During the Voyager 2 encounter with Uranus in January 1986, the Planetary Radio Astronomy (PRA) experiment detected a complex pattern of radio emissions. Two types of emissions were seen: smooth and bursty. The smooth emission has been divided into smooth high-frequency (SHF) and smooth low-frequency (SLF) components which are presumed to come from different sources because of their distinctly different characteristics. The SHF component is considered in this paper. The SHF emission has been modeled by many authors on OTD (offset, tilted dipole (Ness et al., 1986)) L shells ranging from 5 to 40. However, the bursts have been modeled at much higher L shells. We complete an OTD investigation of the SHF emission at high L shells within the range of the bursty source locations, and present a viable high L shell model. This model has fundamentally the same longitudinally symmetric net emission pattern in space as the L shell 5 model presented in Romig et al. (1987) and Barbosa (1988). However, we were unable to produce an acceptable model on intermediate L shell without restricting source longitude.

We discuss the similarities and distinctions between our two models and the models of other authors. We believe that our high L shell model (and others similar to it) cannot account for the observed smoothness and periodicity of the SHF emissions because it has open field lines containing untrapped particles, which should produce more variable emission than that seen in the SHF data. Therefore, we prefer models at L shells less than 18, the boundary for closed field lines (Ness et al., 1986). We then discuss and contrast two models within this boundary: the L=5 model and an L≈12 model by Kaiser et al. (1987) and Farrell and Calvert (1989b). The main distinction between these two models is the longitudinal extent of the source location. The L=5 model has sources distributed 360¿ about the nightside OTD pole, while the L≈2 model has an asymmetric distribution with the sources confined to a small region, nearly a point. We propose several reasons why the symmetric distribution of the L=5 model may be preferable. Finally, we suggest that Miranda, which reaches a minimum L shell at L=5, may be related to the timing of several types of radio emissions. ¿ American Geophysical Union 1990