Simply speaking, updates propagation is to use updates from one newly-updated database (called Master Database, MDB) to revise, improve and correct the content of the other database (Called Client Database, CDB) for ensuring that CDB also has a current representation of the terrain. Currently, updates propagation between spatial databases is usually a complex and time-taking process because of the multi-level potential heterogeneities between them and the special requirement of keeping the updated CDB autonomous, complete, correct, and consistent as much as before. To simplify and improve the propagation of updates, a detailed classification of the possible heterogeneities between MDB and CDB is introduced and their influences on updates propagation are also analyzed. Based on these analyses, an appraoch based on schema matching is proposed to effectively implement updates propagation. The automated solutions to operations involved in this approach are discussed in detail, such as schema matching, entity identification, updates extraction, and updates integration.