Project Blue: Imaging Alpha Centauri Planets

by Paul Gilster on November 15, 2016

We know about an extremely interesting planet around Proxima Centauri, and there are even plans afoot (Breakthrough Starshot) to get probes into the Alpha Centauri system later in this century. But last April, when Breakthrough Initiatives held a conference at Stanford to talk about this and numerous other matters, the question of what we could see came up. For in Alpha Centauri, we’re dealing with three stars that are closer to us than any other. If there are planets around Centauri A and/or Centauri B, are there ways we could image them?

This gets interesting in the context of Project Blue, a consortium of space organizations looking into exoplanetary imaging technologies. This morning Project Blue drew on the work of some of those present at Stanford, launching a campaign to fund a telescope that could obtain the first image of an Earth-like planet outside our Solar System, perhaps by as early as the end of the decade. The idea here is to ignite a Kickstarter effort aimed at raising $1 million to support needed telescope design studies. A $4 million ‘stretch goal’ would allow testing of the coronagraph, completion of telescope design and the beginning of manufacturing.


Project Blue thinks it can bring this mission home — i.e., launch the telescope and carry out its mission — at a final cost of $50 million (the original ACEsat was a $175 million design). The figure is modest enough when you consider that Kepler, which has transformed our view of exoplanets, cost $600 million, while the James Webb Space Telescope weighs in at $8 billion. About a quarter of the total cost, according to the project, goes into getting the telescope into orbit, which will involve partnering with various providers to lower costs.

But Project Blue also hopes to build a public community around the mission to support design and research activities. Jon Morse is mission executive for the project:

“We’re at an incredible moment in history, where for the first time, we have the technology to actually find another Earth,” said Morse. “Just as exciting — thanks to the power of crowdfunding — we can open this mission to everyone. With the Project Blue consortium, we are bringing together the technical experts who can build and launch this telescope. Now we want to bring along everyone else as well. This is a new kind of space initiative — to achieve cutting-edge science for low cost in just a few years, and it empowers us all to participate in this moment of human discovery.”

I go back to last April because it was at Stanford that I saw Eduardo Bendek’s model of a small space telescope called ACEsat, which was conceived at NASA Ames by Ruslan Belikov and Eduardo Bendek and submitted (unsuccessfully) for NASA Small Explorer funding. Belikov had gone on to present the work at the American Astronomical Society meeting in 2015 (see “How to Directly Image a Habitable Planet Around Alpha Centauri with a ~30-45cm Space Telescope,” available here). You’ll recall that Ashley Baldwin wrote up the concept in superb detail on this site in December of that year as ACEsat: Alpha Centauri and Direct Imaging.

Now we have Project Blue, which has connections to the BoldlyGo Institute, its offshoot Mission Centaur, the SETI Institute and the University of Massachusetts Lowell. The aim is to launch a space telescope with a 45-50 centimeter aperture, looking for potentially habitable planets from 0.5 to 1.5 AU within the habitable zones of both Centauri A and B. The ultimate hope, then, is to ‘see blue’ — meaning oceans and atmosphere, a world on which life could emerge. This is Sagan’s ‘pale blue dot,’ only now it’s not our own planet but an Earth 2.0.

The Project Blue space telescope would spend two years in low Earth orbit accumulating image after image — hundreds, thousands, tens of thousands — as a way of teasing out its faint targets. When it comes to ‘another Earth,’ Centauri A and B up the ante on Proxima Centauri. The Proxima planet may well be habitable, but a true Earth analog is not going to be tidally locked to its star, as Proxima b probably is, and it’s not going to orbit a red dwarf.

Neither Belikov or deputy principal investigator Eduardo Bendek are formally connected to Project Blue, but their work in the form of papers and conference presentations feeds directly into the concept driving the project. The original mission now cedes the floor to the private sector, whose job it will be to raise enough cash to support the development of the needed coronagraph to filter out the light of two very close stars, along with other key flight hardware elements. The next step, though, long before building flight hardware, is to finalize the telescope design.

The new Kickstarter campaign will pay for analysis, design, and simulations, but Project Blue has an eye on other partnerships as well as wealthy donors and foundations. Usefully, the project should be able to test coronagraph technologies similar to those being considered on much larger space instruments currently under study by the major space agencies, thus providing a useful testbed for such designs. To make this work, everything must fall into place — the coronagraph for starlight suppression, a deformable mirror to feed the coronagraph and rock-solid stability. No aspect can be allowed to fail if the mission is to achieve its goal.

If the Project Blue planners are correct, we can solve the attendant problems and get this mission into space is as little as 4 to 6 years. The goal is hugely ambitious but it also opens the door to citizen-science, with private donors contributing to an instrument that will not be the result of a government program or a for-profit commercial space effort. The initial Kickstarter campaign is designed to bolster the technical groundwork needed for the telescope, but stretch goals could see publicly funded flight component manufacturing.

Looking for Earth-like planets around other stars is like looking for bioluminescent algae next to a lighthouse. But I keep coming back to that Breakthrough Discuss meeting in Stanford, because I remember Ruslan Belikov telling his audience that the key advantage of Alpha Centauri is how large the habitable zones around its component stars appear in terms of angular size. We would need a significantly larger instrument to attempt something similar around other nearby stars. The Alpha Centauri stars are nature’s gift, and it’s one we would do well to exploit. Check the Kickstarter page for more on this low cost, high impact idea.

For more on the technical background of the ACEsat concept, see Belikov et al., “How to Directly Image a Habitable Planet Around Alpha Centauri with a ~30-45cm Space Telescope” (preprint) and Bendek et al., “Space telescope design to directly image the habitable zone of Alpha Centauri” (preprint).


{ 28 comments… read them below or add one }

DJ Kaplan November 15, 2016 at 13:36

By the time funding is sorted out, the angular separation between the two components should be at their maximum. Perfect!


Alex Tolley November 15, 2016 at 16:24

While Alpha Centauri Bc may exist (but needs to be confirmed?) we don’t even know if it would be a habitable world with oceans. Therefore isn’t Project Blue jumping the gun trying to image a world with an ocean before we are fairly sure one can exist?

Should Project Blue be funded and even launched, is it sensitive enough to image planets at the nearer stars out to some distance so that it isn’t just a single use telescope? Could it be repurposed for other tasks instead as a fallback plan?


DJ Kaplan November 15, 2016 at 20:35

I’d hope that the mission were to find out what’s out there, rather than to confirm hopes and dreams.


Andrew LePage November 16, 2016 at 9:26

> Therefore isn’t Project Blue jumping the gun trying to image a world with an ocean before we are fairly sure one can exist?

Absolutely not! The point of this project is NOT to observe planets we already know exist orbiting Alpha Centauri A or B (and, at this time, there are no confirmed planets known to be orbiting either star). The whole point is to search for planets as small as Earth in or near the habitable zone of these two Sun-like stars. I do not know the details of this proposed mission’s CONOPS, but in principle it could be used to search for planets around any of the nearby stars (although it might not necessarily be able to spot Earth-size planets in the HZ of these other stars).

> While Alpha Centauri Bc may exist (but needs to be confirmed?)

Well, this is news to me! I know that a transit-like signal consistent with an Earth-size exoplanet in a close orbit was observed by Hubble when the telescope was searching for transits of Alpha Centauri Bb (which itself is now suspected of not existing), but the planetary nature of this event has not been confirmed, never mind rise to the point where it would receive a designation like “Alpha Centauri Bc”:

*IF* this planet exists with anything like the size and orbit it is suspected of having, its orbit will be too small by a factor of several for this proposed telescope to observe it. In the near term, photometry will be needed to search for more transit events.


Alex Tolley November 16, 2016 at 13:59

From Project Blue’s website:

Project Blue will build and launch a compact space telescope with a 45-50 centimeter aperture to directly image the habitable zone of the nearest star system to Earth – Alpha Centauri A and B – with the goal of revealing the closest potentially habitable planet beyond our Solar System. Project Blue aims to image “Earth-like” planets (roughly 0.5 to 1.5 times of the size of Earth) orbiting within either star’s habitable zone and possessing an atmosphere that could allow liquid water to exist on its surface. If such a planet were to be just like the Earth, with oceans and such an atmosphere, we might just “see blue”

Possible outcomes:

There is:
1. no planet around either star.
2. a planet[s] but it isn’t in the HZ
3. a planet[s] in the HZ, but it doesn’t have surface water
4. a planet[s] with surface water (and possibly multicellular life)

It seems to me that only outcome 4 would be a successful outcome for the mission planners. Therefore it seems to me that using a cheaper technique to confirm a suitable planet is present should be attempted first before committing funds for a telescope dedicated to this single mission. If outcomes 1-3 are the result, then what? Can this telescope be used for other missions, or will it just be a working white elephant?

While Alpha Centauri Bc may exist (but needs to be confirmed?)

I was going off the Wikipedia entry. If it is wrong, perhaps you could correct it so that we all have more accurate information from this source.


Andrew LePage November 16, 2016 at 15:45

> If outcomes 1-3 are the result, then what?

Then we know something that we did not know before we built and flew this telescope. I honestly don’t see what the problem is here. The whole point of flying this telescope is to see what, if any, exoplanets about the size of the Earth or larger may exist in or around the HZ of these two stars. If we don’t look, then we won’t know if these planets exist or not. You seem to be only willing to fly this mission if we already know it will produce a desired result even though we have no information to make such a prediction unless we fly the mission.

> Can this telescope be used for other missions, or will it just be a working white elephant?

Ask Project Blue. Like I said before, I am not part of this project and I am not privy to their plans, their observation CONOPS, etc., however, *in principle* this same telescope could be used to search for planets orbiting other stars (but not necessarily capable of finding Earth-size planets in the HZs of other potential target stars). Whether or not it is will be used for observing other targets depends on the details of their observing schedule and the capabilities of the spacecraft.


Alex Tolley November 16, 2016 at 18:56

Andrew, you know about astronomy far better than I do. The implication of what you are saying is that there is no other way to detect planets in this system for far lower cost The project is spending $50m to image a planet for just one star system. This seems like a waste of resources to me. Will no other technique work that would at least indicate a suitable planet to image, even a giant planet? I appreciate the funds are private and therefore not a use of tax payer funds, but given the goal for the project, it seems very iffy to me. You imply that it is worth building this even if there is no planet detected. I don’t understand that thinking when that level of funding is 50% of either Breakthrough initiative, either of which will generate far more value in technology even if they fail (admittedly after requiring much more for Starshot).


Andrew LePage November 17, 2016 at 10:54

> The project is spending $50m to image a planet for just one star system. This seems like a waste of resources to me.

I couldn’t disagree more!!! It may be one star system but it contains the two closest Sun-like stars with the best chances of us being able to spot Earth-size planets with a modest instrument. In my mind this is a necessary precursor to any mission we may send to this system. And in the grand scheme of things, $50 million dollars is pretty inexpensive even for an instrument targeting a single star system (which, I like I have earlier, may not be the case – in principle, this telescope could search other nearby stars for exoplanets e.g. maybe Jupiter analogs orbiting Tau Ceti). And there is no reason to believe that spending $50 million on this project will take away resources from completely unrelated programs like Starshot.

> Will no other technique work that would at least indicate a suitable planet to image, even a giant planet?

Radial velocity surveys have already excluded the possibility of Saturn to Jupiter sized worlds orbiting either Alpha Centauri A or B.

But as the episode with Alpha Centauri Bb illustrates, astronomers are reaching the practical limits of what the radial velocity (RV) technique can do with the Alpha Centauri system. While future RV systems with even higher measurement precision are in the works, it looks more and more as though natural noise or “jitter” irregularly sampled in time will not allow astronomers to detect planets all that much smaller than Neptune in or near the HZ of Alpha Centauri. It is becoming increasingly apparent that direct imaging techniques offer us the only practical means of reliably detecting Earth-size planets in the HZ of nearby Sun-like (unlike the transit method which can detect only tiny fraction of such worlds). The Blue Dot project (or something like it) attempting to observe the nearest Sun-like stars is a logical first step in this search.


Ashley Baldwin November 18, 2016 at 14:21

It was because of the narrow focus of this mission that it’s antecedent ,ACEsat , failed in aN Small Explorer bid . With limited budgets ,risk averse Nasa just couldn’t grant the $180 million. ( sadly before Proxima b was discovered which might have helped ) The design is also novel in that the PIAA coronagraph used ( a design that is back up for WFIRST and has the advantage of a much higher light throughout to the optical plane , upwards of 90%, a lot for a small telescope compared to less than 30% for the Lyot design on WFIRST ) has its first two optical elements build in to the rigid silicon carbide telescope wall to provide additional stability on top of more conventional deformable mirrors , which have come on a lot recently . Vitally as stability is a real issue and significant limiting factor for coronagraph based direct imaging .

The original ACEsat design was proposed as having a 35-45cm aperture . I too think the idea is well worth it with the caveat on how realistic is it to build what is still a reasonably sized telescope ( significantly larger than say € 150 CHEOPS) and with a coronagraph ( a piece of state of the art technology that does not come cheap on its own even for a relatively small telescope ) for just $50 million . Not including launch and running costs .

Silicon carbide is much cheaper than conventional fabricants though and can make both high quality mirrors and strong structural material. Nasa don’t seem to enthuse about it but it has already been demonstrated as very effective in the Herschel space telescope and Gaia too with Euclid to come . Private sector efficiencies or not it’s going to need a private philanthropist too .

ACEsat proposed observing at five different wavelengths to look for Raleigh scattering and methane etc, but with no spectrograph . Belikov and Bendek also developed a unique piece of software that allowed each star to be imaged repeatedly as Paul says whilst excluding the corrupting light of the other ” stellar lighthouse ” as each of the binary are examined individually. A year ago they had it at a low Testing readiness level, TRL , of just 3 . Work to be done . A must too , so I wonder if they will contribute this unique but crucial technology and how much ,if at all , it has been developed ( the PIAA has had lots of work already thanks to WFIRST) . In terms of the coronagraph , I know Boldlygo have already extensively tested the PIAA coronagraph for their own proposed 1.8m scope, ASTRO-1. The high throughput of this design is just what you want with a limited aperture when every photon counts . It’s biggest limitations are the precision aspheric mirrors it requires and a small bandwidth though that shouldn’t matter here without a spectrograph .

In terms of other uses smaller ACEsat did have a proposed secondary mission of viewing nearby bright stars like Altair and Procyon . This is where your really philanthropist really comes in as running costs add up over years too ( I’m surprised Breakthrough starshot haven’t partnered up too ) . I’ve heard estimates as high as 30 % of a hab zone planet around one of the stars and I hope that Alpha Centauri Bb does prove false as the closish proximity and high binary eccentricity of the two stars doesn’t allow for many stable orbits or far out for multiple planetary systems round either star though with less than half the luminosity . Alpha Cent B might be a slightly better bet given a hab zone of about 0.48-0.7 AU but not helped if there is indeed a terrestrial planet on top of the star . Seems to have lost a lot of traction though recently though.


Ashley Baldwin November 20, 2016 at 10:41

Apologies . K2 Alpha Centauri B has a relative luminosity 0.5 relative to the Sun , so according to the hab zone equation , Hab zone, H , expressed in AUs = 0.95 ( square root luminosity ) – 1.4 ( square root luminosity )= 0.67 – 0.98 AU. For main sequence stars (of which G2 Alpha Centauri A most definitely is still a member even for its oldest age estimations ) luminosity is thus key at any point ( based around Kasting’s solar system / liquid surface water premis etc ) .

At 1.25 L Sun, Alpha A’ s current hab zone is thus 1.06 AU – 1.56AU. Well within the stable boundary of at least a one planet system though not necessarily the Project Blue observation window . It’s got to be a no brainer if it can be done for ” just ” $50 million. Two sun like , sun aged stars , as near as you can get. Too tempting . What’s the point of peeking at any more distant exosolar planets if you don’t take a look in on the potentially friendly neighbours first ?

The risks are worth the benefits if ANY planets are seen and anything terrestrial in either of the hab zones blows even Proxima b away. ( though I still hope the speculative early plans to adapt SPHERE /ESPRESSO at the VLT for its early characterisation means we don’t have to wait for the ELTs. Watch that space )


FrankH November 16, 2016 at 16:05

The A & B stars are evolving off the main sequence, so for at least the A component, outcomes 3 & 4 may mean that there was a habitable planet that is now dead and bone dry, or that a formerly frozen world is now warm enough to have liquid water, but probably no life.


Ivan Vuletich November 16, 2016 at 20:54

The formerly frozen possibility, would make any Earth-like planet an almost perfect terraforming target.
– Earth size
– Climate allowing liquid water
– No life to worry about (surface life anyway)

While the period of habitability would be too short for life to evolve, it would be more than long enough in human terms, if we are ever able to accomplish such feats.


Andrew LePage November 17, 2016 at 11:03

> The A & B stars are evolving off the main sequence,

No, they are not. The estimated age of the stars in the Alpha Centauri system is about six billion years. Alpha Centauri A still has another billion or so years left on the main sequence and B probably has over ten billion years left. While these stars, like ALL main sequence stars including the Sun, will slowly brighten over time as helium builds up in their cores, they are not evolving off the main sequence… yet.


Geoffrey Hillend November 15, 2016 at 17:54

I like the idea but not the small size of the mission. Why not just build a large telescope with a star shade and ability to image exoplanets in all nearby stars rather than make one that is only good enough for Proxima Centauri. Then we can really find the blue dot somewhere if Promixa B turns out to not be a blue dot.


Neil Stahl November 15, 2016 at 19:58

So how does cost scale with diameter here?


Andrew LePage November 16, 2016 at 9:29

The rule of thumb that is used in my realm (i.e. small to moderate-size optical and IR instruments for space-based remote sensing applications), instrument cost scales approximately with the cube of diameter.


Ashley Baldwin November 18, 2016 at 14:29

Does that apply for silicon carbide though ? It comes out much cheaper than say ULE and conventional construction materials as well as being much lighter thus reducing the size of any launcher .


Andrew LePage November 18, 2016 at 18:11

The original commenter asked for a scaling law. One of the implied assumptions of any such law is that essentially the same materials and technology are being used when scaling from one size instrument to another.


Ashley Baldwin November 20, 2016 at 11:15

Quite correct. Im familiar with Stahl’s comprehensive work in this area in particular . I’m not a physicist or an expert at costing models but I’m lucky enough to know some health economy academic statisticians . One of these has an astrophysicist son. They have looked at the Nasa space telescope costing models in detail and found them to lack the sophistication or accuracy of models in other fields and be somewhat pessimistic cost wise , I.e potentially over estimating the cost / aperture ratio particularly .

That said the perennial issue of “mission creep” always lies around the corner. The issues seem to be in establishing clear and realistic goals AND technological readiness prior to building as was classically not done robustly with the JWST . With disastrous consequences ( and may even have contributed to some of the subsequent conservative space telescope costing models) . That applies here too of course with still no practical experience with the central technology of ultra high contrast coronagraphs, high performance deformable mirrors or binary exoplanet imaging software .

I would also argue however that the telescope aperture equation does not and was not intended to apply to silicon carbide construction materials though given Nasa’s apparent mistrust of the material . ACEsat if successful would have been the first such telescope for Nasa . Hopefully Project Blue might yet allow the US to maintain its space telescope hegemony in this evolving area.


David Herne November 16, 2016 at 1:55

‘The Alpha Centauri stars are nature’s gift…’ Paul, how true! (Even if just a system to fire the imagination.) I haven’t checked the papers, interested to know if the coronagraph is internal or external to the telescope?


David Herne November 16, 2016 at 2:06

Got the answer to my coronagraph question thanks in ‘ACEsat: Alpha Centauri and Direct Imaging’ article from last December. Going back over it.


ljk November 16, 2016 at 14:45

Discovery Of A Nearby Super Earth With Only 5 Times Our Mass

Article Updated: 11 November 2016

by Matt Williams

Red dwarf stars have proven to be a treasure trove for exoplanet hunters in recent years. In addition to multiple exoplanets candidates being detected around stars like TRAPPIST-1, Gliese 581, Gliese 667C, and Kepler 296, there was also the ESO’s recent discovery of a planet orbiting within the habitable zone of our Sun’s closest neighbor – Proxima Centauri.

And it seems the trend is likely to continue, with the latest discovery comes from a team of European scientists. Using data from the ESO’s High Accuracy Radial velocity Planet Searcher (HARPS) and HARPS-N instruments, they detected an exoplanet candidate orbiting around GJ 536 – an M-class red dwarf star located about 32.7 light years (10.03 parsecs) from Earth.

According to their study, “A super-Earth Orbiting the Nearby M-dwarf GJ 536“, this planet is a super-Earth – a class of exoplanet that has between more than one, but less than 15, times the mass of Earth. In this case, the planet boasts a minimum of 5.36 ± 0.69 Earth masses, has an orbital period of 8.7076 ± 0.0025 days, and orbits its sun at a distance of 0.06661 AU.

Full article here:


Harry R Ray December 11, 2016 at 14:55

Since TRAPPIST-1 was mentioned in the above comment, I find it not too OT to mention the following: Last Wedensday, Kepler K2’s campaign 11 ended. If all went according to plan(and I find NOTHING on the internet indicating that it hasn’t)ALL data from campaign 11 should have been downloaded by last Thursday, last Friday, all the protocols for campaign 12 should have been ploaded to the spacecraft, yesterday the spacecraft should have moved into position to be able to lock on to the campaign 12 field. Today and tomorrow, lock-on should be confirmed, and calibration testing should be concluded. That means, Tuesday, ACTUAL OBSERVATIONS OF THE TRAPPIST-1 system should begin. By the end of the week, transits of BOTH TRAPPIST-1b and TRAPPIST-1c should have been recorded! According to one astronomer, the AUTOMATED transit software on Kepler has only a 10% chance perceiving this data as a “HIT”, but, since the times of these transits have been determined with GREAT PRECISION, there is a FAR GREATER chance that the will be detected by EYEBALLING the data, which FINALLY brings me to the following: Gillon et al have obserevd the TRAPPIST-1 system THREE TIMES NOW(in January for 5+ hours, in February for 32+ hours, and in August for 45+ NEAR-CONTINUOUS hours. Why haven’t they published anything except for the b-c double transit prediction on May 4’th? They are noe in SERIOUS DANGER OF BEING SCOOPED BY KEPLER!


Tom Mazanec November 19, 2016 at 21:18

Well, I pledged $13 and posted links to the site on several fora.


EricSECT November 20, 2016 at 6:36

Would the James Webb Space Telescope be up to the task of trying to detect planets around the Alpha Centauri pair?

If so….. Would a better expenditure of resources be to spend the money on JWST observation time vice launching a dedicated satellite?


Ashley Baldwin November 21, 2016 at 14:45

A nice thought. There are speculative plans for it to pick planetary phases out of its spectra of Proxima/ b , but this will take a lot of observation time and seems the realistic limit . Unlike any Alpha Cent A/B planets it’s existence is proven too.

Can’t see $8.5 BILLION JWST’s precious and very expensive operating time per annum (likely considerably more than Project Blue’s $50 million construction alone , assuming an optimistic 10 % of JWST baseline cost pa- ) devoted to Alpha Centauri . It only has a five year primary mission , over a 150 moving parts and practically no hope of being serviced. It does have coronagraphs but they are orders of magnitude away from the necessary contrast , inner working angle and stability for the task of direct imaging many if any planets anywhere . Nor will it have the necessary binary companion light suppression Nyquist software . It was simply never designed with this task in mind . Project Blue is out of necessity only designed for this task. In space no one can hear you scream maybe , but in space , $50 million telescopes are peanuts .


Harry R Ray December 4, 2016 at 14:47

According to Andrew LePage, the Regil Kentaurus(UGH!!!) pair is TOO LUMINOUS in the IR to RISK observation with JWST without filtration, which, if APPLIED, would probably make imagining of ANY planets IMPOSSIBLE! However, I JUST thought of a possible way to get an image(actually TWO images) of Proxima b BEFORE E-ELT comes online! Here’s how: The Interferometer on the Very Large Telescope has ACTUALLY RESOLVED THE DISK of Proxima Centauri, so if they took TWO images with the interferometer EVERY ORBITAL PERIOD AT MAXIMUM SEPARATION on BOTH sides of the star from now to 2020(0r 2021, 2022, or 2023 ONLY if necessary) and added ALL of the images from one side SEPARATELY from adding all of the images from the OTHER side, thus BUILDING UP the images from BOTH SIDES, I would bet ALMOST EVEN MONEY that they could get VERY CRUDE images. The information they could extract from these images would not be very great, but just the HISTORICAL VALUE WOULD BE TREMENDOUS!


Tom Mazanec December 21, 2016 at 14:31

The project failed at just over a third of the goal.


Leave a Comment

{ 3 trackbacks }