Study of overland flow with uncertain infiltration using stochastic tools

Marie Rousseau (1) , Olivier Cerdan (2) , Alexandre Ern (1) , Olivier Le Maitre (3) , Pierre Sochala (2)

1 CERMICS - Centre d'Enseignement et de Recherche en Mathématiques et Calcul Scientifique
2 BRGM - Bureau de Recherches Géologiques et Minières (BRGM)
3 LIMSI - Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur

The saturated hydraulic conductivity is one of the key parameters in the modelling of overland flow water fluxes. In this study, this parameter is defined as a stochastic parameter, idealized as a piecewise constant random field with uniform distribution. This paper aims at investigating the effects of the spatial and temporal scales in uncertainty propagation within overland flow models, and at identifying the localization of the most influential saturated hydraulic conductivity using sensitivity analysis. The results show that the influence of saturated hydraulic conductivity depends on the soil saturation and its spatial localization. For instance, in case of low saturated soils, the most influent parameter is the one located downslope, whereas in case of high saturated soils, the most influent one is either the most infiltrating or the intermediate one. The results indicate where efforts should be concentrate when collecting input parameters to reduce modelling uncertainties.

Mots clés: overland flow; infiltration; saturated hydraulic conductivity; Monte Carlo sampling; stochastic model; sensitivity analysis

Domaines: Modélisation et simulation

Marie Rousseau, Olivier Cerdan, Alexandre Ern, Olivier Le Maitre, Pierre Sochala. Study of overland flow with uncertain infiltration using stochastic tools. 2011. ⟨hal-00612949⟩