≡ Menu

Exoplanet Images: Two Observational Coups

Are we really moving beyond indirect detection methods to being able to produce actual images of extrasolar planets? Apparently so, as witness the first direct images of multiple planets around a normal main sequence star. And on the same day, we have the announcement of a visible light image of a Jupiter-class planet orbiting the star Fomalhaut, one suspected for several years because of the sharply defined inner edge of the dust belt around the star. A planet in an elliptical orbit affecting the debris disk had been thought to be offsetting the inner edge of the belt.

Let’s go to the planets found around the dusty young star HR8799 first. They range from seven to ten times the mass of Jupiter. Bruce Macintosh (Lawrence Livermore National Laboratory), one of the authors of a new paper on the achievement in Science Express, explains its significance:

“Every extrasolar planet detected so far has been a wobble on a graph. These are the first pictures of an entire system. We’ve been trying to image planets for eight years with no luck and now we have pictures of three planets at once.”

Radial velocity methods can tell us much, including not only the presence of a planet but also something about its mass and orbit. But these methods work best when the separation between the star and the planet is relatively small, usually within the range of 5 AU. What we have here are three worlds quite a distance from their primary, at 24, 37 and 67 AU respectively. By comparison, Neptune is about 30 AU from the Sun in our Solar System. The star, a blue A-class object, retains a sizable disk, and its planets are young enough to be throwing a bright infrared signature. The results of observing same appear immediately below:

: Gemini Observatory discovery image using the Altair adaptive optics system on the Gemini North telescope with the Near-Infrared Imager (NIRI). Image shows two of the three confirmed planets indicated as “b” and “c” on the image. “b” is the ~7 Jupiter-mass planet orbiting at about 70 AU, “c” is the ~10 Jupiter-mass planet orbiting the star at about 40 AU. Due to the brightness of the central star, it has been blocked and appears blank in this image to increase visibility of the planets. Credit: Gemini Observatory.

HR8799 is relatively nearby at 130 light years, visible using binoculars or a small telescope (or even via the naked eye, depending on your seeing conditions). Close enough to find, and perhaps image, other planets in this interesting system? It’s a distinct possibility, and with ever more powerful adaptive optics systems going into place via the Gemini Planet Imager, we’ll boost our capabilities a hundred times. Says Macintosh:

“I think there’s a very high probability that there are more planets in the system that we can’t detect yet. One of the things that distinguishes this system from most of the extrasolar planets that are already known is that HR8799 has its giant planets in the outer parts – like our solar system does – and so has ‘room’ for smaller terrestrial planets – far beyond our current ability to see – in the inner parts.”

Interesting indeed, and note that we’re already doing spectroscopy to study these three planetary atmospheres. Bear in mind, too, that the work on this star is part of a survey of eighty young, dusty stars located not far from the Sun. Because the HR8799 planets showed up after observations on only a few, we may discover that Jupiter-class worlds at these separations are not uncommon among more massive stars.

Imaging planets around young stars like this one is by no means an easy task, but as we move to older systems with planets well beyond their early formation stage, the game will only get tougher. HR8799, a star with a dust disk more massive than any star within 300 light years from Earth, is a harbinger of necessary instrumentation tune-ups to come as we extend the search for exoplanetary images from Earth into advanced space missions. But for now, what a job by astronomers using adaptive optics at Gemini North and the Keck Observatory!

As to Fomalhaut b, clearly defined in the Hubble Space Telescope image below, the evidence seems strong indeed that we are looking at a actual planet. James Graham (UC Berkeley), a co-author of the paper on this work, has little doubt: “It will be hard to argue that a Jupiter-mass object orbiting an A star like Fomalhaut is anything other than a planet.”

Image: This 2006 Hubble Space Telescope optical image shows the belt of dust and debris (bright oval) surrounding the star Fomalhaut and the planet (inset) that orbits the star every 872 years and sculpts the inner edge of the belt. A coronagraph on the Advanced Camera for Surveys blocks out the light of the star (center), which is 100 million times brighter than the planet. Credit: Paul Kalas/UC Berkeley; STScI.

Its brightness also argues that Fomalhaut b has a particularly intriguing property. Paul Kalas (also at UC Berkeley), has this to say:

“To make this discovery at optical wavelengths is a complete surprise. If we’re seeing light in reflection, then it must be because Fomalhaut b is surrounded by a planetary ring system so vast it would make Saturn’s rings look pocket-sized by comparison. Fomalhaut b may actually show us what Jupiter and Saturn resembled when the solar system was about a hundred million years old.”

The researchers believe they can constrain the planet’s mass to between 0.3 and 2 Jupiter masses, noting that a more massive object would destroy the dust belt around Fomalhaut. Like HR8799, Fomalhaut is a young star — about 200 million years old — and will have a short lifetime of perhaps a billion years. Sixteen times brighter than the Sun, the star would appear about as bright from the new planet as our Sun does from Neptune even though the distance to HR8799 is four times greater. Fomalhaut b is also a mysterious place in at least one sense: It has dimmed by half a stellar magnitude from 2004 to 2006, perhaps an indication of a hot outer atmosphere affected by convection cells.

So there we are. Not just an exoplanet directly imaged in visible light, but one with a ring system that may well resemble what Jupiter’s must have been like before the large Galilean satellites coalesced. Here again, note the observational constraints. The radial velocity method would not have allowed Kalas and team to have made this detection — the planet is simply too far from its star and too low in mass. And while infrared would seem to be the tool of choice to detect hot young worlds like this one, the planet’s great distance from its star made a visible light image possible.

The Fomalhaut b paper is Kalas et al., “Optical Images of an Exosolar Planet 25 Light-Years from Earth,” Science Express November 13, 2008 (abstract). The paper on the triple find around HR8799 is Marois et al., “Direct Imaging of Multiple Planets Orbiting the Star HR 8799,” Science Express (same date). The abstract is here. This Hubble news release is also helpful.

Comments on this entry are closed.

  • Enzo November 13, 2008, 15:50

    New Scientist has an article on HR8799 and the images look different, taken by a different group :

  • tacitus November 13, 2008, 16:13

    We knew it was coming, but it’s remarkable all the same. Unless I’m mistaken, Hubble is really a “zeroth” generation planet finder — i.e. it wasn’t even expected to be able to image exoplanets, so this discovery is a mighty big bonus (and certainly a coup shared by all the Hubble team members through the years (good times and bad).

    I have every confidence that within a couple of decades we will have a whole catalog of exoplanet images and spectra to study, some of which might even have tantalizing hints about the possibility of life.

    We may not have much need for a Hitchhiker’s Guide to the Galaxy for another few hundred years yet, but at least the creation of the Encyclopedia Galactica has now begun in earnest!

  • Hans Bausewein November 13, 2008, 17:31

    Amazing, indeed!

    It can also help to improve detection strategies: methods that improve visibility of these planets will probably also make it easier to find the more difficult ones.

    And in these systems the now known planet(s) put some constraints on the locations where more can be expected.

  • John Hunt November 13, 2008, 23:18

    I find it amazing!

    We are getting closer and closer to detecting a relatively nearby planet with a size ranging between Mars to Earthx2. If spectra shows CO2, N2, and water vapor in the atmosphere then we’ve got a legitimate target for interstellar colonization. Hopefully it will inspire another Daedalus-type project and even Mars Homestead-like work.

  • ljk November 13, 2008, 23:37

    Fomalhaut’s Debris Disk and Planet: Constraining the Mass of Fomalhaut b From Disk Morphology

    Authors: E. Chiang (UCB), E. Kite (UCB), P. Kalas (UCB), J. R. Graham (UCB), M. Clampin (NASA Goddard Space Flight Center)

    (Submitted on 13 Nov 2008)

    Abstract: Following the optical imaging of the exoplanet candidate Fomalhaut b (Fom b), we present a numerical model of how Fomalhaut’s debris disk is gravitationally shaped by a single interior planet. The model is simple, adaptable to other debris disks, and can be extended to accommodate multiple planets.

    We find that to not disrupt the belt, Fom b must have a mass < 3 Jupiter masses. Previous mass constraints based on disk morphology rely on several oversimplifications. We explain why our constraint is more reliable. It is based on a global model of the disk that is not restricted to the planet’s chaotic zone boundary.

    Moreover, we screen disk parent bodies for dynamical stability over the system age of 100 Myr, and model them separately from their dust grain progeny; the latter’s orbits are strongly affected by radiation pressure and their lifetimes are limited to 0.1 Myr by destructive grain-grain collisions. The single planet model predicts that planet and disk orbits be apsidally aligned.

    Preliminary analysis of Fom b’s space velocity does not bear this out. The disagreement might be resolved by having additional perturbers in the Fomalhaut system, for which there is independent evidence from the star’s anomalous Hipparcos acceleration.

    Our upper mass limit of 3 Jupiter masses for Fom b is not affected by these considerations. The belt contains at least 3 Earth masses of solids that are grinding down to dust. Such a large mass in solids is consistent with Fom b having formed in situ.

    Comments: Accepted to the Astrophysical Journal, 17 pages, 11 figures. Companion theory paper to Science discovery paper. See the movie at this http URL

    Subjects: Astrophysics (astro-ph)

    Cite as: arXiv:0811.1985v1 [astro-ph]

    Submission history

    From: Eugene Chiang [view email]

    [v1] Thu, 13 Nov 2008 18:23:33 GMT (4718kb)


  • ljk November 13, 2008, 23:45

    Optical Images of an Exosolar Planet 25 Light Years from Earth

    Authors: Paul Kalas (1), James R. Graham (1), Eugene Chiang (1,2), Michael P. Fitzgerald (3), Mark Clampin (4), Edwin S. Kite (2), Karl Stapelfeldt (5), Christian Marois (6), John Krist (5) ((1) Astronomy Department, University of California, Berkeley, CA, USA (2) Department of Earth & Planetary Science, University of California, Berkeley, CA, USA (3) Lawrence Livermore National Laboratory, Livermore, CA, USA (4) Goddard Space Flight Center, Greenbelt, MD, USA (5) Jet Propulsion Laboratory, Pasadena, CA, USA (6) Herzberg Institute for Astrophysics, Victoria, BC, Canada)

    (Submitted on 13 Nov 2008)

    Abstract: Fomalhaut is a bright star 7.7 parsecs (25 light years) from Earth that harbors a belt of cold dust with a structure consistent with gravitational sculpting by an orbiting planet.

    Here, we present optical observations of an exoplanet candidate, Fomalhaut b. In the plane of the belt, Fomalhaut b lies approximately 119 astronomical units (AU) from the star, and within 18 AU of the dust belt. We detect counterclockwise orbital motion using Hubble Space Telescope observations separated by 1.73 years.

    Dynamical models of the interaction between the planet and the belt indicate that the planet’s mass is at most three times that of Jupiter for the belt to avoid gravitational disruption. The flux detected at 800 nm is also consistent with that of a planet with mass no greater than a few times that of Jupiter.

    The brightness at 600 nm and the lack of detection at longer wavelengths suggest that the detected flux may include starlight reflected off a circumplanetary disk, with dimension comparable to the orbits of the Galilean satellites. We also observed variability of unknown origin at 600 nm.

    Comments: 25 pages; 4 tables; 4 figures. To appear in Science November
    13, 2008

    Subjects: Astrophysics (astro-ph)

    Cite as: arXiv:0811.1994v1 [astro-ph]

    Submission history

    From: James R. Graham [view email]

    [v1] Thu, 13 Nov 2008 15:28:59 GMT (1709kb)


  • xipehuz November 14, 2008, 6:13

    Have a look at today’s APOD picture (http://antwrp.gsfc.nasa.gov/apod/ap081114.html).

    It’s so exciting to really “see” those planets instead of “reading” them on a graph.

  • Freederick November 14, 2008, 6:45

    Two quotes from the article:
    “If we’re seeing light in reflection, then it must be because Fomalhaut b is surrounded by a planetary ring system so vast it would make Saturn’s rings look pocket-sized by comparison.”

    “Fomalhaut b is also a mysterious place in at least one sense: It has dimmed by half a stellar magnitude from 2004 to 2006, perhaps an indication of a hot outer atmosphere affected by convection cells.”

    If the brightness is due to reflection off the ring system, then any variation caused by convection in the atmosphere would be negligible. The way I see it, there is no mystery to the dimming at all: it is a consequence of our viewing the reflective ring at a different angle w.r. to the light source, as a result of orbital motion. Adding to this is the presence of the dust belt; if it has any circumferential density variations (“spokes”), it could be obscuring the planet to a differing degree at different points.

  • andy November 14, 2008, 7:44

    Yet another planetary system architecture that seems radically different from our own solar system: the traditional models of core accretion find it difficult to form Uranus and Neptune at their current distances from the Sun, yet Fomalhaut has no problems in having a jovian located beyond 100 AU, and HR 8799 has several superjovians in the outer system, despite being substantially metal-poor. Nevertheless the system architecture of HR 8799 does not appear to be hierarchical, so it’s easier to think of these objects as planets rather than sub-brown dwarfs. These are bizarre and alien systems… it’s definitely starting to look like the massive disks around A-type stars allow for modes of planet formation which aren’t seen very often at all in the solar mass domain.

  • James M. Essig November 14, 2008, 9:48

    Hi Folks;

    Since the star around which the planet was detected is 100 million times brighter than the planet in IR light, it would seem that being able to detect Earth-like planets is just around the corner.

    The Sun puts out about 4 x 10 EXP 26 Watts, while the Earth receives about 10 EXP 17 Watts of incident solar power. On average, the Earth must reflect and re-emit as black body radiation just about all of the solar energy that falls on it, otherwise the Earth’s biosphere would be brought to an oven like temperature in short order. Assumming that about half of the Sun’s output is within the IR and about half of the infalling solar radiation on Earth is immeadiately reflected and the other half is ultimately re-emitted as IR blackbody radiation, the Earth would appear about 1/4,000,000,000 as bright as the Sun from the perspective of another star system in IR light.

    Thus, we should in theory be able to detect Earth-like planets around other stars once we are able to detect such planets that are 40 times dimmer relative to their parent star compared to the value of the 100 million times dimmer IR view of the subject planet around Fomalhaut relative to Fomalhaut itself.

    The above simple calculations are not meant to be patronizing nor insulting, but are merely an example of perhaps how close we are to being able to image any Earth-like planets aroung our stellar neigboors.

    I would have never have guessed 20 years ago, that by indirect methods, we would have disovered over 200 extrasolar planets over the following 20 years, and that is exactly what we did. I think the field of extrasolar planetary dynamics is about to heat up in a big way over the following 20 or 30 years. Things should get real interesting.



  • andy November 14, 2008, 10:50

    If the brightness is due to reflection off the ring system, then any variation caused by convection in the atmosphere would be negligible. The way I see it, there is no mystery to the dimming at all: it is a consequence of our viewing the reflective ring at a different angle w.r. to the light source, as a result of orbital motion.

    The problem with that idea is that the planet hasn’t actually travelled very far around its orbit in the time it has been observed, I doubt the change in phase angle of any ring system would cause the observed variations.

  • philw1776 November 14, 2008, 11:40

    Assuming Kepler reaches orbit next year, after under 3 years of observations we should detect many Earth sized planets, some in HZs. Kepler will give us a real statisticly significant audit of sizes of planets and orbital locations for all spectral type stars. It will be interesting to see if there are differing planetary formation mechanisms and planetary system structures for A vs G and M type stars. Detecting Earth sized planets in HZs will encourage further efforts in imagery and spectroscopy. Note that most of Kepler’s detections will be thousands of LY distant, but the statistics will be revealing.

  • ljk November 15, 2008, 10:40

    Sky & Telescope’s report on the new exoplanet images:


  • Ronald November 15, 2008, 20:11

    Further to James’s encouraging comment of November 14th at 9:48, I wonder (again) whether such direct imaging of earthlike planets near relatively nearby stars would be theoretically possible by means of ground-based systems, rendering the more expensive and risky space-based platforms unnecessary.
    For instance something like the planned Extremely Large Telescope (ELT) combined with advanced adaptive optics.

  • Adam November 16, 2008, 5:37

    Hi All

    A pertinent study by Stamatellos & Whitworth…

    The formation of brown dwarfs and low-mass stars by disc fragmentation

    We suggest that a high proportion of brown dwarfs are formed by gravitational fragmentation of massive, extended discs around Sun-like stars. We argue that such discs should arise frequently, but should be observed infrequently, precisely because they fragment rapidly. By performing an ensemble of radiation-hydrodynamic simulations, we show that such discs typically fragment within a few thousand years to produce mainly brown dwarfs (including planetary-mass brown dwarfs) and low-mass hydrogen-burning stars. Subsequently most of the brown dwarfs are ejected by mutual interactions. We analyse the properties of these objects that form by disc fragmentation, and compare them with observations.

    …their simulation and concept isn’t the only one. Several different versions of the same basic idea have been hitting the arXiv in the last year. Seems the time is ripe!

  • John Hunt November 16, 2008, 11:05

    Here is a image of the movement of Fomalhaut b planet over a period of two years.


  • andy November 21, 2008, 8:14

    Seems to be the season for imaging planets around A-type stars: from ESO…

    Beta Pictoris planet finally imaged?

    Like Fomalhaut b, the Beta Pictoris planet has been suspected for a long time. However the separation is much more “normal” – it seems to be located at roughly the same kind of distance from its star as Jupiter (once you scale it to allow for the higher luminosity of Beta Pic)