Title: Constraints on the Architecture of the HD 95086 Planetary System with the Gemini Planet Imager

Authors: Julien Rameau, Eric L. Nielsen, Robert J. De Rosa et al.

Lead Author’s Insititution: Université de Montréal

Status: Accepted for publication in The Astrophysical Journal Letters open access

HD 95086 is one of the more well studied and characterized systems; it hosts planetary, planetesimal, and dust components, which make it quite the intriguing subject to study. Its planet HD 95086 b was directly imaged by GPI in late 2013; the planet is about 4 times the mass of Jupiter and orbits its star at a distance of about 56 AU. The disk in the system is characterized by three components— a 55 K cool component, a 75 K warm component at – each of those two corresponding to a planetesimal belt– and a possible hot component at 300 K, which could suggest activity in the habitable zone of the star (Su et al. 2015). There is also quite a large gap in HD 95086’s disk, extending from about 8 to 80 AU. This system presents a perfect playground for the study of evolution and formation of unusual systems.

Figure 1: The direct images of HD 95086, taken over several epochs, are shown in this figure.

This paper focuses on constraining the orbit of HD 95086 b based on re-analyzation of old images in combination with their newer images. HD 95086 b was originally imaged in late 2013 (Galicher et al. 2014).  Overall, their images spanned four epochs, from December 2013 to March 2016, and provided astromteric measurements for HD 95086 b. From these measurements, the authors found that the orbit of HD 95086 b is face-on and circular. Examples of the direct imaging on the system are shown in Figure  1.

To constraint the orbit of HD 95086 b, the authors used Monte Carlo (MC) techniques that were more efficient than a traditional Markov Chain Monte Carlo. The technique generated parameters from a probability density function, then fit the parameters of the orbits through each epoch. The probability of the orbits generated was then evaluated by comparing the remaining epochs against a uniform random variable, and the orbits were then accepted or rejected by the program. In Figure 2, the generated orbits are shown with the data points. Using the constraints for planet b in the system, the researchers constrained the HD 95086 system as a whole.

Figure 2: The model fitting for HD 95086, overlaid with the measurements, which are color coded by date taken, is shown in this image. The gray areas indicate the approximate location of the planetesimal belts in the system.

In an earlier publication, (Su et al. 2015), several possibilities for the system architecture were presented, and in this paper, these possibilities are explored with the new analysis of data. The authors rule out several scenarios and present the most likely scenarios from their models. The scenarios are as follows:

• Scenario A: HD 95086 b is the only planet in the system and has carved out the large gap in the disk through an eccentric orbit of about 0.7. Verdict: Ruled out with 95% confidence. Astrometric measurements from this paper do not indicate such a high eccentricity

• Scenario B: HD 95086 b has a slight eccentricity of about 0.3, and another more massive, more eccentric planet resides at 16 AU. Verdict: Neither ruled out nor confirmed. The 16 AU planet, if it were there, is undetectable by GPI, and this scenario is not constrained by the observations so a verdict for this one can’t quite be reached.

• Scenario C: Two other planets, slightly larger than HD 95086 b at about 7 Jupiter masses, orbit at 12 and 26 AU. All three planets in this scenario have low (0.3 or less) eccentricities. Verdict: Needs reconfiguration. With a low eccentricity at 26 AU, the planet would have been detected in the observations, but the inner planet at 12 AU would not have been. Its possible that the inner planet is more massive, so parts of this scenario could work with the observations. However, this scenario with two additional planets is unlikely.

• Scenario D: In addition to HD 95086, three large Jupiter-like planets, of all the same mass, orbit at 11, 19, and 34 AU, respectively. Verdict: Ruled out. The third planet, projected at 34 AU, would have been visible in observations if it were there, and this scenario would necessitate all three extra planets to account for the disk configuration.

Certainly, these aren’t the only possibilities, but with the new analysis and data, the scenarios have been considerably confined. It is likely that HD 95086 has two massive planets that have sculpted out the gap in the disk, with one at a closer separation than is currently detectable by direct imaging. Another possibility is that three or four planets are present in the system, but their eccentricities and mass vary greatly, which could explain why they have not been detected by current observations.

In HD 95086, planets have clearly disrupted the disk and the planetesimals, but the exact nature of this contact remains unknown, though now these researchers have found that a two-planet system of moderate eccentricity or possibly an inhomogeneous mix of three or fours planets. The interactions between the planets and the disk can reveal so much about how these systems form and evolve. There’s still some mystery left in the HD 95086 system, but the scientific sleuths of this papers have thoroughly narrowed down the possibilities.

• About the Author

About Mara Zimmerman

Mara is working on her PhD in Astronomy at the University of Wyoming. She has done research with Heartbeat binary stars, and currently works on modeling the composition of debris disks. When not at her computer writing code, she likes hiking, baking scones, and making puns.