Benchmarking ecosystem response models with experimental data from long-term CO2 enrichment experiments

Computer models that connect the earth’s climate system with the cycling of carbon between the atmosphere, ocean, and terrestrial biosphere are prominent in the Intergovernmental Panel on Climate Change’s Fourth Assessment Report, Climate Change 2007. These coupled carbon-climate models are used to predict how human activities will alter the composition of the atmosphere and cause a change in climate. The high profile of this activity makes it especially important that the terrestrial carbon cycle is represented correctly so that we can have confidence in larger-scale model predictions. This working group will advance improvements in how the impact of altered atmospheric CO2 concentration is represented in carbon cycle models. We will use some of the longest and most comprehensive data sets on CO2 impacts on ecosystems from field experiments to provide a standard for evaluating 12 ecosystem process and land surface models. The models, which are being used for predicting terrestrial response to atmospheric and climatic change, will be parameterized with site and weather data from the Duke University and Oak Ridge National Laboratory free-air CO2 enrichment (FACE) experiments. We will evaluate the ability of the models to reproduce the measured processes of the carbon, water, and nitrogen cycles of the Duke and Oak Ridge experimental stands and their responses to elevated atmospheric CO2 concentration. Similarities and differences among the models and their components will provide guidance for improving all of the models. With the experimental data as a benchmark for model performance, the utility of the models for extrapolation to environmental change questions can be demonstrated with increased confidence. This data-model intercomparison project has the potential to provide better scientific outputs for policy making.