Rice greenhouse gas (GHG) emissions in the Red River Delta, Vietnam, were mapped using multiscale satellite imagery and a processed-based biogeochemical model. Multiscale synthetic aperture radar (SAR) and optical imagery were fed into a random forest classifier using field observations and surveys as training data to map rice extent. Time series analysis then generated geospatial information on crop calendar, hydroperiod and cropping intensity to use as parameters for the denitrification–decomposition (DNDC) model to estimate emissions.