Use of cellular automata in physical geography : Use of CA for assessing impacts of morphology on catchment hydrology during hyper-concentrated floods (Northern France)

International audienceHyper-concentrated floods are events regularly affecting during spring and summer the loamy plates of the Paris Basin, in North-West France. Such floods are characterized by sudden triggerings, fast rising times and important specific discharges. According to high rainfall rates, morphology is well-identified as the first-order controlling factor on floods dynamics, while cultivated areas and land use aggravate its influence in a second order. Traditionnal morphometric indexes are generally used to quantify effects of catchment form, slopes or the drainage network on hydrological processes. Therefore, these tools present important drawbacks. Indeed, they do not consider the dynamics generated by the structure of the catchments and they have been created in static planar dimension. Thus, influence of morphology on catchment hydrology has to be re-evaluated and we propose in this study a cellular automata application. RuiCells integrate the catchment geometry, slope patterns and the drainage networks at a cellular scale to simulate the influence of topography on the surface flow dynamics at the overall catchment scale. Results were that: i) the influence of the spatial organisation of cells within the catchment and their connection to the drainage network is more important than the catchment size; ii) simulations reveal the intrinsic relationships between functional morphological areas and water paths; iii) spatial scaling effects and hydrological behaviours are recognized at the global scale but a small catchment part can induce violent hydrological response. Combinations between topography, runoff contributing areas and slopes seem to explain complex concentrated flow distributions in time and space