Characterizing and predicting dynamical instabilities in multiplanet systems

We investigate the dynamical stability of simulated planetary systems with architectures similar to observed high-multiplicity Kepler and K2 exo-planetary systems. Preliminary analysis of a large suite of simulations has revealed that ~20% of such simulated systems are prone to instabilities on timescales shorter than their current estimated ages. A case study of the Kepler-102 system architecture indicates that inward transfer of angular momentum deficit (AMD) due to secular chaos, rather than proximity to mean motion resonances, is critical to the onset of instability on billion orbit timescales. We find that a “spectral fraction” calculated from the power spectrum of a planetary system’s AMD time series from integrations over million-orbit timescales is predictive of stability/instability on billion-orbit timescales; simulated systems with noisy AMD power spectra are less likely to be stable than systems whose AMD power spectrum is dominated by just a few secular frequencies (Volk & Malhotra 2020, AJ 160(3), id.98). We will present an expanded detailed analysis of our large suite of simulations to test the conjecture that AMD transfer is generically the dominant cause of instabilities and to further elaborate on the usefulness of the spectral fraction of a short simulation to predict longer term stability/instability.

We gratefully acknowledge research funding from NASA (grant 80NSSC18K0397).