Introduction to spatiotemporal data models

Space, attribute, and time are three basic features of geographical phenomena and are also three basic data components of GIS databases. The space here refers to the spatial location data and its derived data. "Attribute" refers to the topic property data that has no derivative relationship with the spatial location. "Time" refers to the time-varying information of the time, space, and attribute status.

With the research and application of GIS based on spatial databases in recent years, the changing information has received more and more attention. Therefore, the concept of Temporal GIS (tgis) is proposed. The core of Temporal GIS is the space-time database, while the space-time Data Model is the basis of the space-time database. However, due to the complex relationship and structure of space, attribute, and time, the ideal time-space database and temporal GIS system have not yet appeared. Currently, there are several influential spatiotemporal data models:

① Time-space composite model

Each Independent superposition operation is converted to one-time merging and superposition, and the accumulation of changes forms the smallest unit of change, graphical files composed of the smallest units of change are associated with attribute files that record the change history to express the time-space characteristics of the data. The smallest variation unit is the largest homogeneous unit within a certain time and space range. Its disadvantage lies in polygon fragmentation and over-reliance on relational databases, which may cause many fragments as they change frequently.

② Continuous snapshot model

The continuous snapshot model only records the current data status in the database. After the data is updated, the change value of the old data is no longer retained, that is, the previous state of "forgot. The continuous time snapshot model saves a series of time segment snapshots to reflect the state of the entire space feature. Because snapshots store all unchanged features repeatedly, a large amount of data redundancy is generated. When the event changes frequently and the data volume is large, the system efficiency decreases sharply.

③ Ground state Correction Model

To prevent the continuous snapshot model from repeat the unchanged features, the ground state correction model only stores the data status (ground state) at a certain time point and the amount of changes relative to the ground state. The changed data is stored into the system only when the event or object changes. The temporal resolution scale value exactly corresponds to the time when the event or object changes. The ground state correction model only stores each object once and changes each time. Only a small amount of data needs to be recorded. The ground state correction model is also called the update model, which includes the vector update model and the grid update model. Its disadvantage is that it is difficult to process the spatial relationship between objects at a given time point. When retrieving distant past states, it is very inefficient to read the entire historical state.

④ Spatiotemporal cube Model

The space-time cube model uses ry to represent the process of two-dimensional graphics developing and changing along the time dimension. It expresses the evolution of the real-world plane position over time and marks the time on the spatial coordinate point. Given a time position value, the state of the corresponding section can be obtained from the three-dimensional cube, and the process of time-varying three-dimensional space can be extended. The disadvantage is that as the amount of data increases, operations on cubes become more and more complex, so that they cannot be processed.

⑤ Spatiotemporal Object Model

The space-time object model assumes that the world is composed of space-time atoms (spatio-temporal atom). The space-time atoms are entities with homogeneous time attributes and space attributes. In this model, the time dimension is perpendicular to the space dimension, which indicates the changes of the object in the space and attribute, but does not involve the representation of the gradient object. The disadvantage is that the gradual spatial changes that occur with time cannot be expressed in the time-space object model, and there is no concept to depict changes and processes.

⑥ Object-oriented spatiotemporal data model

The object-oriented method adds time information to the expressions of geometric elements such as nodes, arcs, and polygon, considering the spatial topology and temporal topology. A geographical entity can be modeled as an object no matter how complicated it is. The disadvantage is that there is a lack of explicit definitions of geographical entities or phenomena and basic relationship descriptions without considering the spatial and temporal characteristics and internal relationships of geographical phenomena.

In addition to these, common spatiotemporal data models include the first paradigm (1nf), the second Paradigm (1nf), and the second Paradigm (1nf) relational time-space data model, event-based time-space data model, historical graph model, and so on.