In the context of climate and environmental change, the boreal forest is subject to potential changes in structure and function. Stand-level physiological models can be used to predict these responses over time and to understand the interaction between tree ecophysiological processes and climate variability. We present here a project that aims to develop potential growth scenarios for Canadian boreal forest stands based on the characterization of their ecosystem fluxes using the MAIDEN model. First, we will assess site-level processes using a data-model fusion approach. Using inventory and tree-ring data, and flux towers data, the MAIDEN carbon allocation model will be calibrated and refined to highlight the relationships between climate variables and carbon allocation. Various optimization methods will be used including gradient descent algorithms in Bayesian frameworks. Landscape level simulations will then be performed and validated using MODIS satellite images. The calibrated model will be used to run potential growth scenarios for boreal forest species and determine the evolution of forest ecosystem fluxes over time. We will use climate scenarios to predict the evolution of climate variables, and the results will be analyzed by various statistical methods to determine the involvement of climate in the evolution of the productivity of the species studied. The poster will present the proposed data-model fusion approach, data sources used to inform the modelling, the MAIDEN carbon allocation scheme and preliminary simulations for specific sites. The project will improve understanding of how boreal tree species grow under changing climatic conditions, and the impact of future climate variability on boreal carbon fluxes.