The mast concentrated deposits of SO2 frost on Io occur within a series of large equatorial basins. About 30% of the surface is covered by SO2 outside of the basins, increasing to more than 50% within the basins. This pattern is poorly expressed in the region from longitude 240 degrees to 360 degrees where bright areas are frequently buried by the fallout from the large Pele-type plumes. The fourfold pattern of alternating basins and swells in Io's equatorial region is similar to the heat-flow pattern predicted from tidal heating in a thin, partially molten asthenosphere. However, the topographic pattern is offset from the predicted heat-flow pattern; thus it is unclear whether topographic highs correspond to regions of higher or lower predicted heat flow. These two possibilities imply two very different models for Io's highlands: a thermal-uplift model or a continental-crust model. In the thermal-uplift model, the regions of enhanced asthenospheric heating cause lithospheric thinning and isostatic uplift, perhaps accompanied by uplift due to penetrative magmatism or basaltic underplating. In the continental-crust model, ''continents'' of differentiated crust float on low-density roots, the crust and lithosphere are approximately one and the same, and basal melting controls its thickness. Although both models are plausible, the thermal-uplift model best explains the SO, distribution. Cold trapping must be important for concentrating SO2 frost in optically thick patches; thus either cold traps are preferentially initiated over large basin areas or they are preferentially removed from the highlands. The patchy distribution and approximately 30% SO2 coverage of the highlands show that cold traps are abundant here, but not extensive; thus the SO2 must be preferentially removed and/or buried. Higher heat hows in the highlands should lead to increased volatilization of SO2 frost, and a greater frequency of relatively SO2-poor volcanism should tend to bury frost patches. This model links asthenospheric tidal heating, large-scale heat flow and topography, volcanic activity, and the global distribution of surface SO2, and it leads to several specific predictions for future observations. (C) 1995 Academic Press, Inc.