For spectral-temporal combined analysis, all images are stacked in one data set and classified together, all at once, similar to a multispectral classification with many bands. Changes are identified and labelled. This can be time-saving, but it may be difficult to identify and label the change classes. A complete change matrix cannot be provided.