Spatio-temporal properties (STP) graph model for multiple-entity variables estimation

Modeling the complex interaction and change characteristics of human and environmental systems is one of key issues for environmental variable estimation and integrated assessment. However, a formal spatio-temporal model for integrated environmental system state estimation and assessment is not available. Traditional spatio-temporal data model, such as event based models, is not suited for temporal GIS application and environmental state estimation. This paper has developed a spatio-temporal properties graph model for multiple entities property variables estimation. The geospatial multiple entity-property representation has been discussed based on object entity, field entity, property and relations. A spatio-temporal properties graph model has been designed, which the nodes of the graph represent spatio-temporal property variables and the edges of the graph represent relations among multiple entities properties. Based on the spatio-temporal properties graph model, three methods of multiple entities variables estimation are introduced, i.e., structural equation estimation method, system dynamics estimation method, and Bayesian estimation method. An example of interval flood estimation and warning is developed for the illustration of the purposed approach.