This three-year project proposes to improve our knowledge about the impact of landslides on CRN-derived denudation rates in landslide-prone mountainous environments. The central hypothesis is that under specific conditions in such environments, CRN-derived denudation rates are (1) overestimated compared to the volumetrically-derived denudation rates because landslides are diluting cosmogenic radionuclide concentrations in river sediments, and (2) highly variable due to the stochastic nature of landslides. We verified this hypothesis in a preliminary study (in prep.) On a simplified landscape by iterating the CRN numerical model developed by Yanites et al. (2009), including the stochastic nature of landslides. This proof of concept shows that the stochastic character of landslide might lead to substantial differences in the overestimation of CRN-derived denudation rates, i.e. There is an important variability in the estimated values.
The idea is therefore to properly identify the factors controlling both variability and overestimation, and to eventually provide guidelines for river sediment field sampling strategies for geomorphologists seeking to quantify long-term CRN-derived denudation rates. To this end, we aim to develop a spatially-explicit numerical model of the evolution of in-situ cosmogenic nuclide concentrations in landscapes that are affected by landslides. The modelling effort will be constrained by field and remote sensing data to characterize both CRN concentration replicatibility and the influence of landslide magnitude-magnitude frequency. Because such modelling effort can be tedious and requires validation, the project is designed as a stepwise procedure, progressively increasing the complexity of the model.