Phase II of this study describes and evaluates state-of-the-science BC/EC and OC emission inventories and provides a framework for creating inventories for California. Global BC and OC emission estimates range from 8 – 24 and 33 – 62 Tg/yr, respectively. North American BC emissions accounted for ~6% of the global total, and California BC emissions accounted for < ~0.4% of global emissions. Global inventories are based on fuel use estimates and emission factors taken from published articles and reports. These emission factors vary regionally and depend on the degree of economic development. They do not represent California’s special mixture of fuels, combustion technology, operating conditions, and aggressive emission controls.
The most accurate inventories use a bottom-up approach where emission factors and activities are specified for all stationary, area, and mobile sources. Examples include the California Air Resources Board (ARB) inventory for criteria pollutants in California and the U.S. National Emissions Inventory (NEI). Because such inventories estimate PM emissions, they provide a basis for estimating BC and OC emissions when the BC and OC PM fractions are measured in specific source types. EC and OC mass fractions are included in source profiles that are used to produce speciated PM inventories and for receptor-oriented source apportionment modeling. ARB PM emission factors are based on emission models such as EMFAC2007 and OFFROAD for mobile sources and the Emission Estimation System (EES) model for biomass burning sources. In this study, EMFAC2007 produced reasonable agreement with recently measured values for heavy-duty diesel-fueled vehicles, but it did not capture the large variability in measured gasoline-fueled vehicle emissions. EES provided reasonable estimates for dry litter burning, but it underestimated PM emissions from wet herb and shrub, regen, and wet needles from Ponderosa and Lodgepole Pine trees. ARB emission factors overestimated Chemise (Chaparral) but underestimated rice straw and grass (Grassland) burning. The EES emission model can be updated with more recent emission factor measurements, provide flexibility for estimating specific fire events, and estimate the uncertainty of the emission factor estimates.
Recently measured source profiles were compiled into a database to supplement the U.S. EPA SPECIATE version 4.0 and ARB source profile libraries. Many of the recent studies lack EC and OC measurements, or they applied EC and OC analysis methods that are not compatible with ambient data. A set of the assembled source profiles was applied to the ARB 2006 PM2.5 emission inventory to estimate BC/EC and OC emissions in California. Total BC/EC emissions were 52,084 tons/yr. Major sources included biomass burning (wildfires, managed burning, and residential fuel combustion), and off-road and on-road engine exhaust. Statewide OC emissions (107,979 tons/yr) were twice BC/EC emissions (52,084 tons/yr). BC/EC emissions derived from the 1995 ARB inventory (38,781 tons/yr) were in reasonable agreement (33,281 tons/yr) with those extracted from California’s grid squares from a 1996 global inventory. However, there were large differences for fuel categories (e.g., fossil fuels and biofuels) and source types, indicating that the overall agreement may have been fortuitous.
California BC/EC and OC emission estimates are sensitive to the choice of source profiles used to convert PM2.5 to BC/EC and OC. Recently measured gasoline- and diesel-fueled vehicle exhaust source profiles from the U.S. EPA’s SPECIATE version 4.0 source profile library resulted in twice the mobile on-road BC/EC emissions compared with the profiles drawn from the ARB source profile library. Using ARB gasoline- and diesel-fueled vehicle source profiles provided 17% lower statewide BC emissions. Source profile documentation in both libraries is limited, making selection of appropriate profiles difficult to justify.
ARB can build on its current PM2.5 emission inventory effort by coupling relevant source profiles containing BC/EC and OC abundances with its inventory system. Existing profiles were assembled into a database and documented, and this can serve as a starting point. Examination of these existing profiles indicates that they are insufficient to represent all of the major source types, especially for biomass burning and non-road engine exhaust. More systematic testing of these emissions, using diluted plumes and a common carbon analysis method, are needed to fill in the gaps.
Further traceability is also needed for emission factors and activity databases, specially those used by the local air districts to construct the emissions they submit to the state inventory. Such data are currently not always available but would enable studies to evaluate the sensitivity of BC/EC and OC emissions to variability and uncertainty in these parameters. California has a wealth of speciated PM2.5 measurements from the long-term IMPROVE network operated in its national parks and wilderness areas and numerous special studies conducted in central California, the San Francisco Bay Area, and the South Coast Air Basin. Estimated PM2.5, EC, and OC emission inventories can be evaluated by comparing measured concentrations with those estimated in air quality models.
For questions regarding this research project, including available data and progress status, contact: Research Division staff at (916) 445-0753
Stay involved, sign up with CARB's Research Email Distribution List