Changes in Motor Vehicle Emissions on Diurnal to Decadal Time

Department of Civil and Environmental Engineering, University of California, Berkeley, .... file quarterly international fuel tax agreement returns wh...
0 downloads 0 Views 171KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 5356-5362

Changes in Motor Vehicle Emissions on Diurnal to Decadal Time Scales and Effects on Atmospheric Composition ROBERT A. HARLEY,* LINSEY C. MARR,† JAIME K. LEHNER, AND SARAH N. GIDDINGS Department of Civil and Environmental Engineering, University of California, Berkeley, California 94720-1710

Emissions from gasoline and diesel engines vary on time scales including diurnal, weekly, and decadal. Temporal patterns differ for these two engine types that are used predominantly for passenger travel and goods movement, respectively. Rapid growth in diesel fuel use and decreasing NOx emission rates from gasoline engines have led to altered emission profiles. During the 1990s, onroad use of diesel fuel grew 3 times faster than gasoline. Over the same time period, the NOx emission rate from gasoline engines in California was reduced by a factor of ∼2, while the NOx emission rate from diesel engines decreased only slightly. Diesel engines therefore grew in both relative and absolute terms as a source of NOx, accounting for about half of all on-road NOx emissions as of 2000. Diesel truck emissions decrease by 60-80% on weekends. Counterintuitive responses to these emission changes are seen in measured concentrations of ozone. In contrast, elemental carbon (EC) concentrations decrease on weekends as expected. Weekly and diurnal patterns in diesel truck activity contribute to variability in the ratio of organic carbon (OC) to EC in primary source emissions, and this could be a source of bias in assessments of the importance of secondary organic aerosol.

Introduction Combustion in gasoline and diesel engines is a major source of air pollution (1). Emissions include nitrogen oxides (NOx), carbon monoxide (CO), volatile organic compounds (VOC), and particulate matter (PM). These emissions give rise to a range of air quality problems and human health concerns (2). VOC and NOx emissions play a central role in the formation of tropospheric ozone. Engine exhaust contributes through gas-to-particle conversion processes to secondary particle formation, in addition to primary PM that is emitted directly. Also tropospheric ozone, particulate matter, and carbon dioxide have all been identified as important contributors to climate change (3). While vehicle travel in the U.S. increased during the 1990s, improved emission control technologies and reformulated fuels led to reductions in emission rates for most of the major pollutants during that period (4, 5). Highway tunnel mea* Corresponding author phone: (510)643-9168; fax: (510)642-7483; e-mail: [email protected]. † Present address: Department of Civil and Environmental Engineering, Virginia Tech, Blacksburg, VA 24061. 5356

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 14, 2005

surements and ambient air pollutant ratios indicate that more progress has been made in reducing CO and VOC than NOx emissions (4, 6). As control of light-duty passenger vehicle emissions has progressed, the importance of diesel engines has increased, especially for NOx where diesel emission control efforts through the late 1990s were unsuccessful (5). In this study we estimate motor vehicle emissions using fuel sales data, multiplied by indices that quantify emissions per unit of fuel burned. We focus on NOx emissions, as there are significant contributions from both light-duty gasoline and heavy-duty diesel vehicles to total emissions, with differing temporal patterns on diurnal to decadal time scales. Heywood (7) notes that NOx emissions from diesel engines should be roughly proportional to the mass of fuel injected: even as the overall air/fuel ratio in diesel engines varies, much of the fuel burns under near-stoichiometric conditions regardless of engine load. Laboratory and on-road measurements indicate that NOx emission rates for both gasoline and diesel engines vary less when they are expressed per unit of fuel burned rather than per km driven (5, 8-10). While decadal and longer-term trends are the focus of many assessments, there are interesting phenomena that occur on shorter time scales. For example, weekly cycles in emissions provide an opportunity to test understanding of atmospheric responses to changes in anthropogenic forcing. Weekly cycles in surface-level ozone concentrations were recognized more than 30 years ago (11). Marr and Harley (12) documented the spread of weekday-weekend differences in ozone to an increasing proportion of measurement sites in California. Despite generally lower levels of NOx and VOC on weekends, ozone levels were found to increase or remain unchanged on weekends compared to weekdays. Data from the GOME satellite in sun-synchronous orbit show a continental-scale decrease in NO2 concentrations on Sundays in North America and Europe (13). Decreases in particulate elemental/black carbon on weekends also have been noted (14-16). Forster and Solomon (17) identified weekly cycles in diurnal temperature range; they speculate that this short time-scale anthropogenic influence on climate may be due to indirect effects of aerosols on clouds. In this paper, we describe temporal patterns in emissions from on-road gasoline and diesel engines on diurnal, weekly, and decadal time scales using data from California. We then discuss changes in emissions as they relate to concentrations of primary and secondary pollutants in the atmosphere.

Methods Activity Data. Taxable fuel sales are used to measure onroad vehicle activity (18, 19). Nationally, about one-third of diesel fuel use occurs in off-road engines such as farm tractors, construction and mining equipment, railroad locomotives, etc. (20). Diesel fuel sold for off-road use is dyed red, and its use in on-road engines is prohibited; taxable fuel sales exclude most of the off-road use. Use of gasoline in agriculture, aviation, marine engines, etc. accounts for a small fraction of total fuel usage and is excluded from estimates presented here. A potential problem with using taxable fuel sales is the import and export of diesel fuel in the tanks of trucks that cross state lines. Note however, that these imports and exports are accounted for in taxable fuel sales reports: interstate truckers are required to file quarterly international fuel tax agreement returns which result in taxes being paid based on where fuel is used rather than where it is purchased. To describe diurnal and day-of-week patterns in vehicle emissions, we use weigh-in-motion data from rural and urban 10.1021/es048172+ CCC: $30.25

 2005 American Chemical Society Published on Web 06/09/2005

TABLE 1. On-Road Measurements of Motor Vehicle NOx Emissions nearest city

year sampled

site description

EIa (g kg-1)

ref

California Light-Duty Vehicles Oakland Sacramento Sunnyvale Los Angeles Riverside Los Angeles

hwy 24E Caldecott tunnel hwy 50E ramp to Sunrise I-280N ramp to I-880N I-710N ramp to hwy 91W hwy 91 ramp to 60W SB La Brea Blvd to I-10E averagec

Baltimore, MD Tuscarora, PA Oakland, CA Durham, NC Anaheim, CA

I-95E Fort McHenry tunnel I-76E Tuscarora tunnel hwy 24E Caldecott tunnel I-40 near RTP weigh stn ramp to hwy 91 averagec

1999, 2001 1999 1999 1999 2000 1999, 2001

6.1 4.5b 6.4b 9.0b 8.7 9.3 7.0 ( 2.2

(25) (26) (26) (26) (27) (28)

37 38 42 45 32 39 ( 6

( 8) ( 8) (30) (31) (32)

Heavy-Duty Vehicles 1992 1992 1997 1997 1997

a Emission index in grams of NO (as NO ) emitted per kg of fuel burned. b Listed value applies for 1999; decremented by 0.7 g/kg to account x 2 for 1 year of fleet turnover between 1999 and 2000 (see Figure 1). c Uncertainty estimate provides a 95% confidence interval for the mean.

traffic count sites located in Los Angeles and the San Francisco Bay area and the Central Valley of California (21, 22). Separate traffic counts are available for light-duty vehicles and heavyduty trucks. Patterns of gasoline engine activity are dominated by urban use, whereas diesel engine activity is split more evenly between urban and intercity driving. Based on estimates of population and gasoline and diesel fuel consumption by county (23, 24), we describe 80 and 60% of statewide gasoline and diesel fuel consumption using urban profiles. The balance of fuel use is assigned to hours and days of the week using traffic count data from rural sites. Only annual fuel sales data are required for this analysis; apportionment of total fuel use to finer time scales is done using traffic count data. Emission Factors. To characterize light-duty vehicle NOx emissions, we use on-road measurements from a highway tunnel and freeway ramps. Each site used here encompasses samples ranging from ∼10 000 to 38 000 vehicles. At the entrance and exit of the highway tunnel, online gas analyzers were used to measure NOx, CO, and CO2 concentrations (25); heavy-duty trucks are not allowed in the middle bore of the tunnel where vehicle emissions were measured. At five California freeway ramp sites listed in Table 1 (26-28), infrared and ultraviolet light beams projected at tailpipe height across a single lane of traffic were used to measure pollutant ratios in the exhaust plumes of individual vehicles as they drove by the sensor (29). Emission indices per unit of fuel burned were calculated by carbon balance using measured pollutant ratios. Measurements of diesel vehicle emissions were excluded based on license plate data. Heavyduty diesel vehicle emissions were characterized separately using measurements made in highway tunnels (8, 30) and remote sensing studies (31, 32). Remote sensing capabilities for NOx were developed more recently than CO and hydrocarbons, and questions can be raised about the accuracy of NOx measurements. Ultraviolet remote sensing (29) is thought to be more reliable for NOx than infrared remote sensing is for hydrocarbons (33). This is attributed to a less complicated measurement problem (a well-defined nitric oxide target molecule versus numerous hydrocarbon molecules) and superior signal-to-noise ratios in the ultraviolet vs infrared (33). A small negative bias may arise though, as NO2 is not currently measured as part of NOx in UV-based remote sensors.

Results and Discussion NOx Emission Factors. Measured NOx emission rates are summarized in Table 1 for gasoline and diesel engines. Higher

gasoline engine NOx emission rates were observed at southern California sites compared to the other on-road locations listed in Table 1. The 3 sites with higher NOx emissions also had older average vehicle ages of 7.9-8.8 years vs 7.3-7.4 years for the San Francisco Bay area and Sacramento sites where measured NOx emissions were lower. Differences in annual mileage accrual and the mix of vehicles may augment the effect of site-to-site differences in average vehicle age. Also of note in Table 1 is the diesel NOx emission rate, which is over 5 times the corresponding value for gasoline engines. Chassis dynamometer test data (5) indicate an average heavy-duty diesel NOx emission rate of 34 g kg-1 for vehicles of model years 1975-1997 and show no downward trend over time, so the same NOx emission rate was used here for diesel in both 1990 and 2000. Additional chassis dynamometer test results have been reported recently (34) for California heavy-duty diesel trucks. These data indicate a steady average NOx emission rates of 32 g kg-1 for engines from 1998 and earlier years; the emission index is lower at 21 g kg-1 for 12 more recent engines (1999-2003 model years). Therefore the diesel NOx emission factor shown in Table 1 may be too high for calendar year 2000, as it does not account for the effect of cleaner 1999 and 2000 model year engines being on the road. To estimate NOx emission rates from gasoline engines as of 1990, measurements from the Caldecott tunnel (4, 25) were back-projected as shown in Figure 1. The light-duty vehicle NOx emission rate in 1990 was thereby estimated to be 2.2 ( 0.3 times the level observed in 2000; this scaling was applied to the average gasoline engine NOx emission rate for 2000 shown in Table 1. Also shown in Figure 1 are trends in light-duty vehicle stabilized NOx emission rates estimated using California’s EMFAC model (35). Note the trend line is fit only using measured tunnel data; EMFAC predictions are in good agreement. A change in slope of the EMFAC model predictions can be noted in Figure 1, following the introduction of reformulated gasoline between 1995 and 1996. Though on-road measurements of vehicle emissions in California ca. 1990 do exist, remote sensors at that time did not measure NOx. Results from a 1987 tunnel study conducted in Van Nuys, CA (Los Angeles area) yield an estimated NOx emission rate of ∼11 g kg-1. This was calculated from results for high-speed traffic discussed by Pierson et al. (36), assuming fuel consumption of 12 L per 100 km (equivalent to 20 mpg) and ignoring the presence of diesel trucks. The Van Nuys tunnel result is at the lower end of the range estimated in the present study for gasoline engine NOx emissions as of 1990. CO2 concentrations were not measured; VOL. 39, NO. 14, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5357

FIGURE 1. Trends in California light-duty vehicle NOx emission factors measured at the Caldecott tunnel between 1994 and 2001, with linear extrapolation back to 1990. EMFAC model estimates for stabilized emissions from light-duty vehicles are shown for comparison. The best-fit line was constructed using measured tunnel data only. that information would have permitted more definitive comparisons with the present study. Further measurements made in the Los Angeles area (Sepulveda tunnel) in 1995 and 1996 (37) indicate average NOx emission rates of 8.1 and 6.6 g kg-1, respectively, for driving on a level roadway. The Sepulveda results are lower by 15 and 27%, respectively, than corresponding tunnel emission factors shown in Figure 1, which are for uphill traffic on a 4% grade. In contrast, the remote sensing data in Table 1 suggest NOx emission rates from light-duty vehicles in southern California are higher than elsewhere in the state. However, site-to-site variability in emissions can result from differences in vehicles sampled, roadway grade, vehicle speed and acceleration, ambient temperature and humidity, and air conditioner use. In summary, the rate of light-duty vehicle NOx reduction observed at the Caldecott tunnel may not represent what occurred at other locations, and additional uncertainty arises in extrapolating 4 years prior to the earliest tunnel measurements in 1994. The 1990 NOx emission rate from gasoline engines is less certain than the estimate for calendar year 2000. Emission Inventory. NOx emission rates are combined with taxable fuel sales to estimate emissions as shown in Table 2. Between 1990 and 2000, on-road use of gasoline in California increased by 12%, whereas diesel fuel use increased by 43%. While NOx emissions from light-duty vehicles decreased by a factor of ∼2, this decrease was partly offset by increased emissions from heavy-duty vehicles. The contribution of diesel engines increased from ∼30% of total on-road vehicle NOx emissions in 1990 to over half in 2000. Emission estimates of the current study are compared with predictions from California’s EMFAC model (35) in Table 2. Excess emissions associated with vehicle starting are listed separately, as on-road sampling locations are unlikely to capture vehicles operating in cold start mode. Within the ranges of uncertainty stated in Table 2, EMFAC and fuelbased NOx estimates agree for gasoline engines in both 1990 and 2000 and for diesel engines in 1990. However, EMFAC predicts an increase of 3% in diesel NOx emissions between 1990 and 2000, in contrast to the 43% increase in NOx emissions estimated here over the same time period. There are two offsetting changes predicted by EMFAC for heavyduty diesel engines between 1990 and 2000: a 29% increase in diesel engine activity and a 20% decrease in the NOx emission rate, of which half is attributed to the introduction of reformulated diesel fuel and the rest is due to fleet turnover. Differences between certification and in-use NOx emission 5358

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 14, 2005

rates from new diesel engines sold during the 1990s are already mostly accounted for in the EMFAC model. Figure 2 shows weekly and diurnal patterns of on-road vehicle NOx emissions as of 1990. Emissions were estimated separately for gasoline and diesel engines, as the product of fuel consumption, emission indices, and temporal allocation factors derived from weigh-in-motion data. Clearly apparent in Figure 2 is a large decrease in diesel NOx emissions on weekends: emissions on Saturday and Sunday are lower by 63 and 76%, respectively, compared to the weekday average. On weekdays, diesel engine emissions peak during the middle of the day, in contrast to gasoline engine emissions that show a bimodal distribution linked to commuter peak traffic periods. The total amount of NOx emitted from gasoline engines on weekends is not much different than on weekdays; however, the timing of those emissions is shifted with a single broad peak in the afternoon. The combined total NOx emitted on-road by gasoline and diesel engines is shown in Figure 3 by day and time for both 1990 and 2000. On weekdays as shown in Figure 3, NOx emissions from on-road vehicles decreased by ∼15% between 1990 and 2000, in contrast to weekends where a larger reduction in NOx of 30-35% is apparent. The reason for the different decadal trends on weekdays vs weekends is that increased diesel emissions partially offset reductions in gasoline engine emissions on weekdays. In contrast, on weekends diesel truck emissions are much less important (see Figure 2), so reductions in gasoline engine emissions become the dominant factor. The overall amplitude in the weekly cycle of on-road vehicle NOx emissions grew from a 27% weekend decrease in 1990 to a 43% decrease on weekends in 2000. Though this work focuses on on-road vehicle emissions, variations in other sources of air pollution should not be ignored. Chinkin et al. (38) present field survey data showing variations by day of week in numerous sources of air pollution. The same authors also provide a detailed report of weekly cycles in traffic patterns in southern California. Outside of California, standards for control of gasoline engine NOx emissions have not been as stringent, and larger contributions to total NOx emissions are expected from coal combustion at power plants. Where the role of diesel engines as a source of NOx is less prominent, weekly cycles in air pollution may be less apparent. Effects on Ozone. In Figure 4, the fraction of total sitedays in the Los Angeles area with observed 8-h daily maximum ozone above 85 ppb is plotted by day of week for 1990, 1994, 1998, and 2002. While the frequency and spatial extent of high ozone has decreased overall since 1990, the decrease is not as rapid on weekends as on weekdays, with the result that high ozone was observed disproportionately on Saturdays and Sundays especially in the later years shown in Figure 4. There was rapid progress in reducing ozone in southern California between 1994 and 1998. This coincides with the introduction of reformulated gasoline which, acting together with more durable and robust emission control technologies on new vehicles sold during the 1990s, led to reductions in the mass and reactivity of emissions from gasoline engines (4, 39). At one southern California measurement site (Crestline), high-ozone days continue to be observed fairly uniformly throughout the week. Some of the Crestline exceedances are isolated events (i.e., high ozone is not observed at other sites on the same day). EPA designates entire air basins as nonattainment if even one measurement site exceeds the applicable standard. This can be appropriate from a control perspective, if region-wide strategies are needed to solve the problem. Figure 4 reflects not only the number of days but also the spatial extent of high ozone. Therefore, the figure is not unduly influenced by data from the Crestline mea-

TABLE 2. On-Road Vehicle Fuel Consumption and NOx Emissions in California gasoline

on-road fuel use NOx emission indexa NOx emissionsa,b (this study) EMFAC modelb,c running emissions start/idle emissions

109

yr-1

L g L-1 g kg-1 103 kg d-1 103 kg d-1 103 kg d-1

2000

1990

2000

48 11 ( 4 15 ( 5 1400 ( 500 1448 164

54 5.2 ( 1.6 7.0 ( 2.2 770 ( 240 850 146

7.0 33 ( 5 39 ( 6 630 ( 100 665 19

10. 33 ( 5 39 ( 6 900 ( 140 684 24

a Stated range of uncertainty provides a 95% confidence interval for the mean were emitted as NO2. c EMFAC model estimates (35).

FIGURE 2. Temporal pattern of statewide total NOx emissions from California on-road vehicles as of 1990. Gasoline and diesel engine contributions to the total are shown separately, plotted against a zero baseline in each case. Other weekdays (not shown) are similar to Monday.

FIGURE 3. Change in statewide NOx emissions from California onroad vehicles between 1990 (upper trace) and 2000 (lower trace). Values shown are the sum of gasoline and diesel engine emissions. surement site and is better linked to the potential for population exposure than a single basin-wide attainment determination for each day. Marr and Harley (12) documented the spread of weekdayweekend differences in ozone to an increasing proportion of ground-based measurement sites in California. The measurement sites are generally located in populated areas, as the main purpose is to determine compliance with healthbased air quality standards. From 1980 to 1984, elevated levels of ozone on weekends were observed in major coastal cities and affected only 11% of all monitoring sites operated statewide, whereas from 1995 to 1999 a pattern of higher ozone on weekends was observed at 38% of sites including many inland locations. The spread of higher ozone on weekends is observed despite the fact that weekend NOx

diesel

1990

b

Statewide annual average values; mass reported as if all NOx

FIGURE 4. Fraction of ambient air monitoring site-days in southern California (South Coast air basin) with maximum 8-h average ozone above 85 ppb, shown by day of week for 1990, 1994, 1998, and 2002. concentrations are lower at nearly every measurement site throughout California during the entire period 1980-2000. Various hypotheses have been advanced to explain weekend increases in ozone (40). An important test of the hypotheses is consistency with the spreading of weekdayweekend ozone differences to more observation sites over time. We have described the increasingly important role of diesel trucks as a source of NOx emissions (Table 2) and the increasing amplitude of the weekly cycle in vehicular NOx emissions (Figure 3). During the 1990s, the VOC/NOx emission ratio decreased because VOC emissions from gasoline engines were reduced more than NOx (4, 25), and because diesel NOx emissions increased. Therefore a NOx reduction hypothesis (i.e., that small to moderate reductions in NOx can lead to increased ozone under VOC-sensitive conditions) is consistent with the observed spreading of weekday-weekend ozone differences, and the long-term emission trends described above. Other hypotheses for weekday-weekend ozone differences such as changes in emissions timing, decreased soot leading to higher photolysis rates, and carryover of pollutants from previous days should be assessed in a similar manner. Even if reductions in NOx emissions on weekends contribute to increased ozone at some locations, this does not imply that larger reductions in NOx will be counterproductive for ozone control in the long term. Reynolds et al. (41, 42) found that large reductions in NOx were needed to attain 8-h average ozone air quality standards: 70-90% reduction relative to 1999 baseline in central California and 46-86% reduction relative to 1996 baseline in the eastern United States. VOC reductions alone were found to have only a “modest impact” on ozone in these regions. Other air pollutants of concern such as NO2 decrease in response to reduced NOx emissions (13, 43). Diesel exhaust also contributes to particulate matter concentrations, for which weekly cycles are discussed below. Effects on Particulate Matter. Kim et al. (44, 45) report 24-h average measurements of PM2.5 and PM10 mass and VOL. 39, NO. 14, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5359

FIGURE 6. Daily OC/EC ratios during summer 1995 (average of 24-h filter data from 5 southern California urban sites). OC values were not adjusted to account for oxygen and nitrogen mass. All of the weekly OC/EC peaks occurred on Sundays, except Monday September 4 (Labor Day holiday). Labeled dates along the x-axis are all Mondays.

FIGURE 5. Ratios by day of week of fine particulate organic carbon, elemental carbon, nitrate, and sulfate to seasonal average concentrations for (a) summer and (b) fall 1995. Means and 95% confidence intervals were calculated using data from 5 southern California sites (44, 45). chemical composition using a filter-based sampler developed previously (46). Known sampling artifacts for OC and nitrate were addressed through use of denuders and backup filters. See Kim et al. for discussion of absolute PM2.5 concentrations, spatial and seasonal variations, and the relative importance of various constituents. Here we analyze day-of-week patterns for summer (best data coverage) and fall (highest PM2.5 levels). Samples were collected daily for 8 weeks from July 10 to September 6, 1995. In the fall (October 8 to December 11, 1995), samples were collected once every third day, except for daily samples collected for 2 weeks starting October 8 and 1 week starting November 14. From summer to fall, average concentrations increased from 6.7 to 8.0 µg C m-3 for organic carbon, from 3.4 to 5.9 µg C m-3 for elemental carbon, from 5.5 to 28 µg m-3 for nitrate, and from 4.4 to 6.1 µg m-3 for sulfate. The increase in nitrate was much larger than observed for other PM2.5 constituents. Figure 5 presents ratios of major chemical constituents of PM2.5 to seasonal average values by day of week. For each site and chemical constituent, average concentrations were calculated for summer and fall seasons, and individual daily values were then normalized to seasonal averages. The ratios from five measurement sites (Anaheim, Downtown Los Angeles, Diamond Bar, Fontana, and Rubidoux) were averaged for each day, and then a mean and 95% confidence interval was calculated by day of week as shown in Figure 5. The normalization used here removes the effect of systematic differences in absolute concentrations from site to site. For example, summertime nitrate levels at Rubidoux were much higher than at the other sites. Variations in EC concentrations by day of week are apparent in Figure 5. EC decreases by 10 and 40% below weekly average values on Saturdays and Sundays (only the 5360

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 14, 2005

decrease on Sundays during summer is statistically significant). As the EC decrease is not as large as the 63-76% decrease in diesel truck traffic on weekends, other sources such as gasoline engines and wood burning must contribute to EC concentrations. Figure 5 indicates a gradual buildup of EC through the week, with highest values observed on Fridays. This buildup may be due in part to recirculation of emissions from previous days, not just changes in current day emissions. For the other chemical constituents shown in Figure 5, there are no statistically significant weekly cycles in the data set examined here. Weekly signals that may exist in these pollutants appear to be small compared to variations in EC. Despite the weekend decrease in NOx emissions shown in Figures 2 and 3, there was little change in fine particulate nitrate observed on weekends. This is consistent with findings of Blanchard and Tanenbaum (47), who studied other data sets and other seasons in addition to summer yet found no statistically significant weekly cycle in nitrate levels. Other investigators (16) report a 6% decrease in nitrate concentrations on weekends. Further investigation of this issue is needed, and we recommend examination of newer data sets that provide high time resolution. Common time series analysis methods require at least two samples per cycle of the highest frequency signal being studied (Nyquist sampling criterion). Much of the past record of speciated PM concentration measurements is based on less frequent sampling schedules (e.g., once every sixth day) or on intensive sampling studies of short overall duration. These types of ambient concentration data are not well-suited to the study of weekly cycles. Having two data points per week does not guarantee success in identifying weekly cycles. In the present study the wider confidence intervals shown for the fall season data (see Figure 5) are due in part to having many weeks with once every third day data instead of daily PM2.5 measurements. The episodic nature of fine particulate matter concentrations in the fall also makes it harder to detect weekly signals in the data. Secondary Organic Aerosol. A widely used ratio technique to determine the contribution of secondary organic aerosol (SOA) to total OC concentrations is reviewed by Seinfeld and Pandis (48). The analysis assumes a stable OC/EC ratio in primary source emissions and then attributes increases observed in ambient OC/EC to SOA. While SOA formation does indeed increase OC/EC, there are other sources of variance that complicate determination of a suitable primary emissions ratio (48, 49). As shown in Figure 6 using the PM2.5 data from southern California described above, a strong weekly cycle in ambient OC/EC is observed. The weekend

peaks in Figure 6 are due to decreases in EC, not increases in OC concentrations. The significance of Figure 6 on weekends, and probably within a diurnal cycle on weekdays as well, is that OC/EC variations in primary emissions are large. We therefore recommend caution in attribution of increases in OC/EC observed on weekends and during weekday afternoon rush hour periods when diesel truck traffic and accompanying EC emissions decline precipitously. Though difficult to achieve, a better method to assess SOA contributions to OC would quantify known VOC oxidation products in the particle phase and compute ratios of these SOA markers to total OC rather than EC concentrations.

Acknowledgments The authors thank Joe Avis of the California Department of Transportation for supplying weigh-in-motion traffic count data, Bong-Mann Kim of the South Coast Air Quality Management District for providing PM2.5 data, and Amir Fanai of the Bay Area Air Quality Management District for assistance with EMFAC. This research was supported by the U.S. EPA through a STAR graduate research fellowship (to L.C. Marr).

Literature Cited (1) Sawyer, R. F.; Harley, R. A.; Cadle, S. H.; Norbeck, J. M.; Slott, R.; Bravo, H. A. Mobile sources critical review. Atmos. Environ. 2000, 34, 2161-2181. (2) Lloyd, A. C.; Cackette, T. A. Diesel engines: environmental impact and control. J. Air Waste Manage. Assoc. 2001, 51, 809847. (3) Houghton, J. T.; Ding, Y.; Griggs, D. J.; Noguer, M.; van der Linden, P. J.; Dai, X.; Maskell, K.; Johnson, C. A. Climate Change 2001: The Scientific Basis; Cambridge University Press: Cambridge, 2001; p 37. (4) Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Traverse, M. Impact of California reformulated gasoline on motor vehicle emissions. 1. Mass emission rates. Environ. Sci. Technol. 1999, 33, 318-328. (5) Yanowitz, J.; McCormick, R. L.; Graboski, M. S. In-use emissions from heavy-duty diesel vehicles. Environ. Sci. Technol. 2000, 34, 729-740. (6) Parrish, D. D.; Trainer, M.; Hereid, D.; Williams, E. J.; Olszyna, K. J.; Harley, R. A.; Meagher, J. F.; Fehsenfeld, F. C. Decadal change in carbon monoxide to nitrogen oxide ratio in U.S. vehicular emissions. J. Geophys. Res. 2002, 107(D12), 4140, doi: 10.1029/2001JD000720. (7) Heywood, J. B. Internal Combustion Engine Fundamentals; McGraw-Hill: New York, 1988; p 589. (8) Pierson, W. R.; Gertler, A. W.; Robinson, N. F.; Sagebiel, J. C.; Zielinska, B.; Bishop, G. A.; Stedman, D. H.; Zweidinger, R. B.; Ray, W. D. Real-world automotive emissionsssummary of studies in the Fort McHenry and Tuscarora mountain tunnels. Atmos. Environ. 1996, 30, 2233-2256. (9) Kean, A. J.; Harley, R. A.; Kendall, G. R. Effects of vehicle speed and engine load on motor vehicle emissions. Environ. Sci. Technol. 2003, 37, 3739-3746. (10) Gajendran, P.; Clark, N. N. Effect of truck operating weight on heavy-duty diesel emissions. Environ. Sci. Technol. 2003, 37, 4309-4317. (11) Cleveland, W. S.; Graedel, T. E.; Kleiner, B.; Warner, J. L. Sunday and workday variations in photochemical air pollutants in New Jersey and New York. Science 1974, 86, 1037-1038. (12) Marr, L. C.; Harley, R. A. Spectral analysis of weekday-weekend differences in ambient ozone, nitrogen oxide, and non-methane hydrocarbon time series in California. Atmos. Environ. 2002, 36, 2327-2335. (13) Beirle, S.; Platt, U.; Wenig, M.; Wagner, T. Weekly cycle of NO2 by GOME measurements: a signature of anthropogenic sources. Atmos. Chem. Phys. 2003, 3, 3451-3467. (14) Allen, G. A.; Lawrence, J.; Koutrakis, P. Field validation of a semi-continuous method for aerosol black carbon (aethalometer) and temporal patterns of summertime hourly black carbon measurements in southwest PA. Atmos. Environ. 1999, 33, 817-823. (15) Hies, T.; Treffeisen, R.; Sebald, L.; Reimer, E. Spectral analysis of air pollutants. Part 1: Elemental carbon time series. Atmos. Environ. 2000, 34, 3495-3502.

(16) Motallebi, N.; Tran, H.; Croes, B. E.; Larsen, L. C. Day-of-week patterns of particulate matter and its chemical components at selected sites in California. J. Air Waste Manage. Assoc. 2003, 53, 876-888. (17) Forster, P. M.; Solomon, S. Observations of a “Weekend Effect” in diurnal temperature range. Proc. Natl. Acad. Sci. 2003, 100, 11225-11230. (18) Highway Statistics Summary to 1995. Office of Highway Policy Information, Federal Highway Administration, U.S. Department of Transportation, Washington, DC. (19) Highway Statistics 2000. Office of Highway Policy Information, Federal Highway Administration, U.S. Department of Transportation, Washington, DC. (20) Kean, A. J.; Sawyer, R. F.; Harley, R. A. A fuel-based assessment of off-road diesel engine emissions. J. Air Waste Manage. Assoc. 2000, 50, 1929-1939. (21) Dreher, D. B.; Harley, R. A. A fuel-based inventory for heavyduty diesel truck emissions. J. Air Waste Manage. Assoc. 1998, 48, 352-358. (22) Marr, L. C.; Black, D. R.; Harley, R. A. Formation of photochemical air pollution in central California - 1. Development of a revised motor vehicle emission inventory. J. Geophys. Res. 2002, 107(D6), 4047, doi: 10.1029/2001JD000689. (23) California Motor Vehicle Stock, Travel and Fuel Forecast. Division of Transportation System Information, California Department of Transportation, Sacramento, CA, November 2003. (24) Truck Kilometers of Travel on the California State Highway System, 1986-2001. Division of Transportation System Information, California Department of Transportation, Sacramento, CA, June 2003. (25) Kean, A. J.; Sawyer, R. F.; Harley, R. A.; Kendall, G. R. Trends in exhaust emissions from in-use California light-duty vehicles, 1994-2001. SAE Tech. Pap. Ser. 2002, paper no. 2002-01-1713. (26) Stedman, D. H. On-road motor vehicle emissions monitoring in California; Report to the Steven and Michele Kirsch Foundation; Department of Chemistry and Biochemistry, University of Denver, Denver, CO, February 2000. (27) Pokharel, S. S.; Bishop, G. A.; Stedman, D. H. On-road remote sensing of automobile emissions in the Los Angeles area: year 2; Report to the Coordinating Research Council under contract E-23-4; Department of Chemistry and Biochemistry, University of Denver, Denver, CO, March 2001. (28) Williams, M. J.; Bishop, G. A.; Stedman, D. H. On-road remote sensing of automobile emissions in the La Brea area: year 2; Report to the Coordinating Research Council under contract E-23-4; Department of Chemistry and Biochemistry, University of Denver, Denver, CO, February 2003. (29) Popp, P. J.; Bishop, G. A.; Stedman, D. H. Development of a high-speed ultraviolet spectrometer for remote sensing of mobile source nitric oxide emissions. J. Air Waste Manage. Assoc. 1999, 49, 1463-1468. (30) Kirchstetter, T. W.; Harley, R. A.; Kreisberg, N. M.; Stolzenburg, M. R.; Hering, S. V. On-road measurement of fine particle and nitrogen oxide emissions from light- and heavy-duty motor vehicles. Atmos. Environ. 1999, 33, 2955-2968. (31) Jime´nez, J. L.; McRae, G. J.; Nelson, D. D.; Zahniser, M. S.; Kolb, C. E. Remote sensing of NO and NO2 emissions from heavyduty diesel trucks using tunable diode lasers. Environ. Sci. Technol. 2000, 34, 2380-2387. (32) Bishop, G. A.; Morris, J. A.; Stedman, D. H.; Cohen, L. H.; Countess, R. J.; Countess, S. J.; Maly, P.; Scherer, S. The effects of altitude on heavy-duty diesel truck on-road emissions. Environ. Sci. Technol. 2001, 35, 1574-1578. (33) Bishop, G. A. Department of Chemistry and Biochemistry, University of Denver, Denver, CO, personal communication, April 2005. (34) Clark, N. N.; Gautam, M. California heavy-duty diesel truck emissions characterization; Report to the Coordinating Research Council under contract E-55/E-59; Department of Mechanical and Aerospace Engineering, West Virginia University, Morgantown, WV, August 2004. (35) California Motor Vehicle Emission Factor/Emission Inventory Model, EMFAC 2002 version 2.2. California Air Resources Board, Sacramento, CA, 2002. http://www.arb.ca.gov/msei/ on-road/latest_version.htm. (36) Pierson, W. R.; Gertler, A. W.; Bradow, R. L. Comparison of SCAQS tunnel study with other on-road vehicle emission data. J. Air Waste Manage. Assoc. 1990, 40, 1495-1504. (37) Gertler, A. W.; Sagebiel, J. C.; Dippel, W. A.; O’Connor, M. The impact of California phase 2 reformulated gasoline on realworld vehicle emissions. J. Air Waste Manage. Assoc. 1999, 49, 1339-1346. VOL. 39, NO. 14, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5361

(38) Chinkin, L. R.; Coe, D. L.; Funk, T. H.; Hafner, H. R.; Roberts, P. T.; Ryan, P. A.; Lawson, D. R. Weekday versus weekend activity patterns for ozone precursor emissions in California’s south coast air basin. J. Air Waste Manage. Assoc. 2003, 53, 829-843. (39) Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Hesson, J. M. Impact of California reformulated gasoline on motor vehicle emissions. 2. Volatile organic compound speciation and reactivity. Environ. Sci. Technol. 1999, 33, 329-336.

(45)

(46)

(40) Croes, B. E.; Dolislager, L. J.; Larsen, L. C.; Pitts, J. N. The O3 “Weekend Effect” and NOx Control Strategies. Environ. Manager 2003, 27-35.

(47)

(41) Reynolds, S. D.; Blanchard, C. L.; Ziman, S. D. Understanding the effectiveness of precursor reductions in lowering 8-hr ozone concentrations. J. Air Waste Manage. Assoc. 2003, 53, 195-205.

(48)

(42) Reynolds, S. D.; Blanchard, C. L.; Ziman, S. D. Understanding the effectiveness of precursor reductions in lowering 8-hr ozone concentrationssPart II. The Eastern United States. J. Air Waste Manage. Assoc. 2004, 54, 1452-1470. (43) Marr, L. C.; Harley, R. A. Modeling the effect of weekday-weekend differences in motor vehicle emissions on photochemical air pollution in central California. Environ. Sci. Technol. 2002, 36, 4099-4106. (44) Kim, B. M.; Teffera, S.; Zeldin, M. D. Characterization of PM2.5 and PM10 in the south coast air basin of southern California:

5362

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 14, 2005

(49)

Part 1sspatial variations. J. Air Waste Manage. Assoc. 2000, 50, 2034-2044. Kim, B. M.; Teffera, S.; Zeldin, M. D. Characterization of PM2.5 and PM10 in the south coast air basin of California: Part 2s temporal variations. J. Air Waste Manage. Assoc. 2000, 50, 20452059. Fitz, D.; Chan, M.; Cass, G. R.; Lawson, D. R.; Ashbaugh, L. L. A multicomponent size-classifying aerosol and gas sampler for ambient air monitoring. Paper no. 89-140.1 presented at the 82nd Annual Meeting of the Air & Waste Management Association, Anaheim, California, June 25-30, 1989. Blanchard, C. L.; Tanenbaum, S. J. Differences between weekday and weekend air pollutant levels in southern California. J. Air Waste Manage. Assoc. 2003, 53, 816-828. Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; Wiley-Interscience: New York, 1998; p 712. Shah, S. D.; Cocker, D. R.; Miller, J. W.; Norbeck, J. M. Emission rates of particulate matter and elemental and organic carbon from in-use diesel engines. Environ. Sci. Technol. 2004, 38, 25442550.

Received for review November 21, 2004. Revised manuscript received April 18, 2005. Accepted May 5, 2005. ES048172+