Dynamic Global Vegetation Modeling: Quantifying Terrestrial Ecosystem Responses to Large-Scale Environmental Change.

DGVMs exploit the power of modern computers and computational methods to yield a predictive description of land ecosystem processes that takes account of knowledge previously developed through long histories of separate disciplinary approaches to the study of the biosphere. The degree of interaction between the different scientific approaches still falls far short of optimal; thus, DGVM developers have a responsibility to be aware of progress in several disciplines in order to ensure that their models remain state-of-the-art. We have presented a series of case studies of the evaluation of DGVMs that demonstrate the predictive capability that current models have achieved. Nevertheless, there are plenty of unresolved issues — differences among models that are not well understood, important processes that are omitted or treated simplistically by some or all models, and sets of observations that are not satisfactorily reproduced by current models. More comprehensive "benchmarking" of DGVMs against multiple data sets is required and would be most effectively carried out through an international consortium, so as to avoid duplicating the large amount of work involved in selecting and processing data sets and model experiments. We have also presented a series of case studies that illustrate the power of DGVMs, even with their known limitations, in explaining a remarkable variety of Earth System phenomena and in addressing contemporary issues related to climate and land-use change. These case studies encourage us to believe that the continued development of DGVMs is a worthwhile enterprise. Finally, new directions in Earth System Science point to a range of aspects in which DGVMs could be improved so as to take account of recently acquired knowledge, such as experimental work on whole-ecosystem responses to environmental modification and new understanding of the functional basis of plant traits; complemented by an effort to represent semi-natural and agricultural ecosystems and the impacts of different management practices on these ecosystems; and extended to include processes such as trace-gas emissions, which are important in order to understand the functional role of the terrestrial biosphere in the Earth System. Together, these potential developments add up to an ambitious research program, requiring the economies of scale that only an international collaborative effort can provide. [ABSTRACT FROM AUTHOR]