We are fairly certain the giant planets (Jupiter, Saturn, Uranus and Neptune) formed much quicker than the terrestrial planets (Mercury, Venus, Earth and Mars), and were born in a more compact configuration, closer to one another, than they exist in today. Sometime after their formation, we think that the entire solar system experienced a global orbital instability; wherein the orbits of the planets changed drastically and rapidly. Now referred to as the Nice Model (as in the French city), this evolutionary framework for the solar system was developed in the early 2000's with key contributions from dynamicists including Alessandro Morbidelli, Hal Levison, Rodney Gomes and Kleomenis Tsiganis. Through this process of instability (see example video), the solar system likely lost one or more planets, and the outer solar system was rapidly reshaped from its primordially compact architecture into the diffuse collection of orbits that we observe today. As much of the record of the solar system’s birth was erased in this instability, we must use large numbers of computer simulations (>6,000 in our recent paper: Clement et al., (2021a) to decode what happened during the instability, how it perturbed different objects like asteroids and the tiny terrestrial world, and what the conditions were when the planets were born. This is a bit like trying to understand how to make a smoothie only by tasting the final product. Much of the information regarding the ingredients the went in to the concoction is lost during the mixing process, and you have to try and reverse engineer what the recipe was.
One problem that has plagued instability models for a while now is trying to understand how Jupiter’s orbit got so eccentric and elliptical (non-circular). Historically, simulations that are able to reproduce Jupiter’s orbital shape tend to push Saturn too far out in to the outer solar system, beyond where Uranus is today. We used initial conditions that are consistent with hydrodynamical models of the giant planets forming in gaseous proto-planetary disks to more consistently generate Jupiter and Saturn-like orbits. Specifically our work finds that Jupiter and Saturn likely formed in a 2:1 mean motion resonance (where Jupiter completes exactly two orbits for every one of Saturn's, see examples below), rather than the 3:2 resonance commonly assumed in past studies. Moreover, in Clement et al., (2021b) we found that these initial conditions are largely consistent with disk model predictions that the giant planets' primordial eccentricities might be extremely high (~0.1-0.3) if Jupiter and Saturn are entrapped in the 2:1 resonance. In successful simulations, the eccentricities damp considerably prior to the instability, thus replicating the values tested in Clement et al., (2021a).