Planetesimal Formation and Demographics
Planet formation begins when dust grains in protoplanetary disks assemble into planetesimals a kilometer or more across. I have led a series of simulations of how planetesimals form and what sets their initial masses, sizes, and binarity. These predictions have found strong support in Kuiper Belt objects of the Solar System, advancing our understanding of the earliest stages of planet formation.
We propose an origin story for the cold classical Kuiper Belt, in which it forms in situ at the tail end of the solar nebula's life, when only 2–5% of its original gas remained — reproducing its total mass, characteristic ~100 km sizes, and ratio of prograde to retrograde binaries. A companion 1D global model then shows how an inside-out dispersing gas disk leaves behind a planetesimal disk as its cavity wall recedes outward; the disk ends naturally at the "Kuiper Cliff" near 47 au, with no fine-tuning.
Li, Youdin, & Simon 2019; Li & Youdin 2021
The highest-resolution streaming-instability simulations with self-gravity to date, paired with a new clump-finding tool, show that the initial mass distribution of planetesimals steepens above the masses of ~100 km bodies — mirroring the observed Kuiper Belt size distribution — with evidence of a turnover at the low-mass end. A follow-up survey mapped the metallicity threshold for particle clumping across dust sizes, with fits generalized to other disk conditions, finding that clumping can occur at metallicities as low as 0.4% for optimally sized solids and widening the parameter space in which the streaming instability can form planetesimals.
Nesvorný, Li et al. 2019, 2021
Streaming-instability simulations predict that about 80% of newborn planetesimal binaries should have prograde orbits — matching the observed inclination distribution of Kuiper Belt binaries and providing evidence for the streaming instability. Follow-up N-body simulations of gravitationally collapsing pebble clouds showed that a cloud's angular momentum largely determines the outcome: low-spin clouds produce single planetesimals and tight or contact binaries, while compact high-spin clouds produce widely separated, equal-mass binaries resembling those in the Kuiper Belt.
We develop an analytic theory for the planetesimal initial mass function, based on a Toomre-like instability in the particle midplane limited by diffusion, tides, and shear. The prediction agrees with high-resolution numerical simulations and is consistent with both the "born big" paradigm and the power-law mass functions commonly found in simulations.