Constraining black carbon aerosol over Asia using OMI aerosol absorption optical depth and the adjoint of GEOS-Chem

摘要

Accurate estimates of the emissions and distribution of black carbon (BC) inthe region referred to here as Southeastern Asia (70–150° E, 11° S–55° N) are critical tostudies of the atmospheric environment and climate change. Analysis ofmodeled BC concentrations compared to in situ observations indicates levelsare underestimated over most of Southeast Asia when using any of fourdifferent emission inventories. We thus attempt to reduce uncertainties inBC emissions and improve BC model simulations by developing top-down,spatially resolved, estimates of BC emissions through assimilation of OMI (Ozone Monitoring Instrument)observations of aerosol absorption optical depth (AAOD) with the GEOS-Chem (Goddard Earth Observing System – chemistry)model and its adjoint for April and October 2006. Overwhelmingenhancements, up to 500 %, in anthropogenic BC emissions are shown afteroptimization over broad areas of Southeast Asia in April. In October, theoptimization of anthropogenic emissions yields a slight reduction(1–5 %) over India and parts of southern China, whileemissions increase by 10–50 % over eastern China.Observational data from in situ measurements and AERONET (Aerosol Robotic Network) observations areused to evaluate the BC inversions and assess the bias between OMI andAERONET AAOD. Low biases in BC concentrations are improved or corrected inmost eastern and central sites over China after optimization, while theconstrained model still underestimates concentrations in Indian sites inboth April and October, possibly as a consequence of low prior emissions.Model resolution errors may contribute up to a factor of 2.5 to theunderestimation of surface BC concentrations over northern India. We alsocompare the optimized results using different anthropogenic emissioninventories and discuss the sensitivity of top-down constraints onanthropogenic emissions with respect to biomass burning emissions. Inaddition, the impacts of brown carbon, the formulation of the observationoperator, and different a priori constraints on the optimization areinvestigated. Overall, despite these limitations and uncertainties, usingOMI AAOD to constrain BC sources improves model representation of BCdistributions, particularly over China.