Atmospheric observations reveal Alaskan wetland CH 4 fluxes are up to threefold higher than process-based model estimates

Abstract. Amplified Arctic warming could substantially increase methane (CH4) fluxes from high-latitudes by accelerating soil thaw and releasing soil organic carbon into the atmosphere. Yet, the magnitude and spatiotemporal patterns of current and future high-latitude CH4 fluxes remain highly uncertain. We estimate growing season (May to September) CH4 fluxes in Alaska from 2012 to 2017 using atmospheric observations from two NASA campaigns: the Carbon in Arctic Reservoirs Vulnerability Experiment (CARVE) and Arctic-Carbon Atmospheric Profiles (Arctic-CAP). We couple these observations with an atmospheric transport model and a geostatistical inverse model to estimate the geographic distribution, seasonality, and inter-annual variability of CH4 fluxes. As part of this approach, we evaluate how well different state-of-the-art, process-based flux models compare to the inverse estimate of CH4 fluxes. We find that estimated statewide fluxes are a factor of 1.3 to 2.9 higher compared to different ensembles of process-based models. This discrepancy is largest across the North Slope of Alaska – up to a factor of 6.5 when compared to one of the model ensembles. We do not find any statistically significant inter-annual variability (IAV) during years 2012–2017, and the IAV in our estimate is generally consistent with process-based models. In addition, our analysis reveals that differences in inundation fraction mapping contribute to the magnitude discrepancies in current process-based models, and future improvements in inundation maps are essential to reduce uncertainties in CH4 flux estimates across high latitudes. These results provide an improved atmospheric constraint on regional CH4 flux estimates from northern wetlands under accelerated Arctic warming.