Estimating Methane Emissions from Central California

A three-month time series of continuous CH4 mixing ratio measurements obtained from a tall-tower are applied in an inverse model to estimate regional surface emissions of CH4 in Central California. Simulated CH4 mixing ratios are calculated based on spatially resolved a priori CH4 emission estimates and simulated atmospheric transport. Atmospheric transport and surface influences (footprints) are computed using the Stochastic Time-Inverted Lagrangian Transport (STILT) model driven by customized output from the Weather Research and Forecasting (WRF) model. An uncertainty analysis is performed that propagates measurement and model errors through the inverse model to provide quantitative uncertainties in the estimated CH4 emissions. WRF-STILT predictions of daytime planetary boundary layer (PBL) heights are found to be over-estimated by a factor of 1.33 ± 0.13 (1s) through comparison with PBL heights estimated from a wind profiler located near the tower site for October, 2007. The inverse model estimates of CH4 emissions are performed in three ways. First, geometric linear regressions of modeled and measured CH4 mixing ratios obtains slopes of 0.95 ± 0.14 and 1.12 ± 0.12 for the October and Oct.-Dec., 2007 periods respectively, suggesting that total a priori CH4 emissions estimates are consistent with the atmospheric measurements and the transport model. Second, a Baysian inverse analysis of the Oct.-Dec., 2007 period obtains posterior scaling factors for CH4 suggesting that landfills and crop agriculuture are lower and livestock and petroleum emissions are higher than a priori estimates respectively. Third, a Baysian analysis of uncertainty, using 13 spatial sub-regions as basis functions, shows that the CH4 measurements significantly reduce posterior uncertainties in CH4 emissions relative the a priori assumptions for only the three sub-regions nearest the tower site. This suggests that a regional network of measurement sites will be necessary to provide high-accuracy retrievals of surface CH4 emissions for multiple regions comprising California's CH4 emissions.