Cassini UVIS solar occultations by Saturn's F ring and the detection of collision-produced micron-sized dust

We present an analysis of eleven solar occultations by Saturn's F ring observed by the Ultraviolet Imaging Spectrograph (UVIS) on the Cassini spacecraft. In four of the solar occultations we detect an unambiguous signal from diffracted sunlight that adds to the direct solar signal just before or after the occultations occur. The strongest detection was a 10% increase over the direct signal that was enabled by the accidental misalignment of the instrument's pointing. We compare the UVIS data with images of the F ring obtained by the Cassini Imaging Science Subsystem (ISS) and find that in each instance of an unambiguous diffraction signature in the UVIS data, the ISS data shows that there was a recent disturbance in that region of the F ring. Similarly, the ISS images show a quiescent region of the F ring for all solar occultations in which no diffraction signature was detected. We therefore conclude that collisions in the F ring produce a population of small ring particles that can produce a detectable diffraction signal immediately interior or exterior to the F ring. The clearest example of this connection comes from the strong detection of diffracted light in the 2007 solar occultation, when the portion of the F ring that occulted the Sun had suffered a large collisional event, likely with S/2004 S 6, several months prior. This collision was observed in a series of ISS images (Murray et al., 2008).

Our spectral analysis of the data shows no significant spectral features in the F ring, indicating that the particles must be at least 0.2 μm in radius. We apply a forward model of the solar occultations, accounting for the effects of diffracted light and the attenuated direct solar signal, to model the observed solar occultation light curves. These models constrain the optical depth, radial width, and particle size distribution of the F ring. We find that when the diffraction signature is present, we can best reproduce the occultation data using a particle population with an average effective particle size of less than 300 μm, while occultations without clear diffraction signals are best modeled using a population with an effective particle size larger than 400 μm.