Maritime traffic monitoring needs tools for spatiotemporal decision support. The operators responsible (e.g. the Coast Guard) must monitor vessels that are represented as objects moving in space and time. Operators use maritime tracking systems to follow the evolution of traffic and make decisions about the risks of a situation. These systems are based on Geographic Information Systems (GIS) and OnLine Transaction Processing (OLTP) approaches, which are prohibitively expensive, very slow and produce operational data unsuited to decision-making. Instead, operators require summarized data that is easier for them to produce and use. Therefore, we propose the definition of a geographical decision-making chain that adds a decision-making dimension to current systems. It consists of a carefully assembled set of tools that can automate the three phases of Business Intelligence, namely data loading, modelling and analysis.