Pebble isolation mass variants in N-body simulations fail to reproduce full Kepler super-Earth distributions simultaneously

The study fixed gas-disk parameters and varied only the embryo isolation mass prescription during the pebble-accretion phase, then evolved systems through the giant-impact stage before applying a mock transit survey. Each isolation model reproduced at least one Kepler observable (e.g., migration-feedback matched period ratios), yet none matched all three distributions at once even after reweighting. This mismatch implies that either the isolation-mass treatment or other fixed assumptions—such as disk lifetime or planetesimal accretion efficiency—require revision.

Kepler’s observed dichotomy between compact, low-eccentricity multis and more excited systems has long been attributed to divergent formation pathways. The simulations demonstrate that isolation mass alone cannot generate the full range, pointing instead to stochastic giant impacts or late-stage dynamical instabilities as necessary additional drivers. The result aligns with earlier work by Izidoro et al. (2021) showing that pebble isolation sets embryo mass but does not dictate final architecture.

A key limitation is the absence of explicit gas-disk turbulence or photoevaporation; both alter migration timescales and could relax the tension between models and data. Future runs that couple variable isolation mass with self-consistent disk dispersal will test whether the current shortfall persists or is an artifact of the fixed-disk assumption.

The paper underscores that super-Earth system statistics encode information about the earliest growth phase, tightening constraints on habitable-zone occurrence rates once isolation-mass effects are properly calibrated.