Doppler and transit observations of exoplanets show a pile-up of Jupiter-size planets in orbits with a 3 day period. A fraction of these hot Jupiters have retrograde orbits with respect to the parent star's rotation, as evidenced by the measurements of the Rossiter-McLaughlin effect. To explain these observations we performed a series of numerical integrations of planet scattering followed by the tidal circularization and migration of planets that evolved into highly eccentric orbits. We considered planetary systems having three and four planets initially placed in successive mean-motion resonances, although the angles were taken randomly to ensure orbital instability in short timescales. The simulations included the tidal and relativistic effects, and precession due to stellar oblateness. Our results show the formation of two distinct populations of hot Jupiters. The inner population (Population I) is characterized by semimajor axis a < 0.03 AU and mainly formed in the systems where no planetary ejections occurred. Our follow-up integrations showed that this population was transient, with most planets falling inside the Roche radius of the star in <1 Gyr. The outer population of hot Jupiters (Population II) formed in systems where at least one planet was ejected into interstellar space. This population survives the effects of tides over >1 Gyr and fits nicely the observed 3 day pile-up. A comparison between our three-planet and four-planet runs shows that the formation of hot Jupiters is more likely in systems with more initial planets. Due to the large-scale chaoticity that dominates the evolution, high eccentricities and/or high inclinations are generated mainly by close encounters between the planets and not by secular perturbations (Kozai or otherwise). The relative proportion of retrograde planets seems of be dependent on the stellar age. Both the distribution of almost aligned systems and the simulated 3 day pile-up also fit observations better in our four-planet simulations. This may suggest that the planetary systems with observed hot Jupiters were originally rich in the number of planets, some of which were ejected. In a broad perspective, our work therefore hints on an unexpected link between the hot Jupiters and recently discovered free floating planets.

In this paper, we present a new set of corotational solutions for the 2/1 commensurability, including previously known solutions and new results. Comparisons with observed exoplanets show that current orbital fits of three proposed resonant planetary systems are consistent with apsidal corotations. We also discuss the possible relationship between the current orbital elements fits of known exoplanets in the 2/1 mean‐motion resonance and the expected orbital configuration due to migration. We find that, as long as the orbital decay was sufficiently slow to be approximated by an adiabatic process, all captured planets should be in apsidal corotations. In other words, they should show a simultaneous libration of both the resonant angle and the difference in longitudes of pericenter.

We analyze the distribution of extrasolar planets (both confirmed and Kepler candidates) according to their orbital periods P and planetary radii R.Among confirmed planets, we find compelling evidence for a paucity of bodies with 3 < R < 10R ⊕ , where R ⊕ in the Earth's radius, and P < 2-3 days. We have christened this region a sub-Jovian Pampas. The same trend is detected in multiplanet Kepler candidates. Although approximately 16 Kepler single-planet candidates inhabit this Pampas, at least 7 are probable false positives (FP). This last number could be significantly higher if the ratio of FP is higher than 10%, as suggested by recent studies.In a second part of the paper we analyze the distribution of planets in the (P, R) plane according to stellar metallicities. We find two interesting trends: (i) a lack of small planets (R < 4R ⊕ ) with orbital periods P < 5 days in metal-poor stars, and (ii) a paucity of sub-Jovian planets (4R ⊕ < R < 8R ⊕ ) with P < 100 days, also around metal-poor stars. Although all these trends are preliminary, they appear statistically significant and deserve further scrutiny. If confirmed, they could represent important constraints on theories of planetary formation and dynamical evolution.