Claude Grasland

Spatio-temporal analysis of territorial changes from a multi-scale perspective

The full integration of time dimensions in GIS still represents a research challenge. A number of systems provide ways to visualise harmonised timestamped geographical data onto maps, with interpolated curves representing how these data have changed over time. However, these systems frequently mask the concerns linked to changes in territorial organisations and the harmonisation of data. In fact, the development of a territory should be considered in connection with its neighbourhood, governance and genealogy relationships. This article focuses on the difficulties linked with the ‘change of support problem’, which can arise when conducting spatio-temporal analysis of data. First, we present a data model handling the changing relationships between territories. Then, we illustrate how this model can be instantiated using examples that have taken place in Europe. We show how the model can address questions such as that of change blindness. Finally, we explain how hierarchical and genealogical relationships can be used inside an interactive cartographic tool for spatio-temporal analysis. It provides various views of the same phenomena at multiple scales, through an approach that takes into account the changes in territorial organisation.

HyperSmooth: A System for Interactive Spatial Analysis Via Potential Maps

This paper presents a new cartographic tool for spatial analysis of social data, using the potential smoothing method [10]. The purpose of this method is to view the spread of a phenomenon (demographic, economical, social, etc.) in a continuous way, at a macroscopic scale, from data sampled on administrative areas. We aim to offer an interactive tool, accessible through the Web, but ensuring the confidentiality of data. The biggest difficulty is induced by the high complexity of the calculus, dealing with a great amount of data. A distributed architecture is proposed: map computation is made on server-side, using particular optimization techniques, whereas map visualization and parameterisation of the analysis are done with a web-based client, the two parts communicating through a Web protocol.