≡ Menu

Space Telescopes Beyond Hubble and JWST

Ashley Baldwin tracks developments in astronomical imaging with a passion, making him a key source for me in keeping up with the latest developments. In this follow-up to his earlier story on interferometry, Ashley looks at the options beyond the James Webb Space Telescope, particularly those that can help in the exoplanet hunt. Coronagraph and starshade alternatives are out there, but which will be the most effective, and just as much to the point, which are likely to fly? Dr. Baldwin, a consultant psychiatrist at the 5 Boroughs Partnership NHS Trust (Warrington, UK) and a former lecturer at Liverpool and Manchester Universities, gives us the overview, one that hints at great things to come if we can get these missions funded.

by Ashley Baldwin


Hubble is getting old.

It is due to be replaced in 2018 by the much larger James Webb Space Telescope. This is very much a compromise of what is needed in a wide range of astronomical and cosmological specialties, one that works predominantly in the infrared. The exoplanetary fraternity will get a portion of its (hoped for) ten year operating period. The JWST has coronagraphs on some of its spectrographs which will allow exoplanetary imaging but as its angular resolution is actually lower than Hubble, its main contribution will be to characterise the atmospheres of discovered exoplanets.

It is for this reason that the designers of TESS (Transiting Exoplanet Survey Satellite) have made sure a lot of its most prolonged viewing will overlap with that of the JWST. Its ability to do this will depend on several factors such as the heat (infrared) the planet is giving out, its size and critically its atmospheric depth (the deeper the better) and the proximity of the planet in question. The longer the telescope has to “stop and stare” at its target planet the better, but we already know lots of other experts want some of the telescope’s precious time, so this will be a big limiting factor.

Planet Hunting in Space

The big question is, where are the dedicated exoplanet telescopes? NASA had a mission called WFIRST planned for the next decade, with the predominant aim of looking at dark matter. There was an add on for “micro-lensing” discovery of exoplanets that happened to pass behind further stars, getting magnified by the stars’ gravity and showing up as “blips” in the star’s spectrum. When the National Reconnaissance Mirrors (NRO) were recently donated to NASA, it was suggested that these could be used for WFIRST instead.

Being 2.4 m in diameter they would be much larger than the circa 1.5 m mirror originally proposed and would therefore make the mission more powerful, especially because by being “wide field” they would view far bigger areas of the sky, further increasing the mission’s potency. It was then suggested that the mission could be improved yet further by adding a “coronagraph” to the satellite’s instrument package. The savings made by using one of the “free” NRO mirrors would cover the coronagraph cost.

Coronagraphs block out starlight and were originally developed to allow astronomers to view the Sun’s atmosphere (safely!). Subsequently they have been placed in front of a telescope’s focal plane to cut out the light of more distant stars, thus allowing the much dimmer light of orbiting exoplanets to be seen. A decade of development at numerous research testbeds such as the Jet Propulsion Laboratory, Princeton and at the Subaru telescope on Mauna Kea has refined the device to a high degree. When starlight of all wavelengths strikes a planet it can be reflected directly into space, or absorbed to be re-emitted as thermal infrared energy. The difference between the amount of light emitted by planets versus stars is many orders of magnitude in the infrared compared to the visible, so for this reason telescopes looking to visualise exoplanets do so in the infrared. The difference is still a billion times or so.

Thus the famous “firefly in the searchlight “metaphor. Any coronagraph must cut out infrared to the tune of a billion times or more for an exoplanet to first be seen and then analysed spectroscopically. The latter is crucial as it tells us about the planet and its atmosphere according to the factors described above. This light reduction technique is called “high contrast imaging” with the reduction described according to negative powers of ten. Typically a billion times reduction is simplified to 10e9. This level of reduction should allow Jupiter size planets, ice giants like Neptune and, at a push, “super earths”. To visualise Earth like, terrestrial planets, an extra order of magnitude, 10e10 or better is necessary.


Image: A coronagraph at work. This infrared image was taken at 1.6 microns with the Keck 2 telescope on Mauna Kea. The star is seen here behind a partly transparent coronagraph mask to help bring out faint companions. The mask attenuates the light from the primary by roughly a factor of 1000. The young brown dwarf companion in this image has a mass of about 32 Jupiter masses. The physical separation here is about 120 AU. Credit: B. Bowler/IFA.

The Emergence of WFIRST AFTA

Telescope aperture is not absolutely critical (with a long enough view), with even small metre-sized scopes able to see exoplanets with the correct coronagraph. The problem is the inner working angle or IWA. This represents how close to the parent star its light is effectively blocked, allowing imaging with minimal interference. Conversely, the outer working angle, OWA , determines how far away from the star a planet can be seen. The IWA is particularly important for seeing and characterising planets in the habitable zone (HBZ) of sun-like stars. By necessity it will need to shrink as the HBZ shrinks, as with M dwarfs, which would obviously make direct imaging of any terrestrial planets discovered in the habitable zones of TESS discoveries very difficult. For bigger stars with wider HBZs obviously the IWA will be less of an issue.

So all of this effectively made a new direct imaging mission, WFIRST AFTA. Unfortunately the NRO mirror was not made for this sort of purpose. It is a Cassegrain design, a so-called “on axis” telescope with the focal plane in line with the primary mirror’s incoming light, with the secondary mirror reflecting its light back through a hole in the primary to whatever science analysis equipment is required. In WFIRST AFTA this would mainly be a spectrograph.

The coronagraph would have to be at the focal plane and along with the secondary mirror, would further obscure the light striking the primary. It would also need squeezing between the “spider’ wires that support the secondary mirror (these give the classic ‘Christmas tree star’ images we are all familiar with in common telescopes).

Two coronagraphs are under consideration that should achieve an image contrast ratio of 10 to the minus 9, which is good enough to view Jupiter-sized planets. Every effort is being made to improve on this and to get down to a level where terrestrial planets can be viewed. Difficult and expensive, but far from impossible. Obviously, WFIRST has quite easily the biggest mirror of the options under consideration by NASA and hence the greatest light intake and imaging range. It could also be possible to put the necessary equipment on board to allow it to use a starshade at a later date. The original WFIRST budget came in at $1.6 billion but that was before NASA came under increasing political pressure on the JWST’s (huge) overspend.

An independent review of cost suggested WFIRST would come in at over $2 billion. Understandably concerned about the potential for “mission creep”, seen with the JWST development, NASA put the WFIRST AFTA design on hold until the budgetary statement of 2017, with no new building commencing until JWST launched. So whatever is eventually picked, 2023 will be the earliest launch date. Same old story, but limited costs sometimes lead to innovation. In the meantime, NASA commissioned two “Probe” class alternative back up concepts to be considered in the “light” of the budgetary statement.

Exoplanet Telescope Alternatives

The first of these is EXO-C. This consists of a 1.5 m “off axis” telescope ( the primary mirror is angled so that the focal plane and secondary mirror are at the side of the telescope and don’t obscure the primary, thus increasing its light gathering ability). There are potential imaging issues with such scopes so they cost more to build. EXO-C has a coronagraph and a spectrograph away from the optical plane. The issue for this concept is which coronagraph to choose. There are many designs, tested over a decade or more with the current “high contrast imaging”( see above) level between 10e9 and 10e10. So EXo-C is relatively low risk and should at a push be able to even see some Earth or Super-Earth planets in the HBZs of some nearby stars, as well as lots of “Jupiters”.

The other Probe mission, even more exciting, is EXO-S. This involves combining a self propelled “on axis” 1.1m telescope with a “starshade”. The starshade is a flower-like (it even has petals) satellite — the choice of a flower shape rather than a round configuration reduces image-spoiling “Fresnel” diffraction from the starshade edges. The shade sits between the telescope and the star to be examined for planets. It casts a shadow in space within which the telescope propels itself to the correct distance for observation (several thousands of kms).

Like a coronagraph, the starshade cuts out the star’s light, but without the difficulty of squeezing an extra device into the telescope. The hard bit is that both telescope and shade need to have radio or laser communication to achieve EXACT positioning throughout the telescopic “stare” to be successful, requiring tight formation flying. The telescope carries propellant for between 3 and 5 years. With several days for moving into position, this is around 100 or so separate stop and stares. The shade concept means two devices instead of one although they can be squeezed into one conventional launch vehicle, to separate at a later point in the mission.


Image: The starshade concept in action. Credit: NASA/JPL.

The good news is that with a starshade the inner working angle is dependent on the telescope starshade distance rather than the telescope. The price of this is that the further apart the two are, the greater the precision placement required. The distances involved depend on the size of the starshade. For EXO-S’ 35 m starshade, this is in excess of thirty seven thousand kilometres. EXO-S, despite its small mirror size, will be able to view and spectrographically characterise terrestrial planets around suitable nearby stars and Jupiter-sized planets considerably further out.

Achieving Space Interferometry

“Formation flying” of telescopes is an entirely new concept that hasn’t been tried before, so potentially more risky, especially as its development is way behind that of coronagraph telescopes. If it works, though, it opens the gate to fantastic discovery in a much wider area than EXO-S. This is just the beginning. If you can get two spacecraft to fly in formation, why not 3 or 30 or even more? In the recent review I wrote for Centauri Dreams on heterodyne interferometers, I described how 30 or so large telescopes could be linked up to deliver the resolution of an telescope with an aperture equivalent to the largest gap between the unit scopes of the interferometer (a diluted aperture). The number of scopes increases light intake ( the brightness of the image) and “baselines” , the gap between constituent scopes in the array, delivering detail across the diluted aperture of the interferometer.

We’re in early days here, but this is heading in the direction of an interferometer in space with resolution orders of magnitude larger than any New Worlds telescope. A terrestrial planet finder yes, but more important, with a good spectrograph, a terrestrial planet characteriser interferometer. TPC-I. To actually “see” detail on an exoplanet would require hundreds of large space telescopes spread over hundreds of kilometers, so that’s one for Star Trek. Detailed atmospheric characterisation, however, is almost as good and not so far in the future if EXO-S gets the go ahead and the Planet Formation Imager evolves on the ground before migrating into space. All roads lead to space.

As an addendum, EXO-S has a yet to be described back-up that could best be seen as WFIRST AFTA-S. Here the starshade has the propulsive system, but the telescope is made from the NRO 2.4 m mirror, thus making the device potentially the most potent of the three designs. Having the drive system on the starshade, along with a radio connector to the telescope, is a concept even newer than the conventional EXO-S . But it is potentially feasible. We await a cost from the final reports the design concept groups need to submit.

In the meantime, various private ventures such as the BoldlyGo Institute run by Jon Morse, formerly of NASA, are hoping to fund and launch a 1.8 m off-axis telescope with BOTH an internal coronagraph AND a starshade. Sadly, the two methods have been found not to work in combination, but obviously a coronagraph telescope can look at stars while its starshade moves into position, increasing critical viewing time over a 3 year mission.

By way of comparison, coronagraphs can and have been used increasingly effectively on ground-based scopes such as Gemini South. It is believed that thanks to atmospheric interference the best contrast image achievable, even with one of the new ELTs being built, will be around 10 to the minus 9, so thanks to their huge light gathering capacity, they too might just discover terrestrial planets around nearby stars but probably not in the HBZ.

The future holds exciting developments. Tantalisingly close. In the meantime, it is important to keep up the momentum of development. The two Probe design groups recognise that their ideas, whilst capable of exciting science as well as just “proof of concept”, are a long way short of what could and should be done. The JWST for all its overspend will hopefully be a resounding success and act as a pathfinder for a large, 16 m plus New Worlds telescope that will start the exoplanet characterisation that will be completed by TPC-I. Collapsible, segmented telescopes will be shown to fit into and work from available launch vehicles, such as the upcoming Space Launch system (SLS), or one of the new Falcon Heavy rockets. New materials such as silicon carbide will reduce telescope costs. The lessons learned from JWST will make such concepts economically viable and deliver ground-shaking findings.

How ironic if would be if we discover other life in another star system before we find it in our own !


Comments on this entry are closed.

  • Ashley Baldwin September 12, 2014, 13:31

    For those wanting to know more about star shades the article below by Maggie Turnbull from the Exo-S team is very readable and available on arxiv:

    Also , Dr Aki Roberge from the same team has produced a great lecture on star shades on You Tube:

  • Andrew Palfreyman September 12, 2014, 17:16

    Roughly, if you packed monolayer graphene into a 1U CubeSat (a storage cube a little over 3 cm on the side), that could unfold into a surface of dimension 300×300 metres. It wouldn’t take many CubeSats to make a truly gigantic surface.

  • Jim Early September 12, 2014, 22:54

    Another approach is to use large aperture, very low mass, space telescopes utilizing the Fresnel lens technology being developed by DARPA in their MOIRE program. A good reference is APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014 by Britten et. al.

  • Wojciech J September 13, 2014, 6:13

    Very interesting article. I am personally excited for all these upcoming telescopes and hope these plans will bear fruit.
    Two questions/points:
    1. we heard about plans to create membrane telescopes by DARPA, how would they effect the exoplanet search? I believe Centauri Dreams was going to publish an article on this(unless I missed it)

    2.”To actually “see” detail on an exoplanet would require hundreds of large space telescopes spread over hundreds of kilometers, so that’s one for Star Trek.”. Just recently Centauri Dreams had articles about sending missions composed of hundreds of coordinated probes on micro scale who would combine to carry out science work(A Swarm of Probes to the Stars, Project Dragonfly). Would it be possible for swarms of such coordinated probes to assemble into telescoping arrays(Please note that I am just curious, and wonder if it is technically feasible, and think there might be something hindering this from happening)?

  • Tom Mazanec September 13, 2014, 8:03

    Also coming in a decade or two (hopefully) is ATLAST, the Advanced Technology Large Aperture Space Telescope.

  • ashley baldwin September 13, 2014, 11:30

    The answer to combining lots of smaller telescopes into one huge interferometer and then spatially resolve an exoplanet is yes. But this requires ” formation flying” which is key to the EXO-S mission . This is perceived as cutting edge and rieky though and even the authors admit in thejr report that it requires ” a significant engineering effort” . It will been done for the first time with the launch of the Proba3 ESA solar observatory in 2017 though. Someone has to hold their breath and take the plunge! The name to search for ” hypertelescopes” is the french astronomer , Labeyrie . Dont be despondent meantime though,
    the key to current imaging is spectroscopy . Characterising planetary atmospheres tells you as much as seeing them, almost. The key to spectroscopy is light which comes from bigger telescopes or longer
    observation or ” integration” periods. Exo-s proposes periods of up to 6 months.

    spectroscopy is light, the more the better. So that means a bigger telescope or longer observation , so called ” integration” periods . Exo-s proposes
    some integration periods as long as 6 months!

  • ashley baldwin September 13, 2014, 15:44

    ATLAST or ALAS? IF the ESA had teamed up with NASA over WFIRST and put the 825 million dollars they spent on yet another transit mission,PLATO, and NASA hadnt spent 425 million on ” Insight” ( whatever that is) to Mars , they would have had an extra 1.5billion to turn it into a proper space telescope descendant of Hubble that would have filled those decades till ATLAST. No bigger mirror, but with relatively cheap sophisticated CCDs ,Deformable mirrors and massive computer processing power, low aperture size can be mitigated.The EXO- S team are designing a different , bigger starshade to go with the NRO telescope which should be the most potent Earth hunter of all. Trouble with Exo-s, unlike the other options, is that while the starshade is moving into position, taking perhaps 11days, the telescope just sits idle. That is why the 34m star shade is much smaller and nearer to the telescope than in the arxiv article Maggie Turnbull wrote (50m and 90k km) above.Adding a starshade to the WFIRST-AFTA scope allows the time to be spent using the widefield camera in UV- I R light , as well as astrometry, dark energy and microlensing work. The 2.4 m mirror collects 5 TIMES as much light as the 1.1 of Exo-s, which means the all crucial planetary spectroscopy is done 5 times quicker (allowing more planets to be viewed in the 3 years mission) and at a more sensitive resolution. That said, and considering the limited funding , all the options are fantastic really. Innovation obviously accompanies poverty! We will see some great results , maybe even awesome. This mission is really exciting, especially on the back of TESS , which could be equally as ground breaking and just right to wet the appetite for direct imaging. NASA promised to use the NRO mirror, they cant afford to send up two missions so really they should be pushed into incorporating a starshade or coronagraph into WFIRST. Lobby, lobby, lobby !

  • Michael September 14, 2014, 4:16

    The big issue with thin membranes is the effect of light from the sun, being thin it will distort more easily than a thicker mirror or lens. I personally like the formation flying concept more as more mirrors could be added to the system over time. We could possibly use a few of the mirrors at a time looking in different directions for targets of interest and then combine all the mirrors to get a better look at each of them. I also like the concept of liquid mirrors on the moon, it would also give us a permanent presence on the moon as well.

  • ashley baldwin September 14, 2014, 13:15

    PROBA-3 is the first time formation flying will ever be done. Proof of concept mission. Two satellites with a solar corona experiment including ironically ,a coronagraph, flown up to 250m apart . Launch 2017. Bit of a difference with the 37000 km proposed for EXO-S, but a start and the EXO- S mission proposes using PROBA satellite buses.

  • ashley baldwin September 15, 2014, 2:35

    The idea of membrane based space telescopes, shaped into a parabaloid by posterior electrostatic forces was originally posited by Roger Angel at the University of Arizona in 2000. No one seems to hsve taken it forward as yet. It very much depends on formation flying , so maybe if PROBA-3 and EXO-S showcase the concept , someone will consider the membrane mirrors worth developing. Thats often the way with innovative new technology , it depends on something else to get going.

  • ljk September 15, 2014, 9:13

    ashley baldwin said on September 14, 2014 at 13:15:

    “PROBA-3 is the first time formation flying will ever be done.”

    You sure about that…


  • Jim Early September 15, 2014, 10:57

    Michael, the distortions in thin optics are primarily a problem for reflective optics. Transmissive optics are dramatically less sensitive, particularly lenses with large focal length to diameter ratio (f-number). A transmissive lens can tolerate a distortion that is (4 f/D)squared larger than a reflective optic. Fresnel optics for space would have f-numbers in the 10 – 100 range.

  • spaceman September 15, 2014, 16:47

    Right now exo-planet science seems to be in an interesting “in-between era”– that is, the 20 year anniversary of the 51 Peg announcement is just around the corner and we already have a great statistical understanding of the planet population mostly as a result of Kepler data (let’s also not underestimate the radial velocity and microlensing techniques). I remember the era when we knew of no planets– not even the pulsar planets– existing outside of our solar system. We now seem to have pretty good answers to the following questions:

    What is the percentage of stars with planets?
    Do binary systems have planets?
    What is the planet frequency around stars of a variety of spectral type?

    So, on the one hand, we have come a long way in a relatively short period of time, but it seems like the really detailed information about extrasolar planets coming from spectroscopy will take some more time. I remember reading an article in Astronomy magazine back in the late 1990s about the plans for SIM (Space Interferometry Mission) and TPF (Terrestrial Planet Finder) and in the article they were projecting a 2008 launch for SIM and a 2012 launch for TPF!

    It heartening to read about more near term missions capable of spectroscopy for Jupiter size worlds and maybe even a few Neptune-sized planets, but it seems like we may have to wait until the 4o year anniversary of the 51 Peg discovery before we have a substantial set of detailed spectroscopic observations of exoplanets closer to earth-size.

  • ashley baldwin September 15, 2014, 17:22

    Lfk- Well spotted. I can only say that this is the ESA claim, and they go into great detail on their website. More importantly the EXO-S team heavily cite this mission and propose using a PROBA satellite bus in EXO-S. Im no formation flying historian , but its this mission thats the one for proof of concept. Their interim report in April goes into detail (I would recomend it as a good read) and they are already working with the PROBA company on the technology for the mission. Only over 250m for this so they are obviously very confident of the technology working over the 37000km they propose, and probably 80000 if they use the NRO mirror. Laser metrology . Cross fingers like mad, as formation flying is what is needed for interferometry and the huge space telescopes of the future. I think the PROBA technology will work and well, its just whether NASA will trust itenough to go with it to the tune of $1.6 billion! I believe the WFIRST-AFTA team propose adapting the telescope to work with this technology too for use with a future starshade should they be successful.

  • ashley baldwin September 15, 2014, 17:40

    Looking in more detail, I think the PROBA mission first is about precision formation flying. The telescope has to remain within 1m either way whilst its staring at a planet for up to 6 months while characterising it spectroscopically . This requires close two way radio/laser contact and exquite laser distancing, metrology , between the shade and telescope throughout . Its this that is new i suspect . Unbelievable technology if it works and with potentially great reward results wise. Onthe subject of the military, ive heard from astronomers that they tend to always be ten years ahead technology wise as classically with adaptive optics and the NRO mirrors. If the NRO are giving away two 2.4m mirrors, what are they using now?! Something to be proud of and reassured by in uncertain times.

  • Benjamin Pope September 16, 2014, 1:07

    As far as interferometry goes, a quick baby step will be using single-telescope interferometry in space – JWST will be launching with an aperture mask, which will I expect be quite exciting for high resolution imaging at slightly beyond the diffraction limit and moderately high contrasts. Some of my collaborators have a paper about it here: http://arxiv.org/abs/1406.6882

    Can’t resist a plug for some of my own work on doing single telescope interferometry without even using a mask, in post-processing HST data – this will be equally applicable to enhancing JWST imaging performance. http://arxiv.org/abs/1406.6882

    I’m very excited about the prospects of interferometry in space, but I also think a PFI incarnation on the ground, particularly the many-telescope heterodyne concept, will be a huge step forward for the field.

  • ljk September 16, 2014, 8:49

    Well unless the ESA is stuck on national pride, they would be wise to check these earlier satellite formation flying missions. Considering how successful they apparently were I am sure they could learn a few things from them to save both time and money. But I get the feeling they want this to be all European so they will probably be reinventing the wheel in certain aspects.

    At the height of the Cold War it used to be said the military was twenty years ahead of the civilian efforts, but I am not sure how wide the gap is at this time. Then again with Russia ramping up its production of nuclear weapons we may be back to the twenty year gap soon enough. Though a lot more private companies now make and have access to technology they rarely did back in the day, so perhaps there is only a few selective gaps.

    Though it is interesting to note that DARPA funded the 100 Year Starship effort. I have contended since the beginning that one big reason they did this was to see how much tech knowhow they could get from eager and naive young engineers and scientists who think they are contributing to building the USS Enterprise and will happily give away their inventions to the military for the hope that a real starship will be built.


  • ashley baldwin September 16, 2014, 11:41

    Im sure the EXO-s team will pursue every option if I know Sara Seager and Maggie Turnbull. I think the issue isnt just about formation flying in the sense we know it from aircraft, so much as precision formation flying where inter spacefraft communication and positioning is absolutely critical. In terms of the ESA its more of a pity they couldnt collaborate with NASA and use the M3 money spent on the transiting mission PLATO ( launching about the same tine ironically ) for WFIRST-AFTA. Unfortunately with NASA pulling out of the JUICE mission next decade over budget cuts there may be some residual mistrust. Dont know. All the teams have done a great job of squeezing the most they can out of a pitiful budget and im sure the initial results will be exciting enough to grant the extension where they can push the telescope/coronagraph/starshade to the limit and look for exo Earths. Ten years is a long time , first we hope the budget is granted to even allow any mission to fly . Roll on 2017.

  • stephen September 17, 2014, 13:19

    I think it would be fun to try placing an actual firefly in front of or near an actual searchlight, and try taking a look at it. I’d like to send the idea to Adam Savage and Jamie Hyneman.

  • Dominic Benford September 17, 2014, 14:19

    I believe there is something of a mischaracterization of the present status of WFIRST-AFTA. In particular, the statement “NASA put the WFIRST AFTA design on hold until the budgetary statement of 2017” is incorrect. NASA has been investing substantially in the development of key technologies (mostly the near-infrared detector arrays and the coronagraph) and the overall mission concept. This preparatory effort will significantly reduce mission risk over the course of the next few years, at which time the technology will be mature. At that point, which is anticipated in 2017, the mission will be in a position to transition to its formal development phase. It is true that JWST and WFIRST will not be developed simultaneously (thereby maintaining a stable budget from year to year), which means the WFIRST launch would be in 2024.

    It is also somewhat misleading to compare the cost estimates for WFIRST. The acronym has remained the same, and the core mission goals have remained the same, but the $1.6B version was not the $2B version. The present WFIRST-AFTA mission has a significantly improved capability and costs somewhat more, and the most recent independent review of the project acknowledges this. A large portion of the additional cost is that the bigger and better telescope requires a larger (and hence more expensive) rocket to take it aloft. Adding the coronagraph and its exciting science potential increases the budget a little more, and brings a completely new exoplanet capability.