A feature-based linear data model supported by temporal dynamic segmentation

To effectively serve transportation agencies, GIS-T applications need the backings of linear data models that integrate various linear features with networks. In addition, GIS-T applications need the support of dynamic segmentation that handles temporal as well as spatial aspects of both features and networks.; One such linear data model, Feature-Based Linear Data (FBLD) model, is proposed in the first part of the study. At its core is a Feature-Based Linear Referencing (FBLR) framework capable of communicating both with GIS datasets and with features referenced by various linear Location Reference Methods (LRMs). Discussions are not limited to the architecture of the model. Implementation issues involving the key component designs, reference ambiguity resolutions, and the generation of the FBLR framework, etc. are explored as well.; The second part of the study focuses on extending classic dynamic segmentation into the temporal domain. The foundations on which such extensions are made are first established. These include the definitions of spatiotemporal segments, spatiotemporal join operations, and criteria for identifying segment topological relationships. The proposed extensions involve in two areas of dynamic segmentation. In the area of network maintenance, the composite versioning approach is the primary approach used to track temporal evolutions of networks at different levels. In the area of linear feature processing, operational requirements are specified in the three functional categories: segment geocoding, segment modification, and segment overlay.; Finally, a prototype application on roadway inventory management is presented to prove the validity and feasibility of the proposed linear data model and temporal dynamic segmentation.