N-body simulations of planet formation via pebble accretion. II. How to form various giant planets

Aims. The connection between initial disc conditions and final orbital and physical properties of planets is not well-understood. In this paper, we numerically study the formation of planetary systems via pebble accretion and investigate the effects of disc properties such as masses, dissipation timescales, and metallicities on planet formation outcomes. Methods. We improved the N-body code SyMBA that was modified by taking account of new planet-disc interaction models and type II migration. We adopted the 'two-alpha disc' model to mimic the effects of both the standard disc turbulence and the mass accretion driven by the magnetic disc wind. Results. We successfully reproduced the overall distribution trends of semi-major axes, eccentricities, and planetary masses of extrasolar giant planets. We find that, when planet formation happens fast enough, giant planets are fully grown (Jupiter mass or higher) and are distributed widely across the disc. On the other hand, when planet formation is limited by the disc's dissipation, discs generally form low-mass cold Jupiters (CJs). Our simulations also naturally explain why hot Jupiters (HJs) tend to be alone and how the observed eccentricity-metallicity trends arise. The low-metallicity discs tend to form nearly circular and coplanar HJs in situ, because planet formation is slower than high-metallicity discs, and thus protoplanetary cores migrate significantly before gas accretion. The high-metallicity discs, on the other hand, generate HJs in situ or via tidal circularisation of eccentric orbits. Both pathways usually involve dynamical instabilities, and thus HJs tend to have broader eccentricity and inclination distributions. When giant planets with very wide orbits ('super-cold Jupiters') are formed, we find that they often belong to metal-rich stars, have eccentric orbits, and tend to have (~80%) companions interior to their orbits.