Long-term self-modification of irregular satellite groups

More than 50 irregular satellites revolve around Jupiter and more than 30 around Saturn. There, at least three collisional families were identified. Among these, the Himalia family of prograde irregular moons at Jupiter is characterised by a velocity dispersion of several hundred m/s, inconsistent with a collisional origin (Nesvorný et al., 2003). Here we investigate whether the dispersion could stem from the mutual gravitational interaction among the family members, especially from perturbations by the largest member Himalia. Using long-term N-body simulations, we find that over 1 Gyr, Himalia can disperse its family significantly, particularly in the semimajor axis and eccentricity. By extrapolating our results to 4 Gyr, we show that it is unlikely Himalia's gravity alone is responsible for the observed dispersion. The self-dispersion scenario becomes viable if Himalia is twice as massive as the upper end of current estimates (Brozović and Jacobson, 2017; Emelyanov, 2005). We also find that the collisions with Himalia would have removed  ≳ 60% of an initial population of 10km class family satellites over the last 4 Gyr. During this period, the observed satellites have probably been captured into secular resonances with Himalia (Li and Christou, 2017). These resonances can affect the dispersion through resonance captures/escapes and by restricting close encounter configurations. A similar hypothesis is tested for the putative Phoebe family at Saturn, also of large velocity dispersion (Gladman et al., 2001). Again, we find this effect not sufficient to account for the observed orbital distribution. In all simulations, it is found that the rate of dispersion is decreasing with time, owing to the declining frequency of resonance captures/escapes and close encounters.