An Outline of a Dynamic Solution to Spatio-temporal Windowing in an Urban Gis

In traditional database use it is often desirable to extract a particular “view” or subset of the data in order to work on a particular problem. After the work on the view is completed, the problem becomes how to replace this view, and the modifications made, back in the original main database. The same problem arises in the field of spatio-temporal data in a GIS, where not only do we have desired views that are only a spatial subset of the total (a window), but they may only be concerned with a particular point or period in time. Traditional approaches to this problem usually result in recreating the whole network and topology after a single modification of the spatial database. One reason for this is the fundamental “batch” approach of the pointer structure in a GIS topological spatial data structure, which prevents incremental updating and downdating of the spatial database. With the recent development of dynamic spatial data structures, such as the Voronoi model, it is possible to add and remove individual map objects, but to preserve and maintain the integrity of the topological relationships while doing so. In the scenario of spatio-temporal windowing, however, special operations are needed to update the topology. Since a window is simply a partial view of the whole database, even if the topology in the window is modified and correct it may not be correct when re-merging the window into the main database. This paper outlines solutions towards problems of dynamic windowing with the Voronoi data structure. The problems include the construction of different windows, spatial operations allowed inside a window, and topology checking while merging with the main database. While dynamic windowing is virtually impossible with a “batch” mode GIS, discusses in this paper reveal that these problems can be handled rather easiiy. The principles discussed here could also be applied to design a dynamic distributed urban GIS and to parallelization of processors of dynamic GIS functions.