Predictions of forest ecosystem response to changes in climate and atmospheric CO 2 concentration require hierarchically structured process models. Present forest simulation models have conceptual limitations that restrict their application to climate-change studies. A major drawback of forest succession models is that they often lack physiological details in the simulation of annual tree growth. On the other hand, aggregated ecosystem models assume spatially homogeneous forests, and do not account for successional changes in forest composition and canopy structure. The concept of a new coupled carbon-water-energy-forest vegetation model is presented which attempts to overcome the main limitations of existing models by implementing a modern view of ecological hierarchy and a robust approach for scaling ecological processes in space and time.