"Long-term erosion of mountain belts is driven by discrete landsliding events that deliver sediment to rivers, and flood events that transport sediment away and incise bedrock. Landsliding and flooding events obey different frequency-magnitude distributions, are partially correlated (e.g., storms trigger landsliding and flooding, earthquakes only trigger landsliding), and interact along the drainage network to generate a complex spatio-temporal response still poorly understood at the event scale (i.e. how fast is a landslide deposit eroded, and how does it impact the downstream evolution of a channel ?) and over the long-term (how important are extreme events on river dynamics, landscape morphology and sediment fluxes ?). To progress on this topic, we propose to combine innovative field measurements using Terrestrial Laser Scanner, documenting individual flood and landslide impacts on channel morphology in the extremely active Southern Alps of New-Zealand, and 2D stochastic numerical modeling of landscape evolution to explore the long-term consequences and improve on the hazard assessment of rare extreme events such as landslide dams. The applicant will complement his leading expertise in numerical and experimental modeling of channel dynamics by being trained in landslide dynamics and risk assessment at Dept of Geological Sciences (Univ. Canterbury, N-Z), a leading institution on this topic with unique access to the nearby Southern Alps. During the return phase, he will apply his new skills to implement a stochastic landslide module to the landscape evolution code developed at Geosciences Rennes (CNRS, Univ. Rennes, Fr), validate model outputs against field data, and explore the long-term consequences for mountain erosion of the combined short-term stochasticity in landsliding and flooding. To further improve his career and strengthen the link between Europe and New-Zealand, the fellow will also prepare and submit a Marie-Curie Initial Training Network proposal."