≡ Menu

Where to Look for Rocky Planets

If you were trying to identify the kind of star that should produce Earth-like planets, you’d think the task would be straightforward. Our theories of planet formation focus on a circumstellar disk around a young star out of which planets form, and we’ve already gathered evidence that gas giant worlds are more likely to form around stars that are rich in iron. Since rocky planets are rich in iron and silicon, doesn’t this mean their stars should be rich in metallic elements?

New work out of the Carnegie Institution for Science suggests that the answer is surprisingly complex. As presented by Johanna Teske at the Extreme Solar Systems III meeting in Hawaii, the team’s work has revealed that smaller planets do not require high iron content in their parent stars. In fact, looking at the abundance of 19 different elements in seven stars orbited by at least one rocky planet (these are drawn from the Kepler catalog), the team finds that rocky worlds do not preferentially form around stars rich in metallic elements like iron and silicon.

The data were taken as part of the Kepler Follow-up Observing Program and supplemented by independent observations using the 10-meter Keck telescope and the HIRES echelle spectrometer. From the paper:

The metallicities of the seven stars range from [Fe/H] = −0.30 to +0.15, further demonstrating that small planets form around stars with both low and high metallicities… The abundances of 15 elements are compared to those of a Galactic disk population that includes stars with and without known planets. The abundances of the Kepler planet host stars fall along the Galactic trends, suggesting that stars with small planets have compositions that are typical of the Galactic disk.

Stars with small planets, then, are nothing exceptional in terms of their metallicity. This implies they are common throughout the galaxy. The paper goes on:

Moreover, comparing the abundance distributions of the Kepler planet host stars to a sample of stars with no detected planets, a sample of Neptune- and super-Earth-size planet hosts, and a sample of Jupiter-type giant planet hosts, reveals that the former three – Kepler host stars, stars without known planets, and the Neptune and super-Earth-size planet hosts – may have similar compositions, whereas the Kepler host stars appear to have compositions that are less enhanced than those of Jupiter-type planet hosts.


Image: This figure from the paper shows the abundance of different elements in stars versus their abundances of iron. In each square, you can see a plot of the abundance of one element (represented by [x/Fe]) against the abundance of iron (represented by [Fe/H]). Each red dot, black square, or blue X represents a star. The red dots are the small planet-hosting stars studied in this new work. You can see how they do not stand out from the rest of the stars, which were studied in other publications, some of which host planets and some of which have no known planets. The green dashed lines show these values for our Sun. Credit: NASA Ames/JPL-Caltech/Tim Pyle.

Small, rocky planets appear to form around stars with a wide range of elements, whereas gas giants are more likely around stars rich in iron. Teske adds in this Carnegie news release that the results are exciting “…because they mean that small planets are very common and chemically diverse.” What we don’t yet know is whether the planet formation process depletes stars of some of the elements that become concentrated in the planets. None of the seven stars the team studied show a depletion signature, but if subsequent work does uncover such a trend, then looking for stars with specific chemical depletions could be a useful planet-hunting tool.

Does the architecture of a solar system play a role in these trends? With the exception of two planets, the planets of all the stars investigated have semimajor axes smaller than Mercury’s. As the paper notes, the small sample size here limits the statistical significance of the results, so the obvious next step is generating a larger sample of stars with small planets and performing abundance analyses on this set. We’re fortunate that the Kepler mission has identified numerous small planets whose host stars can now be subjected to this kind of study.

The paper is Schuler et al., “Detailed Abundances of Stars with Small Planets Discovered by Kepler I: The First Sample,” accepted for publication in The Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Eric S. December 9, 2015, 14:14

    I’d actually expect the stars w/ Jovian worlds (especially close in ones) to be enriched in iron, etc. because the gravitational jostling of migrating Jupiters ought to sling more material into the star.

  • Michael December 9, 2015, 14:46

    What strikes me is the grouping in the middle of each plotted graph, it appears that the ‘average’ has more chance of having rocky planets, zinc is a outsider. I wonder if a report has been done about the combination with Oxygen as these tend to form higher melting point materials allowing closer proximity to the star than lower melting point materials. Also not all materials can be treated the same, under UV and light electrostatics play an important part in the planet forming process, some materials are more susceptible to static potentials than others. It does have the good news that rocky worlds are quite common.

  • Alex Tolley December 9, 2015, 14:46

    So is this implying that metallicity has no impact on planet formation, or just on rocky planets? Why so few planets in the study – lack of spectral data from the Kepler list?

  • RobFlores December 9, 2015, 16:35

    Very interesting paper, it’s nice to know some sharp teams are able to
    get some base line data from the Kepler detections.
    Unspoken is the question as pertains to astrobiology is: Do those
    planets formed around low Iron stars have enough iron and formation heat
    to maintain a magnetic field, (assuming they have a significant rotation) at
    HZ distances.

  • Stevo Darkly December 10, 2015, 8:17

    Off-topic, but I wanted to make sure you were aware of this …

    Two new papers claim possible discovery of a largish object (maybe a super-Earth or even a brown dwarf) in the distant outer parts of our solar system. (Submitted to ARXIV but apparently not peer-reviewed yet, if I correctly understand how this works.)



  • Ronald December 13, 2015, 9:13

    This very interesting study reminds me of other work in this field, by Asplund, Melendez, Ramirez et al., e.g. ‘The remarkable solar twin HIP 56948, a prime target in the search for other earths’.
    They argue that, at least in solar twins, a relative depletion of refractory (heavier) elements, such as iron and silicon, relative to volatiles, such as oxygen and nitrogen, may indicate the formation and presence of rocky planets.

  • Ronald December 14, 2015, 3:34

    Eric S., Alex T.: yes, that is exactly the point: giant planet formation is indeed correlated with high metallicity. But for small rocky planets there seems to be no clear correlation.
    This in itself, though very relevant, is not entirely new, several other studies point in the same direction.
    And we already know of many low-metallicity stars with small and medium-sized (superearth/gasdwarf to Neptune size) planets, e.g. Tau Ceti, 82 Eridani.
    And of course many, many stars with a wide range of metallicities and Neptune class planets in compact systems, apparently the most common system by far (?).

    My own very premature and somewhat speculative idea now is that small rocky planets, the true terrestrials, can form around solar type stars in at least 3 different ways:
    – In medium-high metallicity systems, such as ours: from leftover material in the inner system, after the giant planet has absorbed most material.
    – In low-medium metallicity systems, the very common compact systems of gasdwarfs and Neptunes: on the outskirt of the compact inner system, where the protoplanetary dustdisk was not too massive. Relevant question is whether this occurrence can coincide with the Habitable Zone.
    – In very low metallicity systems: throughout the system, from inside to outside, because of absence of heavier planets.

    Future observations will tell.

  • spaceman December 16, 2015, 15:09

    Does anyone remember the HST search for transiting hot Jupiter’s in the very low metallicity globular cluster 47 Tucanae? No HJs were found which is not surprising knowing that close-in giant planets are almost exclusively found in high metallicity systems. This new work described here showing that rocky planets can form in low metallicity environments makes me wonder if there are rocky worlds, perhaps even habitable ones in fair numbers in globular clusters such as 47 Tucanae after all. What do others think?

  • Stan Erickson January 8, 2016, 15:43

    Hasn’t anyone proposed a mechanism for this? If there is high angular momentum in the system, matter would fall in slower and the star would start off burning at a smaller size, but then blast still orbiting matter into farther orbits. Metallic atoms might be differentially pushed farther out, where the relative concentration of them would increase enough to form small rocky planets. You want to find a mechanism to change the relative numbers of high and low Z stuff.