Sunday, March 9, 2014

Small Planets . . . . Then All the Rest

Some people start their day by checking stock prices or sports scores. I usually visit the homepage of the Extrasolar Planets Encyclopaedia (EPE) so I can peruse the latest articles and preprints and check the exoplanetary census. So the other morning I squinted and did a triple take when I saw the number. Overnight the census had jumped from below 1100 to almost 1800 planets.

I understood what had happened: this growth spurt coincided with the circulation of two preprints by the Kepler team, detailing several hundred new planets, most in multiplanet systems (Rowe et al. 2014, Lissauer et al. 2014). But I was surprised by the remarkable swiftness with which EPE incorporated those objects into its catalog.

This latest data dump gives us a whole new picture of the otherworldly cosmos (see a 2013 post for perspective). Transiting objects are now the single largest population of exoplanets, representing almost two-thirds of the full sample. They outnumber planets discovered by radial velocity (RV) by more than two to one.

The RV and transiting populations are now sharply distinct. The RV sample is dominated by gas giants: 78% have minimum masses of at least 0.15 Jupiter masses (0.15 Mjup). This sample also includes planets at a wide range of orbital periods, from less than a day to well in excess of 5,000 days. (By comparison, the period of Jupiter is 4,333 days, or just under 12 years.) Half of them have periods shorter than 365 days.

The transiting sample is dominated by low-mass planets: 78% are smaller than 7 Earth radii (7 Rea). It is also dominated by small orbits. The longest period in the sample is 904 days, while more than 99% have periods shorter than 365 days, and 90% have periods shorter than 50 days.

I’ve graphed the radius distribution of transiting planets a few times over the past two years – check out earlier examples here (2012) and here (2013). The current profile is radically different. What used to be a double-peaked distribution is now a mountain of small planets attended by mere foothills of gas giants. Notably, Earth- and Super Earth-size planets (1-2 Rea) are now the most numerous subgroup. Their abundance appears to contradict the findings of Petigura et al. (2013), who argued that planets of 2-3 Rea were the most numerous in the Kepler data, while largely discounting any selection bias against smaller objects (see Au-delĂ  de neuf cents exoplanètes for more discussion).

Figure 1. Radii of 1117 Transiting Planets < 18 Earth Radii

Data were retrieved from EPE on March 8, 2014. The letters M, E, U, S, and J mark the positions of Mars, Earth, Uranus, Saturn, and Jupiter on the scale of radii.

If we consider low-mass planets to be RV planets with masses below 0.15 Mjup and transiting planets with radii smaller than 7 Rea, then low-mass companions now account for 60% of the combined extrasolar population. In short, small planets rule the galaxy.

As in the past, I find that the RV and transiting samples are most useful for discerning patterns in the full population. That’s because the information available from other search methods (direct imaging, microlensing, pulsar timing) is too patchy for most comparisons. Nevertheless, for completeness, here is how those three samples break down: all the directly imaged objects are either gas giants or brown dwarfs; two-thirds of the microlensed objects are probably gas giants; and two out of five pulsar planets are gas giants, making this tiny sample the only group besides transiting planets with a low-mass majority.

Within the transiting sample, Kepler Mission results dominate at all radii smaller than 9 Rea, which now represent the vast majority. Unfortunately, data on most of these small planets are incomplete. Few are associated with information on star distance, spectral type, planet mass, or orbital eccentricity. Radius data enable rough estimates of the masses of planets smaller than 6 Rea (see Puffy Planets), while stellar equilibrium temperature generally correlates with spectral type. Nonetheless, as many recent studies have shown, the parameters of many – if not most – Kepler target stars need further corroboration, and will likely be revised in the future.

Another challenge in using data on transiting planets obtained from EPE is the frequency of errors in its catalog. For example, when I visited the site today, I noticed that Kepler-87 is presented as a two-planet system, even though four planets have been confirmed and published in a peer-reviewed journal. Kepler-52b, with a radius of 2.13 Rea, is assigned a mass of 0.28 Mjup, similar to the mass of Saturn. Kepler-25b, with a radius of 2.71 Rea, is characterized as a Super Jupiter of 9.6 Mjup! Yet both of these planets must be less massive than Neptune.

For well-informed users of EPE, such glitches are simply annoying. As hinted above, though, the transit data have a more recalcitrant drawback, regardless of source: they are extremely sparse for orbital periods longer than 50 days. If our knowledge of the Solar System were limited to the same parameter space, we would think our Sun hurtled through eternity all alone. This limitation underscores the importance of using RV data to illuminate transit data, and vice versa.

In sum, the latest planet dump provides a brand-new perspective on the ensemble of planetary systems in our region of the Milky Way. Still, some things haven’t changed at all. The only plausible candidate for a habitable planet in the full extrasolar sample remains Kepler-62f. None of the new planets can rival its qualifications. Nor has this avalanche of small planets revealed any new compact, mixed mass systems like the ones I described in my previous post.

My own wish list continues to be dominated by follow-up studies on the most fascinating RV and transiting systems – 55 Cancri, GJ 667C, Tau Ceti, Alpha Centauri B, Kepler-62, and Kepler-90.


Lissauer JJ, Marcy GW, Bryson ST, Rowe JF, Jontof-Hutter D, Agol A, Borucki WJ, Carter JA, Ford EB, Gilliland RL, Kolbl R, Star KM, Steffen JH, Torres G. (2014) Validation of Kepler’s multiple planet candidates. II: Refined statistical framework and descriptions of systems of special interest. Astrophysical Journal 784, 44. Abstract:
Petigura EA, Geoffrey MW, Howard AW. (2013) A plateau in the planet population below twice the size of Earth. Astrophysical Journal 770, 69. Abstract:
Rowe JR, Bryson ST, Marcy GW, Lissauer JJ, Jontof-Hutter D, Mullally F, Gilliland RL, Isaacson H, Ford F, Howell SB, Borucki WJ, et al. (2014) Validation of Kepler’s multiple planet candidates. III: Light curve analysis & announcement of hundreds of new multi-planet systems. Astrophysical Journal 784, 45. Abstract:


  1. I'm surprised how little attention Kepler-62f gets - I've heard Kepler-62e get much more attention, even though it has a larger radius and is probably a super-Venus if it isn't a mini-gas giant (the only thing it has going for it is that it's smaller than the 1.75 Earth-radius mentioned in the Lopez-Fortney paper).

    Sure, it only gets the equivalent of 42% of Earth's sunlight, but it's also bigger than Earth and much more likely to hold on to a magnetosphere and warming atmosphere to make up for it.

    1. Amen to that! From what I see in the astrobiological literature, it's seems easier for planets with lower effective temperatures to warm up (through atmospheric greenhouse effects) than it is for planets with higher effective temperatures to cool down. If Mars were more massive, maybe it could retain enough heat to have liquid water.