Title:Occurrence Rates of Exosatellites Orbiting 3-30M$_{\rm Jup}$ Hosts from 44 Spitzer Light Curves

Abstract:We conduct a comprehensive search for transiting exomoons and exosatellites within 44 archival Spitzer light curves of 32 substellar worlds with estimated masses ranging between 3-30M$_{\rm Jup}$. This sample's median host mass is 16M$_{\rm Jup}$, inclusive of 14 planetary-mass objects, among which one is a wide-orbit exoplanet. We search the light curves for exosatellite signatures and implement a transit injection-recovery test, illustrating our survey's capability to detect $>$0.7R$_{\oplus}$ exosatellites. Our findings reveal no substantial ($>$5$\sigma$) evidence for individual transit events. However, an unusual fraction of light curves favor the transit model at the 2-3$\sigma$ significance level, with fitted transit depths consistent with terrestrial-sized (0.7-1.6R$_{\oplus}$) bodies. Comparatively, fewer than 2.2% of randomly generated normal distributions from an equivalent sample size exhibit a similar prevalence of outliers. Should one or two of these outliers represent a real exosatellite transit, it would imply an occurrence rate of $\eta = 0.61^{+0.49}_{-0.34}$ short-period terrestrial exosatellites per system, consistent with the known occurrences rates for both solar system moons and mid M-dwarf exoplanets. We explore alternative astrophysical interpretations for these outliers, underscoring that transits are not the only plausible explanation. For orbital periods $<$0.8 days, the typical duration of the light curves, we constrain the occurrence rate of sub-Neptunes to $\eta<$0.35 (95% confidence) and, if none of the detected outlier signals are real, the occurrence rate of terrestrial ($\sim$Earth-sized) exosatellites to $\eta<$0.51 (95% confidence). Forthcoming JWST observations of substellar light curves will enable detection of sub-Io-sized exosatellites, allowing for much stronger constraints on this exosatellite population.