A framework for correcting geographical boundary inconsistency

Abstract

Spatial data quality (data accuracy, precision, consistency and so on) is a key issue in Geographic Information System. Geographical boundary inconsistency will directly affect the correctness and efficiency of analysis in GIS application. This paper describes a framework for checking and correcting geographical boundary inconsistency. Two kinds of inconsistency are identified: geometric inconsistency and topological inconsistency. Geometric inconsistency refers to the unequal of point number or coordinates of overlapping boundary. Topological inconsistency means that adjacent boundary is not able to strictly guarantee topological consistent according 9-intersection model, and generate series of gap and fragment area. Different checking and correcting methods are adopted for these two different kinds of inconsistency. In the first case, the generalized algorithm of node snapping is used, and inconsistency is solved by the mathematical methods of repeating points removing, method of averaging, method of projection, etc. In the second case, the inconsistency is solved by buffer operation, Delaunay triangulation, overlay analysis, etc. A complete framework and algorithm procedure are given in this paper to detect and correct the boundary inconsistency problem in GIS. Meanwhile, an application of inconsistency correction to land-use data based on this framework is conducted.