Greenhouse gas emissions from farms can be suppressed in two ways: by curtailing the release of these gases (especially N2O and CH4), and by storing more carbon in soils, thereby removing atmospheric CO2. But most practices have multiple interactive effects on emissions throughout a farm. We describe an approach for identifying practices that best reduce net, whole-farm emissions. We propose to develop a “Virtual Farm”, a series of interconnected algorithms that predict net emissions from flows of carbon, nitrogen, and energy. The Virtual Farm would consist of three elements: descriptors, which characterize the farm; algorithms, which calculate emissions from components of the farm; and an integrator, which links the algorithms to each other and the descriptors, generating whole-farm estimates. Ideally, the Virtual Farm will be: boundary-explicit, with single farms as the fundamental unit; adaptable to diverse farm types; modular in design; simple and transparent; dependent on minimal, attainable inputs; internally consistent; compatible with models developed elsewhere; and dynamic (“seeing”into the past and the future). The Virtual Farm would be constructed via two parallel streams - measurement and modeling - conducted iteratively. The understanding built into the Virtual Farm may eventually be applied to issues beyond greenhouse gas mitigation. Key words: CO2, N2O, CH4, agroecosystems, models, climate change
Elevated salinity and water table drawdown significantly affect greenhouse gas emissions in soils from contrasting land-use practices in the prairie pothole region