I use data-assimilation techniques and remote sensing observations to constrain spatially distributed models of terrestrial carbon cycle. The aim is to better quantify modern land-atmosphere carbon fluxes and the evolution of biomass stocks through the representation of natural and human-driven processes involved in the terrestrial carbon cycle: plant uptake, allocation, residence time, decomposition, respiration, fire, logging. Better understanding the current terrestrial carbon cycle is required to assess the potential of using the land surface in future mitigation efforts.
Due to my background in hydrological sciences, I am also involved in studies on the impact of climate change on agriculture, and looking at climate change and human pressure impacts on water resources in river basins worldwide. I have a keen interest in innovative numerical methods based on machine-learning techniques to explore large observational datasets.