Formation of our solar system's giant planets via pebble accretion

Abstract

The classical core accretion model of gas giant formation requires the formation of a solid core of around 10 $M_\oplus$ to trigger runaway gas accretion, but the growth timescale by pairwise accretion of planetesimals to reach the critical core mass beyond 5 AU is likely too long compared to the typical disk lifetime of a few Myr. Meanwhile, large populations of mm- to cm-sized grains, or pebbles, have been detected in protoplanetary disks, which lays the ground for the notion of pebble accretion. Efforts to incorporate pebble accretion into global simulations of planet formation showed promising results, although planet migration poses challenges in forming Solar System analogue. The present study aims at assembling the giant planets of our Solar System based on the work by Matsumura et al. (2017), which employed a modified version of the $N$-body code SyMBA. SyMBAp (SyMBA parallelized) is built in the course of this work and achieved good scalability to speed up the computation on multi-core system. The planet migration prescription by Ida et al. (2018) is adopted for slower inward migration. Simulations of planetesimal disks with $N>10^3$ are conducted to investigate the effect of viscous stirring. With slower Type-II migration, a few gas giants can indeed form and remain around 5 AU in parts of the parameter space tested. However, these results are generally accompanied by multiple bodies of Earth-mass or lager close to the Sun due to rapid migration. Other attempts to refine the models are also tested in this work including reducing the pebble mass flux, adopting a much thinner pebble disk and using a different disk decay model. These attempts show improvements by producing fewer bodies of Earth-mass or lager while forming a few giant planets beyond 3 AU, but are still not considered as realistic compared to our Solar System.