When we want to detect and classify fires in multispectral satellite images (e.g., various types of forest fires and their consequences), we are faced with the problem of how to recognise and interpret the visible fire patterns, and how to discriminate fires from smoke and clouds (and other less relevant classes). Currently preferred image classification techniques often rely on deep learning (DL) approaches calling for extensive training efforts with lots of training data. Thus, we must apply other techniques if no such data are readily available. In this latter case, we can resort to Latent Dirichlet Allocation (LDA), a dictionary-generated image topic-based classification approach, where no explicit training and learning steps need to be added. Then, the correct assignment of semantic meaning to individual image patches can be accomplished by a priori image knowledge and some off-line image interpretation provided by expert users. In the following, we describe a Latent Dirichlet fire detection and visualisation approach and its performance based on an optimised selection of Sentinel-2 satellite spectral image bands. The combination of automated image classification and multi-colour visualisation seems to be an interesting alternative to deep learning. Thus, our paper supports “data to knowledge”.