Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Dec 2.
Published in final edited form as: Atmos Environ (1994). 2017 Dec 2;175:25–32. doi: 10.1016/j.atmosenv.2017.11.055

The reduction of summer sulfate and switch from summertime to wintertime PM2.5 concentration maxima in the United States

Elizabeth AW Chan a,b,*, Brett Gantt b, Stephen McDow c
PMCID: PMC6134864  NIHMSID: NIHMS947234  PMID: 30220859

Abstract

Exposure to particulate matter air pollution with a nominal mean aerodynamic diameter less than or equal to 2.5 μm (PM2.5) has been associated with health effects including cardiovascular disease and death. Here, we add to the understanding of urban and rural PM2.5 concentrations over large spatial and temporal scales in recent years. We used high-quality, publicly-available air quality monitoring data to evaluate PM2.5 concentration patterns and changes during the years 2000–2015. Compiling and averaging measurements collected across the U.S. revealed that PM2.5 concentrations from urban sites experienced seasonal maxima in both winter and summer. Within each year from 2000 to 2008, the maxima of urban summer peaks were greater than winter peaks. However, from 2012 to 2015, the maxima of urban summertime PM2.5 peaks were smaller than the urban wintertime PM2.5 maxima, due to a decrease in the magnitude of summertime maxima with no corresponding decrease in the magnitude of winter maxima. PM2.5 measurements at rural sites displayed summer peaks with magnitudes relatively similar to those of urban sites, and negligible to no winter peaks through the time period analyzed. Seasonal variations of urban and rural PM2.5 sulfate, PM2.5 nitrate, and PM2.5 organic carbon (OC) were also assessed. Summer peaks in PM2.5 sulfate decreased dramatically between 2000 and 2015, whereas seasonal PM2.5 OC and winter PM2.5 nitrate concentration maxima remained fairly consistent. These findings demonstrate that PM2.5 concentrations, especially those occurring in the summertime, have declined in the U.S. from 2000 to 2015. In addition, reduction strategies targeting sulfate have been successful and the decrease in PM2.5 sulfate contributed to the decline in total PM2.5

Keywords: Air quality, Fine particulate matter, PM2.5, Sulfate, Nitrate, Organic carbon

graphic file with name nihms-947234-f0001.jpg

1. Introduction

Particulate matter (PM) is a heterogeneous mixture of solid and liquid substances suspended in the air. Exposure to PM has been associated with various health effects (Pope and Dockery, 2006). In particular, PM with an aerodynamic diameter of ≤2.5 μm (PM2.5) has been associated with increased risk of total mortality, death from cardiopulmonary disease, and death from lung cancer (U.S. EPA, 2009a). Airborne PM also adversely affects visibility, and major components of PM2.5 that are hygroscopic, such as sulfate and nitrate, can have a substantial impact on light extinction when the ambient relative humidity is high (Chan et al., 1999).

In addition to the nationwide network of monitors measuring PM2.5 mass concentrations, two networks monitor long-term trends in PM2.5 composition in the U.S., the Chemical Speciation Network (CSN, which evolved from the Speciation Trends Network) and the Interagency Monitoring of Protected Visual Environments (IMPROVE) network (Solomon et al., 2014). Both networks have been collecting data since at least 2000. CSN monitors are predominantly located in urban areas, whereas IMPROVE monitors are predominantly located in rural areas. Together, these networks provide nationwide data on the spatial and temporal composition of PM2.5. The purpose of this work is to understand the evolution of seasonal patterns (the time of year maximum and minimum concentrations are observed) and spatial variability in PM2.5 and its constituents at urban and rural sites across the U.S. over the years 2000–2015. Although it is well-understood that the concentration and composition of PM2.5 varies dramatically between the eastern and western U.S., differences in air quality between multi-state regions are less well-defined.

PM2.5 chemical composition data can be used to assess the effcacy of air pollution control efforts, as well as to better understand which major sources of PM2.5 remain. Major constituents of PM2.5 include sulfate, nitrate, and organic carbon (OC). Sulfate typically forms from the oxidation of SO2 emitted by power plants and other industrial failities. Much of the U.S. population resides in urban areas where sulfate has comprised a large portion of PM2.5 mass in the summertime, when the highest sulfate concentrations have historically been observed (Hand et al., 2012b, 2012c; Malm et al., 2002; U.S. EPA, 2009a). Atmospheric ammonia plays a key role in the formation of both sulfate and nitrate (Seinfeld and Pandis, 2016). Nitrate typically forms in the atmosphere from nitrogen oxides emitted by motor vehicles, power plants, and other industrial facilities (U.S. EPA, 2016). Nitrate concentrations are greatest in winter and in urban areas across the U.S., especially in California and the western U.S. (Hand et al., 2012b, 2014, 2012c). OC is directly emitted and can also form in the atmosphere from a complex mixture of natural and anthropogenic sources. OC concentrations are highest in the west, in fall and winter (U.S. EPA, 2009b). In addition, estimates of the portion of organic aerosols that are secondary (formed in the atmosphere from gas phase biogenic and anthropogenic precursors) vary greatly, ranging from 20 to 70% (Baek et al., 2011; Millet et al., 2005; Weber et al., 2007). Determining the seasonal variability and amount of primary (emitted from combustion sources) and secondary organic aerosols could provide information regarding the contribution of various OC sources (Rohr and McDonald, 2016; Schichtel et al., 2017).

To better understand regional-, population-, and temporal-specific differences in U.S. air quality, we used quality-assured PM2.5 mass concentration measurements from 2000 to 2015 to analyze and compare average monthly urban and rural PM2.5 concentrations across the U.S. over time. We then resolved our analyses to more closely examine regional, temporal, and urban/rural PM2.5 changes. Similar analyses were also performed for PM2.5 sulfate and nitrate from 2000 to 2015 and PM2.5 OC from 2009 to 2015, to better understand patterns and changes in the composition of these major PM2.5 components across and within the U.S.

2. Methods

The 24-hour average PM2.5 mass and chemical composition data used in these analyses were obtained from the U.S. EPA’s Air Quality System (AQS) and have a sampling frequency ranging from daily to every 6th day (U.S. EPA, 2015). The PM2.5 mass data collection follows specific quality assurance procedures required for regulatory use. All data from Federal Reference Method (FRM) or Federal Equivalent Method (FEM) monitors were included, regardless of the length of time each site reported data, the presence of an exceptional event flag, or whether they were used for regulatory purposes. PM2.5 data used were site-level (i.e., daily concentrations from the primary monitor per site, if available, or the average of daily concentrations from co-located monitors), 24-hour (midnight to midnight) FRM/FEM average PM2.5 measurements (parameter code 88101) collected in the 50 U.S. states.

The AQS database (https://www.epa.gov/outdoor-air-quality-data) includes various metadata information for each site, including the city, county, state, EPA Region, and location setting. ‘Location setting’ includes the categories ‘rural,’ ‘suburban,’ ‘urban and center city,’ and ‘unknown,’ and are assigned by the local air quality agency that administers the network. Using recent U.S. census data, both ‘urban and center city’ and ‘suburban’ sites were determined to be located in areas of relatively high population, thus the two categories were combined and henceforth termed ‘urban’ (Supplemental Fig. 1). The 0.2% of measurements labeled as ‘unknown’ were excluded from the urban/rural analyses.

Monitor-level, 24-hour average (midnight to midnight) sulfate (parameter code 88403), nitrate (parameter code 88306), and OC (parameter codes 88320 [IMPROVE] and 88370 [CSN]) concentrations from the CSN and IMPROVE networks were also obtained from AQS and converted to the site-level using values from the lowest Parameter Occurrence Code monitor, to avoid the inclusion of multiple measurements from the same site. Although the CSN and IMPROVE networks measure similar parameters, each have slightly different measurement biases (Solomon et al., 2014). For example, the PM2.5 OC measurement biases are known to have changed between 2007 and 2009 for the CSN. In addition, the networks used different blank correction procedures throughout the time period of interest. IMPROVE data were blank corrected following the usual IMPROVE procedure of subtracting a monthly average field blank, and CSN data were not blank corrected. Although not all PM2.5 sites have chemical composition measurements and not all species were measured at each PM2.5 site, grouping measurements by EPA Region (Supplemental Fig. 2) and location setting enables unique analyses of the monthly average trends over time. EPA Regions were selected to divide the U.S. into smaller, more meteorologically-similar areas, and to provide a better understanding of Region-specific trends.

3. Results

3.1. Total PM2.5

To better understand the seasonal patterns and long-term changes in PM2.5 concentrations at monitoring sites across the U.S., we obtained and compiled all daily, site-specific, rural and urban PM2.5 concentration measurements reported by FRM and FEM monitors from 2000 to 2015 in the U.S. Although the majority of PM2.5 monitors are located in urban areas, a significant subset (∼18%) were positioned in rural areas (Supplemental Table 1). Urban monitors were more concentrated in the eastern U.S. and along the west coast, whereas rural monitors were fairly evenly distributed across the U.S. (Supplemental Fig. 3). It should also be noted that the number and distribution of monitors across the U.S. varied slightly each year.

To ensure that monitors collecting data over short periods of time (e.g., 1 year) did not greatly influence the data, we performed a sensitivity analysis using a subset of long-term monitors. We defined long-term monitoring sites as those reporting at least 11 measurements per quarter from all four quarters and from 75% of the years 2000–2015 (Supplemental Fig. 4). Although this reduced the number of monitoring sites by approximately half, the overall PM2.5 concentration trends of urban and rural sites remained consistent in the subset of long-term monitors. The only notable differences were slightly higher (< 1 μg/ m3) overall concentrations and higher summer rural maxima in the subset of monitors, though rural maxima never reached or surpassed summer urban maxima in this subset of monitors.

Monthly urban and rural PM2.5 average concentrations over the 16 years of interest show distinct seasonal patterns (Fig. 1A). Intriguingly, the relative peak heights of monthly average urban PM2.5 concentrations are higher in the summers (June, July, and August) than in the winters (December, January, and February) from 2002 to 2008, then switch so that winter peaks are higher than summer peaks during 2009–2015. Rural PM2.5 concentrations peaked only in the summers between 2000 and 2008, but tended to display small peaks in both the winters and summers from 2009 to 2015. The average concentrations of PM2.5 in rural areas has been consistently lower than in urban areas, and concentrations in both location types have decreased over time.

Fig. 1. Urban and rural PM2.5 concentration patterns and changes in the U.S. from 2000 to 2015.

Fig. 1.

Open in a new tab

A) Annual (thinner lines) and monthly (thicker lines) average concentrations of 24-hour PM2.5 urban (black) and rural (gray) measurements across the U.S. from 2000 to 2015. Faint grid lines indicate increments of 4 μg/m3 on the y-axis and January of each year on the x-axis. B) Average monthly urban (black) and rural (gray) U.S. PM2.5 concentrations. The monthly average, which was separated by year in panel A, was combined with the same months from other years. Seasons are demarcated by vertical dashed lines here, but are applicable to all figures. C) Average monthly urban and rural PM2.5 concentrations of the grouped EPA Regions. EPA Regions were combined by location, with the lightest line representing the eastern Regions (1–4), the medium-gray line representing the central Regions (5–7), and the darkest gray line representing the western Regions (8–10). D) Average national monthly urban and rural PM2.5 concentrations for each of the 16 years. Each year is colored by a different gradient of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line. E) Average regional monthly urban and rural PM2.5 concentrations for each year for each of the 10 EPA Regions. Within each graph, years are colored by a different shade of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line.

To determine monthly urban and rural PM2.5 trends over a typical year, the monthly averages for each month from 2000 to 2015 were pooled, and an average concentration for each month in the year was calculated (Fig. 1B). On average, there were urban PM2.5 concentration maxima of ∼12.5 μg/m3 in summer and winter and concentration minima of 9–10 μg/m3 in spring (March, April, and May) and fall (September, October, and November) across the U.S. Rural PM2.5 concentrations reached 11 μg/m3 in summer, but only ∼9 μg/m3 in winter, and dipped below 8 μg/m3 in spring and fall.

The rural and urban U.S. averages were then grouped by location (Fig. 1C and Supplemental Fig. 2). Urban and rural locations in eastern (Regions 1–4) and central (Regions 5–7) states displayed both summer and winter PM2.5 peaks, with maximum concentration occurring in the summertime. The seasonal pattern in eastern and central states differed greatly from the western states (Regions 8–10). Western urban sites displayed a strong winter peak, whereas rural western sites remained relatively low throughout the year. Rural PM2.5 concentrations display slightly reduced summer peaks as compared to urban, and substantially reduced winter peaks.

When the monthly averages of each of the 16 years were superimposed, the consistent decrease in PM2.5 over time at both rural and urban locations became clear (Fig. 1D). Although in general, the average PM2.5 concentrations decrease over time, a clear distinction in the yearly concentration graphs between 2008 and 2009 is apparent at both rural and urban monitoring sites. The years 2000–2008 and 2009–2015 each group together, leaving a distinct separation that is most apparent in the summertime. Of note, the year 2002 averaged the highest summer concentrations, 2014 averaged the lowest summer concentrations, and 2015 has the lowest spring and fall concentrations of all years analyzed.

To understand how PM2.5 concentrations changed over time within each EPA Region, we compared the average urban and rural monthly concentrations of each year in each EPA Region (Fig. 1E). EPA Regions were selected as a method to subset the country into a manageable number of regions, while also retaining measurements collected in Alaska and Hawaii. While all EPA Regions display a general downward trend over time, substantial reductions in the summer peaks are evident in Regions 1–5 over time at both urban and rural sites. This reduction in summer peaks over time is less pronounced in EPA Regions located in the western U.S. In addition, Regions 1–5 show detectable wintertime peaks and troughs of similar magnitude in the spring and fall. While there are smaller reductions in total PM2.5 over time in Regions 6–10, the overall yearly trends are rather different. PM2.5 concentrations in Region 6 are reasonably consistent throughout all four seasons, whereas Region 7 displays discernible summer and winter peaks of similar magnitude. Regions 8–10 display distinct winter peaks at urban sites, as well as year-specific peaks in late summertime. In contrast, summer peaks of rural PM2.5 are generally small or nonexistent in Regions 8–10. Also, rural concentrations are almost always lower than urban, in agreement with previous analyses (Hand et al., 2012c). Interestingly, the highest monthly average concentration (nearly 40 μg/m3) occured at rural sites in Region 10 in September of 2012, which resulted from a wildfire (see the Discussion section for additional information).

3.2. PM2.5 composition

Sulfate, nitrate, and organic carbon constitute a substantial portion of PM2.5 mass throughout the U.S., although the relative contributions vary by region and season (U.S. EPA, 2009a). Both sulfate and nitrate are known to display seasonal patterns, with sulfate concentrations increasing in the summers and nitrate concentrations increasing in the winters (Hand et al., 2012c). In order to better understand if and how these components contribute to the seasonal variations in PM2.5, the monthly average concentrations of PM2.5 sulfate, nitrate, and OC were evaluated. Please note that the number of monitors reporting PM2.5 sulfate, nitrate, and OC concentration measurements each year is smaller than the number of monitors reporting total PM2.5 concentrations (Supplemental Tables 2–4 and Supplemental Figs. 5–7), though most sites do report a 24-h measurement every 3 or 6 days. In addition, just under 50% of PM2.5 sulfate, nitrate, and OC monitors were located at urban sites, as compared to ∼80% of PM2.5 monitors.

3.3. PM2.5 sulfate

We evaluated urban and rural PM2.5 sulfate concentrations in a similar manner as total PM2.5. In agreement with previous reports (Hand et al., 2012b, 2014, 2012c; Malm et al., 2002), we observed a distinct decrease in the magnitude of summer PM2.5 sulfate peaks over the time period analyzed, especially after 2008 (Fig. 2A and D). In addition, trends from the subset of long-term monitors (sites reporting at least 11 measurements per quarter from all four quarters and from 75% of the years 2000–2015; Supplemental Fig. 8) appeared very similar to trends from all PM2.5 sulfate monitors, even considering that only ∼20% of urban PM2.5 sulfate monitors satisfied the criteria to be considered long-term.

Fig. 2. Urban and rural PM2.5 sulfate concentration patterns and changes in the U.S. from 2000 to 2015.

Fig. 2.

Open in a new tab

A) Annual (thinner lines) and monthly (thicker lines) average concentrations of 24-hour PM2.5 urban (black) and rural (gray) sulfate measurements across the U.S. from 2000 to 2015. Faint grid lines indicate increments of 1 μg/m3 on the y-axis and January of each year on the x-axis. B) Average monthly urban (black) and rural (gray) U.S. PM2.5 sulfate concentrations. The monthly average, which was separated by year in panel A, was combined with the same months from other years. C) Average monthly urban and rural PM2.5 sulfate concentrations of the grouped EPA Regions. EPA Regions were combined by location, with the lightest line representing the eastern Regions (1–4), the medium-gray line representing the central Regions (5–7), and the darkest gray line representing the western Regions (8–10). D) Average national monthly urban and rural PM2.5 sulfate concentrations for each of the 16 years. Each year is colored by a different gradient of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line. E) Average regional monthly urban and rural PM2.5 sulfate concentrations for each year for each of the 10 EPA Regions. Within each graph, years are colored by a different shade of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line. Please note that some CSN monitors did not begin reporting data until the middle or end of 2000, leading to several missing data points in that year from various urban graphs of PM2.5 components in individual regions.

Both urban and rural monthly PM2.5 sulfate concentration averages peaked in the summers, and urban concentrations were higher than rural (Fig. 2B). In addition, summer PM2.5 sulfate concentrations were greater in the eastern and central U.S. (Regions 1–7) than in the western U.S. (Regions 8–10) (Fig. 2C). This result may be related to the distribution of urban monitor locations, which are more concentrated in the eastern U.S. (Supplemental Fig. 5), where sulfate concentrations are higher. In contrast, rural monitors are more evenly distributed across the U.S. It is also important to note that seasonal patterns of PM2.5 sulfate differ between the eastern and western U.S., with summer peaks clearly visible at urban and rural sites in EPA Regions 1–7, but absent at rural monitors in Regions 8–10. Although it appears that rural PM2.5 sulfate concentrations have decreased only slightly over time (Fig. 2D), when combined with Fig. 2C it is more likely that grouping rural areas masks the differences between PM2.5 sulfate patterns observed in the eastern U.S. versus the western U.S. (Supplemental Fig. 5). Interestingly, summer PM2.5 sulfate concentration maxima dramatically decreased in magnitude after 2008 (Fig. 2A, D, and 2E), similar to PM2.5 (Fig. 1A).

3.4. PM2.5 nitrate

Urban and rural PM2.5 nitrate concentrations peaked in the winter months, with urban concentrations always greater than rural concentrations and the magnitude of both urban and rural PM2.5 nitrate concentration maxima remained reasonably consistent over the 16 years analyzed (Fig. 3A and B). This is consistent with previous analyses of rural areas (Hand et al., 2012b, 2012c). Measurements from the subset of long-term monitors (those reporting > 75% of the time for at least 12 years) show that PM2.5 nitrate trends behaved similarly, although urban monitors reached slightly higher maxima (Supplemental Fig. 9). PM2.5 nitrate concentrations varied greatly by urban/rural status, with high wintertime concentrations reported from most urban Region groupings and some individual rural Regions (Fig. 3C, D, and 3E). While a study looking at the Great Plains found an increase in wintertime nitrate ion concentrations (Hand et al., 2012a), our analysis did not show considerable variation in wintertime maxima over time (Fig. 3D). It should be noted that previous analyses have identified that the highest PM2.5 nitrate concentrations occur over a spatial scale smaller than that of EPA Regions, such as the Los Angeles Basin and the San Joaquin Valley (Hand et al., 2012c), and that classifying locations by EPA Regions does not capture this geographical characteristic (Supplemental Table 3 and Supplemental Fig. 6). However, the seasonal patterns of PM2.5 nitrate shown here remain relevant and informative.

Fig. 3. Urban and rural PM2.5 nitrate concentration patterns and changes in the U.S. from 2000 to 2015.

Fig. 3.

Open in a new tab

A) Annual (thinner lines) and monthly (thicker lines) average concentrations of 24-hour PM2.5 urban (black) and rural (gray) nitrate measurements across the U.S. from 2000 to 2015. Faint grid lines indicate increments of 1 μg/m3 on the y-axis and January of each year on the x-axis. B) Average monthly urban (black) and rural (gray) U.S. PM2.5 nitrate concentrations. The monthly average, which was separated by year in panel A, was combined with the same months from other years. C) Average monthly urban and rural PM2.5 nitrate concentrations of the grouped EPA Regions. EPA Regions were combined by location, with the lightest line representing the eastern Regions (1–4), the medium-gray line representing the central Regions (5–7), and the darkest gray line representing the western Regions (8–10). D) Average national monthly urban and rural PM2.5 nitrate concentrations for each of the 16 years. Each year is colored by a different gradient of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line. E) Average regional monthly urban and rural PM2.5 nitrate concentrations for each year for each of the 10 EPA Regions. Within each graph, years are colored by a different shade of gray, from 2000 shown in the lightest gray line to 2015 shown in the darkest gray line. Please note that some CSN monitors did not begin reporting data until the middle or end of 2000, leading to several missing data points in that year from various urban graphs of PM2.5 components in individual regions.

3.5. PM2.5 OC

OC is a significant component of PM2.5 in the U.S. and long-term normalized spatial and seasonal patterns of PM2.5 OC have been reported (Attwood et al., 2014; Hand et al., 2013; Kim et al., 2015; Zhang et al., 2007). OC has many sources, such as wildland fires, various combustion sources, and secondary organic aerosol formation, of which the proportional contributions vary throughout the year (Kleindienst et al., 2007). The methods by which PM2.5 OC were measured differed between the IMPROVE and CSN networks before 2007, but methods were modified in both networks from 2007 to 2009 to make measurements comparable after 2009 (Solomon et al., 2014). Although the full 16 years of data are available from many rural OC monitors from the IMPROVE network (Supplemental Fig. 10), we restricted our primary analysis to urban and rural PM2.5 OC measurements collected from 2009 to 2015 in order to be able to directly compare measurements from both networks (Supplemental Table 4 and Supplemental Fig. 7).

Monthly urban PM2.5 OC concentration averages were higher than rural across the nation, and the two site categories displayed differing seasonal patterns (Fig. 4A and B). Urban PM2.5 OC concentrations peaked in the winters, whereas rural PM2.5 OC peaked in the summers. Seasonal patterns were less clear when only measurements from the subset of long-term PM2.5 OC monitors were included (Supplemental Fig. 11). When considering the average seasonal pattern of each EPA Region over the 7-year period spanning 2009–2015, urban sites in Regions 9 and 10 displayed distinct winter PM2.5 OC peaks, whereas only subtle summer peaks were visible in other Regions (Fig. 4C, D, and 4E). The lack of seasonal PM2.5 OC peaks in Region 4 was interesting because this area is thought to have very high summertime biogenic secondary organic aerosol concentrations (Carlton et al., 2010). Similar to PM2.5 nitrate (Fig. 3), PM2.5 OC concentrations have shown little change in the magnitude of the seasonal maxima and minima over time (Fig. 4D and E).

Fig. 4. Urban and rural PM2.5 OC concentration patterns and changes in the U.S. from 2009 to 2015.

Fig. 4.

Open in a new tab

A) Annual (thinner lines) and monthly (thicker lines) average concentrations of 24-hour PM2.5 urban (black) and rural (gray) OC measurements across the U.S. from 2009 to 2015. Faint grid lines indicate increments of 1 μg/m3 on the y-axis and January of each year on the x-axis. B) Average monthly urban (black) and rural (gray) U.S. PM2.5 OC concentrations. The monthly average, which was separated by year in panel A, was combined with the same months from other years. C) Average monthly urban and rural PM2.5 OC concentrations of the grouped EPA Regions. EPA Regions were combined by location, with the lightest line representing the eastern Regions (1–4), the medium-gray line representing the central Regions (5–7), and the darkest gray line representing the western Regions (8–10). D) Average national monthly urban and rural PM2.5 OC concentrations for each of the 16 years. Each year is colored by a different gradient of gray, from 2009 shown in the lightest gray line to 2015 shown in the darkest gray line. E) Average regional monthly urban and rural PM2.5 OC concentrations for each year for each of the 10 EPA Regions. Within each graph, years are colored by a different shade of gray, from 2009 shown in the lightest gray line to 2015 shown in the darkest gray line. Please note that some CSN monitors did not begin reporting data until the middle or end of 2000, leading to several missing data points in that year from various urban graphs of PM2.5 components in individual regions.

4. Discussion

This analysis presented publicly-available FRM and FEM PM2.5 mass measurements in a novel way that revealed clear, large-scale temporal and spatial patterns. Importantly, these findings demonstrate that summertime PM2.5 sulfate concentrations have dramatically decreased over time. This change likely contributed to both the continued decrease in total PM2.5 and the switch in average PM2.5 concentration maxima from summers to winters in the U.S.

The large summertime maxima in PM2.5 concentrations observed prior to 2009 in the eastern U.S. have been well-characterized and attributed to seasonal differences in meteorological conditions and sulfate oxidation from SO2 (Bell et al., 2007; Hand et al., 2012c; Hidy et al., 1978; Hitchcock, 1976; Malm et al., 2002; Schwab et al., 2004; Wittig et al., 2004). Averaged across the U.S. and in most of the EPA Regions, a transition from an annual PM2.5 maxima in the summer to in the winter occurred between 2000 and 2015 in urban areas (Fig. 1). This can be partially explained by the substantial decrease in summer sulfate during the same period and in similar areas. Since sulfate is formed in the atmosphere by oxidation of SO2, and a strong association between SO2 emissions and sulfate concentrations has been well established in the eastern U.S. (Hand et al., 2012c; Husain et al., 1998; Malm et al., 2002), the decrease in summer sulfate maxima is a predictable consequence of reducing SO2 emissions. Specifically, SO2 emissions from coal-fired power plants, the largest source of U.S. SO2 emissions in 1990, have decreased by > 70% between 1990 and 2010 (Xing et al., 2015) and national SO2 emissions have declined by > 6% per year between 2000 and 2010 (Hand et al., 2012c). Because sulfate accounts for a greater fraction of PM2.5 in the summer, it follows that a decrease in PM2.5 concentrations due to a reduction of sulfate would be most noticeable in summer, resulting in a substantial reduction in the magnitude of summer PM2.5 peaks, as we have observed (Fig. 1A). While not a 1-to-1 relationship, EPA Regions in which summertime PM2.5 concentrations decreased over time also displayed a corresponding decrease in summertime PM2.5 sulfate.

The flattening of the seasonal sulfate concentration pattern may herald the end of summer as the season of highest annual PM2.5 concentrations. The reduction in the summer PM2.5 concentration peaks together with the emergence of winter as the season with the highest nationwide PM2.5 concentrations may have far-reaching implications for human exposure patterns and associations with co-pollutants, such as ozone. The near disappearance of the summer PM2.5 sulfate peaks also provides a vivid visualization of a successful emission control strategy, namely the reduction of power plant SO2 emissions.

Within the 2000–2015 time frame examined here, the highest monthly PM2.5 concentration averages in EPA Regions 1–5 all occurred in July 2002. On July 6, 2002, lightning ignited several wildfires in Quebec, which produced a smoke plume that blanketed New England on July 7. Secondary analyses determined that this singular event was the primary cause of those high PM2.5 concentrations. The highest monthly rural average occurred September 2012 in Region 10, which contributed to the only instance where the rural monthly average surpassed the urban monthly average. The cause of this was a single monitor in Lemhi County, Idaho that reported very high concentrations of PM2.5 through August and September of that year due to wildfire smoke. In fact, late summer maxima that do not decrease over time generally coincide with large wildfires.

Although we show that PM2.5 sulfate concentrations have decreased over time in EPA Regions 1–7 (the eastern and central U.S.), details of regional impacts resulting from these differences were not a focus of this analysis. Instead, the air quality trends highlighted here are those of long-term national scale data, such as national scale decreases in PM2.5 sulfate due to reductions in SO2 emissions and observations of extreme PM events due to wildfires. The dominance of electric power generating units as a national SO2 source, their concentration in the eastern U.S., and the season-specific chemistry of sulfate formation combine to make an analysis based on seasonal and regional classification of a large data set a worthwhile approach for investigating the impact of decreases in SO2 emissions from electric power generation. Similarly, extreme PM2.5 concentrations from fire emissions on a given day have the potential to strongly influence average daily, and even monthly, concentrations across an entire state or region, and can be easily compared to wildfire data to pinpoint the cause.

There have been other major changes in PM2.5 and precursor emissions that apparently did not result in obvious trends in national scale data. The most salient example is the decline in NOX concentration. Total U.S. NOX emissions decreased by 48% between 2002 and 2014 (U.S. EPA, 2017), resulting in a greater contribution from oil and gas production and natural gas emissions relative to coal (U.S. EPA, 2017). However, these emissions reductions and changes in source contributions do not appear to have resulted in major changes in ambient PM2.5 nitrate concentration (Fig. 3), even though NOX is a known precursor of particulate nitrate. Changes in emissions control and fuel composition of both diesel and gasoline vehicles have greatly reduced PM, NOX, and SO2 emissions from automobiles (CRC, 2013). In principle, reductions in sulfate and precursor SO2 emissions from motor vehicles could impact regional ambient concentrations of PM2.5 and PM2.5 sulfate, as they are relatively long-lived can therefore undergo long distance transport. However, in practice SO2 emissions from transportation accounts for less than 5% of national SO2 emissions even before recent reductions in fuel sulfur content (U.S. EPA, 2008).

Combining our analyses with knowledge about seasonal emissions and sources may provide insight into the breakdown of OC into primary and secondary emissions. Combustion sources emitting primary organic aerosol emissions are abundant in winter due to home heating and in summer and fall due to wildfires, whereas industrial, transportation, and cooking sources emit year round. In contrast, increased biogenic precursor emissions and more rapid photochemical reaction rates lead to increased secondary organic aerosol in the summer (Baek et al., 2011; Millet et al., 2005; Weber et al., 2007). As we observed the highest rural PM2.5 OC concentrations in the summertime, it suggests that rural OC predominantly results from wildland fire smoke and secondary organic aerosols. In addition, as urban OC concentrations are highest in the wintertime, a substantial portion of urban OC may come from primary sources like home heating (Baek et al., 2011; Millet et al., 2005). Secondary OC production is enhanced when sulfate is present (Budisulistiorini et al., 2016; Pye et al., 2013), and aerosol water and aqueous processing may influence secondary organic aerosol formation (Adamson et al., 2004; Carlton and Turpin, 2013; Pye et al., 2017).

Although sulfate, nitrate, and OC are the three largest components of PM2.5 mass, other contributors include elemental carbon, crustal material, sea-salt, and ammonium. The contribution of these other components was beyond the scope of this analysis, and are not as likely to contribute to the changes we observed in total PM2.5 because they are not as abundant.

Supplementary Material

Sup 1

Acknowledgments

We would like to thank Drs. Havala Pye, James Kelly, Steven Dutton, and John Vandenberg for technical review and helpful comments on this manuscript. We would also like to thank Dr. Barbara Buckley for her mentorship.

IMPROVE is a collaborative association of state, tribal, and federal agencies, and international partners. The U.S. Environmental Protection Agency is the primary funding source, with contracting and research support from the National Park Service. The Air Quality Group at the University of California, Davis is the central analytical laboratory, with ion analysis provided by Research Triangle Institute, and carbon analysis provided by Desert Research Institute.

Role of the funding source

This research was supported in part by an appointment to the Research Participation Program for the U.S. EPA, Offce of Research and Development, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and EPA. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Abbreviations:

AQS

Air quality system

CSN

Chemical Speciation Network

EPA

Environmental Protection Agency

FEM

Federal Equivalent Method

FRM

Federal Reference Method

IMPROVE

Interagency Monitoring of Protect Visual Environments

NAAQS

National Ambient Air Quality Standard

OC

organic carbon

PM

particulate matter

SO2

sulfur dioxide

U.S.

United States

Footnotes

Declaration of interest

Conflicts of interest

None.

Publisher's Disclaimer: Disclaimer

The study was reviewed by the National Center for Environmental Assessment and the Offce of Air Quality Planning and Standards of the U.S. Environmental Protection Agency. Approval does not signify that the contents necessarily reflect the view and policies of the U.S. Environmental Protection Agency.

References

  1. Adamson IY, Prieditis H, Vincent R, 2004. Soluble and insoluble air particle fractions induce differential production of tumor necrosis factor alpha in rat lung. Exp. Lung Res 30, 355–368. [DOI] [PubMed] [Google Scholar]
  2. Attwood A, Washenfelder R, Brock C, Hu W, Baumann K, Campuzano-Jost P, Day D, Edgerton E, Murphy D, Palm B, 2014. Trends in sulfate and organic aerosol mass in the Southeast US: impact on aerosol optical depth and radiative forcing. Geophys. Res. Lett 41, 7701–7709. [Google Scholar]
  3. Baek J, Hu Y, Odman MT, Russell AG, 2011. Modeling secondary organic aerosol in CMAQ using multigenerational oxidation of semi-volatile organic compounds. J. Geophys. Res. Atmos 116. [Google Scholar]
  4. Bell ML, Dominici F, Ebisu K, Zeger SL, Samet JM, 2007. Spatial and temporal variation in PM(2.5) chemical composition in the United States for health effects studies. Environ. Health Perspect 115, 989–995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Budisulistiorini SH, Nenes A, Carlton AMG, Surratt JD, McNeill VF, Pye HOT, 2016. Simulating aqueous-phase isoprene-epoxydiol (IEPOX) secondary organic aerosol production during the 2013 Southern Oxidant and Aerosol Study (SOAS). Environ. Sci. Technol [DOI] [PMC free article] [PubMed]
  6. Carlton A, Turpin BJ, 2013. Particle partitioning potential of organic compounds is highest in the Eastern US and driven by anthropogenic water. Atmos. Chem. Phys 13, 10203–10214. [Google Scholar]
  7. Carlton AG, Bhave PV, Napelenok SL, Edney EO, Sarwar G, Pinder RW, Pouliot GA, Houyoux M, 2010. Model representation of secondary organic aerosol in CMAQv4.7. Environ. Sci. Technol 44, 8553–8560. [DOI] [PubMed] [Google Scholar]
  8. Chan Y, Simpson R, Mctainsh GH, Vowles PD, Cohen D, Bailey G, 1999. Source apportionment of visibility degradation problems in Brisbane (Australia) using the multiple linear regression techniques. Atmos. Environ 33, 3237–3250. [Google Scholar]
  9. CRC, 2013. Report: Phase 2 of the Advanced Collaborative Emissions Study. Final Report Southwest Research Institute, San Antonio, TX. [Google Scholar]
  10. Hand J, Gebhart K, Schichtel B, Malm W, 2012a. Increasing trends in wintertime particulate sulfate and nitrate ion concentrations in the Great Plains of the United States (2000–2010). Atmos. Environ 55, 107–110. [Google Scholar]
  11. Hand J, Schichtel B, Malm W, Frank N, 2013. Spatial and temporal trends in PM2. 5 organic and elemental carbon across the United States. Adv. Meteorology 2013 https://www.hindawi.com/journals/amete/2013/367674/abs/. [Google Scholar]
  12. Hand J, Schichtel B, Malm W, Pitchford M, Frank N, 2014. Spatial and seasonal patterns in urban influence on regional concentrations of speciated aerosols across the United States. J. Geophys. Res. Atmos 119. [Google Scholar]
  13. Hand J, Schichtel B, Malm W, Pitchford M, Frank N, 2014. Spatial and seasonal patterns in urban influence on regional concentrations of speciated aerosols across the United States. J. Geophys. Res. Atmos 119. [Google Scholar]
  14. Hand J, Schichtel B, Pitchford M, Malm W, Frank N, 2012c. Seasonal composition of remote and urban fine particulate matter in the United States. J. Geophys. Res. Atmos 117. [Google Scholar]
  15. Hidy G, Mueller P, Tong E, 1978. Spatial and temporal distributions of airborne sulfate in parts of the United States. Atmos. Environ 12 (1967), 735–752. [Google Scholar]
  16. Hitchcock DR, 1976. Atmospheric sulfates from biological sources. J. Air Pollut. Control Assoc 26, 210–215. [DOI] [PubMed] [Google Scholar]
  17. Husain L, Dutkiewicz VA, Das M, 1998. Evidence for decrease in atmospheric sulfur burden in the eastern United States caused by reduction in SO2 emissions. Geophys. Res. Lett 25, 967–970. [Google Scholar]
  18. Kim P, Jacob DJ, Fisher JA, Travis K, Yu K, Zhu L, Yantosca RM, Sulprizio M, Jimenez JL, Campuzano-Jost P, 2015. Sources, seasonality, and trends of southeast US aerosol: an integrated analysis of surface, aircraft, and satellite observations with the GEOS-Chem chemical transport model. Atmos. Chem. Phys 15, 10411–10433. [Google Scholar]
  19. Kleindienst TE, Jaoui M, Lewandowski M, Offenberg JH, Lewis CW, Bhave PV, Edney EO, 2007. Estimates of the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at a southeastern US location. Atmos. Environ 41, 8288–8300. [Google Scholar]
  20. Malm WC, Schichtel BA, Ames RB, Gebhart KA, 2002. A 10-year spatial and temporal trend of sulfate across the United States. J. Geophys. Res. Atmos 107. [Google Scholar]
  21. Millet DB, Donahue NM, Pandis SN, Polidori A, Stanier CO, Turpin BJ, Goldstein AH, 2005. Atmospheric volatile organic compound measurements during the Pittsburgh Air Quality Study: results, interpretation, and quantification of primary and secondary contributions. J. Geophys. Res. Atmos 110. [Google Scholar]
  22. Pope CA 3rd, Dockery DW, 2006. Health effects of fine particulate air pollution: lines that connect. J. Air Waste Manag. Assoc 56, 709–742. [DOI] [PubMed] [Google Scholar]
  23. Pye HO, Murphy BN, Xu L, Ng NL, Carlton AG, Guo H, Weber R, Vasilakos P, Appel KW, Budisulistiorini SH, 2017. On the implications of aerosol liquid water and phase separation for organic aerosol mass. Atmos. Chem. Phys 17, 343–369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pye HO, Pinder RW, Piletic IR, Xie Y, Capps SL, Lin YH, Surratt JD, Zhang Z, Gold A, Luecken DJ, Hutzell WT, Jaoui M, Offenberg JH, Kleindienst TE, Lewandowski M, Edney EO, 2013. Epoxide pathways improve model predictions of isoprene markers and reveal key role of acidity in aerosol formation. Environ. Sci. Technol 47, 11056–11064. [DOI] [PubMed] [Google Scholar]
  25. Rohr A, McDonald J, 2016. Health effects of carbon-containing particulate matter: focus on sources and recent research program results. Crit. Rev. Toxicol 46, 97–137. [DOI] [PubMed] [Google Scholar]
  26. Schichtel BA, Hand JL, Barna MG, Gebhart KA, Copeland S, Vimont J, Malm WC, 2017. Origin of fine particulate carbon in the rural United States. Environ. Sci. Technol 51, 9846–9855. [DOI] [PubMed] [Google Scholar]
  27. Schwab JJ, Felton H, Demerjian KL, 2004. Aerosol chemical composition in New York state from integrated filter samples: urban/rural and seasonal contrasts. J. Geophys. Res. Atmos 109. [Google Scholar]
  28. Seinfeld JH, Pandis SN, 2016. Atmospheric Chemistry and Physics: from Air Pollution to Climate Change John Wiley & Sons. [Google Scholar]
  29. Solomon PA, Crumpler D, Flanagan JB, Jayanty RK, Rickman EE, McDade CE, 2014. U.S. National PM2.5 chemical speciation monitoring networks-CSN and IMPROVE: description of networks. J. Air Waste Manag. Assoc 64, 1410–1438. [DOI] [PubMed] [Google Scholar]
  30. U.S. EPA, 2008. Integrated Science Assessment for Sulfur Oxides - Health Criteria
  31. U.S. EPA, 2009a. Air Quality Criteria for Particulate Matter
  32. U.S. EPA, 2009b. Integrated Science Assessment for Particulate Matter [PubMed]
  33. U.S. EPA, 2015. AQS (Air Quality System) User Guide http://www.epa.gov/sites/production/files/2015-09/documents/aqs_user_guide_2015.pdf.
  34. U.S. EPA, 2016. Integrated Science Assessment (ISA) for Oxides of Nitrogen – Health Criteria (Final Report)
  35. U.S. EPA, 2017. Profile of Version 1 of the 2014 National Emissions Inventory https://www.epa.gov/sites/production/files/2017-04/documents/2014neiv1_profile_final_ april182017.pdf.
  36. Weber RJ, Sullivan AP, Peltier RE, Russell A, Yan B, Zheng M, De Gouw J, Warneke C, Brock C, Holloway JS, 2007. A study of secondary organic aerosol formation in the anthropogenic-influenced southeastern United States. J. Geophys. Res. Atmos 112. [Google Scholar]
  37. Wittig AE, Anderson N, Khlystov AY, Pandis SN, Davidson C, Robinson AL, 2004. Pittsburgh air quality study overview. Atmos. Environ 38, 3107–3125. [Google Scholar]
  38. Xing J, Mathur R, Pleim J, Hogrefe C, Gan C-M, Wong D, Wei C, Gilliam R, Pouliot G, 2015. Observations and modeling of air quality trends over 1990–2010 across the Northern Hemisphere: China, the United States and Europe. Atmos. Chem. Phys 15, 2723–2747. [Google Scholar]
  39. Zhang Q, Jimenez J, Canagaratna M, Allan J, Coe H, Ulbrich I, Alfarra M, Takami A, Middlebrook A, Sun Y, 2007. Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophys. Res. Lett 34. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Sup 1
NIHMS947234-supplement-Sup_1.docx (5.7MB, docx)

RESOURCES