2020AJ....159..260R

An intriguing question in the context of dynamics arises: could a moon possess a moon itself? Such a configuration does not exist in the solar system, although this may be possible in theory. Kollmeier & Raymond determined the critical size of a satellite necessary to host a long-lived subsatellite, or submoon. However, the orbital constraints for these submoons to exist are still undetermined. Domingos et al. indicated that moons are stable out to a fraction of the host planet's Hill radius R_H,p, which in turn depend on the eccentricity of its host's orbit. Motivated by this, we simulate systems of exomoons and submoons for 10⁵ planetary orbits, while considering many initial orbital phases to obtain the critical semimajor axis in terms of R_H,p or the host satellite's Hill radius R_H,sat, respectively. We find that, assuming circular coplanar orbits, the stability limit for an exomoon is 0.40 R_H,p and for a submoon is 0.33 R_H,sat. Additionally, we discuss the observational feasibility of detecting these subsatellites through photometric, radial velocity, or direct imaging observations using the Neptune-sized exomoon candidate Kepler 1625b-I and identify how stability can shape the identification of future candidates.