≡ Menu

Upgraded Search for Alpha Centauri Planets

Breakthrough Starshot, the research and engineering effort to lay the groundwork for the launch of nanocraft to Alpha Centauri within a generation, is now investing in an attempt to learn a great deal more about possible planets around these stars. We already know about Proxima b, the highly interesting world orbiting the red dwarf in the system, but we also have a K- and G-class star here, either of which might have planets of its own.


Image: The Alpha Centauri system. The combined light of Centauri A (G-class) and Centauri B (K-class) appears here as a single overwhelmingly bright ‘star.’ Proxima Centauri can be seen circled at bottom right. Credit: European Southern Observatory.

To learn more, Breakthrough Initiatives is working with the European Southern Observatory on modifications to the VISIR instrument (VLT Imager and Spectrometer for mid-Infrared) mounted at ESO’s Very Large Telescope (VLT). Observing in the infrared has advantages for detecting an exoplanet because the contrast between the light of the star and the light of the planet is diminished at these wavelengths, although the star is still millions of times brighter.

To surmount the problem, VISIR will be fitted out for adaptive optics. In addition, Kampf Telescope Optics of Munich will deliver a wavefront sensor and calibration device, while the University of Liège (Belgium) and Uppsala University (Sweden) will jointly develop a coronagraph that will mask the light of the star enough to reveal terrestrial planets.


Image: Paranal at sunset. This panoramic photograph captures the ESO Very Large Telescope (VLT) as twilight comes to Cerro Paranal. The enclosures of the VLT stand out in the picture as the telescopes in them are readied for the night. The VLT is the world’s most powerful advanced optical telescope, consisting of four Unit Telescopes with primary mirrors 8.2 metres in diameter and four movable 1.8-metre Auxiliary Telescopes (ATs), which can be seen in the left corner of the image. Credit: ESO.

According to the agreement signed by Breakthrough Initiatives executive director Pete Worden and European Southern Observatory director general Tim de Zeeuw, Breakthrough Initiatives will pay for a large part of the technology and development costs for the VISIR modifications. Meanwhile, the ESO will provide the necessary telescope time for a search program that will be conducted in 2019. The VISIR work, according to this ESO news release, should provide a proof of concept for the METIS instrument (Mid-infrared E-ELT Imager and Spectrograph), the third instrument on the upcoming European Extremely Large Telescope.


Comments on this entry are closed.

  • DJ Kaplan January 10, 2017, 13:28

    By the time we have all the instruments on line, the two AC elements should be at maximum optical separation, lol.

  • Enzo January 10, 2017, 15:37

    What happened to Deborah Fischer search ?
    It was supposed to be done in 5 years from 2009 :

    For a while there were some updates then….nothing.

    • Debra Fischer June 25, 2017, 12:16

      The stability of the echelle spectrograph at CTIO was not sufficient to obtain radial velocities with needed precision. We then obtained NSF funding to build CHIRON (commissioned in 2012) and started over! This spectrograph was stable over short time baselines, but by 2012, the projected separation of the two stars A and B was so small (about 5 arcseconds) that our spectra for one of the stars was always contaminated by light from the other.

  • DJ Kaplan January 10, 2017, 16:52

    Yes, and ESPRESSO will be a helpful instrument when it comes on line as well.

  • Neil Wingate January 10, 2017, 18:46

    I have a question…. but to set it up….. in considering a trip to Alpha Centauri… in one scenario/ design we use a large solar sail array, powered by a sun orbiting laser (in addition to the sun’s own push). On board is a small unmanned & automated research station… say 2-3 times the size of the space shuttle plus the fuel tank. It has chemical rockets on board for use upon arrival. My question is in regards to the deceleration needs.

    Would the chemical rockets provide a more impressive deceleration per amount of fuel from these speeds than it would acceleration?

    I ask in light of the fact that it takes more and more fuel to accelerate a set amount the faster you are going (Einstein’s ‘the ship gets effectively heavier the faster you are moving). If we are going 20% the speed of light upon arrival at the star system, then would a deceleration from these speeds be more ‘efficient’ using chemical rockets, then the same acceleration would?

    • Paul Gilster January 10, 2017, 20:48

      Neil, you’ve got the same problem either way. Chemical propulsion isn’t workable at speeds like these because of the vast amount of fuel needed, the huge imbalance between fuel and payload becoming absurd. Better to think of deceleration options like magsails, which can brake against a star’s stellar wind. Deceleration is a huge problem no matter what the initial propulsion mechanism is, which is one reason there have been various flyby missions designed that simply blow through the destination system and report back along the way. I believe Icarus intends to use the same fusion engine for deceleration that was used in acceleration, but then, we’re talking fusion there, not chemical rockets. Robert Forward had a ‘staged sail’ concept for deceleration involving separating his huge sail into sections, with one reflecting Earth-based laser light back on the other to slow the sail. Other even more exotic options are possible, but to me the magsail seems most realistic, at least extrapolating forward from our present level of technology.

      • Neil January 10, 2017, 23:26

        Thanks Paul… not sure I got my point/question across… the Alpha Centauri trip is just my story problem for the sake of discussion. Special relativity and its effects on mass at relativistic speeds…. if they are applied to propellant leaving a craft as a deceleration mechanism at relativistic speeds versus acceleration from a stationary status….

        Let me put it this way, lets say 1 pound (earth gravity) of propellant leaving a ship gives it 1 mph of increased velocity at initial fire from a dead stop in space. If the ship is traveling at 10% the speed of light, turns around, and fires that same 1 pound of propellant to decelerate, does it decelerate by 1 mph as well? I expect it does, but I know at these speeds the rules in regards to reaction mass are not as obvious (at least to me)….. Thanks for humoring me in any case.

        • Paul Gilster January 11, 2017, 9:56

          Ah, I see what you’re getting at, Neil. In that case, though, I’ll leave you to some of my readers with more knowledge of relativistic effects than I can muster. It will be interesting to see what they come up with.

        • Ron S January 11, 2017, 12:03

          Neil, deceleration is a not a distinct concept. Acceleration is a vector, which includes a magnitude and a direction. That’s all. Does this help?

          • Neil Wingate January 11, 2017, 17:54

            Thanks Ron, and yes, i think that does it. I was overthinking it (which was part of the fun). So, does this sound right?

            If I go from zero to 1 mile per second using one kg of fuel, then it will also take (all else being equal… overall mass of object, etc.) one kg to go from 18,600 miles per second to 18,599 miles per second.

            Is that statement correct?

            • Ron S January 11, 2017, 22:26

              Yes. Of course when we get into relativistic velocities there is the Lorentz transform to consider with respect to some reference object, however that will only change the measured relative velocity.

              If you accelerate and decelerate to the same local measurement (e.g. 3 g for 5 seconds, then -3 g for 5 seconds) the net velocity change will be zero, regardless of whether you are going fast or slow relative to any external object.

        • mlockmoore January 11, 2017, 13:39

          The relativistic effects like apparent mass increase scale with the Lorentz transformation, and this factor does not get above 2 until you go past 86% of the speed of light. At half the speed of light, the mass increase is only about 15%, that that in itself is pretty manageable.

          As the others said, the real problem is there is just so much speed that you need crazy amounts of mass to shoot out your nozzle to change your momentum enough. If the nozzle exhaust speed is a lot faster than the ship speed, you have a chance, but if your craft is going a decent fraction of the speed of light, that’s not an option. Realistically, chemical rocket exhaust is pretty slow, so you just need way too much propellant.

          • Bruce Mayfield January 12, 2017, 8:55

            Interesting. What is the formula for the apparent mass increase with increasing velocity? Does this have any effect at all at 0.2 & .25c?

            • Bruce Mayfield January 14, 2017, 17:45

              Researched my own questions, learning about the Lorentz factor.

              Effective mass = rest mass x 1/squareroot(velocity squared/c squared).

              So at .2 & .25c the effective masses would be 1.021 & 1.033 times rest mass.

        • Michael January 12, 2017, 18:07

          If you look from the target system at the exhaust of the rocket it will only be going what the chemical energy can give it, namely 4 km/s extra. The effect becomes greater in stronger gravity fields though, but it would have to be a serious g field to be noticeable.

    • Project Studio January 11, 2017, 7:45

      Fun question. Although you have added ‘mass equivalent’ energy to your fuel supply during the acceleration phase, I think that additional relativistic mass is not going to convert to extra chemical energy during the deceleration phase. That additional mass is just going to become extra ‘payload’ that will have to be decelerated along with your research station and fuel tank (at last until it is expelled as exhaust).

  • Ashley Baldwin January 10, 2017, 18:59

    An interesting development , especially as the SPHERE imaging technology and ultra new high res ESPRESSO are still being touted as being adapted for imaging Proxima b , though in the much shorter wavelength visible spectrum. The fore runner of a later than METIS E -ELT visual imaging technology , EPIC.

    The proximity of Alpha Centauri system is certainly making it THE place to image for exoplanets over the coming years, especially if it can be done relatively cheaply and quickly from the ground rather than space .

    The ESO VLT is now clearly setting itself up as a technological demonstrator for the E-ELT. A push to image these planets all the same though with just the limited angular resolution of a single 8.2m Unit telescope . Using the VLT to test out METIS does surprise me a bit as it only at an altitude of 2600 m or so which is far from optimal for imaging in the mid infrared wavelength range of that instrument . The METIS design team couldn’t hide their dismay at the final selection of 3060m Cerro Armaxones for the E-ELT , as imaging beyond 3 microns ( ideally at 5 and 11 microns or the M and N bands of the limited infrared wavelength “Windows” that penetrate the Earth’s atmosphere ) is best done above 5000m where the low atmospheric humidity and cold ambient temperatures make the process of counteracting heat “noise” in the telescope and sky “background” much easier than at “lower” altitudes ( and also applicable to longer still sub millimetre wavelengths and why ALMA is situated at 5200m on the Atacama desert) .

    The METIS team were hoping for the 5000m plus Cerro Macon mountain ridge at Salta in Argentina to be picked as the site for the E-ELT as its “on sky” imaging was optimised for their device’s wavelengths . A good imaging sky isn’t the only selection criteria though for a billion € plus telescope ( accessibility , water /power supply , politics etc ) and the site wasn’t even ultimatelty short listed .

    This Alpha Centauri project is as much about presumably using recently developed techniques ( METIS was conceived as a concept nearly ten years ago ) to try and compensate for this ahead of the E-ELT first light next decade . Why use it at all ? The contrast difference in stellar flux between an Earth mass planet and a Sun like star ( Alpha Centauri A and B) is about a billion times or more at the red end of visible but drops to just 20 million by the N band of METIS. Fifty times better ! That makes for much less stringent requirements on any adaptive optics system ( including the now requisite coronagraph ) though at the price of angular resolution . Even at the optimal “Diffraction limit” ( perfect imaging conditions found in space ) of a telescope , this is expressed as resolution (in metres ) = 1.22 X wavelength (in metres ) all divided by telescope aperture (in metres too) . So the 11 micron N infrared band resolution is 22 times less than say o.5 microns visible light for any given telescope aperture.

    It’s only because Alpha Centauri is so close that an 8m telescope will have any chance at all of resolving any planets, helped by a search for habitable zone explanets at about the same respectable 1AU or more distance for the two sun like constituents of the binary . I’m not sure whether the dedicated binary system imaging software ( Supe Nyquist ) especially developed for the Alpha Centauri concept telescopes ACE-sat & Project Blue might also help and be used here too .

    Looking for habitable Earth like exoplanets around the three nearest stars though. Wow. Inconceivable even twenty years ago.

  • spaceman January 10, 2017, 23:03

    Will the upgraded search look for planets around all three stars of the triple system? Will the upgraded instruments be capable of detecting terrestrial planets the size of earth or just the bigger super-Earth sized worlds?

    • Ashley Baldwin January 11, 2017, 14:45

      No I think that’s the point . Ordinarily an 8m scope couldn’t have the resolution to image “eta Earth” style habitable zone planets as the angular separation would be too small to separate them from their parent star . It’s purely the fact that the three Alpha Centauri star systems are just so close at 2.24-2.34 light years that means they can be resolved and their spectra analysed. Even for Proxima b which is less than 0.05 AU from its star . It’s tight all the same but just on the upper limit of what is achievable . For Alpah Centauri A &B the issue will be blocking out the encroaching light of the other binary constituent while targeting each in turn .

      Analogous in a way to ACEsat /Project blue space telescope concepts where even a 35cm aperture scope ( or ideally nearer 50 cm) can image has zone Alpha Centauri planets with enough time to allow potent post processing . Breakthrough long shot must think adapting the VLT instruments is cheaper and easier than backing Project Blue which had previously surprised me as finding a worthwhile target for any interstellar probe was a must before launching on a long and expensive development project. Now I know why . Despite its relatively cheap $25 million price tag. ( plus at least two years operations and launch costs too)

      Adapting SPHERE and ESPRESSO/ VIRTIS & AO at the VLT in order to image Proxima b also promised imaging planets around other close by stars though I’m not sure as to how far out that applies . Whatever the price of the adaptations ,which I’m sure will be less than the space telescopes, once and made they are also there to last and easily accessible for servicing, unlike any space telescope . Most of the nearest stars are red dwarfs which offers the advantage of less of flux contrast between planet and star but at the price of close angular separation for habzone planets.

      That said , even if you can only image bigger or further out planets than habzone eta Earths in more remote systems you can still get detailed spectra of their atmospheres with ESPRESSO / VIRTIS which isn’t available on either of the space telescopes . The first detailed spectra of temperate terrestrial planets rather than the hot gas giants supplied by Hubble and Spitzer to date .

      That’s obviously felt to be a good return on investment , returning potentially revolutionary science and with some of the cost supported by the ESO who are keen for proof of concept work on their instruments ahead of the E-ELT . This a clever way of extending the useful lifespan of a smaller telescope that has worked so well for the 4.2m William Herschel telescope in the Canaries which now regularly hosts “guest” prototype instruments ahead of their deployment on 8-10m class scopes.

      The small space telescopes cap costs by avoiding any additional even low res spectrograph use and are limited to the cruder characterisation offered by imaging in four discrete wavelengths ( unlike the continuous spectrum of any large ground based high res spectrograph ) to look for things like the blue sky created by Raleigh scattering believed to be typical of an Earth like planetary atmosphere .

      • DJ Kaplan January 12, 2017, 13:34

        “Eta Earth”: I like that. Thanks.

        • Ashley Baldwin January 12, 2017, 16:17

          Too clever for me I’m afraid but I agree that is has a good sound. It’s VISIR rather than VIRTIS too. Exciting times for temperate terrestrial planets though even before the ELTs and WFIRST. SPHERE/ESPRESSO may be a unique combination for Proxima b and VISIR / AO for the other two star systems . Both hopefully for a few other nearby exoplanets as well. High resolution spectra.

  • Ashley Baldwin January 11, 2017, 7:05

    The ESPRESSO spectrograph has the facility for a maximum resolution in excess of an incredible 200000, though only over a narrow bandwidth due to its primary function of RV spectroscopy detection of exoplanets . The plan is to trial it at the VLT before using an enlarged version , CODEX , on the E-ELT. This is where the problem arises that faces all modern, large and high tech telescopes. Cost.

    As telescopes increase in size , so do there instrument suites at a time when they are also becoming more complex too and both of these are leading to ever larger increases in the cost of a new telescope with the instruments often coming out as aimilarly expensive or more so than the actual scope. To this effect, and given the need for high resolution near infrared spectrographs there is a proposal before the ESA Council to merge CODEX and the planned high res NIR spectrograph into one . This will save cost to a degree to create a very expensive single instrument but by reducing the resolution dramatically down towards 100000. Still high , but a lot less so and the reduction will have big implications for the devices capability, especially around exoplanet science . Costs around €18 million have been quoted ( which are invariably lower than turns out ) even for the combined instrument . This is likely to be too much. They are currently deciding what to do !

    It may well be that ESPRESSO ends up the highest resolution spectrograph for some time , but obviously limited to an 8m rather than 40m telescope . This is important to exoplanet science as the promising technique of High definition imaging ,HDI , that can pick the atmospheric spectrum of an exoplanet out of that of its nearby parent star and the Earths atmosphere requires both high spectrographic AND optical resolution to do so .i.e big telescope and high res spectrograph , all in conjunction with a state of the art adaptive optics/imaging system ( very expensive in its own right ) like SPHERE or VIRTIS on the VLT with METIS and EPIC to come on the E-ELT. All of that excluding the necessary development costs of this highly cutting edge technology . It may cost to adapt VIRTIS, SPHERE and ESPRESSO but in the long run it will likely help reduce final costs .

  • Gary D January 11, 2017, 19:48

    Dr. Debra Fischer is now involved with the 100 Earths project, searching out nearby Earth-like planets.


    • Ashley Baldwin January 15, 2017, 7:22

      And very impressive this is too. The EXPRES spectrograph she has developed ( following on from the impressive CHIRON spectrograph ) has RV resolution capabilities on a par with ESPRESSO at the VLT and the same baseline of 10cm/s required to discover Earth sized planets as the title of the project alludes.

      The biggest obstacle in this field now is not spectrographic capability but intrusive background stellar ” noise” . I’m sure this is something all such specialists will now be looking at ways to reduce , which with time they assuredly will allowing a new and exciting round of terrestrial planet discovery around neighbouring stars . ESPRES is a fibre fed , coude focus laser comb spectrograph presumably working like similarly designed ESPRESSO in the visible spectrum ( with a similar but slightly smaller bandwidth ) so it won’t be following the working in the near infrared approach that some observers are championing ( including Gullem Anglada -Escude , discoverer of Proxima b) Just last week a new process of noise reduction was applied to the RV data used to discover Proxima b, not just confirming it but showing an easier way of picking out its signal. So progress is already being made . With two such potent instruments available now I’m sure after a brief hiatus the discovery envelope will be pushed even further . A quick survey of even the fifty nearest stars shows how few have confirmed planets around them ,which doesn’t fit with the wider exoplanet distribution pattern seen in the Kepler data so there is obviously a lot of discovery space still .

      Clever too to deploy the spectrograph at the 4.3m Discovey telescope. Not the world’s biggest ( though still far from small and also modern ) but far easier to get the necessary dedicated observation time required for this time consuming and pain staking work as opposed to fighting for scraps observations on the sought after 8-10m class premier scopes.

  • ljk March 18, 2017, 13:36

    A Journey to Alpha Centauri – Christian Marois (SETI Talks 2017)


    SETI Institute

    Published on March 17, 2017

    The Alpha Centauri star system is ideal to search for habitable planets by various observing techniques due to its proximity and wide range of stellar masses. Following the recent discovery of an Earth-size planet candidate located inside the Proxima Centauri habitable zone, Dr. Marois will discuss this remarkable discovery and the planet’s potential to find life.

    He will also present our current instrument project for the Gemini South observatory, TIKI, to discover similar planets around the two Sun-like pair located 15,000 AU from Proxima Centauri. The Alpha Centauri system is the prime target of the Breakthrough Starshot program, a project to send small quarter-size probes to take resolve images of these new worlds, and to prepare for Humanity’s first step into a new star system.

    Dr Marois completed his Ph.D. at the Université de Montréal in 2004. The main topic of his thesis work was to understand the limits in exoplanet imaging and to design innovating observing strategies. After his thesis, he did postdoctoral researches at the Lawrence Livermore National Laboratory, Univ. of California Berkeley and NRC.

    In 2008, while at NRC, he led the team that took the first image of another planetary system (HR 8799) using the Keck and Gemini telescopes. He is currently pursuing his research at the NRC Herzberg where he is part of the Gemini Planet Imager campaign, and leading the development of instruments for imaging Earth-like planets at Gemini South and at the TMT.