≡ Menu

Exoplanets: Where Will We Be by 2020?

Where will we be in the exoplanet hunt by the year 2020? A few of my own guesses would take this form: We should, within even the next year or two, have detected a terrestrial world in a truly unambiguous position within the habitable zone of a star. That star will doubtless be a red dwarf, like Gliese 581, but we can hope for a result that doesn’t lend itself to so many conflicting interpretations. The detection method will surely be planetary transit, but even by 2020 we may not know if life exists there.

It’s also easy to surmise that by 2020 we’ll have a terrestrial-class world located within a stellar system not completely dissimilar to our own; i.e., one involving a star much like the Sun, orbited by a rocky world in the habitable zone. We can hope that by 2020 the tools will have been put in place to do spectroscopic observations of the planetary atmospheres involved in small rocky worlds, though so much depends on budgets and the needed tuning up of exquisitely sensitive technologies.

I have a number of other guesses that could come into play, but I’m already second-guessing myself. And now I’m drawn up short by Drake Deming (NASA GSFC), who reminded the recent conference at the Space Telescope Science Institute that going back thirteen years instead of forward, we would remember that only the pulsar planets were known to us at the time. Planets around pulsars were incredibly exciting back then, and we know now that PSR B1257+12, the first pulsar involved, actually has at least three planets instead of the two first discovered.

But one thing pulsar planets are not and that is likely homes to anything like life as we know it. Alien to the point of absurdity, they basically proved something about the likelihood of planets being found elsewhere, but who would have guessed from them that we would find such things as ‘hot Jupiters’ or triple star systems with planets of their own? No, making guesses in a field expanding this rapidly is a dangerous game, one I’m nonetheless glad that Deming and the other scientists at this meeting were willing to play.

The conference, which ended on November 15, was titled “Astrophysics 2020: Large Space Missions Beyond the Next Decade.” Its goal was to look at the groundwork that will be laid by emerging technologies in that time. Ponder: The Ares V heavy launch vehicle means that some of the barriers to putting massive astronomical observatories into space will disappear. Moreover, we’re getting better at robotics, leading ultimately to construction and servicing jobs in space that would previously have been impossible. All this was fodder for conference discussion.

I couldn’t be at the STScI conference, but I’m listening to Dr. Deming’s presentation right now as he discusses the possibilities emerging in the field of transit observations. The entire conference is now available, thirty presentations and panel discussions online in PowerPoint as well as low and high-bandwidth streaming or downloadable video. Need I point out what an educational opportunity webcasts like this provide? As more and more conferences take STScI’s lead, the ability for those unable to attend to learn from the conference experience will become the kind of resource we only used to dream about.

I thank STScI’s Ian Jordan for passing along the much appreciated link to these webcasts, thus allowing me to confound my family by watching presentations all weekend. Fortunately, my wife is a patient woman…

And so back to Drake Deming, who is explaining in a video window on my desktop that direct detection of planets will be occurring in the not so distant future — direct detection means separating planetary photons from stellar photons, no easy task but increasingly feasible nonetheless. But even before that goal is reached, there’s a great deal we can discover about planetary atmospheres. Dr. Deming is moving on to discuss how transits can be used to identify habitable planets around M dwarfs, so he’s lighting up practically every synapse I have, obsessed as I am with the M dwarf planetary question.

We have already identified 20 known transiting exoplanets and the discovery rate is accelerating — twelve were announced just this year. As Dr. Deming discusses a temperature inversion in the atmosphere of HD 209458b, I’m reminded that even now we are making the kind of observations of planetary atmospheres (and indeed, constructing the crudest of maps based on temperature data from secondary eclipses as the planet passes behind the star) that no one thought we could do just a few years ago. The ‘hot Neptune’ GJ 436b likewise shows how transits can reveal surface temperatures.

You can see why the lowest mass M dwarfs are so attractive as we shoot for still smaller planets: Here the habitable zone lies close to the star, meaning short orbital periods and a higher probability of transits. And the small size of the star means that rocky planets can more easily be detected (Dr. Deming covers this in detail). M dwarfs also outnumber stars like the Sun by ten to one. “The nearest habitable planet to Earth,” Deming is saying as I write, “probably orbits an M dwarf, just because there are so many of them.”

You’ll want to watch Dr. Deming’s presentation to learn about the MEarth Project (Mt. Hopkins, AZ), using eight 16-inch telescopes to survey the 2000 nearest M dwarfs for rocky planets in their habitable zones, looking for planets that will be good candidates for spectroscopic follow-up (not with Spitzer but via the James Webb Space Telescope) to search for atmospheric biomarkers. We could be looking at extremely interesting data from the atmospheres of such planets as early as 2015.

Does that gibe with your own predictions? Those interested in all branches of astronomy and astrophysics will want to work through all these presentations to see what’s ahead as we put the latest hardware to work and continue to refine our techniques. Well done to the conference presenters for making this treasure trove available!

Comments on this entry are closed.

  • dad2059 November 19, 2007, 10:03

    Honing the skill to detect possible earth-type worlds around the smaller and dimmer M-class red dwarf stars is very practical. I didn’t realize there are ten times as many of them as there is Sol-type stars.

    Since there are so many of them, what are the chances of finding one or two that have significant infrared readings and the system swept clean of rocky planets? I would think that chances of finding Dyson Shells, or semi-shells would be greatly increased.

  • Robin Goodfellow November 19, 2007, 12:51

    Interestingly, 2020 is as distant from today as today is from the dawn of the new age of planet hunting that arrived with the discovery of 51 Peg B in 1995. Given the current pace of discovery and the new methods and instrumentation being piloted and used in the near future I have to think that our knowledge of extrasolar planets in 2020 will be as different from our knowledge today as our knowledge today is different from that of 1994. We will have more exoplanet search programs using a wider variety of instrumentation than today (doppler, astrometry, occultation, etc.). We will have not merely a list of exoplanet examples but a proper catalog with multiple examples of just about every imaginable type of planet (small, large, close, far, eccentric, circular, etc.) And we will start to have two kinds of data which we have just the slightest hint of today. Reliable statistics about planet formation and information about exo-planetary structure (chemical composition, temperature, seasons, etc.) More so we will have a tiny subset of exoplanet discoveries which will change the way most people think about the Universe. A small handful of planets which are twins of Earth in size, composition, and surface temperature and which offer the tantalizing possibility of hosting life as we know it. Examples of dramatic transient phenomena in exoplanetary systems, impact events, ring systems, extraordinary comets, planets and moons caught in the act of forming, etc.

    For millenia mankind has known only the planets in our own Solar System, and we have tried to piece together the methods of planetary formation from those meager examples. Truthfully though it is not a big enough sample set. Planetary science has started in earnest only within the last decade, we are at the start of an amazing era of discovery and a major transformation of our understanding of the Universe and our place in it.

    It boggles the mind that this is happening in parallel with revolutionary changes in our understanding of cosmology (determining the age of the Universe to high accuracy, learning the Universe is expanding due to the presence of dark energy, seeing the afterglow of the big bang with COBE and WMAP, etc.)

  • Lubo November 19, 2007, 14:44

    When we find Earth-like worlds, how do we get there ;)
    I hope that Dr. Tajmar will have some breakthrough in the area of propellantless propulsion by then :)

  • Jyril November 19, 2007, 18:22

    We have already identified 20 known transiting exoplanets and the discovery rate is accelerating — twelve were announced just this year.

    I am happy to inform that this count is already outdated. :D

    There are already 33 published transiting extrasolar planets, 16 which has been found in this year only–and yet this will not be the final number because it is told that first COROT results are to be published on December 10 and they no doubt will reveal some interesting discoveries.

  • John November 19, 2007, 21:04

    What would be the criteria for a type of planet which would be a reasonable target for an interstellar mission to establish humanity there? I figure, when such a planet is found, it will stmulate public interest in supporting the development of a mission to it.

    For me the minimum criteria would be:
    – within 20 ly of Earth,
    – 25% Earth’s gravity,
    – a Mars-density atmosphere to 1.5 Earth-density atmosphere,
    – CO2 in the atmosphere,
    – frozen water
    – no less than 0 degrees avg C – no more than 100 degrees avg C

    A magnetic field would be nice but not essential.

    And could anyone hazard a guess as to when such a plant would be found?

  • Robin Goodfellow November 19, 2007, 21:58

    John, that’s a fascinating question actually. Precisely because the sort of technological capability that would allow for interstellar colonization presupposes a vastly different set of requirements for “habitability” of a target system than, say, ordinary Earth-twin systems. The capability to send thousands of humans across interstellar distances almost guarantees the ability to survive as a technological species without the need for planet-based industry. Meaning, target systems need not have biospheres, or even atmospheres, at all but rather resources of the sort that are exploitable with transported equipment and can be used to bootstrap into a robust infrastructure for an interplanetary civilization. It may be that nearly every stellar system within range ends up being a good target for colonization.

  • Hans Bausewein November 20, 2007, 3:36

    I expect a lot from improved software and computing power, too. There must be a lot to win in noise reduction, especially if you know what to look for.

    An orbiting planet has a speed relative to the parent star that is quite predictable. If measurements are accurate enough to separate the Doppler-shifted features of the planet from those of the parent star, it’s just a matter of getting enough observation time and computing power to get the data out of it.

  • Ronald November 20, 2007, 5:29

    I would agree with almost all your criteria, except the 20 ly limit.
    I understand that the closer is the better, but if we can travel 20 ly to a so-so planet, we can probably also travel 30, 40 or 50 ly to a really good planet ;-)

    Although I agree with Robin Goodfellow that a truly space-traveling civilization may be able to terraform or at least utilize a wide range of targets (even if only as a transit), surely some will be (vastly) more attractive than others, from a human point of view.

    Besides, it may not even be entirely right: terraforming a planet, particularly a very unsuitable one, requires a LOT more energy and time than even stellar travel.

    E.g. using Zubrin’s ‘Entering Space, Creating a Spacefaring Civilization’, it is possible to roughly estimate how much energy it would cost to bring enough comets from the Kuiper belt to Venus to 1) get sufficient water there and 2) to speed up its rotation to approaching earthly day-length.
    The resulting energy demand is so ludicrously high that it probably won’t be done in any foreseeable future, not because we will never be technically able to, but because there would always be better (i.e. cheaper) alternatives (such as terraforming Mars and going to Alpha Centauri, …).

  • Ronald November 20, 2007, 6:33



    a) detection (i.e. any kind of, including astrometry, doppler, transit)
    b) direct imaging
    c) spectroscopic analysis

    of roughly earth-sized terrestrial planets,

    will we always need space-based platforms (in particular interferometers), or can this eventually also be done with ground-based instruments, such as the European Extremely Large Telescope (42 meter diameter), the Keck interferometer, …?.

  • Ronald November 20, 2007, 6:35

    I should have added to ‘of roughly earth-sized terrestrial planets’:

    at roughly earthlike orbits around sunlike stars (say, F7 – K2).

  • Adam November 20, 2007, 8:30

    Hi John

    That minimally interesting planet might be too small. A surface gravity of 0.25 g makes it much smaller than Mars – its radius would be 1/3 Earth’s, just 2/3s of Mars. Thus it would be too small to detect with current and near-term techniques.

    More reasonable would be any planet between about 0.2 and 5 Earth masses – what’s called the ‘habitable mass range’. The radius would be roughly 2/3 Earth to 3/2 Earth radii. About 0.5 g to 2 g surface gravity. Such planets would be big enough to retain geophysical activity and small enough not to become gas giants.

    Getting there will be tricky. One possibility is life-extension via artificial suspended animation – recent surprises in suspended animation research make that possibility more real than ever before. For example a man survived 23 days without food or water in a state of suspension. And mice have been put into a kind of ‘slow metabolism’ mode when exposed to small amounts of hydrogen sulphide. I suspect, based on the mice data, that humans might be able to go into ‘slow mode’ that metabolises about 20 times slower than normal, thus allowing extension of lifespan while in transit between the stars.

    Charles Sheffield fictionalised such a ‘slow mode’ – though his was thousands of times slower than normal.

    If we wanted to and mobilised as a planet a fusion-pulse starship could be built in a couple of decades and launched to Alpha Centauri with a 100 year trip time. A laser-sail could be developed in about 50 years and be able to travel at 0.1-0.3c, provided magnetic-sail technology can be developed to deccelerate the laser-sail.

    But first we need the will…

  • ljk November 20, 2007, 11:43

    Why do people keep assuming that by the time we are able
    to send spaceships to the stars that humans will be an
    essential part of them?

    AI will likely do the job better and cause a lot fewer issues
    in the resources department.

    We all need to think a bit farther ahead when it comes to
    designing a real starship, while at the same time avoiding
    going to the extremes of warp drive and such. That may
    happen some day, but for now it is much more practical to
    focus on the physics we do know to get us to Alpha Centauri.

  • andy November 20, 2007, 13:07

    Why do people keep assuming that strong AI and the Singularity are inevitable?

  • ljk November 20, 2007, 14:43

    Because Vernor Vinge said so.


  • Ronald November 20, 2007, 14:46

    @ljk: sure, robot probes wouls do the job better, safer, faster, cheaper. Particularly for exploration. But ultimately humans will want to go, that’s what it’s all about. And then suspended animation, cryogenics and the like may help a lot to do exactly what you also emphasize: to significanty reduce resource demands and hence the weight of the ship.

    @Adam: I agree, this suspended animation also features prominently in A.C. Clarke’s ‘The city and the stars’ series.

    We need the will…, but first of all I think we need the discovery! I am inclined to believe that once we discover a terrestrial planet near Alpha Centauri with some earthlike characteristics (right temperature, water, atmosphere) budget for further exploration will no longer be a constraint. Even more so, if this planet has the spectral signature of a living planet (biosignature).

    @Lubo: yes, it would be great if Tajmar, Dröscher and others would achive some breakthrough in the application of Heim’s theory, but until then I am afraid we’ll have to do with old-fashioned fusion and laser propulsion ;-)

  • ljk November 20, 2007, 15:04

    What we think of as human in 2007 may be very different by
    2107, or even sooner.

    An enhanced human may make a better starship passenger.

    To think we will remain ourselves while all kinds of technological
    and biological advancements go on around us ala Star Trek is
    just not realistic.

  • ljk November 21, 2007, 17:25

    What Space Telescopes of Tomorrow Will See

    By Dave Mosher

    Staff Writer

    posted: 21 November 2007

    06:47 am ET

    Giant-sized telescopes such as Hubble, Spitzer and Chandra offer unprecedented views of the cosmos, but astronomers are eager to put more powerful tools into orbit around the Earth.

    Without the extra help, said Rachel Somerville, an astronomer at the Max Planck Institute for Astronomy in Germany, it may be impossible to resolve some of the universe’s greatest mysteries.

    “We need better observations to make our models better,” Somerville said, noting her search to understand galaxy formation and mysterious quasars. “… If you just put theorists in a room for the next 15 years with the biggest supercomputer you can find, it will never happen.”

    NASA expects the James Webb Space Telescope (JWST) to launch in 2013, and many scientists are already pondering their future observations of tiny extrasolar planets, elusive black holes and distant galactic arms.

    Somerville and other astronomers laid bare their sky-watching hopes—including telescopes beyond JWST—at the recent Astrophysics 2020 conference, sponsored by Johns Hopkins University and held at the Space Telescope Science Institute in Baltimore.

    Full article here:


  • andy November 22, 2007, 16:58

    Things I would hope we have some handle on by 2020, other than the habitable planet issue which is only one facet of exoplanetary science:

    Are extrasolar planetary systems near-coplanar, or have planet-planet interactions made a significant fraction very messy indeed?

    Do planets ever form around more than one star in a multiple star system?

    Do multi-planet systems exist in close binaries?

    Do circumbinary planetary systems exist, and how frequent are they? What about around contact binary stars?

    Are there planets in other galaxy types than spirals (a question that may be answerable through pixel microlensing surveys)?

    Transit surveys could potentially yield detections of rings or moons around exoplanets. How frequent are these phenomena, particularly around planets that have undergone substantial migration?

    Are “exotic” configurations such as Trojan planets, horseshoe orbits (these two exhibited among the moons of Saturn in our solar system), or the 1:1 eccentric resonance realised in extrasolar systems?

    What is the metallicity dependence of low mass planets?

    What does the distribution of planets look like around intermediate-mass and high-mass stars?

  • philw1776 November 24, 2007, 17:49

    There are strong mis-conceptions about M type stars with planets. First that the planets are necessarily locked with one face perpetually lit, the other dark. There well could be 3:2 and other resonances with Jovian class other planets tens of millions of miles away. With ‘years’ being a couple weeks, there would be multi-week ‘days’ on such planets.

    Secondly astronomers tout that there are 10+ times more M than brighter late F, G and early K stars. True but they overlook the extremely tiny HZ for an M star. It would be a million miles or less wide for a typical M star since they have a fraction of the sun’s luminosity. Therefore the probability of a terrestrial type planet landing exactly on target for an M star is tens of times less than the brighter D stars totally offseting the advantage in M star numbers.

  • andy November 25, 2007, 14:57

    philw1776: however, on a logarithmic scale, the instantaneous habitable zone of an M dwarf is the same size as that for a G dwarf, and planetary systems seem to be roughly logarithmically spaced (at least, for the few systems we know of which have more than 2 planets, and allowing some gaps for additional planets/asteroids/whatever) rather than linearly. It is worth pointing out that both Gliese 581 and Gliese 876 each have two planets inside or close to the habitable zone.

  • Adam November 25, 2007, 17:09

    Hi andy

    Any thoughts on what mechanism spaces planets logarithmically? It’s a fact, seen also among the regular systems of moons of the Jovians, but bit of a puzzle.

  • philw1776 November 26, 2007, 20:59

    So what if it’s approximately logarithmic? The probability of hiting the HZ in a mid-M class star is over 100x less due to it’s luminosity driven tiny size.

  • ljk November 27, 2007, 0:00

    Are there many binary red dwarf systems? Would this improve
    on a HZ and thus the odds for a planet with life?

  • andy November 27, 2007, 15:57

    philw1776: you’d be correct for the probability that a given planet will end up in the HZ, however once you have a system of logarithmically-spaced planets, the probability that one of the planets in the system is in the HZ depends on the logarithmic width of the HZ not on the absolute one.

    In fact, because of the slower rate of stellar evolution, the HZ for an M dwarf of a given age is wider (in logarithmic space) than that for an HZ of a higher-mass star at the same age.

  • Shaun December 2, 2007, 21:11

    philw1776 while the HZ around M dwards may indeed be puny in comparison to sunlike G dwarfs but the focus on HZ alone I think is misplaced. For have you also considered the effects of planets just outside of the HZs of such VLM dwarfs may possess thick atmospheric cover with some grenhouse gases like carbon dioxide and methane? Would not the existence of a thick enough atmosphere and moderate amounts of such greenshouse gases negate be enough to negate the concerns you raised as well as perhaps efficient heat transportation so that temperatures on any part of any a planet just outside of the HZ are essentially the same?

  • Shaun December 2, 2007, 21:19

    As to my own predictions for 2020, I think we may find that amongst the 260 or so exoplanets we alredy have found, quite a few will turn out to be binary planets in reality. I’ll also like to think that additional bnary planets will be locate and identified by 2020. By then we should have found a few planets and brown dwarfs around black holes too. I’m also with andy here on circumbinary planets i.e. if they exist at all that is. And before I’m done, allow me to make a really bold prediction – we will find a bonanza of planets (of all kinds) around a number of the M dwarfs in Sol’s neighborhood including dare I say, a few in orbit around Proxima Centauri.

  • ljk December 5, 2007, 14:18

    Cadence Optimisation and Exoplanetary Parameter Sensitivity

    Authors: Stephen R. Kane, Eric B. Ford, Jian Ge

    (Submitted on 3 Dec 2007)

    Abstract: To achieve maximum planet yield for a given radial velocity survey, the observing strategy must be carefully considered. In particular, the adopted cadence can greatly affect the sensitivity to exoplanetary parameters such as period and eccentricity. Here we describe simulations which aim to maximise detections based upon the target parameter space of the survey.

    Comments: 4 pages, 5 figures, to appear in the Proceedings of the 249th IAU Meeting: “Exoplanets: Detection, Formation and Dynamics” (Suzhou, China)

    Subjects: Astrophysics (astro-ph)

    Cite as: arXiv:0712.0358v1 [astro-ph]

    Submission history

    From: Stephen Kane [view email]

    [v1] Mon, 3 Dec 2007 17:43:34 GMT (106kb)


  • ljk March 6, 2008, 9:41

    THE JAMES WEBB SPACE TELESCOPE (Science Show: 01/03/2008)


  • ljk April 30, 2008, 13:49

    Exoplanets Task Force Report

    Steve Unwin, Editor – stephen.unwin@jpl.nasa.gov

    The Exoplanets Task Force (ExoPTF) has released (in almost final form) its report to the Astronomy and Astrophysics Advisory Committee (AAAC). The ExoPTF was established as a subcommittee to advise NSF and NASA on the future of the ground-based and space-based search for and study of exo-planets, planetary systems, Earth-like planets and habitable environments around other stars.

    In its 160-page Report, the ExoPTF recommended a 15-year strategy to detect and characterize exo-planets and planetary systems, and their formation and evolution, including the identification of nearby candidate Earth-like planets. The (draft) Report may be found at:


    An important finding of the Report is that the diverse methods of observing extrasolar planets, and the systems in which they live, each provide important information that, taken together, advance the field significantly. Accordingly, the Report considers in detail the scientific contributions of each method and their future prospects.

  • ljk July 7, 2008, 23:15

    The Impact of Transiting Planet Science on the Next Generation of Direct-Imaging Planet Searches

    Authors: Joseph C. Carson

    (Submitted on 7 Jul 2008)

    Abstract: Within the next five years, a number of direct-imaging planet search instruments, like the VLT SPHERE instrument, will be coming online. To successfully carry out their programs, these instruments will rely heavily on a-priori information on planet composition, atmosphere, and evolution.

    Transiting planet surveys, while covering a different semi-major axis regime, have the potential to provide critical foundations for these next-generation surveys. For example, improved information on planetary evolutionary tracks may significantly impact the insights that can be drawn from direct-imaging statistical data. Other high-impact results from transiting planet science include information on mass-to-radius relationships as well as atmospheric absorption bands.

    The marriage of transiting planet and direct-imaging results may eventually give us the first complete picture of planet migration, multiplicity, and general evolution.

    Comments: 4 pages, 3 figures, IAU Transiting Planets Proceedings, in press

    Subjects: Astrophysics (astro-ph)

    Cite as: arXiv:0807.0705v1 [astro-ph]

    Submission history

    From: Joseph Carson [view email]

    [v1] Mon, 7 Jul 2008 14:52:33 GMT (1221kb)


  • John Chesnut March 7, 2009, 13:26

    If we discover an exoplanet with habital characteristics and at a “reasonable” distance, is it theoretically possible for humans to reach that planet? Granted, even with the closest planet, the distances are too great for an individual to arrive there during his lifespan. However, with living aboard the spaceship and procreation over several life cycles this might be achievable. What are your thoughts?

  • Administrator March 7, 2009, 15:16

    John, we discuss many strategies for interstellar missions here, ranging from propulsion concepts involving lightsails or fusion runways to ‘worldships’ that take multiple generations to reach their goal using some kind of sail technology. So it’s not an easy question, but you might want to try searching the site under terms like ‘worldship’ or ‘sail’ or ‘antimatter’ to see some of the ideas that are under active study and some that are still purely theoretical. My book Centauri Dreams: Imagining and Planning Interstellar Exploration is devoted to this topic.