≡ Menu

The Largest Solar System Yet

The Kepler mission’s exoplanet discoveries have been so numerous that an extension of the mission seemed all but inevitable. At the same time, bureaucracies can be unpredictable, which is why it was such a relief to have the Senior Review of Operating Missions weighing in with an extension recommendation, one followed up by NASA with extensions not just for Kepler but also for the Spitzer telescope and the US portion of ESA’s Planck mission. Kepler’s extension runs through fiscal 2016 (subject to review in 2014), allowing for plenty of time to home in on Earth-sized planets in the habitable zone around stars like our Sun.

While Kepler’s scheduled mission duration was 3.5 years, the mission was intended to be extendible to 6 years or more and this news is more than satisfying. But of course while we continue to monitor the Kepler work, we’re following numerous other exoplanet stories including the European Southern Observatory’s observations of the prolific star HD 10180, a Sun-like star about 127 light years away in the constellation of Hydrus. The ESO work, performed with the HARPS spectrograph attached to the 3.6 meter telescope at La Silla in the Chilean Andes back in 2010, revealed the presence of at least five and possibly seven planets. Now astronomer Mikko Tuomi (University of Hertfordshire) has performed further data analysis on the HARPS radial velocity data, concluding that the system may contain as many as nine planets.

Image: The circle shows the location of the class G near-solar star HD 10180, which may be orbited by as many as nine planets. Credit: Jim Kaler/UIUC.

The two new worlds Tuomi’s work reveals range from 1.9 to 5.1 Earth masses respectively, allowing them to be classified as probable super-Earths with orbits of 10 and 68 days. If the findings are confirmed, this would make the HD 10180 system more fecund than our own, at least in terms of number of planets. Tuomi also reports he has verified the inner planet signature first announced in 2011, indicating a planet on a 1.18 day orbit with a minimum mass as low as 1.3 Earth masses. He has also revised the orbital parameters of the other six planet candidates around the star. Add it all up and you get nine planets. From the paper:

As noted by Lovis et al. (2011), the star is a very quiet one without clear activity-induced periodicities, which makes it unlikely that one or some of the periodic signals in the data were caused by stellar phenomena. Also, the periodicities we report, namely 9.66 and 67.6 days, do not coincide with any periodicities arising from the movement of the bodies in the Solar system. Therefore, we consider the interpretation of these two new signals of being of planetary origin to be the most credible explanation.

If this is borne out, then we have exceeded our Solar System’s planet count for the first time in exoplanet studies. The paper notes the need for additional high-precision radial velocity studies to confirm these findings. And we may not be through. There appear to be stable orbits for a low-mass companion in or near the habitable zone of HD 10180, one whose mass would be unlikely to exceed 12.1 Earth masses based on Tuomi’s samplings. All that should keep this star among our targets for some time to come as the data are mined for confirmation of these worlds.

The paper is Tuomi, “Evidence for 9 planets in the HD 10180 system,” accepted for publication in Astronomy and Astrophysics (preprint).


Comments on this entry are closed.

  • James D. Stilwell April 11, 2012, 10:13

    250 years for a probe to arrive and 127 years for the report back…
    What will our culture be like 375 years hence…A world culture might work…
    When one world becomes really One…We have more to do than build ships…
    Such is life…
    Keep going…

  • kzb April 11, 2012, 13:38

    These exo-systems sure are strange.
    Seven planets, all probably more massive than Earth, orbiting within 0.5AU (two more further out of course, one of which could be around Saturn size).
    The star HD 10180 is about 1.5 times as luminous as the sun (that innermost planet must be HOT at 0.022AU), and over 7bn years old.

  • Dave Moore April 11, 2012, 17:15

    From a moon around HD 10180 g or h, the night sky would be pretty spectacular, especially just after starset. It would be interesting to see an animation of such a thing.

  • andy April 11, 2012, 18:42

    Most populous yes, largest probably not, at least as measured by the orbit of the outermost known planet.

    It is an interesting question as to whether these compact planetary systems are associated with outer systems that range as far from the star as our own outer solar system. Unfortunately outer solar systems containing only low-mass planets (which may well be in the majority) are very hard to detect.

  • Tom W April 11, 2012, 18:56

    So if we can find 9 planets in one system in such a short period of time then you would expect that there are a lot of systems out there with 10+ planets.

  • Joy April 12, 2012, 2:50

    What andy said. Hubble (maybe) directly imaged Fomalhaut b, which is an (unconfirmed by Spitzer) Jovian class planet with a claimed semi-major axis of 115 AU. Due to the observational bias of Kepler, it excels at discovering compact planetary systems. However, there is no reason to believe that these compact systems are representative of extrasolar systems in general. Whether Fomalhaut b is real or not, I choose to believe that there are many systems considerably larger in radius than the Sol system.

  • kzb April 12, 2012, 8:15

    As I understand it, this is a re-interpretation of existing data, based on a different statistical method. The original interpretation (Lovis 2011) gave six planets.

    Since these things are so open to interpretation, how genuinely scientific are these detections? How much does subjectivity creep in to it?

  • Ronald April 12, 2012, 8:47

    Truly fascinating discovery, article and post! A couple of additional observations;

    If we reasonably assume (loosely following Kasting) that our own HZ extends from 0.95 to (at least) 1.2 AU, HD 10180, with a luminosity of 1.45-1.5 * solar, has a HZ from 1.15 to (at least) 1.8 AU.

    According to the same publication and as Paul mentions, there is a rather great statistical likelihood of another planet with greatest probability around super-earth mass, somewhere between f (at 0.5 AU) and g (at 1.4 AU).
    This additional planet could be in the HZ indeed, though the article shows (fig. 5) that the highest probability is somewhere around an orbital period of 240 – 320 days (peak at 290 days), corresponding with about 0.75 – 1.0 AU, well inside of the HZ (i.e. much too hot). Only toward the highest end of the probability curve the planet would just be within the HZ.

    However, planet g at 1.4 AU is well within the HZ by any standards. So this means that HD10180 has at least one (Neptune class) planet and possibly another (probably super-earth class) planet within its HZ.

    Another interesting observation is the phenomenon, mentioned before by others in previous posts, that many, if not most of these (compact) inner planetary systems around (solar type) stars consist of an rather consistent alternation of a smaller (super-earth class, SE) and a larger (subgiant, Neptune class, N) planet. If we check that for the inner system (up to 1.5 AU) of HD10180, and including the above-mentioned probable hypothetical planet ‘x’ as 5*Me at 0.9 AU, we get the following approximate planetary masses, from inside outward, in Me:
    So the sequence is: SE, N, SE, N, N, SE, N, SE, N

    This confirms previous suspicions that there is a remarkable and almost perfect alternation of super-earth and Neptune class planet, with the only exceptions being the two subsequent Neptunes d and e. However, the ratio of orbital periods and semi-major axes (in AU) of e versus d is significantly larger than for the other inner planets, the latter being 2.1, whereas the average for any two subsequent all these inner planets (excluding the innermost) is 1.5, suggesting that there might even be another, small, planet between these two?

    What is the likely explanation for this fascinating phenomenon, or rather, combination of three phenomena:
    1) a predominance of super-earths (SE) and Neptune class (N) planets
    2) a very compact inner planetary system
    3) an alternation of smaller (SE) and larger (N) planets

    And what stellar parameters is it correlated with? Metallicity (about 15-20% higher than solar)? Stellar mass (about 6% greater than solar)? A combination of both, resulting in a greater mass and density of the primordial dust disc? Or its old age (in the article mentioned as approx. solar at 4.3 gy, but by others as around 7 gy) resulting in more inward migration? Or anything else?

    Intuitively, for what that is worth, it does not seem illogical that a large planet during its formation sweeps clean a larger orbital swath of primordial dust, leaving less planetary material for its neighbors.

    Anyway, it would be very interesting indeed to compare more of these multi-planet systems around solar type stars and plot them against some of those stellar parameters.

  • kzb April 12, 2012, 13:15

    Ronald: there could be a 4th possibility in your list, and that is that the pattern is an artifact of the detection algorithm.

  • amphiox April 13, 2012, 0:10

    Keep in mind that this is essentially, only a detection of an inner solar system. There could still be any number of small outer solar system planets on orbits too wide and long to be detected yet.

  • Ronald April 16, 2012, 7:40

    @kzb: can you please elaborate a bit?

    With all that is known right now, including technical limitations and corrections for observational bias, it seems quite unlikely that the observed patters are just artefacts.
    (Besides, I did not mention 3 different possibilities, but 3 phenomena occurring simultaneously).

  • Ronald April 17, 2012, 4:52

    My other favorite website, Next Big Future, has an interesting post with references about the realistic chances for galactic panspermia through large impacts and resulting ejecta:

    This has really fascinating implications with regard to panspermia. I previously thought that the chance of any meteorite reaching another planetary system was simply too small, because, as with any diffuse radiation, the density of ejecta decreases with the square power of distance.
    This would give it a real chance, although a planetary system being visited by microorganisms that originated outside it is not yet the same as a rock actually reaching a potentially habitable planet and the organisms in it surviving the impact.
    But it does offer a real chance for galactic panspermia.
    And within one planetary system is would present a very real case for panspermia: if Mars has or had life, it may have come from Earth, or vice versa: life on Earth may have originated on Mars. Smaller mass planets have a greater chance of ejecta leaving its gravity well.

  • Ronald April 17, 2012, 6:47

    Further to my previous: another interesting article in this respect is:
    about the search for our solar siblings, i.e. solar type stars that originated in the same stellar nursery cluster, and that may have been seeded with early microbes this way.
    Mentioned in particular are: HIP 87382 and HIP 47399, both presently about 100 ly away.
    And now that I am at it: there is a new very interesting article on the most similar solar twin sofar discovered (in our part of the MW), it was already known as such, but now appears to be even more similar to our sun, the researchers even call the similarity ‘astonishing’:
    It also explores the possibility to predict the presence of terrestrial planets on the basis of detailed chemical analysis.

  • Rob Henry April 17, 2012, 20:04

    Ronald, often algorithms have such subtle quirks that it is difficult to prove them beforehand (except by time consuming Monte Carlo simulations). Suffice it to say that every detection flagged may be given the exactly correct probability (to the limits of ability of that algorithm), but that the probability that it is a false detection is much higher for low mass planet than high mass ones, and a small error in the parameters of a higher mass planet, could be interpreted as a second phantom planet within the limits of noise in the system.

  • Ronald April 20, 2012, 3:32

    @Rob Henry: I get your point, but the publication makes clear that the (results of the) algorithm used and the S/N ration are significant. In other words, though even smaller planets may be hiding in between, those detected are most probably really there. Thus, the combination of at least two of the phenomena I mention is real: a compact system consisting of super-earths and Neptune-class planets, and the third phenomenon, the alteration of these two categories is probably also there.

  • Rob Henry April 20, 2012, 18:00

    Ronald, imagine you were writing an algorithm for planet detection. Suppose that you thought that getting the orbital parameters of each planet correct was of upmost importance, and suppose that you were so skilled at this that you wrote the best possible programme to that effect. The question then arises, how different would the algorithm be that was optimised instead to minimise false detections rather than yours which is fantastically good at giving orbital parameters of a system whose number and mass of planets are already known.

    Now you may be thinking the effect of noise would be insufficient to generate many phantom planets, and for the choice of all systems where we are trying to decide whether we have either one or two planets, you would be right. In a nine planet system, we have nine different sets of parameters that we must adjust for from just one set of data, so the effect is huge.

    I will summarise.
    That hypothetical perfect algorithm could spit out, with complete accuracy, that with high S/N the effect given by the phantom planet is needed. However if we simultaneously readjust several parameters, that need might disappear. The cost is an orbital parameter fit that is a tiny bit lower.

  • ljk April 21, 2012, 11:45



    Posted on January 2, 2012

    by Jon Lomberg

    The Kepler Telescope in space is discovering new planets almost too fast to keep track of. And observatories on Earth like Keck and Subaru on Mauna Kea are confirming the findings and adding additional information about the stars these planets orbit.

    Now we are even hearing about exoplanets roughly the size of Earth, even some in habitable zones around their mother star.

    This heady burst of discovery reminded me of a painting I did many years ago, decades before the first exoplanets were found.

    I was inspired by Stephen Dole’s book Planets for Man that suggested how planets might form and how Earthlike planets might be abundant in the Milky Way Galaxy. I was also thinking of Olaf Stapledon’s visionary suggestion in Starmaker that “a constant stream of touring worlds [was] percolating through the Galaxy.”

    And why limit exploration to just our Universe? In my painting a fleet of Earthlike planets is emerging from a supermassive wormhole into another Universe. Of course they have brought a little of their own spacetime with them to buffer any adverse effects from a different cosmos.

    Compared to that, Kepler’s recent discoveries are practically down the street. That’s literally true if you consider our little Orion Arm as a small side street in a megalopolis of stars. The field of view of Kepler is looking “upgalaxy” along the spiral arm, looking towards our Orion Arm’s intersection with the much larger Sagittarius Arm.

    The Galaxy Garden provides a useful way of visualizing Kepler’s discoveries. A 3-foot long section of 1” plastic PVC pipe scales to one kiloparsec, Kepler’s range at the approximate aperture of the field of view of the telescope.

    This short length of pipe gives these planets a context. They are only one footstep away in the Galaxy Garden. From Earth you can almost reach out and touch them (if your arms are a kiloparsec long) At 600 and 1000 light years the two most promising star systems are close neighbors on our little arm of the Milky Way.

  • Ronald April 23, 2012, 5:15

    Rob, sorry, but I find your summary an bit unclear and hard to understand. Do you mean to say, that the researchers were trying very hard to make the data show planets? As a system analyst I do know a bit about algorithms and statistical probabilities and this just seems unlikely as an explanation of the found S/N ratio peaks in the data using the Bayesian model probabilities. The possibility of more (small) planets hiding in the noise is more likely though.

  • Rob Henry April 23, 2012, 17:59

    Ronald, I want to make it clear that the researchers are doing their job well and with perfect objectivity. I would also like to make it clear that the algorithm that they are working with might be exceptionally good. Unfortunately they always have to be optimised for something, and that will bring subtle weakness in other areas. These might only unveil themselves slowly. What you should look out for is how the cases of proposed systems that best match your theories expectations change as further data on these systems is analysed.

  • Rob Henry April 23, 2012, 18:11

    Oh, I should have also said that this type of fault is always aggravated if you keep looking at new systems – so try to avoid that pitfall.

    And I should have mentioned that you have me intrigued as to whether your on to something, and what its implication might be.

  • Ronald April 24, 2012, 6:11

    Rob: thanks, I get your point now and consider it reasonable and valid.
    With regard to me being on to something: to much honor for me as an ‘informed amateur’, I would say.
    But I really intrigued by the question how certain crucial stellar parameters determine and characterize a planetary system.
    I myself have been and still am convinced that certain (solar type) star’s parameters, foremost (stellar mass and) metallicity, should have great diagnostic and predictive value with regard to this (‘knowing the mother is knowing the children’).
    Recently, with Kepler and HARPS results revealing all those compact inner systems, it seemed, very disappointingly, as if stellar type and metallicity are not enough to characterize a planetary system and that stars of very similar type and metallicity may have very different planetary systems as a result of rather chaotic and therefore inherently unpredictable early planet formation history.
    However, this may eventually appear to be just a need for further fine-tuning of stellar (elemental) abundances instead of overall metallicity.
    I.e., a star’s (relative) abundance (ratio’s) of the various elements.
    In particular the astronomers Melendez and Ramirez et al. have done comprehensive research in recent years into the significance of the relative abundances of elements and have come to the some fascinating preliminary conclusions. Also see the article I mentioned on 17 April about solar-twin-hip-56948 and other publications mentioned in there.
    It is not entirely clear anymore (at least not to me) to what extent these relative abundances are the cause of certain planetary system formation, or a result, but that does not matter too much, because in any way they can be *diagnostic* and hence predictive.
    What is pretty certain and telling, is that a (very) large and/or more than one equal mass large giant planet, leading to high eccentricity and/or inward migration of (at least one of) those giant planets is very detrimental to stable inner terrestrial planets. Also known is a strong correlation between (very) high metallicity and (super) giant planet formation.
    Suspected now also is a correlation between a moderate relative abundance (i.e. slightly depleted, not too abundant) of the higher melting point (refractory) elements, such as Fe, Ni, Ca, Mg, Si, Al, relative to the volatiles such as O and C and the occurrence of an inner terrestrial planet system like our own. It seems that (only) about 15 (10-20) % of solar analogs have this. The question is now, whether the others with those compact super-earth/Neptune inner systems correlate with significantly other ratios.

    I suspect that eventually we will be able to distinguish a limited number of main planetary system types (just as we distinguish a limited number of main planet types: gas giant, sub/ice giant (Neptune class), super-earth, terrestrial). Very (VERY) preliminary and prematurely, and only with regard to inner systems (roughly up to 1.5 AU or so), I would guess something like:
    – no planetary system (just dust and rocks);
    – one or two very large hot giant planets;
    – compact system of intermediate size planets (SE, Neptunes);
    – ‘open’ system of small/terrestrial planets
    And this correlating with (relative) elemental abundances.
    Time will undoubtedly tell.