≡ Menu

TOLIMAN Targets Centauri A/B Planets

We talked about the TOLIMAN mission last April, and the renewed interest in astrometry as the key to ferreting out possible planets around Alpha Centauri A and B. I was fortunate enough to hear Peter Tuthill (University of Sydney), who leads the team that has been developing the concept, rough out the idea at Breakthrough Discuss five years ago; Céline Bœhm (likewise at the University of Sydney) reported on more recent work at the virtual Breakthrough Discuss session this past spring. We now have an announcement from scientists involved that the space telescope mission will proceed.

Eduardo Bendek (JPL) is a member of the TOLIMAN team:

“Even for the very nearest bright stars in the night sky, finding planets is a huge technological challenge. Our TOLIMAN mission will launch a custom-designed space telescope that makes extremely fine measurements of the position of the star in the sky. If there is a planet orbiting the star, it will tug on the star betraying a tiny, but measurable, wobble.”

Work on the mission heated up in April of this year, with scientists from the University of Sydney working in partnership with Breakthrough Initiatives, the Jet Propulsion Laboratory and Australia’s Saber Astronautics. The mission could revolutionize our view of Centauri A and B, according to Tuthill:

“Astronomers have access to amazing technologies that allow us to find thousands of planets circling stars across vast reaches of the galaxy. Yet we hardly know anything about our own celestial backyard. It is a modern problem to have; we are like net-savvy urbanites whose social media connections are global, but we don’t know anyone living on our own block… Getting to know our planetary neighbors is hugely important. These next-door planets are the ones where we have the best prospects for finding and analyzing atmospheres, surface chemistry and possibly even the fingerprints of a biosphere – the tentative signals of life.”

Image: The University of Sydney’s Peter Tuthill, project leader for TOLIMAN. Credit: University of Sydney.

Astrometry tracks the minute changes in the position of a star that are the result of the gravitational pull of a planet. Detection of tiny angular displacements of the star allows the planet’s mass and orbit to be recovered, and unlike the situation with both radial velocity and transit methods, the astrometric signal increases with the separation of the planet and star.

That takes us out to the orbital distance for an Earth-class planet to be in the habitable zone, even though the signal is tiny, in the range of micro-arcseconds for the Alpha Centauri binary. The astrometric signal from an Earth-class planet orbiting in the habitable zone of Centauri A is 2.5 micro-arcseconds; a similar planet around Centauri B is roughly half of that.

TOLIMAN uses what the team calls a ‘diffractive pupil’ lens that spreads out the starlight and allows scientists to eliminate systematic errors and clarify the underlying signal. The flower-like pattern enhances the detection of star movement without the need for field stars as references, eliminating the need for a large aperture (such stars demand a larger collecting area). The pattern also reduces noise levels in the detector. An online description of the TOLIMAN technology explains why the nearest stellar system makes an excellent target for these methods:

With the fortuitous presence of a bright phase reference only arcseconds away, measurements are immediately 2 – 3 orders of magnitude more precise than for a randomly chosen bright field star where many-arcminute fields (or larger) are required to find background stars for this task. Maintaining the instrument imaging distortions stable over a few arcseconds is considerably easier than requiring similar stability over arcminutes or degrees. Alpha Cen’s proximity to Earth means that the angular deviations on the sky are proportionately larger (typically a factor of ~10-100 compared to a population of comparably bright stars).

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.

The same description refers to TOLIMAN as a ‘modest astrometric space telescope,’ and the word ‘modest’ seems to apply in that this is a narrow-field instrument 30cm in diameter, with what proponents estimate is a fast build time on the order of 18 months. We might contrast the mission with existing astrometric missions like the European Space Agency’s space-based GAIA. The latter can make astrometric measurements in the 10s of micro-arcseconds, which basically means it is capable of detecting gas giants. TOLIMAN takes us into the realm of much smaller, rocky worlds. Because it has no need of a large aperture, it is small, inexpensive and, obviously, tightly focused on a nearby system rather than surveying a large star field.

Image: Simulated image of the Alpha Centauri system, as could be viewed by the TOLIMAN telescope. Credit: Peter Tuthill.

TOLIMAN will receive spaceflight mission operations support from Saber Astronautics, including satellite communications and command. Saber’s involvement, says Tuthill, is “a critical part of the mission.” The company has received A$788,000 from an Australian Government International Space Investment: Expand Capability grant for the telescope’s design and construction, and I rather like the spirit in CEO Jason Held’s comment on TOLIMAN:

“TOLIMAN is a mission that Australia should be very proud of – it is an exciting, bleeding-edge space telescope supplied by an exceptional international collaboration. It will be a joy to fly this bird.”

As to when we can expect the bird to fly, Tuthill speaks of launch by 2023. We might know by mid-decade whether an Earth-size rocky planet orbits Centauri A or B. Habitable zone orbits are possible around both stars.

An early description of TOLIMAN is Tuthill et al., “The TOLIMAN Space Telescope,” Proc. SPIE 10701, Optical and Infrared Interferometry and Imaging VI, 107011J (9 July 2018). Abstract.


Comments on this entry are closed.

  • Michael Fidler November 17, 2021, 17:08

    I wonder if something like this could be done from the ground via Lucky Imaging? The smaller amateur telescopes have almost the same level of resolution when high speed planetary imaging cameras are stacked and processed. I’ve been working with an automated telescope control system called ASIAIR Plus that completely automates the imaging and has wifi so you can monitor from inside the comfort of your home.


    All that needs to be added is the ‘diffractive pupil’ lens.

    One problem that needs improvement in the programs and planetarium software used for GOTO in amateur telescopes is that there are no listings or “Tonight’s Best” objects for the nearest stars to earth. The majority of which are M Dwarfs that are not visible to the naked eye. We could be seeing ET optical signals and monitor flaring with such an improvement and would also publicize how little we know or understand about our nearest neighbor’s. A dedicated listing for goto telescopes of nearby stars out to the Trappist 1 planetary system at 40 lightyears would defiantly inspire amateur astronomers to observe and image these stars.

    • Ron S. November 17, 2021, 20:23

      “I wonder if something like this could be done from the ground…”

      The paper Paul linked to (halfway through the article) explains why, and also why existing space based astrometry instruments are inadequate. The resolution is too poor for the former by about 2 orders of magnitude and about 1 order of magnitude for the latter.

      • Michael Fidler November 18, 2021, 15:07

        Ron S. Let see if you are on the same page as what I am suggesting, put a half meter telescope in the Australian outback desert, using the “diffractive pupil’ lens” and stare at Alpha Cen for eight hours each night. Useing lucky imaging to have two hours of perfect Interferometry images each night with maybe AI improving the results. Beating the earths atmospherics distortions is not that difficult in the sub one meter telescopes as has been shown many times with images on par with the Hubble telescope. Ten half meter telescopes would cost little with off the shelf commercial amateur scopes.

        • Ron S. November 19, 2021, 10:39

          My reference to the order of magnitude issue is *with* correction methods to atmospheric distortion. You cannot get blood from a stone — no amount of analysis can derive a signal that isn’t there. There is no path to get from 100 micro-arc second resolution to 2 micro-arc second resolution from ground based instruments, and it may never be possible. We can’t even do it yet, in the general case, for space-based instruments. TOLIMAN appear to be able to do it by focusing on solving a *specific* resolution challenge rather than attempting it for *all* classes of targets.

  • Afterthought November 17, 2021, 22:11

    Good news

  • Thomas R Mazanec November 18, 2021, 9:26

    Time to put TOLIMAN on the list?

    • Paul Gilster November 18, 2021, 9:43

      Yes, its now among the links. Will re-link to it when they have their own website.

      • Antonio November 18, 2021, 14:49

        There are a lot of non-working links, BTW.

        • Paul Gilster November 18, 2021, 20:11

          I assume you mean in the sidebar. Yes, I need to go through those. Have fallen behind, but will try to get to it soon.

  • Ashley Baldwin November 19, 2021, 12:09

    So if TOLIMAN launches and becomes operational in 2023 as planned -with a three year observation mission it will find down to ( and hopefully below) Earth mass planets should any exist around A or B. By 2026. Should one or more be terrestrial sized and in a hab zone of either star , then what next ? A huge discovery but one that has to be followed up.

    The aspirational Decadal survey proposes a bells and whistles 6.5m exoplanet telescope – in thirty years . That’s a long wait.

    So does Bendek ( a big contributor to TOLIMAN) dust off his 0.45m ACESat Alpha Centauri imager for another NASA Medium Explorer bid ? Do Breakthrough dust off their similar but smaller and cheaper CENTaUR ? Rudimentary but effective . An exoplanet imager via the ESA M6 maybe. A combined effort or NASA find the money for a star-shade Roman telescope rendezvous mission ? The best option of the lot as I see it though conspicuously not listed as a potential probe mission in Decadal. If none of those occur what’s to stop a Chinese telescope ?

    Any hab zone terrestrial planet in that system CANNOT be ignored for thirty years. If TOLIMAN comes up trumps there will be big consequences. Scientifically and geopolitically . .

    • FrankH November 22, 2021, 16:19

      The Alpha Cent. system is evolving off the main sequence, so a planet in the HZ of A *now* may have been outside the HZ for most of the star’s life. Same with a planet outside the inner limit of the current HZ; it may have been Earth-like a few 100 million years ago.

      This is less of a concern for B, although if it’s luminosity has increased by a few % in the last few million years, it too may have planets just recently entering or exiting its HZ.

  • Spaceman November 19, 2021, 12:24


    I recently read about how a search for Proxima b using TESS did not find any evidence for transits. Does this result contradict the RV results which show evidence for a terrestrial planet, or, does this result merely mean that the orientation is not favorable for observing a transit of Proxima b?

    • Paul Gilster November 19, 2021, 12:38

      I’ll be writing about the transit finding next week. The result is not contradictory because RV methods can find a planet that does not transit — in other words, transits do indeed depend on the orientation of the planetary orbit so we can see the dip in lightcurve from the star. Only a small percentage of planets will happen to transit, but RV can find many that do not.

    • Ashley Baldwin November 19, 2021, 13:31

      For an exoplanet to transit it’s star, as seen from Earth , the plane of that planet’s orbit must be edge on ( inclined at 90 degrees ) to the plane Earth’s orbit – give or take a degree or so either way. Any transiting planet must block enough stellar radiation to be picked up by a photometer – the so called ‘ transit depth’ or dip seen in the photometer readout . Bigger planets, closer in around smaller stars give deeper transits. Thus leading to a bias towards such planets discovered by this method . Thanks to Kepler and now TESS, this is the leading source of the 4000 plus exoplanets discovered to date. RV spectroscopy , the second leading source of planets, favours a similar planetary population.

      So any interpretations drawn from the current exoplanetary population need to be made with great caution . Not that that has stopped anyone from trying !

      • Michael Fidler November 20, 2021, 7:04

        I’ve heard different figures for the possible percentage of stars that would have transits, something like 7% but do you have a accurate figure. How does it vary from star type, are there more around smaller stars? Do the brighter stars wash the transits out, never heard of transits around O type stars…

        • Paul Gilster November 20, 2021, 8:50

          I believe the figure is something like 2.7 percent of stars that have transits, but that’s just off my own faulty memory, so Ashley or anyone else feel free to correct.

          • Ashley Baldwin November 20, 2021, 12:08

            Thanks Paul –

            the easiest way to look at transit probability is expressed by the equation P=R*/a , where P is the probability of a transit , R* is the radius of the star (expressed in astronomical units, AUs) and a is the exoplanetary semi major axis ( also expressed in AUs ) . This assumes that all exoplanetary orbits are inclined randomly in relation to the Earth’s ( there is some evidence they might not be).

            By this method Proxima b, a= 0.0485 AU , orbiting Proxima Centauri ( 0.1592 Rsun) has a transit probability just short of 1.5%.

            • Henry Cordova November 23, 2021, 8:42

              ” This assumes that all exoplanetary orbits are inclined randomly in relation to the Earth’s ( there is some evidence they might not be).”

              Please tell me more about this, what is the evidence that leads us to this suspicion? Why should the inclinations of exoplanet orbits be affected by or correlated with Earth’s orbital inclination in any way?

              I understand that we don’t see transits unless the exoplanet orbital plane is aligned to our line of sight, but why should our orbital plane inclination have anything to do with that?

              What am I missing here? Is this some kind of selection effect, or are we talking about some causal connection or unexpected correlation?

              • Ashley Baldwin November 23, 2021, 14:59

                Orbital inclinations of currently known exoplanets are not randomly distributed. Transiting planets for the obvious reason that their orbits need to be inclined at 90 degrees or very close as seen from Earth – edge on – to be observed . But less intuitively, for RV photometry discovered exoplanets too.

                The RV method measures the backward and forward gravitational tug exerted on a host star by an orbiting exoplanet – in a straight line as seen from Earth. If the planetary orbital inclination is zero degrees I.e in the plane of the sky – this effect will be zero.Unmeasurable. As the inclination increases towards 90 degrees however, the RV effect will mount until maxing out at 90 degrees. Edge on again. RV spectroscopy has finite sensitivity so it can be seen that the nearer an orbital inclination is to 90 degrees the larger and thus more easily measurable the RV effect. Introducing an observational bias .So planets discovered by this method are ALSO more likely be inclined near 90 degrees and so more likely to transit. As much as 20% above random so far. Bear in mind that the vast majority of the 4000 plus known exoplanets to date were discovered by one or other of the two methods above and you see how the non random inclination trend has arisen.

                Just to further complicate things further , for some reason RV planet masses aren’t randomly distributed either. With clumps occurring around Jupiter mass, Neptune and sub Neptune mass and also Super Earth. Whether this is a true reflection of exoplanet mass distribution or simply another artefact of observation remains to be seen. Hopefully things will become clearer as large microlensing surveys ( a different and the easiest other way – after transit photometry and RV spectroscopy – of compiling a survey of large numbers of exoplanets ) are completed.

                Hope this is helpful.

  • Michael Fidler November 22, 2021, 22:54

    A new AI method for finding Exo Planets called ExoMiner!!!

    New Deep Learning Method Adds 301 Planets to Kepler’s Total Count.


    ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier to Mine Exoplanets.

    “The kepler and TESS missions have generated over 100,000 potential transit signals that must be processed in order to create a catalog of planet candidates. During the last few years, there has been a growing interest in using machine learning to analyze these data in search of new exoplanets. Different from the existing machine learning works, ExoMiner, the proposed deep learning classifier in this work, mimics how domain experts examine diagnostic tests to vet a transit signal. ExoMiner is a highly accurate, explainable, and robust classifier that 1) allows us to validate 301 new exoplanets from the MAST Kepler Archive and 2) is general enough to be applied across missions such as the on-going TESS mission. We perform an extensive experimental study to verify that ExoMiner is more reliable and accurate than the existing transit signal classifiers in terms of different classification and ranking metrics. For example, for a fixed precision value of 99%, ExoMiner retrieves 93.6% of all exoplanets in the test set (i.e., recall=0.936) while this rate is 76.3% for the best existing classifier. Furthermore, the modular design of ExoMiner favors its explainability. We introduce a simple explainability framework that provides experts with feedback on why ExoMiner classifies a transit signal into a specific class label (e.g., planet candidate or not planet candidate).”