A13L-3338:
Assessing the Complementary Constraints on North American Methane Emissions Estimates Provided by Ground-based and Space-based Observations of Methane
Monday, 15 December 2014
Ilya Stanevich1, Dylan B. A. Jones1, Kim Strong1, Wei Lu1, John C Lin2, Arlyn E Andrews3, Doug E.J. Worthy4, Kevin Wecht5, Paul O Wennberg6, Debra Wunch6, Coleen Marie Roehl6 and Manvendra Krishna Dubey7, (1)University of Toronto, Toronto, ON, Canada, (2)University of Utah, Salt Lake City, UT, United States, (3)NOAA Boulder, Boulder, CO, United States, (4)Environment Canada Toronto, Climate Research Division, Toronto, ON, Canada, (5)Harvard University, Cambridge, MA, United States, (6)California Institute of Technology, Pasadena, CA, United States, (7)Los Alamos National Laboratory, Los Alamos, NM, United States
Abstract:
Methane (CH4) is the second most important anthropogenic greenhouse gas. Its global emissions are known to within 15%. However, emissions originating on the regional scale are poorly constrained. Here we present a study focused on assessing the complementarity of the constraints on estimates of North American CH4 emissions provided space-based and ground-based observations of atmospheric CH4. We use data from the Greenhouse gases Observing SATellite (GOSAT) together with discrete air sample measurements from tall towers and ground-based Fourier-Transform InfraRed (FTIR) spectrometers over North America to constrain surface fluxes of CH4. FTIR measurements are made by the Total Carbon Column Observing Network (TCCON) and by the Network for Detection of Atmospheric Composition Change (NDACC), which includes an instrument installed at the University of Toronto Atmospheric Observatory in downtown Toronto. We also employ two inverse modeling approaches. One of them exploits the Stochastic Time-Inverted Lagrangian Transport (STILT) model to link CH4 fluxes with measurements. Updates on emissions are obtained based on simple Bayesian inversion. In the second approach, CH4 emissions are constrained by means of a four-dimensional variational (4D-Var) data assimilation method implemented with the GEOS-Chem global chemical transport model. Comparing the two approaches, we investigate the effect of transport errors and the inversion framework on the inferred CH4 fluxes.
