A Seasonal Perspective on Regional Air Quality in CentralCalifornia - Phase 1

Harley, Robert A. ; Brown, Nancy J. ; Tonse, Shaheen R. ; Jin, Ling

COLLABORATION - UCB

关键词: Pacific Ocean;    Particulates;    Air Quality;    Air Pollution;    Photochemistry;   

DOI  :  10.2172/913278
RP-ID  :  LBNL--62438-(1)
RP-ID  :  DE-AC02-05CH11231
RP-ID  :  913278

美国|英语

来源: UNT Digital Library

【 摘 要 】

Central California spans a wide variety of urban, agricultural, and natural terrain, including the San Francisco Bay area, the Central Valley, and the Sierra Nevada Mountains. Population within this region is growing rapidly, and there are persistent, serious air pollution problems including fine particulate matter (PM{sub 2.5}) and ozone. Summertime photochemical air pollution is the focus of the present study, which represents a first phase in the development and application of a modeling capability to assess formation and transport of ozone and its precursors within Central California over an entire summer season. This contrasts with past studies that have examined pollutant dynamics for a few selected high-ozone episodes each lasting 3-5 days. The Community Multiscale Air Quality model (CMAQ) has been applied to predict air pollutant formation and transport in Central California for a 15-day period beginning on July 24, 2000. This period includes a 5-day intensive operating period (July 29 to August 2) from the Central California Ozone Study (CCOS). Day-specific meteorological conditions were modeled by research collaborators at NOAA using a mesoscale meteorological model (MM5). Pollutant emissions within the study domain were based on CARB emission inventory estimates, with additional efforts conducted as part of this research to capture relevant emissions variability including (1) temperature and sunlight-driven changes in biogenic VOC, (2) weekday/weekend and diurnal differences in light-duty (LD) and heavy-duty (HD) motor vehicle emissions, (3) effects of day-specific meteorological conditions on plume rise from point sources such as power plants. We also studied the effects of using cleaner pollutant inflow boundary conditions, lower than indicated during CCOS aircraft flights over the Pacific Ocean, but supported by other surface, ship-based, balloon and aircraft sampling studies along the west coast. Model predictions were compared with measured concentrations for O{sub 3}, NO{sub x}, NO{sub y}, and CO at about 100 ground observation stations within the CCOS domain. Comparisons were made both for time series and for statistically aggregated metrics, to assess model performance over the whole modeling domain and for the individual air basins within the domain. The model tends to over-predict ozone levels along the coast where observed levels are generally low. Inland performance in the San Joaquin Valley is generally better. Model-measurement agreement for night-time ozone is improved by evaluating the sum of predicted O{sub 3} + NO{sub 2} against observations; this removes from the comparison the effect of any ozone titration that may occur. A variety of diagnostic simulations were conducted to investigate the causes for differences between predictions and observations. These included (1) enhanced deposition of O{sub 3} to the ocean, (2) reduced vertical mixing over the ocean, (3) attenuation of sunlight by coastal stratus, (4) the influence of surface albedo on photochemistry, and (5) the effects of observation nudging on wind fields. Use of advanced model probing tools such as process analysis and sensitivity analysis is demonstrated by diagnosing model sensitivity to boundary conditions and to weekday-weekend emission changes.