[Previous] | [Session 53] | [Next]
I. Mosqueira (Nasa Ames Research Center), P. R. Estrada (Cornell University), J. E. Chambers (Nasa Ames Research Center)
The characteristics of the regular satellite systems of the outer planets have led to the widespread belief that satellites resulted from similar physical processes as planets. For that reason, it has been a common practice to model the formation of satellites starting out with a disk of satellitesimals with sufficient total mass to account for the existing satellite systems (e. g. Richardson 2000). Unfortunately, as was noted early on (Coradini et. al. 1981), this scenario has great difficulty matching the observed masses and positions of the satellites of Jupiter and Saturn, even when those satellites which are more likely to have resulted from capture than from in situ formation are excluded. Furthermore, the overall sizes of the satellite systems might lead one to expect that the characteristics of the disks out of which they arose were directly tied to the satellitesimal feeding mechanism. Motivated by these arguments (and, to a lesser degree, by the long formation timescale implied by Callisto's partially differentiated state) we consider the coupled problem of disk and satellite formation. Our approach is to treat satellite accretion smultaneously with the feeding of satellitesimals into the planetary environment. In order to study this problem, we need to follow the trajectory of planetesimals initially in heliocentric orbits as they are captured by a giant planet. True capture requires close interaction with material bound in orbit around the planet (which, by itself, is insufficient to give rise to the observed satellite systems), or with gas in the planetary environment; either of these mechanisms can prevent the captured object from being ejected after a few circumplanetary orbits due to the conservation of the Jacobi constant. To make progress, we need to integrate the motion of planetesimals as they enter the planet's Hill radius and undergo close encounters with similarly sized objects. A new symplectic code has been developed specifically for this task, and some early results will be discussed.