soot… :-(
Will there be planets in the Alpha-Cen A-B system?
Well, it is not impossible as several planets have been found in close binary systems and several are candidate objects of interest. Further, a couple have been found orbiting the common-centre of gravity of the system they are in, not orbiting any star.
Generally, close binary systems are not , or historically have not been, checked for planets due to the difficulties encountered in studying these systems, and many have felt it unfruitful to search, so the odds of planets in the A-Cen A-B system is unknown as we simply lack enough data to say it’s a 50-50 chance, a 90% chance against or any other guesstimate.
The system is close, which also makes it difficult, but if we can crack the problems here this will undoubtedly assist the search around other binary and more complex systems.
Nature abhors a vacuum, and so it planets are able to form from protoplanetary discs, then it is safe to assume they will form where other forces do not prevent it.
Very good article on planets in binary systems with high scale charts and many links to other sites and articles.
Planets in binary systems.
http://exoplanet.eu/planets_binary/
Very good chart. What is clear, however, is that there are very few stars within 10 AU of their companion. Whether this is due to the limits of observation, like Alpha Cen A&B or something else, it does indicate that planets do form around one star of a close binary and answers my question.
Actualy you should be looking at Gliese 86, which is at about 23 AU distance apart, the average distance of Alpha Centauri A & B. Remember the further away the star systems like our Sun (G-K) the closer the stars are viewed from earth and the harder to see the planets. Plus the brightness is a bigger problem, Keplar 444 shows 5 planets because the second star is a M dwarf. Most of the systems below 35 AU average distance apart are showing only one large planet.
I don’t understand the logic of averaging the stellar distances between the binary stars.
1. As the orbital distances become more extreme, any planet that orbits one star will eventually be disrupted by the other star. It seems that the minimum distance is the more important measurement.
– source
2. Very close binaries will have planets that orbit both stars (circumbinary) as if a single star, rather than around one star. This is a very different situation to non-circumbinary planets.
Depending on star types, for stars in a binary system, there must be a point where decreasing distances result in a transition of planets from non-circumbinary to circumbinary. I would expect that the planets become increasingly restricted to close orbits of their star until the planetary nebula has both stars at the center and the planets form as circumbinary only.
In the table of stars, can you identify which binaries with low separation whether the planets are circumbinary or not?For Alpha Cen A&B, I want to know whether there is any data that shows whether there are non-circumbinary planets where the stars have their closest separation around 10AU or a scaled euivalent for low mass stars. [I have no doubt that astronomers looking for planets in the system have already taken this in account on theoretical grounds. But is there any existing data to support this?]
I think it is possible for a planet to have a stable orbit around ? Centauri A and ? Centauri B if it is close enough to the star, but I think it would have to be a captured rogue planet and not formed in the these two stars protoplanetary disk because they are at the right distance apart to for a ring type gas cloud star system beginning instead and a protoplanetary disk type. With a ring type of beginning, there is no gas and dust the center, but only a ring with two stars on both ends and therefore not any gas and dust to make any planets. The reason for this is the two stars grab all the angular momentum and not any planets at all can form. The two stars are at opposite ends of the ring. Kippenhahn 1983.
A planet that formed around one of these stars would certainly have to be close enough to be tidally locked which might have an effect on any plate tectonics. I agree with the Venus type planet. It possibly might make continental drift impossible since a tidally locked planet can’t have a Moon due to the hill sphere and there would be no fast rotation of the planet and equal distribution of tidal forces from a Moon over the surface.
I do think it is important to look at the atmospheric modeling, chemical composition and geology of a planet around these stars in case the ring theory is wrong or needs some modification. Time will tell, but I would be surprise if the A B Centauri system had planets. I would look close to the star.
Metallicities are higher than our own sun. This alone suggests there are planets there. They could be nice dense planets as well. A planet with a thick crust that does not allow for plate tectonics would make it more Venus-like than Earth-like.
I like the idea that a planet might be in the A B Centauri double star system because at times throughout it’s orbit there would be no night on that planet, yet it’s night would be the same as the distant star which would still give the cooler temperature or Moon light equivalent. At other times there would be a dark night and both stars would appear in the day side.
The Metallicity of the star never takes precedence over general relativity which is the most dominant principle controller of star birth and how all star systems form including double star systems.
First, want to say that new reports ( and perspectives!) on the situation at Alpha Centauri are always welcome.
RE: “binarity”
“Finally, we have yet to explore a larger parameter space, e.g., in mass and radius, but have only benchmarked our analysis with an Earth-sized planet, which would otherwise have an impact on the interior modeling…”
Every once in a while I get a chance to interject on this. Some data on this of my own. Placing an Earth at A at 1.246 AU, the line of apsides rotated in space in an 8,000 terrestrial year cycle with an eccentricity variation of about nil to 0.07 correspondingly. Now granted there are a lot of issues of formation in close Centauran quarters. But the perturbing effect of a K star with e = 0.5 and 80 year period – that’s not going to give you “Venus like stagnation”. The perturbing body is farther than Earth is from Venus, but it’s over 200,000 x as massive.
In the other instance (circum B), since the K star is less luminous, the
Earth analog at nominal HZ position would be more tightly bound. At least judging from the eccentricity variation and line of apsides cycling.
In recent papers about Alpha Cen, I have seen variations in age estimates: Sometimes 6 billion years; others saying roughly the same as our own system. Either way, have to wonder how much variation there has been in Alpha Cen orbital elements as well. If a circum-stellar disk or two could be a significant fraction of a solar mass, then could there have been some evolution of binary elements as well? For example: if each star had a circum stellar disk of 0.1 mass and it dispersed leaving only a few Earths of planetary elements, then I would suspect that the two stars were closer together 4 to 6 billion years ago.
Is there a possibility that planets have been flung out into wide orbits? Proxima is presumably coeval with A and B, and it has been flung out into a very wide orbit. So could the same have happened with gas giants that formed at the same time?
To me, this system of stars is simply the most exciting prospect for scrutiny of the highest fidelity. Close enough that even though currently we couldn’t hope to visit, programs such as Breakthrough Starshot would surely be re-vitalised. Should an Earth-like planet be detected in a stable orbit about one of the stars (A or B) and feature Earth-like environmental parameters, imagine the push to learn as much as possible, perhaps leading to the development of a space-bound optical interferometer or Super Starshot program, what ever that could be. The world could certainly do with such a discovery right now.
The influence of Mars in the 19th and early 20th centuries on space exploration cannot be underestimated. But this was also when it was thought to be habitable or at least partially so. It inspired many rocket pioneers and much science fiction. If we had another such world nearby that was more like the Mars that was, it might create another surge in imagination that could lead to a golden age in interstellar exploration.
I am currently busy working a cloud physics contract but would like to educate myself on the Alpha Centauri system. Anyone here have links to in-depth information/papers including perhaps mathematical treatment of the dynamics of the Centauri A/B system?
David, I’d start with Wiegert and Holman, who first defined stable orbits in the A/B system, and go from there — they’re cited by everyone looking at Alpha Centauri dynamics, so you can move forward to papers by, for example, Greg Laughlin (Guedes JM, Rivera EJ, Davis E, Laughlin G, Quintana EV, Fischer DA. “Formation and detectability of terrestrial planets around ? Centauri B” Astrophysical Journal. 679: 1582-1587. DOI: 10.1086/587799)
The Wiegert and Holman paper is:
Long-Term Stability of Planets in Binary Systems
https://ui.adsabs.harvard.edu/abs/1999AJ….117..621H/abstract
You’ll find lots of references to them in the archives here, which also link to later work on Centauri. For example, Quarles and Lissauer:
Maximizing planet packing in the alpha Centauri AB system
https://ui.adsabs.harvard.edu/abs/2017DDA….4810205Q/abstract
and Jones and Fabrycky, Dynamics of the Triple-Star System Alpha Centauri and its Impact on Habitable Planets
https://ui.adsabs.harvard.edu/abs/2018AAS…23143925J/abstract
Thank you so much Paul. Looking forward to following these links. It is no wonder that you named this site after the Alpha Centauri system.
Yes. The Centauri system has been a preoccupation of mine since childhood.
Any progress report on PROJECT BLUE.
“Project Blue is a science initiative to capture the first photograph of a potential Earth-like planet orbiting another Sun-like star. The mission aims to launch a lightweight space telescope to directly image exoplanets around Earth’s nearest star system, Alpha Centauri. With a budget the fraction of the cost of a mid-size astrophysics mission, and a planned launch by early next decade, this venture represents an ambitious leap forward in low-cost, high-impact space exploration. Through active collaboration between research institutions, universities, private industry and citizens, Project Blue seeks to make space exploration a participatory, collective endeavor and inspire millions worldwide to engage in scientific inquiry.”
https://www.boldlygo.org/project-blue-mission-brief
There has been no updates on how the project is going since 2019, has everyone lost interest, no money or some problem like Covid???
Well, there aren’t any “advanced” civilizations at Centauri, we would have heard them.
We can certainly hope for some incredible luck in finding a planet. A habitable planet would be tantalizing, even if it wouldn’t be habitable for us.
The holy grail is to find another earth, but after Kepler and with TESS well into it’s surveys, NOT ONE earth has been documented.
How rare both earths and life are proving to be. When I was young, it was confidently assumed they would be abundant.
There are a number of implicit biases in your comment Paul. First of all we haven’t really got the sensitivity yet to pick up a lot of Earth sized rocky planets in the habitable zones of many stars. Our planet detection capabilities are still strongly biased towards bigger planets. Also many planets including Earth sized ones are not detectable by the transit method because from our point of view these planets do not pass in front of their host stars. Finally, who says we need an Earth twin to find life, including intelligent species? We have decades if not more to go before we have a decent catalogue of planets in the habitable zones of stars both nearby and far off.
Well, I agree of course about hot jupiters being the low hanging fruit. They are spottable with our primitive equipment.
I would be much more optimistic if it weren’t for the fact that “They” should have found us.
When scientists first listened via radio to the cosmos, they expected to immediately hear civilizations chattering everywhere. Instead total silence. Star Trek envisioned a galaxy swarming with other intelligent creatures. Now the stalwarts hope for perhaps 3 to 5 in the entire galaxy.
Of all the explanations for the Fermi Paradox, I have moved toward the “Rare Earth” theory. Earth as a miraculous confluence of rare, exceedingly rare factors.
I’m all for finding ANY life of any kind! But if we do, it means “life” is plentiful, but the fact that the galaxy is full of life, but not intelligent civilizations is a big problem.
So, we might as well look for “Earths”. We’re alone out here for reasons we can’t fathom.
I should add that I have a huge positive bias towards Earth too :).
Should we explore the Alpha Centauri system first, just because it is the closest to us?
https://bigthink.com/starts-with-a-bang/first-interstellar-target/