Modelling in Geography
- Cet article est une traduction de :
- Printemps de la modélisation
1Mathematical and computer formalization in its early development was experienced as a constraint requiring an explicit representation of the object of study. This was a view held by dedicated and hard-working model builders as well as by those who, profiting from the prestige of the label, dubbed as 'models' what were crude representations with little or no formalization. Recent advances, however, mean that formalization can be an opportunity to expand understanding. New ways of representation using powerful computer tools offer far greater scope than previously for simulation, experimentation and interaction with the infinitely diverse local expressions of more general rules.
2The Groupe de Recherche Libergeo (CNRS) is making a Tour de France of geographical model-builders. Thirty geographers from ten laboratories in ten cities (Aix, Besançon, Dijon, Lille, Metz, Montpellier, Paris, Rouen, Strasbourg, Tours, and more perhaps later) have decided to share their experience and organize at roughly two-monthly intervals a visit to one of the sites. These model builders are noteworthy for having a very broad-based age pyramid: four-fifths of those involved are under 30. Each visit is in two parts: in the morning the models are presented by those who have designed or developed them; in the afternoon they are subject to experimentation. This is an opportunity to acquire if not a full introduction to the models at least a greater awareness of what they can actually achieve and of the problems associated with their application.
3Cellular automata, multi-agent systems, dynamic models and methods of spatial analysis are currently coupled to the resources of geographical information systems. Use of these techniques is increasingly widespread among young geographers. But the new models being developed must not remain confined to single institutions or to the pages of academic theses. They must be compared with other models, providing the basis for empirical testing or improvement, while models that are found to be useful need to be brought to the attention of the wider geographical community. It is also to be hoped that the models will be made available in simplified versions for educational and recreational purposes so that a broader public can become familiar with geographical principles and concepts, transmitted by clear representations and involving an unambiguous interplay between general rules and local exceptions. But there will still be a place for fuzzy logic! (as long as the second of those terms is not lost sight of). In short, modelling as practiced by our young geographers holds great promise for the future.
Pour citer cet article
Denise Pumain, « Modelling in Geography », Cybergeo : European Journal of Geography [En ligne], Éditoriaux, mis en ligne le 12 mars 2000, consulté le 30 avril 2017. URL : http://cybergeo.revues.org/7472Haut de page
© CNRS-UMR Géographie-cités 8504Haut de page