The widespread use of positioning technologies such as GPS has led to the collection of massive amounts of mobility data. This data captures how objects—especially people and vehicles—move through space over time and can provide valuable insights across many application domains. One typical example is tourism analysis in cities. Tourists visit museums, parks, and other attractions, and use services such as hotels, restaurants, and shops. These locations are referred to as places of interest. A tourist trajectory describes the movement of a tourist from one place of interest to another, including stops at certain locations. By collecting and analyzing these trajectories, city planners and businesses can optimize service offerings, design better tourist routes, and improve urban management. 

Another important application area is environmental monitoring in large industrial cities where air quality is a concern. Cities often deploy monitoring stations to measure air pollution at regular intervals. Mobility data from cars, trucks, and buses can be analyzed together with air quality measurements to understand how traffic patterns influence pollution levels. Analysts can also study which populations are exposed to high pollution and at what times. The techniques used in spatial and data warehouse systems are therefore highly relevant for understanding the evolution of environmental conditions and evaluating the impact of policies aimed at reducing pollution. 

Traditional spatial databases and data warehouses represent the spatial characteristics of objects, such as locations or boundaries. Although these spatial features may change over time, the changes are often considered discrete events, such as the merging of land parcels or changes in administrative borders. In contrast, the focus here is on moving objects, whose spatial properties change continuously over time. These objects are typically modeled as moving points, although some applications also involve moving regions, such as tracking pollution clouds or oil spills. 

Trajectories of moving objects can be represented in two main ways: continuous and discrete. A continuous trajectory represents the movement of an object over a time interval, enriched with interpolation functions that estimate its position at any moment within that interval. This allows analysts to approximate the object’s location even between recorded points. A discrete trajectory, on the other hand, consists of a sequence of recorded spatiotemporal positions without reliable interpolation between them. For example, social network check-ins at different times and locations may provide discrete trajectory points. Interpolating movement between these check-ins may not be meaningful for analyzing movement paths, but the data could still be useful for studying presence in specific areas. The distinction between discrete and continuous trajectories depends more on application semantics than on the time gap between recorded points. 

Spatiotemporal databases, also called moving object databases, are designed to store and query the positions of moving objects. They can answer operational queries such as predicting when a train will arrive or determining the speed at which a geographic region is changing. However, they are not optimized for complex analytical queries. Questions such as the total number of deliveries started in a city over a period or the average delivery duration by location are better handled in a data warehouse environment. 

To support such analysis, conventional data warehouses can be extended to include moving object data, resulting in spatiotemporal or trajectory data warehouses. These systems store not only alphanumeric and spatial data but also trajectory information. Trajectories are often analyzed together with other spatial data, such as road networks, and continuous field data like temperature or elevation. This integration allows advanced analysis that combines movement patterns with environmental or contextual information. 

Supporting spatiotemporal data requires specialized data types that capture how values evolve over time. These temporal types extend base and spatial types to represent dynamic changes. Together, these concepts form the foundation for modeling and analyzing moving objects and trajectories in modern data warehouse systems. 

Reference: 

Vaisman, A., & Zimányi, E. (2014). Data warehouse systems: Design and implementation. Springer.