Better prediction and understanding of forest dynamics is important because of the many ecosystem services provided by forests – global climate regulation via carbon sequestration, provisioning of drinking water, flood regulation, erosion control – in addition to the habitat they provide for many species (supporting biodiversity). Climate change is expected to change the geographic distribution of the forest biome, with uncertain consequences with respect to these ecosystem services. Commitments made at the COP21 meeting, in particular, heighten the need to account for the role forests play in the global carbon cycle and anticipate their future behavior. Forests are complex systems that offer many entry points for study, both with respect to sources of data (forest plot, tree ring, flux tower, pollen, functional trait, remotely sensed), and modeling approaches (ecosystem, dynamic vegetation, forest gap, demographic, and physiological models). In this session, we bring together researchers modeling forest dynamics from diverse perspectives – carbon dynamics, ecophysiology, population and community ecology, paleoecology, and forestry. The goal is to span the many kinds of data relevant for inferring forest dynamics, with an emphasis on combining data sources (data assimilation, inverse modeling) to better constrain parameters, and synthesis across modeling approaches to address the many processes and spatio-temporal scales important for forest dynamics.