A Transiting Jupiter Analog

by Paul Gilster on March 4, 2016

David Kipping and colleagues have discovered what they describe as ‘the first validated transiting Jupiter analog,’ a planet orbiting the K-class dwarf KIC-3239945. Kepler-167e is about 90 percent the size of Jupiter and orbits its star at about twice the distance that the Earth orbits the Sun. Given the fact that the star is cooler than the Sun — an orange rather than a yellow dwarf — temperature estimates for the planet are in the 130 K range, only about 20 K warmer than Jupiter. The discovery is discussed on the Cool Worlds YouTube channel, an outreach project launched by the Cool Worlds Lab at Columbia University, and is the subject of a paper submitted to The Astrophysical Journal (citation below).

Kepler-167e isn’t just another ‘hot Jupiter’, a class of worlds that is well populated. ‘Hot Jupiters’ occupy orbits extremely close to the parent star. Finding a true Jupiter analog — i.e., a planet in a close to circular orbit in a position roughly analogous to Jupiter’s in our Solar System — is a considerably more difficult challenge because of the nature of our detection methods. Radial velocity techniques readily detect planets close to their stars but become far more challenging with increasing orbital distance. Transits, meanwhile, have proven their worth through the Kepler mission and elsewhere, but a planet as far from its star as Kepler-167e will show transits separated by years. That demands patient observation, and raw luck.

Consider: Kepler-167e transits its star every 2.9 years. Kipping points out that the primary Kepler mission only looked at this star for 4.3 years. We are lucky, in other words, that we got two transits, when conceivably we might have detected just a single transit right in the middle of the observation period. A single transit, of course, would yield no idea of the orbital period of the planet. But as Kipping goes on to say in the video, the transit method has additional virtues:

“The transit method can actually do a lot more than other methods. We can not only tell [the planet’s] size and its inclination, we can even go look for moons, for rings, we can even measure the atmosphere of exoplanets. And you just can’t do that with other techniques.”

True enough, and of course Kipping’s work at the Hunting for Exomoons with Kepler (HEK) project is all about exomoon detection. We’re now getting into interesting territory indeed, for we know that in our own Solar System, the gas giants Jupiter and Saturn are accompanied by numerous icy moons. If we begin finding more and more Jupiter analogs, we may be upping the stakes for the detection of exomoons, a hunt which has so far been conducted on worlds orbiting their stars closer than the Earth orbits the Sun. Kipping calls Jupiter analogs ‘a promised land for exomoon hunting.’ And let’s not forget the possibility of rings.


Image: David Kipping (Columbia University) delivering a guest lecture at Harvard on “Life as a Planetary Phenomenon.” Click to watch the lecture on YouTube.

Are two transits enough to declare this a planet candidate? I asked Kipping about this in an email, and this is his response:

Yes, historically the standard has been three (or more) transits. I suspect this stems from the fact early transit surveys were exclusively ground-based, which have plenty of data gaps. With just two transits and lots of gaps in between, you can’t be sure the difference in time between the two is the real period, it could a half or a third that value. For Kepler-167e though, we have continuous photometric coverage of the star (from Kepler) in the 1071 day gap between the two events, and so we can rule out the period being some integer fraction of 1071 days.

The paper on this work offers further insights. One issue at stake is the planetary architecture of the system around KIC-3239945. Inside the orbit of Kepler-167e we find three other planets, each of them about twice the size of Earth. In some sense, then, this system resembles ours except that these inner planets are close to their star. We see a compact and transiting inner system, then a large cavity and then a transiting Jupiter analog. The paper speculates that because transit surveys have poor sensitivity to long-period planets, it may be that case that outer planets like Kepler-167e will be found in many compact multi-planet systems.

The paper goes on to suggest that radial velocity surveys targeting the brighter multi-planet systems found by Kepler could eventually resolve this issue. And note this:

Additional non-transiting planets could reside in the Kepler-167 cavity, although the known four planets display remarkable coplanarity and low eccentricities, suggestive of a dynamically cold system. Amongst the inner planets, the planet sizes increase as one approaches the parent star. Using the Chen & Kipping (2016) mass-radius model, we estimate that the inner two planets are most likely gaseous worlds whilst the outer planet is most likely rocky… Whilst this pattern ostensibly jars our anthropocentric prior, as well as the expected outcome of photo-evaporation (Lopez & Fortney 2013), Ciardi et al. (2013) find that there is no preferential ordering of compact Kepler multis for planets R ≲ 3R.

It’s always useful, in my estimation, to give a good jolt to our anthropocentric assumptions! Transiting Jupiter analogs may turn out to be a very helpful kind of exoplanet, leading us to our first exomoons and, of course, offering the opportunity for analysis of their atmospheres through transmission spectroscopy. All in all, Kepler-167e is a satisfying find indeed.

The paper is Kipping et al., “A Transiting Jupiter Analog,” accepted at The Astrophysical Journal (preprint).


Michael March 5, 2016 at 3:48

Here is a neat app that can be downloaded from NASA, it allows you to view the numerous exoplanets out there.


Ron Smith March 5, 2016 at 17:45

I just watched “How Unique is Earth” on “Cool Worlds” and was confused by the statement that the primary Kepler mission has failed to find any Earth analogs orbiting a sun like star. My understanding was that only a fraction of the planets detected by Kepler have been confirmed, and there are still over a thousand candidates awaiting confirmation.
To make such a broad statement seems very premature, unless NONE of the remaining candidates are possible Earth analogs.

Andrew LePage March 7, 2016 at 13:30

Stating that “the primary Kepler mission has failed to find any Earth analogs orbiting a sun like star” is premature (as well as a bit inaccurate) since the detailed analysis of the data from Kepler’s primary mission and making the required followup observations to confirm finds are still taking place as we speak. While no *confirmed* truly Earth-size planets have been found orbiting inside the habitable zone of a Sun-like star *YET*, extrapolations from Kepler’s confirmed finds to date suggest that about one in ten Sun-like star will have such planets.


Harry R Ray March 6, 2016 at 14:15

This is the BEST CHANCE BY FAR for transiting exomoons, because some almost certainly EXIST! However, unless one of them is HUGE, with ONLY TWO TRANSITS, the chances of EXTRACTING one out of the existing data is VIRTUALLY NIL!

Michael March 8, 2016 at 14:42

yip, but they could detect indirect signs such as radio emissions due to interacting ions and planet-moon currents with magnetic fields.


andy March 6, 2016 at 18:11

As noted in the paper, Kepler-167e is a “degenerate world”: it lies in a region where objects with a wide range of masses have very similar radii. The 68.3% credible interval for the mass spans 0.3 to 50 Jupiter masses. Interestingly the system is sufficiently coplanar that the inner planets are also transiting, this might argue for the true mass being at the lower end of the range (a Saturn-mass world may be less likely than a superjovian or brown dwarf to perturb the system into non-coplanar orbits).

RobFlores March 7, 2016 at 12:00

R. Smith. And comments on Jupiter analogue below.

Maybe you missed the dire set back to the Kepler program. Due to
unforeseen extra turbulence on a main sequence stars surface , Kepler CANNOT detect Earth sized objects and smaller in the Habitable ZONE of a sun like stars. It can detect 2x earth’s sized worlds and bigger, but our sized worlds at the HZ are out of it’s reach which is frustrating, since Kepler was designed to detect
earth like worlds from the start.

Regarding the Close in Jupiter analogue: Could this be the basis of
Rule of thumb regarding Jupiter type planets. The rule might say
The closer a Jupiter sized planet is to it’s primary the more compact the system is overall and more likely to have super terrestrials close in. We
would need a few more findings of this type no doubt. But we could reverse
engineer this “rule” by trying to find Jupiters around systems known to have close in Super terrestrial planet close in. ( I does sound as if they were systematically trying to do this for their own ex0-moon detection)

Harry R Ray March 8, 2016 at 10:17

In his Youtube “Cool Worlds” video. Kipping states that he WILL try to find (an)exomoon(s) around Kepler 167e. Good luck to him.

Ron Smith March 9, 2016 at 23:54

Thanks Rob.
I had heard that there were problems resulting from unexpected variability on sun-like stars, but I did not know it was so bad as to exclude detection of Earth analogs. WHAT A BUMMER!

Comments on this entry are closed.