A31B-0040
Evaluating Observational Constraints on N2O Emissions via Information Content Analysis Using GEOS-Chem and its Adjoint
Wednesday, 16 December 2015
Poster Hall (Moscone South)
Kelley C Wells1, Dylan B Millet1, Nicolas Bousserez2, Daven K Henze2, Sreelekha Chaliyakunnel1, Timothy J Griffis3, Edward J Dlugokencky4, Ronald G Prinn5, Simon O'Doherty6, Ray F Weiss7, Geoff Dutton4, James W Elkins4, Paul B Krummel8, Ray L Langenfelds9 and Paul Steele10, (1)University of Minnesota Twin Cities, Minneapolis, MN, United States, (2)University of Colorado at Boulder, Boulder, CO, United States, (3)Univ Minnesota, Saint Paul, MN, United States, (4)NOAA Boulder, Boulder, CO, United States, (5)MIT, Cambridge, MA, United States, (6)University of Bristol, Bristol, United Kingdom, (7)Scripps Institution of Oceanography, La Jolla, CA, United States, (8)CSIRO, Aspendale, VIC, Australia, (9)CSIRO Ocean and Atmosphere Flagship, Aspendale, Victoria, Australia, (10)CSIRO, Aspendale, Australia
Abstract:
Nitrous oxide (N2O) is a long-lived greenhouse gas with a global warming potential approximately 300 times that of CO2, and plays a key role in stratospheric ozone depletion. Human perturbation of the nitrogen cycle has led to a rise in atmospheric N2O, but large uncertainties exist in the spatial and temporal distribution of its emissions. Here we employ a 4D-Var inversion framework for N2O based on the GEOS-Chem chemical transport model and its adjoint to derive new constraints on the space-time distribution of global land and ocean N2O fluxes. Based on an ensemble of global surface measurements, we find that emissions are overestimated over Northern Hemisphere land areas and underestimated in the Southern Hemisphere. Assigning these biases to particular land or ocean regions is more difficult given the long lifetime of N2O. To quantitatively evaluate where the current N2O observing network provides local and regional emission constraints, we apply a new, efficient information content analysis technique involving radial basis functions. The technique yields optimal state vector dimensions for N2O source inversions, with model grid cells grouped in space and time according to the resolution that can actually be provided by the network of global observations. We then use these optimal state vectors in an analytical inversion to refine current top-down emission estimates.
