Airborne flux measurements for validation of VOC emission inventories and source attribution

For accurate prediction and modelling of air quality and climate, it is necessary to understand the emissions of volatile organic compounds (VOCs) from the potpourri of sources that they are emitted from: traffic, industry, households, plants, agriculture, etc. In the past, efforts to understand the magnitude and composition of VOC emissions have often relied on indirect methods – either using bottom-up emission models, or inferring emissions top-down from concentration measurements via chemical transport models. Both approaches rely on a number of assumptions regarding chemical reactions and transport - and thus are subject to large uncertainties.Airborne flux observations provide direct emission and deposition information at landscape scale with a resolution of a few km. We performed airborne eddy covariance measurements of a large range of VOCs on board a Twin Otter aircraft in Los Angeles and the agricultural San Joaquin Valley in California using PTR-ToF-MS. Combining these observations with a footprint model, we matched them with gridded inventories in space and time. The comparison with the inventories showed a good representation of typical traffic VOCs, but a significant underestimation of oxygenated VOCs (likely from volatile chemical products and cooking) and terpenoids by the inventories.Using airborne flux footprints in combination with landcover information of the San Joaquin Valley, we disaggregated the observed VOC emissions by multivariate linear regression and attributed them to their sources. This way, we obtained typical VOC emission rates and composition for dairy farms, citrus crops, citrus processing facilities, oak forests, oil and gas wells, and urban areas.