Boreal–Arctic Wetland Methane Emissions Modulated by Warming and Vegetation Activity
Methane (CH4) accounts for approximately 20–30% of global greenhouse gas radiative forcing and possesses a global warming potential 28–34 times higher than that of CO2 over a 100-year period. Wetlands stand as the largest and most uncertain natural source of global CH4 emissions, closely intertwined with temperature variations. In significant portions of the Boreal-Arctic region, including northern boreal and tundra ecoregions, recent warming has occurred at a rate three to four times faster than the global average, raising concerns due to the positive feedback loop between CH4 emissions and warming. However, the specific regional response of Boreal-Arctic wetland CH4 emissions to long-term environmental changes remains unknown.
To estimate Boreal-Arctic wetland CH4 emissions and to understand the complex interplay between CH4 emissions and environmental factors, this study compiled eddy covariance and chamber observations and used a causality-guided machine learning model to upscale and analyze wetland CH4 emissions. This work significantly reduced the uncertainty in high-latitude wetland CH4 emissions and provided a robust benchmark for global methane budget analysis and the development of a next-generation methane biogeochemistry model.
This work generated a dataset on Boreal-Arctic wetland CH4 emissions during the recent two decades. It offers valuable insights and opportunities for advancing our understanding of the emissions. With long-term and spatially explicit coverage, the dataset provides a robust foundation for improving bottom-up (BU) and top-down (TD) models. By incorporating widespread observations, particularly during the summer season when significant emission trends are observed, this study also addresses the often-overlooked hysteresis characteristics of wetland CH4 emissions. We identify temperature and gross primary productivity (GPP) as key drivers of CH4 emission trends and variability, highlighting the need to refine temperature sensitivity and plant-mediated emission processes in BU models. Additionally, our dataset facilitates better validation and parameterization of biogeochemical models, enhancing confidence in model predictions. Moreover, it serves as a valuable prior for TD transport inversion models, enabling more accurate differentiation between natural and anthropogenic CH4 emissions' impacts on atmospheric concentrations.