≡ Menu

TOLIMAN: Looking for Earth Mass Planets at Alpha Centauri

Why the renewed focus on astrometry when it comes to Alpha Centauri (a theme we saw as well in the previous post on ALMA observations from the surface)? One problem we face with other detection methods is simply statistical: We can study planets, as via the Kepler mission, by their transits, but if we want to know about specific stars that are near us, we can’t assume a lucky alignment.

Radial velocity requires no transits, but has yet to be pushed to the level of detecting Earth-mass planets at habitable-zone distances from stars like our own. This is why imaging is now very much in the mix, as is astrometry, and getting the latter into space in a dedicated mission has occupied a team at the University of Sydney led by Peter Tuthill for a number of years — I remember hearing Tuthill describe the technology at Breakthrough Discuss in 2016.

Out of this effort we get a concept called TOLIMAN, a space telescope that draws its title from Alpha Centauri B, whose medieval name in Arabic, so I’m told, was al-Zulmān. [Addendum: This is mistaken, as reader Joy Sutton notes in the comments. It wouldn’t be until well after the medieval period — in 1689 — that the binary nature of Centauri A and B was discovered by Jesuit missionary and astronomer Jean Richaud. The name al-Zulmān seems to be associated with an asterism that included Alpha Centauri, though I haven’t been able to track down anything more about it. Will do some further digging.]

The name is now fixed — a list of IAU-approved star names as of January 1st, 2021 shows Alpha Centauri A as Rigel Kentaurus and Centauri B as Toliman, so what goes around comes around. Centauri C’s name, according to the IAU, remains the familiar Proxima Centauri.

Image: This is Figure 5 from an online description of this work. Caption: Habitable zones of Alpha Cen A (left) and B (right) in green, along with dynamical stability boundary (red dashed line), 0.4” and 2.5” inner and outer working angles (IWA and OWA) of a small coronagraphic mission. The inset shows the Solar System to scale. Planetary systems of Alpha Cen A & B are assumed to be in the plane of the binary (the likeliest scenario) and orbits of hypothetical Venus-like, Earth-like, and Mars-like planets are shown. Credit: Tuthill et al.

The TOLIMAN Technology

The TOLIMAN mission is designed to use astrometry by means of a diffractive pupil aperture mask that in effect leverages the distortions of the optical system to produce a ‘ruler’ that can detect the changes in position that flag an Earth-mass planet in the Alpha Centauri system. Tuthill’s colleague Céline Bœhm presented the work at the recent Breakthrough Discuss online.

In TOLIMAN, we actually have an acronym — Telescope for Orbit Locus Interferometric Monitoring of our Astronomical Neighborhood — but at least it’s one that’s applicable to the system under study. According to Boehm, there are certain advantages to working with nearby binaries. Astrometry normally uses field stars (stars in the same field as the object being studied) as references, making larger apertures an essential given the faintness of many of these reference points and their distance from the target star, and requiring a wide field of view.

But a bright binary companion 4 arcseconds away, as we find at Alpha Centauri, resolves the problem. Now a small aperture telescope can come into play because we have no need for field stars. The TOLIMAN concept puts a diffractive ‘pupil’ in front of the optics that spreads the starlight out over many pixels, a diffraction caused by features embedded in the pupil. Here stability is at play: We’re trying to eliminate minute imperfections and mechanical drifts in the optical surfaces that can create instabilities that compromise the underlying signal. The pupil mask prevents the detector from becoming saturated and reduces noise levels in the signal.

The astrometric ‘ruler’ used to measure the star’s position is thus created by the light of the stars themselves. The diffraction pattern cast by the pupil positions a reference grid onto the detector plane, which registers precise stellar locations. As a 2018 paper on the mission puts it: “Drifts in the optical system therefore cause identical displacements of both the object and the ruler being used to measure it, and so the data are immune to a large class of errors that beset other precision astrometric experiments.”

Narrow-angle astrometry, where the reference star is extremely close, makes for greater precision than wide-angle astrometry, and angular deviations on the sky are larger. Boehm makes the case that TOLIMAN’s diffractive pupil technology should be effective at finding Earth analogs around Centauri A and B, with a more advanced mission (TOLIMAN+) capable of finding rocky worlds around secondary targets 61 Cygni and 70 Ophiuchi.

Image: This is from Figure 3 of the online description of TOLIMAN referenced above. Caption: Left: pupil plane for TOLIMAN diffractive-aperture telescope. Light is only collected in the 10 elliptical patches (the remainder of the pupil is opaque in this conceptual illustration, although our flight design will employ phase steps which do not waste starlight). Middle: The simulated image observing a point-source star with this pupil. The region surrounding the star can be seen to be filled with a complex pattern of interference fringes, comprising our diffractive astrometric grid. Right: A simulated image of the Alpha Cen binary star as observed by TOLIMAN. Credit: Tuthill et al.

Astrometry and Earth-like Planets

Unlike both transit and radial velocity methods, the signal generated by a potential planet increases with the separation of planet and star, which makes astrometry an ideal probe for habitable zones. We should also throw into the mix the fact that radial velocity detections vary depending on the spectral type of the star, which is why we have had pronounced radial velocity success at Proxima Centauri and Barnard’s Star, but face a steeper climb in doing the same around G- and K-class stars. Astrometry becomes more sensitive with rising luminosity of the host star even as we wrestle with the tiny distortions that a planet like this would produce.

Both HIPPARCOS and Gaia have shown what can be accomplished by moving astrometry into space, and we saw yesterday that Gaia is likely to produce gas giants in the thousands, but the idea behind TOLIMAN is that astrometry now needs a dedicated mission to get down to rocky planets in temperate orbits. Tuthill and team have pointed to Alpha Centauri as an ideal target, one that would yield signals potentially 2 to 10 times stronger than the next best systems for study.

The researchers believe their work yields the optical innovations needed to acquire astrometric data on habitable-zone rocky planets with a telescope with an aperture of 10 cm. To explore the concept further, a CubeSat mission called TinyTol is scheduled to fly later this year, serving as a pathfinder to demonstrate astrometric detection. The full-scale TOLIMAN mission is funded and under development, with launch planned for a 3 year mission in 2023. TOLIMAN+ is envisioned as a 0.5-meter aperture telescope capable of sub-Earth class detections at Alpha Centauri and Earth-mass planets at both 61 Cygni and 70 Ophiuchi.

The most recent paper I know of on this work is Tuthill et al. “The TOLIMAN space telescope: discovering exoplanets in the solar neighbourhood,” SPIE Proceedings Vol. 11446 (13 December 2020). Abstract.


Comments on this entry are closed.

  • Michael Fidler April 30, 2021, 8:26

    Fascinating, I did find a live download site for the paper:

    “The TOLIMAN space telescope: discovering exoplanets in the solar neighbourhood,”


  • Michael Fidler April 30, 2021, 8:42

    I noticed that the inner and outer working angles (IWA and OWA) of a small coronagraphic mission that the IWA would cut out over 50% of the habitable zone of Alpha Cen B. The TOLIMAN mission should be able to see the movement of planets around Alpha Cen B also, but how do they know which system the change in position is coming from A or B ?

    • Marshall Eubanks May 1, 2021, 9:46

      “how do they know which system the change in position is coming from A or B ?”

      I am not sure they can do that separation (or ambiguity resolution, in the jargon) with Toliman, which would produce differential position estimates for A minus B. In other words, I don’t think a periodic motion of the centroid of A could be distinguished from the same motion of the centroid of B.

      In the long run, it may not matter that much, as anything found by Toliman would have to be confirmed, and the confirming techniques (e.g., radial velocity, direct imaging or microlensing) would all reveal which star actually had the planet. (Of course, if each has a planet of roughly the same mass and orbital period, the analysis could get messy, but these are problems you’d like to have.)

      There are two technical papers on this technique, but neither deals with resolving ambiguities.


      • Michael Fidler May 1, 2021, 11:02

        Yes, I see in the pptx presentation above that is one of the limitations.

        n Caveats …
        n Note we only get 1-D data
        n Planets in orbital plane of binary
        n Orthogonal planets invisible
        n Ambiguity: which star is host?

        • Michael Fidler May 1, 2021, 23:59

          There is a good side to this that may work. If two planets or more can be discovered in this way, the odds will be in favor of the planets orbital period, be it from A or B will not have similar periods. Like 100 days and 120 days but set resonance periods. If more then two show up it should be easy to diagnose the two systems with an algorithm like DYNAMITE.


          Exoplanets remain challenging to find and almost all multi-planet systems remain only partially explored. By combining exoplanet demographics with orbital dynamical considerations, Dietrich & Apai (2020, AJ in press) developed the dynamical multi-planet injection tester (DYNAMITE), the first algorithm to predict where yet-to-be-discovered exoplanets likely reside in only partially-explored exoplanet systems.

          DYNAMITE draws from exoplanet population statistics derived from the Kepler exoplanet sample and in multiple tests successfully predicted planets in TESS-discovered multi-planet systems.

          The DYNAMITE has been applied to all TESS multi-planet systems to predict where, within each system, additional planets are most likely to reside and how large these planets. Results from DYNAMITE guide follow-up observations exoplanetary systems.


  • Brett April 30, 2021, 10:35

    The full-scale TOLIMAN mission is funded and under development, with launch planned for a 3 year mission in 2023.

    That’s fantastic news! I hope they find some promising planets – it’s worth having dedicated telescopes just for the nearest stars, because they’re going to be the easiest to follow up on and (down the line) the easiest to do interstellar flyby probe missions with.

  • Joy Sutton April 30, 2021, 15:26

    “Alpha Centauri B, whose medieval name in Arabic, so I’m told, was al-Zulmān.”
    Ok Paul, maybe we have recently changed the definition of the medieval period? Or it was recently discovered that the Arabs had telescopes way back then, which should be headline news? Because the Alpha Centauri A-B binary can’t be resolved with the naked eye. I have quite good binoculars, and at 38 degrees South can see Alpha Centauri on any clear night. But, I can’t resolve the binary. So how is it possible for the Arabs to have two medieval star names for what is a single naked eye object?

    • Paul Gilster April 30, 2021, 20:23

      Yes, good point, as the binary nature of Centauri A/B wasn’t discovered until later. I’m glad you pointed this out — the Arab name turns up associated with an asterism of which Alpha Centauri was the main star, according to what I just now scouted out on the Net about this. I’ll change the text with a pointer to your comment.

      • Joy Sutton April 30, 2021, 23:53

        Thanks Paul, My hand held 15 x 50 imaged stabilised Canon binoculars are not up to the task in my hands, although they are great for Jovian moons. Maybe bigger binoculars on a rock steady tripod (and younger eyes) would be enough? I looked it up too, Father Jean Richaud used a 12-foot long (refracting) telescope to resolve the binary from India, certainly not a toy. Light gathering is not a problem with such a bright object, but resolution is important. You can definitely resolve A-B in the backyard with a quality 8″ compact reflector on a tripod.

        • Patient Observer May 1, 2021, 15:11

          Hmmm, could be more to the story of ancient astronomical knowledge:


          In Mali, West Africa, lives a tribe of people called the Dogon. The Dogon are believed to be of Egyptian decent and their astronomical lore goes back thousands of years to 3200 BC. According to their traditions, the star Sirius has a companion star which is invisible to the human eye. This companion star has a 50 year elliptical orbit around the visible Sirius and is extremely heavy. It also rotates on its axis.

          This legend might be of little interest to anybody but the two French anthropologists, Marcel Griaule and Germain Dieterlen, who recorded it from four Dogon priests in the 1930’s. Of little interest except that it is exactly true. How did a people who lacked any kind of astronomical devices know so much about an invisible star? The star, which scientists call Sirius B, wasn’t even photographed until it was done by a large telescope in 1970.

          There may be other explanations for this prescient knowledge but it is an oddity. Sirius B was discovered in 1862 but its characteristics only determined much later.

          • Hamilton1 May 2, 2021, 18:55

            Hmmm this old chestnut again… what the Dogon were or were not talking about seems to depend on whose translations you believe and the interpretations placed thereupon.

            • Patient Observer May 3, 2021, 7:59

              Some types of chestnuts grow to be mighty oak trees. But, I certainly understand the skepticism.

  • Patient Observer April 30, 2021, 15:43

    One more reason to search for an Earth-like planet in Alpha Centauri systems is to fulfill my boyhood fantasies of being hit by a Zeta beam and be teleported to the Planet Rann to establish a deep friendship with the beautiful Alanna. Per Wikidpedia

    Strange is an archeologist suddenly teleported from Peru, Earth, to the planet Rann through a “Zeta-Beam”. Upon his arrival, he is attacked by one of the planet’s predators and rescued by a dark-haired woman called Alanna. She takes him to her father Sardath, who explains that the Zeta-Beam was transmitted to Earth in the hopes that whatever form of intelligent life lived there would trace it back to Rann, and he also theorizes that in the 4.3 years it took the Zeta-Beam to reach Earth, it was altered by space radiation into a transportation beam. Called on to protect the planet from extraterrestrial threats using Rann’s technology, Strange grew to care for the planet and its inhabitants, especially Alanna. Eventually, the effects of the beam’s transportation wore off, automatically returning Strange to Earth at the exact point of departure—but not before Sardath had given him a schedule of beam firings, allowing him to periodically return to the planet. Using mathematical calculations, he was able to determine the exact time and the precise locations at which the Zeta-Beams would arrive. He travels the world and intercepts them, to defend Rann and be with Alanna.[9]

  • Thomas R Mazanec May 1, 2021, 14:04

    Gotta put this one on your list, Paul.

    • Paul Gilster May 1, 2021, 19:29

      Will do, though waiting for a dedicated website.

      • Jeff Wright May 3, 2021, 4:18

        I might first focus on Gliese 710…referred to as DM 61+366 in the STARFLIGHT HANDBOOK for some reason. Each time they revisit this the passage gets nearer.

        It looks to pass by us by 4,000 AU in a million years or so…but any planets it may have might get even closer…maybe shades of “When Worlds Collide”?

        • Michael Fidler May 3, 2021, 9:56

          Come to think of it, maybe that is where earth came from, a passing star…

  • Henry Cordova May 1, 2021, 21:20

    Astrometry has come a long way. Please forgive me if I take the opportunity to reminisce a bit…

    I originally did two semesters of graduate work in astrometry in the early 1970s at the University of South Florida. I was fortunate to have worked under two giants in the field, Dr George Gatewood, my thesis advisor, and Dr Heinrich Eichorn von Wurmb, the USF Astronomy Department Chairman. I never finished the program–not through any fault of theirs–I found I was not as dedicated or talented as I thought! I was a successful undergraduate, but working at the graduate level was a challenge of another order of magnitude However, the skills I learned there prepared me well for a career in the engineering field: aerial photogrammetry, satellite remote sensing, image processing, computer mapping and geographic information systems.

    It was an exciting time, computers were just coming into the discipline to handle the complex plate reductions and statistical analyses of very large datasets. As undergraduates, my classmates and I had done much work writing FORTRAN code to support the number crunching required to correct for random and systematic errors in star positions measured off glass photographic plates. Astrometrists had previously developed clever numerical approximations to reduce the enormous labor of these calculations to manageable proportions. The computer allowed us to apply more rigorous brute-force solutions to achieve much higher levels of precision.

    In those days, stellar positions on old photographs (most of them decades old, and some dating back to the earliest days of astronomical photography) were determined on a measuring engine, an enormous high-precision binocular microscope that moved the plates about by turning finely machined screws (with 1-micron precision threads) to get the x,y coordinates on the plate. Reference star positions in right ascension and declination on each plate had been determined using other astronomical, methods, and then their positions on the plate were also captured with the microscope. Once these established the plate coordinate system, measured positions from all other stars on the plate could be determined. After a stellar image was centered on the crosshairs, stepping on a foot pedal registered its x,y plate coordinates onto 80 column punch cards. Affine transforms and least-squares analyses on the IBM 360 mainframe then rubber-sheeted the positions on the plate to coordinates on the celestial sphere. In addition, by measuring image sizes and comparing them with the reference stars we could get magnitudes, by comparing plates exposed years apart we could derive proper motions, and by comparing images on red and blue sensitive emulsions we could get color indexes. This was the grunt work on which all stellar astronomy was based. And there were thousands of stars on each plate.

    It was mind-numbing, back-breaking work, but someone had to do it. It turned out grad students were perfect for the job. I guess its all done with scanners and digitizers these days, but back then it was all manual labor. The stone slab the measuring engine was placed on was quarried specially in Georgia with that purpose in mind and the special sand that absorbed vibrations from the nearby highway came from South Africa. And we lubricated the screws that measured stellar positions with spermaceti oil…

    But what I remember most was that scattered among the highly magnified grainy stellar images on the emulsion (they looked like globular clusters!) was the occasional streak where the time exposure had revealed a tiny, distant, probably never before observed asteroid.

    • Alex Tolley May 2, 2021, 21:05

      Live long enough and almost any science or technology discipline will have similar stories. At the time we didn’t know any different so that time-consuming and tedious tasks seemed reasonable. As technology marched on, tasks got easier, or the scope expanded, or the task just disappeared.

      It makes one wonder what tasks we accept today as “cutting edge” and hard will be like in another half-century, and which will become obsolete due to new technology obviating their need entirely. Some we can guess at, but others will likely blindside us when they appear.

      • Michael Fidler May 3, 2021, 10:01

        Yea, but it is getting a little weird when my Android tablet tells me the next morning about everything I looked at on my PC the night before! Pretty soon it will be telling me things before I even thought of it… }

        • Alexander Tolley May 3, 2021, 20:09

          It has been suggested (only half in jest?) that Amazon will predict what you want before you know it and send you the items. With really good predictions, the return costs should be outweighed by the profits. This is the sort of hyper-consumerism satirized by Pohl and Kornbluth back in the mid 20th century.

          I will settle for tools that remove the burden of effort from a task. The best example I have of this over my lifetime was the online search for journal articles that replaced the immensely tedious citation searches and roaming the academic library stacks and even having to get an interlibrary loan for a book or important paper. What was once considered a major problem for experimenters has been almost completely solved. AIs that can read articles and integrate facts from a large number of papers will remove much of the eyeball and attention time still used. No doubt AIs will be able to advise on hypotheses to test, collect the data, do the analysis and present the findings, rather like the Star Trek computers. They will even write the papers and submit them for publication. ;)

    • Patient Observer May 11, 2021, 20:31

      A lot of Quality Control work is done to this day on granite slabs with precision micrometers, “joe” blocks, digital height gauges, go/no go gauges, etc.. This technology is used along with the latest laser scanners and digital “arms”. Frankly, I trust the old technology and insist that the modern scanners and digitizers be calibrated frequently against known standards.

  • Spaceman May 2, 2021, 14:55

    I had never heard of this neat mission until your post, Paul. Thank you for helping to keep us in the know. This mission, in addition to not being too far in the future, is super-exciting!

  • ljk May 3, 2021, 10:25

    Paul Gilster wrote:

    “The name is now fixed — a list of IAU-approved star names as of January 1st, 2021 shows Alpha Centauri A as Rigel Kentaurus and Centauri B as Toliman, so what goes around comes around. Centauri C’s name, according to the IAU, remains the familiar Proxima Centauri.”

    So does this mean we will now have to refer to this blog as Kentaurus-Toliman Dreams? :^)

    Sorry, IAU, but it will always be the Alpha Centauri system to me.

  • Spaceman May 3, 2021, 11:23

    So, Toliman is another effort to ferret out terrestrial-sized exoplanets around alpha centauri A and/or B. Would you expect Toliman to be operational before or after Project Blue?

  • Martin May 3, 2021, 20:25

    Just barely on-topic:
    It’s Rigil Kentauris for alpha centauri A and Toliman for alpha centauri B (Proxima is named rationally).
    How the heck do you pronounce them? I’m saying Rye-jul Kin-tauris, which is confusing as I also say rye-jul for Rigel, the blue supergiant. As for Toliman, I’m saying Tall-E-mon.
    -Martin (who believed in Brontosaurus)

  • ljk May 24, 2021, 13:56

    Venus views from NASA sun probe show potential of hitchhiking science instruments

    By Meghan Bartels 4 days ago