SEARCH

SEARCH BY CITATION

Keywords:

  • planets and satellites: dynamical evolution and stability;
  • planets and satellites: formation;
  • planet-disc interactions – planet-star interactions – planetary systems

ABSTRACT

It is well accepted that ‘hot Jupiters’ and other short-period planets did not form in situ, as the temperature in the protoplanetary disc at the radius at which they now orbit would have been too high for planet formation to have occurred. These planets, instead, form at larger radii and then move into the region in which they now orbit. The exact process that leads to the formation of these close-in planets is, however, unclear and it seems that there may be more than one mechanism that can produce these short-period systems. Dynamical interactions in multiple-planet systems can scatter planets into highly eccentric orbits which, if the pericentre is sufficiently close to the parent star, can be tidally circularized by tidal interactions between the planet and star. Furthermore, systems with distant planetary or stellar companions can undergo Kozai cycles which can result in a planet orbiting very close to its parent star. However, the most developed model for the origin of short period planets is one in which the planet exchanges angular momentum with the surrounding protoplanetary disc and spirals in towards the central star. In the case of ‘hot Jupiters’, the planet is expected to open a gap in the disc and migrate in, what is known as, the Type II regime. If this is the dominant mechanism for producing ‘hot Jupiters’ then we would expect the correct properties of observed close-in giant planets to be consistent with an initial population resulting from Type II migration followed by evolution due to tidal interactions with the central star. We consider initial distributions that are consistent with Type II migration and find that after tidal evolution, the final distributions can be consistent with that observed. Our results suggest that a modest initial pile-up at a ∼ 0.05 au is required and that the initial eccentricity distribution must peak at e ∼ 0. We also suggest that if higher mass close-in exoplanets preferentially have higher eccentricities than lower mass exoplanets, this difference is primordial and is not due to subsequent evolution.