Dynamical History of Planetary Systems
Recent observations suggest that most close-in sub-Neptunes are born in mean-motion resonant chains, the majority of which break apart on timescales of ~100 Myr. Using N-body simulations, I showed that the collisions following the breakup of these chains reproduce the period-ratio distribution of mature systems, including the fine structure near resonances and the phases of transit timing variations. With collaborators, I have also explored what breaks the chains in the first place: small free eccentricities, excited for example by the accretion of a handful of Mercury-sized bodies, can trigger instability on the observed ~100 Myr timescale.
A census of known planetary systems, divided by age, finds that the fraction of neighboring planet pairs within a few percent of a first-order commensurability falls from 70% in young (<100 Myr) systems to 15% in mature (>1 Gyr) ones. This strongly supports "breaking-the-chains" models, in which close-in planets are born in resonant chains that later break apart.
N-body integrations of destabilized resonant chains show that collisions erode but do not eliminate resonant pairs; survivors appear as narrow peaks just wide of commensurability in the period-ratio histogram. Merger products fill the space between the 5:4, 4:3, and 3:2 resonances but not the wider gap between the 3:2 and 2:1, reproducing the observed trough just short of 2:1. Debris from these major mergers drives minor mergers that often widen pairs further, and the overall dynamical activity excites free eccentricities and hence the phases of transit timing variations.
Choksi, Lithwick, Chiang & Li 2026
A two-stage disruption scenario for resonant chains: some mechanism first excites free eccentricities of a few percent — for example, accretion of a handful of Mercury-sized bodies — which then triggers dynamical instability on the observed ~100 Myr timescale, reproducing the observed decline in the incidence of resonance. Impacts from such bodies can also explain why some young systems have period ratios narrow of commensurability. The paper also identifies two new observational trends: multiplicity declines on the same ~100 Myr timescale as resonance, and resonance occupation rises with multiplicity.