Terrestrial planets are thought to be the result of a vast number of gravitational interactions and collisions between smaller bodies. We use numerical simulations to show that practically identical initial conditions result in a wide array of final planetary configurations. This highly chaotic behaviour questions the predictability of different scenarios for the formation and evolution of our solar system and planetary systems in general. However, multiple realisations of the same initial conditions can be used to predict certain global statistics. We present two sets of numerical experiments that quantify this behaviour. Firstly, we demonstrate that simulations with slightly displaced particles are completely divergent after ~500 years, irrespective of initial displacement, particle number, and code accuracy. If a single planetesimal is moved by less than one millimetre, then a different set of planets results — this timescale for chaotic divergence decreases with increasing particle number. Secondly, we show final planetary configurations of initially similar simulations with and without giant planets after evolving them for ~148 Myr. We find that the same simulations including giant planets tend to generate higher mass planets at lower semi-major axes than simulations without gas giants. This prediction can be tested with forthcoming observational programs. By extracting outliers in the observations, we cautiously predict that Kepler-10, Kepler-9, 61 Vir, HD 134060, and HD 51608 may host as yet undetected giant planets.
V. Hoffmann, S. Grimm, B. Moore, et. al.
Thu, 6 Aug 15
Comments: 17 pages, 15 figures, submitted to MNRAS, simulation outputs available at this https URL