Search NASA⌕ Search

Engineering topics

Harley, Robert A

Publications and source records attributed to Harley, Robert A.

Meteorology-induced variations in the spatial behavior of summer ozone pollution in Central California

Cluster analysis was applied to daily 8 h ozone maxima modeled for a summer season to characterize meteorology-induced variations in the spatial distribution of ozone. Principal component analysis is employed to form a reduced dimension set to describe and interpret ozone spatial patterns. The first three principal components (PCs) capture {approx}85% of total variance, with PC1 describing a general spatial trend, and PC2 and PC3 each describing a spatial contrast. Six clusters were identified for California's San Joaquin Valley (SJV) with two low, three moderate, and one high-ozone cluster. The moderate ozone clusters are distinguished by elevated ozone levels in different parts of the valley: northern, western, and eastern, respectively. The SJV ozone clusters have stronger coupling with the San Francisco Bay area (SFB) than with the Sacramento Valley (SV). Variations in ozone spatial distributions induced by anthropogenic emission changes are small relative to the overall variations in ozone amomalies observed for the whole summer. Ozone regimes identified here are mostly determined by the direct and indirect meteorological effects. Existing measurement sites are sufficiently representative to capture ozone spatial patterns in the SFB and SV, but the western side of the SJV is under-sampled.

54 ENVIRONMENTAL SCIENCES↗

Seasonal versus Episodic Performance Evaluation for an Eulerian Photochemical Air Quality Model

This study presents detailed evaluation of the seasonal and episodic performance of the Community Multiscale Air Quality (CMAQ) modeling system applied to simulate air quality at a fine grid spacing (4 km horizontal resolution) in central California, where ozone air pollution problems are severe. A rich aerometric database collected during the summer 2000 Central California Ozone Study (CCOS) is used to prepare model inputs and to evaluate meteorological simulations and chemical outputs. We examine both temporal and spatial behaviors of ozone predictions. We highlight synoptically driven high-ozone events (exemplified by the four intensive operating periods (IOPs)) for evaluating both meteorological inputs and chemical outputs (ozone and its precursors) and compare them to the summer average. For most of the summer days, cross-domain normalized gross errors are less than 25% for modeled hourly ozone, and normalized biases are between {+-}15% for both hourly and peak (1 h and 8 h) ozone. The domain-wide aggregated metrics indicate similar performance between the IOPs and the whole summer with respect to predicted ozone and its precursors. Episode-to-episode differences in ozone predictions are more pronounced at a subregional level. The model performs consistently better in the San Joaquin Valley than other air basins, and episodic ozone predictions there are similar to the summer average. Poorer model performance (normalized peak ozone biases <-15% or >15%) is found in the Sacramento Valley and the Bay Area and is most noticeable in episodes that are subject to the largest uncertainties in meteorological fields (wind directions in the Sacramento Valley and timing and strength of onshore flow in the Bay Area) within the boundary layer.

54 ENVIRONMENTAL SCIENCES↗

Sensitivity analysis of ozone formation and transport for a Central California air pollution episode

CMAQ-HDDM is used to determine spatial and temporal variations in ozone limiting reagents and local vs upwind source contributions for an air pollution episode in Central California. We developed a first- and second- order sensitivity analysis approach with the Decoupled Direct Method to examine spatial and temporal variations of ozone-limiting reagents and the importance of local vs upwind emission sources in the San Joaquin Valley of central California for a five-day ozone episode (29th July-3rd Aug, 2000). Despite considerable spatial variations, nitrogen oxides (NO{sub x}) emission reductions are overall more effective than volatile organic compound (VOC) control for attaining the 8-hr ozone standard in this region for this episode, in contrast to the VOC control that works better for attaining the prior 1-hr ozone standard. Inter-basin source contributions of NO{sub x} emissions are limited to the northern part of the SJV, while anthropogenic VOC (AVOC) emissions, especially those emitted at night, influence ozone formation in the SJV further downwind. Among model input parameters studied here, uncertainties in emissions of NO{sub x} and AVOC, and the rate coefficient of the OH + NO{sub 2} termination reaction, have the greatest effect on first-order ozone responses to changes in NO{sub x} emissions. Uncertainties in biogenic VOC emissions only have a modest effect because they are generally not collocated with anthropogenic sources in this region.

54 ENVIRONMENTAL SCIENCES↗

Size-Resolved Particle Number and Volume Emission Factors for On-Road Gasoline and Diesel Motor Vehicles

Average particle number concentrations and size distributions from {approx}61,000 light-duty (LD) vehicles and {approx}2500 medium-duty (MD) and heavy-duty (HD) trucks were measured during the summer of 2006 in a San Francisco Bay area traffic tunnel. One of the traffic bores contained only LD vehicles, and the other contained mixed traffic, allowing pollutants to be apportioned between LD vehicles and diesel trucks. Particle number emission factors (particle diameter D{sub p} > 3 nm) were found to be (3.9 {+-} 1.4) x 10{sup 14} and (3.3 {+-} 1.3) x 10{sup 15} kg{sup -1} fuel burned for LD vehicles and diesel trucks, respectively. Size distribution measurements showed that diesel trucks emitted at least an order of magnitude more particles for all measured sizes (10 < D{sub p} < 290 nm) per unit mass of fuel burned. The relative importance of LD vehicles as a source of particles increased as D{sub p} decreased. Comparing the results from this study to previous measurements at the same site showed that particle number emission factors have decreased for both LD vehicles and diesel trucks since 1997. Integrating size distributions with a volume weighting showed that diesel trucks emitted 28 {+-} 11 times more particles by volume than LD vehicles, consistent with the diesel/gasoline emission factor ratio for PM{sub 2.5} mass measured using gravimetric analysis of Teflon filters, reported in a companion paper.

54 ENVIRONMENTAL SCIENCES↗

Measurement of Black Carbon and Particle Number Emission Factors from Individual Heavy-Duty Trucks

Emission factors for black carbon (BC) and particle number (PN) were measured from 226 individual heavy-duty (HD) diesel-fueled trucks driving through a 1 km-long California highway tunnel in August 2006. Emission factors were based on concurrent increases in BC, PN, and CO{sub 2}B concentrations (measured at 1 Hz) that corresponded to the passage of individual HD trucks. The distributions of BC and PN emission factors from individual HD trucks are skewed, meaning that a large fraction of pollution comes from a small fraction of the in-use vehicle fleet. The highest-emitting 10% of trucks were responsible for {approx} 40% of total BC and PN emissions from all HD trucks. BC emissions were log-normally distributed with a mean emission factor of 1.7 g kg {sup -1} and maximum values of {approx} 10 g kg{sup -1}. Corresponding values for PN emission factors were 4.7 x 10{sup 15} and 4 x 10{sup 16} kg{sup -1}. There was minimal overlap among high-emitters of these two pollutants: only 1 of the 226 HD trucks measured was found to be among the highest 10% for both BC and PN. Monte Carlo resampling of the distribution of BC emission factors observed in this study revealed that uncertainties (1{sigma}) in extrapolating from a random sample of n HD trucks to a population mean emission factor ranged from {+-} 43% for n = 10 to {+-} 8% for n = 300, illustrating the importance of sufficiently large vehicle sample sizes in emissions studies. Studies with low sample sizes are also more easily biased due to misrepresentation of high-emitters. As vehicles become cleaner on average in future years, skewness of the emissions distributions will increase, and thus sample sizes needed to extrapolate reliably from a subset of vehicles to the entire in-use vehicle fleet are expected to become more of a challenge.

54 ENVIRONMENTAL SCIENCES↗

A Process-Analysis Based Study of the Ozone Weekend Effect

We have used the 3D photochemical model CMAQ to simulate the ozone weekend effect, a phenomenon in which urban areas can have higher ozone concentrations on weekends than on weekdays even though NO{sub x} emissions are usually lower on weekends. A simulation containing a weekend is compared to hypothetical simulations in which the anthropogenic emissions for the weekend have been replaced by weekday emissions. The simulations are identical in all other respects. Process analysis is used to explain the results. We find that the weekend effect can be decomposed into an ozone titration component and an odd oxygen component, each contributing about half of the excess weekend ozone. The titration component simply requires that there be lower weekend NO{sub x} emissions. The odd oxygen component additionally requires that on weekends there be a higher rate of OH + (VOC or CO) reactions, brought about by one or more of lower nitric acid formation, higher OH formation from O{sub 3} photolysis, and higher VOC emissions. This causes higher weekend peroxy radical formation. The odd oxygen component also requires sufficiently high NO concentration even on the lower-NO{sub x} weekends to propagate this higher rate of peroxy formation back to higher weekend OH formation.

54 ENVIRONMENTAL SCIENCES↗