≡ Menu

Axial Tilt, Habitability, and Centauri B

Our fascination with Alpha Centauri doubtless propels at least some of the recent interest in binary star systems, as we ponder the chances for habitable worlds around the nearest stars. But given that the population of binary or multiple star systems in our galaxy is as high as it is (multiple systems are common, and about 50 percent of stars have binary companions), determining the factors that influence habitability in this environment has much broader significance. A new study out of the Georgia Institute of Technology has been looking at the issue by modeling an Earth twin in various binary scenarios.

So how does Alpha Centauri fare? We can find habitable zones in the Centauri A/B system, and into these the researchers introduced a simulated Earth around Centauri B to examine its axis dynamics. They also investigate the dynamical evolution of planets within the habitable zone of either star, generalizing from these results to the larger binary star population. The issue to be addressed: Does the axial tilt, or obliquity, of our planet, apparently a key feature in maintaining a habitable climate, survive the transition into a binary system with the features of Alpha Centauri? The answer is not encouraging for life, at least not around Centauri B.

Image: The Alpha Centauri group is the closest star system outside of our own at a distance of 4.3 lightyears; it can be found in the night sky in the constellation Centaurus. The stars Alpha Centauri A and Alpha Centauri B comprise a binary system, in which the two stars orbit one another, and close by is an additional faint red dwarf, Alpha Centauri C, also called Proxima Centauri. Some astronomers have hoped to someday find an exoplanet capable of harboring advanced life in the system, but a new study lowers those expectations while raising them for the rest of the universe. Credit: NASA/ESA Hubble Space Telescope.

Remember that Centauri B, a K-class dwarf, and the G-class Centauri A orbit each other tightly, with an orbital period of 79.91 years. We’re dealing with an elliptical orbit that varies the distance between the two stars anywhere from 35 AU to 11 AU, the latter being not much further than Saturn is from the Sun in our own system. From the perspective of a planet orbiting Centauri B, Centauri A swings relatively close and then backs away during the 80-year period, causing the simulated Earth to vary considerably in terms of obliquity. The Moon is often cited as a help in stabilizing Earth’s axial tilt, but that doesn’t help here, says lead author Billy Quarles.

“Around Alpha Centauri B, if you don’t have a moon, you have a more stable axis than if you do have a moon. If you have a moon, it’s pretty much bad news.”

From the paper:

…the added spin precession from a Luna-like moon increases the obliquity variation of an Earth-like planet orbiting α Cen B, but a different type of moon, in terms of its mass or semimajor axis, may have a negligible or more beneficial effect.

Image: Astrophysicist Billy Quarles, author of a new study on exoplanet axis tilt, stands with Georgia Tech’s largest telescope housed at its observatory. Credit: Georgia Tech / Rob Felt.

And without a moon? Quarles again:

“The biggest effect you would see is differences in the climate cycles related to how elongated the orbit is. Instead of having ice ages every 100,000 years like on Earth, they may come every 1 million years, be worse, and last much longer.”

The separation of the two stars is the key here. According to this work, Centauri A and B are simply too close for comfort, but a wider separation, which holds in most binary systems, would allow the second star’s effects to be less disruptive to the simulated Earth. I was curious about the effects of a moon, though, and went further into the paper, which notes that the disruptive effects of a moon on a simulated Earth around Centauri B are:

…in contrast to our own Earth-based expectations, where our moon does the opposite (Laskar et al. 1993b). The amount of spin precession from a moon depends on the moon’s mass and semimajor axis, where a Pluto-mass moon at a Luna-like semimajor axis aluna would have a negligible effect. A smaller semimajor axis (0.2 aluna) would allow a Pluto-mass moon to increase the spin precession and allow for larger obliquity variations. The degree to which a moon can increase the overall spin precession depends on many factors, where they are neither needed nor necessarily even desirable to obtain relatively low obliquity variations.

Image: Modeled into an orbit in the habitable zone around Alpha Centauri B, in this artist’s rendering by an author of a new study, our planet appears rather icy and inhospitable to advanced life. Credit: Georgia Tech / Billy Quarles.

Quarles points to Mars as an example of obliquity extremes influencing climate. The axial tilt of Mars varies between 10 and 60 degrees every 2 million years, as opposed to Earth’s axial tilt (between 22.1 and 24.5 degrees over a course of 41,000 years). Earth’s Moon stabilizes our planet’s obliquity, which would otherwise be affected by gravitational influences from the inner planets as well as Mars and Jupiter.

If the precession of Mars’ axis seems to have helped deplete its atmosphere, imagine an Earth precessing the same 60 degrees, which is the figure Quarles deduces for an Earth without its Moon. Clearly, the presence of a moon can have widely varying effects depending on the stellar masses and orbital parameters involved. We learn that the presence of a single large moon is just one factor in questions of habitability, and its effects are not always benign.

But the broader outlook for habitable zone planets in binary star systems seems encouraging. According to the study, a high percentage of such systems can support exo-Earths with axial tilts similarly steady to Earth’s, thus ensuring climate stability. Back to the paper:

We combine our results with population studies of binary stars (Raghavan et al. 2010; Moe & Di Stefano 2017) and find the chance that an Earth-like rotator orbiting the primary star would experience small (< 2.4°) obliquity variations is 87%, 74%, or 54%, depending on the mass of the primary (0.8, 1.0, or 1.2 M⊙ Solar-Type stars, respectively).

All told, those aren’t bad numbers given the number of multiple star systems throughout the Milky Way. The paper shows us that the evolution in axial tilt for a single Earth-like planet in the habitable zone of a binary system depends on the orbital precession induced by the companion star, while also being influenced by orbital distance, nearby terrestrial planets and moons. It’s interesting, too, to see that many disks are misaligned by around 10 percent from the binary plane, which would produce a typical obliquity variation in the neighborhood of 20%. Clearly, orbiting nearly planar with the binary would be the most desirable place to be.

It’s worth noting that the Transiting Exoplanet Survey Satellite (TESS) is expected to uncover ∼500,000 eclipsing binary systems. We’ll have no shortage of candidates for further study. Both TESS and Gaia observations, the authors say, would allow a more robust statistical approach to emerge as we examine the obliquity of planets around Sun-like stellar binaries.

The paper is Quarles et al., “Obliquity Evolution of Circumstellar Planets in Sun-like Stellar Binaries,” Astrophysical Journal Vol. 886, No. 1 (19 November 2019). Abstract / Preprint.


Comments on this entry are closed.

  • Geoffrey Hillend November 21, 2019, 16:58

    If one’s thought experiment about Alpha Centauri begins with an Earth Moon twin already around it, then one has to worry about it affecting the wobble of the exoplanet’s axes. I don’t think that there can be an Earth moon twin in Alpha Centauri. Two stars that close together results is an early planetary evolution much different than our solar system. What happens to the early and late bombardment periods is that they are reduced or split apart into two different systems and made smaller which changes or reduces the probability of collisions. The evolution of the rocky inner planet especially an Earth twin with moon are contingent upon those periods.

    The ring star model of the evolution of two stars keeps all of the angular momentum from the beginning making it impossible for planets for form in such systems which includes Alpha Centauri, Kippenhahn, 1983, 100 Billion Suns. He writes that half of the stars in the galaxy are binary star systems, so only half of them have planets. I think he referring to the ring model and I don’t think all the binary star systems formed like ring systems. Kepler 47 has exoplanets, and these stars are very close together and are inside the orbit of the closest planet, so that all of the planets in Kepler 47 orbit around both stars. I think Kippenhahn is basing his principle on how binary star systems form. I don’t know whether it is correct. I would be surprised if it was. Stars in binary systems that are further apart than Alpha Centauri would be included in the ring formation model.

    • Charlie November 21, 2019, 23:19

      Axial tilt
      Long term
      Using numerical methods to simulate Solar System behavior, long-term changes in Earth’s orbit, and hence its obliquity, have been investigated over a period of several million years. For the past 5 million years, Earth’s obliquity has varied between 22° 2′ 33″ and 24° 30′ 16″, with a mean period of 41,040 years. This cycle is a combination of precession and the largest term in the motion of the ecliptic. For the next 1 million years, the cycle will carry the obliquity between 22° 13′ 44″ and 24° 20′ 50″.[26]

      The Moon has a stabilizing effect on Earth’s obliquity. Frequency map analysis conducted in 1993 suggested that, in the absence of the Moon, the obliquity can change rapidly due to orbital resonances and chaotic behavior of the Solar System, reaching as high as 90° in as little as a few million years (also see Orbit of the Moon).[27][28] However, more recent numerical simulations[29] made in 2011 indicated that even in the absence of the Moon, Earth’s obliquity might not be quite so unstable; varying only by about 20–25°. To resolve this contradiction, diffusion rate of obliquity has been calculated, and it was found that it takes more than billions of years for Earth’s obliquity to reach near 90°.[30] The Moon’s stabilizing effect will continue for less than 2 billion years. As the Moon continues to recede from Earth due to tidal acceleration, resonances may occur which will cause large oscillations of the obliquity.[31]

  • Robin Datta November 21, 2019, 17:13

    This is a thoughtful discussion of yet another set of criteria to be taken into consideration in the quest for “habitability”. Of course that’s habitability by our human standards, and this suggests yet another constraint to finding an intelligent entity with some biological resemblance to earth life.

    And this constraint would operate at the basal level, for our kind of life, with a continuum of habitability, biogenesis (including panspermia), complex life, intelligent life and possibly post-biologic intelligence.

  • Andrew LePage November 21, 2019, 18:09

    For some background on the search for exoplanets orbiting Alpha Centauri A and B (at least up to a couple of years ago), check out the following:


    Hopefully some of the newer programs coming on line will finally be able to spot the illusive planets of this system.

  • Ashley Baldwin November 21, 2019, 19:36

    In terms of direct imaging, an erstwhile WFIRST star-shade rendezvous mission and/or HabEx will both be capable of targeting terrestrial circumstellar planets around the much larger separation ( but still within 5 parsecs of Earth) 61 Cygni binary constituents and the 40 Eridani Primary A
    (Keid) from its binary neighbours B&C. These stars are hundreds of AU apart even at periapsis.

  • Alex Tolley November 21, 2019, 20:24

    I have to wonder what an HZ zone looks like around such stars. For close binaries, the HZ would be highly distorted, possibly not even a complete flat torus.

    Whether obliquity of the planet is unstable, unless the 2nd star is very distant from teh primary, any small change in solar flux is going to have major climatic consequences. Marine life might have the most stable ecosystems, but surface life is likely to be subject to large average temperature ranges over the planet’s orbital period, with longer climate changes over the period of the orbit of the stars about each other. Any life on these worlds would have to be highly adapted to these temperature and climatic ranges. If the HZ is “broken”/disjoint in parts of the orbit, then the planet would not be inhabitable, either freezing or becoming a dry, hot Venus analog.

    The interesting planets would be those where the solar flux varied significantly, but a continuous HZ was maintained, allowing life to evolve to handle the extreme conditions. Adaptations might include mechanisms similar to hibernation or estivation for all the animals in the biosphere, with other adaptations for the plants to endure the heat or cold conditions with dormant spores or seeds.

    • Robin Datta November 21, 2019, 21:42

      Yes sir, they’d be cookin’! Such places could be the incubators of whole biospheres of extremophiles. Add in panspermia for free delivery.

  • Ashley Baldwin November 22, 2019, 14:46

    Simulation, simulation, simulation . As as Sherlock Holmes would say , “ We need more data”. The absence of bone tide planets
    ( Hab zone or not ) in the Alpha Centauri system has not prevented increasingly feverish speculation on its erstwhile incumbents . This work pulls us back the day our old friend the Milankovitch cycle/s. A quick review of these ( orbital eccentricity , axial tilt and precession ) in the Solar system – and their various drivers – Jupiter , Saturn plus the Sun and moon, still shows fluctuating debate on their individual and combined effects on Earth’s climate. At present orbital eccentricity seems the preferred biggest influencer rather than axial tilt though for a long time the reverse was believed . So if there is so much uncertainty even for Earth – after a century of increasingly sophisticated research …and lots of geophysical data providing a physical history of our planet’s climate based on observation rather than just simulation – then I’m going to hold judgement on any planets in our near neighbour star system. At least until we have something tangible to judge.

  • Geoffrey Hillend November 22, 2019, 16:13

    I do agree with the author Astrophysicist Billy Quarles that if there is an Earth like exoplanet around Centauri A or B and it didn’t have a moon, the axial tilt would be large like Mars as an example. It might be difficult for it to have a moon due to changes in the heavy bombardment period which there could be too many or to little collisions.

    I can’t get the ring idea of binary star system evolution out of my mind. It is an older idea, but it makes sense intuitively too me. Two stars near each other with of distance of Sun Saturn apart and larger might change how planets are formed in the protoplanetary disk or make them impossible since the two stars are gravitationally disruptive and their strong gravity uses up most of the matter or gas and dust and there is not a lot left over for any planets?

    • Ashley Baldwin November 22, 2019, 20:06

      Both Alpha Centauri A and B still possess their own circumstellar disks with 10-100 x the mass of the Sun’s. Less expansive now obviously, because of the gravitational effect of the companion star – out to about 2.8 AU for A and 2.5 AU for B. This would imply firstly that they both had much more extensive and more massive disks in their early history. In turn, this also implies that the two stars originally formed much further apart before migrating inwards to their current positions..

      Either way – given that even rocky planets are fully formed via disk accretion within just a few hundred million years of a star entering the main sequence – makes it more than likely that terrestrial planets atleast could have formed around either star . In the inward migration scenario each companion star could also act analogously to Jupiter and Saturn in the solar system during the “Grand Tack” epoch, pushing volatiles in system via cometary bombardment akin to the “late bombardment” . So still plenty of room for optimism .

      With so many previous expectations about exoplanets being proved ( often very) wrong or “counter intuitive” , I fully expect Alpha Centauri to continue this trend when it’s finally reveals its secrets.

  • Geoffrey Hillend November 22, 2019, 22:34

    According to the ring model evolution of binary star systems which I think the Alpha Centauri system fits, the two stars form differently than a single star or binary system with the stars really close so their migration would not matter. If I am apperceiving the ring idea of binary stellar evolution correctly, planets can’t form the usual way since there is no way the gas an dust to form planets to borrow angular momentum from the two stars which form in a wide ring. This is a two body problem because the whole gas cloud is collapsed into two separate gas clouds with common barycenter inside or close to the larger star or gas cloud. The barycenter is the problem so that there is no way for the angular momentum to be borrowed by the gas cloud for any planets, since the angular momentum is used up by the formation of both stars. The barycenter and stars are inside the orbit of Kepler 47 which is inside all the orbits of it’s planets which orbit around the binary system, but not with Alpha Centauri. This is my intuitive attempt to understand what Kippenhahn meant by ring evolution. He states that the stars begin with the ring so the gas and dust is limited to that ring and there is angular momentum in a center so no planets can form in such a ring or disk which is hollow with not much gas and dust in the center point between the two stars in a binary system like Alpha Centauri. Consequently planet’s can form in binary star systems mentioned above.

    This is a rather unorthodox idea and a computer was used to model it, and I don’t expect it to be correct, but we have yet to find any planet in the Alpha Centauri system. Only time will tell. I may not know enough about astrophysics to be able to invalidate Rudolf Kippenhahn’s idea. I am just wondering if this old idea is obsolete and if planets are found around Alpha Centauri or other similar system’s and if they are found, the ring idea will certainly will be invalided.

  • Geoffrey Hillend November 23, 2019, 15:40

    If we include Proxima Centauri, it is technically a three body problem and more unpredictable. According to the ring model of binary stellar evolution, both stars keep the angular momentum, so no planets can form. I don’t know if this model applies to Alpha Centauri since it is three body problem, but I think the principle is supported by astrophysics. Only time well tell if it is correct. If it is correct, then there would be less stars in the galaxy with planets, maybe as much as half of the stars in the galaxy without planets. Ibid.

  • Dave Moore November 23, 2019, 21:28

    Changes in obliquity are the least of the problems for habitability for a planet around ACB. The equators of the two main stars in the Alpha Centauri system are severely misaligned with the plain of their orbits, which means, due to the Kowzai effect, the planets orbit will oscillate from circular to highly elliptical and back again.

    If there is a planet in B’s habitable zone, it’s going to have a wild ride.

    • Ashley Baldwin December 1, 2019, 15:13

      Yes. I fear so. Kozai a well described and likely big player in this system . As well as other neighbouring ,’promising’, tight binaries such as 70 Ophiuchi , 36 Ophiuchi ( not including C) and Eta Cassiopeia .

  • SpaceKnight November 25, 2019, 17:30

    I wonder, with a system like this, could Trojan planets exist? Similar to Jupiter there could be objects trailing or leading Alpha Cen B (or A for that matter?)

  • Harry R Ray November 26, 2019, 10:51

    If you remember 2 years ago, Luger et al stated that GEOLOGICAL FEATURES in the TRAPPIST-1 planets could be resolved during planet-planet occultations using JWST. Then, just last year, Kipping et al stated that such geological features(their term:EXOMOUNTAINS)could be resolved when a Mars-sized exoplanet transits a White Dwarf star using LSST. NOW: Enter The Astronomical Journal Volume 158, Number 6/249(google it and click on the listing at the top of the list page, which is)”Surface Imaging of Proxima b and Other Exoplanets: Albedo Maps, Biosignatures, and Technosignatures.” by S. V. Berdyugina and J. R. Kuhn. No planet-planet occultations or transits necessary, just phase curves using a technique called “EPSI”. WOW!!!!!

    • Alex Tolley November 27, 2019, 16:17

      Very interesting work. It is a pity the authors did not [or were unable to] use real data from telescopes to collect the light intensity from our solar system planets to show that their technique would work with real data, rather than simulated data extracted from existing images. A good test would have been to try Mars over its 2-year orbit to determine if the technique would reproduce the fairly good results from their simulation. Their results, if they work, would provide a better resolution than the best Hubble images of Pluto, which proved reasonably good at finding major features on that body.

    • Mike Serfas November 30, 2019, 18:35

      This was indeed a fascinating and impressive claim, if it can really be done.

      But it mostly just makes me wonder: is there any hope or plan for resolving very small angular features using very small actual _angles_? Suppose we had one of those orbital tethers like you imagine having in orbit to grab a SpaceShip One at the high point of its trip and let it go on an outbound trajectory. But instead of any heavy transportation use, watch how a spinning wire blocks out the exoplanetlight. A parsec is “only” 200,000 AU, so I think a long wire 1 mm wide that is almost 2 AU away from us in some near-antichthonic horseshoe orbit ought to block out about about a 100-meter-per-parsec strip of light reaching a <<1mm sensor from any exoplanet that passes behind its whirling blade. Arrange two such wires moving perpendicularly across the planet at different times and you should be able to do a sort of raster scan of the whole planet … if the precision with which you detect the light and process the signal is impressive enough. (Probably wouldn't hurt to arrange some much closer small object to block out the parent star) Perhaps I exaggerate, but how much?

      • Alex Tolley December 1, 2019, 19:55

        There is a technique (forget the name) where random binary grid masks of n x m pixels between an object and a single-pixel sensor can be used to recreate the object. This might be a more extreme version of your idea. One needs thousand of masks and sensor readings, but these can be generated easily. The differing light intensities received reflect the arrangement of the mask pixels – whether light blocking or not. Integrating the light intensities with the known mask patterns allows the 2D image of teh object to be recreated. Whether this would work on such tiny objects and small noisy, light intensity variations, I don’t know, but the mask pixels would have to be very small to work – but not an issue is the mask is some distance from the sensor, perhaps embedded in a starshade. Noise in teh sensor readings would probably prevent such a system from working, but with enough readings, even that may be overcome.

        • Ron S. December 2, 2019, 15:24

          I think the difficulty is far worse than that. Diffraction would render the tiny “shade” invisible. Any distortion that is conceivably detectable (almost certainly nothing at all) would only affect the same single pixel; consider the angular width of a 1 mm mask at 2 AU distance. Image processing techniques need data to work with and there will be none.

        • Mike Serfas December 3, 2019, 11:02

          For angular resolution within the disc of a distant planet, the starshade needs to be very distant. Unfortunately, this makes it relatively hard to realign at multiple targets. I was hoping a long rotating tether would sweep out obstructions over a large swatch of sky, so that a few tethers and a few sensors could cover a reasonable proportion of nearby stars.

  • Michael Fidler November 27, 2019, 6:54

    Thanks, Harry Ray, it sounds like we are on the threshold of a dream! I hope Paul does a article on this soon and some of the other methods possible. Here our some that could be combined for infinitely higher resolutions.
    1. Quantum Memory interferometers.
    2. UV Flare S/N improvement.
    3. Kipping Whole Earth atmospheric telescope.
    4. Whole Venus, Titan, Jupiter, Saturn, etc. telescopes.
    Combinations of any of the above could be possible in the next 20 to 30 years, especially with SpaceX’s Starship. The need to go to the Sun’s gravity lens may not be required if these combinations of a solar system wide based interferometer are practical.

  • Geoffrey Hillend November 27, 2019, 19:09

    I don’t think we can have moons in Lagrangian points since they are too massive. I like the idea of the ring binary star system formation which the two stars keep all the angular momentum staring with the beginning of their collapse from gas and dust so there is no angular momentum for any planets to form. Even before the stars are born there are not any planets. No planets can form due to the way the binary star system begins. All the gas an dusk is limited to a ring. The collapse of gas is limited to the ring so two stars form opposite each other. There is no gas in the center between the stars or the gas can’t collapse without any angular momentum like a single star in the center of protoplanetary disk systems or binary stars close to each other in the center. Consequently, if the Alpha Centauri system formed by the ring binary star system theory, it has no planets. As already mentioned there are other physics which make this system unstable, so it will be interesting to see if there are any planets there.

  • Harry R Ray December 5, 2019, 11:08

    BREAKING NEWS: A BONIFIDE(i.e., NOT a dwarf planet embedded in a debris field)planet has FINALLY been detected around a BONAFIDE(i.e., NOT a blue subdwarf)solitary WHITE dwarf star! WD J01914+1914 has a radius of ~1 Rearth and a surface temperature of 27,750 K. WD J01914+1914 b orbits its parent stellar reminant at a distance of 0.07 AU in a 10 day orbital period. and has a radius of ~4Rearth. For further details, click on arxiv: 1912.01611 “Accretion of a giant planet onto a white dwarf.” by T. Goensike et al. KEY QUOTE FROM THE ABSTRACT: “The orbit of the planet is most likely the result of gravitational interactions indicating the presence of additional planets in the system. Any ~1 Earth mass planets in the HZ?