Inverse Estimation of California Methane Emissions and Their Uncertainties using FLEXPART-WRF

Methane (CH4) has a large global warming potential and mediates global tropospheric chemistry. In California, CH4 emissions estimates derived from "top-down" methods based on atmospheric observations have been found to be greater than expected from "bottom-up" population-apportioned national and state inventories. Differences between bottom-up and top-down estimates suggest that the understanding of California's CH4 sources is incomplete, leading to uncertainty in the application of regulations to mitigate regional CH4 emissions. In this study, we use airborne measurements from the California research at the Nexus of Air Quality and Climate Change (CalNex) campaign in 2010 to estimate CH4 emissions in the South Coast Air Basin (SoCAB), which includes California's largest metropolitan area (Los Angeles), and in the Central Valley, California's main agricultural and livestock management area. Measurements from 12 daytime flights, prior information from national and regional official inventories (e.g. US EPA's National Emission Inventory, the California Air Resources Board inventories, the Liu et al. Hybrid Inventory, and the California Greenhouse Gas Emissions Measurement dataset), and the FLEXPART-WRF transport model are used in our mesoscale Bayesian inverse system. We compare our optimized posterior CH4 inventory to the prior bottom-up inventories in terms of total emissions (Mg CH4/hr) and the spatial distribution of the emissions (0.1 degree), and quantify uncertainties in our posterior estimates. Our inversions show that the oil and natural gas industry (extraction, processing and distribution) is the main source accounting for the gap between top-down and bottom-up inventories over the SoCAB, while dairy farms are the largest CH4 source in the Central Valley. CH4 emissions of dairy farms in the San Joaquin Valley and variations of CH4 emissions in the rice-growing regions of Sacramento Valley are quantified and discussed. We also estimate CO and NH3 surface fluxes and use their observed correlation with CH4 mixing ratio to further evaluate our CH4 total emission estimates, and understand the spatial distribution of CH4 emissions.