Satellite-measured SO2 mass loadings and ground-based measurements of SO2 emission rate are not directly comparable, with ~40% differences between mean emissions reported by each technique from Tungurahua volcano, Ecuador, during late 2007. Numerical simulations of post-emission processing and dispersal of Tungurahua's SO2 emissions enable more effective comparison of ground- and satellite-based SO2 datasets, reducing the difference between them, and constraining the impact of plume processing on satellite SO2 observations. Ground-based measurements of SO2 emission rate are used as the model input, and simulated SO2 mass loadings are compared to those measured by the Ozone Monitoring Instrument (OMI). The changing extent of SO2 processing has a significant impact on daily variation in SO2 mass loading for a fixed volcanic emission rate. However, variations in emission rate at Tungurahua are large, suggesting that overall volcanic source strength and not subsequent processing is more likely to be the dominant control on atmospheric mass loading. SO2 emission rate estimates are derived directly from the OMI observations using modelled SO2 lifetime. Good agreement is achieved between both observed and simulated mass loadings (~21%), and satellite-derived and ground-measured SO2 emission rates (~18%); a factor of two improvement over the differences found by simple direct comparison. Whilst the balance of emission source strength and post-emission processing will differ between volcanoes and regions, under good observation conditions and where SO2 lifetime is ~24 hours, satellite-based sensors like OMI may provide daily observations of SO2 mass loading which are a good proxy for volcanic source strength.