A11P-05:
Inverse Analysis of North American Methane Emissions Using the CarbonTracker-Lagrange Modeling Framework
Monday, 15 December 2014: 9:00 AM
Joshua Simon Benmergui1, Arlyn E Andrews2, Kirk W Thoning2, Michael Trudeau3, Anna M Michalak4, Vineet Yadav5, Scot M Miller6, Edward J Dlugokencky7, Lori Bruhwiler8, Kenneth A Masarie7, Doug E.J. Worthy9, Colm Sweeney10, Marc Laurenz Fischer11, Thomas Nehrkorn12, Marikate E Mountain12 and Steven C Wofsy13, (1)Harvard University, School of Engineering and Applied Sciences, Cambridge, MA, United States, (2)NOAA Boulder, Boulder, CO, United States, (3)Cooperative Institute for Research in Environmental Sciences, Boulder, CO, United States, (4)Stanford University, Stanford, CA, United States, (5)Carnegie Institution For Scien, Stanford, CA, United States, (6)Harvard University, Department of Earth and Planetary Sciences, Cambridge, MA, United States, (7)NOAA, Boulder, CO, United States, (8)NOAA/ESRL/GMD, Boulder, CO, United States, (9)Environment Canada Toronto, Climate Research Division, Toronto, ON, Canada, (10)NOAA/Earth System Research Lab, Boulder, CO, United States, (11)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (12)Atmospheric and Environmental Research, Lexington, MA, United States, (13)Harvard University, Cambridge, MA, United States
Abstract:
Several recent studies have attempted to quantify methane (CH4) emissions in North America, but large uncertainties remain in the magnitude, spatial and temporal distribution, and source sectors responsible. Estimates of biogenic emissions from wetlands, emissions related to animal husbandry, and the continental scale implications of recent shifts in the oil and gas industry are poorly constrained. We estimate North American CH4 emissions from biogenic and anthropogenic sources using the new CarbonTracker-Lagrange inverse modeling framework. A wide array of measurements taken between 2007 and 2012 provide top-down constraints. These include: in-situ and flask measurements made at surface sites, tall towers, and aboard aircraft; and remote sensing observations. Source region sensitivity is provided by the Stochastic Time-Inverted Lagrangian Transport (STILT) model, driven by meteorological fields from the Weather Research and Forecasting (WRF) model. CH4 emissions and background values are optimized simultaneously through Bayesian and geostatistical inversion. A restricted maximum likelihood estimation procedure is used to derive uncertainties in both emissions and the mole fraction field. The wide temporal coverage, dense network of measurements, and varied sources of data allow for an analysis of spatio-temporal trends in emissions that has not previously been accomplished with top-down constraints.
