There is a widespread perception that the atmosphere and the climate are beginning to change, and that these changes could have profound impacts on the primary productivity of the terrestrial biosphere. The terrestrial biosphere is a dynamic system that interacts with the atmosphere and climate principally through the exchanges of energy, water, and elements. Due to the limitations of equilibrium terrestrial biosphere models, new generation models - dynamic biosphere models, are critically needed for assessing and predicting the primary production and biogeochemical cycles of the terrestrial biosphere in changing global environment. The goal of dynamic biosphere modeling is to model terrestrial ecosystem dynamics induced by natural and anthropogenic disturbances, as well as the interactions of energy, water, and carbon cycles within the terrestrial biosphere and between the terrestrial biosphere and the atmosphere. The critical gaps in developing such a terrestrial biosphere model are not our inability to construct model code but instead the poorly developed links between empiricism and the concepts we used to construct our models, especially a lack of data that would help to make our models mechanistic, an incomplete fundamental knowledge about how complex terrestrial ecosystems work, a poor understanding of how to scale up what we do know and of how to validate such a model. The interaction among data, model structure, parameter sets, and predictive uncertainty will Bead to important progress in the development of dynamic biosphere models.