UPDATED PHOTOCHEMICAL MODELING FOR CALIFORNIA SOUTH COAST AIR BASIN -COMPARISON OF CHEMICAL MECHANISMS AND MOTOR-VEHICLE EMISSION INVENTORIES

Large uncertainties remain in photochemical models used to relate emis sions of VOC and NOx to ambient O-3 concentrations. Bias in motor vehi cle emission estimates for VOC has been a long-standing concern. An im proved Eulerian photochemical model is described and applied to the Au gust 27-28, 1987, period in southern California. The chemical mechanis m used in the model is SAPRC93, which predicts peak ozone concentratio ns 22% higher on average than the LCC mechanism used previously. A rev ised motor vehicle emission inventory was developed using gasoline sal es and infrared remote sensing data for CO and measured ambient NMOC/C O and NOx/CO ratios. On-road vehicle emissions for the South Coast Air Basin in summer 1987 were estimated to be (1800 +/- 400) x 10(3) and (710 +/- 160) x 10(3) kg day(-1) for NMOC and NOx, respectively. These values are 2.4 and 1.0 times, respectively, the corresponding current official inventory estimates (MVEI 7G). Ozone concentrations predicte d using the CIT airshed model matched observations more closely when t he revised inventory was used in place of official emission estimates. If the vehicle fleet in 1987 were operating with no emission controls , NMOC and NO, emissions would have been 5900 x 10(3) and 1200 x 10(3) kg day(-1) respectively. On average, peak predicted ozone concentrati ons for the controlled vehicle fleet operating in 1987 were 43% lower than values predicted for the uncontrolled vehicle fleet. The peak pre dicted ozone concentration with the uncontrolled vehicle fleet was 500 ppb.