Isabelle Thomas, Department of Geography and CORE, UCL, Louvain-la-Neuve, Belgium, Pierre Frankhauser, THEMA
Université de Franche-Comté, Besançon, France
Delineating built-up landscapes by means of fractal indicators - Some empirical and theoretical thoughts (assigned to theme
This paper aims at applying several fractal methods (1) for measuring the morphology of built-up patterns, and (2) for delineating built-up landscapes at a regional scale. We show how far nowadays, traditional built-up landscapes are affected by urban sprawl. Two hypotheses are tested. An empirical one: to what extend do fractal measures help us to distinguish different kinds of settlement patterns, and how these results can be linked to the history and the geography of built-up landscapes? A more theoretical one: what kind of metric/method should best be used in order to classify fractal dimensions? Several fractal measuring methods are applied on all built-up surfaces of an entire European Nuts1 region (Wallonia, Belgium). Fractal parameters are computed for each of the 262 communes. Results are mapped and spatially analysed (full E.S.D.A.) and also compared to other variables commonly used in economic geography (rent, distance, income, planning rules, history) by means of traditional spatial analyses (correlation; clustering; principal component analyses). These analyses seem to be quite useful for describing built-up morphologies and for delineating landscapes by means of the built-up spaces only. These landscapes do not (or little) overlap with the traditional landscapes defined by means of physical criteria only. Fractal dimension is hence a quite promising tool for further simulating/understanding the genesis and planning at a regional scale.
 | Conference organized through conf-vienna
(copyright Gunther Maier) |