Artist's rendering of a hot Jupiter exoplanet around a Solar-type star. Image credit: ESO

Aug 14, 2015 Two New Planets Found in Hot Jupiter System

Sen—Though we now know that super-Earths are the most common exoplanets, early on it was thought that hot Jupiters were most common because they are the easiest planets to find. The first planet discovered around a Sun-like star, 51 Pegasi, is a hot Jupiter, as were the next handful discovered. Because hot Jupiters orbit so close to their stars, they are almost guaranteed to transit deeply. Since the planets are massive, they will also likely induce a large radial velocity signal on their host stars. With both transit and radial velocity measurements possible for most hot Jupiters, we can get both accurate sizes and masses which tell us about the structure and possible atmospheres. 

The hot Jupiter planet, WASP-47 b, was initially discovered in 2012 as part of the Wide Angle Search for Planets (WASP). It orbits a star slightly cooler than the Sun around 650 light-years away. The system was targeted again by Juliette Becker of the University of Michican from November, 2014 to January, 2015 when the Kepler Space Telescope in its new K2 operating mode shifted its field of view towards WASP-47's area of the sky. Analysis of the K2 light curves showed the transits of the WASP-47 b and two other, previously unknown planets with orbits very close to that of WASP-47 b. The WASP-47 system is the first planetary system to have companion planets close-in to a hot Jupiter.

Hot Jupiters, as the name suggests, are gas giant planets in very tight orbits around their stars, usually with orbital periods less than 10 days. The existence of hot Jupiters came as a big surprise, since theories of planet formation show that gas giant planets cannot form in the inner parts of a solar system. That close to the star the majority of the gaseous material should either have accreted onto the star itself at the beginning of the system's life or subsequently be blown outwards by young stellar winds. Only out past the "ice line," where temperatures are cold enough that some ices can form, is the gas density high enough to form Jupiter size planets. 

How exactly hot Jupiters get so close to their stars is still a bit of a mystery. The most widely accepted theory is that a hot Jupiter forms out past the "ice line," like our own Jupiter, and migrates inwards to its current position. The migration of such a large planet would destabilize the orbits of any smaller planets it crossed, essentially sweeping the inner solar system clean of any other planets. In this scenario, it was expected that hot Jupiters would be "lonely" planets in their solar systems. There have only been a few exceptions to the rule of lonely hot Jupiters, and most of them can be explained without radical adjustment of the accepted theory. 

WASP-47 is the only planetary system to date that has two planets in orbits close to a hot Jupiter, with orbital periods of about one day, four days, and nine days. The hot Jupiter actually orbits between the inner super-Earth (WASP-47 c) and the outer Neptune-sized planet (WASP-47 d). This is a unique configuration for a hot Jupiter; to date, hot Jupiters have only been observed as either the innermost or ourtermost planet of the system, with other planets too far away to interact with it.

WASP-47 and its planets were observed with the Kepler Space Telescope operating in its new K2 mode. With only two working reaction wheels to hold the telescope steady while observing, the telescope now uses Solar radiation pressure as its third point of stability. This allows Kepler to observe targets with nearly the same precision as it did in its original mission. K2 now observes targets along the Solar System ecliptic plane, sweeping a band around the sky in a series of campaigns three months long.

While the transit light curves from K2 can measure the orbital periods and transit depths of the planets very precisely, current fitting techniques can only put loose constraints on other orbital parameters. In such a tightly packed system the orbital eccentricity of the planets, how circular or elliptical the orbits are, has a drastic impact on the overall lifetime of the planetary system. At high eccentricity the planets interact too much, which destabilizes the orbits. Only a small range of orbital eccentricities will allow such a system to remain stable over long periods, so Becker and co-authors used this fact to put further constraints on the orbital parameters of each planet.

They first assume, as is typical in astronomy, that they are not observing WASP-47 at a particularly special time in its history. So, whatever the true orbital parameters are, they must keep the WASP-47 system stable for a long time. They then perform computer simulations of the planetary system evolving over 10 million years, taking into account the gravitational interactions of the planets with their star and with each other. These simulations are performed 1,000 times, each with a different set of planet masses and orbital eccentricities. A simulated planetary system is "stable" if all three planets remain in the same orbits after 10 million years of simulation time. After that, it is unlikely that the systems will evolve further.

The stability tests indicate that the WASP-47 planets must all have near-circular orbits in order to be stable over a long time. Eccentricities greater than about five per cent destabilize the orbits after only 1,000 years. The stability of the low eccentricity systems, however, does not appear to depend on the masses of the planets. So, while the simulations can help further constrain the orbits of the planets, they cannot provide any more constraint on the planet masses.

There is another way to get masses for the WASP-47 planets. Since the three planets orbit so close to each other, they tug on each other gravitationally and slightly alter the exact times each planet transits. The strength of these transit timing variations (TTVs) tells us how much each planet gravitationally interacts with the others, which can reveal the masses of the planets. 

TTVs work like this: In a system with a single transiting planet A, each transit we observe would happen at perfectly regular intervals (the orbital period). A second planet in the system, B, can alter the timing of the transits slightly. If B is at a position behind A when A is just about to transit, gravitational pull between A and B decelerates A's orbit. This makes the transit of A happen a few minutes later than it would have without B. Conversely, B will transit a few minutes earlier than it would have because of the forward acceleration from A. Planets A and B will transit earlier and later in an alternating fashion, and when one slows down the other speeds up. This effect can be seen even if one planet is not transiting, and can reveal other unseen planets in a system.

This animation shows the transit timing variations of the inner planet of a two planet system. The gravitational interaction between the inner and outer planet causes the orbital period to change in an oscillating manner compared to the orbital period of the inner planet alone. Credit: NASA Ames/Kepler mission

In the case of WASP-47, the TTVs induced on the hot Jupiter and Neptune-sized planets were strong enough to obtain accurate masses for these planets. The TTVs of the super-Earth were too small to get anything more than an upper limit on the mass. WASP-47 b, the original hot Jupiter, is both slightly larger and slightly more massive than our Jupiter. As the original discovery stated, WASP-47 b "is an entirely typical hot Jupiter." Becker and co-authors find that the outer planet, WASP-47 d, is only slightly smaller than Neptune, but is only about half as massive. Therefore, it must be much less dense than Neptune and perhaps has a small, dense core with an extended "puffy" atmosphere. All they can conclude about the super-Earth WASP-47 c is that it is almost twice the size of Earth, and something less than nine times as massive.

WASP-47 is an anomaly among hot Jupiter systems because it could not have formed via the traditional migration method. WASP-47 b must have migrated "quietly," in a way that either did not disturb the other two planets or brought them along when it migrated. Either way, astronomers now need to go back and find a logical way for this system to exist.

The WASP-47 system would be perfect for follow-up radial velocity measurements of the planetary orbits. Radial velocity measurements would help further constrain the mass of the super-Earth, and also verify the masses of the other two planets. If radial velocity measurements take place in the future, WASP-47 would be the tenth planetary system with radial velocity, transit, and transit timing variation measurements, making it a Rosetta Stone between three major planet-hunting techniques.