Spatial & Temporal Analysis and Visualization

Spatial and Temporal Analysis aims at better understanding the causes and consequences of (human) activities, revealing the structure and organization of social and environmental phenomena over a geographical space (a territory). It relies on the modelling of the geographical space and of its evolution through time. Statistical methods applied to spatially and temporally referenced data support reasoning, diagnosis, and decision. Such spatio-temporal data, called indicators, measure different kindsof phenomena: socio-economical (population, GDP…), environmental (forest, water…), epidemiological (flu, cancer…), etc. Privileged tools for handling such indicators – from the producer (for instance, a national statistical agency) to the consumer (general public) – are Spatial Data Infrastructure (SDI), which commonly embed a set of software, data repositories (databases and/or data warehouses), and
services in order to access, query, analyze and visualize data. W3C, OGC (Open Geospatial Consortium) and ISO provide standards (GML for exchanging geographical features, WMS, WFS, WPS, CSW as geographical web services…) and recommendations for designing, building and assembling such SDI components. The Inspire European directive and OpenData initiatives launched by governments have also pushed forward the development of SDI. In this context, the main goal of STeamer is twofold. First, we aim at providing models, methods and tools that make it possible to build Spatial and Temporal Data Infrastructures, able to handle the evolution through time of: 1) data (i.e. handling time series), 2) indicators (i.e. coping with possible change in their definition); and 3) territorial units upon which such data are collected (i.e. managing changes in contours, frontiers and names). Second, we aim at handling the imperfection of information, considering that spatial data and values of indicators might be missing, imprecise, uncertain, incomplete, and inconsistent. With this respect, STeDI, the SpatioTemporal and evolving Data Infrastructure, designed and developed in the ESPON M4D Pro ject, as well as C. Plumejeaud’s PhD Thesis (2011) are salient results.
We have also investigated the way of representing and visualizing spatiotemporal data. Initially,
our scientific contribution in the field of cartography and visualization of spatiotemporal information has focused on the automatic generation of synchronized interfaces based on spatial (maps, sketches, …), temporal (time-lines, histograms,…) and thematic (tables, charts, multimedia files…) graphical components by promoting the use of several models (data model, presentation model, functionality model, adaptation model, user model…) during the design phase. A. Arnaud’s PhD Thesis (2009) gives the specifications of new cartographic representation modes (interactive, animated, multimedia maps) for displaying heterogeneous and imprecise spatiotemporal data about natural risks. This geographer work has inspired M. Snoussi’s on-going PhD whose goal is to build an interactive visual environment for assessing imperfect spatial and temporal information in the field of natural hazards. P.-A. Davoine’s HDR (2014) provides larger insights about the way spatiotemporal data have to be collected, modelled and visualized for the prevention of natural hazards. A pedagogical perspective and a cognitive approach has been recently brought by R. Balzarini in her PhD Thesis (2013), which sheds a new light in understanding and therefore promoting the use of cartographic tools in Environmental Sciences.