Transport of Atmospheric Fine Particulate Matter
Posted on: Saturday, 9 February 2008, 03:00 CST
By Allen, David T Turner, Jay R
ABSTRACT Air quality field data, collected as part of the fine particulate matter Supersites program and other field measurements programs, have been used to assess the role of aerosol transport, over length scales of approximately 100-1000 km, on fine particulate matter concentrations. Assessment of data from New York, NY; Baltimore, MD; Pittsburgh, PA; Atlanta, GA; Houston, TX; St. Louis, MO; and Fresno, CA, indicates that in virtually all of the regions, transport of aerosol over distances of 100-1000 km has a significant impact on urban particulate matter concentrations and a dominant role in determining rural particulate matter concentrations, though the nature of the regional contributions differs from region to region. This assessment is generally consistent with previous conceptual models of fine particulate matter formation and accumulation in these regions. The nature of the transported aerosol is largely sulfate in Eastern and Midwestern cities and nitrate in the Central Valley of California. In addition to physical transport of aerosol over distances of 100-1000 km, regional transport of aerosol precursors may lead to conditions conducive to large-scale nucleation events. Regional nucleation events have been reported in the East, Midwest, and in California. The events occurred in the morning soon after surface layers coupled with layers aloft, and the events generate ultrafine particles. In some cases, these nucleation events have been correlated with availability of sulfur dioxide and, therefore, may be sulfate formation events.
INTRODUCTION
A critical step in developing air quality management plans for reducing fine particulate matter (PM; PM^sub 2.5^) concentrations is to characterize the relative extents of local, regional, and global source contributions. A variety of observational data can be used to assess the relative extents of these sources, and this series of papers reviews the data emerging from recent air quality field programs that characterize the strengths of local, regional, and global sources.
This paper examines the role that the transport of PM^sub 2.5^, on scales ranging from hundreds to thousands of kilometers, has in accounting for observed concentrations and composition of PM in urban areas and regional domains. An additional paper in this series1 will examine the role that local sources can have in generating intraurban variability in PM^sub 2.5^ concentrations.
The focus in this paper is reviewing the understanding of PM^sub 2.5^ transport that emerges from the observational data collected in recent air quality field studies in North America, especially the PM^sub 2.5^ Supersites program, supported by U.S. Environmental Protection Agency.2 This review will build on the analyses performed as part of the NARSTO Assessment of Fine Particulate Matter Science for Policy-makers.3 In that assessment, conceptual models of PM^sub 2.5^ formation, transport, and accumulation were presented for multiple regions in North America, including several regions in which air quality field studies have subsequently been reported. This review examines each of these regions in which field study data have been reported, reviews the assessment of regional transport of PM in the conceptual model presented by NARSTO, and updates the conceptual model, as appropriate. Data from three regions in which NARSTO developed conceptual models are examined: the Northeastern United States, the Southeastern United States, and the San Joaquin Valley (SJV). Data are also presented for two regions for which the NARSTO assessment did not develop conceptual models: the Texas Gulf Coast and the Central Midwest. Data for the Los Angeles area are presented in the paper describing intraurban transport. A summary of the common elements and differences between regions are presented in the concluding section.
Northeastern United States
The NARSTO review3 concludes that, in the Northeastern United States, "long-range transport of SO^sub 4^^sup -^ is very important, particularly in summer. There is a clear gradient of SO^sub 4^^sup - ^ from the Ohio River Valley to upper New England. However, local or urban production of PM^sub 2.5^ is also important, particularly in winter and along the northeast corridor."4 To quantitatively characterize the role of regional sulfate in the Northeast, the NARSTO assessment cited a source apportionment analysis of Philadelphia, PA. That analysis attributed 50% of the PM^sub 2.5^ mass to regional sulfate, 10% to sulfate generated locally, 25-30% to motor vehicles, and the remainder to soil dust, residual oil combustion, and other sources. Data from the PM^sub 2.5^ Supersites in Baltimore, MD; New York, NY; and Pittsburgh, PA, affirm the significance of regional transport of sulfate and provide more quantitative assessments.
In New York, Liu et al.5 compared PM concentrations and sources at two rural locations in New York State during the summers of 2000 and 2001. The sites, Stockton in Western New York and Potsdam in Northern New York, are dominated by the impacts of transport. Source apportionment using positive matrix factorization (PMF).6-8 identified seven sources at the two sites, six of which were common. The magnitudes of the source attributions, shown in Figure 1, were very similar for these two sites, which are separated by hundreds of kilometers. Potential source contribution function (PSCF) analysis, which combines aerosol data with air parcel backward trajectories, indicated that the source regions extended far into the Ohio River Valley for sulfate and hundreds of kilometers into Canada for many of the other sources. Figure 2 shows mappings of the PSCF values, which are defined as the probability that an air parcel with an air pollution concentration above a threshold value passed through a region.5,9
The highly urbanized region of New York City provides both similarities and contrasts with rural New York sites. Dutkiewicz et al.10 compared PM^sub 2.5^ concentrations and sources at Queens, Pinnacle State Park, and Whiteface Mountain. Sulfate concentrations were correlated over these urban and rural sites, and the highest sulfate concentrations at all three of the sites were associated with air masses that passed through the Ohio River Valley and the area around the Great Lakes Basin. On an annual basis, the authors estimated that 44-55% of the sulfate at Queens and 60% at Whiteface and Pinnacle was transported. Qin et al.11 suggest a higher fraction of PM^sub 2.5^ transported into urban New York, reporting that 69- 82% of PM^sub 2.5^ mass at four urban sites derives from transport. For sulfate, the percentage transported was estimated to be above 93%, and for nitrate the percentage transported was estimated as 54- 65%. Source regions on days with high PM^sub 2.5^ concentrations were the border areas of West Virginia, Ohio, and Pennsylvania.11
Lall and Thurston12 also attribute a higher fraction of sulfate observed in urban New York to transport than the 44-55% reported by Dutkiewicz et al.10 By comparing a rural, forest site near New York City to a site in Manhattan in New York City, more than 90% of sulfate mass at the urban site was attributed to transported PM, and sulfate concentrations in warm months were reported as higher than concentrations observed during cold months, indicating a greater impact of transport during warm seasons.
Although much of the sulfate in urban sites in New York City is associated with transport from the Ohio River Valley, some local sources are significant. Lall and Thurston12 report that more than 90% of the elemental carbon (EC) observed in New York City is of local origin, and both Qin et al.11 and Dutkiewicz et al.10,13 report elevated concentrations of metals associated with a heating oil source at urban sites. In addition to oil burning, local vehicular sources12,14 and emissions from ships11 are significant in New York City.
In Pittsburgh, source attributions have been performed for a central, urban site (the Pittsburgh Air Quality In Pittsburgh, source attributions have been performed for a central, urban site (the Pittsburgh Air Quality Figure 1. Average source mass contribution for (a) Potsdam, NY, and (b) Stockton, NY, in summer 2000 and 2001.5 Allen and Turner Volume 58 February 2008 Journal of the Air & Waste Management Association 255 Study site at Carnegie Mellon University)15,16 and for a site south of the urban center (the National Energy Technology Laboratory [NETL] site).17 These sites were characterized by 1- to 3-day episodes during which PM^sub 2.5^ concentrations exceeded 40 [mu]g/m^sup 3^. These episodes typically involved high sulfate concentrations, with no diurnal variability, associated with long-range transport of pollutants. In contrast, the organic material in the PM^sub 2.5^ was dominated by local sources and exhibited diurnal variability. This result is analogous to the findings from New York City, where inorganics (especially sulfate) were because of regional sources and carbonaceous aerosol was because of local sources.
At the NETL site, there is some evidence that secondary organic aerosol may contribute to overall PM concentrations. Semivolatile organic material (SVOM) concentrations peaked at midday at the NETL site, which is located south of the urban center, and minimum concentrations of nonvolatile organic compounds were observed in the afternoon. This pattern was not observed at the urban site at Carnegie Mellon University, which had no diurnal variation in SVOM, and recorded minimum concentrations of nonvolatile organic compounds in the afternoon and evening.18 Overall, at the NETL site, the three major sources of PM^sub 2.5^ were secondary transported material (dominated by ammonium sulfate) from west and southwest (46%), secondary material formed during midday photochemical processes (21%), and primary emissions from diesel (10%) and gasoline (8%) mobile sources.16 At the Carnegie Mellon University site, annual average source attributions (UNMIX) indicated that approximately two thirds of the aerosol mass was because of regional transport (largely sulfate).18 Measurements of particle size distributions at the Carnegie Mellon University site also indicated that, on approximately one third of all of the days, regional scale nucleation events occurred. 19 These events are characterized as regional because the nucleation was followed by a period of particle growth.20 If nucleation was confined to a small area close by the site measuring the nucleation event, the growing mode would disappear once air parcels began arriving at the site from outside the nucleation zone. So a period of growth following a nucleation event is viewed as evidence that the area experiencing conditions leading to nucleation is not local, but "regional."19,20 These events occurred in all seasons and were associated with elevated SO^sub 2^ concentrations and the effective aerosol surface area available for condensation, suggesting that sulfuric acid is a major component of the nucleated particles. The sites in Baltimore, which were all located within a few miles of the Baltimore Harbor, showed a different pattern, with much greater local source contributions during PM^sub 2.5^ episodes than those observed in Pittsburgh or New York.21,22 In Baltimore, source attributions, performed using PMF analysis, at three urban sites identified 12 major sources: oil- fired power plant emissions, two secondary nitrate sources, local gasoline traffic, coal-fired power plants, secondary sulfate, diesel emissions/bus maintenance, a Quebec wildfire episode, nucleation, incinerator, airborne soil/roadway dust, and steel plant emissions. As shown in Figure 3, the relative strengths of the various sources were highly dependent on wind directions, and, at least on the days for which source attributions have been reported, local sources could dominate the total aerosol mass loadings.
Southeastern United States
For the Southeastern United States, the NARSTO review concludes that "the Southeast has an elevated regional level of fine PM, with higher levels in and downwind of the major cities. While a not insignificant fraction is transported into the region, a large part of the regional cloud is self-generated from biogenic emissions (including wood-burning) and coal-fired electricity generation. Increased levels in the urban areas are because of a variety of activities, particularly vehicle traffic."23 Results from the Atlanta, GA, Supersite and other sampling programs in the Southeast support this assessment.
Several studies have found evidence of significant sources of aerosol generated from biogenic emissions (including wood burning) in the Southeast. Tanner et al.24 used radiocarbon measurements to determine that, during the summer, between 50% and 80% of carbonaceous aerosol at rural sites in the Southeast is nonfossil (modern) carbon. Lewis et al.25 found that 56-80% of the carbonaceous PM^sub 2.5^ in Nashville, TN, during the summer was modern carbon.
Other studies have attempted to identify a broader array of sources. Liu et al.26,27 used PMF to resolve the sources of PM^sub 2.5^ at paired urban and rural sites in Georgia and Alabama. In one analysis,26 correlations between the source attributions identified five common source factors for both urban and rural sites: sulfate (particularly in summer), nitrate (particularly in winter), carbonaceous and sulfate particles associated with coal combustion tracers, soil, and wood smoke. Motor vehicle factors were observed at the urban sites but could not be resolved at the rural sites. In a later analysis,27 the same authors, using speciated rather than nonspeciated organic and EC, showed very similar results but were able to distinguish between gasoline and diesel contributions.27 At all of the sites, the average diesel contribution was higher than the gasoline contribution, typically by a factor of approximately two; both urban sites had higher diesel and gasoline contributions than the corresponding rural sites.
These findings are, however, somewhat dependent on the source resolution method. Marmur et al.28 used an Eulerian chemical- transport model (Community Multiscale Air Quality Model [CMAQ]) and chemical mass balancing techniques to identify the sources of PM^sub 2.5^ at the same four urban and rural sources as Liu et al.26,27 Over a 2-month period (July 2001 and January 2002), the CMAQ analyses predict lower contributions at all of the sites for light- duty motor vehicles and higher concentrations for soil dust than the receptor-based methods. CMAQ-predicted concentrations for wood smoke were higher than the receptor-based methods at three of the four sites. Concentrations attributed to heavy-duty diesel engines were similar for the four sites. The authors attribute the source attribution differences to uncertainties in emission estimates and meteorological inputs used in CMAQ. Although these differences in source attributions can have significant impact on day-to-day variability in source attributions, the findings do not call into question the regional nature of the PM^sub 2.5^ sources in the Southeast. The authors report significant covariation of tracer species between urban and rural sites, suggesting that regional transport of PM^sub 2.5^ is significant for multiple sources.
Texas Gulf Coast
The NARSTO assessment3 did not provide a conceptual model for PM^sub 2.5^ formation, transport, and accumulation along the Texas Gulf Coast; however, results emerging from the PM^sub 2.5^ Supersite in Houston, TX,29 provide the basis for a conceptual model, and the component of that conceptual model addressing regional transport is presented here.
The relative strengths of local and regional emission sources of PM^sub 2.5^ in Southeast Texas were characterized by the spatial variation of mass concentrations on individual days, determined using the Federal Reference Method (FRM; description of the method is available at www. epa.gov/ttnamti1/pmfrm.html). Multiple FRM sampling sites were operated in Houston, and the spatial homogeneity of the PM concentrations is an indicator of the extent of local source strength, because localized sources will tend to introduce spatial gradients in PM concentrations. Figure 4a shows the ratio of 25th percentile FRM mass (over all of the sites) to the FRM mass at the site with the highest concentration for that day plotted versus the maximum FRM mass for the day. These data were taken over sites distributed over a region of several thousand square kilometers.30 A high value of this ratio indicates that at least 75% of monitors were recording similar concentrations, suggesting that the region of high concentrations extended over a wide area. In contrast, a low value of this ratio indicates that only a few monitors were experiencing high FRM mass concentrations, suggesting that high concentrations were more localized. Figure 4b shows the distribution of values of the ratio of the 25th percentile FRM mass to the maximum FRM mass. The data in Figure 4 suggest that, on days with high FRM mass concentrations (>15 [mu]g/m^sup 3^), roughly equal numbers of days have values above 0.7 (regional events) and below 0.5 (more localized events). This suggests that both regional and localized events occur frequently along the Texas Gulf Coast.
Further insight into local versus regional PM^sub 2.5^ events can be gained by plotting air parcel back trajectories. Forty-eight- hour back trajectories starting at midday at a height of 100 m were simulated using the Hybrid Single- Particle Lagrangian Integrated Trajectory (HYSPLIT) model. 31 The HYSPLIT model computes air parcel trajectories using archived meteorological datasets, which are described in the model documentation.31 Trajectories were calculated for days where maximum FRM concentration is above 20 [mu]g/m^sup 3^ and where the ratio on the y-axis of Figure 4a is above 0.7. These dates are expected to be regional PM^sub 2.5^ events, because there is little spatial variation in FRM mass. These trajectories are shown in Figure 5. Trajectories were also calculated for dates where maximum FRM mass concentration is above 20 [mu]g/m^sup 3^ and where the ratio on the y-axis in Figure 4a is less than 0.5. These dates are expected to be local PM^sub 2.5^ events, because there is large spatial variation in FRM mass. These trajectories are shown in Figure 6.30
When FRM mass tends to be high and spatially homogeneous in Southeast Texas (Figure 5), synoptic scale winds tend to come from the east or northeast. These results suggest that high levels of background PM^sub 2.5^ are advected into Southeast Texas from the eastern half of North America. In contrast, when FRM mass is high but the stations recording high mass concentrations are isolated, suggesting local source contributions, synopticscale winds may come from any direction and preferentially come from the south- southeast.
This analysis is consistent with a variety of assessments that have been done to characterize the meteorology on days with high ozone concentrations in Houston, Dallas, San Antonio, and Austin, TX.32 High concentrations and spatially uniform distributions of gas- phase photochemical oxidants tend to be associated with transport from the east and northeast. For Houston and Southeast Texas, however, many of the days with high gasphase oxidant concentrations have only localized regions of high ozone concentrations, and a variety of back trajectories directions are associated with these events, analogous to the results shown in Figure 6. Particle composition provides further insight into the local and regional sources of PM^sub 2.5^. Figure 7 shows the mean concentrations of the major components of the 24-hr averaged PM^sub 2.5^ by site and season for multiple sites in Southeast Texas. Measurements were made using the standard methods for PM^sub 2.5^ speciation, as defined by U.S. Environmental Protection Agency (www.epa.gov/ttn/amtic/ speciepg.html). Figure 7 shows that the bulk composition is generally similar between sites in very different settings and far removed from each other. The relative amounts of the major components to total mass also do not vary considerably by season. However, although average compositions show seasonal and spatial homogeneity, individual days can show great variability from mean values.30
Buzcu et al.33 have performed source allocations for the PM^sub 2.5^ at Houston locations that are both typically urban and industrial source dominated. Organic molecular tracers and chemical mass balances were used as the source apportionment tools. The dominant sources during a summer episode, in decreasing order of importance, were found to be secondary sulfate, diesel vehicles, gasoline vehicles, wood smoke, meat cooking, and vegetative detritus. Wood smoke was found to be highly episodic as a source, and the strength of the source was most dependent on fires occurring within a several-hundred-kilometer radius of Houston. Junquera et al.34 estimated emissions from the fires during the Supersite intensive measurement period considered by Buzcu et al.33 and found that, during periods of intense fire activity, fires can become the dominant primary source of PM^sub 2.5^ in the region.
Buzcu et al.33 and Nopmongcol and Allen35 also report very high sulfate concentrations during the fire episode that occurred during the Supersite intensive measurement period and concluded that the fire particles were catalyzing the formation of additional sulfate. Data emerging from other field programs support the hypothesis that fire particles catalyze soot formation.36 Although the observational evidence is not as compelling, the data during the fire episodes suggest that acid-catalyzed aldehyde condensations also occurred.37
One of the unique features of the Houston location is that it offers an opportunity to examine the importance of urban areas as source regions for regional PM. Unlike all of the other Supersite locations except Los Angeles, CA, Houston often has air entering the region with relatively low concentrations of PM (in this case, originating from the Gulf of Mexico). Unlike Los Angeles, where clean air often enters the region from the Pacific but is confined by the mountains surrounding the Los Angeles Basin, clean air masses entering Houston from the Gulf of Mexico can flow, unimpeded by topographic obstacles, to the north. The Houston Supersite provided both modeling and measurements of regional impacts of PM^sub 2.5^ originating from Houston. Brock et al38 made aircraft measurements of aerosol number concentrations, as a function of particle size in multiple aircraft tracks downwind of Houston. The data, shown in Figure 8, indicate two plumes originating from the Houston area. The western plume originates from a power plant and can be accounted for based on oxidation of SO^sub 2^ emissions to particulate sulfate. The eastern plume originates from the urban and industrial source region in Houston. The reported emissions from this source region do not include sufficient primary PM or SO^sub 2^ to explain the PM^sub 2.5^ observations in the plume; however, the urban and industrial emissions do lead to significant photochemical activity. Russell and Allen39 have modeled the interaction of this plume with biogenic emissions of isoprene and monoterpenes in the forests north of Houston and concluded that ozone reactions with monoterpenes, leading to secondary organic aerosol formation, can account for some of the observed aerosol formation. Estimates of aerosol formation from the gasphase oxidation of isoprene are small; however, Brock et al.38 postulate that acid-catalyzed aldehyde condensation reactions may be occurring in the plume.
These types of processes, which lead to PM formation in urban plumes, may impact downwind regions. Although the phenomenon is easiest to study in Houston, because of its location near the Gulf of Mexico and its lack of topographic features that inhibit transport, it likely occurs to some extent in most cities.
Central Midwest
The NARSTO assessment3 did not provide a conceptual model for ambient PM^sub 2.5^ in the Central Midwest region of the United States. A conceptual model is currently being developed for St. Louis, MO, using 4 yr of monitoring data from the St. Louis-Midwest Supersite together with data collected by state/local agencies. Regional transport is the dominant contributor to ambient PM^sub 2.5^ burdens in St. Louis and throughout the Midwest. Annual average PM^sub 2.5^ mass is dominated by sulfate (2002 annual averages for East St. Louis, IL: 24%), nitrate (12%), and ammonium (11%), which arise from secondary processes with evidence from paired urban and rural monitors that regional transport is significant. Source regions include the Upper Ohio River Basin and some of the Tennessee Valley.40 Carbonaceous matter (38%, expressed as 1.8 times the organic carbon [OC] mass) also has a large regional component.
Paired urban and rural monitors were operated during two 3-month measurement intensives. In each case, the urban monitoring location was in East St. Louis, which is 3 km east of the central business district of the city of St. Louis. The first intensive was conducted mid-August through mid-November 2001 with daily 24-hr integrated sampling at Park Hills, MO, which is ~100 km south of the St. Louis urban core. The second intensive was conducted at Reserve, KS, during the period late August through mid-December 2002 with daily 24-hr integrated sampling for two 6-week periods and semicontinuous (hourly) mass, sulfate, nitrate, and black carbon for the entire study period. Reserve is approximately 500 km west/northwest of St. Louis. Although these two intensives do not reflect annual average behavior, they do provide substantial insight into the regional nature of PM^sub 2.5^.
Figure 9a is a scatter plot of daily 24-hr integrated sulfate at the Park Hills and East St. Louis sites. Surface and synoptic scale winds were typically from the south/ southeast during this period, with the rural Park Hills site being upwind of the urban East St. Louis site. There is excellent day-to-day agreement between sulfate concentrations at these two sites. The geometric mean of the daily urban/rural sulfate ratio was 1.05. One exception was September 5- 6, with elevated sulfate at the rural site on the first day and the urban site on the second day. This period corresponded with an unusual sulfate episode. Typically sulfate episodes in the Central Midwest feature a buildup in sulfate of local origin over several (1- 3) days, followed by sustained high-sulfate concentrations for several (1-3) days until a weather front passes through the region. In this case, however, the sulfate episode as observed in East St. Louis was of very high intensity and short duration (~24 hr) immediately following a moreconventional sulfate event. Figure 10 shows hourly sulfate measured at East St. Louis during this period. Figure 9b is a scatter plot of PM^sub 2.5^ OC (Aerosol Characterization Experiment- Asia method) measured at both sites. The OC mass concentration at the rural site is a lower bound on the OC concentration at the urban site, with significant urban excess observed on several days. Given the uniformity in sulfate concentrations between these sites, it can be inferred that the OC measured at the rural site reflects the regional OC impacting the urban site. The geometric mean of the daily urban/rural undenuded OC ratio was 1.8 for this 3-month study period. This urban excess is consistent with the findings of Rao et al.41 for St. Louis using urban Speciation Trends Network monitors and rural Interagency Monitoring of Protected Visual Environments (IMPROVE) monitors for the period March 2001 to February 2002.
Figure 11 shows a second example of a sulfate episode impacting the Central Midwest.42 Hourly sulfate data are shown for East St. Louis (yellow triangles) and Reserve (blue squares), which are separated by approximately 500 km. Both sites were impacted by a multiday sulfate episode, which originated in the Southeastern United States and migrated northwest. Hourly sulfate at both sites exhibit similar gross features with the East St. Louis site experiencing a few high-sulfate excursions not observed at Reserve, consistent with the Naval Research Laboratory model (www.nrlmry.navy.mil/aerosol_web/Docs/globaer_model.html) predictions.
Regional-scale nitrate events were also observed in the Central Midwest, particularly in the winter and spring. For the rural measurement intensive at Reserve, there was strong temporal coupling between nitrate at East St. Louis and the rural site during nitrate episodes. Lee and Hopke40 also report that nitrate events are influenced by ammonia sources in the North Central Plains. On an annual average basis, the urban excess in nitrate is significant but is small compared with the regional contribution.41
Regional contributions to PM^sub 2.5^ burdens in St. Louis as inferred from urban/rural contrast studies are consistent with source apportionment results. Using PMF, Lee and Hopke40 resolved sulfate and nitrate as distinct factors with PSCF maps identifying the Ohio River Valley and Tennessee Valley areas as potential source regions for ammonium sulfate and the North Central Plains as a potential source region for ammonium nitrate. Lee et al.43 used PMF to apportion 2 yr of daily 24-hr integrated data collected at the St. Louis-Midwest Supersite in East St. Louis. They resolved 10 factors, including distinct sulfate (33%) and nitrate (15%) factors, and also a carbon factor containing some sulfate (20%; mass contributions in parentheses). The carbon factor has been observed in other PMF apportionments of PM^sub 2.5^ data using IMPROVE carbon fractions data, and subsequent refinements to the St. Louis PM^sub 2.5^ mass apportionment suggest that this carbon factor is dominated by regionally transported material.44 A preliminary OC apportionment by Jaeckels et al.45 using PMF and organic molecular marker data resolved eight factors, including secondary organic aerosol and wood combustion as distinct factors. Work is in progress to refine this analysis, which should provide insights into locally generated and regionally transported carbon. Similar to Pittsburgh, the St. Louis area is frequently impacted by region-wide nucleation events as evidenced by the distinct temporal patterns in the size distribution following a nucleation event.20,46 Freshly nucleated particles typically appeared shortly after sunrise (although there were cases where nucleation bursts occurred later in the morning). Particle production was not constrained to a simple burst in the morning, however, but rather it typically persisted for a few hours and sometimes throughout the morning. Nucleation events were most frequent in the spring, summer, and fall (observed on approximately one third of such days) and were rare in the winter.
SJV
For the SJV of Central California, the NARSTO review concludes that "there is a strong urban contribution to PM^sub 2.5^ levels on top of high regional levels. For example on days that exceed the 24- hr National Ambient Air Quality Standards for PM^sub 2.5^, urban areas in the SJV on average had 83 percent higher PM^sub 2.5^ mass than did rural areas, with 4.6 times the vegetative burning material (consistent with domestic fireplace usage) and 2 times the mobile source contribution. Absolute NO3- contributions were similar. The regional distribution of secondary compounds likely results from widespread area sources with a combination of diffusive transport and long-range transport aloft."47 Results from the Central California Regional PM10/PM^sub 2.5^ Air Quality Study (CRPAQS) and related measurement programs support this assessment.
The CRPAQS, which was conducted over 14 months at 38 sites, confirmed the region-wide influence of nitrate and the differences in organic matter between urban and rural sites. Chow et al.48 conclude that winter meteorology and residential wood combustion account for the winter-nonwinter and urban-rural contrasts. The role of wood combustion, along with vehicular sources, in the urban regions was further confirmed with organic molecular tracers.49 Chow et al.48 also conclude that upper air currents are responsible for valley-wide transport of NH4NO3. Herner et al.50 report that, for measurements made in Sacramento, Modesto, and Bakersfield, CA, the ammonium nitrate and sulfate particles that the NARSTO review ascribes to regional sources exist as separate particle populations from the carbonaceous particles, which are attributed to more urban sources.
Analyses of individual episodes have also been reported. Turkiewicz et al.51 examined two multiweek episodes, one in the Northern SJV and the other in the Southern SJV. Both showed the characteristic pattern of regional ammonium nitrate and urban carbonaceous aerosol sources. Watson et al.52 examined episodes with nanoparticle (3-10 nm) and ultrafine (10-100 nm) particle events. The causes of the events were morning nucleation, morning traffic, afternoon photochemical processes, and evening home heating, including residential wood combustion. Although sulfur dioxide emissions in Central California are low, Watson et al.52 conclude that residual amounts are sufficient to initiate nucleation events. As was observed in Pittsburgh and St. Louis, these events were measured in the morning, soon after the surface inversion coupled with layers aloft.
CONCLUSIONS
Assessments of data from New York, Baltimore, Pittsburgh, Atlanta, Houston, St. Louis, and Fresno indicate that, in virtually all of the regions, transport of aerosol over distances of 100-1000 km has a significant impact on urban PM concentrations and a dominant role in determining rural PM concentrations. Recent analyses by U.S. Environmental Protection Agency for additional cities in the Eastern United States support this conclusion.53 Although regional-scale transport of PM^sub 2.5^ is significant in all of these locations, seasonal trends in the extent of transport and the composition of the species transported vary from region to region. The nature of the transported aerosol is largely sulfate in the summertime in Eastern and Midwestern cities and nitrate in the Central Valley of California. In addition to physical transport of aerosol over distances of 100-1000 km, regional transport of aerosol precursors may lead to large-scale nucleation events. Regional nucleation events have been reported in Baltimore, Pittsburgh, St. Louis, and the SJV of California. The events occur in the morning soon after surface layers couple with layers aloft and generate ultrafine particles. In some locations, these nucleation events have been correlated with availability of sulfur dioxide and, therefore, may be sulfate formation events.
Most of the assessments of regional transport of PM^sub 2.5^ examined seasonal or annual average contributions to PM^sub 2.5^; however, episodic instances of regional transport from fires, Saharan dust events, and transport from Asia can also be important.54,55 These continental and globalscale phenomena will be examined in a subsequent paper in this series.
DISCLAIMER
Although the research described in this article has been funded in part by U.S. Environmental Protection Agency through cooperative agreement R-82806201, it has not been subjected to the agency's required peer and policy review and, therefore, does not necessarily reflect the views of the agency, and no official endorsement should be inferred.
IMPLICATIONS
Characterizing the strengths of local, regional, and global sources of fine particulate matter is a critical step in developing air quality management plans for reducing fine particulate matter concentrations. Observational data from the fine particulate matter Supersites program reveal that regional sources have a significant impact on urban particulate matter concentrations and play a dominant role in determining rural particulate matter concentrations; however, the nature of these regional contributions is different in different sections of North America.
REFERENCES
1. Turner, J.R.; Allen, D.T. Transport of Atmospheric Fine Particulate Matter: Part 2-Findings from Recent Field Programs on the Intraurban Variability in Fine Particulate Matter; J. Air & Waste Manage. Assoc., 2008, 58, 196-215.
2. United States Environmental Protection Agency. Background Information on the PM Supersite Program, 2000; available at http:// www.epa.gov/ ttn/amtic/supersites.html (accessed 2007).
3. McMurry, P.H.; Shepherd, M.F.; Vickery, J.S., Eds. Particulate Matter Science for Policy Makers: a NARSTO Assessment; Cambridge University Press: Cambridge, U.K., 2004.
4. McMurry, P.H.; Shepherd, M.F.; Vickery, J.S., Eds. Particulate Matter Science for Policy Makers: A NARSTO Assessment; Cambridge University Press: Cambridge, U.K., 2004; Section 10.7.
5. Liu, W.; Hopke, P.K.; Han, Y.; Yi, S.; Holsen, T.M.; Cybart, S.; Kozlowski, K.; Milligan, M. Application of Receptor Modeling to Atmospheric Constituents at Potsdam and Stockton, NY; Atmos. Environ. 2003, 37, 4997-5007.
6. Paatero, P. Least Squares Formulation of Robust Non-Negative Factor Analysis; Chemometr. Intell. Lab. Systems 1997, 37, 15-35.
7. Paatero, P.; Tapper, U. Analysis of Different Modes of Factor Analysis as Least Squares Fit Problems; Chemometr. Intell. Lab. Systems 1993, 18, 183-194.
8. Paatero, P.; Tapper, U. Positive Matrix Factorization: a Non- Negative Factor Model with Optimal Utilization of Error Estimates of Data Values; Environmentrics 1994, 5, 111-126.
9. Gao, N.; Cheng, M.-D.; Hopke, P.K. Potential Source Contribution Function Analysis and Source Apportionment of Sulfur Species Measured at Rubidoux, CA during the Southern California Air Quality Study, 1987; Anal. Chim. Acta 1993, 277, 369-380.
10. Dutkiewicz, V.A.; Qureshi, S.; Khana, A.R.; Ferraro, V.; Schwab, J;. Demerjian, K.; Husain, L. Sources of Fine Particulate Sulfate in New York; Atmos. Environ. 2004, 38, 3179-3189.
11. Qin, Y.; Kim, E.; Hopke, P.K. The Concentrations and Sources of PM^sub 2.5^ in Metropolitan New York City; Atmos. Environ. 2006, 40, S312-S332.
12. Lall, R.; Thurston, G.D. Identifying and Quantifying Transported vs. Local Sources of New York City PM^sub 2.5^ Fine Particulate Matter Air Pollution; Atmos. Environ. 2006, 40, S333- S346.
13. Dutkiewicz, V.A.; Qureshi, S.; Husain, L.; Khana, A.R.; Schwab, J.; Demerjian, K. Elemental Composition of PM^sub 2.5^ Aerosols in Queens, New York: Evaluation of Sources of Fine- Particle Mass; Atmos. Environ. 2006, 40, S347-S359.
14. Ito, K.; Xue, N.; Thurston, G. Spatial Variation of PM^sub 2.5^ Chemical Species and Source-Apportioned Mass Concentrations in New York City; Atmos. Environ. 2004, 38, 5269-5282.
15. Pekney, N.J.; Davidson, C.I.; Zhou, L.; Hopke, P.K. Application of PSCF and CPF to PMF-Modeled Sources of PM^sub 2.5^ in Pittsburgh; Aerosol Sci. Technol. 2006, 40, 952-961. 16. Pekney, N.J.; Davidson, C.I.; Robinson, A.; Zhou, L.; Hopke, P.; Eatough, D.; Rogge, W.F. Major Source Categories for PM^sub 2.5^ in Pittsburgh Using PMF and UNMIX; Aerosol Sci. Technol. 2006, 40, 910- 924.
17. Eatough, D.J.; Anderson, R.R.; Martello, D.V.; Modey, W.K.; Mangelson, N.F. Apportionment of Ambient Primary and Secondary PM^sub 2.5^ during a 2001 Summer Intensive Study at the NETL Pittsburgh Site Using PMF2 and EPA UNMIX; Aerosol Sci. Technol. 2006, 40, 925-940.
18. Modey, W.K.; Eatough, D.J.; Anderson, R.R.; Martello, V.; Takahama, S.; Lucas, L.J.; Davidson, C.I. Ambient Fine Particulate Concentrations and Chemical Composition at Two Sampling Sites in Metropolitan Pittsburgh: a 2001 Intensive Summer Study; Atmos. Environ. 2004, 38, 3165-3178.
19. Stanier, C.O.; Khlystov, A.Y.; Pandis, S.N. Nucleation Events during the Pittsburgh Air Quality Study: Description and Relation to Key Meteorological, Gas Phase, and Aerosol Parameters; Aerosol Sci. Technol. 2004, 38, 253-264.
20. Shi, Q. Aerosol Size Distributions (3 nm to 2 [mu]m) Measured at the St. Louis Supersite (4/1/01-4/30/02); M.S. Thesis, University of Minnesota, Minneapolis, MN, August 2003.
21. Ogulei, D.; Hopke, P.K.; Zhou, L.; Pancras, J.P.; Nain, N.; Ondov, J.M. Source Apportionment of Baltimore Aerosol from Combined Size Distribution and Chemical Composition Data; Atmos. Environ. 2006, 40, S396-S410.
22. Ondov, J.M.; Buckley, T.J.; Hopke, P.K.; Ogulei, D.; Parlange, M.B.; Rogge, W.F., Squibb, K.S.; Johnston, M.V.; Wexler, A.S. Baltimore Supersite: Highly Time- and Size-Resolved Concentrations of Urban PM^sub 2.5^ and Its Constituents for Resolution of Sources and Immune Responses; Atmos. Environ. 2006, 40, S224-S237.
23. McMurry, P.H.; Shepherd, M.F.; Vickery, J.S., Eds. Particulate Matter Science for Policy Makers: a NARSTO Assessment; Cambridge University Press: Cambridge, U.K., 2004; Section 10.6.
24. Tanner, R.L.; Parkhurst, W.J.; McNichol, A.P. Fossil Sources of Ambient Aerosol Carbon Based on 14C Measurements; Aerosol Sci. Technol. 2004, 38, 133-139.
25. Lewis, C.W.; Klouda, G.A.; Ellenson, W.D. Radiocarbon Measurement of the Biogenic Contribution to Summertime PM^sub 2.5^ Ambient Aerosol in Nashville, TN; Atmos. Environ. 2004, 38, 6053- 6061.
26. Liu, W.; Wang, Y.; Russell, A.; Edgerton, E.S. Atmospheric Aerosol over Two Urban-Rural Pairs in the Southeastern United States: Chemical Composition and Possible Source; Atmos. Environ. 2005, 39, 4453-4470.
27. Liu, W.; Wang, Y.; Russell, A.; Edgerton, E.S. Enhanced Source Identification of Southeast Aerosols Using Temperature Resolved Carbon Fractions and Gas-Phase Components; Atmos. Environ. 2006, 40, S445-S466.
28. Marmur, A.; Park, S.; Mulholland, J.A.; Tolbert, P.E.; Russell, A.G. Source Apportionment of PM^sub 2.5^ in the Southeastern United States Using Receptor and Emissions-Based Models: Conceptual Differences and Implications for Time-Series Health Studies; Atmos. Environ. 2006, 40, 2533-2551.
29. Allen, D.T.; Fraser, M.P. An Overview of the Gulf Coast Aerosol Research and Characterization Study the Houston Fine Particulate Matter Supersite; J. Air & Waste Manage. Assoc. 2006, 56, 456-466.
30. Russell, M.M.; Allen, D.T.; Collins, D.R.; Fraser, M.P. Daily, Seasonal and Spatial Trends in PM^sub 2.5^ Mass and Composition in Southeast Texas; Aerosol Sci. Technol. 2004, 38, 14- 26.
31. HYSPLIT4 (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model. 1997; available from NOAA Air Resources Laboratory at http://www. arl.noaa.gov/ready/hysplit4.html (accessed 2007).
32. McGaughey, G.; Durrenberger, C.; McDonald-Buller, E.; Allen, D.T. Back-Trajectory Analyses on High and Low Ozone Days for the Dallas/Fort Worth and Houston/Galveston/Brazoria Ozone Nonattainment Areas for the Years 2001 through 2005; University of Texas Center for Energy and Environmental Resources: Austin, TX, 2006.
33. Buzcu, B.; Yue, Z.W.; Fraser, M.P.; Nopmongcol, U.; Allen, D.T. Secondary Particle Formation and Evidence of Heterogeneous Chemistry during a Wood Smoke Episode in Texas; J. Geophys. Res. 2006, doi: 10.1029/2005JD006143.
34. Junquera, V.; Russell, M.M.; Vizuete, W.; Kimura, Y.; Allen, D.T. Wild- fires in Eastern Texas in August and September 2000: Emissions and Impact on Photochemistry; Atmos. Environ. 2005, 39, 4983-4996.
35. Nopmongcol, U.; Allen, D.T. Modeling of Surface Reactions on Carbonaceous Atmospheric Particles during a Wood Smoke Episode in Houston, Texas; Atmos. Environ. 2006, 40, S524-S537.
36. Ikegami, M.; Okada, K.; Zaizen, Y.; Makino, Y.; Jensen, J.B.; Gras, J.L.; Harjano, H. Very High Weight Ratios of S/K in Individual Haze Particles over Kalimantan during the 1997 Indonesian Forest Fires; Atmos. Environ. 2001, 35, 4237-4243.
37. Nopmongcol, U.; Khamwichit, W.; Fraser, M.P;. Allen. D.T. Modeling Heterogeneous Formation of Secondary Organic Aerosol during a Wood Smoke Episode in Houston, Texas; Atmos. Environ. 2007, 41, 3057-3070.
38. Brock, C.A.; Trainer, M.; Ryerson, T.B.; Neuman, J.A.; Parrish, D.D.; Holloway, J.S.; Nicks, D.K.; Frost, G.J.; Hubler, G.; Fehsenfeld, F.C.; Wilson, J.C.; Reeves, J.M.; Lafleur, B.G.; Hilbert, H.; Atlas, E.L.; Donnelly, S.G.; Schauffler, S.M.; Stroud, V.R.; Wiedinmyer, C. Particle Growth in Urban and Industrial Plumes in Texas; J. Geophys. Res. 2003, 108 (D3), 4111, doi: 10.1029/ 2002JD002746.
39. Russell, M.M.; Allen, D.T. Predicting Secondary Organic Aerosol Formation Rates in Southeast Texas; J. Geophys. Res. Atmos. 2005, 110, do7s17; doi: 10.1029/2004JD004722.
40. Lee, J.H.; Hopke, P.K. Apportioning Sources of PM^sub 2.5^ in St. Louis, MO Using Speciation Trends Network Data; Atmos. Environ. 2006, 40, S360-S377.
41. Rao, V.; Frank, N.; Rush, A.; Dimmick, F. Chemical Speciation of PM^sub 2.5^ in Urban and Rural Areas. Research Triangle Park and Washington, DC. Personal communication, 2006.
42. Deardorff, N.D.; Turner, J.R. An Investigation of PM^sub 2.5^ Climatology at a Rural Midwest Site. In: Regional and Global Perspectives on Haze: Causes, Consequences and Controversies- Visibility Specialty Conference, Air & Waste Management Association, Asheville, NC, October 25-29, 2004; Paper 78.
43. Lee, J.H.; Hopke; Turner, J.R. Source Identification of Airborne PM^sub 2.5^ at the St. Louis-Midwest Supersite; J. Geophys. Res. 2006, 111, D10S10.
44. Garlock, J.L. Refinements to the PM^sub 2.5^ Mass Apportionment for St. Louis; M.S. Thesis, Washington University in St. Louis, St. Louis, MO, 2006.
45. Jaeckels, J.M.; Bae, M.S.; Schauer, J.J. Positive Matrix Factorization (PMF) Analysis of Molecular Marker Measurements to Quantify the Sources of Organic Aerosols. In Proceedings of the 99th Annual Meeting of A&WMA, New Orleans, LA, June 20-23, 2006; Paper No. 1009.
46. Kulmala, M.; Vehkamaki, H.; Petaja, T.; Dal Maso, M.; Lauri, A.; Kerminen, V.-M.; Birmili, W.; McMurry, P.H. Formation and Growth Rates of Ultrafine Atmospheric Particles: a Review of Observations; J. Aerosol Sci. 2004, 35, 143-176.
47. McMurry, P.H.; Shepherd, M.F.; Vickery, J.S., Eds. Particulate Matter Science for Policy Makers: a NARSTO Assessment; Cambridge University Press: Cambridge, U.K., 2004, Section 10.3.
48. Chow, J.C.; Chen, L-W A.; Watson, J.G.; Lowenthal, D.H.; Magliano, K. L.; Turkiewicz, K.; Lehrman, D. PM^sub 2.5^ Chemical Composition and Spatiotemporal Variability during the California Regional PM10/PM^sub 2.5^ Air Quality Study (CRPAQS); J. Geophys. Res. 2006, 111, D10S04.
49. Rinehart, L.R.; Fujita, E.M.; Chow, J.C.; Magliano, K.L.; Zielinska, B. Spatial Distribution of PM^sub 2.5^ Associated Organic Compounds in Central California; Atmos. Environ. 2006, 40, 290-303.
50. Herner, J.D.; Ying, Q.; Aw, J.; Gao, O.; Chang, D.P.Y.; Kleeman, M.J. Dominant Mechanisms that Shape the Airborne Particle Size and Composition Distribution in Central California; Aerosol Sci. Technol. 2006, 40, 827-844.
51. Turkiewicz, K.; Magliano, K.; Najita, T. Comparison of Two Winter Air Quality Episodes during the California Regional Particulate Air Quality Study; J. Air & Waste Manage. Assoc. 2006, 56, 467-473.
52. Watson, J.G.; Chow, J.C.; Park, K.; Lowenthal, D.H. Nanoparticle and Ultrafine Particle Events at the Fresno Supersite; J. Air & Waste Manage. Assoc. 2006, 56, 417-430.
53. Rao, V; Frank, N; Rush, A.; Dimmick, F. Chemical Speciation of PM^sub 2.5^ in Urban and Rural Areas, National Air Quality and Emissions Trends Report. In A&WMA Air Quality Measurements Conference, November 13-15, 2002, Session 1, Paper 9.
54. Begum, B.A.; Kim, E.; Jeong, C.-H.; Lee, D.-W.; Hopke, P.K. Evaluation of the Potential Source Contribution Function Using the 2002 Quebec Forest Fire Episode; Atmos. Environ. 2005, 39, 3719- 3724.
55. Mendoza, A., M.; Garcia, R.; Vela, P.; Lozaro, F.; Allen, D.T. Trace Gases and Particulate Matter Emissions from Wildfires and Agricultural Burning in Northeast Mexico during the 2000 Fire Season; J. Air & Waste Manage. Assoc. 2005, 55, 1797-1808.
David T. Allen
Center for Energy and Environmental Resources, University of Texas at Austin, Austin, TX
Jay R. Turner
Department of Energy, Environmental and Chemical Engineering, Washington University, St. Louis, MO
About the Authors
David Allen is the Gertz Regents Professor in Chemical Engineering and director of the Center for Energy and Environmental Resources at the University of Texas at Austin. Jay Turner is an associate professor in the Department of Energy, Environmental and Chemical Engineering at Washington University in St. Louis. Please address correspondence to: David T. Allen, Center for Energy and Environmental Resources (R7100), University of Texas at Austin, 10100 Burnet Road, Austin, TX 78758; phone: +1-512-475-7842; fax: +1- 512-471-1720; e-mail: allen@che.utexas.edu.
Copyright Air and Waste Management Association Feb 2008
(c) 2008 Journal of the Air & Waste Management Association. Provided by ProQuest Information and Learning. All rights Reserved.
Source: Journal of the Air & Waste Management Association
User Comments (0)
RSS Feeds