부산대학교 도서관

Accurate prediction of future atmospheric CO2 concentrations is essential for evaluating climate change impacts on ecosystems and human societies. One major source of uncertainty in model predictions is the extent to which global warming will increase atmospheric CO2 concentrations through enhanced microbial decomposition of soil organic carbon. Recent advances in microbial ecology could help reduce this uncertainty, but current global models do not represent direct microbial control over decomposition. Instead, all of the coupled climate models reviewed in the most recent Intergovernmental Panel on Climate Change (IPCC) report assume that decomposition is a first-order decay process, proportional to the size of the soil carbon pool. Here we argue for the development of a new generation of models that link decomposition directly to the size and activity of microbial communities in coupled global models. This process begins with the formulation and validation of fine-scale models that capture fundamental microbial mechanisms without excessive mathematical complexity. These mechanistic models must then be scaled up through an aggregation process and validated at ecosystem to global scales prior to incorporation into global climate models (GCMs). Parameterizing microbial models at the global scale is challenging because some microbial properties such as in situ extracellular enzyme activities are very difficult to measure directly. New parameter fitting procedures may therefore be needed to infer the values of important microbial variables. Validating decomposition models at the global scale is also a challenge, and has not yet been accomplished with the land carbon models embedded in current GCMs. Fortunately new global datasets on soil carbon stocks and fluxes offer promising opportunities to validate both existing land carbon models and new microbial models. If challenges in scaling, parameterization, and validation can be overcome, a new generation of microbially-based decomposition models could substantially improve predictions of carbon–climate feedbacks in the Earth system. Development of new models structures would also reduce any bias due to the assumption of first-order decomposition across all of the models currently referenced in reports of the IPCC.