What are wetlands and where are they? Part 2: Why are wetland areas and methane emissions so different among wetland-methane models and data sets?

Natural wetlands are central to understanding current and future interactions between climate and carbon cycling. They are the world's largest source of methane (CH4) to the atmosphere and their distribution and emissions are sensitive to interannual and longer-term variations in climate. Field observations confirm heterogeneous responses of CH4 emissions to climate variations governed by interacting influences of vegetation, climate, and environmental characteristics that differ among wetland ecosystems. Therefore, improving models of wetland-CH4 emission requires characterizing methane-relevant information across the spectrum of wetland variability. Modeling wetland extent and type--inextricably entwined with predicting methane emissions--remains ad hoc such that improvements in both are needed to increase predictive capability especially under future climate. Wetland distributions from data sets, and simulated or prescribed in wetland-methane models, diverge widely in part because no consensus exists on what and where wetlands are, i.e., wetlands are an ill-defined modeling target. Simulated wetland-methane fluxes also vary widely in magnitude, seasonality and geography due in part to wetland definition but also because few of the ~800 published CH4 flux observations have been used to develop and verify the models. Finally, no approach exists to link methane fluxes of wetland ecosystems represented in the literature to the global distribution of those ecosystems. We diagnose underlying causes for differences in wetland areas and distributions in models and data sets, and quantify their impact on modeled methane emissions. We present initial results from a coordinated effort to codify and amplify methane-relevant wetland data and to link the large body of methane fluxes observed in wetland ecosystems to the global distribution of those ecosystems.