The Ability of Atmospheric Data to Reduce Disagreements in Wetland Methane Flux Estimates over North America

Thursday, 17 December 2015: 13:55
2010 (Moscone West)
Scot M Miller, Carnegie Institution for Science - Stanford, Department of Global Ecology, Stanford, CA, United States, Arlyn E Andrews, NOAA Earth System Research Lab, Boulder, CO, United States, Joshua Simon Benmergui, Harvard University, School of Engineering and Applied Sciences, Cambridge, MA, United States, Roisin Commane, Harvard University, Cambridge, MA, United States, Edward J Dlugokencky, NOAA Boulder, Boulder, CO, United States, Greet Janssens-Maenhout, IES/ Joint Research Centre, Ispra (Varese), Italy, Joe R Melton, CCCma, Victoria, BC, Canada, Anna M Michalak, Carnegie Institution for Science Washington, Washington, DC, United States, Colm Sweeney, NOAA Boulder, ESRL, Boulder, CO, United States and Doug E. J. Worthy, Environment Canada Toronto, Climate Research Division, Toronto, ON, Canada
Existing estimates of methane fluxes from wetlands differ in both magnitude and distribution across North America. We discuss seven different bottom-up methane estimates in the context of atmospheric methane data collected across the US and Canada. In the first component of this study, we explore whether the observation network can even detect a methane pattern from wetlands. We find that the observation network can identify a methane pattern from Canadian wetlands but not reliably from US wetlands. Over Canada, the network can even identify spatial patterns at multi-provence scales. Over the US, by contrast, anthropogenic emissions and modeling errors obscure atmospheric patterns from wetland fluxes. In the second component of the study, we then use these observations to reconcile disagreements in the magnitude, seasonal cycle, and spatial distribution of existing estimates. Most existing estimates predict fluxes that are too large with a seasonal cycle that is too narrow. A model known as LPJ-Bern has a spatial distribution most consistent with atmospheric observations. By contrast, a spatially-constant model outperforms the distribution of most existing flux estimates across Canada. The results presented here provide several pathways to reduce disagreements among existing wetland flux estimates across North America.