Monday, May 4, 2015

Detecting Binary Planets by Transit Observations

In a planetary system hosting three or more gas giant planets, orbital instabilities can disrupt the planets’ orbits and cause two planets to encounter one another. Planet-planet tidal interactions during the encounter can result in the formation of a gravitationally bound pair of gas giant planets, referred to hereafter as binary planets. It is typical for binary planets to be tightly bound, with separations of only 3 to 5 times the sum of their physical radii. A binary planet around a Sun-like star can remain stable over the lifetime of its host star provided it is orbiting further than ~0.3 AU.

Figure 1: Artist’s impression of a gravitationally bound pair of gas giant planets.

Transit photometry appears to be a promising method for detecting binary planets. When a binary planet transits in front of its host star, the transit has a longer duration and a deeper transit depth than in the case of a single planet. Furthermore, the shape of the light curve from the transit of a binary planet is also different. In fact, some of the false positive detections by NASA’s Kepler space telescope and by the French CoRoT space observatory could turn out to be binary planets.

Typical fates of planetary systems containing three or more interacting gas giant planets are - the ejection of planets, planet-star collisions and planet-planet collisions. If planet-star tidal interactions are taken into account, most of the “planet-star collisions” result in the formation of hot Jupiters (i.e. gas giant planets that orbit very close to their host stars). Correspondingly, if planet-planet tidal interactions are taken into account, some of the “planet-planet collisions” result in the formation of binary planets.

Figure 2: Artist’s impression of a gravitationally bound pair of gas giant planets.

Assuming ~10 percent of all planetary systems host multiple gas giant planets in eccentric orbits, and ~10 percent of these planetary systems eventually form binary planets, then ~1 percent of all planetary systems should host binary planets. Additionally, if binary planets are on average ~0.5 AU from their host stars, then there is a ~1 percent probability that they will transit. As a result, one in ~10,000 stars should host transiting binary planets. A key characteristic of the transit light curve of a binary planet is that it varies from transit-to-transit depending on the positions of the two planets during the transit event. 4 cases have been identified.

Case A: During the transit event, one planet mutually eclipses the other. This causes the light curve to feature a “bump” due to the reduction in the total amount of starlight both planets block (Figures 3 and 4). Such a “bump” in the light curve cannot happen when a single planet transits its host star, unless the planet transits in front of a starspot that happens to be present on the surface of its host star.

Case B: The binary planet transits without any one planet mutually eclipsing the other. As a result, the transit light curve shows a deep dip near mid-transit (transit by both planets) and the deep dip is flanked on both sides by relatively shallower transit depths (transit by one planet) (Figure 4).

Case C: The binary planet transits its host star and one planet either ingresses into transit first or egresses out of transit first. This causes the light curve to feature a “step”. The transit depth steps from shallow to deep when one planet ingresses into transit first, or steps from deep to shallow when one planet egresses out of transit first (Figure 4).

Case D: If the orbital separation of the binary planet is larger, one planet can enter the transit around the same time the other planet leaves the transit. This can create a pair of side-by-side transits or two overlapping transits that overlap around a “bump” in the middle of the transit light curve (Figure 5).

Figure 3: Transit light curve of a binary planet illustrating “Case A”, where one planet mutually eclipses the other during the transit event. The 4 transit light curves shown here involve - a binary planet with each planet having the same mass as Jupiter and twice the diameter of Jupiter (solid purple line); a single planet with twice the mass and 2.8 times the diameter of Jupiter (dotted red line); a binary planet with each planet having the same mass and diameter as Jupiter (dashed blue line); and a single planet with twice the mass and 1.4 times the diameter of Jupiter (dash-dotted light-blue line). Note that a planet with 2.8 times the diameter of Jupiter is physically quite unlikely since the most inflated planet known does not exceed 2 times the diameter of Jupiter. K. Lewis, H. Ochiai, M. Nagasawa and S. Ida (2015).

Figure 4: Transit light curves of a binary planet illustrating “Case A”; “Case B”, where the binary planet transits without any one planet mutually eclipsing the other; and “Case C”, where the binary planet transits and one planet either ingresses into transit first or egresses out of transit first. The 4 curves in each panel are the same as the ones described in Figure 3. K. Lewis, H. Ochiai, M. Nagasawa and S. Ida (2015).

Figure 5: Transit light curve of a binary planet illustrating “Case D”, where one planet enters the transit around the same time the other planet leaves the transit. Here, the orbital separation of the binary planet is larger than in Figures 2 and 3. The 2 transit light curves shown here involve a binary planet with each planet having the same mass as Jupiter and twice the diameter of Jupiter (solid purple line); and a single planet with twice the mass and 2.8 times the diameter of Jupiter (dotted red line). K. Lewis, H. Ochiai, M. Nagasawa and S. Ida (2015).

The variations in shapes of the transit light curve from transit-to-transit may be the most noticeable signal of planet binarity. Since the transits of binary planets are generally longer in duration and deeper in depth than for single planets, some of the known hyper-inflated gas giant planets or light curves classified as false positive detections, may turn out to be binary planets. It is quite likely that binary planets are present in the large dataset of light curves collected by Kepler and CoRoT.

Reference:
K. Lewis, H. Ochiai, M. Nagasawa and S. Ida (2015), “Extrasolar Binary Planets II: Detectability by Transit Observations”, arXiv:1504.06365 [astro-ph.EP]