The Congo Basin hosts the second largest expanse of tropical rainforests in the world. As such, it plays a critical role in our changing Earth System. Yet, there is a striking discrepancy between the paramount importance of Congo Basin forests on the one hand and their scientific coverage on the other. This explains why the scientific community has not developed, as of today, reliable modelling tools capable of projecting the impact of future changes on the forest state and functioning. In this project, I propose to fill this gap by combining a state-of-the-art Land Surface Model with the most up-to-date observations of the carbon cycle and functional diversity in the Congo Basin. More specifically, I will develop a vegetation model (ED2) using multiple data sources from intact and disturbed forests, upscale the model simulations, and validate model outputs using recent remote sensing products. Data for model calibration and validation include (1) dynamics of carbon and functional diversity observed in multiple chronosequences, (2) repeated inventories of dozens of intact forests, (3) eddy covariance data from the first fluxtower in the Congo Basin and (4) remote sensing observations of human-induced disturbance, forest structure, productivity and biomass. I will then use the model to examine forest resilience under current and future climate and land-use change scenarios and to detect any potential large-scale tree cover loss in the Cong Basin by the end of this century.