≡ Menu

Focus on Alpha Centauri

At Palo Alto’s superb Amber India, I was thinking about Alpha Centauri. There are several Amber India locations in the Bay area, but the Palo Alto restaurant dishes up, among other delights, a spicy scallop appetizer that is searingly hot and brilliantly spiced. Greg and Jim Benford were at the table, Claudio Maccone and my son Miles. It was the night before Breakthrough Discuss convened. And while the topics roamed over many aspects of spaceflight, it was that star system right here in our solar neighborhood that preoccupied me.

How lucky could we be to have not one but two stars this close and so similar to our own? Centauri A is a G-class star, Centauri B a K, and if we hit the jackpot, we could conceivably find planets orbiting both. Then there is Proxima Centauri, an M-dwarf that is the closest star of all to the Solar System. The presence of so many astronomers on the Breakthrough Discuss roster made it clear we’d get the latest on the hunt for planets here, a vital factor as we assessed prospects for the Breakthrough Starshot mission. A nice blue target world would help.

The Binary Star Factor

Growing up, I would haunt the local library for books on astronomy, learning that despite its tantalizing proximity, Alpha Centauri was likely devoid of planets because Centauri A and B were so close to each other. After all, at times they close to within 11 AU — how could planets exist in such a gravitationally unsettled region? The question is still unanswered, as is the question of whether Proxima Centauri is truly part of a triple star system or simply shares a common motion with A and B. But our picture of Alpha Centauri has changed radically since my days in the local library, and the system is under scrutiny as never before.


Image: Gathering outside Stanford’s Arrillaga Alumni Center with coffee in hand as we waited for the first session of Breakthrough Discuss to begin.

Paul Wiegert and Matt Holman discussed stable orbits around Centauri A and B back in 1997, work that led to a generally accepted belief that out to a distance of perhaps 2.5 AU, small planets of Earth-like radius or a bit larger could exist. Giant planets are seemingly ruled out by radial velocity studies, although Jared Males (University of Arizona) would note that we might save James Cameron’s Polyphemus, a gas giant in the film Avatar, by postulating a large radius, low mass planet. But he was quick to add that the prospect was not likely.

The session, titled Exoplanet Detection Programs Focused on Alpha Centauri, was led by Olivier Guyon, but it was Michael Endl (University of Texas at Austin) who presented the overview of Alpha Centauri work so far. The issue of planet formation is far from settled, and as Centauri Dreams readers know, the question is not so much one of stable orbits but whether planet formation can deliver an intact planet in the first place. Key work here has been done by Philippe Thébault and Javiera Guedes, who have reached opposite conclusions, with Guedes arguing for small planet formation, and Thebault arguing against the proposition because of planetesimal encounters and the impact velocities at which they would occur.


Image: One of Michael Endl’s slides, this one discussing planet formation around Alpha Centauri.

The argument will be settled, as Endl pointed out, by our increasingly powerful ability to deploy new technologies. Radial velocity methods are pushing toward the region we’ll need to study, but even now we would need to work at a 10-12 centimeters per second level to find an Earth mass planet, a feat that Endl noted was orders of magnitude below what today’s best instruments can deliver. Remember, too, that the planet we thought we had found around Centauri B probably isn’t there, now considered to be a false positive probably caused by activity on the star itself. In fact, let me quote Xavier Dumusque on this, as he was at the conference and was on the team that performed the original Centauri B work:

We worked with 20,000 spectra on Centauri B taken over four years and found a small 50 cm per second signal that seemed to be a planet in a 3.2 day orbit. Subsequent papers have shown that the signal is in the data, but it is probably not due to a planet. All our techniques are at the limit of their capabilities, which means we should use all the techniques we have, so that if one tells us we have a planet, another can assure us it is real.

The problems Alpha Centauri presents, particularly right now, are manifest. Spectral contamination means that when you’re trying to tease a Doppler signal out of the light from a star like Centauri B, you get light mixing in from Centauri A, for at this point in their orbits, the two stars have closed to their closest point as viewed from Earth. The work Dumusque referred to, drawn from HARPS spectroscopic data at the European Southern Observatory’s La Silla Observatory, may well have been affected by magnetic effects on Centauri B’s surface. But right now we’re in that period when the primary Centauri stars are very hard to analyze.


I’ll remind readers that there has been one possible transit detected. Motivated by the HARPS work on the possible 3.2-day period planet, the search turned up what looked like the transit of an Earth-like planet with a period of less than 12 days, but if it was a transit, it did not recur. As to Proxima Centauri, we still have no planets there, but we can rule out larger worlds, while allowing the possibility of planets of two to three Earth masses in the habitable zone. We’ve followed the work of the Pale Red Dot project that has collected new spectra using the HARPS instrument in these pages and are awaiting the data analysis with great interest.

Image: Natalie Batalha (NASA Ames) raising a point during the panel discussion that followed the Alpha Centauri planet detection talks.

Bringing New Methods to Bear

The point that Michael Endl made, and it was echoed by other speakers, is that we need to throw everything we have at this intractable problem. One way forward is to keep improving radial velocity precision, but we also need to do x-ray monitoring of Centauri A and B to look for activity cycles, and consider the possibilities of astrometry and even direct imaging. Thomas Ayres (University of Colorado) noted the dramatic changes to Alpha Centauri A in x-ray imaging — the star goes dark at x-ray wavelengths in 2005 with an unprecedented darkening by a factor of 50, and is now showing a return to activity levels close to that of our own Sun.

The scientists at Breakthrough Discuss were generally upbeat about the prospects of finding planets in the Alpha Centauri system, though the feeling was not quite unanimous, with Peter Tuthill (University of Sydney) saying he found the likelihood of planets there in the range of 20 percent. Adding “I’ve just put myself out of a job with that comment,” he went on to explain JAM, the JWST Aperture Mask, which would use astrometric methods to look for the tiny stellar motion that a planet tugging either Centauri A or B would induce. A separate mission called TOLIMAN (a medieval name for Alpha Centauri) would use a diffractive pupil aperture mask, with the distortions optical systems produce becoming a ‘ruler’ that detects such motion.


Image: A view across the way from the alumni center during a break. After the intense sessions, it was a pleasure to walk outside for a few minutes to rest my eyes.

And what about observing Alpha Centauri planets from the ground? We’re moving into the era of enormous ground observatories with apertures from 20 meters up to 100 meters across in the works. These Extremely Large Telescopes (ELTs), like the European Extremely Large Telescope (Chile), the Thirty Meter Telescope (Mauna Kea, Hawaii) and the Giant Magellan Telescope (Chile) point toward future instruments as enormous as Colossus, a 100 meter telescope concept that could become the world’s largest optical and infrared instrument.

Needless to say, such instruments can become major tools for studying planetary systems around nearby stars. But as Markus Kasper (European Southern Observatory) explained, we can also perform upgrades on existing instruments — the Very Large Telescope (VLT), Magellan (Chile) and Gemini (sites in Hawaii and Chile) — to perform pathfinder work at thermal infrared wavelengths for future imaging with the giant instruments to come. Thus ground-based instruments become a complement to space telescopes for actual exoplanet imaging.

I was interested in Bruce Macintosh’s presentation on direct imaging from space, because years ago I talked to Webster Cash (University of Colorado) about the prospects of using a starshade, in which the optics for a mission are separated. Rather than using a coronagraph to block out the light of the central star, you create a starshade whose shape is precisely determined to block the same light, with the starshade operating some 25000 kilometers away from the telescope.

Macintosh (Stanford University) said the problem of seeing a planet next to the blazing star that it circles was akin to looking for bioluminescent algae next to a lighthouse, which is why we need a coronagraph or a starshade in the first place. The WFIRST mission (Wide-Field Infrared Survey Telescope), scheduled for launch late in the next decade, will carry an advanced coronagraph, but a starshade would also be compatible with this instrument.


Image: The starshade in position far from the space telescope observing the light, with the central star effectively masked. Credit: University of Colorado.

But maybe we don’t have to wait that long. During a break I had the chance to talk to Cash, who had been a huge help with my original Centauri Dreams book. Cash has been working with starshade concepts for a long time, but even he was surprised when his team began testing small starshades in the atmosphere. In a field of view that included the bright star Sirius, the star would simply disappear. While continuing work on a space telescope/starshade concept called the Aragoscope (after French optical scientist Francois Arago), Cash and team began testing an airborne starshade that could be observed by a telescope on the ground.

All of this could lead to serious results at Alpha Centauri. Cash hopes to use an airborne starshade no more than a meter across that will be observed by a balloon-lofted telescope several hundred kilometers away to probe the habitable zone of Alpha Centauri. “Anything you can do on ground, you should do on ground,” Cash explained. “If we can do it remotely with big telescopes, it’s not a key part of payload that actually goes to Alpha Centauri.”

I’m running out of time today, so I’ll start tomorrow with an Alpha Centauri observing platform called ACEsat, a dedicated space observatory, and move from there into some of the more speculative thoughts of the attendees on what we might find around these stars.


Comments on this entry are closed.

  • Alex Tolley April 19, 2016, 10:46

    Cash hopes to use an airborne starshade no more than a meter across that will be observed by a balloon-lofted telescope several hundred kilometers away to probe the habitable zone of Alpha Centauri.

    That is intriguing. How do they keep both the telescope and star shade aligned and stable enough in the atmosphere, even in the very thin air at observing altitude? Is there any link to an article that explains this?

    • Paul Gilster April 19, 2016, 16:27

      This is as much as I’ve got right now, Alex, but if I can track down background info, I’ll pass along the link.

  • Alex Tolley April 19, 2016, 10:48

    Amber India – arguably the best Indian restaurant[s], if expensive, in the Bay Area. They have a way with spices that blows away other Indian restaurants.

  • DJ Kaplan April 19, 2016, 13:04

    Thank you (and the conference) for bringing the focus back on the primary target!

    The fact that the components of AC orbit one another, and that AC is fairly close, inspires a question: As Alpha Centauri A and Alpha Centauri B orbit one another, would that tend to bring the orbits of their exoplanets around eventually to orientations from which they could be detected using the transit method?

    With most star systems, with a single star and the usual long distance away, we would be stuck with whatever orbital inclination assumed by their exoplanets. Would AC be different? Please let me know if I’ve overlooked something.

  • Ashley Baldwin April 19, 2016, 13:04

    There is an article in arxiv today by Lissauer et al on the long term stability of orbits around the binary constituents of Alpha Centauri, both circumbinary ( beyond 80AU of the common centre of gravity ) and circumstellar , most especially the habitable zones of A and B . These are proven to have stable orbital options out to at least 2 AU with retrograde orbits more stable as well as some eccentric orbits (and surprisingly also some inclined to the plane of the binary as well ). No deal breakers importantly for future imaging telescopes and hopefully habitable planets discovery .

    Accurate positioning is going to be key too with dedicated precision astrometry telescope concepts like NEAT and Theia discovering and positioning any planets with the exsuiteky high levels of precision required for any future flyby mission . An innovative combination of the Gaia and WFIRSTs’s observations should provide high level astrometry meantime out to ten parsecs or so. The key as ever is extreme thermomechanical stability of both telescope and its observing environment .

    One of the specific advantages of the ACEsat dedicated design is in its single target with no need for significant manoeuvring during its two years of observation. It just stares at its single target non stop from an Earth trailing orbit , a fact that Kepler exploited as well to achieve the ultra sensitive transit photometry for Earth mass planet observation.

  • don wilkins April 19, 2016, 14:19

    Re: the terrestrial starshade. Would the balloon stay in position long enough to obtain useful imagery?

    I think a UAV would be better. The UAV would fly away from the telescope. The starshade would collapse as the distance between the UAV and the telescope increases. If the telescope is suitably equipped it might be possible to outfit the UAV with lasers to correct for optical distortions.

  • Michael April 19, 2016, 14:58

    There was mentioned the use of the moons crater rims or building an occultation device to aid a moon based telescope.

  • RobFlores April 19, 2016, 16:28

    I know that current theories posit that A&B centauri system formed
    together. And some think this hinders the possibility of planet formation
    for the A & B stars. But maybe all is not as it seems.

    Could the reality be a bit more complicated and interesting.
    Is it plausible that A & B formed in close proximity but not as close as
    they are now. This would make planetary formation less chaotic and
    it would give a chance for terrestrial planets to form.
    I we assume the above scenario could perturbations by the cloud they were formed in just a few millions years after large proto planets form
    bring A&B to the tighter orbit they currently have around each other.
    Or are we 98% certain the A&B pair were pretty close to current arrangement when formed.

  • Harry R Ray April 19, 2016, 16:34

    I just got through reading the Drew Ex Machina post, “Habitable Planet Reality Check: Kepler 186f Revisited. Andrew LePage is STILL convinced that it is potentially habitable, and maybe the only SOLID candidate we have EVER discovered! What he did NOT include in his post was the NEW paper by Jingjing Chen that claims that “super-earths” really don’t exist AT ALL, and there are only FOUR categories of planet: Subterran, terran, neptunian, and jovian! This paper states that the REAL cutoff between terran and neptunian is at 2 earth masses AND 1.23 earth radii(BOTH are SUFFICIENT, but a grey aria exists as to whether ONLY ONE is NECESSARY). Kepler 186f comes VERY VERY close to the 1.23 Re cutoff point, at 1,17Re. I would like his thoughts on this, as to whether the 1.23 Re “tipping point” starts a GRADUAL shift to neptunianism, gathering steam through 1.5Re, and then becoming exponential(AND IRREVERSABLE) at the MORE COMMONLY ACCEPTED VALUE of 1.6Re! NOW: How does this apply to the Alpha Centauri triple star system? I propose an ENTIRELY NEW CLASS OF PLANET: Cold Venus, which could be the ONLY PERMANENTLY habitable kind of planet around a star like Proxima Centauri(AND TRAPPIST-1, Which I have commented on A LOT). Since Kepler 186f IS so close to the NEW Cheng cut-off point, the atmospheric pressure at the surface could very well be EQUIVALENT to VENUS’S! The starlight at the surface would be very similar to the sunlight at the surface of Titan(i.e. NOT VERY MUCH AT ALL)! Archea would flourish, but NO photosynghtsis would be IMPOSSIBLE! If you were to MAGICALLY transfer Kepler 186f to a stellar flux EQUIVALENT orbit around Proxima Centauri, even THIS thick(relatively, of course) atmousphere would probably boil away in 1 to 2 billion years. You would probably need to START with a planet like Kepler 432b to ensure that after 4 or so billion years you wouls STILL have a VIABLE atmosphere.

    • Andrew LePage April 20, 2016, 9:21

      > Habitable Planet Reality Check: Kepler 186f Revisited – Latest on potentially habitable exoplanet discovered 2 years ago

      Well, this discussion is a bit off-topic since it has nothing to do with Alpha Centauri but since you brought it up, I *DID* include a mention of the new paper by Chen & Kipping’s, “Probabilistic Forecasting of the Masses and Radii of Other Worlds”, because they specifically *DID* cite Kepler 186f as an example in their paper. In my new article I stated:

      “A more recent analysis of the mass-radius relationship with a much larger collection of exoplanetary data by Jingjing Chen and David Kipping (Columbia University) suggests that Kepler 186f has a mass of 1.74 +1.31/-0.60 ME, roughly in keeping with an Earth-like composition. Based on a statistical analysis of known exoplanetary radii and masses as well as the current uncertainties in the properties of Kepler 186f, they estimate that Kepler 186f has a 59% probability of being a rocky planet like the Earth.”

      Here is a link to my new article on Kepler 186f for those who my want to read it:


      > I would like his thoughts on this, as to whether the 1.23 Re “tipping point” starts a GRADUAL shift to neptunianism, gathering steam through 1.5Re, and then becoming exponential(AND IRREVERSABLE) at the MORE COMMONLY ACCEPTED VALUE of 1.6Re

      The analysis by Chen & Kipping and their resulting model explicitly includes the prospect that the transition from rocky to volatile rich worlds is a gradual one (in fact, the empirical evidence seems to suggest just such a gradual transition although statistically the data are too spare and poor to characterize this transition with any real precision). The value of 1.6 Re you cite comes I assume comes from the work of Leslie Rogers but it is not a “hard” value. Based on the data she had when she wrote her paper, she clearly stated that the transition seems to occur at 1.5 Re and, when taking into account the large measurement uncertainties in her limited data set, that there was a 95% chance that the transition from rocky to volatile-rich world occurs at a value of no greater than 1.6 Re. We still have a way to go before we have enough data with sufficient accuracy to get a good handle on this transition and its nature.

  • Michael April 20, 2016, 3:27

    I wonder if there are still any geostationary satellites out there that could act as starshades, it would be nice to reuse old technologies.

  • Michael Spencer April 20, 2016, 6:56

    Disappointing not to see more women in the mix.

    • Andrew Palfreyman April 20, 2016, 14:34

      It seems that the generational starship concept provides the biggest attraction for women in this domain of interstellar dreaming.

  • Harry R Ray April 20, 2016, 9:16

    Sorry, I meant “…but NO photosynthesis would be POSSIBLE…”.

  • spaceman April 20, 2016, 12:13

    Hi Paul,

    Would it be fair to say that there is more agreement in the astronomical community about the possibility of planets remaining stable once they form than there is about planets forming in the first place around either alpha centauri A or B?

    Based on your examination of the evidence, do you think either of these two stars have terrestrial planets currently orbiting within 2 AU? Is there anything other than theory and/or simulations to guide us to a preliminary answer as to the likelihood of planets existing in the centauri system? The kind of observational clues I have in mind would be multiple system proto-planetary disks at various stages of growth and other close separation binary systems with known planets…

    • Paul Gilster April 20, 2016, 12:56

      Yes, I’d say the idea that there are stable orbits around the Centauri stars is pretty widely accepted, whereas there is still some doubts about the formation of planets in the first place, though most astronomers I talked to in Palo Alto are optimistic about finding them there. I don’t think we have the kind of observational clues you’re looking for re planets in the habitable zone. Not yet, but it shouldn’t be long.

  • DJ Kaplan April 20, 2016, 12:49

    Lisa Kaltenegger’s remark brought up an interesting image. In stereotype sci-fi, we have astronauts step out onto foreign planets, remove their helmets, inhale and say “oh good, there’s oxygen here. We can live here”. Haha, the joke’s on you; the ET air-borne viruses will get you, in ways that you cannot predict. Can you say ET allergies?

    Yes, checking out Tabby’s star using the sun’s gravitational lens could answer some nagging questions.

    • Alex Tolley April 20, 2016, 13:03

      Only if the life forms use the exact same DNA and RNA chemistry, structure and coding. Even the protein coats have to work to penetrate our cells. It is just as likely that we might be shadow biologies to each other, with minimal interaction. Far more likely a multicellular predator beastie will try to eat you, than that the micro-organisms can do you any damage.

      Our colonialist astronauts might well have to bring a complete terrestrial biosphere with them (one value of a world ship) to successfully colonize a living alien world. Of course the moment they land, the galactic cops will write them a citation for interfering with an living world and that they must appear before the local court for a hearing. Failure to do so will result in… ;)

      • Eniac April 22, 2016, 19:15

        I am not totally convinced alien biology would be totally harmless. Alien macromolecules could be immunogenic, for example, very likely so in my opinion. That predator beastie would get a severe case of indigestion, at least, from eating the intrepid astronaut, as would astronauts from sampling the local flora and fauna.

        While viruses would obviously not work, some alien autotrophs might find the water and certain other small molecules in our bloodstream to be a good growth medium. Some of the many mechanisms by which our immune system keeps such freeloaders at bay could work with aliens, such as clumping together by antibody cross-linking and engulfment by phagocytes followed by lysis in acid. Others might not work, and that could be a real problem.

  • Robert G. Clark April 20, 2016, 17:07

    I’ve been thinking of ways we can get such nanocraft to link up through
    self-assembly and form larger structures that can do more detailed
    observations and experiments. This could work even for visits to far off
    destinations still in the Solar System such as Kuiper belt objects like
    Pluto or the Oort cloud.

    The main problem is getting the many objects flying independently and
    getting further apart the further out they go to gradually be drawn to each
    other and link up. Once they link up, I don’t it would be to difficult to
    then get them to do self-assembly.

    But it’s that drawing together step that is the sticking point.

    Bob Clark

  • Ronald April 21, 2016, 8:14

    Harry, very interesting paper you quote and I will definitely check it out.
    And Andrew LePage’s already wrote his reply to you (I wrote a similar comment last night but now that is redundant).

    My only remaining additional comment here: one could argue that there are only 3, or even only 2, natural planetary categories: terrestrials and gas planets (maybe there is also a real distinction and a gap between Neptunes and gas giants, justifying a 3rd category).

    I say this, because the only essential *natural* factor in planetary categorization seems to be the presence/absence of a massive H/He envelope, which in turn largely depends on a threshold planetary mass. All other categorization now seems to be artificial and arbitrary.
    So, there is probably a continuum from gas dwarfs and mini Neptunes, via Neptunes to gas giants.

    This should also be visible in the data: a gradual continuum in mass/radius for the sub-terrestrials and terrestrials, then a sudden transition, then again a gradual continuum in mass/radius for the gas planets.
    All this may also show up in planetary abundance data, i.e. prevalence of planets per radius and mass category.

  • andy April 21, 2016, 17:33

    A recent paper on the arXiv today on the influence of close binary star companions on planet occurrence rates: Kraus et al. “The Impact of Stellar Multiplicity on Planetary Systems, I.: The Ruinous Influence of Close Binary Companions

    There appears to be a cut-off for binary semimajor axes at 49 +59/-23 AU (are subscripts/superscripts allowed in the HTML comment markup here?) below which planet occurrence drops to about a third of that found in wider binaries. A planetless Alpha Centauri system would be a highly disappointing result, but does not appear to be totally outside the realms of possibility.