Air Pollution, the Automobile, and Public Health (1988)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Ambient Levels of Anthropogenic Emissions and Their Atmospheric Transformation Products T. E. GRAEDEL A TAT Bell Laboratories Concentrations of Atmospheric Trace Constituents / 134 Temporal and Spatial Patterns of Primary Gases / 134 Selected Particle Constituents / 139 Photochemical Products and Unregulated Emittants / 143 Indoor Concentrations / 147 Principal Trace Gases / 147 Minor Emittants and Products / 149 Emittants with Potential Global Influence / 151 Carbon Dioxide / 151 Carbon Monoxide / 152 Methane / 152 Summary / 153 Data Adequacy / 153 Trends / 153 Concentrations / 153 Summary of Research Recommendations / 156 Air Pollution, the Automobile, and Public Health. (it) 1988 by the Health Effects Institute. National Academy Press, Washington, D.C. 133

134 Anthropogenic Emissions and Their Atmospheric Transformation Products Concentrations of Atmospheric Trace Constituents Tllc arnounl ana sometimes tne type or effect produced by an atmospheric constit- uent on a receptor- a human being, vege- tation, a building material, for example depends on its concentration. Relating the observed effect to its original cause, pre- dicting effects in new circumstances, and controlling causes to limit the effects all require, in one way or another, an under- standing of concentrations. In a compre- hensive description, the concentration of a species reflects its emission flux Johnson, this volume; Russell, this volume) modi- fied by transport and turbulent diffusion (Sampson, this volume), and physical and chemical transformation and removal (Ar- kinson, this volume). An observed or statis- tically or mathematically predicted concen- tration is the starting point for estimating the flux delivered to a surface or to a tissue during respiration (Sexton and Ryan; Schle- singer; Sun, Bond, and Dahl; Overton and Miller; Ultman; all in this volume). This paper reviews what is known about the measured concentrations of selected at- mospheric trace constituents emitted or evolved from human activity, particularly the operation of motor vehicles. Two cat- egories of environments, within which most of the exposure of human beings and much of the exposure of susceptible mate- rials takes place, are emphasized: outdoor air in urban areas and indoor air within residential and nonindustrial buildings. Actual concentrations are neither uni- form over all space nor constant for all time, although some may be relatively uni- form and stable for long times over large regions whereas others, especially concen- trations of reactive species, those emitted intermittently, and those emitted at high concentration from a few widely spaced points, may be highly variable and/or non- uniform. Every actual measurement effec- tively averages the concentration over some finite volume and finite time interval, and every set of measurements is a distri- bution of such individual measurements at a sample of space/time regions distributed within the overall region of interest. -Or When the effect of an atmospheric species is approximately proportional to concen- tration as well as to duration of exposure, as it is in many cases, average concentration is a good measure of effective concentra- tion. If long-term averages are measured, fewer measurements are required and they are less likely to be contaminated by occa- sional extreme conditions. For these rea-  sons, long-term averages often year-long averages- are considered where possible. In some cases, however, ambient air qual- ity standards are established for extreme values (ozone is an example), so data ap- propriate to the standard are presented. In other cases, only a few "spot" measure- ments rather than an average of any sort are available. Since these give at least some idea of typical concentrations, they are pre- sented with appropriate comments about their restricted validity. Finally, since the health effects of an atmospheric species may be a complex function of time and concen- tration, a table of extremes, ranges, and . .. . . . c ~str~out~ons Is given. Temporal and Spatial Patterns of Primary Gases The primary gases are the principal gases other than water vapor and carbon dioxide (CO2) that are directly emitted from com- bustion sources. Ambient air quality stan- dards have been established in the United States and in several other countries for each primary gas. As a result, extensive mea- surement programs have been initiated to monitor concentrations for most of them, and substantial amounts of data are available. The quantity reported is usually the one named in the standard, most often a long- term average. For some species of gases and some purposes, however, that form may not be the best. For toxicologic purposes, for example, it is often important to know peak short-term concentrations whereas for other purposes it may be necessary to know the variability of concentrations from place to place and from time to time. These peaks and variations may go unreported in data sum- maries from governmental monitoring sta- tions, yet they may be crucial to assessments of air quality effects on people and materials.

T. E. Gracdel 135 Even so, the data for the primary emitted gases are more extensive than those for any other atmospheric constituents. Despite oc- casional inadequacies, the data base is quite sufficient to represent atmospheric concen- trations and predict their consequences. At- mospheric concentrations of several spe- cies, their ranges at different monitoring sites, and their long-term trends have been illustrated with box plots of data from the U. S. National Air Monitoring Stations (NAMS). The technique is shown in figure 1, where the 5th, 10th, and 25th percentiles of the data depict the concentration at the "cleaner" monitoring sites, the 75th, 90th, and 95th percentiles depict it at the "dirtier" sites, and the median and average describe the "typical" concentration. Al- though the average and the median both characterize typical behavior, the median has the advantage of not being affected by a few extremely high or low observations. The major products of motor vehicle fuel combustion are, of course, water and CO2, but their direct impact on human life is not significant. The totality of man-made CO2 emissions, of which motor vehicle emis- sions make up only a small part, do have a measurable global effect, however, which is discussed in a later section. ~ 95th percentile in_ L - - qNth r,rrcentile 75th percentile Composite average ~Median .. -.1 ' 25th Percentile I - 10th percentile l" 5th percentile Figure 1. Technique used for box plots in subse- quent figures in this paper. The 5th, 10th, and 25th percentiles depict the "cleaner" monitoring sites; the 75th, 90th, and 95th depict the "dirtier" sites; and the median and average represent typical concentration. (Adapted from U. S. Environmental Protection  Agency 1985.) 25 1975 1976 1977 1978 1979 1980 1981 ~982 1983 · YEAR Figure 2. Trends in annual second highest nonover- lapping 8-fur average CO concentrations at 174 sites during 197~1983. On this figure and others of its type, NAAQS indicates the U.S. National Ambient (outdoor) Air Quality Standard for the gas species. See figure 1 caption for explanation of plotting tech- nique. (Adapted from U.S. Environmental Protection Agency 1985.) Carbon Monoxide. About two-thirds of all carbon monoxide (CO) emissions come from transportation activities, with the combustion of solid waste and fuel provid- ing most of the remainder (U.S. Environ- mental Protection Agency 1985~. As a re- sult, any reduction in CO emissions from automobiles is reflected directly in the mea- sured CO concentrations. Figure 2 shows the distribution of CO concentrations at 179 sites in the United States for nine years. The quantities re- ported are the second highest 8-fur averages measured in the over 1,000 non-overlap- ping 8-fur periods covering the year. Over the period 1975 to 1983, the median de- creased from about 12 parts per million by volume (ppm) to about 7 ppm, well below the national ambient air quality standard (NAAQS) of 9 ppm. The concentrations at a number of sites exceed the standard, however, with a few at or above 15 ppm. For the foreseeable future, it appears likely that reductions in CO emission will be roughly offset by increases in the total number of vehicles and in the miles trav- eled per vehicle, so that little change in the CO concentration distribution is antici

136 Anthropogenic Emissions and Their Atmospheric Transformation Products 70~ 60 50 Q Q o ~ 40 z UJ o 30 8 Cal 0 20 10 1975 1976 1977 1978 1979 1980 1981 1982 1983 YEAR Figure 3. Trends in annual mean NO2 concentra- tions at 177 sites during 197~1983. See figure 1 caption for explanation of plotting technique. (Adapted from U. S. Environmental Protection Agency 1985.) pated (U. S. Environmental Protection Agency 1985~. Oxides of Nitrogen. The emission flux of oxides of nitrogen (NOX) is approximately equally divided between motor vehicles and stationary combustion activities (U.S. Environmental Protection Agency 1985~. The total NOX (NO + NO2) emissions increased slightly in the late 1970s, de- creased slightly in the early 1980s, and are now relatively stable. As a consequence, major changes in atmospheric concentra- tions of NOX are not anticipated over the next few years. Boxplots of annual mean nitrogen diox- ide (NO2) concentrations at 174 sites in the United States for nine years are shown in figure 3. The quantity reported is the an- nual mean concentration. The 95th, 90th, and 75th percentiles increase from 1975 to 1979, followed by a decrease from 1979 to 1983. The trend is less evident in the mean values of the annual concentrations and disappears entirely for the lower percentiles of the data. Most sites have annual mean concentrations of NO2 in the range of 20-30 parts per billion by volume (ppb). Nearly all are below the NAAQS of 53 ppb. (Los Angeles is an exception; its NO2 levels are in the upper 5 percent of values [not shown in the figure].) The detection technique generally used for NO2 is sensi- tive to several other nitrogenous com- pounds as well, so the data are properly regarded as upper bounds to the true con- centrat~ons. Hydrocarbons. The hydrocarbons (HC) found in the atmosphere comprise an ex-  tremely numerous and chemically diverse group of atmospheric compounds. They are of interest not only for their intrinsic properties but also because, with NOX, they are precursors to ozone (03) and a variety of other atmospheric oxidants. Methane, the most abundant of the HCs, is of limited reactivity; as a consequence, data are often given for nonmethane hydrocar- bons (NMHC) rather than for total HCs including methane. Although there is an NAAQS for NMHC 240 ppb carbon by volume (ppbC), maximum ~9 a.m. concentration, to be exceeded no more than once per year the standard was established to serve only as a guide in assessing HC emission reductions needed to achieve O3 standards. As such it has not been enforced, and only limited routine monitoring of NMHC has been performed. Thus the data on HC concentrations are less extensive than the data for some of the other species men . . . . . tlonec . 1n this section. Typical concentrations of total NMHC in urban areas are about 1-2 ppmC (Graedel and Schwartz 1977; Tilton and Bruce 1980~. Rather than examining total concentrations, however, it is often more instructive to examine concentrations for individual compounds or groups of com- pounds that comprise the NMHC. Atmo- spheric HCs are often grouped into three classes for convenience of discussion: alkanes (aliphatic HCs characterized by a straight or branched carbon chain, generic formula CnH2n+2), alkenes (aliphatic HCs having one or more double bonds), and aromatic compounds (unsaturated cyclic HCs containing one or more rings). For the more reactive alkenes and aromatics, dia- grams of typical concentrations in various

T. E. Graedel 137 1 000 r 100 1 Q 1 0 Q 1 00 Q 1 10 4 Alkenes 2 ~33 ~2 ~3 ~2 5 ~3 MS LS FT D S G F M O U REGIME Figure 4. Approximate concentration ranges of alkenes in the following different atmospheric regimes: U = urban; 0 = oceanic; M = marshland; F = forest; G = grassland; S = steppes and mountains; D = desert (all of the previous measured within the boundary layer); FT = free troposphere (~5 km altitude); LS = lower stratosphere (1~20 km altitude); MS = middle stratosphere (~25 km altitude) (Graedel et al. 1986). Within the flags, the segments indicate the typical fraction of total concentration due to alkenes with the number of carbon atoms shown. For example, oceanic alkenes are typically comprised of about 65 percent ethylene (C = 2) and 35 percent propylene (C = 3). T 10 con E ~ ~ 1 _- 100 ,[ Q Q 1 100 10 Aromatics  ~6 ,777 8 =17 ~6 ~7 ~6 MS LS FT D S G ~7 ~6,7 _ ~6 ~7 =16 . i O U 8 REGIME F M Figure 5. Approximate concentration ranges of the aromatic HCs benzene (6), toluene (7), and xylenes and ethylbenzene (8) in different atmospheric regimes. The regime code and the segmented division of the flags are explained in the caption of figure 4. atmospheric regimes are available; they are reproduced in figures 4 and 5. In each case, concentrations in urban areas can be as high as about a thousand ppbC, and concentra- tions of several hundred ppbC are com- mon. In more remote regions the measured concentrations are sharply lower, reflecting lack of proximity to the principal sources as well as diminution of concentrations as a consequence of atmospheric reactions. The U. S. Environmental Protection Agency (EPA) (1985) estimates that 37

138 Anthropogenic Emissions and Their Atmospheric Transformation Products as, 30 ~ as o 20 tar he LU o 15 o Cat 0 10 CO 5 n NAAOS LL _ r _ 1975 1976 1977 1978 1979 1980 1981 1982 1983 YEAR Figure 6. Trends in annual mean SO2 concentra- tions at 286 sites during 197~1983. (Adapted from U.S. Environmental Protection Agency 1985.) percent of the volatile organic carbon com- pounds in the atmosphere come from mo- tor vehicles, 37 percent from industrial activities, 15 percent from solid waste and miscellaneous, and 10 percent from volatil- ization of organic solvents. The total flux of these emissions decreased slightly over the period 1975-1983 and is now believed to be relatively stable. Sulfur Dioxide. Sulfur dioxide (SO2) is a trace gas of substantial concern because of its acid-forming potential in the atmo- sphere and because of its potential health effects. Its emission is dominated by fossil fuel combustion, with industrial activity being much less important and with motor vehicles emitting only a few percent of the total SO2 flux. As increased controls and cleaner fuels have been used over the past decade, SO2 concentrations have steadily decreased. This pattern is illustrated in fig- ure 6 for annual mean concentrations mea- sured at NAMS. As of 1983, typical annual mean concentrations were about 9 ppb. Spatial and Temporal Variations in Con- centration. Any effort to assess the effects of atmospheric trace gases on animate or inanimate objects must take into account the very great spatial and temporal variabil- ity of the trace gas concentrations. If the  effects are cumulative and roughly propor tional to concentration and duration of exposure, the problem may not be severe. If, on the other hand, the effects are strong functions of peak intensity, some quantita- tive knowledge of variation from the mean value is necessary. The present review can offer no general rules to relate, for example, peak intensities to annual averages. What can be done is to illustrate some typical variations to provide some perspective on the magnitude of the problem. Throughout the world, the concentra- tions of atmospheric trace gases demon- strate the degree to which sources of emis . . . . . ~ salons are present In t he vicinity ot t he monitoring sites, to what degree emissions from those sources are controlled, and to what degree the local meteorological situ- ation influences the measured values. In figure 7, annual averages of SO2 concentra- tions at various cities throughout the world are displayed. The ordinate on figure 7 is logarithmic, and the figure shows that the ratio of average SO2 concentration at Milan (the site with the highest average annual concentration) and at Aukland (the site with the lowest) is about 16. The concen- tration also varies widely within cities be- cause of local meteorological conditions. In a typical situation described by Johnson et al. (1973), circulation patterns in an urban street canyon concentrate emissions at cer- tain locations within the canyon. The result is concentration patterns in which assess- ments made on one side of a city street can easily differ by 50 percent or more from similar assessments made on the other side. The variation of concentration with time, as in space, can also be significant. In figure 8, the seasonal patterns in total HC (NMHC plus methane) concentration in Camden, New.lersey, are shown. The dif- ference between the June value (the lowest) and the September value (the highest) is about 0.4 ppmC, or more than 20 percent of the mean value of about 1.9 ppmC. Diurnal differences can be important as well. In figure 9, 10 years of August total HC data are presented in hour-by-hour format. The difference between the 7 a.m. high value and the 4 p.m. low value is about 1.0 ppmC, that is, half the total concentration. Seasonal and diurnal fluctu- ations in concentration vary with location,

T. E. Graedel 139 500 300 Q Q - O 100 at UJ o o ~ 10 50 30 3 l Auckland Amsterdam Y Montreal London Figure 7. The range of annual averages of SO2 concentrations measured at multiple sites within cities throughout the world. Several cities are identified here; further information is available in Bennett et al. (1985). The asterisk within each bar is the composite average for the city. (Adapted with permission from Bennett et al. 1985, and the American Chemical Society.) species, and time of year, but those illus- trated here are typical. Because of the degree to which micro- meteorology or meso- or synoptic-scale meteorology can influence species concen- trations at any particular location, many species concentrations will tend to be in phase with each other. This situation has important consequences for studies at- tempting to link epidemiologic effects to atmospheric concentrations, for it requires that all potentially significant atmospheric variables be monitored simultaneously. Field measurements of such complexity have seldom been accomplished by epide . . mlo. OglStS. Recommendation 1. No analytical in- strument is readily available for routine  monitoring of NMHC concentrations and concentration trends, although several techniques are available for potential incor- poration into such an instrument. It is extremely important to achieve agreement on a satisfactory monitoring technique for NMHC (or some significant fraction thereof) and to begin to acquire data on a . ~ . routine Casts. Selected Particle Constituents Elemental Carbon. Carbon comprises 10 to 20 percent by weight of urban aerosols (Countess et al. 1980; Wolff et al. 1982b). Of this amount, nearly half is present as elemental carbon (soot), a consequence of incomplete combustion of fossil fuels. Die- sel engines are the most significant of all the sources of elemental carbon (Wolff and Klimisch 1982~. Extensive monitoring data are not available for elemental carbon, but the limited studies that have been per- formed suggest concentrations ranging from 1 to 35 ,ug/m3 and averaging about 7 ,ug/m3 (see, for example, Countess et al. 1980; Wolff 1985~. The principal concerns with regard to soot are its efficient reduc- tion of atmospheric visibility (Rosen et al. 1978), its potential as a catalytic oxidizer of SO2 and other compounds in atmospheric droplets (Brodzinsky et al. 1980), and its ability to adsorb and concentrate toxic or- ganic compounds and carry them into the lung (Sun, Bond, and Dahl, this volume). Organic Carbon. Organic compounds are also a significant fraction of the urban

140 Anthropogenic Emissions and Their Atmospheric Transformation Products atmospheric aerosol. The data of Grosjean and Friedlander (1975), Countess et al. (1980), and Wolff (1985) indicate that or- ganics comprise between 2 and 40 ,ug/m3 annually (mean value of 1 ~15 ,ug/m3) of the aerosols in urban areas (see also Na- tional Academy of Sciences 1972~. Motor vehicle emissions are responsible for perhaps somewhat less than half this amount. The diurnal patterns of reactive aerosol constituents in urban areas have been dem E Q c,' 0.2 en LL Or G C, 0 0. 1 a: o IL o o us J z <t: 0 - 0.1 cr At: c, - z -0 2 o o G ~ -0.3 onstrated by the data of Grosjean and Friedlander (1975~. In figure 10, these pat- terns are shown for a day in Pasadena, California. During this period, the change in concentration of total suspended partic- ulates followed that of O3, suggesting that smog chemistry is directly involved in the generation of the urban aerosol. Two fea- tures in the data are worth noting. The first is that the organic component of the aero- sols is the largest of the four reactive com Monthly midmean of daily mean TOO 4  1 . ..... Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec MONTH Figure 8. Derived seasonal dependence of the monthly midmeans of daily mean total organic carbon (TOC) concentrations (ppmC) in Camden, New Jersey, from 1968 to 1977. These data, with the long-term trend removed, are grouped into monthly sets, and the midmean of each set is determined. That value is plotted as a horizontal line. For each month, the excursions from the midmean are then determined for each of the years in the data sample and plotted as lines perpendicular to the midmean. The overall mean is approximately 1.9 ppmC. The seasonal trend for the entire time series can thus be seen, as well as the trend from year to year during each month. (Adapted with permission from Graedel and McRae 1982.)

T. E. Graedel 141 0.6 LU C) G by lo: ILL - ~ O G I o lL G ~-0.2 C] -0.4 Monthly midmeans of hourly TOO (August) [ill 1 ~ ~ ~ ~ ~ ~ I I I I I 1 1 1 1 'I I I I ~ I I I  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 HOUR OF DAY Figure 9. Diurnal pattern of the seasonal component of total organic carbon (TOC) data (ppmC) in Camden, New Jersey, for August days from 1968 to 1977. The overall August midmean is about 2.0 ppmC. The plotting technique is explained in the caption to figure 1. (Adapted with permission from Graedel and McRae 1982.) portents measured. (It is not always the largest component in all cities but is nearly always one of the major components.) The second is that the four aerosol constituents, all of which are produced by smog chem- istry, comprise very large fractions of the total particle mass. The remainder of the mass, which largely consists of soot and soil dust, is important only during periods unfavorable for atmospheric chemistry. Metals. The concentrations of metals in atmospheric aerosol particles are monitored routinely in the United States and in other areas throughout the world. Most of the metals are primarily by-products of various industrial processes. Motor vehicles, how- ever, have historically been major sources of lead (Pb) because of the use of Pb compounds as antiknock additives to gas- oline. Manganese compounds are now be- coming used for this purpose as well. It is therefore of interest to examine the atmo- spheric concentrations and trends of these two metals. An NAA`2S exists for Pb, and the result has been a careful study of its atmospheric abundance. Figure 11 shows the NAMS Pb

142 Anthropogenic Emissions and Their Atmospheric Transformation Products r I I , . . . . 100 ~ o O 8C UJ jig 6C o he 2C NH4 ~ @~ so24 ~ _ NO3 ~~ l Organics ~ _ 6:30 12:30 16:30 TIME (PDT) _~___ O Figure 10. Diurnal patterns of nitrate, sulfate, and ammonium ions, total aerosol organics (as weight percent of total dry aerosol), and O3 (-I-, ppm) in Pasadena, California, on July 25, 1973. The aerosol constituents were determined from seven 1-fur sam- ples taken in late morning and throughout the after- noon. (Adapted with permission from Grosjean and Friedlander 1975, and the Air Pollution Control Association. ) data for the past decade. It is easy to see that the substantial reductions in leaded gasoline that have occurred during this period have been reflected in sharply decreasing atmo- spheric concentrations. As of 1983, the mean Pb level was about 0.3 ,ug/m3. Nitrate. The inorganic nitrate component of atmospheric aerosols is directly related to emissions of gaseous NOX. Stationary combustion sources and motor vehicles may thus be supposed to be about equally responsible for aerosol nitrate; within ur- ban areas, that attributable to motor vehi- cles probably dominates. Typical concen- trations of nitrate fall within the range 1-10 ,ug/m3, with a mean annual value of about 4 ,ug/m3 (Graedel and Schwartz 1977; Har- rison and Pio 1983~. The diurnal behavior of urban aerosol nitrate is shown in figure 10. The largest concentrations tend to oc- cur in the late morning, a circumstance that Grosjean and Friedlander (1975) attribute to a combination of rush-hour emissions of NOX and rapid smog chemistry.  Sulfate. Urban aerosol sulfate concentra- tions are typically in the range 1-20 ,ug/m3, with a mean annual value of about 7 ,ug/m3 (Graedel and Schwartz 1977; Harrison and 0.6 Pio 1983~. The sulfate concentration in o urban areas often increases during the day °~1, (see figure 10) as a result of the conversion ~of SO2 to sulfate during periods of high 0.2 ~photochemical activity. As with SO2, sta tionary-source combustion of fossil fuel is the primary cause of aerosol sulfate in urban areas. Ensemble Measurements of Suspended Particulate Matter. Total suspended par ticulate matter (TSP) in the atmosphere (that is, the atmospheric aerosol) is the most noticeable of the emittants from sources of atmospheric constituents. As such, it has been the object of substantial mitigation and reduction efforts over the years. As of 1983, emission fluxes from industrial processes and fuel combustion comprised about 30 percent each of the total emission flux. Motor vehicles and "solid waste and miscellaneous" contrib uted about 20 percent each (U.S. Environ mental Protection Agency 1985~. Concentrations and trends in TSP over the past decade are illustrated in figure 12. A very significant reduction in the higher - E ~ i,1.~ ~ 2.( 6 E ~ 1.( ~ o .ol 1982 1983 YEAR Figure 11. Trends in maximum quarterly Pb levels at 61 sites, 197~1983. See figure 1 caption for expla- nation of plotting technique. (Adapted from U.S. Environmental Protection Agency 1985.)

T. E. Graedel 143 percentiles of the data is seen, with the result that virtually all U. S. sites now meet the NAAQS for annual concentrations. The mean and percentiles near the median have fallen proportionately somewhat less than the higher percentiles; although, over the last two years represented in the data, noticeable reductions in these TSP concen- trations occurred as well. A feature that makes TSP data far from ideal indicators of health effects is that they are strongly influenced by high concentra- tions of very large particles (diameters 15 lam), yet such particles are too large to be readily inhaled. Accordingly, many recent observations have been made with instru- ments designed to reject particles larger than 10 or 15 ,um in aerodynamic diameter. The remaining particles are designated "in- halable particles" (IP). During normal breathing, these particles may travel to the bronchi and be retained there. Often the IP are further differentiated experimentally, the particles with aerodynamic diameter < 2.5 ,um being known as the "respirable particles" (RPs). RPs travel as far as the lung parenchyma during normal breathing and may be retained there. Extensive IP and RP data are not avail- able, but it is clear that the chemical com ~o Coo 90 me 80 Cal `3 70 He ,0 60 6 z 50 ° 40 o o So In 20 10 o LN0ASA ~ I 1975 1976 1977 1978 1979 1980 1981 1982 1983 YEAR Figure 12. Trends in annual geometric mean total suspended particulate (TSP) concentrations at 1,510 sites, 197~1983. See figure 1 caption for explanation of plotting technique. (Adapted from U. S. Environ- mental Protection Agency 1985.) position of larger particles is substantially different from that of the smaller ones. Elements residing on larger particles are emitted mainly by natural processes such as  crustal erosion and sea spray. Those on smaller particles are commonly generated by high-temperature anthropogenic pro- cesses such as welding or soldering, smelt- ing of metals, and combustion of fossil fuels (Milford and Davidson 1985~. Thus sulfate (Wolff et al. 1985a), organic carbon (Wolff et al. 1982a), and elemental carbon (Wolff et al. 1982b) are among the species concentrated on RPs. The potential health effects of the RPs are much more significant than those of IPs, because of their chemical differences as well as the deeper respiratory system penetration that is characteristic of the smaller particles. · Recommendation 2. A difficult but es- sential job is to monitor the chemistry of atmospheric aerosol particles in much more detail than is now being done, concentrat- ing especially on chemical differences as a function of particle size. Detailed organic analyses are particularly important. Photochemical Products and Unregulated Emittants The atmospheric species discussed thus far include those that are directly emitted from combustion sources and extensively moni- tored, either because they are the subject of a standard or as possible preparatory efforts in the establishment of a standard. In this section an attempt is made to discuss other compounds that may be of concern. One cannot begin to comment on all possible atmospheric compounds, however, be- cause the total number thus far identified exceeds 2,800 (Graedel et al. 1986) and many are demonstrably unimportant. To render the discussion tractable, the follow- ing criteria have been used to select com- pounds for attention: · High chemical reactivity · Positive toxicologic test results · Known effects, other than toxicologic, on humans · Known effects other than those con- nected directly with human organisms.

144 Anthropogenic Emissions and Their Atmospheric Transformation Products 2s0r Q Q o 15C an C) 100 8 so o 1 1975 1976 1977 1978 1979 1980 1981 1982 1983 YEAR Figure 13. Trends in annual second highest daily maximum 1-fur O3 concentrations at 176 sites, 197~1983. See figure 1 caption for explanation of plotting technique. (Adapted from U.S. Environmen- tal Protection Agency 1985.) Inorganic Compounds. O3 is the chemical species of most extensive interest in this group. It is not emitted directly from sources but is formed in the atmosphere by the gas-phase chemistry of NOX, small organic molecules, and sunlight. O3 is itself toxic and is the major oxidizing gas in photochemical smog mixtures. The concentrations of O3 at U.S. mea- surement sites are indicated in figure 13. The O3 standard specifies the concentration at the second highest daily maximum dur- ing a year; therefore, the data in the figure correspond to near extreme, rather than average, values. The data suggest some decrease in concentrations since 1975 (per- haps as a result of a calibration change in O3 monitors in 1979), but the increase in 1983 makes it clear that no monotonic trend exists. It is particularly significant that the majority of sites included in the data set have continued to exceed the NAAQS for o3. The influence of anthropogenic emit- tants on O3 can be seen in the remote troposphere, far from populated regions. A careful data analysis by Logan (1985) sug- gests an increase of 20-100 percent in the annual average O3 concentration in those regions since the 1940s. The increase is attributed to photochemical production as- sociated with increased levels of NOX, HCs, and CO. Other unregulated inorganic compounds worth comment are the oxy acids of nitro- gen, which are among the final products of the chemistry of combustion-related emis- sions. Nitric acid (HNO3) is the more abundant, with concentrations of a few ppb being common in urban areas. HNO3 lev- els as high as a few tens of ppb are occa- sionally observed (Miller and Spicer 1975; Tuazon et al. 1981~. Nitrous acid (HNO2) concentrations are lower: they are often below 1 ppb, but occasionally peak near 10 ppb (Harris et al. 1982; Sjodin and Ferm  1985~. A fourth inorganic atmospheric species of great interest is hydrogen peroxide (H2O2), a product of atmospheric hydro- carbon chemistry. Its importance is due to its strong oxidizing capacity, which is thought to be of crucial importance to the chemistry of atmospheric water droplets. This species is exceptionally difficult to measure in the gas phase, so extensive data are not available. It appears, however, that urban concentrations of a few ppb are typical during the summer months (Kok et al. 1986~. Aldehydes. Formaldehyde and other al- dehydes are common atmospheric constit- uents. They irritate the eyes and upper respiratory tract, but not generally at the outdoor concentrations experienced in ur- ban areas. The aldehydes are directly emit- ted from a number of sources, including motor vehicles, but are also produced eff~- ciently from small HCs by atmospheric reactions involved in smog chemistry. Measurements of the atmospheric con- centrations of formaldehyde are reasonably extensive, although no routine monitoring programs exist. Typical formaldehyde con- centrations in urban areas in the summer are 5-20 ppb, although higher concentra- tions are occasionally observed (National Academy of Sciences 1981~. Acetaldehyde concentrations are generally about a third those offormaldehyde (Hoshika 1977; Tan- ner and Meng 1984~. Outdoor urban concentrations of form

T. E. Graedel aldehyde appear to reflect direct sources as well as atmospheric photochemistry. This is seen in figure 14, which shows diurnal formaldehyde concentrations in Newark, New Jersey, together with those for CO (an indicator of motor vehicle emissions) and O3 (an indicator of photochemical ac- tivity). The signatures of both indicators appear to be reflected in the formaldehyde data, as is also the situation with other data sets illustrated by Cleveland et al. (1977~. The published information also demon- strates clearly that formaldehyde has a strong seasonal pattern, though not as strong as that of O3 (see figure 15~. A similar seasonal pattern exists for acetal- dehyde (Tanner and Meng 1984~. Polynuclear Aromatic Hydrocarbons. Poly- nuclear aromatic hydrocarbons (PAH), a major fraction of atmospheric polycyclic 15 12 Q UJ C] ~ 9 I 9 ~ 6 o 3 _ O _ 7.5r 6.0t ~= 4.5 o C' 3.0 1 5 o _ ~ 0 4 145 25 10 Q Q A o ~ - o it IIJ  O 40 8 30 20 Formaldehyde Bayonne O . i': ~ ~ . Newark 1 1 1 11 11 1 1 00 Newark , (_ d - 80 I \~; 1 i i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 MONTH Figure 15. Monthly plots of daily averages of form- aldehyde and O3 in the New Jersey cities of Bayonne and Linden, respectively. Each index number cor- responds to a month: 1 = May 1972; 2 = June 1972, . . . , 17 = September 1973. Unlike the box- plots shown earlier, the 5th and 95th percentiles and the composite average are omitted here. (Adapted from Cleveland et al. 1977.) \ ~ jo3 \ i. 1 ~ , 1 1 8 12 16 20 24 HOUR OF DAY Figure 14. Formaldehyde, CO, and O3 diurnal con- centrations at Newark, New Jersey, for days within the periods dune 1, 197~August 31, 1973 end dune 1, 1970August 31, 1974 on which the daily maximum O3 concentrations exceeded 100 ppb. (Adapted from Cleveland et al. 1977.) 1 organic matter (POM), can be formed and/or directly released from any combus- tion process involving hydrocarbons. Power generation and refuse burning are the two largest sources; motor vehicle emissions are thought to contribute no more than a few percent of the total flux. In urban areas, of course, the motor vehicle contribution is proportionately higher. Because of their intermediate vapor pres- sures, PAHs are generally present in the 0 0 atmosphere in gaseous form and as particle == components. Gas-phase PAH, of which 0 ~ fluorene, phenanthrene, and fluoranthene are among the more abundant, typically are  present in urban air at concentrations of a few parts per trillion by volume (ppt) (Krstulovic et al. 1977; Cautreels and Van Cauwenberghe 1978~. In the solid phase, where pyrene, benzo~a~pyrene, chrysene, and benzotyhilperylene are among the more abundant compounds, the PAH con centrations in urban areas are typically 10-20 ng/m3 (the equivalent of about 1-2 ppt for a PAH of molecular weight 200~.

146 Anthropogenic Emissions and Their Atmospheric Transformation Products Routine monitoring and speciation of PAHs are seldom performed, and any as- sessment of their abundance in specific ur- ban areas generally requires special mea- surement programs. Alcohols. Alcohols are of interest in this review because of the potential for rapidly increasing use in gasohol fuel blends for motor vehicles. If such use should occur, the unburnt alcohols that are emitted will participate in a number of atmospheric chemical processes including the produc- tion of organic acids. To this point, few determinations of atmospheric alcohol concentrations have been performed. Those that have suggest a typical concentration of perhaps ~10 ppb in urban areas (Bellar and Sigsby 1970; Hanst et al. 1975; Holzer et al. 1977; Snider and Dawson 1985~. The alcohol concentra- tions do not follow regular diurnal pat- terns, and Hanst and his colleagues (1975) suggest that they are related to random emission from direct sources rather than to generation by atmospheric photochemis- try. It is important to generate baseline data on the atmospheric compositions of the most abundant alcohols now, before the alcohol emission fluxes from motor veh~- cles increase markedly (if they do). Nitro Compounds and Organic Nitrates. Nitro compounds are those in which the NO2 group is attached to an organic mol- ecule. If the NO2 group is attached to an oxygen atom, the compound is an organic nitrate. Many of these species are of interest because of their toxicity (National Acad- emy of Sciences 1977) and because they have been shown to be harmful to certain types of vegetation (Taylor 1969~. The most widely measured of this group is peroxyacetyl nitrate (PAN). It is formed by chemical reactions involving small HCs and NOX and its concentrations therefore tend to follow, in pattern if not in magni- tude, those of indicators of smog activity such as O3 (National Academy of Sciences 1977~. Typical urban concentrations of PAN are ~10 ppb (Spicer 1977; Tuazon et al. 1981~. Since PAN is a mutagen and powerful eye irritant and can be measured by relatively straightforward experimental techniques, its regular monitoring would seem to be advisable. A number of other gas-phase nitrogen . . . . containing organlcs occur In t he atmo- sphere (Graedel et al. 1986), but their con- centrations are very low. Most atmospheric nitro compounds occur in the solid phase at concentrations of a few tenths of a ng/m3 in  urban areas (Pitts et al. 1985b). Several direct mutagens are included in the species that comprise the solid-phase nitrated poly- nuclear aromatics. Heterocyclic Organic Compounds. Het- erocyclic organic compounds are those in which an atom other than carbon is present in one or more ring structures. A number of such compounds have been shown to be potent mutagens (see Graedel et al. 1986), so their atmospheric concentrations are of interest. Few analyses have been made, however, either in the gas or aerosol phases. Heterocyclic nitrogen compounds, of which the most abundant are perhaps pyridine, indole, and quinoline, have urban concentrations of a few ppt (gas phase) or a few ng/m3 (particle phase) (Dong et al. 1977; Ketseridis and Jaenicke 1978~. Het- erocyclic oxygen compounds, of which the most abundant appear to be furan, dioxane, coumarin, and their derivatives, have urban concentrations of about 10 ppt in the gas phase, and 0.10 ng/m3 in the aerosol phase (Louw et al. 1977; Ketseridis and Jaenicke 1978; Harkov et al. 1983~. In all cases, these concentration estimates are based on very limited data. If major concerns should arise concerning these substances, much addi- tional information would be needed to en- sure that a satisfactory picture of their urban concentrations was at hand. Recommendation 3. An effort to de- velop a monitor for either formaldehyde or the aldehyde group with appropriate sensi- tivity (100 ppt or less), reliability, and an appropriately low cost, is a high priority for atmospheric chemists, particularly those addressing problems in urban air. · Recommendation 4. More attention should be given to monitoring unregu

T. E. Graedel 147 lated, but potentially hazardous species, since extensive data will be required should regulatory action prove desirable. Among the species to which increased effort might be directed are PAN, nitric acid, hydro- chloric acid, nitro-PAHs, and heterocyclic organics. ~ Recommendation 5. Given the possibil- ity of sharply increased use of methanol and ethanol in motor vehicle fuels over the next decade or two, it is important to begin promptly to establish a satisfactory baseline against which future changes in atmo- spheric alcohol concentrations could be as- sessed. Indoor Concentrations Principal Trace Gases Most people spend most of their lives in- doors in homes, offices, automobiles, and other enclosed places. As a consequence, the quality of indoor air may be of more concern from a public health standpoint than that of outdoor air quality. Many vulnerable materials electronic compo- nents, fabrics, or works of art are exposed to the atmosphere, mainly under cover. Nonetheless, the data base for assessment of indoor air quality is smaller than that for ambient air. Citywide average indoor air quality levels are clearly not appropriate or meaningful, because the available data in- dicate quite strongly that the concentra . , . . . . lions ot trace species m 1nc oor ocat1ons vary greatly within cities, within neighbor- hoods, and within buildings. For many trace gases, the concentrations in indoor air reflect indoor sources rather than infiltration of air from outdoors. For a few, injection of outdoor air appears to control indoor concentrations. In this sec- tion, measured indoor concentrations of species of interest are presented, as well as a summary of what is known about the controlling sources. Further information is available in the detailed and authoritative book Indoor Pollutants (National Research Council 1981), as well as in reports by Fanger and Valbjorn (1979), Aurand et al. (1982), Spengler et al. (1982), and Berglund et al. (1986~. Carbon Dioxide. CO2 is a natural con- stituent of the atmosphere, with typical concentrations of about 350 ppm in remote areas and higher concentrations near popu- lation centers. It is a principal product of fossil fuel combustion, and any indoor combustion source that is not well vented will give rise to heightened CO2 levels. Differences in building construction and  indoor sources result in wide variations in CO2 indoor concentrations. The minima generally correspond to that outdoors (Yocom 1982; Spengler and Sexton 1983), whereas maxima of 3,000 ppm are not too uncommon. Occasionally the peak values may reach the Occupational Safety and Health Administration exposure standard of 5,000 ppm (World Health Organization 1983~. Carbon Monoxide. Indoor concentrations of CO tend to be lower than those out- doors, but not markedly so (Ott and Flachsbart 1982~. Urban indoor CO con- centrations of a few ppm are thus common, and higher concentrations are sometimes seen. Two types of indoor environments are particularly likely to have high CO concentrations: indoor garages or buildings with attached indoor parking areas (Ott and Flachsbart 1982) and residences with improperly ventilated space heating equip- ment (National Research Council 1981~. In these situations, concentrations as high as 100 ppm have been measured. High con- centrations can also be produced as a con- sequence of cigarette smoking. If obvious indoor sources are absent, indoor CO concentrations tend to follow outdoor CO concentrations but to be somewhat lower. The exposure of the pop- ulation to CO thus has a lower limit set by the outdoor concentrations of CO. The concentration variations indoors lag behind those outdoors in time. Oxides of Nitrogen. NOx exists indoors as a consequence of both indoor and out- door sources. Indoors, these sources consist largely of poorly ventilated products of -7

148 Anthropogenic Emissions and Their Atmospheric Transformation Products burnt gas. Gas stoves are by far the most as common problem. Somewhat elevated NOX levels may also be seen in buildings with unvented space heaters, gas-fired hot water heaters, or gas-fired clothes dryers. Many of the concentration dependencies have been studied by Moschandreas et al. (1980), whose data show that gas-heated residences generally have NOX concentra tions greater than those outdoors, whereas electrically heated residences tend to mimic outdoor NOX conditions. Typical NOX concentrations are 1~0 ppb (Spengler et al. 1979~. Office buildings tend to be effi ciently ventilated and to have heating sys tems well removed from office areas, so that the type of heating is often not a factor in their indoor NOX concentrations. Hydrocarbons. Indoor environments are rich sources of HCs and oxidized organic molecules, with several hundred different organic compounds known to be present in certain indoor environments Jarke et al. 1981; Graedel et al. 1986~. The most abun dant HC compounds, at least in residences, are alkanes, alkylbenzenes, and terpenes (M¢lhave and M¢ller 1979~. Indoor concentrations of gaseous NMHCs span a measured range of 0.00~0 ppm, with average concentrations in newer homes being of the order of 12 ppm and those in older homes being of the order of 1 ppm (M¢lhave and M¢ller 1979; Wallace et al. 1982~. Building materials and a host of other indoor sources appear to be responsible for many of the observed organic substances. In view of the observed infiltration of outdoor species such as CO and NOX, it appears likely that outdoor concentrations will gen erally establish a lower bound to indoor NMHC concentrations of a few tenths of a ppm or higher in central urban areas. Ozone. Except for occasional generation by electric arcing, photocopiers, and in door ultraviolet light sources (Committee on Indoor Pollutants 1981), any O3 indoors is present as a result of the injection or infiltration of outside air. O3 iS readily absorbed and destroyed on surfaces and, as a result, its indoor concentrations tend to be low. The National Research Council E 0-4 Q O 0.3 an 0.2 o CO o 0.1 _~\ _ O _  I 1330 1 400 it, ---1 ~ ~Indoor // \\ / ~ / ~-_ _ 1 1 1430 1500 1530 TIME (PDT) Figure 16. Relative indoor and outdoor concentra- tions of O3 in the Spaulding Laboratory, California Institute of Technology, July 22, 1975. Boxed areas indicate times when the building's charcoal filtering system was in operation. (Adapted from Committee on Indoor Pollutants 1981.) (1981) suggested an anticipated range for indoor O3 of 20-200 ppb, with low values in this range being much more common than high ones. The relationship between indoor and outdoor O3 during smoggy conditions is shown in figure 16. This figure demon- strates that indoor O3 reflects outdoor con- centrations, but at a lower level, and that charcoal filtration is effective in removing O3 from indoor environments. Similar findings in a number of geographical loca- tions have been reviewed by Yocom (1982~. Sulfur Dioxide. In the absence of indoor sources, SO2 concentrations are normally 0.3 to 0.7 times those outdoors, although air conditioning or filtration can reduce this factor substantially (Yocom 1982~. Within a residence in which sulfur-containing fuel (such as coal, wood, or kerosene) is burned, however, inadequate venting of exhaust gases can result in high local SO2 concentrations. Indoor SO2 levels of a few ppb are typical (Spengler et al. 1979), but concentrations as high as several hundred ppb have been observed in special cases (World Health Organization 1983~. As with outdoor SO2, there is little or no evidence for a significant contribution by motor vehicle emissions.

T. E. Graedel 149 Concentrations of Principal Trace Gases Within Motor Vehicles. A farce number of people spend one or more hours a day within automobiles, buses, and other mo- tor vehicles. Notwithstanding this fact, in- formation on concentrations of trace gases within motor vehicles is sparse indeed. Some data on CO within automobiles in- dicate that these concentrations can exceed 100 ppm if vehicle exhaust products are allowed to intrude into the passenger com- partment. Similar tendencies are suggested by the data of figure 17, which show high levels of interior CO in a vehicle moving slowly or halted in dense traffic. HCs show similar behavior (Mucke et al. 1984), and concentrations of NOx and small particles would be expected to mimic those of CO as well. Additional data are needed to assess more fully the contribution of exposure in the passenger compartments of motor ve- hicles to total exposure burdens. Personal Air Quality Monitors. A recent development in the assessment of exposure to atmospheric trace species is the introduc- tion of personal exposure monitors. These instruments are designed to be carried or worn throughout one's normal activities, and thus are limited in power, size, and 50 40 at ON 30 z us 8 20 10 weight. In table 4 of their chapter in this volume, Sexton and Ryan summarize the direct personal monitoring studies that have been performed in the United States. The potential benefits from the use of such instruments is substantial, but much re- search is needed to develop suitable tech- niques for other species of interest and to . . . . miniaturize t le equipment. Recommendation 6. Personal exposure monitors have important future roles to play in health effects research. Instrument development is required, however, to im- prove portability, reliability, and the quan- titative detection of many atmospheric spe- cies of interest not presently capable of being monitored in this way.  Minor Emittants and Products Inorganic Compounds. As with outdoor air, one would like to have information on the indoor concentrations of nitric and ni- trous acids, hydrogen peroxide, and per- haps other inorganic gases and aerosol con- stituents. To my knowledge, however, no such measurements have ever been made in unperturbed indoor environments. A re- lated study of interest is that of Pitts et al. Morning commute PASS GAS LINES 1495 ~ . _ ROUTE 664 ROUTE 215t MANASSAS,] 0EVANS ~ ~ ,~ 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 HOUR OF DAY Evening commute WALK - IN EPA GARAGE DRIVE- I N GA R. A GE COLLECTION BOOTH SW FREEWAY 1495 I ROUTE 66 IN OFFICE ~ MANASSAS, VA Figure 17. Daytime CO exposure profile of an Environmental Protection Agency employee. The employee's location in the Washington, D.C., area is given for points of interest on the diagram. The detector was not operated during the periods indicated by dotted lines. (Adapted with permission from Ott 1985, and the American Chemical Society.)

150 Anthropogenic Emissions and Their Atmospheric Transformation Products (1985a), who injected NO2 into a humid indoor space and measured the formation of nitrous acid. The results suggested that indoor spaces with high concentrations of NO2, such as can occur near gas stoves, might also be areas in which a few ppb of HNO2 can be found. Much more study of the indoor nitrogen cycle in the presence of strong sources is needed. Some data are available on the concen- trations of major ions on indoor aerosols. Within fine particles (those with diameters <2.5 lam), Sinclair and coworkers (1985) found ammonium, sulfate, and nitrate to be predominant, whereas within coarse parti- cles, they found calcium and nitrate con- centrations to be highest. Typical average concentrations and peak values in office buildings were SO42-, 0.1 and 0.3 ,ug/m3; NO3-, 0.1 and 0.2 ,ug/m3; NH4+, 0.2 and 0.3 ,ug/m3; and Ca2+, 0.05 and 0.2 ,ug/m3. Aldehydes. Formaldehyde (and to a much smaller extent, other aldehydes) is one of the major indoor air quality concerns. The current understanding of aldehyde sources is described in a recent report by the Na- tional Research Council (1981~: "The primary sources (of formaldehyde and other organic substances) are in the indoor environment itself building mate- rials, combustion appliances, tobacco smoke, and a large variety of consumer products. A buildup of formaldehyde may be exacerbated in buildings that have been subjected to energy-eff~ciency measures in- tended to reduce infiltration and, thus, en- ergy consumption. Emission rates for formaldehyde and other organic pollutants emitted in the indoor environment are gen- erally unknown." More information on sources of aldehydes and their effects has been reported by the National Academy of Sciences (1981~. The indoor concentrations of formalde- hydo the only aldehyde for which any significant amount of data is available vary greatly. They can be negligibly small in buildings that contain few or none of its common indoor sources. Conversely, in buildings such as new, well-insulated mo bile homes, concentrations may be as high as several ppm (National Research Council 1981; Hanrahan et al. 1985), although sev- eral tenths of a ppm is a more typical value (Bundesgesundheitsamt 1985; Sexton et al. 1986~. Acetaldehyde concentrations are typically much lower (Wang 1975; DeBor- toli et al. 1984~. There is no indication that formaldehyde from infiltrated outdoor air plays any significant role in establishing  indoor formaldehyde concentrations unless there are no indoor sources whatever. Alcohols. Very limited data exist on the indoor concentrations of alcohols. Wang (1975) detected several alcohols in a college classroom and deduced that the sources were indoor rather than outdoor. Berglund and coworkers (1982) measured butanol concentrations in a school and attributed the presence of butanol to the vaporization of solvent from building materials. In both cases, concentrations of 5-50 ppb were observed. Nitro Compounds and Organic Nitrates. Few studies of indoor nitro compounds have been conducted. Thompson and co- workers (1973) examined indoor and out- door concentrations of gaseous peroxy- acetyl nitrate (PAN) at several sites in the Los Angeles Basin. They found PAN levels of a few ppb, always lower indoors than outdoors, and attributed them to the infil- tration of outdoor air. Seifert and co- workers (1984) detected amines and nitro- samines, the former at levels as high as a few hundred ppt, the latter at levels 10 to 100 times smaller. It has been suggested that the nitrosamines are formed in kitch- ens when NO and amines are simulta x neously trapped in air-cleaning units. Since condensed-phase nitro compounds are common outdoor constituents, one would expect to find them indoors as well, perhaps at much reduced concentrations. No studies of such species indoors have been performed. Heterocyclic Organic Compounds. The only study identifying indoor heterocyclic compounds is that of Jarke et al. (1982~. At homes in the Chicago area, they found

T. E. Graedel 151 many organic compounds, including furan, dioxane, and indole. No quantification of the concentrations was attempted, but the authors estimated their detection limit for these compounds was about 0.5 ppb. The sources of the compounds were not de- termined but might be supposed to be either the infiltration of outdoor air or the by-products of indoor fossil fuel combus- tion. Polynuclear Aromatic Hydrocarbons. PAHs are readily detectable in the indoor environment, as a consequence of infiltra- tion of outdoor air as well as from indoor combustion sources. Their combined con- centrations indoors total perhaps 5-10 ng/m3 (Butler and Crossley 1979; Sexton et al. 1985~. Given PAH vapor pressures, these data imply as well that indoor equi- librium gaseous PAH concentrations will be around 1 ppt. Such levels are similar to, or slightly smaller than, outdoor PAH con- centrations. Suspended Particulate Matter. The con- centration of suspended particulate matter in buildings without air filtration appears to be generally higher than it is outdoors. The National Research Council (1981) states a range of indoor TSP of 10-500 ,ug/m3. Typical levels within most buildings are about 15-50 ,ug/m3 (Sexton et al. 1984, 1985). The chemical constituents of the indoor aerosols bear substantial resemblance to those outdoors. Organic carbon com- pounds make up perhaps 4-25 ,ug/m3 of the total (Sexton et al. 1985~; these compounds include phthalates, alkalies, fatty acids, and other oxygenated species (Weschler 1980, 1984; Weschler and Fong 1984~. Elemental carbon accounts for some 10 percent of the total aerosol mass, or about 2-7 ,ug/m3 (Sexton et al. 1985~. Heavy metals, includ- ing aluminum, iron, copper, and zinc, are present at concentrations of a few tens or hundreds of ng/m3 (Tosteson et al. 1982; Sexton et al. 1985~. The ions common to outdoor aerosols are also found indoors (Sinclair et al. 1985~. The principal sources of many of the organic and metallic com- pounds appear to be located within the buildings. In the case of Pb, however, most is present on fine particles and exists in- doors as a result of infiltration of outdoor air. Recommendation 7. Air quality re- searchers have only the most general ideas of indoor fluxes of trace species, the relative importance of indoor and outdoor sources to indoor species concentrations, and total  exposures. A major effort should be made to acquire such data, without which no epidemiologic studies can hope to be au- thoritative. Special effort should be directed to the passenger compartments of automo- biles, where total exposure is potentially quite high. Emittants with Potential Global Influence Carbon Dioxide CO2 is one of the principal products of the combustion of fossil fuels. Its emissions from motor vehicles are substantial but are small fractions of the global emission flux. CO2 is not toxic at or near atmospheric concentrations, but its presence in the at- mosphere has major effects on biogenic life cycles on the earth because it is a major absorber of the infrared radiation emitted toward space from the earth's surface. As a result, it is crucial to the establishment and maintainence of the planetary temperature structure. The concentration of CO2 at a remote atmospheric site is shown in figure 18. The upward trend is readily evident. It has been estimated that the atmospheric CO2 con- centration will double by the year 2030 or thereabouts, producing a global average temperature increase of about 1.5 to 4.5°C (National Research Council 1983~. Despite considerable study, it appears unlikely that any global program for the reduction of CO2 emissions will prove feasible. As a result, motor vehicle CO2 emissions are unlikely to be controlled by law. It is possible that increasing amounts of CO2, together with other radiation-absorbing gases, will change the total environment of

152 Anthropogenic Emissions and Their Atmospheric Transformation Products 350 c=` 345 o 340 335 z 330 c, 325 o 320 o 315 () 310 58 60 62 64 66 68 70 72 74 76 78 80 82 YEAR Figure 18. Concentration of atmospheric CO2 at Mauna Loa Observatory, Hawaii, expressed as a mole fraction in parts per million of dry air. The dots depict monthly averages of visually selected data that have been adjusted to the center of each month. The curve represents the fit simultaneously to an expo- nential function, a spline function, and a linearly increasing seasonal cycle. (Adapted with permission from Bacastow et al. 1985, and the American Geophy- sical Union.) the planet within two or three generations, and it may be prudent to keep that change as small as possible. Carbon Monoxide Most of the trace molecules emitted into the air are ultimately removed from it by reaction chains initiated by the hydroxyl (HO ~ radical (Atkinson, this volume). It appears from the results of photochemical atmospheric models that the most impor UJ 120 8 100_ ___- JAN. JAN. JAN. JAN. 1979 19430 19~1 1982 Figure 19. CO concentrations at Cape Meares, Or- egon. Monthly concentrations are formed from the 440 to 2,200 measurements each month, and these averages are then combined to form 12-month mov- ing averages. In this approach, seasonal cycles of a year or less disappear. The solid line is the trend calculated by linear least-squares techniques. (Adapted with permission from Khalil and Rasmussen 1984, @) 1984 by AAAS.) tent reactant in controlling the global HO- concentration is CO, because of its abun- dance and its rapid reaction with HO.. As with CO2, the atmospheric concentrations of CO are steadily increasing (figure 19~. As was pointed out above, motor vehicles are responsible for about two-thirds of all CO emissions and are thus major factors in the global CO increase. Photochemical model studies (Levine et al. 1985) suggest that over the past 35 years the average HO- concentra- tion has decreased by 25 percent. As a result, the ability of the atmosphere to cleanse itself has become increasingly inhibited.  Methane Methane is an absorber of infrared radiation as well as a factor in controlling the atmo- spheric abundance of the HO- radical. As a consequence, its long-term trend is also of interest. A summary of atmospheric con- centration measurements of methane over the past several years is given in figure 20. As with CO2 and CO, methane concentra- tions are increasing, at about 1.2 percent per year. About a thousandth of the annual methane emissions are attributable to mo- tor vehicles (Ehhalt and Schmidt 1978~. 1.65 Q Q - z 1.60 o z MU of 8 1.55 I ; 1.50 I I I I I_ 78 79 80 81 82 83 84 YEAR Figure 20. Average worldwide tropospheric con- centrations of methane during the period 1978-1983. The solid line represents a least-squares fit with an increase of 0.018 ppm per year. (Adapted with per- mission from Blake and Rowland 1986, and D. Reidel Publishing Company.)

T. E. Graedel 153 Summary Data Adequacy Data adequacy is taken here to mean that sufficient data are available to permit rea- sonable assessments of the effects of a given atmospheric constituent on animate and inanimate objects, and not that the concen- trations are known on every street corner. By this definition, sufficient outdoor data are available for the "criteria species" CO, N02' S02, 03, Pb, and TSP. For total or NMHCs the amount of available data is marginally adequate, in large part because . . . . . . no satlstactory routine monitoring 1nstru ment is available. Inside buildings a similar situation exists, with at least the approxi ~ . . . mate range ot criteria species concentra- tions having been established. Within the passenger compartments of automobiles, very few concentration data have appeared in the literature. In the case of atmospheric species for which ambient standards have not been established, the available outdoor data are generally inadequate to do more than infer order of magnitude concentrations and to produce some idea of the relative strengths of the potential sources of the compounds. For example, the data on formaldehyde and other small aldehydes are from very few sites and are not now being enhanced by any regular monitoring. As with NMHC, this is partly because no routine, reliable monitoring instrument is available. For methanol, ethanol, and manganese, species that may soon be emitted from motor vehicles at much higher rates, the data are extremely sparse. It is important that these compounds be included soon in routine monitoring programs in order to establish baseline concentrations for future refer- ence. Other species for which more data are needed are those generally present in aero- sol form which possess positive bioassay characteristics. Such compounds include the nitro derivatives of PAHs and several heterocyclic species. Indoors, measurements have been re- stricted largely to the criteria species and to formaldehyde. Much greater characteriza . , . .. . . . . lion or trace species In the 1nc boor envlron ment is needed. Trends For CO, CO2, and methane, the atmo- spheric concentrations show an increasing trend at global background locations, and stable or slightly decreasing peak levels in urban areas. For SO2, Pb, and TSP, de- creasing trends are seen. The concentra- tions of O3 and NO2 in most U.S. urban  regions appear to be roughly stable on an annual basis (Los Angeles is the exception, having shown a 25 percent decrease over the last eight years); at global background sites the concentration of 03, at least, ap- pears to be increasing, thus implying also an increase in NO2. Concentrations As is evident from the information above, concentrations of airborne species of interest show wide variations from site to site and time to time. Notwithstanding this complex- ity, it is useful to tabulate information from the literature on typical values of average and peak concentrations. Such data appear in table 1 for 21 species. In each case, an attempt is made to indicate the approximate state of measurement technology currently required. The ranges of values for urban areas are annual averages, if available. Peak values are given for urban outdoor environments, for . . . mc oor nonmanu actunng environments, and for the passenger compartments of auto- mobiles. The availability of data generally decreases from left to right in the table. Many more data are extant on emitted gases and TSP than for other species shown. For gaining a quick perspective on typical concentrations of trace species, graphical dis- plays are often convenient. Such displays are presented here for species for which sufficient data are available to establish typical ranges of concentrations. Trace gases are displayed in figure 21. In most cases, the concentrations in remote areas are the lowest, those in outdoor urban air next highest, and those indoors highest of all. (The exception is 03, which has roughly the same peak values in all three regimes.) The ordinate on figure 21 is loga- rithmic; remote and indoor concentrations differ by as much as four or five orders of . . magmtuc .e In some cases.

154 Anthropogenic Emissions and Their Atmospheric Transformation Products Table 1. Typical Ranges and Peak Values for Gaseous and Particulate Species Species Measure ment Ca pability Concentrations Urban Range Urban Indoor Auto Peak Peak Peakb References Emitted gases (ppb) CO RM (3-15) x 103 CO2 RM (3-6) x 105 4 x 104 1 X 105 3 x 104 National Research Council 6x 105 3x 106 NOr RM 10 - 50 800 500 1 x 103 - .~ NMHC ET (1-5) x 103 SO2 RM 3-20 Product gases (ppb) o3 X 104 3 x 104 2 x 103 300 20 RM 90-210 350 HNO2 ES 0.2-4(s) HNO3 PAN H202 HCHO ES ES ES ET Particle species (,ug/m3) Pb RM  EC OC NOT so2 PAH 1-5 30 5-10 0.2-2(S) 3-60 25 50 1 x 103 0.1-0.7 1.0 0.1 ET ET RM RM 1-15 5-20 1-10 1-20 ES (5-10) x 10-2(s) Nitro-PAH ES (1-3) x 1o-4(S) 3 x 10-4 (Table continued next page.) (1981); U.S. Environmental Protection Agency (1985); Mucke et al. (1984) McRae and Graedel (1979); Spengler and Sexton (1983) U.S. Environmental Protec- tion Agency (1985); Mucke et al. (1984); Spengler and Sexton (1983) DeBertoli et al. (1984); Mucke et al. (1984); Tilton and Bruce (1980) 200 35 40 150.7 30 ~_ 0.110 National Research Council (1981); U.S. Environmental Protection Agency (1985)  National Research Council (1981); U.S. Environmental Protection Agency (1985); Tuazon et al. (1981) Harris et al. (1982); Pitts et al. (1985b); Sjodin and Ferm (1985) Spicer (1977); Tuazon et al. (1981) Tuazon et al. (1981) Kok et al. (1986) National Research Council (1981); National Academy of Sciences (1981); Tuazon et al. (1981) U.S. Environmental Protec- tion Agency (1985); Toste- son et al. (1982) Countess et al. (1980); Wolff (1985) Countess et al. (1980); Wolff (1985) Graedel and Schwartz (1977); Graedel et al. (1986) Graedel and Schwartz (1977); Graedel et al. (1986) Lahmann et al. (1984); Mos- chandreas et al. (1981); Sei- fert et al. (1983) Gibson (1982); Pitts et al. (1985a)

T. E. Graedel 155 Table 1. Continued S. penes Measure ment Ca pability Concentrations Urban Range Urban Indoor Auto Peak Peak Peakb References TSP RM IP RP 30-75 RM 5-80 120 RM 10-75 210 400 500 500 150 Bennett et al. (1985); U.S. Environmental Protection Agency (1985); Spengler and Sexton (1983) Lioy et al. (1983); Wolff et al. (1985b) Budiansky (1980); National Research Council (1981); DeBortoli et al. (1984) a Averaging times are annual for urban range except shorter where noted by (s), daily for urban peak, an hour or two for indoor and automobile interior data. b Concentrations as measured within the passenger compartment of an automobile. NOTE: Abbreviations and symbols not elsewhere defined are EC = elemental carbon; 0C = organic carbon; RM = routine monitoring; ET = event monitoring by technician; ES = event monitoring by scientist. 103 Con  o z C) o C' 1'°° 10 1 100 _ a, GO 10 l 1 CO NO2 NMHC SO2 O3 HCHO TRACE GASES Figure 21. Typical concentration ranges of selected atmospheric trace gases in remote areas (R), urban areas (U), and indoors (I). Data for this display are from Noxon 1975; Spengler et al. 1979; M0lhave 1982; U.S. Environmental Protection Agency 1985; Wallace et al. 1985; and Graedel et al. 1986. Because ofthe form ofthe ambient air quality standards for CO and 03, the urban data for those compounds represent upper extreme values (extracted from an annual data set) rather than mean values. - ~ 10 of o  z 10-1 10-2 _ _ $, 1 ~ TSP SO42- NO3- Pb AIRBORNE PARTICLE CONSTITUENTS Figure 22. Typical concentration ranges of selected atmospheric airborne particle constituents in remote areas (R), urban areas (U), and indoors (I). The data for this display are from Rhodes et al. 1979; Graedel 1980; Sinclair et al. 1985; and the references given in table 1.

156 Anthropogenic Emissions and Their Atmospheric Transformation Products A similar display for particulate matter is given in figure 22. In general, the data for the three regimes are much more similar than was the case for the gases. Two cave- ats are worth noting, however: adequate trace metal data, except for Pb, are not extensive, and few indoor data exist for other than TSP. Limited studies of trace metal compositions throughout the world suggest that urban and remote concentra- tion differences are substantial. Summary of Research Recommendations HIGH PRIORITY Recommendation 7 The proportion of time spent indoors by most people is high, yet Indoor Air Quality the available data for indoor air quality is quite sparse. This is particularly true of nonresidential environments such as automobile interiors, subway platforms, and the like. Air quality researchers have only the most general ideas of indoor fluxes of trace species, the relative importance of indoor and outdoor sources to indoor species concentrations, and total exposures. A major effort should be made to acquire such data, without which no epidemiologic studies can hope to be authoritative. Special effort should be directed to the passenger compartments of automobiles, where total exposure is potentially quite high. . Recommendation 3 The alkanic aldehydes are unusual atmospheric constituents in Aldehyde Monitor the sense that they are directly emitted by sources as well as being produced in the atmosphere by gas-phase chemistry. Their effects on humans and animals could be significant. The determination of atmospheric aldehyde levels is a difficult task, but the limited data available show that much insight into atmospheric processes is likely to be derived from carefully collected, extended data records. An effort to develop a monitor for either formaldehyde or the aldehyde group with appropriate sensitivity (100 ppt or less), reliability, and an appropriately low cost, is a high priority for atmospheric chemists, particularly those addressing problems in urban air. MODERATE PRIORITY Recommendation 2Measurements of the total particulate loading of the atmosphere Unregulatedgive only the crudest possible idea of the condensed-phase aerosol Species Particlechemistry. A difficult but essential job is to monitor the chemistry Phaseof atmospheric aerosol particles in much more detail than is now being done, concentrating especially on chemical differences as a function of particle size. Detailed organic analyses are particularly important. It is likely that the most interesting and complex air quality problems in the next decade will relate to condensed-phase species; enhancing the level of analysis and depth of study of these species are thus matters of critical concern. Recommendation 1Hundreds of NMHCs are present in the atmosphere. They are NMHC Monitorinvolved in the formation of 03, PAN, formaldehyde, and other

T. E. Graedel 157 lachrymators, some of which are toxic, and some of which serve as precursors for such potentially hazardous compounds as the nitro- aromatics. Notwithstanding this central role, no analytical instru- ment is readily available for routine monitoring of their concentra- tions and concentration trends, although several techniques are available for potential incorporation into such an instrument. It is extremely important to achieve agreement on a satisfactory mon- itoring technique for NMHCs (or some significant fraction thereof) and to begin to acquire data on a routine basis. Recommendation 4 From the bioassay perspective, the complex chemical products of Unregulated the compounds emitted to the atmosphere are often of the most Species Gas Phase concern. Perhaps 98 percent of monitoring efforts are directed at criteria species, however. It is important to recognize that many ~ _ , assessments having to do with atmospheric species cannot go forward unless supporting data for them are available. A few examples of crucial species listed earlier include PAN, nitric acid, hydrochloric acid, and others. It is important for atmospheric chemists to focus their thinking on unmet needs in species diversity and geographical diversity. LOW PRIORITY Recommendation 5 Atmospheric alcohols have not often been studied in the atmo Alkanic Alcohol sphere, but the limited data suggest that in urban areas, at least, Monitoring their concentrations rival those of many better known organic compounds. Once present in the atmosphere, the alcohols will react to produce aldehydes and organic acids, two groups of compounds potentially involved in a number of injurious interac tions with animate and inanimate surfaces. Given the possibility of sharply increased use of methanol and ethanol in motor vehicle fuels over the next decade or two, it is important to begin promptly to establish a satisfactory baseline against which future changes in atmospheric alcohol concentrations could be assessed. The tech niques now available, if perhaps not optimum, are satisfactory at least for a limited screening program. Recommendation 6 The ultimate concern of the epidemiologist dealing with the Personal Exposure effects of atmospheric species is not species concentrations at a Monitors monitoring site, but those encountered by human beings. Personal exposure monitors thus have important future roles to play in health effects research. Instrument development is required, how ever, to improve portability, reliability, and the quantitative detec tion of many atmospheric species of interest not presently capable of being monitored in this way.

158 Anthropogenic Emissions and Their Atmospheric Transformation Products Acknowlecigments I thank R. S. Freund, l. D. Sinclair, and C. l. Weschler for useful reviews of this paper, and K. Sexton and B. Seifert for providing considerable assistance in locating reference material dealing with indoor air quality measurements. References Aurand, K., Seifert, B., and Wegner, J. 1982. Luft- qualitat in Innenraumen, Gustav Fischer Verlag, Stuttgart. Bacastow, R. B., Keeling, C. D., and Whorf, T. P. 1985. Seasonal amplitude increase in atmospheric CO2 concentration at Mauna Loa, Hawaii, - 1951-1982. J. Geophys. Res. 90:1052~10540. Bellar, T. A., and Sigsby, J. E., Jr. 1970. Direct gas chromatographic analysis of low molecular weight substituted organic compounds in emissions, Envi- ron. Sci. Technol. 4:150-156. Bennett, B. G., Kretzschmar, J. G., Akland, G. G., and de Koning, H. W. 1985. Urban air pollution worldwide, Environ. Sci. Technol. 19:298-304. Berglund, B., Johansson, I., and Lindvall, T. 1982. A longitudinal study of air contaminants in a newly built preschool, Environ. Int. 8:111-115. Berglund, B., Berglund, U., Lindvall, T., Spengler, J., and Sunden, J. (eds.) 1986. Indoor air, Environ. Int..12: whole issue (nos. 10) pages 1-494. Blake, D. R., and Rowland, F. S. 1986. World-wide increase in tropospheric methane, 197~1983, J. Atmos. Chem. 4:43~2. Brodzinsky, R., Chang, S. G., Markowitz, S. S., and Novakov, T. 1980. Kinetics and mechanism for the catalytic oxidation of sulfur dioxide on carbon in aqueous suspensions,_~. Phys. Chem. 84:335~3358. Budiansky, S. 1980. Indoor air pollution, Environ. Sci. Technol. 14:1023-1027. Bundesgesundheitsamt, Bundesanstalt fur Arbeitss- chutz, and Umweltbundesamt. 1985. Formaldehyde. Medizin Verlag, Munchen. Butler, J. D., and Crossley, P. 1979. An appraisal of relative airborne sub-urban concentrations of poly- cyclic aromatic hydrocarbons monitored indoors and outdoors, Sci. Total Environ. 11:53-58. Cautreels, W., and Van Canwenberghe, K. 1978. Ex- periments on the distribution of organic pollutants between airborne particulate matter and the corre- sponding gas phase, Atmos. Environ. 12:1133-1141. Cleveland, W. S., Graedel, T. E., and Kleiner, B. 1977. Urban formaldehyde: observed correlation with source emissions and photochemistry, Atmos. Environ. 11 :357-360. Countess, R. J., Wolff, G. T., and Cadle, S. H. 1980. Correspondence should be addressed to T. E. Grae- del, AT&T Bell Laboratories, 600 Mountain Avenue, Murray Hill, NJ 07974. The Denver winter aerosol: a comprehensive chem- ical characterization, J. Air Pollut. Contr~l Assoc. 30:119~1200.  DeBortoli, M., Knoppel, H., Pecchio, E., Peil, A., Rogora, L., Schauenburg, H., Schlitt, H., and Vissers, H. 1984. Integrating "real life" measure- ments of organic pollution in indoor and outdoor air of homes in northern Italy. Proc. 3rd Int. Conf: Indoor Air Quality and Climate 4:21-26. Dong, M. W., Locke, D. C., and Hoffmann, D. 1977. Characterization of aza-arenes in basic organic por- tion of suspended particulate matter, Environ. Sci. Technol. 11 :612018. Ehhalt, D. H., and Schmidt, U. 1978. Sources and sinks of atmospheric methane. Pure Appl. Geophys. 116:452-464. Fanger, P. O., and Valbjorn, O. (eds.) 1979. Indoor Climate, Danish Bldg. Res. Inst., Copenhagen. Gibson, T. L. 1982. Nitro derivatives of polynuclear aromatic hydrocarbon in airborne and source par- ticulate matter, Atmos. Environ. 16:2037-2040. Graedel, T. E. 1980. Atmospheric photochemistry, In: Handbook of Environmental Chemistry (O. Hutzinger, ed.), Vol. 2A, Springer, Heidelberg: pp. 107-143. Graedel, T. E., and McRae, J. E. 1982. Total organic component data: a study of urban atmospheric patterns and trends, Atmos. Environ. 16:11101132. Graedel, T. E., and Schwartz, N. 1977. Air quality reference data for corrosion assessment, Mater. Per- form. 16(8):17-25. Graedel, T. E., Hawkins, D. T., and Claxton, L. D. 1986. Atmospheric Chemical Compounds: Sources, Oc- currence, and Bioassay, Academic Press, Orlando, Fla. 732 pp. Grosjean, D., and Friedlander, S. K. 1975. Gas- particle distribution factors for organic and other pollutants in the Los Angeles atmosphere, J. Air Pollut. Control Assoc. 25:1038-1044. Hanrahan, L. P., Anderson, H. A., Dally, K. A., Eckmann, A. D., and Kanarek, M. S. 1985. Form- aldehyde concentrations in Wisconsin mobile homes.J. AirPollut. ControlAssoc. 35:1164-1167. Hanst, P. L., Wilson, W. E., Patterson, R. K., Gay, B. W., Jr., Chaney, L. W., and Burton, C. S. 1975. A spectroscopic study of California smog, EPA Report 650/~75-006, U.S. Environmental Protec- tion Agency, Washington, D.C. Harkov, R., Kebbekus, B., Bozzelli, J. W., and Lioy, P. J. 1983. Measurement of selected volatile organic compounds at three locations in New Jersey during the summer season, J. Air Pollut. Control Assoc. 33: 1177-1183. Harris, G. W., Carter, W. P. L., Winer, A. M., Pitts, J. N., Jr., Platt, U., and Perner, D. 1982. Obser- vations of nitrous acid in the Los Angeles atmo- sphere and implications for predictions of ozone- precursor relationships, Environ. Sci. Technol. 16: 414-419. Harrison, R. M., and Pio, C. A. 1983. Major ion composition and chemical associations of inorganic atmospheric aerosols, Environ. Sci. Technol. 17: 169-174. Holzer, G., Shanfield, H., Zlatkis, A., Bertsch, W., Jaurez, P., Mayfield, H., and Liebich, H. M. 1977.

T. E. Graedel Collection and analysis of trace organic emissions from natural sources,J. Chromatogr. 142:755-764. Hoshika, Y. 1977. Simple and rapid gas-liquid-solid chromatographic analysis of trace concentrations of acetaldehyde in urban air, J. Chromatogr. 137:455- 460. Jarke, F. H., Dravnieks, A., and Gordon, S. M. 1981. Organic contaminants in indoor air and their rela- tion to outdoor contaminants, ASHRAE Trans. 81 (Part 1):153-165. Johnson, W. B., Ludwig, F. L., Dabberdt, W. F., and Allen, R. J. 1973. An urban diffusion simulation model for carbon monoxide, J. Air Pollut. Control Assoc. 23:49~498. Ketseridis, G., and Jaenicke, R. 1978. Organische Beimengungen in atmospharisher Reinluft: Fin Bei- trag zur Budget-Abschatzung, In: Organische Verun- reinigungen in der Umwelt-Erkennen, Berwerben' Ber- mindern (K. Aurand et al., eds.) E. Schmidt, Berlin: pp. 379-390. Khalil, M. A. K., and Rasmussen, R. A. 1984. Carbon monoxide in the earth's atmosphere: in- creasing trend, Science 224:54-56. Kok, G. L., Heikes, B. G., and Lazrus, A. L. 1986. Gas and aqueous phase measurements of hydrogen peroxide, paper presented at the National Meeting of the American Chemical Society, New York, N.Y., April 14, 1986. Krstulovic, A. M., Rosie, D. M., and Brown, P. R. 1977. Distribution of some atmospheric polynu- clear aromatic hydrocarbons, Am. Lab. 9(7):11-18. Lahmann, E., Seifert, B., Zhao, L., and Bake, D. 1984. Immissionen von polycyclischen aromatis- chen kohlenwasserstoffen in Berlin (West), Staub- Reinhalt. Luff 44:149-157. Levine, J. S., Rinsland, C. P., and Tennille, G. M. 1985. The photochemistry of methane and carbon monoxide in the troposphere in 1950 and 1985, Nature 318:254-257. Lioy, P. J., Daisy, J. M., Reiss, N. M., and Harkov, R. 1983. Characterization of inhalable particulate matter, volatile organic compounds and other chemical species measured in urban areas in New Jersey. I. Summertime episodes, Atmos. Environ. 17:2321-2330. Logan, J. A. 1985. Tropospheric ozone: seasonal behavior, trends, and anthropogenic influence, J. Geophys. Res. 90:10463-10482. Lonw, C. W., Richards, J. F., and Faure, P. K. 1977. The determination of volatile organic compounds in city air by gas chromatography combined with standard addition, selective subtraction, infrared spectrometry, and mass spectrometry, Atmos. Envi- ron. 11:703-717. McRae, J. E., and Graedel, T. E. 1979. Carbon dioxide in the urban atmosphere: dependencies and trends,J. Geophys. Res. 84:5011-5017. Milford, J. B., and Davidson, C. I. 1985. The sizes of particulate trace elements in the atmosphere-a re- view, J. Air Pollut. Control Assoc. 35:1249-1260. Miller, D. F., and Spicer, C. W. 1975. Measurement of nitric acid in smog, J. Air Pollut. Control Assoc. 25:940-942. M01have, L. 1982. Indoor air pollution due to organic  159 gases and vapours of solvents in building materials, Environ. Int. 8:117-127. M01have, L., and M011er, J. 1979. The atmospheric environment in modern Danish dwellings mea- surements in 39 flats, Proc. 1st Int. Indoor Climate Symp. 1:171-186. Moschandreas, D. J., Zabransky, J., and Pelton, D. J. 1980. Comparison of Indoor-Outdoor Concentra- tions of Atmospheric Pollutants, GEOMET Report ES-823, Electric Power Research Institute, Palo Alto, Calif. Moschandreas, D. J., Pelton, D. J., and Berg, D. R. 1981. The effects of woodburning on indoor pol- lutant concentrations. Paper No. 81-22.2 presented at the 74th Annual Meeting of the Air Pollution Control Assoc., Philadelphia, Pa. Mucke, W., Jost, D., and Rudolf, W. 1984. Luftve- runreinigungen in kraftfahrzeugen, Staub-Reinhalt. Luft 44:374-377. National Academy of Sciences. 1972. Particulate Poly- cyclic Organic Matter, Washington, D.C., 361 pp. National Academy of Sciences. 1977. Ozone and Other Photochemical Oxidants, Washington, D.C., 719 pp. National Academy of Sciences. 1981. Formaldehyde and Other Aldehydes. National Academy Press, Wash- ington, D.C., 340 pp. National Research Council, Committee on Indoor Pollutants. 1981. Indoor Pollutants, National Acad- emy Press, Washington, D.C. National Research Council, Carbon Dioxide Assess- ment Committee. 1983. Changing Climate, National Academy Press, Washington, D.C., 496 pp. Noxon, J. F. 1975. Nitrogen dioxide in the strato- sphere and troposphere measured by ground-based absorption spectroscopy, Science 189:547-549. Ott, W. R. 1985. Total human exposure, Environ. Sci. Technol. 19:88~886. Ott, W., and Flachsbart, P. 1982. Measurement of carbon monoxide concentrations in indoor and out- door locations using personal exposure monitors, Environ. Int. 8:295-304. Pitts, J. N., Jr., Sweetman, J. A., Zielinska, B., Winer, A. M., and Atkinson, R. 1985a. Determination of 2-nitrofluoranthene and 2-nitropyrene in ambient par- ticulate organic matter: evidence for atmospheric re- actions, Atmos. Environ. 19:1601-1608. Pitts, J. N., Jr., Wallington, T. J., Biermann, H. W., and Winer, A. M. 1985b. Identification and mea- surement of nitrous acid in an indoor environment, Atmos. Environ. 19:763-767. Quackenboss, J. J., Kanarek, M. S., Spengler, J. D., and Letz, R. 1982. Personal monitoring for nitrogen dioxide exposure: methodological considerations for a community study, Environ. Int. 8:249-258. Rhodes, R. C., Fair, D. H., Frazer, J. E., Long, S. J., Loseke, W. A., Wheeler, V. A., and Walling, J. F. 1979. Air Quality Data for Metals 1976 from the National Air Surveillance Networks. Report EPA- 600/4-79-054. U. S. Environmental Protection Agency, Research Triangle Park, N.C. Rosen, H., Hansen, A. D. A., Gundel, L., and Novakov, T. 1978. Identification of the optically absorbing component in urban aerosols, Appl. Op- tics 17:3859-3861.  Seifert, B., Ullrich, D., and Schmahl, H. J. 1983.

160 Anthropogenic Emissions and Their Atmospheric Transformation Products Occurrence of carcinogenic organic substances in kitchen air, In: Proc. 6th World Cong. Air Quality 2:177-179. Seifert, B., Presher, K.-E., and Ullrich, D. 1984. Auftreten anorganischer und organischer substan- zen in der luff von kitchen and anderen wohn- raumen. Inst. fur Wasser-, Boden-, und LuEthy- giene des Bundes gesundheitsantes, Berlin. Sexton, K., Spengler, J. D., and Treitman, R. D. 1984. Effects of residential wood combustion on indoor air quality: a case study in Waterbury, Vermont, Atmos. Environ. 18:1371-1383. Sexton, K., Liu, K.-S., Treitman, R. D., Spengler, J. D., and Turner, W. A. 1986. Characterization of indoor air quality in wood-burning residences, En- viron. Int. 12:265-278. Sexton, K., Liu, K. -S., and Petreas, M. X. 1986. Formaldehyde concentrations inside private resi- dences: a mail-out approach to indoor air monitor- ing, J. Air Pollut. Control Assoc. 36:698-704. Sinclair, J. D., Psota-Kelty, L. A., and Weschler, C. J. 1985. Indoor/outdoor concentrations and indoor surface accumulations of ionic substances, Atmos. Environ. 19:31~323. Sjodin, A., and Ferm, M. 1985. Measurements of nitrous acid in an urban area, Atmos. Environ. 19:985-992. Snider, J. R., and Dawson, G. A. 1985. Tropospheric light alcohols, carbonyls, and acetonitrile: concen- trations in the southwest United States and Henry's law data, J. Geophys. Res. 90:3797-3805. Spengler, J. D., and Sexton, K. 1983. Indoor air pollu- tion: a public health perspective, Science 221:9-17. Spengler, J. D., Ferris, B. G., Jr., Dockery, D. W., and Speizer, F. E. 1979. Sulfur dioxide and nitrogen dioxide levels inside and outside homes and the implications on health effects research, Environ. Sci. Technol. 13:1276-1280. Spengler, J., Hollowell, C., Moschandreas, D., and Fanger, O. (eds.) 1982. Indoor air pollution, Envi- ron. Int. 8:1-534. Spicer, C. W. 1977. Photochemical atmospheric pol- lutants derived from nitrogen oxides, Atmos. Envi- ron. 11 :1089-1095. Tanner, R. L., and Meng, Z. 1984. Seasonal varia- tions in ambient atmospheric levels of formalde- hyde and acetaldehyde, Environ. Sci. Technol. 18:723-726. Taylor, O. C. 1969. Importance of peroxyacetyl nitrate (PAN) as a phytotoxic air pollutant, J. Air Pollut. Control Assoc. 19:347-351. Thompson, C. R., Hensel, E. G., and Kats, G. 1973. Outdoor-indoor levels of six air pollutants, J. Air Pollut. Control Assoc. 23:881-886. Tilton, B. E., and Bruce, R. M. 1980. Review of Criteria for Vapor-Phase Hydrocarbons. Report EPA-600/8-80-045. U. S. Environmental Protection Agency, Washington, D.C. Tosteson, T. D., Spengler, J. D., and Weker, R. A. 1982. Aluminum, iron, and lead content of respira- ble particulate samples from a personal monitoring study, Environ. Int. 8:26~268. Tuazon, E. C., Winer, A. M., and Pitts, J. N., Jr. 1981. Trace pollutant concentrations in a multiday  smog episode in the California South Coast Air Basin by long path length Fourier transform infra- red spectroscopy, Environ. Sci. Technol. 15:1232- 1237. U.S. Environmental Protection Agency. 1985. Na- tional Air Quality and Emissions Trends Report, 1983. Report EPA-450/~84-029, Research Triangle Park, N.C. Wallace, L., Zweidinger, R., Erickson, M., Cooper, S., Whitaker, D., and Pellizzari, E. 1982. Monitor- ing individual exposure. Measurements of volatile organic compounds in breathing-zone air, drinking water, and exhaled breath, Environ. Int. 8:269-282. Wallace, L. A., Pellizzari, E. D., Hartwell, T. D., Sparacino, C. M., Sheldon, L. S., and Zelon, H. 1985. Personal exposures, indoor-outdoor relation- ships, and breath levels of toxic air pollutants mea- sured for 355 persons in New Jersey, Atmos. Envi- ron. 19:1651-1661. Wang, T. C. 1975. A study of bioeffluents in a college classroom, ASHRAE Trans. 81 (Part 1):32-44. Weschler, C. J. 1980. Characterization of selected organics in size-fractionated indoor aerosols, Envi- ron. Sci. Technol. 14:428-431. Weschler, C. J. 1984. Indoor-outdoor relationships for nonpolar organic constituents of aerosol parti- cles, Environ. Sci. Technol. 18:648-652. Weschler, C. J., and Fong, K. J. 1984. Characteriza- tion of organic species associated with indoor aero- sol particles, Proc. 3rd Int. Con~ Indoor Air Quality and Climate, 2:203-208. Wolff, G. T. 1985. Characteristics and consequences of soot in the atmosphere, Environ. Int. 11 :259-269. Wolff, G. T., and Klimisch, R. L. (eds.) 1982. Partic- ulate Carbon: Atmospheric Life Cycle, Plenum, New York, N.Y. Wolff, G. T., Ferman, M. A., Kelly, N. A., Stroup, D. P., and Ruthkosky, M. S. 1982a. The relation- ships between the chemical composition of fine particles and visibility in the Detroit metropolitan area, J. Air Pollut. Control Assoc. 32:121~1220. Wolff, G. T., Groblicki, P. J., Cadle, S. H., and Countess, R. J. 1982b. Particulate carbon at various locations in the United States, In: Particulate Carbon: Atmospheric Life Cycle (G. T. Wolff and R. L. Klimisch, eds. ), Plenum, New York, N. Y., pp. 297-315. Wolff, G. T., Korsog, P. E., Kelly, N. A., and Ferman, M. A. 1985a. Relationships between fine particulate species, gaseous pollutants and meteoro- logical parameters in Detroit, Atmos. Environ. 19:1341-1349. Wol~, G. T., Korsog, P. E., Stroup, D. P., Ruth- kosky, M. S., and Morrissey, M. L. 1985b. The influence of local and regional sources on the con- centration of inhalable particulate matter in south- eastern Michigan, Atmos. Environ. 19:305-313. World Health Organization. 1983. Indoor Air Pollut- ants: Exposure and Health Effects, EURO Reports and Studies 78, Copenhagen, 42 pp. Yocom, J. E. 1982. Indoor-outdoor air quality rela- tionships: a critical review, J. Air Pollut. Control Assoc. 32:50~520.