Development and implementation of a spatio-temporal data model for parcel-based land information systems

This thesis discusses spatial and temporal aspects of the cadastral system, functional requirements for temporal parcel-based Land Information System (LIS), design of a spatio-temporal data model and associated functions, and implementation for verification.; The nature of cadastral data in the U.S. is dynamic in a temporal perspective and measurement-based with contradictions in a spatial perspective. The fundamental characteristics have been comparatively ignored in the development of LIS. Functional requirements with respect to the characteristics are investigated based upon literature review and interviews of professionals in six target groups: (1) surveyors; (2) title insurance workers; (3) the register of deed office; (4) property assessors; (5) planners; and (6) GIS applications developers and users. The investigation ended up with six major functional requirements: (1) incorporation of measurement data; (2) identification of spatial discrepancy; (3) retention of historical data; (4) temporal topology; (5) snapshot; and (6) incorporation of document image data. Based on the findings concerning functional requirements, a spatio-temporal data model is proposed with investigations of some issues such as time semantics and temporal topology. In addition, associated functions—least squares adjustment, blunder detection, automatic coordinate update, and geometric process for identifying spatial discrepancies—are presented for temporal parcel-based LIS. The proposed data model and required functions were implemented using Dane county LIS data. A pilot system was developed using prolog, XPCE, and C++, and titled Temporal Land Information System (TLIS).; The author presents the proposed data model with associated functions for a potential solution, in order to extend the capability of current parcel-based LIS and to support more users' needs.