Evaluation of a lanscape evolution model to simulate stream piracies: Insights from multivariable numerical tests using the example of the Meuse basin, France
Résumé
In landscapes dominated by fluvial erosion, the landscape morphology is closely related to the hydrographic network system. In this paper, we investigate the hydrographic network reorganization caused by a headward piracy mechanism between two drainage basins in France, the Meuse and the Moselle. Several piracies occurred in the Meuse basin during the past one million years, and the basin's current characteristics are favorable to new piracies by the Moselle river network. This study evaluates the consequences over the next several million years of a relative lowering of the Moselle River (and thus of its basin) with respect to the Meuse River. The problem is addressed with a numerical modeling approach (landscape evolution model, hereafter LEM) that requires empirical determinations of parameters and threshold values. Classically, fitting of the parameters is based on analysis of the relationship between the slope and the drainage area and is conducted under the hypothesis of equilibrium. Application of this conventional approach to the capture issue yields incomplete results that have been consolidated by a parametric sensitivity analysis. The LEM equations give a six-dimensional parameter space that was explored with over 15,000 simulations using the landscape evolution model GOLEM. The results demonstrate that stream piracies occur in only four locations in the studied reach near the city of Toul. The locations are mainly controlled by the local topography and are model-independent. Nevertheless, the chronology of the captures depends on two parameters: the river concavity (given by the fluvial advection equation) and the hillslope erosion factor. Thus, the simulations lead to three different scenarios that are explained by a phenomenon of exclusion or a string of events.