Saturday, August 29, 2015

The Blazing WASP-47s

Figure 1. The three transiting planets of WASP-47 are shown at their relative sizes. The number inside each sphere refers to the planet’s estimated mass in Earth units (Mea). The value for planet e is a guesstimate within narrow limits; the other two are supported by radial velocity and transit timing data.

[Portions of this posting were updated on October 14
to reflect the published version of Becker et al. 2015]

A few weeks ago WASP-47b was just another “entirely typical” Hot Jupiter (Hellier et al. 2012), similar to hundreds of others in the exoplanetary catalog. Its primary is a G-type star almost identical in mass to our Sun, while the planet itself is a gas giant almost identical in mass to Jupiter (though its radius is a bit larger and its “year” is only 4 days, compared to 4,333 days for Jupiter). As of early August, 20 years of observations had detected very few planetary companions to Hot Jupiters (defined as planets of 60 Mea or more with bulk compositions dominated by hydrogen and helium, traveling on orbital periods shorter than 10 days). Those few companions were always other gas giants on much wider orbits. Not even the Kepler Mission identified a Hot Jupiter with a nearby Hot Neptune or Super Earth. All available evidence indicated that Hot Jupiters were inimical to the compact, low-mass systems detected in such abundance by Kepler.

Two weeks ago, everything changed. Juliette Becker and colleagues announced that the K2 Mission has confirmed two-low mass planets within hailing distance of WASP-47’s blazing giant: one just inside its orbit, the other just outside it (Figure 1, Table 1). Even more than the other known mixed-mass systems, this one provides critical evidence about how planetary systems form. 

Table 1. WASP-47 System Parameters (updated)
Column 2 shows the radius in Earth units (Rea); column 3, the mass in Earth units (Mea); column 4, the semimajor axis in astronomical units (AU); column 5, the orbital period in days. Becker et al. 2015 note that all orbital eccentricities are smaller than 0.05 and describe their radius values as accurate within 4%. Semimajor axes for planets d and e are estimates based on planets with similar masses, orbits, and host stars, since Becker et al. provide no semimajor axis for either one. The data in this table have been updated to reflect the published version of Becker et al. 2015.
The host star is located about 200 parsecs (650 light years) away, which happens to be a typical distance for Kepler systems. The star is slightly more massive than our Sun (1.04 ±0.08 Msol), slightly larger (1.15 ±0.04 Rsol), a little cooler (5576 K, compared to 5778 K for the Sun), and much richer in metals, with [M/H] equal to +0.36. In other words, the star is just as typical of Hot Jupiter hosts as the planet is typical of Hot Jupiters.

That is, except for those two unexpected companions.

The inner planet, WASP-47e, is a likely Super Earth with a radius of 1.82 Rea. This value falls within the “zone of avoidance” noted by Wolfgang & Lopez (2015), who characterize the zone in terms of radii between 1.7 and 2.0 Rea. Within this range, current theoretical models have difficulty distinguishing between planets made of rock and planets made of rock plus hydrogen/helium (H/He). Wolfgang & Lopez suggest that objects of this size might incorporate some fraction of astrophysical ices in their bulk composition, depending on their mass. 

Before publication, Becker and colleagues indicated a maximum mass of 8.9 Mea for this planet, but in the final published version of their study, that upper limit increased to 22 Mea. Unfortunately, the new value provides no insight into the composition of WASP-47c: the maximum mass for an iron-rich planet of this radius, assuming maximum collisional stripping, is only 15 Mea (Zeng & Sasselov 2013). With an iron/silicate composition like Earth’s, WASP-47c would be 8.5 Mea (Dressing et al. 2015). Adding more and more ice would result in lower and lower mass values. 

The nature of WASP-47d, the outer planet, is a bit easier to discern, even though its mass isn’t especially well constrained. At any mass within the range provided (8.2-22.2 Mea), this planet must support an H/He envelope. The mass preferred by the discovery team (15.2 Mea) implies a smaller, warmer, and denser version of Uranus. Smaller and larger values, respectively, would lead to a more or less puffy planet. 

Notably, the orbital period of planet d places it just outside a 2:1 mean motion residence with planet b, the Hot Jupiter. Similar relationships are observed in a significant fraction of Kepler systems, with the ratio of the orbits always a little larger (rather than a little smaller) than a value consistent with resonance (Steffen & Hwang 2015). Such a relationship between two adjacent planets has often been interpreted as evidence that the planets underwent convergent migration to the vicinity of their present orbits, but then escaped a resonant lock through the effects of orbital eccentricities, stellar tides, interactions with a remnant planetesimal disk, or other factors (Batygin 2015, Chatterjee & Ford 2015). 

why this discovery changes everything 

Shortly after the first Hot Jupiters were announced in the mid-1990s, theorists began to explain their bizarre characteristics by invoking a process of core accretion followed by Type II migration. In this model, gas giants form by very fast (“runaway) accretion of H/He envelopes onto rocky cores in cool outer regions of the protoplanetary nebula – that is, beyond radii of about 2 astronomical units (AU). Once they reach a certain mass (in the approximate range of Saturn and Jupiter), these baby giants are rapidly carried by interactions with the nebula to the immediate vicinity of their parent stars. According to this scenario, Hot Jupiters would likely scatter or engulf any planets that had already formed in the inner nebula, but their passage to the threshold of the star would also bring additional mass in protoplanets into these warm regions and encourage the accretion of Super Earths and perhaps Exo-Neptunes, both inside and outside the Hot Jupiter’s final orbit (Raymond et al. 2006; Fogg & Nelson 2006, 2007; Mandell et al. 2007). Of necessity, the orbits of such companion planets would be mutually well-aligned (“coplanar”) with those of the Hot Jupiter.

Yet none of these hypothetical familiars were ever detected by ground-based or space-based telescopes, regardless of method (radial velocity, transit, or transit timing). At the same time, observations of the Rossiter-McLaughlin effect during Hot Jupiter transits demonstrated that many of these objects are dramatically misaligned with the spin axis of their parent stars (Triaud et al. 2010). Since extreme misalignment appears to be inconsistent with smooth Type II migration, many theorists began to argue that Hot Jupiters often, or even primarily, form by high-eccentricity migration (Nagasawa et al. 2008, Naoz et al. 2012, Mustill et al. 2015). In this scenario, a gas giant planet on a relatively cool orbit is perturbed by another massive object (either another gas giant in the same system or an eccentric binary companion of the host star) into an elliptical orbit that carries it within a few million kilometers of the stellar chromosphere at its closest approach. Stellar tides then circularize the gas giant’s orbit until it arrives at its final period of a few days.

Even when dynamic interactions are insufficiently violent to create a Hot Jupiter, other research has shown that planet scattering in systems containing three gas giants tends to destroy any terrestrial planets that managed to form in the inner system (Matsumura et al. 2013). The same process typically eliminates one of the three giants, and often two of them. Thus it appears that whenever gas giants gang up, mayhem ensues, and the smallest planets within reach bear the brunt of the damage (see Dwarfs vs. Giants, Round Two).

The architecture of WASP-47 demonstrates that in some fraction of planetary systems, dwarfs and giants coexist amicably at close quarters, even on the burning threshold of their parent star. In such peaceable systems, Hot Jupiters actually can form by a smooth process of Type II migration, and their interactions with the primordial nebula permit the formation and survival of nearby low-mass planets.

Evidence from the original Kepler Mission suggests that such systems are very rare – much rarer than compact low-mass systems that lack any gas giants. But maybe they are no rarer than systems containing rocky planets with tenuous atmospheres traveling in the vicinity of the local habitable zone, plus one or two gas giants following stately orbits at a cooler distance (i.e., systems like home). 

implications for planetary systems across the galaxy 

One simple formalism for understanding system architectures considers the masses and radii of adjacent planets. Assuming that we have two types of planets (gas giants and low-mass) and accepting three planets as the minimum architectural unit for a multiplanet system, we get eight possible configurations: 

Figure 2. Triplanetary units of system architecture
Blue circles indicate low-mass planets; red circles indicate gas giants.
According to the schema in Figure 2, WASP-47 is a perfect 3, while the eight planets of our Solar System comprise six successive units: 1-1-2-5-7-4 (i.e., Merc-Venus-Earth; Venus-Earth-Mars; Earth-Mars-Jup; Mars-Jup-Sat; Jup-Sat-Ura; Sat-Ura-Nep). Among other well-studied systems, the four planets of Mu Arae constitute a 5-8, while the four planets of GJ 876 make a 5-7. Upsilon Andromedae, HIP 14810, HD 82943, and HD 37124 are each a perfect 8. Kepler-11 is a double 11 (1-1-1-1); Kepler-20, Kepler-62, and Kepler-186 are each 1-1-1; and Kepler-289 is a 2.

It’s immediately apparent that some configurations are more common than others. Type 1 is extremely prevalent in Kepler data, while 3 is rare across all catalogs. Apart from WASP-47, only two other examples of that configuration are known: Kepler-87 (2-3) and Kepler-89 (also 2-3). Type 4 remains unattested outside our Solar System, while type 6 is completely hypothetical.

Theories of planet formation and system evolution need to account for every attested configuration, as well as for their relative frequency in known system architectures. As I’ve pointed out before, virtually all existing theories ignore types 2, 3, and 4. When these occur in compact architectures, as in WASP-47, Kepler-87, Kepler-89, Kepler-289, and HD 219134, in situ models fail (e.g., Hansen & Murray 2012, Chiang & Laughlin 2013), since the gas giants in these configurations could not have formed on their present orbits. Even a recent theory that incorporates orbital migration along with core accretion is described by its authors as inapplicable to systems with configurations 2 through 4 (Izidoro et al. 2015). So according to me, at least, exoplanetary theorists have their work cut out for them.

Even though earlier studies predicted that Hot Jupiters undergoing Type II migration would acquire low-mass companions, none of the many numerical simulations undertaken to model those scenarios produced a system with a gas giant, a Super Earth, and a Hot Neptune orbiting within 0.1 AU. As such, WASP-47 poses a truly fundamental mystery. 

Note: If Figure 2 looks familiar from somewhere else, it should. A similar recombination of two variables in sets of three units each is the basis of the 64 hexagrams of the I Ching or Book of Changes, the Chinese classic on divination. In Chinese, the trinary building blocks of the hexagrams are collectively known as Ba Gua, usually translated Eight Trigrams. Figure 3 illustrates an arrangement of the Ba Gua matched to the eight triplanetary units in Figure 2: 

Figure 3. The Eight Trigrams or Ba Gua of the I Ching
Solid lines are Yang; broken lines are Yin. In each trigram, lines are read from the bottom up.
Oddly enough, the triple Yin configuration known as Kun, which corresponds to #1 in Figure 2, is translated in English as Earth. Is this a subtle hint from the Taoist Immortals that we are most likely to find another Earth in a compact system of low-mass planets? 

Postscript, September 20, 2015 

On Thursday, a team led by Roberto Sanchis-Ojeda reported ground-based observations of a very recent transit of WASP-47b, the system’s Hot Jupiter. Their measurements indicate that the planet’s orbit is well aligned with the spin of the host star. The stellar obliquity is consistent with zero, plus or minus 24 degrees. The authors confirm the rarity of the system architecture of WASP-47 by surveying current exoplanetary catalogs. The suggest a family resemblance between WASP-47 and 55 Cancri, both of which have an ultra-short period planet orbiting interior to a gas giant with a period shorter than 15 days.

Sanchis-Ojeda's group notes that all available evidence supports the conclusion that compact multiplanet systems are well-aligned with their host stars. They also note in passing that Neveau-VanMalle and colleagues have radial velocity data indicating a second gas giant in the WASP-47 system, with a much wider orbit than the planets reported to date. (A manuscript is in preparation but not yet available for general consumption.) This news puts a whole new wrinkle in our picture of the system architecture. As Sanchis-Ojeda’s group reports, all Hot Jupiters previously known to be accompanied by another planet have a second gas giant on a much wider orbit. For a moment it looked like WASP-47 was the exception. Yet despite its weirdness in so many other ways, this system evidently conforms to the norm, at least in this regard. 

Postscript, October 5, 2015 

Last week, Neveu-VanMalle and colleagues posted their preprint reporting the discovery of a second gas giant orbiting WASP-47, which they named WASP-47c. It has a minimum mass of 1.24 Mjup, a semimajor axis of 1.36 AU, and an orbital period of 572 days. Its orbital eccentricity is poorly constrained, with estimates ranging from 0.03 to 0.23. Although the inner planets are co-planar, we have no information on the inclination of this new planet.

To my surprise, WASP-47c is very similar in mass WASP-47b. Theoretical studies often find that gas giants of equal mass are more likely to engage in scattering than those of unequal mass. For example, theories that seek to explain the relatively large eccentricities of most extrasolar gas giants on periods longer than 10 days have typically invoked dynamical interactions among three equal-mass giants (Chatterjee et al. 2008, Matsumura et al. 2013). In the Solar System, the relatively smooth inward and outward migration proposed for Jupiter and Saturn in the Grand Tack scenario is a function of the two planet’s unequal masses (Pierens & Raymond 2011). Given that context, it seems odd to find such a placid, well-organized system of four planets featuring giants that are almost twins. 

Figure 4. Architectures of mixed-mass systems
Solid circles represent secure radial velocity detections; open circles represent transit detections. A few of the radial velocity planets have also been observed in transit. ME = Earth mass; RE = Earth radius. Objects more massive than 55 ME or larger than 8 RE are assumed to be gas giants.
Figure 4 presents all systems reported to date with at least two low-mass planets and at least one gas giant. For some systems, only radial velocity data are available; for others, only transit data; but for five systems in this figure (Kepler-89, WASP-47, Kepler-68, Kepler-48, and 55 Cancri), at least one planet was observed in transit and at least one was detected by radial velocity measurements. (HD 219134 might qualify as a sixth such double-duty system, pending confirmation of the single reported transit.)

Remarkably, the new planet proposed for WASP-47 fills the only remaining gap in my I Ching-inspired deconstruction of system architectures (Figure 2). Just a few weeks ago, type 6 (“Li”) was unknown. Now we see that this system features a giant-dwarf-giant configuration within 2 AU of the central star. True, we might eventually find additional low-mass planets orbiting in the space between WASP-47d and WASP-47e, but for now we see the equivalent of a smartcar trapped between two 18 wheelers.

It looks like the abundance of planetary systems in the Milky Way can match every physical configuration we can imagine. I’d say that’s a good thing.

Batygin K. (2015) Capture of planets into mean-motion resonances and the origins of extrasolar orbital architectures. Monthly Notices of the Royal Astronomical Society 451, 2589-2609.
Becker J, Vanderburg A, Adams F, Rappaport S, Schwengler H. (2015) WASP-47: A Hot Jupiter system with two additional planets discovered by K2. Astrophysical Journal Letters 812, L18. Abstract: 2015arXiv150802411B
Chatterjee S, Ford EB, Matsumura S, Rasio F. (2008) Dynamical outcomes of planet-planet scattering. Astrophysical Journal 686, 580-602.
Chatterjee S, Ford EB. (2015) Planetesimal interactions can explain the mysterious period ratios of small near-resonant planets. Astrophysical Journal 803, 33.
Chiang E, Laughlin G. (2013) The minimum-mass extrasolar nebula: In situ formation of close-in super-Earths. Monthly Notices of the Royal Astronomical Society 431, 3444-3455. Abstract: 2013MNRAS.431.3444C
Dressing C, Charbonneau D, Dumusque X, Gettel S, Pepe F, Collier Cameron A, and 29 others. (2015) The mass of Kepler-93b and the composition of terrestrial planets. Astrophysical Journal 800, 135. Abstract:
Fogg MJ, Nelson RP. (2007) On the formation of terrestrial planets in hot-Jupiter systems. Astronomy & Astrophysics 461, 1195–1208. doi: 10.1051/0004-6361:20066171
Hansen B, Murray N. (2012) Migration then assembly: Formation of Neptune-mass planets inside 1 AU. Astrophysical Journal 751, 158.
Hellier C, Anderson DR, Collier Cameron A, Doyle AP, Gillon M, Jehin E, and 12 others. (2012) Seven transiting hot-Jupiters from WASP-South, Euler and TRAPPIST: WASP-47b, WASP-55b, WASP-61b, WASP-62b, WASP-63b, WASP-66b & WASP-67b. Monthly Notices of the Royal Astronomical Society 426, 739-750. Abstract: 2012MNRAS.426..739H
Izidoro A, Raymond SN, Morbidelli A, Hersant F, Pierens A. (2015) Gas giant planets as dynamical barriers to inward-migrating super-Earths. Astrophysical Journal Letters 800, L22.
Lopez ED, Fortney JJ. (2014) Understanding the mass-radius relation for sub-Neptunes: Radius as a proxy for composition. Astrophysical Journal 792, 1. Abstract: 2014ApJ...792....1L
Mandell A, Raymond S, Sigurdsson S. (2007) Formation of Earth-like planets during and after giant planet migration. Astrophysical Journal 660, 823-844. Abstract.
Matsumura S, Ida S, Nagasawa M. (2013) Effects of dynamical evolution of giant planets on survival of terrestrial planets. Astrophysical Journal 767, 129. Abstract:
Mustill AJ, Davies MB, Johansen A. (2015) The destruction of inner planetary systems during high-eccentricity migration of gas giants. Astrophysical Journal 808, 14.
Naoz S, Farr WM, Rasio FA. (2012) On the formation of Hot Jupiters in stellar binaries. Astrophysical Journal Letters, 754:L36. doi:10.1088/2041-8205/754/2/L36
Nagasawa M, Ida S, Bessho T. (2008) Formation of hot planets by a combination of planet scattering, tidal circularization, and the Kozai mechanism. Astrophysical Journal 678, 498-508. Abstract.
Neveu-VanMalle M, Queloz D, Anderson DR, Brown D, Collier Cameron A, Delrez L, Díaz RF, Gillon M, Hellier C, Jehin E, Lister T, Pepe F, Rojo P, Ségransan D, Triaud A, Turner OD, Udry S. (2015) Hot Jupiters with relatives: discovery of additional planets in orbit around WASP-41 and WASP-47. Astronomy & Astrophysics, in press. Abstract: 2015arXiv150907750N
Pierens A, Raymond SN. (2011) Two phase, inward-then-outward migration of Jupiter and Saturn in the gaseous Solar Nebula. Astronomy & Astrophysics 533, A131. Abstract:
Sanchis-Ojeda R, Winn JN, Dai F, Howard AH, Isaacson H, Marcy GW, Petigura E, Sinukoff E, Weiss L, Simon Albrecht S, Hirano T, Rogers L. (2015) A low stellar obliquity for WASP-47, a compact multiplanet system with a Hot Jupiter and an ultra-short period planet. Astrophysical Journal Letters 812, L11. Abstract: 2015ApJ...812L..11S
Steffen JH, Hwang JA. (2015) The period ratio distribution of Kepler’s candidate multiplanet systems. Monthly Notices of the Royal Astronomical Society 448, 1956–1972. doi:10.1093/mnras/stv104
Triaud A, Collier Cameron A, Queloz D, et al. (2010) Spin-orbit angle measurements for six southern transiting planets: New insights into the dynamical origins of hot Jupiters. Astronomy & Astrophysics 524, A25.
Raymond SN, Mandell AM, Sigurdsson S. (2006) Exotic Earths: forming habitable planets with giant planet migration. Science 313, 1413-1416.
Wolfgang A, Lopez E. (2015) How rocky are they? The composition distribution of Kepler's sub-Neptune planet candidates within 0.15 AU. Astrophysical Journal 806, 183. Abstract:
Zeng L, Sasselov D. (2013) A detailed model grid for solid planets from 0.1 through 100 Earth masses. Publications of the Astronomical Society of the Pacific 125, 227-239. Abstract: 2013PASP..125..227Z