1. Estimating anthropogenic methane emissions with GOSAT satellite retrievals and ground-based observations.
- Author
-
Maksyutov, Shamil, Tsuruta, Aki, Janardanan, Rajesh, Wang, Fenjuan, Ito, Akihiko, Sasakawa, Motoki, Machida, Toshinobu, Morino, Isamu, Yoshida, Yukio, Kaiser, Johannes, Janssens-Maenhout, Greet, Dlugokencky, Ed, and Matsunaga, Tsuneo
- Subjects
- *
METHANE , *BIOMASS burning , *PHOTOSYNTHETICALLY active radiation (PAR) , *ARTIFICIAL satellites , *SOLAR radiation management , *SATELLITE DNA - Abstract
To estimate anthropogenic methane emissions localized around large cities, we use theLagrangian particle dispersion model FLEXPART to model local tracer transport at 0.1˚ spatial resolution and compare the model simulations of local methane enhancements toGOSAT observations made in 2009-2015. The satellite-observed methane columnenhancements are aggregated according to discrete simulated enhancement levels. Theobserved and simulated enhancements compare well for the global domain and showdifferences for several large regions such as the US and East Asia. To extend the analysis andaccount for large scale transport and influence of natural fluxes, we perform globalhigh-resolution methaneflux inversion to estimate global methane emissions usingatmospheric methane data collected at global in-situ network, which is archivedat WDCGG, and GOSAT satellite retrievals. FLEXPART is coupled to a globalatmospheric tracer transport model (NIES-TM). Prior fluxes at 0.1˚ resolution wereprepared for anthropogenic emissions (EDGAR 4.3.2), biomass burning (GFAS), andwetlands (VISIT). The inverse model NIES-TM-FLEXPART-VAR (NTFVAR)applies variational optimization to two categories of fluxes: anthropogenic andnatural (wetlands). Bi-weekly emissions are estimated for years 2009 to 2017. Toreduce GOSAT retrieval biases, the monthly mean difference between GOSAT dataand the inversion-optimized forward simulation is estimated for each 5º latitudeband and then it is subtracted from GOSAT retrievals before including them in theinversion. The bias correction is designed to remove large scale biases in GOSATretrievals, while retaining local scale variability that contains most information onanthropogenic emissions. Estimated anthropogenic emissions over large regions (US,China, India) are comparable to GCP-CH4 top-down estimates. The sensitivity of theestimated emissions to prior fluxes is checked by making inverse modeling with prioremissions adjusted to match national reports to UNFCCC for selected large countries. [ABSTRACT FROM AUTHOR]
- Published
- 2019