United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Research Triangle Park NC 27711
EPA-600/8-82-029b
December 1982
FINAL
Research and Development
Air Quality
Criteria for
Particulate Matter
and Sulfur Oxides
Volume II
-------
EPA-600/8-82-029b
December 1982
Air Quality Criteria
for P^tieulate Matter
and Sulfur Oxides
Volume II
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Environmental Criteria and Assessment Office
Research Triangle Park, NC 27711
-------
NOTICE
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
ii
-------
Preface
This document is Volume II of a three-volume revision of Air Qual1ty
Criteria for Particulate Matter and Air Quality Criteria for Sulfur Oxides,
first published in 1969 and 1970, respectively. By law, air quality criteria
documents are the basis for establishment of the National Ambient Air Quality
Standards (NAAQS). The Air Quality Criteria document of which this volume is
a part has been prepared in response to specific requirements of Section 108
of the Clean Air Act, as amended in 1977. The Clean Air Act requires that the
Administrator periodically review, and as appropriate, update and reissue
criteria for NAAQS. "' •-'
As the legally prescribed basis for deciding on National Ambient Air
Quality Standards, the present document, Air Quality Criteria for Particulate
Matter and Sulfur Oxides, focuses on characterization of health and welfare
effects associated with exposure to particulate matter and sulfur oxides and
pollutant concentrations which cause such effects. The major health and
welfare effects of particulate matter and sulfur oxides are discussed in
Chapters 8 through 14 in Volume III of this document. To assist the reader in
putting the effects into perspective with the real-world environment, Chapters
2 through 7 in the present volume (Volume II) have been prepared. The
chapters of Volume II discuss essential points regarding: physical and chemi-
cal properties; air monitoring and analytical measurement techniques; sources
and emissions; transport, transformation, and fate; and observed ambient
concentrations of the pollutants. Also, Chapter 7 in this volume introduces
the reader to the contemporary problem of acidic deposition and potential
contributions of sulfur oxides to acidic deposition phenomena.
Volume I introduces the criteria document, explains the rationale behind
combining the evaluation of criteria for particulate matter and sulfur oxides
in a single document and briefly summarizes the content of the entire criteria
document. However, for a fuller understanding of the health and welfare
effects of particulate matter and sulfur oxides, both Volumes II and III of
the document should be consulted.
The Agency is pleased to acknowledge the efforts of all persons and
groups who have contributed to the preparation of this document. In the last
analysis, however, the Environmental Protection Agency accepts full respon-
sibility for its content.
n i
-------
Page Intentionally Blank
-------
VOLUME II
CONTENTS
2. PHYSICS AND CHEMISTRY OF SULFUR OXIDES AND PARTICULATE MATTER..... 2-1
2.1 INTRODUCTION .• 2-1
2.2 ATMOSPHERIC DOMAIN AND PROCESSES. 2-3
2.3 PHYSICS AND CHEMISTRY OF SULFUR OXIDES 2-8
2.3.1 Physical Properties of Sulfur Oxides in the Gas
Phase 2-8
2.3.2 Solution Physical Properties 2-8
2.3.2.1 Sulfur Dioxide 2-8
2.3.2.2 Sulfur Trioxide and Sulfuric Acid 2-12
2.3.3 Gas-Phase Chemical Reactions of Sulfur Dioxide 2-12
2.3.3.1 Elementary Reactions....... 2-14
2.3.3.2 Tropospheric Chemistry of Sulfur Dioxide
Oxidation 2-15
2.3.4 Solution-Phase Chemical Reactions 2-22
2.3.4.1 S(IV)-Q2 - H20 System . 2-23
2.3.4.2 S(IV) - Catalyst - 02 - H20 System 2-27
2.3.4.3 S(IV) - Carbon Black - 02 - H20....... 2-35
2.3.4.4 S(IV) - Dissolved Oxidants - H20 2-35
2.3.4.5 The Influence of Ammonia. 2-37
2.3.5 Surface Chemical Reactions 2-38
2.3.6 Estimates of S02 Oxidation 2-40
2.4 PHYSICS AND CHEMISTRY OF PARTICULATE MATTER 2-41
2.4.1 Definitions 2-42
2.4.2 Physical Properties of Gases and Particles 2-45
2.4.2.1 Physical Properties of Gases 2-4S
2.4.21.2 Physical Properties of Particles 2-46
2.4.3 Dynamics of Single Particles 2-60
2.4.4 Formation and Growth of Particles 2-62
2.4.4.1 Growth Dynamics 2-65
2.4.4.2 Sulfuric Acid - Water Growth Dynamics.. 2-67
2.4.4.3 Dynamics of Growth by Chemical Reaction 2-67
2.4.4.4 Dynamics of Desorption 2-68
2.4.5 Characterization of Atmospheric Aerosol 2-69
2.4.5.1 Distribution... 2-69
2.4.5.2 Composition of Particles 2-75
2.4.6 Particle-Size Spectra Evolution. 2-80
2.4.6.1 General Dynamics Equation (GDE) 2-80
2.4.6.2 Application of the GDE. 2-81
2.5 REFERENCES 2-86
3. TECHNIQUES FOR THE COLLECTION AND ANALYSIS OF SULFUR OXIDES,
PARTICULATE MATTER, AND ACID PRECIPITATION 3-1
3.1 INTRODUCTION 3-1
3.2 MEASUREMENT TECHNIQUES FOR SULFUR DIOXIDE 3-2
3.2.1 Introduction 3-2
3.2.2 Manual Methods. 3-2
-------
CONTENTS (continued)
Page
3.2.2.1 Sample Collection 3-2
3.2.2.2 Calibration 3-3
3.2.2.3 Measurement Methods 3-4
3.2.3 Automated Methods. 3-12
3.2.3.1 Sample Collection 3-12
3.2.3.2 Calibration 3-12
3.2.3.3 Measurement Methods 3-13
3.2.3.4 EPA Designated Equivalent Methods 3-17
3.2.4 Summary 3-21
3.3 PARTICULATE MATTER (PM). 3-24
3.3.1 Introduction 3-24
3.3.2 Gravimetric PM Mass Measurements 3-30
3.3.2.1 Filtration Samplers 3-32
3.3.2.2 Impactor 'Samplers 3-49
3.3.2.3 Dustfall Sampling 3-54
3.3.3 Nongravimetn'c Mass Measurements 3-54
3.3.3*. 1 Filtration and Impaction Samplers 3-54
3.3.3.2 In Situ Analyzers. 3-60
3.3,4 Particle Composition.. 3-62
3.3.4.1 Analysis of Sulfates 3-63
3.3.4.2 Ammonium and Gaseous Ammonia Determination.. 3-70
3.3.4.3 Analysis of Nitrates 3-71
3.3.4.4 Analysis of Trace Elements 3-75
3.3.4.5, Analysis of Organic Compounds 3-79
3.3.4.0 Analysis of Total Carbon and Elemental
Carbon 3-80
3.3.5 Particle Morphology Measurements 3-81
3.3.6 Intercomparison of Particulate Matter Measurements... 3-81
3.3.7 Summary 3-83
3.4 MEASUREMENT TECHNIQUES FOR ACIDIC DEPOSITION. » .... 3-85
3.4.1 Introduction 3-85
3.4.2 U.S. Precipitation Studies 3-86
3.4.3 Analytical Techniques 3-8i
3.4.3.1 Introduction 3-89
3.4.3.2 Analysis of Acidic Deposition Samples 3-89
3.4.4 Interlaboratory Comparisons 3-93
3.5 REFERENCES..., 3-96
APPENDIX 3-A..... ......... 3-120
4. SOURCES AND EMISSIONS 4-1
4.1 INTRODUCTION , 4-1
4.2 OATA SOURCES AND ACCURACY 4-2
4.3 NATURAL SOURCES AND EMISSIONS 4-3
4.3.1 Terrestrial Dust. ......... 4-4
4.3.2 Sea Spray .... 4-7
4.3.3 Biogenic Emanations 4-7
4.3.4 Volcanic Emissions...... 4-9
4.3.5 Wildfires 4-10
4.4 MANMADE SOURCES AND EMISSIONS 4-11
4.4.1 Historical Emission Trends 4-11
4.4.2 Stationary Point Source Emissions. 4-13
VI
-------
CONTENTS (continued)
Page
4.4.2.1 Fuel Combustion 4-24
4.4.2.2 Industrial Processes 4-27
4.4.3 Industrial Process Fugitive Particulate Emissions 4-30
4.4.4 Nonindustrial Fugitive Particulate Emissions 4-33
4.4.5 Transportation Source Emissions 4-35
4.5 SUMMARY 4-36
4,6 REFERENCES 4-38
ENVIRONMENTAL CONCENTRATIONS AND EXPOSURE 5-1
5.1 INTRODUCTION 5-1
5.2 AMBIENT MEASUREMENTS OF SULFUR DIOXIDE. 5-2
5.2.1 Monitoring Factors. 5-4
5.2.2 Sulfur Dioxide Concentrations 5-5
5.2.3 Sulfur Dioxide Concentration by Site and Region 5-7
5.2.3.1 Analyses by Various Site Classifications 5-7
' 5.2.3.2 Regional Comparisons 5-7
5.2.4 Peak Localized Sulfur Dioxide Concentrations 5-12
5.2.4.1 1978 Highest Annual Average Concentrations... 5-12
5.2.4.2 1978 Highest Daily Average Concentrations 5-12
* 5.2.4.3 Highest 1-Hour Sulfur Dioxide Concentra-
tions-1978 National Aerometric Data Bank
(NADB) Data 5-12
5.2.5 Temporal Patterns in Sulfur Dioxide Concentrations.... 5-13
5.2.5.1 Diurnal Patterns 5-13
5.2.5.2 Seasonal Patterns 5-16
5.2.5.3 Yearly Trends 5-16
5.3 AMBIENT MEASUREMENTS OF SUSPENDED PARTICULATE MASS 5-22
5.3.1 Monitoring Factors 5-23
5.3.1.1 Sampling Frequency., 5-23
5.3.1.2 Monitor Location 5-27
5.3.2 Ambient Air TSP Values 5-27
5.3.3 TSP Concentrations by Site and Region 5-30
5.3.3.1 TSP by Site Classifications 5-31
5.3.3.2 Intracity Comparisons 5-31
5.3.3.3 Regional Differences in Background
Concentrations — . 5-33
5.3.3.4 Peak TSP Concentrations 5-33
5.3.4 Temporal Patterns in TSP Concentrations 5-35
5.3.4.1 Diurnal Patterns 5-35
5.3.4.2 Weekly Patterns 5-35
5.3.4.3 Seasonal Patterns 5-37
5.3.4.4 Yearly Trends 5-37
5,4 SIZE OF ATMOSPHERIC PARTICLES 5-46
5.4.1 Introduction 5-46
5.4.2 Size Distribution of Particle Mass 5-47
5.5 FINE PARTICLES IN AIR 5-57
5.5.1 Sulfates , 5-58
5.5.1.1 Spatial and Temporal Variations. 5-58
5.5.1.2 Urban Variations 5-64
5.5.2 Nitrates 5-73
5.5.3 Carbon and Organics 5-77
vi i
-------
CONTENTS (continued)
Pa
5.5,3.1 Physical Properties of Participate Qrganies..
5.5.3.2 Carbon and Total Organic Mass.... 5-79
5.5.3.3 Chemical Composition of Participate Organic
Matter 5-85
5.5.4 Metallic Components of Fine Particles. 5-87
5.5.4.1 Lead...... 5-92
5.5.4.2 Vanadium, Nickel, and Other Metals 5-92
5.5.5 Acidity of Atmospheric Aerosols 5-96
5.6 COARSE PARTICLES IN AIR. 5-99
5.6.1 Introduction 5-99
5.6.2 Elemental Analysis of Coarse Particles ,.. 5-100
5.6.3 Evidence from Microscopical Evaluation of Coarse
Particles 5-103
5.6.4 Fugitive Dust 5-106
5.6.5 Summary 5-109
5.7 SOURCE-APPORTIONMENT OR SOURCE-RECEPTOR MODELS 5-109
5.8 FACTORS INFLUENCING EXPOSURE. 5-115
5.8.1 Introduction... 5-115
5.8.2 Indoor Concentrations of Sulfur Dioxide 5-117
5.8.3 Particle Exposures Indoors 5-118
5.8";3.1 Introduction. 5-118
5.8.3.2 Coarse-Particle Concentrations Indoors 5-122
5.8j3.3 Fine Particles Indoors 5-127
5.8,4 Monitoring and Estimation of Personal Exposures 5-131
5.9 SUMMARY OF- ENVIRONMENTAL CONCENTRATIONS AND EXPOSURE 5-136
5.10 REFERENCES 5-139
6. ATMOSPHERIC TRANSPORT, TRANSFORMATION, AND DEPOSITION 6-1
6.1 INTRODUCTION 6-1
6.2 CHEMICAL TRANSFORMATION PROCESSES 6-1
6.2.1 Chemical Transformation of Sulfur Dioxide and
Particulate Matter 6-3
6.2.2 Field Measurements on the Rate of Sulfur Dioxide
Oxidation 6-3
6.3 PHYSICAL REMOVAL PROCESSES 6-6
6.3.1 Dry- Deposition. 6-7
6.3.1.1 Sulfur Dioxide Dry Deposition 6-8
6.3.1.2 Particle Dry Deposition 6-10
6.3.2 Precipitation Scavenging 6-17
6.3.2.1 Sulfur Dioxide Wet Removal 6-19
6.3.2.2 Particle Wet Removal 6-20
6.4 TRANSPORT AND DIFFUSION 6-23
6.4.1 The Planetary Boundary Layer 6-23
6.4.2 Horizontal Transport and Pollutant Residence Times 6-27
6.5 AIR QUALITY SIMULATION MODELING 6-30
6.5.1 Gaussian Plume Modeling Techniques 6-31
6.5.2 Long-Range Air Pollution Modeling 6-32
6.5.3 Model Evaluation and Data Bases 6-36
6.5.4 Atmospheric Budgets....... 6-37
6.6 SUMMARY 6-38
6.7 REFERENCES 6-39
viii
-------
CONTENTS (continued)
7. ACIDIC DEPOSITION
7.1 INTRODUCTION
7.1.1 Overview of the Problem
7.1.2 Ecosystem Dynamics
7.2 CAUSES OF ACIDIC PRECIPITATION.
7.2.1 Emissions of Sulfur and Nitrogen Oxides....
7.2.2 Transport of Nitrogen and Sulfur Oxides
7.2,3 Formation
7.2.3.1 Composition and pH of Precipitation
7.2.3.2 Geographic Extent of Acidic Precipitation
7.2.4 Acidic Deposition
7.3 EFFECTS OF ACIDIC DEPOSITION ,
7.3.1 Aquatic Ecosystems....
7.3.1.1 Acidification of Lakes and Streams..
7.3.1.2 Effects on Decomposition
7.3.1.3 Effect on Primary Producers and Primary
Productivity
7.3.1.4 Effects on Invertebrates
7.3.1.5 Effects on Fish
7.3.1.6 Effects on Vertebrates other than Fish..
7.3.2 Terrestrial Ecosystems
7.3.2.1 Effects on Soils
7.3.2.2 Effects on Vegetation,
7.3.2.3 Effects on Human Health
' 7.3.2.4 Effects of Acidic Precipitation on Materials,
7.4 ASSESSMENT OF SENSITIVE AREAS
7.4.1 Aquatic Ecosystems
7.4.2 Terrestrial Ecosystems
7.5 SUMMARY
7.6 REFERENCES
-------
CONTENTS (continued)
FIGURES
FIGURE
2-1 • The global sulfur cycle, showing the major reservoirs, pathways,
and forms of occurrence of sul fur ................. , .............. 2-4
2-2 Interrelations of pathways, processes, and properties of
sulfur oxides and particulate matter and effects... ---- ..... ....... 2-7 !
2-3 The distribution of species for the S02 • H20-HS03-SO|~ system
as a function of pH. Also, the ratio of the concentrations of
to the total quantity dissolved in water is shown.,...' ..... 2-11
2-4 Schematic of the polluted atmospheric photooxidation cycle ........ 2-18
2-5 The theoretical rate of reaction (percent per hour) of various
free-radical species on S02 is shown for a simulated sunlignt-
irradiated (solar zenith angle of 40°) polluted atmosphere..... ___ 2-20
2-6 Percentage conversion at midday of sulfur dioxide to sul fate
by HO and by HO, H02, and CH302 radicals as a function of °N
1 ati tude i n summer and wi nter ....................... . ......... .... 2-21
2-7 Solubility diagram for the H+-NHt-SO|~-H20 system at
equi 1 ibrium (30°C) ..... .... ..... . ................ . ............... 2-49
2-8 Growth of H+-NHt~SQi~ particles as a function of RH ....... ....... 2-50
2-9 Condensational growth and evaporation of (NH4)2S04 particles as
a function of relative humidity at 25°C .......................... 2-52
2-10 The equilibrium size of sul f uric acid solution droplets as a
function of relative humidity .................................... 2-54
2-11 NHs and HN03 partial pressures as a function of droplet' s nitrate
(C..Q-) and sulfate (Ccg2~) concentrations at 85 percent relative
humidity, 25°C .................. .... ....... . ____ . ......... . ...... 2-58
2-12 Frequency plots of number, surface, and volume distributions for
1969 Pasadena smog aerosol ............................ . .......... 2-71
2-13a Idealized size distribution for particles found in typical urban
aerosols (mainly from anthropogenic sources) under varying
weather conditions ......................... . . .................... 2-73
2~13b Idealized size distribution for atmospheric particles from
anthropogenic sources ........... ........... ...................... 2-73
2-13c Idealized size distribution for atmospheric particles from
natural sources in a marine setting .............................. 2-74
2-13d Idealized size distribution for atmospheric particles from
natural sources in a continental setting ......... ................ 2-74
2-14 Idealized representation of typical fine- and coanse-particle
mass and chemical composition distribution in an urban aerosol — 2-76
3-1 Respiratory deposition models used as patterns for sampler •
cutpoints ............................................... : T.~ ."..'... 3-25
3-2 Plots illustrating the relationship of particle number,
surface area, and volume distribution as a function of
particle size ...................... . ................... -. ......... 3-27
3-3 Typical ambient mass distribution data for particles
up to 200 pm ..................... . ............................... 3-28
3-4 Sampling effectiveness of a Hi-Vol sampler as a function of
-------
CONTENTS (continued)
Figure , Page
3-5 Sampling effectiveness of the dichotomous sampler inlet as a
function of windspeed , 3-33
3-6 Sampling effectiveness of the Wedding IP inlet. 3-34
3-7 Sampling effectiveness of UM-LBL IP in-let 3-35
3-8 Effect of sampler flowrate on the performance of a Hi-Vol for
30 pm particles at a windspeed of B km/hr 3-38
3-9 Separator efficiency and wall losses of the dichotomous
sampler at 2.5 jjm 3-41
3-10 Sampling effectiveness for the 3.5 jam cutpoint CHESS
cyclone sampler 3-43
3-11 Fraction of methylene blue particles deposited in a cyclone
sampler as a function of the aerodynamic particle diameter,...... 3-45
3-12 Sampling effectiveness for the size-selective inlet Hi-Vol
sampl er 3-46
3-13 Effect of windspeed upon cutpoint size of the size-selective
inlet 3-47
3-14 Effect of sampler flowrate on the sampling effectiveness of
the size-selective inlet Hi-Vol for a particle size of
15.2 pm and windspeed of 2 km/hr 3-48
3-15 An example of mass size distribution obtained using a cascade
impactor 3-50
'3-16 Fractional particle collection of the CHAMP fractionator inlet
at a sampler flowrate of 1133 liters/min under static windspeed
conditions 3-52
3-17 Efficiency of the single impaction stage of the CHAMP Hi-Vol
sampl er 3-53
3-18 Sampling effectiveness of the inlet alone and through the
entire flow system of the British Smoke Shade sampler 3-56
3-19 Response of a Piezoelectric Microbalance to relative humidity
for various particle types 3-60
3-20 Light scattering and absorption expressed per unit volume of
aerosol . , 3-61
5-1 Distribution of annual mean sulfur dioxide concentrations across
an urban complex, as a function of various spatial scales 5-3
5-2 Histogram delineating annual average sulfur dioxide concentrations
for valid continuous sampling sites in the United States in 1978. 5-6
5-3 Characterization of 1974-76 national SQ^ status is shown by
second highest 24-hour average concentration 5-10
5-4 Composite diurnal pattern of hourly sulfur dioxide concentrations
are shown for Watertown, Massachusetts, for December 1978........ 5-14
5-5 Monthly means of hourly sulfur dioxide concentrations are shown
for St. Louis (city site 26-4280-007, "Broadway & Hurck") for
February 1977 and 1978 5-15
5-6 Monthly means of hourly sulfur dioxide concentrations are
shown for Steubenville, Ohio (NOVAA site 36-6420-012) for
June 1976 and July 1977 5-17
5-7 Seasonal variations in S02 levels are shown for Steubenville,
St. Louis, and Watertown. 5-18
-------
CONTENTS (continued)
Figure Page
5-8 Annual average sulfur dioxide concentrations are shown for 32
urban NASN stations 5-19
5-9 Nationwide trends in annual average sulfur dioxide concentrations
from 1972 to 1977 are shown for 1233 sampling sites 5-20
5-10 Distribution shows the number of TSP observations per valid site
in 1978; total of 2882 sites 5-24
5-11 The 95 percent confidence intervals about an annual mean TSP
concentration of 75 ug/m3 is shown for various sampling
frequencies 5-26
5-12 Distribution of mean and 90th percentile TSP concentrations is
shown for valid 1978 sites 5-28
5-13 Histogram of number of sites against concentration shows that
over one-third of the sites had annual mean concentrations
between 40 and 60 ug/m3 5-29
5-14 Histogram of mean TSP levels by neighborhood shows lowest levels
in residential areas, higher levels in commercial areas, and
highest levels in industrial areas , — 5-32
5-15 Average estimated contributions to nonurban levels in the East,
Midwest, and West are most variable for transported secondary
and continental sources 5-34
5-16 Severity of TSP peak exposures is shown on the basis of the
90th percentile concentration. Four AQCR's did not report 5-36
5-17 Seasonal variations in urban, suburban, and rural areas
for four size ranges of particles 5-38
5-18 Monthly mean TSP concentrations are shown for the Northern Ohio
Valley Air Monitoring Headquarters, Steubenville, Ohio. No
cl ear seasonal pattern 1 s apparent. 5-39
5-19 Annual geometric mean TSP trends are shown for selected NASN
sites 5-40
5-20 (Top) Nationwide trends in annual mean total suspended
particulate concentrations from 1972 to 1977 are shown for
2707 sampling sites. (Bottom) Conventions for box plots 5-42
5-21 Regional trends of annual mean total suspended particulate
concentrations, 1972-1977, Eastern states 5-44
5-22 Regional trends of annual mean total suspended particulate
concentrations, 1972-1977, Western states 5-45
5-23 Linear-log plot of the volume distributions for the four
background distributions 5-49
5-24 Linear-log plot of the volume distributions for two urban
aerosols and a typical distribution measured in the Labadie
coal-fired power plant plume near St. Louis. Size distri-
butions measured above a few hundred meters above the
ground generally have a rather small coarse particle mode 5-50
5-25 Incursion of aged smog from Los Angeles at the Goldstone
tracking station in the Mojave Desert in California 5-51
5-26 Sudden growth of the coarse particle mode due to local dust
sources measured at the Hunter-Liggett Military Reservation
in California. This shows the independence of the
accumulation and coarse particle mode. 5-52
5-27 Inhalable particle network sites established as of
March 19, 1980 5-54
5-28 Contour maps of sulfate concentrations for 1974 are shown for:
(a) annual average; (b) winter average; (c) summer average 5-59
-------
CONTENTS (continued;
Figure Page
5-29 Intensive Sulfate Study area in eastern Canada shows the
geometric mean of the concentration of particulate soluble
sulfate during the study period. Units are micrograms of
sulfate per cubic meter , 5-61
5-30 Map of SURE regions shows locations of ground measurement
stations , 5-62
5-31 Cumulative plots show the frequency of sulfate concentrations
in the SURE region on the basis of the 1974-75 historical data... 5-63
5-32 Map shows the spatial distribution of number of days per month
that the sulfate concentration equaled or exceeded 10 MS/1"3 5-65
5-33 1977 seasonal patterns of S02 emissions and 24-hr average S02
and S04 ambient levels in the New York area are normalized to
the annual average values 5-66
5-34 Monthly variation in monthly mean of 24-hour average sulfate
concentration at downtown Los Angeles is compared with monthly
mean 1973 Los Angeles County power plant S02 emissions 5-67
5-35 Map shows annual mean 24-hr average sulfate levels in micrograms
per cubic meter in the New York area, based on 1972 data from
Lynn et al. (1975). Squares are locations of three CHAMP site
stations. The fourth station is at the tip of Long Island
about 160 km from Manhattan 5-70
5-36 Distribution of annual average sulfate concentration in
micrograms per cubic meter in the greater Los Angeles area
based on 1972-1974 data.... 5-71
5-37 Map shows U.S. mean annual ambient nitrate levels in micrograms
per cubic meter. ,...*.... 5-74
5-38 Mean nitrate concentrations in micrograms per cubic meter at
nonurban sites in the U.S. based on valid annual average from
1971 through 1974...... 5-75
5-39 Calculated distributions of aerosol constituents for two aerosol
samples taken in the Los Angeles Basin 5-82
5-40 Benzo(a)pyrene seasonality and trends (1966-75) in the
50th and 90th percentiles for 34 NASN urban sites 5-84
5-41 Seasonal patterns and trends in quarterly average urban lead
concentrations 5-94
5-42 Regional trends in the 90th percentile of the annual averages
for vanadium 5-95
5-43 Seasonal variation in quarterly averages for nickel and
vanadium at urban sites in the northeast 5-97
5-44 Elemental compositions of some coarse particle components 5-102
5-45 Diurnal variation- of particle concentrations and Plymouth
Avenue traffic volume at Falls River, Mass., during March
through June 1979 (weekdays only), shows contribution from
reentrained particles 5-108
5-46 Types of receptor source apportionment models 5-110
5-47 Source contributions at RAPS sites estimated by chemical element
balance are illustrated 5-112
5-48 Monthly averages of size fractionated Denver aerosol mass and
composition for January and May, 1979 5-113
5-49 Aerosol source in downtown Portland, annual stratified
arithmetic average. Does not include the 17%, on the average,
of material collected with the standard Hi-Vol sampler which
was not collected and characterized with the ERT-TSP sampler 5-114
xm
-------
CONTENTS (continued)
Page
"Retention of Au198 labeled FegOa particles from human lungs—
comparison of 9 non-smoking subjects with three smoker
subjects 5-116
5-51 Annual S02 concentrations averaged across each community's
indoor and outdoor network (May 1977-April 1978). 5-119
5-52 Monthly mean S02 concentrations averaged across Watertown's
indoor and outdoor network (November 1976-April 1978) 5-120
5-53 Monthly mean S02 concentrations averaged across Steubenville's
indoor and outdoor network (November 1976-April 1978) 5-121
5-54 Annual respirable particulate concentrations averaged across
each community's indoor and outdoor network (May 1977-
Apri 1 1978) 5-130
5-55 An example of personal exposure to respirable particles 5-132
5-56 Normalized distribution of personal (12-hour) exposure samples
((jm/m3) for non-smoke exposed and smoke exposed samples 5-134
5-57 Daily mean indoor/outdoor and personal concentrations (ng/m3)
of respirable particles. Daily means averaged over 24 homes
and outdoor locations and up to 46 personal samples. Samples
col 1 ected during May and June 1979 5-135
6-1 Pathway processes of airborne pollutants. 6-2
6-2 Predicted deposition velocities at 1 |j for 1^=30 cm s and
-3
particle densities of 1,4, and 11.5 g cm 6-16
6-3 Basic factors influencing precipitation scavenging 6-18
6-4 Relationship between rain scavenging rates and particle
size. 6-24
6-5 Percentages of aerosol particles of various sizes removed by
precipitation scavenging 6-25
6-6 Estimated residence times for select pollutant species and
their associated horizontal transport scale 6-29
6-7 Trajectory modeling approaches are shown 6-34
7-1 Schematic representation of the nitrogen cycle, emphasizes human
activities that affect fluxes of nitrogen 7-10
7-2 Law of tolerance 7-12
7-3 Historical patterns of fossil fuel consumption in the
United States 7-15
7-4 Forms of coal usage in the United States 7-16
7-5a Trends in emissions of sulfur dioxide 7-17
7-5b Trends in emissions of nitrogen oxides 7-17
7-6 Characterization of U.S. SO emissions density by state 7-18
7-7 Characterization of U.S. NO emissions density by state 7-19
7-8 Trends in mean annual concentrations of sulfate, ammonium,
and nitrate in precipitation 7-24
7-9 Comparison of weighted mean monthly concentrations of sulfate in
incident precipitation collected in Walker Branch Watershed,
Tennessee (WBW) and four MAP3S precipitation chemistry monitoring
stations in New York, Pennsylvania, and Virginia 7-27
7-10 Seasonal variations in pH (A) and ammonium and nitrate
concentrations (B) in wet-only precipitation at Gainesville,
Florida 7-28
xiv
-------
CONTENTS (continued)
Seasonal variations of precipitation pH in the New York
Metropol Han Area 7-31
7-12 History of acidic precipitation at various sites in and
adjacent to State of New York 7-32
7-13 pH of rain sample, as measured in the laboratory, used in
combination with the reported amount of precipitation 7-35
7-14 Annual mass transfer rates of sulfate expressed as a percentage
of the estimated total annual flux of the element to the forest
floor beneath a representative chestnut oak stand 7-37
7-15 Schematic representation of the hydrogen ion cycle 7-39
7-16 pH and calcium concentrations in lakes in northern and northwestern
Norway sampled as part of the regional survey of 1975, in lakes in
northwestern Norway sampled in 1977 (o) and in lakes in southernmost
and southeastern Norway sampled in 1974 (o) 7-43
7-17 The pH value and sulfur loads in lake waters with extremely
sensitive surroundings (curve 1) and with slightly less
sensitive surroundings (curve 2) 7-44
7-18 Total dissolved Al as a function of pH level in lakes in
acidified areas in Europe and North America., • 7-45
7-19 pH levels in Little Moose Lake, Adirondack region of New York
State, at a depth of 3 meters and at the lake outlet 7-47
7-20 Numbers of phytoplankton species in 60 lakes having different
pH values on the Swedish west coast, August 1976 are compared 7-51
7-21 Percentage distribution of phytoplankton species and their
biomasses. September 1972, West Coast of Sweden. ... 7-52
7-22 The number of species of crustacean zooplankton observed in
57 lakes during a synoptic survey of lakes in southern Norway 7-56
7-23 Frequency distribution of pH and fish population status in
Adirondack Mountain lakes greater than 610 meters elevation 7-60
7-24 Frequency distribution of pH and fish population status in 40
Adirondack lakes greater than 610 meters elevation,
surveyed during the period 1929-1937 and again in 1975 7-61
7-25 Norwegian salmon fishery statistics for 68 unacidified and 7
acidified rivers 7-62
7-26 Showing the exchangeable ions of a soil with pH 7, the soil
solution composition, and the replacement of Na by H from
acid rain 7-71
7-27 Regions in North America with lakes that are sensitive to acid-
ification by acid precipitation by virtue of their underlying
bedrock characteristics 7-96
7-28 Soils of the eastern United States sensitive to acid rainfall
are mapped. 7-100
xv
-------
CONTENTS (continued)
TABLES
Table Page
2-1 Estimates of environmental sulfur annual fluxes (tg/year) 2-5
2-2 Characteristic times and lengths for observation of effects.... 2-6
2-3 Dilute sulfur dioxide-water system. 2-10
2-4 Relative strengths of acids in water solution (25°C) 2-13
2-5 Rate constants for hydroxyl, peroxyl, and methoxyl radicals 2-15
2-6 Investigations of S02 - 02 aqueous systems 2-24
2-7 Values of k and kh for reaction type 1 2-26
2-8 Values of 1C for reaction types 2 2-27
2-9 Investigations of S02 - manganese - 02 aqueous system 2-29
2-10 Rate expression for the manganese-catalyzed oxidation 2-30
2-11 Investigations of SQ2 ~ iron - 02 aqueous system. 2-31
2-12 Rate expression for the iron-catalyzed oxidation 2-32
2-13 Investigations of S02 - copper - 02 aqueous systems .... 2-34
2-14 Estimates of S02 oxidation rates in well-mixed troposphere 2-40
2-15 Estimate of global tropospheric particulate matter production
rates 2-43
2-16 Particle shapes and source types 2-46
2-17 Deliquescence and efflorescence points of salt particles... 2-51
2-18 Sulfuric acid solution values (25°C) 2-55
2-19 Conditions for the single-particle regime 2-60
2-20 Mass transport parameters for air 2-63
2-21 Dependence of particle behavior on air temperature, pressure,
and viscosity 2-64
2-22 Classification of major chemical species associated with
atmospheric particles 2-77
2-23 Application of GDE to describe particle size evaluation 2-82
3-1 Temperature effect on collected S02-TCM samples (EPA reference
method) 3-6
3-2 Performance specifications for EPA equivalent methods for S02
(continuous analyzers) 3-18
3-3 List of EPA designated equivalent methods for S02 (continuous
analyzers) 3-19
3-4 Interferent test concentrations (parts per million) used in the
testing of EPA equivalent methods for S02 3-20
3-5 Comparison of EPA designated equivalent methods for S02
(continuous analyzers) 3-22
3-6 Recommended physical /chemical parameters for analysis 3-90
3-7 Results of WMO intercomparisons on synthetic precipitation
samples. 3-94
3-8 Coefficients of variation of WMO intercomparisons on
synthetic precipitation samples 3-95
4-1 Two EPA estimates of 1977 emissions of particulate matter,
and sulfur oxides (106 metric tons per year) 4-2
4-2 Summary of natural source particulate and sulfur emissions 4-5
4-3 Aerosol enrichment factors relative to Al , 4-6
4-4 Summary of estimated annual manmade emissions (1978) 4-11
4-5 (a) National estimates of particulate emissions (106 metric
tons per year) 4-13
(b) National estimates of SO emissions (106 metric tons per
year) 4-13
xvi
-------
CONTENTS (continued)
Tab! e Page
-"4-6 1978 estimates of participate and sulfur oxide emissions
from stationary point sources 4-14
4-7 State-by-State listing of total particulate and sulfur oxide
emissions from stationary point sources (1977), population,
and density factors 4-16
4~8 Examples of uncontrolled particulate emission characteristics.. 4-20
4-9 Size-specific particulate emissions from coal-fired boilers. ., 4-23
4-10 Trace element air emissions vs. solid waste: percent from
conventional stationary fuel combustion sources, and total
(metric tons per year) 4-25
4-11 Uncontrolled industrial process fugitive particulate emissions. 4-31
"4-12 Estimated annual particulate emissions from nonindustrial
fugitive sources.... 4-34
4-13 Estimated particle size distributions for several
nonindustrial fugitive source categories in California's
south coast air basin 4-34
5-1 Crosstabulation of annual mean SQ% concentration by method
(bubbler or continuous) for population-oriented and for
source-oriented center-city sites. ,.... 5-8
5-2 Continuous S02 monitor results by region, ug/m3 5-9
5-3 Eleven S02 monitoring sites with the highest annual mean
concentrations in 1978 (valid continuous sites only) 5-11
5-4 Comparison of frequency distribution of S02 concentration (ppm)
during 1962-67 and during 1977. j-',\.
5-5 Range of annual geometric mean concentrations in areas with
high TSP concentrations in 1977 5-33
5-6 Regional summaries of TSP values from valid monitors 5-43
5-7 Fine and coarse aerosol concentrations from some urban
measurements compared to clean areas 5-53
5-8 Fine fraction and coarse fraction dichotomous sampling by
Environmental Science Research Lab, US EPA in four locations... 5-55
5-9 Recent dichotomous sampler and TSP data from selected sites—
arithmetic averages 5-56
5-10 Some characteristics of pollution in the New York and
Los Angeles areas ^ . . 5-M
5-11 Primary ranking of variables for correlating airborne S0|
in two cities based on a stepwise linear reg--p . , ..A *v;
15 variables from CHAMP and related monitor ing b ,«.,.,
5-12 Typical values of aerosol concentration for different
geographic areas (annual averages) 5-81
5-13 Annual averages of organic fractions in TSP, New York City,
dispersion normalized 5-83
5-14 Composition of the organic fraction of airborne PM
collected in Detroit 5-85
5-15 Comparison of urban and nonurban annual average concentrations
for selected metals, 1970-74 (ug/m3) 5-88
5-16 Ratios of urban (U) to suburban (S) concentrations in air,
Cleveland, Ohio, area 5-09
5-17 Correlations of chemical content with particle sive , ... .. S-OO
5-18 Particulate analyses from selected urban locations
5-19 Trends in reported urban metal concentrations" and their possible
causes 5-93
xvi i
-------
CONTENTS (continued)
Table Page
5-20 Coarse particle silicon, aluminum, calcium, and iron 5-101
5-21 Relative amounts of fine, coarse, and super-coarse particles at
selected sites 5-104
5-22 Fourteen-city study - microscopical identification of coarse
particles collected in urban atmospheres 5-105
5-23 Summary of indoor/outdoor (I/O) PM monitoring studies by
method 5-123
5-24 Measurements in principal room of study 5-128
5-25 Measurements in various closed rooms.. 5-128
5-26 Respirable particulate concentrations outdoors and indoors by
amount of smoking , 5-129
6-1 Field measurements on the rates of S02 oxidation in plumes..... 6-4
6-2 Average dry deposition velocity of S02 by surface type 6-9
6-3 Laboratory measurements of deposition velocities of particles.. 6-11
6-4 Field measurements of deposition velocities of particles ... 6-13
6-5 Predicted particle deposition velocities . 6-17
6-6 Field measurements of scavenging coefficients of particles 6-21
6-7 Summary of long range transport air pollution models 6-35
7-1 Composition of ecosystems. 7-8
7-2 Mean pH values in the New York metropolitan area 7-30
7-3 Storm type classification 7-30
7-4 Chemical composition (Mean ± standard deviation) of acid lakes
(pH <5) acidic precipitation (pH <4.5), and of soft-water lakes
in areas not subject to'highly acidic precipitation
(pH >4.8) 7-41
7-5 pH levels identified in field surveys as critical to
long-term survival of fish populations 7-63
7-6 Changes in aquatic biota likely to occur with increasing
acidity 7-67
7-7 Summary of effects on aquatic organisms associated with a
range in pH -. 7-68
7-8 Potential effects of acid precipitation on soils 7-72
7-9 Types of direct, visible injury reported in response to acidic
wet deposition 7-81
7-10 Thresholds for visible injury and growth effects associated with
experimental studies of wet deposition of acidic substances 7-84
7-11 Lead and copper concentration and pH of water from pipes
carrying outflow from Hinckley Basin and Hanns and Steele
Creek Basi n, near Amsterdam, New York 7-91
7-12 Composition of rain and hoarfrost at Headingley, Leeds 7-93
7-13 The sensitivity to acid precipitation based, on: buffering
capacity against pH-change, retention of H , and adverse
effects on soils 7-98
xvm
-------
2. PHYSICS AND CHEMISTRY OF SULFUR OXIDES AND PARTICULATE MATTER
2.1 INTRODUCTION
This chapter presents the current state of knowledge of the chemistry and physics of
sulfur oxides (SO ) and particulate matter (PM) that is pertinent to tropospheric phenomena,
effects, and sampling methodology. The 1970 Air Quality Criteria for Sulfur Oxides (National
Air Pollution Control Administration, 1970) and the 1969 Air Quality Criteria for Particulate
Matter (National Air Pollution Control Administration, 1969) at the time of their publication
adequately described the existing knowledge of the ambient chemistry and physics of sulfur
oxides and particulate matter. However, significant progress has been made since that time in
understanding tropospheric properties and processes. While this chapter focuses mainly on
advances made in the past decade, earlier work is mentioned for the sake of comprehensiveness.
This chapter is organized into three principal parts, as follows:
A. Atmospheric domain and processes
--global sulfur cycle
--atmospheric sulfur cycle
—pathways and processes
B. Physics and chemistry of sulfur oxides
--gaseous physical properties
—solution physical properties
—gas-phase chemical reactions (elementary rate consists for S0« oxidation; influence
of volatile organics and nitrogen oxides)
—solution-phase chemical reactions (reactions kinetics for oxidation, by 0,, H20P,
and .catalysts; limitations of reported studies)
—surface chemical reactions (metal oxides and carbon)
C. Physics and chemistry of particulate matter
—definitions of aerosol science terms
—physical -properties of gases and particles (size, shape, density, morphology, charg-
ing, adhesion, vapor pressure; optics are discussed in Chapter 9)
—dynamics of single particles (sedimentation, impaction, diffusion, electrodynamics)
--formation and growth of particles (nucleation, coagulation, condensation, gas-
particle chemical reaction)
--characterization of atmospheric aerosol (size, distribution by number, area,
volume, and mass distributions; composition of fine and coarse particle mass
fractions)
Available evidence leads to the following conclusions:
A. The physics and chemistry of S0?
2-1
-------
i. me tnermoaynamtc properties, molecular structure and bonding, and electro-
magnetic absorption spectra in air and dissolved in water are well established, except for
hygroscopic/deliquescent properties of internally mixed salts.
2. Of the homogeneous gas-phase SOp oxidation reactions, only three have been
identified as likely being significant in the troposphere:
a. HO radical attack on S02
b. HOp radical attack on SOp
c, CHgO,, radical attack on SOp
3. The auto-oxidation (uncatalyzed) reaction of SOp dissolved in liquid water is
too slow to be an important reaction in the troposphere.
4. The Mn(II)- and Fe(III)-catalyzed oxidation of SOp dissolved in liquid water
may be an important reaction in the troposphere. However, there is serious doubt regarding
the rate expression for th Mn(II)-catalyzed oxidation.
5. The effectiveness of Cu(II), V(V), V(IV), Ni(II), Zn(II), and Pb(II) as cata-
lysts for the oxidation of SOp dissolved in liquid water are unknown.
6. Lack of data on the effectiveness of dissolved organics and of bicarbonate ion
(HCO») as inhibitors prevents the confident use of aqueous-phase SOp catalyzed oxidation rate
expressions for tropospheric model prediction.
7. Elemental carbon (soot) particles coated with an aqueous film may be important
catalysts for S0? oxidation in the troposphere.
8. The reaction rate expressions of dissolved 03 and dissolved N0« with dissolved
SOp species are known, but these reactions appear to be ineffective for sulfate formation in
the troposphere.
9. The rate expression for dissolved HpOp and dissolved SO, species is known and
appears to be a highly effective reaction for sulfate formation in the troposphere.
10. SO, reactions with solid-particle surfaces are not effective for sustained
2-
SO* formation in the troposphere.
B. The physics and chemistry of particulate matter
1. The physical properties of gases that affect aerosol behavior are well known.
2. The physical characteristics of tropospheric particles are highly variable, but
the physics through which they influence aerosol behavior is well known.
3. The particle mass distribution function (AM/A log- Diam. v. log Diam.) for
tropospheric aerosols over land is often multimodal. The fine particles (diameter less than
about 2 |jm) may have two (or more) modes, usually at about 0.02 urn and at about 0.2 pm. The
coarse particles (greater than 2.5 pm) generally have one mode in the range 5 to 50 |jm.
4. The mass composition of the coarse particles is dominated by minerals whose
direct source types are well-known.
2-2
-------
5. The mass composition of the fine particles Is dominated by so|~, organfcs,
elemental carbon (soot), NO- and NH, whose direct and indirect source types are quantitatively
well known.
2~ -
6. The detailed chemical pathways for forming the SO, , organics, and NO- found in
the fine particles have not been established.
7. Strong acids are often found in the fine mass fraction, while bases are found
in the coarse mass fraction.
8. The molecular composition of the organic components (generally found in the
fine mass fraction) is not well characterized.
9. Water is the major constituent of the particle mass, but the deliquescence and
hygroscopic properties of mixed salts cannot be predicted reliably.
10. The dynamics of motion of a particle with a diameter of less than 10 \im can
best be described in terms of the physical characteristics of the particle, the force fields
present, and the motion of the suspending gas. However, this ability does not adequately
extend to larger particles, especially in the presence of nonsteady force fields and motions
of the suspending gas. This limitation requires empirical design of inlets or the use of
non-aspirating methods for particles with diameters greater than about 10 jjm.
11. Fundamental questions remain about the aerosol processes of nucleation, con-
densation, and coagulation, but these processes are understood sufficiently to explain and
predict the qualitative behavior of aerosols in the troposphere.
12. Atmospheric observations confirm theoretical predictions that condensation,
gas-particle reactions, and coagulation are important processes only for the growth of fine
particles as opposed to coarse particles.
2.2 ATMOSPHERIC DOMAIN AND PROCESSES
Scientists are interested in the physical properties and chemistry of SO and PM because
sulfur has an important natural cycle (see Figure 2-1) in the environment. It undergoes
various oxidation and reduction reactions and trans!ocations among the atmosphere, biosphere,
hydrosphere, pedosphere, and lithosphere. Human activity (especially fossil-fuel combustion)
has added a major perturbation to the natural cycle, and possibly modified natural rates and
reservoirs. The fluxes of sulfur translocation between reservoirs have been estimated for the
paths numbered in Figure 2-1. Several estimates of annual fluxes are presented in Table 2-1.
The agreement among the reported values is not good; for example, the estimates of annual
anthropogenic sulfur fluxes to the atmosphere range from 11 to 45 percent of the total sulfur
involved in the atmospheric balance. The global cycles of carbon and nitrogen and their
mutual inter-actions with the sulfur cycle are important, but are too complex to present here.
While the global sulfur cycle (Figure 2-1) and the annual fluxes of sulfur between com-
partments (Table 2-1) provide a broad view of the processes that may lead to adverse impacts
upon mankind and ecological systems, the global scale is clearly beyond the scope of this
2-3
-------
ATMOSPHERIC :
Figure 2-1. The global sulfur cycle, showing the major reservoirs, pathways, and forms of
occurrence of sulfur. Figures enclosed in circles (e.g., 1) refer to the individual fluxes and
correspond to figures in column 1, Table 2-1.
Source: Moss (1978).
2-4
-------
TABLE 2-1. ESTIMATES3 OF ENVIRONMENTAL SULFUR ANNUAL FLUXES (TG/YEAR)
rsj
tn
Source of sulfur
Biological decay (land)
Biological decay (ocean)
Volcanic activity
Sea spray (total)
To ocean
To land
Anthropogenic
Precipitation (land)
Dry deposition
Absorption (vegetation)
Precipitation and dry
deposition (ocean)
Absorption (ocean)
Total sulfur involved in
atmospheric balance
Atmospheric balance
Land -> sea
Sea -» land
Fertilizer
Rock weathering
Pedosphere •* river runoff
Total river runoff
Flux number
in Figure 2-1
1
2
3
4
4i
42
5
6
7
8 ,
9
10 '
'
11
12
13
14
Eriksson
(1960,1963)
110
170
45
(40)
(5)
40
65
100
75
100
100
365
-10
5
10
15
55
80
Robinson
and Robbins
(1968,1970)
68
30
--
44
—
—
70
70
20
26
71
25
212
+26
4
: 11
14
J48
73
Kellogg
et al., (1972)
90
1.5
47
(43)
(4)
50
86
10
15
72
—
183
* \
• ' +5
- 4
—
.
—
"*™*
Friend
(1973)
58
48
2
44
(40)
(4)
65
86
20
15
71
25
217
+8
4
26
42
89
136
Granat
et al. (1976)
5
27
3
44
(40)
(4)
65
43
yo
i.O
70
/ -3
144
+18
17
—
—
__
122
Note; The numbers in parentheses for sub-pathways 4-. and 4- are estimates of the decomposition of the total pathway 4.
Sources: As cited in each column and, in part, Friend (1973, Table 4).
-------
document. Also, the sulfur and particulate matter emissions are not uniformly distributed
over the land mass of the United States, nor is the time scale of one year adequately sensi-
tive to relate emissions to effects.
Cause-effect relationships are described in terms of length and time scales. Conse-
quently, any discussion of the atmospheric physics and chemistry of sulfur oxides and par-
ticulate matter must be in terms of those length and time scales. The characteristic time and
length scales for typical effects are shown in Table 2-2; also given are the parameters that
control the functions relating the effects to pollutants. The relationships of emissions to
effects, as shown in Table 2-2, require that we understand the physics and chemistry of sulfur
oxides and particulate matter on time scales of one hour to decades and length scales of 1 cm
to thousands of kilometers. With these constraints, our attention is focused on that portion
of the global sulfur cycle that consists of the perturbed atmosphere over land surfaces. Most
of the natural and anthropogenic emissions of sulfur and PM are contained within the tropo-
sphere, which is the layer of air contained in the zone from ground level to a height of 12 ±
5 km. This zone contains most of the pollutants emitted into the atmosphere.
Thus, in order to understand the .relationships between sources and effects, it is
necessary to have detailed knowledge of the pathways, properties, and processes that are shown
in Figure 2-2.
TABLE 2-2. CHARACTERISTIC TIMES AND LENGTHS FOR OBSERVATION OF EFFECTS
Types of Effects
Function
Time
Length
Damage to ecosystems and
materials due to S02 and
particulate mass deposition
(see Chapters 7, 8 and 10)
Loss of visual quality
(see Chapter 9)
Climate modification
(see Chapter 9)
Damage to human lungs
due to S02 and particulate
inhalation/deposition (see
Chapters 11 & 12)
Acid flux (= acid and Hours to 10-10 km
S02 mass concentration years
x deposition velocity)
Mass concentration, Hours to 10-10 km
particle size days
distribution, and
composition
Atmospheric burden Decades Global
of particle mass,
particle size, and
composition
Mass concentration, Hours to < 1 cm
particle size years
distribution,
composition
2-6
-------
SOURCES
DISPERSION
AND
TRANSPORTATION
so2
- PHYSICAL
PROPERTIES
- CHEMISTRY
/ i I i
DRY
REMOVAL
ECO-SYSTEMS
TRANSFORMATIONS
WET
REMOVAL
EFFECTS
1
PARTICULATE MATTER
- PHYSICAL
PROPERTIES
- DYNAMICS
-CHEMISTRY
DRY
REMOVAL
WET
REMOVAL
ECO-SYSTEMS
Figure 2-2. Inter-relations of pathways, processes, and properties of sulfur oxides and particulate mat-
ter and effects.
2-7
-------
This chapter discusses the state of knowledge of physical properties, chemistry, and
gas-to-aerosol transformations. The dry and wet removal pathways are discussed in Chapter 6,
which also addresses modeling of atmospheric dispersion, transport, transformation, and
removal.
2.3 PHYSICS AND CHEMISTRY OF SULFUR OXIDES
Knowledge of the physics and chemistry of sulfur oxides is necessary for designing satis-
factory samplers and monitors, understanding relationships between sources and effects, and
understanding important tropospheric processes, e.g., chemical transformations and deposition.
In Section 2.3, the physical properties and reaction chemistry of sulfur oxides in the
gas and solution phases are reviewed. Current knowledge in these important areas is sum-
marized at the end of each of the subsections.
2.3.1 Physical Propertiesof Sulfur Oxides in the Gas Phase
The four known monomeric sulfur oxides are sulfur monoxide (SO), sulfur dioxide (S0?),
sulfur trioxide (SO,), and disulfur monoxide (S^O). Of these, only SO, is present at signi-
ficant concentrations in the gas phase of the troposphere. S0~ is emitted directly into the
atmosphere by combusion and manufacturing sources and is formed in the atmosphere by oxidation
of SOy- However, because of its high affinity for water (H^O), it reacts within milliseconds
to form sulfuric acid (H2SO,). Polymeric sulfur oxides are known to exist, but they are not
stable in the presence of H,0 vapor and are not found in the atmosphere.
Since the standard enthalpy of formation AH° of S02 is -70.9 kcal/mole (25°C), S02 is
thermodynamically stable with regard to its formation from the elements (Glasstone, 1947).
SO- is capable of being oxidized to SO- (AH° = -94.4 kcal/mole), which yields H_SQ, in the
atmosphere (the important tropospheric reactions are discussed in Sections 2.3.3 to 2.3.5).
S02 is also capable of being reduced by reaction with H,S to form elemental S (the Clauss
Reaction). This reaction is important commercially but is not thought to be important in the
troposphere for removing S09; however, it is the likely formation pathway for formation of the
elemental S found in urban particles. The physical properties of SQ_, including its molecular
structure and bonding, vapor pressure of liquid and solid phases, electromagnetic absorption
(ultraviolet, visible, and infrared) spectra, and thermodynamic constants are well established.
Extensive descriptions and references to original work are given by Schenk and Steudel (1968)
and Schroeter (1966).
The physical properties of gaseous S02 are well known.
2.3.2 Solution Physical Properties
Knowledge of the physical properties of dissolved S02 solution species and sulfates is
required for sampler design, interpretation of laboratory measurements of S02 oxidation, and
modeling SO, oxidation in particles, fog, and rain.
£ y-
2.3.2.1 Sulfur Dioxide—SO^ dissolves in H2G to form these species: S02«H20, HS03, and S03 .
Although the formation of sulfurous acid, H2S03$ is often postulated instead of SQ2-H20, it
2-8
-------
has not been observed (Lyons and Nickless, 1968). The electronic absorption spectra (Hayon et
al., 1972), redox potentials (Valensi et al., 1966), and structure and bonding (Lyons and
« *?»•
Nickless, 1968) of HSO_ and S0» are known. The formation of these specie's in water^occurs
through the equilibrium reactions given in Table 2-3. Eigen et al. (1961) measured -'the
forward (k+,) and reverse (k ,) rate constants at 20°C for reaction:
S02-H20 * H+ -t- HSOg ' ' (2-1)'
and found that
k+1 = 3.4 x JLO^'1 , ;
k.j = 2 x loW1, . ' " ""b
which are in good agreement with the thermodynamic value of K., in Table 2-3. These measure-
"-I- ,
ments are important because they demonstrate that the SCU'hLO - HS03 reaction will achieve
equilibrium within 1 |js of a perturbation. The rate constants k ? and k_? for the reaction
HSQ~ $ H+ + SO^" ' • ' (2-2)
i - J,,f~ > ~*
are unknown. It is reasonable that the value of the protonation rate constant (k_«) is less
than the theoretical diffusion limit (~5 x 10 Ms), but greater than k ,. The expected
3 -1 2-
range of k+2 is therefore (0.008-2) x 10 s , which means that SOt will achieve equilibrium
concentration within 0.5-125 ms of a perturbation. Thus, the equilibrium distribution of
2-
SOp'HpQ, NSC,, and SO- is expected to be achieved with a relaxation time of 0.5 to 125 ms.
This time is too short to impact SO, formation rates in particles, mists, and rain; that is,
equilibrium conditions can be assumed to be continuously satisfied in these liquid systems.
However, the relaxation times need to be considered in interpreting the kinetics of rapid
oxidation^ such as measured in flash photolysis and flash radiolysis experiments.
,. o— 4. '< f
An important feature of the SQ^'HJ) - HSO, - S03 system is the influence of H in govern-
ing the distribution of these species and the ratio of S0?-- .. to total dissolved S(IV) specfes
concentrations, as shown in Figure 2-3. The oxidation rate of this system is often pH de-
pendent, indicating different oxidation rates for the three species. The oxfdation reactions
are discussed in Section 2.3.4.
i
Sulfite ion forms stable complexes with many metal ions, especially those in Periodic
Group VIII (Lyons and Nickless, 1968), The possible formation and significance' of stable
transition metal ion-sulfite complexes have been suggested (Eatough et al., 1978; Hilton et
al., 1979), but contrary evidence has been reported (Dasgupta et al., 1979). At this time,
the issue is unresolved (Eatough et al., 1979). The formation of the stable complex di-
chlorosulfonatomercurate ion is the basis of the West-Gaeke method for determining S02 in the
air (see Chapter 3).
The physical properties of dissolved SO- and its water-association products are well
known.
2-9
-------
TABLE 2-3. DILUTE SULFUR DIOXIDE-WATER SYSTEM
Reaction Constant (25°C)
S09, , + H90,0, 5 S09-H90 H = 0.0332
£\.y J *•" \.-**.jF £. €m
S02-H20 * H+ + HSO~ KA1 = 1.39 x 10"2
pKA1 = 1.86
HSOg ? H"1" + S02~ KA2 = 4 X 10"8
pKA2 = 7-4°
Notes:
1. H = Henry's law constant (dimensionless)
= ^^2(a"i mo^ar concentration)/(SO--HpO molar concentration)
Source of value: Hales and Sutter (1973)
2. K«, = dissociation constant, mole/liter
-H^] (for dilute solutions)
where
a. = activity of species i, mole/liter
[i] = concentration of species i, mole/liter
Source of value: Huss and Eckert (1977)
3. K«2 = dissociation constant, mole/liter
(for dilute solutions)
Source of value: Salomaa et al. (1969)
4, The ratio of dilute concentrations of S0?r v to the total quantity dissolved
in water is given by:
2(g)
CS(IV)]
[S02-H20] + [HS033 + [SOg I
2-10
-------
' MOLAR CONCENTRATION \ \
.* RATIO \ »
/ I I I I IV \l
1 234567 8
10'° —
10
,-7
Z
o
LU
O
Z
o
o
o
10 11 12
Figure 2-3. The distribution of species for the SO2-H2O—HSO3~—SO32"
system as a function of pH. Also, the ratio of the concentrations of SO2(g) to
the total quantity dissolved in water is shown; see Table 2-3, Note 4.
2-n
-------
2.3.2.2 Sulfur Tri oxide and Sulfur ic Acid-- Know! edge of the properties of SO, and H^SO. are
important for the design of samplers and monitors, and for understanding the behavior of
ambient particles, fogs, and rain.
SO, has a high affinity for water and is not present at significant concentrations in the
atmosphere. Free SO- molecules quickly react with water molecules and droplets to form H?SO,
water solution droplets. The vapor pressure of S03 over H^SO. water solutions is extremely
low; the vapor pressure of hLQ over hLSO* water solution is an important parameter governing
nucleation of particles and the size and pH of water droplets in the atmosphere. (See dis-
cussion in Section 2.4.4.)
H2$0. is the only important strong sulfur oxy acid in the troposphere. Its solution
properties are well known. H^SO. is a strong dibasic acid that reacts with water:
(2-3)
HSO~ * H* + SO^" (2-4)
For water systems likely to be present in the troposphere, the first dissociation can be
considered to be complete. The pK. of the first dissociation is -3 and of the second disso-
ciation is ~2 (Robinson and Stokes, 1970). The relative strengths of acids likely to be found
in tropospheric particles are shown in Table 2-4. Thus, for pH >3, the hLSO. - H00 system can
+ p- '£.**£.
adequately be described in terms of H and SO, ; for lower pH's, it is often necessary to
consider the presence of HSO».
Sulfuric acid is a strong dibasic acid; however, it is not a very strong oxidizing agent.
Metals below hydrogen in the electromotive series are not oxidized by cold concentrated or
dilute HnSO*. The properties of H^SO, are well known (Cotton and Wilkerson, 1967; Robinson
and Stokes, 1970; Gillespie, 1968).
Host sulfate salts are soluble; the only important exceptions in the troposphere are
CaSO. and PbSO.. The properties of tropospheric aerosols are influenced by NhLHSO. and
(NH4)2S04 (see Section 2.4.4).
2.3.3 Gas-Phase Chemical Reactions of Sulfur Dioxide
The chemical transformation of sulfur dioxide in the atmosphere has been studied ex-
tensively over the past 20 years. Recent reviews (Calvert et al., 1978; Middleton et a!.,
1980; and Holler, 1980), which consider analysis of laboratory and field data as well as
theoretical studies, indicated that S0? oxidation may proceed through both yas- and liquid-
phase ractions. The oxidation of SOg in the atmosphere is of considerable importance, in that
it represents a major pathway for particle production through the formation of sulfates.
Although the mechanism of SO- oxidation is not completely understood, it appears to proceed
via four pathways: homogeneous gas phase reactions; heterogeneous gas-solid interface re-
actions; and catalyzed and uncatalyzed liquid phase reactions. Homogeneous gas phase re-
actions are by far the most extensively studied and best understood quantitatively.
2-12
-------
TABLE 2-4. RELATIVE STRENGTHS OF ACIDS IN WATER SOLUTION (25°C)
Acid Dissociation
HI 2 H* ^
HBr * H* H
HCL ? H* H
H2S04 ? H+<
HN03 * H+<
H20-S02 2 H* H
HSO~ ? H* H
H3P04 * H+H
HF J H*H
HN02 ? H* ^
H2C03 '? H* H
H2S * H4" H
HSO" ? H+ H
HCN +• H* H
NH^ ^ H* ^
HCO; ? H+-
H20 ? H* -
- r
H Br"
^ CL~
h HS04
3
hHS03
H SflJ"
H H2p0;
H F"
h N02
3
i- HS"
H SO|"
i- CN"
- NH3
^ cof
i- OH"
pKft (= -Log KA)
-10
-9.5
-7
-3
-1.3
1.86
1.92
2.12
3.2
3.3
6.38
7.0
7.4
9.21
9.25
10.3
14.0
Reference
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Robinson & Stokes (1970)
Robinson & Stokes (1970)
Huss & Eckert (1977)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Salomaa et al. (1969)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
Cotton & Wilkerson (1980)
2-13
-------
The homogeneous gas-phase chemistry of oxidation in the clean and the polluted tropo-
sphere is reviewed in this section. The current status of knowledge is presented for the
elementary oxidation reactions of SO, and' the importance of volatile organic -and nitrogen
oxides as generators of free radical oxidizers. This review will show that the photochemical
oxidation of SO,, may be a significant pathway for tropospheric sulfate formation. The three
most important oxidizers of SOp are: hydr*oxyl radical, HO; peroxyl radical, H0?; and methyl-
peroxyl radical, CH^O,- At this time^ only the reaction rate constant for HO is well estab-
lished. The pathways of formation of the oxidizer radicals for the unpolluted troposphere can
be explained in terms of the photochemistry of the NO-CH.-CO-O., system. In polluted
atmospheres, volatile organics and oxides of -nitrogen act together to produce additional
radicals and accelerate overall radical production. There is also evidence that a dark re-
action among 0.,, alkenes, and SO, is effective in-oxidizing-SOp,
2.3.3.1 elementary Reactions—The elementary chemical reaction's of SOp in air'have been the
subject of intense investigation. Studies prior to 1965 have been critically reviewed by
Altshuller and Bufalini (1971), and more recently by Calvert et al. (1978). Calvert et al.
(1978) systematically examined the rate constants and significance of SOp elementary reactions
known to occur in the troposphere. Identified as generally unimportant reactions were:
photodissociation; photoexcitation; reaction with singlet delta oxygen, 09('A ); re'action with
3 ^9
atomic oxygen, 0( P); reaction with ozone, 03; reaction with nitrogen oxides (NO-, N03, NpO,-);
reaction with tert-butylperoxyl radical, (CH3)3C02; and reaction with acetylperoxyl radical,
CH-COO-. The only SQp reactions in the troposphere that were identified as important were
those due to HO, HOp, and CH^Op. The rate constants recommended by Calvert et al. (1978) for
these three reactions are given in Table 2-5. More recent work is in conflict with the rate
constants for HO- and CH3Qp that have been recommended by Calvert et al. (1978). Graham et
al. (1979) and Burrows et al. (1979) have reported rate constants for the HOp reaction that
are much lower than that recommended by Calvert et al. (1978). Also, Sander and Watson (1981)
have reported a rate constant for the CHgOp reaction that is much lower than that recommended
by Calvert et al. (1978). These more recent values are given in Table 2-5. The reasons for
the discrepancies in these two rate constants 'are unknown, and there is no basis to recommend
preferred values.
Although the dark reaction of SOp + 03 is too slow to be important in the troposphere,
the addition of alkenes greatly enhances the-oxidation rate. The experimental work of Cox and
Penkett U971a,b; 1972) and flcNelis et al. (1975) has been reviewed and reevaluated by Calvert
et al. (1978). The reaction system is too complex to discuss here, but Calvert et al. (1978)
report results of their calculations for total alkenes = 0.10 ppm, [0,] = 0.15 ppm, and [$02]
= 0.05 ppm; they estimated that the disappearance rate of S0? is 0.23 and 0.12 percent/h at 50
and 100 percent relative humidity (25°C), respectively. The reaction mechanism for the 0, •*•
2-14
-------
TABLE 2-5. RATE CONSTANTS FOR HYDROXYL, PEROXYL, AND METHOXYL RADICALS
Reaction
Second grde$ rate.. ,
constant, cm-mole s
Source
HO + S0£ •» HOSO£
H0£ + S02 •+ HO
S02 •* CH30 + S03
(1.1 ±0.3) x 10
-12
>(8.7 ± 1.3) x 10
v-18
-16
x 10
<2 x 10
-17
(5.3 ±2.5) x 10
5 x 10" 17
-15
Calvert et al. (1978)
Calvert et al. (1978)
Graham et al. (1979)
Burrows et al. (1979)
Calvert et al. (1978)
Sander and Watson
(1981)
Temperature = 25°C.
alkene + S02 system is not known, but studies by Niki et al. (1977) and Su et al. (1980)
indicate that the reactive species may be the biradical (RCHOO) formed by the decomposition of
the original molozonide,
In summary:
1. Three gas-phase tropospheric oxidation reactions have been identified as being
possibly important:
a. HO radical. The rate constant appears to be well established.
b. H02 radical. The rate constant is not well established.
c, CH302 radical. The rate constant is not well established.
2. SO, + 0~ + alkenes may be an important dark reaction.
2.3.3.2 Tropospheric Chemistry. of Sulfur Dioxide Oxidation— The chemistry of the clean tropo-
sphere and its mathematical simulation have been studied extensively by Levy (1971), Wofsy et
al. (1972), Crutzen (1974), Fishman and Crutzen (1977), Chameides and Walker (1973, 1976) and
Stewart et al. (1977).
The photochemistry of the unpolluted troposphere develops around a chain reaction
sequence involving NO, CH, , CO and 0,. The photochemical reaction chain sequence in the
1
troposphere is initiated by HO formed from the interaction of 0(0), the product of photolysis
of 0, in the short wavelength portion of the solar spectrum, with HJ3.
hv(A. < 310 nm) •* 0(AD)
0(1D) + H20 * 2HO
(2-5)
(2-6)
Z-15
-------
The HO produced reacts with CH» and CO present in the clean troposphere, resulting in the
generation of peroxyl radical species, H02, CH^Og.
HO + CH4 •» CH3 + H20 (2-7)
HO !+ CO •»• H + C02 _ (2-8)
CH3 + 02 + M -» CH302 + M (2-9)
H + 02 + M -> H02 + M (2-10)
In turn, the peroxyl radicals participate in a chain propagating sequence which converts
nitric oxide (NO) to nitrogen dioxide (NOp) and in the process produces additional hydroxyl
and peroxyl radical species.
CH302 + NO •» CH30 + N02 (2-11)
H0£ + NO -» HO + N02 (2-12),.
CH30 + 02 -» H02 + H2CO (2-13)
H£CO + hv(A < 370 nm) •* H + HCO (2-14)
HCO + 0£ •+ H02 + CO (2-15)
H202 + HO •+ H20 + H02 (2-16)
The major chain terminating steps include:
HO + N02 + M •»• HON02 + M (2-17)
H02 + H02 •»• H202 + 02 (2-18)
The reaction sequence for 03 production involves converting NO to N02 at a rate suffi-
ciently high to maintain a N02/N0 ratio to sustain the observed background levels of 0,.
H02 + NO •»• HO + N02 (2-12)
N02 + hv •»• NO + 0 (2-19)
0 + 02 + M-»03 + M (2-20)
NO + 03 •*• N02 + 02 (2-21)
HO + CO -» H + C02 (2-8)
In general, reactions (19) through (21) govern the 03 concentration levels present in the
sunlight irradiated well-mixed atmosphere at any instant and to a first approximation the
steady state relationship, Leighton (1961),
(N02) k,g
f J-J = (n \
(NO) k21 tu3j
2-16
-------
provides an accurate estimate of 0, given the ratio of (NO?)/(NO) and k,q/k?,. The photolytfc
rate constant k-,g is directly related to the integrated actinic solar flux over the wavelength
range 290 - 430 nm.
The paths for CU net destruction in the troposphere include the reaction sequence
H02 + 03 -» HO + 202 (2-22)
HO + 03 -» H02 + 02 (2-23)
Hydroxyl radical abundances predicted by the tropospheric photochemical models, 10 to
c «. o
10 molec cm , are in qualitative agreement with recent measurements by Davis et al. (1976),
Perner et al. (1976), and Campbell et al. (1979) and inferred HO levels based on measured
trace gas abundances in the troposphere by Singh (1977).
In the case of the chemistry of polluted atmospheres, extensive discussions on the
mechanism of photochemical smog and its computer simulation have been presented by Demerjian
et al. (1974), Calvert and McQuigg (1975), Niki et al. (1972), Hecht et al. (1974), and Carter
et al. (1979).
Perturbations of the photochemical oxidation cycle within the atmosphere introduced by
anthropogenic emissions are predominately caused by two classes of comppunds, volatile organic
compounds and nitrogen oxides. The reaction chain sequence discussed earlier for the clean
troposphere has now been immensely complicated by the addition of scores of volatile organic
compounds which participate in the chain propagating cycle. Figure 2-4 depicts a schematic of
the polluted atmospheric photooxidation cycle. The addition of volatile organic compounds
(VOC) in the atmosphere introduces a variety of new peroxyl radical species.
In its simplest form, the photochemical oxidation cycle in polluted atmospheres is
governed by the following basic features. Free radical attack on atmospheric VOC's is ini-
tiated by a select group of compounds, which are for the most part activated by sunlight.
Formaldehyde and nitrous acid, in particular, show high potential as free radical initiators
during the sunrise period. After initial free radical attack, the VOC's decompose through
paths resulting in the production of peroxyl radical species (HO,,, R0?, R'0», etc.) and par-
tially oxidized products, which in themselves may be photoactive radical-producing compounds.
The peroxyl radicals react with NO, converting it to N02, and in the process produce hydroxyl/
alkoxyl radical species (OH, RO, R'O, etc.). Alkoxyl radicals can be further oxidized, form-
ing additional peroxyl radicals and partially oxidized products, thereby completing the inner
loop reaction chain illustrated in Figure 2-4; or they may attack, as would be the major path
for hydroxyl radical, the VOC pool present in the polluted atmosphere, thereby completing the
outer loop reaction chain. The resultant effect in either case is the conversion of NO to NG*2
with a commensurate oxidation of reactive organic carbon, shortening of the hydrocarbon chain,
and production of organic particles, C02 and HgO.
The complex mixtures of organic compounds present in the polluted atmosphere react at
different rates depending upon their molecular structures, resulting in varying yields of free
radical species, ozone, N02, peroxyacetyl nitrate (PAN), and other partially oxidized organic
products as a function of VOC composition and VOC-NO levels.
A
2-17-
-------
FREE RADICAL INITIATORS
NO2 + hv
03 + hv
MONO + hv
RCHO + hv
PAN + hv/AT
03 +C-C
VOC = VOLATILE ORGANIC COMPOUNDS
PAN - PEROXYACETYL NITRATE
Figure 2-4. Schematic of the polluted atmospheric photooxidation cycle.
2-18
-------
Hydroxyl radical reactions seem to be the dominant mechanism by which CO, hydrocarbons,
N02 and SO, are consumed in the atmosphere (Niki et al., 1972; Demerjian et al., 1974; Calvert
et al., 1978). This highly reactive transient species, quite contrary to its organic free
radical counterparts, does not show appreciable change in concentration with atmospheric VOC
and NO variation, a result readily explainable upon review of the free radical production and
X
consumption sources. In the case of HO, ambient concentration conditions which enhance its
production tend also to consume the radical at an equivalent rate. The result is a faster
cycling in the VOC-NO oxidation chain (i.e., increased chain lengths) but very little pertur-
A
bation in the HO steady state concentration. In contrast, organic free radicals, mainly
peroxyl species, are consumed by alternate pathways which are less competitive and result in
increased steady state concentration.
Applying this basic knowledge of the photochemistry of the lower atmosphere, Calvert et
al, (1978) determined theoretical rates of SO, oxidation via attack of various free radical
species whose concentrations were estimated from computer simulations of the chemical reaction
mechanisms for clean and polluted atmospheres.
Based on limited rate constant data for the SO,,-free radical reactions, Calvert deter-
mined that HO dominated the rate of SO, oxidation in the clean troposphere, while in polluted
atmospheres the rate of SO- oxidation showed equivalent contributions from HO, HO,, and CH,0,
radicals. Figure 2-5 depicts the estimated time dependent rates of SO,, oxidation by free
radical species in a polluted air mass. Recent laboratory measurements suggest that the rate
of reaction of SO, with HO, and CHqO? may not be as great as estimated by Calvert et al.
(1978) (see discussion in Section 2.3.3.1). During July at mid-northern latitudes, typical
rates of S0? oxidation were of the order of 1.5 percent/h and 4.0 percent/h for clean and
polluted atmospheres, respectively. The major difference in rates was a result of higher
concentration levels of free radicals in the hydrocarbon rich polluted atmospheres. In a
2-
similar manner, Altshuller (1979) predicted the rates of homogeneous oxidation of S0? to SO-
in the clean troposphere using concentration predictions of the pertinent free radicals from a
two dimensional global model by Fishman and Crutzen (1978). A sample result from this study
showing the latitudinal and seasonal dependence of the rate of S0? oxidation is presented in
Figure 2-6, the variability in rate being predominantly due to availability of ultraviolet
solar intensity which drives the free-radical production process. The solar radiation depend-
ence of SO, conversion rate has also been observed in field measurements within power plant
plumes (Husar et al. , 1978), but should be viewed cautiously in light of the complicating
factors introduced by the dispersion and local chemistry of the primary source emissions.
The most important impact on SO, homogeneous gas phase reactions has come from recent
experimental determinations of the reaction rate constants of SO, with HO, by Graham et al.
(1979) and by Burrows et al. (1979) and S02 with CH302 by Sander and Watson (1981). As a
result of these recent determinations, H0? and CH,0, must be considered questionable as con-
tributing sources to the oxidation of S02 in the atmosphere. Therefore, in the theoretical
2-19
-------
30 60 80
IRRADIATION TIME, min
120
Figure 2-5. The theoretical rate of reaction (percent per hour) of
various free-radical species with SO2 is shown for a simulated sunlight-
irradiated (solar zenith angle of 40°) polluted atmosphere. The initial
concentrations (in ppm) were as follows: 862, 0.05; NO, 0.15; NC^,
0.05; CO, 10; CH4,1.5; CH20,0; CH3CHO, 0. The relative humidity
was 50 percent, and the temperature was 25° C.
Note: The rate constants for HO£ and CHgO2 radical reactions with
SO2 are not well established. See Table 2-4 and its discussion.
Source: Calvertetal. (1978).
2-20
-------
O JULY,HO,HO2,CH3O2
• JULY, MO
D JANUARY, HO, HO2,CH3O2
• JANUARY, HO
LATITUDE, °N
Figure 2-6. Percentage conversion at mid-day of sulfur dioxide to
sulfate by HO and by HO, HO2. and CH3O£ radicals as function of
°IM latitude in summer and winter.
Source: Altshuller (1979).
2-21
-------
estimates of S02 oxidation rates, by Calvert et al. (1978), and by Altshuller (1979), only the
hydroxyl radical portion of the contribution is now accepted as established, in view of these
recent experimental rate constant determinations. This results in maximum established SCL
oxidation rates of the order of 1.5 percent/h for both clean and polluted atmospheres during
July at mid-northern latitudes, a factor of 2.5 less than previous theoretical estimates for
polluted atmospheres. The revised rate is equivalent to a diurnally averaged rate of the
order 0.4 percent/h. Field measurements on the rates of SO* oxidation, discussed in Chapter
6j indicate that maximum S0? oxidation rates of the order of 10 percent/h are typical of many
atmospheric pollution scenarios. Our present knowledge of homogeneous SOp gas-phase reactions
does not sufficiently account for the rates observed. Smog chamber studies have demonstrated
that some species other than HO radical oxidizes S02 (Kuhlman et al., 1978; McNelis et al.,
1975). Alternate homogeneous gas reaction oxidation pathways are being studied (Su et al.,
1980), but certainly the role of heterogeneous and liquid phase SQp oxidation pathways should
not be overlooked in attempts to resolve this discrepancy.
In summary:
1. HO radical dominates the gas-phase oxidation of SOp in the clean troposphere. A
typical rate is on the order of 1.5 percent/h at noon during July at mid-northern
latitudes.
2. HO radical accounts for about 1.2 percent/h of the SO- oxidation in the polluted
troposphere. The combined contribution of H0« and CH^O, radical reactions may be as
great as about 2.8 percent/h, but their rate constants are not well established.
2.3.4 Solution-Phase Chemical Reactions
2-
The reactions of the aqueous 502^20^503-50.3 system is important to understanding the
processes of H,SO* formation in tropospheric particles, mists, fogs, and rain. This section
reviews the oxidation reaction of dissolved SQp species, including the auto-oxidation, metal-
ion catalyzed oxidation, carbon catalyzed oxidation, and reactions with the dissolved oxidants
N02, 03, and HgOg.
The state of knowledge of aqueous oxidation rates of dissolved SQp, HSO-, and SO, is
inadequate for simple systems and is extremely poor (or nonexistent) for complex systems that
include dissolved nitrogen and carbon compounds. Unfortunately, most of the studies are not
definitive because the investigators: (1) did not provide sufficient descriptions of experi-
mental procedure (especially the purification of the water and reagents), (2) did not select a
proper reactor design, and (3) worked at concentration levels that were orders of magnitude
greater than possible for ambient atmospheric aqueous systems. Trace quantities (at the
part-per-billion level) of catalytic metal ions are capable of enhancing the reaction veloci-
ties by orders of magnitude over the auto-oxidation rate, while similar trace quantities of
organics inhibit the rate. The characteristics of the chemical reactor govern the range of
the half-life that can be investigated and may influence the observed rate of oxidation.
2-22
-------
Two-phase air-water reactors (e.g., bubblers and supported droplets) may have reaction charac-
teristics that are dependent upon: (1) the mass transfer rate of the reactants through the
air-water interface, and (2) the mixing rates within the gas and water phases (Carberry, 1976;
Freiberg and Schwartz, 1981). Unless an adequate characterization of the two-phase reactor
was performed, it is not recommended that the implied elementary rate constant be accepted,
although in many cases the results may be correct. Supported droplets may present an addi-
tional problem: radical chains are efficiently terminated at liquid-solid interfaces, thereby
reducing the observed rate. Therefore, supported droplet measurements are not defensible
unless it is established that the oxidation is not a free-radical mechanism. Several notable
reviews of the oxidation of dissolved S0? and its hydration products in simple systems have
been published recently (Schroeter, 1963; Hegg and Hobbs, 1978).
This review will show that:
1. The auto-oxidation (uncatalyzed) reaction is very slow compared to the other re-
actions.
2. Mn(II) and Fe(III) are significant catalysts for the oxidation. The kinetic rate
expression is in doubt for the Mn(II) reaction, but several independent investi-
gators agree on a rate expression for Fe(III).
3. The catalytic effectiveness of these ions is unknown: Cu(II), V(¥), V(IV), Ni(II),
Zn(II), and Pb(II).
4. Elemental carbon (soot) with a water film may be an effective oxidation catalyst.
5. Dissolved HN02 and 03 oxidation rates are known and appear to be too low to be
effective.
6. The kinetics of the dissolved H202 oxidation of dissolved S02 species are known and
appear to be effective for forming SO. in particles, mists, fogs, and rain.
2.3.4.1 S(IV)-00 - H.,0 System—The simple S(IV) - 0» auto-oxidation (see glossary) has been
iL. "£,, " " ""'"
-------
TABLE 2-6. INVESTIGATIONS OF S02 - 02 AQUEOUS SYSTEMS
Investigators
Bigelow (1898)
Titoff (1903)
Lumiere and Seyewetz (1905)
Milbaur and Pazourek (1921)
Reinders and Vies (1925)
Haber and Wansbrough- Jones (1932)
VoVfkovick and Belopol'skii (1932)
Backstrom (1934)
Fuller and Crist (1941)
Riccoboni et al. (1949)
Abel (1951)
Winkelmann (1955)
van den Heuvel and Mason (1962)
Schroeter (1963)
Schwab and Strohmeyer (1965)
Rand and Gale (1967)
Scott and Hobbs (1967)
McKay (1971)
Miller and de Pena (1972)
Brimblecombe and Spedding (1974a)
Beilke et al. (1975)
Horike (1976)
Larson (1976)
Huss et al. (1978)
Larson et al. (1978)
Type of system
Bubbler
Bulk
Bulk
Bulk
2-phase bulk
Bulk
Bubbler
Theoretical
Bubbler
Bulk
Theoretical
Bulk
Supported droplet
Bubbler
Bulk
Bulk
Theoretical
Theoretical
Supported droplets
Bubbler
Bulk
Bubbler
Bubbler
Bulk
Bubbler
*
Comment
1,2,3
2,3
2,3
2,3
1,2,3
2,3
1,3
--
1
2,3
--
2
1,3
1
2
2,3
—
__
1
1
2
1,3
1
1,3
I
1. Incompletely characterized 2-phase system; results may be incorrect since
the investigators did not completely account for mass transfer rate or
demonstrate that no concentration gradients existed.
2. Purity of water is uncertain; results cannot be considered reliable.
3. Rate expression not reported.
2-24
-------
Termination
SO. + inhibitor -» (non-reactive species) (2-28)
radical + radical •* (non-reactive species). (2-29)
Brimblecombe and Spedding (1974b) propose an alternative scheme that does not include the
SCL radical -ion; in their scheme, equation (2-26) is replaced by:
S0~ + SOg" + SOj + SOg" (2-30)
•* 2 soj~ (2-31)
and equation (2-28) is absent,
Hegg and Hobbs (1978) have discussed most of the investigations identified in Table 2-6,
and they summarized the rate expressions, rate constants, and important features of the studies.
The observations can be classified into three types of rate expressions:
1. The type first reported by Fuller and Crist (1941),
2-
d[SO» ] n co ?_
- gf- = (ka + kb [H ]U'bU) [SOf ] (2-32)
2, The type first reported by Winkelmann (1955),
d[So!~] „
- gj- = kc [SO^ ] (2-33)
3. The type observed by Beilke et al. (1975).
o|~] + _0 ,6 2_
gf- = kd [H+] °'16[S02 ] (2-34)
d[So~]
The values that have been reported for k and k. for Type 1 reaction (equation 2-32) are
oE D
shown in Table 2-7. Schroeter (1963) has argued that his values obtained* for air (20 percent
Op) are in excellent agreement with those obtained by Fuller and Crist (1941) for pure Qy,
assuming a first-order dependence of the rate on dissolved [0?]. In fact, no study has demon-
strated an [Op]-dependence other than zero; hence, these two studies differ by a factor of ~5.
The values reported for k for Type 2 reaction (equation 2-33) are shown in Table 2-8.
These values, many of which were obtained by Beilke and Gravenhorst (1978) through a re-
analysis of the original investigators' reported data and expressions, are in good agreement.
For Type 3 reaction (equation 2-34), Beilke et al. (1975) observed a value of krf = 1.2 x
10"4 sec'V"16 (pH = 3-6, T = 25°C). It is presently unresolved as to which type of rate ex-
pression is correct. Doubt is cast on "Type 3" found by Beilke et al. (1975) because of the
2-25
-------
TABLE 2-7. VALUES3 OF ka AND kb FOR REACTION TYPE 1
Investigator
Fuller and Crist (1941)
Schroeter (1963)
K
McKay (1971)°
Larson et al. (1978)
k^sec"1
1.3 x 10"2
2.9 x 10"3
_•}
1.3 x 10 4
4.8 x 10"3
k^sec'V*
6.6
32
57
8.9
pH
5.1 - 7.8
7.0 - 8.2
5.1 - 7.8
4-12
T,°C
25
25
25
25
a Adapted from Hegg and Hobbs (1978)
McKay used a more recent value for K.n(=6,26 x 10 M) in expressing k. for Fuller
and Crist's (1941) data. w D
TABLE 2-8. VALUES3 OF k FOR REACTION TYPE 2
Investigator
Winkelmann (1955)
Schroeter (1963)b
Scott and Hobbs (1967)c
Hi Her and de Pena (1972)
Brimblecombe and.
Spedding (1974a)a
kc, sec'1
3. 5x10" 3
M6~0.6)xlO~3
1.6x10" 3
3xlO"3
(3.7-0.6)xlO~3
pH
7
7-8
•v.6.2 - 6.9
2-4
4-6
T,°C
25
25
25
25
25
3 Adapted in part from Beilke and Granvenhorst (1978) and Hegg and Hobbs (1978).
b Determined by Beilke and Gravenhorst (1978) from Schroeter1s (1963) data for
pH = 7-8.
c Determined by Scott and Hobbs (1967) from Van den Heuvel and Mason (1963) data.
Determined by Beilke and Gravenhorst (1978) by transforming the rate constants
reported for pH = 4-6 by Brimblecombe and Spedding (1974a).
2-26
-------
use of a plastic vessel that could have introduced trace organic inhibitors into the system.
All of the other studies (yielding "Types 1 and 2") were performed witn two-phase systems
whose mass transfer properties were insufficiently reported.
The auto-oxidation is inhibited by trace concentrations of organic species. The classes
of organic species capable of serving as inhibitors include alcohols, glycols, aldehydes,
ketones, phenols, amines, and acids. Backstrom (1934) first demonstrated that the inhibition
of sulfite oxidation can be expressed as:
] 2_
g— = {A/(B + m)} k35 [SO* ] (2-35)
where:
koc = the uninhibited rate constant
A,B = constants that are functions of the inhibitor
m = molar concentration of the inhibitor.
The influence of inhibitors on the rate has been extensively studied by Schroeter (1963), and
more recently by Altwicker (1979). According to Schroeter (1963), A and B are usually on the
order of 10 molar, which means that inhibitor concentrations greater than 10 molar are
effective. The form of the rate equation (Equation 2-35) suggests that the mechanism involves
a bimolecular reaction between an inhibitor molecule and a radical in the chain.
In summary:
1. The auto-oxidation reaction is very slow.
2. The rate is extremely sensitive to the presence of catalysts and inhibitors.
3. The rate is first order in sulfite.
4. No reaction mechanism has been satisfactorily demonstrated to account completely for
the observations of the more reliable studies (e.g., the dependence of the rate on
[H*]°-5 found by Fuller and Crist, 1941 and by Larson et al., 1978).
2.3.4.2 S(IV)- catalyst - 00 - H00 System—It is well established 'that some metal cations
4 2-
catalyze the oxidation of HSO, and SO, by molecular 0~. Of particular interest to the issue
of atmospheric sulfate formation in particles, mist, fog, and rain is possible catalytic
activity of Mn(II), Fe(III), Cu(II), Ni(II), V(IV), and V(V). General features of the cata-
lyzed reaction include: (a) inhibition by oxidizable organic molecules, (b) inhibition by
metal ion-complexing molecules (inorganic and organic), (c) exhibition of an induction time of
several seconds to several minutes, (d) detection of metal ion - S(IV) complexes, (e) no
dependence of rate on dissolved 0? concentration, and (f) dependence of the rate on the in-
verse of the initial H concentration (i.e., the rate is independent of pH change after the
reaction has been initiated). While the catalytic reaction mechanisms are unknown, they are
thought to be a modification of the initiation step of the auto-oxidation free radical mecha-
+ —9
nism (Equations 2-24 through 2-29); instead of M being a trace concentration (<10 M) of
2-27
-------
metal ion or a reactive wall, it i- a reagent present at concentrations >10 M. The rate
expressions for the various catalysts have different forms, suggesting different types of
initiation mechanisms (e.g., simple redox reactions or the formation of stable, reactive
complexes). The agreement among independent investigators is generally poor, indicating the
likelihood of mass transfer limitations of the rate or the presence of contaminants. A large
percentage of the investigations were conducted with two-phase reactors for which the mass
transfer characteristics were not adequately reported; therefore, those results, which may be
correct, are not considered .for estimating the elementary rate constant and for determining
the reaction order. Also, results of investigations using supported droplets may be biased
due to radical chain termination at the liquid-solid interface.
The Mn(II) catalyzed reaction kinetics have been investigated for over 75 years. The
9-
studies pertinent to the formation of SO- in the troposphere are presented in Table 2-9. One
of the first critics of Mn(II) catalysis studies was Titoff (1903), who remarked: "In
Bigelow's (1898) work the reaction occurred between two phases, and the retardation could be
determined by a change in the,boundary layer or by a decrease in the solution rate of oxygen."
Unfortunately, that comment applies to all but three of the Hn studies shown in Table 2-9,
which are: Hoather and Goodeve (1934), Neytzell-de Wilde and Taverner (1958), and Coughanowr
and Krause (1965). It is odd that none of these investigators presented rate expressions and
rate constants derived from their data. Instead, they left to the reader the task of extract-
ing that information. Estimates of their rate expressions are presented in Table 2-10. There
is agreement that the Mn(II) catalyzed rate is independent of dissolved 00, S00, HSO~, and
9*. *- *- *^
S03 concentrations.
Clearly, Hoather and Goodeve (1934) and Coughanowr and Krause (1965) are in agreement.
However, Neytzell-de Wilde and Taverner (1958) observed a first-order dependence on [Mn(II)].
There seems to be no basis to discount any of the three investigations, yet it appears that
serious errors may have been made. The results of Neytzell-de Wilde and Taverner (1958) are
slightly preferred because: (1) they measured the' rate"'df!"~disappearance of S(IV) by direct
chemical means, and (2) the period of observation (10-100 minutes) of the experimental runs
were sufficiently long that it is reasonable that the rate of oxidation was measured after the
establishment of the radial chains, and not during the induction period.
2-
The Fe(III) catalyzed reaction studies that are pertinent to the formation of SO^ in the
troposphere are identified in Table 2-11. The only studies not using two-phase systems (sub-
ject to mass transport limitations) are those of Neytzell-de Wilde and Taverner (1958),
Karraker (1963), Brimblecombe and Spedding (1974a), and Fuzzi (1978). Hegg and Hobbs (1978)
have pointed out that Karraker (1963) did not investigate the catalyzed oxidation in which
dissolved 02 is the oxident, but instead investigated the redox system associated with the
couple Fe(III) + e •* Fe(II) in an oxygen-free system. Thus, Karraker's work is not appli-
2-
cable. Neytzell-de Wilde and Taverner (1958) reported that the SO, formation rate was second
order for [S(IV)], but Karraker (1963) has reanalyzed their data and has shown instead that
2-28
-------
TABLE 2-9. INVESTIGATIONS OF S02 - MANGANESE - 0£ AQUEOUS SYSTEM
Investigators
Type of System
Comment
Titoff (1903)
Johnstons (1931)
Hoather and Goodeve (1934)
Bassett and Parker (1951)
Johnstons and Coughanowr (1958)
Neytzell-de Wilde and Taverner (1958)
Johnstone and Moll (1960)
Coughanowr and Krause (1965)
Bracewell and Gall (1967)
Matteson et al. (1969)
Cheng et al. (1971)
Bulk
Bubbler
Bulk
Bulk
Supported droplet
Bulk
Free droplets
Bulk and flow
Bubbler
Free and supported droplets
Supported droplets
2 •
1,2
2
2
1,2
2
2
2
1
3
1
1. Incompletely characterized 2-phase sys.tem; results may be incorrect since
the investigators did not completely account for mass transfer rate or
demonstrate that no concentration gradients existed. ,
2. Rate expression not reported.
3. Results are biased due to continued reaction of (supported) droplets on
filter of sampler; rate expression cannot be considered reliable.
2*29
-------
TABLE 2-10. RATE EXPRESSION FOR THE MANGANESE-CATALYZED OXIDATION
Expressiona'b'c
Investigators
~
= 44
n + n
[S(IV)]U [H f
3-4
Adapted from
Hoather and
Goodeve (1934)
= 1.7 x 10 J CMn(II)] [S(IV)]
°
dt
= 8
[S(IV)]
-2.2
-3-4
Adapted from
Neytzell-de Wilde
and Taverner
(1958)
Adapted from
Coughanowr and
Krause (1965);
dependence on
pH not reported
a.
b.
The units are: liter, mole, second.
-.0,
Concentrations shown with zero power (e.g., [S(IV)] ) indicate that the
investigators found the rate to be independent of those species. Note that
any concentration to the zero power is equal to unity.
c. The term [H+1 indicates that the rate is dependent only on the inverse of the
initial H ion concentration; changes in H
in progress do not affect the rate.
concentration after the reaction is
2-30
-------
TABLE 2-11. INVESTIGATIONS OF S02 - IRON - 02 AQUEOUS SYSTEM
Investigators
Reinders and Vies (1925)
Bassett and Parker (1951)
Higgins and Marshall (1957)
Johnstone and Coughanowr (1958)
Junge and Ryan (1958)
Neytzell-de Wilde and Taverner (1958)
Johnstone and Moll (1960)
Danilczuk and Swinarski (1961)
Karraker (1963)
Bracewell and Gall (1967)
Brimblecombe and Spedding (1974a)
Brimblecombe and Spedding (1974b)
Freiberg (1974)
Lunak and Veprek-Siska (1975)
Barrie and Georgii (1976)
Fuzzi (1978)
Type of System
Bulk
Bulk
Bulk
Support droplet
Bubbler
Bulk
Free droplets
Bulk
Bulk
Bubbler
Bubbler
Not reported
Theoretical
Flow
Supported droplet
Bulk
*
Comment
2
2
2
1
1,2
2
2
2
3
1
1
4
--
5
1
™* *"*
*
1. Incompletely characterized 2-phase system; results may be incorrect since
the investigators did not completely account for mass transfer rate or
demonstrate that no concentration gradients existed,
2. Rate expression not reported. '•••-. i« -i' i *.:..».>
3. 0?-free system; results not applicable to tropospheric SO,, oxidation.
4. Insufficient details reported to determine if the results should be
considered reliable.
5. Photochemical initiation.
2-31
-------
the order Is unity. NeytzeTl-de Wilde and Taverner (1958) did not present a-rate expression
and constant for the Fe(III) system. An estimate derived from their paper is presented in
Table 2-12. Brimblecombe and Spedding (1974a) reported a rate expression and constant
measured at a constant pH = 4; unfortunately, they used a plastic reaction vessel, which could
have released organic inhibitors into the system, causing the rate to be diminished. (At pH =
4, their rate is 0.25 of that of Neytzell-de Wilde and Taverner, 1958, and 0.1 of that of
Fuzzi, 1978.) Fuzzi (1978) did not note the similarity of his observations to those of
Neytzell-de Wilde and Taverner (1958), especially the dependence of the rate on the initial
inverse H concentration for pH < 4.0. Fuzzi's (1978) rate expression has been modified by
incorporating the dependence on [H ] and is presented in Table 2-12. Note that Fuzzi's
(1978) modified rate constant is 2.5 times greater than that of Neytzell-de Wilde and Taverner
(1958), which is good agreement for this type of measurement; these two studies appear to be
definitive for the Fe(III) system, and there is no basis to prefer one over the other. Fuzzi
(1978) has clearly demonstrated the change in the reaction order of [$(!¥)] from 1 to 2 as pH
increases from 4 to 5. The change in kinetics is due to the formation of colloidal Fe(OH),
for pH > 4, which provides an, explanation for the disagreement among earlier investigators.
•p •>•»•».',,* J"M*~ i£i ""'. ' ""* * ' ' '
Because of the formation of the Fe(OH)- colloid, it is unlikely that a meaningful Fe(III)
catalyzed rate expression for use in tropospheric sulfate formation can be stated for con-
ditions in which pH > 4.
TABLE 2-12. RATE EXPRESSION FOR THE IRON-CATALYZED OXIDATION
Expression9'
pH
Investigators
dt
M
= 0.04 [Fe(III )] [SUV)]
= 100 [Fe(III)] [S(lv)]
~
dt =0.1 [Fe(III)] [S(IV)]
dt
<4
Adapted from
Neytzell-de
Wilde and
Taverner (1958)
Adapted from
Brimblecombe and
Spedding (1974a)
Adapted from
Fuzzi (1978)
The units are: liter, mole, second.
K 4- — 1
The term [H ] indicates that the rate is dependent on the inverse of the
•!• +
initial H ion concentration; changes in H concentration after the reaction
is in progress do not affect the rate.
2-32
-------
The Cu catalyzed reaction kinetics have been described in the early work of Titoff
(1903). The pertinent investigations are identified in Table 2-13. As with the Mn and Fe
studies, most of the Cu studies were performed with incompletely characterized systems.
Fuller and Crist (1941) point out that the prior work is unreliable because of the likely
presence of contaminants. However, the investigations of Fuller and Crist (1941) were carried
out in a two-phase reactor whose mass transfer characteristics were not completely described;
no one since has conducted a more definitive study of this stystem. The reagent concentra-
tions used by Barron and O'Hern (1966) are orders of magnitude too large, and the pH range
(>8) used by Mishra and Srivastava (1976) is not applicable. For that reason, no rate ex-
pression can be recommended as reliable for use in calculating sulfate formation rates due to
Cu catalysis in the troposphere. Also, because of interference from the electric motors in
sampling devices, reported airborne Cu concentrations may be .unreliable (Patterson, 1980).
Vanadium catalysis has been reported in only one study (Bracewell and Gall, 1967); a
bubble reactor was used, and its mass transfer characteristics were inadequately reported.
Therefore, no rate expression can be recommended as reliable. However, Bracewell and Gall
(1967) did observe qualitatively " that "V~W was 'Srdefs crf'yagnituVfess1 effective 'than Mn and,
Fe; unfortunately, they did not study V(IV). It is likely'that V(V) catalysis is unimportant
for sulfate formation in the troposphere. "There are also no definitive studies for Cr(III),
Ni(II), Zn(II), and Pb(II), but from the qualitative work of Bracewell and Gall it appears
that these catalytic reactions are unimportant. , .
Barrie and Georgii (1976) have demonstrated qualitatively that Mn(II) and Fe(III) exhibit
a synergistic rate for the catalysis of S(IV) oxidation. Their rate expression cannot be con-
sidered reliable, since they use a supported droplet.
In summary:
1. S(IV) oxidation rates are significantly increased t?y Mn(II) and'Fe(III). There is
serious doubt regarding the rate expression for Mn(II), but the agreement among
independent studies is much better for Fe(III).
2, These systems are presently inadequately characterized: Cu(II), V(V), Ni(II),
Zn(II), Pb(II), and Cr(III). No studies of V(IV) have been reported.
3. There are no quantitative studies of metal ion-metal ion synergism.
4. The ability of atmospheric organic compounds to inhibit the catalysis is unknown.
5. All studies have been performed in the absence of HCO.,; however, the reactions
so + HcOg -» Hco3
HO + HCOg -* HC03 + OH , .
may be important. It is possible that such reactions may occur, and if, so, they
would prevent the oxidation radical chain from establishing, since HCOg is not a
powerful oxidizer (Hoigne and Bader, 1978).
In general, the rate expressions for catalytic oxidation to form HgSO^ are not well
established.
2-33
-------
TABLE 2-13. INVESTIGATIONS OF S02 - COPPER - 02 AQUEOUS SYSTEMS
Investigators
Titoff (1903)
Reinders and Vies (1925)
Alyea and Backstrom (1929)
Johnstone (1931)
Albu and von Schweinitz (1932)
Fuller and Crist (1941)
Riccoboni et al. (1949)
Bassett and Parker (1951) . >.,.- ._A«,,>.
Higgins and Marshall (1957)
Johnstone and Coughanowr (1958)
Junge and Ryan (1958)
Barren and O'Hern (1966)
Bracewell and Gall (1967)
Cheng et al. (1971)
Veprek-Siska and Lunak (1974)
Barrie and Georgii (1976)
Huss et al. (1978)
Mishra and Srivastava (1976)
Type of System
Bulk
Bulk
Bulk
Bubbler
Bulk
Bubbler
Bulk
., Bulk
Bulk
Supported droplet
Bubbler
Flow
Bubbler
Supported droplet
Flow
Supported droplet
Bulk
Flow
*
Comment
2
2
2
1
2
1
2
2
2
1
1
—
1
1
2
1
2
*"•*
1. Incompletely characterized 2-phase system; results may be incorrect since
the investigators did not completely account for mass transfer rate or
demonstrate that no concentration gradients existed.
2. Rate expression not reported.
2-34
-------
2.3.4.3 S(IV) - Carbon Black - 0, - H,0—The catalysis of the oxidation of dissolved S00 by
' """"** '" ' ' ' £. ——^— ^
carbon particles suspended in the water,has been studied by Chang et al. (1979) and by Eatough
et al. (1979). -It was found by Chang et al. (1979) that the oxidation rate of dissolved SOp
.species was:
- d[S(IV)] = k36[C] [02]°-69[S(IV)]° expC-E^/RT) (2-36)
with an activation energy of E = 11.7 kcal/mol over -the pH range of 1.45 to 7.5 for the
carbon studied, which was Nuchar-190. (The investigators demonstrated that Nuchar-190 behaved
• • 5
similarly to soot from acetylene and natural,gas flames.) An average value of k = 1.17 x 10
mol ' x liter ' /g-sec was reported. «T.he rate-limiting step has been suggested to be the
formation of an activated complex 'between-molecular oxygen and the carbon surface (Chang et
al., 1979;.Eatough et al., 1979).
Chang et al. (1979)-have estimated that for 10 ug of their fine carbon soot suspended in
0.05 g of liquid water and dispersed in 1 m of air, the atmospheric sulfate production would
be about 1 ug/hr. Heavy hydrocarbons are ads'orbed on the surfaces of atmospheric soots and
may inhibit the carbon-surface catalyzed oxidation of dissolved SQp, At this time,- it remains
to be demonstrated that the laboratory soots used by Chang et al. (1979) correspond to'those
3
present in the atmosphere or that the suspension of soot at ambient levels (<10 ug/m ) in
aerosols, cloud droplets, or rain is similar to the laboratory system,
2-3.4.4 S(IV) - Dissolved Oxidants - H,,0--Hydrogen peroxide, 0, and NOp may be important in
the oxidation of SOp in aqueous aerosols and fogs. These compounds are not highly reactive
with SO- in air, but their reactivity is enhanced in the liquid phase. Again, caution is
advised in accepting the results of studies of two-phase systems in which the investigators
have not completely accounted for the possibility of the mass transport limitation of the
oxidation rate. Therefore, only the recent results for single-phase systems are discussed
here.
Martin et al. (1981) used a stopped-flow reactor to investigate the kinetics of oxidation
of aqueous SO^ species by aqueous NO, NO™, and NO,. Over the pH range of 0.6 to 3.2, they
found*for NO and N03 that the disappearance of S(IV) species is:
- dCS(IV)] = k3?[NO or NQ~] [S(IV>] (2-37)
k3? = 0.01 mole I'V"1.
9-
Hpwever, for the same conditions, the reaction with NOp is rapid and the formation of SO. can
be expressed as:
d[S04~] 4-05
3? = k38CH J [HN02 + N02] [S02*H2° * HS03] (2"38)
k3g = 142 (liter/mole)1'V1
2-35
-------
The N09 is reduced quantitatively in this reaction to N90. Martin et al. (1981) also observed
9+
that this reaction is not catalyzed by Fe(III), Mn(II), or VO . It is unlikely that tropo-
spheric HN02 concentrations are high enough for this reaction to be important for HUSO, forma-
tion.
The oxidation of dissolved S02 by 03 has been investigated with stopped-flow systems.
Penkett (1972) and Penkett et al. (1979) have interpreted their work in terms of a decomposi-
tion of 0- to initiate a free-radical chain reaction involving OH, HSOo, and HSOr radicals,
after Backstrom (1934). Penkett et al . (1979) suggested that the rate expression is
+ ,
•• 1
where k-g = 71 sec . Erickson et al. (1977) reported the fractional contributions of the
oxidation of the three sulfur oxide species by ozone at various pH values; their rate ex-
pressions are:
= k40[SVH20:i [03]
dt = k41[HS03] [03] (2-41)
d[S02~] 2_
dt = k42tS03 1 t°3^ <2-42>
where k4Q = 590 liter/mol -sec, k41 = 3.1 x 10 liter/mol -sec, and k42 = 2.2 x 10 liter/
mol -sec.
Penkett et al. (1979) used a stopped-flow reactor to determine the kinetics of oxidation
of dissolved S02 species by H202- It was found that the rate of sulfate formation is given
by:
d[S02~] _ +
dt/ = k43[H202] [HS03] [H ] + k43a[H202] [HS03] [HA] (2-43)
where k43 = 2.6 x 10 liter /mol2-sec, with k43 and k43g being the third-order rate constants
for the catalysis by hydronium ions and proton-donating buffers (HA), respectively. At
pH £ 4, it is found that ^43^43 > 3200. Therefore, the second term is probably not
important for acid aerosols and fogs. It is of great significance that the reaction rate
increases as the solution becomes more acidic, which is in contrast to aqueous oxidation by
metal ions and by 03. The activation energy and the effect of ionic strength on the reaction
have been measured by Penkett et al. (1979). Dasgupta (1980) has criticized the presentation
of Penkett et al. (1979); use of the rate expression (Equation 2-43) takes into account
Dasgupta1 s (1980) points. Martin and Damschen (1981) have found that:
2-36
-------
[S0-H0]/{0.1 + [H ]} (2-44)
dt 4422 22
4 ~1
where k». = 7.2 x 10 s ; their expression is applicable over the range 0 dissolves in the water,
NH3(g) l NH3(aq) +
2. The dissolved NI-L/ ^ reacts with H , which raises the pH
NH3(aq) + H+ ^ <
Therefore, the ambient pathways of auto-oxidation, Mn(II)- and Fe(III)-catalyzed oxida-
tion, and 03 oxidation would have their rates enhanced by absorption of NH^. However, the
ambient pathways of H,,Q2 and HN02 would have their rates retarded by NH3 absorption. The rate
for soot would not be influenced.
NH3 can play other important roles. Reinders and Vies (1925) observed qualitatively that
Cu(II) was complexed by NH3 and rendered noncatalytic. At high pH's (>9) where NHg, ^ is the
dominant form, NhU may be oxidized by 03 and free-radicals (Hoigne and Bader, 1978).
In summary, the role of NH~ is explained in terms of its influence on the pH of the water
system; NHL is not a catalyst.
2-37
-------
2.3.5 Surface Chemical Reactions
Industrial emissions of solid particles (e.g., fly ash) and fugitive dust (e.g., wind-
blown soil and minerals) provide a solid-surface that may chemisorb SCL and yield sulfate
ions. This section will review investigations of the S02 oxidation on the surfaces of metal
oxides, fly ash, charcoal, and soot. Reaction kinetics have not been reported, but two
general types of processes have been recognized: a capacity-limited reaction for S02 removal
and a catalytic SCL oxidation process. The initial contact of SOp with the solid produces a
rapid loss of SO, from the gas phase; the reaction rate decreases with time. For the
capacity-limited reaction, the rate slowly approaches zero; for the catalytic precess, the
rate levels off for a time and then approaches zero. The latter phenomenon is attributed to a
pH decrease caused by hLSO, formation.
Urone et al. (1968) and Smith et al. (1969) found a number of solids to be effective in
removing SCL. In Urone's studies, S02 was admitted to a flask containing a powder that was
allowed to react with no mixing, and the product and remaining SO- were determined. Only the
average reaction rates can be calculated from these experiments; more importantly, with this
experimental procedure the rates may be diffusion-limited. The highest rate determined was
for SO, with Fe,SO,; the value was >75 percent per minute. Other materials found to be
slightly less reactive than Fe^SO, were Fe304, PbO, PbOp, CaO, Al-03. The rate for the ferric
oxide experiment was for 20 mg of Fe203 in a 2-liter flask; the Fe-0, concentration would thus
be 10 ug/m • Assuming a direct proportionality between rate and particle concentration, the
SO, removal rate in the atmosphere would be calculated to be 0.04 percent per hour for 100
3
ug/m of particles with the same reactivity as ferric oxide. However, since the mass transfer
characteristics of the reactor were not reported, these results cannot be considered reliable
for estimating rates.
Smith et al. (1969) did not focus on sulfate formation kinetics; instead, they
illustrated through a novel experiment the ability of solid particles to adsorb S0? and to
release SO, during passage through a tube with a wall that adsorbs SO,. They measured the
number of SQy monolayers absorbed on suspended Fe,0, as a function of SO, partial pressure.
The monolayer coverage data reported in their Table I are irr error by a factor of 100 too
•y
large; e.g., the number of monolayers at 1.13 ppm should be 0.38 x 10 .
Chun and Quon (1973) measured the reactivity of Fe203 to S02, using a flow system in-
volving a filter containing suspended particles. They determined a removal rate constant of
"~^ —i —i
9.4 x 10 ppm min (-din 8/dt), where 0 is the fraction of surface sites available for
reaction. Extrapolating this to an atmospheric particle concentration of 100 yg/m with an
o
equivalent reactivity and an SO,, concentration of 0.1 ppm (260 yg/m ), the data project an
atmospheric removal rate of 0.1 percent per hour.
Stevens et al. (1978) report total iron concentrations in six U.S. cities ranging between
0.5 and 1.3 yg/m . Other species such as manganese, copper, or vanadium had total concentra-
o
tions usually below 0.1 jjg/m . Thus actual ambient air concentrations are approximately
2-38
-------
l/50th those assumed by the authors in the above papers. A reactive particle concentration of
Q
2 ug/m would yield a predicted SO, removal rate of no more than 0.002 percent per hour.
Therefore, surface reactions are probably not important except in sources prior to or immedi-
ately after emission.
The most comprehensive studies to date on S0? removal by pure solids were made by Siege!
et al. (1974) and Judeikis et al. (1978). A tubular flow reactor, in which solids were
supported on an axial cylinder, was used to measure reactivities of MgO, F^o^S' ^2^3' ^n^?'
PbO, NaCl, charcoal, and fly ash. They found that the rates of SO, removal diminished with
exposure until the solids completely lost ability to react with SO,. The relative humidity
was important in determining the total capacity for S0? removal, but not the initial rate of
uptake; total capacity increased as relative humidity increased. The capacity for SO, could
be extended by exposure to NH,. This type of behavior is consistent with the formation of
H?SQ. on the surfaces.
Because carbonaceous matter is so common in ambient air paniculate samples, various
studies have been made of the S0? removal rate by carbon. A comparison of the results is
difficult because of the varieties of carbon available for study, such as activated charcoal,
graphite, acetylene flame products, and combustion products of diesel oil and heating oil. In
regard to investigations that deal with the gas-solid reaction of S0~ with carbon, Novakov et
al. (1974) performed laboratory experiments that showed that graphite and soot particles
2-
oxidize SOg in air. The soot exposed to humidified air produced more SO. than that exposed
only to dry air. For downtown Los Angeles, they observed a strong correlation between con-
2~
centrations of ambient carbon and SO. , which supports their hypothesis that carbon (soot)
2-
oxidation of SO, is the major pathway for SOI formation (see Section 2.3.4.3).
Tartarelli et al. (1978) studied the interaction of SO, with carbonaceous particles
collected from the flue ducts of oil-burning power stations. They concluded that the amount
of adsorption is increased by the presence of oxygen and water in the gas stream. Reaction
rates were not determined in this study.
Liberti et al. (1978) studied the absorption and oxidation of S0£ on various particles,
including soot from an oil furnace and various atmospheric particulate samples. They con-
cluded that the main interaction between the S0? and PH is adsorption, with most catalytic
reactions occurring at high temperatures, near the combustion source. Their experiments with
atmospheric particulate samples led them to the conclusion that any heterogeneous nonphoto-
chemical sulfate formation is strongly dependent on the reactivity of the particle surface,
and hence the history (aged, freshly emitted), of the aerosol.
In summary:
1, Surface reactions are capacity-limited. Those that involve catalysis in liquid
films can be extended by the absorption of NH~.
2. The initial rates may be large, but quickly approach zero.
2-39
-------
3. Except for the carbon (soot) reaction, solid surface reactions do not seem to be
effective pathways for hLSO. formation in the troposphere.
2.3,6 Estimates of SQgOxidation
At this point it is interesting to compare the rates of SO, oxidation by the more impor-.
tant reactions identified in the previous sections of this chapter. The important reactions
for gas-phase and aqueous-phase oxidation are listed in Table 2-14, and rates of SO, oxidation
for an assumed set of conditions are present. These calculations ignore the nonhomogeneous
nature of the troposphere and assume that all of the reactants are well mixed. (The more
general case is treated in Chapter 6.)
TABLE 2-14. ESTIMATES OF S02 OXIDATION RATES IN WELL-MIXED TROPOSPHERE
I.
II.
Reaction
Gas Phase
HO radical
H02 radical
CHgOa radical
Aqueous Phase, pH:
Mn(II) catalysis
Fe(III) catalysis
C (soot) catalysis
03 (40 ppb)
03 (120 ppb)
H202 (1 ppb)
H202 (10 ppb)
_-i
Rate, % h """
0.3 - 1.3
0.4 - 2.0
0.3 - 1.5
135
1E-1 1E+1 1E+3
5E-5 5E-1 5E+3
3E+1 3E+1 3E+1
2E-8 2E-6 2E-4
6E-8 6E-6 6E-4
2E-2 3E-2 3E-2
2E-1 2E-1 3E-1
Discussion
Section
2.3.3.2
2.3.3.2
2.3.3.2
2.3.4.2
2.3.4.2
2.3.4.3
2.3.4.4
2.3.4.4
2.3.4.4
2.3.4.4
*
Comments
a
a,b
a,b
b.c.d
c.e.i
f,i
c,g
c,g
c,h
c,h
,i
NOTE:
*
a.
b.
c.
"E" denotes "exponent to
Typical range for daytime
This reaction rate is not
Assumed that liquid water
the base 10" (e.g., 3E-1
at northern midlatitudes
= 3 X 10"1)
during the summer.
well established; see discussion section.
volume of aerosol = 50 x
10"12m3/m3, [50,1
= 10
ppb
(or 26 ug/m3).
d. 'Assumed that Mn(II) mass concentration = 20 ng/m3; also, the Mn(II) is assumed
to be uniformly dissolved in the liquid water of the aerosol (CMn(II)] = 8.9 x
10 M). Rate calculation used the expression of Neytzell-de Wilde and Taverner
(1958); see Table 2-7.
e. Assumed that Fe(III) mass concentration = 2 M9/m3"» also, the Fe(III) is assumed
to be uniformly dissolved in the liquid water of the aerosol ([Fe(III)] =0.9 M).
Rate calculation used the expression of Neytzell-de Wilde and Taverner (1958);
see Table 2-9.
f. Assumed that C mass concentration = 10 |jg/m3 and behaves as the soots studied by
Chang et al. (1979), whose expression was used for this calculation (Equation 2-36).
g. Rate calculation was based on Equation 2-39.
h. Rate calculation was based on Equation 2-43.
i. Influence of inhibitors has been ignored, but they are likely to suppress the
rate by orders of magnitude.
2-40
-------
2
For this comparison, it has been assumed that the S0? concentration is 10 ppb (26 [jg/m )
-12 3 3
for all of the reactions and the liquid water content of the aerosol is 50 x 10 m /m .
The gas-phase rates were calculated from the material in Section 2.3.3.2, while the
aqueous-phase rates were derived from Sections 2.3.4.2-4. Unbased assumptions include:
3
1, The ambient mass concentration of 20 ng/m for Mn is reasonable, but: (a) it is not
known if the predominant form is Mn(II), and (b) it is unlikely that Mn is uniformly
distributed and dissolved. Inhibitors have been ignored, but they likely suppress
the rate by orders of magnitude.
2. Likewise, the ambient concentration of 2 (ag/m for Fe is reasonable, but: (a) it is
not known if Fe(III) is the predominant form, and (b) it is unlikely that Fe is
uniformly distributed and dissolved. Inhibitors have been ignored, but they likely
suppress the rate by orders of magnitude.
3. There is no basis to assume that the rate equation observed for laboratory-generated
carbon (soot) applies to atmospheric carbon. Inhibitors may be important.
4. The rates for the H0? and CI-LOp reactions recommended by Calvert et al. (1978) are
not well established.
It is very likely that the inhibitor-free rates estimated for Mn(II) catalysis, Fe(III)
catalysis, and C (soot) catalysis are gross overestimates. Also, the HO, and CH.,0? rates may
be too high.
Uncritically accepting all of the rates, at a pH = 3, and [H^O^] = 10 ppb, the SO^ con-
version rate would exceed 40 percent/h. However, if only the well-established rates are
considered, the S02 conversion rate becomes vL.l percent/h.
In summary:
1. The gas-phase reaction rate of HO and the aqueous-phase reaction of H?0p are well
established, but are expected to account for only about 1.1 percent/h (under the
conditions given in Table 2-14).
2. The Mn(II), Fe(III), and C (soot) catalyzed reactions have sufficient rates to
dominate S0? oxidation in the troposphere, but the assumptions discussed above may
not be reasonable.
2.4 PHYSICS AND CHEMISTRY OF PARTICIPATE MATTER
Knowledge of the physics and chemistry of particulate matter is necessary for design of
satisfactory samplers and monitors, understanding the relationships between sources and
effects, and understanding important processes in the troposphere that involve chemical trans-
formations and removal.
In Section 2.2, the global cycle and annual budget for sulfur were presented to aid in
establishing the goals" and limitations of this document's treatment of sulfur oxides. That
discussion is incomplete, in that particulate matter is related to cycles of numerous elements
and their interactions. Among the most important cycles of elements are: sulfur, nitrogen,
carbon, hydrogen, boron, oxygen, sodium, aluminum, silicon, phosphorous, chlorine, potassium,
2-41
-------
calcium, vanadium, manganese, iron, mercury, and lead. Since it is beyond the scope of this
document to deal with the details of these cycles, a perspective can be obtained form a budget
estimate of the particulate mass injected into the troposphere. The estimate by Hidy and
Brock (1971) of the daily particulate mass emitted or formed in the troposphere is presented
in Table 2-15. Globally, the anthropogenic contribution is about 6 percent; however, the
nonhomogeneous distribution and the type of emissions pose serious problems. (See Chapter 4
for a discussion of sources in the United States.) Figure 2-2 showed the general interrela-
tions of pathways, processes, and properties of sulfur oxides and particulate matter and
effects. Section 2-3 treated the SO, physical properties and chemistry (including trans-
formation chemistry), which are indicated in Figure 2-2. Section 2.4 will discuss the physics
and chemistry of particulate matter that are related to particle properties, single-particle
dynamics of motion, formation and growth, and aerosol system dynamics.
The budget in Table 2-15 indicates that secondary particulate matter formation dominates
the rates. This important source will be discussed in Section 2.4.
2.4.1 Definitions
Aerosol science spans chemistry, physics, engineering, meteorology, and the biological
sciences. Unfortunately, the lack of communication among workers in these diverse disciplines
has impeded the unification of their ideas. One of the results has been a lack of universally
accepted definitions of the terms "aerosol" and "particle" and the terms for classification of
aerosol systems. The definitions used here are consistent with general usage by atmospheric
scientists.
Particle: Any object having definite physical boundaries in all directions, without any limit
with respect to size (Cadle, 1975). In practice, the particle size range of interest is
used to define "particle." In atmospheric sciences, "particle" usually means a solid or
liquid subdivision of matter that has dimensions greater than molecular radii (~10 nm);
there is also not a firm upper limit, but in practice it rarely exceeds 1 mm.
Aerosol: A disperse system with a gas-phase medium and a solid or liquid dispersed phase
(Fuchs, 1964). Often, however, individual workers modify the definition of "aerosol" by
arbitrarily requiring limits on individual particle motion or surface-to-volume ratio
(e.g., see Hidy and Brock, 1970). Aerosols are formed by (a) the suspension of particles
due to grinding or atomization, or (b) condensation of supersaturated vapors (Fuchs,
1964).
An aerosol is not the halocarbon vapor used as the propellant in pressured cans (commonly
referred to as "aerosol cans" and "aerosol bombs"). Improper use of the term "aerosol" by
marketers of foams, gels, sprays, etc., has caused the lay public to associate incorrectly
environmental issues of suspended particulate matter with the issue of halocarbon impact Dn the
stratospheric ozone layer. In the context of this document, the term "aerosol" is not related
to the impact of halocarbons on the stratospheric ozone layer.
2-42
-------
TABLE 2-15. ESTIMATE OF GLOBAL TROPOSPHERIC PARTICULATE MATTER PRODUCTION RATES3
Source • % by Weight of Totalb
A. Natural Sources
1. Primary
Windblown dust 9.3
Sea Spray 28
Volcanoes 0.09
Forest Fires 3.8
2. Secondary
Vegetation 28
Sulfur Cycle 9.3
Nitrogen Cycle . 14.8
Volcanoes (gases) 0.009
SUBTOTALS:
a Source: Hidy and Brock (1971)
Production rate = 10.7 x 10 metric tons/day
Not 100% because of round-off errors.
SUBTOTALS: 93
B. Manmade Sources
1. Primary
Combustion and Industrial 2.8
Dust from Cultivation 0.009
2. Secondary
Hydrocarbon Vapors 0.065
Sulfates 2.8
Nitrates 0.56
Ammonia 0.028
TOTAL: 99C
2-43
-------
Traditionally, workers in various scientific fields have classified the aerosol systems
to reflect their origin, physical state, and range of particle size. The meanings of these
classifications are not universally accepted; however, the following- definitions 'are con-
sistent with general usage by atmospheric scientists.
Particulate Mass; A generic classification in which no distinction is made on the basis of
origin, physical state, and range of particle size (Dennis, 1976). (The term "par-
ticulate" is an adjective, but often it is incorrectly used as a noun.)
Dust: Dispersion aerosols with solid particles formed by comminution or disintegration,
without regard to particle size (Fuchs, 1964; Dennis, 1976; Hidy and Brock, 1970).
Typical examples include (a) natural minerals suspended by the action of wind, and (b)
solid particles suspended during industrial grinding, crushing, or blasting.
Smokes: Dispersion aerosols containing both liquid and solid particles formed by condensa-
tion from supersaturated vapors (Fuchs, 1964; Hidy and Brock, 1970). Generally, the
particle size is in the range of 0.1 urn to 10 pm. A typical example is the formation of
particles due to incomplete combusion of fuels.
Fumes: Condensation aerosols containing liquid or solid particles formed by condensation of
vapors produced by chemical reaction of gases or sublimation (Dennis, 1976'). Generally,
the particle size is in the range of 0.01 ym to 1 urn. Distinction between the terms
"smokes" and "fumes" is often difficult to apply.
Mists; Suspension of liquid droplets formed by condensation of vapor or atomization; the
droplet diameters exceed 10 iim and in general the pariculate concentration is not high
enough to obscure visibility (Hidy and Brock, 1970).
Fogs: Same as "mists", but the particulate concentration is sufficiently high to 'obscure
visibility (Hidy and Brock, 1970). [Dennis (1976) proposes alternate definitions that
distinguish "mists" and "fogs" on the basis of particle size.]
Haze: An aerosol that impedes vision (Dennis, 1976) and may consist of a combination of water
droplets, pollutants, and dust (Hidy and Brock, 1970). ,
Smog: A combination of "smoke" and "fog." Originally, this term referred to episodes in
Great Britain that were attributed to coal burning during persistent foggy conditions
(Chambers, 1976). In the United States "smog" has become associated with urban aerosol
formation during periods of high oxidant concentrations.
Cloud: A free aerodisperse system of any type having a definite size and form and without
regard to particle size (Fuchs, 1964).
Primary particles (or primary aerosols): Dispersion aerosols formed from particles that are
emitted directly into the air and that do not change form in the atmosphere (MAS, 1977).
Examples include windblown dust and ocean salt spray.
Secondary particles (or secondary aerosols): Dispersion aerosols that form in the atmosphere
as a result of chemical reactions, often involving gases (HAS, 1977). A typical example
2-
is SO. produced by photochemical oxidation of SO,,.
3-44
-------
In addition to classifying aerosol systems by their properties (origin, physical state,
size), systems are classified according to the performance characteristics of the sampler or
analyzer. Some of the more common classifications used are the following:
Aitken nuclei: Those particles and ions measured by means of an instrument in which water-
vapor is made to condense on particles by supersaturating the vapor (White and Kassner,
1971). In order to eliminate condensation on light ions, the supersaturation should not
exceed 270 percent (White and Kassner, 1971). The term "condensation nuclei" is often
used synonymously with the term "Aitken nuclei."
Total suspended particulate (TSP) mass: The particulate mass that is collected by the high-
volume sampler. (The system is classified in terms of the operational characteristics of
the sampler. See the discussion in Chapter 3.)
Coarse and fine particles: These two fractions are usually defined in terms of the separation
diameter of a sampler. For the dichotomous sampler (see Chapter 3), the separation
diameter has usually been set at 2.5 urn. Thus, for the dichotomous sampler, the "coarse
particles" are those collected by the sampler with aerodynamic diameters greater than 2.5
pm; the "fine particles" are those collected by the sampler with diameters less than 2.5
[jffl. (NOTE: separation diameters other than 2.5 urn have been used.)
Additional definitions that relate to particle size, particle size distributions, and
particle motion will be provided in the context of the material discussed .in the following
sections.
2.4.2 Physical Properties of Gases and Particles
To understand the behavior of an aerosol, it is necessary to know the physical properties
of the gases and particles. Such knowledge is necessary to designing particle samplers,
understanding the effects of aerosols (e.g., loss of visual quality), understanding aerosol
processes (such as coagulation, growth, deposition), and modeling the effects and dynamics of
aerosols.
2.4.2.1 Physical Properties of Gases—An aerosol consists of two principal components: the
gas-phase medium and the solid or liquid dispersed phase. The behavior of aerosol systems can
be described in terms of the behavior and interaciton of these two components.
For tropospheric aerosols, the gas of interest is "air." The molecular and fluid pro-
perties of air are well established and will not be reviewed here (see Hirschfelder et al.,
1954; Bird et al., 1960). The fluid motion of air,' especially laminar flow, is adequately
understood. Presently, turbulent flow is formulated in statistical descriptions, and often
the flow fields in complex geometry cannot be satisfactorily predicted. This limitation in
theory has seriously affected our ability to describe the tropospheric microscale motion of
particles with diameters greater than 10 (jm; specific problems include the performance of
particle samplers and the formulation of particle dry deposition.
2-45
-------
In summary:
1. The physical properties are adequately known.
2. Laminar flow of air is adequately understood.
3. Turbulent flow must be described in terms of a random fluctuating component; this
limitation seriously affects our ability to'describe particle motion, especially for
particles with diameters greater than 10 pm.
2.4.2.2 Physical Properties of Particles—The physical properties of particles that influence
behavior are divided into three types (Billings and Gussman, 1976): physical configuration,
bulk material properties, and surface properties.
2.4.2.2.1 Physical configuration. The shape, structure, and density are physical configura-
tion properties that are very important parameters in the equations of motion for particles.
The shape of particles is highly varied. Tropospheric particles have been reported to
have the following types of shapes: spherical, irregular, cubical, flake, fibrous, and conden-
sation floes. Particle shape is related to source type, as shown in Table 2-16.
TABLE 2-16. PARTICLE SHAPES AND SOURCE TYPES3
Shape Examples
Spherical Smoke, pollen, fly ash
Irregular Cinder
Cubical Mineral
Flakes Mineral, epidermis
Fibrous Lint, plant fiber
Condensation floes Carbon, smoke, fume
a Whitby et a!., 1957.
The physical dimensions of particles are usually expressed in terms of an equivalent
statistical diameter. For such a measure to be meaningful for nonspherical particles, it must
be applied as an average to a statistically significant number of particles (Cadle, 1975).
For sizing collected particles, the most widely used "diameters" for irregular particles are:
1, Martin's diameter: The distance between opposite sides of the particle, measured
crosswise of the particle, on a line that bisects the projected area and that is
parallel to a reference line. (For examples, see Cadle, 1975; McCrone and Delly,
1973.)
2. Feret's diameter: The distance between two tangents on opposite sides of the par-
ticle and parallel to a reference line. (For examples, see Cadle, 1975; McCrone and
Delly, 1973.)
3. The maximum horizontal intercept: The longest diameter from edge to edge of the
particle, parallel to the reference line (McCrone and Delly, 1973).
2-46
-------
4, The projected area diameter (British Standard method); Found by comparing the
projected area of the particle with the areas of reference circles on an ocular
graticule (McCrone and Delly, 1973).
These four "diameters" are used in powder technology. However, they are less useful for
applications relating to particle dynamics. They describe the shape in terms of one or two
dimensions (i.e., projected surface area). The dynamics of particle motion are formulated in
terms of the diameters of three-dimensional spheres. The relation between the "diameters"
measured for projected surface areas of irregular particles and the "diameter" meaningful for
particle motion and light scattering is not obvious. For regular-shaped particles (e.g.,
cubes, cylinders, oblates), Fuchs (1964) has derived dynamic shape factors that permit their
representation as equivalent spheres.
The density (mass per unit volume) of particles is important because it affects motion
and behavior. The density of particles that are spheres, cubes, and other regular geometries
is the same as the density of the bulk material. However, many particles are agglomerates of
smaller particles of various composition. A large percentage of the volume of such agglome-
rates is voids or air-filled pores. Such a structure has the appearance of a cluster of,
grapes. The sum of the volume of the small solid/liquid particles plus the void volume is)
defined as the "apparent" volume of the agglomerate. The "apparent density" of agglomerates
is defined as the ratio of solid/liquid mass to the apparent volume (Fuchs, 1964), and it i«
often 2-10 times lower than the density of mass that excludes the pore volume (Hesketh, 1977)J
Because many tropospheric particles are irregular or agglomerates and have unknown dem
sity, it is common practice to represent the shape, structure, and density of particles in
terms of dynamically equivalent spheres of unit density. Hence, the following definition!;
Aerodynamic diameter: The diameter of a sphere of unit density (1 g/cm ) that attains the
same terminal velocity at low Reynolds number in still air as the actual particle under con-
sideration.
2.4.2.2.2 Bulk material properties. The bulk material properties that affect aerosol be-
havior include chemical composition, vapor pressure, hygroscopicity and deliquescence, and
index of refraction. These properties are of interest because they control (a) the physical
state and growth, and (b) the scattering and absorption of light by tropospheric particles.
The chemical composition of tropospheric particles will be discussed briefly in Section
2.4.5 and in more detail in Chapter 5. It is sufficient to point out here that the particles
2- + +
with diameters less than approximately 2.5 urn contain most of the SO. , H , and NH. and a
significant fraction of the NCL and Cl ; therefore, these particles interact with H^O vapor
much more strongly than larger particles (Meszaros, 1971; Char!son et al., 1978).
The most important systems are those of HgSO^, NH-HSO^, and (NH^SO^. The most im-
portant aqueous systems are those containing H+s NH*, So| , N03, and Cl since these species
are usually present in sufficient mass to control the liquid water concentration and the phase
transition points of particles as a function of relative humidity. Presently, phase diagrams
2-47
-------
for this multi component system are not available for conditions relevant for tropospheric
+ + 2-
particles. However, the phase diagram for the H -NI-L-SO. -HJ3 system at equilibrium is shown
in Figure 2-7. On this diagram, the dry, pure crystals (NH,)2S04, (NH.)3(HSO.)2, and
(NH.)HSO. are indicated as points A, 8, and C, respectively. If letovicite, (NH4)3H(S04)2, is
exposed to relative humidities that start at 0 and increase to 100 percent, its behavior can
be described in terms of the locus BO in Figure 2-7. From 0 percent r.h. , the salt immedi-
ately enters the 3-phase zone consisting of (NH.KHCSQ.)-, (NH^pSO., and some liquid solution
of H+, NH4, and SO. . At point D, the locus intersects a phase boundary for (NH4)3H(S04>2 and
a partial deliquescence occurs. Between the point D and the intersection with curve EE,,
solid (NH.^SO. remains; however, at the intersection of EE,, a second and complete de-
liquescence occurs. From EE-, to point 0, only the solution phase is present. In similar
fashion, equilibrium trajectories for salts subjected to varying composition and relative
humidity can be traced. The locus EE2 demonstrates the dependence of the complete de-
liquescence point r.h. on the weight percent H»,SO • as the system's acid composition changes
from 0 to 35 percent, the complete deliquescence point r.h. changes from 80 to 39 percent.
Thus, it is obvious that NI-U plays a key role in governing the phase transition points.
The H2S04, NH4HS04, (NH4)3H(,S04)2> and (NH4>2S04 particle systems have been characterized
recently by Char! son et al. (1978), Tang et al. (1976), Tang and Munkelwitz (1977), and Tang
(1980a). Charlson et al. (1978)» used an apparatus in their studies that measured the light
scattering coefficient of the aerosol as a function of the relative humidity. They obtained
good agreement between theory and experiment in observing the hygroscopic behavior of H?S04.
However, they observed no deliquescence point for NH4HS04 particles, and one at ^38 percent
r.h. for (NH4).jH(S04)p particles. The bulk salts have a deliquescence point at 39 percent and
69 percent r.h., respectively. They suggested that the deliquescence point for NH4HS04 par-
ticles may not have been exhibited because the initial droplets which were only dried to 15
percent r.h. entered a hysteresis loop, forming supersaturated solution droplets. Their
observation of a growth point for (NH4)3H(S04)2 particles at ^38 percent r.h. but not the
deliquescence point at 68 percent r.h. is difficult to explain, but it may have been due also
to the particles forming supersaturated droplets and entering a hysteresis loop. If that were
the case, then the true deliquescence point at 69 percent r.h. would not be exhibited, and the
observed transition at ~38 percent may be an efflorescence point. Efflorescence points of
solution droplets exposed to decreasing relative humidity are not sharp and usually occur at
relative humidities more than 30 percent below the deliquescence point. Tang (1980a) used a
system in which the salt aerosol was first dried, and then passed through a controlled humi-
dity chamber; the particles were sized with a single-particle optical analyzer. His data on
deliquescence points and hygroscopic growth agreed well with theory, as is shown in Figure
2-8, and he concluded that for the NH4HS04-H2Q systems that the equilibrium size of mixed-salt
droplets may be adequately predicted from bulk solution properties. (The Kelvin effect, which
2-48
-------
100
8
M
*"?r
z
o
Hi
30
20
10
30 40
WEIGHT % H2SO4
70
Figure 2-7. Solubility diagram for the H+—NH4*-SO42-—H2O system at
equilibrium (30°C).
A = solid phase of (
B = solid phase of (
C = solid phase of (NH4)HSO4
= liquid solution phase
a^ = activity of water
= fractional relative humidity
y = mole fraction of (IMH^gSC^
The numbers in parentheses are the fractional relative humidities for the complete
deliquescence points that are indicated.
Source: Tang (1980a).
2-49
-------
PO
2.4
2.2
2.0
UJ
a 1.8
u
N
1.6
o 1.4
E
1.2
1.0
50
COMPOSITION, wt%
(NH4)2S04 H2S04
• 100
A 88.3
O 83.3
0
60 70 80
RELATIVE HUMIDITY, %
90
Figure 2-8. Growth of H*—NH4*—SQ42~ particles as a function of a relative
humidity. The solid curves represent theory and the points are experimental
observations.
Source; Tang(1980a).
-------
may limit growth, is discussed in Section 2.4.4.) Tang (1980a) also demonstrated the hysteresis
phenomenon for (NH.^SCL particles; it can be seen in Figure 2-9 that as the particles were
subjected to decreasing relative humidity, solid crystals did not form at the deliquescence
point (79.5 percent r.h.), but formed at about 30 percent r.h, (in agreement with the observa-
tions of Orr et al., 1958).
The deliquescence and efflorescence points of salt particles relevant to environmental
investigations are presented in Table 2-17.
TABLE 2-17. DELIQUESCENCE AND EFFLORESCENCE POINTS OF SALT PARTICLES6
Composition
NaCl
KC1
NaCl-KCl
(NH4)2S04
(NH4)3H(S04)2
NH,HSO,,
4 4
2NH4NQ3-(NH4)2S04
Deliquescence, % r.h.
75.7
84.3
73.8
79.5
69.0
39.0
56.4
Efflorescence, % r.h.
43
53
38
•v36
Adapted from Tang (1980a).
It will be pointed out in Section 2.4.5.2 and in Chapter 5 that the composition of tropo-
spheric particles may consist of many species. In general, it is not known if the individual
particle is essentially a pure compound in a population of particles of various compounds
(external mixture) or if the individual particle contains a mixture of compounds (internal
mixture) (Winkler, 1975). The composition of the individual particles governs their size as a
function of the relative humidity. For two-component droplets smaller than 0.05 (jm, the
surface curvature affects the vapor pressure of the droplets, and their size is related to the
relative humidity through the modified Kelvin-Gibbs equation (Nair and Vohra, 1975):
ln(p/pQ) - In
2Mcr ri x dp 3x da-,
RTpT Li p dx " 25 dxj
(2-47)
where
p = vapor pressure over the curved surface
p = vapor pressure over a flat surface
a° = surface tension
M = molecular weight of water vapor
p = density
x = solute mass fraction
2-51
-------
2.4
o
51
13
*
111
o
01
N
a
u
oc
2
2.0
1.8
1.6
1,2
1.0
20
THEORETICAL
EXPERIMENTAL
.•Op-
n —.
30
40
50 60 70
RELATIVE HUMIDITY,%
80
90
100
Figure 2-9. Condensational growth and evaporation o
humidity at 25°C.
Source: Tang(1980a).
particles as a function of relative
2-52
-------
R = universal gas constant
T = temperature
r = radius
a = activity of water
w
Nair and Vohra (1975) added the terms in the square brackets on the right side of Equa-
tion 2-47 to correct the Kelvin-Gibbs equation for the variation of surface tension and
density as a function of concentration. The behavior of particles that are external and
internal mixtures can be described by Equation 2-47. However, the Nair and Vohra (1975)
modification has been recently challenged (Renninger et al., 1981) on the basis that nuclea-
tion theory may have been used incorrectly to derive the thermodynamic expression Equation
2-47. Presently, the issue has not been resolved (Doyle, 1981) and the validity of Equation
2-47 is not known. In general, the activity of water (a ) is known only for pure salt solu-
tions, with notable exceptions, such as ocean water, which makes the equation (if correct)
immediately applicable only to external mixtures. There have been attempts (Fitzgerald, 1975;
Hanel and Zankl, 1979; Sangster and Lenzi, 1974) to calculate a for multicomponent electro-
lyte solutions without resorting to detailed theory. Hanel and Zankl (1979) compared the
results for two mixture rules: (a) the mass of water condensed on the mixed electrolyte is
equal to the sum of the masses of water condensed on the separated pure components, and (b)
the practical osmotic coefficient is equal to the molality-weighted practical osmotic co-
efficient of the separate^ pure components. The first mixture rule gave the better acuracy,
but to limit errors to 10-15 percent, Hanel and Zankl advised that it be used only for water
activities larger than 0.85 to 0.9, which corresponds to relative humidities of 85 to 90
percent. Thus, for known internal mixtures, the change in droplet size as a function of
relative humidity can be calculated adequately only for high humidities (greater than 85 to 90
percent).
Ideally, the growth behavior (especially at deliquescence points) of particles as a
function of relative humidity offers a means of distinguishing internal/external mixtures and
of identifying the predominant salts. However, both the likelihood that atmospheric particles
also exhibit hysteresis by forming supersaturated salt solutions and our inability to predict
accurately the activity, of water for multicomponent mixtures for relative humidities below
about 85-90 percent militate against this approach.
-19 -12
The growth of H?SO. droplets with dry mass in the range of 10 to 10 g and their
concentrations (normality) are shown in Figure 2-10, which is based on Equation 2-47 and the
data given in Table 2-18 (Nair and Vohra, 1975). However, as previously pointed out in this
section, the validity of Equation 2-47 has been questioned; thus, the information in Figure
2-10 must also be questioned until the issue of the validity of Nair and Vohra1 s (1975) modi-
fication is resolved. As Figure 2-10 shows, in the absence of atmospheric NHn, H^SO^ droplets
are highly acidic. For example, at relative humidities <40 percent, the concentration of H£SO
2-53
-------
100
n 80 —
£
Q
LU
1
Ul
cc
20 —
0
Figure 2-10. The equilibrium size of su If uric acid solution droplets as a function of relative
humidity. The mass of sulfuric acid (from 10"^ to 10"^ gram) is indicated on the growth
curves (solid). The normality (N) of the solution is marked for selected values from 1 N to
21 N (dashed lines).
2-54
-------
TABLE 2-18. SULFURIC ACID SOLUTION VALUES (25°C)
Mass %
x
0.5
1.0
5.0
10.0
20.0
25.0
40.0
50.0
66.0
85.0
P
g cm"
1.000
1.004
1.030
1.064
1.136
1.175
1.299
1.391
1.560
1.773
dp/dx
x 103
8.1
7.4
6.7
7.0
7.5
7.9
8.8
9.8
11.3
8.7
a
dyn cm
72.0
72.1
72.3
72.6
73.7
74.3
75.8
76.8
75.2
68.7
do/dx
x 102
12.0
8.6
5.5
8.0
11.4
11.3
10.0
6.4
-30.0
-72.0
aw
0.998
0.996
0.982
0.958
0.880
0.823
0.555
0.340
0.075
0.001
H20/H2S04
mole ratio
1081. 0
537.8
103.2
48.9
21.7
16.3
8.1
5.4
2.8
1.0
Normality
0.102
0.206
1.03
2118
4.65
6.00
10.7
14.2
21.0
30.6
Source: Adapted from Nair and Vohra (1975)
2-55
-------
is >14N. Increases in the relative humidity up to 100 percent cause great reductions in H,SOA
-16
concentrations for droplets with dry mass >10 g. At 100 percent r, h. , these droplets grow
without bounds and become infinitely dilute. However, droplets with HpSO. dry mass <10 g
have their growth restricted by the Kelvin-Gibbs effect, which is expressed in the right-hand
side of Equation 2-47. Such droplets have a critical radius at relative humidity >100 percent
~19
that prevents unbounded growth. As an example, the droplet with dry mass of 10 g H?SO.
attains a radius of 0.004 urn and concentraiton of 6N at 100 percent r.h.; it has a critical
radius of M).005 pm which can be attained at ~114 percent r.h. As long as the relative humid-
ity does not exceed 114 percent, the droplet1 s size is governed by the portion of the curve to
the left of the critical radius (0.005 pm); if the relative humidity becomes >114 percent, and
the critical radius is reached, the droplet will then grow without bounds as long as the
relative humidity is XIOO percent. For ambient air and breathing, the relative humidity does
not exceed percent. Thus, from Figure 2-10, it is easy to see that small H9SO, droplets (with
-16
dry mass <10 g) might pose serious threats to health, as hypothesized by Stauffer (1974).
If inhaled, such droplets would attain H7SO. concentrations of 2 to 6N for dry masses of 10"
-19
to 10 g, respectively, if not neutralized by NH.,.
Cl and NO, may be' displaced as HC1 and HNO- from solution droplets or thin aqueous films
on particles by HpSO,. Although the order of the relative acid strengths is HC1 > HoSO. >
HN03 (see Table 2-4), their vapor pressures do not have the same order. The acid-gas vapor-
pressures over aqueous solutions are given by;
PHC1 = '
PHN03
5 x 10"6 atm M~2 (aH+)(ac1-)
2 x 10"7 atm M~2 (aH+)(aNQ-)
+3
PH SO = % SO ^aH SO * = % SO )(K1,H SO * (aH+)(aHSQ-) (2-50)
24 2424 24 24 4
= 3 x 1Q~U atm M~2 (aH*)(aH*)(aSo2"^
i _^-a
The Henry's Law constant H,,»-|(= atm M ) was calculated from the free energy change for
HC1, £ HC1, y (Cotton and Wilkinson, 1980). The value of HHNQ (= 4.8 x 10"6 atm M"1) has
been reported by Schwartz and White (1981). Only recently has H,,SO», * vapor pressure become
available for aqueous solutions (Roedel, 1979; Chu and Morrison, 1980). The value of HH cn
-R -i • n?iU4
(=3 x 10 atm M ) was estimated from Roedel 's (1979) data for 82 percent (mass) H2S04> Thus,
from these vapor pressure equations it can be seen that the vapor pressure of HpSO,/ % will
2-56
-------
be orders of magnitude less than HMO-, -, and HC1, , for tropospheric particles. In acidic
solutions, hLSQ. is effective in volatilizing HNCL and HC1 , causing these acids either to
reside in the gas , phase or to recondense on less acidic particles (Harker et al., 1977;
Hitchcock et al, , 1980). . , - ,
The influence of NO, is an important consideration. Tang (1980b) has performed detailed
•* J
qualitative calculations of the partial pressures of NH-, and HNO^ over the NH--HNO--HoSO»-H?0
system at 25°C. His calculations for the partial pressures PNUO and -Pimno over this system at
85 percent relative humidity^are shown in Figure 2-11. The solution composition ranges from
pure H2S04 ([NH*]/[SC)|~] = 0) to pure, (NH4)2$04 ([NH*]/[$o|~] = 2).- The PRN03 has almost a
linear dependence on [NH.]/[SO. ], while P^LIO^S not sensitive to this parameter, but depends
only on solution pH. Figure 2-11 demonstrates the complex interactions among NH,, HNOo, and
HpSO.. He also studied the effects, of relative humidity and pH on these partial pressures and
deduced that: (1) the HNO- .partial pressure depends strongly on both the relative humidity
and droplet pH, and (2) the , NH, partial pressure varies only slightly with humidity but
+
inversely with H concentration. Tang (1980b) remarked that the strong dependence of the HMO,
partial pressure on relative humidity may affect the nitrate content of 'particles that are
sampled, leading to biases in the> determination of ambient NCU.
Charlson et al. (1978) have reviewed the potential use of the difference in the indices
of refraction for differentiating ambient particles of HpSO, and (NH.)?SO». The ratio of
backward hemispheric scattering to the total scattering was measured near St. Louis in 1973
during periods that were classified as H2$Q. or (NH.^SQ. dominated events. While the ob-
served differences were generally in the right direction, the average value of the backward/
total scatter ratio did not agree with prediction, which led Charlson et al. (1978) to con-
clude that refractive index is too complex a variable to be used as an analytical tool for
2-
differentiating the types of SO, systems.
In summary:
1. Bulk material properties are adequate in most cases for describing the state of
tropospheric particles.
2. However, there is a paucity of thermodynaraic data to permit prediction of deliquesc-
ence and hygroscopic behavior and vapor pressures of multicomponent systems, especi-
ally for relative humidities below about 90 percent.
2.4.2.2.3 Surface properties. The surface properties of particles provide means of detec-
tion, measurement, and collection; and may increase persistence of droplets in trie atmosphere.
Some of the more important surface properties are: electrostatic charging, adhesion, and the
influence of surface films. " "
A number of identifiable mechanisms can lead to' electrostatic charging of, particles
including contact charging, photoionization, field emission charging, and gaseous ion capture.
For practical applications in the troposphere, gase'ous ion capture is the most important of
these mechanisms. Contact also produces charging, as in the triboelectric charging of dust
Z-S7
-------
1000
100
•a
a
LU
cc
1
Ul
K
a.
re
<
a.
10
0.1
0.01
0.5 1.0
MOLAR RATIO CNH^
1.5
2.0
Figure 2-11. NHo and HNOg partial pressures as a function of
droplet's nitrate fc|yOo"^ anc* su'^1*6 ^~^ concentrations
at 85% relative humidity, 25°C.
Source: Tang(1980b).
2-53
-------
blown by wind along the earth's surface. Bursting bubbles may produce charged sea spray
aerosols over the ocean.
Reviews of experiment and theory for gaseous ion capture by aerosol particles may be
found in papers by Bricard (1977) and by Whitby and Liu (1966). Two recognized capture
mechanisms are field charging and diffusion charging. Field charging denotes the process in
which an ion is captured by a particle through the influence of an external electrical field.
Diffusion charging is a process in the absence of an external electrical field. Field charg-
ing of aerosol particles is used in particle control technology for the operation of electro-
static precipitators. Diffusion charging is employed in classification or sizing of aerosol
particles according to their electrical mobility.
The rate of gaseous ion capture by an aerosol particle depends upon a number of para-
meters including the particle size and shape, the dielectric constant of the particle, the
number of charges already on the particle, and the mean free path and mobility of the gaseous
ion, plus, for field charging, the external electrical field strength.
The charging of an aerosol has been shown by Boisdron and Brock (1970) to be a stochastic
process. Inherent difficulties in the use of particle charging as an aerosol detection method
have been shown by Marlow (1978a and 1978b) and by Porstendorfer and Mercer (1978). These
studies indicate that the polydispersity of the aerosol, the dielectric constant of the par-
ticles, and humidity in the presence of trace gases lead to uncertainties in aerosol particle
charges.
Particles are removed from the troposphere by diffusion to or impaction against surfaces.
Particles also collide with each other and stick together. The forces of adhesion that hold
particles to surfaces and to each other include electrostatic forces, capillary forces in the
presence of a liquid, and London-van der Waals forces. In general, for uniform conditions,
the efficiency of particle removal from surfaces by air flow decreases as the particle size
decreases for dry, solid particles (Corn, 1976), While the types of forces are known, the
magnitude of these forces usually cannot be predicted precisely.
The influence of surfacefl1ms on aqueous droplets has been recognized for many years
(Bradley, 1955; Eisner et al., 1960). Chang and Hill (1980) have reviewed some of the studies
on droplet stabilization by surface films. They have also demonstrated that the products of
the reaction between 0, and 1-decene in humid air contain species that adsorb on water
droplets and retard the evaporation rate, Chang and Hill (1980) suggest that photochemical
reactions may produce similar products that would retard the evaporation of urban fogs, and
perhaps extend their duration by hours. However, they report no kinetic data. Eisner et al.
(1960) investigated the kinetics of evaporation of droplets with fatty alcohols on the
surface. They were able to increase the lifetime of an evaporating 10-pm droplet only by a
factor of about 250, which corresponds to about 2.5 minutes. (Another likely cause of stable
fogs is the formation of supersaturated droplets.) At this time, the influence of photo-
chemically produced organic condensates on the kinetics of droplet evaporation is not known.
2-59
-------
2.4.3 pynarm' csof SI ngl e Part id es ,-.,'••*.
The behavior af atmospheric aerosols depends upon" the physical" properties' of the suspend-
ing gas, the particles,< gas-particle interactions, particle-particle interactions, and !the
fluid motion of the gas. Knowledge'of these properties and interactions is essential to our"
understanding of atmospheric phenomena, our ability to formulate predictive models of pol-
lutant particle concentrations and effects, and our ability to sample and measure particles.
In this section, the conditions will be presented for which aerosol systems can be described
in terms of the dynamics of single particles.
According to Hidy and Brock (1970), particles may be considered to be independent of each
other and the dynamics of single particles may be applied if the conditions in Table 2-19 are
satisfied. If coagulation or deposition are important processes, and the stystem satisfies
all but the last condition, they r.efer to the system as being in the "quasi-single particle
regime." As is seen in Table 2-19, the conditions for the "single particle regime" are
generally satisfied for the troposphere.• , ,, , , ,
TABLE 2-19. CONDITIONS FOR THE SINGLE-PARTICLE REGIME3
Conditions
1.
2.
3.
4.
5.
n
X
R
n
Q
i
G
i
i
1
/n
n
n
i
,1
G
1/3
1/3
1/3L
Q
j
VI
K/kT
« 1
« 1
« 1
« 1
« 1
for
for
for
for
for
Range in Troposphere
• > <••(,*•• %
all, i , , :, , - ., ~ 10"13 to 10"19
all i ~ 10~3 to 10"5
all i,j " ~ 10"1 to 10"7
all i ~ 0
all i,j * ^ • - ~ 10"2 to 10"6 ''
a
b
From
Hidy
and Brock (1970).
n.
Ri
Lvi
19 "3
= number concentration of air molecules (~ 10 cm )
= mean free path of air molecules-(~ 10 cm)
= radius of particle, cm , . • »
= characteristic distance (cm) associated with change in numb,er concentration,
temperature, and velocities
= electrostatis charge, esu
K = Debye reciprocal length (effective distance of Goulombic interaction), cm
k = Boltzmann' constant (= 1.38 x 10 erg/K-mole)
T = temperature, K
-1
2-60
-------
The dynamics of single particles include sedimentation, impaction, diffusion, coagula-
tion, electrodynamics, and filtration (Fuchs, 1964; Hidy and Brock, 1970; Friedlander, 1977).
In general, the dynamics can be described adequately for all particle sizes in calm air.
According to Fuchs (1964), complete (~99 percent) entrainment of particles by eddy fluctua-
tions occurs if
t/tL < 0.02 (2-51)
where
T = particle relaxation time
g
V = initial velocity of particle in the absence of external forces, cm/sec
V = terminal (steady) settling velocity of particle in still air, cm/sec
s>
g = gravitational constant
t, = the Lagrangian period of eddy fluctuations, sec
= 0.5rtX/u.
A
X = scale length of eddy fluctuations, cm
u, = root mean square (rms) eddy velocity corresponding to fluctuations on a scale
< X, cm/sec.
Using an rms eddy velocity of 30 cm/sec, which is a typical value ~1 'm above the ground,
and X ~ 40 cm at X/u. ~ 1; in this case, t/t. ~ 0.01 for a particle (unit density) with a
~" A U
diameter of 10 urn. That means particles with diameter < 10 urn will be >99 percent density
entrained in the atmospheric eddy fluctuations. As the rms eddy velocity increases above 30
cm/sec and the particle approaches nearer to the surface or a large obstacle, t., decreases, t
remains constant, and thus the particle diameter corresponding to ~99 percent entrainment must
decrease from 10 urn. It is important to note that larger particles with diameters >1 mm have
< 2 percent entrainment in the eddy fluctuations for these conditions (Fuchs, 1964), which is
adequate justification for ignoring the influence of atmospheric turbulence on their motion.
Thus, the following practical limits for considering atmospheric motion can be stated:
a. Diameter <10 pm: The particles follow the eddy motion with 99 percent entrainment
b. 10 Mm < diameter < 1 mm: The particles lag behind the eddy motion
c. Diameter > 1 mm: The particles do not follow the eddy motion.
These practical limits suggest that in turbulent atmospheres the dynamic behavior of particles
with diameter <10 urn can be described in terms of viscous flow mechanics with superimposed
eddy flow. The dynamic behavior of particles with diameters in the range 10 ym to 1 mm cannot
be similarly described (see Soo, 1967*, Fuchs, 1964), and satisfactory approaches are still
fertile areas for research. The inability to describe completely the motion of such particles
in turbulent air has had an inhibiting effect on the design and use of aspiration samplers for
particles with diameter >10 jim (May et al., 1976). A practical method to reduce biases in
aspirating samplers caused by the inertial effects of flowing large particles is to sample
2-61
-------
1soklneticany, which means that the air streamlines neither converge nor diverge upon enter-
ing the sampler. Isokinetic sampling is attempted by matching the inlet flow velocity of the
sampler to the local air flow velocity. However, even under ideal conditions, the probe
extends a disturbance upwind of the inlet, causing entrance biases for particles having
appreciable inertia (Fuchs, 1964). In the turbulent atmosphere, it is not practical to
attempt isokinetic sampling which would require: (a) a fast-response realigning inlet that can
maintain its axis parallel to the local, rapidly fluctuating wind vector, (b) a fast-response
pumping system that can maintain the inlet flow speed equal to that of the wind's, and (c) a
thin-walled inlet (Belyaev and Levin, 1974). The most common atmospheric particle samplers
(e.g., high volume samplers, dichotomous samplers; see Chapter 3) are operated anisokinetic-
ally. For real, fluctuating turbulent atmospheres, actual trajectories for unit density
particles greater than ~10 urn diameter cannot be calculated; instead, the trajectories are
estimated by ignoring turbulence. While turbulent wind tunnel tests of sampler inlet effi-
ciencies can be performed under steady conditions (Wedding and Weigand, 1980; Liu and Pui,
1981; McFarland et al., 1977), the turbulence in the wind tunnels in general is much lower than
atmospheric levels. The degree of correspondence between wind-tunnel characterizations and
performance of aspirating inlets in turbulent atmospheres is not known; however, for particles
in the low portion (10-15 urn) of the nonviscous flow range, significant differences in the
inlet's entrance efficiency are unlikely. For particles in the complete nonviscous range (10
urn - 1 mm), no practical inlet has been demonstrated at atmospheric turbulent conditions. To
date, it appears that the only reliable atmospheric particle size data for diameter >10 urn has
been obtained with Rotorod and similar samplers that draw the impactor stages through the air
instead of aspirating air through a fractionating device (May et al., 1976; Noll and Pilat,
1971; Johnson, 1976).
The influence of the variety of elevations and temperatures in the United States on
particle formation, growth, and motion must be considered. The important mass transport
parameters for air that influence particle dynamics and that change with altitude and tempera-
ture are atmospheric pressure (or density) and air viscosity. Typical values for several
elevations and temperatures are shown in Table 2-20.
As shown in Table 2-20, viscosity is independent of altitude over the range considered.
Within the contiguous 48 States, the acceleration due to gravity (g) varies within the range
p
of 9,790 to 9,809 m/sec ; this variation is so small that it will have significant effects
only on precise fundamental investigations. The general shift in particle behavior as a
function of air temperature, pressure, and viscosity is given in Table 2-21. The dependence
of particle mechanics on these variables is well known; for many instruments and samplers that
use particle mechanics for sizing and separating, the effects of changes in temperature,
pressure, and viscosity must be considered.
2.4.4 Formation andGrowth ofParticles
Particles are formed by two processes: (1) grinding or atomization of matter, and (2)
nucleation of supersaturated vapors. The particles formed in the first process may be emitted
2-62
-------
TABLE 2-20. MASS TRANSPORT PARAMETERS FOR AIR
Elevation, km
0
1.52
(5,000 ft)
3.05
(10,000 ft)
Standard
Pressure3
kPa
10.33
(1.00 atm)
8.56
(0.83 atm)
7,07
(0.68 atm)
Density ,
-30°C
1.56
1.30
1.06
kg m
20°C
1.29
1.07
0.88
Molecular
Free Path
-30°C
54.0
64.8
79.2
Mean-
, nm
20°C
65.3
78.8
95.4
Viscosity0,
-30°C
1.54
1.54
1.54
_-i _-i
kg m sec
20°C
1.81
1.81
1.81
a Fairbridge (1967)
b Weast (1976)
c Bretsznajder (1971)
NOTE: -30°C = -22°F; 20°C = 68°F.
directly into the atmosphere. However, the particles formed in the second process usually
result from reactions of gases in the atmosphere to yield compounds with low vapor pressures;
when such species reach sufficiently high supersaturation, they nucleate to form particles.
The dynamics of nucleation, which are still understood only incompletely, have been exten-
sively reviewed by Hidy and Brock (1970), who discuss the two types described below:
1. Homogeneous nucleation is the formation of particles by the molecular agglomeration
of supersaturated vapors in the absence of foreign particles and ions. Important
examples include the formation of particles by H2SQ. molecules produced by the
reaction of HO radical with S02, and carboxylic acids formed by the reaction of 0,
and olefins.
2. Heterogeneous nucleation is the condensation of molecules of a supersaturated vapor
onto foreign particles or ions. Important examples include the condensation of
hydrocarbon vapors onto Pb halide and carbon particles during cooling of automobile
exhaust, and the condensation of HpSO. molecules onto fly ash during the cooling of
plumes form power plants burning fossil fuels. Heterogeneous nucleation occurs when
foreign nuclei are plentiful and may suppress the critical supersaturation pressure
below the critical value required for homogeneous nucleation.
Particle growth in the atmosphere occurs through gas-particle interactions, which will be
discussed in this section, and particle-particle interactions (coagulation), which are well
understood and mentioned above in Section 2.4.3.
2-63
-------
TABLE 2-21. DEPENDENCE O'F PARTICLE BEHAVIOR ON AIR TEMPERATURE, PRESSURE, AND VISCOSITY
Behavior
Dependence for variation within the ranges given in Table 2-17
Temperature, T Pressure, p Viscosity, n
Nucleation rate3
Condensation growth rate3
a. Nonvolatile species
b. Volatile species
Sedimentation, velocity in
calm air
Impaction parameter, and
stop distance
Diffusion coefficient
Electrical mobility
increases in com-
plex fashion as
T increases
increases as a
function of T
.1.5
complex; may de-
crease due to in-
crease in partial
vapor pressure of
volatile species
dependence appears
through n; as T
increases, n
increases
none
increases in a
complex fashion
as T increases
dependence appears
through n;, as T
increases, n
increases
probably none
none
depends on p
depends on p
-1
none
none
none for viscous depends on
flow; slight de-
crease for non-
viscous flow (i.e.,
diameter = 100 urn)
as p increases
-1
none for
viscous flow
decreases in a
complex fashion
as p increases
increases in
complex fashion
as p increases
depends on
depends on
depends on
-1
-1
-1
a. See Hidy and Brock (1970) for general discussion.
b. See Fuchs (1964) for equations with dependence on T, p (or mean free path length), and
n-
2-64
-------
Gas-particle interactions include the absorption and the adsorption of pollutant gases,
such as SO,, NOp, hydrocarbons, 0.,, and H2CL, followed by their chemical reactions to yield
products such as SO* , NO.,, and organic compounds. Also included is the condensation of low
vapor-pressure molecules formed in gas-phase reactions, such as HpSO, and organic compounds.
An important limitation on the accumulation of chemical species on submicrometer particles is
the Kelvin-Gibbs effect; The vapor pressures of the solvent and solute (or surface-absorbed
species) are increased as surface curvature is increased. For a condensing species being
formed by gas-phase reactions, there will be a minimum particle size below which condensation
will not occur; this value is determined in part by the supersaturation reached by the species
(Friedlander, 1977).
2.4.4.1 Growth Dynamics--Knowledge of the mass transfer process is necessary to understanding
the growth of particles due to gas-particle reactions in the troposphere and in laboratory
studies. Since the dynamics of transfer of gases to particles has been presented"in great
detail by Hidy and Brock (1970), only general features will be reviewed here. The controlling
mass transfer processes can be identified through their relaxation times (see glossary), which
are:
a. the interparticle diffusion relaxation time, T
lpp - ^i.air"1 <2-52>
which is a measure of the time required for a cloud of particles to attain a uniform
vapor concentration of species i, where only molecular diffusion is important.
b. the diffusion to a single particle, t
*sp = •fy.alr"1' Kn<<1 (2'53)
or
Tsp = Rv."1, Kn « I (2-54)
which is a measure of the time required for transport of species i to a single
particle in a stagnant gas, where only molecular diffusion is important.
c. the momentum to a single particle, T
tm = R u^1 (2-55)
which is the characteristic time for transport on momentum to a particle by fluid
motion.
tsp • V1 = Pe (2-56)
where Pe is the Peclet number and is given by
Pe = fluid velocity x R x D. .Jl, (2-57)
1 9 31 I
2-65
-------
The Peclet number represents the ratio of convective and diffusion transport of gases to
a particle. When this value is « 1, the mass transport relaxation times are governed by
molecular diffusion; however, for larger values, fluid convection, is significant and t and
T will be reduced,
where
-3
N = particle number concentration, cm
2
D. . = gaseous diffusion coefficient of species i in air, cm /sec
I, ai r
R = radius of particle, cm
v. = the molecular mean speed of species i in air, cm/sec
u = the relative speed between the particle and air, cm/sec
Kn = Knudsen number = mean free path of air/R.
ft ~~"%
For particles with radius R < lOpm and' a-maxfmum number concentration N < 10 cm , the relaxa-
tion times are
Tsp $ Tm ^ *pp * 10~3sec»
which has this meaning: The magnitude of a perturbation in the concentration of gaseous
species i will be reduced by 1/e at-time ='t. Thus, in the case stated above, which is reason-
able for lower tropospheric aerosols, stationary conditions in the gas phase will.be achieved
-3
in less than 10 sec.
The relaxation time T, for adsorption-and,solution at clean surfaces may be on the order
a
of that for relaxation of internal molecular energy, which is very small compared to T
However, as mentioned in 2.4.2.2.3, surface films- may significantly retard mass transfer
between the air and aqueous phase. The mechanism and the magnitude of the effect is not
established; probably, the phenomenon is due to the organic molecules in the film aligning to
present their hydrophobic (paraffinie) parts to«the air interface. Thus, the droplet takes on
a "paraffinic" type surface, which will exhibit a much lower condensation coefficient than
clean water surfaces to polar molecules such as HpO, NHj, and S0_. The only relevant studies
of the influence of organic surface films on T have been conducted for the mass transfer of
Q
HgO molecules from the aqueous solution droplets to air (see Section 2.4.2.2.3). The signi-
ficance of such films in reducing T for SOv, and NH, have not been studied. However, the
a L. &
effect has been proposed by Junge and Scheich (1971) to explain their observations of the
simultaneous presence of H?SO, droplets and NH, in London. If organic surface films are able
to reduce T to >5 sec for absorption and subsequent reaction of NH-, in the breath by H«SOA
a ^ o £. *T
droplets, then their inhalation and deposition in the lungs may pose serious health threats to
humans.
For chemical reactions occurring in aqueous droplets., the diffusional transfer of the
reactants must be considered. The diffusion" relaxation time for reacting species i in the
absence of-internal circulation is (Lamb, 1945):
p
Ton = £ /D. (with no internal circulation) (2-58)
•*
-------
where
£ = the thickness of the aqueous film on the particle, or the
diameter of the aqueous droplet
D. = diffusion coefficient of species .i in the aqueous,solution.
i ,aq
Since atmospheric particles are normally not, subjected ,to forces that would maintain
oscillation or deformation of small droplets,-.internal .circulation needs to be considered only
for droplets with diameter > 10 jjm. For droplets with internal circulation;
~4 ~1
T£D = ^ cm sec ^ (with internal circulation). (2-59)
Thus, the diffusional .relaxation time of reactants for likely tropospheric droplets is
expected to be !„„ < 0.1 sec.
For aqueous particles with diameters less than 1 pm, the mass- transport processes of
reacting gases have relaxation times T < 10 sec, except possibly for t for transfer through
""* a
an organic film-air,, interface. If the reactant half-lives are much less than the diffusional
relaxation time, then the droplets may be viewed as homogeneous reactors whose feedstock rate
is controlled by the mass transfer through the interface. Otherwise, the effects of chemical
gradients in the droplet and in the surrounding air must be included (Satterfield, 1970).
2.4.4.2 Sulfuric Acid-Water Growth Dynamics—The growth rate of 100 percent HpSO. submicro-
meter droplets suddenly exposed to humidified gas streams has been measured (Carabine and
Haddock, 1976); it was found that the growth ended when the water vapor equilibrium was
reached in <6 sec. Calculations that ignore the dissipation of the heat of dilution of the
-5 -2
droplet predict growth times of 10 to 10 sec. Gentry and Brock (1968) have performed
mass/heat transfer calculations for 0.1 (j™ diameter, l.OM H-SCK droplets and found that the
heat transfer dominated the growth rate, causing it to be much less than expected from assum-
ing isothermal conditions. Azarniouch et al. (1973) have performed similar calculations for
supermicrometer H-SO. droplets and have also deduced that heat transfer significantly reduces
the growth rate. Thus, it is reasonable to expect that inhaled concentrated H2$0^ droplets
may require a period of <6 sec. to attain their terminal size (governed by the relative humi-
dity in the lungs) and final dilution.
2.4.4.3 Dynamics of Growth by Chemical Reaction—Few studies have been reported on the
chemical reaction rate for gases and particles that are of interest in the lower troposphere.
Attempts to measure the SO^ oxidation rate in free and supported droplets were discussed in
Section 2.3.4. Because of the attendant problems of the inability to separate collected
droplets from the outflow stream of the reactor for free-droplet studies and possible radical
termination at solid surfaces for supported droplets, those studies cannot be accepted as
reliable for estimating mass transfer rates. At this time, the only reaction kinetics studies
for gases and suspended droplets that are applicable to the lower troposphere are those for
2-67
-------
the NH~<- N - HgSO^ droplet system (Robbins and Cadle, 1957; Cadle and Robbins, I960;
Huntzicker et al., 1980). 'Using 98 percent and 12 percent H^SO, droplets with diameters 0.2
to 0.9 pra, Robbins and Cadle (1957) observed that the reaction of NH., with the 98 percent
HgSO, droplets was not diffusion controlled, whereas the reaction with 12 percent H^SO, was.
They interpreted the slower reaction rate for the 98 percent H?SO. droplets (10 percent of the
collisions of NH3 were effective) in terms of surface adsorption followed by the formation of
an activated complex that transmitted the NH~ through the interface. For the 12 percent
HgSO., 100 percent of the NH3 collisions were effective. They did not consider the influence
on growth time of heat transfer and formation of solid (NH4)2S04 as a surface crust; both of
these should be expected to be important in reducing the growth time of concentrated H2SO,
droplets. Huntzicker et al. (1980) investigated the reaction for 0.3 to 1.4 (jm H2SO, droplets
at 8 to 80 percent relative humidity; they found that the rates were between 21 to 70 percent
of the diffusion-limited theoretical rate. They interpreted this range of rates, in addition
to the observed decrease in rate as a function of time for low relative humidity, to indicate
the accumulation reaction product on the surface. Possible heat transfer efforts were not
considered. Cadle and Robbins (1960) also observed a reaction between N02 and Nad solution
droplets that was too fast to measure; the products were NaN03 droplets and HClx *.
No kinetic studies relevant to tropospheric conditions have been reported for gases
reacting with solid particles. In Section 2.3.4, the works of Smith et al. (1969), Chun and
Quon (1973), Siegel, et al. (1974), and Judeikis et al. (1978) were mentioned. The work of
Judeikis and coworkers is of most interest because their reactor design permitted them to
estimate the effectiveness of gas collisions for reacting. They investigated the collision
effectiveness $ of S02, N02, 03, and CO on a variety of materials, including metal oxides,
salts, charcoal, cement, fly ash, and sand. S02 and N02 exhibited high reaction rates with
most materials. Unfortunately, the highest collision efficiency leading to reaction that they
-4
could measure was $ < 10 . (The value for a water molecule condensing on a water droplet is
$ < 10 ; for such values of $ > 10 , there is little difference for gas-particle reaction
rates in the lower troposphere if ~10 <$
-------
In summary:
1. The transfer of reactive gases to lower tropospheric and urban particles is known,
with the exception of the possible role of organic films.
2. The growth rate of hLSQ, droplets suddenly exposed to high humidity has been demon-
strated by experiments and theory to be dependent on the heat transfer rate. The
nonisothermal growth rates are several orders of magnitude greater than those pre-
dicted assuming heat transfer to be unimportant,
3. Few gas-particle chemical reactions relevant to the lower troposphere have been
reported, which is hindering the testing of theory.
4. No quantitative investigations of S02 or NH3 desorption from particles under condi-
tions relevant to the lower troposphere have been reported. One study indicates
that S0? may be carried (as sorbed species) on particles through tubes with reactive
walls in mcuh greater quantities than gaseous S0? can penetrate.
2.4,5 Characterization ofAtmospheric Aerosol
Significant advances have been made in the past decade in regard to elucidating the
nature of the tropospheric particle size, area, volume, and mass distribution functions and
the chemical composition. This section will discuss general aspects of distribution func-
tions, the observed behavior of urban particle distributions, and chemical composition.
Evidence will be presented for the existence of multimodal mass distributions and the differ-
ence in composition of urban coarse and fine particles separated at about 2 urn, but it is not
a sharp division of the chemical composition. One of the major reasons for selecting 2.5 pm
as the separation point is that it occurs in the region of the minimum between the accumu-
lation (fine) mode and the coarse particle mode in volume (or mass) distributions of urban
particles. However, nonurban particles may not have volume (or mass) distributions that
resemble those of urban environments; in such cases, a separation at 2.5 pm may not yield
differences in composition (see Figure 1-1),
2.4.5.1 PistributIon—The multimodal nature (to be discussed below) of tropospheric particle
surface and mass distribution functions remained unrecognized until the early 1970's largely
because of the methods used to present number, size and mass distribution data.
Tropospheric particles are polydisperse. For reason of convenience, particle number (N),
area (A), volume (V), and mass (M) concentration data are usually expressed in terms of a
mathematical distribution function of diameter (D). Such functions are ordinarily charac-
terized by two parameters. The fraction of the total number of particles having diameters
which lie between D and dD is:
dN = f(D)dD (2-60)
with the normalization condition:
/ °° f(D)dD = 1. (2-61)
2-69
-------
The curve representing the function f(D) is called the "number frequency distribution" or the
"number differential" curve. Similarly, the area and volume frequency distributions are:
and
dA = f(D2)dD (2-62)
dV = f(D3)dD (2-63)
where the proper normalization is taken into account.
Prior to the 1970's, atmospheric scientists employed two predominant types of frequency
2 3
distributions [i.e., functional forms of f(D), f(D ), and f(D )]. They were (a) Junge's
(1955) power law distribution for particle size and (b) the log-normal distribution for mass.
Investigators who used instrumentation to measure particle size spectra usually reported
their data in terms of Junge's (1955) power law, which is:
{}} = AD~k (2-64)
where A and k are constants. Clark and Whitby (1967) found that this power law was a reason-
able fit to the number distribution, but it was an inadequate model of the surface and volume
distribution.
Also, investigators who used cascade impactors to determine the mass distribution spectra
usually reported their data in $erms of the log-normal distribution (see Fuchs, 1964; Cadle,
1975). While multimodal log-normal distributions can be recognized from standard plots of
data on log-normal graph paper, three effects combined to mask the multimodal character of
urban particle mass distribution: (1) the cascade impactors had serious (and unknown) inlet
biases against particles larger than about 5 pm, (2) the cascade impactors did not have
operational characteristics that permitted mass fractionation below about 0.2 \im, and (3)
particle bounce distorts the mass distribution in cascade impactors. A warning by Fuchs
(1964) that the log-normal distribution is an adequate model only if the particles are sampled
perfectly and the sampler provides adequate fractionation points went unnoticed.
Whitby et al. (1972) were responsible for a major advance in recognizing that Junge's
(1955) power law and the log-normal distribution function were inadequate models of experi-
mental data for urban aerosols. Instead of seeking other functional forms to express the
number, area, and volume distribution, they plotted dN/d log D, dA/d log D, and dV/d log D
versus log D. The result is seen in Figure 2-12. This type of plot has a convenient feature:
The area under the curve is proportional to the quantity (N, A, or V) between two diameters.
The particle volume between the diameters D-^ and D2 is:
log D?
V(D,, D?) = J ^ (dV/d log D) d log D (2-65)
log D
2-70
-------
in
o
tr
ui
CD
12
— 1.0
JH 0.8
Q.
Q
9
r— 3 0.6
(J
< 0.4
— 0.2
6
J^ 4
D
ca
_O
LU
I
. — 1
I 11 Mllll TTT
NUMBER
__ SURFACE
— • — VOLUME
0.01
01 1
PARTICLE DIAMETER,/
\ —
.-1 H.L.
10
Figure 2-12. Frequency plots of number, surface, and volume distribu-
tions for 1969 Pasadena smog aerosol.
Source: Whitby, in National Academy of Sciences (NAS), Airborne
Particles, (1977).
2-71
-------
For small values of A log D, Equation 2-47 becomes
V(Dr D2) = (AV/A log D) x A log D m (2-66)
where (AV/A log D) is the average value in the interval between log DI and log D«.
The peaks in the three types of distribution plots are called modes. As is evident in
Figure 2-12, there is usually one number mode, one or two surface modes, and two volumemodes
for urban aerosols; sometimes an additional volume mode is observed in the range from 0.005 to
0.05 urn when a strong source of fresh nuclei is close to the sampling site. In strict usage,
a mode is a single point which is a maximum in a frequency distribution; however, aerosol
investigators have often used it to represent the integral of the distribution between the
minima on each side of the maximum. For example, the^ particle volume distribution in Figure
2-12 has maxima (modes) at 0.3 and 8 urn. However, the integral of the particle volume fre-
quency distribution with a maximum at 0.3 urn and minima at ~0.02 nm and ~2 pm is commonly
called the accumulation mode. The use of "mode" to denote integrals over specific limits is
unfortunate, but now so common that change is unlikely. A reader must determine from the
context of an article whether the author uses "mode" to mean a single point or an integral.
The urban particle volume (mass) distribution generally has 2 or 3 modes. The integrals
associated with these modes are: (1) the coarsemode, which usually extends from ~2-3 urn to
~100 urn, and has a maximum in the range 5-50 pm; (2) the accumulation mode, which usually
extends from ~0.02 to ~l-5 urn, and has a maximum in the range 0.1-1.0 |jm; and (3) the Aitken
(or nuclei) mode, which is sometimes observed near a strong source of fresh nuclei and extends
from ~0.005 to 0.05 urn. These modes result from the differences in major source types. The
coarse mode consists mainly of primary particles such as mineral dust. The accumulation and
Aitken modes consist of primary combustion smoke and secondary particles. The limitations of
particle growth processes cause these particles to "accumulate" in the size range from 0.2 to
2 urn. The minimum between the coarse mode and the accumulation mode is generally not at zero
on the distribution curve because the mineral dust size extends below 1 (jm, and the growth of
secondary particles extends above 3 |jm. Also, the minima may shift in the region of 1-5 pm.
In spite of this behavior, the minimum between the coarse and accumulation modes offers an
attractive position to fractionate particle samples for health and welfare considerations.
However, it must be recognized that this mode of the urban particle volume (or mass) distribu-
tion is an idealization. The chemical separation between the accumulation mode (integral) and
the coarse mode (integral) is not sharp. Figure 2-13 shows the more general picture of the
volume (or mass) distributions for a variety of types of locations and conditions that range
from urban to mountain background. It is important to note the influence of combustion
sources, secondary sources, natural sources, windspeed, and turbulence on the volume distri-
bution. It is obvious that the distributions observed at a site will be a composite of these,
and it may exhibit a large degree of fluctuation among the components. Thus, the relative
2-72
-------
f\J
I
irf*
10"
E
o
o» 1°2
o
-FINE
I I IliilJ I I I
COARSE-
L
x
0
%
2
O
tu
O
oc
10
,-2
10
,-1
10U
10'
102
10J
PARTICLE DIAMETER (D),/m>
10S
10"
10
10
ID
'2
10
'3
11 I III
- b
11 nun) 11 uniij i i iuiii| i Mi]
FOREST FIRE:
— "CLEAN"
COMBUSTION
- //'
/
MECHANICAL-
/ \
/FLYASH\
U/S
/ 1 1 '
/fit
/I
x '/ '
>•' / / /
/ /
-FINE-
-j COARSE »~ —
i mill 111 mill 111 mill i i nun! i iimiil i i mi
10"
10'
102
PARTICLE DIAMETER (D),/um
Figure 2-13a,Idealized size distribution for particles found in
typical urban aerosols (mainly from anthropogenic sources)
under varying weather conditions. Note bimodal distribu-
tion under usual conditions and shift in distribution (in-
creasing fine-mode particles, decreasing coarse-mode par-
ticles) under stagnation (1) and serious "smog" conditions,
(2), respectively.
Source: Adapted from Slinn (1976).
Figure 2- 13b,Idealized size distribution for atmospheric par-
ticles from anthropogenic sources, showing fine particle con-
tributions from "clean" high-temperature combustion, and
coarse particle contributions from "dirty" fly ash sources,
forest fires, and crushing and grinding operations. Note
change in distribution near sources (1) and at increasing
distances (2,3,4) from sources.
Source; Adapted from Slinn (1976).
-------
IN*
I
10°
10"
103
.a
I
01
o
u
D
P 10°
O
I 10'1
a
10'2
10
,-3
=l MNlllj i I Illllj I f11 III! j Milltllj | I ||llll| I I
~* c
HURRICANE
//"V-
,^-V \ \\ i
NORTH-ATLANTIC
i mill 1
E_j COARSE— »
\ii ii i mill 111 mill 111 mill 11
10'1
10
,-2
10
,-1
10U
10
,1
10*
PARTICLE DIAMETER (D),Aim
a
O)
Ul
104
105
10"
1015
= l llllll] I IIHH!) II 11 lilt) I 11 Hill] I II Hlli
M*_* J
DUST STORM f
I
o
E
5 102
10'
n-1
10"'
,-3
10
10"
I
/
/,
//
I I IIS
\
\ —
ATYPICAL CONTINENTAL
~ (WITHIN MIXED LAYER)
I-
!~
f|
Iz
.'ill
\ll -
MOUNTAIN
ELEVATIONS
(ABOVE MIXED LAYER)
I II
I I
BACKGROUND
-FINE—j—COARSE
il 1111 mi! 11 n mil i 11 mill
linn
10
.-3
10''
10
,-1
10'
,0
10
PARTICLE DIAMETER (D),M«n
Figure 2-13c.ldealized size distribution for atmospheric par-
ticles from natural sources in a marine setting. Note, in com-
parison to typical background levels over open ocean, in-
creasing levels of coarse-mode particles ranging from those
found in sea spray (1,2) to the extreme cases of storms (3)
and hurricanes (4).
Source: Adapted from Slinn (1976).
Figure 2-13d.Idealized size distribution for atmospheric par-
ticles from natural sources in a continental setting. Note, in
comparison to usual background profiles over typical con-
tinental and high-elevation mountain areas, increasing con-
tributions of coarse-mode particles from wind-blown dusts
(1,2,3), ranging to the extreme case of a dust storm (4).
Source: Adapted from Slinn (1976).
-------
magnitudes of the Aitken, acculation, and coarse mode integrals will vary in an uncorrelated
manner, as will the degree of goodness of the chemical separation of these mode integrals.
Samplers have been devised to collect the particles into two size fractions (coarse and
fine) with a separation diameter in the range of 1-5 pm. The coarse fraction is the mass with
aerodynamic diameter between the separation and the inlet cutoff diameters. Therefore, the
coarse fraction is the portion of the coarse mode that lies between these aerodynamic dia-
meters. The fine fraction is the mass with aerodynamic diameter less than the separation
point; it usually corresponds approximately to the sum of the accumulation and Aitken modes.
A variety of separation diameters in the range of 1-5 urn and inlet cutoff diameters greater
than 10 urn have been used.
Multimodal distributions generally are observed for urban aerosols but may not be de-
tected in other cases, e.g., marine environments or areas dominated by a strong source.
In summary:
1. The particle volume (or mass) frequency function (AV/A log D versus log D) is often
multimodal. The fine-volume fraction may have two or more modes at ~0.02 and ~0.2
urn. The coarse fraction generally has one mode within the range ~5 - 50 Mm-
2. The types of sources that contribute particles to the fine and to the coarse frac-
tions are relatively well known.
3. The particle volume frequency functions for the fine and for the coarse fractions
often behave independently.
2.4.5.2 Composition of Particles—Upon elucidating the multimodal behavior of particle dis-
tributions through the use of the forms AN/A log 0, AS/A log D, and AV/A log D, it was
recognized that the chemical composition of urban particles in the coarse fraction is differ-
ent from that in the fine fraction (with a separation diameter of 1-3 pin). Evidence is cited
2- + +
in Table 2-22, which shows that tropospheric secondary particles containing SO, , NH«, H , and
organics are in the fine fraction, while primary particles consisting principally of basic
minerals are in the coarse fraction. The composition of the fine and coarse fractions are
shown in Figure 2-14, which is an idealization of the bimodal mass distribution. As pointed
out in Section 2.4.5.1, the accumulation and coarse mode integrals are not sharply divided.
Thus, the separation point for the fine and coarse modes does not neatly divide*the particles
by chemical composition.
Investigations of the chemical composition of the fine and coarse particles for urban
aerosols indicate that chemical species may be distributed mainly in the fine or coarse frac-
tion, or both, as is shown in Table 2-22. The major components of the fine fraction of urban
o- + -
particles are S0| , NH», N03, Pb compounds, elemental C (soot), and condensed organic matter.
In Sections 5.5 and 5.6, the composition of the fine and coarse fractions and their acidity
characteristics are discussed in more detail.
2.4.5.2,1 Elemental carbon (soot) and organics. The carbon in fine particles consists of an
elemental component (such as graphite or soot) and an organic component of low volatility.
2-75
-------
-------
TABLE 2-22. CLASSIFICATION OF MAJOR CHEMICAL SPECIES ASSOCIATED WITH
ATMOSPHERIC PARTICLES
Both Fine
Fine Fraction Coarse Fraction and Coarse Fractions Variable
S0^~, C (soot), Fe, Ca, Ti, Mg, NQ~, Cl~ In, Cu, Ni, Mn
organic (con- K, P0^~, Si, Al, Sn, Cd, V, Sb
densed vapors), organic (pollen,
Pb, NH*, As, spores, plant parts),
-j.^"
Se, H , acids bases
Sources: Lee and Goranson (1976); Patterson and Wagman (1977); Durham et al. (1975);
Rahn et al. (1971); Akselsson et al. (1975); Hardy et al. (1976); Gladney
et al. (1974); Lundgren and Paul us (1975); Lee et al. (1968); Lee et al.
(1972).
There are significant differences in the optical properties of elemental and organic carbon
components. Elemental carbon is formed during the combustion of fossil fuels and is emitted
as primary particles (~0.1 urn), which strongly absorb light. The organic component consists
of primary hydrocarbons emitted in combustion exhaust and of secondary organics formed by
photochemical reactions. These primary hydrocarbons and secondary organic vapors either
nucleate or condense on existing aerosols. They do not strongly absorb light, but do con-
tribute to light scattering in urban hazes.
There are only limited data on the mass ratio of elemental/(primary + secondary organics)
for a few cities. Appel et al. (1978, 1979) found that for a, 4-day period in July 1975
elemental carbon was the most abundant carbon species in Pasadena, Pomona, and Riverside.
Also, the concentration of secondary organic carbon was usually twice that of primary hydro-
carbons. Of the secondary organics, hexanedioic and pentanedioic acids were among the most
abundant products; most likely they were oxidation products of cyclohexene and cyclopentene
emitted by motor vehicles. The composition of the organic component retained on filters
varied with the length of the sampling period. The retention of less polar organics (e.g.,
hydrocarbons) was favored by longer sampling time, apparently because of adsorption of such
organics on previously collected material. From total carbon, benzene soluble organics, and
hydrogen analyses of fine particles collected in Denver in November 1971, it was estimated
3
that the elemental carbon was 2.3-3.6 |jg/m for the episode days observed; using Pb concen-
tration as a tracer, it was suggested that in November 1973 in Denver the elemental carbon in
•>
fine particles was 1.7 - 4.4 jjg/m (Durham et al. , 1979). Also, for Denver in November 1973,
2-77
-------
Pierson and Russell (1979) estimated from high-volume samples the total elemental carbon to be
2.9 - 27.6 ug/m3.
Although atmospheric measurements of carbon-containing particles are less complete than
those of sulfates, available results suggest that carbon-containing particles in many loca-
tions, both urban and nonurban, are the second most abundant fine-particle species after
sulfates. At some western urban locations where SO emissions have been small, carbon-con-
)\
taining aerosols have made the largest contribution to fine-particle mass. The concentration
of primary carbonaceous particles is likely to have been even higher in the past in the
Eastern United States when coal was more widely used as a fuel. With the growing use of
industrial coal and wood combustors for home heating, carbonaceous particle concentrations are
likely to increase.
Grosjean (1977) has extensively reviewed the methods of primary and secondary organic
particle identification, and the physical and chemical aspects of their formation. Primary
organics emitted into the atmosphere by industrial sources, motor vehicles, agricultural
activities, and natural sources include: linear and branched alkanes and alkenes, substituted
benzenes and styrenes, quinones, acridines, quinolines, phenols, cresols, phthalates, fatty
acids, carbonyl compounds, polyaromatic hydrocarbons, terpenes, and pesticides. Secondary
organic particles are formed by tne oxidation reactions of the primary organics, ozone, and
nitrogen oxides. Typical products that have been identified are: aliphatic organic nitrates,
dicarboxylic acids, benzoic and phenylacetic acids, and terpene products such as pinonic acid
(Grosjean and Friedlander, 1975; Miller et al., 1972; Schuetzle et a!., 1975). By using
computer-controlled high-resolution mass spectrometry and thermal analysis Schuetzle et al.
(1975) and Cronn et al. (1977) obtained diurnal variations of primary and secondary organics
from 2-hour size-resolved samples.
In an attempt to understand the atmospheric oxidation pathways that yield secondary
organic particles, simple mixtures have been investigated in laboratory chamber studies. As
discussed in more detail by Grosjean (1977), the following trends have been observed by
chamber researchers: (a) most paraffins do not generate aerosols during irradiation, (b)
acetylenes do not form aerosols, (c) all other unsaturated compounds with six or more carbon
atoms can form organic aerosols, (d) cyclic olefins and diolefins form more aerosol than their
1-alkene analogs, (e) conflicting results have been reported on the aerosol-forming ability of
aromatics, (f) carbonyl compounds do not generate aerosol, and (g) mechanical stirring
inhibits particle formation. Cyclic olefins are the most efficient class of organic particle
precursors, due mainly to their high gas-phase reactivity and their ability to form nonvola-
tile dicarboxylic acids.
The chemical composition of organic particles generated in smog chambers is not well
established for suspected important aerosol precursors. Functional group analyses for the
products of olefins, benzene and benzene-substituted compounds, and terpenes that have reacted
with ozone and nitrogen oxides show that the bulk consists of highly oxygenated compounds,
2-78
-------
which Include carbonyls, carboxylic acids, and nitrate esters. Only a few studies of species
Identification have been reported. Detailed aerosol product identification has been reported
for the ozone-1-butene reaction (Lipeles et al., 1973); the NO -toluene, NO -cyclohexene, and
NO -crpinene photoreactions (Schwartz, 1974), and the NO -cyclopentene, NO -cyclohexene, and
1-7-octadlene photoreactions (Grosjean, 1977), Good agreement was indicated by Grosjean
(1977) with Schwartz (1974) for the NO -cyclohexene photoreaction, except that Grosjean ob-
J\
served hexanedioic acid to be the major product (not reported by Schwartz). It is significant
that most of the polyfunctional compounds identified (see the cited papers and the HAS report
for details) have also been Identified as important constituents in ambient aerosols.
The secondary particles formed form alkenes having seven or more carbons (cyclic olefins,
diolefins, and terpenes) grow into the light-scattering range and produce appreciable visibi-
lity reduction. For example, particles formed from cyclic olefins and diolefins have particle
sizes between 0.1 and 0.3 (jm. For such systems, the gas-to-particle conversion process con-
sists of the formation of the supersaturation of the gas phase and subsequent condensation on
preexisting particles.
The rates of conversion of precursor organic vapors to organic particles in Los Angeles
have been estimated to average 1 to 2 percent per hour. This moderate rate of conversion Is
consistent with the observation (see Grosjean, 1977) that organics account for an important
fraction of the fine particles under conditions of intense photochemical activity, while only
a small part of the precursor organic vapors are converted to particulate matter.
2.4.5,2.2 Nitrates. Nitrogen oxide gases are oxidized in the atmosphere to yield HNO,, which
accumulates as nitrate in both fine and coarse particles. Because the topics related to the
transport of nitrogen oxides and their transformation to gaseous and particulate nitrates are
discussed in the document Air Quality Criteria for Oxides of Nitrogen (U.S. EPA, 1982), they
will not be repeated here. (These topics include visibility, environmental transport and
transformation, and acidic precipitation.) Atmospheric nitrates most likely result from
photochemical reactions involving the oxidation of NO and N0? to yield HNO., and organic
nitrates (Demerjian et al., 1974). The measurement of ambient nitrate particles has been
recognized to be subject to significant sampling errors, which are discussed in Chapter 3.
In summary:
1. The composition of the coarse fraction of continental tropospheric particles is
dominated by primary minerals.
2. The composition of the fine fraction of continental tropospheric particles is
2- - + +
dominated by secondary particles that consist mainly of SO, , NO,, NH», H , and
organics, plus primary elemental soot.
3. The fine fraction is often acidic, and the coarse fraction is often basic.
4. The chemical pathways for forming organics and NO- particles are not fully under-
stood.
2-79
-------
2.4.6 Particle-Size Spectrum Evolution
The evolution of the atmospheric particle-size distribution spectra can be related in
principle to pollution sources through the "general dynamic equation" (GDE). The form of the
GDE is complex, requiring the application of sophisticated numerical techniques to obtain solu-
tions through the use of large computers. Since the application of the GDE requires the de-
tailed knowledge of all processes significantly influencing particle formation, growth, and
removal, its main application to date has been- to, simu-late simple model systems. In a few
cases, simulation results have been compared to smog chamber and atmospheric observations.
While this type of research is active, progress has been slowed due to the lack of knowledge
of the important pathways and chemical rates for forming sulfate, nitrate, and organic par-
ticles in the atmosphere. Another limitation is imposed by the incomplete knowledge of the
relation between atmospheric turbulence and particle dynamics. In this section, the GDE will
be presented and recent studies identified in which it has been applied.
2.4.6.1 General DynamicEquation (GDE)—The evolution of the particle size/composition distri-
bution spectra is given by the nonlinear, partial integro-differential equation (Brock, 1976;
Friedlander, 1977; Gelbard and Seinfeld, 1978):
8n./8t + ¥-n. v = ^
* K K~"~
+ I SN - 7-cnk (2-67)
where
n., = the number concentration of particles of type "k" at a specific point r in.space at
K time t, (r,t)
v = the fluid velocity
K = eddy diffusivity tensor
SnC^ rate of input at (r,t) of k-type particles from primary source P
r
eN. = rate of production of particles at (r,t) by homogeneous nucleation of the i-th
1 chemical species
c = particle velocity resulting from external force field (such as gravity).
2-80
-------
In words, the equation has the following meaning:
rate of change in particle transport
composition or In + with velocity v
size distribution (advection)
particle transport rate of change (growth)
= due to dispersion + due to condensation
(convection) and chemical reactions
in particles
rate of change rate of change
+ due to coagulation + due to input from
primary sources
rate of change due rate of change from
+ to homogeneous + external force field
nucleation (sedimentation).
2.4,6.2 Application of the GDE—Through the use of the GDE, it is possible to describe simul-
taneously and quantitatively particle interactions, gas-to-particle conversions, nucleation,
and sedimentation. The GDE can be used as a module in atmospheric K-theory transport/trans-
formation models that may include: (a) primary point, line, and area sources, and (b) gas-
phase photochemistry, producing low-vapor pressure products that condense on existing par-
ticles or nucleate new ones. Because of computer limitations and lack of knowledge of all the
pathways, the GDE has usually been applied by making many simplifications. Many of the
studies and the features of the GDE that they retained are shown in Table 2-23. These studies
have been classified as (a) Growth Laws, (b) Complex Simulations, (c) Comparisons with Chamber
Observations, and (d) Comparisons with Atmospheric Observations.
The "Growth Laws" studies deal with the differences in evolution of the size distribution
due to condensation, chemical reactions on the particle surface, and chemical reactions in the
particle. While the results of these calculations are enlightening, caution must be exercised
in using them to infer atmospheric gas-to-particle pathways. The remaining processes that are
ignored may not be stationary (especially "sources") and may significantly influence the par-
ticle size evolution.
The "Theoretical Complex Simulations" studies are those in which the investigators
realized that processes other than "growth" of particles are important to the size distribu-
tion evolution. These studies demonstrated through theoretical calculations the relative
roles of the processes considered in influencing the evolution. Again, caution is suggested
in using the results to infer atmospheric pathways from simulations that do not incorporate
processes known to be important in the atmosphere.
The "Comparisons with Chamber Observations" studies are useful for developing the growth,
coagulation, and nucleation components of the investigators' GDE model. Table 2-23 references
only those studies in which the investigators predicted the particle size evolution using the
GDE. Takahashi (1970), using a non-chemically reactive system, obtained good agreement for
2-81
-------
Investigators
TABLE 2-23. APPLICATION OF GDE TO DESCRIBE PARTICLE SIZE EVOLUTION
Condensation Coagulation Sources Nucleation External Forces Advection Convection
A. GROWTH LAWS
Brock (1972) V
Seinfeld & Ramabhadran (1975) V
B. THEORETICAL COMPLEX SIMULATIONS
Huang et al. (1970)
Burgmeier et al. (1972)
Wadden et al. (1974)
Burgmeier & Blifford (1979)
Chu & Seinfeld (1975)
Ramabhadran et al. (1976)
Middleton & Brock (1976)
Kiang & Middleton (1977)
Middleton & Brock (1977)
Sheih (1977)
Suck et al. (1977)
Suck & Brock (1979)
Crump & Seinfeld (1980)
Tambour & Seinfeld (1980)
Tsang & Brock (1982)
V
V
V
(continued)
-------
TABLE 2-23 (continued)
DO
OJ
Investigators Condensation Coagulation Sources Nucleation External Forces Advection Convection
C. COMPARISONS WITH CHAMBER
OBSERVATIONS
Takahashi (1970) V V
Heisler & Friedlander (1977) V
Gel bard & Seinfeld (1979) V V V V
McMurry (1980) V V V
D. COMPARISONS WITH ATMOSPHERIC
OBSERVATIONS
Husar et al. (1972) V
Heisler et al. (1973) V
Husar & Whitby (1973) V
Gartrell & Friedlander (1975) V
Heisler & Friedlander (1977) V
Suck et al. (1978) ' V V V V
Eltgroth & Hobbs (1979) V V V V V V
McMurry et al. (1981) V
McMurry and Wilson (1982) V
-------
the evolution of the size distribution in a continuous-stirred tank reactor. Gelbard and
Seinfeld (1979) were able to predict the condensation growth of pre-existing particles for
HcMurry's (1977) SO- photochemical oxidation/particle evolution measurements, but they were
unable to predict new particle formation using the theory of binary nucleation. McMurry
(1980) has reported good agreement between modeling results and the photochemical oxidation/
particle evolution measurements of Clark (1972). He considered the condensation of monomers
on small clusters of molecules and cluster coagulation. Heisler and Friedlander (1977)
s'tudied the gas-to-particle conversion of organics in photochemical smog. They concluded that
their chamber results for particle evolution are explained by the gas-phase oxidation of
organic gases to form low-vapor pressure molecules that condense on pre-existing particles.
Their mechanism included the Kelvin effect, which restricted" the growth to particles above a
critical particle size. Heisler and Friedlander (1977) reported that this particle evolution
mechanism yielded computed particle size distributions which agreed well with atmospheric
measurements. Unfortunately, to date, GDE modeling comparisons with observations for more
complex chemically reactive aerosol systems have not been reported.
The "Comparisons with Atmospheric Observations" studies referenced in Table 2-23 include
only those for which the investigators predicted the particle size evolution using the GDE.
Host of the studies included only particle evolution due to condensation growth and are not
reviewed here. Suck et al. (1978) have used a 3-dimensional K-theory model with primary area
sources to describe the transport and dry deposition of dust in Maricopa Co., AZ. They re-
ported good agreement between predicted and observed suspended mass concentrations. McMurry
et al. (1981) have used a 1-dimensional model to infer the evolution of the averaged cross-
wind particle size distribution. For the data they analyzed, they reported that about 80 per-
cent of the particle volume formation could be accounted for by condensation growth. The most
elaborate application of the GDE has been by Eltgroth and Hobbs (1979) for the evolution of
particle size in coal-fired power plant plumes. They have combined a trace-gas chemistry
scheme for SO , NO , hydrocarbons, and oxidants (35 reactions) with a particle scheme (GDE)
t\ if\ *
including condensation, coagulation, gas-particle reactions, and sedimentation. These two
schemes were used in a K-theory dispersion model to predict the 3-dimensional concentrations
and size distributions. The model predicted the essential features of the plume reactions,
including enhanced reactivity at the outer boundaries. They concluded that diffusion, coagula-
tion, sedimentation, and condensation growth all are important to the particle size distribu-
tion evolution.
The GDE offers an attractive pathway to develop atmospheric aerosol models based on
physico-chemical processes. An active area of research is to formulate such models on large
computers and reduce their size through simplified representation of parameters to obtain
versions that can be operated on smaller, more available computers. This derivative approach
defines the useful range of parameters in relation to the phenomenologically correct parent
model. The alternate pathway of directly formulating reduced parameter models to fit a
2-84
-------
limited number and type of field observations has been used, but those studies will not be re-
viewed here. It is difficult to define the useful range of the parameters of such models.
In Summary:
1. The evolution of atmospheric particle-size distribution spectrum can be described by
the GDE.
2. The application of the GDE requires detailed knowledge of all important processes.
Often, such information is not known.
3. The GDE is suitable for use as a module in K-theory type dispersion models.
4. Most applications of the GDE have been made with extensive simplifications; however,
comparisons with observations of several smog chamber and atmospheric studies have
indicated good agreement.
5. The application of the GOE to relate sources and particle size distributions for the
ultimate use in planning control strategies is an active area of research.
2-85
-------
2.5 REFERENCES
Abel, E. Theory of the oxidation of sulfite to sulfate by oxygen. Monatsh. Chem. 82:815-834,
1951. ~~
Akselsson, R., J. W. Nelson, and J. W. Winchester, Proton scattering for analysis of atmo-
spheric particulate matter. In: Proceedings of the Conference on Nuclear Cross Sections
and Technology, NBS Special Publication 425. Bull. Am. Phys. Soc. 2:484, 1975.
Albu, H. W., and H. D. von Schweinitz. Autoxidations. Vol. V. Formation of dithionate by
the oxidation of aqueous sulfite solutions. Ber. Dtsch. Chem. Ges. 65:729-737, 1932.
Altshuller, A. P. Model predictions of the rates of homogeneous oxidation of sulfur dioxide
to sulfate in the troposphere. Atmos. Environ. 13:1653-1662, 1979.
Altshuller, A. P., and J. J. Bufalini. Photochemical aspects of air pollution: A review.
Environ. Sci. Techno!. 5:39-64, 1971.
Altwicker, E. R. Oxidation of sulfite ion in aqueous solution. In: Control of Emissions
from Stationary Combustion Sources: Pollutant Detection and Behavior in the Atmosphere.
AICHE Symp. Ser. 75 (188):145-150, 1979.
Alyea, H. N., and H. L. J. Backstrom. The inhibitive action of alcohols on the oxidation of
sodium sulfite. J. Am. Chem. Soc. 51:90-109, 1929.
Appel, B. R., E. M. Hoffer, E, L. Kothny, S. M. Wall, M. Haik, and R. L. Knights. Diurnal and
spatial variations of organic aerosol constituents in the Los Angeles Basin. In: Pro-
ceedings: Carbonaceous Particles in the Atmosphere, National Science Foundation and Others
Berkeley, California, March 20-22, 1978. T. Novakov, ed., LBL-9037, University of
California, Lawrence Berkeley Laboratory, Berkeley, CA, June 1979. pp. 84-92.
Appel, B, R. , E. M. Hoffer, E., L. Kothny, S. M. Wall, M. Haik, and R. L. Knight. Analysis of
carbonaceous material in southern California atmospheric aerosols. II. Environ. Sci.
Techno!. 13:98-104, 1979.
Azarnioueh, M. K., A. J. Bobkowicz, N. E. Cooke, and E. J. Farkas. Growth of a sulphuric acid
droplet exposed to water vapor. Can. J. Chem. Eng. 51:590-5, 1973.
BackstrSm, H. L. J. The chain mechanism in the auto-oxidation of sodium sulfite solutions.
Z. Phys. Chem. 625:99-121, 1934.
Barrie, L. A., and H. W. Georgii. An experimental investigation of the absorption of sulphur
dioxide by water drops containing heavy metal ions. Atmos. Environ. 10:743-749, 1976.
Barren, C. H., and H. A. O'Hern. Reaction kinetics of sodium sulfite oxidation by the rapid-
mining method. Chem. Eng. Sci. 21:397-404, 1966.
Bassett, H. , and W. G. Parker. The oxidation of sulphurous acid. J. Chem. Soc.
1951:1540-1560, 1951.
Belike, S., and G. Gravenhorst. Heterogeneous sulfur dioxide oxidation in the droplet phase.
Atmos. Environ. 12:231-9, 1978.
Beilke, S., D. Lamb, and J. Milller. On the uncatalyzed oxidation of atmospheric SO, by oxygen
in aqueous systems. Atmos. Environ. 9:1083-1090, 1975.
2-86
-------
Belyaev, S. P. and L. M. Levin. Techniques for collection of representative aerosol samples.
Atmos. Environ. 5:325-38, 1974.
Bigelow, S. L. Katalytische Wirkungen auf die Geschwindigkeit der Oxydation des
Natriumsulfits durch den Sauerstoff der Luft. [Catalytic action on the rate of oxidation
of sodium sulfite by the oxygen in the air.] Z. Phys. Chem. 26:493-532, 1898.
Billings, C. E., and R. A. Gussman. Dynamic behavior of aerosols. In: Handbook on Aerosols.
R. Dennis, ed. , Report No. ERDA TID-26608, U.S. Energy Research and Development Admini-
stration, Washington, DC, 1976. pp. 40-65.
Bird, R. B., W. E. Steward, and E. N. Lightfoot. Transport Phenomena. John Wiley and Sons,
Inc., New York, NY, 1960.
Boisdron, Y., and J. R. Brock. On the stochastic nature of the acquisition of electrical
charge and radioactivity by aerosol particles. Atmos. Environ. 4:35-50, 1970.
Bracewell, J. M,, and D. Gall. The catalytic oxidation of sulphur dioxide in solution at con-
centrations occurring in fog droplets. In: Air Pollution, Proceedings of the Symposium
on the Physico-chemical Transformation of Sulphur Compounds in the Atmosphere and the
Formation of Acid Smogs, Mainz, Germany, June 1967. Organization for Economic
Co-operation and Development, 1967. pp. 17-26.
Bradley, R. S. The rate of evaporation of microdrops in the presence of insoluble monolayers.
J. Colloid Sci. 10:571-575, 1955.
Bretsznajder, S. Prediction of Transport and Other Physical Properties of Fluids. Pergamon
Press, New York, NY, 1971.
Bricard, J. Physique des aerosols. 2eme Partie, Nucleation, Condensation, Ions,
Electrisation, Proprietes Optiques. [Physics of aerosols. Part 2. Nucleation,
Condensation, Ions, Electrification, Optical Properties.] Rapport CEA-R-4832(2),
Institut de Protection et de Surete Nucleair Departement de Protection, Centre d' Etudes
Nucleaires de Fontenay-aux-Roses, Gif-sur-Yvette, France, 1977.
Brimblecombe, P., and D. J. Spedding. The catalytic oxidation of micromolar aqueous sulphur
dioxide. I. Atmos. Environ. 8:937-945, 1974a.
Brimblecombe, P., and D. J. Spedding. The reaction order of the metal ion catalyzed oxidation
of sulphur dioxide in aqueous solution. Chemosphere 3:29-32, 1974b.
Brock, J. R. Condensational growth of atmospheric aerosols. J. Colloid Interface Sci., 30:
32-36, 1972.
Burgmeier, J. W. and I. H. Blifford, Jr. A reinforced coagulation-sedimentation model for
stratospheric aerosols. Wat. Air Soil Pollut. , 5_:133-147, 1975.
Burgmeier, J. W., I. H. Blifford, Jr., and 0. A. Gillette. A reinforced coagulation-sedimenta-
tion aerosol model. Wat. Air Soil Pollut., 2:97-104, 1972.
Burrows, J. P., D. I. Cliff, 6. W. Harris, B. A. Thrush, and J. P. T. Wilkenson. Atmospheric
reactions of the H02 radical studies by laser magnetic resonance spectroscopy. Proc. R.
Soc. London. A368:463-481. 1979.
Cadle, R. D. The Measurement of Airborne Particles. John Wiley and Sons, Inc., New York, NY,
1975.
Cadle, R. D., and R. C. Robbins. Kinetics of atmospheric chemical reactions involving
aerosols. Discuss. Faraday Soc. 3£:155-61, 1960.
2-87
-------
Calvert, J, G., and R. D. McQuigg. The computer simulation of the rates and mechanisms of
photochemical smog formation. In: Proceedings of the Symposium on Chemical Kinetics
Data for the Upper and Lower Atmosphere, Stanford Research Institute, Warrenton,
Virginia, September 15-18, 1974. Int. J. Chem. Kinet, Symp. (1):113-154, 1975.
Calvert, J. G., F. Su, J. W. Bottenheim, and 0. P. Strausz. Mechanism of the homogeneous
oxidation of sulfur dioxide in the troposphere. In; Sulfur in the Atmosphere,
Proceedings of the International Symposium, United Nations Environmental Program and
Others, Dubrovnik, Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:197-226, L978.
Campbell, W. J., J. C. Sheppard, and B. f. Au. Measurement of hydroxyl concentration in
boundary layer air by monitoring carbon monoxide oxidation. Geophys. Res. Lett.
6:175-178, 1979.
Carabine, M. D. and J. E. L, Haddock. The growth of sulphuric acid aerosol particles when
contacted with water vapor. Atmos. Environ. 10:735-42, 1976.
Carberry, J. J. Chemical and Catalytic Reaction Engineering.- McGraw-Hill, New York, NY,
1976. pp. 194-305.
Carter, W. P. L., A. C. Lloyd, J. L. Sprung, and J. N. Pitts, Jr. Computer modeling of smog
chamber data: progress in validation of a detailed mechanism for the photooxidaticn of
propene and n-butane in photochemical smog. Int. J. Chem. Kinet. 11:45-111, 1979.
Chambers, L. A. Classification and extent of air pollution problems. In: Air Pollution.
Volume I. A. C. Stern, ed., Academic Press, New York, NY,' 1976. pp. 1-21.
Chameides, W. L., and J. C. Walker. Photochemical theory of tropospheric ozone. J. Geophys.
Res. 78:8751-8760, 1973.
Chameides, W. L., and J. C. Walker. A time-dependent photochemical model for ozone near the
ground. J. Geophys. Res. 81:413-420, 1976.
Chang, D. P. Y., and R. E. Hill. Retardation of aqueous droplet evaporation by air pol-
lutants. Atmos. Environ. 14:803-807, 1980.
Chang, S.-G., R. Brodzinsky, R. Toossi, S. S. Markowitz, and J. Novakov. Catalytic oxidation
of S02 on carbon in aqueous solutions. In: Proceedings: Carbonaceous Particles in the
Atmosphere, National Science Foundation and Others, Berkeley, California, March 20-22,
1978. T. Novakov, ed., LBL-9037, University of California, Lawrence Berkeley Laboratory,
Berkeley, CA, June 1979. pp. 122-130.
Charlson, R. J., A. H, Vanderpol, D. S. Covert, A. P. Waggoner, and N. C. Ahlquist.
H2S04/(NH4)2S04 background aerosol: optical detection in the St. Louis region. Atmos.
Environ. 8:1257-1267, 1974.
Charlson, R. J., D. S. Covert, T. V. Larson, and A. P. Waggoner. Chemical properties of trop-
ospheric sulfur aerosols. Jji: Sulfur in the Atmosphere, Proceedings of the Inter-
national Symposium, United Nations Environment Program and Others, Dubrovnik, Yugoslavia,
September 7-14, 1977. Atmos. Environ. 12:39-53, 1978.
Cheng, R. T., M. Corn, and J. 0. Frohliger. Contribution to the reaction kinetics of water
soluble aerosols and S0? in air at ppm concentrations. Atmos. Environ. 5:987-1008, 1971.
Chu, K. S. and R. Morrison. The vapor pressure and latent heat of aqueous sulfuric acid at
phase equilibrium. In: Environmental and Climatic Impact of Coal Utilization,
[Proceedings of the Symposium], Williamsburg, Va., April 17-19, 1979. J. J. Singh and A.
Deepak, eds., Academic Press, NY, 1980. pp. 293-307.
2-88
-------
Chun, K. C. , and J. E. Quon. Capacity of ferric oxide particles to oxidize sulfur dioxide in
air. Environ. Sci. Technol. 7:532-538, 1973.
Clark, W. E. Measurements of aerosols produced by the photochemical oxidation of S02 in air.
Ph.D. thesis. Mechanical Engineering Department, University of Minnesota, Minneapolis,
MM, 1972.
Clark, W. E. , and K. T. Whitby. Concentration and size distribution measurement of
atmospheric aerosols and a test of the theory of self-preserving size distributions. J.
Atmos. Sci. 24:677-687, 1967.
Corn, M. Nonviable particles in the air. In: Air Pollution. Volume I: Air Pollution and
Its Effects. A. C. Stern, ed., Academic Press, New York, NY, 1976. pp. 47-94.
Cotton, F. A., and G. W. Wilkinson. Advanced Inorganic Chemistry: A Comprehensive Text. 4th
Edition. John Wiley and Sons, Inc., New York, NY, 1980.
Coughanowr, D. R. , and F. E. Krause. The reaction of S09 and 05 in aqueous solution of MnSOA.
Ind. Eng. Fund. 4:61-66, 1965. ^ ^ ^
Cox, R. A. , and S. A. Penkett. Oxidation of atmospheric S0? by-products of the ozone-olefin
reaction. Nature (London) 230:321-322, 1971a.
Cox, R. A., and S. A. Penkett. Photooxidation of atmospheric SO,. Nature (London) 229:
486-488, 1971b.
Cox, R. A., and S. A. Penkett. Aerosol formation from sulfur dioxide in the presence of ozone
and olefinic hydrocarbons. J. Chem. Soc. Faraday Trans. I 68:1735-1753, 1972.
Cronn D. R. , R. J. Charlson, R. L. Knights, A. L. Crittenden, and B. R. Appel. A survey of
the molecular nature of primary and secondary components of particles in urban air by
high-resolution mass spectrometry. Atmos. Environ. 11:929-937, 1977.
Crump, J. G. and J. H. Seinfeld. Aerosol behavior in the continuous stirred tank reactor.
AIChE J., 26:610-616, 1980.
Crutzen, P. J. Photochemical reactions initiated by and influencing ozone in unpolluted tropo-
spheric air. Tell us 26:47-57, 1974.
*3 *3 r\
Danilczuk, E,, and A. Swinarski. The complex ion [Fe (111) (SO™) ] . Roczniki Chem.
35:1563-1572, 1961. d n
Dasgupta, P. K. The importance of atmospheric ozone and hydrogen peroxide in oxidizing
sulphur dioxide in cloud and rainwater—further discussion. Atmos. Environ. 14:620-621,
1980.
Dasgupta, P. K. , P. A. Mitchell, and P. W. West. Study of transition metal ion - S(IV)
systems. Atmos. Environ. 13:775-782, 1979.
Davis, D. D. , W. Heaps, and T. McGee. Direct measurements of natural tropospheric levels of
hydroxyl via an aircraft-borne tunable dye laser. Geophys. Res. Lett. 3:331-333, 1976.
Demerjian K. , J. A. Kerr, and J. Calvert. Mechanism of photochemical smog formation. In;
Advances in Environmental Science and Technology. Volume 4. J. N, Pitts and R. J.
Metcalf, eds., John Wiley and Sons, Inc., New York, NY, 1974. pp. 1-262.
Dennis, R., ed. Handbook on Aerosols. Report No. ERDA TID-26608, U.S. Energy Research and
Development Administration, Washington, DC, 1976, 142 pp.
2-89
-------
Doyle, G. J. Response to the comment on self-nucleation In the sulfuric acid-water system.
J. Chem. Phys. 75:1585-1586, 1981.
Durham, J. L., R. K. Patterson, and R. G. Draftz, Carbon in Denver's urban plume. In: Pro-
ceedings: Carbonaceous Particles in the Atmosphere, National Science Foundation and
Others, Berkeley, California, March 20-22, 1978. T. Novakov, ed. , LBL-9037, University
of California, Lawrence Berkeley Laboratory, Berkeley, CA, June 1979. pp. 102-106.
Durham, J. L., W. E. Wilson, T. G. Ellestad, K. Willeke, and K. T. Whitby. Comparison of
volume and mass distributions for Denver aerosols. Atmos. Environ. 9:717-722, 1975.
Eatough, D. J., W. P. Green, and L. D. Hansen. Oxidation of sulfite by activated charcoal.
In: Proceedings: Carbonaceous Particles in the Atmosphere, National Science Foundation
and Others, Berkeley California, March 20-22, 1978. T. Novakov, ed., LBL-9037, University
of California, Lawrence Berkeley Labojatory, Berkeley, CA, June 1979. pp. 131-132.
Eatough, D. J., T. Major, J. Ryder, M. Hill, N. F. Mangelson, N. L. Eatough, and L. D. Hanson,
R. G. Meisenheimer, and J. W. Fischer. The formation and stability of sulfite species in
aerosols. In: Sulfur in the Atmosphere, Proceedings of the International Symposium,
United Nations Environmental Program and Others, Dubrovnik, Yugoslavia, September 7-14,
1977. Atmos. Environ. 12:263-272, 1978.
Eatough, D. J., L. D. Hansen, P. K. Dasgupta, P. A. Mitchell, and P. W. West. Discussions of:
"Study of transition metal ion - S(IV) systems" by Dasgupta et a!., Atmos. Environ.
13:775-782. Atmos. Environ. 13:1725-1729, 1979.
Eigen, M., K. Kustin, and G. Maass. Die Geschwindigkeit der Hydration von S05 in Wassriger
LSsung, Z. Phys. Chem. 30:130-136, 1961. ^
Eisner, H. S., B. W. Quince, and C. H. Clark. The stabilization of water mists by insoluble
monolayers. Discuss. Faraday Soc. 30:86-95, 1960.
Eltgroth, M. W. and P. V. Hobbs. Evolution of particles in the plumes of coal-fired power
plants—II. A numerical model and comparisons with field measurements. Atmos. Environ.,
13:953-975, 1979.
Erickson, R. E., L. M. Yates, R. L. Clark, and C. M. McEwen. The reaction of sulfur dioxide
with ozone in water and its possible atmospheric significance. Atmos. Environ.
11:813-817, 1977.
Eriksson, E. The yearly circulation of chloride and sulfur in nature; meteorological,
geochemical, and pedological implications. Tell us 11:375403, 1960.
Eriksson, E. The yearly circulation of sulfur in nature. J. Geophys. Res. 68:4001-4008,
1963,
Fairbridge, R. W., ed. The Encyclopedia of Atmospheric Sciences and Astrogeology. Encyclo-
pedia of Earth Sciences Series, Volume II. 'Van Nostrand Reinhold Publishing Corp., New
York, NY, 1967.
Fishman, J., and P. J. Crutzen. A numerical study of tropospheric photochemistry using a one-
dimensional model. J. Geophys. Res. 82:5897-906, 1977.
Fishman, J., and P. J. Crutzen. The distribution of the hydroxyl radical in the troposphere.
Atmospheric Science Paper No. 284, Colorado State University, Fort Collins, CO, 1978.
2-90
-------
Fitzgerald, J. W. Approximate formulas for the equilibrium size of an aerosol particle as a
function of its dry size and composition and the ambient relative humidity, J. Applied
Meteorol, 14:1044-9, 1975.
Freiberg, J. Effects of relative humidity and temperature on iron-catalyzed oxidation of SOp
in atmospheric aerosols. Environ. Sci. Techno!. 8:731-734, 1974.
Freiberg, J. E., and S. E. Schwartz. Oxidation of SOp in aqueous droplets: mass-transport
limitation in laboratory studies and the ambient atmosphere. Atmos. Environ.
15:1145-1154, 1981.
Friedlander, S. K. Smoke, Dust and Haze. Fundamentals of Aerosol Behavior, John Wiley and
Sons, Inc., New York, NY, 1977.
Friend, J. P. The global sulfur cycle. In: Chemistry of the Lower Atmosphere. S. I.
Rasool, ed., Plenum Press, New York, NY, 1973. pp. 177-201.
Fuchs, N. A. The Mechanics of Aerosols. Pergamon Press, New York, NY, 1964.
Fuller, E. C., and R. H. Crist. The rate of oxidation of sulfite ions by oxygen. J. Am.
Chem. Soc. 63:1644-1650, 1941.
Fuzzi, S. Study of iron (111) catalyzed sulphur dioxide oxidation in aqueous solution over a
wide range of pH. Atmos. Environ. 12:1439-1442, 1978.
Gartrell, G., Jr. and S. K. Friedlander. Relating particulate pollution to sources: The 1972
California aerosol characterization study. Atmos. Environ., 9:279-299, 1975.
Gelbard, F. and J. H. Seinfeld. The general dynamic equation for aerosols = Theory and appli-
cation to aerosol formation and growth. J. Colloid Interface Sci., 68:363-382, 1979.
Gentry, J. W. and J. R. Brock. A study of the lifetime of aerosol particles. J. Colloid
Interface Sci. 27:691-701, 1968.
Gillespie, R. J. Sulphuric acid as a solvent system. In: Inorganic Sulphur Chemistry. G.
Nichless, ed., Elsevier Publishing Co., New York, NY, 1968, pp. 563-586.
Gladney, E. S. , W. H. Zoller, A. G. Jones, and G. E. Gordon. Composition and size distribu-
tions of atmospheric particulate matter in the Boston area. Environ. Sci. Techno!.
8:551-557, 1974.
Glasstone, L. Thermodynamics for Chemists. Van Nostrand, Princeton, NJ, 1947. pp. 72-73.
Graham, R. A., A. M. Winer, R. Atkinson, J. N. Pitts, Jr. Rate constants for the reaction of
H09 with HO,, SO,, SO N,0, trans-2-Butene, and 2,3,Dimethyl-2-Butene at 300 K. J. Phys.
ChSm. 83:1563-1567, 1979T
Granat, L., H. Rodhe, and R. 0. Hallberg. The global sulfur cycle. In: Nitrogen, Phosphorus
and Sulfur-Global Cycles. SCOPE Report 7. Ecol. Bull. (22):89-134, 1976.
Grosjean, D. Aerosols. In: Ozone and Other Photochemical Oxidants. National Academy of
Sciences, Washington, DC, 1977. pp. 45-125.
Grosjean, D. , and S. K. Friedlander. Gas-particle distribution factors for organic and other
pollutants in the Los Angeles atmosphere. J. Air Pollut. Control Assoc. 25:1038:1044,
1975.
2-91
-------
Haber, F. , and 0. H. Wansbrough-Jones. Autoxidation (VI) action of light on sulfite solutions
in absence and presence of oxygen. Z. Physik. Chem. 618:103-123, 1932.
Hales, J. M. , and S. L. Sutter. Solubility of sulfur dioxide in low concentrations. Atmos.
Environ. 7:997-1001, 1973.
Hanel, G. and 8. Zankl. Aerosol size and relative humidity: water uptake by mixtures of
salts. Tellus 31:478-486, 1979.
Hardy, K. A., R. Akselsson, J. W. Nelson, and J. W. Winchester. ElementaJ constituents of
Miami aerosol as a function of particle size. Environ. Sci. Techno!. 10:176-182, 1976.
Harker, A. B., L. W. Richards, and W. E. Clark. The effect of atmospheric S02 photochemistry
upon observed nitrate concentrations in aerosols. Atmos. Environ. 11:87-91, 1977.
Hayon, E., A. Treinin, and J. Wilf. Electronic spectra, photochemistry, _and autoxidation
mechanism of the sulfite-bisulfite-pyrosulfite systems. The SQ?, SO,, SO., and SO^
radicals. J. Am. Chem. Soc. 94:47-57, 1972. L s <\ a
Hecht, T. A., J. H. Seinfeld, and M. C. Dodge. Further development of generalized mechanisms
for photochemical smog. Environ. Sci. Technol. 8:327-339, 1974.
Hegg, D, A., and P. V. Hobbs. Oxidation of sulfur dioxide in aqueous systems with particular
reference to the atmosphere. Atmos. Environ. 12:241-253, 1978.
Heisler, S. L. and S. K. Friedlander. Gas-to-particle conversion in photochemical smog:
Aerosol growth laws and mechanisms for organics. Atmos. Environ., 11:157-168, 1977.
Heisler, S. L., S. K. Friedlander, and R. B. Husar. The relationships of smog aerosol size
and chemical element distributions to source characteristics. Atmos. Environ.,
7:633-649, 1973.
Hesketh, H. E. Understanding and Controlling Air Pollution. Ann Arbor Science Publishers,
Ann Arbor, MI, 1977.
Hidy, G. M., and J. R. Brock. An assessment of the global sources of tropospheric aerosols.
In: Proceedings of the Second International Clean Air Congress, International Union of
Air Pollution Prevention Associations, Washington, DC, December 6-11, 1970. H. M. Englund
and W. T. Beery, eds., Academic Press Inc., New York, NY, 1971. pp. 1088-1097.
Hidy, G. M., and J. R. Brock. The Dynamics of Aerocolloidal Systems. Pergamon Press, New
York, NY, 1970.
Higgins, W. C. E., and J. W. Marshall. Equivalence changes in oxidation-reduction reactions
in solution: some aspects of the oxidation of sulphurous acid. J. Chem. Soc. 447-458,
1957.
Hilton, C. M., J. J. Christensen, D. J. Eatough, and L. D. Hansen. Fe (III) - S(IV) aerosol
generation and characterization. Atmos. Environ. 13:601-5, 1979.
Hirschfelder, J. 0., C. F. Curtiss, and R. B. Bird. Molecular theory of gases and liquids.
John Wiley and Sons, Inc., New York, NY, 1954.
Hitchcock, D. R., L. L. Spiller, and W. E. Wilson. Sulfuric and aerosols and HC1 release in
coastal atmospheres: evidence of rapid formation of sulfuric acid particulates. Atmos.
Environ. 14:165-182, 1980.
Hoather, R. C., and C. F. • Goodeve. The oxidation of sulphurous acid. III. Catalysis of
manganous sulphate. Trans. Faraday Soc. 30:1149-1156, 1934.
2-92
-------
Hoigne, J., and H. Bader. Ozonation of water: kinetics of oxidation of ammonia by ozone and
hydroxyl radicals. Environ. Sci. Techno!. 12:79-84, 1978.
Horike, N. R. The Rate of Oxidation of Aqueous Solutions of Sodium Sulfite: Influence of
Temperature, Oxygen Partial Pressure, and pH. MS Thesis, University of Washington,
Seattle, WA, 1976.
Huang, C. M. , M. Kerker, and E. Matijevic. The effect of Brownian coagulation, gradient
coagulation, turbulent coagulation and wall losses upon the particle size distribution of
an aerosol, J. Colloid Interface Sci., 33:529-538, 1970.
Huntzicker, J. J. , R. A. Gary, and C. -S. Ling. Neutralization of sulfuric acid aerosol by
ammonia. Environ. Sci. Techno1. 14:819-824, 1980.
Husar, R. B. , K. T. Whitby, and B. Y. H. Liu. Physical mechanisms governing the dynamics of
Los Angeles smog aerosol. J. Colloid Interface Sci., 39:211-224, 1972.
Husar, R. B. and K. T. Whitby. Growth mechanisms and size spectra of photochemical aerosols.
Environ. Sci. Technol., 7:241-247, 1973.
Husar, R. B., D. E. Patterson, J. D. Husar, N. V. Gillani, and W. E. Wilson. Sulfur budget of
a power plant plume. Iri: Sulfur in the Atmosphere, Proceedings of the International
Symposium, United Nations Environment Program and Others, Dubrovnik, Yugoslavia,
September 7-14, 1977. Atraos. Environ. 12:549-568, 1978.
Huss, A., Jr., and C. A. Eckert. Equilibria and ion activities in aqueous sulfur dioxide
solutions. J, Phys. Chem. 81:2268-2270, 1977.
Huss, A., Jr., P. fa Lim, and C. A. Eckert. On the uncatalyzed oxidation of sulfur (IV) in
aqueous solutions. J. Am. Chem. Soc. 100:6252-6253, 1978.
Johnson, D. B. Ultragiant urban aerosol particles. Science (Washington, DC) 194:941-942,
1976. ~"~
Johnstone, H, F. Metallic ions as catalysts for the removal of sulfur dioxide from boiler
furnace gases. Ind. Eng. Chem. 23:559-561, 1931.
Johnstone, H. F., and D. R. Coughanowr. Absorption of sulfur dioxide from the air. Ind. Eng.
Chem. 50:1169-1172, 1958.
Johnstone, H. F. , and A. J. Moll. Formation of sulfuric acid in fogs. Ind. Eng. Chem. 52:
861-863, 1960.
Judeikis, H. S., T. B. Steward, and A. G. Wren. Laboratory studies of heterogeneous reactions
of S00. Atmos. Environ. 12:1633-1642, 1978.
£. "~
Junge, C. E. The size distribution and aging of natural aerosols as determined from electric
and optical data on the atmosphere. Am. Meteorol. Soc. J. 12:13-25, 1955.
Junge, C., and T. G. Ryan. Study of the S0? oxidation in solution and its role in atmospheric
chemistry. Q. J. R. Meteorol. Soc. 84T46-55, 1958.
Junge, C., and G. Scheich. Determination of the acid content of aerosol particles. Atmos.
Environ. 5:165-75, 1971.
Karraker, D. G. The kinetics of the reaction between sulphurous acid and ferric ion. J.
Phys. Chem. 67:871-874, 1963.
2-93
-------
Kellogg, W. W., R. D. Cadle, E. R. Allen, A. L. Lazrus, and E. A. Mar-tell, The sulfur cycle.
Science (Washington, DC) 175:587-596, 1972.
Kiang, C. S. and P. Middleton. Formation of secondary sulfuric acid aerosols in urban
atmosphere. Geophy. Res. Lett., 4:17-20, 1977.
Kuhlman, M. R., D. L. Fox, and H. E. Jeffries. The effect of CO on sulfate aerosol formation.
Atmos. Environ. 12:2415-2423, 1978.
Lamb, H. Hydrodynamics. Dover, NY, 1945.
Larson, T. V. The Kinetics of Sulfur Dioxide Oxidation by Oxygen and Ozone in Atmospheric
Hydrometers. Ph.D. Thesis, University of Washington, Seattle, WA, 1976.
Larson, T. V., N. R. Horike, and H. Harrison. Oxidation of sulfur dioxide by oxygen and ozone
in aqueous solution: a kinetic study with significance to atmospheric rate processes.
Atmos. Environ. 12:1597-1611, 1978.
Lee, R. E., Jr., and S. Goranson. National Air Surveillance Cascade Impactor Network. III.
Variations in size of airborne particulate matter over three-year period. Environ. Sci.
Techno!. 10:1022, 1976.
Lee, R. E. , Jr., S. S. Goranson, R. E. Enrione, and G. 8. Morgan. National Air Surveillance
Cascade Impactor Network. II. Size distribution measurements of trace metal components.
Environ. Sci. Techno!. 6:1025-1030, 1972.
Lee, R. E., Jr., R. K. Patterson, and J. Wagman. Particle-size distribution of metal compo-
nents in urban air. Environ. Sci. Technol. 2:288-290, 1968.
Leighton, P. A. Photochemistry of Air Pollution. Academic Press, New York, NY, 1961.
Levy, H., II. Normal atmosphere: large radical and formaldehyde concentrations predicted.
Science (Washington, DC) 173:141-143, 1971.
Liberti, A., D. Brocco, and M. Possanzini. Adsorption and oxidation of sulfur dioxide on
particles. In: Sulfur in the Atmosphere, Proceedings of the International Symposium,
United Nations Environment Program and Others, Dubrovnik, Yugoslavia, September 7-14,
1977. Atmos. Environ. 12:255-261, 1978.
Lipeles M., C. S. Burton, H. H. Wang, E. P. Parry, and G. M. Hidy. Mechanisms of Formation
and Composition of Photochemical Aerosols. EPA-650/73-036. Rockwell International
Science Center, Thousand Oaks, CA, 1973.
Liu, B. Y. H. and D. Y. H. Pui. Aerosol sampling inlets and inhalable particles. Atmos.
Environ. 15:589-600, 1981.
Lumiere, A. L., and A. Seyewetz. Sur 1'antioxydation des solutions de sulfite de sodium et
sur les antioxydants. [On the antioxidation of sodium sulfite solutions and on the
antioxidants,] Bull. Soc. Chim. France 33:444-454, 1905.
Lunak, S., and J. Veprek-Siska. Photochemical autooxidation of sulphite catalyzed by iron
(III) ions. Coll. Czech. Chem. Commun. 41:3495-3503, 1975.
Lundgren, D. A. , and J. J. Paulus. The mass distribution of large atmospheric particles. J.
Air Pollut. Control Assoc. 25:1227-1231, 1975.
Lyons, D., and G. Nickless. The lower oxy-acids of sulphur. In: Inorganic Sulphur
Chemistry. G. Nickless, ed., Elsevier Publishing Co., New York, NY, 1968. pp. 509-533.
2-94
-------
Marlow, W. H. Unipolar aerosol diffusion charging. Part I. Particle dielectric constant and
ion mobility distribution effects. J. Colloid Inter. Sci. 64:543-548, 1978a.
Marlow, W. H. Unipolar aerosol diffusion charging. Part II. Ion and aerosol polydispersi-
ties: the "diffusion charging mobility analysis" hypothesis. J. Colloid Inter. Sci.
64:549-554, 1978b.
Martin, L. R. , and D. E. Damschen. Aqueous oxidation of sulfur dioxide by hydrogen peroxide
at low pH. Atmos. Environ. 15:1615-1621, 1981.
Martin, L. R., D. E. Damschen, and H. S. Judeikis. The reactions of nitrogen oxides with SO,
in aqueous aerosols. Atmos. Environ. 15:191-195, 1981.
Matteson, M. J., W. Stober, and H. Luther. Kinetics of the oxidation of sulfur dioxide by
aerosols of manganese sulfate. Ind. Eng. Chem. Fundam. 8:677-687, 1969.
May, K. R. , N. P. Pomeroy, and S. Hibbs. Sampling techniques for large windborne particles.
J. Aerosol Sci. 7:53-62, 1976,
McCrone, W. C. , and J. G. Delly. The Particle Atlas. 2d Edition. Volume I: Principles and
Techniques. Ann Arbor Science Publishers, Ann Arbor, MI, 1973. pp. 264-266,
McFarland, A. R., J. B. Wedding and J. E. Cermak. Wind tunnel evaluation of a modified
Anderson Impactor and an all-weather sampler inlet. Atmos. Environ. 11:535-42, 1977.
McKay, H. A. C. The atmospheric oxidation of sulphur dioxide in water droplets in the
presence of ammonia. Atmos. Environ. 5_:7-14, 1971.
McMurry, p. H. On the Relationship Between Aerosol Dynamics and the Rate of Gas-to-Particle
Conversion. Ph.D. Thesis, California Institute of Technology. Pasadena, CA, 1977.
McMurry, P. H, Photochemical aerosol formation from sulfur dioxide: A theoretical analysis
of smog chamber data. J. Colloid Interface Sci., 78:513-527, 1980.
McMurry, P. H., D. J. Rader, and J. L. Stith. Studies of aerosol formation in power plant
plumes—I. Growth laws for secondary aerosols in power plant plumes: Implications for
chemical conversion mechanises, Atmos. Environ., 15:2315-2327, 1981.
McMurry, P, H. and J. C. Wilson. Growth laws for the formation of secondary ambient aerosols:
Implications for chemical conversion mechanisms. Atmos. Environ., 16:121-134, 1982.
W
McNeil's, D. N. , L. A. Ripperton, W. E. Wilson, P. L. Hanst, and B. W. Gay. Gas phase
reactions of ozone and olefin in the presence of sulfur dioxide. In: Removal of Trace
Contaminants from Air, A Symposium at the 168th Meeting, American Chemical Society, 1974.
ACS Symp. Ser. (17):187-200, 1975.
Meszaros, E. The size distribution of water soluble particles in the atmosphere. Idojaras
75:308-314, 1971.
Middleton, P. B. , and J. R. Brock. Simulation of aerosol kinetics. J. Colloid Interface
Sci., 54:249-264, 1976.
Middleton, P. B., and J. R. Brock. Modeling the urban aerosol. J. Air Pollut. Control Assoc.
27:771-775, 1977.
Milbaur, J. , and J. Pazourek. Oxidation of sulfites in concentrated solutions. Chem. Listy
15:34-38, 1921.
2-95
-------
Miller, J. M. , and R. G. de Pena. Contribution of scavenged sulfur dioxide to the sulfate
content of rain water, J. Geophys. Res. 77:5905-5916, 1972.
Miller, M. S., S. K. Friedlander, and G. M. Hidy. A chemical element balance for the Pasadena
aerosol. J. Colloid Interface Sci. 39:165-176, 1972.
Mishra, G. C., and R. D. Srivastava. Homogeneous kinetics of potassium sulfite oxidation.
Chem. Eng, Sci. 31:969-971, 1976.
Holler, D. Kinetic model of atmospheric S0? oxidation based on published data. Atmos.
Environ. 14:1067-1076, 1980. i
Moss, M. R. Sources of sulfur in the environment; the global sulfur cycle. In: Sulfur in
the Environment. Part I: The Atmospheric Cycle. J. 0. Nriagu, ed. , John Wiley and
Sons, Inc., New York, NY, 1978. pp. 23-50.
Nair, P. V. N., and K. G. Vohra. Growth of aqueous sulphuric acid droplets as a function of
relative humidity. J. Aerosol Sci. 6:265-271, 1975.
National Research Council. Subcommittee on Airborne Particles, Airborne Particles, University
Park Press, Baltimore, MD, 1977.
National Air Pollution Control Administration. Air Quality Criteria for Particulate Matter.
- NAPCA-Pub-AP-49, U. S. Government Printing Office, Washington, D.C., 1969.
National Air Pollution Control Administration. Air Quality Criteria for Sulfur Oxides.
NAPCA-Pub-AP-50, U.S. Government Printing Office, Washington, D.C., 1970.
Neytzell-de Wilde, F. G., and L. Taverner. Experiments relating to the possible production of
an oxidizing acid leach liquor by auto-oxidation for the extractions of uranium. In:
Proceedings of the 2d United Nations International Conference on the Peaceful Uses of
Atomic Energy, Volume 3, Geneva, Switzerland, September 1-13, 1958. United Nations,
Geneva, Switzerland, 1958. pp. 303-317.
Niki, H., E. E. Doby, and B. Weinstock. Mechanisms of smog reactions. Adv. Chem. Ser. 313:
16-57, 1972.
Niki, H., P. D. Maker, C. M. Savage, and L. P. Breitenbach. Fourier transform IR spectro-
scopic observations of propylene ozonide in the gas phase reaction of ozone-cis-
butene-formaldehyde. Chem. Phys. Lett. 46:327-330, 1977.
Noll, K. E. and M. J. Pilat. Size distribution of atmospheric giant particles. Atmos.
Environ. 5:527-540, 1971.
Novakov, T., S. G. Chang, and A. B. Harker. Sulfate as pollution particulates: Catalytic
formation on carbon (soot) particles. Science (Washington, DC) 186:259-261, 1974.
Orr, C., F. K. Hurd, and W. J. Corbert. Aerosol size and relative humidity. J. Colloid
Interface Sci. 13:472-482, 1958.
Patterson, R. K. Aerosol contamination from high-volume sampler exhaust. J. Air Poll.
Control Assoc., 30:169-171, 1980.
Patterson, R. K., and J. Wagman. Mass and composition of an urban aerosol as a function of
particle size for several visibility levels. J. Aerosol Sci. 8:269-279, 1977.
Penkett, S. A. Oxidation of S02 and other atmospheric gases by ozone in aqueous solution.
Nature (London) Phys. Sci. 240:105-106, 1972.
2-96
-------
Penkett, S. A., F. J. Sandalls, and B. M. R. Jones. PAN measurements in England—analytical
methods and results. VDI-Ber. 270:47-54, 1976.
Penkett, S. A., B. M. R. Jones, K. A. Brice, and A. E. J. Eggleton. The importance of atmos-
pheric ozone and hydrogen peroxide in oxidizing sulphur dioxide in cloud and rain-
water. Atmos. Environ. 13:123-137, 1979.
Perner, D. D., D. H. Ehhalt, H. W. Paetz, U. Platt, E. P. Roeth, and A. Voltz. Hydroxyl
radicals in the lower troposphere. Geophys. Res. Lett. 3_:466-468, 1976.
Pierson, W. R. , and P. A. Russell. Aerosol carbon in the Denver area in November 1973.
Atmos. Environ. 13:1623-1628s 1979.
Porstendorfer, J., and T. T. Mercer. Influence of nuclei concentration and humidity upon the
attachment rate of atoms in the atmosphere. Atmos. Environ. 12:2223-2228, 1978.
Rahn, K. A., R. Dams, J. A. Robbins, and J. W. Winchester. Diurnal variations of aerosol
trace element concentrations as determined by nondestructive neutron activation analysis.
Atmos. Environ. 5:413-422, 1971.
Ramabhadran, T. E. , T. W. Peterson, and J. H. Seinfeld. Dynamics of aerosol coagulation and
condensation. AIChE J., 22:840-851, 1976.
Rand, M. C., and S. B. Gale. Kinetics of the oxidation of sulfites by dissolved oxygen. In;
Principles and Applications of Water Chemistry. S. D. Faust and J. V. Hunter, eds., John
Wiley and Sons, Inc., New .York, NY, 1967. pp. 380-404.
Reinders, W., and S. I. Vies. Reaction velocity of oxygen with solutions of some inorganic
salts. Reel. Trav. Chim. »44:249-268, 1925,
Renninger, R. G. , F. C. Hiller, and R. C. Bone. Comment on self-nucleation in the sulfuric
acid-water system. J. Chem. Phys. 75:1584-5, 1981.
Riccoboni, L., A. Foffani, and E. Vecchi. Studies of the chemical kinetics of the process of
autoxidation in dilute liquid phases. Note II. Autoxidation of sodium sulfite. Gazz.
Chim. Ital. 79:418-442, 1949.
Robbins, R. C. and R. D, Cadle: Kinetics of the reaction between gaseous ammonia and sulfuric
acid droplets in an aerosol. J. Phys. Chem. 62:469-71, 1958.
Robinson, E., and R. C. Robbins. Sources, abundance, and fate of gaseous atmospheric
pollutants. SRI Project Report PR-6755, American Petroleum Institute, New York, NY,
1968. '
Robinson, E,, and R. C. Robbins. Gaseous sulfur pollutants from urban and natural sources.
J. Air Pollut. Control Assoc. 20:233-235, 1970.
Robinson, R. A., and R. H. Stokes. Electrolyte Solutions. 2nd Edition. Butterworths,
London, England, 1970.
Roedel, W., and W. Junkerman. High volume sampling of sulfur dioxide for isotropic studies by
gas absorption in small droplets. Atmos. Environ. 12:2399-2402, 1978.
Roedel, W. Measurement of sulfuric acid saturation vapor pressure; implications for aerosol
formation by heteromolecular nucleation. J. Aerosol Science, 10:375-386, 1979.
Salomaa, P., R. Hakala, S. Vesala, and T. Aalto. Solvent deuterium isotope effects on
acid-base reactions. III. Relative acidity constants of inorganic oxyacids in light and
heavy water. Kinetic applications. Acta Chem. Scand. 23:2116-26, 1969.
2-97
-------
Sander, S. P., and R, T. Watson. A kinetics study of the reaction of S09 with CH-09. Chem.
Phys. Lett. 77:473-475, 1981. ^ d l
Sangster, J. and F. Lenzi. Choice of methods for the prediction of the water activity and
activity coefficient for multicomponent^aqueous solutions. Can. J. Chem. Eng. 52:392-6,
1974.
Satterfield, C. N. Mass Transfer in Heterogeneous .Catalysis. M.I.T. Press, Cambridge, MA,
1970.
Schenk, P. W., and R. Steudel. Oxidesr of« su.lphur. In: Inorganic Sulphur Chemistry. G.
Nickless, ed., Elsevier Publishing Co., New York, NY, 1968. pp. 365-418.
Schmidkunz, H. Chemiluminescence of .Sulfite Axidation. Ph.D. 'Theses, Joljann Wolfgang Goethe
University, Frankfurt, Germany, 1963.
Schroeter, L. C. Kinetics of air, oxidation of sulfurous acid salts. J. Pharm. Sci.
52:559-563, 1963.
Schroeter, L. C. Sulfur Dioxide: .Application in ,f.oods>, Beverages, and Pharmaceuticals.
Pergamon Press, Oxford, England, 1966.
Schuetzle, D., D. Cronn, A. L. Crittenden, and R. J. Charlson. Molecular ,composition of
secondary aerosol and its , possible origin. Environ. Sci. Techno!. 9_;838-845, 1975.
Schwab, G. M., and M. Stwahmeyer.. On the kinetics of autoxidation of benzaldehyde by sodium
sulfite in the absence of light. Z. Physk. Chemie, New Series 7_:132-157, 1965.
Schwartz, S. E. and W. H. White, Solubility equilibria of the nitrogen oxides and oxyacids in
dilute aqueous solution. In: Adv- Environ. Sci. Eng. J. R. Pfafflin and E. N. Ziegler,
eds., Gordon and Breach, New York, NY, in press, 1981.
Schwartz, W., P. W. Jones, C. J. Riggle, and D. F. Miller. Chemical Characterization of Model
Aerosols. EPA-650/3-74-011, U.S. Environmental Protection Agency, Research Triangle
Park, NC, August 1974.
Scott, W. D., and P. V. Hobbs. Formation of sulfate in water droplets. J,. Atmos. Sci.
24:54-57, 1967.
Seinfeld, J. H. and T. E. Ramabhadran. Atmospheric aerosol growth by heterogeneous condensa-
tion. Atmos. Environ., 9:1091-1097, 1975.
Sheih, C. M. Mathematical modeling of particulate thermal coagulation and transport
downstream of an urban area source. Atmos.. Environ., 11:1185-1190, 1977.
Siege!, S., H. S. Judeikis, and C. C. Babcock. The Role of Solid-Gas Interactions in Air
Pollution. EPA-650/3-74-Q07, U.S. Environmental Protection Agency, August 1974.
Singh, H. B. Atmospheric halocarbons: evidence in favor of reduced average hydroxy! radical
concentration in the troposphere. Geophys. Res. Lett. 4:101104, 1977.
Smith, B. M., J. Wagman, and B. R. Fish. Interaction of airborne particles with gases.
Environ. Sci. Techno!. 9:558-562, 1969.
Soo, S. L. Fluid Dynamics of Multiphase Systems. Blaisdell Publishing Co., Waltham, MA,
1967.
Stauffer, D. On the theory of lung deposition of .very small H90-H,S04 aerosols. Health
Physics 26:365-366. 1974.
2-98
-------
Stevens, R, K. , T. G. Dzubay, G. Russwurm, and D. Rickel. Sampling and analysis ot
atmospheric sulfates and related species. In: Sulfur in the Atmosphere, Proceedings of
the International Symposium, United Nations Environment Program and Others, Dubrovnik,
Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:55-68, 1978,
Stewart, R. W., S. Hameed, and J. P. Pinto. Photochemistry of tropospheric ozone. JGR J.
Geophys.Jies. 82:3134-3140, 1977.
Su, F., J. G. Calvert, and J. H. Shaw. A FT-IR spectroscopic study of the ozone-ethene
reaction mechanism in 02-rich mixtures. J. Phys. Chem. 84:239-246, 1980.
Suck, S. H. and J. R. Brock. Evolution of atmospheric aerosol particle size distributions via
Brownian coagulation: Numerical simulation. J. Aerosol Sci., 10:581-590, 1979.
Suck, S. H., E. C. Upchurch, and J. R. 'Brock. Dust transport in Maricopa County., Arizona.
Atmos. Environ., 12:2265-2272, 1978.
Suck, S. H., P. B. Middleton, and J. R. Brock. On the multimodality of density functions of
pollutant aerosols. -Atmos. Environ., 11:251-255, 1977.
Suck, S. H., E. C. Upchruch, and J. R. Brock. Dust transport in Maricopa County,, Arizona.
Atmos. Environ. 12:2265-2271, 1978.
Takahashi, K. Changes in particle size distribution of aerosols flowing through vessels.
Tech. Repts. Eng. Res. Inst., Kyoto Univ., No. 149, 10 pp., 1970.
Tambour, Y. and J. H. Seinfeld. Solution of the discrete coagulation equation. J. Colloid
Interface Sci., 74:260-272, 1980.
Tang, I. N. Phase transformation and growth of aerosol particles composed of mixed salts. J.
Aerosol Sci. 7:361-371, 1976.
Tang, I. N. Deliquescence properties and particle size change of hygroscopic aerosols. In;
Generation of Aerosols and Facilities for Exposure Experiments, K. Willeke, ed., Ann
Arbor Science Publishers, Ann Arbor, MI, 1980a. pp. 153-167.
Tang, I. N. On the equilibrium partial pressures of nitric acid and ammonia in the atmos-
phere. Atmos. Environ. 14:819-828, 1980b.
Tang, I. N., and H. R. Munkelwitz. Aerosol growth studies. III. Ammonium bisulfate aerosol
in a moist atmosphere. J. Aerosol Sci. 8:321-330, 1977.
Tartarelli, R., P. Davini, F. Morel!i, and P. Corsi. Interactions between SO, and carbon-
aceous particulates. lr\: Sulfur in the Atmosphere, Proceedings of the International
Symposium, United Nations Environment Program and Others, Dubrovnik, Yugoslavia,
September 7-14, 1977. Atmos. Environ. 12:289-293, 1978.
Titoff, A. Contributions to the knowledge of negative catalyses in a homogenous system. Z.
Phys. Chem. 45:641-683, 1903.
Tsang, T. H. and J. R. Brock. Aerosol coagulation in the plume from a cross-wind line source.
(Accepted for publication by Atmospheric Environment, 1981).
Drone, P., H. Lutsep, C. M. Noyes, and J. F. Parcher. Static studies of sulfur dioxide
reactions in air. Environ. Sci. Techno!. 2:611-618, 1968.
U.S. EPA. Air Quality Criteria for Oxides of Nitrogen. Draft final, EPA-600/8-82-026. U.S.
Environmental Protection Agency, Research Triangle Park, NC, September 1982.
2-99
-------
Valensi, G., J. Van Muylder, and M. Pourbalx. Sulphur. In: Atlas of Electrochemical Equili-
bria in Aqueous Solutions. M. Pourbaix, ed., Pergamon Press, Oxford, England, 1966. pp.
545-553.
van den Heuvel, A. P., and B. J. Mason. The formation of ammonium sulphate in water droplets
exposed to gaseous sulphur dioxide and ammonia, Q. J. R. Heteorol. Soc. 89:271-275,
1963.
Veprek-Siska, J., and S. Lunak. The role of copper ions in copper catalyzed autoxidation of
. sulfite. I. Naturforsch 29b:689-690, 1974.
Vol'fkovick, S. I., and A. P. Belopol'skii. Oxidation of sulfites. Report No. 1. J. Appl.
Chem. 5:509-528, 1932.
Wadden, R. A., J. E. Quon and H. M. Hulburt. A model of a growing, coagulating aerosol.
Atmos. Environ., 8:1009-1028, 1974.
Weast, R. C., ed. Handbook of Chemistry and Physics. 57th Edition. CRC Press, Cleveland OH,
1976.
Wedding, J. B. • and M. Weigand. Sampling effectiveness of the inlet to the dichotomous
sampler. Environ. Sci. Techno!. 14:1367-70, 1980.
Whitbys K. T., A. B. Algren, R. C. Jordan, and J. C. Annis. The American Society of Heating
and Ventilating Engineers and airborne dust survey. J. Air Pollut. Control Assoc.
7:157-165, 1957.
Whitby, K. T., and B. Y. H. Liu. The electrical behavior of aerosols. In; Aerosol Science.
N. Oavies, ed., Academic Press, New York, NY, 1966. pp. 59-86.
Whitby, K. T., R. B. Husar, and B. Y. H. Lui. The aerosol size distribution of Los Angeles
smog. J. Colloid Interface Sci. 39:177-204, 1972.
White, D. R., and J, L. Kassner, Jr. Experimental and theoretical study of the sign
preference in the nucleation of water vapor. J. Aerosol Sci. 2:201-206, 1971.
Winkelmann, D. Die elektrochemische Messung der Oxydationsgeschwindigkeit von NapSQ-j durch
gelosten Sauerstoff. [The electrochemical measurement of the rate of oxidation or Na?SO-
by free oxygen.] I. Elektrochem. 59:891-895, 1955. ^
Winkler, P. Chemical analysis of Aitkin particles (<0.2 |jm radius) over the Atlantic Ocean.
Geophys. Res. Lett., 2:45-48, 1975.
Wofsy, S. C., J. C. McConnell, and M. B. McElroy. Atmospheric methane, carbon monoxide, and
carbon dioxide. J. Geophys. Res. 77:4477-4493, 1972,
2-100
-------
3. TECHNIQUES FOR THE COLLECTION AND ANALYSIS OF
SULFUR OXIDES, PARTICULATE MATTER, AND ACIDIC PRECIPITATION
3.1 INTRODUCTION
The 1969 and 1970 Air Quality Criteria documents for particulate matter (PM) and sulfur
oxides (SO ), respectively, (National Air Pollution Control Administration, 1969, 1970) pro-
A
vided a reasonably thorough review of measurement techniques available at that time. Subse-
quent advances in measurement technology for these pollutants have resulted in several new
techniques and more information on the quality of data collected by older methods. This chap-
ter provides a review and, where possible, a critical assessment of both the earlier tech-
niques used in historical monitoring efforts and the newer techniques on which much of the
information gathered in the next few years will be based.
Methods were selected for inclusion in this review based primarily on frequency of their
use in past or current studies. These include routine monitoring applications used in demon-
strating compliance with air quality standards; in support of effects studies, especially epi-
demiology; and in examining long-term trends for the evaluation of control strategy effective-
ness. More widely used research measurement methods that have been used to collect important
ancillary data, such as particle size distributions for aerosols, are also discussed but in
less detail.
Measurement techniques for SO , PM, and acidic precipitation are governed by the chemical
and physical properties of the substances to be measured. Since the chemistry and physics of
SO and ~PM are discussed in detail in Chapter 2, and those of acidic precipitation in Chapters
6-8, only the measurement methods per £e are discussed in this chapter. Chemical analysis
methods for PM and acidic precipitation for constituents such as sulfates are described fol-
lowing the sections on methods of sample collection. The relationship of particles to visi-
bility and their related measurements are discussed in Chapter 9.
Discussion of each sampling and analytical method covered in this chapter includes a
general description, a discussion of the utility and applicability of the method, and, where
information is available, a critical assessment of the method's capabilities. The capabili-
ties described include accuracy, precision, measurement range, sensitivity to interferences,
and reliability. The last parameter (reliability) is strongly influenced by competency of the
operator and completeness of accepted procedure documents. Except in very specific cases, it
is difficult to evaluate these factors and make conclusions about the general usefulness of
the method. Hence, an assessment of the quality of historical data based on reliability of
the method alone is virtually impossible. Many important earlier studies did not collect cer-
tain quality assurance information now shown to be important in field monitoring (Von Lehmden
and Nelson, 1977). In other cases, supporting data were collected, but are no longer avail-
able. Therefore, critical assessment of the methodology will focus on those areas that are
the most important to the general usefulness of the method, except in cases where specific
problems of a selected study were quantified in the open literature.
3-1
-------
3.2 MEASUREMENT TECHNIQUES FOR SULFUR DIOXIDE
3.2.1 Introduction
Atmospheric SO originates from both natural and manmade sources. Sulfur diaxidertS09)
"• &>
is the predominant SO in the atmosphere. This section discusses commonly used techniques for
>/\
determining atmospheric concentrations of S0_.
Manual methods for determination of S02 are those in which sample collection, prepara-
tion, and analysis, or some combination thereof, are performed by hand. Automated methods
are those in which sample collection and analysis are performed continuously and automatically
by devices generally referred to as continuous analyzers.
This section briefly describes each method, emphasizing measurement principle and method
characteristics, such as detection limits and interferences. Sample collection and method ca-
libration are discussed for both manual,and automated methods. The most widely used manual
methods are discussed first. Sulfation methods are presented last because they measure "sul-
fation rate" rather than ambient SO, concentration per se. The discussion of automated
methods follows a semichronological order, with earlier continuous analyzers described first.
Much of the descriptive information in the section is based on a review by Tanner et al.
(1978). Also discussed in this section are various continuous analyzers designated by EPA as
equivalent methods for the measurement of atmospheric S0« to determine compliance with Nation-
al Ambient Air Quality Standards (NAAQS).
3.2.2 Manual Methods
3.2.2.1 Sample Collection—A number of methods use aqueous solutions for collection of S0?.
The efficiency of mass transfer of SO, from air to the solution phase depends on the gas-
liquid contact time, diffusion coefficients of S02 in the gas and liquid phases, bubble size,
concentration of SO,, and solubility of S0? in solution. Calvert and Workman (1960) describe
a method to predict the efficiency of various bubbler designs in collecting S02. Their method
is predominantly qualitative, but it can serve as a useful guide. The most efficient designs
include that of Wartburg et al. (1969); the Greenberg-Smith impinger (Smith et al., 1961);
midget impingers (Jacobs et al., 1957); Drechsel bottles (British Standards Institution,
1963); and packed columns (Bostrom, 1966), which are useful where low flow rates are involved.
In using such devices, care must be taken to prevent carryover of solution at high flow rates
and to compensate for solvent losses by evaporation.
Collection efficiency depends in part on the solution in which SQ« is actively dissolved
and stabilized. One current method involves stabilization of S0« as the sulfite anion in an
aqueous solution of sodium or potassium tetrachloromercurate, with which the sulfite anion com-
plexes. To prevent conversion of sulfite to sulfate, the temperature of the collecting solu-
tion must be maintained below 20°C. Failure to maintain temperature control of samples during
collection, shipment, and storage leads to underestimation of SOp levels in the atmosphere,
particularly during summer months. Another approach involves collection in aqueous solution
and conversion to the sulfate anion by oxidizing agents such as H-,0,,. Although stabilization
3-2
-------
of SOp as the sulfate anion can be effective, some of the soluble sulfate in the atmospheric
aerosol is collected (unless removed by a particle filter) and added to the sample; thus, dis-
crimination between SO- and sulfate may be impossible.
Several methods employ alkaline solutions for the absorption of SO,,. Although their col-
lection efficiency is quite high, alkaline solutions rapidly oxidize the collected sulfite
anion to sulfate unless some means are available for the direct complexation and stabilization
of the sulfite anion.
Another collection technique uses filter papers or tapes impregnated with an alkaline re-
agent such as potassium hydroxide, triethanolamine, or potassium carbonate, together with
small amounts of glycerol as an humectant (Lodge et a!., 1963; Huygen, 1963; Pate et a!,,
1963; Forrest and Newman, 1973). The collected SO, is supposed to be maintained as a sulfite,
but it may be oxidized to sulfate. Although laboratory tests have shown that such oxidation
can be negligible, field tests have produced very erratic results. Typical PM contains traces
of transition metal ions, which promote rapid oxidation of sulfite to sulfate. Prefilters
should be used to eliminate PM. Oxidation of the collected sulfite to sulfate prior to
analysis is also recommended. As an alternative, an analytical technique that measures the
sum of sulfite and sulfate may be employed.
Some of the earlier methods for estimating ambient S0? concentrations (sulfation methods)
are based on the reaction of SO^ with lead dioxide to form lead sulfate (Wilsdon and McConnell,
1934). The SO^ is stabilized in the form of a sulfate, eliminating the problem of oxidative
conversion; however, any particulate matter containing sulfate species that comes into con-
tact with the collection surface will lead to errors.
Occasionally, samples of ambient air are collected in a gas-tight syringe or other suit-
able container for later analysis. The reactivity of S09 is a major problem, however.
®
Natusch et al. (1978) have reported extensive adsorption losses of SO,, on thick-walled Mylar
® ®
laminates, Tygon , Teflon , and stainless steel container walls.
3.2.2.2 Calibration—The relationship between true pollutant concentration and the measured
value by any method is determined by calibration. For methods that measure relative exposure
to sulfur species (e.g., sulfation methods), no calibration is usually attempted. With these
methods, use of uniform reagents, equipment, and procedures is essential to compare exposure
data over time and space. Methods involving direct collection of air samples for later analy-
sis or collection of the S02 in an air sample by absorption or adsorption require calibration
of both the sample volume measurement and the analytical measurement.
Devices used for sample volume measurement generally are calibrated against reliable vol-
ume standards. The analytical measurement often is calibrated statically, using a known
amount of the sulfite or sulfate anion in solution. Static calibration is a rapid and simple
method for checking the analytical procedure, but does not subject the overall measurement
method to scrutiny since the process of S02 collection is circumvented. Dynamic calibration
of these methods has an advantage over the static approach because it scrutinizes the total
3-3
-------
measurement, but it is time consuming and therefore not used routinely. This approach,
described in more detail in Section 3.2.3 on automated methods, uses synthetic atmospheres
containing the pollutant in known concentrations to define the response of the method.
3.2.2.3 Measurement Methods—This section deals with the principal manual methods for deter-
mining S02 in the air.
3.2.2.3.1 Colori'metric method: pararosaniline. The West-Gaeke method is probably the most
widely used colorimetric procedure for SCL determination in ambient air (West and Gaeke,
1956). It is also the basis of the EPA reference method for measurement of S0? in the atmos-
phere (U.S. Environmental Protection Agency, 1979). In the West-Gaeke method, air is bubbled
into fritted bubblers containing 0.1 M sodium tetrachloromercurate (TCM) solution, which forms
a stable complex with SO,,. This complex, which resists air oxidation, was thought to be the
dichlorosulfitomercurate (II) ion. Recently, however, Dasgupta and DeCesare (1981) have
clearly demonstrated that the SO., group is bonded to mercury through the sulfur atom rather
than through one of the oxygen atoms and that the complex is actually a monochlorosulfonato-
mercurate (II) ion. The SO^-TCM complex is reacted with acid-bleached pararosaniline and
formaldehyde to form red-purple pararosaniline methanesulfonic acid. The optical absorbance
of the solution is measured spectrophotometrically at 560 nm and is, within limits, linearly
proportional to the concentration of SO,,. The method is applicable to the measurement of SO,,
in ambient air using sampling periods from 30 minutes to 24 hours. The lower limit of
detection of S09 in 10 ml of TCM absorbing solution is approximately 0.5 ug, representing a
3
concentration of 13 |jg 50,,/m (0.005 ppm) in an air sample of 38.2 liters. Ozone, nitrogen
dioxide, and heavy metals were negative interferents in early versions of this method.
An improved version of the West-Gaeke method was adopted by the EPA in 1971 as the refer-
ence method for determining atmospheric S0? (U.S. Environmental Protection Agency, 1979).
Several important parameters were optimized, resulting in greater sensitivity and reproduci-
bility, as well as adherence to Beer's Law throughout a greater working range. In the EPA
method, S02 is collected in impingers containing 0.04 M potassium tetrachloromercurate. A
20-minute wait before analysis allows ozone, a potential interferent, to decompose. Sulfamic
acid is then added, followed by a 10-minute wait, to -remove interference from nitrogen oxides.
Interference by heavy metals is eliminated by use of phosphoric acid in the dye reagent and
the disodium salt of ethylenediaminetetraacetic acid (EDTA) in the TCM absorbing solution.
The complex is then reacted with a purified pararosaniline dye reagent and formaldehyde to
form the colored pararosaniline methanesulfonic acid. Absorbance is measured at 548 nm. Ac-
curacy depends on rigid control of many critical variables: pH, temperature, reagent purity,
color development time, age of solutions, and concentrations of some atmospheric interferents
(Scaringelli et al., 1967). Because temperature affects rate of color formation and color
fading, a constant-temperature bath is recommended for maximum precision. Highly purified re-
agents, especially the pararosaniline dye, are vital for acceptable reproducibility. The pre-
cision of the "EPA reference method analytical procedure was estimated using standard sulfite
3-4
-------
samples (Scaringelli et a!., 1967) and reported to be 4.6 percent at the 95-percent confidence
level. The lower limit of detection of S0« in 10 ml of TCM absorbing solution was 0.75 ug,
tf
representing a concentration of 25 ug S02/m (0.01 ppm) in an air sample of 30 liters.
A collaborative study (McCoy et a!., 1973) of the 24-hour EPA reference method indicated
the following: method repeatability (day-to-day variability within an individual laboratory)
varies linearly with S02 concentration from ± 18 ug/m (0.007 ppm) at concentration levels of
100 ug/m (0.04 ppm) to ± 51 ug/m (0.019 ppm) at concentration levels of 400 jjg/m3 (0.15 ppm);
method reproducibility (day-to-day variability between two or more laboratories) varies
linearly with S09 concentration from ± 37 ug/m3 (0.014 ppm) at 100 ug/m3 to ± 104 ug/m3 (0.040
3
ppm) at 400 ug/m . The method has a concentration-dependent bias. This bias becomes
3
significant (95-percent confidence level) at the 400 pm/m level. Observed values tend to be
lower than the expected S0? concentration level.
Results of the above collaborative study and other investigations (Blacker et al., 1973;
Bromberg et al. , 1974; Foster and Beatty, 1974) suggest that pararosaniline methods tend to
underestimate S0? concentrations by 5 to 20 percent. In the Bromberg study, simulated 24-hour
bubbler samples were analyzed by 134 laboratories throughout the United States. Observed neg-
3 3
ative biases ranged from -3 percent for a 45 ug/m sample to -16 percent for a 767 ug/m sam-
ple, but reasons for the negative biases have not been determined. Based on the Bromberg
study results, EPA recommended that intralaboratory quality control programs be upgraded and
improved in laboratories that routinely analyze SO?-TCM samples. EPA also recognized the need
for and promoted development of standard reference samples for use in laboratory quality con-
trol programs.
More recent information, on the reliability of pararosaniline analytical procedures has
been obtained through EPA's ambient air audit program. In this program, freeze-dried mixtures
of sodium sulfite and TCM are sent to participating laboratories for analysis. These
simulated field samples represent ambient S09 concentrations ranging from about 10 to 200
3
pg/m (0.004 to 0.076 ppm). EPA audit results from 1976-1978 summarized by Bromberg et al.
(1979, 1980) indicate no apparent problems with bias (accuracy) in the analytical portion of
the pararosaniline methods.
Subsequent to promulgation of the S0« reference method, effects of temperature on the
method have been studied (Kasten-Schraufnagel et al. , 1975; Sweitzer, 1975). Fuerst et al.
(1976) showed that collected SO^-TCM samples decay at a temperature-dependent rate. Table 3-1
indicates that sample collection at 25°C results in a 1.1 percent loss in SO, during the 24-
hour sampling period, but further exposure of the collected sample for 4 days at this tempera-
ture leads to a 10 percent loss in S0?. Significant decay can occur during collection of am-
bient samples and during shipment and storage of collected samples when TCM solutions are ex-
posed to temperatures above 20°C. Under typical field conditions, temperature exposure is
quite often extreme, especially during the summer months at sites with relatively little pro-
tection from the elements (e.g., rooftops).
3-5
-------
TABLE 3-1. TEMPERATURE EFFECT ON COLLECTED S09~TCM SAMPLES
(EPA REFERENCE METHOD) *
°c
15
20
25
30..
35
40
op
59
68
77
86
95
104
' •
At end
of sampling
99.8
99.6
98.9
97.4
95.1
87.6
Percent
2
99.0
97.8
94.4
87.4
74.1
50.8
SOp remaining
Days of Exposure
4
98.2
96.1
90.2
78.5
57.9
29.5
• 6
97.4
94.3
86.1
70.4
45.2
17.2
Source: Fuerst et al. (1976).
Measures to minimize these temperature effects have been investigated by Martin (1977),
who recommends use of thermostatted shelters to house sampling equipment during sample collec-
tion. The temperature of samples during shipment can be controlled with cold-pack shipping
containers. When samples are stored before being analyzed, refrigeration at 5°C minimizes
further decay. Temperature control procedures are currently being incorporated in the EPA
reference method.
A variation of the pararosaniline method that eliminates the use of the toxic mercuric
chloride has recently been reported (Dasgupta et a!., 1980). In this method a dilute solution
of formaldehyde, buffered at pH 4 with potassium hydrogen phthalate, is used to collect and
stabilize atmospheric SO, as hydroxymethanesulfonic acid. Sulfite, liberated from the
compound by base, is added to acidic pararosaniline for color development. The optical absor-
bance of the colored solution is measured at 580 nm. The method is comparable to the estab-
lished pararosaniline methods in absorption and recovery efficiency, sensitivity, and
precision. No unusual interferences are observed due to 0.,, N0?, and transition metal ions,
except Mn (II). The stability of collected S02 samples is significantly greater than that
observed for samples collected in TCM. The reported decay rates are 0.033 percent/day and 0.3
percent/ day at room temperature and 37°C respectively. No photochemical degradation of
collected samples was observed after 8 hours of exposure to bright sunlight.
Under the provisions of EPA's "Ambient Air Monitoring Reference and Equivalent Methods"
regulations (U.S. Environmental Protection Agency, 1979b), two additional pararosaniline
methods have been designated as equivalent methods (U.S. Environmental Protection Agency,
1975). These methods are identified as:
3-6
-------
1. EQS-0775-001, "Pararosaniline Method for" the Determination of Sulfur
Dioxide in the Atmosphere - Technicon I Automated Analysis System."
2. EQS-0775-002, "Pararosaniline Method for the Determination ofSulfur
Dioxide in the Atmosphere - Technicon II Automated Analysis System."
These methods employ the same sample collection procedure used in the EPA reference method and
an automated analytical measurement based on the colorimetric pararosaniline method.
3.2.2.3.2 Titrimetric method: hydrogen peroxide. The British Standard Method uses a single-
day or 7-day sampling instrument for the measurement of smoke and SO, (British Standards
Institution, 1963). Air is drawn through a filter paper and into a Drechsel bottle containing
~0.3 percent hydrogen peroxide solution adjusted to pH 4.5. The H^Op oxidizes the atmospheric
S02 to HoSO,, which is subsequently titrated with standard sodium borate using the mixed
indicator of the British Drug House (gray at pH 4.5). The method is capable of measuring SO,
3
concentrations from about 25 to 25,000 jjg/m (0.01 to 10 ppm) using a 24-hour sampling period.
Since the method measures total acidity rather than S0? specifically, any strong acids
that are collected produce positive errors. Normally the concentration of such substances is
low relative to that of SO,, and the measurement is generally accepted as a good approximation
of the actual S02 concentration. Ammonia will neutralize the HpSO, and give negative errors.
When the presence of ammonia is suspected, a portion of the absorbing solution can be analyzed
for dissolved ammonia and the S0? measurement adjusted accordingly.
An instruction manual on the use of the hydrogen peroxide method in the British National
Survey was issued in 1966 (Warren Spring Laboratory, 1966). The manual discusses the quality
of the water used for reagent preparation and states that it need not be free of carbon
dioxide. Martin and Barber (1971), however, reported that use of water rich in carbon dioxide
can lead to significant negative errors in the method. During sample collection and
subsequent standing, sufficient carbon dioxide can be evolved from the absorbing solution to
cause low titers and, on some occasions, to result even in alkaline solutions. The instruc-
tion manual also discusses the problem of alkaline contamination in the glassware required in
the method. The Drechsel bottles used during sample collection and sample storage bottles
need to be conditioned with absorbing reagent prior .to use. Likewise, alkaline contamination
in other glass parts of the sampling apparatus can lead to underestimation of ambient SO™
levels.
Evaporation of absorbing reagent during sampling can result in overestimation of ambient
S02 levels with the hydorogen peroxide method (Fry, 1970). If evaporation occurs, the pH of
the solution is lowered and a portion of the standard alkali added during the subsequent
titration is required to compensate for this effect alone. The effect is likely to be more
prevalent in the summer months and can lead to overestimation of S0« levels by up to about 15
3
M0/m (0.006 ppm). Fry reports that this source of error can be overcome either by making up
the absorbing solution to its original volume prior to the titration or by making a mathemati-
cal correction to the titration result based on the final volume of absorbing solution after
collection.
3-7
-------
Uncertainty in the tltration endpoint and rounding-off of the volume of alkali required
in the titration to the nearest 0.1 ml each introduce errors of up to about ±5 ug/m (0.002
ppra) (Warren Spring Laboratory, 1975). Other potential sources of error in the method include
inaccurate air sample volume measurements and conversion of ambient SO* to sulfate on the
smoke filter used in the sampler.
Warren Spring Laboratory (1962) has reported the reproducibility of the hydrogen peroxide
method, based on results from five comparative studies using duplicate sampling apparatus.
The coefficients of variation were on the order of 15 to 20 percent for SO, concentrations
3
ranging from about 15 to 250 (jg/m (0.006 to 0.095 ppm) and 5 to 10 percent for concentrations
o
ranging from about 100 to 800 ug/m (0.033 to 0.305 ppm). The same reagents and analytical
apparatus were used to service all the samplers in each study, thus obviating a further poten-
tial source of error.
In a more recent investigation (Barnes, 1973), duplicate S0? measurements were obtained
with the British Standard Method at a residential site where ambient levels were low (18 to 84
|jg/m ) (0.007 to 0.032 ppm). Nineteen sets of observations were made from two samplers with
a common inlet using the same supply of reagents and glassware and a further 18 sets of obser-
vations were made using a separate supply for each. Differences in measured concentrations
using the two samplers on individual occasions ranged up to 31 percent of the mean of the two
separate values. Most of these differences, however, did not exceed 13 percent of the mean.
Titration error was cited as the single most common source of variation between the samplers
in these experiments. An error in titration of 0.1 ml would result in an error in the
3
measured S0? concentration of 7 ug/m (0.003 ppm). When measuring low concentrations, such
errors could represent a difference of 100 percent from the true concentration. Barnes
concludes from these observations that measurement of low SO, concentrations with the method
require great care on the part of the operator, more than might be expected of most operators.
3.2.2.3.3 lodimetric methods. Several iodimetric methods have been used for measuring S02 in
the atmosphere. In one version, an absorbing solution containing soluble starch, potassium
iodide, dilute H^SO,, and standard 0.01 N iodine solution is prepared (Katz, 1950). SO^ in
the air sample reacts with this 8 x 10 N iodine solution to decolorize the blue iodine-
starch complex. The reduction in color intensity is measured spectrophotometrically. The
range of applicability is 25 to 2600 ug S0?/m (0.01 to 1 ppm), depending upon the volume and
concentration of absorbent solution and the volume of air sampled. In a modification of this
method, the excess iodine is titrated with a standard thiosulfate solution (Katz, 1969).
Oxidizing gases interfere to give low results; reducing agents interfere to give high
results. Interference from high concentrations of nitrogen oxides or 03 can be removed by
introducing hydrogen into the air sample and passing the mixture over a platinum catalyst at
100°C (Bokhaven and Niessin, 1966).
In another version, air is bubbled through a sodium hydroxide solution that absorbs SO™
(Jacobs, 1960). After acidification of the solution, the liberated sulfurous acid is titrated
3-8
-------
with standard iodine solution, using starch as an indicator. Because sulfite oxidizes to sul-
fate in the alkaline absorbent solution, samples cannot be stored. Oxidizing agents, NO?J and
0_ interfere, resulting in an underestimation of the S0? concentration. Hydrogen sulfide
(H?S) and other reducing agents result in an overestimation. For an 850-liter air sample
1
collected at 30 liters/minute, the lower limit of detection is 25 ug SOp/m (0.01 ppm)
(Terraglio and Manganelli, 1962).
3.2.2.3.4 Impregnated filter paper methods. Filter papers, impregnated with alkali plus a
humectant to keep them moist, will absorb S02 from air samples (Lodge et al., 1963; Huygen,
1963; Pate et al., 1963; Forrest and Newman, 1973). Two solutions commonly used to impregnate
papers are a mixture of 20 percent potassium hydroxide and 10 percent triethanolamine, and a'
mixture of 20 percent potassium carbonate and 10 percent glycerol. The treated filters are
inserted into filter holders, and air is aspirated through them. An untreated prefilter is
generally recommended to remove particulate matter. Absorbed S0? can be extracted from the
papers and determined colorimetrically by the West-Gaeke method. The alkali must be neutral-
ized exactly to attain the proper acidity prior to color development. Alternatively, the ex-
tract solution may be treated with an oxidizing agent, such as l-LOp to convert sulfite to sul-
fate, followed by a sulfate analysis (Johnson and Atkins, 1975; Forrest and Newman, 1973).
Efficiency of SO, absorption is better than 95 percent under average weather conditions
but decreases rapidly below 25-percent RH and above 80-percent RH. The error may be minimized
by using two filter papers in series (Forrest and Newman, 1973). Elimination of glassware and
reagents during sampling removes the possibility of spillage or breakage during transport.
Sampled papers may be stored conveniently for long periods before being analyzed.
Sulfur dioxide S02 may be sampled on Whatman No. 17 filter papers impregnated with tetra-
chloromercurate TCM solution containing mercuric chloride, sodium chloride, ethyl alcohol, and
glycerol in water (Axelrod and Hansen, 1975). Sampled filters are extracted with TCM, and the
West-Gaeke procedure generally follows. Capacity of the 47-mm filters is 13 mg of SO™, after
which collection efficiency decreases. Samples 'collected at very low RH (10 percent) cannot
be stored more than 1 day before exhibiting losses. Filters sampled at 40 percent RH may be
safely stored for 1 week. No interference is observed for N02 and H2S, but 0- at 175 yg/m
(0.09 ppm) causes negative errors.
A method that uses nondispersive X-ray fluorescence to measure ambient S0? collected on
sodium carbonate-impregnated membrane filters has been developed by Hardin and Shleien (1971).
After collection, the sample filter is irradiated with a one millicurie iron-55 source. The
resulting 2.3 kev sulfur X-rays are counted by a proportional counter with a beryllium window.
A minimum detectable quantity of 30 M9 S09 can be detected by the counter, equivalent to 25
3
(jg/m (0.01 ppm) using a collection time of 1 hour and a sampling rate of 20 liters/minute.
Chlorine gas is collected to a significant degree and since its characteristic X-ray cannot be
distinguished from that of sulfur, it may interfere to produce elevated results if not removed
prior to encountering the treated filter.
3-9
-------
3.2.2.3.5 ChemiTuminescencemethod. The basis for this method is the chemiluminescence
produced when a sulfite solution is oxidized (Stauff and Jaeschke, 1975). Ambient S02 is
absorbed in 50 ml of tetraehloromercurate solution to form the monochlorosulfonatomercurate
ion. Five milliliters of 2 x 10"5 N KMn04 in 10"3 N H£S04 is added. Oxidation of the
absorbed sulfite is accompanied by a chemiluminescence, which is detected by a photomultiplier
tube. The total light yield, measured by a photon counting system, is proportional to the
3 3
oxidizable sulfite. By sampling 1 m of air, 0.5,|jg/m (0.2 ppb) of S02 may be detected with
an error of less than 10 percent.
3.2.2.3.6 Ion exchange chromatographic method. Small et al. (1975) have described an ion
exchange chromatographic system that separates ionic species and effectively neutralizes the
eluant, allowing a conductometric measurement of the ion. A commercial instrument based on
the above system is now available (Dionex Corporation, 1975) for use in trace anion analysis.
In this system, a strong base anion exchanger of low capacity, agglomerated onto a surface-
sulfonated DVB resin, is used as the analytical column. This is followed by a high capacity,
strong acid exchange column that converts the eluant (typically 0.003 M NapC03 + 0.024 M
NaHCO,) into a nonconducting carbonic acid solution, after which the separated ions are
«5
monitored with a high sensitivity, multirange conductivity meter. Although the method is not
totally free from ambiguity, careful selection of eluant and ion chromatographic exclusion
steps can effectively separate ionic species of interest.
Mulik et al. (1978) have developed a method for collection and ion exchange chromato-
graphic analysis of atmospheric SO,. The method uses dilute (0.6 percent) HgOg to collect the
ambient S02. The resultant sulfate ion is analyzed by ion exchange chromatography. When a
prefilter is used in the sampling train to remove aerosol sulfates, there are no apparent
interferences. Collection 'efficiency is approximately 100 percent over the range of the
method, 25 to 1300 ug S02 /m3 (0.01 to 0.5 ppm).
A novel approach for the ion chromatographic determination of atmospheric SQ2 has
recently been suggested by Dasgupta (1981). Sulfur dioxide is collected and stabilized as
hydroxymethanesulfonic acid in a dilute solution of formaldehyde buffered at pH 4 with
potassium acid phthalate (KHP). The sample is analyzed by an ion chromatographic procedure
using KHP as the eluant. The hydroxymethanesulfonate ion elutes as a very sharp peak and the
analysis is facilitated by the fact that both the sample and the eluant have the same ionic
background of KHP, thus minimizing any undesirable phthalate peak or solvent dip.
3.2.2.3.7 Sulfation methods. Sulfation methods are based on the reaction of gaseous S02 in
air with lead dioxide (Pb02) paste to form lead sulfate (PbSO^). They are cumulative methods
for estimating average concentrations over extended periods. In the lead dioxide gauge method
(Department of Scientific and Industrial Research, 1933) and the lead candle method (Wilsdon
and McConnell, 1934), the paste is prepared by mixing Pb02, gum tragacanth, alcohol, and
water. The paste is applied to a piece of cotton gauze wrapped around a cylinder 10 cm in
circumference and 10 cm high. After drying, the cylinder is exposed to the atmosphere in a
3-10
-------
sheltered location. After exposure, the gauge and sulfated lead dioxide are treated with
sodium carbonate solution, and the dissolved sulfate is then determined gravimetrically or
turbidimetrically. Measurements with the method are reported as sulfation rates (mg S0,/100
2 " o
cm /day). In the sulfation plate method (Huey, 1968), a similar paste containing glass filter
fibers is poured into a plastic petri dish 48 mm in diameter. After drying, the plate is
exposed to the atmosphere and analyzed for sulfate.
Sulfation methods have the advantage of being inexpensive, but their accuracy is subject
to many physical and chemical variables and interferents. For example, the rate of sulfate
formation is proportional to atmospheric SO, concentration up to 15-percent conversion of the
lead dioxide (Wilsdon and McDonnell, 1934). Reaction rate increases with temperature and with
humidity. Other factors affecting rate of sulfation are purity of the lead dioxide, its
particle size and shape, wind velocity, and shape of the shelter (Bowden, 1964). Positive
errors are contributed by hydrogen sulfide and sulfate aerosols. Methyl mercaptan is a poten-
tial negative interferent.
Huey et al. (1969) compared the sulfation plate method with the sulfation candle method
at some 250 sampling sites nationwide. A correlation coefficient of 0.95 was obtained, con-
firming that both methods are measuring the same species. The results also indicated that
sulfation plates are 10-percent less reactive than sulfation candles.
Various attempts have been made to correlate sulfation methods with more specific methods
for estimation of S0? concentrations. In 1962, as part of the establishment of the British
National Survey, measurements with the lead dioxide gauge were compared to simultaneous meas-
urements with the hydrogen peroxide method (Warren Spring Laboratory, 1967). The correlation
between 829 pairs of results from 20 sites over 4 years was highly significant, showing that
both methods were predominantly affected by the same pollutant, SOp. The Warren Spring Labor-
atory concluded, however, that there was no generally applicable conversion factor for
relating lead dioxide and hydrogen peroxide results. The conversion from lead dioxide to
hydrogen peroxide reading was not recommended except to give a rough indication of the levels
of concentration concerned.
Stalker et al. (1963) compared the lead dioxide method and the pararosaniline method to
measure SO™ at 123 stations in Nashville, Tennessee. The lead dioxide method was considered
good for estimating mean SO, levels in communities during months with arithmetic mean concen-
3
trations of at least 65 ug/m (0.025 ppm). The reliability of these mean estimates was esti-
mated to be within ± 25 percent. Seasonal effects were noted, however, and the lead peroxide
estimates of SO, (using an average factor of 0.031 for conversion of sulfation rate in mg
2
S03/100 cm /day to S02 concentration in ppm) during the spring season of low SOg levels were
about twice as high as simultaneous 24-hour colorimetric measurements of SOg.
Huey et al. (1969) compared ambient S02 measurements by conductometric, coulometric, and
colorimetric methods with sulfation results. They concluded that sulfation data in mg SO.,/100
2
cm /day could be converted to S02 concentrations in ppm by multiplying by 0.03. They also
3-11
-------
determined that 95 percent of the time this approximation from a single sulfation value will
lie within a factor of about 3 of any single measurement using the other techniques.
3.2.2.3.8 Other manual methods. Other manual methods that have been used for the measurement
of ambient concentrations of S0? include the barium perchlorate-thorin titrimetric method
(Fritz and Yamamura, 1955), the barium sulfate turbidimetric method (Volmer and Frohlich,
1944), the barium chloranilate colorimetric method (Bertolacini and Barney, 1957), and the
silica gel reduction method (Stratmann, 1954).
3.2.3 Automated Methods
3.2.3.1 Sample Collection—In continuous S0_ analyzers, sample collection is an integral part
of the total automated measurement process. The sample line leading from the sample manifold
to the inlet of the analyzer should be constructed of an inert material such as Teflon . The
sample line dimensions (length and internal diameter) should be selected to minimize the resi-
dence time without creating a significant pressure drop between the sample manifold and the
analyzer inlet. The use of an inert particle filter at the inlet of the analyzer should
depend on the analyzer's susceptibility to interference, malfunction, or damage due to PM.
Heavy loading of PM on the filter may lead to erroneous SO, measurements; therefore, it may be
necessary to change the filter frequently.
3.2.3.2 Calibration—The relationship between true pollutant concentration and the response
of a continuous analyzer is best determined by dynamic calibration. In dynamic calibration,
zero air and standard atmospheres containing known concentrations of S0? are introduced into
the analyzer to define the analyzer response over the full measurement range. Dynamic cali-
bration provides evidence that all components of the instrument are functioning properly.
Standard atmospheres required for calibration purposes may be generated using permeation
tubes (Q'Keeffe and Ortman, 1966), (i.e., sealed Teflon tubes containing liquified gas). Gas
diffuses through the walls at a low, constant rate at constant temperature. The gas is then
diluted with zero air at accurately known flow'rates to obtain S0? concentrations over the
required range. Permeation tubes with certified permeation rates are available from the
National Bureau of Standards (NBS) as Standard Reference Materials (SRM's) or from commercial
suppliers. Dynamic calibration may also be carried out using known concentrations of SO^ in
high-pressure cylinders. To ensure stability, they are usually prepared in high concen-
trations and dynamically diluted to the desired level. Traceability of such standards to NBS
SRM's may be established by the gas standard manufacturer or by the user.
Static calibration techniques are possible for several of the continuous S02 analyzers
described below. Static calibration introduces a stimulus to measure instrumental response
under no sample air flow conditions. Typical stimuli are electrical signals, solutions chemi-
cally equivalent to the pollutant, or solutions producing comparable physical effects upon
properties by which the pollutant is detected, such as optical density or electrical conduc-
tivity. Static calibration is a rapid and simple method for checking various components of
the instrument, but does not scrutinize total instrument performance.
3-12
-------
3.2.3,3 Measurement Methods — The principal automated methods (continuous analyzers) for
determining S0? in the air are discussed in this section in chronological order.
3.2.3.3.1 Conductometric analyzers. Conductometric analyzers were the first commercially
available instruments for continuously monitoring atmospheric SO,. They are still used
today. In their operation, air is brought into contact with an absorbing solution, which
dissolves SCL. The ions formed by SOp dissolution increase the conductivity, which is propor-
tional to the concentration. The absorbent may be either deionized water or acidified
hydrogen peroxide solution. When water is used, conductance is increased by formation of the
sulfite and bisulfite ions:
S02 •*• H20 t SQ2-H20 S H+ + HSO~ (3-1)
HSO~ ? H+ + S0~2
Hydrogen peroxide solution oxidizes S0? to form sulfuric acid:
S02 + H202 •> H2S04 t H+ + HS04 (3_2)
Conductance is measured by a pair of inert (platinum) electrodes within the cell. To increase
accuracy, comparison is made to a reference cell, which measures conductance of unused absor-
bent. The response characteristics of Conductometric analyzers include lower detection limits
3
ranging from 0.005 to 0.04 ppm (10 to 100 jjm/m ), lag times (time interval from change in
input concentration to change in output signal) ranging from 5 to 200 seconds, and response
times (time interval from change in input concentration to 90 percent of maximum output
signal) ranging from 1 to 4 minutes (Lawrence Berkeley Laboratory, 1972).
The major disadvantage of conductometric analyzers is their susceptibility to interfer-
ence by any species that either forms or removes ions from solution and changes the
conductivity of the solution. The degree of interference depends on humidity, temperature,
S0? concentration, and the particular instrument. The worst interferents are chlorine, hydro-
chloric acid, and ammonia (Rodes et a!., 1969); nitrogen dioxide and carbon dioxide interfere
to a lesser extent. Airborne particles, especially oceanborne salt aerosols, are potentially
damaging. Several methods have been used to minimize these problems. Chemical scrubbers,
which selectively remove gaseous interferents, have been incorporated into some Conductometric
analyzers. Particle filters have also been employed.
3.2.3.3.2 Colon' metric analyzers. Colorimetric analyzers are based upon reaction of SOg with
solutions of organic dyes to form colored species. Optical absorbance of the resulting solu-
tion, measured spectrophotometrically, is within limits linearly proportional to the concen-
tration of the colored species in accordance with Beer's Law. Most instruments use modified
versions of the manual pararosaniline method developed by West and Gaeke (1956). Automation
of the West-Gaeke method per s>e does not ensure a practical continuous monitoring instrument
since some solutions require daily preparation.
3-13
-------
The response characteristics for some commercially available instruments include lower
3
detection limits ranging from 0.002 to 0.01 ppm (5 to 25 jjm/m ), lag times ranging from 0.6 to
25 minutes, and response times ranging from 5 to 30 minutes (Lawrence Berkeley Laboratory,
1972). Advantages of these instruments include good sensitivity and, with proper control,
good specificity. Interferences by nitrogen oxides may be controlled by using a reagent
containing sulfamic acid. Heavy metals may be complexed with EDTA in the scrubbing solution
or with phosphoric acid in the dye solution. Ozone interference may be controlled by use of a
delay coil downstream from the absorber to allow time for ozone to decay, but this results in
longer lag and response times. Major disadvantages of these instruments are the need for re-
agent and pump tubing replacement and frequent recalibration.
3,2.3.3.3 Coulometric andamperpmetric analyzers. Coulometric analyzers are based on the
reaction of S02 with a halogen, formed directly by electrolysis of a halide solution. The
current necessary to replace the depleted halogen is proportional to the amount of SO^
absorbed in the solution, and hence to the SO, concentration in the air.
In one common coulometric system, an inner chamber, into which air is introduced, is con-
tiguous with an outer chamber (Treon and Crutchfield, 1942). Both contain a solution of
potassium bromide and bromine in dilute sulfuric acid. Potential difference between chambers,
relative to a reference potential, is measured by the reference electrodes. As absorbed SO,
reduces the Br2 concentration in the inner chamber, the amplifier produces a current to
restore the Br, content in the inner chamber until the potential difference is again zero. In
a second system, the change in halogen concentration is detected as a current change rather
than a potential difference. The cell is filled with a potassium iodide solution, buffered to
pH 7. At the platinum anode, a constant current source continuously generates iodine, which
is subsequently reduced at the cathode. An equilibrium concentration of iodine is estab-
lished, and no current is generated at an activated-carbon bipolar reference electrode,
connected in parallel. Reaction with SO,, decreases the equilibrium concentration of iodine,
which cannot transport the charge generated by the constant-current source. Part of the
current is diverted through the reference electrode; this flow is proportional to the SO,
concentration in the air sample. The response characteristics of modern coulometric analyzers
include lower detection limits ranging from 0.002 to 0.05 ppm (5 to 130 jjm/m ), lag times
ranging from 2 to 120 seconds, and response times ranging from 2 to 5 minutes (Lawrence
Berkeley Laboratory, 1972).
Interferent species are those able to oxidize halides, reduce halogens, or complex with
either. They consist primarily of sulfur compounds (hydrogen sulfide, mercaptans, and organic
sulfides and disulfides) with sensitivities comparable to or greater than that of SOg. Other
potential interferents, at lower sensitivities, are ozone, nitrogen oxides, chlorine, olefinic
hydrocarbons, aldehydes, benzene, chloroform, other nitrogen- or halogen-containing compounds,
and other hydrocarbons (DeVeer et a!., 1969; Schulze, 1966; Thoen et a!., 1968; Washburn and
Austin, 1952). Interferences can be minimized by selective filters, which are sometimes built
3-14
-------
Into the instrument or offered as optional features. For example, a heated silver gauze
filter js reported to remove hydrogen sulfide, ozone, chlorine, nitrogen oxides, carbon disul-
fide, -ethylene; aldehydes, benzene, and chloroform, but will not remove mercaptans (Philips
Electronic Instruments, undated).
Minimal maintenance is the major advantage of a coulometric analyzer (reagent may need
only monthly replacement; electrodes may require annual cleaning). Also, reagent consumption
is negligible because of halide regeneration, and evaporated water is replaced by condensation
from air or from a reservoir.
3.2.3.3.4 Flame photometric analyzers. The flame photometric detector (FPD) is based on the
—. ^
measurement of the band emission of excited S? molecules during passage of sulfur-containing
compounds through a hydrogen-rich (reducing) flame. The emitted light passes through a narrow-
*
pass optical filter, "which isolates the 394 nm $2 band, and is detected by a photomultiplier
tube (PMT). PMT output is proportional to the square of the sulfur concentration; hence, an
electronic system to "linearize" output is a desirable feature. Application of the FPD to the
detection of SQ~ was first made by Crider (1965) and analyzers using FPD have been widely
accepted for ambient S0? monitoring. The response characteristics of continuous flame photo-
metric SO, analyzers include lower detection limits ranging from 0.002 to 0.010 ppm (5 to"25
3
fjm/m ), lag times ranging from 1 to 5 seconds, and response times ranging from 10 to 30
seconds (Lawrence Berkeley Laboratory, 1972).
Although the FPD is insensitive to nonsulfur species, it will detect sulfur compounds
other than S0». Particle filters will remove troublesome aerosol sulfates and selective
filters may be used to reduce1 interference from other gaseous sulfur compounds (e.g., an H?S
filter is used on most commercial instruments). Interference by C02 can be minimized by main-
taining ambient levels of C0? in the calibration and sample matrices.
Gas chromatographs with flame photometric detectors (GC-FPD) are also available commer-
cially. GC-FPD can separate individual sulfur compounds and measure them individually
(Stevens et al,, 1971). The temporal resolution of GC-FPD data, however, is limited by the
chromatographic elution time of SO- and other gaseous sulfur compounds.
Disadvantages of FPD systems include the need for a source of compressed hydr^g'-<\ and
sensitivity to all sulfur compounds. Advantages of FPD systems include low maintener e, gocd
sensitivity, very fast response, and good selectivity for sulfur compounds. No reagents are
necessary, other than compressed hydrogen.
3.2.3.3.5 Second-derivative spectrometric analyzers. The second-derivative spectrometer
processes the transmission-versus-wavelength function of a spectrum to produce a signal
proportional to the second derivative of this function (Hager and Anderson, 1970). The signal
amplitude is proportional to the concentration of the gas in the absorption path. These
instruments center on the shape characteristics rather than basic intensity changes of
molecular band spectral absorption. The slope and curvature characteristics are often large,
specific, and independent of intensity. Because these shape characteristics are large but
specific to individual compounds, resolution of component gases is possible.
3-15
-------
In the operation of a second-derivative spectrometer, radiant energy from a UV or visible
source is directed into a monochromator, where it is dispersed by a grating to provide mono-
chromatic light to the sample cell. The wavelength of this light is modulated with respect to
time in a sinusoidal fashion by an oscillating entrance slit. The angular position of the
grating sets the center wavelength (299.5 nm) coming out of the monochromator into the multi-
pass cell. The sample is continuously drawn through the cell by a pump. Output from the
4
photomultiplier tube is electronically analyzed to develop the second derivative of the
absorbance.
Sensitivity is enhanced because the output is an AC signal of known wavelength and phase,
adaptable to high-gain electronic amplification. Uniqueness of the curvature of a given
molecular band enables this type of instrument to be highly specific. A theoretical
assessment by Ratzlaff and Natusch (1977) indicates that precision may be a problem with
spectrometric techniques of this type. Measurements are independent of sample flowrate, but
relatively high flowrates (4 liters/minute) are necessary to achieve reasonable response
times. The response characteristics for one commercially available instrument include a lower
3
detection limit of 0.01 ppm (25 um/m ), lag time of I minute, and response time of 8 minutes
(U.S. Environmental Protection Agency, 1979a).
3.2.3.3.6 Fluprescence analyzers. Fluorescence analyzers are based on detection of the
characteristic fluorescence released by the sulfur dioxide molecule when it is irradiated by
ultraviolet light (Okabe et a!,, 1973). This fluorescent light is also in the ultraviolet
region of the spectrum, but at a different wavelength than the incident radiation. Wave-
lengths between 190 and 230 nm are used for excitation and the fluorescent wavelengths usually
monitored are between 300 and 400 nm. In this region of the spectrum, there is relatively
little quenching of the fluorescence by other molecules occurring in ambient air. The light
is detected by a photomultiplier tube that, through the use of electronics, produces a voltage
proportional to the light intensity and SO, concentration. The fluorescent light reaching the
photomultiplier tube is usually modulated to facilitate the high degree of amplification
necessary. Some analyzers mechanically "chop" the incident irradiation before it enters the
sample cell. Other instruments electronically pulse the incident light source at a constant
rate. The response characteristics of fluorescence analyzers include lower detection limits
of 0.005 ppm, lag times of about 30 seconds, and response times of about 5 minutes (U.S.
Environmental Protection Agency, 1979a).
Potential interferences to the fluorescence technique include any species that either
quenches or exhibits fluorescence. Both water vapor and oxygen strongly quench the fluores-
cence of SOp at some wavelengths. Water vapor can be removed by a dryer within the instrument
or the water interference can be minimized by careful selection of the incident radiation
wavelength. The effect of oxygen quenching can be minimized by maintaining identical oxygen
concentrations in the calibration and sample matrices.
3-16
-------
Aromatic hydrocarbons such as naphthalene exhibit strong fluorescence in the same spec-
tral regions as S0? and are major interferents. These aromatics must be removed from the
sample gas stream by an appropriate scrubber upstream of the sample cell. The scrubbers may
operate at ambient or elevated temperature. Certain elevated-temperature scrubbers, however,
have the potential for converting ambient FLS (which normally does not interfere with the
fluorescence technique) into SOy- In these cases, the hydrocarbon scrubber must be preceded
by a scrubber for H?S.
3,2.3,3.7 Other automated methods. Other automated methods (continuous analyzers) that have
been used for the measurement of ambient concentrations of SO, include: voltammetry (Chand
and Marcote, 1971); correlation spectroscopy (Barringer Research, Ltd., 1969; Moffat et a!.,
1971); and differential lidar (Johnson et a!., 1973).
3.2.3.4 EPA Designated Eguivalent Methods—-Under provisions of EPA's "Ambient Air Monitoring
Reference and Equivalent Methods" regulations (U.S. Environmental Protection Agency, 1979b),
several commercial continuous analyzers have been designated as equivalent methods for deter-
mining compliance with National Ambient Air Quality Standards for SO^. These analyzers have
undergone the required testing and meet EPA's performance specifications for automated
methods, summarized in Table 3-2. A list of S0? analyzers designated from the promulgation of
the regulations in 1975 to December 31, 1980, is given in Table 3-3. Information on designa-
tion of these analyzers as equivalent methods may be obtained by writing the Environmental
Monitoring Systems Laboratory, Methods Standardization Branch (MD-77), U.S. Environmental Pro-
tection Agency, Research Triangle Park, North Carolina 27711.
Review of performance data submitted in support of the designations listed in Table 3-3
indicates that these modern analyzers exhibit performance better than that specified in Table
3-2. For the analyzers tested, noise levels were typically 3 ppb or less. The zero drift re-
sults (12- and 24-hour) were all less than 5 ppb and typically less than 3 ppb. The span
drift results (at 20 and 80 percent of the full scale range of 0 to 0.5 ppm) were all less
than 5 percent and typically 2 to 3 percent. The precision results (at 20 and 80 percent of
the full scale range of 0 to 0.5 ppm) indicate a typical precision of 1 to 2 ppb. Lag times
were typically less than 1 minute. Response times (rise and fall times) for the various types
of analyzers were typically as follows: flame photometric, 1 minute or less; fluorescence, 5
minutes; coulometric, 3 minutes; conductometric, 0.5 minute; second-derivative spectrometric,
8 minutes. For analyzers of the same type (e.g., flame photometric), interference test
results for a given potential interferent were somewhat variable. The concentration of S0?
during the tests was 0.14 ppm and the interferent concentrations were as indicated in Table
3-4. The interference equivalent for each interferent must not exceed ±20 ppb and the total
interference equivalent (sum of the absolute values of the individual interference equiva-
lents) must not exceed 60 ppb. Interference equivalents of 5 ppb or less were obtained in
each case except for the following: flame photometric-negative CO- interference equivalents
3-17
-------
TABLE 3-2. PERFORMANCE SPECIFICATIONS FOR EPA EQUIVALENT METHODS FOR S02
(CONTINUOUS ANALYZERS)3
Performance parameter
Range
Noise
Lower detectable limit
Interference equivalent
Each interferent
Total interferent
Zero drift, 12- and 24-hour
Span drift, 24- hour
20 percent of upper range limit
80 percent of upper range limit
Lag time
Rise time
Fall time
Precision
20 percent of upper range limit
80 percent of upper range limit
Units
ppm
ppm
ppm
ppm
ppm
ppm
percent
percent
minutes
minutes
minutes
ppm
ppm
Specification
0-0.5
0.005
0.01
±0.02
0.06
±0.02
±20.0
±5.0
20
15
15
0.01
0.015
Note: 1 ppm S02 = 2620 \jtg/m .
aSource: U.S. Environmental Protection Agency (1979b).
3-18
-------
TABLE 3-3. LIST OF EPA DESIGNATED EQUIVALENT METHODS FOR SO
(CONTINUOUS ANALYZERS)
Designation
number
EQSA-1275-005
EQSA-1275-006
EQSA-0276-009
EQSA-0678-010
EQSA-0876-011
EQSA-0876-013
EQSA-0877-024
EQSA-0678-029
EQSA-1078-030
EQSA-1078-032
EQSA-0779-039
EQSA-0580-046
EQSA- 1280-049
Manufacturer
Lear Siegler
Meloy
Thermo Electron
Philips
Philips
Monitor Labs
ASARCO
Beckman
Bendix
Meloy
Monitor Labs
Meloy
Lear Siegler
Model
SM1000
SA185-2A
43
PW9755
PW9700
8450
500,600
953
8303
SA285E
8850
SA700
AM2020
Measurement principle
Second-derivative spectrometric
Flame photometric
Fluorescence
Coulometric
Coulometric
•B
Flame photometric
Conductometric
Fluorescence
Flame photometric
Flame photometric
Fluorescence
Fluorescence
Second-derivative spectrometric
The four digits in the middle of each number indicate the month and
year of designation.
3-19
-------
TABLE 3-4. INTERFERED TEST CONCENTRATIONS {PARTS PER MILLION)® USED IN THE TESTING
OF EPA EQUIVALENT METHODS FOR SO
Analyzer type
Flame photometric (FPD)
Gas chroaatography-FPD
Spectrophotonetric-wet
chemical (pararosaniline
reaction)
Electrochemical
Conductivity
Spectrophotometric-gas phase
Hydro-
chloric Am-
acid morn a
,.
—
0,2 0.1C
0.2 O.lc
0.2 O.lc
—
Hy-
drogen
sulfide
0.1
0.1
0.1
0.1
_.
--
Sulfur
dioxide
0.14d
0.14d
0.14d
0.14d
0.14d
0.14d
Nitro-
gen Nitric
dioxide oxide
..
—
0.5
0.5 0.5
0.5
0.5 0.5
Carbon Eth-
di oxide ylene
780
750
750
0.2
750
-.
Carbon
m- Water mon-
Ozone Xylene vapor oxide
20,000C 50
20,000C 50
0.5
0.5 -- 20,000C
--
0.5 0.2
Concentrations of interferent" listed must be prepared and controlled to ± 10 percent of the stated value.
Analyzer types not listed will be considered by the EPA Administrator as special cases.
cDo not mix with pollutant.
Concentration of pollutant used for test. These pollutant concentrations must be prepared to ± 10 percent of the stated value.
eSource: U.S. Environmental Protection Agency (1979b).
-------
of about 10 ppb were typical; coulometric-positive 0,, interference equivalents" of about 8 ppb
were typical.
As part of required equivalency testing by manufacturers, all continuous S0? analyzers
designated by EPA as equivalent methods have demonstrated a consistent relationship with the
reference method. A consistent relationship is demonstrated when the differences between (1)
measurements made by the test analyzer, and (2) measurements made by the reference method are
less than or equal to the allowable discrepancy specifications prescribed in the equivalency
regulations, when both methods simultaneously measure S0? concentrations in a real atmosphere.
All the equivalent methods listed in Table 3-3 have demonstrated this consistent relationship
with the reference method and the observed differences between simultaneous measurements were
generally well within the required specifications.
A comparison study using EPA designated equivalent methods for SCU was recently conducted
by EPA in an urban/industrial/commercial area of Durham, North Carolina (U.S. Environmental
Protection Agency, 1979a). Eight continuous S0? analyzers were compared over 150 days under
more or less typical air monitoring conditions. During the study, the analyzers
simultaneously measured ambient air sampled from a common manifold. The ambient sample was
occasionally augmented with artificially generated pollutant to allow for analyzer comparisons
at higher concentrations. A statistical comparison of hourly averages for each test analyzer
with the average of the hourly averages (for corresponding hours) from the other test
analyzers is presented in Table 3-5. Each test analyzer is identified in the table by manu-
facturer, model number, and measurement principle. The data clearly indicate that these con-
tinuous SCL analyzers are capable of excellent performance (high correlation with one another,
small mean differences).
3.2.4 Summary
Methods for measuring of SO- can be classified as: (1) manual methods, which involve
collection of the sample over a specified time period and subsequent analysis by a variety of
analytical techniques, or (2) automated methods, in which sample collection and analysis are
performed continuously and automatically.
In the. commonly used manual methods, the techniques used for the analysis of the
collected sample are based on colorimetric, titrimetric, turbidimetric, gravimetric, X-ray
fluorescent, chemiluminescent, and ion exchange chromatographic measurement principles.
The most widely used manual method for the determination of atmospheric SO, is the
pararosaniline method developed by West and Gaeke. An improved version of this colorimetric
method, adopted as the EPA reference method in 1971, is capable of measuring ambient S0? con-
3
centrations as low as 25 yg/m (0.01 ppm), with sampling times ranging from 30 minutes to 24
hours. The method has acceptable specificity for S02> but samples collected in tetrachloro-
mercurate (II) are subject to a temperature-dependent decay, which can result in an under-
estimation of the ambient SO, concentration. Temperature control during sample collection,
shipment, and storage effectively minimizes this decay problem. A variation of the
3-21
-------
TABLE 3-5. COMPARISON OF EPA DESIGNATED EQUIVALENT METHODS FOR SO, (CONTINUOUS ANALYZERS)0
Analyzer
Measurement
principle
Correlation
coefficient
Mean
difference
ppb
Std. dev.
of diff.
ppb
Max. abs.
diff.
ppb
No. of abs.
diff.
>20 ppb
No. of data
pairs
Meloy Monitor Labs
SA185-2A 8450
Flame Flame
photometry photometry
0.999 0.999
-3.695 -0.006
3.925 4.555
19.3 19.2
0 0
3302 3186
Heloy
SA285E
Flame
photometry
0.999
-0.251
3.243
15.5
0
3306
Thermo Electron Beckntan Lear Siegler
43 953 SH1000
Second-
derivative
Fluorescence Fluorescence spectronetry
0.997 0.998 0.936
-0.177 5.108 4.924
8.300 6.901 20.712-
29.9 25.4 100.9
49 21 427
2170 1594 1820
Philips
PW9755
Couloietry
0.998
5.775
4.631
25.6
13
3070
Bendix
8303
Flame
photometry
0.998
-3.278
4.392
21.0
1
1984
a8etween subject analyzer and average of other test analyzers (for corresponding hours).
Source: U.S. Environmental Protection Agency (1979a).
Note: 1 ppb S02 = 2.62 vg/m3.
-------
pararosaniline method uses a buffered formaldehyde solution for sample collection and is
reported to be less susceptible to the temperature-dependent decay problem.
In Great Britain, a titrimetric method based on collection of SOp in dilute hLOp followed
by titration of the resultant HpSO. with standard alkali is the standard method and is used
extensively. Although simple to perform, the method requires long sampling times (24 hours)
and is subject to interference from atmospheric acids and bases. Additional sources of error
with this method include evaporation of reagent during sampling, titration errors, and alka-
line contamination of glassware.
Methods that collect SO, with alkali-impregnated filter papers for subsequent analysis as
sulfite or sulfate by a variety of techniques have been developed. Most of these methods
involve an extraction step prior to analysis, although nondispersive X-ray fluorescence has
been used for the direct measurement of S0? collected on sodium carbonate-impregnated membrane
filters.
Two of the most sensitive methods available use measurement principles based on chemi-
luminescence and ion exchange chromatography. In the chemiluminescence method, SOp is
absorbed in a tetrachloromercurate solution and subsequently oxidized with potassium permanga-
nate. The oxidation of the absorbed SO, is accompanied by a chemi luminescence that is
detected by a photomultiplier tube. One method, using ion exchange chromatography to
determine ambient levels of SQp that have been absorbed into dilute HpQp and oxidized to
sulfate, has been developed. Another ion chromatographic approach, using a buffered formalde-
hyde absorbing reagent, has been reported.
Sulfation methods, based on reaction of SQp with lead dioxide paste to form lead sulfate,
have commonly been used to estimate ambient SOp concentration over extended periods. The
accuracy of sulfation methods is subject to many physical and chemical variables and interfer-
2
ents. Sulfation rate (mg SQ.,/100 cm /day) is commonly converted to a rough estimate of SO,
concentration (ppm) by multiplying the sulfation rate by the factor 0.03.
Automated methods for measurement of ambient levels of SO, have gained widespread use in
the air-monitoring community. Some of the earliest continuous SO- analyzers were based on
conductivity and coulometry. These first-generation analyzers were subject to interference by
a wide variety of substances present in typical ambient atmospheres. However, more recent
commercially available analyzers using these measurement principles exhibit improved speci-
ficity for SOp through the incorporation of sophisticated chemical and physical scrubbers.
Early continuous colorimetric analyzers using West-Gaeke type reagents and having good sensi-
tivity and acceptable specificity for SO, were fraught with various mechanical problems,
required frequent calibration, and never gained widespread acceptance.
Continuous SOp analyzers using the techniques of flame photometric detection, fluor-
escence, and second-derivative spectrometry have been developed over the past 10 years and are
commercially available from a number of air-monitoring instrumentation companies. Flame
photometric detection of ambient SO, is based on measurement of the band emission of excited
3-23
-------
*
$2 molecules formed from sulfur species in a hydrogen-rich flame. The FPD analyzers exhibit
high sensitivity and fast response, but must be used with selective scrubbers or coupled with
gas chromatographs when high specificity is required.
Fluorescence analyzers are based on detection of the characteristic fluorescence of the
SOg molecule when it is irradiated by UV light. These analyzers have acceptable sensitivity
and response times, are insensitive to sample flowrate, and require no support gases. They
are subject to interference by water vapor (due to quenching effects) and certain aromatic
hydrocarbons, and therefore must incorporate some means to minimize these species or their
effects.
Second-derivative spectrometry is a highly specific technique for measuring atmospheric
S0o> and continuous analyzers based on this principle are available commercially. The
analyzers are insensitive to sample flowrate and require no support gases, but relatively high
sample flowrates are required to achieve reasonable response times. Excessive electronic
noise and inherent lack of precision can be problems with these analyzers.
Continuous analyzers based on many of the above measurement principles (conductivity,
coulometry, flame photometry, fluorescence, and second-derivative spectrometry) have been
designated by EPA as equivalent methods for the measurement of SQ~ in the atmosphere. Testing
of these analyzers by the manufacturers prior to designation demonstrated adequate performance
for use when an EPA reference or equivalent method is desired or required. EPA testing of
these methods verified their performance and has also demonstrated excellent comparability
among these designated methods under typical monitoring conditions.
3.3 PARTICULATE MATTER
3.3.1 Introduction
As described in Chapter 2, particulate matter (PM) suspended in ambient air presents a
complex multiphase system consisting of a spectrum of aerodynamic particle sizes from below
0.01 micrometer (urn) up to 100 urn and larger. Fine particles (below ~ 2 pm) tend to remain
suspended in air unless removed by external processes such as rainfall. Coarse particles
(above ~ 2 urn) have appreciable settling velocities and tend to settle unless kept in
suspension by high windspeeds-or turbulence. The sources and characteristics of the particles
in both size ranges are generally quite different, and depending on the objectives of the sam-
pling, measurements are often made that consider only a selected size fraction. Samplers used
to identify fine and coarse particle fractions typically are designed to have inlet and sub-
stage cutpoints that are as sharp as possible. Samplers used to simulate the deposition
pattern of particles in the respiratory system have well-defined but more gradual cutpoints.
Lippmann (1970) summarized samplers and deposition patterns in the 1-10 (jm range proposed by
several organizations. As Figure 3-1 shows these include models of the American Conference of
Governmental Industrial Hygienists (ACGIH), British Medical Research Council (BMRC), and U.S.
Atomic Energy Commission (called the "Los Alamos" curve) (Lippmann, 1970). Miller et al.
(1979) proposed a sampler cutpoint of 15 pm related to respiratory system deposition but did
3-24
-------
100
UJ
oc
O
O
O
p
cc
h-
Ul
Z
IU
a.
40
20
02468
DIAMETER UNIT DENSITY SPHERE, pm
Figure 3-1. Respiratory deposition models used as patterns for
sampler outpoints.
Source: Lippmann (1970).
3-25
-------
not recommend a desirable cutpoint sharpness. Particle deposition in the respiratory system
is discussed in more detail in Chapter 11.
The aerodynamic diameter is one of the most important physical parameters when consider-
ing particle deposition in the atmosphere or the human respiratory system. Suspended
particles are rarely spherical, and characterizing particle size with a single physical
dimension is often difficult. Aerodynamic diameter is not a direct measurement of size but is
the equivalent diameter of a spherical particle of unit density that would settle at the same
rate. This definition inherently considers such factors as particle density and shape without
requiring their direct measurement. Aerodynamic diameters are used in this chapter unless
stated otherwise. Sampling methods using collection or separation techniques based on the
inertia or settling properties of particles are classified according to the aerodynamic size.
In general, all sampling methods that draw the particles into an inlet or opening perform an
aerodynamic size segregation. However, particles with unusual geometries, such as long fibers
may not be separated as effectively as more spherical particles, since the orientation of the
fiber at the point of separation has a substantial impact on the effective diameter.
As Fuchs (1964) pointed out, particle size distributions can be examined in several ways.
Separate distributions of volume, surface area, and number of particles, as shown in Figure
3-2, can be measured to provide detailed information especially useful in studying particle
transport and transformation. The particle size distribution by mass is perhaps the most
important characteristic of an aerosol to consider with the majority of current sampling
methods. A mathematical integration of the mass distribution function over the effective
aerodynamic collection range of the sampler directly provides the total mass collected per
unit volume of air sampled. This information can be obtained indirectly from volume, surface
area, or number distribution, but an estimate of the average particle density must be included
in the calculations.
The most common aerosol measurement made in conjunction with health and welfare effect
studies is the mass concentration measurement. Direct measurement of the mass concentration is
made by collecting particles on a substrate, such as a filter, gravimetrically determining the
mass of the particles, and dividing the mass by the volume of air sampled. Ideally, the par-
ticles reaching the ^substrate have been segregated by an efficient sampling mechanism that
provides a defined portion of the ambient size distribution of particles to be collected.
Airborne [Particles (National Research Council, 1979) stated that ". . . integral methods used
are always sensitive to the modification of the size distribution by the sampling inlets and
transport lines used in the technique." This reference notes that Lundgren (1973), using
special samplers to produce mass size distributions as shown in Figure 3-3, showed that most
mass sampling methods truncate the true ambient particle distribution, thereby giving
concentrations less than those actually existing. If these less-than-perfect samplers
operated consistently in all conditions, the mass collected would always be a consistent
proportion of the true ambient size distribution, assuming a constant distribution function.
3-26
-------
o
*"
X
a
O
K
1U
eo
I 1.2
— 1.0
h-Ji 0.8
HgO.6
3
ui
— < 0.4
u.
cc
— 0.2
o
x
~"Q.
Ul
1
l
——. NUMBER
_ —_ SURFACE
— »— VOUIME
A
0.01
0.1 1
PARTICLE DIAMETER, urn
10
Figure 3-2. Plots illustrating the relationship of particle number, surface area, and volume
distribution as a function of particle size.
Source: Whitby (1975).
3-27
-------
180 —
165
150
135
m
1 «0
a.
Oft 105
a
3 go
I
^ 75
60 —
45
30
15
0
0.01
MASS CONCENTRATION £lg/m3
Q—.—i,,.. 192»
. _ __ 93
— 71
0_.._ 60
«<_..._ 26
•FUGITIVE DUST EPISODE
0.1
1.0
10
100
PARTICLE DIAMETER (Dp),/um
Figure 3-3. Typical ambient mass distribution data for particles up to 200 ^m.
Source: Lundgren (1973).
1000
3-28
-------
•However, it has been determined that many sampler inlets are substantially affected by wind-
speed and, in some cases, wind direction. Some of the commonly used PM samplers employing
direct mass measurement techniques include the TSP Hi-Volume Sampler, the dichotomous sampler,
cascade impactors, and cyclone samplers.
Mass concentrations of particles can be estimated using methods that do not employ direct
weighing. These indirect measurements use analytical techniques other than direct weighing
for assessing integral properties (other than mass) of particles. Typically, an empirical
relationship with a gravimetric method is developed and pseudo-mass concentrations reported in
lieu of the integral property measurement. Seta-ray attenuation by the particles on a filter
and optical reflectance of the darkening of a filter by collected particles are examples of
indirect measurement techniques. Li situ methods, which examine particles still suspended in
the airstream, include a wide variety of techniques such as the light-scattering measurements
of the integrating nephelometer and the size classification capability of optical particle
counters.
Analytical measurement of the chemical composition of particles can be strongly in-
fluenced by the sampling method. Surface measurements, such as X-ray fluorescence spectro-
scopy, require a filter that retains particles on the surface rather than allowing penetration
as into a fiber filter. The composition and impurities in the collection substrate can be
critical, especially in the analysis of trace elements. Selected substrates can also interact
with ambient gases to produce artifact particulate matter. Section 3.3.4 contains descrip-
tions of the most common analytical methods and Section 3.3.5. briefly discusses particle
morphology measurements by microscopic examination.
Measurement technology for aerosols has advanced significantly in the past 10 years,
especially in the area of size-specific measurements for larger particles. Before the advent
of specially designed wind tunnels into which specific aerosol sizes and types can be
injected, determination of sampling accuracy (effectiveness) under conditions similar to field
sampling had rarely been attempted. For these tests, effectiveness is measured as the percent
of the particle mass reaching the collection substrate of the sampler compared to results
obtained by isokinetic sampling in the wind tunnel.
It has been recognized (Cermak, 1974) that the length and time scales of fluid motion in
a wind tunnel can be vastly different from those existing in the atmospheric boundary layer.
While ultimate agreement between the performance characteristics of an inlet for large parti-
cles tested in a wind tunnel and in the atmosphere must be substantiated, the use of a wind
tunnel provides a tool to characterize inlets over a wide range of test parameters under con-
trolled conditions. Tests conducted in the atmosphere will only provide benchmarks for com-
parison because of the lack of consistency of test conditions. The comparability of artifi-
cial and real particles and the turbulent macroscale on the motion of particles must be
addressed when considering wind tunnel simulation. While no models exist for predicting the
behavior of large particles in the atmosphere, such models are not necessarily required to
3-29
-------
substantiate the validity of wind tunnel data. Wedding et al. (1980) developed a rigorous,
analytically based model for predicting accurately the results of wind tunnel data for the
Sierra dichotomous sampler inlet. The turbulence of the flow approaching the sampler is not
considered by this model, which describes the behavior of particles after entering the inlet.
In both the wind tunnel and the atmosphere, the sizes of the turbulent eddies are usually
larger than the characteristic dimension of the inlet opening and are much greater than the
particles' stopping distance. Even though atmospheric eddies are much larger than those in
the wind tunnel, the inlet performances for the particle types tested could be expected to be
similar in either regime. Wind tunnel test results, therefore, provide controlled testing
environments for intercomparison of inlets, but to date have not yet been used to predict the
mass concentrations realized in atmospheric situations.
Researchers such as McFarland and Ortiz (1979), Wedding et< al. (1980), and Liu and Pui
(1980) have designed and built such test facilities for -characterizing aerosol samplers. From
these tests, it is now recognized that ambient windspeed and direction can have a profound
effect on particle sizes reaching the point of collection or measurement within the sampler.
Without knowledge of these and related sampler characteristics, an accurate interpretation of
the aerometric data is impossible. This section describes important characteristics for
commonly used sampler types, so that the usefulness of aerometric data discussed in subsequent
chapters can be assessed.
3.3.2 Gravimetric PM MassMeasurements
Techniques that employ direct gravimetric weighing of particles collected on a substrate
are discussed separately here. Sampling techniques that fall into this category are
extractive rather than i_n situ, in that the particles are removed from the airstream for
subsequent analysis. Typically the ambient air is drawn into an inlet, transported to the
collection substrate, often after one or more stages of particle size separation, and then
deposited on a substrate by either filtration or impaction. In addition to the effect of
internal separation stages, the particle size range collected by a filtration sampler depends
on other parameters such as inlet geometry, internal wall losses, and the efficiency of the
filter material. The high-volume sampler defined in the previous Air Quality Criteria for
Participate Matter (National Air Pollution Control Administration, 1969) and in the reference
method for TSP (U.S. Environmental Protection Agency, 1979c) was considered to have captured
all sizes of particles up to 100 inn (aerodynamic diameter). However, recent sampler
characterization testing by Stevens and Dzufray (1975), Wedding et al. (1977), and McFarland
and Rodes (1979) has shown that the gable roof used as an inlet and weather shield actually
provides a D™ (the particle size at which 50 percent of the particle mass is passed on to the
filter) of only 25 to 50 ym, depending on the windspeed. As shown by the data of McFarland
and Rodes in Figure 3-4, the sampling effectiveness of the hi-vol sampler for large particles
is substantially affected by ambient windspeed. Lundgren (1973) has examined the mass
distribution of large particles up to 200 jjm in the atmosphere as shown in Figure 3-3.
3-30
-------
2
i.
w
u
100
80
60
40
20
WIND
SPEED,
km/hr
50
2 45
8 44
24 30
\
I I t I I I
I
I I I
6 8 10 20 40
AERODYNAMIC PARTICLE DIAMETER,/urn
60
Figure 3-4. Sampling effectiveness of a Hi-Vol sampler as a function of wind
speed. Sampler rotated at 1 RPM and operating at 1A m3/min.
Source: McFarland and Rodes (1979).
3-31
-------
Comparison of hi-vol sampler collection efficiency data in Figure 3-4 with these particle size
distributions shows that the hi-vol sampler does not provide a true measure of the large
particles in the atmosphere. Because particle mass increases as a cubic function of diameter
for particles with constant density, the sampling of large particles must be treated carefully
when considering a broad size distribution.
Size-specific sampler inlets designed to limit the particles collected to a certain size
range are a relatively new technology for particles larger than 10 urn. Since these larger
particles are difficult to transport in quantity, a sharp cutoff for large particles is not
easily obtained except at high sampler flowrates using multiple stages of separation. The ef-
ficiency of a single stage inlet designed in 1977 (Stevens and Dzubay, 1978; see Appendix
q
Figure 3A-1) to provide a 15 urn cutoff for a low flowrate sampler operating at 1.0 m /hr,
(Wedding et al. 1977), is shown in Figure 3-5. Note that the D™ for this inlet is very
windspeed dependent and varies from 9 to 22 urn. More advanced" inlets (see Appendix Figure
3A-2 for diagram) for this flowrate have been designed by Wedding (1980) and Liu et al. (1980)
and have reduced windspeed sensitivities and sharper cutpoints, as shown in Figures 3-6 and
3-7, respectively. The geometric standard deviations of the sampling effectiveness curves (a
measure commonly reported in the literature of the sharpness of the size cut-off and denoted
as o ) for these inlets vary from approximately 1.2 to 1.5 as compared to an ideal
step-function inlet with a cr of 1.0. The use of a should be interpreted as only an estimate
y * y
of the slope, since it implies that the effectiveness versus particle size relationship is
log-normal, which is rarely the case. Values of o typically are determined either by
dividing the particle diameter associated with an effectiveness of 84 percent into the D™ or
by dividing the D™ into the diameter corresponding to 16-percent effectiveness.
After particles pass through the sampler inlet they caft be lost from the flowstream
before collection or measurement by attraction to or impaction on the internal surfaces of the
sampler. Minimizing internal loss, especially for larger particles, requires careful design
of the sample transport system geometry as well as consideration of factors such as surface
charge dissipation. Wedding et al. (1977) reported internal wall losses in a prototype
size-specific sampler to exceed 40 percent for particles greater than 15 urn. Loo et al.
(1979) reported that recent improvements in the dichotomous sampler reduced internal particle
losses to less than a few percent.'
3.3.2.1 Filtration Samplers—The most commonly used method for direct gravimetric measurement
involves collection of the particles suspended in a known volume of ambient air on a
preweighed filter. The size distribution of particles reaching the filter are affected by the
characteristics of the inlet, the transport system, and the separation stages, operating at
the sampler flowrate. The performance of a sampler is also substantially affected by the
filter characteristics. The efficiency of the filter medium used can influence the total mass
collected if very small particles are not retained on the filter, or if very large particles
bounce from the filter to subsequent stages. The collection efficiencies over a range of
3-32
-------
152
100
LU
Ul
6
u.
LU
80
60
40
20
(SIERRA 244EINLET)
AVERAGE OF ALL TESTS
O 5 km/hi
A 15km/hr
O 40km/hr
I I I M
1I I
3 5 7 10 15
AERODYNAMIC PARTICLE DIAMETER, M
20 25 30
Figure 3-5. Sampling effectiveness of the dichotcmous sampler inlet as a
function of wind speed.
Source: Wedding et al. (1980).
3-33
-------
CO
W
UJ
IU
I
IU
u.
U.
UI
(9
Z
120
110
100
AERODYNAMIC DIAMETER, (Jtm
Figure 3-6. Sampling effectiveness of the Wedding IP inlet.
Source: Wedding et al. (1980).
3-34
-------
-------
particle sizes for a wide variety of filter materials, face velocities, and effective porosi-
ties have been determined by Liu et al. (1978a) for clean filters and by John and Reischl
(1978) for exposed filters. Appendix Table 3A-1 tabulates selected fractional efficiency data
(8
for the commonly used TSP hi-vol sampler glass fiber filter, the Teflon membrane filter used
by the dichotomous sampler, and the cellulose fiber filter material (Whatman No. 1) used by
the BS sampler. The latter filter shows some inefficiency at the smallest particle sizes,
/a
while the glass fiber and nominal 2 [jm porosity Teflon filters are highly efficient for all
particle sizes. The relationship of flowrate through the filter to the pressure drop across
it is also a very important mechanical consideration since this determines the available
operating flowrate range for a vacuum pump of any given size. Membrane filter samplers,
because of the rapid increase in pressure drop as particles deposit, require lower flowrates
than fiber filter samplers. This results in substantially less PM being collected during a
sampling interval and necessitates the use of a much more sensitive weighing device (balance).
3.3.2.1.1 TSP high-volume sampler. The hi-vol sampler is the EPA reference method for TSP.
3
It is intended to operate at flowrates from 1.1 to 1.7 m /min, drawing air through a 200 x 250
mm (8 x 10 in) glass fiber filter. The mass of particles collected on the filter is deter-
mined from the difference between weights before and after exposure. The mass concentration
is averaged over the sampling interval and is normally expressed in (jg of mass collected per
3 3
ra of air sampled ((jg/m ).
Although materials such as quartz fiber can be used, glass fiber is by far the most com-
monly used filter medium for this sampler and is nearly 100-percent efficient for 0.3 urn
particles (Liu et al., 1978a). As noted by Friedlander (1977), this size particle is the most
difficult to capture, since the collection of smaller and larger particles is accomplished by
diffusion and impaction, respectively. This filter material is not prone to rapid overloading
as is a membrane substrate and permits sampling over 24-hr periods in ambient TSP concentra-
2
tions in excess of 300 to 400 ug/m . Glass fiber filters, although available in a variety of
types, do not generally provide a chemically inert surface, and the surface impurities and
basic pH may interfere with some measurements. The fibrous nature of the filter also makes
surface measurements, such as X-ray fluorescence, impractical except for high atomic number
elements such as lead.
Ine hi-vol is relatively simple to operate and reasonably inexpensive to purchase. The
original method description in the Federal Register (U.S. Environmental Protection Agency,
1979c) was recognized to be an inadequate description of the procedure, and a much more
detailed document was prepared by EPA (Smith and Nelson, 1973) to improve the quality of TSP
data.
As shown in Appendix Figure 3A-3, the inlet is formed by the overhang of a gable roof
which serves as a rainshield for the filter. The inlet effectiveness, as already discussed,
does not produce a sharp particle size cutoff and is sensitive to windspeed. The collection
efficiency of the hi-vol is also affected by sampler orientation (i.e., it is somewhat sensi-
tive to wind direction) as described by Wedding et al. (1977). The average sampler flowrate
3-36
-------
is determined either by averaging single measurements before and after collection using an ex-
ternal flowmeter or by integration of a flow recorder trace. The effect of sampler flowrate
on the sampling effectiveness for large particles as shown in Figure 3-8 is not substantial;
however, use of a flow controller provides the most accurate sampler performance.
The absolute accuracy of ambient particle measurements such as those made by the hi-vol
sampler cannot be determined directly with current technology. On the other hand, estimates
of components of the overall accuracy can be determined, including the collection effective-
ness of the sampler inlet and filter media and the accuracy of the flow measurement system.
Two commonly used flow measurement devices on hi-vol samplers are the rotameter and the
orifice meter with a pressure recorder. The rotameter is used to measure the initial and
final flowrates from which an average is calculated. The pressure recorder provides a
continuous trace of the orifice pressure drop that can be integrated for a more accurate
measurement. Smith et al. (1978) using hi-vol samplers with both types of devices noted that
the pressure recorder produced smaller errors (2 to 4 percent) when compared with a reference
flow device than the rotameters (6 to 11 percent).
The precision of the hi-vol sampler as determined from collocated sampler measurements
under field conditions and expressed by the coefficient of variation (cv) have been reported
by several investigators. McKee et al. (1971) determined the cv for a measurement by a single
analyst to be 3.0 percent while the same measure among multiple analysts in a collaborative
test was 3.7 percent.
The design of the gable roof provides a settling chamber above the filter for larger
particles blown in during periods when the sampler is not operational. McFarland and Rodes
(1979) have determined this deposition experimentally as a function of particle size and
ambient windspeed. Interpreting these relationships, however, requires knowledge of the
existing ambient size distribution of particle mass. For a typical distribution, the amount
of mass added to a hi-vol sampler filter during 5 days of exposure when it was not operational
was predicted to be 6 to 8 percent. This effect has been measured in a field situation by
S.ides and Saiger (1976) and Lizarraga-Rocha (1976), who measured weight increases from 3 to 12
percent. Errors from this effect can be reduced by equipping the sampler with a mechanical
device that keeps the filter covered during nonsampling periods. Timely installation and
retrieval of filters will also minimize the problem.
As shown by Coutant (1977), Spicer and Schumacher (1979) and Appel et al. (1979), arti-
fact PM can be formed by, oxidation of acidic gases (e.g., SO^, NOp) or by retention of gaseous
nitric acid on the surface of alkaline (e.g., glass fiber) filters and other filter types.
The effect is a surface-limited reaction and, depending on the concentration of the acidic
gas, should be especially significant early in the sampling period. The magnitude of the
resulting error depends on such factors as the sampling period, filter composition and pH, and
the RH. The magnitude and the significance of artifact mass errors are variable and dependent
3-37
-------
100
80
I 60
tf
tu
z
01
p 40
u
20
I i i I I I I I I 1 1 I I I I I i I I I
O--
I I I I I I I 1 I 1 I I I I I I I I I I
OS 1.0 1.5
VOLUMETRIC FLOW RATE, m3/min
2.0
Figure 3-8. Effect of sampler flow rate on the performance of a Hi-Vol for
30 nm particles at a wind speed of 8 km/hr.
Source: McFarland and Rodes (1979).
3-38
-------
on local conditions. Excluding the uncertainty associated with the collection and retention
of organic particulate matter with appreciable vapor pressure, artifact mass primarily
reflects the sum of the sulfates and nitrates formed by filter surface reactions with S03 and
nitric acid gas, respectively. The ambient concentration of SO, is primarily dependent on
fossil fuel combustion, while the nitric acid concentration is dependent on atmospheric
photochemistry and, possibly, reactions in suspended water droplets (Orel and Seinfeld, 1977).
A laboratory study by Coutant (1977) reported artifact SO,, for 24-hour samples from 0.3 to 3
3 3
ug/m , Appel et al. (1978) observed up to 5 pg/m artifact SO. on alkaline glass fiber
3
filters in 24 hr laboratory exposures, and up to 3.2 ug/m artifact sulfate in atmospheric
trials at two California sites. Stevens et al. (1978) similarly found 2.5 pg/m average
artifact SO, based on sampling at 8 sites around St. Louis, Missouri; and Rodes and Evans
3
(1977) noted 0.5 pg/m artifact sulfate in West Los Angeles, California.
Artifact particulate nitrate values on glass fiber filters ranging from 1.9 to 26.4 ug/m
(mean 10.6 ±6.9 ug/m , n = 13), were reported by Spicer and Schumacher (1979) in Upland,
California. These values were obtained by comparison with nitrate concentrations measured
simultaneously with quartz fiber filters. The likelihood of negative sampling artifacts on
quartz fiber filters, as discussed below, make these artifact nitrate measurements upper limit
values only. Appel et al. (1980) reported that artifact particulate nitrate on glass fiber
filters is limited only by the gaseous nitric acid concentration. Such filters approximated
total inorganic nitrate samplers, retaining both particulate nitrate and HNO., even when the
latter was present at very high atmospheric concentrations (e.g., 20 ppb). Nitric acid was
found to represent from approximately 25 to 50 percent of the total inorganic nitrate at
Pittsburgh, Pennsylvania, and Lennox and Claremont, California. Based on an estimate of the
most probable 24-hour artifact sulfate error (3 (jg/m ), and of the most probable artifact
particulate nitrate (8.2 ug/m in the Los Angeles, California, Basin and 3.8 pg/m elsewhere)
typical errors in mass due to SO, plus nitrate artifacts are estimated at 11.2 ug/m in the
3
Los Angeles Basin and 6.8 pg/m elsewhere.
Nitrate salts can rapidly be lost from inert filters (e.g., Teflon, quartz) by volatili-
zation (Appel et al., 1980; Forrest et al., 1980), and by reactions with acidic materials
(Harker et al., 1977; Forrest et al. 1979). Loss of atmospheric nitrate from glass fiber
filters occurs slowly. For example Smith et al. (1978) observed a 25-percent decrease in
nitrate over 3 months in storage at room temperature accompanied by a corresponding loss of
ammonium ion. Colovos et al. (1977) noted loss of up to 1.5 pg/m NH, after 30 days storage.
Immediate analysis after collection would minimize the significance of such loss.
In general, the hi-vol sampler data have been shown to be reproducible (3 to 5 percent),
if an orifice meter and flow recorder are used. The sampling effectiveness for larger
particles is windspeed dependent and, based on the data in Figures 3-3 and 3-4, the effect of
windspeed could be estimated to produce as much as a 10-percent day-to-day variability for the
same ambient concentration for typical conditions. The effect of the sums of the reported
3-39
-------
positive and negative artifact related to the glass fiber filter could be expected to add 6 to
7 M9/m to the collected mass.
3.3,2.1.2 Dichotomous sampler. The dichotomous sampler collects two particle size fractions,
typically 0 to 2.5 urn and 2.5 to about 15 jam, the latter cutoff depending on the inlet. This
bimodal collection approximately separates the fine particles from the coarse particles as
described in Chapter 2 to assist in the identification of particle sources. Since the fine
and coarse fractions collected in many locations tend to be acidic and basic, respectively,
this separation also minimizes potential particle interaction after collection.
The particle separation principle used by this sampler was described by Hounam and
Sherwood (1965) and Conner (1966). As illustrated in a simplified fashion in Appendix Figure
3A-4, the separation principle involves acceleration of the particles through a nozzle, after
Which 90 percent of the flowstream is drawn off at right angles. The small particles follow
the right angle flowstream, while the larger particles, because of their inertia, continue
©
toward the collection nozzle. A separate 37 mm Teflon filter is used for each fraction. The
sharpness of separation is shown in Figure 3-9 from data by Loo et al. (1976) for a design
cutpoint at 2.5 urn. Although cutpoints below 2.5 pm are mechanically impractical with
dichotomous separators, 3.5 urn units are available commercially with equivalent cutpoint
sharpness. A 2.5 urn cutpoint for the separator was recommended by Miller, et. al.(1979)
because it provided good chemical separation between size fractions, satisfied the
requirements of health researchers, and was mechanically practical. Inherent in the
dichotomous separation technique is a contamination of the coarse particle fraction with a
small percentage of the fine particles in the total flowstream. This is not considered a
substantial problem for mass measurements and a simple mathematical correction as described by
Dzubay et al. (1977) can be applied.
Teflon membrane filters with porosities as large as 2.0 (jm can be used in the sampler
and have been shown to have essentially 100-percent collection efficiency for 0.3 urn particles
(Liu et al . , 1978a). Filters with smaller porosities, such as 0.5 and 1.0 urn, a^e also highly
efficient, but are prone to much more rapid clogging as loading increases, accompanied by
3
rapid decreases in sampler flowrate. Because the sampler operates at a flowrate of 1 m /hr
(16.7 i/min) and collects sub-milligram quantities of particles, a microbalance with a 1 ug
resolution is required for filter weighing. Removal of the stickier fine particles causes the
collected coarse particles to have a greater tendency to fall off the filter if care is not
taken during filter handling and shipments (Shaw et al . , 1979).
Dichotomous samplers are significantly more complicated to operate than single size frac-
tion samplers and therefore are more prone to operator errors. As with the low flowrate
cyclone samplers, the small mass collected on each filter requires careful weighing on a
microbalance to provide reproducible results. The Beckman inlet currently available for this
sampler is shown in Appendix Figure 3A-1. Testing has shown that this inlet, as well as the
essentially identical Sierra inlet, are significantly windspeed sensitive, as shown in Figure
3-40
-------
120
SEPARATOR EFFICIENCY
3 4567 10
PARTICLE SIZE (DJ.um
20
Figure 3-9. Separator efficiency and wall losses of the dtchotomous sampler at 2 5 p.m.
Source: Loo et al. (1976).
3-41
-------
3-5. As the windspeed changes, the D™ changes, resulting in variable collection of the
larger particles. As noted earlier, inlets have recently been developed that have sharper
cutpoints and are less windspeed dependent.
Automated versions of this sampler can automatically change the sampler filters to
provide unattended operation. Depending on atmospheric concentrations, short-term samples of
as little as 4 hours are possible with the automatic samplers to provide diurnal pattern
information. The mass collected during such short sample periods, however, is extremely small
*
and high variability of the results could be expected. With an inlet sampling effectiveness
at 15 km/hr as described in Figure 3-5, the total mass collected would be 5 to 10 percent
lower than the concentration collected by an ideal (a = 1.0) inlet for a typical ambient size
3 "
distribution such as the 60 |jg/m case shown in Figure 3-3. The overall reproducibility of
dichotomous mass measurements is somewhat dependent on the care taken during filter handling
and weighing, but could be expected to be about ± 10 percent.
3.3.2.1.3 Cyclone samplers. Ambient cyclone samplers are simple to operate and only moder-
ately complex to build. Lippmann and Chan (1979) summarized the available cyclones for
ambient particle sampling below 10 jam and noted that the separation effectiveness of cyclones
can be designed to match respiratory deposition curves closely (see Figure 3-1). The cyclone
separation principle can be applied to larger particle cutpoints, as demonstrated by Wedding
et al. (1980) for a 15 urn sampler inlet. The small size of some cyclones makes them useful
for personnel dosimetry sampling, if a suitably small pump and flow control system are
employed. The Dorr-Oliver hydroclone, which is 10 cm in length and 10 mm inside diameter,
matches the ACGIH curve (American Industrial Hygiene Association, 1970) and can be used for
personal sampling. This cyclone has also been used in ambient field studies including the
Harvard 6-City Study (Lioy, et al., 1980).
A cyclone sampler used in the Community Health Environmental Surveillance Studies (CHESS)
(Barnard, 1976) is shown in Appendix Figure 3A-5. This sampler, as characterized in Figure
3-10, provides a relatively sharp separation with a D™ of 3.5 p.m. The inlet of the sampler
is the cyclone inlet, and a single 0 to 3.5 urn particle fraction is collected on the filter.
The filter medium used in the CHESS network was glass fiber.
At an operational flowrate of 9.0 liter/minute, a typical fine fraction concentration of
a
30 ug/m would result in the collection of only 390 pg of PM on the filter in 24 hours. At
this level, Barnard (1976) determined the reproducibility of this sampler to be 13 percent.
The effectiveness of the cyclone inlet for particles 3.5 urn and smaller should be nearly 300
percent. Use of the glass fiber filter would cause artifact mass problems similar to those
identified with the hi-vol sampler.
Lippman (1970) discussed the effect of sample flowrate on the performance of cyclone
samplers. Knight and Lichti (1970) compared the performance of the 10mm cyclone sampler to
that of horizontal elutriators and noted that the results were comparable if the appropriate
flowrates were used. Cap!an et al. (1977) noted that five different flowrates, from 1.4 to
3-42
-------
100
90
80
S 70
g
i
FECTIVENESS,
en e>
o o
u.
UJ
0 40
13
a.
< 30
20
10
0
(
I ! I I I I I I _J £
-«•**"" ""
-
X
ox
/
/
o
1
- /
I
1
1
(61111111
J12 3456789 1(
AERODYNAMIC PARTICLE SIZE,microns
Figure 3-10. Sampling effectiveness for the 3.5-jum outpoint CHESS cyclone sampler.
Source: Barnard (1976).
3-43
-------
2.8 liter/minute, have been recommended by researchers since 1962 in order for this cyclone
sampler to meet the ACGIH curve. They also noted that these small samplers are unaffected by
ambient air velocity, dust loading, mass loading, orientation, or aerosol charge. The
reproducibility of this sampler has not been given in the literature, but the low sampler
flowrate and proportionately small aerosol mass collected may result in values greater than ±
10 percent.
Collection of the larger particles excluded by a cyclone sampler on a removable subsbrate
is difficult, but alternative approaches, such as that designed by John et al, (1978) shown in
Appendix Figure 3A-6, are available to provide a "total" sample dependent on the effectiveness
of the inlet. The efficiency data for this cyclone as a function of sampler flowrate are
shown in Figure 3-11 and indicate that sharp cutpoints are possible with current state-of-
©
the-art units. A neutral pH Teflon filter medium was recommended to minimize artifact mass
formation. The inlet normally used for this sampler is the dichotomous sampler inlet shov/n in
Appendix Figure 3A-1. This inlet was designed to operate at 16,7 £/min. The windspeed
influence on sampling effectiveness is shown in Figure 3-5, Reproducibility data for this
sampler are not available but would be expected to be approximately 10 percent.
3,3.2.1.4 Hi-vol sampler withsize selective inlet. To meet the monitoring requirements for
Inhalable Particles (IP) as proposed by Miller et al. (1979), EPA commissioned the design of a
size-selective inlet for existing TSP hi-vol samplers to provide a single 0 to 15 urn particle
size fraction. This inlet is mounted on a conventional hi-vol sampler in place of the gable
roof inlet. (See Appendix Figure 3A-7.) It has been tested by McFarland and Ortiz (1979) and
has an inlet effectiveness as shown in Figure 3-12 and a sensitivity to windspeed as shown in
Figure 3-13. Dry particle bounce and reentrainment were also reported to be insignificant at
3
the sampler flowrate of 1.1 m /min.
The filter materials used are the same as those used for TSP hi-vol samplers, thereby
presenting the same potential for artifact mass formation. This sampler, as with any size
fractionating device, is somewhat sensitive to sampler flowrate for larger particles, as shown
in Figure 3-14. However, these data suggest that special flow controlling measures are not
necessarily required to maintain consistent collection efficiencies over a range of sampler
flowrates.
The inlet effectiveness data shown in Figure 3-12 would indicate reasonably accurate
particle collection with minimal windspeed dependence. The influence of artifact mass on the
3
total mass collected could be expected to add about 6 to 7 (jg/m . The reproducibility should
be similar to the 3 to 5 percent of the TSP hi-vol.
3.3.2.1.5 Elutriator samplers. The British Medical Research Council (BMRC) (Orenstein, 1960)
defined a respiratory system particle deposition curve (see Figure 3-1). This deposition
curve is defined by the performance of a horizontal elutriator consisting of multiple parallel
plates (Hamilton and Walton, 1961). A schematic diagram of this elutriator is shown in
Appendix Figure 3A-8. This sampler has been used in Great Britain for ambient air sampling
3-44
-------
AERODYNAMIC DIAMETER, fig
Figure 3-11. Fraction of methylene blue particles deposited in a cyclone
sampler as a function of the aerodynamic particle diameter. Curves are
labeled with flowrate in liters/min.
Source; John et al. (1978).
3-45
-------
100
80
60
en
sa
ill
Z
u
ui
ui
40
20
I I I I I I I
2 4 6 8 10 20 40
AERODYNAMIC PARTICLE DIAMETER, Mm
Figure 3-12, Sampling effectiveness for the size-selective inlet Hi-Vol sampler.
Source; McFarland and Ortiz (1970).
3-46
-------
20
15 —
I
o
IB
O
LU
N
O
a.
o
10
I
_L
04 8 12 16 20
WIND VELOCITY, km/hr
Figure 3-13. Effect of wind speed upon outpoint size of the size selective inlet.
Source. Me Far I and and Ortiz (1979).
24
3-47
-------
S.
ta
ui
z
100
80
60
o
S 40
20
! I I I I i I I i I I I I I i i I I I I
nl I I I I I I I I I I I I I I I I I I I I
0.8 1.0 1.5
VOLUMETRIC FLOW RATE,m3/min
2.0
Figure 3-14. Effect of sampler flow rate on the sampling effectiveness of
the size selective inlet Hi-Vol for a particle size of 15.2 jim and wind speed
of 2 km/hr.
Source: McFarland and Ortiz (1979).
3-48
-------
and during mining operations in the United States as an occupational exposure sampler. Corn
et al. (1967) successfully used a horizontal elutriator to collect ambient particles below 3
urn selectively for optical examination on glass slides. Hamilton and Walton (1961) noted that
reentrainment of coarse particles can be a problem in an elutriator if mechanical vibration
exists. Because current ambient or wind tunnel test data on these samplers are not available,
the reproducibility or accuracy cannot be estimated.
3.3.2.2 Impactor Samplers—Impactor samplers provide a means of dividing an ambient particle
sample into subfractions of specific particle sizes. As shown in Appendix Figure A-9 for a
cascade impactor the jet of air is directed toward a collection surface, which is often coated
with an adhesive or grease to enhance collection. Large, high-inertia particles are unable to
turn with the airstream and consequently impact against the collection surface. Smaller
particles follow the airstream and can be directed either to another stage of impaction or
collected on a filter. Use of multiple stages, each with a different particle velocity,
provides collection of particles in several size ranges. Particle size distributions are
constructed using impactor sampler data. (See Figure 3-3.)
Impactor samplers use removable impaction surfaces for collecting particles. Impaction
substrates are weighed before and after exposure and typically are metal foil plates or glass
fiber filters. The selection and preparation of these substrates have a significant effect on
the impactor performance. Improperly coated or overloaded surfaces can cause particle bounce
to lower stages, resulting in substantial cutpoint shifts (Dzubay et al.,r!976). Marple and
Willeke (1976) showed the effect of various impactor substrates on the sharpness of the stage
cutpoint. Glass fiber substrates can also cause particle bounce and are subject to the forma-
tion of artifact particles similar to those on hi-vol sampler filters.
3.3.2.2.1 Cas cade impactors. Cascade impactors typically have 2 to 10 stages, and commercial
3
low-volume version flowrates range from about 0.01 to 0.10 m /minute. Lee and Goranson (1972)
2
modified a commercially available 0.03 m /minute low-volume impactor sampler and operated it
o
at 0.14 m /minute to obtain larger mass collections on each stage. Cascade impactors have
also been designed to mount on a hi-vol sampler and operate at flowrates as high as 0.6 to 1.1
o
m /minute. A hi-vol sampler with a single impactor stage, shown in Appendix Figure 3A-10, was
o
used in the Community Health Air Monitoring Program (CHAMP) and operated at 1.1 m /minute.
Particle size cutpoints for each stage are 'dependent primarily on sampler geometry and
flowrate. The smallest particle size cutpoint routinely used is approximately 0.3 urn,
although special low-pressure impactor samplers, such as that described by Hering et al. (1978)
are available with cutpoints as small as 0.05 urn. A high-efficiency filter typically is used
after the last impaction stage to .collect the small particles ,not impacted previously. The
masses collected on each stage plus the backup filter mass collection are often reported, as
shown in Figure 3-15 from data by Lee (1972).' This cumulative d-istribution format permits
determination of the Mass Median Diameter (MMD), at which point 50 percent of the mass is
smaller than the indicated size. Use of straight line plotting techniques (as shown in Figure
3-49
-------
10,0
5.0
w
O
DC
U
CC
111
UJ
s
5
UJ
o
CC
IX
2.0
1.0
0.5
0.2
0.1
i i i i r~
CHICAGO, III. 1970 AVERAGE
I I I
12 5 10 20 30 40 50 60 70 80 90 95 98 99
CUMULATIVE PERCENT MASS <£ PARTICLE DIAMETER
Figure 3-15. An example of a mass size distribution obtained using a cascade impactor.
Source: Lee (1972).
3-50
-------
3-15) implies a log normal mass distribution, which can result in misinterpretation of the
mass median diameter.
Cascade impactors are not normally operated in routine monitoring networks because of the
manual labor required for sampling and analysis. Although impactor sampling systems are not
extremely complex, careful operation is required to obtain reliable data, especially if coated
collection surfaces are used. Analysis of constituents other than mass to obtain size
distributions of species such as sulfates are possible, but require careful analytical
techniques or compositing by stage with other samples to obtain an adequate quantity of
material for analysis. Impactor stages that use grease coatings may prove undesirable for
certain analyses because the grease may interfere with the method. Natuseh and Wallace (1976)
investigated the errors associated with impactor sampling and concluded that even under very
unfavorable conditions the HMD can be determined to well within 25 percent of the true value.
The inlet characteristics of most impactors have not been determined, resulting in uncer-
tainty about the size range of particles sampled. McFarland (1980) examined the inlet of the
NASN low-volume (0.14 m /min) cascade impactor and determined that particles larger than 10 urn
were unlikely to reach the collection stages because of substantial wall losses. Willeke and
McFeters (1975) characterized the CHAMP hi-vol sampler inlet under static windspeed
conditions, as shown in Figure 3-16. If the characteristics of the impactor inlet are known,
the total mass collected by the sampler can be used for comparison with other similar size-
specific measurements.
The particle separation of an impactor stage can be very sharp, and mathematical models
are available to permit stage sizing at selected cutpoints. The single impaction stage of the
CHAMP hi-vol sampler designed to be 3.5 \an was characterized by Ranade and Van Osdell (1978)
and demonstarted a reasonable agreement with theory (Figure 3-lf). Note, however, that solid
particles above 5 urn deviate from the relationship, indicating possible particle bounce
effects.
3.3.2.2.2 Rotary inertial impactors. Whereas cascade impactors draw the fluid stream to the
impactor surfaces, rotary inertial impactors move the impactor surfaces through the fluid.
The impactor surfaces, which are spun by an electric motor, are coated with a sticky film and
collect particles with diameters greater than about 1 ura by inertial impaction. The
collection efficiencies of such samplers, which include the Rotorod (Balzer, 1972) and the
Noll rotary impactor (Noll, 1970), are based on the Stokes number (Fuchs, 1964).
Using rectangular glass slide collectors, at a rotational speed of at least 35 m/sec (78
mph), Noll (1970) measured efficiencies of 85 to 100 percent for particles from 6 to 108 urn.
The collection efficiencies are windspeed independent if the rotational velocity of the
impactor surfaces is large compared to the ambient windspeed.
Microscopic counting techniques are used to determine the particle distribution on the
collector. The sample volume is the volume swept by the stage during the sampling interval.
Variations of this design principle include exposed wires and plates extended from moving
vehicles such as boats and airplanes.
3-51
-------
co
a
Ul
a
3
I
Ul
CO
Ul
_J
u
p
e
<
a.
1
o
<
DC
U.
100
90
80
70
60
50
40
30
20
10
20.0
25.0
30.0
35.0
40.0
AERODYNAMIC DIAMETER, fim
Figure 3-16. Fractional particle collection of the CHAMP
fractionator inlet at a sampler flow rate of 1133 liters/min.
under static windspeed conditions.
Source: Willeke and (WcFeters (1975).
3-52
-------
o
a.
100
90
80
70
60
z
ui
0
E 50
u.
ui
2 40
U
IU
3 -
u
20
10
CALCULATED FROM MARPLi'S THEORYH970)
AMMONIUM FLUORESCEIN
(SOLID)
O DI-OCTYL PHTHALATE
(LIQUID)
I I I I I I
1 2 3 456789 10
AERODYNAMIC DIAMETER, pn
Figure 3-17. Efficiency of the single impaction stage of the CHAMP Hi-Vol sampler.
Source: Ranade and Van Osdell (1978).
15
20
3-53
-------
The advantage of this type of sampler is that it eliminates inlet biases associated with
aspirated samples, especially for particles > 10 urn in turbulent air. Because of the amount
of labor required to measure large-particle size distributions, they are not practical for
routine use in networks.
3.3.2.3 Dustfall Sampling—Since very large suspended particles have appreciable settling
velocities, they are collected by deposition in a dustfall container and weighed as described
by the American Society for Testing and Materials (1981a). Although a cylindrical jar might
be expected to collect the equivalent of the dust content of an air column of its own diameter
extending to the top of the atmosphere, the aerodynamic effects of the jar, the angle of ap-
proaching windflow, the mounting brackets for the jar, and adjacent structures tend to compli-
cate the collection pattern. As noted by Nader (1958), it is difficult to interpret the mean-
ing of dustfall data and the significance of correlations with other measurements. The most
recent evaluation of available dustfall measurement techniques was reported by Kohler and
Fleck (1966), who noted that typical coefficients of variation were less than ±10 percent.
3.3,3 Non-gravimetric Mass Measurements
A variety of particle measurement techniques other than direct weighing are available.
Many of these techniques collect the particles on a filter substrate, followed by an analysis
that measures an integral property of the deposited particle other than the total mass.
Examples include light reflectance, light transmittance, and beta-ray attenuation. Other i_n
situ measurements are also used, which do not deposit the particles on a filter but measure a
characteristic of the suspended particles, such as light scattering. Host of these
alternative methods are less' expensive per sample and provide more rapid collection and
analysis of data than gravimetric analysis. Some measurements are not generally useful
because they depend heavily on site-dependent particle characteristics, such as color or
density. In most cases, a scientifically-based physical model relating the integral measure-
ments to mass is not available, thereby providing no basis for regression analysis. A
site-by-site best-fit regression must then be considered, which provides questionable accuracy
in predicting the true mass concentration, especially if the composition of the local PH
changes with time.
3.3.3.1 Filtration and Impaction Samplers—Samplers in this category collect particles on a
substrate and then use an alternative analytical technique as a surrogate to direct weighing.
3.3.3.1.1 British Smoke Shade sampler (BSS). The design of the currently used BSS sampler is
based in part upon early work by Hill (1936),, who used transmitted light to assess the
darkness of the stain resulting from particle collection on the filter paper. This sampler
draws an airstream upward through an inverted funnel and 3 meters of nominal one-quarter-inch
diameter plastic tubing to an inverted filter holder containing a Whatman No. 1 cellulose
fiber filter. As noted earlier, the Whatman filter medium has been shown to be somewhat
3-54
-------
inefficient when sampling very small particles (Liu et a!., 1978a). A schematic diagram of a
version of this sampler designed to collect sequential samples for 8 days is shown in Appendix
*
Figure 3A-11. A bubbler is often used downstream of the filter holder for subsequent SCL
measurements. The sampler is operated at approximately 1.5 liters/minute, which is verified
by a dry test meter built into the sampler. The filter holder can be 25, 50, or 100 mm in
diameter to collect a spot of the proper darkness range for subsequent measurement by
reflectance.
Since the early 1900's, many studies have been conducted in order to establish
relationships between smoke shade reflectance readings and gravimetric measurements. A 1964
study supported by the OECD established the currently used relationships between smoke shade
reflectance measurements and gravimetrically determined particulate concentrations. These
data were accepted by the WHO (1976) and compiled into a standard operating procedure for
reporting smoke shade measurements in equivalent ug/m . These equivalent mass concentrations
are not determined by weighing the smoke shade sampler filter but through comparison with a
collocated gravimetric sampler. The gravimetric measurements that were made for OECD and
compared to the smoke shade measurements were called "high-volume sampler" readings, but were
not taken with the U.S. TSP hi-vol sampler. The OECD gravimetric "high-volume sampler" as
described by the British Standards Institution (1964) operates at approximately 60
2
liters/minute, compared to the 1.5 m /minute of the U.S. hi-vol sampler. The OECD hi-vol was
designed to be aerodynamical^ similar to the smoke shade unit but has not been characterized
for aerosol collection effectiveness. Prior to 1964, various calibration curves were
3
published relating smoke shade reflectance to mass estimates in nominal pg/m units, but.it is
difficult to compare these earlier relationships to the OECD version.
McFarland (1979) examined the aerosol collection properties of the smoke shade sampler
and produced the effectiveness plot shown in Figure 3-18, which shows that the D™ for
particles reaching the filter (entire system) is only about 4.5 jjm. A comparison of the
"entire system" data with the deposition models given in Figure 3-1 shows a coincidental
agreement within 1 urn of -the ACGIH and Los Alamos curves. Most large particles are either
rejected at the inlet or lost in the inlet line, although some particles as large as 10 pm do
reach the filter. Because the size range of particles collected by the smoke shade sampler is
substantially less than that collected by the TSP hi-vol sampler, results of comparisons
between the methods could be expected to vary.
Ball and Hume (1977) and Waller (1963) noted that consistent relationships can be
developed, but are site, season, and particle-source dependent. Lee et al. (1972) noted, from
collocated TSP hi-vol and smoke shade sampler comparisons made at various sites in England,
that the overall correlation coefficients between these measurements for all sites was 0.618.
However, the individual coefficients ranged from 0.936 (good correlation) to 0.072 (no
correlation). Bailey and Clayton (1980) showed that smoke shade measurements correlated more
closely with soot carbon content than with gravimetric mass. Recent work by Edwards (1980)
3-55
-------
100
80
&
tf
ui
ui
o
iu
u.
u.
UI
1
SO
40
20
1 I I I I I 1 1
"O
I
INLET ALONE
O 2 km/hr
• 8 km/hr
ENTIRE SYSTEM
A 2 km/hr
I
J I
5 7 10 30
AERODYNAMIC PARTICLE DIAMETER, ftm
Figure 3-18. Sampling effectiveness of the inlet alone and through the entire flow system of the
British Smoke Shade sampler.
Source; McFarland (1979).
3-56
-------
has shown that smoke shade reflectance measurements can be related to the absorption
coefficient of the atmosphere. This work also showed that smoke shade measurements can be
converted to approximate COH measurements made by the American Iron and Steel Institute (AIS!)
tape sampler (see Section 3.3.3.1.2) using the absorption coefficient relationships. As
several investigators noted, (e.g., Lodge, et al. 1981), if a relationship could be developed
between optical and gravimetric measurements, it would be site specific but still variable
because of seasonal and long-term differences in the sources of collected particle size
fractions and their carbon content. The smoke shade sampler is relatively simple and
inexpensive to use for routine monitoring. The British Standards Institution (1964) reported
the reproducibility of collocated smoke shade sampler measurements as 6 percent. The accuracy
of a given relationship's predicted mass concentrations from reflectance measurements cannot
be discussed in general terms because of -the previously mentioned confounding influences.
Subsequent use of smoke shade results elsewhere in this document (Chapter 14) will discuss the
accuracy of the measurements relative to the specific studies evaluated there.
3.3.3.1.2 Tapesampler. A variation of the optical measurement of',€pot darkness is the use
of a continuous filter tape and an automatic tape advancing system. A sampler; using this
approach, developed by Hemeon (1953) for the AISI, samples at a'flowrate of approximately 7
liters/minute using Whatman No. 4 filter paper and collects particles on a 25 mm filter spot.
f *
The spot darkness is read either by a transmittance or reflecta'nce measurement. Transnffttance
measurement is most popular in the United States.
The AISI tape sampler typically collects particles in selected time intervals of 1 to 4
hours, and then advances to an unexposed clean portion of/the tape. Optical measurements are
referenced to an unexposed filter area and can be made either external to the sampler after
sample collection or with a continuous readout self-contained in the sampler.
•6
Transmittance measurements are converted to optical density through a Beer's Law
relationship and then to CoH units per 1000 linear feet of air sampled. A CoH is defined as
the quantity of PM on the paper tape" that produces a change in optical density of 0.01. The
alternate Reflectance Unit of Dirt Shade (RUDS) is equivalent to 0.1 CoH units per 1000 feet
(American Society for Testing and Materials, 1981b).
As stiown in Appendix Figure 3A-12, this sampler uses a funnel inlet and a small diameter
transport tube nearly identical to the British Smoke Shade sampler. Although the .two samplers
operate at different flowrates, the particles reaching the filter tape could be expected to
have a size range similar to that illustrated in Figure 3-18.
The utility of the sampler to estimate mass concentrations has been investigated by many
researchers, usually in comparison with the TSP hi-vol sampler. Since these two samplers do
not collect similar particle size ranges, such comparisons could be expectedvto vary unless
only a small proportion of coarse particles are present. Regan et al. (1979,) as well as
others have shown with field data that the correlation improves substantially when the tape
3-57
-------
sampler data are compared with smaller particle fractions such as the 0-2.5 urn fine fraction.
As noted in the discussion of the smoke shade sampler, the accuracy of a relationship between
AISI readings and mass concentration for a given data set is difficult to predict.
3.3.3.1.3 Beta-ray attenuation. Beta-ray attenuation is another technique for estimating the
mass of particles collected on a filter, without direct weighing. An exposed filter is placed
between a low-energy beta-ray source, such as Ni, C, or Pm, and a beta detector is used
to measure the amount of attenuation caused by the particle as compared to a clean filter. A
set of gravimetrically prepared standards are used to relate the results to units of mass.
This method is useful because it can be automated to handle a large number of samples
(Goulding et al., 1978; Loo et al., 1978). Real time mass measurements are also feasible
(Macias and Husar, 1976).
Investigators (Macias and Husar, 1976; Goulding et al., 1978) have studied the dependence
of the beta-ray absorption coefficient on elemental composition of the sample. Goulding et
al. (1978) found the dependence on composition to be reasonably small for the ranges of
average compositions that occur in aerosol samples. In a recent inter!aboratory comparison of
aerosol sampling and measurement methods (Camp et al., 1978), it was demonstrated that labora-
tory beta gauge measurements of ambient aerosols collected at one site by dichotomous samplers
on Teflon filters compared favorably in precision and accuracy with concurrent gravimetric
analyses. Principal sources of error in the method are possible changes in the orientation of
the filter substrate between the pre- and post-sampling measurements and changes in attenua-
tion because of absorption of water from the atmosphere by the filter material or the
collected particulate matter (Lawrence Berkeley Laboratory, 1975).
Jaklevic et al. (1981), after an extensive evaluation of the technique, have concluded
2
that the precision is typically better than 5 ug/cm for a variety of sample types, assuming
reasonable care in laboratory' manipulation. The precision and accuracy are significantly
better for samples of fine particles only, since one of the major causes of deviation from
ideal behavior is the presence of a significant number of larger particles. The possibility
of automating the process makes it particularly attractive for laboratories handling large
numbers of samples. Especially considering the latter advantage, it is not seriously inferior
to gravimetric techniques. In some cases it may be superior in precision and accuracy to
weighing very lightly loaded samples by hand on a microbalance, since it is insensitive to
such errors as those caused by loss of small fragments from the edges of the filters.
3.3.3.1.4 Piezoelectric microbalance. The piezoelectric microbalance technique collects par-
ticles on an oscillating quartz crystal, either by impaction or electrostatic precipitation.
The frequency change of the crystal oscillation is proportional to the mass collected and the
rate of change in frequency is proportional to the mass concentration (Woods, 1979). Advan-
tages of the piezoelectric detection principle, as-noted by Lundgren et al. (1976), include
extreme sensitivity and real time response. The technique can also be applied in a multiple
stage impactor form, using crystals as the collection plates. This approach provides rapid
determination of particle size distributions.
3-58
-------
Disadvantages of this detection principle include the need for frequent cleaning of the
crystal and severe interference from large RH changes and nonlinearity of crystal response to
large particle concentrations. As shown by the data of Lungren et al. (1976) in Figure 3-19,
the effect of humidity depends on particle type and is, therefore, nearly impossible to pre-
dict. Daley and Lundgren (1974) studied the potential errors in piezoelectric detection and
noted that although not currently used as a routine monitoring method, it can be used sucess-
fully for short-term studies when realistic operating limits are observed.
3.3.3.2 In Situ Analyzers—Instead of collecting particles on a filter before analysis,
certain aerosol characteristics can be examined while the particles are still suspended in the
airstream.
3.3.3.2.1 Integrating nephe1ometer. The integrating nephelometer measures the optical
scatter caused by suspended particles in the airstream. The initial designs for this
technique were made by Beuttell and Brewer (1949) and subsequently improved by Ahlquist and
Charlson (1967). The nephelometer is calibrated using the known scattering coefficients of
gases such as Freon 12 (CCl-F,). Light scattering is at a maximum for particles in the 0.3 to
0.8 urn size range as shown in Figure 3-20. Accumulation mode particles are the primary catise
of light scattering, which is only slightly affected by particles in the nucleation or coarse
particle modes (Waggoner, 1973). Visual range can also be calculated from the scattering
coefficient using a relationship developed by Koschmieder (1924).
Correlation of optical scattering with hi-vol suspended particle mass would be possible
only at sites where coarse particle mass concentrations are low or correlated with
accumulation mode particle mass. From theoretical models (Waggoner et al., 1973), there
should be a high site-independent correlation between fine particle mass loading and optical
scattering. Waggoner and Weiss (1980) and Groblicki et al. (1980) have reported correlations
above 0.95 between gravimetric fine particle mass and nephelometric measurement of scattering
extinction. Their measurements found the same ratio of scatter to fine mass at an urban and a
rural Colorado site, as well as at a Pacific coast site in Washington. Nephelometric
scattering extinction appears to be a useful indicator for fine particle mass but will provide
erratic results when compared to any particle measure that contains coarse particle mass. A
comprehensive discussion of tha nephelometer as a visibility monitor is contained in Chapter 9.
3.3.3.2.2 Condensationnuclei counter. The condensation nuclei counter measures the total
light scatter of submicron particles whose size has been increased by condensing vapor onto
their surface in a cloud chamber. This device is of interest in determining the number con-
centration of particles in the nuclei mode but is not normally used for particle sizes above
about 0.5 urn (Perera and Ahmed, 1978). It is often used in conjunction with prior size
separation stages to obtain a particle size distribution for submicron size particles. The
condensation nuclei counter is rarely used for routine monitoring.
3-59
-------
URANINE
10
20 30 40 BO 60 70 80
RELATIVE HUMIDITY, percent
90
Figure 3-19. Response of a Piezoelectric Microbalance to relative humidity for
various particle types.
Source: Lundgren et al. (1976).
3-60
-------
10
.3 3-
£^"5
•F ~
* E
\ • ' i ' '"I I ' ' I ' '"tT
SCATTERING / \
I »
/ %
\ I l I | | |yf I I t II IIII
10"2 2 5 10"1 2
2 4.00
DiAMiTER, pm
Figure 3-20 Light scattering and absorption expressed per unit volume of aerosol.
Source: Charlson et al. (1978).
3-61
-------
3.3.3,2.3 Electrical aerosol analyzer (EAA). The electrical aerosol analyzer, as described
by Whitby and Clark (1966), measures the electrical mobility of particles as related to their
size. This device provides a detailed number size distribution over the range of 0.01 to 0.5
urn, approximately (Liu and Pui, 1975). The analyzer must be empirically calibrated to obtain
the relationship to aerodynamically sized particles. The size range sampled does not include
the entire'0 to 2.5 (jm fine fraction and, comparision between the measurements must include
extrapolations.
3.3.3.2.4 Diffusion battery. The diffusion battery, as described fay Sinclair et al. (1979),
is a set of parallel tubes or plates through-which the airstream flows to produce selected
differential particle removal as a-function of parti die size-by diffusion to the walls. A
condensation nuclei counter is us'ed as the particle counter. This diffusion separation prin-
ciple is useful in the 0.01 to. 0.3 (jm .range.
3.3.3.2.5 Optical particle counters. Optical particle counters direct the -flow stream
through a small nozzle into a narrow collimated light beam so that the light scatter from
single particles can be..measured. The signal produced by this scatter is mathematically
related to the geometric size of a spherical particle of a specified refractive index which
scatters an equal amount of light. These devices are sensitive to geometric particle
diameters from about 0.5 to 10 pm (Whitby and Willeke, 1979). Calibration with monodispersed
particles is required. For sampling of particles larger than 10 (jm, modification of
commercially available devices is required. Mass concentrations for specific size ranges can
be estimated by selecting an appropriate particle density. These devices can be used for
routine operations, but their usefulness in estimating mass concentrations is limited by the
accuracy and consistency of the selected average particle density and index of refraction.
3.3.3.2.6 Long path optical measurement. >Long-path (typically >1 km) optical measurement de-
vices for ambient air are available. They examine one of several possible aspects of
visibility over a defined distance. Transmissometers measure the attenuation of transmitted
light resulting from scattering and absorption in the atmosphere. These devices are similar
to their in-stack counterparts, requiring either a light source and receptor or light source,
retro-reflector, and receptor at separate locations. Telephotometers measure the contrast
caused by brightness differences between a distant object and its surroundings. These devices
appear to have promise as visibility monitors (see Chapter 9), but estimates of ambient mass
concentration have not been made from their data.
3.3.4 Particle Composition
Particles collected from ambient air contain a wide range of metallic elements and inor-
ganic and organic compounds. Their identification and determination usually involves the col-
lection of the particles upon a substrate (e.g., glass fiber filters in a high-volume sampler)
with subsequent chemical analysis in a laboratory. Most methods for analyzing the inorganic
fraction of particulate matter have focused on elemental and ionic composition. Atomic
absorption spectrometry has been the technique most used for the determination of metallic
3-62
-------
elements, although multielement analytical tools, such as optical emission spectroscopy, X-ray
fluorescence spectroscopy, and neutron activation analysis, have been successfully applied to
+ -?
the analysis of elements. For the most part, inorganic io-nic species (e.g., NH., SO. , NO-,
etc.) have been analyzed with wet-chemical spectrophotometric techniques. The organic frac-
tion of PM contatns aliphatic and aroma-tic hydrocarbons, acids, bases, and other organic
compounds, such as those containing nitrogen. Methods for analyzing organics generally
involve solvent extraction, some form of chromatographic separation and detection based on a
selected physical or chemical property of the specific compound.
Due to the complex chemical composition of the atmospheric particles and the wide variety
of compounds likely to be present, it is not practical to review all the possible methods for
their analysis and characterization. Only methods pertinent to the primary objectives of this
document are reviewed here in detail. Primary emphasis is placed on methodology for measuring
particulate sulfur compounds with lesser emphasis on metallic elements and other inorganic
ionic species. Detailed information concerning the analysis of airborne particles is
contained in a recent monograph edited by H. Malissa (1978), among others.
3.3.4.1 Analysis of Sul fates—Analytical techniques for determining trace amounts of sulfates
in clean, uncomplicated solution matrices are numerous (Forrest and Newman, 1973). Howev,er,
application of these techniques to complex atmospheric particles is not so straightforward,
Quantitative transfer from the collection medium and homogeneous dispersion in the analysis
medium—without contamination, chemical alteration, or cotransfer of analytical
interferents—is required.
A detailed critical review of the state of analytical methodologies for aerosol sulfur
compounds has been compiled by Tanner et al. (1978}. Tanner's review includes methods for
total aerosol sulfur, for total water-soluble sulfates, and for quantitative differentiation
of aerosol sulfur compounds of various oxidation states, as well as a definitive review of
methods for speciation of aerosol sulfate. Much of the discussion in this section is taken
from Tanner's review, with emphasis on the more widely used methodology. Where information is
available, a critical assessment of each method's capabilities is provided.
3.3.4.1.1 Total water-soluble sulfates. A comprehensive review of wet chemical methods has
been compiled by Hoffer and Kothny (1974), providing background information on the principal
methods for determinating trace sulfates in aqueous extracts of particulate matter collected
on filters. Sulfate measurements made with these methods, particularly when applied to the
analysis of samples collected with alkaline filter media, are vulnerable to error due to
"artifact sulfate" formation caused by the absorption and subsequent oxidation of ambient SOy
in the presence of the basic components of the filter media. With the use of common glass
fiber filters under normal high-volume sampling conditions, this error has been estimated to
range from 0.3 to >3 ug/m (0.1 to > 1 ppb), depending on the ambient SO, levels at the time
of sampling (Coutant, 1977; Pierson et al., 1980). This potential error should be considered
when assessing data collected using any of the methods for water-soluble sulfates discussed
below.
3-63
-------
Currently, the methods most widely:employed for soluble sulfate determinations are the'
BaSO. turbidimetric, methylthymol blue, thorin, and ion exchange chromatographic procedures'.
These methods have inherent analytical* precisions,1 accuracies, detection limits, linear work-
ing ranges, and other operational characteristics'which will be discus'sed in the following sec1*'
tions. Each method has an analytical 'lower detection limit (LDL) and working range for sul-
fate in the aqueous extract expressed in'jJg/ml. The'corresponding detection limit and working '
3
ranQe of each method for measuring sulfate in ait4 expressed in ug/m depend upon the analy-
tical LDL and the sample size (volume of air sampled, fraction of filter taken for extraction,
etc.)- For example, if an analytical technique'having a working range of 1 to 10 yg SQ./ml is
• 3
used to measure sulfate from a typical hi-vol'-sample (20.3 cm x 25.4 cm filter, 1.4 m /min
flow rate, 24-hr sample, 1.9 cm x 20.3s cm strip extracted in 50 ml), the working 'range for
Q "
sulfate in air would be D.3"to '2.9>ug SQ./rh .' l£ should'be "noted that the upper range limit
can usually be "extended to higher concentrations"by dilution of the aqueous extract prior to
analysis, •
3.3.4.1.1.1 BaSO* turbidiroetry.' Sulfate in the aqueous extract from a particulate sam-
ple* is precipitated by addition of barium chloride. The resulting"BaSO. turbidity is "measured
spectrophotometrically or'hephelometrically and compared to a standard curve prepared by mea-
suring the absorbance of standard solutions of sulfate. Numerous versions and modifications
of methods based on this principle appear in the 1'iterature (Kol'thoff'et a!., 1969; Techm'cori
Corp., 1959; American Pubfic Health Association, 1971; Appel et a!., 1979a). Appel, et al.
* » »
(1979a) described and evaluated a procedure applicable to the measurement of sulfate in
aqueous extracts from 24-hour hi-vol particulate samples. They reported an analytical working
3
range from 10 to 70 ug SO^/ml (2.9 to 20.3 ug S04/m for a typical hi-vol sample), an accuracy
within 4 percent, and an average* precision of 3.8 percent '(coefficent of variation) of the
working range. They reported that extract background turbidity and color interfere with the
procedure but are minimized by means of blanks. Sulfur compounds converted to sulfate by air
oxidation also interfere. The apparatus required for turbidimetric sulfate determination is
relatively inexpensive and if proper care is taken, the procedure is capable of producing re-
liable data.
3.3.4.1.1.2 Methylthymol blue (MTB). A reagent containing equimolar amounts of barium
ions and MTB, at a pH of 2.8, is added to the aqueous extract from a particulate sample. Sul-
fate in the solution is precipitated as BaSO. and the'pH of the solution -is raised to 12.4 by
addition of NaOH. The remaining barium combines with the anionic MTB and leaves an amount of
free MTB equivalent to the sulfate. The MTB is measured spectrophotometrically at 460 nm and
compared to a standard curve of absorbance versus concentration. Lazrus et al. (1966) de-
scribed an automated version of this method and reported that the reagent is oxidized in air
when made alkaline, thus limiting the use of the method to a closed system.
An evaluation by Appel et al. (1979a) of automated MTB methods examined two procedures;
one covering an analytical range of" 0 to 100 ug SO./ml developed by Midwest Research Institute
(HRI) (Bergman and Sharp, 1979) for EPA and the other'covering a range of 0 to 10 ug S04/ml
3-64
-------
developed by Colovos et al. (1976). The results for the MRI procedure indicated a working
range of 17 to 90 M9 S0,/ml (5.0 to 26.5 ug SQ./m for a typical hi-vol sample), an accuracy
of 1.06 (mean observed/ theoretical) with an average coefficient of variation of 2.8 percent
for analyses of filter strips spiked with known amounts of sulfate. Similarly, the results
for the Colovos et al., (1976) procedure indicated a working range of 2 to 10 ug S0«/ml (0.6
to 2.9 g SO^/m for a typical hi-vol sample), an accuracy of 0.98 and a precision of 1.3
percent. The samples must be treated with an ion exchange resin to remove metal ions which
may also react with MTB. No significant sources of interference were found in this evalua-
tion. The automated versions of the MTB procedure are widely used and are capable of produc-
ing reliable results. Large sample loads can be analyzed in relatively short periods of time.
The equipment is relatively expensive, but can be used for other analyses.
3.3.4.1.1.3 Thorin. Titrimetric methods for sulfate using barium ion and thorin indica-
tor for visual or spectrophotometric detection of the endpoint are popular (Akiyama,'-1957;
American Society for Testing and Materials, 1981c; Bakas, 1956; Oubois et al., 1969; Fritz,
1955 and 1957; Menis, 1958; Rayner, 1966). These procedures provide for titration of aqueous
sulfate with a solution of barium ion to precipitate barium sulfate (BaSO.). When the sulfate
is completely reacted, excess barium complexes with thorin to produce a pink color indicating
the endpoint of.the titration. The samples must be treated with a cation exchange resin to
remove metal ions that also complex with thorin.
A recent modification of this technique by Brosset and Perm (1978) allows rapid determi-
nation of sulfate by employing an automatic pipetting system. Aqueous sulfate extract is
treated with a solution containing an amount of barium in excess of the anticipated sulfate
and BaSO. is precipitated. Then, a solution of thorin indicator is added, which combines with
the remaining barium to form a colored complex. The absorbance of the solution is measured at
520 nm and compared to a standard curve obtained from sulfate standards. The absorbance of
the solution is inversely proportional to the sulfate concentration. This procedure has been
evaluated by Appel et al. (1977) who reported an effective working range of 3 to 13 ug SO./ml
3
(0.8 to 3.8 ug SO./m for a typical hi-vol sample), an accuracy of 1.04 (mean observed/theore-
tical), and a precision of 5 to 9 percent (coefficient of variation). No significant source
of interference is reported, but the samples must be corrected for background turbidity and
color. The Brosset modification employs an relatively expensive automatic pipet.
3.3.4.1.1.4 Ion exchange chromatography. The principle of the ion exchange chromato-
graphic technique is described briefly under Section 3.2.2.3.6. Stevens et al. (1978) de-
scribed the use of this technique for analysis of sulfate as well as other ions. Appel et al.
(1979a) evaluated a procedure for sulfate analysis using a system manufactured by Dionex Corp.
(1975). This procedure showed a working range of 7 to 130 ug SO./ml, an accuracy of 1.08
(mean observed/theoretical), and a precision of 6.2 percent (coefficient of variation). A
small interference from nitrate ion was also reported. Apparatus for this procedure is rela-
tively expensive and requires a skilled operator. Nevertheless, the procedure is considered
to be reliable and specific. Other ionic species can be determined simultaneously.
3*65
-------
3,3.4.1,2 Total aerosol sulfur. Nearly 100 percent of aerosol sulfur mass is present in the
form of sulfate (Forrest and Newman, 1973a). This was experimentally demonstrated by Stevens
et al. (1978), who also showed that most of the data on air-borne sulfur concentrations can be
accurately described as total sulfur calculated as sulfate or total soluble sulfate. X-ray
fluorescence is the primary and most practical technique for measuring total aerosol sulfur
collected on filters. This technique can be used to analyze many elements besides sulfur.
It is nondestructive and can be automated to facilitate the analysis of large numbers of
ambient aerosol samples. A particulate sample, collected on an appropriate filter (usually
©
Teflon ) is irradiated with photons (X-rays, gamma rays, etc.), protons, or other charged
particles, and the intensity of the fluorescent X-rays induced is measured as a function of
wavelength or energy to determine the amounts of the constituent elements present. Qualita-
tive and quantitative analysis can be obtained when the system is properly calibrated. This
calibration step is difficult, since few standards of known elemental composition are availa-
ble in disks of known thickness in an appropriate matrix (Adams and Van Grieken, 1975),
However, recent work by Dzubay et al. (1977) has shown that calibration standards can be
prepared to an accuracy of ± 5 percent.
The most extensive set of aerosol sulfur data were reported by Stevens et al. (1978) and
Loo et al. (1978) using a energy nondispersive X-ray fluorescence spectrometer designed by
Goulding and Jaklevic (1973). Stevens et al. (1978) reported sulfur and 18 other elements
from dichotomous samplers operated in New York City, New York; Philadelphia, Pennsylvania;
Charleston, West Virginia; St. Louis, Missouri; Portland, Oregon; and Glendora, California.
Loo et al. (1978) reported sulfur concentrations determined from samples collected over a
2-year period from a network of 10 automated dichotomous samplers operated in and around St.
Louis, Missouri, during the Regional Air Pollution Study (RAPS), They reported a detection
2. 3
limit of 0.034 \ig/cm of filter, which corresponds to a concentration value of <0.1 ug/m
sulfur for a 2-hour sample collected at 50 liters/minute on a 37-iran filter and is adequately
sensitive for a 1-hour time discrimination at ambient sulfur levels.
Proton-induced X-ray emission spectroscopy, which has the advantages of lower
bremsstrahlung background and focusing properties of the exitation beam, is a useful tool when
short-time resolution of ambient sulfur levels is desired (Johansson et al., 1975; Courtney
et al., 1978). However, it takes substantially more energy to produce an X-ray with charged
particles than with photons, and in some cases vaporization or decomposition of the sample may
occur (Shaw and Willis, 1978). A related approach to nondestructive aerosol sulfur analysis
based on cyclotron in-beam gamma-ray spectroscopy has been reported by Macias (1977). Gamma
rays induced by proton or a-irradiation are detected by a Li-drifted Ge detector and used to
determine S and other light elements such as Mg and C in aerosol samples. This technique is
less sensitive for S than X-ray emission methods. In spite of the aforementioned limitations,
3-66
-------
it is clear that these induced X-ray and gamma-ray methods will continue to be important tools
for determining total sulfur in large numbers of ambient aerosol samples.
Other techniques that have been applied to the determination of total sulfur in aerosol
particles include: (1) electron spectroscopy for chemical analysis (ESCA) (Novakov, 1973; and
Novakov et al., 1974); (2) various applications of flame photometric detectors (FPD) (Crider
etal., 1969; Kittelson etal., 1978; Huntzicker eta!., 1976, 1977, 1978; Tanner etal.,
1978, 1980); and, (3) an isotope tracer technique using Ag tracer (Forrest and Newman,
1977). ESCA is sensitive to surface composition of samples, which is an advantage for
surface-oriented studies but not for ambient aerosol samples whose elemental composition is
likely to be heterogeneous. A comparison of ESCA to wet chemical sulfate measurements by
Appel et al. (1976) showed agreement only within a factor of two. Direct flame photometry has
potential as a sensitive total aerosol sulfur analyzer, but-its application is complicated
because S0~ must be removed and the FPD response varies with the chemical form of the aerosol
sulfate. Recent work by Huntzicke'r et al. (1978) and Tanner et al. (1980) has shown that
direct flame photometry can not only provide a sensitive total aerosol analysis but, when
combined with thermal volatilization, can provide semicontinuous measures of H^SO,, ammonium
sulfate, and metal sulfates.
3.3.4.1.3 Sulfuric aciddetermination. Most of the efforts to determine the species composi-
tion of sulfate in airborne particles have concentrated on development of a specific analyti-
cal method for H?SO. in air. Despite the substantial efforts of several groups, the existence
of free aerosol H2SO, in the ambient atmosphere has been unequivocally established in only a
few cases. Interference problems and difficulties in sample preservation have contributed
markedly to the lack of valid HpSCL measurements. The procedures discussed below have been
applied primarily by research analysts, are vulnerable to error both in sampling and analysis,
and are not generally applicable to routine monitoring.
Procedures for determining H2SO. and other sulfate species include thermal volatilization
and solvent extraction techniques, gas phase ammonia (NH-) titration, infrared and visible
spectrometry, flame photometry, and electron microscopy. The determination of H?SO. by its
selective thermal volatilization from filters has been reported by several workers
(Scaringelli and Rehme, 1969; Dubois et al., 1969a; Maddalone et al., 1975; Thomas et al,
1976; Leahy et al., 1975). This technique generally suffers from poor H^SO. recoveries, poor
reproducibility, and interferences from ammonium sulfate salts. The most successful approach
to thermal volatilization of H?S(L. in ambient aerosol samples was reported by Mudgett et al.
, si
(1974). Aerosol samples are collected on Fluoropore filters, the HgSO. subsequently volati-
lized by passage of heated (~150°C), dry N? in the reverse direction through the filter and
released HpSO. determined with a flame photometric detector. Lamothe and Stevens (1976) re-
ported that laboratory aerosol samples of as little as 0.25 pg H?SO» may be determined with
reasonable precision. However, serious difficulties were encountered in removing HgSO^
quantitatively in the presence of ammonium bisulfate (NH.HSO,). They observed that HnSQ, is
3-67
-------
totally removed at 180°C, but NH4HS04 is also partially volatilized; at 140°C» NH4HS04 is not
volatilized but H^SO. is incompletely volatilized.
A solvent extraction procedure to remove collected hUSO. aerosol selectively in the pre-
sence of other aero,sol sulfates was first reported by Barton and McAdie (1971). They con-
cluded that aerosol collectfon on Nuclepore filters, followed by extraction with 2-propanol
for subsequent analysis by the chloranilate procedure, was selective for airborne H^SCK. Sub-
sequent work by Barton and McAdie (1973) reported reduction of interference by buffer control
of the 2-propanol extract and also reported development of an automated instrument for the
extraction procedure. Leahy et al. (1975) reported, however, that 2-propanol will also
extract NH^HSO. quantitatively and partially extract other bisulfates and that it should not
be considered a selective extractant for H?SO.. They demonstrated that benzaldehyde is a
selective extractant for H,SO. in the presence of bisulfates and sulfates. Subsequent radio-
35
chemical experiments (Tanner et al., 1977) with H9 SO. have established that H9SO. may be
• ft ®
reproducibly removed from a variety of filter media (Mitex , Fluoropore, H,P(K-treated
quartz) with recoveries varying from 75 percent for 10 ug H~SO, samples to 95 percent for LOO
pg samples. The selectivity of the benzaldehyde extraction technique has been confirmed in a
study by Barrett et al. (1977) employing laboratory aerosol H?SO* samples as low as 5 ug in
the presence of bisulfate and sulfate.
As discussed briefly in the previous section, systems for measuring ambient levels of
H,SO. ahd other sulfate aerosols by flame photometry have been developed recently by Huntzicker
et al. (1978) and Tanner et al. (1980). A heated denuder for SO- removal allows the direct
measurement \)f total sulfate aerosol and selective thermal volatilization allows the discrimi-
nation of semicontinuous measurements of H^SO. ammonium sulfates and metal sulfates. Con-
version 6ff the H2SO. or the aerosol sulfates to ammonium sulfate [(NH,)2SO,] by addition of
NH- v'i initiates the problem of FPD response variations with the chemical form of the aerosol
sulfate.
Since -atmospheric sulfate »can be associated with various cations, the compounds of sul-
fate can ^sometimes be inferred by measuring the cation. If the ions in a series of samples
are measured and Ithe ammonium (NH.) content is highly correlated with the sulfate content,
* 4*
then it can be inferred that vari6us NH- salts of H2$0- are probably present. Brosset and
Perm (1978) and Stevens et al. (1978) described a Gran titration procedure for hydrogen ion
(H ) and a procedure using ion selective electrode for NH, in aqueous extracts of aerosols
i". T"
collected on Teflon filters. Stevens et al. (1978) applied such techniques to aerosols
collected at Research Triangle Park, North Carolina, during the summer of 1977 and 1978 and
•f-4- 4- 4-
found a stoichiometric balance between sulfate ion (S04 ) and the sum of H and NH4 ion con-
centrations. The acidity was found to range from none [(NhL^SQ.J to that of NhLHSQ,.
Dzubay (1979) developed and used a sensitive radiolabeling technique for measurements of
acid sulfate aerosols. Liu et al. (1978b) described a new technique that uses an aerosol
3-68
-------
mobility chromatograph 'for the detection of sulfuric acid aerosols. Several semiquantitative
methods for estimating sulfate species have been investigated. These include gas phase
ammonia titration techniques (Dzubay et al. , 1974), humidographic techniques (Covert et al.,
1980), methods based on infrared spectroscopy (Blanco et al., 1968, 1972; Cunningham et al.,
1974; Cunningham and Johnson, 1976), and microscopy techniques (Heard and Wiffen, 1969; Lodge
et al., 1960; Frank and Lodge, 1967; Mamane and de Pena, 1978).
3.3.4.1.4 Filter sampling problems related to sulfate analysis. Ideally, the measurement of
aerosol sulfate species in PM requires that the sulfate-containing particles from the air be
quantitatively collected on a filter surface that does not permit chemical or physical trans-
formations and that does not lead to spurious sulfate particle formation from SO, present in
the gas stream. The particles must then be transferred to the analysis medium under the same
constraints. Sampling for airborne sulfate is especially difficult, since acidic sulfate
species react with many common filter materials (e.g., "neutral" glass fiber filters and many
plastic filters—Nuclepore , Acropore , Millipore ) as well as basic particles in the sample-
The result is neutralization of the -acid sulfate and alteration of the composition from that
extant in ambient air. Many of the historical data on sulfate species are questionable due to
insufficient consideration of the above sampling difficulties.
id
Several filter materials made of Teflon , have been found to be inert and suitable for
nonreactive collection of aerosols, including acid sulfates. The most widely used are backed
Teflon membranes, Fluoropore, and Mitex . A modified quartz filter material has been
developed (Tanner et al., 1977) from which impurities are removed by preheating to 750°C, and
reactive basic sites are removed by treatment with hot, concentrated phosphoric acid. After
rinsing and drying, the quartz filters may be used for high-volume, high-efficiency particle
collection without interfering with acid determinations of the collected particles at the
fractional microequivalent level.
Two additional problems have been identified in filter sampling for airborne sulfate
analysis. Sulfur dioxide may be converted to sulfate by adsorption on and catalytic oxidation
by the filter material (Lee and Wagman, 1966) and/or by previously collected particles (Coffer
et al., 1974). Recent studies by Forrest and Newman (1973a) seem to indicate that active
catalytic sites on the filter material are the likely culprits. Experiments by Tanner et al.
(1978) with high- and low-level S0?-spiked ambient air passed through preloaded and clean
hLPQ.-treated quartz filters at high and low linear flow velocities failed to find any
evidence of artifact sulfate formation for this filter material. This work was confirmed b>
the work of Pierson and coworkers (1976) from whose data it is clear that the low sodium
content of the Pallflex GAO quartz is the probable reason for the negligiblv low artifact
sulfate formation.
A second problem results from potential neutralization of acidic sulfate particles by NH,
in the gas stream traversing the filter. Neutralization by NFL and oxidation of SO^ may both
3-69
-------
be reduced by "diffusion-denuding" the stream of these gases (without removing significant
particles), but this is a cumbersome process, especially for high-volume sampling.
The most effective way to -reduce particle formation and transformation reactions on the
filter is to collect the smallest airborne particle sample that, is compatible with available
analytical methods. Minimum-quantity sampling also reduces collision-induced interaction of
particles on the filter surface and thus real-life chemical inhomogeneities in ambient parti-
cles are more likely to be unaffected by the sampling process. The only way to eliminate
confounding interparticle interactions on filters totally is to determine the sulfate in situ
without filter collection.
3.3.4.2 Ammonium and Gaseous Ammonia Determination—Gaseous NhU and ammonium ion (NH.)
measurements are important in understanding speciation of sulfate in airborne particles.
Ambient ammonia is by far the most important neutralizing agent for acid sulfate; its concen-
tration is directly related to the chemical form of sulfate in the ambient air. Measurements
of ammonia along with the set of species which, by mutual interaction, determine the chemical
form of sulfate, are crucial to understanding such pervasive problems as acid rain, visibility
degradation and the effects of dry deposition. Ammonium ion is found predominantly in the
optical-scattering size range or below and is presumed to be secondary in origin, being formed
in the neutralization of acidic sulfate particles. The high correlation of NH, content with
soluble sulfate, in both urban (Tanner et a!., 1977a) and rural (Tanner et a!., 1977) aerosol
samples, and the identification by X-ray diffraction of (NH4)2S£L in dried aqueous extracts of
airborne particles would tend to confirm the above hypothesis (Brosset et al., 1975).
Ammonium ion in particulate matter is almost always determined by collection on filters,
extraction into an appropriate leach solution, and determination by one of two methods. The
first is a concentration measurement by an ion-selective electrode sensitive to either NH*
(Beckman electrode) or NH~ (Orion or- Markson electrodes). The limit of detection is
determined by the equilibration time of the electrode, a representative value being 5 to 7
minutes for 20 ppb NH, concentration in water (Eagan and DuBois, 1974; Gilbert and Clay,
^ 4.
1973). This is marginally sensitive for high-volume samples of rural ambient air where NH,
3
may be as low as 0.3 ug/m . A later development, the air gap electrode (Ruzicka and Hansen,
1974; and Ruzicka et al., 1974), eliminates the problems of electrode contamination by sensing
of the NH_-water equilibrium across an air gap between the analyte solution and the electrode
surface.
The second commonly used method for NH. traces in aqueous solution is the indophenol
colorimetric method based on the color-producing reaction of phenol and hypochlorite in the
presence of NHg. Modifications most analytically useful for determination of NH, in aqueous
leaches were reported by Bolleter et al. (1961) and by Tetlow and Wilson (1964). Automated
procedures have been proposed by Lazrus et al. (1968) and by Keay and Menage (1970). The
latter method is capable of a lower detection limit of 0.05 (jg/ml (as nitrogen), requires only
a few minutes of analysis time, and has a minimum sample volume of 2 ml.
3-70
-------
Methods for determination of free atmospheric NH~ can be divided into direct methods and
methods in which NH, is first immobilized on acid-treated filters, leached, and determined as
+
NH* by one of the methods described above. Ammonia has been analyzed directly by quantitative
conversion to nitric oxide (NO) over a hot catalyst and determination by chemiluminescence of
the NO (Hodgeson et al.s 1971. and Baumgardner et a!., 1979). This method is marginally
3
sensitive enough for ambient levels of NH, (limit of detection = ~ 1 )jg/m ) and must be care-
fully zeroed in the NhL scrubber mode to eliminate interference from atmospheric oxides of
nitrogen (NO ). Filter pack methods using either KHSOA (Eggleton and Atkins, 1972) or oxalic
/> T* »
acid impregnants have been used to collect ambient levels of NH~, but they are fraught with
blank and contamination problems and may not collect ambient levels of 0.5 to 5 ppb NH-, with
reproducible efficiency under commonly observed conditions of temperature and relative
humidity. Shendrikar and Lodge (1975) applied the ring oven technique to the determination of
ammonia collected from ambient air on filter tapes impregnated with ethanolic oxalic acid.
The range for quantitative determination is reported to be 0.1 to 1.00 pg of ammonia.
It has also been proposed that gaseous NH3 could be determined at or below ambient levels
by gas phase reaction with HC1 vapor. The resultant ammonium, chloride (NH.C1) aerosol
particles would be measured by a condensation nuclei counter (CMC). Unfortunately, there are
several difficulties that severely limit the usefulness of this technique. The concentration
of HC1 and the relative humidity must be carefully controlled to attain proportionality
between number of particles and NH-, concentration. In addition, it is necessary to provide an
ionization source (a corona discharge or a UV light source) in the airstream just prior to HC1
vapor addition in order to approximate precise, proportional CNC response. However, this
method has potential for extremely high sensitivity and real time operation. McClenny and
Bennett (1980) developed a semi-real time detection technique for ambient NH, based on inte-
8
grative collection on Teflon beads, followed by thermal desorption and detection by either
chemiluminescence or photoacoustics. Perm (1979) and Braman and Shelley (1980) reported col-
lection of NH~ on diffusion tubes. Perm used oxalic acid as a coating which is rinsed from
+
the tube at the end of a 24-hour run and analyzed for NH. by ion selective electrode tech-
niques. Braman and Shelley used a tungsten oxide coating for 20 minute samples and released
the NH- into a chemiluminescence analyzer by thermal desorption. Hoell, et al. (1980) deter-
mined vertical concentration profiles by interpretation of infrared solar spectra obtained
with a heterodyne radiometer. Abbas and Tanner (1981) reported work on the continuous deter-
mination of gaseous NH-, using fluorescence derivatization. These recent advances in the
development of new techniques for measuring NH., will be helpful in determining the role of NH,
in conversion of H?SO. to less harmful materials.
3.3.4.3 AnalysIs o'f Nitrates—Nitrate analyses have been performed routinely for many years,
and a large number of chemical methods have been reported. In typical monitoring for nitrate
in air, a portion of a particulate filter is subjected to aqueous,extraction and the water-
soluble nitrate is analyzed by one of the methods discussed below.
3-71
-------
3.3.4.3.1 Measurement techniquesfor nitrates. The oldest procedures, for analyzing nitrate
used brucine (Intersociety Committee, 1977b) or phenol disulfonic acid (Intersociety Commit-
tee, 1977c). Many other methods for analyzing nitrates have been reported, including: (1) the
nitration of chromotropic acid (West and Ramach.andran, 1966) and coumarin analogs (Laby and
Morton, 1966; Skujins, 1964); (2) the quenching of the fluorescence after nitration of fluor-
escein (Axelrod et al. 1970); (3) reduction with Devarda alloy to ammonia (Kieselbach, 1944;
Richardson, 1938); and (4) the use of ion-selective electrodes (DiMartini, 1970; Driscoll et
al., 1972; Gordievskii et al., 1972). Microscopic techniques also allow identification and
size estimation of individual nitrate particles (Bigg et al., 1974).
One of the most extensively used techniques to analyze nitrates in atmospheric particu-
late extracts involves reduction of the nitrate to nitrite by zinc (Chow and Johnstone, 1962),
cadmium (Morris and Riley, 1963; Strickland and Parsons, 1972; Wood et al., 1967), or hydra-
zine (Mullin and Riley, 1955). Measurement of the nitrite produced is accomplished by a
sensitive diazotization-coupling reaction CSaltzman, 1954). Automated versions of this tech-
nique provide much better results because critical reduction parameters such as temperature,
surface contact area, and reaction time can be precisely controlled (Technicon, 1973). Lazrus
et al. (1968) used an automated system, in which nitrate was reduced to ammonium and deter-
mined by the indophenol method. Another extensively used technique to analyze nitrate in
atmospheric particulate matter extracts involves the nitration of xylenols and separation of
the nitro-derivative by extraction or distillation. A comparison of a 2,4-xylenol procedure
(Intersociety Committee, 1977d) with the automated copper-cadi urn reduction and diazotization
method in samples collected near high density vehicular traffic, demonstrated a negative in-
terference in the former up to a factor of 3 (Appel et al. 1977#). ,
Small et al. (1975) report an application of ion exchange chromatography to the measure-
ment of a wide variety of cations and anions including the nitrate and nitrite ions. The
novel feature of this method is the use of a second ion exchange "suppressor" column (after a
conventional separating column) that effectively eliminates the ions of the eluting medium.
Since the chromatographically separated species of interest leave the suppressor column in a
background of deionized water, concentration determinations may be made by a simple and
sensitive conductometric technique. Mulik et al. (1976) report the application of this
technique to measurement of water-soluble nitrate on hi-vol filters. The separator column,
containing a strong basic resin, separates anions in a background of carbonate eluant. The
suppressor column, containing a strong acid resin, converts the sample ion and the carbonate
eluant to nitric and carbonic acid, respectively. Since carbonic acid has low conductivity
and partially decomposes to carbon dioxide and H-0, the nitrate ion alone is effectively
measured in a conductivity detector. Under the experimental conditions, sensitivity of 0.1
ug/ml was reported. The relative standard deviation was 1 percent (95-percent confidence
3-72
-------
level) for 10 replicate injections at the 5 itg/ml level. At this concentration, no
interferences were found from fluoride, chloride, nitrite, sulfite, sulfate, silicate, or
carbonate. Positive interferences were found for bromide and phosphate, but the authors
suggest techniques for eliminating these.
In other work, Glover and Hoffsommer (1974) and Ross et al, (1975) reported a technique
for assay of aqueous nitrate and nitrite extractions by conversion to nitrobenzene. Both
techniques involve the nitration of benzene in the presence of H?SO, to form nitrobenzene, a
relatively stable compound, followed by gas chromatographic analysis. Careful calibration is
required in both procedures, since a significant fraction of the nitrobenzene formed may be
lost to the acid layer. Ross et al. (1975) recommended a calibration procedure, whereby a
standard is subjected to the same procedures as the unknown, while Glover and Hoffsommer used
internal calibration with added nitrotoluene. The lower detection limits reported by Ross et
—1 ?
al. (1975) are in the range of 10 g nitrobenzene in a 1 |jl sample. Conversion efficiences
for KN03, KN02, and HNO_ were reported as 90.3 ± 7.9, 100.4 ± 4.2, and 99.9 ± 5.2 percent, re-
spectively. Glover and Hoffsommer report similar recovery rates for KNQ~ and KNO?.
3.3.4.3.2 Filter sampling problems related to nitrate analysis. Serious difficulties have
been reported to be associated with the routine analysis of nitrates in PM collected using
glass fiber filters. In a study of nitrate in auto exhaust, Pierson et al. (1976) reported
that glass fiber filters collected about twice as much nitrate as quartz fiber filters.
Nitrate also was found on glass fiber filters that were inserted downstream of either quartz
or glass fiber primary filters, providing additional evidence of artifact formation from
gaseous constituents. Spicer (1976) reported that glass fiber filters completely removed
gaseous nitric acid (HN03) in low concentrations in gas streams, while Teflon and quartz
filters showed no corresponding effect. O'Brien et al. (1974) described very unusual particle
size distribution determinations for photochemical aerosol collected in the Los Angeles Basin.
This study used a cascade impactor, and all particle size fractions were collected on glass
fiber filters. The authors speculated that conversion of gaseous nitrate precursors on the
filter masked the true nitrate size distribution. Qkita et al. (1976) reported that untreated
glass fiber filters collect nitric acid vapor with a highly variable collection efficiency (0
to 56 percent), suggesting erratic nitrate artifact formation in urban atmospheres containing
HN03.
In an intensive laboratory investigation of interferences in atmospheric particulate
nitrate sampling, Spicer et al, (1978) concluded that all five types of glass filters investi-
gated exhibited serious artifact formation due to collection of gaseous HN03 and, to a small
extent, N0? as nitrate. Cellulose acetate and nylon filters were also reported to collect
(8
HNQ.,. Negligible interferences were reported for polycarbonate and Teflon filters. Collec-
tion of N0? on quartz fiber filters varied with the filter type, with ADL Microquartz showing
the least effect. Artifact nitrate formed on the Gelman AE filter was calculated to be less
than 2 (jg/m during a standard 24-hour hi-vol measurement. This estimate was derived from
3-73
-------
3 3
drawing air samples of about 1 m containing 4512 pg/m (2.4 ppm) N02 through the filters.
The relative humidity was 30 ± 10 percent.
Spicer and Schumacher (1977) also reported the results of a comparison of nitrate concen-
trations in samples collected on various filter types in Upland, California, during October
and November 1976. During the experiment, meteorological conditions varied from warm and hazy
to hot, dry, very clean desert wind conditions. Nitrate analyses were performed by ion ex-
change chromatography. All filter types used had comparable particle collection efficiencies
according to the manufacturers' specifications. The ratio of nitrate collected on glass fiber
filters to that collected simultaneously with identical hi-vol samplers on quartz filters
ranged from 2.8 to 49.
More recently, Appel and coworkers (1979, 1980) conducted several studies bearing on both
positive artifact formation and loss of nitrate from a variety of filter media. They con-
cluded that gaseous HNQ, is the principal source of artifact nitrate formation; NO, collection
only became substantial at high ozone levels. Ambient particulate nitrate values (at San Jose
and Los Alamitos, California) differed by up to a factor of 2.4 depending upon filter medium
and sampling rate, in contrast to the much larger differences reported by Spicer and
Schumacher, 1977. They also reported that at low HNO~ levels, nitrate on glass filters indi-
cated (within 3 percent) total nitrate, (i.e., particulate nitrate plus HNQ., rather than
particulate nitrate alone). They concluded that the degree of error associated with glass
fiber filter media could be expected to vary with location, time of year, and time of clay,
paralleling changes in HNO-, levels.
Laboratory studies by Harker et al. (1977) and ambient studies by Pierson et al. (1980)
have suggested that reaction of particulate nitrates with acidic particulate sulfates (e.g.,
H?SOfl) can result in negative errors in the determination of nitrate on filters. Formation
and subsequent loss of gaseous HNO, was presumed to be the mechanism. Recent studies by Appel
and Tokiwa (1980) support these observations and indicate that atmospheric particulate nitrate
00
levels obtained with Teflon filters may be only a small fraction of the true values. Similar
studies with Gelman A glass fiber filters showed insignificant nitrate losses.
Another mechanism for the loss of nitrate from particulate samples collected on inert
filters is the dissociation of NHJKU and loss of the resulting HN03. The equilibrium vapor
pressures of NH, and HNQ-, above solid NH.NO, are appreciable and very sensitive to temperature
(Stelson et al., 1979). In a laboratory study, Appel et al. (1980) observed losses of up to
®
50 percent of the particulate nitrate when NHL- and HNOv-free air was passed through Teflon
filters loaded with about 200 ug NH4N03 (<0.5 pm .particle size) at 20 liter/minute for 6
hours. These results suggest that the volatilization of NH.NO,, can be a major source of
©
negative error in sampling particulate nitrate with Teflon filters. The presence of
relatively high NH- and HNO, levels in ambient air or high humidity may decrease the error,
while elevated temperatures should increase it.
3-74
-------
These results point to the conclusion that most of the existing data on urban ambient ni-
trate concentrations are of doubtful validity. Furthermore, it is unlikely that any of these
data can be corrected even if mechanisms for artifact formation or nitrate loss are clarified
in the future, since HNO«, which appears to play a significant role in each mechanism, is not
routinely monitored.
3.3,4.4 Analysisof Trace Elements—Over the years, a variety of techniques have been applied
to analysis for elements. Presently, the most commonly used techniques use some type of
spectroscopic detection. By definition, these techniques respond only to the presence of the
elements in the PM and do not provide information concerning the chemical compounds present.
For the most part, techniques for analyzing trace elements have not provided information con-
cerning the oxidation state of the elements, although Braman et al. (1977) reported attempts
at such analysis for arsenic. ,
3.3.4.4.1 Atomic absorption spectrometry. Atomic absorption spectrometry (Morrison, 1965)
has been widely employed for quantitative analysis of a large number of elements in particles.
In principle, a beam of light of a wavelength that is characteristic of an electronic
transition for the element of interest is made to traverse a region of space with a constant
intensity and to impinge on a detector. The element of interest is atomized in a portion of
the beam of light. The amount of light absorbed by the atoms of interest in the sample can be
related to the amount of that element present. Any element can be determined if a lamp is
available to produce the characteristic light.
A variety of techniques can be used for atomizing and introducing the element into the
light beam. Generally, a flame or a heated carbon rod atomizer is used. Flame techniques are
most commonly used for atmospheric PM. An extract of the PM is prepared and aspirated into
the flame, which volatilizes the sample and produces a sufficiently large population of ground
state atoms for absorption. An example of this kind of application is the EPA reference
method for lead (U.S. Environmental Protection Agency, 1979d). If the concentration of the
element of interest is too low for flame application, however, or if an extremely limited
amount of sample is available, an electrically heated atomizer can be used to volatilize atoms
into the light beam. In this application, solutions can be used, or a small portion of soiled
filter without any other preparation may be examined directly. In the latter case the filter
substrate must be oxidizable and the representativeness of the sample may be questioned.
Properly applied, atomic absorption spectrometry is generally specific for the analysis
of the elements desired. The instrumentation can be inexpensive relative to other instru-
mental techniques for the analysis of trace elements and is generally available as standard
equipment in most analytical laboratories. However, it can only analyze one element at a
time. Additional elements must be determined serially. This can be a severe disadvantage
when a number of elements need to be determined on the same sample, both from the standpoint
of the resources required to obtain the information and the limitations of the volume of
3-75
-------
extract available to perform the analysis. Although the technique is useful, there are sig-
nificant interference problems in some determinations. Spectral interferences from other
elements absorbing at the same wavelength can be a problem but can usually be avoided by
judicious choice of wavelengths. However, interference effects on the element of interest
caused by other substances present in the material introduced into the spectrometer, refered
to as matrix effects, can be much more difficult to resolve. These effects can differ sig-
nificantly, ranging from the effect of viscosity on the amount of material which can be
atomized to effects due to the presence of refractory compounds containing the element of
interest, which may not be completely volatilized at the flame or atomizer temperature.
Matrix effects can adversely affect the concentration of atoms in the beam and result in
significant errors in the measurement of an element. The literature is replete with
discussions of these difficulties and provides both general and specific techniques for
overcoming them. One of the more convenient ways to keep abreast of developments in this area
is through the use of a continually updated bibliography with convenient indexing, such as
that provided by "Atomic Spectroscopy." (S. Slavin, ed.) Bibliographies for this publication
are routinely updated and appear each January and July.
3.3.4.4.2 Optical emission spectrometry. Optical emission spectrometry is a technique which
can simultaneously determine the amounts of numerous elements. The advantage of this
technique are obvious in situations of limited sample availability or limited time and
resources with which to do a measurement. Conventional arc or spark-excited optical emission
spectrometry has been used extensively on atmospheric PM (Scott et a!., 1976). In most
applications of this technique an extract of the PM is excited by a spark or arc discharge.
This decomposes any substances present and excites the atoms to other than their ground
electronic states. In the de-excitation to the ground state, light is emitted at a character-
istic wavelength. The intensity of the light emitted is an indication of the quantity of the
element present. Most conventional optical emission spectrometers are capable of simul-
taneously analyzing 20 to 30 elements.
Conventional arc or spark-excited optical emission spectrometers were never very popular,
partly because of cost, partly because the photographic readout was complicated and gave
rather approximate answers, and partly because of high detection limits for a number of
species. The development of optical emission spectrometers based on plasma excitation
(Boumans and DeBoer, 1975} has resulted in significant improvements. Although there are
several kinds of plasma excitation, the commercially available optical emission spectrometers
with inductively coupled argon plasma excitation has proven most advantageous. In this
technique, an extract of the sample of PM is aspirated into an inductively coupled argon
plasma whose very high temperature decomposes the materials and excites the atoms. The light
emitted when these excited species fall back to the ground state is collected and monitored
just as before. However, this approach has numerous advantages not available with the older
excitation techniques. The technique is capable of using the same acid extract used in atomic
3-76
-------
absorption. It is more free of matrix effects than atomic absorption; it requires,, for. a
single multielement determination on a given sample, about the same amount of time and
solution as a single element determination on atomic absorption; it usually has a much wider
linear range than atomic absorption; and it has detection limits equal to or lower than flame
atomic absorption (Fassel, 1978). If an acid extract of atmospheric PM is to be analyzed,
inductively coupled argon plasma optical emission spectrometry is usually the technique of
current choice.
3.3.4.4.3 Spark source mass spectrometry. Spark source -mass spectrometers are high-
resolution magnetic sectoring mass spectrometers which usually use photographic, emulsion
detection and very high resojution densitometry for quantitative analysis. They are uncommon
and very expensive. The material to be analyzed is incorporated into two small (usually
graphite) electrodes, which are placed in the spectrometer with a well-controlled gap between
them. A spark is passed across the electrodes, vaporiz-ing them and ionizing the material, in
them. The ions are subsequently led into; the mass analyzer portion of the spectrometer
(Ahern, 1972).
The electrodes used in the spark source mass spectrometer can be fabricated either fr,om
PM that has been separated from a filter or. an extract of the PM. The technique 'is not
suitable for the generation of large data bases because only a few samples can be analyzed-on
any given day. The time required to prepare the instrument and to obtain a set of spectra
necessary for quantitation is substantial. Double ionization of elements is common and so are
ionized oligomers of carbon. Therefore, high resolution, detection and complex interpretation
are the rule rather than the exception. The advantage of spark source mass spectrometry is
that it can simultaneously estimate the quantity of all nonvolatile-elements in the periodic
table and can do so with roughly equal sensitivity. „
3.3.4.4.4 Neutron activation analysis. Neutron activation analysis (Morrison, 1965) implies
a variety of distinct procedures, all of which produce unstable atomic nuclei .which then,emit
high-energy radiation or particles. The intensity of a specific emission and its energy are
monitored as an indicator of. the element and its quantity.
Instrumental thermal neutron activation analysis is the technique most commonly applied
to atmospheric PM. With this approach, a nuclear reactor is used to produce neutrons, which
bombard the samples and produce the unstable nuclei. The emitted gamma radiation is detected
by a Li-drifted Ge detector whose output is processed to produce the gamma-ray spectrum of the
irradiated PM. The method has low detection limits, can simultaneously determine up to,about
25 elements in a given sample, and particulate matter can be analyzed directly as received,-on
a very small portion of the filter surface. The-technique has been successfully applied with
the glass fiber used in hi-vol samplers (Lambert e± a!., 1979). The 'time required for
analysis is short. However, data are usually-not available until 2 to 3 weeks after the sample
3-77
-------
is irradiated because there is a significant delay period between the irradiation and the
collection of gamma-ray spectra for certain long-lived isotopes. Some important elements
(e.g., S and Pb) are not practically measured using this method. These limitations and the
need for very complex, highly specialized and expensive equipment are the main disadvantages
of the neutron activation technique.
3.3.4.4.5 X~ray f1uorescencespectrpmetry. X-ray fluorescence spectrometry is a multi-
element, nondestructive technique that can simultaneously determine numerous trace elements in
PM, directly on the filter media. It involves the excitation of tightly bound electrons in
the atoms by an X-ray generator arjd observation of the X-ray emissions that occur as the
de-excitation of the electrons proceeds (Dzubay, 1977).
X-ray fluorescence spectrometers use either energy-dispersive or wavelength-dispersive,
detection. Spectrometers using energy-dispersive detection simultaneously collect all emitted
quanta with a silicon-lithium detector and, through subsequent processing, can analyze about
30 elements. Spectrometers using wavelength-dispersive detection monitor carefully
preselected wavelengths that are characteristic of the de-excitation emissions of the,elements
of interest. With wavelength-dispersive detection, about 20 elements can be determined simul-
taneously on a single sample and interelement effects are minimal due to the high resolution
capability of the instrumentation. With energy-dispersive detection, all wavelengths are
simultaneously collected and interelement corrections must be handled in the data reduction
process.
Good detection limits and the ability to handle a sizable number of samples nondestruc-
tively with minimal sample preparation are clear advantages of the X-ray fluorescence
technique. In order to analyze the sample directly, however, it must be of uniform surface
texture, and it is best if the particulate layer is very thin. This obviously places some
limitations on the kind of sample that can be analyzed without preparation. Even in the most
ideal samples, concern with special corrections must exist (Gould et al., 1976). The
techniques has been applied extensively to analysis of filters from dichotomous samplers
tDzubay and Stevens, 1975). ~
3,3.4.4.6 Electrochemical Methods. Electrochemical methods have been used to a limited ex-
tent to determine a small number of elements in airborne particles. These methods include
potentiotnetry with ion selective electrodes, polarography, and anodic stripping voltammetry
(Morrison, 1965). Electrochemical techniques have few advantages for airborne particulate
analysis, aside from their low initial capital equipment cost compared to, other techniques.
While the methods are usable (Ryan and Siemer, 1976) there appears to be fairly little use of
such techniques at present, except in the area of ion-selective electrodes.
3.3.4.4.7 Chemical methods—In the past, many classical wet chemical procedures were employed
for trace element analysis of airborne particles. In general, a col or-forming reagent was in-
volved. The amount of the given element present is determined by the extent of color develop-
ment. Perhaps the best known of these procedures is based on the use of diphenylthiocarbazone
3-78
-------
(dlthizone) as the colorlmetric reagent for lead (Snell, 1978). Wet chemical procedures are
labor intensive and slow compared with spectral techniques. Sample preparation and interfer-
ences are also usually a problem. As a result, laboratories with heavy sample loads tend to
use instrumental methods, particularly spectral techniques,
3.3.4.5 Analysis of Organic Compounds—Numerous papers have appeared dealing with the
characterization of organic compounds in airborne particles. The following discussion was
taken primarily from the monograph edited by H. Malissa (1978) and describes the principal
methods used in this field and some typical examples mentioned in the literature.
Organic compounds significantly contribute to total PM in urban aerosols. Organic
contents of up to 43 percent of the total particle mass have been reported (Hidy, 1975).
Characterization of organic compounds in ufban aerosols generally involves trace separation
and identification by gas and liquid chromatography, with detection methods having sensitivity
in the nanogram range. The sample amount needed to allow analysis of substances in parts per
million concentrations is in the milligram range (Cautreels and Van Cauwenberghe, 1976;
Ketseridis, et a!., 1976). Usually, high-volume samplers with glass fiber filters are used to
provide the needed sample sizes.
One of the earlier simple and extensively used methods for estimating the organic content
of PM was called "benzene soluble organics." Filters were simply refluxed with benzene in a
soxhlet extractor for several hours. The benzene was evaporated, and the weight of the
residue was measured and reported. Benzene soluble organic data were recorded in the National
Aerometric Data Bank for many years. Because of the purported hazard of benzene, this method
has not been used on a national scale by any single laboratory for roughly a decade.
Extraction efficiencies of 25 different solvents and 24 binary mixtures were investigated by
Grosjean (1975). Grosjean determined that extraction with benzene or other nonpolar solvents
usually leads to serious underestimation of aerosol organics, especially of the polar
secondary (photochemical) products like carbonyl compounds, organic nitrates, or carboxylic
acids. The use of binary mixtures for extraction or a nonpolar and a polar solvent for
successive extractions was strongly recommended. This leads to a higher organic carbon
•i
extraction efficiency (in comparison to benzene as solvent) than with both single polar and
nonpolar solvents.
In the area of compound-specific analysis, a large amount of early work techniques
focused on the measurement of polycyclic aromatic hydrocarbons (PAH). Numerous measurement
techniques have been proposed to analyze quantitatively for many of the polycyclic aromatics.
Chromatographic techniques (Sawicki, 1964; Thomas et al., 1960) were used in much of the
earlier work; more recently, frozen solution fluorimetry (Bacon et al., 1978) and matrix
isolation spectroscopy (Wehry and Mamantov, 1979) have been explored. High-pressure liquid
chromatography (HPLC) is a promising technique for separation of high molecular weight PAH.
The development of bonded octadecylsilyl (ODS) columns of micro particle size allowed Fox and
3-79
-------
Stal.ey (1976) to accomplish the near baseline separation of the benzo[a]pyrene (BaP) arid its
isomer, benzo[e]pyrene. A significant increase in sensitivity over other methods was achieved
by use of fluorescence spectroscopy for on-line detection. Perhaps the most extensive data
base has been concerned only with BaP (Swanson et a!., 1978) utilizing a thin layer
chromatographic technique with fluorescence detection. Gas chromatographic (GC) separation of
organic extracts of airborne particles requires the application of preseparation steps, such
as thin layer chromatography (Zoccolillo et a!., 1972) or liquid-liquid extraction (Cautreels
and Van Cauwenberghe, 1976; Ketseridis et al. , 1976). Primary extraction is generally carried
out by means of single solvents such as benzene, cyclohexane," or others. A typical procedure
including solvent extraction for preseparation is described by Ketseridis et a1.(1976).
The application of gas chromatography coupled with mass spectrometry for the analysis of
benzene"extractable compounds in airborne particles is described in detail by Cautreels and
Van Cauwenberghe (1976). This work led to the identification of more than 100 compounds in
urban aerosols. The benzene-extractable compounds (5.8 percent of total particles) were sepa-
rated into neutral, acidic, and basic substances. The acidic fraction was converted to the
methylated derivatives for GC analysis. In the neutral fraction, 22 saturated aliphatic
hydrocarbons, 36 polynuclear hydrocarbons, and 13 polar oxygenated substances were identified.
In the acidic fraction, 19 fatty acids and 19 aromatic carboxylic acids were identified. In
the basic fraction, 15 peaks of nitrogen-containing analogs of the PAH were identified.
Interest in the organic content of atmospheric particles ranges from particulate carbon
(Rosen and Novakov, 1978) to any other possible organic substance. A variety of techniques
have been used in an effort to solve this problem (Fox and Jeffries, 1979). However, it is
clear that this area is large, exceedingly complex, and will need a great deal of develop-
mental effort.
3.3.4,6 Analysis of Total Carbon and Elemental Carbon—The most 'widely used technique for
measurement of total carbon in collected PM is by combustion to carbon dioxide followed by
detection of this gas; or, after C0? to CH* reduction, of an equivalent quantity of CH,.
Several commercial instruments are available for this purpose. Johnson and Huntzicker (1979)
have designed an instrument for analysis of organic and elemental carbon o'n filter samples of
ambient particles. The organic carbon is volatilized or pyrolyzed in an inert atmosphere,
oxidized to C02, and converted to methane, which is detected with a flame ionization detector.
Elemental carbon is determined after oxidation to C0? and chromatographic separation of the
CO- from 0-. Cadle et al. (1980) have discussed a similar system using non-dispersive infra-
red detection of the C0? and automation of the analysis procedure.
Nondestructive analysis of carbon in collected PM involving the interaction of light with
the particles (i.e., changes in reflection, transmission, and absorption of the filter) is
used to estimate elemental carbon loading. Delumyea et al. (1980) used a reflectance tech-
nique having a tungsten filament lamp as a light source. Lin et al. (1973) described an
®
apparatus to integrate light scattered by particles collected on Nuclepore filters. Since
3-80
-------
elemental carbon is the major absorbing species in ambient PM, this technique has been used by
other researchers to estimate the elemental carbon content of collected PM. Yasa et al.
(1979) have shown a linear relationship between photoacoustic measurements and optical
attenuation measurements of the type done by Rosen et al. (1978), to establish that the
optically absorbing component of urban aerosol particles is graphitic carbon. Photoacoustic
measurements have also been used for in-line measurements of diesel particulate emissions
(Faxvog and Roessler, 1979). (See Chapter 9 for a more thorough discussion of the role of
light absorbing particles in visibility reduction.)
Macias et al. (1978) used the gamma-ray analysis of light elements (GRALE) technique to
measure carbon in ambient aerosols deposited on quartz or Teflon filters. The GRALE technique
involves the in-beam measurement of gamma rays emitted from an aerosol sample during the
inelastic scattering of 7-MeV protons accelerated in a cyclotron.
Calcium carbonate also is commonly found in airborne particles. Mueller et al. (1971)
described a technique to distinguish carbonates from elemental carbon by acid evolution of
co2.
3.3.5 Particle Morphology Measurements
Visual examination of particles collected on a filter or impaction substrate can provide
extremely useful information concerning the sources and transport of airborne particles. A
reticle-equipped light microscope can be used to examine particles larger than about 0.5 urn.
Use of transmission and scanning electron microscopes can improve the resolution for particles
as small as 0.001 (jm. The effective ranges of microscopes and their utility are described by
McCrone (1973) and shown in Appendix Figure 3A-13. Particle size distributions by number can
be generated using statistically valid counting procedures. By applying an average density an
estimate of the size distribution by mass can be made.
Microscopic identification and analysis requires a high degree of skill and experience
plus extensive quality assurance to provide meaningful information. Critical factors in
effective use of these methods are the selection of sampling substrates, allowable particle
loadings, and sample handling. In addition, particle interactions and structure changes on
the collection surface must be minimized if accurate size distributions and characterizations
are to be obtained. In a study of ambient particles collected on hi-vol filters, Bradway
et al. (1976) examined the ability of multiple microscopists to characterize particles in
specific categories. Significant misidentification and misassignment problems were noted,
which made it difficult to compare results. Multiple microscopists and blind replicates were
recommended as standard procedures for quantitative optical characterization studies.
3.3.6 Intercomparison of PM Measurements
The intercomparison of particle sampling methods is not straightforward because of the
complex nature of ambient particulate matter. As noted earlier in this chapter, gravimetric
mass measurement methods can differ dramatically in the particle size ranges collected and the
3-81
-------
sensitivity of the sampler to external factors, such as windspeed. Also, nongravimetric mass
measurement methods examine only specific portions of the particle size spectrum and, in
addition, measure selected integral properties of particles rather than mass.
Comparison of gravimetric mass measurement methods can be made in the field through
collocated sampling or in the laboratory by examining the sampling effectiveness of the inlet
or substage for various particle sizes. Because it is difficult to simulate the character of
real suspended particles, however, the final intercomparison test must be performed at
selected field locations. The choice of the number and types of field locations is important,
since the local particle sources can have a substantial impact on the sampler performance,
especially if coarse particles dominate the size distribution.
The most recent and comprehensive intercomparison of gravimetric mass measurement
particle samplers was reported by Camp et al. (1978). Eleven different types of samplers were
compared for mass and other analyses including sulfates, nitrates, and elemental composition.
The most salient observation of the study was the difficulty of intercomparison of the
samplers because of differences in inlet or substage particle size cutpoints. Since there is
no reference sampler, all measurements of the same size fraction were averaged as a comparison
measurement. Some duplicate samplers were present, permitting reproducibility measurements.
The coefficient of variation for the automated dichotomous samplers was determined to be 11
percent for the coarse fraction and 3 percent for the fine fraction. The same values for a
manual dichotomous sampler were 18 percent and 1 percent, respectively. The high-volume
impactor used in the CHAMP network gave values of 15 percent and 5 percent, respectively, for
the equivalent size fractions. The results from this study should be considered "best case,"
since the sampler operations were monitored continuously by highly skilled individuals. In
some cases, the operators were the developers of the sampling method. It is expected that
routine field sampling by less qualified personnel would produce larger variabilities. The
reproducibility of certain chemical analyses were reported to be better than the mass mea-
surements, such as elemental sulfur, which averaged ±3 percent for all size fractions.
Overall, the study showed that comparable results could be obtained by different particle
samplers if appropriate quality assurance steps were taken and identical size fractions were
compared.
Miller and DeKoning (1974) compared the TSP high-volume sampler with several commercially
available cascade impactors. None of the impactors gave results comparable to the high-volume
sampler, but the two types of samplers did correlate reasonabley well. The agreement among
cascade impactors for mass median diameter (HMD) was very poor. The HMD often differed by
more than a factor of 2.
Comparison of gravimetric mass measurements with indirect mass measurements should only
be attempted to determine correlation or to test a physical model relating the measurements.
The literature contains many intercomparison studies attempting to relate the TSP hi-vol with
surrogate techniques such as: the British Smoke Shade sampler (Commins and Waller, 1967; Lee
3-82
-------
et a!., 1972; Pashel and Egner, 1981); the integrating.nephelometer (Charlson etal., 1968;
Kretzschmar, 1975); and the AISI tape sampler (Lee et al., 1972; Ingram and Golden, 1973).
Comparisons are also available between other direct and indirect mass measurements, such as
the dichotomous sampler with the AISI tape sampler (Regan et al, 1979) and with the nephe-
lometer (Waggoner and Weiss, 1980). In many cases, a simple regression was fitted between
measurements rather than attempting to establish a physical basis for the comparison for test-
ing with empirical data. Most of these comparisons were attempts to demonstrate the useful-
ness of a nongravimetric sampling method for predicting the mass concentrations that would
have been measured by a gravimetric mass method. However, there is currently no indirect
technique that has gained acceptance as a general-purpose surrogate for direct mass concen-
tration measurements.
Mulholland et al. (1980) compared the estimated mass concentrations calculated from Elec-
trical Aerosol Analyzer (EAA) measurements with direct gravimetric analyses. It was noted
that the errors are in the ± 20 to 30 percent range for spherical particles, and for non-
spherical particles the errors are as high as ± 60 percent. Therefore, great care should be
taken in attempting to predict gravimetric mass concentration from nongravimetric particle
measurements.
3.3.7 Summary
Particulate matter suspended in ambi'ent air contains a range of particle sizes and
shapes. Separating particles according to aerodynamic size groups particles that behave alike
in more situations of interest to human health and welfare (other than visibility) than any
other measure of particle size. Samplers can be designed to collect specific size fractions
or match specific particle deposition patterns through carefully designed inlets and substage
fractionators. Mass concentration derived from gravimetric analysis is the most common
measure of PM. High-volume samplers, dichotomous samplers, cascade impactors, and cyclone
samplers are the most common examples of this type of measurement. Carefully collected size
distributions of ambient particle mass have shown that most particle samplers underestimate
the concentration of particles in the air because of external factors such as windspeed or
because their particle transport systems are not effective for the larger particle sizes.
Mass concentrations can be estimated using methodologies that measure an integral
property of particles such as optical reflectance. Empirical relationships between mass con-
centrations and the integral measurement have been developed and are used to predict mass con-
centration. Without a valid physical model relating the measurements plus empirical data to
demonstrate the model, these techniques have a limited ability to estimate mass concentra-
tions. These conditions are poorly met in the case of reflectance or transmission tape
samplers, fairly well met in the integrating nephelometer, and very well met in the case of
beta-ray attenuation analysis.
Sampling accuracy is difficult to determine directly, since the measurement requires the
production of very accurately known concentrations of particulate matter of a wide variety of
3-83
-------
sizes. Instead, accuracy is estimated by determining the accuracy of sampler flow rate
measurement or control, and inlet sampling effectiveness. These separate measurements provide
a means of intercomparing methods in the absence of a reference measurement technique. Recent
interest in larger particle sampler cutpoints (e.g., 15 pm) have resulted in wind tunnel test
procedures for particles that determine sampling effectiveness under controlled conditions.
Such measurements have added significantly to the ability to estimate particle sampling
accuracy.
Recent evaluations show that the high-volume sampler collects a smaller particle size
range than that reports in the 1969 criteria document (National Air Pollution Control Admini-
stration, 1969). The sampling effectiveness of the hi-vol inlet also is wind speed-sensitive
for larger (>10 pm) particles. Sampling biases caused by typical wind speeds could be
expected to cause no more than a 10-percent day-to-day variability for the same ambient
concentration. The hi-vol is one of the most reproducible particle samplers currently in
use, with a coefficient of variation of 3 to 5 percent. A significant problem associated with
the glass fiber filter used on the hi-vol is the formation of artifact mass caused by the
2
presence of acid gases in the air. These artifacts can add 6 to 7 pg/m to a 24-hour sample.
The dichotomous sampler was designed to collect the fine and coarse aiftbient particle
®
fractions, usually providing a separation at 2.5 pm. This sampler uses Teflon filters to
minimize artifact mass formation and is available in versions for manual or automatic field
operation. The earlier inlets used with this sampler were very windspeed dependent, but newer
versions are much improved. The dichotomous sampler collects submilligram quantities of
particles because of low sampling flowrate and requires microbalance analyses, but is capable
of reproducibilities of ± 10 percent, or better.
Cyclone samplers with cutpoints in the vicinity of 2 urn have been used for years to
separate the fine particle fraction. A recent development has coupled the cyclone sampler to
a sampler inlet to give a 15 pm cutpoint. Cyclone samplers can be designed to cover a range
of sampling flowrates and are available in a variety of physical sizes. A 10mm version is
available for personnel dosimeter sampling. Cyclone sampling systems could be expected to
have coefficients of variations similar to that of the dichotomous sampler.
The Size Selective Inlet (SSI) hi-vol collects samples containing particles less than 15
Miu for comparison with TSP. This sampler is identical to the TSP hi-vol except for the inlet
and is expected to have the same basic characteristics.
Cascade impactors have been used extensively to obtain mass distribution by particle
size. Because care 'must be exercised to prevent errors, such as those caused by particle
bounce between stages, these samplers are normally not operated as routine monitors. A
comparison of impactors showed inconsistencies in the MMD and in total mass collections
compared with the hi-vol.
Samplers that derive mass concentrations using analytical techniques other than direct
weight have been used extensively. One of the earliest was the British Smoke Shade sampler,
3-84
-------
which measures the reflectance of particles collected on a filter and uses empirical relation-
ships to predict mass concentration. These relationships have been shown to be more sensitive
to carbon concentrations than mass; hence, they are very difficult to interpret as total mass.
Another optical technique, the AISI tape sampler, uses transmittance instead of reflectance to
predict mass. It has been shown that this sampler correlates favorably with gravimetric
measurements limited to the smaller particle sizes. Several researchers have also reported
good correlation between the integrating nephelometer and gravimetric fine particle mass. The
EAA, however, was shown to have difficulties in reliably predicting gravimetric mass measure-
ments. - -
Optical particle morphology techniques are very useful for identifying the character and
sources of collected particles. Researchers have noted, however, that these techniques are
dependent on the skill of the microscopist and stressed the need for careful quality assurance
procedures.
An extensive number of analytical techniques are available for analyzing particles col-
lected on a suitable substrate. Many of the analytical techniques, such as those for ele-
mental sulfur, have been demonstrated to be more precise than the analyses for gravimetric
mass concentration. Methods are available to provide reliable analyses for sulfates, ni-
trates, organic fractions, and elemental composition (e.g., sulfur, lead, silicon). Not all
analyses can be performed on all particle samples because of factors such as incompatible sub-
strates and inadequate sample size. Misinterpretation of analytical results can occur when
samples have not been appropriately segregated by particle size and when artifact mass is
formed on the substrate rather than collected in a particle form. Positive artifacts are
particularly likely in sulfate and nitrate determination, and negative nitrate artifacts also
occur.
Sampling technology is available to meet specific requirements, such as providing sharp
cutpoints, cutpoints which match particle deposition models, separate collection of fine and
coarse particles, automated sample collection capability, collection of at least milligram
quantities of particles, minimal interaction of the substrate with the collected particles,
ability to produce particle size distribution data, low purchase cost, and simple operating
procedures. Not all these sampling requirements may be needed for each measurement study.
Currently, there is no single sampler which meets all requirements, but samplers are available
that can meet most typical requirements.
3.4 MEASUREMENT TECHNIQUES FOR ACIDIC DEPOSITION
3.4.1 Introduction
Studies designed to monitor precipitation first appeared in the literature around the
turn of the century. Many small-scale networks were organized in the United States and Europe
between the 1920's and the 1950's. Knight (1911), Wilson (1926), and Collison and Mensching
(1932) reported the earliest U.S. precipitation chemistry studies at local sites. The physi-
cal size of the networks changed during the 1950's from single or dual site studies to large-
3-85
-------
scale national/International studies. During the 1950's, Barrett and Brodin (1955) organized
a European monitoring network and Junge and Gustafson (1956) established the first continental
U.S. network. The World Meteorological Organization (WHO) reported (1971) the formation of an
international network to monitor global trends and changes in the chemical composition of acid
precipitation. In addition to the existing WHO network, various local, regional, and national
acid precipitation studies currently operate in the industrialized nations.
Ecologists and biologists were among the first scientists to become concerned with the
causes and effects of acid precipitation. These two groups were responsible for most of the
acid precipitation studies conducted before 1970. During the period 1974-1976, multidiscipli-
nary study of acidic precipitation phenomena expanded. International scientists from every
scientific field focused their attention on the potential affects of acidic precipitation.
Symposia, committees, and groups were organized to examine every" facet of past and on-going
studies (Dochinger and Seliga, 1976; Kronebach, 1975; Likens et al., 1972; Likens, 1976).
Network siting, sampling procedures and analysis schemes were critically reviewed. Special
committees were formed to develop techniques to describe statistically the quality of the pre-
cipitation chemistry data being reported by the various international laboratories. During
this period, concern about the effects of particulate deposition on vegetation increased. As
a result, improved wet/dry collection devices were designed to give a better understanding of
total acidic deposition problems.
Many new studies resulted from this increased emphasis on acidic deposition, including
the development of a long-term continental U.S. monitoring network (Galloway and Cowling,
1978). Comprehensive reviews of past and current studies aife provided by Niemann et al.
(1979); Kennedy (1978); and the U.S. Atomic Energy Commission (1974).
3.4.2 U.S.Precipitation Studies
Past U.S. precipitation chemistry studies can best be described as ad hoc (U.S. Atomic
Energy Commission, 1974). New studies were randomly developed without adequate consideration
of either past or current proposals. General characteristics of past U.S. studies include:
1. Overall study objectives varied among projects.
2. Pre-1970 studies were short-term, lasting only 1 to 2 years.
3. Sites were randomly selected at locations, of convenience. Siting with respect to
program objectives or standard siting criteria were rarely considered.
4. Sampling/storage procedures and the extent of sample chemical analysis varied among
studies.
Each of these deviations in study design/protocol obviously affects the existing data and
precludes any simplistic consolidation or correlation of past study results. Of these varia-
tions, the differences in sampling/storage procedures are the most difficult to resolve. In
general, sampler collection efficiencies, sample representativeness, and sample integrity at
the time of chemical analysis (i.e., does the sample reflect what was collected in the field
or have chemical changes occurred?) can only be speculative for the earlier studies.
3-86
-------
Since 1900, techniques used to collect and store precipitation samples have changed. The
most significant change in wet-only or bulk precipitation sampling is in the collector itself.
In earlier reports, monthly bulk samples were manually collected in glass devices. Plastic
devices became the collection medium during 1950-1960. Galloway and Likens (1976) note that
plastic collectors are preferred today for inorganic species, whereas glass collectors are
currently used for organic sampling. Automated wet/dry samplers are replacing the manual
wet-only and bulk collection techniques. The height of the collection device above ground
level has varied throughout the years. Earlier studies placed collectors at ground level
while current studies commonly place the collector 1 to 3 meters above ground level. Other
study-dependent sampling variations observed in wet-only and bulk sampling procedures include:
filtering versus non-filtering of the sample before chemical analysis; the use of biocides to
preserve the sample and retard biological growth; and the storage techniques used after sample
collection and before chemical analysis. Although the degree to which each of these sampling
and storage differences affect the sample remains unanswered, it is generally agreed that the
study data will be related to the study sampling techniques employed.
When reviewing past studies, the extent to which these common variations biased the
resulting data must be determined. For example, the effects of sample evaporation, splash
contamination, loss of initial (usually most concentrated) rainfall, and contamination from
insects, leaves, etc., have been commonly reported. How the authors addressed these problems
differ. Some deleted the questionable data, others did not, and still others stated that
these effects would be averaged out over the length of the study.
The data analyst must also know if the samples were filtered before analysis. Past study
data indicates that the inclusion of particulate "wash-out" material in bulk and wet-only
samples as well as dry deposition in bulk samples changes the overall chemistry of the sample.
Several studies routinely filtered the sample before analysis. ' This filtration of particulate
matter, depending on technique and actual time when filtering occurred, could possibly change
the resulting sample chemical composition and related analytical data. Whether glass or plas-
tic collection devices were used could also affect the data. Galloway and Likens (1976) note
the leaching of inorganic species into and out of glass collection devices and the loss of
organic species with plastic devices. Other authors (Kadlecek and Mohnen, 1976; Norwegian
Institute for Air Research, 1971) report similar findings. Metal ion losses in dilute samples
have been repeatedly reported. To minimize this potential metal loss, various authors
acidified a representative aliquot of the sample immediately following collection. Again,
others did not. Sample storage techniques also vary among studies. Larger networks usually
kept the sample in a cool, dark place. Some smaller networks either froze the sample or
refrigerated the sample at 4°C. Galloway and Likens (1976) indicate that significant changes
do not occur when samplers are stored at 4°C. Unfortunately, this storage technique is cost
prohibitive for large national/international networks. How long a sample is stored before
3-87
-------
chemical analysis may also affect its integrity. A review of the literature reveals that the
reported length of time from sample collection to sample analysis has varied by as much as 60
days. Some authors used chemical biocides to retard green algae growth in samples collected
in warmer climates because the presence of algae changes the sample chemical composition.
Obviously, the storage technique affects sample integrity at the time of chemical analysis.
Although procedures for collection of dry deposition are not as well documented, similar
sampling variations are expected depending on technique (i.e., dust-fall buckets, ambient air
monitors, or automated dry-only collection devices). Siting of the dry deposition collector
is crucial; in particular, the height of the collector above ground level. Dry deposition
samples 0 to 1 meter above ground level have been reported to be heavily influenced by contri-
butions from the local terrain, as well as bird and vegetation contamination.
Data from special studies is also available for summation. Behrmann (1975), Pickerel!
et al. (1979), Anderson (1978), Cooper et al. (1976), and Gatz et al. (1971) report special
one-of-a-kind samplers to monitor the change in the chemical composition of precipitation
events at single sites. Special sampling procedures have been developed for the collection of
snow (Galloway and Likens, 1976; Hagen et al., 1973), fog (Waller, 1963; Mrose, 1966), indivi-
dual raindrops and canopy throughfall (McColl and Bush, 1978). Understanding these special
sampling techniques is essential before these data are compiled or summarized.
The data analyst must also consider the various collection periods reported in past
studies. Over the years, bulk samples have typically been collected on a monthly basis.
However, wet deposition collection periods have ranged from event sampling to monthly
sampling. Although clearly defined in terms of showers and thunderstorms, the definition of
the beginning and ending of an event during a large frontal system varies from author to
author. Monthly sampling is common in larger networks designed to monitor the trends in
chemical composition over time. • Daily or more frequent sampling is typically reported at
individual sites with the objective to determine exact chemical loadings at specific sites.
Weekly sampling is currently recommended by Galloway and Cowling (1978) as the minimum
allowable sampling frequency to obtain usable wet deposition results. Dry deposition is
commonly collected on a 1-to 2-month basis, as recommended by Galloway and Cowling (1978).
Each of the sampling and storage variations addressed above can affect the sample
integrity and resulting data. In addition, site meteorology and collector efficiency also
affect the sample. Summers and Whelpdale (1976) stress the existing need to document both
scavenging and collection mechanisms involved with acid precipitation. Initial reviews of the
most commonly used collectors have been conducted and are provided by Niemann (1979) and
Galloway and Likens (1976). Additional comprehensive collector evaluations, including dry
deposition collector efficiencies and species collected, along with a reevaluation of the
meteorological mechanisms involved in acidic precipitation processes are needed. Before any
past data summarization can be developed, a careful analysis of the sampling and storage pro-
cedures and collection mechanisms must be performed.
3-88
-------
3-4.3 Analytical Techniques
3.4.3.1 Introduction—Analytical methodologies currently employed to analyze precipitation
samples are, for the most part, state-of-the-art freshwater or natural water analytical tech-
niques. Typical procedures are presented in Methods of Air Sampling and Analysis (Interso-
ciety Committee, 1977a), Standard Methods for the Examination of Water and Waste Water (Ameri-
can Public Health Association, 1971) and U.S. Environmental Protection Agency's "Methods for
Chemical Analysis of Water and Waste" ( U.S. Envrionmental Protection Agency, 1979d). Theore-
tical detection limits and the quality of the data in terms of precision/accuracy for each
technique are specified in the literature. Rainwater, however, is a dilute solution of chemi-
cal species, and represents extremely pure freshwater or natural water. Chemical analyses on
rainwater yield results at or below the published analytical detection limits (Miller and High-
smith, 1976). Added precautions must be taken to minimize field/laboratory contamination
(Likens, 1972), which is analytically indistinguishable from, and sometimes larger than,
natural rainwater species contributions. To preclude changes in the chemical composition of
the sample, analyses should occur within 24 hours after sample collection. Although operator
and instrumentation dependent, laboratory analyses should meet or exceed the analytical preci-
sions and accuracies presented in Table 3-6. Operator and instrumentation biases must be mini-
mized through supporting internal and external quality assurance programs. Before 1975,, no
mechanism existed to externally evaluate the quality of the precipitation chemistry data being
reported by the international precipitation laboratories. WHO instituted such an interna-
tional quality assurance program in 1975. Potential errors in past data have subsequently
been noted by Ridder (1978), Galloway et. a!,, (1979), Tyree et al. (1979), and Tyree (1981).
3.4.3.2 Analysis of Acidic Deposition Samples
3.4.3.2.1 Sample preparation. Wet deposition samples are allowed to equilibrate to room
temperature before chemical analyses. Sample pH and conductivity are initially measured.
Filtering or centrifuging of the sample may follow. A representative portion of the sample
may then be acidified (HN03) to preserve the metal ion concentrations. Between analyses,
samples are either stored in a dark, cool place or at 4°C. Dry deposition samples are dis-
solved with a known quantity of distilled water (typically 50 ml). Analytical procedures for
these dissolved dry deposition samples are identical to the wet deposition analytical proce-
dures described below.
3.4.3.2.2 Volume. Direct volume measurements, accuracy + 3 percent, are made on the wet depo-
sition sample with Class A graduated cylinders. Care must be taken to ensure that t,he sample
is not contaminated by the labware used in this procedure. Indirect volume techniques include
weighing the collection container before and after the sampling period or. measuring the col-
lection in a standard rain gauge (cylinder, tipping bucket, or weighing). Standard rain gauge
accuracies are + 0.02 in or better, depending on the manufacturer. The weighing rain gauge is
3-89
-------
TABLE 3-6. RECOMMENDED PHYSICAL/CHEMICAL PARAMETERS FOR ANALYSIS
Parameter
Volume (field)
pH (field + lab)
Conductivity
(field + lab)
so4
N03
Cl
NH4
K
Ca
Na
Mg
Acidity
Unit of
report
inches
PH
|jS/cm
mgS/1
mgN/1
mgCl/1
mgN/1
mgK/1
mgCa/1
mgNa/1
mgMg/1
ueq/1
Expected
range
0.00-10.00
2.00-9.00
0.1-200.0
0.1-10.0
0.1-10.0
0.1-10.0
0.05-10.0
0.01-5.00
0.01-5.00
0.01-10.00
0.01-2.00
1.0-500.0
Suggested
precision
±.02"
±.01pH
±5%
±2%
±2%
±2%
±3%
±2%
±2%
±5%
±2%
±10%
Suggested
accuracy
+.02"
±.01pH
±5%
±2%
±2%
±2%
±2%
±2%
±2%
±3%
±2%
±5%
preferred since it offers minimal evaporation loss and a higher degree of reliability over the
tipping bucket rain gauge during intense storms.
3.4.3.2.3 jaH. The pH is the measurement of the hydrogen ion activity (i.e., pH = -log [H+]),
commonly referred to as the free acid content of the solution. The pH of a typical United
States wet deposition sample ranges between 4.0 and 5.0 pH units. Samples collected in more
arid regions may range as high as pH 8.0, whereas samples collected in the Northeastern United
States typically range from pH 3.5 to pH 4.5. Precipitation pH measurements were not reported
prior to 1962. Instead, methyl orange indicator solution (endpoint pH = 4.4) (Skoag and West,
1965) was added to assess the sample acidity. Cogbill and Likens (1974) and Granat (1972)
indirectly calculated and reported the pH's of precipitation samples taken prior to 1962.
Currently, pH is electrometrically determined with a standard pH meter in conjunction with
either glass/reference calomel electrodes or a combination electrode at 25°C. Three certified
buffer solutions are used to calibrate the pH meter/electrode system in the pH range 3.50 to
7.50. Direct pH measurements are made on a representative portion of the sample. Measure-
ments of pH are dependent on operator techniques and the condition of the electrode(s).
Galloway et al. (1979); Ridder (1978); and Tyree et al. (1979) noted potential sources of
3-90
-------
errors in pH measurements. With proper care and quality equipment, pH results with a pre-
cision of ± 0.02 pH units and accuracy of ± 0.05 pH units should be obtainable (American
Public Health Association, 1971).
3.4.3.2.4 Conductivity. The specific conductance of a wet deposition sample indicates the
capability for that sample to carry an electrical current. Conductivity is related to the
total concentration of dissolved ions, is directly determined on the sample with a standard
Wheatstone bridge in conjunction with a calibrated conductivity cell at 25°C. Wet deposition
sample conductivities normally range from ca. 10 to 200 uS/cm. Daily calibration of the con-
ductivity meter and cell with freshly prepared standard KC1 solution is required for accurate
measurements. Operator techniques, the condition of the conductivity cell, and the quality of
the standard KC1 solutions determine the quality of data reported. Under careful supervision,
conductivity measurements with precision and accuracy of 5 percent are obtainable (American
Public Health Association, 1971).
3.4.3.2.5 Acidity. An acidity measurement indicates the capacity of the wet deposition
sample to donate protons from both strong and weak acids. Numerous techniques have been
reported to measure sample acidity. Each technique can yield a slightly different result
(Tyree, 1981). Acidity values from -200 ueg/liter to +200 ueg/liter are routinely reported.
The precision and accuracy of any acidity measurement dependents on the analytical technique
employed and the ability of the operator to standardize and titrate minute quantities of
highly diluted strong base. Tyree (1981) notes that the operator is the key to good acidity
measurements. Both the presence of substances such as carbon dioxide, aluminum, iron, and
ammonium, and the method endpoint pH influence the results.
3.4.3.2.5.1 p_H. See Section 3.4.3.2.3. Depending on the actual sample pH, acidity
based on a pH measurement yields either the strong acid proton component (sample pH <4.5) or
the strong acid component plus some undetermined contribution from the weak acid component
(sample pH >4.5). Acidity based on pH alone is not considered conclusive.
3.4.3.2.5.2 Titrimetric. Various titrimetric endpoint procedures are available. ,,
Phenolphthalein Endpoint. The rainwater sample is titrated with standardized 0.01 to
0.001 N strong base (NaOH) to the phenolph\halein endpoint (8.0 to 9.6 pH units per Skoag and
West, 1965). Precision and accuracy (+ 5 percent) depend on the ability of the operator to
standardize and deliver minute quantities of dilute base and the repeatability of the endpoint
color (American Public Health Association, 1971).
WHO. A prescribed quantity of strong acid (H^SO») is added to the sample, lowering the
immammr-r- £ i\,
sample pH to less than 4.0 and thus removing the C02- The sample is titrated against standar-
dized base until the sample pH, monitored via a pH meter, reads pH 5.6. Precision (+ 0.02 pH
units) and accuracy (+ 0.05 pH units) are dependent on the quality of the pH meter and stan-
dardized base, the condition of the electrodes, and operator technique.
AmericanPub!ic Health Association. Strong acid is added to lower the sample pH below
4.0. Hydrogen peroxide is then added. The sample is boiled to eliminate COg, cooled to room
temperature and titrated electrometrically with strong base to pH 8.3 (EPA method endpoint =
3-91
-------
8,2). The U.S. Environmental Protection Agency (1979) reports a standard deviation of 1 to
2 mg CaCO,/liter and bias of ca. + 20 percent for sample measurements in the 10 to 120 mg
CaCQ,/liter range.
Likens (1972). Nitrogen is bubbled through the sample to eliminate any CO, interference.
Samples are titrated to pH 9.00 with standard base. The accuracy of the pH meter is reported
to be + 0.03 pH units (1972). Following this technique, Hendry and Brezonik (1980) titrated
samples to pH 7.00 endpoint.
Coulometric/Potentiometric. The sample is titrated with cathodically generated hydroxyl
ion (OH ) [i.e., (-) reference electrode/test solution/glass electrode (+)] as outlined by
Liberti et al. (1972) and Askne et al. (1973). Gran plots (Gran, 1952) are interpreted to
determine the strong and weak acid contributions. Liberti et al. (1972) report a + 5 percent
standard derivation and 0.1 mg/liter ^SQ. sensitivity when analyzing strong acids. The
Norwegian Institute for Air Research (NAIR) (1971) reported a 2 to 5 ueq acidity/liter stan-
-4 -5
dard deviation in rainwater samples having 10 -10 acidity concentrations. Askne et al.
(1973) observed "exact agreement" in strong acid analyses, but only "reasonable-to-good
agreement" with samples containing various concentrations of strong and weak acids. Tyree et
al. (1979) and Krupa et al. (1976) also observed difficulties in determining strong/weak acid
contributions in rainwater samples using this technique.
3.4.3.2.5.3 Ion Balance. Granat (1972) and Cogbill and Likens (1974) reported a tech-
nique to calculate sample pH's based on the total ionic strength. In this technique, the
charge difference in favor of the anion concentration is related to the sample hydronium ion
concentration. Possession of accurate analytical data for the individual principal ions is
essential. The overall precision and accuracy of this technique is no better than the summa-
tion of precisions and accuracies of the analytical methods used to determine the individual
ionic species. Tyree et al. (1979) states that this technique could possibly be used to de-
termine the strong acid contribution in samples with observed pH 5.6 or below.
3.4.3.2.6 Sulfate (S0^.~). Analytical procedure for sulfate analysis are described in Section
3.3.4.1.1. Typical wet deposition samples contain 0.1 to 5.0 mg SO."/liter.
.1 '
3.4.3.2.7 Ammonium (NH. ). Ammonium concentration of 0.1 to 1.0 mg N/liter are normally
observed in wet deposition samples. Two techniques (ion selective electrode and indophenol
colorimetry) are discussed in Section 3.3.4.2. Manual Nesslerization techniques (American
Public Health Association, 1971) are commonly used. The Nessler reagent is thoroughly mixed
with the sample (ca. 30 minutes). The characteristic yellow color is photometrically deter-
mined (425 mu with 1 cm cell path). Analyses of samples containing 0.2 mg N/liter typically
produce results with + 0.12 mg N/liter standard deviation and + 18 percent bias (U.S. Environ-
mental Protection Agency, 1979).
3.4.3.2.8 Nitrate (NO, ). The rainwater sample's nitrate (normal concentration range 0.1 to
5.0 mg N/liter) is quantitatively reduced to nitrite by the addition of hydrazine sulfate
3-92
-------
(Kamphake et al., 1967) or by passing the sample through a copper-cadmium column (U.S. Envi-
ronmental Protection Agency, 1979d). The addition of sulfanil amide and N-(l-napthyl)-ethylene
diamine dihydrochloride yields a highly colored azo dye measureable colorimetrically at 520
nm. Automated techniques (U.S. Environmental Protection Agency, 1979c) minimize operator
error and increase ,the sample throughput. A second analysis without the nitrate reduction
step is required to correct for sample nitrite concentration. Precision and accuracy of + 5
percent are expected with samples above 1 mg N per liter. Butler et al. (1978) and Tyree et
al. (1979) report comparable sensitivity, precision, and accuracy using an automated 1C tech-
nique (see Section 3.3.4.2.2.4).
3.4.3.2.9 Chloride (Cl ). Various manual and automated procedures are used to determine
chloride in rainwater in the concentration range 0.1 to 10.0 mg Cl /liter with precision ca. +
0.2 mg Cl/liter. The WMO method adds mercuric nitrate and diphenylcarbazone-bromophenol blue
to the sample forming mercuric chloride. The excess mercury complexes with the indicator to
form a blue-violet dye measured photometrically at 525 nm. Zall et al. (1956) displace the
thiocyanate ion (SCN ) in Hg(SCN)? and form HgCl~. In the presence of excess iron, the highly
colored dye [Fe (SCN)+] is formed and can be photometrically measured at 460 |jm. The
automated ferricyanide procedure (U.S. Environmental Protection Agency, 1979d) is preferred
over the manual methods since operator/standard solution errors are minimized. Automated 1C
techniques (Butler et al., 1978; Tyree et al., 1979 and Section 3.3.4.1.1.4) yield comparable
results.
3.4.3.2.10 Fluoride (F ). Fluoride in wet deposition (range 0.01 to 0.1'mg F/liter) is
generally determined by the ion selective electrode technique. The condition of the fluoride
ion selective electrode is critical. Analysis of synthetic samples containing 0.85 mg F/liter
yielded results with a 3.6 percent relative standard deviation and 0.7 percent relative error
(American Public Health Association, 1971). Automated 1C techniques (Butler et al. , 1978;
Tyree et al., 1979; and Section 3,3.4.1.1.4) yield similar results.
3.4.3.2.11 Trace Metals. Techniques used to determine trace metal concentration in rainwater
are described in Section 3.3.4.3. Observed metal concentration ranges generally approximate
the lower detection limit for flame atomic absorption metal analysis.
3.4.4 Interlaboratory Comparisons
WMO (1975) instituted an international inter!aboratory program to describe the quality of
the wet deposition chemistry data being reported by the various WMO laboratories. Participa-
tion in this program is voluntary. The results of three comparisons on synthetic rainwater
samples (WMO,, 1976; Thompson, 1978; WMO, 1980) have been reported previously. The United
States Department of Energy (DOE) sponsored a similar round robin (Battelle Pacific Northwest
Laboratories, 1979) on both simulated and composited rainwater samples. Tyree et al. (1979)
and Tyree (1981) discuss the WMO and DOE results.
Three physical analyses (pH, conductivity, and acidity) are most frequently reported and
compared. Conductivity and pH data have been extracted from the three WMO reports and are
3-93
-------
TABLE 3-7. RESULTS OF HMO INTERCOMPARISONS OH SYNTHETIC PRECIPITATION SAMPLES
pH (pH Units)
Sample _k
Session type
1976
1978
1980
A
B
C
D
A
B
C
D
71
72
,73
"74
x
5.45
5.53
1.53
5.56
5.66
5.77
5.60
5.65
4.21
4.02
5.58
3.91
.74
.76
.52
.41
.49
.51
.44
.39
.16
.11
.19
.16
.j
Nd
17
18
17
18
25
25
25
25
26
26
26
24
High X
6.40
7.22
6.20
6.10
6.54
7.07
6.58
6.64
4.55
4.17
6.02
4.30
Sample
Low x Session type
3.65 1976
3.85
4.10
4.20
4.61 1978
4.65
4.54
4.79
3.80 1980
3.72
5.17
3,60
A
B
C
D
A
B
C
0
71
72
73
a74
X
6.9
17.3
56.3
109.3
6.7
17.9
52.4
103.6
29.1
80.4
195.9
'64.4
Conductivity {yS/cm)
°x
4.3
4.9
4.8
8.1
3.0
8.8
12.0
23.9
4.0
9.1
21.6
8.0
N
17
18
17
18
25
25
25
25
25
25
25
23
Acidity (peq/1)
Sample
High x Low x Session type x 0X N
22.0
28.0
62.0
119.0
19.0
57.0
67.5
132.0
37.3
95.6
224.0
81.5
3.9 1976
20.0
44.0
84.0
3.9 1978
9.0
19.1
37.9
18.3 1980
57.3
132.0
48.0
NOT REPORTED BY WHO
NOT REPORTED BY WHO
71 70.0 36.7 22
72 106.3 43.6 22
73 25.3 63.7 18
a74 10.9 16.1 22
High x Low x
206.0 18.9
260.6 29.2
202.0 -37.0
68.0 4.7
Sample contains only
b = sample mean
a = standard deviation =
/
_
= number of laboratories.
-------
provided in Table 3-7. WMO did not summarize the acidity measurement results in the first and
second analysis sessions. The synthetic samples used in the first two sessions contained only
weak acids. The laboratory results indicated that the WMO laboratories, as a whole, could not
perform acidity measurements on samples containing only weak acids. Four new samples were
used in the third WMO analysis session. Three of the third analysis session samples (samples
71, 72, 73) contained both weak and strong acids. Sample type 74 contained only strong
sulfuric acid. The results of the acidity measurements reported in the third analysis session
have also been extracted. Table 3-8 lists the among laboratories percent coefficient of
variation (% cv) by session for'those chemical analyses routinely performed by the WMO
participants, where
cv =
VI(di"5)2 ' x 100% (3-3)
N-l
Improvement in the analysis for a given constituent by the WMO laboratories, as a whole, is
indicated by decreasing percent cv. from session 1 to session 3. In general, the WMO labora-
tories showed improvement from session 1 to session 3.
WMO and DOE intercomparison results highlight the difficulties encountered -in analyzing
dilute precipitation samples. When comparing data from various studies, the data analyst must
include the appropriate biases resulting from the laboratory's sampling and storage techniques
as well as the ability of the laboratory to perform chemical analyses on the sample.
TABLE 3-8. COEFFICIENTS OF VARIATION OF WMO INTERCOMPARISONS ON
SYNTHETIC PRECIPITATION SAMPLES
Among laboratories percent coefficient of variation (% cv.)
% cv by session
Constituent 1976 1978 1980
PH
Conductivity
SO.
NHA
NO,
Cl3
Ca
K
Ma
Na
10
15
13
38
79
58
27
32
21
29
8
22
24
32
64
24
25
30
8
27
3.4
12
34
33
74
25
25
22
99
19
3-95
-------
3.5 REFERENCES
Abbas, R. and R. L. Tanner. Continuous determination of gaseous ammonia in the ambient,
atmosphere using fluorescence derivatization. Atmos. Environ. 15:277-81, 1981.
Adams, F. C. , and R. E. Van Grieken. Absorption correction for X-ray fluorescence analysis of
aerosol loaded filters. Anal. Chem. 47:1767-1773, 1975.
Ahern, A. E. Trace Analysis by Mass Spectrometry, Academic Press, New York, NY, 1972.
Ahlquist, N. C., and R. J. Charlson. A new instrument for evaluating the visual quality of
air. J. Air Pollut. Control Assoc. 17:467-469, 1967.
Akiyama, T. Direct titration of sulfate. Kyoto Yakka Daigaku Gakuho 5:48-50, 1957.
American Industrial Hygiene Association. Guide for respirable dust sampling. Am. Ind. Hyg.
Assoc, J. 31:133, 1970.
American Public Health Association. Standard Methods for the Examination of Water and Waste-
water, 13th ed. American Public Health Association, Washington, DC, 1971.
American Society for Testing and Materials. Standard method for collection and analysis of
dustfall (settleable particulates): ASTM D 1739-70. In: Annual Book of ASTM Standards.
Part 26: Gaseous Fuels; Coal and Coke; Atmospheric Analysis. American Society for
Testing and Materials, Philadelphia, PA, 1981a. pp. 521-525.
American Society for Testing and Materials. Standard method for particulate matter in the
atmosphere (optical density of filtered-material): ASTM D 1704 78. In: Annual Book of
ASTM Standards. Part 26: Gaseous Fuels; Coal and Coke; Atmospheric Analysis. American
Society for Testing and Materials, Philadelphia, PA, 1981b. pp. 513-520.
American Society for Testing and Materials. Standard test methods for sulfate ion in
water: ASTM D 516-80. Method C - volumetric method. In: Annual Book of ASTM
Standards. Part 31: Water. American Society for Testing and Materials, Philadelphia,
PA, 1981c. pp. 602-605.
Anderson, D. E. Continuous Measurement of pH Using An Automatic Precipitation Collector and
Analyzer. University of Maryland, Publication 78-178, College Park, MD, May 1978.
Appel, B. R., and Y. Tokiwa. Atmospheric particulate nitrate sampling errors due to reactions
with particulate and gaseous strong acids. Atmos. Environ. 15:1087-1089, 1981.
Appel, B. R., E. M. Hoffer, E. L. Kothny, and S. M. Wall. Interference in 2,4-xylenol proce-
dure for nitrate determination in atmospheric aerosols. Environ. Sci. Techno!. 11:189,
1977a.
Appel, B. R., E. L. Kothny, E. M. Hoffer, and J. J. Wesolowski. Comparison of Wet Chemical and
Instrumental Methods for Measuring Airborne Sulfate. EPA-600/7-77-128,' U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, November 1977.
Appel, B. R., S. M. Wall, Y. Tokiwa, and M. Haik. Interference effects in sampling particu-
late nitrate in ambient air. Atmos. Environ. 13:319-325, 1979.
Appel, B. R., S. M. Wall, Y. Tokiwa, and M. Haik. Simultaneous nitric acid, particulate
nitrate and acidity measurements in ambient air. Atmos. Environ. 14:549-554, 1980.
3-96
-------
Appel, B. R. , Y. Tokiwa, S. M. Wall, E. M. Hoffer, M. Haik, and J. J. Wesolowski. Effects of
Environmental Variables and Sampling Media on the Collection of Atmospheric Sulfates.
Final Report to the California Air Resources Board, Contract #5-1032, Air and Industrial
Hygiene Laboratories, Calif. Dept. of Health Services. Berkeley, CA, 1978.
Appel, B. R., E. M. Hoffer, M, Haik, W. Wehrmeister, E. L. Kothny, and J. J. Wesolowski, Im-
provement and Evaluation of Methods for Sulfate Analysis, EPA-600/4-79-028, U.S.
Environmental Protection Agency, Research Triangle Park, NC, April 1979a.
Appel, B. R., J. J. Wesolowski, E, Hoffer, S, Twiss, S. Wall, S. G. Chang, and V. T. Novak. An
intermethod comparison of X-ray photoelectron spectroscopy (ESCA) analysis of atmospheric
particulate matter. Int. J. Environ. Anal. Chem. 4:169-181, 1976.
Askne, C., C. Brosset, and M. Fern. Determination of the Proton-Donating Property of Airborne
Particles. NL Publication B157, Goteborg, Sweden, 1973.
Axelrod, H. D. , and S. G. Hansen. Filter sampling method for atmospheric sulfur dioxide at
background concentrations. Anal. Chem. 47:2460-2461, 1975.
Axelrod, H. D. , J. E. Bonelli, and J. P. Lodge, Jr. Fluorimetric determination of trace
nitrates. Anal. Chim. Acta 51:21-24, 1970.
Bacon, J. C., M. E. Edelson, and G. J. Small. Fluorescence line narrowing spectrometry in
organic glass containing PPB levels of polycyclic aromatic hydrocarbons. Anal. Chem.
50:1394-1397, 1978. -
Bailey, D. L. R. , and P. Clayton. The Measurement of Suspended Particulate and Carbon Concen-
trations in the Atmosphere Using Standard Smoke Shade Methods. Warren Springs Laboratory
Report LR325(AP), Stevenage, Hertfordshire, England, 1980.
Ball, D. J., and R. Hume. The relative importance of vehicular and domestic emissions of dark
smoke in Greater London in the mid-1970's, the significance of smoke shade measurements,
and an explanation of the relationship of smoke shade to gravimetric measurements of
particulate. Atmos. Environ. 11:1065-1073, 1977.
Balzer, J. L. Inertia! and gravitational collectors. In; Air Sampling Instruments for
Evaluation of Atmospheric Contaminants, 4th ed. , American Conference on Governmental
Industrial Hygienists, Cincinnati, OH, 1972. pp. 16-17.
Bakas, E. Fast titrimetric method for the determination of sulfate ions. Magy. Kern.
Foly. 62:135-139, 1956.
Barnard, W. F. , ed. Community Health Environmental Surveillance Studies (CHESS) Air Pollution
Monitoring Handbook: Manual Methods. EPA-600/1-76-011, U.S. Environmental Protection
Agency, Research Triangle Park, NC, January 1976.
Barnes, R. A. Duplicate measurements of low concentrations of smoke and sulphur dioxide using
two "National Survey" samplers with a common inlet. Atmos. Environ. 7:901-904, 1973.
Barrett, E., and G. Brodin. The acidity of Scandinavian precipitation. Tellus 7(2):25-57,
1955.
Barrett, W. J. , H. C. Miller, J. E. Smith, Jr., and C. H. Gwin. Development of a Portable
Device to Collect Sulfuric Acid Aerosols. EPA-600/2-77-027, U.S. Environmental
Protection Agency, Research Triangle Park, NC, February 1977,
Barringer Research Limited. A report to Department of Health, Education, and Welfare of
Optical Measurements of S0« and NO, Air Pollution Using Barringer Correlation Spectrome-
ters. Technical Report TR-69-112r, Barringer Research, Ltd., Rexdale, Ont., Canada,
December 1969.
3-97
-------
Barton, S. C. , and H. G. McAdie, A specific method for the automatic determination of ambient
HjSCL aerosol. In: Proceedings of the 2nd International Clean Air Congress, Interna-
tional Union of Air Pollution Prevention Associations, Washington, D.C. , December 6-11,
1970, H. M. Englund and W. T. Berry, eds, , Academic Press, New York, NY, 1971. pp.
379-382.
Barton, S. C. , and H. G. McAdie. An automated instrument for monitoring ambient H«SO,
aerosol. In: Proceedings of the Third International Clean Air Congress, Internationa*
Union of Air Pollution Prevention Associations, Dusseldorf , Germany, October 8-12, 3973.
Verein Deutscher Ingenieure, Dusseldorf, Germany, 1973. pp. C25-C27.
Battelle Pacific Northwest Laboratories. MAP3S Precipitation Chemistry Network. PNL-2829.
U.S. Department of Energy, Washington, DC, 1979.
Baumgardner, R. E. , W. A. McClenny, and R. K. Stevens. Optimized Chemi luminescence System for
Measuring Atmospheric Ammonia. EPA-600/2-79-028, U.S. Environmental Protection Agency,
Research Triangle Park, NC, 1979.
Behrmann, J. J. Design and Testing of a Continuing Monitor of Atmospheric Precipitation.
Master's Thesis, North Carolina State University, Department of Chemical Engineering,
Raleigh, NC, 1975.
Bergman, F. J. , and M. C, Sharp. Measurement of Atmospheric Sulfates: Evaluation of the
Methylthymol Blue Method Using an Autoanalyzer II. EPA Contract No. 68-02-1728, U.S.
Environmental Protection Agency, Research Triangle Park, NC, March, 1979.
Bertolacini, R. J. , and J. E. Barney. Colorimetric determination of sulfate with barium
chloranilate. Anal. Chem. 29:281, 1957.
Beuttell. R. G. , and A. W. Brewer. Instruments for the measurement of the visual range. J.
Sci. Instrum. Phys. Ind. 26:357-359, 1949.
Bigg, E. K. , A. Ono, and J. A. Williams. Chemical tests for individual submicron aerosol
particles. Atmos. Environ. 8:1-13, 1974.
Blacker, J. H. , R. G. Confer, and R. S. Brief. Evaluation of the reference method for deter-
mination of sulfur dioxide in the atmosphere (pararosaniline method). J. Air Pollut.
Control Assoc. 23:525-527, 1973.
Blanco, A. J. , and G. B. Hoidale. Microspectrophotometric technique for obtaining the intra-
red spectrum of microgram quantities of atmospheric dust. Atmos. Environ. 2:327-330,
1968.
Blanco, A. J. , and R. G. Mclntyre. An infra-red spectroscopic view of atmospheric particu-
lates over El Paso, Texas. Atmos. Environ. 6:557-562, 1972.
Bokhoven, C. , and H. J. Niessin. The continuous monitoring of traces of SO- in air on the
basis of discolouration of the starch-iodine reagent with prior eliminations of inter-
fering compounds. Air Water Pollut. 10:233-243, 1966.
Bolleter, W. T. , C. J. Bushman, and P. W. Tidwell. Spectrophotometric determination of
ammonia as indophenol. Anal. Chem. 33:592-594, 1961.
Bostro'm, C. E. The absorption of low concentrations (pphm) of hydrogen sulfide in a ^
suspension as studied by an isotopic tracer method. Air Water Pollut. 10:435-441, 19i5B.
Boumans, P. W. J. M. , and F. J. Le Boer. Studies of an inductively-coupled high-frequency
argon plasma for optical emission spectrometry - II. Compromise conditions for simul-
taneous multi-element analysis. Spectrochem. Acta. 308:309-334, 1975.
3-98
-------
Bowden, S. R, Improved lead dioxide method of assessing sulphurous pollution of the atmos-
phere. Air Water Pollut. 8:101, 1964.
Bradway, R. M., F. A. Record, and T. G. Pace. Application of polarizing microscopy to the
characterization of ambient suspended particulate. In: Proceedings Annual Meeting of
the Federation of Analytical Chemists and Spectroscopists Society, Philadelphia, PA,
, ; 1976. Available from Standard Oil Company, Cleveland, OH, 1976.
Braman, R. S., and T. J. Shelley. Tungsten VI Oxide for Pre-Concentration and Determination
of Gases and Particulate Ammonia and Nitric Acid in Ambient Air. EPA-600/7-80-179, U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1980.
Braman, R. S., D. L. Johnson, C. C. Foreback, J. M. Ammons, and J. L. Bricker. Separation and
determination of nanogram amounts of inorganic arsenic and methylarsenic compounds.
Anal. Chem. 49(4):621-625, 1977.
British Standards Institution. Methods for the Measurement of Air Pollution. Part 3: Deter-
mination of Sulphur Dioxide. B.S. 1747, British Standards Institution, London, England,
1963.
British Standards Institution. Methods for-the Measurement of Air Pollution. Part 2: Deter-
mination of Concentration of Suspended Matter. B.S. 1747, British Standards Institution,
London, England, 1964.
Bromberg, S. M., G. G. Akland, and J. C. Puzak. Survey of laboratory performance. Analysis
of simulated ambient S00 bubbler samples. J. Air Pollut. Control Assoc. 24:1073-6 1973,
, 1974. l ~
Bromberg, S. M., R. L. Lampe, and 8. I, Bennett. Summary of Audit Performance: Measurement
of S02, NQ2, CO, Sulfate, Nitrate, Lead, Hi-Vol Flow Rate - 1977. EPA-600/4-79-014, U.S.
Environmental Protection Agency, Research Triangle Park, NC, February 1979.
Bromberg, S. M.» R. L. Lampe, and B. I. Bennett. Summary of Audit Performance: Measurement
of SO,, NQp, Sulfate, Nitrate, Lead, Hi-Vol Flow Rate - 1978. EPA-600/4-80-017, U.S.
Environmental Protection Agency, Research Triangle Park, NC, June 1980,
Brosset, C., and M. Ferm. Man-made airborne acidity and its determination. Atmos. Environ.
12:909-916, 1978.
Brosset, C., K. Andreasson, and M. Ferm. The nature and possible origin of acid particles
observed at the Swedish west coast. Atmos. Environ. 9:631-642, 1975.
Butler, F. E., R. H. Jungers, L. F. Porter, A. E. Riley, and F. J. Toth. Analysis of air par-
ticulates by ion chromotography: comparison with accepted methods. In: Ion Chromato-
graphic Analysis of Environmental Pollutants. E. Sawicki, J. Mulik, and E. Wittgenstein,
eds., Ann Arbor Science, Ann Arbor, MI, 1978.- pp. 65-76.
Cadle, S. H., P. J. Groblicki, and D. P. Stroup. An Automated Carbon Analyzer for Particulate
Samples. Research Publication GMR-3306, ENV f77, General Motors Research Laboratories,
Warren, MI, 1980.
Calvert, S., and W. Workman. Estimation of efficiency for bubbler-type gas absorber. Talanta
4:89-100, 1960.
Camp, D. C. , A. L. Van Lehn, and B.,W. Loo. Interconoparison of Samplers Used in the Determi-
nation of Aerosol Composition. EPA-600/7-78-118, U.S. Environmental Protection Agency,
Research Triangle Park, NC, July 1978.
3-99
-------
Caplan, K. J., L, J. Doemeny, and S. D. Sorenson. Performance characteristics of the 10mm
cyclone respirable mass sampler: Part II. Am. Ind. Hyg. Assoc. J. 38:162-173, 1977.
Cautreels, W., and K. Van Cauwenberghe. Determination of organic compounds in airborne par-
ticulate matter by gas/chromatography - mass spectrometry. Atmos. Environ,, 10:447,
1976.
Cermak, J. E, Applications of fluid mechanics to wind engineering, Freeman Scholar Lecture.
Presented at Winter Annual Meeting of the American Society of Mechanical Engineers. New
York, November 17, 1974.
Chand, R., and R. V. Marcote. Evaluation of portable electrochemical monitors and associated
stack sampling for stationary source' monitoring. Presented at the 68th National Meeting
of the American Institute of Chemical Engineers, Houston, TX, February 28-March 4, 1971.
Charlson, R. J., N. C. Ahlquist, and H. Horvath. On the generality of correlation of atmos-
pheric aerosol mass concentration and light scatter. Atmos. Environ. 2:455-464, 1968.
Charlson, R. J., A. P. Waggoner, and J. F. Thielke. Visibility Protection for Class I Areas:
the Technical Basis. CEQ517714498, Council on Environmental Quality, Washington, DC,
August 1978.
Chow, T. J., and M. S. Johnstone. Determination of nitrate in seawater. Anal. Chim. Acta
27:441-446, 1962.
Coffer, J. W., C. McJilton, and R. J. Charlson. SCL oxidation to sulfate due to particulate
matter on a high volume air sample filter. Paper no. ANAL 102. Presented at 167th
National Meeting, American Chemical Society, Los Angeles, CA, April 1-5, 1974.
Cogbill, C. V., and G. E. Likens. Acid precipitation in the northeastern United States.
Water Resour. Res. 10:1133-1137, 1974.
Collison, R. C., and J. F. Mensching. Lysimeter investigations. II. Composition of rain
water at Geneva, N.Y. for a ten-year period. N.Y. Agr. Expt. Sta., Tech. Bull. 193:1-19,
1932.
Colovos, G., M. R. Panesar, and E. P. Parry. Linearizing the calibration curve in determina-
tion of sulfate by the methylthymol blue method. Anal. Chem. 48:1693-1696, 1976.
Colovos, G., E. P. Parry, and C. E. Rodes. Comparison of the degradation of ammonium ion on
high volume glass fiber filters and on membrane filters. In: The Los Angeles'Catalyst
Study Symposium, U.S. Environmental Protection Agency, Raleigh, NC, April 12-13, 1977.
EPA-600/4-77-034, U.S. Environmental Protection Agency, Research Triangle Park, NC, June
1977. pp. 265-279.
Commins, B. T., and R. E. Waller. Observations from a ten year study of pollution at a site
in the city of London. Atmos. Environ. 1:49-68, 1967.
Conner, W. D. An inertial-type particle separator for collecting large samples. J. Air
Pollut. Control Assoc. 16:35-38, 1966.
Cooper, H. B. H., J. A. Lopez, and J. M. Demo. Chemical composition of acid precipitation in
central Texas. In; First International Symposium on Acid Precipitation and the Forest
Ecosystem, U.S. Department of Agriculture and Others. Columbus, Ohio, May 12-15, 1975.
Water, Air, Soil Pollut. 6:351-359, 1976.
Corn, M., F. Stein, and N. Esmen. The Shape of Atmospheric Particles: Progress Report.
Public Health Service Research Grant AP00431-02, University of Pittsburgh, Air Pollution
Laboratory, Pittsburgh, PA, November 1967.
3-100
-------
Courtney, W. J., S. Rheingrover, J. Pilotte, H. C. Kaufmann, T. A. Cahill, and J. W. Nelson.
Continuous observation of particulates during the General Motors Sulfate Dispersion
Experiment. J. Air Pollut. Control Assoc. 28:224, 1978,
Coutant, R, W. Effect of environmental variables on collection of atmospheric sulfate.
Environ. Sci. Techno!. 11:873-878, 1977,
Covert, D. S. , A. P. Waggoner, R. E. Weiss, N. C. Ahlquist, and R. J. Charlson. Atmospheric
aerosols, humidity, and visibility. In: The Character and Origins of Smog Aerosols. G,
M. Hidy, P. K. Mueller, D. Grosjean, B. R. Appel, and J. J. Wesolowski, eds. , J. Wiley
and Sons, New York, NY, 1980. pp. 559-581.
Crider, W. L. Hydrogen flame emission spectrophotometry in monitoring air for sulfur dioxide
and sulfuric acid aerosol. Anal. Chem. 37:1770-1773, 1965.
Crider, W. L., N. P. Barkley, M. J, Knott, and R. W. Slater, Jr. Hydrogen flame chemilumines-
cence detector for sulfate in aqueous solutions. Anal. Chim. Acta 47:237-241, 1969.
Cunningham, P. T., and S. A. Johnson. "Spectroscopic observation of acid sulfate in atmos-
pheric particulate samples. Science (Washington, D.C.) 191:77-79, 1976.
Cunningham, P. T. , S, A. Johnson, and R. T. Yang. Variations in chemistry of airborne parti-
culate material with particle size and time. Environ. Sci. Techno!. 8:131, 1974.
Daley, P. S., and D. A. Lundgren. The performance of piezoelectric crystal sensors used to
determine aerosol mass concentrations. Presented at American Industrial Hygiene Confer-
ence, Miami, May 12, 1974.
Dasgupta, P. K., and K. B. DeCesare. Stability of sulfur dioxide in formaldehyde and its
anomalous behavior in tetrachloromercurate (II). Anal. Chem., in press 1981.
Dasgupta, P. K., K. DeCesare, and J. C. Ullrey. Determination of atmospheric sulfur dioxide
without tetrachloromercurate (II) and the mechanism of the Schiff reaction. Anal. Chem.
52:1912-22, 1980.
Delumyea, R. G., L, C. Chu, and E. S. Macias, Determination of elemental carbon component of
soot in ambient aerosol samples, Atmos. Environ. 14, 647-652, 1980.
Department of Scientific and Industrial Research. The Investigation of Atmospheric Pollution,
1931-1932. 18th Report. H. M. S. 0., London, 1933.
De Veer, S, M., H. J. Brouwer, N. V. Philips, and H. Zeedijk. A remotely controlled SO,
monitor for operation in air pollution networks. Presented at the 62nd Annual Meeting,
Air Pollution Control Association, New York, NY, June 1969. Paper no. 69-6.
DiMartini, R. Determination of nitrogen dioxide and nitric oxide in the parts per million
range in flowing gaseous mixtures by means of the nitrate specific-ion electrode. Anal.
Chem. 42:1102-1105, 1970.
Dionex Corporation, Dionex Analytical Ion Chromatographs. Technical Brochure, Dionex Corp.,
Palo Alto, California, 1975.
Dochinger, L. S., and R. A. Seliga, Eds. Proceedings of the First International Symposium on
Acid Precipitation and the Forest Ecosystem, U.S. Department of Agriculture and Others,
Columbus, Ohio, May 12-15, 1975. NEFES/77-1, U.S. Department of Agriculture, Forest
Service, Upper Darby, PA, August 1976.
Driscoll, J. M. , A. W. Berger, J. M. Becker, J. T. Funkhouser, and J. R. Valentine. Determi-
nation of oxides of nitrogen in combustion effluents with a nitrate ion selective
electrode. J. Air Pollut. Control Assoc. 22:119-122, 1972.
3-101
-------
Dubols, L., C. J. Baker, T. Telchman, A. Zdrojewski, and J. L. Monkman. The determination of
sulphuric acid in air: a specific method, Mikrochim.-Acta 2:269-279, 1969.
Dubois, L., R. S. Thomas, T. Teichman, and J. L. Monkman. A general method of analysis for
high volume air samples. I. Sulphate and sulphuric acid. Mikrochim. Acta. 6:1268-1275,
1969.
Dzubay, T. G., ed. X-ray Fluorescence Analysis of Environmental Samples. Ann Arbor Science,
Ann Arbor, MI, 1977.
14
Dzubay, T. G. Aerosol acidity determination by reaction with C labelled amine. Atmos.
Environ. 13:1209, 1979.
Dzubay, T. G., and R. K. Stevens. Ambient air analyses with dichotomous sampler and X-ray
fluorescence spectrometer. Environ. Sci. Tech. 9:663-668, 1975.
Dzubay. T. G., H. L. Rook, and R. K. Stevens. A chemiluminescent approach to measurement of
strong acid aerosols. In: Analytical Methods Applied to Air Pollution Measurements.
R. K. Stevens and W. F. Herget, eds., Ann Arbor Science Publishers, Inc., Ann Arbor, MI,
1974. pp. 71-83.
Dzubay, T. G. , L. E. Mines, and R. K. Stevens. Particle bounce errors in cascade impactors.
Atmos. Environ. 10:229-234, 1976.
Dzubay, T. G., P. J. Lamothe, and H. Yoshuda. Polymer films as calibration standards for
X-ray fluorescence analysis. Adv. X-Ray Anal. 20:411, 1977.
Eagan, M. L., and L. Dubois. The determination of ammonium ion in airborne particulates with
selective electrodes. Anal. Chim. Acta 70:157-167, 1974.
Edwards, Joseph Q., A Comparison of the Bristish Smoke Shade Method and the Integrating Plate
Method, Masters Thesis, .Department of Civil Engineering, University of Washington,
Seattle, WA, 1980.
Eggleton, A. E., and D. H. Atkins. Results of the Tees-side investigation. U.K. Atomic
Energy Authority, Research Group Report AERE-R6983, 1972.
Fassel, V. Quantitative elemental analyses by plasma emission spectroscopy. Science (Wash-
ington, D.C.) 202:183-191, 1978.
Faxvog, F. R. , and D. M. Roessler. Optoacoustic Measurements of Diesel Particulate Emissions.
Research Publication GMR-2987, General Motors Research Laboratories, Warren, MI, 1979.
Perm, M. Method for determination of atmospheric ammonia. Atmos. Environ. 13:1385-1393,
1979. ~~
Forrest, J., and L. Newman. Ambient air monitoring for sulfur compounds—a critical review.
J. Air Pollut. Control Assoc. 23:761-768, 1973.
Forrest, J., and L. Newman. Sampling and analysis of atmospheric sulfur compounds for isotope
ratio studies. Atmos. Environ. 7:561-573, 1973a.
Forrest, J., and L. Newman. Silver-110 microgram sulfate analysis for the short time resolu-
tion of ambient levels of sulfur aerosols. Anal. Chem. 49:1579, 1977.
3-102
-------
Forrest, J., R. L. Tanner, D. Spandau, T. D'Qttavio, and L. Newman. Determination of atmos-
pheric HMO., with NaCl-impregnated filters at high volume flow rates. In: Current
Methods to Measure Atmospheric Nitric Acid and Nitrate Artifacts. R. K. .Stevens, ed.,
EPA-600/2-79-051, U.S. Environmental Protection Agency, Research Triangle Park, NC, March
1979. pp. 51-61.
Forrest, J. , R. L. Tanner, D. Spandau, T. D'Qttavio, and L. Newman. Determination of total
inorganic nitrate utilizing collection of nitric acid on sodium chloride-impregnated
filters. Atmos. Environ., 14:137-144, 1980.
Foster, J. E. , and G. H. Beatty. Inter!aboratory Cooperative Study of the Precision and
Accuracy of the Measurement of Sulfur Dioxide Content in the Atmosphere Using ASTM Method
D2914. ASTM Data Series Publication DS-55-S1, American Society for Testing and
Materials, Philadelphia, PA, 1974.
Fox, D. L., and H. E. Jeffries. Air pollution. Anal. Chem. 51(5):27R, 1979.
Fox, M. A., and St. W. Staley. Determination of polycyclic aromatic hydrocarbons in atmos-
pheric particulate matter by high pressure liquid chromatography coupled with fluores-
cence techniques. Anal. Chem. 48:992, 1976.
Frank, E. R. , and J. P. Lodge, Jr. Morphological identification of airborne particles with
the electron microscope. J. Microsc. 6:449, 1967.
f
Friedlander, S. K. Smoke, Dust, and Haze. Wiley Interscience Publication, New York, NY,
1977. p. 154.
Fritz, J. S. , and S. S. Yamamura. Rapid microtitration of sulfate. Anal. Chem. 27:1461-1464,
1955.
Fritz, J. S. , S. S. Yamamura, and M. J. Richard. Titration of sylfate following separation
with alumina. Anal. Chem. 29:158-161, 1957.
Fry, J. D. Determination of Sulphur Dioxide in the Atmosphere by Absorption in Hydrogen
Peroxide Solution—the Effect of Evaporation of the Solution During Sampling.
SSD/SW/M.246, Central Electricity Generating Board, Southwestern Region, England, April
1970.
Fuchs, N. A. The Mechanics of Aerosols, Pergamon Press, New York, NY, 1964. p. 64.
Fuerst, R. G, , F. P. Scaringelli, and J. H. Margeson. Effect of Temperature on Stability of
Sulfur Dioxide Samples Collected by the Federal Reference Method. EPA-600/4-76-024, U.S.
Environmental Protection Agency, Research Triangle Park, NC, May 1976.
Galloway, J. N. , and E. B. Cowling. The effects of precipitation on aquatic and terrestrial
ecosystems: A proposed precipitation network. J. Air Pollut. Control Assoc. 28:229-235,
1978.
Galloway, J. N., B. S. Cosby, and G. E. Likens. Acid precipitation: measurement of pH and
acidity. Limnol. Oceanogr. 24(6):1161, 1979.
Galloway, J. N., and G. E. Likens, Calibration of collection procedures for the determination
of precipitation chemistry. In: First International Symposium on Acid Precipitation and
the Forest Ecosystem, U.S. Department of Agriculture and Others, Columbus, Ohio, May
12-15, 1975. Water Air Soil Pollut. 6:241-258, 1976.
Gatz, D. f., and A. N. Dingle. Trace substances in rain water: concentration variations dur-
ing convective rains, and their interpretation. Tellus 23:14-17, 1971.
3-103
-------
Gilbert, T. R., and A, M. Clay. Determination of ammonia in aquaria and in sea water using
the ammonia electrode. Anal. Chem. 45:1757-1759, 1973.
Glover, D, J., and J. C. Hoffsommer. Gas chromatographic analysis of nitrate and nitrite ions
in microgram quantities by conversion to nitrobenzene. J. Chromatogr. 94:334-337, 1974.
Gordievskii, A. V., A. Ya, Syrchenkov, V. V. Sergievskii, and M. I. Savvin. Nitrate-selective
membrane electrodes. Elektrokhim. 8:520-521, 1972.
Gould, R. W., C. S. Barrett, J. B. Newkirk, and C. 0. Ruud, eds. Adv. in X-ray Anal. |9:305
et seq., 1976.
Goulding, F. S. , and J. M. Jaklevic. X-ray Fluorescence Spectrometer for Airborne Particulate
Monitoring. EPA-R2-73-182, U.S. Environmental Protection Agency, Research Triangle Park,
NC, April 1973.
Goulding, F. S., J. M. Jaklevic, and B. W. Loo. Aerosol Analysis for the Regional Air Pollu-
tion Study. EPA-600/4-78-034, U.S. Environmental Protection Agency, Research Triangle
Park, NC, July 1978.
Gran, G. Determination of the equivalence point in potentiometric titrations. Part II.
Analyst 77:661-671, 1952.
Granat, L. Relation between pH and the chemical composition in atmospheric precipitation.
Tell us 24(6):550, 1972.
Groblicki, P. J., G. T. Wolf, and R. J. Countess. Visibility Reducing Species in the Denver
"Brown Cloud," Part I, Relationships Between Extinction and Chemical Composition.
General Motors Research Laboratories publication #GMR-2417, Env # 81, MI, 1980.
Grosjean, D. Solvent extraction and organic carbon determination in atmospheric particulate
matter: the organic extraction-organic carbon analyzer technique. Anal. Chem. 47:797,
1975.
Hagen, A., and A. Langeland. Polluted snow in southern Norway and the effect of the meltwater
on freshwater and aquatic organisms. Environ. Pollut. S:45-57, 1973.
Hager, R. N., Jr., and R. C. Anderson. Theory of the derivative spectrometer. J. Opt. Soc.
Am. 60:1444-1449, 1970.
Hamilton, R. J. , and W. H. Walton. The selective sampling of respirable dust. In: Inhaled
Particles and Vapours. C. N. Davies, ed., Pergamion Press, London, England, 1961. pp.
465-475.
Hardin, J., and B. Shleien. Determination of sulfur dioxide in ambient air by non-dispersive
X-ray fluorescence. In: Proceedings of Symposium on Nuclear Techniques and Measurement
and Control of Environmental Pollution. International Atomic Energy Agency, Vienna,
Austria, 1971.
Harker, A,, L. Richards, and W. Clark. Effective atmospheric S0? photochemistry upon observed
nitrate concentrations. Atmos. Environ. 11:87-91, 1977.
Heard, M. J. , and R. D. Wiffen. Electron microscopy of natural aerosols and the identifica-
tion of particulate ammonium sulphate. Atmos. Environ. 3_:337-340, 1969.
Hemeon, W. C. L., G. F. Haines, Jr., and H. M. Ide. Determination of haze and smoke concen-
trations by filter paper samples. Air Repair 3:22-28, 1953.
3-104
-------
Hendry, C. D., and P. L. Brezonik. Chemistry of precipitation at Gainesville, Florida.
Environ. Sci, Techno!. 14(7):843-49, 1980.
Hering, S. V. , R. C. Flagan, and S. K. Friedlander. Design and evaluation of new low-pressure
impactor. I. Environ. Sci. Technol. 12:667-73, 1978.
Hidy, G. M. Summary of the California aerosol characterization experiments. J. Air Pollut.
Control Assoc. 25:1106, 1975.
Hill, A. S. G, Measurement of the optical densities of smokestains on filter papers. Trans.
Farad. Soc. 32:1125-1131, 1936,
Hodgeson, J. A., J. P. Bell, K. A. Rehme, K. J. Krost, and R. K, Stevens. Applications of a
Chemiluminesence Detector for the measurement of total oxides of nitrogen and ammonia in
the atmosphere. Presented at Joint Conference on Sensing of Environmental Pollutants,
AIAA, New York, NY, 1971. Paper no. 71-1067.
Hoell, J. M., C. N. Harwarel, and B. S. Williams. Remote infrared heterodyne radiometer
measurements of atmospheric ammonia profiles. Geophys. Res. Lett., 7.313, 1980.
Hoffer, E., and E. L. Kothny. A Micromethod for Sulfate in Atmospheric Particulate Matter.
Report No. 163, Air and Industrial Hygiene Laboratory, Berkeley, CA, July 1974.
Hounam, R. E., and R. J. Sherwood. The cascade centripeter: a device for determining the
concentration and size distribution of aerosols. Am. Ind. Hyg. Assoc. J. 26:122-131,
1965.
Huey, N. A. The lead dioxide estimation of sulfur oxide pollution. J. Air Pollut. Control
Assoc. 18:610-11, 1968.
Huey, N. A., M. A. Wallar, and C. D. Robson. Field evaluation of an improved sulfation
measurement system. Presented at the 62nd Annual Meeting, Air Pollution Control Associa-
tion, New York, NY, June 22-26, 1969. Paper no. 69-133.
Huntzicker, J. J., L. M. Isabelle, and J. G. Watson. The continuous monitoring of particulate
sulfate by flame photometry. In: International Conference on Environmental Sensing and
Assessment, vol. 2. Institute of Electrical and Electronics Engineers, Inc., New York,
NY, 1976. Paper no. 23-4.
Huntzicker, J. J., L. M. Isabelle, and J. G. Watson. The continuous monitoring of particulate
sulfur compounds by flame photometry. Presented at the 173rd National Meeting, American
Chemical Society, New Orleans, LA, March 20-25, 1977. Paper no. ENVR-39.
Huntzicker, J. J. , R. S. Hoffman, and C. S. Ling. Continuous measurement and speciation of
sulfur containing aerosols by flame photometry, Atmos. Environ. 12:83-88, 1978.
Huygen, C. The sampling of sulfur dioxide in air with impregnated filter paper. Anal. Chim.
Acta 28:349, 1963. }
Ingram, W., and J, Golden. Smoke curve calibration, J. Air Pollut. Control Assoc. 23:110-115,
1973.
Intersociety Committee. Methods of Air Sampling and Analysis, 2nd ed. American Public Health
Association, Washington, DC, 1977a.
Intersociety Committee. Tentative method of analysis for nitrate in atmospheric particulate
matter (brucine method). In: Methods of Air Sampling and Analysis, 2nd ed. American
Public Health Association, Washington, DC, 1977b. pp. 521-523.
3-105
-------
Intersociety Committee. Tentative method of analysis for total nitrogen oxides as nitrate
(phenoldisulphonic acid method). In: Methods of Air Sampling and Analysis, 2nd ed.
American Public Health Association, Washington, DC, 1977c. pp. 534-538.
Intersociety Committee. Tentative method of analysis for nitrate in atmospheric participate
matter (2,4-xylenol method (1)). In: Methods of Air Sampling and Analysis, 2nd ed.
American Public Health Association, Washington, DC, 1977d. pp. 518-520.
Jacobs, M. B. The Chemical Analysis of Air Pollutants. Wiley-Interscience, New York, NY,
1960.
Jacobs, H. B., H. H. Braverman, and S. Hochheiser. Ultramicrodetermination of sulfides in
air. Anal. Chem. 29:1349-1359, 1957.
Jaklevic, J. H., R. C. Gatti, F. S. Goulding, and B. W. Loo. A p-gauge method applied to
aerosol samples. Environ. Sci. Techno!. 15:680-686, 1981.
Johansson, T. B., R. E. Van Grieken, J. W. Nelson, and J. W. Winchester. Elemental trace
analysis of small samples by proton induced X-ray emission. Anal. Chem. 47:855-860,
1975.
John, W. G., and G. P. Reischl. Measurements of the filtration efficiencies of selected
filter types. Atmos. Environ. 12:2015, 1978.
John, W., G. Reischl, and J. Wesolowski. Size Selective Monitoring Techniques for Particulate
Matter. Report No. ARB A5-00487, California Department of Health, Air and Industrial
Hygiene Laboratory, Berkeley, CA, February 1978.
Johnson, D. A., and D. H. F. Atkins. An airborne system for the sampling and analysis of
sulphur dioxide and atmospheric aerosols. Atmos. Environ. 9:825-829, 1975.
Johnson, R. L., and J. J. Huntzicker. Analysis of volatilizable and elemental carbon in
ambient aerosols. In: Proceedings: Carbonaceous Particles in the Atmosphere, National
Science Foundation and Lawrence Berkeley Laboratory, Berkeley, California, March 20-28,
1978. T. Novakov, ed., LBL - 9037, Lawrence Berkeley Laboratory, University of Califor-
nia, Berkeley, CA, June 1979. pp. 10-13.
Johnson, W. B., R. J. .Allen, and W. E. Evans. LIDAR Studies of Stack Plumes in Rural and
Urban Environments. EPA-650/4-73-002, U.S. Environmental Protection Agency, Meteorology
Laboratory, National Environmental Research Center, Research Triangle Park, NC, October
1973.
Junge, C. E., and P. E. Gustafson. Precipitation sampling for chemical analysis. Bull. Ara.
Meteq.ro!. Soc. 37: 244, 1956.
Kadlecek, J. A., and V. A. Mohnen. Time dependence of the pH of rain. |n: Proceedings of
the First International Symposium on Acid Precipitation and the Forest Ecosystem, U.S.
Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. L.S. Dochinger
and T. A. Seliga, eds., NEFES/77-1, U.S. Department of Agriculture, Forest Service, Upper
Darby, PA, August 1976. p. 207.
Kamphake, L., S, Hannah, and J. Cohen. Automated analysis for nitrate by hydrazine reduction.
Water Res. 1:205-16, 1967.
Kasten-Schraufnagel, P., D. L. Ehman, and D. J. Johnson. Telmatic Study Phase II: Effects of
Collection and Handling Conditions on the Stability of the Dichlorosulfitomercurate (II)
Complex Formed in Sampling for S0? by the Modified West-Gaeke Method. Texas Air Control
Board, Air Quality Evaluation Division, January 1975.
3-106
-------
Katz, M. Photoelectric determination of atmospheric sulphur dioxide employing dilute starch-
iodine solutions. Anal. Chem. 22:1040-1047, 1950.
Katz, M. Measurement of Air Pollutants—Guide to Selection of Methods. World Health Organi-
zation, Geneva, Switzerland, 1969.
Keay, J. , and P. M. A. Menage. Automated determination of ammonium and nitrate in soil
extracts by distillation. Analyst 95:379, 1970.
Kennedy, V. Research and Monitoring of Precipitation Chemistry in the U.S. 1/m Present and
Future Needs, U.S. Geological Service, Reston, VA, July 1978.
Ketseridis, G., J. Hahn, R. Jaenicke, and C. Junge. The organic constituents of atmospheric
particulate matter. Atmos. Environ. 10:603, 1976."
Kieselbach, R. Microdetermination of nitrates by the Devarda method. Ind. Eng. Chem. Anal.
Ed. 16:764-766, 1944.
Kittelson, D. B. , R. McKenzie, M. Vermeersch, f. Dorman, D. Pui, M. Linne, B. Liu, and K.
Whitby. Total sulfur aerosol concentration with an electrostatically pulsed flame photo-
metric detector system. Atmos, Environ. 12:105, 1978
Knight, G. and Lichti, K. Comparison of cyclone and horizontal elutriator size selectors.
Am. Ind. Hyg. Assoc. J. 31:437-441, 1970.
Knight, N. Nitrogen in rain and snow. Proc. Iowa Acad. Sci. 18:75-77, 1911.
Kohler, A., and W. Fleck. Comparative dust-precipitation and dust-concentration measurements,
Staub Reinhalt. Luft 26(3):21-28. 1966.
Kolthoff, I. M. , E. B. Sandell, E. J. Meehan, and S. Bruckenstein. Quantitative Chemical
Analysis, 4th ed., MacMillan, Toronto, Canada, 1969. pp. 992-994.
Koschmieder, H. Beitr. Phys. Freien Atmos. 12:33-55, 171-181, 1924.
Kretzschmar, K. Comparison between three different methods for the estimation of the total
suspended matter in urban air. Atmos. Environ. 9j931-935, 1975.
Kronebach, G. W. WHO letter 31,675/R/lOP, Report of the Expert Meeting on Wet and Dry Deposi-
tion (Toronto, 17-21 November 1975)(WADEM), December 12, 1975.
Krupa, S. V., M. R. Coscio, Jr., and F. A. Wood. Evidence for multiple hydrogen-ion donor
systems in rain. In: First International Symposium on Acid Precipitation and the Forest
Ecosystem, U.S. Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975.
Water Air Soil Pollut. 6:415-422, 1976.
Laby, R. H., and T. C. Morton. Estimation of nitrates by nitration of 7-hydroxy-4, 8-dimethyl
coumarin. Nature (London) 210:298-299, 1966.
Lambert, J. P. F., and F. W. Wilshire. Neutron activation analysis for simultaneous determi-
nation of trace elements in ambient air collected on glass-fiber filters. Anal. Chem.
51:1346-1350, 1979.
Lamothe, P. J. , and R. K. Stevens. Sulfuric acid analysis using low temperature volatiliza-
tion. Presented at the 171st National Meeting, American Chemical Society, New York, NY,
April 4-9, 1976. Paper no. ENVR-38.
3-107
-------
Lawrence Berkeley Laboratory, Instrumentation for Environmental Monitoring, Air. LBL-1, Vol.
1, Part 1, University of California, Berkeley, CA, May 1972.
Lawrence Berkeley Laboratory. Instrumentation for Environmental Monitoring, Air. LBL-1, Vol.
1, Part 2, University of California, Berkeley, CA, July 1975.
Lazrus, A. L., K. C. Hill, and J. P. Lodge. In: Proceedings of the 1966 Technicon Symposium
on Automation in Analytical Chemistry. Mediad, Inc., New York, NY, 1966. p. 291.
Lazrus, A., E. Lorange, and J. P. Lodge, Jr. New automated microanalyses for total inorganic
fixed nitrogen and for sulfate ion in water. In: Trace Inorganics in Water, R. F.
Gould, ed., American Chemical Society, Washington, DC, 1968. pp. 164-171. Advances in
Chemistry Series 73.
Leahy, D., R. Siegel, P. Klotz, and L. Newman. The separation and characterization of sulfate
aerosol, Atmos. Environ. 9:219-229, 1975.
Lee, R. E. Size distribution of suspended particulates in air. Research/Development
23(6):18-21, 1972.
Lee, R. E., and S. Goranson. National Air Surveillance Cascade Impactor Network. I. Size
distribution measurements of suspended particulate matter in air. Environ. Sci. Techno!.
6:1019-1024, 1972.
Lee, R. E. , Jr., and J. Wagman. A sampling anomaly in the determination of atmospheric sul-
fate concentration. Am. Ind. Hyg. Assoc. J. 27:266-271, 1966.
Liberti, A., M. Possanzini, and M. Vicedamini. Determination of the nonvolatile acidity of
rainwater by a coulometric procedure. Analyst 47:352, 1972.
Likens, G. E. Acid precipitation. Chem. Eng. News 54(48):29-44, 1976.
Likens, G. W., F. H. Bormann, and N. M. Johnson. Acid rain. Environment 14:33-40, 1972.
Lin, C. I., M. Baker, and R. J. Charlson. Absorption coefficient of atmospheric aerosol: a
method for measurement. Appl. Opt. 12, 1356-63, 1973.
Lioy, P. J., J. G. Watson, and J. D. Spengler. APCA Specialty Conference Workshop on Baseline
Data for Inhalable Particulate Matter. J. Air Pollut. Control Assoc. 30:10, 1980. pp.
1126-1130.
Lippmann, M. Respirable dust sampling. Am. Ind. Hyg. Assoc. J. 31:138-159, 1970.
Lippmann, M., and T. L. Chan. Cyclone sampler performance, Staub Reinhalt. Luft 39:7-11,
1979.
Liu, B, Y, H., and D. Y. H. Pui. On the performance of the Electrical Aerosol Analyzer. J.
Aerosol Sci. 6:249-264, 1975.
Liu, B. Y. H, D. Y. H. Pui, K. L. Rubow, and G. A. Kuhlmey. Research on Air Sampling Filter
Media: Progress Report - Grant Report R804600, University of Minnesota, Minneapolis, MN,
May 1978a.
Liu, B. Y. H. , D. Y. H. Pui, K. T. Whitby, D. B. Kittelson, Y. Kousaka, and R. L. McKenzie.
The aerosol mobility chromatograph, a new detector for sulfuric acid aerosols. Atmos.
Environ. 12:99, 1978b.
3-108
-------
Liu, B. Y. H., D. Y. H. Pui, and B. W. Loo. A Sampling Inlet for Inhalable Particles (in
press). Particle Technology Laboratory, University of Minnesota, 1980.
Liu, B. Y. H., and D. Y. H. Pui. Aerosol sampling inlets and inhalable particles. In: Pro-
ceedings: Advances in Particle Sampling and Measurement, U.S. Environmental Protection
Agency, Daytona Beach, FL, October 1979. EPA-600/9-80-004, U.S. Environmental Protection
Agency, Research Triangle Park, NC, January 1980. pp. 382-408.
Lizarraga-Rocha, J. A. The Effect of Exposure of Filters During Non-operating Periods in the
High Volume Sampling Method. M.S. Thesis, University of North Carolina, Chapel Hill, NC,
1976.
Lodge, J. P., Jr., A. P. Waggoner, D. T. Klodt, and C. N. Grain. Non-health effects of air-
borne particulate matter. Atmos. Environ. 15:431-482, 1981,
Lodge, J. P., Jr., J. Ferguson, and B. R. Havlik. Analysis of micron-sized particles: Deter-
mination of sulfuric acid aerosol. Anal. Chem. 32:1206-1207, 1960.
Lodge, J. P., Jr., J. B. Pate, and H. A. Huitt. The use of impregnated filters to collect
traces of gases in the atmosphere. Am. Ind. Hyg. Assoc. J. 24:380-387, 1963.
Lodge, J. P., Jr., K. C. Hill, J. B. Pate, E. Lorange, W. Basbergill, A. L. Lazrus, and G. S.
Swanson. Chemistry of United States Precipitation: Final Report on the National Precip-
itation Sampling Network. National Center for Atmospheric Research, Boulder, CO, August
1968.
Loo, B. W., J. M. Jaklevic, and F. S. Goulding. Dichotomous virtual impactors for large-scale
monitoring of airborne particulate matter. In: Fine Particles: Aerosol generators,
measurement, sampling, and analysis. B. Y. H. Liu, ed., Academic Press, New York, NY,
1976.
Loo, B. W. , R. S. Adachi, C. P. Cork, F. S. Goulding, J. M. Jaklevic, D. A. Landis, and W. L.
Searles. A second generation dichotomous sampler for large scale monitoring of airborne
particulate matter, LBL-8725, Lawrence Berkeley Laboratory, Berkeley, CA, January 1979.
Loo, B. W, , W. R. French, R. C. Gatti, F. S. Goulding, J. M. Jaklevic, J. Llacer, and A. C.
Thompson. Large scale measurement of airborne particulate sulfur. Atmos. Environ.
12:759-771, 1978.
Lundgren, D. A. Mass Distribution of Larger Atmospheric Particles. Ph.D., Thesis, University
of Minnesota, Minneapolis, MN, 1973,
Lundgren, D. A. , L. D. Carter, and P. S. Daley. Aerosol mass measurement using piezoelectric
crystal sensors. |n: Fine Particles: Aerosol Generation, Measurement, Sampling, and
Analysis. B. Y. H. Liu, ed. Academic Press, New York, NY, 1976. pp. 485-510.
Macias, E. S. Instrumental analysis of light element composition of atmospheric aerosols.
In: Methods and Standards for Environmental Measurement. W. Kirchhoff, ed. , National
Bureau of Standards Special Publication 464. Department of Commerce, Washington, DC,
1977. pp. 179-188.
Macias, E. S. , and R. B. Husar. Atmospheric particulate mass measurement with beta attenua-
tion mass monitor. Environ. Sci. Techno!. 10:904-907, 1976.
Macias, E. S. , C. D. Radcliffe, C. W. Lewis, and C. R. Sawicki. Proton induced X-ray analysis
of atmospheric aerosols for carbon, nitrogen, and sulfur composition. Anal. Chem.
50:1120-1124, 1978.
3-109
-------
Maddalone, R. F., G, L. McClure, and P. W, West, Determination of sulfate by thermal reduc-
tion of perimidylammonium sulfate. Anal. Chem. 47:316, 1975,
Malissa, H. Analysis of Airborne Particles by Physical Methods. CRC Press, Inc., Cleveland,
OH, 1978.
Mamane, Y., and R. G. de Pena. A quantitative method for the detection of individual sub-
micron size sulfate particles. In: Sulfur in the Atmosphere, Dubrovnik, Proceeding of
the International Symposium, Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:69-82,
1978. , ~
Harple, V. A., and K. Willeke. Impactor design. Atmos. Environ. 10:891-896, 1976.
Harple, V. A., and K, Willeke. Inertia! Impactors. Aerosol Measurement, University Presses
of Florida, Gainesville, FL, 1979.
Martin, A., and F. R. Barber. Some measurements of loss of atmospheric sulphur dioxide near
foliage. Atmos. Environ. 5:345-52, 1971.
Martin, B. E. Sulfur Dioxide Bubbler Temperature Study. EPA-600/4-77-040, U.S. Environmental
Protection Agency, Research Triangle Park, NC, August 1977,
McClenny, W. A., and C. A. Bennett, Jr. Integrative technique for detection of atmospheric
ammonia. Atmos. Environ. 14:641-645, 1980.
McColl, J. G., and D. S. Bush. Precipitation and throughfall chemistry in the San Francisco
Bay Area. J. Environ. Qua!. 7:352-357, 1978.
McCoy, R. A., D. E. Camann, and H. C. McKee. Collaborative Study of Reference Method for
Determination of Sulfur Dioxide in the Atmosphere (Pararosaniline Method) (24-Hour
Sampling). EPA-650/4-74-027, U.S. Environmental Protection Agency, Research Triangle
Park, NC, December 1973.
McCrone, W. C., and J. G. Delly. The Particle Atlas Edition Two. Volume I: Principles and
Techniques. Ann Arbor Science Publishers, Inc., Ann Arbor, MI, 1973. p. 259.
McFarland, A. R. Wind Tunnel Evaluation of British Smoke Shade Sampler: Test Report. Air
Quality Laboratory Report, 3565/05/79/ARM, Texas A&M University, College Station, TX, May
1979.
McFarland, A. R., and C. A. Ortiz. Aerosol Characterization of Ambient Particulate Samplers
Used in Environmental Monitoring Studies: Progress Report. Texas A&M University
Research Foundation, College Station, TX, October 1979.
McFarland, A. R., and C. E. Rodes. Characteristics of aerosol samplers used in ambient air
monitoring. Presented at 86th National Meeting, American Institute of Chemical
Engineers, Houston, TX, April 2, 1979.
McFarland, A. R. , C. A. Ortiz, and C. E. Rodes. Evaluation of the NASN Cascade Impactor
Ambient Air Monitor. Report 3565/07/78/ARM, Texas A&M, Air Quality Laboratory, College
Station, TX, 1978. Atmos. Environ., in press, 1980.
McKee, H. C., R. E. Childers, and 0. Saenz, Jr. Collaborative Study of Reference Method fpr
the Determination of Suspended Particulates in the Atmosphere (High Volume Method).
APTD-0904, U.S. Environmental Protection Agency, Research Triangle Park, NC, June 1971.
Menis, 0., D. L. Manning, and R. G. Ball. Automatic spectrophotometric titration of fluoride,
sulfate, uranium, and thorium. Anal. Chem. 30:1772-1776, 1958.
'3-110
-------
Miller, ,F. J. , D. E. Gardner, J. A. Graham, R. E. Lee, Jr., W. E. Wilson, and J. D. Bachmann.
Size considerations for establishing a standard for inhalable particles. J. Air Pollut.
Control Assoc. 29:610-615, 1979.
Miller, J. M. , and V. R. Highsmith. The WMO Regional Precipitation Chemistry Network in the
United States. NOAA-TM-ERL ARL-60, U. S. Department of Commerce, National Oceanic and
Atmospheric Administration, Silver Spring, MD, October 1976.
Miller, K. , and H. W. DeKoning. Particle sizing instrumentation. Presented at 67th Annual
Air Pollution Control Association meeting, June 9, 1974, Denver, CO. Paper no. 74-48.
Moffat, A. J. , J. R. Robbins, and A. R. Barringer. Electro-optical sensing of environmental
pollutants. Atmos. Environ. E>:511-25, 1971.
Morris, A. W. , and J. P. Riley. The determination of nitrate in seawater. Anal. Chim. Acta
29:272-279, 1963.
Morrison, G. H., ed. Trace Analysis: Physical Methods. John Wiley and Sons, New York, NY,
1965. pp. 245-263, 271-310, 325-360, 377-425.
Mrose, H. Measurements of pH, and chemical analyses of rain- snow-, and fog-water. Tell us
18:266-70, 1966.
Mudgett, P. S. , L. W. Richards, and J. R. Roehrig. A new technique to measure sulfuric acid
in the atmosphere. In: Analytical Methods Applied to Air Pollution Measurement.
R. K. Stevens and W. F. Herget, eds., Ann Arbor Science Publishers, Inc., Ann Arbor,
MI, 1974.
Mulholland, G. W. , D. Y. H. Pui, A. Kapadia, and B. Y. H. Liu. Aerosol number and mass con-
centration measurements: - A comparison of the EAA with other measurement techniques. J.
, Colloid Interface Sci. 77(l):57-67, 1980.
Mulik, J. , R. Puchett, D. Williams, and E. Sawicki. Ion chromatographic analysis of sulfate
and nitrate in ambient aerosols. Anal. Lett. 9(7): 653-663, 1976.
Mulik, J. D. , G. Todd, E. Estes, R. Puckett, E. Sawicki, and D. Williams. Ion Chromatographic
Determination of Atmospheric Sulfur Dioxide. In: Ion Chromatographic Analysis of Envi-
ronmental Pollutants. E. Sawicki, J. D. Mulik, and E. Wittgenstein, eds., Ann Arbor
Science Publishers, Inc., Ann Arbor, MI, 1978. pp. 23-40.
Mueller, P. K. , R. W. Mosley, and L. B. Pierce. Carbonate and noncarbonate carbon in atmos-
pheric particles. In: Proceedings of the 2nd International Clean Air Congress, Inter-
national Union of Air Pollution Prevention Associations, Washington, DC, December 6-11,
1970. H. M. Englund and W. T. Beery, eds., Academic Press, New York, NY, 1971. pp.
532-539.
Mullin, J. B. , and J. P. Riley. The spectrophotometric determination of nitrate in natural
waters, with particular reference to seawater. Anal. Chim. Acta 12:464-480, 1955.
Nader, J. S. Dust-retention efficiencies of dust fall collectors. J. Air Pollut. Control
Assoc. 8:35-38, 1958.
National Research Council, Subcommittee on Airborne Particles. Airborne Particles. Univer-
sity Park Press, Baltimore, MD, 1979.
National Air Pollution Control Administration. Air Quality Criteria for Particulate Matter.
NAPCA Publication No. AP-49, U.S. Department of Health, Education, and Welfare,
Washington, DC, 1969.
3-111
-------
National Air Pollution Control Administration. Air Quality Criteria for Sulfur Oxides, NAPCA
Publication No. AP-50, U.S. Department of Health, Education, and Welfare, Washington, DC,
1970.
Natusch, D. F. S., and J. R. Wallace. Determination of airborne particle size distributions:
Calculations of cross sensitivity and discreteness effects in cascade impaction. Atmos.
Environ. 10:315-324, 1976.
Natusch, D. F. S., C. F. Bauer, and A. Loh. Collection and analysis of trace elements in the
atmosphere. In; Air Pollution Control, Part III. W. Strauss, ed., John Wiley and Sons,
Inc., New York, NY, 1978.
Niemann, B. L., J. Root, N. Van Zwalenburg, and A. L. Mahan. An Integrated Monitoring Network
for Acid Deposition: A Proposed Strategy. Report R-023-EPA-79, Teknekron Research [nc.,
Berkeley, CA, November 1979.
Noll, K, E. A rotary inertia! impactor for sampling giant particles in the atmosphere.
Atmos. Environ. 4:9-20, 1970.
Norwegian Institute for Air Research (NIAR). LRTAP 5/71. Coulometric Titration of Strong
Acid in Precipitation, 1971.
Novakov, T. Chemical characterization of atmospheric pollution particulates by photoelectron
spectroscopy. In: Proceedings of the 2nd Joint Conference on Sensing of Environmental
Pollutants, Instrument Society of America, Washington, DC, December 10-12, 1973. p. 197.
Novakov, T. , S. G. Chang, and A. B. Harker. Sulfates as pollution particulates: catalytic
formation on carbon (soot) particles. Science (Washington, DC) 186:259-261, 1974.
O'Brien, R. J. , J. R. Holmes, R. J. Reynolds, J. W. Remoy, and A. H. Bockian. Analysis of
photochemical aerosols in Los Angeles basin according to particle size. Presented at the
67th Annual Meeting, Air Pollution Control Association, Denver, Colorado, 1974. Paper
no. 74-155.
Okabe, H., P. L. Splitstone, and J. J. Ball. Ambient and souce S0» detectors based on a
fluorescence method: J. Air Pollut. Control Assoc. 23:514-516, 1973.
O'Keeffe, A. £., and 6. C. Ortman. Primary standards for trace gas analysis. Anal. Chem.
38:760-63, , 1966.
Okita, T. , S. Morimoto, M. Izawa, and S. Konno. Measurement of gaseous and particulate
nitrates in the atmosphere. Atmos. Environ. 10:1083-1089, 1976.
Orel, A. E. and J. H. Seinfeld. Nitrate formation in atmospheric aerosols. Environ. Sci.
Techno!. 11:1000-7, 1977.
Orenstein, A. J., ed. Proceedings of the Pneumoconiosis Conference, the University of Wit-
watersand, Johannesburg, South Africa, February 9-24, 1959, Little, Brown, Boston, 1960.
p. 619.
Pashel, G. E., and D. R. Egner. A comparison of ambient suspended particulate matter concen-
trations as measured by the British smoke sampler and the high volume sampler at 16 sites
in the United States. Atmos. Environ. 15:919-27, 1981.
Pate, J. B., J. P, Lodge, Jr., and M. D. Neary. The use of impregnated filters to collect
traces of gases in the atmosphere. Anal. Chim. Acta 28:341, 1963.
3-112
-------
Perera, F. P. , and A. K. Ahmed. Respirable Particles: Impact of Airborne Fine Participates
on Health and the Environment. National Resources Defense Council, Inc., Washington, DC,
October 1978.
Philips Electronic Instruments. Coulometric amperometric analyzers. Bulletin. WA-4-C13, un-
dated.
Pickerell, D., T. Hook, T. Dolzine, and J. K. Robertson. Intensity-weighted sequential sam-
pling of precipitation: a technique for monitoring changes in storm chemistry during a
storm. Ion Chromatogr. Anal. Environ. Pollut. 2:289,94 1979.
Pierson, W. R., R. H. Hammerle, and W. W. Brachaczek. Sulfate formed by interaction of sulfur
dioxide with filters and aerosol deposits. Anal. Chem. 48:1808-1811, 1976.
Pierson, W. R., W. W. Brachaczek, T. J. Korniski, T. J. Truex, and J. W. Butler. Artifact
formation of sulfate, nitrate, and hydrogen ion on backup filters: Allegheny Mountain
experiment. J. Air Pollut. Control Assoc. 30:30-34, 1980.
Ranade, M. B. , and D. VanOsdell. Quality Assurance Evaluation of the CHAMP Aerosol Sampler.
Project Report 43U-1487-59, Research Triangle Institute, Research Triangle Park, NC, July
1978.
Ratzlaff, K. L., and D. F. S. Natusch. Theoretical assessment of precision in dual wavelength
spectrophotometric measurement. Anal. Chem. 49:2170-2176, 1977.
Rayner, A. C. A rapid titrimetric method for sulfate in lead dioxide exposed to atmospheric
sulfur pollution. J. Air Pollut. Control Assoc. 16:418-421, 1966.
Regan, G. F., S. K. Goranson, and L. L. Larson. Use of tape samplers as fine particulate
monitors. J. Air Pollut. Control Assoc. 29:1158-1160, November 1979.
Richardson, H. L. The nitrogen cycle in grassland soils; with special reference to the Roth-
amsted Park grass experiment. J. Agric, Sci. 28:73-121, 1938.
Ridder, T. B. On the Chemical Composition of Rain Fall: Comparison of Measurement Results.
KNMI-WR-78-4, Royal Netherlands Meteorological Inst., DeBilt, The Netherlands, 1978.
Rodes, C. E. , and G. F. Evans. Summary of LACS Integrated Measurements. EPA-600/4-77-034,
U.S. Environmental Protection Agency, Research Triangle Park, NC, June 1977.
Rodes, C. E., H. F. Palmer, L. A. Elfers, and C. H. Morris. Performance characteristics of
instrumental methods for monitoring sulfur dioxide. J. Air Pollut. Control Assoc.
19:575-584, 1969.
Rosen, H. , and T. Novakov. Identification of primary particulate carbon and sulfate species
by ramon spectroscopy. Atmos. Environ. 121:923-926, 1978.
Rosen, H., A. D. A. Hansen, L. Gundel, and T. Novakov. Identification of the optically absorb-
ing component in urban aerosols. Appl. Opt. 17:3859-61, 1978.
Ross, W. D. , G. W. Buttler, T. G. Duffy, W. R. Rehg, M. T. Wininger, and R. E. Sievers.
Analysis for aqueous nitrates and nitrites and gaseous oxides of nitrogen by electron
capture gas chromatography. J, Chromatogr. 112:719-727, 1975.
Ruzicka, J. , and E. H. Hansen. New potentiometric gas sensor. Air-gap electrode. Anal.
Chim. Acta. 69:129-41, 1974.
3-113
-------
Ruzicka, J., E. H. Hansen, P. Bisgaard, and E. Raymann. Determination of the ammonium content
in waste waters by means of the air-gap electrode. Anal. Chim. Acta. 72:215-219, 1974.
Ryan, M. D., and D. D. Siemer. Anodic stripping voltammetry of atmospheric samples. Pre-
sented at 3rd Annual Meeting Federation Analytical Chemistry and Spectroscopy Society,
1976.
Saltzman, B. E. Colorimetric microdetermination of nitrogen dioxide in the atmosphere. Anal.
Chem. 26:1949-1955, 1954.
Sawicki, E. Separation and analyses of polynuclear aromatic hydrocarbons present in the human
environment. Chemist. Analyst 53:24-26, 28-30, 56-S2, 88-91, 1964.
Scaringelli, F. P., and K. A. Rehme. Determination of atmospheric concentrations of sulfuric
acid aerosol by spectrophotometry, coulometry, and flame photometry. Anal. Chem.
41:707-13, 1969.
Scaringelli, F. P., B. E. Saltzman, and S. A. Frey. Spectrophotometric determination of
atmospheric sulfur dioxide. Anal. Chem. 39:1709-1719, 1967.
Schulze, F. Versatile combination ozone and sulfur dioxide analyzer. Anal. Chem. 38:748-752,
1966.
Scott, D. R., W. A. Loseke, L. E. Holboke, and R. J. Thompson. Analysis of atmospheric parti-
culates for trace elements by optical emission spectrometry. Appl. Spectrosc.
30(4):392-405, 1976.
Shaw, R. W., T. G. Dzubay, and W. J. Courtney. Reproducibility of mass measurements by dicho-
tomous sampler. Technical Note. U.S. Environmental Protection Agency, 1979.
Shaw, R, W., and R. D. Willis. X-ray emission analysis: sample losses during excitation.
In; Electron Microscopy and X-Ray Applications. P. A. Russell and A. E. Hutchings,
eds., Ann Arbor Science Publishers, Inc., Ann Arbor, MI,,1978.
Shendrikar, A. D., and J. P. Lodge, Jr. Microdetertnination of ammonia by the ring oven tech-
nique and its application to air pollution studies. Atmos. Environ. 9:431, 1975.
Sides, J. 0., and H. F. Saiger. Effects of Prolonged Static Exposure of Filters to Ambient
Air on High Volume Sampling Results. Kansas State Department of Health and Environment,
Bureau of Air Quality and Occupational Health, Topeka, KS, May 1976.
Sinclair, D., R. J. Countess, B. Y. H. Liu, and D. Y. H. Pui. Automatic analysis of submicron
aerosols. I_n: Aerosol Measurement, a Workshop, University of Florida, Gainesville, FL,
March 24-26, 1976. D. A. Lundgren, M. Lippmann, F. S. Harris, Jr., W. E. Clark, W. H.
Marlow, and M. D. Durham, eds. , University Presses of Florida, Gainesville, FL, 1979.
pp. 544-563.
Skoag, D. A., and D. M. West. Analytical Chemistry. Holt, Rinehart and Winston, New York,
NY, 1965.
Skujins, J. J. Spectrophotometric determination of nitrate with 4-methylumbel!iferon. Anal.
Chem. 36:240-241, 1964.
Small, H., T. S. Stevens, and W. C. Bauman. Novel ion exchange chromatographic method using
conductimetric detection. Anal. Chem. 47:1801-1809, 1975.
Smith, A. F., D. G. Jenkins, and D. E. Cunningworth. Measurement of trace quantities of
hydrogen sulfide in industrial atmospheres. J. Appl. Chem. 11:317, 1961.
3-114
-------
Smith, F., and A. C. Nelson, Jr. Guidelines for Development of a Quality Assurance Program.
Reference Method for the Determination of Suspended Particulates in the Atmosphere (High
Volume Method). EPA-R4-73-028b, U.S. Environmental Protection Agency, Research Triangle
'Park,' NC, June 1973.
Smith, F., P. S. Wohlschlegel, R. S. C. Rogers, and D. J. Mulligan. Investigation of Flow
Rate Calibration Procedures Associated with the High Volume Method for Determination of
Suspended Particulates. EPA-600/4-78-047, U.S. Environmental Protection Agency, Research
Triangle Park, NC, June 1978.
Snell, F. D. Photometric and Fluorimetric Methods of Analysis Metals, Part I. John Wiley and
Sons, New York, NY, 1978. p. 1-25.
Spicer, C. W. The Fate of Nitrogen Oxides in the Atmosphere. EPA-600/3-76-030. U.S. Envi-
ronmental Protection Agency, Research Triangle Park, NC, March 1976.
Spicer, C. W., and P. M. Schumacher. Interferences in sampling atmospheric particulate
nitrate. Atmos. Environ. 11:873-876, 1977.
Spicer, C. W., and P. M. Schumacher. Particulate nitrate: laboratory and field studies of
major sampling interferences. Atmos, Environ. 13:543-552, 1979.
Spicer, C. W., P. M. Schumacher, J. A. Kouyoumjian, and D. W. Joseph. Sampling and Analytical
Methodology for Atmospheric Particulate Nitrates. EPA-6QO/2-78-067, U.S. Environmental
Protection Agency, Research Triangle Park, NC, April 1978.
Stalker, W. W., R. C. Dickerson, and G. D. Kramer. Atmospheric sulfur dioxide and particulate
matter: a comparison of methods of measurements. Am. Ind. Hyg. Assoc. J. 24:68-79,
1963.
Stauff, J., and W. Jaeschke. A chemiluminescence technique for measuring atmospheric trace
concentrations of sulfur dioxide. Atmos. Environ. 9:1038-1039, 1975.
Stelson, A. W., S. K. Friedlander, and J. H. Seinfeld, A note on the equilibrium relationship
between ammonia and nitric acid and particulate ammonium nitrate. Atmos. Environ.,
13:369, 1979.
Stevens, R. K, , and T. G. Dzubay. Dichotomous Sampler - A Practical Approach to Aerosol Frac-
tination and Collection. EPA-600/2-78-112, U.S. Environmental Protection Agency,
Research Triangle Park, NC, June 1978.
.Stevens, R. K., T. G. Dzubay, G. Russwurm, and D. Rickel. Sampling and analysis of atmos-
pheric sulfates and related species. In: Sulfur in the Atmosphere, Proceedings of the
International Symposium, United Nations, Dubrovnik, Yugoslavia, September 7-14, 1977.
Atmos. Environ. 12:55-68, 1978.
Stevens, R. K., J. D. Mulik, A. E. O'Keeffe, and K. J. Krost. Gas chromatography of reactive
sulfur gases in air at the parts-per-billion level. Anal. Chem. 43:827-831, 1971.
Stevens, R. K., A. R. McFarland, J. B. Wedding. Comparison of a virtual dichotomous sampler
with a Hi-Volume sampler. (Environ. Sci. Res. Lab., EPA, Research Triangle Park, NC,
Prepr. Pa. Natl. Meet,, Div. Environ. Chem., Am. Chem. Soc. 1976.
Stratmann, H. Microanalytical method for determining sulfur dioxide in the atmosphere.
Microchim. Acta, 668-78, 1954.
Strickland, J. D. H., and T. R. Parsons. A Practical Handbook of Seawater Analysis, 2nd ed.
Bulletin 167, Fisheries Ressearch Board, Ottawa, Canada, 1972.
3-115
-------
Summers, P. W., and D. H. Whelpdale. Acid precipitation in Canada. Ini: First International
Symposium on Acid Precipitation and the Forest Ecosystem, U.S. Department of Agriculture
and Others, Columbus, Ohio, May 12-15, 1975. Water, Air, Soil Pollut. 6:447-455, 1976.
Swanson, D., C. Morris, R. Hedgecoke, R. Jungers, R. J. Thompson, and J. E. Bumgarner. A
rapid analytical procedure for the analysis of BaP in environmental samples. Trends
Fluoresc. 1:22-27, 1978.
Sweitzer, T. A. The evaluation of gas bubbler field performance. Presented at the 32nd
Annual Meeting of the East Central Section, Air Pollution Control Association, Dayton,
OH, September 17-19, 1975.
Tanner, R. L., J. Forrest, -and L. Newman. Determination of atmospheric gaseous and particu-
late sulfur compounds. In: Sulfur in the Environment: Part I, The Atmospheric Cycle.
J. 0. Nriagu, ed., John Wiley & Sons, Inc., New York, NY, 1978. pp. 371-452.
Tanner, R. L., T. D'Qttavis,, R. Garber, and L, Newman. Determination of ambient aerosol
sulfur using a continuous flame photometric detection system. Part I, sampling system
for aerosol sulfate and sulfuric acid. Atmos. Environ. 14:121-127, 1980.
Tanner, R. L., R. Cederwall, R. Garber, D. Leahy, W. Marlow, R. Meyers, M. Phillips, arid L.
Newman. Separation and analysis of aerosol sulfate species at ambient concentrations.
Atmos. Environ. 11:955-966, 1977.
Tanner,' R. L., R. W., Garber, W. H. Marlow, B. Leaderer, D. Eatough, and M. A. Leyko. Chemical
composition and size distribution of sulfate as a function of particle size in New York
City aerosol. Presented at the Industrial Hygiene Conference, New Orleans, Louisiana,
May 22-27, 1977a. Paper no. 140.
Technicon Corporation. Sulfate Method VIb Via Turbidimetry. Technicon Corporation, Tarry-
town, NY, 1959.
Technicon Industrial Systems, Method 100-70W, Tarrytown, NY, 1973.
Terraglio, F. P., and R. M. Manganelli. Laboratory evaluation of sulfur dioxide methods and
v the influence of ozone oxides of nitrogen mixtures. Anal. Chem, 34:675-677, 1962.
Tetlow, J. A., and A. L. Wilson. An absorptiometric method for determining ammonia in boiler
feed-water. Analyst 89:453-465, 1964.
Thoen, G. N., G. G. Dehaas, and R. R. Austin. Instrumentation for quantitative measurement of
sulfur compounds in kraft gases. Tappi 51:246-249, 1968.
Thomas, J. F., B. D. Tebbens, E. N. Sanborn, and J. M. Cripps. Fluorescent spectra of
aromatic hydrocarbons found in polluted atmosphere. Int. J. Air Pollut. 2:210, 1960.
Thomas, R. L., V. Dharmarajan, G. L. Lundquist, and P. W. West. Measurement of sulfuric acid
aerosol, sulfur trioxide, and the total sulfate content of the ambient air. Anal. Chem.
48:639-642, 1976.
Thompson, R. J. Second Analysis on References Precipitation Samples. U.S. Environmental Pro-
tection Agency, private communication 78, Research Triangle Park, NC, May 22, 1978.
Freon, J. F., and W. E. Crutchfield, Jr. Rapid turbidmetric method for determination of sul-
fates. Ind. Eng. Chem., Anal. Ed. 14:119-121, 1942.
3-116
-------
Tyree, S. Y. Rainwater acidity measurement problems. Atmos. Environ. 15:57-60, 1981.
Tyree, S. Y., M. Ballinger, S, E. Delos, and J. M. Stouffer. Analysis of Rainwater for Acci-
dents - Standards for and Methods of. EPA Grant Report R8Q449801. College of William
and Mary, Williamsburg, VA, 1979.
U.S. Atomic Commission, Division of Biomedical and Environmental Research. Chemist-Meteoro-
logist Workshop, Fort Lauderdale, FL, January 14-18, 1974. Available as WASH-1217-74
from National Technical Information Service, Springfield, VA.
U.S. Environmental Protection Agency. Ambient air monitoring reference and equivalent
methods. Notice of equivalent method designation. Fed. Regist. 40:34024-34025, 1975.
U.S. Environmental Protection Agency. National Primary and Secondary Ambient Air Quality
Standards. Appendix A - Reference method for the determination of sulfur dioxide in the
atmosphere (pararosaniline method). 40 CFR 50:6-11, July 1, 1979.
U.S. Environmental Protection Agency. Summary of Performance Test Results and Comparative
Data for Designated Equivalent Methods for S0«. QAD/M-79.12, Environmental Monitoring
Systems Laboratory, Research Triangle Park, NC,^December 1979a.
U.S. Environmental Protection Agency. Ambient Air Monitoring Reference and Equivalent
Methods. 40 CFR 53:952-986, July 1, 1979b.
U.S. Environmental Protection Agency. National Primary and Secondary Ambient Air Quality
Standards. Appendix B - Reference method for the determination of suspended particulates
in the atmosphere (high volume method). 40 CFR 50:12-16, July 1, 1979c.
U.S. Environmental Protection Agency. Methods for Chemical Analysis of Water and Wastes. EPA
600/4-79-020, U.S. Environmental Protection Agency, Environmental Monitoring and Support
Laboratory, Cincinnati, OH, March 1979d.
Volmer, W. , and F. Z. Frohlieh. Determination of small quantities of sulfate by turbidity
measurement. Z. Anal. Chem. 126:401-404, 1944.
von Lehmden, D. J. , and C. Nel.son. Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume I - Principles. EPA-600/9-76-005, U.S. Environmental Protection Agency,
Research Triangle Park, NC, May 1977.
Waggoner, A. P., and R. E. Weiss. Comparison of fine particle mass concentration and light
scattering in ambient aerosol. Atmos. Environ. 14:623-626, 1980.
Waggoner, A. P., M. B. Baker, and R. J. Charlson. Optical absorption by atmospheric aerosols.
Appl. Opt. 12:896, 1973.
Waller, R. E. Acid droplets in town air. Air Water Pollut. 7:773-8, 1963.
Waller, R. E., A. G. F. Brooks, and J. Cartwright. An electron microscope study of
particles in town air. Air Water Pollut. 7:779-86, 1963.
Warren Spring Laboratory. Measurement of Atmospheric Smoke and Sulphur Dioxide: Reproduci-
bility of Results. Stevenage, England, August 1962.
Warren Spring Laboratory. National Survey of Smoke and Sulphur Dioxide: Instruction Manual,
Stevenage, England, 1966.
Warren Spring Laboratory. The Investigation of Atmospheric Pollution 1958-1966. Thirty-
second Report. HMSO, London, England, 1967.
3-117
-------
Warren Spring Laboratory. National Survey of Air Pollution, 1961 - 1971. Vol. 5: Scotland -
Northern Ireland. Chapter 15: Accuracy of Data and Representativeness of National
Survey Data. Stevenage, England, January 1975. pp. 111-118.
Wartburg, A. F., J. B. Pate, and J. P. Lodge, Jr. An improved gas sampler for air pollutant
analysis. Environ. Sci. Technol. 3:767-768, 1969.
Washburn, H. W. , and R. R. Austin. Air Pollution. In: Proceedings of the U.S. Technical
Conference on Air Pollution. McGraw-Hill, New York,"NY, 1952.
Wedding, J. B. Ambient Aerosol Sampling: History, Present Thinking, and a Proposed Inlet for
Inhalable Particulate Matter. Presented at the 73rd Annual APCA Meeting, Montreal,
Canada, June 1980.
Wedding, J. B., A. R. McFarland, and J. E. Cernak. Large particle collection characteristics
of ambient aerosol samplers. Environ. Sci. Technol. 11:387-390, 1977.
Wedding, J. B., M. Weigand, W. John, and S. Wall. Sampling effectiveness of the inlet to the
dichotomous sampler. Environ. Sci. Technol. 14:1367-1370, 1980.
Wehry, E. L., and G. Mamantov. Matrix isolation spectroscopy. Anal. Chem. 51(6): 643A-636A,
1979.
West, P. W., and G. C. Gaeke. Fixation of sulfur dioxide as disulfitoroercurate (11) and sub-
sequent colorimetric estimation. Anal. Chem. 28:1816-1819, 1956.
West, P. W., and T. P. Ramachandran. Spectrophotometric determination of nitrate using chrom-
otropic acid. Anal. Chim. Acta 35:317-324, 1966.
Whitby, K. T., W. E. Clark, V. A. Marple, G. M. Sverdrup, G. J. Sem, K. Willeke, B. Y. H. Liu,
and D. Y. H. Pui. Characterization of California aerosols—I. Size distributions of
freeway aerosol. Atmos. Env. 9:463-482, 1975.
Whitby, K. T., and W. E. Clark. Electrical aerosol particle counting and size distribution
measuring system for the 0.015 to 1.0 mm size range. Tellus. 18:573-586, 1966.
Whitby, K. T. and B. Y. H. Liu. Advances in Instrumentation and Techniques for Aerosol Gener-
ation and Measurement. Particle Lab. Pub. No. 216; avaiable from Mech. Eng. Dept., Univ.
of Minn., Minneapolis, MN, 1973.
Whitby, K. T., and K. Willeke. Single particle optical counters: principles and field use.
In; Aerosol Measurement, a Workshop. University of Florida, Gainesville, FL, March
24-26, 1976. D. A. Lundgren, M. Lippmann, F. S. Harris, Jr., W. E. Clark, W. H. Marlow,
and M. D. Durham, eds. , University Presses of Florida, Gainesville, FL, 1979. pp.
145-182.
Willeke, K. , and J. J. McFeters. Calibration of the CHAMP Fractionator, Particle Technology
Publication No. 252, University of Minnesota, Minneapolis, MN, March 1975.
Wilsdon, B. H., and F. J. McConnell. The measurement of atmospheric sulfur pollution by means
of lead peroxide. J. Soc. Chem. Ind., 53:385-81, 1934.
Wilson, B. D. Nitrogen and sulfur in rainwater in New York. J. Am. Soc. Agron, 18:1108-1112,
1926.
Wood, E. D. , F. A. J. Armstrong, and F. A. Richards. Determination of nitrate in seawater by
cadmium-copper reduction to nitrate. J. Mar. Biol. Assoc. U.K. 47:23-31, 1967.
3-118
-------
Woods, D. C. Measurement of particulate aerosol mass concentration using a piezoelectric
crystal microbalance. In: Aerosol Measurement, D. A. Lundgren, and M. Lippmann, eds.
[Workshop]. University Presses of Florida, Gainesville, FL, 1979. pp. 119-130.
World Health Organization. World Health Organization Selected Methods of Measuring Air Pol-
lutants. WHO Offset Publication No. 24, World Health Organization, Geneva, Switzerland,
1976.
World Meteorological Organization, Manual No. 299, WMO Operations Manual for Sampling and
Analysis Techniques for Chemical Constituents in Air and Precipitation. WMO Geneva,
Switzerland, 1971.
World Meteorological Organization. Letter PR-2569. The Operation of the Precipitation Refer-
ence Laboratory for WMO Members, June 30, 1975.
World Meteorological Organization. Environmental Pollution Circular No. 6—Results of the
First Analysis of Reference Precipitation Samples. Geneva, Switzerland, October 25,
1976.
World Meteorological Organization. Report of the Third Analysis on Reference Precipitation
Samples. Geneva, Switzerland, April 10, 1980.
Yasa, Z. , N. M. Amer, H. Rosen, A. D. A. Hansen, and T. Novakov. Photoacoustic investigation
of urban aerosol particles. Appl. Opt. 18:2528, 1979.
Zall, D. M., D. Fisher, and M. Q. Garner. Photometric determination of chlorides in water.
Anal. Chem. 28: 1665-1668, 1956.
Zoccolillo, L., A. Liberti, and D. Brocco. Determination of polycylic hydrocarbons in air by
gas chroroatography with high efficiency packed columns. Atmos. Environ, 6:715, 1972.
3-119
-------
APPENDIX 3A
FILTER: Gelman Type A, glass fiber
AP, cm Hg .
V, cm/sec
D , pm
0.035
0.10
0.30
1.0
FILTER: Ghia S2
1
11.2
<0.0001
<0.0001
<0.0001
<0.0001
37PJ 02, teflon
1.5
16.9
PENETRATION
<0.0001
<0,0001
<0.0001
<0.0001
membrane, 2.0 nn
3
32.7
<0.0001
<0.0001
<0.0001
<0.0001
n pore
10
108
0.0008
0.00054
<0. 00007
<0.0002
&?, cm Hg
V, cm/sec
Dni Mm
P
0.035
0.10
0.30
1.0
FILTER: Whatman
1
23.4
<0.0002
<0. 00006
<0. 00007
<0. 00007
No. 1, cellulose
3
64.1
PENETRATION
0.0011
0.00008
<0. 00007
<0. 00009
fiber
10
187
0.0005
<0. 00024
<0. 00022
<0. 00008
AP, cm Hg
V, cm/sec
Dp, Mm
. 0.035
0.10
0.30
1.0
1
6.1
0.56
0.46
0.16
0.019
3
17.4
PENETRATION
0.52
0.43
0.044
0.034
10
47.6
0.34
0.13
0.0049
0.0044
30
102
0.058
0.0071
0.00051
0.00042
Table A-l. Fractional penetration by particle size and face velocity for three
selected filter types (Liu et al., 1978a).
3-120
-------
SCALE, in SCALE, cm
0
3-
4-J
-6
-8
-10
FLOW
BUG SCREEN
16X16 MESH
SHIELD
FLOW TO DICHOTOMOUS SAMPLER
Figure 3A-1, Early inlet for the dichotomous sampler.
Source: Stevens and Dzubay (1978).
3-121
-------
COLLECTING SURFACE-~
8 DIRECTIONAL.
VANES
I
ENTRANCE I
PLANE •—
I
44»H I i I
f
a
• INLET
HOUSING
i PARTICLE PATH
EXIT PLANE
Figure 3A-2. Wedding IPM inlet, section view, not to scale.
Source: Wedding (1980).
3-122
-------
FLOW
FILTER/
FLOW
CONTROLLER
FLOW
RECORDER
INLET COVER
Wt.~65 Ibs.
Figure 3A-3. TSP Hi-VoI.
3-123
-------
16.67 1/min
TOTAL F LOW, Q
SEALED HOUSING
ACCELERATION
NOZZLE
FRACTIONATION
ZONE
COLLECTION
NOZZLE
LARGE PARTICLE
FLOW.fQ
LARGE PARTICLE
COLLECTION FILTER"
1
r
SMALL PARTICLE
FLOW(l-f)Q
SMALL PARTICLE
-COLLECTION FILTER
TO F LOWM ETE R
AND PUMP
1.67 l/min
TO F LOWM ETE R
AND PUMP
15.0 l/min
Figure 3A-4. Dichotomous sampler separator.
Source: Loo etal. (1979)
3-124
-------
MAST SUPPORT
AND VACUUM LINE
CRITICAL ORIFICE
9.0 LITER/WIN.
RUBBER
VACUUM HOSE
CONNECTIONS
CYCLONE SEPARATOR
Figure 3A-5. Chess cyclone sampler and shelter assembly.
Source: Barnard (1976).
3-125
-------
AIB INLET
TO PUMP
47 mm
AFTER-FILTER
CYCLONE
47 mm
TOTAL FILTER
TO PUMP
Figure 3A-G. Assembly for sampling with a total filter and cyclone in parallel,
, Source: John et al. (1978)
3-126
-------
INLET
FILTER
FLOW
CONTROLLER
FLOW
RECORDER
FLOW
STANDARD HI-VOL
SAMPLER
Figure 3A-7. Size-Selective Inlet (SSI) Hi-Vol.
3-127
-------
AIR FLOW LINES
DUST FREE AIR
PENETRATING
DUST CLOUD
DUST TRAJECTORIES
(a)
75cm
DUCT FLOOR AREA 8400 cnT
AIR FLOW RATE 76 I/mm
24 TRAYS
(W
Figure 3A-8. The horizontal elutriator designed to match the BWIRC deposition curve.
Source: Hamilton and Walton (1961).
3-128
-------
STAGE 1 <
STAGE 2<
STAGE
AFTER
FILTER
NOZZLE
JET EXIT
PLANE
IMPACT1ON
PLATE
FILTER
TO VACUUM PUMP
Figure 3A-9. Schematic diagram of a cascade impactor.
Source: Marple and Willeke (1979).
3-129
-------
LARGE PARTICLE
FRACTIONATOR
IMPACTOR UNIT-—
SPACER
HANDLE
IMPACTOR
PLATE
\ \
FLOW SENSOR
COVER
X
.
INLET
T.
WATER DRAIN
,GLASS FIBER
IMPACTION SURFACE
FINAL FILTER
VACUUM PUMP
Figure 3A-10. Cross section schematic of the CHAMP aerosol sampler.
Source: Ranade and Osdell (1978).
3-130
-------
ROTARY SEQUENCING VALVE
SUCTION PUMP
Schematic arrangement of sampling apparatus.
Sampler capable of sight-day sequential operation.
Figure 3A-11. British smoke shade sampler.
3-131
-------
PUSH TO TEST VALVE
J~A
THREi WAY VALVE
-------
10"3g, mg
SEASAND-
HUMAN HAIR-
o
p
cc
<
a.
LL
O
CD
O
O
10"9g, ng
1(T12g,pg
10"15g,fg
10'18g, ag
RAGWEED POLLEN.
POTATO STARCH-
CORN STAHCH-
RED BLOOD fELLS •
BACTERIA-
10A
100A
0.01 fan
lOOnm 0.01mm
0.1 jim 1 fim 10 nm
I I '
LOG PARTICLE DIAMETER (p = Ig/oc)
0.1 mm
10
1.0 mm
TRANSMISSION ELECTRON MICROSCOPE'
STEREOBINOCULAR
MICROSCOPE
MONO OBJECTIVE
OPTICAL MICROSCOPE
SCANNING ELECTRON MICROSCOPE
CHOICE OF MICROSCOPE FOR PARTICLE SIZE MEASUREMENT
Figure 3A-13. Relationship between particle size, diameter and number of
atoms for the light and electron microscope range.
Source: McCrone and Delly (1973).
3-133
-------
-------
4. SOURCES AND EMISSIONS
4.1 INTRODUCTION
This chapter highlights the magnitude and characteristics of natural and anthropogenic
sources and emissions of particulate matter and sulfur oxides. Natural emissions are defined
as those not caused by human activities, such as volcanoes and the biosphere. Manmade sources
include stationary point sources (e.g., utility power plants, industrial facilities, etc.),
fugitive industrial and non-industrial sources (e.g., roadway dust), and transportation
sources (e.g., vehicle exhausts). Each of these emissions categories is discussed further in
this chapter.
Chapters 4, 5, and 6 present the information concerning the relationship between
emissions and ambient air concentrations. The information in Chapter 4 concerning the sources
and emissions of particulate matter and sulfur oxides relates directly to the chapters in this
document which discuss pollutant effects on visibility, acidic deposition, and health.
Chapter 5 summarizes measured ambient pollutant concentrations and characteristics. Chapter 6
presents what is known about the complex processes that alter and disperse the emitted
substances as they move through the atmosphere.
The issue of the relationship between emission intensity and possible effects on humans
is important, but the proximity of emissions to humans can often be more important than
relative intensity. For example, mass emissions from residential fuel combustion (home
heating) and transportation sources are minor on a national level. Since they are emitted in
highly populated areas and close to ground level, however, they are more likely to affect
human health and welfare. On the other hand, dust from unpaved roads appears significant in
many areas, but unpaved roads are more prevalent in rural areas, and their influence tends to
be highly localized. Conversely, although some natural source emissions can be fairly intense
(volcanic ash or sulfur from marshlands, for example), their effects are lessened, in gereral,
because they tend to be distributed fairly broadly nationwide. Consequently, simple
comparisons of total national tonnages of manmade vs natural emissions will seldom reflect the
impact that localized manmade sources can have on an area's air quality. For these reasons,
certain manmade emission sources, particularly stationary point sources, have been given a
greater share of the attention in this chapter.
A number of other issues are discussed here briefly or not at all. Predictions of future
emissions trends have not been presented because of the complexities of supporting
assumptions. Documents in which adequate discussions of assumptions can be found are
referenced. In many cases, data on the particle size distribution and chemical composition of
particulate emissions are incomplete or inadequate. Available data have been briefly
summarized. Also, discussion of the effects of control devices on emission particle size
distributions has been limited. Documents that discuss these effects thoroughly are
referenced.
4-1
-------
4.2 DATA SOURCES AND ACCURACY
The most important information presented in this chapter concerns emission quantities and
characteristics. Though this information was gathered from the best and most recent
literature available, problems are still apparent. Specifically, estimates of emission
quantities vary, as do those of emission characteristics.
Emission quantities are typically estimated using emission factors. (The impossible
alternative would be direct measurement of pollutants from each emission point.) Emission
factors relate the quantity of pollutants emitted to an indicator of activity such as pro-
duction capacity or quantity of fuel burned. Because emission factors are statistical
averages, they do not precisely reflect emissions from individual sources. When a large
number of sources are considered, however, a reasonable estimate of total emissions can be
obtained. Therefore, the percentage error associated with emission estimates decreases as the
geographic area studied increases (from the local to the regional to the national scale).
Table 4-1 illustrates the variability of specific emission estimates that can be
associated with the use of different estimation approaches. In the table, "estimates" refers
to National Air Pollutant Emission Estimates, 1970-1978 (U.S. Environmental Protection Agency,
1980a). Emissions totals in that document were obtained from one calculation performed at the
national level by use of total national activity levels and national average emission factors
for each source category. "NEDS" (National Emissions Data System) refers to the 1977 National
Emissions 'Report (U.S. Environmental Protection Agency, 1980b). NEDS national emissions
totals were obtained by adding the emissions from individual facilities and may be affected by
mistakes, biases, and omissions. Therefore, the "Estimates" are judged by the U.S.
Environmental Protection Agency to be the most reliable national estimates presently
available. The NEDS State emissions totals, presented in Section 4.5.2 are the best available
at the State level. However, neither of these totals should be regarded as precise, for the
reasons given above. Nevertheless, such emission estimates do provide useful indications of
the relative contributions from various source categories.
TABLE 4-1. TWO EPA ESTIMATES OF 1977,-EMISSIONS OF
PARTICULATE MATTER AND SULFUR OXIDES (10° METRIC TONS PER YEAR)
Source Category
Fuel combustion
Industrial processes
Solid waste disposal
Parti cul ate
Estimates
4.8
6.4
0.5
Matter
NEDS
3.
3.
0.
Sulfur
Estimates
6
9
4
22
4
0
.2
,2
.0
Oxides
NEDS*-
22.
5.
0.
,1
.1
,0
4-2
-------
Attempts to obtain more than rough estimates of fugitive industrial and nonindustrial
participate emissions also present problems. Industrial process fugitive particulate
emissions, or process fugitives, include most industrial particulate emissions not passing
through a stack or another identifiable emission point. Process fugitive emissions are
difficult to estimate because of the lack of engineering data and adequate information on
emission factors. In one reference (Zoller et al., 1978) these emissions were estimated at
3.4 x 10 metric tons per year. However, the U.S. Environmental Protection Agency's
"Estimates" include "rough estimates of fugitive particulate emissions from industrial
processes." Therefore, the particulate emission estimates alluded to under the industrial
processes category discussed in this .chapter probably include part of the emissions listed as
process fugitives in Table 4-4.
Estimates of nonindustrial fugitive particulate emissions vary quite significantly in
some cases. Cooper et al. (1979) estimated annual emissions from entrainment of dust from
unpaved and paved roads at 290 x 10 metric tons and 7.2 x 10 metric tons, respectively.
The U.S. Environmental Protection Agency (1980b) estimated emissions from the same categories
at 35 x 10 metric tons and 4.7 x 10 metric tons, respectively. The differences can probably
be attributed to the use of different assumptions and methods of calculation.
Finally, emission estimates by particle size and chemical composition also vary depending
on the specific information and estimation method used. Most of these estimates are based on
data from emissions sampling and analysis studies. Although these studies probably exhibit a
high degree of accuracy on a case-by-case basis, data from a small number of individual
sources should not be used to make generalizations. Emission characteristics, as well as
emission quantities, are highly dependent on a number of source-specific factors, such as
source and fuel characteristics and operating conditions. For example, the size distribution
of particulate emissions from a given utility boiler can be altered significantly by changing
the boiler load. Therefore, the emission characteristics from a particular source could vary
from the information presented in this chapter.
4.3 NATURAL SOURCES AND EMISSIONS
Knowledge of natural sources and emissions of particulate matter and sulfur, including
sulfur oxides, is important for understanding air pollution. Baseline concentrations in
continental and marine air represent natural exposure levels and thus provide a reference for
comparing concentrations in air polluted by emissions from manmade sources.
Significant natural sources of particulate matter and sulfur, including reduced sulfur
which can become oxidized to sulfur oxides in the atmosphere (see Chapter 6), are terrestrial
dust, sea spray, the biosphere, volcanoes, and wildfires. Estimates of emissions from these
natural sources in the U.S. are described in more detail in subsequent sections. Table 4-2
presents a summary of natural source emission totals and characteristics.
4-3
-------
4.3.1 Terrestrial Oust
Terrestrial dust is transferred to the atmosphere by the action of wind- on the earth's
soils and crustal materials. Theoretical and experimental studies (Gillette, 1974) indicate
that sand grains, produced by the weathering of rocks and soils and moved by wind, cause the
pulverization of soil minerals, as in sandblasting, to produce particles. These particles may
become airborne and may be transported through the atmosphere for considerable distances. For
example, dust from the Sahara Desert may be carried by air currents across the Atlantic Ocean
as far as Florida and Barbados (Delany et a!., 1967; Junge, 1957).
The amounts of global terrestrial dust have been estimated at 180 x 10 metric tons per
fi *
year (Robinson and Robbins, 1971) and 100-500 x 10 metric tons per year (National Research
Council 1979). Calculations by Vandegrift et al. (1971), based on soil conservation data,
resulted in estimated U.S. natural dust emissions of 57 x 10 metric tons per year.
Terrestrial dust in the atmosphere is composed primarily of seven major elements, -
silicon, aluminum, iron, sodium, potassium, calcium, and magnesium; organic material; and
trace elements (Miller et al., 1972). The major elements are present in aerosol samples to
nearly the same extent as in earth crustal material (Miller et al., 1972; Lawson and
Winchester, 1979a). Atmospheric concentrations of many trace elements, however, are
10- to 1000-fold higher than would be expected from physical dispersion of soil materials.
These anomalous trace element enrichments have been observed in many parts of the world,
including northern Canada (Rahn, 1974), the South Pole (Zoller et al., 1974), and South
America (Adams et al., 1977). Table 4-3 summarizes geometric mean enrichment factors,
relative to aluminum, for various elements according to Rahn's compilation of all published
data up to 1976 (Rahn, 1976).
The atmospheric enrichment sources of these elements are unknown, but transport from
polluting industries (Rahn, 1974), natural rock volatility (Goldberg, 1976), and biogenic
emanations (Barringer, 1977) have all been suggested. In general, not enough is known about
element ratios in the natural atmosphere to detect a pollution component, thus a high
anomalous content in particulate material cannot be related arbitrarily to air pollutants.
Most terrestrial dust particles are greater than 2 pm in diameter (U.S. Environmental
Protection Agency, 1979). The major element constitutents of terrestrial dust also occur
principally as coarse particles. Size-fractionated particle samples indicate that more than
90 percent of the mass occurs on the first three impactor stages, representing particle sizes
of >4, 4-2, and 2-1 urn aerodynamic diameter (Winchester et al., 1979). The low relative
abundance of submicron silicon, iron, and other major dust constituents reflects the greater
*
Includes unknown amounts of indirect manmade contributions.
4-4
-------
TABLE 4-2. SUMMARY OF NATURAL SOURCE PARTICULATE AND SULFUR EMISSIONS"
Estimated U.S. <
(106 metric ton;
Source category Parti cul ate
Terrestrial dust 57
f*
Sea spray 5.5
Biosphere 20
Volcanoes Variable
Wildfires 0.5 - 1.0
TOTAL -V-84+
^missions
; per year). Parti cul ate characteristics
Sulfur Size range data Chemical composition
10% <1 fjrn Al, Ca, Fe, K, Mg, Na,
Si, organics, trace
elements
22% <3 (jm Seawater, organics
1.2 - 5.5 Unknown Organic aerosols, trace
metals
Variable ^5% <1 urn Al , Ca, Fe, K, Mg, Si,
Trace elements
80% <1 |jm Organics, trace minerals
., * A _
All data are referenced in the text.
When oxidized, one metric ton of sulfur equals 2 metric tons of sulfur dioxide, SO,,.
clncluding 0.7 x 10 metric tons of sulfate aerosol.
Predominantly reduced sulfur compounds.
-------
TABLE 4-3. AEROSOL ENRICHMENT FACTORS RELATIVE TO Al
EFa Elements
0.7-7 Li, Na, K, Rb
Be, Mg, Ca, Sr, Ba
Sc, Y, lanthanides
Al, Ga, Tl
Si, Ti, Zr, Hf, Th, U
Mn, Fe, Co, Nb
F, P
7-70b Cr, Cs, V, W, B, Ni, Ge
70-400b H, In, Cu, Mo, Bi, Zn, As
400-4000b I, Hg, S, Cl, Au, Ag, Sn, Sb,
Pb, Br, Cd, Te, Se, C, N
aGeometric means of element ratios to Al, relative to geochemical average earth
crustal material.
EF= (e1emen^A1 Aerosol
(el ernent/A 1) c r us t
Anomalously enriched elements arranged in order of increasing EF.
Source: Based on Rahn (1976).
4-6
-------
amount of energy needed In order for fine particles to be generated from soils by the
wind-driven sandblasting mechanism. This energy is normally not provided by the atmosphere
near the ground.
4.3.2 Sea Spray
Aerosol droplets are generated at the ocean surface by the action of wind, principally
through a process whereby air bubbles become entrained and rise to burst at the surface.
Robinson and Robfains (1971) estimated global emissions of particles from sea spray at 900 x
10 metric tons per year, including 120 x 10 metric tons of sulfate aerosol per year
(Eriksson, 1959, 1960; Robinson and Robbins, 1968). Assuming 10 percent of the annual
production penetrates continental areas (Eriksson, 1959), and assuming the impact on the U.S.
3 3
is proportional to the ratio of the U.S. to global coastline (12 x 10 miles: 200 x 10
miles), approximately 5.5 x 10 metric tons of sea spray particulate per year (including
0.7 x 10 metric tons of sulfate aerosol per year) impact on U.S. coastal areas.
Sea spray is composed of seawater, organic materials, and surface-active materials which
may be concentrated into the 0.05 to 0.5 um thickness of bubble surface (Maclntyre, 1974).
The surface-active material may be of natural or pollutant origin and may include organic
molecular films and organic and inorganic particles including viruses, bacteria, and other
microscopic organisms (Blanchard and Parker, 1977; Duce and Hoffman, 1976), Such materials,
by becoming components of sea spray aerosol droplets, may be carried through the atmosphere
far from the point of origin. The potential for virus transfer from coastal waters to the
atmosphere and transport by winds inland to inhabited areas has been demonstrated (Baylor et
al., 1977), although the process is not clearly understood.
Because of differences in the mechanics of droplet formation, differences ih chemical
composition may exist (Berg and Winchester, 1978). For example, some droplets may or may not
contain the surface-active particulate matter scavenged from the water column by the rising
bubbles. Chloride, bromide, and iodide also may be present (Moyers and Duce, 1972a,b), The
general processes through which sea spray droplets are formed and transported have been
described by a number of authors, including Blanchard and Woodcock (1957) and Wallace and
Hobbs (1977). The size distribution of sea spray particles by weight percent, as documented
by Taback et al. (1979), is as follows: >10 um, 24 percent; 3-10 (am, 54 percent; 1-3 urn, 20
percent; <1 um, 2 percent.
4.3.3 Biogenic Emanations
Plants emit particulate matter in the form of organic aerosols, trace metals, and
nutrients. Global emissions of volatile organic compounds released from plants have been
estimated at 200 x 10 metric tons per year (Went, 1960). The U.S. total would probably be
less than 20 x 10 metric tons per year. Isoprene derivatives such as terpenes, caroteniods,
and other compounds are believed to predominate and are likely to be partially oxidized,
resulting in blue haze and submicron condensation nuclei (Went, 1960; Went et al., 1967;
Rasmussen and Went, 1965; Schnell and Vali, 1972, 1973).
4-7
-------
Trace metals have long been known to occur in fluids secreted by plants. Radiotracer
strontium is transferred from plant foliage to the atmosphere, presumably in particles (Moorby
and Squire, 1963) which may be affected by electric fields (Fish, 1972). Transpiration caus.es
the transfer of both cations and anions to the atmosphere (Nemeryuk, 1970). Twenty-seven
trace elements have been identified in exudates from coniferous trees (Curtin et al., 1974).
Radiotracer experiments using zinc and lead show that particles greater than 5 pm in diameter
contain most of the metals released (Beauford et al., 1975, 1977). Sulfur, potassium, and
phosphorus have also been associated with tropical forests and occur in large aerosol
particles (Lawson and Winchester, 1979b). The metal content of plant-derived aerosols is so
high that several investigators suggest that it might serve as an indicator for geochemical
prospecting (Barringer, 1977; Curtin et al., 1974).
The terrestrial and marine biospheres, while not direct sources of sulfur dioxide, are
significant sources of reduced sulfur compounds. Volatile reduced sulfur compounds are
released to the atmosphere via microbiological processes and may become oxidized to S0? and
sulfate. The compounds released included hydrogen sulfide (H?S), dimethyl sufide (DMS),
dimethyl disulfide (DMDS), carbon disulfide (CS2), carbonyl sulfide (COS), and methyl
mercaptan (CH-SH) (Lovelock et al., 1972; Rasmussen, 1974; Lovelock, 1974; Adams et al.,
1979a). Transformation to H-S and thence to SCL and SOT is predicted (McElroy et al., 1980),
as well as direct oxidation to S02 and/or SO,.
The classic method of estimating biogenic sulfur emissions has been to identify the net
difference between input from known sources and removal by scavenging processes as being
indicative of an unmeasured source, namely a widespread source in the biosphere. A number of
previous estimates of global emissions of reduced sulfur compounds range from 64 x 10 metric
tons per year (land) and 27 x 10 metric tons per year (ocean) (Robinson and Robbins, 1968) to
3 x 10 metric tons per year (land) and 34 x 10 metric tons per year (ocean) (Granat et al.,
1976). Granat's estimate, scaled down to the U.S., would result in 0.2 x 10 metric tons per
year (land) and about 2-5 x 10 metric tons per year (ocean) (based on Galloway and Whelpdale,
1980). These estimates, however, were derived indirectly as balances for other sulfur fluxes.
Results of recent field monitoring studies conducted by Maroulis and Bandy (1977),
McClenny et al. (1979), and Adams et al. (1979a) yield slightly different estimates. Adams et
— o ™ 1
al. (1979b) calculated a mean annual sulfur flux of 0.02 g S m yr , weighted over a number
of Eastern U.S. soil types, including marshes. The entire U.S. would probably average
_p -i
0,02-0.05 g S m yr (Adams, 1980). When these numbers are applied to the entire earth's
land area (about 56 million square miles) the result is about 3-7 x 10 metric tons per year.
The land area of the U.S. (3.6 million square miles) emits about 0.2-0.5 x 10 metric tons per
-2 -1
year based on 0.02-0.05 ig S m yr . The impact of marine biogenic activity would be limited
primarily to coastal areas. Sulfur emissions from marine biogenic activity are probably cm
the order of 1 x 10 metric tons per year (based on Galloway and Whelpdale, 1980). These
lower estimates of sulfur emissions on a national scale do not preclude significant localized
4-8
-------
biogenic sulfur emissions, especially in areas where wetlands are prevalent (Henry and Hidy,
1980). Marshes and tidal flats may have high local sulfur-gas production (Hitchcock, 1976;
Hitchcock et al., 1980), but since the total area in the U.S. covered by such features is
relatively small, the contribution to total background sulfur is modest. Thus, the inventory
of terrestrial biogenic sulfur emission should not overemphasize the wetland areas as a
source. In the western U.S., more arid soils would be expected to have a much reduced sulfur
emission rate, but detailed study in this area is lacking. ,'
4.3.4 VolcanicEmissions
Emissions from volcanic eruptions and fumaroles may contribute to global atmospheric
background levels of particulate matter and sulfur. Volcanoes are one of the few sources of
atmospheric particles and sulfur whose effects can be felt at great distances. Plumes from
volcanoes intense enough to inject material into the upper troposphere or lower stratosphere
(about 10 to 15 miles above the earth's surface) have been tracked great distances before
removal (Fegley et al., 1980). The famous eruption of Krakatoa in 1883 injected enough dust
and sulfur into the stratosphere to cause brilliant sunsets thousands of kilometers away and a
global reduction of incoming solar radiation (Wexler, 1951a,b).
Until recently, volcanic activity has been relatively insignificant in the United States.
The Mt. St. Helens eruption was only the second this century in the contiguous United States.
Mt. Lassen, California, in 1915, was the first. About 20 volcanic eruptions have occurred in
Hawaii and Alaska since 1900.
With the 1980 eruption of Mt. St. Helens in the Pacific northwest, considerable attention
has been focused on the potential impact of volcanoes on the atmosphere and air quality. Away
from the immediate downwind area, volcanic impacts can probably be related to a cycle that
starts with injection into the stratosphere of dust and sulfur gases, oxidation and reaction
of the sulfur to form particulate compounds, and finally injection of the particles into the
troposphere, where they are scavenged. Since injection from the stratosphere occurs mainly in
low pressure systems, it is likely that precipitation scavenging predominates.
The average global volcanic emission rates of particles and sulfur compounds have been
estimated by a number of investigators. Robinson and Robbins (1971) estimated the average
global emission rate of small particles (the persistent fraction) at 3.6 x 10 metric tons per
year. Airborne measurements and observations made during the 1976 eruption of the St.
Augustine volcano (Alaska) led to particulate emissions estimates for a 1-year period for that
particular volcano of 6 x 10 metric tons for particles of 0.01 to 66 (jm in size and 0.25 x
10 metric tons for particles 0.01 to 5 (am in size (Stith et al., 1978).
Estimates of global volcanic sulfur emissions, as documented by Granat et al. (1976),
range from 0.75 to 3.75 x 10 metric tons per year. Emissions of SO, for a 1-year period at
St. Augustine were estimated at 0.1 x 10 metric tons (or 0.05 x 10 metric tons of sulfur)
(Stith et al., 1978). The St. Augustine volcano also emitted lesser quantities of H,S. Stith
£ ^
et al. (1978) estimated global volcanic emissions of H2S at 1 x 10 metric tons per year.
4-9
-------
Particles collected from the St. Augustine eruptions were composed primarily of silicon,
aluminum, magnesium, calcium, and iron. Trace amounts of potassium, titanium, and sulfur were
also present (Stith et a!., 1978). Samples of Mt, St. Helens ash contained mainly silicon,
aluminum, iron, calcium, sodium, magnesium, and potassium. Titanium, phosphorus, and
manganese, as well as traces of sulfur, chlorine, strontium, barium, vanadium, zirconium, and
zinc (among others) were also found (Fruchter et al., 1980).
Based on the St. Augustine particulate emissions (6 x 10 metric tons total, 0.25 x 10
metric tons less than 5 urn), less than 5 percent of the particles were smaller than 5 urn in
size (Stith, et al., 1978). According to preliminary airborne studies of Mt. St. Helens ash,
significant amounts of particulate matter between 1 and 2 (jm have been emitted to the
atmosphere (Hobbs et al., 1981). Other preliminary studies of Mt. St. Helens ash place the
fraction less than 3.5 urn at around 2 percent (Fruchter et al., 1980).
4.3.5 Wildfires
The three major types of large scale fires are: wildfires, prescription fires in natural
areas, and agricultural burning. The latter two types are exclusively caused by human
activities. Wildfires, defined by the U.S. Department of Agriculture Forest Service as "any
fire that burns uncontrolled in vegetative or associated flammable material," are treated
here, as in the literature, as a natural emission source even though man's activities cause
about 90 percent of their total number; only 10 percent are truly "natural," resulting from
lightning (U.S. Department of Agriculture, Forest Service, 1979).
Wildfire particulate emissions calculations typically have been based on three numbers:
wildfire acreage, fuel burned per acre, and emissions per unit mass of fuel. Robinson and
Robbins (1971) estimated yearly particulate emissions from torest fires in the U.S. as 0.7 x
10 metric tons, based on 4.5 x 10 acres burned, 18 tons of fuel per acre, and 17 pounds of
particulate per ton of fuel. Yamate (1973) arrived at 0.5 x 10 metric tons per year,
assuming numbers similar to those used by Robinson and Robbins.
In recent research, however, particulate emissions per unit mass of fuel were estimated
at 17-67 pounds per ton (GEOMET, 1978) and 80 pounds per ton (Radke et al., 1978), based on
airborne sampling studies in Oregon and Washington. Because emissions from fires are
dependent on fuel conditions and fire behavior (GEOMET, 1978), the estimates should probably
be averaged. If an emission rate of 40 pounds per ton, a U.S. wildfire acreage of 3.15 x 10
in 1977 (U.S. Department of Agriculture, Forest Service, 1979), and a U.S. average of 17 tons
of fuel per acre (Yamate, 1973) are assumed, U.S. particulate emissions from wildfires total
1.0- x 10 metric tons per year.
Chemical analysis of particulate matter from temperate forest burning indicates
approximately 50 percent benzene-soluble organic matter, 40 percent elemental carbon, and 10
percent mineral matter (Ryan and McMahon, 1976). Another analysis suggests 55 percent tar, 25
percent soot, and 20 percent ash (Vines et al., 1971). About 80 percent of the mass of smoke
particles from forest fires is less than 1 urn in diameter, with the average size being 0.1 pm
(GEOMET, 1978; Radke et al., 1978; Vines et al., 1971).
4-10
-------
Wildfires contribute varying amounts of other pollutants to the atmosphere. Carbon
monoxide and hydrocarbons are the most significant. Wildfires are not, however, considered to
be a source of sulfur oxides (Radke et al. , 1978; Yamate, 1973; Vines et al., 1971).
4.4 SOURCES AND EMISSIONS
A number of definable source categories emitting particulate matter and sulfur oxides can
be attributed solely to man and are the subjects of this section. Representative estimates of
emissions from these source categories in the United States are summarized in Table 4-4,
Manmade emissions of particulate matter result primarily from stationary point sources (fuel
combustion and industrial processes), industrial process fugitive particulate emission
sources, nonindustrial fugitive sources (e.g., roadway dust from paved and unpaved roads, wind
erosion of cropland), and transportation sources (e.g., automobiles). The data in Table 4-4
show that nonindustrial fugitive emissions are significant on a mass basis. However, the
relative impact of these emissions is somewhat lessened by their coarse size and the fact that
fugitive dust sources (e.g. unpaved roads) are more prevalent in rural areas.
TABLE 4-4. SUMMARY OF ESTIMATED ANNUAL MANMADE EMISSIONS (1978)
Source category
Stationary point sources
Industrial process fugitives
Nonindustrial fugitives
Transportation sources
TOTAL
Emissions (10
Particulate matter
10,5
3.3a
110-370
1.3
-125-385
metric tons)
Sulfur oxides
26.2
_
0.8
27.0
NOTE: Approximately half of the 3.3 x 10 metric tons of particulate matter
from process fugitives are probably included in the 10.5 x 10 metric
tons from stationary point sources. See Section 4.2 for explanation.
Source: U.S. Environmental Protection Agency, 1980a
Manmade emissions of sulfur oxides result almost exclusively from stationary point
sources. The combustion of fossil fuels by electric utilities causes most sulfur oxide
emissions. Transportation sources also contribute a small amount of sulfur oxide emissions.
4.4.1 Historical Emission Trends
Economic conditions and the degree to which air pollution control devices are used are
the two factors having the most impact on emissions totals, especially from stationary point
sources (fuel combustion and industrial processes). Economic conditions affect the amounts of
goods produced and, therefore, the amounts of emissions generated. The economics of relative
fuel prices also affect emissions; that is, higher prices on oil and natural gas cause
increased use of coal, which generally emits more particulate matter and sulfur oxides per
4-11
-------
unit energy than oil or natural gas. Increased use of control devices has resulted from the
enactment of regulations such as New Source Performance Standards and State Implementation
Plans.
Historical trends in emissions of particulate matter (not including fugitive emissions,
which have not been documented) and sulfur oxids are shown in Table 4-5. Data for the years
1940, 1950, and 1960 are from the U.S. Environmental Agency (1978b); data for 1970 through
1978 are from the U.S. Environmental Protection Agency (1980a). Emissions estimates from the
latter are considered more accurate. It should be noted that local emission trends might not
necessarily coincide with national emission trends.
Nationwide emissions of particulate matter (not including fugitive emissions) have
generally decreased since 1950 after a slight increase from 1940 to 1950, These emissions
have resulted primarily from stationary fuel combustion (utility and industrial) and
industrial processes. Particulate emissions from stationary fuel combustion decreased fairly
consistently from 1940 to 1978. From 1940 through the early 1970's this decrease was probably
due to "increased use of oil and natural gas. Even though the oil embargo of 1973-74 caused
increased use of coal, conservation efforts by industry and the installation of control
equipment resulted in further reductions in particulate emissions through 1978.
Industrial process emissions of particulate matter increased from 1940 to 1960, then
declined steadily through 1978. Increases were attributed to expanding production; decreases
were attributed to installation of controls.
Nationwide emissions of sulfur oxides have increased overall since 1940. As with
particulate matter, stationary fuel combustion (primarily utility and industrial) and
industrial processes (primarily ore smelting) have been the main contributors. Coal
combustion was the largest stationary fuel combustion source, although coal use by industrial,
commercial/institutional, and residential users has declined, corresponding with a decrease in
sulfur oxide emissions from those categories. Increased coal use by electric utilities has
snore than offset this decrease. Sulfur oxide emissions from electric utilities account for
more than half the total emissions. Flue gas desulfurization (FGD) systems have seen only
limited use to date and have not had a major impact on emissions. About 11 percent of U.S
coal-fired electrical generating capacity is presently fitted with FGD (U.S. Environmental
Protection Agency, 1980c).
Increased industrial production caused most of the sulfur oxide emission increases
through 1970. Since that time, however, significant emission reductions from nonferrous
smelters and sulfuric acid plants have occurred. For smelters, byproduct recovery of sulfuric
acid has significantly reduced sulfur oxide emissions. Sulfur oxide emissions from copper,
lead and zinc smelters have decreased from 4 x 10 metric tons per year in 1970 to about 2 x
10 metric tons per year in 1978.
Future emission trends are subject to a number of assumptions concerning economic
climate, fuel use, environmental policy, and control technology. These considerations are
beyond the scope of this document. (See U.S. Department of Energy, 1978; 1979.)
4-12
-------
TABLE 4-5 (a).gNATIONAL ESTIMATES OF PARTICIPATE EMISSIONS'1
(10 metric tons per year)
SOURCE CATEGORY
Stationary fuel
combustion
Industrial processes
Solid waste disposal
Transportation
Miscellaneous
TOTAL
1940
8.7
9,9
0.5
0.5
5.2
24.8
1950
8.1
12.6
0.7
1.1
3.7
26.2
1960
6.7
14.1
0.9
0.6
3.3
25.6
1970
7.2
12.8
1.1
1.1
1.0
23.2
1975
5.1
7.4
0.5
1.0
0.6
14.6
1978
3.8
6.2
0.5
1.3
0.7
12.5
Table 4-5 (b). NATIONAL ESTIMATES OF SULFUR OXIDE EMISSIONS
(10 metric tons per year)
SOURCE CATEGORY 1940 1950 1960 1970 1975 1978
Stationary fuel 15.1 16.6 15.7 22.7 20.9 22.1
combustion
Industrial processes
Solid waste disposal
Transportation
Miscellaneous
TOTAL
3.4
0.0
0.6
0.4
19.5
4.1
0.1
0.8
0.4
22.0
4.8
0.0
0.5
0.4
21.4
6.2
0.1
0.7
0.1
29.8
4.5
0.0
0.8
0.0
26.2
4.1
0.0
0.8
0.0
27.0
aDoes not include industrial process fugitive particulate emissions, and non-
industrial fugitives from paved and unpaved roads, wind erosion, construction
activities, agricultural tilling, and mining activities.
Includes forest fires, agricultural burning, coal refuse burning, and structural
fires.
Sources; U.S. Environmental Protection Agency (1978b)
U.S. Environmental Protection Agency (1980a)
4.4.2 Stationary Point Source Emissions
In this analysis of sources and characteristics of particulate and sulfur oxide emissions
from stationary point sources, the two major source categories are fuel combustion and
industrial processes. A third but minor category is solid waste disposal. Table 4-6 lists,
calculated estimates of 1978 emissions from these source categories. Based on these
estimates, fuel combustion contributed 36 percent of the particles and 84 percent of the
sulfur oxides emitted by stationary point sources in 1978. Industrial processes emitted 59
percent of the particulate matter and 16 percent of the sulfur oxides. Solid waste disposal
contributed 5 percent of the particulate matter.
4-13
-------
TABLE 4-6. 1978 ESTIMATES OF PARTICULATE AND SULFUR
OXIDE EMISSIONS FROM STATIONARY POINT SOURCES
Source category
Fuel combustion
Utility
Coal
Oil
Gas
Industrial
Coal
Oil
Gas
Other fuels
Commercial /Institutional
Coal
Oil
Gas
Residential
Coal
Oil
Gas
Industrial processes
Metals
Iron and steel
Primary smelting
Iron foundries
Other
Mineral products
Cement
Asphalt
Lime
Crushed rock
Other
Petroleum
Refining
Natural gas production
Chemicals
Sulfuric acid
Other
Other
Grain processing
Pulp and paper
Other
Solid waste disposal
TOTAL
Emissions
Parti cul ate
Matter
2,350
140
10
700
90
40
280
20
60
10
20
20
30
830
480
140
120
780
150
150
1,340
910
70
0
0
190
730
240
60
500
10,460
(10^ metric tons)
Sulfur oxides
15,900
1,720
0
1,890
1,150
0
150
40
900
0
60
260
0
110
1,960
0
0
670
0
0
0
30
900
140
220
0
0
80
0
0
26,180
Primarily wood/bark waste.
Source: U.S. Environmental Protection Agency (1980a).
4-14
-------
An unknown percentage of particulate matter and sulfur oxides is emitted as primary
sulfates. Primary sulfates consist of gaseous sulfur trioxide (SO,), sulfuric acid (H2S04),
and particulate sulfates. These primary sulfates are of increasing concern because of their
potential impacts, especially on health, but estimates of primary sulfate emission quantities
from major sources have not been generated to date. Secondary sulfates can be formed in the
atmosphere, following oxidation of S0», hours or even days after its release. The principal
reactions are described in Chapter 2; field measurements attempting to trace the actual
production and fate of secondary sulfates under the complex influences and diverse
combinations of meteorological variables are discussed in Chapter 6.
Varying amounts of particulate matter and sulfur oxides are emitted in different
geographic regions of the United States. Table 4-7 presents State and regional estimates of
1979 population, PM and SO emissions Abased on the 1977 NEDS inventory), emission densities,
)\, ,
and percentage contributions to total U.S. point source emissions. Based on this information,
Regions III through VI accounted for over 70 percent of the particulate matter and sulfur
oxides emitted by stationary sources in the U.S. In Region III, utility and industrial fuel
combustion contributed most of the particulate matter and sulfur oxides. The mineral products
industry also contributed heavily to particulate emissions. In Regions IV and V, utility fuel
combustion and the mineral products industry contributed most of the particulate emissions,
while utility fuel combustion contributed most of the sulfur oxide emissions. The mineral
products industry and total fuel combustion caused most of the particulate emission in Region
VI. The primary metals and petrochemical industries, along with fuel combustion, contributed
most of the sulfur oxides emissions in Region VI.
In other regions grain processing (Region VII) and mineral products (Region IX) emitted
large amounts of particulate matter. Fuel combustion (Regions II and VII) and the primary
metals industry (Region IX) contributed significant amounts of sulfur oxides.
i
Several factors affect the quantity and characteristics (size and composition) of
particulate matter emissions from stationary sources. Examples of such factors are source
type, operating conditions and practices, fuel characteristics (if the source is a fuel
combustion source), and type of emission control equipment, if any. The chemical composition
of emitted particles can determine possible reactions that occur during transport and the
final effects upon receptors (see Chapters 5 and 6). Particle size affects suspension time
and transport distance and is also an important factor in determining any possible health
effects (see Chapters 11-14).
Table 4-8 presents a summary of particle size and chemical composition data for
uncontrolled particulate emissions from stationary sources. These data demonstrate the strong
influence of control devices on the particle size distribution of emissions. Table 4-9
illustrates that, for coal-fired boilers, most control devices are more efficient at removing
larger particles. Therefore, even though the total mass of smaller particles decreases, the
percentage increases. [Refer to U.S. Environmental Protection Agency (1980b) for further
4-15
-------
TABLE 4-7. STATE-BY-STATE LISTING OF TOTAL PARTICULATE ANOULFUR OXIDE EHISSIOHS
FROM STATIONARY POINT SOURCES (1977),a
POPULATION, AND DENSITY FACTORS
Region and state
Region I
Connecticut
Maine
Massachusetts
New Hampshire
Rhode Island
Vermont
TOTAL
Percent of U.S.
Region II
New Jersey
New York
Puerto Rico
Virgin Island
TOTAL
Percent of U.S.
Region III
Delaware
Population
(1000 's)
3,099
1,091
5,774
871
935
487
12,257 *
5.5
7,327
17,748
2,712
62
27,849
12.6
582
District of Columbia 674
Mary! and
Pennsylvania
Virginia
West Virginia
TOTAL
Percent of U.S.
4,144
11,862
5,032
1,821
24,143
10.9
Total garea*"
5,009
33,215
8,257
9,304
1,214
9,609
66,608
1.8
7,836
49,576
3,435
133
60,980
1.7
2,057
67
10,577
45,333
40,817
24,181
123,032
3.4
Population
density,
(people/mi )
619
33
704
94
770
51
184
-
935
358
790
466
457
-
283
10,060
392
259
126
77
196
.
Parti cul ate matter
Total emissions
(10 aetric tons)
17.6
37.7
38.2
7.7
1.7
2.2
105.2
1.51
46.8
171.6
55.2
11.3
284.9
4.08
31.0
1.7
40.3
652.9
101.9
174,4
1,002.2
14.36
Emission
density,
(tons/mr)
3.5
1.1
4.6
0.8
1.4
0.2
-
-
6.0
3.5
16.1
84.6
-
-
15.1
25.1
3.8
14.4
2.5
7.2
-
-
State
emissions
(X of U.S.)
0.25
0.54
0.55
0.11
0.02
0.03
.
-
0.67
2.46
0.79
0.16
-
-
0.44
0.02
0.58
9.36
1.46
2.50
-
-
Sulfur oxides
Total emissions
(10 metric tons)
53.4
115.2
206.7
100.8
10.2
0.9
487.3
1.85
213.7
749.7
280.1
3.6
1,247.2
4.73
115.5
15.0
289.2
2,149.5
372.0
1,084.7
4,025.8
15.28
Emissions
density „
(tons/Mi
10.7
3.5
25.0
10.8
8.4
0.1
-
~
27.3
15.1
81.5
27.4
-
-
56.2
223.5
27.3
47.4
9.1
44.9
-
-
State
emissions
) (% of U.S.)
0.20
0.44
0.78
0.38
0.04
<0.01
-
-
0.81
2.85
1.06
0.01
-
-
0.44
0.06
1.10
8.16
1.41
4.12
-
-
-------
TABLE 4-7. (continued)
Participate matter
Region and state
Region IV
Alabama
Florida
Georgia
Kentucky
Mississippi
North Carolina
South Carolina
Tennessee
TOTAL
Percent of U.S.
Region V
Illinois
« Indiana
i- Michigan
*"* Minnesota
Ohio
Wisconsin
TOTAL
Percent of U.S.
Region VI
Arkansas
Louisana
New Mexico
Oklahoma
Texas
TOTAL
Percent of U.S.
. _ Population
Population Total area density
(1000's) (mi ) (people/nrr)
3,742
8,594
5,084
3,498
2,404
5,577
2,918
4,357
36,174
16.4
11,243
5,374
9,189
4,008
10,749
4,679
45,242
20.5
2,186
3,966
1,212
2,880
13,014
23.2S8
10.5
51,609
58,560
58,876
40,395
47,716
52,586
31,055
42,244
383,041
10.6
56,400
36,291
58,216
84,068
41,222
56,154
332,351
9.2
53,104
48,523
121,666
69,919
267,338
560,550
15.5
73
147
86
87
50
106
94
103
94
-
199
148
158
48
261
83
136
-
41
82
10
41
49
41
-
Total emissions
(10 metric tons)
248.
182.
88.
369.
135.
159.
79.
132.
1,395.
20.
448.
313.
209.
116.
529.
231.
1,848.
26.
105.
295.
77.
71.
367.
917.
13.
5
5
6
1
8
6
1
3
5
0
5
4
3
5
3
8
8
49
8
7
6
0
1
2
14
Emission
density
(tons/mi )
4.82
3.12
1.50
9.14
2.9
3.02
2.6
3.13
-
-
8.0
8.6
3.6
1.4
12.8
4.1
-
-
2.0
6.1
0.6
1.0
1.4
-
-
State
emissions
(X of U S. )
3.56
2.62
1.27
5.29
1.95
2.29
1.13
1.90
.
-
6.43
4.49
3.00
1.67
7.58
3.32
-
-
1.52
4.24
1.11
1.02
5.26
-
,
Sulfur oxides
Emissions State
Total emissions density, emissions
(10Jinetric tons) (tons/mi ) (% of U.S.)
898.
861.
597.
1,442.
211.
510.
260.
1,135.
5,917.
22.
1,462.
1,572.
1,064.
231.
2,840.
568.
7,740.
29.
103.
290.
511.
91.
1,141.
2,138.
8.
0
6
5
4
1
6
7
3
1
46
1
9
0
6
6
8
1
38
1
4
8
1
8
3
12
17.4
14.7
10.1
35.7
4.4
9.7
8.4
26.9
-
-
25.9
43.3
18.3
2.8
68.9
10.1
.
-
1.9
6.0
4.2
1.3
4.3
-
-
3.41
3.27
2.27
5.48
0.80
1.94
0.99
4.31
-
-
5.55
5.97
4 04
0.88
10.78
2.16
-
-
0.39
1.10
1.94
0.35
4.33
-
-
-------
TABLE 4-7. (continued)
Particulate matter
Region and state
Region VII
Iowa
Kansas
Missouri
Nebraska
TOTAL
Percent of U.S.
Region VIII
Colorado
Montana
North Dakota
South Dakota
Utah
Wyoming
5 TOTAL
Percent of U.S.
Region IX
Arizona
California
Hawaii
Nevada
Guam
TOTAL
Percent of U.S.
Population
(1000 's)
2,896
2,348
4,860
1,565
11,669
5.3
2,670
785
652
690
1,307
424
6,528
3.0
2,354
22,294
897
660
85
26,317
11.9
Population
Total area density Total emissions
(mi4) (people/mr) (Mr metric tons)
56,290
82,264
69,686
77,227
285,467
7.4
104,247
147,138
70,665
77,047
84,916
97,914
581,927
16.1
113,909
158,693
6,450
110,540
212
389,880
10.8
51
29
70
20
41
-
26
5
9
9
15
4
11
-
21
140
139
6
401
68
-
192.0
157.8
112.4
110.1
572.3
8.20
34.2
22.8
19.2
37.3
72.3
144.4
330.2
4.73
40.1
185.7
15.8
85.3
4.5
331.5
4.75
Emission State
density emissions
(tons/mr) (.% of U.S.)
3.4
1.9
1.6
1.4
-
-
0.3
0.2
0.3
0.5
0.9
1.5
-
-
0.4
1.2
2.5
0.8
21.3
-
-
2.75
2.26
1.61
1.58
-
-
0.49
0.33
0.28
0.53
1.04
2 07
-
-
0.58
2.66
0.23
1.22
0.06
-
-
Sulfur oxides
Emissions State
Total emissions density, emissions
(104metric tons) (tons/nr) (X of U.S.)
288.9
157.4
1,296.1
41.5
1,783.9
6,77
91.4
177.7
97.3
34.6
173.6
146.4
721.0
2.74
1,097.9
498.0
55.6
295.2
53.6
2,000.4
7.59
5.1
1.9
18.6
0.5
-
-
0.9
1.2
1.4
0.4
2.0
1.5
-
-
9.6
3.1
8.6
2.7
252.9
-
-
1.10
0.60
4.92
0.16
-
-
0.35
0.67
0.37
0.13
0.66
0.56
-
-
4.17
1.89
0.21
1,12
0.20
-
-
-------
TABLE 4-7, (continued)
Region and state
Region X
Alaska
Idaho
Oregon
Washington
TOTAL
Percent of U.S.
U.S. TOTALS
. Population
Population Total area density
(1000' s) (rnr) (people/mi^)
403
878
2,444
3,774
7,499
3.4
220,951
589,757
83,557
96,981
68,192
838,487
23.0
3,620,000
1
11
25
55
9
-
61
Particulate matter
Emission State
Total emissions density emissions
(104 metric tons) (tons/mi ) (% of U.S.)
12.7 <0.1 0.18
15.1 0.2 0.22
67.1 0.7 0.9S
96.2 1.4 1.38
191.0
2.74
6,978.7
Sulfur oxides
Emissions State
Total emissions density, emissions
(10 metric tons) (tons/mi ) (% of U.S.)
5.6
34.2
23.3
217.5
280.7
1.07
26,341.8
<0.1 0.02
04 0.13
0 2 0.09
3.2 0.83
-
-
"Source; U.S. Department of Commerce (1979).
cSource: The World Almanac and Book of Facts 1980 (1980).
Ja.
*"*
-------
TABLE 4-8. EXAMPLES OF UNCONTROLLED PARTICULATE EMISSION CHARACTERISTICS
I
f\3
o
Particle size data3
(weight % less than stated size)
Source category 15 urn 2.5 urn 1.0 urn
Fuel combustion
Utility
Coal 15-90 5-70 1-15
Oil 95 70-95 5-20
Industrial
Oil -- — 65-95
Gas — — 100
Commercial/Institional/
Residential
Oil
Chemical composition data
Major elements
and compounds
A1,Ca,Fe,Si,
sul fates .organics
Al,Ca,Fe,Mg,Na,
sul fates, organics
Al,Fe,Mg,Si,
sul fates, organics
Cl.Na, sul fates,
organics
Al,Ca,Mg,Zn,
sul fates
Trace elements
(less than 1% by weight)
As.BjBa.Be.Cd.CljCo^r,
Cu,FfHg,K,Mg,Mn,Na,Nf,
P,Pb,S,Se,Ti,V,Zn,Zr
As,Ba,Br,Co,Cr,Cu,K,
Mn,MosNi,Pb,Se,SrJi,V
As,Ba,Ca,Cd,Co,Cr,Cu,Hg,
K.Mo.Ni.Ph.Se.Sr.Ti.V.Zn
As,Ba,Cd,Cr,Cu,Hg,K,
Ni.Pb.Sb.C
Gas
100
Cl,Na,sulfates
organics
-------
TABLE 4-8. (continued)
Particle size data
Source category
(weight % less than stated size)
15 |jm 2.5 (Jin 1.0 pm
Chemical composition data
Major elements
and compounds
Trace elements
(less than 1% by weight)
Industrial processes
Metals
Iron and steel
Primary aluminum 90
Primary copper —
Primary lead —
Primary zinc —
Iron foundries 70-95
Mineral products
Cement 80
Asphalt 10-15
Lime —
Gypsum --
35-99 30-95
75 35-45
20-95 70
80
90-98
65-90
30
65
5-30
1-2 < k
25-50 5
-- 20
A1,C,Ca,Cr,Fe,K,Mg, Ag.As.Br.Cd.Cs.Cu.F.I,
Mn,Pb,Si,Zn, Mo.Ni.Rb.Se.Sn.Sr.V.Zr
sul fates,organ!cs
Al,C,Ca,F,Fe,Na
Cu.Pb.S.Zn Ag,Al,As,Cd,Hg,Sb,Se,
Si.Te
Pb,Zn As.Cd.SeJe
Cd,Fe,Pb,S,Zn Cu.Hg.Mn.Sn
Al ,0,03,01 ,K,Mg, Ag.Ba.Cd.Cr.Cu.F.Fe^n,
Na.Si.carbonates, MosNi,Pb,Rb,Se,Ti,Zn
sulfates
Al.C.Ca.Fe.K.Mg, Ag,As,Ba,CrJi
Si,sulfates
Ca,Fe,Mg,Se,Si,
carbonates
Al.C.CA.Mg.Na, As,Ba,Br,Cd,Cl,Cr,Cu,
sulfates Fe^.Mn.Mo.Ni.Pb.Se,
Sr.Y.Zn
-------
TABLE 4-8. (continued)
Source category
Crushed rock
Petroleum
Particle size data3
(weight % less than stated size)
15 urn 2.5 urn 1.0 urn
1-2
50-90
Cheiical coiposition data
Major elements
and compounds
Ca.Si.P
Asphalt, coke dust,
Trace elements
(less than 1% by weight)
'§a,Cu,Fe5K,Mn,Sr
Chemicals
sulfuric acid
Others
40-95
10-55
sulfuric acid mist,
flyash, soot
Sulfuric acid mist
-p.
1
ro
PO
Grain processing
Pulp and paper
Solid waste disposal
Incinerators
15
90-95
45
1 0
70-80
35
Organics
Ca,Mg,Na, carbonates,
sul fates
..
b
Since a number of references were cited, some characterizing different processes, discrepancies
may exist in the ranges shown.
Elements and compounds listed were included in at least one of the references cited.
Sources: Surprenant et al. (1979); Taback et al. (1979); U.S. Environmental Protection Agency
(1980c); U.S. Environmental Protection Agency (1980d); Plemons and Parnell (1981).
-------
TABLE 4-9. SIZE-SPECIFIC PARTICULATE EMISSIONS FROM
COAL-FIRED BOILERS
Control device
ESP
Wet scrubber
Fabric filter3
Inlet size distribution
(uncontrolled)
(Mass percent less than)
15 p* 2.5.|jm
15-50 5-20
30-95 10-70
55-65 20-45
Outlet size distribution Removal efficiency (%)
(controlled)
(Mass percent less than)
15 Mm 2.5 |jm 15 pi
70-95 15-70 65-99+
80-95 50-90 75-95
-------
discussion on the effects of control devices on emissions characteristics.] Therefore, the
application of the particle size percentages representing uncontrolled emissions shown in
Table 4-8 to the emission quantities from controlled sources listed in Tables 4-6 and 4-7
would probably result in an underestimation of the finer particle fractions.
As a further example of the difference between controlled and uncontrolled conditions,
control devices have helped reduce the mass flow of particulate emissions in California's
South Coast Air Basin by 95 percent or more from what prevailed under uncontrolled conditions.
However, over 90 percent of the remaining emissions (both point sources and miscellaneous area
sources) have particle sizes less than 10 pm (Taback et a!., 1979).
A final point with respect to Table 4-8 is that the particle size and chemical
composition data represent an overall source category. Therefore, in the iron and steel
industry, for example, not all of the many different processes emitting particulate matter
would necessarily have emissions exhibiting the exact- characteristics shown. Further
information can be obtained from the cited documents4<
The same factors mentioned earlier may affect the quantity and characteristics of sulfur
oxide emissions. By volume, over 90 percent of total national sulfur oxide emissions are in
the form of sulfur dioxide, SO,,. Primary sulfates account for most of the other 10 percent.
Little is known about primary-'sul fates, but combustion of coal and oil is thought to be a
major source. Primary sulfates are of increasing concern because of their potential impacts
on visibility, acidic deposition, and health.
4.4.2.1 Fuel Combustion—Stationary fuel combustion includes all boilers, heaters, and
furnaces found in utilities, industry, and commercial/institutional and residential
establishments. In the utility and industrial sectors, coal, and to a lesser degree, oil
combustion contribute most of the particulate and sulfur oxides emissions (see Table 4-6),
Oil combustion causes most of these emissions from commercial/institutional and residential
establishments.
Coal is a slow-burning fuel with a relatively high ash content. Coal combustion
particles consist primarily- of carbon, silica, alumina, and iron oxide (See Table 4-8).
Particulate sulfates and trace elements are also included. A large percentage of the trace
elements in raw coal remains in the solid waste or bottom ash, as shown in Table 4-10 (based
on 1974 emissions data). The roughly 940,000 metric tons of trace elements emitted to the
atmosphere represent about 15 percent of total particulate emissions.
Uncontrolled, the quantity and particle size distribution of coal fly ash depend on the
amount and type of coal burned, the unit type,-and the ash content of the coal. - Cyclone and
pulverized-coal furnaces, typically used in utility boilers, discharge finer particles than
stoker-fired boilers, used mainly by industry. The combustion of low-ash coal produces less
particulate matter than the combustion of high-ash coal. High-sodium lignite causes less
combustion particulate formation than does low-sodium lignite (U.S. Environmental Protection
Agency, 1977).
4-24
-------
TABLE 4-10. TRACE ELEMENT AIR EMISSIONS VS. SOLID WASTE; PERCENT FROM CONVENTIONAL
STATIONARY FUEL COMBUSTION SOURCES, AND TOTAL (METRIC TONS PER YEAR)
Air emissions (fly
Element
As
Ba
Be
B
Br
Cd
C1
Cr
Co
Cu
F
Fe
Pb
Mn
Hg
Ni
Se
Ti
U
V
Zn
Zr
Util
90
88
89
90
84
61
83
84
63
72
83
77
92
89
81
60
85
89
86
63
89
78
Indust
8
9
9
9
13
21
13
11
23
16
13
20
7
10
14
21
13
9
10
20
10
20
Com/Inst
2
3
2
1
2
18
2
5
14
12
2
3
1
1
3
19
2
2
4
17
1
2
ash)
Res Total
<1 2,990
<1 2,770
<1 240
<1 4,990
1 6,080
<1 300
2 644,100
<1 1,630
<1 460
<1 2,540
2 33,570
<1 154,200
<1 1,180
<1 4,630
2 50
<1 7,350
1 790
<1 56,250
<1 1,540
<1 9,980
<1 2,090
<1 2,090
939,820
Solid waste (bottom
Util
89
83
83
85
0
83
0
75
69
78
0
87
81
98
78
81
76
83
84
84
83
86
Indust
9
13
12
13
0
14
0
12
9
12
0
9
15
1
20
12
22
13
13
12
13
11
Com/Inst
1
2
2
1
0
1
0
5
8
4
0
2
2
1
2
3
1
2
1
1
2
1
ash)
Res
1
2
3
2
0
2
0
8
14
6
0
2
3
<1
<1
4
1
2
2
2
3
2
Total
12,250
15,970
740
16,240
0
110
0
5,040
1,920
4,280
0
1,369,900
2,530
12,520
10
4,700
370
178,700
4,510
8,450
6,890
17,240
1,662,370
Source: Surprenant et al. (1976).
-------
In the combustion of most coals (most commonly bituminous), more than 90 percent of the
coal sulfur is converted to gaseous S0?; about 1 to 2 percent of the emitted sulfur oxides are
in the form of primary sulfates (Homolya and Cheney, 1978a; Homolya and Cheney, 1979).
Lignite is used where it is plentiful at relatively low cost. The alkali content (mostly
sodium) of lignite ash has a major effect on the amount of coal sulfur retained in bottom ash.
A high-sodium lignite may retain over 50 percent of the available sulfur, and a low-sodium
lignite may retain less than 10 percent (U.S. Environmental Protection Agency, 1977).
Several factors can affect the formation of primary sulfates from coal-fired boilers.
The higher excess oxygen levels commonly used in industrial boilers increase the oxidation of
SOg to SCL and HoSO* (Homolya and Cheney, 1978a; Bennett and Knapp, 1978). Most gaseous SO-
is hydrated to gaseous or aerosol H-SQ* before exiting the boiler stack (Homolya and Cheney,
1979). Dirty equipment also may increase primary sulfate emissions from coal-fired boilers,
since boiler deposits can act as catalysts in the oxidation of SO,, to sulfates. Conversely,
the relatively low flame temperatures used in most coal-fired boilers lessen the formation of
SO, from S0«.
w £.
After coal, oil combustion in the utility and industrial sectors results in the next
largest amount of emissions. In direct contrast to coal, however, oil is a fast burning, low
ash fuel. The low ash content results in formation of less particulate matter, but the size
of particles formed by oil combustion is generally smaller than that of particles formed by
coal combustion (see Table 4-8). Also, although coal combustion contributes most of the trace
elements associated with particulate emissions, oil combustion is the source of 50 to 80
percent of cadmium, cobalt, copper, nickel, and vanadium emissions (Surprenant et al., 1976).
Oil-fired boilers generally convert over 90 percent of available fuel sulfur to gaseous
SO, emissions. However, high flame temperatures used in the combustion of oil exacerbate the
formation of primary sulfates. Tests have shown that about 7 percent by weight of sulfur
oxide emissions from oil combustion is emitted as primary sulfates (Homolya and Cheney,
1978b). Increasing excess oxygen and increasing the oil vanadium content will increase the
formation of primary sulfates in the gas (Homolya and Cheney, 1978b; Bennett and Knapp, 1978;
Oietz et al., 1978). As the emissions disperse and cool to ambient temperature, vanadium's
catalytic action becomes insignificant (see Chapter 2).
Low sulfur oil and natural gas are the fuels typically used for space heating in the
commercial/institutional and residential sectors. Total emissions are minor compared with the
utility and industrial sectors. However, most commercial/institutional and residential
sources are in areas of high population density and release emissions at OP near ground level,
thereby providing for high population exposure (Surprenant et al., 1979). Also, emissions are
concentrated primarily during the winter heating season.
Currently, homeowners, particularly those in the northern forested areas of the U.S., are
turning to wood as a heating fuel. In 1976, a total of 16 x 10 metric tons was combusted in
wood stoves and furnaces, auxiliary heating devices, and fireplaces (deAngelis et al., 1980).
4-26
-------
The emissions of participate matter from wood combustion primarily consist of organic matter
that is rich in phenols, derivatives of benzaldehyde, and furfural, and a variety of compounds
in the general category of polycyclic organic matter (POM). Measurements of the filterable
(with collection taking place at 116°C) particles emitted during residential wood combustion
typically show levels of approximately 3 g per kg of fuel. Condensable (at 0°C) organic
emissions are on the order of 8 g/kg. Of critical importance is the fact that measurements of
ROM's show values of 0.02 to 0.3 g/kg and 0,03 g/kg for stoves and fireplaces, respectively
(deAngelis et al., 1980). On the national scale, exclusive of aircraft emissions, 35 percent
of all POM's are emitted from residential wood combustion (Peters, 1981). In 1980, the CO
emissions from home wood combustion accounted for about 3 percent of the national total;
emissions of sulfur oxides are at least two orders of magnitude less than the particulate
emissions. Although emissions from residential wood burning are a small fraction of national
totals, they are injected practically at ground level and can be expected to become an
increasingly significant component in urban air pollution.
4.4.2.2 Industrial Processes—Major industrial process sources of particulate and sulfur
oxide emissions include the metals, mineral products, petroleum, and chemicals industries.
Others are grain processing and pulp and paper production (See Table 4-6).
The most significant emission sources in the metals industry are iron and steel and
primary smelting operations. The iron and steel industry involves coke, iron, and steel
production. Coking is the process of heating coal in a low-oxygen atmosphere to remove
volatile components, which are recovered. Coke is used in the production of iron. Both
particles and sulfur oxides result from the charging of coal to the hot ovens, door and
topside leaks, underfiring, pushing (removal of hot coke), and quenching. Some fine particles
consist, at least partly, of condensed organic components.
Particulate emission sources of iron production include the combustion gases, tapping
operations, and blast furnace slips (operations that require bypassing the control device).
The emitted particles are probably all fine particles that either escape the control device or
result from tapping (see Table 4-8). Blast furnace flue dust is composed primarily of iron,
siljcon dioxide, and aluminum oxide, among others.
Steel is produced in several different ways. The basic oxygen furnace produces steel
from a furnace charge composed of about 70 percent molten pig iron and 30 percent scrap. A
stream of commercially pure oxygen is used to oxidize impurities, principally carbon and
silicon. The tremendous agitation produced by the oxygen lancing produces high dust loadings
consisting mostly of iron and small amounts of fluorides. Most of the particles are less than
5 urn in size (U.S. Environmental Protection Agency, 1977).
From 1960 to 1975, steel production in open hearth furnaces declined from 90 percent of
the U.S. total to 20 percent (Desy, 1978). Open hearth furnaces are being replaced by basic
oxygen furnaces that produce 272 or more metric tons of steel per hour compared with the 27 to
54 metric tons per hour typically produced in open hearth furnaces. The composition of
4-27
-------
particulate emissions is similar to those from the basic oxygen furnace. Most of the
emissions before control are less than 5 urn, and probably 90 percent are fine particles after
control.
Two types of electric furnaces, the arc furnace and the induction furnace, are used to
produce steel. The arc furnace is used to produce high-alloy steel, as well as a considerable
amount of mild steel. The emissions, most of which are fine particles, consist primarily of
oxides of iron, manganese, aluminum, calcium, magnesium, and silicon. Particulate fluorides
are also emitted. The induction furnace produces primarily specialty and high-alloy steels
and has no major emission problems.
The primary metals industry includes the smelting of copper, lead, and zinc, along with
aluminum production. Sulfur in unprocessed copper, lead, and zinc ores is converted to S0~ in
the smelting processes (U.S. Environmental Protection Agency, 1977). A relatively small
portion of the sulfur is emitted as particulate sulfate and sulfuric acid. The bulk of SO- is
formed in the roasting, smelting, sintering, and converting processes (U.S. Environmental
Protection Agency, 1974). Particulate matter emitted from the same processes is mostly fine
particles, less than 2.5 urn in diameter.
Aluminum production involves mainly bauxite grinding, calcining, and reduction.
Particulate emissions are primarily alumina with about 25 percent particulate fluoride (U.S.
Environmental Protection Agency, 1977). Before control, 35 to 44 percent of the particles
were below 1 urn in diameter.
Emissions from the mineral products industry result primarily from the production of
Portland cement, asphalt, and crushed rock and to a lesser extent, lime, glass, gypsum, brick,
fiberglass, cleaned coal, phosphate rock, and potash. Emission points such as crushing,
screening, conveying, grinding, drying or calcining, and loading are common to most mineral
products industries. Fugitive dust from most of these processes tends to be larger than 15
urn, although drying and calcining produce relatively finer particles. The composition of
particulate emissions is similar to the mineral being processed.
The more than 30 raw materials used to make cement can be grouped into four basic
categories: lime (calcareous), silica (siliceous), alumina (argillaceous), and iron
(ferriferous). The kiln and associated clinker cooler are potentially the largest sources of
particulate and sulfur oxides emissions (U.S. Environmental Protection Agency, 1977). Kiln
emissions also include primary sulfates (Dellinger et a!., 1980). Probable particle size
distribution and chemical composition are shown in Table 4-8.
Asphalt concrete is a mixture of aggregate, asphalt cement, and occasionally mineral
filler. Commonly, asphalt concrete is produced in conjunction with crushed and broken stone
production facilities. The rotary dryer typically used to dry and heat the aggregate is
potentially the largest particulate emission source (U.S. Environmental Protection Agency,
1977).
4-28
-------
Emissions from the production of crushed rock result primarily from the processes
mentioned earlier. The chemical composition of particulate emissions is similar to the
material processed. Usually, the particles emitted are relatively coarse.
The major sources of sulfur oxide emissions in the petroleum industry are the catalytic
cracking and sulfur recovery processes and off-gas flares (U.S. Environmental Protection
Agency, 1977); Dickerman et al., 1977). Sulfur dioxide is emitted during the catalyst
regeneration step of the catalytic cracking process.
Major sour gas streams are usually treated in a sulfur plant. Most sulfur plants utilize
a modified Claus process that consists of multistage oxidation of hydrogen sulfide to
elemental sulfur. The sulfur recovery efficiency of these sulfur plants ranges from 92 to 97
percent depending on the number of catalytic stages. Sulfur plant tail gas is usually
incinerated so that most of the remaining sulfur species are oxidized to SOp. Some plants
have installed tail gas cleanup systems to reduce SOp emissions further. These units along
with a sulfur plant can achieve up to 99.8 percent sulfur recovery.
Minor off-gas streams and recovered vapors are often combusted in flares. Most of the
sulfur species present in these vapors are oxidized to S0?.
A variety of processes are used by the chemical production industry. Chemical process
industries that contribute significant amounts of sulfur oxide emissions are sulfuric acid
plants, elemental sulfur plants, and explosives manufacturing.
Sulfuric acid is manufactured primarily by the contact process. The three types of raw
materials charged to sulfuric acid plants are elemental sulfur, spent acid and hydrogen
sulfide, and sulfide ores and smelter gases. The amount of S0? emissions in acid plant exit
gases is an inverse function of the sulfur conversion efficiency of the process (U.S.
Environmental Protection Agency, 1977). Sulfuric acid mist is generated by the process S0?
absorbers. The quantity and size distribution of the acid mist are dependent on the type of
sulfur feedstock used, the strength of the acid produced, and the conditions in the absorber.
The manufacture of TNT and nitrocellulose explosives produces emissions of SOp and
sulfuric acid mist. Sulfuric acid is a major raw material in the production of these
explosives. Sulfuric acid concentrators, exhaust from the preparation of sodium sulfite/
sodium hydrogen sulfite (Sellite), and incinerators are the major sulfur oxides sources in
these processes. Sulfur oxide emissions may vary considerably depending on the efficiency of
the process and the operating conditions (U.S. Environmental Protection Agency, 1977).
Particulate emissions from grain processing typically result from handling, cleaning,
drying, and milling (U.S. Environmental Protection Agency, 1977). Grain processing particles
are normally coarse and composed of the parent organic material.
Chemical wood pulping by the kraft or sulfite processes involves cooking wood chips under
pressure to dissolve the lignin that binds the cellulose fibers, in addition to washing,
milling, bleaching, and drying (U.S. Environmental Protection Agency, 1977). Particulate
emissions occur primarily from the recovery furnace (used to recover cooking chemicals)\and
4-29
-------
the lime kiln (lime is used in cooking). Sulfur dioxide emissions result mainly from
oxidation of reduced sulfur compounds in the recovery furnace.,
4.4.3 Industrial Process Fugitive Particulate Emissions
Fugitive dust emissions result from wind erosion of storage piles and unpaved plant roads
and from vehicular traffic over plant roads. Fugitive process emissions result from
industry-related operations such as materials handling, loading, unloading, and transfer
operations. Point sources that are incompletely controlled, such as furnace charging and
tapping, and equipment that is maintained poorly, such as leaking furnaces and coke oven
doors, are also fugitive process emission sources.
Process fugitive emissions are not emitted from a definable point such as a stack. They
are difficult to collect, measure, and control. A given industry generally has a large number
of fugitive particulate emission sources. For example, 20 separate sources have been
identified for foundries (Jutze et al., 1977). In terms of total emissions, however, one or
two of these sources may predominate.
Even though fugitive particulate emission totals may appear small when compared with
totals from large point sources, they may take on importance because of the concentration of
control efforts on point source emissions. In the integrated iron and steel industry, where
fugitive particulate emissions are characterized relatively well, fugitive emissions are
estimated to account for about 10 percent of all uncontrolled emissions. However, since
fugitive particulate emissions are poorly controlled, they account for more than 60 percent of
total controlled emissions (Spawn, 1979). Also, in situations where point sources are we'll
controlled or use high stacks, fugitive particulate emissions exert a major effect on local
air quality. Extremely high suspended particulate matter levels have been measured in areas
where process fugitive emissions are predominant (Lynn et al., 1976; Lebowitz, 1975).
Table 4-11 presents estimates of uncontrolled industrial process fugitive particulate
emissions. Particle size and composition characteristics are also presented. Unfortunately,
many of the emission factors used to estimate process fugitive emissions are based on
engineering judgment or extrapolation from similar processes. Often, few test data are
available to support these estimates since process fugitive emissions are difficult to
measure. Therefore, the accuracy of these estimates is questionable. Also, some of the
emissions presented in Table 4-11 may have already been accounted for in Table 4-6 (Section
4.4.2). This overlap results from the use of different references (see Section 4.3).
As is evident from Table 4-11, three broad categories account for nearly all of the
potential process fugitive emissions in the United States. They are mineral products, food
and agriculture, and primary metals. In the mineral products industries, fugitive particulate
emissions tend to reflect the composition of the parent materials. The limited amount of
particle size data indicates that most particles are relatively coarse.
Grain elevator operations account for the fugitive particulate emissions in the food and
agriculture industry. These emissions consist almost entirely of grain dust from loading and
4-30
-------
TABLE 4-11. UNCONTROLLED INDUSTRIAL PROCESS
FUGITIVE PARTICULATE EMISSIONS3
Source category
Annual uncontrolled
fugitive particulate
,, emissions
Size
(10 metric tons) characteristics
Major components
Mineral products
Crushed rock
Extraction,
Surface coal mining
Portland cement
Asphalt concrete
Lime manufacturing
Concrete batching
Food and agriculture
Grain elevators
Primary metals
Coke/i ron/steel
Foundries
730
700
100
50
30
1,250
250
125
10-50% < 10 pm
1-2% < 1 pm
10-151 < 10 pm
50-70% < 4 pm
45-70% < 5 pm
10-20% < 5 pm
40% < 10 pm
Coke mfg: 27-80%
< 10 pm, 15-26%
< 2 pm; iron mfg:
1-10% < 5 pm;
Steel mfg: 50%
< 5 pm
50% < 15 pm
Same as parent
material (important
for toxic minerals
such as asbestos,
beryllium, silica)
Limestone, clay,
shale, gypsum,
iron-bearing and
siliceous materials
Sand, crushed stone,
1imestone, hydrated
lime
Limestone, lime
Cement dust
Grain dust
Polycyclic organic
matter, coal dust,
coke dust, iron
oxide dust, kish
(graphite material),
metal fume (pri-
marily iron oxide),
plus trace amounts of
As, Be, Pb, Cr, Cd,
Se, Co, Ni, and
fluorides
Metal oxide fume
(primarily oxides of
silicon and iron),
fine carbonaceous
fume, plus trace
amounts of Pb, Cr,
Ag, Co, and Ni
4-31
-------
TABLE 4-11. (continued)
Source category
Annual uncontrolled
Fugitive particulate
2 emissions Size
(10 metric tons) characteristics
Major components
Aluminum
60
Copper
40
Lead
15
Secondary metals
10
Wood products
TOTAL
50-90% < 10 urn
10-50% < 5
10-90% < 5 |jm
80-100% < 5
40-90% < 10
Particulate fluo-
rides, alumina
(AlpO,), carbon
dust, condensed
hydrocarbons, tars
Cu, Fe, S, SiO,
from ore concen-
trate; metal fume
consisting of
oxides of As, Pb,
In, Cu, Cd, plus
trace amounts of
Se and Ag
Metal fume consist-
ing of oxides of
Pb, Cd, In, Sb, plus
trace amounts of As,
Se, and Ag
Oxides of Al,Cu,Pb,
Sn,Zn; oxides of
alkali metals;
A1CU, NH.Cl.NaCl,
ZnClX; fluorides,
and carbonaceous
materials, plus
trace amounts of
Cr, Cd, and Ni
Sawdust
Note: Emissions are based on data for the mid-1970's and may differ from
current levels. « -
aSources: Taback et al. (1979)
Zoller et al. (1978)
Jutze et al. (1977)
Norman et al. (1977)
4-32
-------
unloading, drying and cleaning operations, conveyor belts, and transfer points. Only about 29
percent of these particles are less than 10 pm in size.
Primary metals production encompasses six separate industries. Fugitive particulate
emissions in this category result from the handling and transporting of raw materials and from
the smelting and refining of these raw materials into their finished metal products. Although
emissions of the first type are not well characterized, emissions of the second type often
consist of fine metal fumes. This finding is particularly significant because of the
quantities of toxic trace metals that can be concentrated and volatilized in metal melting
operations. Some of these trace toxic components of particulate emissions are identified in
Table 4-11.
The remaining two categories, secondary metals and wood products, account for less than 1
percent of the national total of industrial process fugitive particulate emissions. Fugitive
metal fume particles from secondary metal melting operations also include toxic components
and are listed in Table 4-11.
4.4.4 Nom'ndustrial Fugitive Particulate Emissions
Nom'ndustrial fugitive particulate emissions, or fugitive dusts are caused by traffic
entrainment of dust from public paved and unpaved roads, agricultural operations, construction
activities, surface mining operations, and fires. With the exception of fires, all of these
sources may be classified as open-dust sources; that is, they involve dust entrainment by the
interaction of machinery with aggregate materials and by the forces of wind on exposed
materials.
A number of factors can affect emissions from open sources but they can generally be
classified under three headings: material, equipment, and climate. Material factors
encompass such influences as silt and moisture content. For example, increasing the silt
content and decreasing the moisture content of unpaved road material would probably result in
more dust being generated. Equipment factors generally refer to vehicle weight and speed.
For example, increasing the speed or weight of a vehicle travelling over an unpaved road would
tend to increase emissions. Climatic factors are windspeed and precipitation. Increased
windspeed and decreased precipitation would both tend to increase emissions from any open-dust
source.
Estimated U.S. annual particulate emissions from nonindustrial fugitive dust sources are
difficult to estimate accurately. As shown in Table 4-12, fugitive dust emissions from
unpaved roads tend to be quite significant. The two available estimates, however, vary by
almost an order of magnitude. Fugitive dust from wind erosion of cropland and construction
activities as documented by Cooper et al. (1979) also appears significant. However, no other
estimates are available for comparison purposes. Estimated total fugitive emissions range
f~ /-
from approximately 112 x 10 metric tons per year to 369 x 10 metric tons per year. The
lower figure assumes the U.S. Environmental Protection Agency (1980b) estimate of fugitive
emissions from paved and unpaved roads; the higher figure assumes the estimate of Cooper et
4-33
-------
TABLE 4-12. ESTIMATED ANNUAL PARTICULATE EMISSIONS FROM
NONINDUSTRIAL FUGITIVE SOURCES
Source category
Estimated Emissions
c _,
(10 metric tons/year)
Cooper et al. (1979)
U.S. Environmental
Protection Agency
(1980b)
Unpaved roads
Paved roads
Wind erosion of cropland
Agricultural tilling
Construction activities
Minerals extraction
Mineral tailing
Prescribed fires
290
7.2
40
2.9
25
3
0.7
0.4
35
4.7
-
-
-
-
-
0.2
Particles less than 30 pm in diameter.
^Includes prescribed forest burns and agricultural burning.
TABLE 4-13. ESTIMATED PARTICLE SIZE DISTRIBUTIONS FOR SEVERAL
NONINDUSTRIAL FUGITIVE SOURCE CATEGORIES IN CALIFORNIA'S
SOUTH COAST AIR BASIN
Source category
Unpaved road dust
Agricultural tillage dust
Road building and construction dust
Agricultural burning
Weight
>10 pin
54
40
36
<1
jpercent
3-10 (jm
16
21
24
2
in size
1-3 urn
12
17
16
8
range
<1 pm
18
22
24
90
Source: Taback et al. (1979).
4-34
-------
al. (1979). Because of the disparity between comparable estimates, the estimated emissions
presented in Table 4-12 should be considered with a degree of caution.
Information on particle size distribution is also limited. Some limited data are
presented in Table 4-13, but they are representative only of California's South Coast Air
Basin and should not be extrapolated to the Nat-ion as a whole. Dust from unpaved roads,
agricultural tilling, construction, and road building is composed primarily of silicon,
phosphorus, aluminum, iron, calcium, and potassium. Trace elements include barium, cobalt,
copper, lead, manganese, nickel, titanium, vanadium, and zinc (Taback et al., 1979).
Finally, it is estimated that fugitive dust emissions exceed particulate emissions from
stationary point sources in 90 percent of the Air Quality Control Regions that are not meeting
the ambient standards for total suspended particulates (Carpenter and Weant, 1978). However,
the impact of fugitive dust emissions on populated areas may be somewhat lessened because a
major portion of these emissions consists of large particles that settle to the ground a short
distance from the source and because many fugitive dust sources, like unpaved roads, exist
mainly in rural areas (U.S. Environmental Protection Agency, 1980b).
4.4.5 Transportation Source Emissions
Transportation source emissions may be divided into two categories: engine-related
emissions from vehicle exhaust, and other highway vehicle-relate'd particles from tire wear and
clutch and brake lining wear (Bradow et al., 1979; U.S. Environmental Protection Agency,
1978b; Dannis, 1974; and Jacko and DuCharme, 1973). Total transportation source emissions for
1978, including emissions from highway vehicles, aircraft, railroads, and vessels, were
estimated at 1.3 x 10 metric tons (particulate matter) and 0.8 x 10 metric tons (sulfur
oxides) (U.S. Environmental Protection Agency, 1980a). About "75 percent of the particulate
emissions and 50 percent of the sulfur oxide emissions in 1978 were from highway vehicles.
Engine-related particulate emissions from transportation sources are composed primarily
of lead halides, sulfates, and carbonaceous matter (including absorbed organics). Highway
vehicle-related particles are emitted at the rate of about 0.01 to 0.30 grams per mile by
gasoline engines and 0.5 to 3.0 grams per mile for diesel engines (Bradow et al., 1979). The
major components are lead (except for vehicles using diesel fuel or unleaded gasoline),
carbon, organics, and sulfates. Vehicles burning leaded gasoline also emit inorganic
compounds of lead (mostly PbBrCl), bromine, and chlorine (Springer, 1978). Particulate matter
from catalyst-equipped vehicles using unleaded gasoline is dominated by sulfate and
carbonaceous material.
Diesel exhaust particles can include traces of iron, copper, calcium, lead, and zinc,
along'with carbon,"organics, and sulfates (Lee and Duffield, 1979). The particle emission
rate and composition for diesel engines are sensitive to many factors, including vehicle size,
operating conditions (speed, load), and fuel characteristics. Normally, carbon-containing
species dominate, including a material similar to lubricating oil (Black and High, 1978).
4-35
-------
Engine-related particles are mostly under 1 pm in diameter. For vehicles burning leaded
gasoline, the available data indicate a mass median diameter of about 0.25 pn (Moran et a!.,
1971). Due to the predominance of sulfates, catalyst-equipped vehicles burning unleaded
gasoline emit smaller particles having a mass mean diameter of about 0.05 jjm (Groblicki,
1976). The size distribution of diesel particulate matter suggests a mass median diameter of
about 0.2 urn (Dolan and Kittelson, 1979).
Very few data exist on nonengine particulate emissions from highway vehicles. About 40
percent of particles from tire wear are less than 10 |jm (about 20 percent are less than 1 |jm);
they are composed primarily of carbon (Taback et a!., 1979). Particles from brake lining
attrition are all less than 1 urn and are composed mainly of asbestos (80 percent) and carbon
(Taback et a!., 1979).
4.5 SUMMARY
Participate matter and sulfur oxides are emitted into the atmosphere from a number of
sources, both natural and manmade. Natural source emissions include terrestrial dust, sea
spray, biogenic emanations, volcanic emissions, and emissions from wildfires. The predominant
manmade sources are stationary point sources, industrial and nonindustrial fugitive sources,
and transportation sources. Annual U.S. emissions from natural sources are estimated at 84 x
10 metric tons of particulate matter and 5 x 10 metric tons of sulfur (the equivalent of 10
x 10 metric tons of sulfur dioxide). Manmade sources emit roughly 125 to 385 x 10 metric
tons of particulate matter per year and 27 x 10 metric tons of sulfur oxides per year in the
U.S. Because of the assumptions and approximations inherent in emissions calculations, the
numbers quoted above should not be considered more than estimates. Section 4.2 further
discusses the problem involved with emissions estimates.
The characteristics of particulate matter emissions vary according to source type and a
number of other factors. Particulate emissions from natural sources tend to be rather coarse.
(For the purposes of this chapter, coarse refers to particles with a diameter greater than 2.5
urn.) Particulate matter generated by nonindustrial fugitive sources (e.g., unpaved roads and
wind erosion of cropland) is quite significant on a mass basis. However, only about 50 and 20
percent is less than 10 and 1 urn, respectively. Most of the particulate matter emitted by
stationary sources and transportation sources, on the other hand, is relatively fine, or less
than 2.5 urn in diameter. Adding control devices further concentrates particulate emissions in
the finer ranges because most control devices are more efficient at removing larger particles.
Therefore, the estimated 10.5 x 10 metric tons of particulate matter generated in 1978 by
stationary point sources probably consists largely of finer particles, since that estimate was
arrived at assuming the application of control devices. In addition, the finer particles
emitted by stationary point sources tend to include a greater variety of toxic substances than
do emissions from natural or manmade fugitive sources.
Virtually all of the manmade sulfur oxide emissions result from stationary point sources.
Over 90 percent of these manmade sulfur oxide emissions are in the form of sulfur dioxide.
4-36
-------
The balance consists of sulfates in various forms. Most natural sulfur is emitted as reduced
sulfur compounds, but these compounds are probably oxidized in the atmosphere to sulfur
dioxide and sulfates.
4-37
-------
4.6 REFERENCES
Adams, D. F. Personal communication [Telephone conversation with E. Robinson], October 1980.
Adams, D. F., S. 0. Farwell, M. R. Pack, and W. L. Bamesberger. Preliminary measurements of
biogenic sulfur-containing gas emissions from soils. J. Air Pollut. Control Assoc.
29:380-383, 1979a.
Adams, D. F., S. 0. Farwell, M. R. Pack, and E. Robinson. Estimates of natural sulfur source
strengths. Presented at the Gatlinburg Conference on Sulfur, Gatlinburg, TN, October
1979b.
Adams, D. F., S. 0. Farwell, E. Robinson, M. R. Pack, and W. L. Bamesberger. Biogenic Sulfur
Strengths. Environmental Science and Technology 15:1493-1498, 1981.
Adams, F., R, Dams, L. Guzman, and J. W. Winchester, Background aerosol composition on
Chacaltaya Mountain, Bolivia. Atmos Environ, 11:629-634, 1977.
Barringer, A. R. AIRTRACE—an airborne geochemical exploration technique. |n: Proceedings
of the First Annual William T. Pecora Memorial Symposium, American Mining Congress, Sioux
Falls, South Dakota, October 1975. Geological Survey Professional Paper 1015, U.S.
Department of the Interior, Washington, D.C., September 1977. pp. 231-251.
Baylor, E. R., M. B. Baylor, D. C. Blanchard, L. D. Syzdek, and C. Appel. Virus transfer from
surf to wind. Science (Washington, D.C.) 198:575-580, 1977.
Beauford, W., J. Barber, and A. R. Barringer. Heavy metal release from plants into the
atmosphere. Nature (London) 256:35-37, 1975.
Beauford, W., J. Barber, and A. R. Barringer. Release of particles containing metals from
vegetation into the atmosphere. Science (Washington, D.C.) 195:571-573, 1977.
Bennett, L., and K. T. Knapp. Sulfur and trace metal particulate emissions from combustion
sources. In: Workshop Proceedings on Primary Sulfate Emissions from Combustion Sources,
Vol. 2: Characterization,' U.S. Environmental Protection Agency, Southern Pines, North
Carolina, April 24-26, 1978. EPA-600/9-78-020b, U.S. Environmental Protection Agency,
Research Triangle Park, NC, August 1978. pp. 165-183.
Berg, W. W. Jr., and J. W. Winchester. Aerosol chemistry of the marine atmosphere. In:
Chemical Oceanography, Vol. 7. 2nd Edition. J. P. Riley and R. Chester, eds., Academic
Press Inc., London, England, 1978. pp. 173-231.
Black, F., and L. High. Diesel hydrocarbon emissions, particulate and gas phase. Presented
at Symposium on Diesel Particulate Emissions Measurements and Characterization, Ann
Arbor, HI, May 1978.
Blanchard, D. C., and B. C. Parker. The freshwater to air transfer of microorganisms and
organic matter. In: The Aquatic Microbial Community. J. Cairns, Jr., ed., Garland
Publishing, New York, NY, 1977. pp. 625-658.
Blanchard, D. C., and A. H. Woodcock. Bubble formation and modification in the sea and its
meteorological significance. Tellus 9:145-158, 1957.
Bradow, R. L,, F. M. Black, J. N. Braddock, C. T. Hare, M. Ingalls, and J. M. Kawecki. Study
of Particulate Emissions from Motor Vehicles. A Report to Congress. Draft Report. U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1979.
4-38
-------
Carpenter, B. H. , and G. E. Weant, III. Parttculate Control for Fugitive DUst.
EPA-600/7-78-071, U.S. Environmental Protection Agency, Research Triangle Park, NC, April
1978.
Cooper, D. W. , J. S. Sullivan, M. Quinn, R, C. Antonelli, and H. Schneider. Setting
Priorities for Control of Fugitive Particulate Emissions from Open Sources.
EPA-600/7-79-186, U.S. Environmental Protection Agency, Washington, D.C., August 1979.
Cooper, J. A. Environmental Impact of Residential Wood Combustion Emissions and Its
Implications. J. Air Pollut. Control Assoc. 30:855-861, 1980.
Curtin, G. C., H. D. King, and E. L. Moiser. Movement of elements into the atmosphere from
coniferous trees in subalpine forests of Colorado and Idaho. J. Geochem. Explor.
3:245-263, 1974.
deAngelis, D. G., D. S. Ruffin, J. A. Peters and R. B. Reznik. Source assessment; residential
combustion of wood. EPA-600/2-80-Q42b. U.S. Environmental Protection Agency, IERL,
Research Triangle Park, N.C. 1980.
Dannis, M. L. Rubber dust from the normal wear of tires. Rubber Chem. Technol. 47:1011-1037,
1974. ~~
Delany, A. C. , A. C. Delany, D. W. Parkin, J. J. Griffin, E. D. Goldberg, and B. ,E. F.
Reimann. Airborne dust collected at Barbados. Geochim. Cosmochim. Acta 31:885-909,
1967.
Dellinger, B. , G. Grotecloss, C. R. Fortune, J. L. Cheney, and J. B. Homolya. Sulfur dioxide
oxidation and plume formation at cement kilns. Environ. Sci. and Techno!., 14:1244-1249,
1980. ~~
Desy, D. H. Iron and Steel. Bureau of Mines Mineral Commodity Profiles No. MCP-15, U.S.
Department of the Interior, Washington, D.C., July 1978.
X
Dickerman, J. D., T. D. Raye, J. D. Colley, and R. H. Parsons. Industrial process profiles
for environmental use. In: Petroleum Refining Industry. EPA-600/2-77-023c, U.S.
Environmental Protection Agency, Research Triangle Park, NC, January 1977.
Dietz, R. N., R. F. Wieser, and L. Newman. Operating parameters affecting sulfate emissions
from an oil-fired power unit. In: Workshop Proceedings on Primary Sulfate Emissions
from Combustion Sources, Vol. 2: Characterization, U. S. Environmental Protection
Agency, Southern Pines, North Carolina, April 24-26, 1978. EPA-600/9-78-02Qb. U.S.
Environmental Protection Agency, Research Triangle Park, NC, August 1978. pp. 239-270.
Dolan, D. F. , and D. B. Kittelson. Roadway Measurements of Diesel Exhaust Aerosols. SAE
technical paper no. 790492, Detroit, MI, March 1979.
Duce, R. A. , and E. J. Hoffman. Chemical fractionation at the air/sea interface. Annu. Rev.
Earth Planet, Sci. 4:187-228, 1976.
Eriksson, E. The yearly circulation of chloride and sulfur in nature; meteorological,
geochemical and pedological implications. Part I, Tellus, 11:375-403, 1959; Part II,
Tell us, 12:63-109, 1960.
Fegley, R. W., H. T. Ellis, and J. L. Heffter. Volcanic contributions to the stratospheric
sulfate layer. J. Appl. Heteorol. 19:683-690, 1980.
4-39
-------
Fish, B. R. Electrical generation of natural aerosols from vegetation. Science (Washington,
D.C.) 175:1239-1240, 1972.
Fruchter, J. S., D. E. Robertson, J. C. Evans, K. B. Olsen, E. A. Lepel, J. C. Laul, K. H.
Abel, R. W. Sanders, P. 0. Jackson, N. S. Wogman, R. W. Perkins, H. H. Van Tuyl, R, H.
Beauchamp, J. W. Shade, J. L. Daniel, R. L. Erikson, G. A. Sehmel, R. N. Lee, A. V.
Robinson, 0. R. Moss, J. K. Briant, and W. C. Cannon. Mount St. Helens ash from the 18
May 1980 eruption: chemical, physical, mineralogical, and biological properties.
Science (Washington, D.C.) 209:1116-1125, 1980.
Galloway, J. N., and D. M. Whelpdale. An atmospheric sulfur budget for eastern North America.
Atmos. Environ. 14:409-417, 1980.
GEOMET, Incorporated. Impact of Forestry Burning Upon Air Quality. A State-of-the-Know!edge
Characterization in Washington and Oregon. EPA-910/9-78-052, U.S. Environmental
Protection Agency, Seattle, WA, October 1978.
Gillette, D. A. On the production of soil wind erosion aerosols having the potential for long
range transport. J. Rech, Atmos. 8:735-744, 1974.
Goldberg, E. D. Rock volatility and aerosol composition. Nature (London) 260:128-129, 1976.
Granat, L., H. Rodhe, and R. 0. Hallberg. The global sulfur cycle. Ecol. Bull. (Stockholm)
22:89-134, 1976.
Groblicki, P. J. General Motors Sulfate Dispersion Experiment: Aerosol Sizing Measurements.
GMR-2127, General Motors Corporation, Warren, MI, April 1976.
Henry, R. C., and G. M. Hidy. Potential for atmospheric sulfur from microbiological sulfate
reduction. Atmos. Environ. 14:1095-1103, 1980.
Hitchcock, D. R. Atmospheric sulfates from biological sources. J. Air Pollut. Control Assoc.
26(3): 210-215, 1976.
Hitchcock, D. R., L. L. Spiller, and W. E. Wilson, Sulfuric acid aerosols and HC1 release in
coastal atmospheres: Evidence of rapid formation of sulfuric acid particulates. Atmos.
Environ. 14:165-182, 1980.
Hobbs, P. V., L. F. Radke, M. W. Eltgroth, and D. A. Hegg. Airborne studies of the emissions
from the volcanic eruptions of Mt. St. Helens. Science 211:816-818, 1981.
\
Homolya, J. B., and J. L. Cheney. An assessment of sulfuric acid and sulfate emissions from
the combustion of fossil fuels. I_n: Workshop Proceedings- on Primary Sulfate Emissions
from Combustion Sources, Vol. 2: Characterization. U.S. Environmental Protection
Agency, Southern Pines, North Carolina, April 24-26, 1978. EPA-600/9-78-020b, U. S.
Environmental Protection Agency, Research Triangle Park, NC, August 1978a. pp.3-11.
Homolya, J. B., and J. L. Cheney. An Assessment of Sulfuric Acid and Sulfate Emissions from
the Combustion of Fossil Fuels. In: Workshop Proceedings on Primary Sulfate Emissions
from Combustion Sources, Vol. 2, Characterization. EPA-600/9-78-0206. U.S.
Environmental Protection Agency, Research Triangle Park, NC, 1978b.
Homolya, J. B., and J. L. Cheney. A study of primary sulfate emissions from a coal-fired
boiler with FGD. J. Air. Pollut. Control Assoc. 29:1000-1004, 1979.
Jacko, M, G., and R. T. DuCharme. Brake Emissions—Emission Measurements from Brake and
Clutch Linings from Selected Mobile Sources. APTD-1557, U.S. Environmental Protection
Agency, Ann Arbor, MI. March 1973.
4-40
-------
Junge, C. E. Chemical analysis of aerosol particles and of gas traces on the Island of
Hawaii. Tell us 9:528-537, 1957.
Jutze, G. A., J. M. Zoller, T. A. Jansen, R. S. Amide, C. E. Zimmer, and R. W. Gerstle.
Technical Guidance for Control of Industrial Process Fugitive Particulate Emissions.
EPA-450/3-77-01Q. U.S. Environmental Protection Agency, Research Triangle Park, NC,
March 1977.
Lawson, D. R., and J. W. Winchester. A standard crystal aerosol as a reference for elemental
enrichment factors. Atmos. Environ. 13:925-930, 1979a.
Lawson, D. R. , and J. W. Winchester. Sulfur, potassium, and phosphorus associations in
aerosols from South American tropical rain forests. J. Geophys. Res. 84:3723-3727,
1979b.
Lee, R. E. , Jr., and F, V. Duffield. Sources of environmentally important metals in the
atmosphere. In: Ultratrace Metal Analysis in Biological Sciences and Environment, A
Symposium, 174th Meeting America! Chemical Society, Chicago, Illinois, August 29-30,
1977. T. H. Risby, ed. , Advances in Chemistry Series 172, American Chemical Society,
Washington, D.C., 1979. pp. 146-171.
Lebowitz, M. F. Short-term testing for fugitive dust effect. Presented at the 68th Annual
Meeting, Air Pollution Control Association, Boston, MA, June 15-20,1975. Paper No.
75-25.4.
Lovelock, J. E. CSp and the natural sulfur cycle. Nature (London) 248:625-626, 1974.
Lovelock, J. E., R. J. Maggs, and R. A. Rasmussen. Atmospheric dimethyl sulphide and the
natural sulphur cycle. Nature (London) 237:452-453, 1972.
Lynn, D. A., G. L. Deane, R. C. Galkiewicz, and R. M. Bradway. National Assessment of the
Urban Particulate Problem. Vol. 1: Summary of National Assessment. EPA-450/3-76-024,
JJ. S. Environmental Protection Agency, Research Triangle Park, NC, June 1976.
Maclntyre, F. Chemical fractionation and sea-surface microlayer processes. In: The Seas:
Ideas and Observations on Progress in the Study of the Seas, Vol. 5: Marine Chemistry.
John Wiley and Sons, Inc., New York, NY, 1974. pp.245-299.
Maroulis, P. J., and A. R. Bandy. Estimate of the contribution of biologically produced
dimethyl sulfide to the global sulfur cycle. Science (Washington, D.C.) 196:647-648,
1977.
McClenny, W. A. , R. W. Shaw, R. E. Baumgardner, R. Paur, A. Coleman, R. S. Braman, and J. M.
Ammons. Evaluation of Techniques for Measuring Biogenic Airborne Sulfur Compounds:
Cedar Island Field Study 1977. EPA-600/2-79-004, U.S. Environmental Protection Agency,
Research Triangle Park, NC, January 1979.
McElroy, M. B. , S. C. Wofsy, and N. D. Sze. Photochemical sources for atmospheric H?S.
Atmos. Environ. 14:159-163, 1980.
Miller, M. S., S. K. Friedlander, and G. M. Hidy. A chemical element balance for the Pasadena
aerosol. J. Colloid Interface Sci. 39:165-176, 1972.
Moorby, J. , and H. M. Squire. The loss of radioactive isotopes from the leaves of plants in
dry conditions. Radiat. Bot. 3:263-167, 1963.
Moran, J. B., 0. J. Manary, R. H. Fay, and M, J. Baldwin. Development of Particulate Emission
Control Techniques for Spark-ignition Engines. APTD-0949, U.S. Environmental Protection
Agency, Ann Arbor, MI, July 1971.
4-41
-------
Meyers, J. L., and R. A. Duce. Gaseous and participate bromine in the marine atmosphere. J.
Geophys. Res. 77:5330-5338, 1972a.
Hoyers, J. L., and R. A. Duce. Gaseous and particulate iodine in the marine atmosphere. J.
Geophys. Res. 77:5229-5238, 1972b.
National Research Council, Subcommittee on Airborne Particles. Airborne particles.
University Park Press, Baltimore, HD, 1979,
Nemeryuk, G. E. Migration of salts into the atmosphere during transpiration. Sov. Plant,
Physio!. 17:560-566, 1970.
Norman, B. M., C. B. Parnell, and R. V. Avant. Characterization of particulate emissions from
'grain sorghum storage and handling installations. Paper 77-3516. American Society of
Agricultural Engineers, St. Joseph, MI 1977.
Peters, J., A. POM emissions from residential wood burning; an environmental assessment.
Presented at International Conference on Residential Solid Fuels, Portland, OR, June 1-4,
1981. Sponsored by Oregon Graduate Center, Beaverton, OR., 1981.
Plemons, 0. S. and C. B. Parnell. Comparison of physical and chemical characteristics of
grain dust from five different grain species. Paper SWR181-4Q4 American Society of
Agricultural Engineers, St. Joseph, MI. 1981.
Radke, L. F., J. L. Stith, D. A. Hegg, and P. V. Hobbs. Airborne studies of particles and
gases from forest fires. J. Air Pollut. Control Assoc. 28:30-34, 1978.
Rahn, K. A. Sources of Trace Elements in Aerosols—An Approach to Clean Air. Ph.D. Thesis,
University of Michigan, Ann Arbor, MI, 1974.
Rahn, K. A. The Chemical Composition of the Marine Aerosol. Technical Report, University of
Rhode Island, Kingston, RI, July 1976.
Rasmussen, R. A. Emission of biogenic hydrogen sulfide. Tellus 26:254-260, 1974.
Rasmussen, R. A., and F, W. Went. Volatile organic material of plant origin in the
atmosphere. Proc. Natl. Acad. Sci. USA 53:215-220, 1965.
Robinson, E., and R. C. Robbins. Sources, Abundance, and Fate of Gaseous Atmospheric
Pollutants. Stanford Research Institute, Menlo Park, CA, 1968.
Robinson, E., and R. C. Robbins. Emissions, Concentrations, and Fate of Particulate
Atmospheric Pollutants. Stanford Research Institute, Menlo Park, CA, 1971.
Ryan, P. W., and McMahon. Some Chemical and Physical Characteristics of Emissions from Forest
Fires, Paper 76-2.3, presented at the 69th Annual Meeting of the Air Pollution Control
Association, Portland, OR, June 22-July 1, 1976.
Schnell, R. C., and G. Vali. Atmosphere ice nuclei from decomposing vegetation. Nature
(London) 236:163-165, 1972.
Schnell, R. D., and G. Vali. World-wide source of leaf-derived freezing nuclei. Nature
(London) 246:212-213, 1973.
Spawn, P. Letter Report submitted to U. S. Environmental Protection Agency, Industrial
Environmental Research Laboratory, under Contract No. 68-02-2687, Technical Directive No.
007. GCA Corporation, Bedford, MA, September 20, 1979.
4-42
-------
Springer, J. J, Current and Future Trends of Exhaust Emissions from HD Highway Vehicles.'
Presentation to Federally Coordinated Program of Highway Research and Development,
Federal Highway Administration, College Park, MO, October 1978.
Stith, J. L., P. V. Hobbs, and L. F. Radke. Airborne particles and gas measurements in the
emissions from six volcanoes. J. Geophys. Res. 83:4009-4017, 1978.
Surprenant, N,, R. R. Hall, S. Slater, T. Susa, and M. Sussman. Preliminary Emissions
Assessment of Conventional Stationary Combustion Systems. Vol. II: Final Report.
EPA-600/2-76-046b. U.S. Environmental Protection Agency, Research Triangle Park, NC,
March, 1976.
Surprenant, N. , R. R. Hall, K. T. McGregor, and A. S. Werner. Emissions Assessment of
Conventional Stationary Combustion Systems. Vol. 1: Gas and Oil-Fired Residential
Heating Sources. EPA-600/7-79-029b. U.S. Environmental Protection Agency, Research
Triangle Park, NC, May 1979.
Taback, H. J. , A. R. Brienza, J. Macko, and N. Brunetz. Fine Particle Emissions from
Stationary and Miscellaneous Sources in the South Coast Air Basin. Prepared for
California Air Resources Board by KVB, Inc., Tustin, CA, February 1979.
U.S.. Department of Agriculture, Forest Service. 1977 Wildfire Statistics. U.S. Department of
Agriculture, Washington, D.C., April 1979.
U.S. Department of Commerce. Bureau of the Census^. Statistical Abstract of the United
States: 1979. 100th Edition. U.S. Department of Commerce, Washington, D.C., September
1979.
U.S. Department of Energy. National Energy Plan II--Appendix: Environmental Trends and
Impacts. U.S. Department of Energy, Washington, D.C., May 1979.
U.S. Department of Energy. National Environmental Impact Projection No. 1. U.S. Department
of Energy, Washington, D.C., April, 1978.
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards.
Background Information for' New Source Performance Standards: Primary Copper, Zinc, and
Lead Smelters. Vol. 1: Proposed Standards. EPA-450/2-74-OQ2a. U.S. Environmental
Protection Agency, Research Triangle Park, NC, October 1974.
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards.
Compilation of Air Pollutant Emission Factors 3rd Edition, (including Supplements 1-7).
Parts A and B. AP-42, U.S. Environmental Protection Agency, Research Triangle Park, NC,
August 1977.
U.S. Environmental Protection Agency. Carcinogen Assessment of Coke Oven Emissions. External
review draft. EPA-600/6-82-003. U.S. Environmental Protection Agency, Research Triangle
Park, NC, April 1978a.
U.S. Environmental Protection Agency. National Air Pollutant Emission Estimates, 1940-1976.
EPA-450/1-78-003, U.S. Environmental Protection Agency, Research Triangle Park, NC, July
1978b.
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards.
Protecting Visibility. EPA-450/5-79-008. U.S. Environmental Protection Agency, Research
Triangle Park, NC, October 1979.
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards. National
Air Pollutant Emissions Estimates, 1970-1978. EPA-450/4-80-002. U.S. Environmental
Protection Agency, Research Triangle Park, NC, January 1980a.
4-43
-------
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards. 197)
National Emissions Report. EPA-450/4-80-OQ5. U.S. Environmental Protection Agency,
Research Triangle Park, NC, March 1980b.
U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards. Control
techniques for particulate emissions from stationary sources (preliminary draft). U.S.
Environmental Protection Agency, Research Triangle Park, NC, July 1980c.
U.S. Environmental Protection Agency. Industrial Environmental Research Laboratory. Summary
data received from G. Johnson, Fine Particle Emissions Information Data System. U.S.
Environmental Protection Agency, Research Triangle Park, NC, September 1980d.
Vandegrift, A. E., L. J. Shannon, P. G. Gorman, E. W. Lawless, and E. E. Sallee. Particulate
Pollutant System Study. Volume I - Mass Emissions. APTD-0743. U.S. Environmental
Protection Agency, Research Triangle Park, NC, May 1971.
Vines, R. G., L. Gibson, A. B. Hatch, N. K. King, and D. A. MacArthur. On the Nature,
Properties, and Behavior of Bush-fire Smoke. Commonwealth Scientific and Industrial
Research Organization, Australia, 1971.
Wallace, J. M., and P. V. Hobbs. Atmospheric Science. Chapter 4. Academic Press, New York,
NY, p. 150, 1977.
Went, F. W. Organic matter in the atmosphere, and its possible relation to petroleum
formation. Proc. Natl. Acad. Sci. USA. 46:212-221, 1960.
Went, F. W. j D. E. Slernmons, and H. N. Mozingo. The organic nature of atmospheric
condensation nuclei. Proc. Natl. Acad. Sci. USA, 58:69-74, 1967.
Wexler, H. On the effects of volcanic dust on insolation and weather. Bull. Am. Meteorol,
Soc. 32:10-15, 1951a.
Wexler, H. Spread of the Krakatoa volcanic dust cloud as related to the high-level
circulation. Bull. Am. Meteorol. Soc. 32:48-51, 1951b.
Winchester, J. W., R. J. Ferek, D. R. Lawson, J. 0. Pilotte, M. H. Thiemens, and L. E. Wangen.
Comparison of aerosol sulfur and crustal element concentrations in particle size
fractions from continental U.S. locations. Water Air Soil Pollut., 12:431-440, 1979.
World Almanac and Book of Facts 1980. Grosset and Dunlap, New York, NY, 1980.
Yamate, G. Development of Emission Factors for Estimating Atmospheric Emissions from Forest
Fires. EPA-450/3-73-009, U.S. Environmental Protection Agency, Research Triangle Park,
NC, October 1973.
Zoller, J., T. Bertke, and J. Janzen. Assessment of Fugitive Particulate Emission Factors for
Industrial Processes. EPA-450/3-78-107. U.S. Environmental Protection Agency, Research
Triangle Park, NC, September 1978.
Zoller, W. H., E. S. Gladney, and R. A. Duce. Atmospheric concentrations and sources of trace
metals at the South Pole. Science (Washington, D.C.) 183:198-200, 1974.
4-44
-------
5. ENVIRONMENTAL CONCENTRATIONS AND EXPOSURE
5.1 INTRODUCTION
This chapter has two objectives: (1) to delineate the concentrations of S0« and par-
ticles suspended in the air to which human populations, other organisms, and manmade objects
are exposed; and (2) to show how various sources of air pollutants contribute to these ex-
posures.
The first goal is to describe ambient air concentrations of these two pollutants in a way
which is relevant to the effects they might cause. Measurements of S0?, TSP, and some
chemical components of PM in the ambient air have been made for a long time, mostly with
imperfect methods and procedures. In Chapter 3, the most current information relative to
sources of error in measurement are covered in detail. Here, only those issues that influence
interpretation are mentioned and then only briefly. The reader is advised to consult Chapter
3 for more detail.
Despite imperfections in measurement methods, State and local monitoring data stored by
EPA are the largest available source of information on long-term trends in pollutant concen-
trations and on the geographical distributions of the pollutant levels. Therefore, the exist-
ing monitoring data are presented first to provide an overall perspective regarding S0« and PM
concentrations encountered in the ambient air.
Particulate matter, as a pollutant class, is exceedingly complex both in regard to its
physical properties and its chemical composition. In Chapter 2, those characteristics of
particles generally observed in most atmospheres are discussed in detail. Consequently, in
this section only those features of chemical composition and physical size are treated that
influence data interpretation. The reader is directed to Chapter 2 for more detail on these
subjects.
Recently, particle-measurements have been collected and analyzed to estimate the relative
contributions of important sources. In this case, the elemental and chemical complexity of
the particles proves to be valuable since many source types have identifiable characteristic
chemical signatures. Consequently, it is often possible to make at least approximate assign-
ments of the relative amounts of suspended PM derived from road dust, power plants, auto-
mobiles, and other common sources, provided that an adequate description of the source
signature is available. Since this technique is still new, only a few studies are available,
some of which are discussed to show the approximate source contributions in representative
cases. For a more complete description of particle emission factors and inventories, Chapter
4 should be consulted.
Ultimately, the importance of the ambient air measurements of pollutant concentrations is
in identifying and predicting undesirable effects. When the effect considered is visibility
reduction, the important factors are concentrations of light scattering and absorbing
particles over the geographical scale of several miles. In materials damage, concentrations
5-1
-------
of SO, and of soiling particles are important over time scales of months or years. Wher
considering effects other than visibility, the important matter is the dose. Dose incor-
porates concentration, time, uptake and loss, and the relationship among these parameters.
Throughout their lifetime, people inhale a complex mixture of gases and particles. Other
living things, vegetation and animals, are also exposed to the same complex mixtures, the
composition of which varies with time at any given location because of changing ataiospheric
conditions and source contributions. The biological effects of air pollution are functions of
dose delivered to the receptor and the ability of the receptor to cope with the resultant
stress. In humans, the stress experienced by a critical organ or receptor tissue from
particle inhalation depends on particle size, composition, morphology, acidity or alkalinity,
and other physicochemical properties of the aerosol. The delivered dose is also a function of
the anatomical features of the receptor as well as the manner of breathing, breathing rate,
and integrity of bodily defense systems.
It is almost impossible to measure directly the air pollution dose to a population or
even to an individual, except in the laboratory. As an alternative to direct measurement of
dose, exposure can and often must be used as an approximation of dose for studies on air pol-
lution risk and effects. The exposure-response relationship for air pollution is most impor-
tant for establishing standards. Unfortunately, to extrapolate from measurements of ambient
levels at a few locations to individual or population exposure levels is a very difficult task
at present. The contribution of outdoor air to indoor concentrations is still being investi-
gated. The additional exposures to gases and particles from nonoccupational indoor sources
are not adequately known.
Indoor air quality and activity patterns complicate air pollution exposure estimates and
are discussed later in this chapter. First, the ambient outdoor concentrations of SO, and PM
are examined.
5.2 AMBIENT MEASUREMENTS OF SULFUR DIOXIDE
Ambient concentrations of $02 are determined by the following factors:
1. Density of emission sources.
2. Source characteristics such as stack height, exit velocity, and source strength.
3. Local meteorological conditions.
4. Local topography and surrounding buildings.
5. Reaction rate for oxidation of SO^.
6. Removal rates by precipitation, deposition at surfaces, and other reactions.
These factors interact in such a way that in urban and industrialized areas with high
densities of S02 emissions, the S02 concentrations are much higher than in surrounding rural
areas. It is quite common to find gradients in S02 concentration within these industrialized
areas, with a central core area reporting the highest S02 concentrations. This pattern is
shown diagrammatically in Figure 5-1.
5-2
-------
UJ
o
CM
O
to
iu
MICRO AND MIDDLE
WITH MAJOR POINT SOURCES
RURAL AREAS
SUBURBS |
I
URBAN CORE
CITY LIMITS
I SUBURBS
I
RURAL AREAS
Figure 5-1. Distribution of annual mean sulfur dioxide concentrations across an
urban complex, as a function of various spatial scales,
Source: Ball and Anderson (1977).
5-3
-------
Where SCL emissions are dominated by a single source or a few point sources, the patten
of SOp concentrations could be different from the pattern displayed in Figure 5-1. Depending
on topography, meteorology, and source characteristics, the concentration patterns may bi
asymmetrical, and the temporal distribution may be skewed to low mean values with a fev
intermittent high peaks. These differences in concentration patterns may be important ir
relation to the types of effects experienced in exposed human populations.
Host urban areas have experienced dramatic improvements in air quality as a result of
restrictions on sulfur in fuel, better controls on new and existing sources, displacement of
old sources and building of new sources in less populated regions, and construction of taller
stacks.
This section presents SO, concentration data for specific locations and areas where
levels are currently high. The national status of SCL concentrations is reviewed, along with
temporal trend data. A comparison is made between SCL levels in six cities in the early
1960's and concentrations in the late 1970's. Insights into important determinants of popu-
lation exposures are presented in the discussion of diurnal and seasonal SCL concentration
patterns. Since SCL can be measured by a variety of methods (see Chapter 3), a brief
discussion of SCL monitoring and instruments precedes the substantive sections on con-
centrations.
5.2.1 Monitoring Factors
The EPA is now in the process of revising Federal, State, and local air monitoring net-
works. By 1981, States will be operating a selected number of sites in the National Air
Monitoring Station (NAMS) Network. These sites are to be located in densely populated areas
with the highest pollutant concentrations. They are designed to serve in assessing pollutant
trends and progress in meeting standards. By 1983, State and local agencies are to be
operating the State and Local Air Monitoring Station (SLAMS) Network. This network is
designed to be part of each State's implementation plan. It is expected that this will mean
fewer sites than are currently in operation; however, the Federal coordination of air monitor-
ing should provide much-needed quality control. The trend toward reduction in the number of
stations is already apparent in the 1977 SCL data. There were 117 fewer monitoring sites
reporting data in 1977 than in 1976 (2365 versus 2482). Many States terminated all or most of
their 24-hour West-Gaeke bubbler sampling in 1978, and most remaining bubbler stations are
being fitted with temperature controls to avoid sample degradation (see Chapter 3). However,
state and local agencies are relying primarily on continuous monitoring equipment whenever
possible.
Nationally, SCL monitoring is not as extensive as TSP monitoring. In 1978 there were 947
sites with continuous monitoring equipment and 1298 bubbler sites. Every State conducted SCL
monitoring. All reported sites produced useful information on short-term (1- to 24-hour) SOg
concentrations. However, only those sites reporting a specified number of hourly or daily
observations per year are considered valid in terms of their annual mean. For the EPA,
5-4
-------
minimum criteria for a valid annual mean are 6570 hourly values from a continuous monitor or
five 24-hour values in each quarter from a bubbler monitor. It is in the number of S02 sites
with valid annual means that national coverage appears inadequate. Only 99 of the 1298
bubbler sites (or 7.6 percent) had valid annual means in 1978; only 385 of the 947 (or 40.7
percent) continuous sites were considered valid. There were seven States with no valid annual
SO- data for 1978. The EPA is currently taking steps to improve the quality of SO- data and
to increase the number of representative sites reporting valid data.
For valid bubbler sites, the average number of 24-hour observations in 1978 was 60. The
number of observations per site ranged from 28 to 322. For the valid continuous sites, the
mean number of observations was 7806 hourly measurements. This ranged from a minimum of 6578
hours to a maximum of 8755 hours.
5.2.2 Sulfur Pioxide Concentrations
Although there are natural sources of SOg such as vofcanoes (see Chapter 4) that can be
important in proximity to the source,- they are usually unimportant on an urban scale. Sulfur
dioxide has a rather short half-life in the troposphere (see Chapter 6), and background levels
are often below the monitor's detection limit. Therefore, it is not surprising that the re-
3
ported annual mean S09 concentration is 3 ug/m in some nonurban locations. It may be lower;
3
most monitoring techniques have detection limits close to 10 ug/m , and apparent zero values
are commonly recorded as half the detection limit.
Monitoring in urbanized areas near industrial sources that use sulfur-bearing fuels shows
rather high concentrations of SO,. In 1978 the annual mean concentrations obtained by S0?
bubblers ranged from 3 to 79 yg/m . The valid continuous monitors registered 1978 annual mean
3
concentrations ranging from 3 to 152 ug/m .
The concentration of SO,, is affected by meteorological variables influencing transport,
dispersion, and removal, as well as by topography and configuration of sources. Spatial and
temporal variations in these parameters are reflected in the range of maximum and 90th per-
cent! le concentrations reported across the Nation. For bubbler sites, the lowest 24-hour
3 3
maximum value reported by a site was 3 ug/m ; the highest was 907 ug/m . For the valid con-
2
tinuous sites, the spread of 24-hour maximum values was greater, ranging from 10 ug/m at one
site to 2512 jjg/m at another site. Among all continuous sites reporting in 1978, the extreme
24-hour value was 3931 pg/rn .
Figure 5-2 presents the distribution of annual averages for all valid continuous mopi-
toring sites in 1978. On this time scale, the most commonly measured values fell between 20
3 3
and 30 \ig/m , with most values below 60 ug/m . Most monitoring stations were situated
specifically to detect higher urban or source-specific levels of SO-, however, and the data in
Figure 5-2 may be judged more nearly representative of populated areas or areas influenced by
specific sources rather than the entire U.S. land area. The following section discusses the
effect of site location.
5-5
-------
at
ui
CO
1
u.
o
DC
UI
m
3
2
120
110
100
90
80
70
60
50
40
30
20
10
0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
ANNUAL AVERAGE CONCENTRATION
Figure 5-2. Histogram delineating annual average sulfur dioxide concentrations
for valid continuous sampling sites in .the United States in 1978.
Source: SAROAD.
5-6
-------
5.2.B Sulfur Dioxide Concentration By Siteand Region
5.2.3.1 Analyses by Various Site Classifications—In this section, the distributions of annual
mean and 90th percentile S0? concentrations by site descriptors are presented for bubbler and
for continuous sampling methods. A two-descriptor code has been assigned to each site.
Distributions for every combination of Type 1 (population, source, background) and for Type 2
(central-city, suburban, rural, remote) are not presented. In some cases the designations are
contradictory, such as "population-remote" or "background-central city." The purpose of
presenting these distributions is to permit comparison of these two categories of sampling
methods and to examine the SCL concentrations as a function of location.
In Table 5-1, a cross-tabulation of mean concentrations by method is presented for
center-city sites that are primarily either population oriented or source oriented. The third
and fourth numbers in each cell are the column percentage and total percentage of sites having
concentrations within the designated range. Examination of each of these numbers reveals that
bubblers are, on the average, reporting lower concentrations than the continuous instruments,
an expected result because of the method biases reported in Chapter 3. In Table 5-1, 14.4 per-
cent of the population-oriented continuous monitors reported mean concentrations above 62
•3
|jg/m , whereas only 1.6 percent of the bubblers reported such concentrations. Of the source-
oriented sites, a higher percentage (7.1 percent) of the bubblers were above 62 pg/m , but
this was still less than the 12.5 percent of the continuous monitors in this category.
5.2.3.2 Regional Comparisons—Regional differences in SOn concentrations are not striking.
In part, this is the result of the location of the monitor. In Section 5.2.4, it is shown
that high SO- levels are found around smelters in otherwise clean areas. In the eastern and
northern States, most continuous S00 monitoring is in urbanized areas. In 1978, mean concen-
3
trations across all continuous monitors in Regions I, IIS III, IV, and V ranged from 23 ug/m
to 51 p.g/m (see Table 5-2). The maximum annual mean among the valid sites in these regions
•3 Q
ranged from 59 |jg/m in Region I to 140 pg/m in Region III. In the less industrialized or
less populated regions (VI through X), the mean annual concentration across all sites in each
region ranged from 8 ug/m to 40 M9/m •
Even with the summary of the 1978 continuous SO- data, it is difficult to speculate on
regional differences in SOp concentrations. Sulfur dioxide monitors are not systematically
sited for population exposure monitoring purposes; •sometimes SO^ instruments are used for
monitoring the local influence of strong point sources (e.g., the smelters noted above).
Therefore, better indicators of regional differences in SO^ concentrations and population
exposures are sulfur emission patterns (see Chapter 4).
The data base used in compiling Figure 5-3, collected between 1974 and 1976, offers finer
spatial resolution of national SO, concentrations on a county scale. The second highest 24-
hour average S0? concentration by county is displayed. Some areas in the West with extremely
high concentrations were still problem areas in the late 1970's (see Table 5-3). Several
counties and cities are still reporting high concentrations; however, one should not infer
that the reported concentration prevails throughout the county. High readings may exist at
5-7
-------
TABLE 5-1. GROSSTABULATIQN OF ANNUAL MEAN S02 CONCENTRATION BY OR CONTINUOUS)
FOR POPULATION-ORIENTED AND FOR SOURCE-ORIENTED CENTER-CITY SITES
01
00
Purpose of Site
Annual mean
S02 concentration,
Mff/ni3
2-7
Number of sites
Percent of row
Percent of column
Percent of total
7-18
Number of sites
Percent of row
Percent of column
Percent of total
18-33
Number of sites
Percent of row
Percent of column
Percent of total
^33-62
'Number of sites
Percent of row
Percent of column
Percent of total
>62
Number of sites
Percent of row
Percent of column
Percent of total
Column total
Number of sites
Percent of total
Bubbler
139
90.3
27.6
17.7
159
83.2
31.6
20.2
106
55.2
21.1
13.5
91
45.3
18.1
11.6
8
16.3
1.6
1.0
503
63.9
Population
Continuous
15
9.7
5.3
1.9
32
16.8
11.3
4.1
86
44.8
30.3
10.9
no
54.7
38.7
14.0
41
83.7
14.4
5.2
284
36.1
Row Total
154
—
—
19.6
191
__
—
24.3
192
__
—
24.4
201
—
—
25.5
49
__
—
6.2
787
100.0
Bubbler
9
75.0
16.1
8.7
18
64.3
32.1
17.3
12
46.2
21.4
11.5
13
46.4
23.2
12.5
4
40.0
7.1
3.8
56
53.8
Source
Continuous
3
25.0
6.3
2.9
10
35.7
20.8
9.6
14
53.8
29.2
13.5
15
53.6
31.3
14.4
6
60.0
12.5
5.8
48
46.2
Row Total
12
—
—
11.5
28
—
—
26.9
26
--
—
25.0
28
—
—
26.9
10
—
—
9.6
104
100.0
Note: 1 ppm S02 = 2620 ug/nf
Source: SAROAD.
-------
TABLE 5-2. CONTINUOUS S02 MONITOR RESULTS BY REGION,
Region Type
I Valid
All
II Valid
An
III Valid
An
IV Valid
All
V Valid
All
en
r
*° VI Valid
All
VII Valid
All
VIII Valid
All
IX Valid
All
X Valid
All
Number of
sites
22
72
51
87
26
108
100
203
111
254
13
32
13
38
12
49
19
52
18
29
Min.
6665
185
6597
140
657B
94
6678
421
6580
129
6631
1669
6769
334
6741
373
6857
105
6651
625
Number of
observations
per site
Mean Max. s.d.
7519
4582
7540
5815
7562
4381
8305
5754
7640
5512
7443
5461
7540
4676
7739
4694
7952
4973
7854
6158
8416
8416
8697
8697
8638
8638
8755
8755
8715
8715
8452
8452
8325
8325
8624
8624
8638
8638
8677
8677
567
2720
546
2380
670
2534
574
2848
625
2327
607
2072
439
2624
514
2358
507
2525
464
2526
Arithmetic Means
Min. Mean Max.
16
8
15
15
12
7
5
3
7
3
3
3
6
4
3
3
3
3
13
13
33
39
37
41
51
45
23
23
36
37
13
12
31
25
40
34
8
24
34
33
59
138
78
94
140
140
63
77
84
192
31
56
47
82
152
152
29
87
78
78
s.d.
12
23
16
19
23
21
12
13
16
25
1
13
14
20
47
39
6
20
18
17
90th Percenti le
Min. Mean Max.
34
14
35
35
34
14
9
'3
10
5
3
3
13
5
3
3
3
3
35
29
65
77
72
78
97
86
54
49
70
73
31
29
62
52
100
89
16
49
90
72
147
340
159
173
282
282
135
180
167
501
69
160
94
155
488
488
48
213
150
150
s.d.
27
52
30
33
46
40
27
29
30
50
19
38
25
41
146
113
12
49
38
38
Note: 1 ppm S02 = 2620 ug/m .
Source: SAROAD.
-------
en
Figure 5-3. Characterization of 1974-76 national SO2 status is shown by second highest 24-hr.
average concentration. Asterisks denote counties for which this level exceeded 36§ /jg/m3.
(The current 24-hr, primary standard is 365 ftg/m3, which is not to be exceeded more than once
per year. Alaska and Hawaii reported no such exceedences.)
Source: Monitoring and Reports Branch, Monitoring and Data Analysis Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection Agency.
-------
TABLE 5-3. ELEVEN S02 MONITORING SITES WITH THE HIGHEST ANNUAL MEAN
CONCENTRATIONS IN 1978 (VALID CONTINUOUS SITES ONLY)
Annual Means,
Location pg/m
Helena, Deer Lodge
Co., MT
Pittsburgh, PA
Helena, Deer Lodge
County, MT
Magna, Salt Lake Co. ,
UT
Toledo, OH
Pittsburgh, PA
Buffalo, NY
Kellogg, Shoshone Co.,
ID
Shoshone Co. , ID
New York City, NY
Mingo Junction, OH
152
140
95
93
84
79
78
78
77
77
76
Maximum 24- hr,
pg/m Description
2512
602
1450
811
915
376
267
294
493
296
329
Rural mine smelter
Center-city industrial
Rural industrial,
1.6 miles east of
smelter
Suburban industrial
Center-city industrial
Suburban industrial
Suburban industrial
Suburban residential
Suburban industrial
Center-city residential
Center-city industrial
Note: 1 ppm S02 = 2620 ug/m~
Source: SAROAD.
5-11
-------
one or more monitoring sites (for example, Deer Lodge County, MT), but it is likely that there
are substantial gradients across the county and almost certainly across an air quality control
region (AQCR).
5.2.4 Peak Localized Sulfur Dioxide Concentrations
5.2.4.1 1978 Highest Annual Average Concentrations—The reasons for variability in ambient
air S02 levels have been mentioned in earlier sections. This section examines United States
locations with highest annual averages and highest maximum concentrations and analyzes the
distribution of SCL concentrations nationally by site descriptors. However, because of the
differences in S02 concentrations between bubbler and continuous monitoring, the distributions
by site descriptor will be treated separately for each method.
Table 5-3 lists annual mean and maximum 24-hour concentrations for the 11 valid con-
tinuous monitoring sites in the United- States with the highest annual means in 1978. The
highest annual mean concentration was 152 (jg/m , at a site in Montana 2 miles northeast of a
smelter. That site also had the highest 24-hour concentration (2512 (jg/m ) of any valid
o
continuous monitor. The maximum hourly level was 7205 (jg/m , and the second highest hourly
level was 6026 (jg/m at this site. Of the highest 11 sites, 5 were associated with smelters,
5 were associated with industrialized areas or towns, and one (New York City) was a densely
populated city. In New York City, S02 emissions from space heating, power plants, and a
variety of industrial sources resulted, in a high annual mean concentration. These were peak
reporting sites; the urban sites do not .typify an entire city. Conversely, even higher
concentrations may exist in unmonitored neighborhoods.
5.2.4.2 1978 Highest Daily Average Concentrations—About 50 monitoring sites in the United
States have consistently reported maximum 24-hour average S0? levels in excess of 300 (jg/m in
recent years. Almost all of these have had very high second and third highest values also.
Many of the sites having high daily averages were located near specific industrial sources
such as smelters, steel plants, and paper mills. Monitors around smelters have frequently
o
reported 24-hour values of 1000 to 3000 (jg/m , the highest levels in the United States.
In 1978, high 24-hour S02 values occurred in 17 States encompassing all major regions of
the country. Ten of the highest sites were .in Montana, six were in Wisconsin, and six in
Minnesota. Most of these were close to intense sources. However, several urban sites,
especially center-city sites in industrialized, communities such as Philadelphia and
Pittsburgh, PA; New York,„ NY; Toledo, OH;- and Hammond, ID, still have high maximum 24-hour
3
values, above 250 to 300 (jg/m .
5.2.4.3 Highest 1-hour Sulfur Dioxide Conce.ntrations-1978 National Aerometric Data Bank (NADB
Data—Single hourly S02 values greater than 1000 (jg/m (0.4 ppm) have been measured in about
100 cities and counties in 28 States in recent years.. Such values were very widespread across
the country; Maine, Florida, Montana, Texas,, Arizona, and Washington all had sites in this
category. Of these top 100 sites, all except 15 also had second highest values in excess of
1000
5-12
-------
Hourly measurements this high are comparatively infrequent and for most of these 100 high
sites, less than 1 percent of hourly values were in this category. But for a few sites,
notably those close to metal-smelting operations in a few cities, such values were observed up
to 5 percent of the time. Highest 1-hour values were found in Deer Lodge County, MT, where
2
several measurements over 5000 pg/m were recorded at two sites in 1978. Anaconda, MT; Miami
and San Manuel, AZ; Newark, DE; Buffalo County, WI; and St. Charles County and North Kansas
City, MO, all reported at least one value in excess of 4000 pg/m .
5.2.5 Temporal Patterns in Sulfur Dioxide Concentrations
5.2.5.1 Diurnal Patterns—In some locations, SO- concentrations have distinct temporal
patterns. These patterns depend on the variability of meteorological factors and on the vari-
ability of source emissions.
Diurnal variations in SO^ concentrations reflect the changing dispersion characteristics
of. the lower atmosphere and variations in mixing height. If emissions are predominantly from
low-level sources such as residential and institutional space heating, the highest hourly con-
centrations will frequently occur at night and in the early-morning hours. At these times,
low mixing height and decreased windspeeds lead to higher concentrations. During the day more
vertical mixing usually occurs and windspeeds increase, diluting low-level emissions. Figure
5-4 gives the composite diurnal pattern of hourly concentrations for SCL for the month of
December 1978 in Watertown, MA. The pattern just described is apparent.
In locations where SO- emissions from taller stacks are the major SOp source, a different
diurnal pattern can occur. In these situations, typical of power plants and smelters, the
highest concentrations usually occur in the morning hours just after sunrise. Levels can be
almost zero at night if the source is emitting into a stratified region above a lower level
inversion. Upon inversion breakup, when heating at the surface causes vertical mixing, an
elevated plume can be mixed to the ground. Fumigation conditions lasting from several minutes
to several hours can occur. Montgomery and Coleman (1975) analyzed the effects of tall stacks
on the peak-to-mean ratios for different averaging times and discussed the influence of inver-
sion breakup. In essence, even with tall stacks, inversion breakup that catches the plume and
brings it to the surface can occur. So the peak-to-mean ratio is almost independent of stack
height. The frequency of occurrences of fumigation, on the other hand, would most likely be
less with taller stacks.
Some similarity can be found in comparing the diurnal pattern of hourly averages for
Watertown, MA (Figure 5-4) and St. Louis, MO (Figure 5-5). In February 1977, a major local
source of S0? was still in operation in St. Louis. The midmorning and late night maxima were
again associated with diurnal variations of meteorological factors. By February 1978, the
source had shut down, and the reported SO,, levels at the monitoring stations reflected this
fact. The absence of low-level stacks emitting into a stable layer of air near the surface at
night was noticeable. Concentrations did not build up at night as in the previous year.
5-13
-------
0.040
0.030
I
cc
Ul
o
o
u
CM
O
0,020
0.010
1 I I I I I I I
I I
lilt
I I
I
I I
8 10 12 14 16 18 20 22
HOUR
24
Figure 5-4. Composite diurnal pattern of hourly sulfur dioxide concentrations are shown for
Watertown, MA, for December 1978.
Source: Spengler (1980).
5-14
-------
0.200
0.100
0.090
0.080
0.070
a 0.060
z*
2 0.050
0.040
ui
o
§ 0.030
CM
O
Ui
0.020
0.010
I I I I I I I I I I I I I I I I
I I I I I
FEBRUARY 1977
FEBRUARY 1978
I I I I I I I I I I I I I I I I I I I I I
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
HOUR
Figure 5-5. Monthly means of hourly sulfur dioxide concentrations are shown for St. Louis (city site
no. 26-4280-007, "Broadway & Hurck") for February 1977 and 1978.
Source: Spengler (1980).
5-15
-------
Diurnal patterns of hourly SO, values for the industrialized river valley town of
Steubenville, OH, are shown in Figure 5-6. In June 1976, a distinctive maximum in the diurnal
pattern appeared. In July 1977, there was no apparent variation across the hours.
In conclusion, it can be said that the variations in hourly concentrations are influenced
by source configuration and meteorological dispersion. Therefore, it is difficult to general-
ize about diurnal patterns of hourly S0~ concentrations. Although there may be some
similarities, the daily patterns in SCL concentrations are different for different locations
and can change in time for a given location.
5.2.5,2 Seasonal Patterns—Concentrations of SO, display seasonal variability. The vari-
ability is most pronounced in areas in which there is strong seasonal variation in the
emission-source strength or in meteorological conditions. Obviously, in urban areas where
space heating is the major source of S02, the levels will be much higher during the heating
season. Figure 5-7 illustrates just such situations in Watertown, MA and Steubenville, OH.
The highest monthly mean concentrations occur in the winter months.
Figure 5-7 also shows the data for St. Louis, MO where the seasonal pattern is different.
Here a local industrial s'ource dominates SQ~ concentration patterns around the monitor. The
higher monthly mean concentrations occur in the months with the higher frequency of southerly
winds. The source is to the south of the monitoring station. Any increase in SO^ concen-
trations as a result of the winter heating season is not apparent.
5.2.5.3 Yearly Trends—The SQ2 levels in most urban areas in the United States have improved
steadily since the mid-I9601s. The trend of decreasing S02 concentrations can be resolved
into three distinct periods. From 1964 to 1969, the improvement was gradual. In the middle
period, between 1969 and 1972, the improvement in most urban areas was more pronounced. Since
1973 the improvement has again become slower. The 1977 EPA trends report states: "In most
urban areas, this is consistent with the switch in emphasis from attainment of standards to
maintenance of air quality; that is, the initial effort was to reduce pollution to acceptable
levels followed by efforts to maintain air quality at these lower levels." From 1972 through
1977 annual averaged SQy levels dropped by 17 percent, or an annual improvement rate of about
4 percent per year. Figure 5-8 summarizes the annual average SQy concentrations for 32 urban
National Air Sampling Network (NASN) stations for the years 1964 through 1971. In this
figure, the first-two periods are apparent. In Figure 5-9, the national trends in annual
average SO, concentrations from* 1972 through 1977 at 1233 sampling sites are displayed. In
this figure, the diamond symbolizes the composite annual average concentration; the triangle
is the median value, while the circles are extreme values and the thick band covers the 25th
to 75th percent!"le range.
Over the period of 1970 through 1977, SQ2 emissions have decreased only slightly (U.S.
Environmental Protection Agency, 1978). In 1970 the estimated annual manmade S02 emissions
were 29.8 million metric tons. By 1977 this was reduced to 27.4 million metric tons. The
improvement in the ambient air quality levels for S02 reflected the displacement of sources
from urban areas to rural areas, restriction of sulfur content of fuels used by low-level
5-16
-------
0.080
0.070
0.060
Q.
Q.
O 0.050
UJ
u
0.040
o 0.030
CM
8
0.020
0.010
I I
I I I I
I
_L
10
12 14
HOURS
16
18
20
22
24
Figure 5-6. Monthly means of hourly sulfur dioxide concentrations are shown for Steubenville, OH
(NOVAA site 36-6420-012) for June 1976 and July 1977.
Source: Spengler (1980).
5-17
-------
0.040
Q.
Q.
%
o
St
EC
0.030
0.02O
U]
u
O
u
CM
O
U)
0.010
CITY SITE
STEUBiNVtLLE 36 6420 012
ST. LOUIS 26 4280 007 Q—-—Q
WATERTQWN 22 2380 003
I
I
I
I
I
JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC
MONTH
Figure 5-7. Seasonal variations in sulfur dioxide levels are shown for Steubenviile, St. Louis, and
Watertown.
Source: Spengler (1980).
5-18
-------
1964
1972
Figure 5-8. Annual average sulfur dioxide concentrations
are shown for 32 urban IMASN stations.
Source: National Academy of Sciences (197i).
5-19
-------
1972
90TH PERCENTILE
c_ O> •* COMPOSITE AVERAGE
56
48
to
-I
OI
a, 40
g
cc
z 32
UJ
o
z
o
o
111
S 24
X
o
o
cc
3
^3 16
3
CO
8
n
C
-
—
_
_
I I I A •+ MEDIAN
5
0 I
O
A
T
(>« 10TH PERCENTILE
0
0
0 n -
—
•»> . * , '
i I mil a 1 | —
^^ >V yv
o o
—
A A A A
A
1
—
* T T T I
1973
1974 1975
YEAR
1976
1977
Figure 5-9. Nationwide trends in annual average sulfur dioxide concentrations from 1972 to 1977 are
shown for 1233 sampling sites.
Source: U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards (1978a).
5-20
-------
area sources, building of sources with taller stacks, and source emission controls.
The first air quality criteria document for S0x, published in 1969, presented the fre-
quency distributions for SOp levels in selected American cities for 1962 through 1967 (National
Air pollution Control Administration, 1970). The 1960's data came from the Continuous Air Mon-
itoring Project (CAMP), which operated continuous monitors in a few of the largest U.S. cities:
Chicago, Philadelphia, St. Louis, Cincinnati, Los Angeles, and San Francisco. Improvements in
SOg levels in each of the six cities are demonstrated by comparing 1962 to 1967 data with data
for 1977 (Table 5-4). In each city there is more than one continuous monitor now operating.
The station reporting the highest levels in 1977 was used in order not to overemphasize any
improvement. The comparison is only an approximation because the locations of the monitors
and the instrumental methods used were not the same- as those reported in the 1969 document.
In each city the peak concentration decreased. In most cities -the 1977 peak was less
than one-half the earlier values. The only exception was St. Louis, where the earlier peak
was 0.72 ppm and the 1977 peak was 0.67 ppm. The result was not unexpected in that the earlier
summary was a composite frequency distribution of 5 years of monitoring.
TABLE 5-4. COMPARISON OF FREQUENCY DISTRIBUTION OF S02 CONCENTRATION (P£M)
DURING 1962-673 AND DURING 1977°
Frequency Distribution
City
Chicago
Philadelphia
St. Louis
Cincinnati
Los Angeles
San Francisco
Year
1962-67
1977
1962-67
1977
1962-67
1977
1962-67
1977
1962-67
1977
1962-67
1977
Percenti le 30
0.03
0.01
0.03
0.01
0.02
0.01
0.01
0.01
0.01
0.01
::
50
0.08
0.02
0.05
0.02
0.03
0.01
0.02
0.02
0.01
0.02
0.001
70
0.17
0.03
0.09
0.04
0.05
0.03
0.03
0.02
0.02
0.02
0.01
0.01
9,0
0/32
0.06
0.-.21
0.08
0.11
0.10
0.07
0.04
0.04
0.03
0.03
0.01
of SO,, (ppm)
99
0.65
0.12
0.45
0.23
0.26
0.37
0.18
0.08
0.08
0.05
0.07
0.03
Maximum
0.95
0.25
0.85
0.44
0,72
0.67
0.53
0.29
0.25
0.09
0.17
0.03
Concentrations from CAMP stations as reported in Air Quality Criteria for
Sulfur Oxides, National Air Pollution Control Admini strati on (1970).
Concentrations from National Aerodynamic Data Base (1977).
3
Note: 1 ppm S02 = 2620 ug/m .
5-21
-------
More stable indicators of improved air quality are provided by the 50th, 70th, and 90th
percentile concentrations. In Chicago, the 50th percentile concentration dropped from 0.08
ppm to 0.022 ppm. In Philadelphia, the levels improved substantially; the 50th, 70th, and
90th percentile concentrations were less than one-half the earlier values. St. Louis and
Cincinnati showed modest improvement; their 50th percentile concentrations were lower in 1977
than they were in the mid-1960's. The highest concentrations occurred as frequently in St.
Louis as they did earlier, but in Cincinnati they occurred less frequently. Review of the St.
Louis SO, data showed improved air quality for most of the city. The high concentrations
reported in 1977 were typical of only a small section of the city. Los Angeles showed im-
provement in reducing the high concentrations, but the 50th percentile concentration was
actually slightly higher in 1977 than it was in the previous decade. Similarly, San Francisco
trimmed the peaks but had a very low median value.
In summary, the frequency of peak levels has been reduced in most urban areas. The
steady improvement of S09 ambient air quality has been slowed somewhat in recent years. Only
3
1 percent of the S02 monitoring sites show levels above 80 ug/m , the current annual NAAQS.
In 1974, the annual mean S0_ standard was exceeded in 3 percent of the monitoring stations (31
of 1030), compared with 16 percent in 1970. In 1977 and 1978, 2 percent of the sites reported
violations of the 24-hour standard. In 1974, this standard was exceeded in 4.4 percent of the
reporting stations (99 of 2241), compared with 11 percent in 1970. Many of these sites re-
porting violations of the 24-hour standard are in remote areas near large point sources.
5.3 AMBIENT MEASUREMENTS OF SUSPENDED PARTICULATE MASS
The general character of airborne matter designated as atmospheric suspended particles
— Q
has been described in Chapter 2. These particles range in size from about 5 x 10 m, roughly
-4
corresponding to agglomerates of a few tens or hundreds of molecules, up to about 10 m,
specks of material discernable to the human eye. A useful division of these particles by size
into fine and coarse fractions occurs in the range of 1 to 2 x 10 m or 1 to 2 urn, as was dis-
cussed in Chapter 2.
The mass of suspended particles, generally concentrated in particles above about 0.3 urn,
is usually estimated by filtration of known volumes of air. The goal of this filtration
process is the separation of the gas phase from liquid and solid condensed phases of atmos-
pheric aerosol. Thus, the mass of material accumulated on a filter is taken to represent the
volume of aerosol filtered, and results are presented as |jg of PM/cubic meter of aerosol,
abbreviated ug/m .
Chapter 3 discusses some of the complications of commonly used filtration methods,
including retention of reactive gases such as SO- and HN03 and loss by evaporation of water
and other moderately volatile substances. While commonly used filtering media are highly
efficient for collection of fine particles, samplers of coarse-particle concentrations have
had major design defects. Despite these complications, which were discussed in detail in
Chapter 3, the largest body of information on the distribution of suspended PM in time and
5-22
-------
space has been obtained with the hi-vol, or TSP method. Routine monitoring information is
available from the NADB maintained by EPA for many sites. Some are EPA sites; many are
operated by State and local agencies.
The following discussion relies mainly on NADB data; additional analyses are to be found
in the 1976 and 1977 National Air Quality and Emissions Trends reports, in Trends in the
Quality of the Nation's Air reports in 1980, and in the document Deputy Assistant Admini-
strator' s Report cm Ambient Monitoring Activities - Air Portion. (U.S. Environmental Pro-
tection Agency, 1977; 1978; 1980a; 1980b).
5.3.1 Monitoring Factors
The accuracy and precision of PM monitoring are limited by three general considerations:
sampling methods, including instrumentation, analytical methods, and quality assurance; samp-
ling frequency; and location of monitors. Chapter 3 discusses the first of these considera-
tions; the second and third are discussed in this chapter. Sampling frequency affects the
confidence limits on mean TSP concentrations and annual or seasonal trends. It is appropriate
to discuss this limitation at the beginning of this section before the 1978 national TSP data
base is presented. The siting of particle monitors significantly influences the levels
measured and, hence, the*' interpretation of data. These considerations are presented with
examples in several sections of this chapter.
5.3,1.1 Samp1ing Frequency--1n 1978, there were 4105 TSP monitoring sites in the United
States and its territorie*s that reported data to the NADB of the U.S. EPA. Of these, only
2882 had enough observations per quarter and per year for the data to be considered valid for
estimating annual averages. The number of sites reporting valid data ranged from zero in
Delaware and American Samoa to 318 in Ohio. The most populous states, California and New
York, had 60 and 236, respectively.
The U.S. EPA has established a uniform sampling schedule to be followed by all State and
local agencies. It requires a 24-hour sample (midnight to midnight) every sixth day. Hence,
in 1 year there are 60 or 61 possible sampling days from which to derive the mean value and
distribution, and to determine attainment of current standards. In 1978, 14 percent ofxall
reporting sites had 60 or more observations.
Sampling days are missed and samples must be voided for a variety of reasons. Therefore,
a minimum requirement has been established for considering the data from any site as valid in
determining an annual average: There must be at least five observations during each quarter
of a calender year. Of the Federal, State, and local TSP sites reporting data to NADB, 70
percent met this requirement.
The distribution of observations for the 2882 valid monitoring sites in 1978 is shown in
Figure 5-10. Of these sites, 10 percent had less than 47 observations and 50 percent had more
than 56 observations. However, 90 percent of the monitoring sites collected fewer than 60
samples. Three percent of the monitors sampled at an equivalent frequency of 1 day in 3, and
fewer than 2 percent collected samples at a frequency of 1 day in 2.
5-23
-------
1000
900
800
700
600
§ SOO
ui
S 400
D
Z
Z
K
Ul
VI
CO
O
cc
X
u.
O
cc
Ul
CD
z
300
g 200
u.
CO
O
100
90
80
70
60
50
40
30
20
10
I I
I I I I I I I
1 I
IT
I I I I III
I
I
I I I I I I
I
_L
_L
I L
l
0.01 0.1 0.5 1 2 5 10 20 30 40 SO 60 70 80 90 95 98 99 99.8 99.9 99.99
% WITH NUMBER OF OBSERVATIONS LESS THAN
Figure 5-10. Distribution shows the number of TSP observations per valid site in 1978;
total is 2882 sites.
Source: SAROAD.
5-24
-------
The current NAAQS for TSP consists of an annual geometric mean and a once-per-year daily
value. Frequency of monitoring is a fundamental parameter of the aiir quality data used fov
comparison with the standards. The period of determining an annual average for comparison is
a calendar year. If the number of 24-hour observations is less than 365, then the true mean
concentration for the year can only be expressed as residing within a range of values. On the
assumption that the actual distribution of values is log normal, confidence intervals can be
calculated from the geometric mean and the geometric standard deviation. Figure 5-11 shows
the effect of sample size on the 95-percent confidence intervals for a hypothetical site whose
true annual geometric mean is equivalent to 75 pg/m , the current annual standard. From this
example, we can only conclude that with 95 percent probability the annual geometric mean is
3
between 67 and 85 |jg/m , if the mean for that year was determined from a sample size of 61.
Increasing the sampling frequency to 1 day in 2 reduces the level of uncertainty so that the
annual mean is known to lie between 70 and 81 ug/m . Thus, by increasing the sampling fre-
quency by a factor of three, the width of the 95-percent confidence interval for the true
annual geometric mean has been cut by a factor of 1.7 (square root of 3).
A critical factor in evaluating compliance with once-a-year standards is the effect of
sampling frequency. Figure 5-10 shows that in 1978 the majority of valid sites (80 percent)
had fewer than 60 sampling days. The sites with more frequent sampling had a greater chance
of sampling the higher concentrations (Figure 5-11). Assuming that there are a number of
days on which the observations are above the standards, the probability of selecting 2 or more
days on which standards are exceeded is a function of sampling frequency. If there are 10
days above the standards, there is only a slightly better than 50-percent chance of actually
monitoring on 2 of those days given a sampling frequency of 61 out of 365 days. When the sam-
pling frequency is doubled to 122 sampling days, the probability of capturing 2 days out of 10
that exceed the standards increases to 80 percent. In actuality, samples are not taken ran-
domly; they are taken systematically, usually at a rate of once every 6 days. The probability
of capturing the highest period is further complicated in that the log normal distribution of
TSP concentrations does not apply uniformly to all sites.
An additional complication in determining compliance or trends occurs when the meteoro-
logical regimes affecting the TSP concentrations are considered. Attainment of standards may
depend on the number of "clean" sampling days versus the number of "dirty" sampling days.
Watson (1979) exemplifies this problem with Portland, OR, TSP data. The annual geometric mean
TSP data show a decreasing trend from 1973 through 1975, with a significant increase in 1976.
If these data are reexamined and weighted by the meteorological regimes actually sampled in
each year, the conclusions are changed. The stratified mean TSP values show a large drop in
concentrations occurring between 1974 and 1975; the levels are constant for the years before
and after. Since these means are determined from a sample set varying from 49 samples in 1974
to 79 samples in 1976, a statistical test is required to determine whether the means of any
year are significantly different from those of any other year. The 95-percent confidence
5-25
-------
95
90
OJ
* 85
Z
o
a 80
z
UJ
u
1 75
a.
70
65
D
ABOVE THE STANDARD
AREA OF<
UNCERTAINTY
BELOW THE STANDARD
61 91 122 183 365
NUMBER OF SAMPLING DAYS PER YEAR
Figure 5-11. The 95 percent confidence intervals about an annual
mean TSP concentration of 75 fig/m? is shown for various sampling
frequencies {assume the geometric standard deviation equals 1.6).
Source: Curran and Hunt (1975).
5-26
-------
Intervals for all stratified means do overlap. Watson concludes, then, that there is a rea-
sonable chance that the true means do not really vary from year to year.
5.3.1.2 Monitor Location—The choice of sampling location can obviously affect the concen-
trations measured. Remotely located monitors typically record low concentrations; urban moni-
tors characteristically record higher concentrations. The positioning of a monitor at a
chosen location can also affect measured concentrations. For example, at a specific location,
the height of the monitor above the ground influences sample concentrations. If the monitor
is elevated above surface sources, lower concentrations of coarser particles might be
measured. Some studies clearly indicate that TSP concentrations decrease with increasing
monitor elevation (Record and Bradway, 1978; Record et al., 1979; Lioy et al., 1980a; Pace et
a!., 1977), with distance from a roadway, and with distance from other nearby sources.
The inferences drawn about air quality levels, trends, and population exposures from the
TSP data presented in this chapter are made in full knowledge of the following limitations of
TSP monitoring:
1. Sampling sites are not standardized.
2. Frequency of sampling is quite varied.
3. The majority of sites reporting have fewer than 60 sampling days per year.
4. The frequency of sampling is not weighted with respect to meteorological conditions.
5. No spatial averaging is used in analyzing or reporting data
6. Though the ambient air monitor -is stationary, the population it is intended to
represent is highly mobile and spends some time indoors.
5.3.2 Ambient Air TSP Values
The distribution of 1978 annual arithmetic means for valid TSP monitoring sites is
plotted in Figure 5-12. Half of all the nation's sites had annual arithmetic mean values less
3 3
than 60 ug/m . Annual mean values range from 9 to 256 ug/m . Only 14 valid sites had annual
3
mean concentrations equal to or less than 16 ug/m . These lower values were recorded in
remote monitoring sites. Two background sites, Glacier National Park, MT, and Acadia
3
National Park, ME, had 1977 "annual averages of 11 and 21 ug/m , respectively. At the other
3
end of the distribution, 25 percent of sites had annual means greater than 76 ug/m , and 10
percent were greater than 96 ug/m . Higher annual concentrations were found in many populated
3
and industrialized areas. About 30 sites reported annual averages in 1978 above 150 ug/m .
Topping the list were four central-city sites in commercial, residential, or industrial
3
settings. A Phoenix, AZ, site (0136) had the highest annual mean of 256 ug/m , followed by a
3 3
site in Calexico, CA, at 201 ng/m and an industrial site in Granite City, IL, at 197 H9/m -
These extremely high annual TSP concentrations were found in commercial and industrial
locations. Of the 30 highest sites, 15 were industrial. Many of the higher concentrations
(19 of 30) were found at central-city locations. Only four were classified as rural sites,
most of which were also residential areas. It is also likely that arid climates and dusty
conditions in the vicinity of some monitoring sites might have led to suspension of surface
material. However, it is impossible to ascertain the contribution of fugitive or resuspended
5-27
-------
1000
900
800
700
600
500
400
300
"H 200
g
g
EC
21 100
g 90
1 80
0 70
2 60
P 50
40
30
20
IT I I I I I I
O O 90TH PERCENT! LE
D Q MEAN
T I I I T
1 T
I I
I I I I i i I
J I
J I
I I
0.01
0.1
0.5 1 2 5 10 20 30 40 50 60 70 SO 90 95 98 99 99.8 99.9 99.99
% OF SITES REPORTING ANNUAL MEAN
Figure 5-12, Distribution of mean and 90th percentile TSP concentrations is shown for valid 1978 sites.
Source: SAROAD.
5-28
-------
20
40
60
80
100 120 140 160 180 200 220 240 260 280
TSP CONCENTRATION,,
Figure 5-13. Histogram of number of sites against concentration shows that
over one-third of the sites had annual mean concentrations between 40 and
60 uglm3 in 1978.
Source: SAROAD.
5-29
-------
dust to the concentrations measured at these 30 sites without more detailed analysis. The
histogram of sites against concentrations (Figure 5-13) shows that • over a third of all
monitoring sites had annual mean values between 40 and 60 ug/m . Slightly less than another
3
third had annual averages between 60 and 80 ug/m .
The distribution of 90th percentile values is also plotted in Figure 5-12. Half of all
Q
valid monitors had a 90th percentile value in excess of 97 ug/m . For 10 percent of the moni-
tors, 10 percent of the observations exceeded 160 (jg/m , For one monitor, 10 percent of the
3
observations exceeded 600 ug/m .
Daily, or 24-hour, TSP concentrations have a wide range. In remote areas such as the
Pacific Islands, daily values may be as low as a few micrograms per cubic meter. Over the
2
continental United States, concentrations from 5 to 20 pg/m are routinely reported. In other
locations, daily TSP values can exceed 10 times the levels found in remote areas, on occasion
3 3
exceeding 3000 ug/m . Values exceeding 1000 ug/m are observed in remote arid regions as well
as in populated urban areas. Daily TSP levels approaching these higher values, 500 to 1500
o
fjg/m , are frequently associated with adverse meteorological conditions: low-level inversion,
stagnation, or high winds resuspending surface material.
o
Thirty valid TSP monitoring sites report highest 24-hour values above 600 |jg/m . Only a
few of these sites are in the top 30 in annual average. In cities like Topeka, KS, and Libby,
HT, which are not densely populated or industrialized, these high concentrations may result
from chance occurrences, such as fires or dust storms. In other cities like El Paso, TX, and
Granite City, IL, which are industrialized, the maximum concentrations are more likely to be
related to persistent sources of pollution.
5.3.3 TSPConcentrations by Site andRegion
Airborne particles measured by the hi-vol TSP method arise from many sources, and these
are described in Chapter 4. Important source categories include:
1. Fugitive dust emissions stirred up by mechanical action or the wind as in dust
storms.
2. Large point sources such as smokestacks.
3. Chemical reactions in the atmosphere that transform gaseous substances, such as
sulfur and nitrogen oxides, ammonia, and volatile organic species, to PM substances.
4. Low-level widely dispersed combustion sources such as automobiles and trucks,
residential furnaces, fireplaces, and wood stoves.
5. Occasional or sporadic sources that' can in some instances cause very high TSP
values. Examples are agricultural tilling and burning, wildfires in forests, grass-
lands or even cities, and roadgrading.
The relative contributions of these sources can vary widely and are influenced by location of
TSP monitors relative to potential sources and meteorological conditions. Hence, it is not
surprising to see a wide variation in daily measurements at a single site, variation over an
urban area, or variation among regions of the country. This section will, by illustration,
examine these differences in TSP concentrations by location.
5-30
-------
5.3.3.1 TSP by Site Classifications—Evidence that fugitive dust contributes significantly to
both western arid sites and many urban sites is quite extensive. Discussion of the general
influence of fugitive dust may be found in section 5,6.4, below.
The general differences in annual TSP concentration among locations are seen in Figure
5-14, These differences reflect the character of the neighborhoods where the monitors are
located. This figure summarizes the mean concentrations from 154 sites in 14 cities. Resi-
3
dential neighborhoods in and near cities have TSP levels between 50 and 70 |jg/m . Commercial
3
sites have a wider range of concentrations (60 to 110 yg/m ). Industrial locations generally
range between 80 to 150 ug/m .
The 1978 data base also has been analyzed on the basis of two additional sets of descrip-
tors. One description scheme classifies monitoring sites by their purpose: population expo-
sures, source receptors, or background sites. The other scheme identifies sites by the amount
of development: central-city, suburban, rural, or remote. These classifications are not
mutually exclusive.
When sites are grouped by descriptors, a distinct weighting becomes apparent. Almost 80
percent of the sites are population oriented; approximately 15 percent are source related, and
less than 6 percent are background monitors. The distribution by development also reflects
its population emphasis. Of the total monitors, 83 percent are at either central-city or
suburban sites, 15 percent are at rural sites, and 2 percent are at remote sites. In these
data, 38 percent of the background sites had median values less than or equal to 21 ug/m ,
whereas only 4.4 percent of all sites had these low values. Only 30 percent of the background
3 3
sites had median values above 44 ug/m (none had values above 97 ug/m ), whereas 75.5 percent
3
of source sites had median values above 44 M9/m • The pattern is consistent for the distri-
bution of the 90th percentiles cross-tabulated by site purpose. Cross-tabulations of site
median values and site 90th percentile values with the development-related site descriptors is
further confirmation of the influence of location on measured TSP concentrations. Rural and
remote sites have lower median values and lower 90th-percentile values. The suburban sites
reflect the overall national distribution. The central-city category has proportionately more
sites in the higher concentration ranges.
5.3.3.2 Intracity Comparisons—Because of the strong neighborhood influence on TSP concen-
trations, it is not unusual to find considerable variation in peak and mean concentrations
across a community. Examination of intracity differences illustrates the difficulty in esti-
mating population exposures to TSP.
Data on the nine cities having the highest annual TSP concentrations in 1977 are given in
Table 5-5. Only sites having enough observations per quarter to report an annual mean are
used. Although TSP concentrations in these cities were generally high, in 1977 the less
developed or less industrialized areas in each city had annual geometric mean concentrations
3
below 75 ug/m , (currently the annual primary NAAQS), with the exception of Granite City, IL.
5-31
-------
150
en
a.
H
100
a:
z
Ui
u
o
u
Q.
CO
50
RESIDENTIAL
COMMERCIAL
INDUSTRIAL
Figure 5-14. Histogram of mean TSP levels by neighborhood shows
lowest levels in residential areas, higher levels in commercial areas,
and highest levels in industrial areas.
Source: Office of Air Quality Planning and Standards (1976).
5-32
-------
TABLE 5-5. RANGE OF ANNUAL GEOMETRIC MEAN CONCENTRATIONS IN
AREAS WITH HIGH TSP CONCENTRATIONS IN 1977
Number of sites Annual
Number with annual average range,
City of sites >75 pg/m3 yg/m3
Tucson, AZ
Pocatello, ID
Chicago, IL
Granite City, IL
Taos County, NM
Middletown, OH
Cleveland, OH
Youngstown, OH
El Paso, TX
7
4
25
8
1
3
23
5
14
3
3
12
8
—
2
13
4
10
67-156
65-218
50-170
85-185
168
64-192
48-152
66-172
60-158
Range of
maximum 24- hr
value, jjg/m3
178-591
344-1371
152-1106
227-485
577
157-707
128-705
163-602
205-691
il mean concentration
:leanest sites within
Regional Differences
for
the
in
the dirtiest sites
same city.
can be two to four
Background Concentrations — It has been
times higher tha
demonstrated th
concentrations can vary across an urban area and among cities with different sources and
meteorology. In addition, there may be regional differences in the natural or transported
fraction of TSP concentrations. Figure 5-15 shows the contribution of these sources to
nonurban levels. It was assumed that the global and local contributions in the average would
be similar. The greatest difference among regions is the contribution from "continental" and
transported emissions. These two categories of particles contribute in such a way that
nonurban sites in the West typically report annual geometric means of 15 ug/m ; in the
Midwest, 25 ug/m ; and in the East, 35 pg/m . Except for the Acadia National Park site (18
ug/m ) and Millinocket (23 pg/m ), all sites in Maine had 1977 annual geometric means above 30
|jg/m . Nonurban sites in Wisconsin had mean TSP levels less than 25 pg/m . Nonurban sites in
Montana had levels less than 20 ug/m in 1977; the individual means were Big Horn County, 17
3 3 3
ug/m ; Custer County, 15 ug/m ; and Powder River County, 14 ug/m . Such an analysis does not
exist for coastal areas in the far West, where densely populated areas cluster thickly and the
city-to-city transport component is large.
5.3.3.4 PeakTSP Concentrations—To indicate the severity of TSP ambient exposures, the 90th
percentile concentration of the 24-hour measurements was examined for all 4008 sites in the
5-33
-------
50
40
30
<
oc
UJ
Z 20
O
O
10
LOCAL
TRANSPORTED PRIMARY
TRANSPORTED SECONDARY
CONTINENTAL
GLOBAL
MIDWEST
EAST
Figure 5-1S. Ave*fiiQe estimated contributions to nonurban
levels in the East, Midwest, and West are most variable for
transported secondary and continental sources.
Source: Lynn et al'. (1976).
5-34
-------
1977 NADB. The concentrations of TSP and other air pollutants had been widely reported to be
log normally distributed (Larsen, 1971). This statistical relationship, however, appeared
inappropriate at the high and low ends of the distribution (Mage and Ott, 1978). Because the
extreme values at the high end are subject to wide scatter, the 95th or 99th percentile was
found to be less representative of the severity of high TSP levels. The 90th percentile was
therefore chosen as being a more stable indicator, the TSP value which is exceeded 36 days of
the year.
Figure 5-16 shows the number of AQCR's where at least one of the monitors had its 90th
percentile TSP concentration within the various categories. Of the country's 247 AQCR's, only
q
21 reported 90th percentile values for all data collected below 100 |jg/m . There were 142
2
AQCR's that had 90th-percentile values at one or more sites, between 100 to 200 ug/m . These
data suggest that most of the U.S. population might experience ambient TSP concentrations
exceeding 100 (jg/m for at least 36 days of the year.
5.3.4 Temporal Patterns in TSP Concentrations
5.3.4.1 Diurnal Patterns--TSP concentrations vary with local emission strength, meteorologi-
cal conditions, and changes in contributions from background particles. The particle
emissions loadings to the atmosphere generally increase during the day and decrease at night.
The atmosphere undergoes greater vertical mixing during the day, and windspeeds near the sur-
face increase as a result. Greater vertical mixing coupled with increased source emissions
cause particle mass loadings to increase. ' At night, decreased mixing and the resultant
decreased surface winds permit settling of larger particles. With increased atmospheric sta-
bility, local elevated sources are not as likely to mix to the ground. Unfortunately, diurnal
cycles are not well established because the standard sampling procedure for TSP measurements
yields a 24-hour sample, midnight to' midnight.
Trijonis et al. (1980) found no clear diurnal trend in sub-15 urn particle mass in 6-hour
samples from the St. Louis Regional Air Pollution Study (RAPS). Stevens et al. (1980) have
found slightly higher daytime levels of sub-15 urn particle mass in a remote site in the Smoky
Mountains; however, Pierson et al. (1980) noted no significant diurnal pattern in a forested
region in Pennsylvania.
It is likely that day-night patterns are somewhat obscured by averaging times. Heisler
et al. (1980) found peaks in light scattering and in particle mass corresponding to rush hours
in Denver in the winter of 1978; minimum values were found in mid afternoon corresponding with
mixing height maxima.
5.3.4.2 Weekly Patterns—Since human activity follows distinct weekly cycles, it is reason-
able that anthropogenic sources of particles will also have weekly patterns. The most
distinct weekly patterns are weekdays versus weekends. Trijonis et al. (1980) examined the
St. Louis TSP and dichotomous data base for weekend-weekday differences in particle loadings.
They concluded that there was only a slight (-9 percent) difference between weekend TSP values
and weekday values for the average of five urban sites in St. Louis. For three suburban
sites, the difference was -5 percent and for two rural sites the difference was -12 percent.
5-35
-------
150
140
130
120
£ ii°
o
< 100
It
0 90
cc
LU
1 BO
z 70
60
SO
40
30
20
10
0
21
142
32 48
<100
101-200
201-260
90TH PERCENT1LE TSP CONCENTRATIONS, jug/m3
Figure 5-16. Severity of TSP peak exposures is shown on the
basis of the 90th percentile concentration. Four ACQR's did
not report.
Source: Environmental Protection Agency (1978).
5-36
-------
The urban difference was dominated by readings from one monitor in a heavily industrial and
commercial area.
5.3.4.3 Seasonal Patterns—Analyzing temporal patterns can frequently provide insight into
the nature and source of PM. Meteorological parameters affect the generation and dispersion
of particles. These parameters include (among others) degree-days, mixing height, ventilation
factors, frequency of calms and stagnations, and precipitation. There are also seasonal
patterns in some source emissions.
Because meteorological parameters are so important, it is likely that seasonal patterns
in one area cannot be generalized to other areas. Trijonis et al. (1980) found a modest sea-
sonal pattern of higher TSP concentrations in the summer months in St. Louis. In support of
this observation, Figure 5-17 compares the TSP monthly mean values and the data from dichoto-
mous samplers. The really distinct seasonal pattern was in the fine aerosol fraction. Summer
fine-particle concentrations are twice as great as winter values. As discussed later, aerosol
sulfates make up most of the fine-fraction particles and show a distinct seasonal pattern.
To illustrate the geographic specificity of these seasonal cycles, 3 years of monthly
averaged TSP data are presented in Figure 5-18. The data came from Steubenville, OH, an
industrialized site in the upper Ohio River Valley. Each monthly mean was derived from 20 or
more sampling days. The TSP concentrations were considerably higher than the St.- Louis
values. The months with the highest TSP in Steubenville were March, April, and May in 1977;
July, August, September, and November in 1978; and February and June in 1979. No clear
seasonal pattern emerges from this 3-year period.
5.3.4.4 Yearly Trends—In 1957, a National Air Sampling Network (NASN) began to operate rou-
tinely on a national basis. The U.S. Public Health Service, with cooperation from "State
health departments, operated 231 urban and 37 nonurban stations. Some of these stations
operated every other year, so in a given year there were 143 urban and 37 nonurban TSP hi-vol
monitoring sites in operation. These sites collected one 24-hour sample every other week for
a total of 26 samples per year. In 1977, over 4000 stations, most of them in State and local
networks, reported TSP values to the National Aerometric Data Bank (NADB) of the U.S. EPA.
Not only has the number of sites greatly increased, but the sampling frequency has been 1 day
in 6 since 1971. In some cities, TSP monitoring data has been recorded for more than 20
years. Although the sites may not be in exactly the same locations for every city, general
trends in TSP concentrations can be obtained. Figure 5-19 plots the annual geometric mean TSP
concentrations for three groups of cities. In 1958, the five cities classified as industrial
had annual mean TSP concentrations between 140 and 170 (jg/m . By 1974 the annual mean concen-
trations had dropped to between 80 and 110 jjg/m . Similarly, three of four cities classified
as moderately industrialized showed substantial decreases. Only the Denver station recorded
an increase, and that is only for a single year. The four cities classified as lightly
industrialized showed less overall change.
5-37
-------
TSP (HI VOL)
••• DURBAN
SUBURBAN
RURAL
DICHOT COARSE
N (2515pm)
URBAN
4J SUBURBAN
RURAL
JAN FEB ,,MAR U-»APR,, MAY JUN JUL AUG SEPT OCT MOV DEC
MONTH
Figure 5-17. Seasonal variations in urban, suburban, and rural areas for four
size ranges of particles. The data were obtained from a relatively small num-
ber of monitoring sites.
Source: After Trijonis et al. (1980).
5-38
-------
en
a
o
§
cc
o
o
o
a.
Ui
>
o
200
150
100
SO
I I I I I I I I I I I I I I I I i I I I I I [ I IT I I I I I I I I I I
I I I i i I I i I I I I I f I i i I I I i i i i i I i i i i I i i i i i
J F M A M J JASONDJFMAMJ JASONDJ FMAMJ J A S O N D
1977 1978 1979
YEAR AND MONTH
Figure 5-18. Monthly mean TSP concentrations are shown for the Northern Ohio Valley Air .Monitoring
Headquarters, Steubenville, OH. No clear seasonal pattern is apparent.
Source: Spengler (1980).
5-39
-------
to
220
200
180
160
140
M 120
100
80
60
40
20
0
I T
O BALTIMORE
D CINCINNATI
A CLEVELAND
O PHILADELPHIA
|| ST LOUIS
HEAVILY INDUSTRIALIZED CITIES
I I I I I I I
I I
I I I I
j I
1957
1960
1970
1974
.6
"o>
3,
CHATTANOOGA
A DENVER
PROVIDENCE
• SEATTLE
I I I I I I I I I I I I
1957
1960
19G5
1970
1974
YEAR
160
140
120
100
80
60
40
20
0
I I
1 I I I I I I I I
T 1 f
O MIAMI
Q OKLAHOMA CITY
A SAN FRANCISCO
• WASHINGTON, D C
LIGHTLY INDUSTRIALIZED CITIES
I t I I I I I I j I 1 I J I I I I L
1967
1B60
1U65
YEAR
1970
1974
Figura 6-19. Annual geometric mean TSP trends are shown for selected NASN sites
S
-------
Examination of the expanded TSP data set from hi-vol samplers shows that for 2707 sites
o
the composite median concentration has remained about 60 M9/m between 1972 and 1977. The
geometric mean over this period has decreased by approximately 8 percent. The decrease in the
90th percentile of the annual average concentrations is most pronounced over this period (see
Figure 5-20). Lowering the TSP concentrations in locations with very high levels has been a
target of State air pollution control strategies. In addition, relocating sources to rural
regions, building new sources with taller stacks, converting to cleaner fuels, paving of
streets and roads, and restricting open burning have decreased the number of locations
3
experiencing annual concentrations of over 100 ug/m •
For the period 1970 to 1977, EPA reported an almost 50 percent reduction in particulate
emissions. Most of this reduction occurred in the early 1970's as State air pollution control
programs started many major emitters on compliance schedules. The rather modest composite
overall reduction of 8 percent in annual TSP levels may be explained by the fact that direct
emissions from stationary sources contribute only a fraction of the TSP loadings in the atmos-
phere.
Another perspective on regional differences is gained from observations of the 1978 data.
Table 5-6 provides a statistical summary for the 50th and the 90th percentiles for valid moni-
tors. Region IX ranked highest for the mean and maximum 50th and 90th percentiles, followed
by Region VII, Region VI, and Region V. Regions I and II had consistently lower values.
The column presenting the standard deviations of the mean values for the 50th and 90th
percentiles is also of interest. Smaller standard deviations suggest more uniformity in re-
ported concentrations among monitoring sites. Because Regions I, II, and IV had less variance
among sites than other regions, it could be interpreted that these regions had either more
uniform distribution of pollution levels or more uniform placement of monitoring sites. The
larger standard deviations in other regions, particularly in the West, probably mean that
there is greater variation in pollution levels.
There are distinct regional differences in the trends of TSP concentrations. The distri-
bution of site means and the actual rate of change in TSP levels differed among regions of the
country. These trends are shown in Figures 5-21 and 5-22; the differences between years and
even over the entire period were not tested for significance. Therefore, intraregion and in-
terregion comparisons are presented qualitatively.
In the Eastern United States, in EPA Regions I and II, the composite average across sites
3 3
decreased from 60 ug/m to approximately 55 ug/m . The range of concentrations was much
narrower in Regions I and II than it was in the more industrialized Regions III, IV, and V.
In Region III, the composite average decreased from 78 to 60 ug/m , with the 90th per-
centile in the distribution of annual mean concentrations decreasing from slightly over 100
33
ug/m to about 90 jjg/m . However, it has remained relatively stable or has even increased
3
slightly since 1975. In Region V, the composite average decreased from 80 to 70 ug/m , and
3
the 90th percentile decreased from 100 to 85 ug/m , reflecting the effectiveness of point
source control.
5-41
-------
160
140
M
"S 120
§ 10°
< 80
w 60
g 40
a.
{2 20
n
I I
9 9 (
i i
I I
*i ?
X X
6 0
i I •
I
1
6
i
I
?-
6 -
i
1972 1973 1974 1975 1976 1977
YEAR
I
O
A
A.
90TH PERCENTILE
75TH PERCENTILE
COMPOSITE AVERAGE
MEDIAN
25TH PERCENTILE
10TH PERCENTILE
Figure 5-20. (Top) Nationwide trends in annual mean total suspended
particulate concentrations from 1972 to 1977 are shown for 2707
sampling sites. (Bottom) Conventions for box plots.
Source: U.S. Environmental Protection Agency (1978).
5-42
-------
TABLE 5-6. REGIONAL SUMMARIES OF TSP VALUES FROM VALID MONITORS
en
i
Number
of sites
128
315
300
534
781
294
136
152
89
113
15
9
10
2882
Median
Region
I
II
III
IV
V
VI
VII
VIII
IX
X
Alaska
Hawaii
Puerto.Rico
Total
Minimum
14.0
10.0
27.0
22.0
13.0
12.0
34.0
7.0
16.0
11.0
11.0
25.0
32.0
7.0
Mean
49.2
43.2
59.8
55.3
64.1
65.0
69.7
54.8
76.5
60.3
48.1
39.7
54,3
58.9
Maximum
100.0
114.0
171.0
137.0
189.0
166.0
154.0
164.0
226.0
129.0
94.0
70.0
85.0
226.0
soa
14.6
15.3
20.3
17.2
22.0
20.9
20.2
32.8
38.0
24.5
22.7
13.7
13.9
22.8
Minimum
32.0
29.0
52.0
41.0
26.0
37.0
58.0
18.0
37.0
23.0
35.0
40.0
66.0
18.0
90th percent! le
Mean
87.3
85,0
105.7
93.5
122.4
110.4
123.6
107.8
133.4
123.6
137.1
63.6
90.7
107.9
Maximum
181.0
286.0
296.0
256.0
383.0
436.0
359.0
412.0
381.1
381.1
250.0
99.0
134.0
436.0
SDa
27.7
30.6
42.1
30.9
42.8
45.8
44.2
64.0
66.0
52.5
68.7
18.9
18.9
44.9
.SO, Standard deviation of the median and 90th percentile values.
Including American Samoa and Guam.
Source: SAROAD.
-------
U S EPA AIR QUALITY CON1ROl REGIONS, EASTERN STATES
160
148
120
lioo
S ID
Si 60
*" 4D
20
0
REGION 1
I I I I I I
1972 1973 1974 1975 1976 1977
160
140
120
"e 100
REGION 2
I I I I I I
1372 1973 '1974 1975 1976 1977
REGION 3
i ii i I
197Z 1973 1974 197S 1976 1977
REGION a _
t. i I
REGION S
I Ii r I I
1S72 1973 1974 197S 1976 1i77
1972 1173 1974 1976 1978 1977
Fifluro 5-21. Regional trends of annual mean total suspended participate concentrations, 1972-1977, Eastern states.
Source: U.S. Environmental Protection Agency (1978).
5'44
-------
U S EPA AIR ttUALITV CONTROL REGIONS, WESTERN STATES
160
140
120
1-100
"I 80
85'BO
*- 40
20
0
REGIONS
I I 1
1972 1973 1974 1975 1976 1977
REGION? -
I
I I
I
REGIONS -
II I I
1872 1973 1974 1975 1976 1977
1972 1973 1974 1975 1978 1977
160
140
120
I 100
2 60
P 40
20
0
REGION 9
I i i i I
REGION 10 -
1972 1973 1974 1975 197S 1977
1972 1973 1974 1975 1976 1977
YEAR
Figure 5-22. Regional trends of annual mean total suspended paniculate concentrations, 1972-1977, Western states.
Source: U.S. Environmental Protection Agency (1978).
5-45
-------
The Western States make up Regions VI through X. In Region VI, the composite average re-
3 '
mained at approximately 75 pg/m , and the 90th percentile increased slightly from the 1973
2
reading to about 100 ug/m . Industrial, utility, and related growth in this area, as well as
in Region IV, was probably responsible for keeping TSP concentrations from decreasing. In
Region VII, the composite average was almost constant, varying only slightly between 80 and 75
3 3
pg/m . The 90th percentile varied between 110 and 100 pg/m . Region VIII showed wide distri-
3
bution in the concentrations. The 10th percentile, at about 20 jjg/m , was the lowest among
o
all regions. The 90th percentile, approximately 100 jjg/m , was roughly equal to the highest
concentrations in most regions. The composite average varied over the 6-year record but re-
o
mained essentially the same, approximately 80 |jg/m , in 1977 as it was in 1972. The back-
ground air quality in the upper States of this region (Montana, North and South Dakota, and
2
Wyoming) was among the best in the country. Thus, some of the low levels (20 pg/m and below)
represented some of the lowest background concentrations measured in the United States. The
high composite average and high 90th percentile levels reflected the impact of locating
monitors near industrial sources such as smelters and the fugitive dust emissions from wind-
3 3
blown soils. Region IX had a composite average of 100 pg/m , which was up from 90 pg/m in
3
the early 1970's. The 90th percentile was also high, at 120 ug/m . Thus, Region IX had some
of the highest levels in the country. Region X had a composite average of approximately 70
3 3
ug/rn , up slightly from a low of 60 |jg/m in 1975. The 90th percentile varied between 90 and
100 ug/m3.
The overall trend in improvement from 1972 through 1975 was followed by a reversal in
some regions in 1976. Despite this short-term reversal in 1976, 60 percent of the sites
showed long-term improvement from 1972 to 1977. For those sites at which TSP concentrations
violated the current annual standard, 77 percent showed long-term improvements. Approximately
25 percent of these sites reported their lowest annual values in 1977. Possibly, the short-
term reversal in 1976 was due to unusually dry weather, resulting in windblown dust that may
have contributed to elevated TSP levels throughout the Central Plains, Far West, Southwest,
and Southeast.
5.4 SIZE OF ATMOSPHERIC PARTICLES
5.4.1 Introduction
In Chapter 2, the general features of size distributions of atmospheric particles were
discussed in some detail. In recapitulation, atmospheric particles tend to be more prevalent
in certain particle size bands or modes than in others. Particles that have grown from the
gas phase, either because of condensation or atmospheric transformation or combustion, occur
initially as very fine nuclei 0.05 (jm or smaller. These tend to grow rapidly to accumulation
mode particles around 0.5 pm in size, which are relatively stable in the air. Because of
their initially gaseous origin, this range of particle sizes includes inorganic ions such as
O— w. -i.
S0£ , NO-, and NH., combustion-formed carbon, organic aerosols from photochemical conversions,
and a variety of trace elements associated with combustion sources.
5-46
-------
Airborne particles of soil or dust mostly result from entrainment by the motion of the
air or from other mechanical action, and most of the mass of these materials is in particles
larger than 5 urn. While the relative amounts of these two particle types are highly variable
in both time and place, there is almost always a clearly observable minimum or gap in atmos-
pheric mass distribution occurring in the particle size range of roughly 1 to 3 micrometers.
In this range, there are only minor percentages of the total mass, and this material appears
to be overlap from the two major categories. The larger particles frequently contain clay
minerals, bits of local rocks, limestone aggregate from roadways, fly ash from power plants,
and other substances ranging from insect parts, pollen, and sawdust to liquid globules of
acidic smut blown from boiler tubes (Draftz and Severin, 1980). The elemental analysis of
these larger particles is usually dominated by silicon, aluminum, magnesium, calcium, and.
iron, all components of soil and of fly ash (see Chapter 4).
In the last several years, a general perception has been growing that not all PM is
equally damaging to the environment (see Chapters 8 & li). For this reason, information on
the mass of PM in various size categories has been gradually accumulating and is here sum-
marized. Furthermore, a national network of sampling stations equipped with size-selective
sampling devices is currently being set up. While some tabulated data are available from this
network and are summarized here, no detailed interpretative analysis has been published yet,
nor are any chemical analytical data available. The analysis of monitoring results from the
national inhalable particle (IP) network must wait for subsequent revisions of this document.
In the following discussion, the major chemical components of atmospheric aerosols are
organized by the size mode or particle category in which they are most frequently observed.
2-
This is not to say that SCL ion, for example, is exclusively a component of fine particles,
Sometimes, e.g. in the vicinity of a cement manufacturing facility, there can be substantial
Own
amounts of coarse CaSO, . However, the relationship between size and composition of particles
is so general that more reason is served by this organization than by any other.
Since the finer particles seem to have less diversity and since measurements of the major
anion components of this fraction have been made for a long time, this group is discussed
first. The more complicated coarse fraction has not been very well defined and, indeed, may
not be definable by chemical analysis alone. There is considerable interest in this size
range currently, though, and studies of these materials are cited in Section 5.6.
5.4.2 Size Distribution of Particle Mass
Evidence from chemical analysis and physical theory (Chapter 2) strongly suggests that
atmospheric aerosols commonly occur in two distinct modes. The fine or accumulation mode is
attributed to growth of particles from the gas phase and subsequent agglomeration. The coarse
mode is made up of mechanically abraded or ground particles. Therefore, it is not surprising
to find atmospheric PM distributed among fine and coarse particles with a rather clear inter-
val of demarcation in between.
5-47
-------
Unfortunately, gravimetric data by size fraction were sparse until comparatively recent-
ly. Furthermore, most were obtained with impactors, which are influenced by particle "bounce"
(see Chapter 3). Several works suggested the existence of a distinct minimum in the mass-size
distribution in the 1968 to 1970 time period. Lee et al. (1968) observed only 14-percent of
the aerosol mass between 2 and 4 urn in three samples from Fairfax, OH. Lundgren (1970) found
only 10-percent of aerosol mass in this range in 10 Riverside, CA, aerosols samples ranging
o
from 47 to 144 vg/m , O'Donnell et al, (1970) found only 10-percent in the 2 to 4 pi range in
one Pittsburgh, PA, sample. Lee and Goranson (1972) and Lee et al. (197-2) reported many
impactor size distributions for six cities obtained in 1970, all indicating 12- to 15-percent
of aerosol mass between 2 to 5 pm. However, many of these data were clouded by bounce and
entry losses and probably were biased toward low coarse-mode distributions.
More recently, evidence from electrostatic sizing equipment has confirmed this general
trend. Figures 5-23 through 5-26 show the distribution of particle volume by size. These
data differ from mass distributions because particle density (mass/volume) was not measured as
a function of size. Figures 5-23 and 5-24 present distributions in and around St. Louis, MO,
for a variety of conditions. Generally these distributions show distinct minimum values in
the vicinity of 1 to 2 pm.
However, the combined influence of nearby sources and aerosol aging can produce major
shifts in volume and, presumably, mass distribution. For example, Figure 5-24 shows a third,
very fine "nuclei" mode of particles centering around 0.05 urn. This mode can be attributed to
the presence of nearby automotive traffic. Also shown in Figure 5-24 is the rather narrow
size distribution from a coal-fired power plant adding to the fine aerosol burden. However,
the mass-size distribution from large sources can vary dramatically among sources depending on
the type and efficiency of control equipment (see Chapter 4).
There can be major shifts in the relative proportions of fine and coarse particle mass as
an aerosol ages (i.e., moves with the wind). Figures 5-25 and 5-26 show dramatic examples of
this phenomenon obtained during the 1972 California Aerosol Characterization Experiment
(ACHEX). In the first case, aged aerosol was transported in the wind to the site from the Los
Angeles area; during this process coarse particles settled out. In the second case, local
winds stirred up dust, shifting the distribution toward larger sizes (Whitby, 1980).
A summary of mass data calculated from electrostatic size distributions for several
environments is shown in Table 5-7. Here, the dramatic variations in coarse and fine particle
fractions found in practice are clear.
More recently, a number of studies have been done with dichotomous samplers designed to
obtain mass samples of the 0 to 2.5 |jm fine fraction and a 2.5 to 15 urn coarse fraction. Re-
searchers from EPA's Environmental Science Research Laboratory have measured coarse and fine
aerosol mass concentrations in several locations: Dzubay et al. (1977) report on 18 days of
5-48
-------
60
SO
40
I
30
20
10
BACKGROUND AEROSOLS NEAR ST. LOUIS, MO.
—— URBAN PLUME INFLUENCED
___ BACKGROUND AVERAGE
_._ AUTO INFLUENCED
__.._ CLEAN BACKGROUND
Figure 5-23. Linear-log plot of Die volume distributions for the four background distributions.
Notice how much the urban plume adds to the accumulation mode of the background.
Source: Cantrell and Whitby (1978).
5-49
-------
70
50
S» 30
20
10
URBAN AEROSOLS
" LABADIE POWER PLANT
PLUMB
___ URBAN
__ . __ URBAN AUTO
INFLUENCED
0
0.003
.01
0.1
1.0
10
50
Figure 5-24, Linear-log plot of the volume distributions for two urban aerosols and a typical
distribution measured in the Labadie coal-fired power plant plume near St. Louis. Size distri-
butions measured above a few hundred meters above the ground generally have a rather small
coarse particle mode.
Source: Cantrell and Whitby (1978).
5-50
-------
o
M~
O
^
<3
0.1
1.0
PARTICLE DIAMETER, i
10
Figure 5-25. Incursion of aged smog from Los Angeles at the Goldstone tracking station in the
Mojave Desert in California. Note the buiidup in the accumulation mode.
Source: Whitby (1980).
-------
"
10 —
1.0
PARTICLE DIAMETER, jrni
10
Figure 5-26, Sudden growth of the coarse particle mode due to local dust sources measured at the
Hunter-Liggett Military Reservation in California. This shove the independence of the accumulation
and coarse particle mode.
Source; Whitby (1980).
5-52
-------
TABLE 5-7. FINE AND COARSE AEROSOL CONCENTRATIONS FROM
SOME URBAN MEASUREMENTS COMPARED TO CLEAN AREAS
Concentration (pg/m )a
Location
St. Louis
Los Angeles
Los Angeles
freeway
Denver
Go Ids tone
Mil ford, Mich.
Pt. Arguello
(seaside)
Condition
Very polluted
Grand average
Wind from
freeway
Grand average
Clean
Very clean
Marine air
Fine
particles
296.0
37.0
77.0
16.6
1.5
1.03
1.1
Coarse
particles
94.0
30.0
59.0
23.2
3.0
0.82
53.0
Calculated from volume distribution using assumed particle
density, pp = 1 gm/cm3.
Source; National Research Council (1979).
summer sampling in St. Louis; Stevens et al. (1979) report on 2 months of summer sampling in
Houston^ TX; Stevens et al. (1980) discuss results of an extensive sampling for a week in the
Great Smoky Mountains. Courtney et al. (1980) discuss the early results from winter sampling
at two locations in Denver, CO. Table 5-8 summarizes their reported findings.
In another short-term study, Lewis and Macias (1980) sampled atmospheric aerosols for 21
3
days in Charleston, WV. The fine-fraction average was 33.4 pg/m , and the coarse fraction
3
average was 27.1 pg/m .
Because of the influence of particle size on adverse effects such as health, visibility,
and soiling (see later effects chapters), EPA is establishing a network of size-selective PM
monitors. Ultimately this grid will include 250 stations to be established over a 3-year
period. During the period from April 1, 1979, to June 30, 1980, 94 stations were established.
A map showing current sampler locations is shown in Figure 5-27 (U.S. Environmental Protection
Agency, 1981). Since dichotomous samplers are used in this network, together with hi-vols, it
is possible to obtain a general conception of the relationship between TSP (0 •*• 60 (j<")> di-
chotomous total or "inhalable" particle mass (0-15 pm), and the fine and coarse fractions de-
fined above.
A total of 1960 dichotomous fine- and coarse-mass measurements and 2675 TSP measurements
are now in this data base; hi-vol measurements with a size-selective inlet are now also being
3
made. In this data base, daily TSP values range from 33.2 |jg/m in Litchfield, CT to 474.4
5-53'
-------
2*—.
(
BOSTON 82^
BU FF ALOB 4 HARTFOB0 • 2 i
ARTFOR0*2^»'
|RKCITY«4*
, J . ™^_ . ]
\.*NFOS OUSTER ft i (
I I MINNEAPOLIS* 2
l^^fSK'-l i -f ^^^^i^^'ssi^
/ «IT LAKE CITY* 2 • . I DESMOINES«2 %'n't'm"J (AKHON"2 »3P|TTS PpHILADELPHI/
/ SALT LAKE CITY w* . J__ , \ STEU^ENVILLE l( 1 jy«SH"iBALTIMORE 2
2^!»NFRANSI5CO / .' I *i \ ' WASHINGTON, D C f 2 ' *
Z»S»NI=BANtISCO / l i. ^ ciNCINNATlkl > „" ~
»z y x X ^
| ST LOUIST. \.^.~* - -
I ^ 1 • «* _
10
NFOS » NATIONAL FOREST OZONE STUDY
'NPS = NATIONAL PARK SERVICE
SMALL NUMBERS REPRESENT NUMBER OF SAMPLING SITES AT EACH LOCATION
Figure 5-27. Inhalable-particle network sites established as of March 19,1980.
Source: U.S. Environmental Protection Agency (1981).
5-54
-------
TABLE 5-8. FINE FRACTION AND COARSE FRACTION DICHOTOMOUS SAMPLING
BY ENVIRONMENTAL SCIENCE RESEARCH LAB, USEPA IN FOUR LOCATIONS
Location Period Days Comments
Concentration
Fi-ne (ug/m3) Coarse ((jg/m3)
St. Louis
Houston
Denver
Smoky Mtns
Summer
Summer
Winter
Winter
Spring
Fall
18
28
19
19
28
7
Urban
Rural
Urban
Urban Site D
Urban Site N
Urban Site D
Urban Site N
Urban
Urban
Remote Day
Remote Night
29
26
52.2
18.1
25.4
23.2
26.4
26.5
16.1
26.4
22.0
22
15
39.8
22.5
23.4
33.0
26.5
27.1
9.8
6.2
4.9
Source: Courtney et al. (1980)
3
to 474.4 (jg/m in Dallas, TX. Maximum dichotomous sampler totals (fine + coarse) ranged from
28.7 ug/m in Pearl City, HI, to 267.5 ug/m in Rubidoux, CA (U.S. Environmental Protection
Agency, 1981).
Because of the limited time period available for analysis (April 1979 to June 1980), it
would be unwise to consider analysis of these data as indicative of geographical or seasonal
trends in particle size. But some additional general factors associated with particle size
can be seen from inspection of the data summary in Table 5-9. (The ratios in the table are
averages of ratios of individual sample pairs and thus will not equal ratios of the average
concentrations given.)
On the average the dichotomous sampler total mass was about 67 percent of the TSP (Pace,
1980), but this ratio varied widely across the country, from about 0.4 to almost 1 in Port-
land, OR and Litchfield, CT (five samples). However, most Portland and all Litchfield samples
were collected during the winter months when rainfall or snow cover could have materially
reduced dust levels.
The fraction of fine and coarse components was even more variable. The coarse mass
fraction of the total sub-15 urn mass ranged from about one-fifth to two-thirds in this
selected set and was even higher for individual days. Particularly striking were the average
values for Dallas and El Paso, TX. At both sites, the sub-15 urn mass was only about half the
TSP mass. However, in Dallas only 27 percent of this was in the 2.5 to 15 urn range while in
El Paso, 64 percent was "coarse."
5-55
-------
TABLE 5-9. RECENT DICHOTOMOUS SAMPLER AND TSP DATA
FROM SELECTED SITES—ARITHMETIC AVERAGES
Location
No. of TSP DTOTAL/TSP Coafse
3 3
observations pg/m (# pairs) MS/m
Fine Coarse,
Mg/m3 DTOTAL
Northeast
Buffalo, NY
Erie Co. , NY
Litchfield, CN
Philadelphia, PA
Southeast
Birmingham, AL
Midwest
Minneapolis, MN
Cincinnati, OH
Southwest
Dallas, TX
El Paso, TX
Far West
Los Angeles, CA
Portland, OR
Pearl City, HA
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
TSP
DTotal
28
40
41
44
5
5
102
109
38
40
44
41
51
48
22
24
29
26
43
50
37
36
27
25
93.7 0.70 (21) 25.2
32.8 0.64 (25) 5.1
18.9 0.86 (1) 6.6
45.1 0.83 (40) 13.3
60.8 0.68 (23) 15.0
50.1 0.61 (26) 15.6
53.6 0.77 (26) 14.4
94.9 0.47 (21) 9.8
86.5 0.51 (7) 46.3
68.4 0.53 (18) 21.3
66.7 0.90 (19) 42.3
33.0 0.43 (11) 7.9
25.9 0.50
16.2 0.24
13.3 0.33
22. 5 0,38
24.4 0,38
16.4 0.46
25.2 0.35
24.1 0.27
11.7 0.64
24.6 0.47
22.0 0.60
8.4 0.47
Source: U.S. Environmental Protection Agency (1981).
5-56
-------
Pace et al, (1981) analyzed some of the general features of these preliminary IP network
data. These authors concluded that regional scale fine particle mass ranged from 6 to 13
3 " 3
in the Western States and from 15 to 23 (J§/m in the eastern United States, comparing
only four sites in each category. They also found strong influence of local sources around
o
monitors in urban sites. "Concentration changes averaging 16 ug/m or 31 percent were found
in sites separated by as little as 1 km distance," according to this study. No strong
seasonal trends were found in various regions of the United States, although there appeared to
be slightly higher fine-particle mass in the summer than in spring in the eastern United
States.
It appears that chemical analysis and several years more data from the IP network could
considerably increase understanding of temporal and spatial distribution of fine- and coarse-
particle fractions.
5.5 FINE PARTICLES IN AIR
Sulfate, ammonium ions, organics, carbon, and combustion-associated metals are widely
recognized to be the major components of fine PM. Few studies of aerosol composition have
attempted material balance, and fewer still have done so with size fractionation.
Nevertheless, a great deal has been learned about the chemical and elemental composition
of airborne particles since the early experiments in the 1950' s by Junge in Germany, Massa-
chusetts, Hawaii, and various sites in Florida (Junge 1952). Junge's observation that sulfate
and ammonium ions appear predominantly in the fine-particle fraction has been confirmed in
independent field observations, both in urban and rural areas (Lewis and Macias, 1980; Dzubay
and Stevens, 1975). In analyzing the St. Louis, MO, dichotomous sampler data by x~ray
fluorescence, Dzubay found 75 percent of the zinc, sulfur, .bromine, arsenic, selenium, and
lead occurred in the fine particles and at least 75 percent of, the silicon, calcium, titanium,
and iron in the coarse fraction (Dzubay, 1980).
In studies of Charleston, WV, particles, Lewis and Macias (1980) reported material
balances of fine and coarse particles accounting for 69 percent and 60 percent of the mass,
respectively. Eighty-five percent of the sulfate and ammonium ions were in the fine particles
where they accounted for 30 and 12,8 percent of the mass, respectively. Carbon, both
elemental and organic, was mainly in the fine aerosol (61 percent) where it accounted for 18.2
percent of the mass.
Stevens et al. (1978) reviewed size-fraction analyses for St. Louis, MO and Charleston,
WV and for four other sites including New York, NY; Portland, OR; Philadelphia, PA; and
Glendora, CA. They conclude that sulfate ion is predominantly a fine component (70 percent)
*
that usually accounts for 40 percent of the mass of that fraction and occasionally up to 50
2-
percent. The S04 must be present as ammonium salts or as I^SO* since metallic sulfate could
be only 10 to 30 percent of the total at maximum (Stevens et al., 1978).
In one site in the Great Smoky Mountains, 89 percent, of the fine PM was identified
(Stevens et al., 1980). Sulfate accounted for 61 percent; ammonium ion, 12 percent; elemental
carbon, 5 percent; and organic carbon, 10 percent. Trace elements, mainly lead, made up the
5-57
-------
balance. In this study, only organic carbon was also a significant component of the coarse
particles.
Studies in a number of sites in California produced similar results. Flocchini et al.
(1978) and Cahill et al, (1977) reported size-fraction distributions for sulfur (presumably
o«
SO, ) in three districts of California. In all areas, sulfur was present almost exclusively
in the sub-3.6 pm fraction. In dry weather, sulfur was found in sub-0.65 (jm fractions, while
under humid conditions it appeared in the 0.65 to 3.6 pm cut.
2- +
Since SO- , NhL, elemental carbon, and organics are the major components of the fine
aerosol, analytical data relating thereto, whether size fractionated or total, will be
discussed together in this section.
5.5.1 Sulfates
The term "atmospheric sulfates" describes a several sulfur compounds, including ammonium
sulfate, NH.HSO., H2S04, letovicite, calcium sulfate, and a variety of metal salts. Most of
the historic data on atmospheric concentrations of sulfates are based on the water-soluble
extract of TSP filters and measurements of the sulfate ion. These samples were subject to
artifact formation on the glass fiber filters used in the early NASN measurements. For a
complete discussion of these issues, see Chapter 3. It is now generally accepted that TSP
2-
S0a measurements taken before 1974 or 1975 using the traditional glass fiber filters may have
3
overestimated sulfates by as much as 2 pg/m or more in areas where ambient S0« concentrations
were high.
Ow. O
Annual average TSP SOA concentrations range from less than 1 pg/m in some States to
3
almost 20 (jg/m in urban industrial areas of the Northeast. For 24-hour average concen-
2_ 3
trations, S0| concentrations range from near zero to more than 80 pg/m .
Sulfate, particularly ammonium sulfate, appears to account for the majority of fine PM in
many locations (Dzubay, 1980; Stevens et al., 1980; Watson, 1979; Flocchini et al., 1978;
Stevens et al., 1978; Pierson et al., 1980). Although some of this material may be emitted
directly from sources, the majority appears to be secondary (i.e., formed by chemical re-
actions in the atmosphere) (Friedlander, 1973; Grosjean and Friedlander, 1975).
2-
5.5.1.1 Spatial and Temporal Variations—The spatial distribution of measured SO, concen-
trations for 1974 is displayed in Figure 5-28. Figure 5-28(a) presents the annual average
3
concentrations. An area having an annual average of more than 15 pg/m extended from the
lower Ohio Valley through the upper Ohio Valley, including major portions of Kentucky, West
3
Virginia, Ohio, and western Pennsylvania. The areas with annual averages exceeding 10 pg/m
included almost all of the United States east of the Mississippi, except for the South
Atlantic States and upper New England.
3
Through the Central Midwest area, values of 4 to 9 pg/m were reported. The Far West
2- 3
States and the Pacific Northwest experienced annual S0| levels below 2 to 3 ug/m , except for
the Los Angeles area. The Los Angeles levels are not shown in this figure, but a 1975 National
Academy of sciences report on air quality and stationary source emission controls indicated
o
that they were between 7 and 13 pg/m (National Academy of Sciences, 1975).
5-58
-------
Figure 5-28. Contour maps of sulfate concentrations for 1974 are
shown for: (a) annual average; (b) winter average; (c) summer
average.
Source: National Research Council (1978a).
5-59
-------
O—
Seasonal variations in SO, concentrations are shown in Figures 5-28(b) and 5-28(c) for
2-
the winter months and the summer months respectively. The area of elevated SO, greatly ex-
pands during the summer months. As demonstrated by several regional studies on atmospheric
2- 2~
SQ* transport, SO. concentrations can be elevated over large geographical regions under
certain meteorological conditions (Eliassen, 1978; Lyons and et al., 1978; Perhac, 1978;
Whelpdale, 1978). This supports the idea of transport and conversion beyond the source
regions of S09 emissions. As these contour maps clearly show, a sizeable portion of the U.S.
2- 3
population is exposed to annual SO. concentrations of more than 10 u/m in the ambient air.
In view of the rising S00 emissions from increased use of coal throughout the United States,
2-
particularly in the South Central States, the area of maximum S0| levels might expand and
shift to the lower Ohio Valley and the Southeast.
2-
In a large-scale study of atmospheric SOt in eastern Canada, Whelpdale (1978) reported
3
mean levels of 10 pg/m over southern Ontario. The mean levels of sulfates dropped to less
than 2.5 (jg/m above the 49th parallel. Figure 5-29 displays these values for the period of
study. During episodic conditions primarily affecting the lower Great Lakes region, 24-hour
3
concentrations were reported as high as 40 to 50 ug/m . Such episodic conditions are assoc-
iated with a high-pressure cell over eastern Canada with southwest flow occurring on the back
side of the high pressure. This synoptic situation favors transport of SQ? and sulfates from
the high SO- source regions of the industrialized northeastern United States.
Recently, new information on the interrelationship of SOp, N0p» 0,, TSP, sulfates, and
nitrates has become available from a large-scale regional study. The Electric Power Research
Institute (EPRI) Sulfate Regional Experiment (SURE) involves intensive monitoring from some 54
rural stations and an aircraft sampling program. The area being studied is 2400 by 1840 kilo-
meters; it extends from Kansas to the Atlantic coast and from mid-Alabama to southeastern
Canada (Hidy et al., 1979) (see figure 5-30).
Mueller et al. (1979) reported on the earlier SURE data collected in 1974 and 1975 and
presented the preliminary results of an intensive field study made during July 1977 through
February 1978. Using the limited historical data base, they indicated that the rural stations
2~
experienced a frequency of occurrence of 24-hour average SO, concentration similar to that
observed around large metropolitan areas such as New York City. As seen in Figure 5-31, 24-
3
hour values greater than 10 ug/m occurred in approximately half the data, and the occurrence
2~ 3
of 24-hour SO. levels exceeding 20 ug/m was about 10 to 12 percent.
3
Based on concentrations of 10 to 20 yg/m as an indicator of elevated exposure, the
average concentrations over the entire SURE network area were estimated by a linear interpo-
lation procedure with a resolution of 80 by 80-km grids. Episodes of elevated sulfates were
2-
extensive; during an episode in early August 1977, the area where SO, levels exceeded 20
3 2
ug/m expanded to more than 500,000 km. Two regional episodes occurred in January and early
?-• 3
February 1977. In August, 39 percent of the SO, values exceeded 10 yg/m ; in January the
3
figure was 30 percent. Five percent of the values exceeded 20 jjg/m . In October, 20 percent
3 3
of the values exceeded 10 pg/m , and less than 1 percent of the values exceeded 20 ug/m .
5-60
-------
5 If I MOUNT FOREST f »
\
MICHIGAN
Figure 5-29. Intensive Suifate Study area in Eastern Canada shows the geometric mean
of the concentration of soluble particulate sulfate during the study period. Units are
micrograms of sulfate per cubic meter.
Source: Whelpdale (1978).
5-61
-------
Figure 5-30. Map of SURE region shows locations of ground
measurement stations.
Source: Hidy et ai. (1379).
5-62
-------
100
50
E
Z
UJ
u
3
(a
20
10
0.01
O HIVERHEAD, NY (champs!
A BRONX, NY (champs)
O ROCKPORT. (sure I)
• SCRANTON ( sure I)
I I
O
4
RANGE OF OCCURRENCE
FOR SURE I AND NYC
CHAMP STATIONS
10
30
eo
80 90 95
99
99.8
99,9
Figure 5-31. Cumulative plots show the frequency of sulfate concentrations in the SURE region on
the basis of the 1974-75 historical data.
Source: Mueller et al. (1979).
5-63
-------
Figure 5-32 shows the estimated number of days exceeding 10 ug/m for August 1977 and January
2~
to February 1978. In August, almost the entire Northeast had at least 10 days with SO. con-
3 3
centrations greater than 10 |jg/m . The area having 20 or more days with more than 10 ug/m
involved Ohio, West Virginia, Maryland, Pennsylvania, and New York. By contrast, in the
O™.
winter months the area of prolonged elevated SO. concentrations shifted toward the West and
Southeast. The upper Ohio Valley remained high, and an increase in the number of days with
more than 10 pg/m also occurred over Tennessee, Alabama, and Georgia,
Studies of seasonal variations have reported elevated concentrations in the summer months
P—
(Hitchcock, 1976; Hidy et al., 1978). The summer monthly mean concentrations of SO,, in some
Z""
regions can be twice those for the winter months. The seasonal variation in SO. concen-
trations in Southeastern and Midwestern cities is less distinct than the variation in New York
2—
City or Los Angeles (see Figures 5-33 and 5-34). Elevated summertime SO. concentrations are
generally reported to be the result of increased homogeneous and heterogeneous oxidation of
anthropogenically produced S09. However, oxidation of biologically produced hydrogen sulfide
2-
has been offered as an explanation for some high SO. concentrations in isolated areas
(Hitchcock, 1976, 1977; Hitchcock, Spiller, and Wilson, 1980). (See also Chapter 2 relative
to H,S oxidation.)
Lavery et al. (1979) postulated the existence of two meteorological conditions that
2- 3
result in regional accumulation of particulate SO* concentrations above 20 ug/m in the
northeastern United States:
The first regime consists of cases where widespread stagnation
occurs with a large high pressure area slowly moving eastward over
the midwestern and eastern United States. Zones of polluted air
collect over areas within 100-300 kilometers of high sulfur dioxide
emissions sources. These zones maintain themselves over periods of
one to four days in warm, moist air, with light winds, around the
southern and western parts of the high pressure area. The second
regime appears to be conducive to long-range (greater than 500 km)
sulfate transport and involves a channeling of air flow between the
west side of the Appalachian Mountains and weak cold fronts approxi-
mately oriented west-southwest to east-northeast and traveling
south-eastward. The channeling appears to be combined with capped
vertical mixing associated with subsidence around the frontal
system. These episodes can last up to four days.
2-
5.5.1.2 Urban Variations—The preceding discussion of spatial and temporal variations of SO.
was derived for the most part from widely spaced rural monitoring stations. It is of interest
O™.
to note spatial variations on the much smaller scale of a metropolitan area. The SO,
measured on this scale may consist of a natural background component, a long-distance trans-
ported component, a component formed locally in the atmosphere, and/or an artifact formed on
2-
the filter. Hidy et al. (1978) compared urban SO. distributions from the previously reported
works of Lynn et al. (1975) for the New York City area, and Kurosaka (1976) did the same for
the Los Angeles area. These areas differ in meteorology and climate, but the population and
total S02 emissions are similar.
5-64
-------
B
250
Figure 5-32. Maps show the spatial distribution of number of days
per month that the sulfate concentration equaled or exceeded
10 jug/m3. (A) January-February 1978 (31 days); (B) August 1977
(31 days).
Source: Mueller et at. (1979).
5-65
-------
200
190
180
170
160
150
uj 140
D
I
Z
<
u.
o
130
120
110
100
90
80
70
60
50
40
30
20
10
0
1 I I I I I T
I I I I
•SOZ AMBIENT LEVEL
I I I
I I I I I I I I
567
MONTH
10 11 12
Figure 5-33. 1977 seasonal patterns of SO2 emissions and 24-hr
average SO2 and 804 ambient levels in the New York area are
normalized to the annual average values.
Source: Lynn et al. (1975).
5-66
-------
300
JAN MAR
MAY
JUL
SEP
NOV
Figure 5-34. Monthly variation in monthly mean of 24-hr average
sulfate concentration at downtown Los Angeles is compared with
monthly mean 1973 Los Angeles County power plant SO2
emissions.
Source:.Hidy et al. (1978).
5-67
-------
The population density of New York City is greater than that of Los Angeles (see Table
5-10). In addition, the emission patterns are dissimilar. As seen in Figure 5-35, there was
2-
a significant difference in SO, concentrations across the New York urban area, with the
highest values observed in a strip from Staten Island northeast into Brooklyn. This finding
2~
may have been biased by the wintertime emissions; in summer, fairly uniform SO. concen-
trations have been found in the New York metropolitan area. The highest density of SCL
emissions were in eastern New Jersey, Staten Island, Brooklyn, and Manhattan. Within a
2-
distance of 10 to 50 km from the sources of highest S0? concentration, the SO* concentrations
decreased by 30 to 40 percent from their maximum values.
As shown in Figure 5-36, the mean annual average concentrations derived from 24-hour
values in Los Angeles showed a relatively uniform distribution across the Los Angeles basin
area. A weak maximum was found near Burbank and another in the San Bernardino area. The
areas of major S02 emissions were El Segundo, Long Beach, and Fontana. One similarity to
New York was found in the Los Angeles area; at distances exceeding 50 km from highest concen-
2-
tration areas, the SO. levels dropped off significantly.
Spengler and Dockery (1979) measured sulfates in particles less than 3.5 urn in diameter
using a network of 10 to 12 sites in each of six cities for periods of up to 2 years.
Analysis of variance showed no significant variation among sites within the cities of Topeka,
KS; Portage, WI; Kingston, TN; and Watertown, MA. Some slight variations occurred among the
sites in St. Louis, and significant variations occurred among the sites in Steubenville, OH.
Only the Carondolet area of southeast St. Louis was monitored, not the entire city. There are
a coke plant and a lead pigment plant nearby, which cause large SO, gradients and perhaps also
?-•
SO. gradients. In Steubenville, the TSP and SO,, values near the river were approximately
twice the concentrations 5 km to the west of the river. For sulfates in this size range, the
pattern was similar and the gradient was not as pronounced, but the differences among sites
were significant.
2~
An attempt was made to'explain the variability in SO, data for both the Los Angeles area
and the New York City area by means of stepwise linear regression. Table 5-11 displays the
three principal independent variables and the r values associated with them in explaining the
2-
variance of the daily SQ4 concentrations. The results were very consistent in both areas
except for Vista, CA, a community about 100 miles southeast of Los Angeles.
The results indicated that the most important variables were the 24-hour 0,, level, the
2-
midday RH, and the total particulate mass concentration, minus the SO. and nitrate fraction.
Hidy et al. (1978) also suggested that these three factors were important in determining the
2-
daily variations of SO, concentrations. The 0,, or oxidant levels are an indication of photo-
chemical oxidation, the RH is an indication of water content of the air mass, and TSP is an
indication of reactions involving PM.
5-68
-------
TABLE 5-10. SOME CHARACTERISTICS OF POLLUTION IN THE
NEW YORK AND LOS ANGELES AREAS
Parameter
Surface area considered, km
Population estimate (1970)
2
Population density, no. /km
S02 emissions, tons/yr
2
SOp emission density, kg/km /yr
Maximum temperature, °CC
Minimum temperature, °CC
Relative humidity, %e
Normal precipitation, cm
Mean wind speed, m/sec
Mixing height, m^
2
Ventilation, m /sec
so2, ug/m3 h
Water-soluble sulfate (S042"), ug/m3 h
N0£, ug/m3 h
Water-soluble nitrate (NO, ), ug/m
3 h
Qg, Mg/m
Total PM concentration (TSP) 3
less SO* and nitrate, TSPM, |jg/m
Los Angeles
21,000
9,000,000
430
238,000
10,300
22.8 (5.5)d
10.8 (4.6)
50.2 (17.0)
36
3.3 (1.4)
849 (472)
2690 (2160^)
12.5 (19.9,)
10.1 (7.9)
83.9 (44.3)
9.1 (7.7)
52 (34)
64.5 (27.4)
New York
17,000
12,000,000
710
266,000
14,200
15.0 (7.4)
9.3 (8.4)
59.6 (16.5)
106
5.8 (2.3)
1290 (906)
7460 (6200)
42.9 (45.0)
8.9 (5.7)
67.6 (36.0)
2.6 (2.1)
20 (22)
40.4 (19.9)
aGreater metropolitan areas; Los Angeles, South Coast Air Basin; New York, tri-
.state metropolitan area.
Based on EPA Air Quality Control Regions.
•Annual mean of daily maximum or minimum hourly temperature.
Numbers in parentheses are standard deviations.
-rAnnual mean of daily minimum humidity.
Annual mean of noon wind speed at surface.
^Defined by annual mean of daily midday radiosonde sounding.
Annual mean of 24-hr averaged values, 1974-75; Los Angeles, seven stations,
New York, four stations.
Source: Hidy et al. (1978).
5-69
-------
NEW JERSEY
JERSEY
CITY
13
ATLANTIC OCEAN
km
Figure 5-35. Map shows annual mean 24-hr average
sulfate levels in micrograms per cubic meter in the
New York area, based on 1972 data from Lynn et al.
(1975). Squares are locations of three CHAMP site
stations. The fourth station is at the tip of Long Island
about 160 km from Manhattan.
Source: Hidy et al. (1978).
5-70
-------
LOS ANGELES CO.
zusa
A
West Covina
"1*1 «
•12.5
**-/
West Los Anelas
'MB
1
Santa Monica
.foil \
Anaheim
RIVERSIDE CO.
.CIRCLED NUMBERS; STATION DATA
A/ CHAMP STATIONS
UNCERTAIN
DISCREPANCIES L
BETWEEN AGENCY ANALYTICAL METHODS
Figure 5-38. Distribution of annual average sulfate concentration in micrograms per cubic meter in the
greater Los Angeles area is based on 1972-74 data.
Source: Kurosaka (1976).
5-71
-------
TABLE 5-11. PRIMARY RANKING OF VARIABLES FOR CORRELATING AIRBORNE
IN TWO CITIES BASED ON A STEPWISE LINEAR REGRESSION OF 15
VARIABLES FROM CHAMP AND RELATED MONITORING STATIONS
A. Los Angeles area
Garderr West
Santa Thousand
Anaheim Grove Covina Glendora Monica Oaks Vista
Variable
1
2
3
Correlation
coefficient (R)
B. New York
Variable
1
2
3
Correlation
coefficent (R)
°3 °3
TSPM TSPM
RH RH
0.71 0.77
Brooklyn Queens
TSPM TSPM
RH RH
°3 °3
0.60 0.63
03 TSPM 03 TSPM Tm1n
TSPM RH RH RH QX
RH 03 TSPM 03 RH
0.79 0.79 0.79 0.72 0.56
Bronx Riverhead, L.I.
TSPM TSPM
°3 RH
RH 03
0.54 0.62
Located 50 km north of San Diego and 16 km inland from the coast,
RH = Relative humidity.
0- = 1-hr daily maximum ozone value.
TSPM - Mean TSP.
T - = Minimum temperature.
Source: Hidy et al. (1978).
5-72
-------
The local SOp concentrations did not enter into the correlation sequence as one of the
three principal variables. The findings of Spengler et al. (1979) are not inconsistent with
these results, since the only city with a significant spatial variation among sites for sul-
fates also had a variation among sites for the respirable particles and TSP,
5.5.2 Nitrates
Nitrate aerosols make up a varying amount of the TSP. Although widely reported to be
2-
significantly less than the SO, fraction, nitrates nevertheless represent an important con-
stituent. Most nitrates in the atmosphere are formed in gas-to-aerosol reactions, principally
involving nitrogen dioxide and nitric oxide. These reactions may yield HNCU (gas or aerosols),
ammonium nitrate, sodium nitrate, and lesser amounts of other compounds. A minor fraction of
the NO- aerosols measured in the atmosphere can be attributed to wind erosion of soil and re-
suspension of fertilizers (National Research Council, 1978b). These sources may be more im-
portant locally near fertilizer plants, transfer facilities, or munition factories (National
Academy of Sciences, 1977).
Measurement of participate NO,, has proved especially difficult (see Chapter 3). It has
been clearly shown that gaseous HNQ- is absorbed by glass-fiber filters (Pierson et al., 1980a;
Spicer et al., 1978; Appel et al., 1979) and, consequently, NO., values measured, with such
filters might be erroneously high when interpreted as reflecting particle composition. On the
other hand, ammonium nitrate has significant vapor pressure and could be Tost from some media
by evaporation or by acid attack (Pierson et al., 1980a). There exists considerable-contro-
versy over the interpretation of available analytical data for NO,, but the weight of current
opinion appea'rs to favor the idea that glass-fiber filter N03 may be the sum of particulate
nitrate and gaseous HNOg. This issue is not yet settled, nor could any sort of consensus be
said to exist. Still, the available data do indicate those regions with elevated levels of
some NO, species, either gaseous or particulate, previously measured as TSP mass.
*5
Mean NO- aerosol concentrations from urban and nonurban NASN sites are summarized in
3 ™,_~™,
Figures 5-37 and 5-38, respectively. The annual average concentrations shown are in micro-
grams per cubic meter, as measured from hi-vol samples. Concentrations in urban air were sub-
stantially higher than those in nonurban air. A zone of high urban concentrations exceeding 4
o
ug/m extended eastward from Chicago through the industrialized Northeast through Pennsylvania
to the Philadelphia area. Other high NO- zones were in southern Louisiana, around Birmingham,
,3
AL, and near Little Rock, AK. In general, a zone of high urban nitrate concentrations 3 jjg/m
and higher extended from southeastern Texas through the Midwest and into the Northeast. Of
course, a major emission source may cause high NO- gradients in the surrounding area. For
example, a study in Chattanooga, TN, (Helms et al., 1970; National Academy of Sciences, 1977)
showed an average NO- concentration of 48.9 M9/m 'for a site close to the Volunteer Army
Ammunition Plant. This is more than three times the NASN maximum station average for 1965
(13.5 ug/m ). This station average was 15 to 20 times higher than those of the four other
5-73
-------
Figure 5-37. Map shows U.S. mean annual ambient nitrate levels in micrograms per cubic meter.
Source: Akland (1977).
5-74
-------
Figure 5-38. Mean nitrate concentrations in micrograms per cubic meter at nonurban sites in the
U.S., based on valid annual averages from 1971 through 1974.
Source: U.S. Environmental Protection Agency (1977).
-------
Chattanooga sites presumably not influenced directly by the munitions plant. Their averages
ranged from 2.4 to 3.8 ug/m . While the artifact phenomenon may discredit the absolute
values, the ratios among sites have more credibility.
It is obvious from these figures that the data base is quite incomplete for the West
Coast. No data were reported for the Los Angeles area, nor for the large metropolitan areas
of San Francisco, Seattle, and Portland.
A few studies have sought information on NCL concentrations by composition and particle
size. Orel and Seinfeld (1977) compared the formation, sizes, and concentrations of ambient
2-
SO. and NO, particles. Unlike HLSO., the HMO, that is formed tends to remain in the gaseous
phase, although it may be an important component of acid precipitation. The EPRI SURE Project
_ O-
(Kneip et al., 1979) reported NO., ion concentrations one-tenth the concentration of SO^ ions.
3
The monthly mean values for August and October 1977 were less than 0.6 (jg/m ammonium nitrate
at three locations across the Northeast. On a few occasions the daily levels exceeded 1.5
ug/m .
2-
Data from the SO. and N03 data base of the California Aerosol Characterization Experi-
ment are reported by Appel et al. (1978). Summer measurements for 1972 and 1973 from five
fixed and one mobile site indicate that the mass median diameter for nitrates was between 0.3
and 1.6 urn. Twenty-four-hour averaged concentrations of NO, ion varied across the Los Angeles
3 3
basin, from a low of 4 ug/m in Dominguez Hills to a high of 31 ug/m in the eastern community
2-
of Rubidoux. In contrast to SO. the diurnal pattern for NO, often had a maximum during the
morning close to the maximum for gas-phase nitrogen oxides. The authors concluded that the
ratios of ionic constituents and ambient NHQ levels suggested that ammonium salts were the
2- -
principal form of SO. and NO,.
Until recently such high NO, levels were not suspected in other parts of the United
States. However, the EPA early analyses from a sampling program in Denver, CO, found 24-hour
NO, levels, primarily in the fine fraction, that often exceeded 10 ug/m (Courtney et al.,
1980).
Japanese workers have been investigating atmospheric nitrates for some time. Kadawaki
(1977) found a bimodal distribution of nitrates in the Nagoya area of Japan. Submicrometer
particles (0.4 to 0.6 urn in diameter) were ammonium nitrate; the coarse particles (3 to 5 urn
in diameter) were sodium nitrate. Background nonurban levels as low as 0.8 to 0.9 ug/m' on
the outer islands of Japan have been reported (Kito, 1977). Maximum average concentrations in
the city of Kawasaki were reported to be as high as nearly 7 ug/m (Terabe, 1977).
In summary, our knowledge of nitrates in the atmosphere is rather limited. No compre -
hensive data set exists. The NASN measures NO, ion every 12 days at relatively few sites;
spatial and short-term temporal variations cannot be discerned. In fact, there are many
cities for which no measured values of nitrates have been reported. Furthermore, historic
data before 1977 are in doubt because of the artifact formation on the filters.
5-76
-------
There are spatial patterns In NO, concentrations. Cities tend to have higher levels of
nitrates than do rural regions. Some studies indicate that localized areas may have substan-
tially higher NO, levels. This raises the concern that available data on NO, concentrations
may underestimate the actual population exposure. In the near future, new sampling and
analysis techniques should expand our knowledge of NO, aerosols, HNO,, and other nitrogen
compounds.
5.5.3 Carbon and Organics
A variety of carbon-containing compounds often account for a large but highly variable
portion of fine PM. Elemental carbon, which is emitted from a variety of combustion sources,
is a significant component also, usually accounting for 10 to 20 percent of urban aerosol
mass.
Organic substances of biological origin occur in particles. Evidence from microscopic
examination of particles, cited later in this chapter, has demonstrated the presence of wood,
paper, insect parts, pollen, bits of leaf, and textile fibers in coarse particle fractions.
For the most part, these materials are made up of cellulose and protein, biologically derived
substances that are insoluble in common • organic solvents and that decompose to form carbon
residues under thermal treatment. Analysis techniques capable of distinguishing this bio-
logically fixed form of carbon from combustion-formed soot are only now becoming available
(Huntzicker et al., 1982; Novakov, 1982; Wolff et al., 1982).
Organics in particles usually have been determined by extraction in organic solvents or
by thermal evolution of organic vapors, procedures that concentrate relatively low molecular
weight organic components. Consequently, there is little chemical analytical information
relative to the organic high-polymer composition of particles, either those of biological or
manmade origin. In the following discussion, the reader is cautioned that most studies have
concentrated on these simpler organic molecules and do not entirely account for organic
material in particles.
Particles emitted from combustion sources frequently have fuel-derived organic substances
sorbed on their sources (see Chapter 4), and such materials are commonly found in the
atmosphere. Combustion processes often alter fuel molecules considerably and combustion
product mixtures often contain substances not in the fuel. Some examples include products of
coal and wood pyrolysis, oxidized or nitrated hydrocarbons in motor vehicle exhausts, and
synthesis of polynuclear aromatic hydrocarbons in rich flames. Chapter 4 covers many of these
processes.
Photochemical reactions are also capable of generating substantial quantities of organic
particles, and high concentrations of solvent-extractable PM are found associated with high 0^
levels. It is now believed that atmospheric oxidation of some volatile organic species leads
to formation of bifunctional molecules, especially dicarboxylic acids, of very much lower
vapor pressure than their precursors. These reactions are discussed in considerable detail in
Chapters 2 and 6.
5-77
-------
In addition to the organic species accounting for aerosol mass, there are also present in
the particles very much smaller amounts of polynuclear aromatic hydrocarbons, components of
special interest because several are known to be carcinogenic. One of these compounds,
benzo(a)pyrene was conventionally measured in NASN TSP samples. A complete discussion of
these substances is contained in the EPA Health Assessment Document for Polycycljc Organic
Hatter (Santodonato et al., 1979) to which the reader is referred.
Several comprehensive reviews of airborne-organic PM have appeared recently (National
Academy of Sciences, 1972; Duce, 1978; Daisey, 1980; Hahn, 1980; Lamb et al., 1980). The
subject has also been discussed in other reviews (National Academy of Sciences, 1976; Perera
and Ahmed, 1978; Grosjean, 1977).
5.5.3.1 Physical Properties of Particulate Organics—Many atmospheric organic compounds are
distributed between the vapor and particulate phases of the aerosol (De Wiest and Rondia,
1976; Krstulovic et al., 1977; Cautreels and Van Cauwenberghe, 1978), and, presumably, this
distribution can vary with temperature. Because of this volatility, there can be substantial
losses of low molecular weight compounds during sampling (Cautreels and Van Cauwenberghe,
1978; Krstulovic et al., 1977; De Wiest and Rondia, 1976; Katz and Chan, 1980; Schwartz et
al., 1981). At the high temperatures found in combustion sources, larger proportions of the
emitted organic compounds will be present in the vapor phase. These compounds will condense
on the surface of PM as the emissions cool and, thus, be enriched at the surface. Natusch
(1976) found that this occurs when PAH (polynuclear aromatic hydrocarbons) is emitted from
powerplant stacks. Such surface enrichment can affect the biological impact of polycyclic
organic matter (POM). While there is the possibility that POM may exist as particles formed
by self-condensation, most POM is probably adsorbed on the surface of other particles, much of
it presumably associated with soot particles (Thomas et al., 1968). The effect of the
substrate upon which POM is adsorbed upon the chemical and biological reactivity of these
compounds is almost entirely unknown. Korfmacher and coworkers (1980) recently reported that
photodegradation of some PAH compounds proceeds much more slowly when the compounds are
adsorbed on coal fly ash than when adsorbed on other substrates such as silica gel.
The distribution of organic particles between vapor and particulate phases is also
profoundly influenced by chemical reactions in the atmosphere. Grosjean and Friedlander
(1975) found that during photochemical oxidant incidents organic substances are converted from
volatile to relatively nonvolatile species. In this process, the fraction of organic mass in
the particle phase (relative to the gas) can grow from very low values, 1 percent or so, to
about 6 percent of the vapor and particle total.
While both mass and size distribution of organic substances in particles is clouded by
their volatility, there have been some attempts to establish the fine/coarse ratios. Some of
the heavier polycyclic components are known to be predominantly fine-particle components
(Mueller et al., 1964; De Maio and Corn, 1966; Kertesz-Saringer et al., 1971; Pierce and Katz,
5-78
-------
1975; De Wiest, 1978; Van Vaeck and Van Cauwenberghe, 1978, 1980). In Los Angeles oxidant
incidents, virtually all the organic particles were found to be smaller than 2.5 urn (Schuetzle
et al., 1975; Mueller et al., Hidy et al., 1975). Van Vaeck and Van Cauwenberghe (1978)
reported that aliphatic hydrocarbons and carboxylic acids were predominantly (90 percent) in
fine particles in European samples. Since organic compounds are generally distributed
disproportionately in the fine fraction aerosols, it is not surprising that they represent an
important fraction of the mass. Steigerwald (1975) estimated that organic substances are
between 25 and 47 percent of the fine particle fraction in the United States. However, there
are also reports of significant fractions of organic substances in coarse particles in rural
samples (Stevens et al., 1982). No consistent study of organic species or classes by particle
size exists in the literature. Therefore, it is currently impossible to trace the origins and
fate of organic particle components or area or timewise distributions on the basis of ambient
air measurements.
5.5.3.2 Carbon and Total Organic Mass — There are limited historical data on the mass fraction
of elemental carbon in atmospheric aerosols, but very recent work is contributing information
in this area (Rosen et al., 1982; Novakov, 1982; Wolff et al., 1982; Huntzicker et al., 1982;
Stevens et al., 1982; Lewis and Macias, 1980; Stevens et al., 1980). Techniques currently
employed detect both organics and carbon by optical absorption and selective combustion
techniques.
Novakov (1982) found elemental and organic carbon in over 1000 samples collected from a
variety of urban sites. In New York City, the principal chemical species present was
elemental carbon, accounting for two- thirds or more of the carbon mass; the balance was
organic. In Denver, about 60 percent was elemental carbon, while in Los Angeles about 70
percent was organic and the balance, elemental carbon.
Wolff et al. (1982), using a somewhat different technique, reported carbon concentrations
a
in 10 U.S. locations. "Apparent" elemental carbon was reported to range from 1.1 ug/m in a
remote South Dakota location to 13.3 M9/m in New York City. These values covered a range of
4 to 11 percent of the TSP.
Stevens et al. (1980, 1982) and Lewis and Macias (1980) reported total carbon values of
8.4 ug/m (14 percent of total dichotomous sample mass) in Charleston, WV, while in the Great
' 3 3
Smoky Mountains and Shenandoah Valley only 1.2 MS/m and 1-5 ug/m » respectively (4 percent of
o
total dichotomous sample mass in both cases), was elemental carbon and 3.4 ug/m and 0.9
ug/m , respectively (12 percent and 3 percent of total dichotomous sample mass), was organic
carbon.
The mass of organic substances present in atmospheric aerosols was at one time
approximated by solvent-extraction with benzene in routine NASN hi-vol samples. Other
solvents have also been used in such determinations. Unfortunately, such determinations were
terminated in 1970 and, except for a few intensive studies (Daisey, 1980; Grosjean and
Friedlander, 1975; Wesolowski et al., 1980), there has been no extensive data base on organic
5-79
-------
extractables covering the past 10 years.
Some typical values for particle number and mass concentrations of organic substances are
listed in Table 5-12. The organic fraction of the mass concentration as measured by the
*
benzene-soluble component is also listed, with the benzo(a)pyrene fraction for comparison.
In the organic fraction, a variety of organic compounds have been identified, including some
materials classified as PAH (Corn, 1968). However, the identified fraction represents only
10% of the organic components of the urban aerosol. Although the total aerosol number
concentration is often very large in cities, the mass concentration varies less and rarely
exceeds about 200 ug/m in the United States. The benzene-soluble fraction of this is about
10-20 percent of the total mass, and the concentration of benzo(a)pyrene is far lower. Even
in remote areas, there is a contribution of organic material.
A limited number of samples have been collected in unpolluted atmospheres. Levels in
remote areas and in marine air for the ether-soluble fraction of organic particulate matter
have been as low as 0.51 (0.18-0.84) ug/m STP. Marine air with a continental influence had
averages of 0.93 (0.48-1.38) ug/m3 and continental air 1.2 (0.69-1.71) ug/m3. Similar
concentrations have been observed at Barrow, AK, a remote' site in the Arctic, for cyclohexane-
and dichloromethane-soluble POM (Daisey et al., 1981).
Variations in the concentration of organic particulate matter by location, meteorological
conditions, season and by time of day have been observed repeatedly (Hidy et al., 3975;
Gordon, 1976; Calvert, 1976). By way of illustration, Figure 5-39 shows the differing
contributions of the organic fraction in samples obtained in two cities in southern California
(National Academy of Sciences, 1972). In both instances, however, the organic fraction
represents a sizeable portion of total suspended particulate material.
A pattern of elevated wintertime concentrations of organic particulate matter has been
observed in New York City and in Mainz, West Germany. Winter samples in Mainz of
ether-extractable organic material averaged 27 ug/m for TSP concentrations averaging 150
o
{jg/m . Winter samples collected in February, 1977 in New York City had a total extractable
3 3
organic fraction of 22 ug/nr for a TSP average of 96 ug/m . The August, 1976 levels were 13.3
pg/ra3 for a TSP average of 86 ug/m3 (Kneip et al., 1979).
A 1971 study of Colucci and Begeman is an example of a more detailed short-term urban
survey of PAH than is available from NASN data. From 1964 to 1965, Colucci and Begeman (1971)
found that the concentrations of benzo(a)pyrene and benz(a)anthracene were 4 1/2 times greater
in central Los Angeles than at two suburban sites. However, the suburban site downwind of the
downtown area (on the average) appeared to have systematically higher benzo(a)pyrene
concentrations than the upwind site. Daily concentrations reported in the Los Angeles area
3 ^
ranged from 0.1 ng/ra to over 10 ng/m , depending on the season. Benz(a)anthracene
concentrations were 1 1/2 times larger than the benzo(a)pyrene concentrations. Annual average
benzo(a)pyrene concentrations were similar to the NASN data for downtown Los Angeles. The PAH
The benzene-soluble extract is not necessarily equivalent to the tota.l amount of organic
material in the sample, but it is taken to be representative of such a fraction.
5-80
-------
TABLE 5-12. TYPICAL VALUES OF AEROSOL CONCENTRATION FOR DIFFERENT
GEOGRAPHIC AREAS (ANNUAL AVERAGES)3
Number of ,
Location particles/cm^
Nonurban
Continental
General
California
Oregon
Colorado
Indiana
Maine
New York
So. Carolina
Maritime
General
Pacific offshore
Oahu, Hawaii
Urban
Continental
General
Los Angeles
Portland
Denver
Minneapolis
Chattanooga
New York «
Greenville, SQ
Maritime
Honolulu, Hawaii
San Juan, Puerto
Rico
103-
103 -
3
HT -
-
-
-
-
iof-
10; -
io3-
f\
io3-
103 -
*\
103-
103 -
-
-
-
103-
"
wl
IO4
A
IO4
fl
10}
IO4
IO4
10}
IO4
f)
103
IO3
IO4
Mass
concen-
tration,
MS/m^
20 - 80
39
47
14
39
18
29
40
-
19 - 146
10 - 49°
MOO
93
72
110
70
105
105
76
40
77
Benzene
soluble
fraction,
|jg/ms
1.1 - 2.2
2.8
0.9
1.1
2.1
1.2
1.8
2.7
— A
1.5 - e.rj
0.7 - 6.3°
7
12.5
6.6
9.0
6.1
6.9
3.9
7.4
2.3
6.9
BaP
fraction,
ng/m^
-
0.48
0.09
0.11
0.25
0.12
0.25
0.43
-
-
-
-
1.87
2.60
2.52
1.18
4.18
3.63
7.49
0.59
1.42
Data based on 1969 NASN observations, except for
maritime data.
Aitken nuclei
"Geometric means.
-i
Short-term data.
5-81
-------
I
J I I 1 I -1 1 I I 1 } I I I I I I I II I I I I I I I I I
NH/'(3%1
S04!M4W
NO (i
(WATER (16%)
ORGANICS (43%)
Mill IN M I I I I I I I II I I I I IN I I I
10
20
30
40
50
60
PASADENA, 9/20/72; TWO HOUR SAMPLE OVER 1200-1400 PST; LOW OXIDANT, TOTAL
MASS CONCENTRATION, 79 /ng/m3.
I MITT IT! I I I M i M M
NH4+ (10%)
WATER (10%)
S04" (1354!
i ORGANICS (24%)
«• C26%!
INOjf (26%!
I I M I I I M I I I I I I I.
10
20
30
40
50
60
POMONA, 10/24/72; SAMPLES FROM 1200-1400 PST; MODERATE OXIDANT, TOTAL MASS
CONCENTRATION, 178 pg/m3.
Figure 5-39. Calculated distribution of aerosol constituents for two aerosol samples
taken in the Los Angeles basin.
Source: National Academy of Sciences (1972).
5-82
-------
concentrations increased substantially in winter. Benzo(a)pyrene concentrations were higher
at night, in contrast with those of other pollutants. All pollutants were higher on weekdays
than on weekends. Benzo(a)pyrene concentrations were found to be correlated with carbon
monoxide and lead concentrations, with coefficients ranging from 0.6 to 0.9. Benzo(a)pyrene
concentrations were also significantly related to those of hydrocarbon vapors, oxides of
nitrogen, and vanadium (a nonautomotive pollutant). Despite the strong relation to lead, the
statistics in the study failed to reveal a clear identification of BaP emissions with
automotive or stationary combustion sources.
Trends of BaP concentrations as measured at 34 NASN urban sites are displayed in Figure
5-40. It is indeed encouraging to see the steady decline in BaP concentrations that has
occurred since the mid-19601 s. The 90th percent! le of quarterly measurements fell
3 3
dramatically from near 7 ng/m to less than 2 ng/m . These changes reflect both emission
controls and shifting of sources. Incomplete combustion of fossil fuels, especially coal, is a
primary source of BaP. Major point and area sources include residential coal-fired furnaces,
coal-fired utilities and industrial boilers, coke ovens, petroleum refineries, and
incinerators (see Chapter 4). Shifts away from coal for residential, commercial, and light
industrial use have made a substantial contribution to the reduction of urban BaP
concentrations. To a lesser extent, the control of particulate emissions has also helped to
lower concentrations.
The national trends in benzene-soluble particulate matter and BaP as reported fay Faoro
(1975) may not be true everywhere. Indeed, specific organic fractions may show opposite
trends. Daisey (1980) discussed benzene soluble organic trends for New York City; annual
averages for the New York University station, normalized to account for year-to-year
meteorological variations, are reported in Table 5-13.
TABLE 5-13. ANNUAL AVERAGES OF ORGANIC FRACTIONS IN TSP
NEW YORK CITY3, DISPERSION NORMALIZED
Year
1968
1969
1977-78
TSP,
3
jjg/m
95.7
129
59.8
Organic fraction
pg/m
10.2
10.8
8.8C
Percent organics
in TSP
10.6
8.4
14.7
aNYU Medical Center Station.
b
total of nonpolar (benzene-soluble) and polar (acetone-soluble) organics.
€Respirable (£3.5 u) organics only.
Source: Daisey (1980).
5-83
-------
10
•
DC
Z
ui
u
z
8 4
CQ
. 90PERCENTILE OF
QUARTERLY
MEASUREMENTS
•50PERCENTILE OF
OUARTERI.Y
MEASUREMENTS
1
1966 67 68 69 70 71
TIME, year
72
73
74
75
Figure 5-40, Benzo{a)pyrene seasonality and trends (1966 to 1975) in the 50th and 90th
percontiles for 34 NASN urban sites.
Source: U.S. Environmental Protection Agency, 1979.
5-84
-------
Although TSP had decreased by 40% between 1968 and 1978, the POM fraction decreased by only
10%.
5.5.3.3 ChemicalComposition of Particulate Organic Matter—Participate organic matter has
often been fractionated by means of acid-base extractions followed by column chromatography
(Hueper et a!., 1962; Tabor et al., 1958; Hoffmann and Wynder, 1977; Asahina et al., 1972).
The composition data for Detroit PM found in Table 5-14 are fairly typical and present the
general proportions of various broad classes of compounds present. Hoffmann and Wynder (1977)
found that the fraction containing PAH was principally responsible for the tumorigenic pro-
perties of POM in mice.
TABLE 5-14. COMPOSITION OF THE ORGANIC FRACTION
OF AIRBORNE PM COLLECTED IN DETROIT3
Fraction Percent of total extractable
organic matter
Aliphatic hydrocarbons 48.3
Aromatic hydrocarbons 3.6
Neutral oxidized hydrocarbons 20.8
Acidic compounds 14.8
Basic compounds 0.55
Insolubles 10,8
a3 jjg/m annual average benzene-soluble organics.
blncluding PAH.
TSP not reported.
Source: Hoffman and Wynder (1977).
Specific classes of compounds identified in the organic fraction of airborne PM include
PAH,aromatic and aliphatic hydrocarbons, aza-arenes, aliphatic and aromatic aldehydes and
ketones, quinones, phenols, polyols, phthalic acid esters, sulfur heterocyclics, aryl and
alky! halides, chlorophenols, nitro compounds, and alkylating agents (Hoffmann and Wynder,
1977; Daisey, 1980; Lamb et al., 1980). Of all the airborne organic compounds, the most
information exists for the classes of POM. The greatest attention has been focused on the
subclasses of PAH and the polycyclic heterocyclic compounds such as the aza-arenes, because
many of the compounds in this class are potent carcinogens in animals. Some of the polycyclic
5-8$
-------
hydrocarbons identified are pyrene, BaP, benzo(e)pyrene, benz(a)anthracene, perylene,
chrylene, chrysene, coronene, fluoranthene, benzo(ghi)perylene, and alkyl derivates of these
compounds (Sawicki et al., 1962; Sawicki et a"!., 1965).
Benzo(a)pyrene was one of the earliest compounds in this mixture of organic matter to be
identified and routinely measured. Some measurements for BaP in the United States date to the
» •*.« '^
early 1950's. Sawicki and coworkers in the 1960's extracted and identified many organic
compounds. Today there is a renewed effort, using more sophisticated techniques and
attempting to answer the many questions still remaining on the biological significance,
variations and concentrations, specific source contributions, and the reactivity of airborne
organic matter.
In 1972, the National Academy of Science published an extensive report on the biologic
effects of airborne matter entitled Particulate Polycyclic Organic Matter (National Academy of
Sciences, 1972). According to the report, emission source data for airborne organic
substances are generally expressed in terms of estimated BaP emissions. Benzo(a)pyrene is
used as a surrogate for detecting the presence of airborne organic pollutants because it
appears to be a prominent constituent of POM. Benzo(a)pyrene is 'also a known animal
carcinogen and the best documented of all the polycyclic organic compounds (National Academy
of Sciences, 1972). It cannot be regarded as a perfect indicator of polycyclic aromatic
hydrocarbons in the air nor of their carcinogenic properties; however, because better data are
generally not available, BaP is presently used as an indicator of the potential catcino-
genicity of general air pollution (Bridbord, et al., 1976).
Despite much work on certain subfractions of POM, such as the polycyclic organic
fraction, other compound classes such as the oxidized hydrocarbons remain relatively
unexplored. Sawicki (1976) has estimated that "over 99% (sic) of the organic pollutants in
the air have never been determined."
In photochemical incidents, volatile hydrocarbons are converted to very large quantities
of 5 to 7 carbon bifunctional carboxylic acids (Schuetzle et al., 1975; Grosjean and
Friedlander, 1975; Cronn et al., 1977). Schuetzle et al. (1975), in a report of a 1972
incident, state that alkanes and alkyl naphthalenes accounted for 1.5 to 3 percent of the fine
particle mass, and bifunctional compounds amounted to about 11 percent. In addition to
glutaric, adipic, and pimelic acids, the corresponding hydroxy carboxylic acids and a variety
of their nitrate and nitrite ester derivatives were reported.
Cronn et al. (1977) confirmed those findings in a series of sub-3.5 pm samples taken
during the 1973 California Air Characterization Experiment. These authors found levels of
3 3
organic particulate matter up to 65 jjg/m out of a fine-particle loading of 230 pg/m . These
substances included small amounts of alkanes, alkyl naphthalenes, and piperidines (up to 12
ug/m ) and much larger quantities of CK to C-, dicarboxylic acids, hydroxy-acids, and amides.
3
Grosjean and Friedlander (1975) have found organic extractables of 141 jjg/m during an
incident in 1973; one-half to one-third of this mass was polar organics. These organic
5-86
-------
substances together with ammonium sulfate and nitrate accounted for 95 percent of the
secondary aerosol during photochemical incidents. Therefore, there is substantial evidence
that organic particles can be influenced in a very major way by photochemistry.
Classes of biologically active compounds other than PAH and related polycyclic organics
have been identified in airborne PM. This includes alkylating agents, N-nitrosamines (Kneip
et al. , 1979), nitro derivatives of PAH (Jager, 1978), and the compounds responsible for the
mutagenic activity of POM in the Ames assay (Kneip et al.} 1979; Talcott and Wei, 1977; Pitts
et al., 1977; Daisey et al.s 1979). In addition, there may also be unstable compounds present
in the aerosol, such as epoxides and lactones (Van Duuren, 1972), that are significant for
human health but decompose when collected by conventional sampling techniques. There is a
need to identify specific compounds such as these, to evaluate their significance for human
health, and to determine their sources and concentrations in the ambient atmosphere. There is
also a need to identify sampling artifacts and develop improved sampling techniques for
organic compounds in the aerosol.
5.5.4 Metallic Components of Fine Particles
It is useful to study not only the chemical but also the elemental composition of
airborne particles. Many trace elements are known to be toxic and can act as catalysts in
atmospheric reactions. Table 5-15 indicates the mean and maximum concentrations of several
elements found in urban and nonurban areas in the United States from 1970 to 1974. Certain
trace elements tend to be enriched in urban airborne particles relative to their concentration
in fuels. For example, Table 5-16 lists the ratios of urban to suburban concentrations of
trace elements in three groups. Since the ratio of urban to suburban TSP was 2.8, those
elements with higher ratios (antimony through bismuth) were concluded to be mainly of urban
origin. The middle group (mercury through calcium) had ratios near that of TSP and therefore
occur equally abundantly in city and suburban particles. The remaining group (silicon through
bromine) appeared mainly in suburban suspended particulate matter. As Table 5-17 indicates,
they were not homogeneously distributed among the various particle size fractions. There was
also spatial variability in the composition of the aerosol, as indicated by Table 5-18. These
intercity differences reflect the difference in industries and types of fuel used in these
urban areas.
Hi-vol filter samples from the NASN have been routinely analyzed for certain metals since
the early 1960's. The data for certain metals for the years from 1965 to 1974 have been
summarized in an EPA report (Faoro and McMullen, 1977). This report presents the composite
national and regional trends for nine trace metals, fuel-related metals (lead, vanadium,
nickel, and titanium) and industry-related metals (cadmium, chromium, copper, iron, and
manganese). These trends were derived from samples collected from 92 urban and 16 nonurban
NASN hi-vol stations.
The instrumental techniques for detecting metals changed in 1970, significantly improving
the lower limits of detection. This report, in addition to Akland (1976) describes the
5-87
-------
TABLE 5-15. COMPARISON OF URBAN AND NONURBAH ANNUAL AVERAGE CONCENTRATIONS FOR SELECTED HETALS, 1970-74
Cd Cr Co Fe Pb Hn N1 TJ_
IT NUb U NU U NU U NU U NU U NU U NU U NU U NU U NU
1970
Tiximuiii 2.9 .24 ,099 .0001 .130 .07S .014 LD 14.2 1.62 5,83 1.471 2.10 .068 .277 .097 .26 .093 1.222 .112
ArithiM-
tic mean L0q ID .003 .0001 .008 .003 LD ID 1.7 .38 1.19 .088 .07 .015 .015 .005 .05 .013 .052 .008
Std. devi-
ation 0.2 .03 .007 -- .011 .009 .001 ~ 1.3 .27 .80 .190 .12 .013 .028 .014 .03 .011 .116 .019
1971
TixidtttS 0.7 .24 .215 .0001 .171 .061 .085 LO 16.0 2.80 6,31 1,134 1.95 .102 .347 .083 .51 .069 1.325 .209
Arithme-
tic mean 0.1 .01 .004 .0001 .009 .004 .001 LD 2.1 .51 1.23 .047 .08 .018 .015 .003 .04 .017 .041 .007
Std. devi-
ation 0.1 .03 .016 -- .014 .007 .003 -- 1.6 .38 .87 .155 .11 .015 .028 .011 .05 .020 .108 .024
tn
s> 19.72
» TiaxiiMi LO LO .112 ,0001 .143 .039 .042 LD 6.4 1.15 6.88 1.048 .86 .046 .268 .082 .48 .092 .858 .205
Arithme-
tic mean - LO .002 .0001 .006 .002 LD LD 1.2 .25 1.13 .139 .04 .007 .011 .004 .04 .027 .022 .004
Std. devi-
ation — — .007 -- .010 .004 ,002 -- .8 .22 .78 .169 .06 .009 .023 .012 .03 .022 .056 .019
1973
"Maximum LD LO .032 .0001 .228 .066 .027 LD 6.9 1.19 5.83 .939 .56 .030 .439 .280 .23 .084 .393 .035
Arithme-
tic mean LD LD .001 .0001 .007 .003 .001 LD 1.1 .19 .92 .110 .04 .004 .014 .011 .04 .028 .016 .002
Std. devi-
ation LD --- .003 -- .015 .009 .002 ~ .8 .18 .64 .149 .05 .005 .037 .037 .03 .021 .034 .005
1974
"Maximum 0.2 LD .077 ,0002 .073 .009 .029 LD 6.2 .69 4.09 .534 .35 .033 .639 .026 .22 .066 .248 .023
Arithme-
tic mean LD LD .002 .0002 .006 .002 LD LD 1.1 .24 .89 .111 .04 .006 .009 .002 .04 .020 .019 .002
Std. devi-
ation - — .001 — .006 .002 .001 — .7 .17 .57 .111 .04 .007 .029 .004 .03 .017 .037 .004
__, ,
Expressed in ng/m?; U = urban; NU = nonurban; CLD = less than detectable. Source: G. Akland (1976).
-------
TABLE 5-16. RATIOS OF URBAN (U) TO SUBURBAN (S) CONCENTRATIONS IN AIR,
CLEVELAND, OH, AREA
Enriched in cities
Ratio similar to TSP
Enriched in suburban par-
ticles
Element U/S
Antimony
Chloride
Beryllium
Chromi urn
Cobalt
Bismuth
6.9
6.5
6.1
5.6
3.4
3.3
Element U/S
Mercury
Iron
Cadmium
Sodium
Magnesium
Manganese
Calcium
3.0
2.8
2.5
2.4
2.4
2.2
2.0
Element U/S
Silicon, tin 1.8
Copper, vanadium 1.8
Aluminum 1.7
Zinc 1.6
Arsenic 1.4
Selenium 1.3
Bromine 1.2
Note:
Mean TSP ratio = 2.8
Source: Economic Commission for Europe's Working Party on Air Pollution Problems
(1977).
5-89
-------
TABLE 5-17. CORRELATIONS OF CHEMICAL CONTENT WITH PARTICLE SIZE
a) Predominant particlesize for various substances
Normally fine (formally coarse Normally bimodal Variable
Iron, calcium
Sulfates
Organic (con-
densed vapors)
Lead
Arsenic
Selenium
Hydrogen ion
Ammonium salts
Soot
Titanium
Magnesium
Potassium
Phosphate
Silicon
Aluminum
Chloride
Nitrate
Nickel
Tin
Vanadium
Antimony
Manganese
Zinc
Copper
b) Ratios of element distribution between fineandcoarse particles
(St. Louis urban aerosol, 18-day average, Aug. to Sept. 1975)
Predominantly fine
Predominantly coarse
Element Fine/coarse
Sulfur 8.90
Lead 3.67
Element
Calcium
Silicon
Iron
Potassium
Titanium
Fine/coarse
0.09
0.13
0.29
0.33
0.55
Source: (a) Miller et al. (1979).
(b) Dzubay et al. (1977).
5-90
-------
TABLE 5-18. PARTICULATE ANALYSES FROM SELECTED URBAN LOCATIONS
i
tc
Suspended particles
Antimony
Beryl 1 i urn
Bismuth
Cadmium
Chromium
Cobalt
Copper
Iron
Manganese
Nickel
Tin
Titanium
Vanadium
Zinc
Atlanta, 6A
97.
0.000
0.000
0.000
0.017
0.002
0.000
0.04
1.2
0,06
0.007
0.02
0.03
0.001
0.52
Birmingham, AL Baltimore, MD
142.
0.000
0.000
0.000
0.008
0.005
0.000
0.06
1.7
0.15
0.004
0.01
0.03
0.003
1.09
146.
0.000
0.000
0.000
0.003
0.018
0.000
0.06
0.8
0.08
0.034
0.01
0.01
0.071
0.34
Albuquerque, NM
120.
0.000
0.000
0.000
0.000
0.001
0.000
0.07
a
0.03
0.000
0.01
0.01
0.001
0.00
Not analyzed.
Arithmetic mean values for 1966 expressed as micrograms per cubic meter.
Copyright by the American Association for the Advancement of Science, 1970.
Source: Corn (1976).
-------
methods used and the implications for trends analysis. In general, the data presented
indicate changes in atmospheric concentrations of various metals occurring in different
regions of the country. For the most part, the reported trends are consistent with changes in
emission patterns (due to industrial source control) and in fuel use.
Similar to trends in urban TSP concentrations, metals concentrations declined in most
urban areas, with the exceptions of copper, titanium, and possibly chromium. Table 5-19
summarizes metal trends and possible causes for these trends. Both beryllium and cobalt had
such very low concentrations that trends could not be identified with any certainty. Trends
in other metals, such as vanadium and nickel, parallelled air pollution control regulations
mandating the use of low-sulfur fuels. There was a drop in vanadium and nickel, particularly
in the Northeast, because the desulfurization process of petroleum also removes these
impurities. Titanium, on the other hand, may have increased because of the rise in coal use
by utilities. Decreases in iron, manganese, and cadmium concentrations were probably the
result of reduced particulate emissions From steel plants and related industries. Improved
incineration methods and use of sanitary landfills instead of incinerators may also have
contributed to the decrease. No trends were apparent for copper, but it is felt that the high
concentrations were caused by contamination from the commutators of the high-volume samplers.
5.5.4.1 Lead—The seasonal patterns and trends in the quarterly averaged urban lead
concentrations are displayed in Figure 5-41. The national composite 50th percentile of lead
3 3
concentrations fell from about 1.1 ug/m in 1971 to 0.84 ug/m in 1974, approximately a
24-percent decline. This is attributed to decreased lead in gasoline and the drop in premium
gasoline sales since 1970. Premium gasoline has, on the average, a higher lead content, than
regular gasoline. Lead concentrations are expected to continue to decrease in the future
because of increased use of unleaded gasoline in new cars equipped with catalytic converters.
This national decrease is not equally evident throughout the country, however, because of
differences in growth rates and vehicle miles travelled. These results should be used with
caution because of the small number of stations used in determining the trends. An NASN
location in downtown Los Angeles experienced the highest lead concentrations, averaging
o o
between 4 and 5 ug/m until 1971; the concentrations decreased to about 2 ug/m in 1974.
Again, while the reduction of lead content in gasoline and the increased use of lead-free
gasoline may have contributed to this decline, the decrease in vehicle miles travelled in
downtown urban areas also contributed to the decline. There is some evidence that rural sites
have shown either stable or slightly increasing patterns in ambient air lead content in the
United States.
5.5.4.2 Vanadium,Nickel.and Other Metals—Figure 5-42 shows, for the five broad geo-
graphical areas of the United -States, the 90th percent!les of the annually averaged vanadium
concentration. The Northeast had substantially higher vanadium concentrations than any other
area of the United States. Over this 10-year period the concentrations of vanadium decreased
5-92
-------
TABLE 5-19. TRENDS IN REPORTED URBAN METAL CONCENTRATIONS
AND THEIR POSSIBLE CAUSES
Metal
Observed trends
Possible causes
Fuel combustion-related
metals:
Beryllium
Lead
Nickel
Titanium
Vanadium
Industry-related
metals:
Cadmium
Chromium
Cobalt
Copper
Iron
Manganese
Unknown
Down; expected to
decrease further
Down
Up
Down
Down
No trend
Unknown
No trend
Down
Down
Lower lead content in
gasolines after 1969
Reduction of Ni in residual
oils
Increasing use of coal in
electric utilities
Reduction of V in residual
oils
Controls in metal industry
and improved incineration
practices
Unknown
Contamination from hi-vol
commutator
Improved incineration or waste
burning practices, fuel switch-
ing, controls in steel industry
Controls in metals industry
Source: Faoro and McMullen (1977).
5-93
-------
3.0
M
2.0
en
z
01
u
O
u
01
1.0
0
1965 66 67 68
I I
72
73 74
69 70 71
TIME, year
Figure 5-41. Seasonal patterns and trends in quarterly average urban lead concentrations.
Source: Faoro and WlcWIullen, 1977.
75
5-94
-------
0.10
"I
o
LU
o
O
o
5
I
0.01
0.001
/ X .
/ <* s
/ .V N, S
-r .• —\->
'\ / x-
* 4
\
N.
NORTHEAST (29 SITES)
SOUTH (IS SITES)
WEST (15 SITES)
NORTH CENTRAL (14 SITES)
MIDWEST {19 SITES)
I I I I I I I I I I
1965 66 67 68 6S 70 71 72 73 74
TIME, year
Figure 5-42. Regional trends in the 90th percentile of the annual averages for vanadium. (A indi-
cates value below lower discrimination limit.)
Note: 1971-1974 90th percentile below lower discrimination limit of 0.003 = jug/m3in the midwest.
Source: Faoro and McMullen (1977).
5-95
-------
3 3
74 percent, from 0.35 ug/m in 1969 to 0.09 ug/m in 1974. Most of this drop occurred between
1971 and 1972. The slight increase apparent in the South was caused mainly by two or three
stations showing relatively high readings in 1972 to 1974; this was not characteristic of
other sites in the region. For both vanadium and nickel, pronounced and regular high winter
and low summer seasonal variations in both the 50th and 90th percentiles occurred in the
Northeast. This is shown in Figure 5-43. These variations are attributed to the metal
contents in the fuels used for space heating. The decrease in the 50th and 90th percentiles
of these two metals was caused by the decrease-in the sulfur content in petroleum used in this
area. This decrease is exemplified by the approximately 70-percent decrease in the sulfur
content of residual oil in the New York City-Westchester County area since 1979. The vanadium
concentrations decreased between 70 and 80 percent over the same time period at the New York
City NASN site.
5.5.5 Acidityof Atmospheric Aerosols
Along with size and chemical composition, acidity of fine atmospheric aerosols is an
important property. Measurement of acidity by titration is preferred to pH measurements
(Junge and Scheich, 1971). To date, measurements of the strong acid content of atmospheric
aerosols have indicated that it is quite variable. Around certain sources such as cement and
lime kilns, the airborne particles may be basic, whereas around other sources such as HpSO^
plants and coke plants the particulate emissions may be very acid.
Weak and strong acids exist in the atmosphere in both gaseous and particulate form.
Organic acids were reported by Ketseridis et al. (1976) in rural, urban, and marine
3 3
atmospheres ranging between 0.3 ug/m to 10 ug/m . Nitric acid has been found in the
2
atmosphere in concentrations ranging from < 2 to nearly 100 (jg/m (Spicer, 1977). Vapor-phase
HC1 may be present (Rahn et al., 1979), although quantitative data are sparse. These organic
and inorganic gases can condense or be adsorbed on particles either jn situ or during filter
sampling. As pointed out previously, filters are capable of absorbing HNOg directly.
Host of the strong acid found in aerosol particles is chemically associated with the fine
O«*
particulate SOI aerosol mass. Charlson et al. (1978) and others cited numerous measurements
of approximate chemical balance between ammonium cations and sulfate anions. Thus the major
?—
form of SOT is ammonium sulfate (that is, HgS04 fully neutralized by ambient ammonia).
However, on occasion urban and rural aerosols can be acidic., Brosset et al. (1975), Brosset
(1978), Hitchcock et al. (1980), Cobourn et al. (1978), Pierson et al. (1980), Lioy et al.
(1980), Leaderer et al. (1982), Stevens et al. (1979), and Tanner et al. (1977, 1979) all
demonstrated that strong acid in the form of NH*HSO* and, less frequently, H-SO^, may exist at
significant levels in the ambient atmosphere. Strong acid levels equivalent to > 15 ug/m of
HgSO, have been observed for periods :> 6 hours.
In urban atmospheres, sulfate anion usually appears primarily in the form of ammonium
+ ?-
sulfate or ammonium bisulfate (NH^ /S0| molar ratios between 1 and 2) as reported by Lioy
5-96
-------
0.800
I I I I I I
VANADIUM
90th PERCENTILE _
73 74
Figure 5-43. Seasonal variation in quarterly averages for nickel and vanadium it
urban sites in the northeast.
Source: Faoro and McMullen (1977).
5-97
-------
et al, (1980) for New York City, Leaderer et al. (1982) for New Haven and New York City, and
Cobourn et al. (1978) for St. Louis. Presumably this greater extent of neutralization of
2~
urban SO. aerosols is due to additional NH- sources in urban areas, although it may be due in
part to analytical interferences from coarse, basic particles such as resuspended cement dust
present in large amounts in urban aerosols.
On the basis of recent experimental and theoretical work, Huntzicker et al. (1980)
2~
indicate that SO, aerosol more acidic than (NH-)HSO. should occur only when SO- is being
oxidized rapidly and where the ratios of [SO,] to [NH,]'are high or when the equilibrium vapor
€, 3
pressure of NH3 over the partially neutralized H^SO. droplet exceeds the ambient NH, partial
pressure. The situation is more complicated in ambient aerosols in which partially ammoniated
p~
SO, is present in mixtures with NO,, carbonaceous, and other aerosol components in solid or
liquid form which may affect its neutralization rate. In particular, some data (Tanner,
1980a) suggest tha.t the degree of mixing in the "well-mixed" boundary layer is inadequate to
prevent vertical stratification of strong acid levels since NH3 is largely emitted (and HNO-
largely removed by dry deposition) at the earth's surface. Further information on the
vertical distribution of strong acid and related species is needed before emission and
neutralization rates may be used to predict acid levels in urban areas.
Cobourn et al. (1978) demonstrated that acid aerosol episodes could occur in urban areas
as suggested by Tanner et al. (1977, 1979) and Brosset (1978). Cobourn et al. (1978), using a
2-
continuous SO. monitor to distinguish H^SO* species from ammonium sulfate, recorded two acid
aerosol episodes lasting 3 days or longer in St. Louis, MO. Both episodes were reported to
have occurred simultaneously with a regional haze, one in July 1977, the other in February
1978. Cobourn ascribed the city occurrence of H?SO. to conditions where atmospheric NI-L
concentrations were exceptionally low (NH, was temporarily depleted from the atmosphere).
?—
The temporal variations of the acid fraction of the SOf aerosol in St. Louis displayed
patterns similar to those reported by Cunningham and Johnson (1976) in Chicago and Tanner
(1980b) for the New York area. The aerosol acidity often changed drastically within a few
hours. Cobourn et al. (1978) noted a diurnal pattern with highest acid levels in midafternoon
and lowest at night. Leaderer et al. (1982), taking 6-hour samples, reported increased
aerosol acidity in the noon to 6 p.m. samples taken at High Point, NJ and Upton, Long Island
(west and east of New York City, respectively), compared to other sampling times. At this
time the relative contribution of SO, oxidation chemistry, temporal variations in NH, and SO,
emission rates, diurnal variations in turbulent mixing rates, and varying height of the mixing
layer to the diurnal patterns were not known. Dzubay (1980) observed that while sulfur and
lead dominated the fine particles, they were insignificant in the coarse particles. The
similar composition of the rural and urban aerosols indicated that the urban materials were
transported to the rural areas.
5-98
-------
Measurements of acidity in eastern United States aerosol samples indicate that strong
acids are present more frequently and in larger amounts in rural and semirural samples than in
the urban samples. Pierson et al. (1980) reported 12-hour concentrations of H ions expressed
as HpSCL as high as 17 ug/m in the Allegheny Mountains of Pennsylvania in July and August
1977. It is likely that strong acid concentrations were substantially higher for periods less
than 12 hours.
Lioy et al. (1980) and Leaderer et al. (1982) characterized the aerosol acidity in the
region surrounding New York City during the summer of 1977. The samples taken at High Point,
NJ (west-northwest of the city), were more acidic (17.8 jjg/m , 6-hour average maximum H ions
expressed as HpSO*) than samples taken in New Haven, CT, and Brookhaven, Long Island, NY.
Part of the period studied by Lioy et al. (1980) was coincident with the research of
Pierson et al. (1980). Using the combined data set in conjunction with air parcel trajec-
tories and haze analyses, Lioy et al. (1980) suggested the presence of a regional acidic aero-
sol distribution encompassing an area at least 200 miles in diameter during the period August
3 to 9, 1977.
As pointed out earlier, atmospheric particles are nearly neutral for the most part, pre-
sumably because of the presence of NHg in the atmosphere. However, instances of particulate
acidity have been reported in several credible studies done in the East, in both winter and
summer and in rural and urban sites.
5.6 COARSE PARTICLES IN AIR
5.6.1 Introduction
In Chapter 2 and in earlier sections of this chapter it was shown that air in cities
usually contains large amounts of particles larger than 1 to 3 pm in size. Particles larger
than 10 to 20 pm tend to settle out of air suspension under the force of gravity. Yet in many
areas these very large dust particles are also present in substantial quantity. This material
is quite commonly deposited as dust on window ledges and silt on roadways. For potential
effects of the ambient concentrations described below, the reader is directed to Chapter 10,
Section 10.3.2, for descriptions of soiling and to Chapter 11, Section 11.2, for information
on deposition of particles in human airways. Coarse-particle mass contributions are
substantial and important in the context of air pollution effects.
The composition and sources of coarse particles are not as thoroughly studied as those of
fine particles. One reason is that they are more complex. For example, it is possible to
recognize dozens of particle types in ambient air samples; these range from soil particles,
limestone road aggregate, fly ash and oil soot to cooking oil droplets, pollen, wood ashes,
and even instant coffee (McCrone, 1968; Draftz and Seven"n, 1980). Man's industry and
activity stirs up dust, quite a lot of dust in arid climates. Unfortunately, the chemical
composition of many kinds of coarse particles can be very similar, at least as determined by
elemental analysis. Consequently, touch of the evidence on large-particle composition has been
obtained from deductions based on microscopical examination.
5-99
-------
5.6.2 Elemental Analysis of Coarse Particles
Measurable elements constituting the major portion of coarse-particle mass in cities are
silicon, aluminum, calcium, and iron (Akselsson et a!., 1975; Lewis and Macias, 1980; Camp et
a!,, 1978; Stevens et al., 1980; Dzubay, 1980; Stevens et al., 1978; Cahill et a!., 1977;
Hardy, 1979; Trijonis et al., 1980), Although these elements do exist in the fine fraction to
a minor degree,-they are everywhere substantially enriched in the coarse fraction.
Occurrence of some elements in coarse particles is time and place dependent, though, and
Table 5-20 shows some data illustrative of this point. There appear to be substantial
differences across the country in the fraction of these elements occurring in coarse par-
ticles. The presence of local sources dominates both the mass and composition of coarse par-
ticles. However, Cooper and Watson (1980) have graphically demonstrated the similarity in
elemental distribution for a variety of coarse-particle sources as Figure 5-44 shows. Here,
the most that can be said is that rock-grinding operations produce remarkably similar
coarse-particle elemental compositions, whether the mechanical action is intentional or inci-
dental to other activities. Cement dust and limestones (not surprisingly) also have similar
elemental composition (Draftz, 1979).
Even greater evidence of localized influence on coarse-particle concentrations can be
seen with other elements. In the Smoky Mountains, titanium and chlorine are greatly enriched
in the coarse particles (Stevens et al., 1980) while in St. Louis, titanium is mainly a
fine-particle component and chlorine is about evenly distributed between coarse and fine
particles (Stevens et al., 1978; Akselsson et al., 1975). In the case of titanium, the
emissions from a paint plant greatly influence the fine-particle titanium (Rheingrover, L981).
Chlorine appears to originate in fine automotive particles at inland sites (Winchester et al.,
1967) but in coarse sea-salt particles near the coast (Hardy, 1979; Draftz, 1979).
A variety of carbon-containing species, particularly organics and carbonates, sometimes
can be found in substantial quantities in coarse particles. For example, De Wiest (1978)
found 30 to 50 percent of extractable organics in 2 to 10 \*m particles. Lewis and Macias
(1980) found 40 percent of the carbon (presumably mostly in organic compounds) in dichotomous
sampler coarse fractions in Charleston, WV, while Stevens et al. (1980) found about one-third
of the organic compound mass in the coarse fraction in the Great Smoky Mountains. Mueller et
al. (1970) were able to differentiate between carbonates and elemental carbon by acid
evolution of CO,, but this technique, unfortunately, has not been applied to coarse-particle
analysis. Considering that calcium carbonate has often been found as a major component of
urban coarse-particle samples (Graf et al., 1977; Draftz, 1979), it is surprising that direct
analyses for carbonate have not been reported. Elemental carbon appears to be uncommon in
coarse particles^ As mentioned previously, most water-soluble inorganic ions are found in
fine fractions.
5-100
-------
TABLE 5-20. COARSE PARTICLE SILICON, ALUMINUM, CALCIUM, AND IRON
Location
Charleston, WV
Smoky Mountains,
TN
New York, NY
Y1 Philadelphia, PA
1— j
2
St Louis, MO
Portland, OR
Glendora, CA
St. Louis, MO
Dates
08/25-
9/14/76
05/11-
05/19/77
09/21-
09/26/78
02/77
03/77
12/75
02/77
03/77
08/-
09/76
Coarse
mass
27.1
43
5.6
42.6
17.5
NA
27.6
NA
28.0
Mfl/m
Si
2.8
7.7
0.50
2.0
1.8
4.3
2.8
1.0
4.5
3
Al Ca
1.1 0.96
NA 2.2
0.20 0.32
0.84 1.15
0.64 0.94
8.7 1.9
1.2 0.76
0.3 0.44
1.2 2.8
Coarse/fine
Fe
0.59
1.4
0.12
0.96
0.69
1.0
0.95
0.36
1.2
Si
6.8
7
15
5.6
7
10
31
5.3
10
Al
15
NA
10
6.5
13
4.4
5.6
>6
6
mass ratios
Ca
9.7
7.4
20
3.2
6
16
11
4.5
21
Fe
4
4
4
2.5
3.2
3
5.0
3
4.4
Reference
Lewis and Mac i as,
1980
Camp et al . ,
1978
Stevens
et al . , 1980
Stevens
et al . , 1978
Stevens
et al . , 1978
Stevens
et al., 1978
Stevens
et al., 1978
Stevens
et al . , 1978
Dzubay, 1980
-------
ui
IU
o-
IUU
10.0
1.0
0.10
ft f|<{
U.U i
100
10.0
1.0
0.10
0 01
-
-Ni
mat
SOIL
Si
K C.T.
Mn
F*
•Ml
i UT
b
—
Zn
ROCK CRUSHER
d
Si
r-
Cl
fl
Fa
S Mr
JCI v_Crl
inn
IUU
10.0
z
111
g 1.0
Ul
Q.
0.10
n ni
^^
—
Cu
I'd*
Si
_ f
Naj
-
-
S
Cl
IUU
10.0
1.0
0.10
0.01
100
10.0
1.0
0.10
n ni
ROAD DUST
Si
~~ r™^ Fo
_pl ,iiflTI
S "•"•
Cl
Mn
/
—
Pb
^m
Zn
Al "
Nil
ASPHALT PRODUCTION
Si
— AJJ Fe _
I /*„
• j«*
S
Cl
Mn
/
—
^^ "
COAL PLY ASH
C. fL
n ''
—
Zn ~
ut, Mjj n pb
RI »J lB/n
Figure 5-44. Elemental composition of some coarse particle components.
Source: Cooper and Watson (1980).
5-102
-------
5.6,3 Evidence from Microscopical Evaluation of Coarse Particles
Efforts to understand the importance of coarse particles in air have been hampered by the
inability of simple chemical analyses (so very successful with fine particles) to reveal much
of their nature. However, estimates of mass balances have suggested for a long time that
locally generated coarse particles must constitute a substantial part of the suspended parti-
cle burden.
As an example, the most recent data from the EPA network of dichotomous samplers and
hi-vols could be interpreted as demonstrating significant amounts of particles larger than 15
pm in the air, if the difference between the dichotomous sampler total and TSP is taken as
representing supercoarse particles. Table 5-21 displays some of these data for selected sites
for illustrative purposes. Most sites have two-thirds or more of the TSP in coarse and
supercoarse particles. For the more arid and dusty parts of the country, rough estimates of
this kind and common sense have suggested to pollution control officials that TSP must be
dominated by locally generated coarse particles.
Since these larger objects can be readily inspected with an optical microscope, a sub-
stantial body of information has been accumulated by visually inspecting particle samples,
such as hi-vol filters or impactor stages. The largest of these studies involved evaluation
of 300 filters from 14 U.S. cities (Bradway and Record, 1976). Table 5-22 presents composite
analyses for all filters from each of the cities. A wide variation was found in these filters
ranging from virtually all dust in Denver and Oklahoma City to mostly dust with considerable
fly ash and soot in most of the industrialized cities. Chattanooga was anomalous, in that
extremely large amounts of plant materials were found, including pollen, fibers, fragments of
leaves, and other tissues.
Similar investigations were combined with emissions inventory, modeling, and control
studies in Phoenix, AZ, in 1977 (Richard, 1977; Richard and Tan, 1977; Richard et a!.,
1977a,b; Graf et al., 1977). In that city, 90 percent of the TSP was found to be mineral dust
apparently entrained in air by automotive traffic over 1100 miles of unpaved roads in the area
and by intense construction activities. Suck et al. (1978, 1979) found, through meteoro-
logical modeling, that very little motion of this coarse PM occurs in the wind. Since wind
velocities are characteristically low, agricultural influences are minor and coarse particles
stay more or less where they are generated.
Microscopic evaluations of Miami and St. Louis particles have been conducted both on
total filter samples and on impactor plates (Oraftz, 1979; Draftz and Severin, 1980). In
Miami, calcite (calcium carbonate) was the principal component of the coarse particles. There
was evidence that a small part of the calcite was recrystallized from ocean spray. However,
most of the calcite appeared to be roadway aggregate suspended in the air. There were also
significant quantities of halite (NaCl) and other trace elements characteristic of sea salt.
5-103
-------
TABLE 5-21. RELATIVE AMOUNTS OF FINE, COARSE, AND SUPERCOARSE
PARTICLES AT SELECTED SITES
Phoenix, AZ
El Paso, TX
Dallas, TX
Portland, OR
Los Angeles, CA
Akron OH
Philadelphia, PA
Hartford, CN
Fine
<2.5 pm
34
16
26
32
36"
49
51
34
Weight percent
Coarse
2. 515pm
6
51
10
64
31
26
32
32
Supercoarse
60
33
64
4
33
25
17
34
Note: The term supercoarse refers to the difference between the
hi-vol TSP concentration and total dicnotomous sampler concen-
tration.
Source: U.S. Environmental Protection Agency (1981).
5-104
-------
TABLE 5-22, FQURTEEN-CITY STUDY - MICROSCOPICAL IDENTIFICATION OF COARSE PARTICLES
COLLECTED IN URBAN ATMOSPHERES
Wt. % of Component
Location
Oklahoma City
Denver
Miami
St. Louis
Washington, DC
Baltimore
Birmingham
Philadelphia
Providence
Seattle
San Francisco
Cincinnati
Cleveland
Chattanooga
Minerals
88
8}
79
75
70
69
66
64
64
60
52
51
51
36
Combustion
products
8
7
9
21
23
25
22
33
22
27
29
44
40
35
Biological
material
1
1
1
>1
5
3
2
1
1
3
3
1
1
16
Miscellaneous
(rubber tire
debris)
4
11
12
4
2
3
10
2
13
10
16
4
8
13
Source: Bradway and Record (1976).
5-105
-------
The general picture from these studies is that coarse particles are contributed from
numerous local sources and vary dramatically from place to place. It is likely that dust and
roadway aggregate suspended by traffic are major sources of coarse particles. However, in in
industrialized cities especially, there is still some evidence of combustion source
contributions.
5.6.4 Fugitive Dust
In the discussion of coarse particles, evidence was presented that a substantial portion
of TSP, usually more than half, is accounted for by coa'rse'-br s'upercoarse particles, and a
great portion of this mass is mineral dust, also called "crustal material" in many of the
papers reporting chemical analytical data. There is growing opinion that this major component
of TSP is contributed almost entirely by agitation of soil in some way, and this component is
commonly called fugitive dust. One of the major sources for fugitive dust generation is
vehicle traffic; the motion of vehicles can reentrain silt (fine soil particles) that has been
deposited by settling from the air, washing from nearby areas in rain, or falling from vehicle
tires. All these mechanisms were significant in the Phoenix, AZ, study cited previously
(Richard, 1977; Richard and Tan, 1977; Richard et al., 1977a,b). These emission sources and
related ones are discussed in some detail in Chapter 4.
In an assessment of particle source influence on TSP in western States, reentrained dust
from paved and unpaved roads accounted for 10 to 75 percent of TSP emissions and 16 to 49 per-
cent of TSP concentrations at critical receptor locations in 20 inland western cities
(Axetell, 1980). In most of these locations, 40 to 60 percent of TSP emissions were from
roadways. One critical feature of this observation, apparently mainly derived From
microinventory studies around TSP monitors, was the traffic volume on unpaved roads. Cowherd
et al. (1979) reported emission factors from unpaved roads of about 300 grams/vehicle-km
traveled. Consequently, even rather small traffic volumes can generate substantial TSP
contributions. There are very large State-to-State variations in both the remaining number
and length of unpaved roads and the traffic volume such roadways carry. Carpenter and Weant
(1978) analyzed the available information on roadway use and found that while unpaved roads
were fairly common in the western mountains, deserts, and Great Plains, other areas of the
country, the Southeast and New England especially, also had many miles of unpaved roads and
substantial roadway contributions to TSP. In industrial areas, truck traffic over access
roads can be a major source of TSP emissions (Cowherd et al., 1979).
The wind alone can be a major source when velocities are high and the soil aggregates are
small. For example, Gillette (1978) estimated soil fluxes for six test soils in a wind
tunnel. He concluded that windspeed and crust play major roles in wind entrainment. A sur-
face crust effectively eliminated fine-particle entrainment and greatly reduced coarse
particle entrainment.
Wilson et al. (1979) found that car and truck traffic produced large amounts of dust on
unpaved mining roads in northeastern Minnesota. Particle sizes were mainly in the 6 to 30 urn
5-106
-------
range near the roadway, but large particles were found at about one-fifth the roadside level
400 or 500 meters downwind. There was visual evidence of dust coating roadside foliage and
gusts of wind caused major short-term pulses in particle concentrations downwind of the road.
Davidson and Friedlander (1978) measured deposition of coarse particles on Avena, the common
wild-oat grass of the Far West. Dry deposition on the stems of such plants was reported to be
a significant removal mechanism for particles larger than about 7 urn.
Reentrainment of road dust has been found a major source of coarse particles in central
business districts. In a study of several sites in Philadelphia, Record and Bradway (1978)
concluded that entrainment of dust from roadways contributed the majority of street-level
coarse particles and very significant levels at rooftops, 11 meters above the street.
Rainfall, if there was enough of it, reduced the dust levels significantly (e.g., about 20
percent). However, attempts to flush the street with water redistributed the fine particles
and increased the observed coarse-particle level.
In a study of one site in Massachusetts, Record et al. (1979) found coarse-particle
levels highly correlated with traffic volume as shown in Figure 5-45. In this study, very
large contributions of roadway salt, used for winter snow control, were found in the coarse
particles.
Yocom et al. (1981) estimated, by analysis of a variety of records, the contribution of
fugitive dust to areawide particle burdens in Allegheny County, PA. They found both
industrial sources and roadways to be significant contributors, though in widely varying pro-
portions. In the 12 study sites, roadway dust contributed from a^low of 4 percent to a high
of 45 percent of annual geometric mean TSP. In most sites, 15 to 20 percent of these
particles came from traffic. Industrial fugitive particle emissions were more significant in
this area, although the general range was similar, 5 to 40 percent of the annual level. In
most sites, industrial fugitive dust contributed 20 to 30 percent of the TSP and was greater
than roadway dust. These two sources, together with the general area background, accounted
for 80 to 90 percent of the TSP burden in Allegheny County.
Clearly, fugitive dust is a major contributor to TSP in most U.S. monitoring sites.
There is evidence that unpaved roadways and commercial streets can be major sources. Since
reentrained dust appears to be mainly coarse mode particles, monitor-siting considerations,
and especially monitor height or slant distance relative to roadways, can markedly alter the
observed TSP level (Pace et al., 1977; Record et al., 1979).
How important this fugitive dust might be in assessing exposure relative to potential
effects is not clear. Of those cited in this section, only the Record et al. (1979) study
reported particle size data; further, the whole issue of the relationship of outdoor concen-
trations to exposures is under serious question. (Vide infra, §5.8)
It is also interesting to note in Table 5-22 that there is always some PM of biological
origin in atmospheric particles. Occasionally, especially during pollen season, this material
can account for a significant fraction of the coarse-particle mass (Draftz et al., 1980).
5-107
-------
CO
Ul
o
u.
o
DC
Ul
a
2
z
2000
1800
1600
1400
1200
1000
800
600
400
200
0
I 111 I I I I I
I I
TRAFFIC VOLUME
8 10 12 14 16
TIME OF DAY (START HOUR)
18
20
100
90
80
70
60
50
40
30
20
10
0
22 24
Ul
o
I
Figure 5-46. Diurnal variation of particle concentrations and Plymouth Avenue traffic
volume in Fall River, Mass,, during March through June 1979 (weekdays only), shows
contribution from reentrained particles.,
Source: Record et al. (1979).
5-108
-------
Draftz et al. (1980) also identified cornstarch as a major contributor to particle mass in
some nonattainment areas. These authors also found a few other reproductive stages of
organisms such as conidia spores. It has been hypothesized that bacteria occur in coarse
aerosols, perhaps even legionella pneumophilia (Eraser and McDade, 1979). However, the
presence of airborne infectious organisms has usually been deduced by disease incidence rather
than by direct measurement, and quantitative techniques are largely lacking. Considering the
importance of viable particles in allergies, for example, this area could be a desirable one
for future research.
5.6.5 Summary
The wind, traffic, construction, mining, and general industrial activity are the major
causes of coarse particles suspended in the air. Dry climates, intense construction activity,
lack of paving, and salt from icy streets or the sea can all be contributing factors. The
quantitative assignment of particular sources to the coarse and fine particle burdens has been
addressed in a cursory and introductory fashion in this section. In the next part, more
formal systems for this source apportionment will be addressed.
5.7 SOURCE-APPORTIONMENT OR SOURCE-RECEPTOR MODELS
For quite a long time, the goal of quantitatively determining the contributions of
particle sources to ambient PM concentrations has been pursued. Recently, Cooper and Watson
(1980) and Gordon (1980) reviewed the current status of calculational systems or models to
estimate source contributions to atmospheric particle concentrations. Cooper and Watson
described several methods in a systematic sense, and Figure 5-46 shows their analysis of
models capable of yielding at least semi-quantitative information. Many microscope-based
conclusions were discussed in the previous section and the work of Yocom (1981) is basically
an example of series analysis. However, chemical mass balance and multivariate models have
been used quite effectively recently and a few examples of these approaches are cited below.
The results from three cities (St. Louis, MO, Denver, CO, and Portland, OR) illustrate the
contrast in fractional contributions of PM from different sources.
In analyzing the St. Louis Regional Air Pollution Study (RAPS) dichotomous sampler data
by x-ray fluorescence, Duubay and Stevens (1975) found 75 percent of the zinc, sulfur,
bromine, arsenic, selenium, and lead occurred in the fine particles and at least 75 percent of
the silicon, calcium, titanium, and iron in the coarse fraction. Using groupings of elements
to represent sources, Dzubay (1980) postulated the sources making fractional contributions to
2-month summer mean concentrations at several sampling locations in St. Louis. Approximately
50 to 70 percent of the concentration of fine particles was made up of ammonium sulfates. The
next largest identifiable source was motor vehicle emissions, followed by shale and other
sources.
5-109
-------
OPTICAL
MICROSCOPIC
METHODS
SEM
AUTOMATED
SEM
ENRICHMENT
FACTORS
TIME SERIES
ANALYSIS
CHEMICAL
MASS
BALANCE
SPECIAL SERIES
ANALYSIS
ADVANCED
MULTIVAR1ATE DATA
ANALYSIS METHODS
Figure 5-46. Types of receptor source apportionment models.
Source: Cooper and Watson (1980).
5-110
-------
Pace and Meyer (1979) resolved the constituents found in the St.Louis dichotomous sampler
data to demonstrate the relative contributions of sources to the urban and nonurban concentra-
tions (see Figure 5-47). As might be expected, the vehicle emission component was much
2-
smaller in the rural samples than it was in the urban samples. The SO- fraction in the
nonurban sites made up a larger proportion of the total fine particulate mass than did the
2-
SO^ fraction in the urban samples. The crustal component and the "nondescribed components
remained about the same in both sets of data. Looking at the coarse fraction, it is
interesting that the crustal component was much larger in the urban sites than it was in the
nonurban sites.
The monthly averages of size fractionated Denver aerosol mass were compared for 2 months,
2-
January and May 1979, by Dzubay (1980). The SO^ component was smaller than in St. Louis and
the motor vehicle component was larger (see Figure 5-48). The winter concentrations were
higher for both the fine and coarse fraction. Much of this difference appeared to be in the
excess carbon and NO., component. The coarse fraction contained road salt particles in the
winter. Presumably, accompanying sand could account for part of the unidentified mass. In
2-
Denver samples, unlike east.ern aerosol samples, the summer SO* concentrations appeared to be
lower than in winter. The asterisks (*) indicate that some or all of the component could be
due to excess carbon. In this case "excess carbon" was determined by the amount of carbon
observed less the amount accounted for by the identified sources. In the winter, the carbon
3 •
concentration in the fine fraction was 17 ug/m . It was not determined for the coarse
3
fraction. In May, the unfractionated carbon was 7.2 ug/m . Wood burning and vehicle exhaust
are believed to be the important carbon sources in Denver.
Chemical element balance techniques were applied to TSP and fine-fraction aerosols col-
lected in Portland, OR, in a year-long study. Cooper et al. (1979) describe the experiment
and results. Figure 5-49 summarized the resulting source allocation. As in several other
findings, soils and road dust were important components of TSP. The study revealed that
burning vegetation was the second most important source, contributing almost-15 percent of the
TSP mass and 20 percent of the .fine-particle mass. Sulfate was not the most abundant
component of the fine-particle mass. In fact, it was only 8 percent of the mass, fourth after
auto exhaust, volatilizable carbon, and aerosols from vegetation burning. The study showed
the contribution of residential wood burning to ambient aerosol concentrations.
Studies resolving the source components are great aids in resolving the fractional
contribution of local versus distant sources. Resolution of this question may await the
application of receptor modeling to other cities and other regions of the country.
Source-receptor techniques, which rely on measurements of PM and composition, can often
be combined with studies of wind direction or of synoptic air mass motion to further elucidate
source effects on ground pollutant levels. For example, Rodhe et al. (1972) studied the soot
2-
and SO, concentrations at four coastal sites in Sweden. By combining wind direction data
5-111
-------
•FINE
COARSE
(27 tig/m3)
RAPS URBAN SITES
(103,10S, 106,108,,112)
*2-MONTH AVERAGE CONCENTRATIONS
FINE
RAPS NON-URBAN SITES
(115,118,120,122,124)
COARSE
(21 jiB/m3)
Figure 5-47. Source contributions at RAPS sites for July-August 1976
estimated by chemical element balance.
Source: Pace and Meyer (1979).
•5-112
-------
FINE FRACTION
COARSE FRACTION
MOTOR VEHICLES I (NH4»2SO4
T OM%
EFUSE 0,2%
LT1%
LIMESTONE
105%
ALE 22%
__(NH4)2S04 2.6%
NO3 OJ%
T 1%
RE FUSE 0.8%
LIMESTONE 2.6%
JANUARY
JANUARY
27 MD/m3
-------
(VEGETATIVE BURN
(14.6%)
SOIL AND ROAD DUST
[39.0%7
VOLATILIZABLE
CARBON (8.1%)
RESIDUAL OIL
(0.8%)
MARINE
(3.8%)
SOIL AND ROAD DUST
(43%)
NONVOLATtZABLE CARBON
(4.0%)
PRIMARY INDUSTRIAL (3.0%)
• STEEL PRODUCTION (1.0%)
• ALUMINUM PRODUCTION (0.72%)
• HOG FUEL BOILERS (0.48%)
« SULFIT1 PROCESS (0.39%)
.UNIDENTIFIED (21.3%)
{NH4, H2O, ate)
TOTAL
INONVOLATILIZABLE CARBON
<2.2%T
(NH4,H2O),8tc)
RIMARY INDUSTRIAL (4.9%)
* CARBIDE FURNACE, PROCESS EMISSIONS (2.0%)
* ALUMINUM PRODUCTION (1.35%)
• STEEL PRODUCTION (0 34%)
* HOG FUEL BOILERS (0.22%)
• SULFIDE PROCESS (0 18%)
» SULFITE PROCESS (0.18%)
FERROMANGANESE PRODUCTION
CO.18%)
FINE
ESIDUAL OIL
(1.4%)
MARINE (3.2%)
CARBIDE FURNACE, PROCESS EMISSIONS (0.6%)
Figure 5-49. Aerosol source in downtown Portland, annual stratified arithmetic average. Does not
include the 17%, on the average, of material collected with the standard hi-vol sampler which was not
collected and characterized with the ERT-TSP sampler. Volatilizable and non-volatilizabie carbon are
operational definitions which approximately correspond to organic and elemental carbon, respectively.
Source: Cooper at al. (1979a).
5-114
-------
2~
with observations of high SO, and soot loadings, these authors deduced the existence of
long-range transport of air pollutants from industrialized areas of northern Europe, Trijonis
(1980) used similar analysis in detailing local source contributions to PM at specific
monitoring sites in St. Louis.
Samson (1981) developed a more elaborate model based mainly oh the trajectory analysis
procedure of Heffter et al. (1975). This model was used to plot area source contributions to
O—
ground-level SO, values reported in the literature for seven rural sites in the northeastern
United States. Samson concluded that high concentrations of sulfates appear to be
consistently associated with upstream (upwind) stagnation. Further he found that
incorporation of the magnitude of S09 emissions and the upstream mixing height into the model
2~
did not improve the correlation between measured SO, and upwind stagnation as estimated by
the inverse of windspeed.
This class of models appears to offer some promise for future applications, but like the
somewhat more complex Lagrangian trajectory models discussed in Chapter 6, it has some
problems. Samson (1981) pointed out that a complete investigation of sulfur budget (or that
for any other pollutant, for that matter) would require measurements of the vertical profiles
of pollutants over a very wide area. "Clearly, these data requirements are not likely to be
met by experiments in the near future. The reader is directed to Chapter 6 for further
discussion of deterministic models and their data requirements, successes, and shortcomings.
5.8 FACTORS INFLUENCING EXPOSURE
5.8.1 Introduction
To this point only outdoor concentrations of S0? and PM have been considered in the
present discussion. Outdoor concentrations are of major concern in estimating air pollution
effects on visibility, and ecological and materials damage. However, people spend the
majority of their time inside buildings or other enclosures; they breathe mostly indoor air
and, therefore, indoor concentrations dominate average exposure. To the extent that indoor
concentrations are different from the outdoors, population exposures are different from those
estimated by outdoor monitors.
In the United States the population is highly mobile. Many persons in their daily travel
pass through areas of both high and low pollution levels within a city. Others work or play
outdoors to a greater degree than the general population. Therefore, individual exposures to
SQy and PM vary more widely than measurements from stationary outdoor monitors suggest
(Spengler et al.., 1979).
Furthermore, individual variations in respiratory anatomy, illness, or smoking habits can
exert important influences on the dose of a pollutant retained by individuals receiving the
same exposure. For example, Cohen et al. (1979) found that smokers retain test particles
longer than nonsmokers.. Figure 5-50 presents the results of a study of 12 subjects, 3 of
5-115
-------
100
468
POST INHALATION, months
10
12
Figure 5-50. Retention of Au1 labeled fe^O^ particles from
human lungs; comparison of 9 non-smoking subjects with
three smoker subjects.
Source: Cohen et al. (19791.
5-116
-------
whom were smokers. Ten months following exposures to a known quantity of metallic dust, the
nonsmokers had cleared 85 to 95 percent of the dust from their lungs. At the same time,
smokers had retained about half of their original dose Unfortunately, there are few other
studies that help in understanding these individual variations.
In this section the two major factors influencing human exposure to SO, and PM, indoor
exposures and activity variations, are presented because they are important in understanding
health effects. First the systematic differences between indoor and outdoor concentrations of
SO- and fine and coarse particles are discussed. Then the limited evidence is presented for
varying exposures of individuals depending on their activities.
5-8-2 Indoor Concentrations of SulfurDioxide
Indoor concentrations of SO, are invariably lower than outdoors, usually by a factor of
two (Spengler et a!., 1979). Since indoor sources of SO- (such as matches, natural gas odor-
ants) are usually negligible, virtually all SO, indoors originated outdoors. Lower indoor
concentrations are commonly attributed to SO- removal by contact with wall coatings,
furniture, flooring and carpets, air conditioning filters, and the like.
Removal of SO, inside chambers and rooms has been shown to be a function of the material
present and the relative humidity (RH). Cox and Penkett (1972) measured the decay rate of SO-
inside containers. Reaction rates were found to be first order in SO,,, and irreversible
absorption occurred on the walls. The removal rates were very sensitive to the RH. As
relative humidity increased, so did SO, removal, approaching a maximum value at slightly above
80 percent RH (Cox and Penkett, 1972). Spedding studied SO, sorption by indoor surfaces
(Spedding and Rowlands, 1970; Spedding, 1970; Spedding et al., 1971). The surface finish on
wallpaper affected sorption rates of SO-. Conventional wallpaper showed better uptake than
polyvinyl chloride (PVC) wallpaper. Hard woods sorbed SO, better and to a greater depth than
did softer woods. Sulfur dioxide sorption was also measured for leather surfaces. The rate
of absorption seemed to be influenced by the tanning process and the dyes used.
Walsh et al. (1977) measured sorption of SO, by typical indoor surfaces including wool
carpets, wallpaper, and paint. Absorption rates, as measured by deposition velocities, were
lower for carpets with an acidic pH than those which were either neutral or alkaline.
Sorption of SO, appeared to be irreversible. When carpets were preexposed to an SO,
3
concentration equivalent to 27 years of exposure at 30 yg/m , the amount of S02 uptake was
reduced by a factor of three. Fresh emulsion paints had the highest deposition velocity or
SO, absorption rates, and vinyl wallpaper had the lowest. It was concluded that the most
effective sorbing materials likely to be present inside homes are cellulose wallpaper,
furniture fabrics, and wool carpets. Therefore, most studies report lower levels of SOg
indoors than outdoors.
Anderson (1972) reports that indoor SO- concentrations averaged 51 percent of the outdoor
concentrations over a 7.5-month period of paired 24-hour sampling inside and outside a single
room. The correlation coefficient was only 0.52 (Anderson, 1972). Biersteker et al. (1965)
5-117
-------
analyzed over 800 paired samples from the living rooms and exteriors of 60 Rotterdam homes.
Indoor SO* levels averaged 20 percent of the outdoor levels and were lower for newer homes
than for older homes. This may imply longer air turnover times in the newer homes and/or a
"fresher" surface area for S02 absorption (Biersteker etal., 1965). Weatherly (1966)
measured S02 and smoke levels inside and outside a building in central London in early 1960.
Indoor SO- levels were always lower than the corresponding ambient conditions, averaging 40
percent. Spengler et al. (1979) reported on paired SO, monitoring inside and outside at least
3
10 homes in each of 6 cities. Figure 5-51 displays the annual S02 concentration in pg/m
averaged across each community's indoor/outdoor network of monitors (May 1977 to April 1978).
The cities were: Portage, WI; Topeka, KS; Kingston and Harriman, TN; Watertown, MA;
St. Louis, MO; and Steubenville, OH. Where ambient levels were high, the indoor
concentrations were 30 to 50 percent of the ambient levels. In Kingston, many of the indoor
levels were less than the minimal detectable level and were averaged in as zeros. This was
not done for Portage and Topeka, where ambient levels were very low; hence the indoor SOj,
levels in these cities appear to be reduced by only 20 percent of the outdoor concentrations.
The seasonal indoor/outdoor pattern for each city depends on the S02 sources in each city
and the use of air conditioning. These differences can be seen by comparing the monthly mean
indoor and outdoor S02 concentrations for Watertown and Steubenville, as shown in Figures 5-52
and 5-53. In Watertown, SO, was primarily derived from sulfur-laden fuels. The ambient
levels rose in the winter as more residual and distillate oil was used for space heating. The
indoor/ outdoor ratios became small because homes were sealed more tightly. In the summer,
ambient levels decreased, but the indoor/outdoor ratio approached unity because of increased
ventilation. In Steubenville, the summer SO™ levels were not substantially reduced from
winter values, since residential and commercial space heating was not the primary source of
SO, in this area. Yet the reduced indoor levels continued since more air conditioning was
used in Steubenville (50 percent of homes sampled). Even in homes without air conditioning,
summer levels were reduced by 30 to 40 percent.
While most information supports the idea of lower indoor S02 concentrations, exceptions
are known. Yocom et al. (1971) found one home, heated by a leaky coal furnace, in which the
indoor SO, level was periodically 10 times the outdoor level. Bierstecker et al. (1965) also
found leaking flue gas contributions indoors.
However, the principal body of evidence suggests that indoor exposures are generally
about half that found outdoors. Consequently, highest exposure levels are likely to be
incurred by people who spend time outdoors near local S02 sources.
5.8.3 Particle Exposures Indoors
5.8.3.1 Introduction—Available data on indoor particle levels were collected by a wide
variety of measurement procedures ranging from dustfall to condensation nuclei counting
Earlier in this chapter and in Chapter 3, it was noted that the various particle measurement
5-118
-------
60
SO
40
I,
1,30
O
20
10;
|§%f OUTDOOR
j [ INDOOR
»p <0.05
~ m
WA I ifM^-. m
vm I ^h~i w "~i
1
1
M^HMI
i
i
I
=v. ._
MMPMI
PORTAGE* TOP1KA' KING-
STON*
WATER- ST. LOUIS* STEUBEN-
TOWN* VILLE*
Figure 5-51. Annual sulfur dioxide concentrations averaged across each community's
indoor and outdoor network (May 1977 — April 1978).
Source: Spengler et al., 1979.
5-119
-------
"i
10b
96
84'
72
60
48
36
24
12
0
WATERTOWN
O INDOOR
O OUTDOOR
NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP QCT NOV DEC JAN FEB MAR APR
1976 1977 1978
Figure 5-52. Monthly mean SO? concentrations averaged across Water-town's indoor and outdoor
figure o-o*:. mommy mean auo concei
network (November 1976 - Aprir1978).
Source; Spongier et al. (1980).
5-120
-------
i
&
i
108
96
84
72
60
48
36
24
12
0
I I i I
STEUBENVILLE
NOV DEC JAN FEBMAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR
1976 1977 1978
Figure 5-53. Monthly mean SOo concentrations averaged across Steubenville's indoor "and outdoor
network (November 1976 —April! 978).
Source: Spengler (1980).
5-121
-------
procedures have definite particle size biases. For example, dustfall and TSP mass measure-
ments are dominated by coarse particles, while light scattering and nuclei counting, on the
one hand, and smokeshade and coefficient of haze on the other, are better measures of fine
particles and participate carbon mass (one fine particle component), respectively.
Evidence from these techniques has produced a consistent view of indoor and outdoor
particle concentrations and sources, and this view is presented below by separate con-
sideration of coarse- and fine-particle studies. Table 5-23 summarizes available studies by
particle measurement technique.
5.8.3.2 Coarse-Particle Concentrations Indoors—Particles larger than 5 or 10 urn tend to
settle from the air, and two studies, using dustfall collection techniques suggest that these
particles are greatly reduced indoors. Whitby et al. (1957) studied dustfall in offices,
laboratories, and homes. Average indoqr dustfall was only 15 to 20 percent of the outdoor
level. No significant differences were found among residential or business locations.
Schaefer et al. (1972) found indoor dustfall about one-eighth outdoor values in a study of 30
residential sites in four cities. There was little correlation between indoor and outdoor
levels. Dustfall was higher in homes where windows were open.
Yocom et al. (1971) studied TSP in public buildings, offices, and homes using a
scaled-down version of the hi-vol sampler. As mentioned previously, the mass of such filter
samples contains both coarse- and fine-particle fractions. Indoor levels were about half out-
door levels on the average. In summer and fall, private homes had almost the same interior
daytime TSP values as those found outside although night interior levels were much lower than
outside. In the same study, indoor/outdoor ratios in air-conditioned office buildings
differed seasonally. In summer and winter, indoor TSP was about half the outdoor values, but
in fall, when increased volumes of outdoor air were used in air conditioning, indoor and
outdoor TSP values were about equal.
Yocom et al. (1970) also obtained some cascade impactor size distributions of indoor and
outdoor particles in six structures in the Hartford, Conn., area in the fall and winter. The
mass of particles larger than 2,5 pm was always greater outdoors than indoors. However, the
mass of sub-2.5 fjm particles was mostly greater indoors than outdoors, and the indoor/outdoor
ratio varied from 0.63 to 2.6. In a subsequent report, Yocom et al. (1971) reported substan-
tially increased organic particle levels indoors, confirming similar findings by Goldwater et
al. (1961). This result was attributed to smoking and cooking, indoor activities that could
also increase fine-particle mass.
Alzona et «-al. (1979) reported elemental analyses for calcium and iron, normally
coarse-particle components, and for zinc, lead, and bromine, components of fine particles. In
these studies, an experimental room was cleared of airborne PM and then allowed to come to
equilibrium under controlled penetration of outdoor ambient air. Experiments were carried out
with windows "cracked" open and wide open and with windows artd/or room surfaces covered with
plastic sheets. Filter samples drawn throughout the experiment were analyzed by X-ray
5-122
-------
TABLE 5-23. SUMMARY OF INDOOR/OUTDOOR (I/O) PM MONITORING STUDIES BY METHOD
Method
Dustfall3
Total sus-.
pended PMD
n
-t
o
o
Smoke
Author
Whitby'et al.
(1957).
Schaefer et
al. (1972)
Yocom et al.
(1971)
Whitby et al.
(1957)
Goldwater
et al.
(1961)
Location
Minneapolis
Chicago,
Washington,
Atlanta,
Austin-San
Antonio
Hartford, CT
Minneapolis
Louisville
New York
Building
type
Residential
Lab & office
Residential
Public
Office
Residential
Lab & office
Residential
Lab & office
Residential
Lab & office
Residential
Number
of
sites
12
30
2
2
2
__
_-
__
--
12
18
Month
or
season
Annual
Annual
Mar.-
Aug.
Summer
Fall
Winter
Summer
Fall
Wi nter
Summer
Fall
Wi nter
__
__
__
--
Feb.-
Mar.
Feb,-
Mar,
Sampling
period
_«
— ~
—
12 hr
12 hr
12 hr
12 hr
12 hr
12 hr
12 hr
12 hr
12 hr
__
--
__
--
__
__
Mean
Indoor
0.54
0.47
0.44
59
58
63
53
31
42
62
53
43
42
46
101
121
149
164
Outdoor
2.86
2.86
3.48
111
119
277
108
50
122
67
77
97
74
74
124
124
263
179
Number
of I/O
samples ratio
0.19
0.16
26 0.12
0.53
0.49
0.23
0.49
0.60
0.35
0.94
0.69
0.44
0.57
0.62
0.81
0.98
12 0.57
18 0. 91
-------
TABLE 5-23 (continued)
Method Author
Jacobs et al.
(1962}
Weatherly
(1966)
Biersteker
et al.
(1965)
Berdyev
et al.
(1967)
1 Anderson
£ (1972)
Respirable . Binder et al.
particless0 (1976)
Particle Parvis
counts0 (1952)
Ishido
et al.
(1956)
Location
New York
London
Rotterdam
Dushambee
U.S.S.R.
Arhus,
Denmark
Ansonia, CT
Italy
Osaka
Building
type
Residential
Office
Residential
Residential
1st floor
Residential
2nd floor
Classroom
Smoking
homes
Non-
smoking
homes
Residential
Apartment
Number
of
sites
17
— «.
1
60
1
1
1
11
9
5
1
Month
or
season
Apr.-
May
Jan.-
Mar.
Winter
Summer
Summer
Sept. -
Apr.
Sept. -
Dec.
Sept.-
Dec.
—
— •*
March
May
June
Nov.
Sampling
period
-_
1 hr
24 hr
—
24 hr
24 hr
24 hr
—
~—
24 hr
24 hr
24 hr
24 hr
Mean
Indoor
239
195
153
1270
660
27
132
93
45.7
1000
1287
978
738
--
Outdoor
226
205
184
960
960
34
58
58
97.6
1036
1528
1047
752
1897
Number
of
samples
17
800
8
9
150
11
9
__
"""*
—
—
—
-—
I/O
ratio
1.06
0.95
0.84
1.32
0.60
0.81
2.28
1.60
0.47 CN
0.97 PC
0.84
0.91
0.98
—
-------
TABLE 5-23 (continued)
1
I—«
r\>
cji
Method Author Location
Ishido Osaka
(1959)
Jacobs New York
et al.
(1962)
Megaw England
(1962)
Lefcoe and
Inculet (1975)
Coefficient Whitby et al. Minneapolis
of haze9 fl9^7)
Louisville
Carey et al. Cincinnati
(1958)
Shephard Cincinnati
(1959)
Number
Building of
type sites
Apartment
Residential
Hospital
School
Office & lab
Homes
Test building
Homes
(AC & ESP)
Residential
Lab & office
Residential
Lab & office
Residential
Residential
1
1
1
I
12
18
1
2
—
—
--
— -
__
9
9
9
9
—
—
—
—
9
9
9
9
Month
or Sampling
season period
__
—
—
— — ~
—
—
—
Annual
—
—
—
--
--
Oct.-
Dec.
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Mean
Indoor
706
662
1611
2382
424
705
—
510d
156
0.6
1.0
2.4
2.8
2.1
2.2
2.0
1.6
1.6
--
0.8
0.8
1.3
0.8
1.5
1.7
1.8
Outdoor
619
678
1595
2346
481
472
—
227f
•59e
1.05
1.05
2.6
2.6
3.8
2.7
2.3
1.8
1.7
1.0
0.9
0.8
0.8
0.5
0.9
1.1
1.5
Number
of I/O
samples ratio
1. 14
0.98
1.01
1.02
0.88
1.49
5 0.66
1.46
2.60
0.57
0.95
0.92
1. 06
0.55
0.86
0.89
0.85
0.94
-— -~
0.88
1.00
1.63
0.80
1.15
0.89
1.06
-------
TABLE 5-23 (continued)
ro
en
Number
Building of
Method Author Location type sites
Yocom et al. Hartford, Public
(1971) CT
Office
Homes
Piezoelectric Repace and Metropolitan Public
microbalance Lowrey (1980) Washington, (no smoking)
DC
Public
(smoking)
2
2
2
3
13
Month
or
season
Summer
Fall
Winter
Summer
Fall
Winter
Summer
Fall
Winter
March-
June
March-
June
Sampling Mean
period Indoor
0.32
0,33
0.36
0.29
0.20
0.37
0.41
0.28
0. 32
20 min. 29
to,:
55r
2 min (I) 86
5 min (0) to-
697T
Outdoor
0.36
0.34
0.51
0.41
0.26
0.54
0.42
0.30
0.39
40
to.
55T
22
to.
63T
Number
of
samples
__.
—
-—
-_
--
-_
-_
—
--
3
4-25 (I)
13 (0)
I/O
ratio
0.90
0.97
0.69
0.70
0.78
0.88
0.98
0.93
0.82
0.66
to
1.38
1.56
to
11.62
?Measured as g/mVmonth.
Measured as pg/m3.
^Measured as number/cm3.
Particles >0.3 urn.
fParticles >0.5 pm.
Particles >1.0 pm.
Measured as COH/1000 linear ft.
Note: 1 ppm S02 = 2620 ug/m3.
1 ppm N02 = 1885 ug/m3.
-------
excitation for elements of known outdoor origin (Fe, Zn, Pb, -Br, Ca). Within several hours,
equilibrium was reached in which the indoor/outdoor ratio was typically 0.3 (see Tables 5-24
anti 5-25). On the basis of the indoor/outdoor element ratios, they conclude that remaining
indoors with doors and windows partially closed reduces outdoor dust exposure by two-thirds.
The indoor/outdoor ratios for the coarse-particle components, calcium and iron, were lower
than for the fine-particle components zinc, lead and bromine. Therefore, it appears that
tracer components of coarse particles do not penetrate any of these structures as readily as
the fine components.
5.8.3.3 Fine Particles Indoors—In addition to the cascade impactor studies mentioned earlier
in conjunction with the coarse-particle discussion, there have been several recent reports of
sub-3.5 |jm particle mass measurements indoors and outdoors.
Repace and Lowrey (1980) reported even larger indoor/outdoor contrasts, which they
attribute to smokers. Using a piezoelectric microbalance, they sampled fine particles (0.01
to 3 jjm) inside and outside a variety of public places, many of them restaurants. In the
absence of smokers, the indoor/outdoor ratios were in the range of 1:1, comparable with ratios
reported by investigators using other methods. With smokers present, indoor/outdoor ratios
ranged to over 11:1 (Table 5-23).
Binder et al. (1976) used hi-vol air samplers outdoors and personal samplers equipped
with a 3.5 urn cutoff device. The personal monitors were carried by school children who spent
60 to 80 percent of their time indoors. In homes where there were smokers, the indoor
fine-particle mass was almost twice the outdoor TSP.
Spengler and Dockery have measured indoor and outdoor levels of sub-3.5 urn particulate
mass using cyclone-equipped membrane filter samplers in the same six cities noted in Section
5.8.2 (Spengler et al., 1981; Dockery and Spengler, 1981; and Dockery, 1979). Figure 5-54
presents annual average values for all sites in the six cities. In all cities except
Steubenville, Ohio, the indoor fine-particle level was higher than outdoors. Steubenville, an
industrialized community, also had the highest annual average outdoor level. Table 5-26
presents arithmetic averages for all homes in this study stratified by numbers of smokers in
the household (Dockery, 1979). It is apparent that in the absence of smoking, indoor and out-
door levels of fine-particle mass are almost the same. However, smoking contributes very
significantly to indoor levels. Dockery (1979) calculated that a one pack/day-smoker con-
tributes about 18 ug/m to inside fine-particle mass, and this level is increased by the use
of air conditioning, presumably because of recirculation, to 43 pg/m .
5-127
-------
TABLE 5-24. MEASUREMENTS IN PRINCIPAL ROOM OF STUDY3
Case
J
K
L
M
N
P
Number
of runs
3
2
1
2
6
3
Conditions
Normal
Plastic over windows
Window wide open
Window cracked open
All surfaces plastic covered
All but windows plastic covered
Ca
0.10
0.10
0.52
0.20
0.02
0.10
Fe
0.17
0.15
0.81
0.16
0.12
0.15
I/O
Zn
0.52
0.71
0.93
0.69
0.24
0.58
Pb
0.49
0.17
1.2
0.55
0.15
0.57
Br
0.33
0.17
1.0
0.53
0.20
0.32
Source: Alzona et al. (1979).
The study room in all these c;
listed below in Table 5-25 as case J.
a 2
The study room in all these cases was the same 12 m , old university building room
TABLE 5-25. MEASUREMENTS IN VARIOUS CLOSED ROOMS
Number
Number of
Case of runs Type of room windows Ca
A
B
C
D
E
F
G
H
I
J
Average
1
1
1
1
1
1
1
2
2
3
10 m2,
50 m2,
30 m2,
20 m2,
20 m2,
new univ.
old univ.
bedroom,
bldg.
bldg.
tight home
attic, tight home
bedroom,
leaky home
Chevrolet Vega
Datsun
20 m2,
20 m2,
»| *} m*"*
(except A)
440
old chem.
new univ.
old univ.
lab
bldg.
bldg.
0
2
8
2
2 0.05
6
6 ""*""
(3) 0.08
3 0.15
(3) 0.10
0.10
Fe
<0
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.1
33
27
10
33
27
09
13
54
17
24
I/O
Zn Pb
-------
TABLE 5-26. RESPIRABLE PARTICULATE CONCENTRATIONS OUTDOORS AND INDOORS
BY AMOUNT OF SMOKING3
Location
Outdoor
Indoor, no smokers
Indoor, 1 smoker
Indoor, 2+ smokers
Number of
homes
74
38
22
9
Number of
samples
2S98
1328
712
323
Mean concentra-
tion, ug/m3
22.3
24.0
42.8
74.5
Standard
deviation of
home means
12.7
11.4
22.2
37.9
Data averaged across network of samples in six communities for 1977.
Source: Dockery (1979).
2-
Spengler et al, (1981) also reported that indoor SO. (a fine-particle component) levels
were significantly lower than outdoor levels, but that gas cooking stoves can increase levels
by about 1 pg/ra (Dockery and Spengler, 1981). Yocom et al. (1971) found that lead
indoor/outdoor ratios were greater than TSP ratios. Taken with the previously mentioned
Alzona et al. (1979) study, it appears that most fine particle components analyzed are found
in high proportion indoors.
A number of studies have reported indoor measurements of smokeshade or COM, both
estimates of fine-particle carbon. Whitby et al. (1957), Shephard et al. (1958), and
Weatherly (1966) all found smokeshade values inside and outside buildings to be nearly equal.
Goldwater et al. (1961) found indoor smoke about 75 percent of the outside levels in 30 New
York sites, but that difference was not significant. Jacobs (1962) found no indoor/outdoor
differences in a followup New York study. Anderson (1972) found nearly equal and highly
correlated smoke values in a classroom and outdoors in Denmark. Biersteker et al. (1965), on
the other hand, found no correlation between indoor and outdoor smoke in a winter study of 60
homes in Rotterdam. Whitby et al. (1957) and Yocom et al. (1971) both reported that indoor
COH values were much closer to outdoors than either dustfall or TSP. Apparently, the fine
carbon particles measured by these techniques effectively penetrate buildings.
A similar conclusion is reached in indoor/outdoor light-scattering studies. Since
scattering of visible light is caused by the range of particles from about 0.2 to 1.0 pm,
measurements using this technique provide another index of fine-particle mass. Indoor and
outdoor light-scattering values were found to be the same and highly correlated in Japan
(Ishido, 1959; Ishido et al., 1956), in Italy (Parvis, 1952; Romagnol, 1961), and in New York
(Jacobs et al., 1962).
5-129
-------
\
50
40
30
20
10
—
|||{| OUTDOOR
\ \ INDOOR
«p
-------
Therefore, fine particles readily penetrate buildings and occur inside to about the same
extent as outdoors. Indoor activity adds incrementally to outdoor levels and, frequently,
somewhat higher levels of fine particles are observed indoors. Smoking adds very materially
to indoor levels.
5,8.4 Monitoring and Estimation of PersonalExposures
In previous sections of this chapter, the spatial and temporal variations in the
concentrations of S0? and of fine and coarse particles and their components were summarized
for both outdoor and indoor exposures. However, looking forward to health effects summarized
in Chapters 11 to 14, there is still one element of exposure remaining for discussion. In
addition to the particle concentrations measured by long integrating-time monitors (i.e.,
long-term doses of pollutants), people are exposed to short-term high concentrations.
Unfortunately, sufficient data do not exist to establish the relative importance of
concentration and time of exposure. There is, however, evidence (cited in Chapter 12) for
gaseous S0« and particles that long-term exposures can cause adverse health effects. There is
also evidence that a short burst of pollutant exposure can cause adverse health effects.
Therefore, these peaks in exposure are likely to be important, and there is some evidence that
peaks occur both indoors and outdoors. For example, earlier in this chapter it was noted that
high levels of S0? occur periodically close to intense sources. Obviously, people passing
through such an area, even though they are not resident there, receive a high short-term dose.
On roadways, particle concentrations tend to be very high because of resuspension of road
dust. Clearly, travelers experience such concentrations at least for the time they are in
traffic. As pointed out earlier in this section, an individual's daily activities, the places
visited, and activities in the home all play a role in that person's exposure.
For example, Repace et al. (1980) followed an individual's daily exposure with a portable
particle monitor and correlated these measurements with activities. Figure 5-55 shows time
series plots of PM concentrations to which James Repace, the principal investigator, was
exposed on October 16, 1979. Sharp peaks were evident in traffic, indoor smoking areas, and
his own home, particularly in the kitchen. Obviously, controlling outdoor air pollutant
levels would have little influence on his exposure to short-term doses of particles except for
those incurred on roadways. There have been other recent reports of statistical studies of
the relationships among personal, indoor, and outdoor particle concentrations.
In a personal monitoring study designed to test the relationships between outdoor concen-
trations and personal exposures and to estimate activity concentrations, Spengler et al.
(1980) collected 12-hour respirable particle samples for 15 days on 45 individuals in Topeka,
Kans. Particle concentrations experienced by monitored individuals were 2.5 times greater
than average outdoor levels for this time interval. Further, there was no correlation between
the outdoor level and the personal exposure of individuals. Variation in outdoor concentrations.
5-131
-------
Ul
M
OJ
o
z
Hi
O
o
u
280
260
240
220
200
180
160
140
120
100
80
60
40
20
I I I I I 1 I I I I I I ! I I II 1 1 I ! ! !
• INDOORS
• IN TRANSIT
O OUTDOORS
CAFiTERIA, SMOKING SECTOR
BEHIND SMOKY DIESEL TRUCK
COMMUTING I
BEDROOMjqf
STREET SUBURBS, OUTDOOR ,
WELL-VENTILATED KITCHEN
OUTSIDE CIGAR
SMOKER'S OFFICE
CAFETERIA, NONSMOKING
SECTION
SIDEWALK
SUBURBS IN BUS EXHAUST
VEHICLE CITY
OFFICE
LIBRARY, UNOCCUPIED CAFETERIA
CITY, OUTDOOR
COMMUTING
SUBURBS
JOGGING
LIVING
ROOM
LIVING'
ROOM
I I I I I I I I I I I I I I | | I I
12 1
MIDNIGHT
234
567
A.M.
9 10 11 12 1
NOON
TIME OF DAY
3 4
567
1 P.M.
8 9 10 11 12
Figure 5-55. An example of personal exposure to respirable particles.
Source: Repace et al. (1980).
-------
explained only 4 percent of the variability in personal concentrations experienced. On the
other hand, indoor respirable particle levels explained a fair amount of the variability in
personal exposure; for men, indoor levels accounted for 25 percent of their exposure variation
while for women 50 percent was explained by this variation. It appears that somewhere in the
daily activities of these individuals, their exposure was substantially greater than that
measured by outdoor monitors and, further, ttie variation in exposure is not measured
adequately by fixed indoor monitors either. Passive smoke exposure accounted for a
significant portion of the increment above outdoor levels. Figure 5-56 plots the histograms
of concentrations for both volunteers who reported no passive smoke exposure during the day
and those who reported some exposure to passive smoke. The means were 20 ug/m for
3
nonsmoke-exposed samples versus 40 ug/m for smoke-exposed samples.
In Figure 5-57, the daily mean concentrations for all outdoor, indoor, and personal
samples are presented. There is the suggestion that the variation in outdoor concentrations
causes variations in indoor and personal concentrations. However, variations in indoor con-
centration cause considerable variance in individual exposures.
As an alternative to direct measurement (monitoring), typical personal exposures may be
estimated on the basis of information on indoor and outdoor concentrations and human activity
patterns.
The exposure to particles and gases that one experiences will be ultimately determined by
location and activity. Certainly, locational and activity patterns are very complex in our
society. They are functions of age, sex, social, economic, and educational factors. While a
limited data base exists on activity patterns within our population and on the distribution of
smokers, housing stock, and various other building factors, an exhaustive discussion is not
appropriate for this document.
Time budget studies of the U.S. population indicate that on the average, 90 percent of an
individual's time is spent indoors. Between 5 and 10 percent of the time is spent in transit
in a vehicle. Considering these figures, the indoor environment is very important in
determining the time-weighted average exposure.
However, the time-weighted average is only one measure of pollution exposure. Time spent
outdoors is variable. The time outdoors varies by the time of day and year, among regions of
the country and among different categories of people. Therefore, in regard to the concern for
indoor pollution, the fact that short-term peak ambient concentrations may be an important
component of exposure should be remembered.
Much work remains to be done on personal exposures to gases and particles. Based on
current understanding, the following qualitative statements can be made:
1. Depending on spatial gradients in ambient air, personal exposures to SO- should be
less than the outdoor concentrations.
5-133
-------
20
18
„ 16
e
§
I 14
Z
2 12
§
1
NONSMOKING EXPOSED
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
NORMALIZED MEAN FINE PARTICLE CONCENTRATION
(<3.5 /urn)
14
I 12
fc*
1 10
o
P 8
5
3
0. 6
O
D.
1
g 2
o
—
—
*"""""* ••••»
::::::=K
::::::::1
lmm=|
nfll 1
MM*
^mm^ .
F^—>
•"^
^^"
fln[
SMOKING EXPOSED
nlln,-,^
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
—
—
—
„,„,,„
95
NORMALIZED MEAN FINE PARTICLE CONCENTRATION
(<3.5fJm)
Figure 5-56* -Normalized distribution of personal (12-hour) exposure samples
exposed and smoke exposed samples.
Source: Spengler and Tosteson
) for non-smoke
5-134
-------
Z
o
111
o
o
o
LU
O
p
DC
<
a.
LU
m
<
cc
Z
sa
ui
CC
40
35
30
25
20
15
10
n m
O PERSONAL
Q INDOOR
A OUTDOOR
I I I
Th Sa Tu Th Sa Tu Th Sa Tu Th Sa Tu Th Sa Tu Th Sa Tu
WEEK1 WEEK 2 WEEKS WEEK 4 WEEK 5 WEEK 6 WEEK?
DAILY AVERAGE CONCENTRATIONS FOR THE ENTIRE GROUP OF 46 SUBJECTS
IN THE TOPEKA STUDY
Figure 5-57. Daily mean indoor/outdoor and personal concentrations
of respirable particles. Daily means averaged over 24 homes and outdoor
locations and up to 46 personal samples. Samples collected during May and
June 1979.
Source: Spengleretal. (1980).
5-135
-------
2. Depending on activity times and building characteristics, longer term exposure could
be less than half the ambient concentrations.
3. For estimates of personal exposure to particle mass concentration, the ambient
measurement appears to be a poor predictor. While ambient concentrations exert an effect,
personal activities and indoor concentrations cause personal exposures to vary substantially.
4. Tobacco smoke is an important contributor to indoor and personal exposures,
5. Personal exposures to the components of suspended PM of outdoor origin and contained
in the micrometer and submicrometer size fraction may be estimated by ambient measurements.
The smaller size particles of toxic trace elements (V, Cd, Ni, Br, Se, etc.) and some organic
o— -
and inorganic compounds (SOt , NO,), which are exclusively of outdoor origin, penetrate the
indoor environment in a predictable way. Outdoor measurements of primary and secondary
fine-fraction aerosols in nonindustrialized communities may prove to be adequate to
characterize population exposures and trends. This last statement assumes no important indoor
sources for this typical outdoor component. This question certainly needs verification and
quantification in field studies.
5.9 SUMMARY OF ENVIRONMENTAL CONCENTRATIONS AND EXPOSURE
The purpose of this chapter is to document the existing concentrations of SO and PM in
the environment. Since the damage caused by these pollutants to man, other living things, and
valuable objects varies with time, place, and other circumstances, a wide variety of exposure
conditions are relevant for these pollutants.
Sulfur oxide concentrations in the air have been markedly reduced over the past 15 years
because of fuel sulfur restrictions, control technology implementation on major sources,
redistribution of power plants to regions outside cities, and the use of taller stacks. There
are still some areas with very high S09 concentrations, though, and hourly values of 4000 to
•i £•
6000 ug/m are rather common near large smelters. In about 100 U.S. locations, maximum hourly
o
values above 1000 ug/m are occasionally found, but much of the nation is basically in com-
pliance with the current NAAQS for SQ2.
After a downward trend from 1970 to 1974, total suspended PM concentrations have changed
very little in recent years despite major reductions in stationary source emissions
inventories. Dusty arid regions of the country still have high TSP as do industrialized
cities in the East and Far West. Ninetieth-percentile values of 24-hour TSP (the values that
o
are exceeded 10 percent of the time) are above 85 ug/m in every part of the country except
Q O
Alaska. Regional mean TSP values range from about 50 ug/m in EPA Region I to 77 ug/m in EPA
Region IX.
Ambient airborne particles exist in two distinct size ranges, fine particles below about
1 urn, and coarse particles above about 3 urn. Rather little mass is in intermediate sizes.
Except that both sizes are captured on filters, the two kinds have very little in common.
5-136
-------
Fine and coarse particles differ in origin, chemical composition, geographical distribution
and physical behavior.
Fine particles are composed mainly of (1) sulfate, nitrate, and ammonium ions, (2)
organic substances from atmospheric photochemical conversions, and (3) carbon, organic matter,
and metallic components directly emitted from combustion sources. Sulfate, most often in the
form of neutral ammonium sulfate, but sometimes in association with acidity as NH»HSO» or
hLSQ*, is the principal component, often accounting for 40 percent of fine-particle mass.
Sulfate and NO, ions are present in high concentrations during both summer and winter episodes
3
over very large sections of the eastern United States. This area experiences 10 |jg/m or
p-
greater SO, levels for one or two periods 'up to a month or more every year. The affected
region is so large in scope that no real background levels of fine particles are available for
measurement east of the Mississippi. Sulfat'e and fine-particle levels are nearly the same in
i
cities and in rural areas. Southern California experiences high levels of sulfates and
nitrates, particularly during photochemical incidents. In that area, high levels of fine
o
organic aerosols are also found, often exceeding 100 jjg/m .
Toxic organic particulate matter and imetals are mainly emitted from combustion and
industrial sources, and their concentrations are highest in cities. Trends in fine-particle
components have been mostly downward because of control measures taken, such as lead
reductions in gasoline.
Coarse particles in air are stirred up by the wind and by machinery. Since these parti-
cles settle fairly rapidly, they tend to be high close to sources. In most cases the coarse
particles account for two-thirds of TSP in dry regions like Phoenix, Oklahoma City, El Paso,
or Denver in the summer. The overwhelming cause of high TSP is local dust, but in
industrialized cities there is evidence that contributions of soot, fly ash, and industrial
fugitive emissions are also present.
Coarse particles are mainly composed of silica, calcium carbonate, clay minerals, and
soot. Chemical constituents found in this fraction include the elements silicon, aluminum,
potassium, calcium, and iron together with other alkaline earth and transition elements. Or-
ganic substances are also found in coarse particles, although their source is unknown.
Much of this coarse material is road dust suspended by traffic action, and street levels
of resuspended dust can be very high. Traffic on unpaved roadways can generate huge amounts
of dust, which deposits on vegetation and can be resuspended by wind action. Rain and snow
cover can reduce these emissions, but one study suggests that salting of roadways can be a
major source of winter TSP. Probably, associated sand is also important. Industrial fugitive
emissions can be even greater local sources of coarse particles, particularly from unpaved
access roads, construction activity, rock crushing, and cement manufacturing.
The problem of tracing existing levels of particles to sources is being solved in part by
a number of calculational methods generally categorized as source apportionment or
source-receptor models. The results from chemical element balance calculations or factor
5-137
-------
analysis are available now for several cities. Apportionments for these cities are presented
as examples of results to be expected in the future by application of these powerful methods.
Although outdoor concentrations of pollutants can be measured at particular sites, our,
highly mobile population can be exposed to either higher or lower values than community moni-
tors show. Indoor values of S02 tend to be lower than outdoor levels because walls, floors,
and furniture absorb SOp. Indoor particle levels can be high because of smoking, cleaning
operations, or normal activities. Exposure of individuals to SO and PM can vary more than
community monitors show.
5-138
-------
5.10 REFERENCES
Akland, G. G. Air Quality Data for Metals 1970 through 1974 from the National Air Surveil-
lance Networks. EPA-600/4-76-041, U. S. Environmental Protection Agency, Research
Triangle Park, NC, August 1976.
Akland, G. G. Air Quality Data for Nonmetallic Inorganic Ions 1971 through 1974 from the
National Air Surveillance Networks. EPA-600/4-77-003, U. S. Environmental Protection
Agency, Research Triangle Park, NC, January 1977.
Akselsson, R., C. Orsinc, D. L. Meinert, T. B. Johansson, R. E. Van Grieken, H. C. Kaufman, K.
R. Chapman, J. W. Nelson, and J. W. Winchester. Application of proton-induced x-ray
emission analysis to the St. Louis regional air pollution study. Adv. X-ray Anal.
18:588-597, 1975.
Alzona, J., B. L. Cohen, H. Rudolph, H. N. Jow, and J. 0. Frohliger. Indoor-outdoor relation-
ships for airborne particulate matter of outdoor origin. Atmos. Environ. 13:55-60, 1979.
Anderson, I. Relationships between outdoor and indoor air pollution. Atmos. Environ. 6:275-
278, 1972.
Appel, B. R., E. L. Kothny, E. M. Hoffer, G. M. Hidy, and J. J. Wesolowski. Sulfate and
nitrate data from the California Aerosol Characterization Experiment. Environ. Sci.
Techno!. 12:418-425, 1978.
Appel, B. R., S. M. Wall, Y. Tokiwa, and M. Ha'ik. Interference efects in sampling particulate
nitrate in ambient air. Atmos. Environ. 13:319-325, 1979.
Asahina, S. J. Andrea, A. Carmel, E. Arnold, Y. Bishop, S. Joshi, 0. Coffin, and S. S.
Epstein. Carcinogenicity of organic fractions of particulate pollutants collected in New
York City and administered subcutaneously to infant mice. Cancer Res. 32:2263-2268,
1972.
Axetell, K. Assessment of non-traditional source impacts in the Western United States. In:
Specialty conference on the Technical Basis for a Size Specific Particulate Standard.
Proceedings: Danvers, Massachusetts and Seattle, Washington, March 24-26 and April 23-
25, 1980. E. R. Frederick, ed., Air Pollution Control Association, Pittsburgh, PA, 1980,
pp. 223-229.
Ball, R. J., and G. E. Anderson. Optimum Site Exposure Criteria for S02 Monitoring. EPA-
450/3-77-013, U.S. Environmental Protection Agency, Research Triangle Park, NC, April,
1977.
Berdyev, Kh. B. , N. V. Pavlovich, and A.A. Tu'zhilina. Effect of motor vehicle exhaust gases
on atmospheric pollution in dwelling and in a main street. Gig. Sanit. ^2:424-426, 1967.
Biersteker, K. , H. de Graaf, and C. A. G. Nass. Indoor air pollution in Rotterdam homes. Air
Water Pollut. 9:343-350, 1965.
Binder, R. E. , C. A. Mitchell, H. R. Hosein, and A. Bouhuys. Importance of the indoor environ-
ment in air pollution exposure. Arch. Environ. Health 31:279, 1976.
Bradway, R. M., and F. A. Record. National Assessment of the Urban Particulate Problem. Vol.
II - Particle Characterization. EPA-450/3-76-025, U. S. Environmental Protection Agency,
Research Triangle Park, NC, July 1976.
5-139
-------
BHdbord, K., J, F. Finklea, J. K. Wagoner, J. B. Moran, and P. Caplan. Human exposure to poly-
nuclear aromatic hydrocarbons. In: Carcinogenesis: vol. I: Polynuclear Aromatic Hydro-
carbons: Chemistry, Metabolism and Carcinogenesis, a Symposium, U.S. Environmental Pro-
tection Agency and Battelle Columbus laboratories, Columbus, Ohio, October 1975. R.
Freudenthal and P. W. Jones, eds., Raven Press, New York, NY, 1976. pp. 319-324.
Brosset, C., K. Andreasspn, and M. Perm. The nature and possible origin of acid particles
observed at the Swedish west coast. Atmos. Environ. 9:631-642, 1975.
Brosset, C. Water-soluble sulphur compounds in aerosols. In: Sulfur in the Atmosphere, Pro-
ceedings of the International Symposium, United Nations Environmental Program, Dubrovnik,
Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:25-38, 1978.
Cahill, T. A., L, L. Ashbaugh, J. B. Barone, R. A. Eldred, P. Feeeney, R. G. Flocchine, C.
Goodart, D. J. Shadoan and G. W. Wolfe. Analysis of respirable fractions in atmospheric
particulates via sequential filtration. J. Air Pollut. Control Assoc. 27:675-678, 1977.
Calvert, J. G. Hydrocarbon involvement in photochemical smog formation in Los Angeles atmos-
phere. Environ. Sci. Techno!. 10:256-262, 1976.
Camp, D. C., A. L. Van Lehn, and B. W. Loo. Intercomparison of Samplers Used in the Determi-
nation of Aerosol Composition. EPA-600/7-78-118, U. S. Environmental Protection Agency,
Research Triangle Park, NC, July 1978.
Cantrell, B., and R. T. Whitby. Aerosol size distributions from the Labodie power plant
plume. Atmos. Environ. 12:324, 1978.
Carey, G. C. R., J. J. Phair, R. J. Shephard, and M. L. Thomson: The effects of air pollution
on human health. Am. Ind. Hyg. Assoc. J. 19:363-370, 1958.
Carpenter, B. H., and G. E. Weant, III. Particulate Control for Fugitive Dust.
EPA-600/7-78-071, U.S. Environmental Protection Agency, Research Triangle Park, NC, April
1978.
Cautreels, W., and K. Van Cauwenberghe. Experiments on the distribution of organic pollutants
between airborne particulate matter and the corresponding gas phase. Atmos. Environ.
12:1133-1141, 1978.
Charlson, R. J., D. S. Covert, T. V. Larson, and A. P. Waggoner. Chemical properties of
tropospheric sulfur aerosols. In: Sulfur in the Atmosphere, Proceedings of the Inter-
national Symposim, United Nations Enviornmental Program, Dubrovnik, Yugoslavia, September
7-14, 1977. Atmos. Environ. 12:39-53, 1978.
Cobourn, W. G,, R. B. Husar, and J. D. Hysar. Continuous jji situ monitoring of ambient parti-
culate sulfur using flame photometry and thermal analysis. In: Sulfur in the Atmos-
phere, Proceedings of the International Symposium, United Nations Environmental Program,
Dubrovnik, Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:89-98, 1978.
Cohen, D., S. F. Arai, and J. D. Brain. Smoking impairs long-term dust clearance from the
lung. Science (Washington, DC) 204:514-517, 1979.
Colucci, J. M., a nd C. R. Begeman. Polynuclear aromatic hydrocarbons and other pollutants in
Los Angeles air. In: Proceedings of the Second International Clean Air Congress, Inter-
national Union of Air Pollution Prevention Associations, Washington, DC, December 6-11,
1976. H. M. Englund and W. T. Berry, eds., Academic Press, Inc., New York, NY, 1971. pp.
28-35.
5-140
-------
Cooper, D. W., J. S. Evans, M. Quinn, R. C, AntoneTIi, and M. Schneider. Setting Priorities
for Control of Fugitive Participate Emissions from Open Sources. EPA-600/7-79-186, U. S.
Environmental Protection Agency, Research Triangle Park, NC, August 1979.
Cooper, J. A. , J. G. Watson, and J. J. Huntzicker. Summary of the Portland Aerosol Charac-
terization Study. Presented at the 72nd Annual Meeting, Air Pollution Control Assoc-
iation, Cincinnati, OH, June 24-29, 1979a. Paper 79-24,4.
Cooper, J. A., and J. G. Watson, Jr. Receptor oriented methods of air particulate source
apportionment. J. Air Pollut, Control Assoc. 30:1116-1125, 1980.
Corn, M, Aerosols and the primary air pollution - nonviable particles: their occurrence,
properties, and effects. In: Air Pollution. Vol. I: Air Pollutants, Their Transforma-
tion and Transport. A. C. Stern, ed. , Academic Press, New York, NY, 1976. pp. 78-168.
Corn, M. Nonviable particles in the air. In: Air Pollution. Vol. I: Air Pollution and its
Effects. A. C. Stern, ed., Academic Press, 2nd Edition. New York, NY, 1968, pp. 47-94.
Courtney, W. J., J. W. Tesch, R. K. Stevens, and T. G, Dzubay. Characterization of the Denver
aerosol between December 1978 and December 1979. Presented at the 73rd Annual Meeting,
Air Pollution Control Association, Montreal, Quebec, Canada, June 22-27, 1980. Paper no.
80-58.1.
Cowherd, C., Jr., R. Bonn, and T. Cuscino, Jr. Iron and Steel Plant Open Source Fugitive
Emission Evaluation. EPA-600/2-79-103, U. S. Environmental Protection Agency, Research
Triangle Park, NC, May 1979.
Cox, R. A., and S. A. Penkett. Effect of relative humidity on the disappearance of ozone and
sulphur dioxide in contained systems. Atmos. Environ. 6:365-368, 1972.
Cronn, D. R., R. J. Charlson, R. L. Knights, A. L. Crittenden, and B. R. Appel. A survey of
the molecular nature of primary and secondary components of particles in urban air by
high-resolution mass spectrometry. Atmos. Environ. 11:929-937, 1977.
Cunningham, P. T., and S. A. Johnson. Spectroscopic observation of acid sulfate in atmos-
pheric particulate samples. Science (Washington, D.C.) 191:77-79, 1976.
Curran, T. C., and W. F. Hunt, Jr. Interpretation of air quality data with respect to the
national ambient air quality standards. J, Air Pollut, Control Assoc. 25:711-714, 1975.
Daisey, J. M. Organic compounds in urban aerosols. In: Aerosols: Anthropogenic and
Natural, Sources and Transport, Proceedings of a Conference, New York Academy of Sciences
and Air Pollution Control Association, New York, NY, January 9-12, 1979. Ann, N.Y. Acad.
Sci. 338:50-69, 1980.
Daisey, J. M., I. Hawryluk, T. J. Kneip, and F. Mukai. Mutagenic activity in organic
fractions in airborne particulate matter. In: Proceedings: (Conference) Carbonaceous
Particles in the Atmosphere, National Science Foundation and Lawrence Berkeley Labora-
tory, University of California, Berkeley, California, March 20-22, 1978. LBL-9037, T.
Novakov, ed. , University of California, -Lawrence Berkeley Laboratory, Berkeley, CA,
1979. pp. 187-192.
Daisey, J. M., R. J. McCaffrey, and R. A. Gallagher. Polycyclic aromatic hydrocarbons and
total extractable particulate organic matter in the Arctic aerosol. Atmos. Environ. 15:
1352-1363, 1981.
5-141
-------
Davidson, C. I., and S. K. Frledlander. A filtration model for aerosol dry deposition:
application to trace metal deposition from the atmosphere. J. Geophys. Res.
83:2343-2352, 1978.
De Maio, L, and M. Corn. Polynuclear aromatic hydrocarbons associated with participates in
Pittsburgh air. J. Air Pollut. Control Assoc. 16:67-71, 1966.
De Wiest, F. Any factors influencing the dispersion and the transport of heavy hydrocarbons
associated with airborne particles. Atmos. Environ. 12:1705-1711, 1978.
De Wiest, F., and D. Rondia. Sur la validite des determinations du benzo(a)pyrene atmospheric
pendant les mois d'ete, [on the validity of the determination of atmospheric
benzo(a)pyrene in the summer.] Atmos. Environ. 10:487-489, 1976.
Dockery, 0, W., and J. D. Spengler. Indoor-outdoor relationships of respirable sulfates and
particles, Atmos. Environ. 15:335.-343, 1981.
Dockery, D. W. Personal Exposure to Respirable Particulates and Sulfates: Measurement and
prediction. Ph.D. Dissertation, Harvard School of Public Health, Boston, MA, 1979.
Draftz, R. G. Aerosol Source Characterization Study in Miami, Florida: Microscopical Analy-
sis. EPA-600/3-79-097, U. S. Environmental Protection Agency, Research Triangle Park,
NC, September 1979.
Draffz, R. 6., and K. Severin. Microscopical Analysis of Aerosols Collected in St. Louis,
Missouri. EPA-600/3-80-027, U, S. Environmental Protection Agency, Research Triangle
Park, NC, February 1980.
Draftz, R. G., I. Graf, and E. Arnold. Preliminary assessments of traditional and non-
traditional source impacts. In: Specialty Conference on the Technical Basis for a Size
Specific Particulate Standard. Proceedings, Danvers, Massachusetts and Seattle, Wash-
ington, March 24-26, and April 23-25, 1980. E. R. Frederick, ed. , Air Pollution Control
Association, Pittsburgh, PA, 1980. pp. 94-106.
Duce, R. A. Speculations on the budget of particulate and in vapor phase non-methane organic
carbon in the global troposphere. Pure Appl. Geophys. 116(23):244-273, 1978.
Dzubay, T. G. Chemical-elemental balance method applied to dichotomous sampler data. In;
Aerosol; Anthropogenic and Natural Sources and Transport, Proceedings of a Conference,
New York Academy of Sciences and Air Pollution Control Association, New York, New York.
January 9-12, 1979. Ann., N.Y. Acad. Sci. 338:126-144, 1980.
Dzubay, T. G. , and R., K. Stevens. Ambient air analysis with dichotomous sampler and x~ray
fluorescence spectrometer. Environ. Sci. Techno!. 9:663-668, 1975.
Dzubay, T. G., R. K. Stevens, and C. M. Peterson. Application of the dichotomous sampler to
the characterization of ambient aerosols. In: X-Ray Fluorescence Analysis of Environ-
mental Samples. T. G. Dzubay, ed., Ann Arbor Science Publishers, Inc., Ann Arbor, MI,"
1977, pp. 95-106.
Dzubay, T. G., R. K. Stevens, W. J. Courtney, and E. A. Drane. Chemical element balance
analysis of Denver aerosol. In: Electron Microscopy and X-ray Applications to Environ-
mental Health Analysis. Volume 2. P. A. Russell, ed. , Ann Arbor Science, Ann Arbor, MI,
1981. pp. 23-42.
Economic Commission for Europe's Working Party on Air Pollution Probfems. ECE Task Force on
Fine Particulate Pollution. World Health Organization, Geneva, Switzerland, December
1977.
5-142
-------
Eliassen, A. The OECD study of long range transport of air pollutants: Long range transport
modelling. In: Sulfur in the Atmosphere, Proceedings of the International Symposium,
United Nations Environmental Program, Dubrovnik, Yugoslavia, September 7-14, 1977.
Atmos. Environ. 12:479-487, 1978.
Faoro, R. B. Trends in concentrations of benzene-soluble suspended particulate fraction and
benzo(a)pyrene. J. Air Pollut. Control Assoc. 25:638-640, 1975.
Faoro, R. D. , and T. B. McMullen. National Trends in Trace Metals in Ambient Air, 1965-1974.
EPA-450/1-77-003, U.S. Environmental Protection Agency, Research Triangle Park, NC,
February 1977.
Flocchini, R. G. , T. A. Cahill, R. A. Eldred, L. L. Ashbaugh, and J. B. Barone. Sulfur size
distribution by season and site in California. Presented at the 71st Annual Meeting, Air
Pollution Control Association, Houston, TX, June 25-30, 1978. Paper no. 78-39.6.
Fraser, D. W., and J. E. McDade. Legionellosis. Sci. Am. 241:82-99, 1979.
Friedlander, S. K. Chemical element balances and identification of air pollution sources.
Environ. Sci. Tech. 7:235-240, 1973.
GEOMET, Incorporated. Indoor-Outdoor Pollution Levels: A Bibliography. EPRI EA-1025,
Electric Power Research Institute, Palo Alto, CA, March 1979.
Gillette, D. A wind tunnel simulation of the erosion of soil: effect of soil texture,
sandblasting, wind speed, and soil consolidation on dust production. Atmos. Environ.
12:1735-1743, 1978.
Goldwater, L. J. , A. Manoharan, and M. B. Jacobs. Suspended particulate matter, dust in
"domestic" atmospheres. Arch. Environ. Health 2:511-515, 1961.
Gordon, G. E. Receptor models. Environ. Sci. Technol. 14:792-800, 1980.
Gordon, R. J. Distribution of Airborne Polycyclic Aromatic Hydrocarbons throughout Los
Angeles. Environ. Sci. Technol. 10:370-373, 1976.
Graf, J., R. H. Snow, and R. G. Draftz. Aerosol sampling and analysis - Phoenix, Arizona.
EPA-600/2-77-015, U.S. Environmental Protection Agency, Washington, DC, 1977.
Grosjean, D. Aerosols. In: Ozone and Other Photochemical Oxidants. National Academy of
Sciences, Washington, DC, 1977. pp. 45-125.
Grosjean, D., and S. K. Friedlander. Gas-particle distribution factors for organic and other
pollutants in the Los Angeles atmosphere. J. Air Pollut. Control Assoc. 25:1038-1044,
1975.
Hahn, J. Organic constituents of natural aerosols. In: Aerosols: Anthropogenic and Natural
Sources and Transport, Proceedings of a Conference, New York Academy of Sciences and Air
Pollution Control Association, New York, New York, January 9-12, 1979. Ann. N.Y. Acad.
Sci.338:359-376. 1980.
Hardy, K. A. Aerosol Source Characterization Study in Miami, Florida: Trace Element
Analysis. EPA-600/7-79-197, U.S. Environmental Protection Agency, Research Triangle
Park, NC, September 1979.
Heffter, J. L. , A. 0. Taylor, and G. J. Ferber. A Regional-Continental Scale Transport,
Diffusion, and Deposition Model. NOAA Technical Memorandum ERL-ARL-50, U. S. Department
of Commerce, National Oceanic and Atmospheric Administration, Washington, DC, 1975.
5-143
-------
Heisler, S. L., R. C. Henry, J. G. Watson, and G. M. Hidy. The 1978 Denver Winter Haze Study,
vol. II - Final Report. ERT Document No. T-5417-1, Environmental Research and Tech-
nology, Inc., Westlake Village, CA, March 1980.
Helms, G. T., J. H. Southerland, K. R. Woodard, I. J. Hindawi, and D. H. Coventry, Chat-
tanooga, Tennessee — Rossville, Georgia Interstate Air Quality Study 1967-1968.
APTD-0583, U.S. Department of Health, Education, and Welfare, National Air Pollution
Control Administration, Durham, NC, October 1970.
Hidy, G. M., B. R. Appel, R. J. Charlson, W. E. Clark, S. K. Friedlander, D. H. Hutchison, T.
B. Smith, J. Suder, J. J. Wesolowski, and K. T. Whitby. Summary of the California
Aerosol Characterization Experiment. J. Air Pollut. Control Assoc. 25:1106-1114, 1975.
Hidy, G. H., P. K. Mueller, and E. Y. Tong. Spatial and temporal distributions of airborne
sulfates in parts of the United States. In: Sulfur in the Atmosphere, Proceedings of
the International Symposium, United Nations Environmental Program, Dubrovnik, Yugoslavia,
September 7-14, 1977. Atmos. Environ. 12:735-752, 1978.
Hidy, G. M., P. K. Mueller, V. Dio, and K. C. Detor. Study and implementation of the Sulfate
Regional Experiment (SURE). Proc., 4th Symposium on Turbulence and Diffusion and Air
Pollution, Aerican Meteorological Society, Reno, Nevada, January 1518, 1979.
Hitchcock, D. R. Atmospheric s'ulfates from biological sources. J. Air Pollut. Control Assoc.
26:210-215, 1976.
Hitchcock, D. R. Biogenic Sulfur Sources and Air Quality in the United States. NSF/RA-770276,
National Science Foundation, Washington, DC, August 1977.
Hitchcock, D. R., L. L. Spiller, and W. E. Wilson. Sulfuric acid aerosols and HC1 release in
coastal atmospheres: evidence of rapid formation of sulfuric acid particulates. Atmos.
Environ. 14:165-182, 1980.
Hoffmann, D., and E. L. Wynder. Organic particulate pollutants—chemical analyses and
bioassays for carcinogenicity. In: Air Pollution. A. C. Stern, ed., 3d ed., Academic
Press, Inc., New York, NY, 1977. pp. 361-455.
Hueper, W. C., P. Kotin, E. C. Tabor, W. W. Payne, H. Falk, and E. Sawicki. Carcinogenic
bioassays on air pollutants. Arch. Pathol. 74:89-116, 1962.
Huntzicker, J. J., R. A. Cary, and C-S. Ling. Neutralization of sulfuric acid aerosol by
ammonia. Environ. Sci. Technol. 14:819-824, 1980.
Huntzicker, J. J., R. L. Johnson, J. J. Shah, and R. A. Cary. Analysis of organic and ele-
mental carbon in ambient aerosols by a thermal-optical method. In: Particulate Carbon:
Atmospheric Life Cycle, Proceedings of an International Symposium, General Motors
Research Laboratories, Warren, MI, October 13-14, 1980. G. T. Wolff and R. L. Klimisch,
eds., Plenum Prss, New York, NY, 1982. pp. 79-88.
Ishido, S. Variations in indoor and outdoor dust intensities. Bull. Dept. Home Econ., Osaka
City University, 6:53-59, 1959.
Ishido, S., K. Kamada, and Nakagawa: Free dust particles and airborne microflora. Bull.
Dept. Home Econ., Osaka City Univ., 21:31-37, 1956.
Jacobs, M. B., L. J. Goldwater, and A. Fergany: Comparison of suspended particulate matter of
indoor and outdoor air. Air Water Pollut. 6:377-380, 1962.
5-144
-------
Jager, J. Detection and characterization of nitro derivatives of some polycyclic aromatic
hydrocarbons by fluorescence quenching after thin-layer chromatography: application to
air pollution analysis. J, Chromatogr. 152:575, 1978.
Junge, C. Die (Constitution des Atmospharischen Aerosols. Ann. Meteorol. 5 (Beiheft) 1-55,
1952.
Junge, C. and G. Scheich. Determination of the acid content of aerosol particles. Atmos.
Environ. 5:165-175, 1971.
Kadawaki, S. Size distribution and chemical composition of atmospheric participate nitrate in
the Nagoya area. Atmos. Environ. 11:671-675, 1977.
Katz, M., and C, Chan. Comparative distribution of eight polycyclic aromatic hydrocarbons in
airborne particulates collected by conventional high-volume sampling and by size fraction-
ation. Environ. Sci. Technol. 14:838-843, 1980.
Kertesz-Saringer, M., E. Meszaros, and T. Varkovni. On the size distribution of carbon-
containing particles in urban air. Atmos. Environ. 5_: 429-431, 1971.
Ketseridis, G., J. Hahn, R. Jaenicke, and C. Junge. The organic constituents of atmospheric
particulate matter. Atmos. Environ. 10:603-610, 1976.
Kito, T. Survey of air pollution on outlying islands: An attempt to collect the background
data as to urban air pollutions. Kuki Seijo 15:10-21, 1977.
Kneip, T. J., M. A. Leyko, M. T. Kleinman, M. Lippmann, and J. M. Daisey. Organic matter in
New York City TSP: comparisons of seasonal variations and relationships to source
tracers. In: Proceedings: Conference on Carbonaceous Particles in the Atmosphere,
National Science Foundation and Lawrence Berkeley Laboratory, University of California,
Berkeley, California, March 20-22, 1978. LBL-9037, T. Novakov, ed., University of
California, Lawrence Berkeley Laboratory, Berkeley, CA, June 1979. pp. 79-83.
Korfmacher, W. A. , E. L. Wehry, G. Mamantov, and D. F. S. Natusch. Resistance to photo-
chemical decomposition of polycyclic aromatic hydrocarbons vapor—adsorbed on coal fly
ash. Environ. Sci. Technol. 14:1094-1099, 1980.
Krstulovic, A. M., D. M. Rosie, and P. R. Brown. Distribution of some atmospheric polynuclear
aromatic hydrocarbons. Am. Lab. 9:11-12,14,16-18, 1977.
Kurosaka, D. Sulfate Concentrations in the South Coast Air Basin. DTS-76-1, California Air
Resources Board, Sacramento, CA, January 1976.
Lamb, S. I., C. Petrowski, I. R. Kaplan, and B. R. T. Simoneit. Organic compounds in urban
atmospheres: a review of distribution, collection and analysis. J. Air. Pollut. Control
Assoc. 30:1098-1115, 1980.
Larsen, R. I. A Mathematical Model for Relating Air Quality Measurements to Air Quality
Standards. AP-89, U.S. Environmental Protection Agency, Research Triangle Park, NC,
November 1971.
Lavery, T. F. , G. M. Hidy, R. L. Basket, and J. Thrasher. Occurrence of long-range transport
of sulfur oxides in the northeastern United States. In: Proceedings: 4th Symposium on
Turbulence Diffusion and Air Pollution American Meteorological Society, Reno, Nevada,
Jan. 15-18, 1979.
5-145
-------
Leaderer, B. P., R. L. Tanner, and T. H. Holford. Diurnal variations, chemical composition
and relation to meteorological variables of the summer aerosol in the New York sub-
region. Atmos Environ. 16:2075-2087, 1982.
Lee, R. E., Jr., and S. S. Goranson. National Air Surveillance Cascade Impactor Network. I.
Size distribution measurements of suspended particulate matter in air. Environ. Sci.
Techno!. (5:1019-1024, 1972.
Lee, R. E., Jr., S. S. Goranson, R. E. Enrione, and G. B. Morgan. National Air Surveillence
Cascade Impactor Network. II. Size distribution measurements of trace metal components.
Environ. Sci. Techno!. 6:1025-1030, 1972.
Lee, R. E., Jr., R. K, Patterson, and J. Wagman. Particle-size distribution of metal
components in urban air. Environ. Sci. Techno!. 2:288-290, 1968.
Lefcoe, N. H., and I. I. Inculet. Particulates in domestic premises: II. Ambient levels and
indoor-outdoor relationships. Arch. Environ. Health 30:565-570, 1975.
Lewis, C. W., and E. S. Macias. Composition of size-fractionated aerosol in Charleston, West
Virginia. Atmos. Environ. 14:185-194, 1980.
Lioy, P. J., P. J. Samson, R. L. Tanner, B. P. Leaderer, T. Minnich, and W. Lyons. The distri-
bution and transport of sulfate "species" in the New York metropolitan area during the
1977 summer aerosol study. Atmos. Environ. 14:1391-1407, 1980.
Lioy, P. J., R. P. Mallon, and T. J. Kneip. Long-term trends in total suspended participates,
vanadium, manganese, and lead at near street level and elevated sites in New York City.
J. Air Pollut. Control Assoc. 30:153-156, 1980a.
Lundgren, D. A. Atmospheric aerosol composition and concentration as a function of particle
size and of time. J. Air Pollut. Control Assoc. 20:603-608, 1970.
Lynn, D. A., B. S. Epstein, and C. K. Wilcox. Analysis of New York City Sulfate Data. Final
Report. EPA Contract No. 68-02-1337, GCA Corp., Bedford, MA, 1975.
Lynn, D. A., G. L. Deane, R. C. Galkiewicz, R. M. Bradway, and F. R. Record. National Assess-
ment of the Urban Particulate Problem. Volume I. Summary of the National Assessment.
EPA-450/3-76-024, U. S. Environmental Protection Agency, Research Triangle Park, NC, July
1976. p. 53.
Lyons, W. A., J. C. Dooley, Jr., and K. T. Whitby. Satellite detection of long range pol-
lution transport and sulfate aerosol haze. In: Sulfur in the Atmosphere, Proceedings of
the International Symposium, United Nations Environmental Program, Dubrovnik, Yugoslavia,
September 7-14, 1977. Atmos. Environ. 12:621-631, 1978.
Mage, D. T., and W. R. Ott. Refinements of the lognormal probability model for analysis of
aerometric data. J. Air Pollut. Control Assoc. 28:796-798, 1978.
Maunsell, K. Concentration of airborne spores in dwellings under normal conditions and under
repair. Int. Arch. Allergy Appl. Immunol. 5:373-376, 1954.
HcCrone, W. C. Morphological analysis of particulate pollutants. In: Air Pollution. 2nd
Ed., Vol. II: Analysis, Monitoring, and Surveying. A. C. Stern, ed., Academic Press,
Inc., New York, NY, 1968. pp. 281-301.
McFadden, J. E., J. H. Beard, JI !.,,.„ and D. J. Moschandreas. Survey of Indoor Air Quality
Health Criteria and Standards. EPA*-600/7-78-027, 'U.S. Environmental Protection Agency,
Research Triangle Park, NC, March 1978.
5-146
-------
Megaw, W. S. The penetration of iodine into buildings. Air Water Pollut. 5:121-128, 1962.
Miller, F. J., D. E. Gardner, J. A. Graham, R. E. Lee, Jr., W. E. Wilson, and J. D. Bachman.
Particle Size Considerations for Establishing a Standard for Inhaled Particles, Draft,
U.S. Environmental Protection Agency, Research Triangle Park, NC, January 1979.
Montgomery, T. L. and J. H. Coleman, Empirical relations between timeaveraged sulfur dioxide
concentrations. Environ. Sci. Techno!. 9:9537, 1975.
Moschandreas, D. J., J. W. Winchester, J. W. Nelson, and R. M. Burton. Fine particle resi-
dential indoor air pollution. Atmos. Environ. 13:1413-1418, 1979.
Mueller, P. K., H. L. Helwig, A. E. Alcocer, W. K. Gong, and E. E. Hones. Concentration of
fine particles and Pb in car exhause. Am. Soc. Testing Mater., Spec. Tech. Publ. No.
352:6073, 1962.
Mueller, P. K., R. W. Mosely, and L. B. Pierce. Carbonate and noncarbonate carbon in atmos-
pheric particles, jri: Proceedings of the Second International Clean Air Congress,
International Union of Air Pollution Prevention Associations, Washington, DC, December 6-
11, 1970. H. M. Englund and W. T. Beery, eds., Academic Press, Inc., New York, NY, 1971.
pp. 532-539.
Mueller, P. K., R. W. Mosley, and L. B. Pierce. Chemical composition of Pasadena aerosol by
particle size and time of day. IV. Carbonate and noncarbonate carbon content. J.
Colloid Interface Sci. 39:235-239, 1972.
Mueller, P. K., G. M. Hidy, T. F. Lavery, K. Warren, R. L, Bassett. Some early results from
the Sulfate Regional Experiment (SURE). Proceedings: 4th Symposium on Turbulence Diffu-
sion and Air Pollution, American Meteorological Society, Reno, Nevada, Jan. 15-18, 1979,
pp. 322-329.
National Academy of Sciences. Air Quality and Stationary Source Emissions Control. Serial
No. 94-4, U.S. Senate, Committee on Public Works, Washington, DC, March 1975.
National Academy of Sciences. Nitrogen Oxides. National Academy of Sciences, Washington, DC,
1977.
National Academy of Sciences. Vapor-Phase Organic Pollutants. National Academy of Sciences,
Washington, DC, 1976.
National Air Pollution Control Administration. Air Quality Criteria for Sulfur Oxides. NAPCA
pub. no. AP-50. U.S. Government Printing Office, Washington, DC, 1970. /
/
National Aerometric Data Base. U. S. Environmental Protection Agency, Research Triangle Park,
NC, 1977.
National Academy of Sciences. Particulate Polycyclic Organic Matter. National Academy of
Sciences, Washington, DC, 1972.
National Research Council. Nitrates: An Environmental Assessment. National Academy of
Sciences, Washington, DC, 1978b. ^
National Research Council. Sulfur Oxides. National Academy of Sciences, Washington, DC,
1978a.
National Research Council. Airborne Particles. University Park Press, Baltimore, MD, 1979.
5-147
-------
Natusch, D. F. A. Characterizations of atmospheric pollutants from power plants. |n:
Proceedings of the Second Federal Conference on the Great Lakes, 1975. Great Lakes Basin
Commission, Ann Arbor, MI, 1976. pp. 114.
Novakov, T. Soot in the atmosphere. In: Particulate Carbon; Atmospheric Life Cycle,
Proceedings of an International Symposium, General Motors Research Laboratories, Warren,
MI, October 13-14, 1980. G. T. Wolff and R. L. Klimisch, eds., Plenum Press, New York,
NY, 1982. pp. 19-41.
O'Donnell, H., T. L. Montgomery, and M. Corn. Routine assessment of the particle size-weight
distribution of urban aerosols. Atmos. Environ. 4:1-7, 1970.
Office of Air Quality Planning and Standards. U. S. Environmental Protection Agency, Research
Triangle Park, NC, 1976.
Office of Air Quality Planning and Standards. U. S. Environmental Protection Agency, Research
Triangle Park, NC, 1979.
Orel, A., and J, Seinfeld. Nitrate formation in atmospheric aerosols. Environ. Sci. Techno!.
11:1000-1007, 1977.
Pace, T. G. Ambient particulate baseline conditions - sources and concentrations. In:
Specialty Conference on the Technical Basis for a Size Specific Particulate Standard.
Proceedings, Danvers, Massachusetts and Seattle, Washington, March 24-26 and April 23-25,
1980. E. K. Frederick, ed. , Air Pollution Control Association, Pittsburgh, PA, 1980.
pp. 26-39.
Pace, T. G., and E. L. Meyer, Jr. Preliminary characterization of inhalable particulates in
urban areas. Presented at the 72nd Annual Meeting, Air Pollution Control Association,
Cincinnati, OH, June 24-29, 1979. paper no. 79-47.2.
Pace, T. G., W. P. Freas, and E, M. Afify. Quantification of the relationship between monitor
height and measured particulate levels in six U.S. urban areas. Presented at the 70th
Annual Meeting, Air Pollution Control Association, Toronto, Ontario, Canada, June 1977.
Paper 77-13.4.
Pace, T. G., J. G. Watson, and C. E. Rodes. Preliminary Interpretation of Inhalable Particu-
late Network Data. Presented at the 74th Annual Meeting, Air Pollution Control Assoc-
iation, APCA Paper no. 81-5.2. Philadelphia, PA, June 21-26, 1981.
Parvis, D. Condensation nuclei in the air of artificially heated environments. Ann. Sanita
Pub!ica 13:1569-1581, 1952.
Perera, F. P., and A. K. Ahmed. Respirab,le Particulates: Impact of Airborne Fine Parti-
culates on Health and the Environment. Natural Resources Defense Council, Inc., New
York, NY, October 1978.
Pierce, R. C., and M. Katz. Dependency of polynuclear aromatic hydrocarbon content on size
distribution of atmospheric aerosols. Environ. Sci. Techno!. 9:347353, 1975.
Pierson, W. R., W. W. Brachaczek, T. J. Korniski, T. J. Truex, and J. W. Butler. Artifact
formation of sulfate, nitrate, and hydrogen ion on backup filters: Allegheny Mountain
Experiment. J. Air Pollut. Control Assoc. 30:30-34, 1980a.
5-148
-------
Pierson, W. R., W. W. Brachaczek, T. J. Truex, J. W. Butler, and T. J. Korniski. Ambient
sulfate measurements on Allegheny Mountain and the question of atmospheric sulfates in
the northeastern United States. In: Aerosols: Anthropogenic and Natural Sources and
Transport, Proceedings of a Conference, New York Academy of Sciences and Air Pollution
Control Association, New York, New York, January 9-12, 1979. Ann. N.Y. Acad. Sci. 338:
145-147, 1980b.
Pitts, J. N., Jr., D. Grosjean, T. M. Mischke, V. F. Simmon, and D. Poole. Mutagenic activity
of airborne particulate organic pollutants. Toxicol. Lett. 1:65-70, 1977.
Rahn, K. A., R. D. Borys, E. L. Butler, and R. A. Duce. Gaseous and particulate halogens in
the New York City atmosphere. Ann. N.Y. Acad. Sci. 322:143-151, 1979.
Record, F. A., and R. M. Bradway. Philadelphia Particulate Study. EPA-903/ 9-78-003, U. S.
Environmental Protection Agency, Philadelphia, PA, June 1978.
Record, F. A., K. W. Whiltsee, Jr., and R. M. Bradway. Program to measure, analyze, and
evaluate total suspended particulates (TSP) in Massachusetts. GCA-TR-79-44-G, GCA Cor-
poration, Bedford, MA, September 1979.
Repace, J. L., and A. H. Lowrey. Indoor air pollution, tobacco smoke, and public health.
Science 208:464-472, 1980.
Repace, J. L. , W. R. Ott, and L. A. Wallace. Total human exposure to air pollution.
Presented at the 73rd Annual Meeting, Air Pollution Control Association, Montreal,
Quebec, Canada, June 2.2.-21, 1980. Paper no. 80-61.6.
Rheingrover, S. W. A statistical model of titanium pollution transport and dispersion in the
atmosphere of St. Louis. J. Appl. Meterol. 20:693-705, 1981.
Richard, G. An Implementation Plan for Suspended Particulate Matter in the Pheonix Area.
Vol. IV - Control Strategy Formation. EPA-450/3-77-021d, U. S. Environmental Protection
Agency, Research Triangle Park, NC, June 1977.
Richard, G., and R. Tan. An Implementation Plan for Suspended Particle Matter in the Phoenix
Area. Vol. I - Air Quality Analysis. EPA-450/3-77-021a, U. S. Environmental Protection
Agency, Research Triangle Park, NC, November 1977.
Richard, G. , J. Avery, and L. Baboolal. An Implementation Plan for Suspended Particulate
Matter in the Phoenix Area. Vol, III - Model Simulation of Total Suspended Particulate
Levels. EPA-450/3-77-021c, U. S. Environmental Protection Agency, Research Triangle
Park, NC, August 1977a.
Richard, G., R. Tan, and J. Avery. An Implementation Plan for Suspended Particulate Matter in
the Phoenix Area. Vol. II - Emission Inventory. EPA-450/3-77-021b, U. S. Environmental
Protection Agency, Research Triangle Park, NC, December 1977b.
Rodhe, H. , C. Persson, and Akesson. An investigation into regional transport of soot and
sulfate aerosols. Atmos. Environ. 6:675-693, 1972.
Romagnol, G. Studies on the climatic conditions in some elementary classrooms of Novara.
Ital. Rev. Hyg. 21:410-419, 1961.
Rosen, H. , A. D. A. Hansen, R. L. Dod, and T. Novakov. Soot in urban atmospheres: deter-
mination by an optical absorption technique. Science (Washington, D.C.) 208:741-744,
1980.
5-149
-------
Samson, P. 0. Trajectory analysis of summertime sulfate concentrations in the northeastern
United States. 0. Appl. Meteorol. 19:1382-94, 1981.
Santodonato, J., P. Howard, D. Basu, S. Lande, J. K, Selkirk, and P. Sheehe. Health Assess-
ment Document for Polycyclic Organic Matter. EPA-6QO/9-79-008, U.S. Environmental Pro-
tection Agency, Research Triangle Park, NC, December 1979.
SAROAD (Storage and Retrieval of Aerometric Data), U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, NC.
Sawicki, E, Analysis of atmospheric carcinogens and their cofactors. In: Environmental
Pollution and Carcinogenic Risks. Proceedings of a Symposium, International Agency for
, Research on Cancer and French National Institute of Health and Medical Research, Lyon,
France, November 3-5, 1975. C.\Rosenfeld and W. Davis, eds., IARC Scientific Publi-
cations No. 13, INSERM Scientific Publications No. 13, French National Institute of
Health and Medical Research (INSERML Paris, France, 1976. pp. 297-354.
Sawicki, E. The genotoxic environmentalXpollutants, Paper presentedXat the Symposium in
Management of Residuals from Synthetic Fuel9 Production, Denver, CO,\May 25, 1976.
Sawicki, E., T. R. Hauser, W. C. Elbert, F. T. Fox, and J. E. Meeker. Polynuclear aromatic
hydrocarbons: Composition of the atmosphere in some large American Vities. Am. Ind.
Hyg. Assoc. 0, 23:137-144, 1962.
Sawicki, E., S. P. McPherson, T. W. Stanley, J. Meeker, and W. C. Albert. Quantitative com-
position of the urban atmosphere in terms of polynuclear aza-heterocyclic compounds and
aliphatic and polynuclear aromatic hydrocarbons. Air Water Pollut. 9:515-/524, 1965.
Schaefer, V. S., V. A. Mohnen, and V. R. Veirs: Air quality of American homes. Science
(Washington, D.C.) 175:133-175, 1972. ,
Schuetzle, D., D. Cronn, A. L. Crittenden, and R. J. Charlson. Molecular composition of
secondary aerosol and its possible origin. Environ. Sci. Techno!. 9:838, 1975.
Schwartz, G,, J. M. Daisey, and P. J. Lioy. Effect of sampling duration on the concentration
of particulate organics collected on glass fiber filters. Am. Ind. Hyg. Assoc. J. 42:
258-263, 1981.
Shephard, B. J. Topographic and meteorological factors influencing air pollution in
Cincinnati. AMA Arch. Ind. Health 19:44-54, 1959,
Shepherd, M. B., G. C. R. Carey and, J. J. Phair: Critical evaluation of a filter-strip smoke
sampler used in domestic premises. AMA Arch. Ind. Health 17:236-252, 1958.
Spedding, D. J. Sorption of sulphur dioxide by indoor suraces — II. Wood. J, Appl. Chem.
20:226-28, 1970.
Spedding, D. J., and R. P. Rowlands. Sorption of sulphur dioxide by indoor surfaces — L
Wallpapers. J. Appl. Chem. 20:143-146, 1970.
s,
Spedding, D. J., R. P. Rowlands, and J. E. Taylor. Sorption of sulphur dioxide by indoor
surfaces — III. Leather. J. Appl. Chem. 21:68-70, 1971.
Spengler, J. D. [Unpublished data from the ''six-cities study."] Harvard School of Public
Health, Department of Environmental, Health Science(s, Boston^ MA, 1980.
5-150
-------
Spengler, J. D., and 0. W. Dockery, Long term measurements of respirable sulfates and
particles inside and outside homes. 1979 Electro Professional Program, New York, NY,
April 25, 1979. Paper no. 2413.
Spengler, J. D., and T. Tosteson. Personal exposures to respirable particles. Presented at
Environmetrics 81, SIAM Institute for Mathematics and Society, Washington, DC, April
8-10, 1981.
Spengler, J. D., B. G. Ferris, Jr., D. W. Dockery, and F. E. Speizer. Sulfur dioxide and
nitrogen dioxide levels inside and outside homes and the implications on health effects
research. Environ. Sci. Techno!. 13: 1276-1280, 1979.
Spengler, J. D., D. W. Dockery, M. P. Reed, T. Tosteson, and P. Quinlan. Personal exposures
to respirable particles. Presented at the 73rd Annual Meeting, Air Pollution Control
Association, Montreal, Quebec, .Canada, June 22-27, 1980. Paper no. 80-61.5b.
Spengler, J. D., D. W. Dockery, W. A. Turner, J. M. Wolfson, and B. G. Ferris, Jr. Long-term
measurements of respirable sulfates and particles inside and outside homes. Atmos.
Environ. 15:2330, 1981.
Spicer, C. W. Photochemical atmospheric pollutants derived from nitrogen oxides. Atmos.
Environ. 11:1089-1095, 1977.
Spicer, C. W., P. M. Schumacher, J. A. Kouyoumjian, and D. W. Joseph. Sampling and Analytical
Methodology for Atmospheric Particulate Nitrates. EPA-600/2-78-067, U. S. Environmental
Protection Agency, Research Triangle Park, NC, April 1978.
Steigerwald, B. J. Particulate control strategy. Office of Air Quality Planning and
Standards, EPA, memorandum to Roger Strelow, Asst. Administrator for Air and Waste
Management, July 2, 1975; submitted by EPA to the Subcommittee, House Committee on
Interstate and Foreign Commerce.
Sterling, T. D. , and D. M. Kobayashi. Exposure to pollutants in enclosed "living spaces."
Environ. Res. 13:1-35, 1977.
Stevens, R. K., T. G. Dzubay, G. Russwurm, and D. Rickel. Sampling and analysis of
atmospheric sulfates and related species. In: Sulfur in the Atmosphere, Proceedings of
the International Symposium, United Nations Environmental Program, Dubrovink, Yugoslavia,
September 7-14, 1977. Atmos. Environ. 12:55-68, 1978.
Stevens, R. K., T. G. Dzubay, R. W. Shaw, Jr., W. A. McClenny, C. W. Lewis, and W. E. Wilson.
Characterization of the aerosol in the Great Smoky Mountains. Environ. Sci. Technol.
14:1491-1498, 1980.
Stevens, R. K.» R. J. Paur, C. W. Lewis, T. G. Dzubay, and W. J. Courtney. Houston aerosol
characterization: EPA measurements. In: Proceedings of a Specialty Conference, Air
Pollution Control Association, Houstonj, Texas, October 14-17, 1979. E. R. Frederick,
ed., Air Pollution Control Association, Pittsburgh, PA, 1979. pp. 212-220.
Stevens, R. K., W. A. McClenny, T. G. Dzubay, M. A. Mason, and W. J. Courtney. Analytical
methods to measure the carbonaceous content of aerosols. In: Particulate Carbon:
Atmospheric Life Cycle, Proceedings of an International Symposium, General Motors Re-
search Laboratories, Warren, MI, October 13-14, 1980. G. T. Wolff and R. L. Klimisch,
eds. , Plenum Press, New York, NY, 1982. Pp. 111-129.
Suck, S. H., E. C. Upchurch, and J. R. Brock. Dust transport in Maricopa County, Arizona.
Atmos. Environ. 12:2265-2271, 1978.
5-151
-------
Suck, S. H., E. C. Upchurch, and J. R. Brock. Dust Transport in Maricopa County, Arizona.
EPA-600/3-79-082, U. S. Environmental Protection Agency, Research Triangle Park, NC,
September 1979.
Tabor, E. C., T. E. Hauser, J. P. Lodge, and R. H. Burttschell. Characteristics of the
organic particulate matter in the atmosphere of certain cities. AMA Arch. Ind. Health
17:58-63, 1958.
Talcott, R., and E. Wei. Airborne mutagens bioassayed in Salmonella typhimurjym. J. Natl.
Cancer Inst. 58:449-451, 1977.
Tanner, R. L. An ambient .experimental study of phase equilibrium in the atmospheric system:
aerosol H+, NH.+, SO, , NO,—NH,(q), HN03(g), Presented at the 180th National Meeting,
American Chemical Society, Las Vegas, NV, August 24-29, 1980a.
Tanner, R. L. Sulfur and nitrogen compounds in urban aerosols. In: Aerosols: Anthropogenic
and Naturan Sources and Transport, Proceedings of a Conference, New York, NY January
9-12, 1979. Ann. N.Y. Acad. Sci. 338:39-49, 1980b.
Tanner, R. L., R. Cederwall, R. Garber, D. Leahy, W. Marlow, R. Meyers, M. Phillips, and L.
Newman. Separation and analysis of airborne sulfate at ambient concentrations. Atmos.
Environ. 11:955-966, 1977.
Tanner, R. L., W. H. Marlow, and L. Newman. Chemical composition correlations of size frac-
tionated sulfate in New York City aerosol. Environ. Sci. Techno!. 13:75-78, 1979.
Terabe, M. Air pollution in Japan: present state area trends. Kuki Seijo 15:1-10, 1977.
Thomas, J. F., M. Mukai, and B. D. Tebbens. Fate of airborne benzo(a)pyrene. Environ. Sci.
Technol, 2:33-39, 1968.
Trijonis, J., J. Eldon, J. Gins, and G. Berglund. Analysis of the St. Louis RAMS Ambient
Particulate Data. Vol. I - Final Report. EPA-450/4-80-006a, U.S. Environmental Pro-
tection Agency, Research Triangle Park, NC, February 1980.
U.S. Environmental Protection Agency. Deputy Assistant Administrator's Report on Ambient
Monitoring Activities—Air Portion. U.S. Environmental Protection Agency, Washington,
DC, 1980a.
U.S. Environmental Protection Agency. Inhalable Particulate Network Annual Report: Opera-
tion and Data Summary. (Mass concentrations only), April 1979 - June 1980. EPA-600/4-
81-037, U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory
Research Triangle Park, NC, May 1981.
U.S. Environmental Protection Agency. National Air Quality, Monitoring and Emissions Trends
Report, 1977. EPA-450/2-78-052, U.S. Environmental Protection Agency, Research Triangle
Park, NC, December 1978.
U.S. Environmental Protection Agency. National Air Quality and Emissions Trends Report, 1976.
EPA-450/1-77-002, U.S. Environmental Protection Agency, Research Triangle Park, NC,
December, 1977.
U.S. Environmental Protection Agency. Trends in the Quality of the Nation's Air—A Report to
the People. OPA16/9, U.S. Environmental Protection Agency, Office of Public Awareness,
Washington, DC, October 1980b.
Van Duuren, B. L. Epoxides, hydroperoxides, and peroxides in air pollution. Int. J. Environ.
Anal. Chem. 1:233-241, 1972.
5-152
-------
Van Vaeck, L., and K. Van Cauwenberghe. Cascade impacted measurements of the size distri-
bution of the major classes of organic pollutants in atmospheric particulate matter.
Atmos. Environ. 12:2229-2239, 1978.
Van Vaeck, L. , and K. Van Cauwenberghe. Differences in particle size distributions on major
organic pollutants in ambient aerosols in urban, rural and seashore areas. Environ. Sci.
Techno!. 13:1494-1502, 1980.
Walsh, M., A. Black, and A. Morgan. Sorption of SOp by typical indoor sources including wool
carpets, wallpaper, and paint. Atmos. Environ. 11:1107-1111, 1977.
Watson, J. G., Jr. Chemical Element Balance-Receptor Methodology for Assessing the Sources of
Fine and Total Suspended Particulate Matter in Portland, Oregon. Ph.D., Thesis, Oregon
Graduate Center, Beaverton, OR, 1979.
Weatherly, M. L. [Untitled.] Air Water Pollut. 10:404-409, 1966.
Wesolowski, J. J. , C. P. Flessel, S. Twiss, J. Cheng, R. Chan, L. Garcia, J. Ondo, F. Fong,
and S. Lum. The chemical and biochemical characterization of particulate matter as part
of an epidemiological cancer study. Proceedings: 1980 Conference on Aerosols in
Science, Medicine and Technology — Physical and Chemical Properties of Aerosols.
Gesellschaft fgr Aerosolforschung, Schmallenberg, West Germany, 1980.
Whelpdale, D. M. Large scale atmospheric sulfur studies in Canjida. Atmos. Environ.
12:661-670, 1978.
Whitby, K. T. Aerosol formation in urban plumes. Ann. N. Y. Acad. Sci. 338:258-275, 1980.
Whitby, K. T., A. B. Algren, R. C. Jordan, and J. C. Annis. The ASHAE air-borne dust survey.
Heat. Piping Air Cond. _29:185-192, 1957.
Wilson, J. C., K. T. Whitby, V. A. Marple, and J. E. McCormack. Characteristics and origins
of coarse particles in the air of northeastern Minnesota. Particle Technology Laboratory
Publication No. 390, University of Minnesota, Minneapolis, MN, April 1979.
Winchester, J. W., W. H. Zoller, R. A. Duce, and C. S. Benson. Lead and halogens in pollution
aerosols and sndw from Fairbanks, Alaska. Atmos. Environ. 1:105-119, 1967.
Wolff, G. T., P. J. Groblicki, S. H. Cadle, and R. J. Countess. Particulate carbon at various
locations in the United States. In: Particulate Carbon: Atmospheric Life Cycle,
Proceedings of an Internation Symposium, General Motors Research Laboratories, Warren MI,
October 13-14, 1980. G. T. Wolff and R. L. Klimisch, eds., Plenum Press, New York, NY,
1982. pp. 297-315.
Yocom, J. E. , E. T. Brookman, R. C. Westman, and 0. Ambardar. Determining the contributions
of traditional and non-traditional sources of particulate matter. J. Air Pollut. Control
Assoc., 31:17-23, 1981.
Yocom, J. E. , W. A. Cote, and W. L. Clink. Summary Report—Study of Indoor-Outdoor Air
Pollution Relationships. National Air Pollution Control Administration Contract No.
CPA-22-69-14, TRC—The Research Corporation of New England, Hartford, CT, May 1970.
Yocom, J. E., W. L. Clink, and W. A. Cote. Indoor/outdoor air quality relationships. J. Air
Pollut. Control Assoc. 21:251-259, 1971.
5-153
-------
-------
6. ATMOSPHERIC TRANSPORT, TRANSFORMATION, AND DEPOSITION
6.1 INTRODUCTION
The preceding chapters discussed the physical and chemical properties of sulfur oxides
and PM (Chapter 2); methods of measuring them (Chapter 3); their sources and emissions
(Chapter 4); and measurements of ambient levels in urban and rural environments (Chapter 5).
These chapters also discussed information relevant to parts of this chapter; whenever pos-
sible, reference is made to resource material in previous chapters. Emissions, which can be
viewed as a model input, are discussed only in the context of their relevancy to air quality
simulation modeling in the final section.
This chapter reviews our knowledge of the physical and chemical processes that contribute
to the transport and diffusion, transformation and deposition of PM and sulfur oxides in the
atmosphere and discusses the theoretical approaches for integrating these processes with
source emission contributions through the use of mathematical models. Such integrating
approaches help improve understanding of the complex processes that operate in polluted
atmospheres. These source-receptor relationships provide a credable scientific basis for
determining the nature and extent of emission control required to meet specified ambient air
quality levels.
The concentration of a pollutant species at a fixed point in time and space after it has
been emitted from a source at another given point depends on four fundamental factors: (1)
emission—the rate of pollutant emitted and the configuration of its source; (2) transforma-
tion—the chemical and physical reaction processes that convert one pollutant species to
another; (3) transport and diffusion—the movement and dilution of a pollutant species through
time and space as a result of various meteorological variables; and (4) deposition—the re-
moval of pollutant species through their interaction with land and water surfaces (dry deposi-
tion) and through interaction with precipitation or cloud droplets (wet deposition).
Figure 6-1 schematically illustrates the principal process pathways of airborne pollut-
ants. Ideally, each of these processes should be treated explicitly in any air quality simu-
lation model, but generally this is not the case.
The modeling approaches discussed here include explicit treatments of the dynamic physi-
cal and chemical atmospheric processes that simulate relationships between pollutant emissions
and ambient air quality. More implicit statistical-empirical approaches, which deduce source
contribution through analysis of empirical information only, are not within the purview of
this chapter. A brief discussion of source-apportionment techniques that have shown consider-
able promise in developing source-receptor relationships for particulate matter is presented
in Section 5.7 of Chapter 5.
6.2 CHEMICAL TRANSFORMATION PROCESSES
Chapter 2 presents a detailed discussion of the chemistry of S0« and other gases that
react to form PM in the atmosphere. Section 6.2.1 summarizes the results of the atmospheric
6-1
-------
FREETROPOSPHERIC
EXCHANGE
VERTICAL
DIFFUSION
AEROSOL
"T BY WIND*1 COMPENSATION
COAGULATION
CHEMICAL REACTIONS
SEDIMENTATION
AS AEROSOL
ABSORPTION IN
CLOUD ELEMENTS
NATURAL
SOURCES
DRY DEPOSITION ON »//////, 777/7,,77///7
THE GROUND ////^/.V'/////'///.'//
O
ANTHROPOGENIC
SOURCES
ABSORPTION IN
PRECIPITATION
WASHOUT IN PRECIPITATION
Figure 6-1. Pathway processes of airborne pollutants.
Source: Adapted from Drake and Barrager(1979).
6-2
-------
chemical transformation processes of SCL and PM presented in Chapter 2. Section 6.2,2 reviews
the status of field measurements of the rate of S0? oxidation in industrial and urban plumes,
and their contribution to elucidating the transformation pathway processes of SO- oxidation.
6.2.1 Chemical Transformation of Sulfur Dioxide and Particulate Hatter
Present understanding of the homogeneous gas phase reactions of S0? indicates that the
rate of S0? oxidation in the atmosphere is dominated by free radical reaction processes. The
free radical species important to the SOp oxidation process are HO, HOp, CHgO-, and other
organic peroxyl species (RCL, R'0?, etc.). The concentrations of these radicals in the atmos-
phere depend on many factors, the more important of which are the concentrations of volatile
organic compounds and nitrogen oxides (NO and NCL) in the atmosphere, temperature and solar
intensity. Theoretical estimates have shown that maximum S0? oxidation rates of 4.0 percent/h
are possible in polluted atmospheres. Recent experimental rate constant determinations for
the H0? and CH,0, reactions with S0?, however, indicate that these processes may not be as
important as previously thought, and that the maximum,possible homogeneous S0? oxidation rate
under optimum atmospheric conditions may only be of the order of 1.5 percent/h. This rate is
a result of S0? reaction with HO radical only.
Present knowledge of heterogeneous pathways to S02 oxidation in the atmosphere indicates
2+ 3+
that the liquid phase catalyzed oxidation of S0? by Mn , Fe , and carbon are potentially
important processes, as is oxidation by hydrogen peroxide. Theoretical estimates of atmo-
spheric S02 oxidation rates by these processes are of the order of 10 percent/h. Unfortun-
ately, the actual availability of these catalyzing substances in ambient fine PM is uncertain.
The quantitative determination of rates of S0? oxidation by these processes has never been
demonstrated under actual atmospheric conditions.
Processes that form organic and nitrate particles are thought to be dominated by homo-
geneous gas phase reactions. In the case of atmospheric nitrates, a significant production
pathway is through reaction between HO free radical and N0_, resulting in nitric acid (MONO,)
formation. The fate of nitric acid in the atmosphere is not well understood, though a portion
of gaseous nitric acid is known to enter into an equilibrium with NH^ to form particulate
NH.NO_. Information on the production rates and mechanism details of organic particulate
matter is limited. Available product information indicates that oxidation reactions involving
the interaction of ozone, nitrogen oxides, and HO free radicals with higher molecular weight
organics represent a major pathway to organic particle production.
6.2.2 Field Measurements on the Rate of Sulfur Dioxide Oxidation
The majority of atmospheric S0? oxidation studies have been carried out only in recent
years, and most have involved power plant plumes. One reason for the late start in this re-
search was the lack of adequate measurement technology for particulate sulfur, but recent
developments (e.g., Huntzicker et al., 1978; Cobourn et al. , 1978) seem to have alleviated
this problem (see Chapter 3). Table 6-1 summarizes S0? oxidation rates, based on field
measurements in power plant, smelter, and urban plume studies carried out from 1975 to 1980.
6-3
-------
TABLE 6-1. FIELD MEASUREMENTS ON THE RATES OF S02 OXIDATION IN PLUMES"
Plume type/
location
SO,, oxidation
rate (percent/h)
Method
Reference
Power plants
KeybLone
(Pennsylvania)
L abaci ie
(Missouri)
0-10
0.41-4.9
Four Corners 0.27-0.84
(New Mexico)
Labadie and
Portage des Sioux
(Missouri)
Muscle Shoals
(Alabama)
Kycjer Creek
(Ohio)
LabaiUe
(Missouri)
Four Corners
(New Mexico)
Labadie
(Missouri)
Cumberland
(Tennessee)
Great Canadian
Oil Sands
(Alberta, Canada)
Keystone
(Pennsylvania)
Centralia
(Washington)
Four Corners
(New Mexico)
Four Corners
(New Mexico)
0-5
0-3
2-8
0-4
0-7
0-3
0-5
0-6
0.15-0.5
32 34
S/ S,ratio, change with
oxidation
Total change in particle
volume
Submicron sulfate and
S02 - change of ratio
with time
Newman et al. (1975)
Cantrell and Whitby (1978)
Ursenbach et al. (1977)
Particulate sulfur to total Forr%st and Newman (1977)
sulfur ratio
Particulate sulfur to total Gillani et al. (1978)
sulfur ratio
CCNb production (CCN to S02 Pueschel and Van Valin (1978)
ratios)
Particulate sulfur to total Husar et al. (1978)
sulfur ratio
Particulate sulfur to total Meagher et al. (1978)
sulfur ratio
Particulate sulfur to total Lusis et al. (1978)
sulfur ratio
Particulate sulfur to total Dittenhoefer and dePena (1978)
sulfur ratio
Total change in particle
volume
Hobbs et al. (1979)
CCN production (CCN to S02 Mamane and Pueschel (1980)
ratios)
6-4
-------
TABLE 6-1. (continued)
Plume type/
location
S0« oxidation
rate (percent/h)
Method
Reference
Leiand-Qids
(North Dakota)
Sherburne County 0-5.7
(Minnesota)
Big Brown
(Texas)
Smelters
INCO Nickel 0-7
(Copper Cliff,
Canada)
INCO Nickel 1.2-5.2
(Copper Cliff,
Canada)
MT ISA Mines 0.25C
(MT ISA, Australia)
Urban
Total change in particle
volume
Particulate sulfur to total
sulfur ratio
Particulate sulfur to total
sulfur ratio
Particulate sulfur to
lead ratio
Hegg and Hobbs (1980)
Lusis and Wiebe (1976)
Forrest and Newman (1977)
Roberts and Williams (1979)
Los Angeles
(California)
St. Louis
(Missouri)
St. Louis
(Missouri)
1 2~13 Particulate sulfur to total
sulfur ratio
7-12.5 Particulate sulfur to total
sulfur ratio
3.6-4.2 Particulate sulfur to total
sulfur ratio
Roberts and Friedlander (1975)
Alkezweeny and Powell (1977)
Chang (1979)
Adapted in part from Hegg and Hobbs (1980)
Cloud condensation nuclei
cDiurnal average rate
6-5
-------
The rates of SO, oxidation in industrial plant plumes consistently range from 0 to 10 per-
cent/h, with urban plumes showing only a slightly greater maximum rate of 13 percent/h. The
pre-1975 studies (Gartrell et al., 1963; Dennis et al., 1969; Weber, 1970; and Stephens and
McCaldin, 1971), which observed conversion rates an order of magnitude larger than more recent
observations, are suspect due to possible artifact formation in the sulfate analysis technique
and limitations in the analytical methods in general.
Newman (1980)' reviewed the majority of the power plant and smelter plume studies pre-
sented in Table 6-1 and arrived at the following conclusions:
1. The diurnal average oxidation rate of S09 to sulfate is probably less than 1 per-
cent/h. £
2. Little or no oxidation of SCL occurs from early evening to early morning.
3, Maximum oxidation rates of S0? to sulfate of 3 percent/h can occur under midday con-
ditions.
4. The separate contribution of homogeneous and heterogeneous mechanisms to S0? oxida-
tion in plumes cannot be deduced from the present studies.
The reported SO,, oxidation rates are estimates based on analyses of measured physical and
chemical parameters and in many instances have incorporated certain simplifying assumptions
that are not totally substantiated. Also, present understanding of S0? chemical transforma-
tion processes indicate that SCL oxidation rates can vary significantly as a result of differ-
ences in the composition of source plumes and the air masses into which the plumes enter.
Typical experimental uncertainties in measured S0? oxidation rates reported in Table 6-1 are
50 percent, but may be greater if inappropriate assumptions have been used. Even with these
uncertainties, the overall consistency in the observed range of S0? oxidation rates is grati-
fying.
6.3 PHYSICAL REMOVAL PROCESSES
The removal of PM and gases from the atmosphere generally occurs through two physical
processes: (1) dry deposition—the removal of chemical species from the atmosphere at the air-
surface interface; and (2) precipitation scavenging—the removal of chemical species from the
atmosphere by interaction with various types of precipitation such as rain, snow, etc. These
processes have both a positive and negative impact on environmental air pollution issues. On
one hand, they constitute the major mechanisms by which the polluted atmosphere cleanses it-
self, lowering ambient air concentrations of pollutant species and thereby reducing health-
related risks. On the other hand, the deposited pollutant materials may constitute increased
risks to our terrestrial and aquatic ecosystems.
Since wet and dry removal processes significantly affect the lifetime of SO, and PM in
the atmosphere and thereby affect the distance traveled and the concentration of these species,
understanding these processes is essential for a proper assessment of their environmental
significance. The removal of pollutant species by dewfall has not been studied, and it
remains for future research to determine whether this process is an important removal
mechanism for atmospheric contaminants.
6-6
-------
In the sections to follow, dry deposition and precipitation scavenging are discussed with
emphasis on experimental data bases and theoretical treatment.
6.3.1 Dry Deposition
Sehmel (1980), HcMahon and Dem'son (1979), Chamberlain (1980), and Garland (1978) re-
viewed particle and gas dry deposition. The dry deposition of S0? and PM, like other atmos-
pheric species, is governed by three major components: meteorological variables, properties of
the depositing pollutant, and surface variables. These components are influenced by specific
parameters that interact in complex ways, which in many instances are not completely under-
stood.
The most important meteorological processes affecting dry deposition are transport-
related phenomena, which are governed by the wind and temperature profiles; eddy diffusion;
and sedimentation across the boundary layer to the vegetation canopy or the surface. Two
meteorological parameters strongly influence these processes: the friction velocity (u*) and
the aerodynamic surface roughness (z ). Both of these parameters are used to describe the
windspeed profile above a given surface under given conditions of atmospheric stability.
Typically, these two variables are determined empirically by fitting optimal curves to wind-
speed data as a function of height. The strong diurnal dependence of dry deposition is linked
to the formation of a stable layer of air at the earth's surface at night (nocturnal inver-
sion) that effectively inhibits the vertical transport of pollutant species to the canopy or
the surface. The formation of the nocturnal inversion and its effect on other atmospheric
processes is discussed in the section on transport and diffusion.
Solubility in water is an important property influencing the dry deposition of a pollut-
ant. Other important characteristics of PM are size distribution, density, morphology, and
composition. Among the important surface properties are: (1) the moisture content of the sur-
face which, in conjunction with the solubility of the pollutant species, govern the overall
sticking efficiency of the deposited material; and (2) the physiological state of the vegeta-
tion surface, especially the opening and closing of stomatal pores, where the rate of pollut-
ant uptake is thought to be strongly governed.
Chamberlain and Chadwick (1953) introduced a convenient way to express the rate of dry
deposition of both gases and particles in terms of velocity. Dry deposition velocity (V )
(defined as the downward flux (F) of the species, divided by its ambient concentration (x) at
some specified height [typically 1 to 2 meters above the surface]), is the standard form in
which all measured deposition rates are reported. Dry deposition velocities typically are
reported in units of cm sec
v = "F
g *
Dry deposition velocity is positive by convention and therefore requires a minus sign on
F, the downward flux, which is defined as negative.
6-7
-------
Hicks et al. (1980) reviewed and evaluated the measurement techniques for the dry deposi-
tion of pollutant species. They sorted measurement methods into three major categories: (1)
estimates of accumulation, (2) flux monitoring, and (3) flux parameterization.
Though none of the experimental techniques has proven to be useful in all dry deposition
measurements, a general consensus has been reached on the overall accuracy of the methods and
their suitability for specific applications. Based on Hicks et al. (1980), the three cate-
gories are described briefly, with general comments on their limitations.
Estimates of accumulation may be considered using atmospheric radioactivity or mass
balance methods. Radioactive techniques compare ambient concentrations of selected radioac-
tive species with concentrations in water bodies, vegetation, etc., to evaluate the rate at
which material enters the ecosystem over long periods. The technique generally is limited to
small particle uptake of long-lived species and has difficulty distinguishing between dry and
wet removal and resolving short-term variations. Mass balance studies attempt to measure the
various inflow and outflow processes in the ecosystem, with the exception of dry deposition,
which is then determined through a budget calculation. The method's major limitation is that
dry deposition is inferred by indirect measurements, which in themselves are difficult to make
accurately.
Flux monitoring considers the direct measurement of total deposition over a well-defined
surface for set periods. Several types of deposition surfaces have been used with this
general category, including open pots, flat filters, flat plates and shallow pans, fiber
filters, and sticky films. Overall, the methods are limited due to their lack of standardiza-
tion, dissimilarity to natural surfaces, and potential for contamination by locally re-
suspended particles.
Flux parameterization includes a variety of methods, one of which, eddy correlation,
shows promise as a measurement standard for dry deposition of gases. Eddy correlation re-
quires the simultaneous measurement of the concentration of pollutant species and the vertical
component of the wind velocity at a sufficiently fast rate to determine the turbulent flux of
the pollutant. The lack of adequate fast-response instruments for many of the pollutant
species of interest significantly limits the technique. In addition, particle flux due to
gravitational settling is not detected, and this can produce invalid results if significant
particle resuspension occurs. Laboratory methods, including chamber and wind tunnel studies
are also considered tinder the flux parameterization category. In these controlled experi-
mental studies, plants, leaves, simulated canopy surfaces, etc., are exposed to known pollut-
ant concentrations. Measurements of the change in concentration, which can be accomplished by
a variety of methods, are then used to determine pollutant uptake.
6.3.1.1 Sulfur Dioxide Dry Deposition—The dry deposition of SO, onto grass, crops, forests,
soil, and building surfaces has been reviewed by Sehmel (1980), McMahon and Denison (1979),
Chamberlain (1980), and Garland (1978). McMahon and Denison (1979) presented compilations of
dry deposition laboratory and field measurements of S0?. A review of these results indicates
6-8
-------
measured dry deposition velocities ranging from 0.04 to 3.7 cm sec , but with the majority
in the range from 0,3 to 1.6 cm sec . The apparent wide range of dry deposition values
is not particularly disconcerting when the variety of surfaces, meteorological conditions, and
experimental methods are considered. Table 6-2 summarizes the average dry deposition velocity
by surface type.
TABLE 6-2. AVERAGE DRY DEPOSITION VELOCITY OF S02 BY SURFACE TYPE
Surface
Alfalfa
Grass
Wheat
Forest
Sandy Soil
Clayey Soil
Soil
Land
Water (Fresh)
Ocean
Snow
Laboratory measurement,
v (cm sec )
1.2 (2)
__
—
—
0.6 (2)
0.8 (2)
--
—
--
—
..
Field measurement,
v (cm sec )
1.6 ( 2)
1.1 (14)
0.4 ( 3)
1.4 ( 5)
--
—
1-2 ( 4)
1.2 ( 4)
1.1 ( 6)
0.5 ( 2)
0.3 ( 2)
Note; Values in parentheses indicate the number of separate studies used to obtain the
average deposition velocity.
Source: McMahon and Denison (1979).
After reviewing the same set of data, Garland (1978) concluded that the mean deposition
velocities for S0? over surfaces ranging from water and soil through short grass to forest
^ -^
were very similar and suggested that a value of about 0.8 cm sec was applicable to large
areas of Europe.
In a more detailed effort to estimate dry deposition velocities of S0? and particulate
sulfate over the eastern half of the United States, southern Ontario, and nearby oceanic
regions, Sheih et al. (1979) computed deposition velocities as a function of land use
characteristics, surface roughness scale lengths, and,surface resistances to pollutant uptake.
Gridded dry deposition velocity maps of sulfur dioxide and sulfate corresponding to half
degree increments of longitude and latitude were computed for a range of atmospheric stabili-
ties. The results indicate that deposition velocity distributions for SO, are not uniform for
the less stable atmospheric conditions. For very unstable atmospheric conditions (Pasquill
category A) dry deposition velocities over the eastern United States ranged from 0.4 to 0.9 cm
sec (excluding water surfaces) for SO,, with a mean areawide dry deposition velocity of
-1
approximately 0.6 cm sec . Under the same conditions, the deposition velocity for sulfates
ranged from 0.7 to 0 9 cm sec with a mean value of approximately 0.8 cm sec . Sheih et al.
(1979) noted that, under nearly calm conditions at night, stability classification schemes do
6-9
-------
not adequately represent the nocturnal inversion formed at the surface and recommend that a
dry deposition velocity of 0.07 cm sec be assumed for both SCL and sulfate particles.
6.3.1,2 Particle Dry Deposition—As a measurement for pollutant species of interest, the dry
deposition of particulate matter is less understood. In Sehmel (1980) and McMahon and Denison
(1979), deposition velocities for particle species are compiled for both artificial and
natural surfaces. Unfortunately, with the exception of lead particles from automotive ex-
haust, virtually no data exist for other particulate pollutants, such as sulfates, nitrates,
and carbon-containing particles. The interpretation of the relationship between deposition
measurements with fallout collectors and dry deposition rates on natural surfaces creates an
additional problem. Fallout collectors, which were used in a significant portion of the
measurements reported, generally do not have the characteristics of the surfaces they are
attempting to simulate.
Tables 6-3 and 6-4 from the study by McMahon and Denison (1979) compile literature values
for the deposition velocities of particles measured under laboratory and field conditions,
respectively. The data cover a range of surface variables, particle sizes and composition,
and meteorological conditions. A review of the data indicates:
1. Deposition in nature varies considerably through processes that are not totally
understood.
2. The minimum deposition velocity for particles occurs at diameters from 0.1 to 1.0
urn.
3. Deposition velocities are often reported for particle diameters and size distribu-
tions that do not reflect typical atmospheric PM characteristics.
The experimental uncertainties associated with particle dry deposition velocity measure-
ment have stimulated the development of theoretical models for simulating the dry deposition
process and predicting dry deposition velocities given specific meteorological data (Sehmel,
1980; Slinn, 1978, 1977; Davidson and Friedlander, 1978). The models describe only the physi-
cal processes of bringing the particle to the depositing surface. Particle shapes other than
spherical, particle composition, or surface properties have not been considered with regard to
particle retention. Particle size, an important property in the aerodynamic flow of particles
to surfaces, is considered in these studies. The typical model result shows that predicted
deposition velocities increase as surface roughness and/or friction velocity increases and are
nearly independent of atmospheric stability. The deposition velocity for particles passes
through a minimum in the 0.1 to 1 ^m diameter particle range. Figure 6-2 (Sehmel, 1980) pre-
sents deposition velocities predicted by one model at 1 m from the surface, for U* = 30 cm
sec and particle densities of 1, 4, and 11.5 g/cm .
Table 6-5 shows, the range of predicted deposition velocities at a height of 1 m for two
particle size regions and for a range of aerodynamic surface roughness lengths, mean wind-
speeds, and calculated friction velocities. These results are based on the model of Sehmel
(1980) and should be representative of most meteorological and surface conditions.
6-10
-------
TABLE 6-3. LABORATORY MEASUREMENTS OF DEPOSITION VELOCITIES OF PARTICLES
Author
(date)
Chamberlain (1967)
Moller and Schumann
(1970)
Chamberlain and
Chadwick (1972)
Clough (1973)
Sehmel (1973)
Sehrael and Sutter (1974)
Belot and Gauthier
(1975)
Klepper and Craig (1975)
Craig et al. (1976)
Reference
v height
(cm sec ) (m)
0.03
0.03
0.1
0.8
2/3
V « D '
g
v =0.06 u*
V9=0.12 u*
0.005
0.003 0.1
0.3
2
2 x 10"3-10 0,01
5 x 10~3-29 0.01
vga"4
vgd
0.0035
0.01
Particle
diameter
((jm) Surface
0.1
i Grass
5
20-30 Cereal crops
0.08
0. 5 Filter paper
5
20
0.1-28 Smooth brass
0.2-30 Water
1-10 Shoots of pine
and oak trees
0.8 Bean leaves
0.1 1 Smooth
Comment
~—
D = diffusion coefficient ~ ,
2 x 10"^ > D > 2 x 10 cnTsec
Dry Includes wind tunnel
Wet and field data
v to copper also measured
v^ found to be a function of
* windspeed
—
—
u = wind speed
d = particle diameter
—
Wind tunnel
Wedding et al. (1976)
Deposition rate on pubescent
leaves of sunflower was nearly
7 times that of the non-
pubescent leaves of tulip
poplar.
-------
TABLE 6-3. (continued)
Author
(date)
Little and Wiffen (1977)
Little (1977)
vg_
(cm sec )
0.11
0.02
0.5
0.04
0.3
0.9
0.1
0.3
1.5
0.3
0.8
Reference Particle
height diameter
(m) (pro)
5 x 10"2
0.2
o 71:
5
8.5
Surface
Short grass
Nettle
Beech
White Poplar
Nettle
Beech
White Poplar
Nettle
Beech
White Poplar
Comment
__
These data are for whole
shoots for..wind speeds of
2.5 m sec . Data for other
wind speeds and separate
plant surfaces are given in
reference.
Source: McMahon and Denison (1979).
I
t—»
rsa
-------
TABLE 6-4. FIELD MEASUREMENTS OF DEPOSITION VELOCITIES OF PARTICLES
cr>
i
Author
(date)
Chamberlain (1953)
Eriksson (1959)
Small (1960)
Neuberger et al. (1967)
White and Turner (1970)
Esmen and Corn (1971)
_#
Chamberlain and
Chadwick (1972)
Pierson et al. (1973)
vg_
(cm sec )
2.1
1.1
0.5
0.7
1.6
0,5
(0.2-3.4)
5.6
4.7
3.0
7.1
0.8
v = 0.50
9
v = 0.06 M*
vj= 0.12 M*
0.1-0.6
Reference
height Particle
(m) composition
0.3-0.9 16 pm*
0.3-0.9 16 Mm*
0.3-0.9 16 Mm*
Ragweed
Na
K
Ca
Mg
P
0.1-10 pro*
20-30 M"i*
Surface
Grass
Ocean
Land
1 anrl
L-u 1 114
Coniferous
forest
Mixed
deciduous
woodland
Filter paper
Millipore filter
Glass slide
Cereal crops
Land
Comment
u = 9.2 m/sec
u = 3.2 m/sec
u = 1.1 m/sec
Chloride over Scandinavia
— —
Radioactive particles over
Norway
80 percent ragweed pollen re-
moved from air by forest
1. Probable overestimation of
aerosol income, hence v .
2. Standard deviation varied
between 65 and 95 percent of
mean v .
g
—
Dry, Includes wind tunnel and
Wet* field data.
v estimated for 23 trace
g
elements based on several
years of data
-------
TABLE 6-4. (continued)
en
i
Author
(date)
Cawse (1974)
Hart and Parent (1974)
Clough (1975)
vg_
(cm sec )
1.3
0.22
(0.45)
0.50
(0.50)
1.1
0.56
(0.45)
0.30
(1.0)
0.29
0.62
3.4
7.3
11
61
100
0.74
1.1
0.75
12.7
Reference
height Particle
(ra) composition
Al
As
Cd
Cr
Cu
Fe
Mn
Ni
Pb
Ti
V
Zn
Na, Ca,
Mg, K,
P, N03
30 jjm*
4 pi"*
3jb
|jm*
Surface
Douglas fir
and junipers
Grass
Grass
Grass
Dry moss
Wet moss
Grass
Grass
Dry moss
Comment
Extracted from Gatz (1975).
Values in parentheses were
estimated by Gatz from a
relationship between particle
size and v .
g
Deposition ratio:
beneath trees _ ~ ,,.
open terrain
Dry u* = 37 cm sec ,
Dry u* = 87 cm sec ,
Wet u* = 87 cm sec
Dry u,< = 37 cm sec
Dry u* = 37 cm sec
*particle diameter
-------
TABLE 6-4. (continued)
Author
(date)
Abrahamsen et al. (1976)
Dovland and Eliassen
(1976)
Fritschen and Edmonds (1976)
Prahm et al. (1976)
Krey and Toonkel (1977)
Wesley et al. (1977)
vg_
(cm sec )
0.16
0.68
0.07
0.46
0.4
0.5
0.6
Reference
height Particle
(m) composition
S°4?
Atmospheric
aerosol
3 (jm*
Atmospheric
aerosol
5 0.05-0.1 |jm*
Surface
Spruce and
pines
Snow
Douglas fir
Atlantic
ocean
Bare soil
and grass
Comment
Deposition ratio:
beneath trees _ 0
open terrain
Lead
SO. : upper bound value
so;2
90S : HASL wet-dry collector
u < 2 m sec : Eddy correlatio
method.
Source: McMahon and Denison (1979).
-------
= 111 HIM) I I I IMII| I I I IIIIIj I I I II
UPPER LIMIT
NO RESISTANCE BELOW AND
ATMOSPHERIC DIFFUSION FROM
1 cm TO 1 m
\
STABLE ATMOSPHERE
WITH ROUGHNESS
HEIGHT, cm
p = PARTICLE DENSITY —
0 = ROUGHNESS HEIGHT ~
= FRICTION VELOCITY
I I I I Hill S\ I 1 Hill 1 I I Hill I I 1 I Hill I I I 111II
10
10"' 1
PARTICLE DIAMETER,
Figure 6-2. Predicted deposition velocities at 1 m for /n. = 30 cm s and particle
"3
densities of 1,4, and 11.5 g cm""1.
Source: Sehmel (1980).
6-16
-------
TABLE 6-5. PREDICTED PARTICLE DEPOSITION VELOCITIES3
Deposition Velocity Range
Z U/UK
cm
0.1
10
0.1
10
m sec
2.
1.
11.
5.
/cm sec
3/10
2/10
5/50
8/50
0.
1.5x1-
9 0x10
2 0x10
1 0x10
1 to
cm
-2 _
-2
~2 _
-1
Particle Diameter
1
sec
5
1.
5,
2.
pro
-1
0x10
5x10
5x10
0x10
-2
-1
-2
-1
1 M
5.
1,
5.
2.
to 10 ur
-1
cm sec
0x10"? -
5xlO~2 -
Oxio"1 -
n
4
4
4
4
.Based on model predictions in.,Sehmel (1980)
Particle density of 11.5 g/cm
6.3.2 Precipitation Scavenging ,
As with dry deposition, precipitation scavenging or wet removal results from a series of
complex physical and chemical interactions involving properties of the scavenging media and
the species removed. For the past 30 years, research in the area has focused on removing from
the atmosphere radioactive debris introduced by nuclear weapons testing (Bowen, 1960;
Engelmann, 1968; Volchok et a!., 1971) and in conjunction with material balance or budget
studies on the removal of various elemental species from the atmosphere (Robinson and Robbins,
1970; Rasmussen et al. , 1975; Junge, 1972, 1974).
Engelmann (1968), Postma (1970), Hales (1972), SI inn et al. (1978) and SI inn (1981) have
researched and reviewed the theory of precipitation scavenging. Though our understanding of
the details of the complex processes operating in precipitation scavenging is incomplete," sig-
nificant progress has been made in deducing the general scavenging pathways and in developing
appropriate parameters for their quantitative treatment. As pointed out by SI inn (1981) and
others, the removal of trace constituents from the atmosphere by precipitation scavenging
depends on: (1) the position of the trace consitituent relative to the scavenging media; (2)
the physical form of the scavenging media; (3) the chemical and physical properties of the
trace constituent; and (4) the specific physical/chemical process that is operative These
basic factors are schematically illustrated in Figure 6-3.
A convenient practice in evaluation of precipitation scavenging is to distinguish between
below-cloud and in-cloud scavenging processes. Unfortunately, use of the terms "rainout" for
in-cloud and "washout" for below-cloud scavenging, has led to confusion. It is difficult to
clarify the contribution of these processes to the total scavenging during precipitation.
Washout, which is easier to study, has received more scientific attention. In many experi-
mental studies, the distinction between the two processes has been ignored and only the total
precipitation scavenging has been considered. The theoretical approaches discussed in the
6-17
-------
CTl
I
\. > 1
IN-CLOUD SCAVENGING
(RAIN OUTI '
PRECIPITATION MEDIA: e.g., RAIN, SNOW
POLLUTANT PARAMETERS: e.g., GAS, PARTICLE, SOLUBILITY
SCAVENGING PROCESSES: e.g., GAS-LIQUID MASS TRANSFER
DIFFUSION, IMPACTION
"\XV^>\^
n
\ \\ BELOW-CLOUD SCAVENGING
Figure 6-3. Basic factors influencing precipitation scavenging.
Source: Adapted from Slinn (1981).
-------
following section are for washout processes only, while the empirical parameterizations con-
sider total gas and particle scavenging.
The parameterization of the precipitation scavenging process has generally taken the form
of a loss rate per unit volume and has evolved from various assumptions applied to the conti-
nuity equation (SI inn, 1977) Parameterizations for the removal of gases by rain and the re-
moval of particles by rain,and by snow are considered in the following sections. The formal-
ism and technical rigor used in their development are discussed elsewhere (Hales, 1972, 1978;
Slinn, 1977) and are beyond the scope of the present discussion.
6.3.2.1 Sulfur Dioxide Wet Removal—The removal of SO- from the atmosphere by rain is governed
by basic physical processes of absorption and desorption of the SO- molecules from the hydro-
meteor (Hales, 1972, 1978) and by a series of chemical reactions (Postma, 1970; Hill and
Adamowicz, 1977; Barrie, 1978) that account for the liquid phase oxidation of SO-. As a
physical process, the rate of scavenging of a gas by rain is a function of the size spectrum
of the rain droplets, the fall path of the rain droplet to the ground, the rainfall rate, and
the solubility of the gas. Hales (1978) and Slinn et al. (1978) developed general expressions
for computing the scavenging of SO,, and other gases, given simplifying assumptions on the
character of the precipitation and solubility of the gas. As pointed out in Chapter 2, the
liquid phase oxidation of SO- is complex and not thoroughly understood. The uptake of SO- by
rain droplets proceeds through the dissolution of SO,, and a subsequent series of dissociation
reactions. The chemical equilibria and associated reaction rates for the dilute sulfur
dioxide-pure water system are well known and the. reactions sufficiently fast that thermo-
dynamic equilibrium between gas phase and liquid droplet phase can be assumed. Therefore, the
physical dissolution of SO- in rain droplets follows a Henry's Law relationship that predicts
reversible absorption to a degree that depends on the physical and chemical state of the rain
droplet and the ambient atmosphere through which it travels (Hales, 1972). Treatment of these
processes is relatively straightforward, but there is uncertainty in the SO- wet removal
process because of the effect of trace substances on the SO- dissociation equilibria and the
effective oxidation rate of SO- in solution. In this light, several recent studies by Hill
and Adamowicz (1977), Barrie (1978), Garland (1978), and"Gravenhorst et al. (1978) have con-
sidered the 50,,-bisulfite oxidation process in predicting wet removal rates of SO,, under
various atmospheric conditions.
Barrie (1978) modeled the washout of SO- from a plume under varying meteorological and
SO- concentration conditions. He assumed that SO- oxidation in the raindrops could be
neglected due to the limited time for reaction (0 to 5 min), the low pH's encountered (Beilke
et al. 1975), and the dissolution of the SO- governing pH of the raindrop. He concluded that
the fractional plume washout rate (percent/mm rain) is inversely related to the plume concen-
tration and thickness. That is, for a given precipitation rate, the plume washout rate
(percent/h) increases with decreasing plume concentration or decreasing plume thickness. For
heavy rain (25 mm/h), washout from a 1000 ppb(v) SO- plume of 20 m thickness occurs at a rate
6-19
-------
of 56 .percent/h, while under drizzle conditions (0.5 mm/h) for a 300 ppb(v) S0? plume of 50 m
thickness the rate was 2 percent/h.
In a more explicit treatment of S0? washout, Hill and Adamowicz (1977) accounted for the
effects of S02 oxidation within rain droplets and of the pH of precipitation on the SO,, wash-
out process. They indicated that pH can be quite variable (over six orders of magnitude in
hydronium ion concentration, equivalent to six pH units) at SO,, ambient levels of 10 ppb and
less. As SO- levels increase, the variability in background pH decreases. The SO,, oxidation
rate of 3.6 percent/h used in the calculations is based on the catalytic oxidation studies of
Brimblecombe and Spedding (1974). In a typical calculation of the rate of SO,, washout, Hill
and Adamowicz (1977) assumed various ambient SO,, concentrations, well mixed through a layer 1
km in depth, and a rainfall rate of 1 mm/h, with a predominant drop radius of 0.5 mm and a pH
of 7. Calculated washout rates of SO,, under these conditions were 2.6 percent/h and 0.8 per-
cent/h for ambient S02 levels of 10 ppb and 100 ppb (30 and 260 (jg/m3). respectively.
A convenient empirical expression for the wet removal of gases takes the form of an expo-
nential decay process, where the time constant for decay (scavenging coefficient for the gas,),
determined in field and laboratory studies, is a function of the rainfall intensity. The ex-
pression takes the form
(-At)
xt = V
where xt and x_ are the atmospheric concentrations of the gas at time t and zero, respective-
v O j
ly, and A is the scavenging coefficient for the gas. Chamberlain (1953), Beilke (1970),
Hales et al. (1971), Dana et al. (1975) and others have reported estimates of the-scavenging
coefficient for sulfur dioxide. Calculated scavenging rates of S02 using these coefficients
can range typically from 2 percent/h to 22 percent/h.
6.3.2.2 Particle Wet Removal—The study of precipitation scavenging of particles has focused
on theoretical studies (Slinn, 1977; Grover et al., 1977; Wang et al., 1978), but emphasis in
experimental work has taken hold in recent years (Dana and Hales, 1976; Radke et al. , 1980;
Gatz, 1977). The wet removal of sulfate PM in the ambient environment has been of particular
interest (Scott, 1978; Hales, 1978; Dana, 1980) due to the acidity of many of these particles
and the increased concern for the phenomenon termed "acid rain." Acidic precipitation and its
associated scientific issues are discussed in Chapter 7.
As with gases, the size spectrum of the rain droplets, the fall path of the-rain droplet
to the ground, and the rainfall rate, as well as the size distribution and composition of the
particulate matter affect the particle scavenging rate by precipitation. Slinn (1977)
developed general expressions for computing the scavenging of particles given certain simpli- .>
fying assumptions, as has been done with gases.
A practical approach in predicting the wet removal of particles, as mentioned previously
for gases, has been through the measurement of empirical scavenging coeFficients. McMahon and
Denison (1979) compiled a comprehensive list of field measurements of wet scavenging
coefficients of particles, which is presented in Table 6-6. A cautionary note is in order.
6-20
-------
TABLE 6-6. FIELD MEASUREMENTS OF SCAVENGING COEFFICIENTS OF PARTICLES
Author
(date)
Kalkstein et al . (1959)
Georgii (1963)
Banerji and Chatter jee (1964)
Makhon'ko (1964)
Shirvaikar et al. (1960)
Makhon'ko and Dmitrieva
(1966)
Makhon'ko (1967)
Wolf and Dana (1969)
Bakulin et al. (1970)
Burtsev et al . (1970)
Dana (1970)
Perkins et al. (1970)
Peterson and Crawford
(1970)
Esmen (1972)
AP
(sec"1)
2 x 10~|?
2 x 10"5
4 x 10~|?
22 x 10";?
4 x 10
0.4 x 10"5
2 x 10~j?
< 1 x 10"3
7 x 10"5
20 x 10"5
7 x 10"5
0.5 x 10"5
3 x 10"5
15 x 10~|?
20 x 10
13 x 10"5
300 x 10"5
16 x 10"5
0.4 x 10~5
Particle
composition
Cl^NO™-'
Dissolved
inorganic
contaminant
Radon
Fission
products
Atmospheric
dust
Fission
products
Atmospheric
dust
J 0.5 urn*
J0'l ^0.2 urn*
JU'D -0.2 urn*
J 7.5,3 urn*
Atmospheric
aerosol
,0.5 ,- *
J 5 urrr
Atmospheric
aerosol
Comment
Rainout As-. calculated by
Washout Makhon'ko (1967)
Rainout
Washout
Rainout As calculated by
Makhon'ko (1967)
Rainout
Washout
Rainout As calculated by
Makhon'ko (1967)
Rainout
Rainout plus washout
Snowout, see also Knutson
and Stockham (1977)
Pb; washout from
thunderstorm
Washout
Rainout
Uranine and rhodamine
particles respectively
Rainout
Based on Engelmann's
data (1965)
Includes rainout
6-21
-------
TABLE 6-6. (continued)
Author
(date)
Rodhe and Grandell (1972)
Acres-AESC (1974)
Graedel and Franey (1975)
Hicks (1976)
Graedel and Franey (1977)
Radke et al. (1977)
AP
(sec"1)
Particle
composition
Comment
Suggest A proportional
to rainfall intensity
0.7
snow
50
19
18
28
43
65
92
x 10
= 25-50Ar
x 10"5
x 10 R
x 10"^
x 10"?
x 10 ^
x 10~?
x 10 D
Atmospheric
aerosol
•ain O-4"1 Mm*
< 1 M"1*
0.3-0.5 urn*
0.5-0.7 |jm*
0.7-0.9 Mm*
0.9"1.5 M"1*
1.5-3 Mm*
Includes rai
nout
See SI inn (1976)
Rainout
Condensation
Snow
See Fig. 6.4
nuclei
0 = rainfall intensity in mm/h
^particle diameter
Source: HcHahon and Denison (1977).
6-22
-------
The scavenging coefficients are dependent on the rainfall rate, the mean raindrop radius, and
the particle size. When these factors are considered, the scavenging coefficients reported in
Table 6-6 show reasonable consistency, as demonstrated by Figure 6-4.
Airborne measurements by Radke et al. (1980) on precipitation scavenging of aerosol par-
ticles greater than 0.01 urn diameter in aged air masses, coal-fired power plant plumes, a
kraft paper mill, and a plume from a volcanic eruption supported theoretical estimates of wet
removal for aerosol particles greater than 1.0 jjm. Marked differences were observed in the
submicrometer particle region, where measured scavenging efficiencies for submicrometer aero-
sol particles were typically an order of magnitude greater than theoretical predictions. The
scavenging gap, that portion of the aerosol particle size range where scavenging collection
efficiencies are at a minimum, was narrower than theoretically predicted, Radke et al. (1980)
offer some explanations for the discrepancies, including deliquescent growth and nucleation
scavenging of the submicrometer particles in convective clouds. Considering the varied aero-
sol particle sources and precipitation studied, the measurements showed marked continuity (see
Figure 6-5).
6.4 TRANSPORT AND DIFFUSION
Pollutant substances emitted into the atmosphere are transported and diffused as a result
of a series of complex physical interactions that result in the mean motion of air and its
fluctuating components. Transport and diffusion are associated with spatial and temporal
scales. The spatial or temporal domain directly influences what specific physical phenomena
will most affect transport and diffusion.
Studies of air pollution transport and diffusion fall broadly into two categories, de-
pending on the extent of the horizontal scale studied. High pollutant concentrations that
occur in the vicinity of a major emission source are dominated by physical processes that
operate on a local horizontal scale of the order of 1 to 5 km, or approximately 1 hour of
transport. Since the majority of criteria pollutants are emitted directly into the atmosphere
by major sources, this has been the primary area of interest in air pollution regulation.
However, as air pollution issues are raised with regard to pollutants of a more ubiquitous
nature that have appreciably longer lifetimes and, in some cases, form through secondary re-
action processes, the horizontal scale of interest expands considerably. Sulfur dioxide and
particulate matter span a horizontal scale ranging from local to global. A brief review of
the physical processes contributing to transport and diffusion is presented in Section 6.4.1,
while Section 6.4.2 considers residence times of pollutants and their long distance transport.
6.4.1 The Planetary Boundary Layer
The mean wind within approximately the first 1000 meters above the earth's surface car-
ries most of the pollutants within the atmosphere. The mean wind is determined primarily by
the interaction of three forces governed by thermodynamic and mechanical processes: (1) the
force due to the horizontal pressure gradient produced by differential solar heating of the
6-23
-------
100
X
s
UJ
tr
o
z
5
Ul
u
V)
X
10
0.01
I
—I RADKE ET AL. (1977)
1 BUHTSEV ET AL. (1970S
2 HICKS (1976)
3 DANA (1970)
4 PETERSON & CRAWFORD (1970)
I
0.1 1 10
EQUIVALENT PARTICLE DIAMETER, p
Figure 6-4. Relationship between rain scavenging rates and particle size.
Source: McMahon and Denison (1979).
6-24
-------
tu
K
u
I-
cc
a.
ui
O
o
a-
Ul
Q-
100
o
1 50
z
tu
1
m ,
Q
tu
DATE SOURCE OF AEROSOL PARTICLES
-MAY 13, 1974 PT. TOWNSEND PAPER MILL, WA
-MAR. 25,1976 "NATURAL," NEAR CENTRAL1A, WA
-MAY 10,1976 CENTBALIA POWER PLANT, WA
(a) 400 s AND (b) 265 s SOAVEMGING--*-
TIME y x
/'
(a)
tijii
.....I
10
,-2
10
,-1
iou
10'
DRY AEROSOL PARTICLE DIAMETER, j
100
50
I I I I I III! I I 1 I I llll| I I I I I III]
PATE SOURCE OF AEROSOL PARTICLES
-JUL. 1,1976 "NATURAL" (a) 3 km MSL AND
(b) 2,5 km MSL AT MILES CITY, MT
-APR. 21, 1977 VOLCANIC MAAB, AL
-JUN. 29,1977 FOUR CORNERS POWER PLANT, NM
i ml
i *i 1 1 1 1 ii
10
t-2
10
,-1
10"
101
DRY AEROSOL PARTICLE DIAMETER, jum
(b)
Figure 6-5, Percentages of aerosol particles of various sizes
removed by precipitation scavenging.
Source: Radkeetal. (1980).
6-25
-------
earth's surface; (2) the Con'olis force due to the earth's rotation; and (3) the friction
force due to the texture of the earth's surface. The planetary boundary layer is that portion
of the atmosphere within which surface frietional effects have a substantial impact on the
mean wind. Typically, this layer is hundreds of meters deep and varies diurnally.
Diffusion in the planetary boundary layer, which governs the spreading of pollutants per-
pendicular to the transport flow, is regulated by turbulence. Turbulence, which comprises a
complex spectrum of fluctuating motion superimposed on the mean wind, is generated through the
interaction of directional and speed differences (shear) in large-scale atmospheric motions
and perturbations introduced into the mean flow by the roughness of the earth's surface, as
well as by solar heating.
The theory of the mean vertical structure of the planetary boundary layer is fairly well
understood (Haugen, 1973) and can be characterized by measuring the basic meteorological para-
meters. The description of the turbulent properties within the mean motion, which govern dif-
fusion, is more elusive. Detailed theoretical approaches to turbulence are difficult to solve
because they have more unknown parameters than equations. Higher order closure techniques
apply assumptions that permit new unknowns to be expressed in terms of others in such a way as
to allow solution of the equation set. However, practical application is limited because of
their intensive computer and data requirements. Consequently, the practical treatment of
atmospheric diffusion to air pollution-related processes is based on highly parameterized
theories that depend strongly on basic experimental data sets.
Practical approaches to the treatment of atmospheric diffusion have been derived from
statistical theory, similarity theory, and gradient transport or K-theory (Pasquill, 1974). A
brief description of these approaches and their usefulness in air pollution-related problems
is presented below. More detailed discussions of the theories can be found in the cited
references.
Statistical theory considers the time history of the motion of a single fluid "particle,"
relative to a fixed coordinate axis (Taylor, 1921), and of groups or clusters of such parti-
cles relative to their centroid (Batchelor, 1953). The theory provides the basis for the
development of the Gaussian diffusion formula and provides an effective means of correlating
empirical dispersion data. As a result, diffusion equations for various emission source types
have been developed (Gifford, 1968, 1975; Turner, 1970; Pasquill, 1974, 1975). A practical
limitation of this approach is that it makes the fundamental assumption of turbulence homo-
geneity, whereas boundary-1ayer turbulence is inhomogeneous, especially in the vertical
dimension.
The similarity theory of diffusion relates the mean position and other properties of dif-
fusing clouds and plumes to the characteristic parameters of the surface layer, by dimensional
reasoning. Results (Monin and Yaglom, 1971) are reasonably complete for the surface layer,
but extension to the entire boundary layer introduces further parameters, which limits their
practical use.
6-26
-------
The gradient-transport, or K-theory, of diffusion is the oldest, originating with Pick
(1855) and Boussinesq (1877). Atmospheric applications have been most successful at large
scales, including global diffusion. At boundary-layer scales the behavior of K is quite com-
plicated. Useful results can be obtained (Pasquill, 1974; Yaglom," 1975; Csanady, 1973), but
the mathematics tend to be elaborate. The essential problem is to account for the strong
space-time scale dependence of eddy-diffusivity, which was first demonstrated by Richardson
(1926). Berlyand (1974) based a comprehensive system of air pollution analysis entirely on a
form of K-theory.
Obukhov (1941) showed that the parameterization of atmospheric diffusion follows a form
of Richardson's law of diffusion, where the total amount of turbulent energy dissipation and
the pollutant spreading is proportional to the diffusion time to the 3/2 power. Data on the
instantaneous values of the spreading of plumes and puffs (i.e., on relative diffusion) shows
that this law describes diffusion up to t on the order of an hour (Gifford, 1976a). On the
other hand, the time-average spreading of plumes was shown by Taylor (1921) to obey the asymp-
totic laws a « t (where a = the standard deviation of the horizontal spreading of the
1/2
pollutant cloud and t = time) for small t values and a
-------
mean wind; and (3) the time of day and height at which the pollutant is emitted into the
atmosphere.
Residence times for pollutants are governed by the extent of wet and dry removal and
chemical transformation the pollutant species undergoes in the atmosphere. Figure 6-6 esti-
mates residence times for typical pollutants and their associated characteristic horizontal
meteorological scale. Average windspeeds of 5 m/sec were assumed in approximating the dis-
tance scale. Residence time estimates are based on the work of Junge (1972, 1974).
Transport scales for pollutants such as S0?, which have appreciable dry deposition veloc-
ities, are sensitive to the relative height at which the pollutant is emitted. Pollutant dis-
tributions are also sensitive to the stability of the atmosphere, which governs the extent of
vertical mixing to the surfaces.
Industrial facilities emitting large quantities of S0~ have tried to take advantage of
these natural meteorological phenomena to reduce ambient levels of SO, in the vicinity of
their stacks. IJy building taller effluent stacks, emitting facilities injected SCL at higher
levels in the atmosphere, allowing the pollutant more time to be dispersed and transported
before reaching ground level, thereby effectively reducing ground-level concentrations of S0?.
A great deal of controversy arose over whether this approach circumvents the intent of the
Clean Air Act. As a result of the Clean Air Act Amendments of 1977, EPA requires that the
degree of emission limitation necessary for control of any air pollutant cannot be achieved
through the construction of stacks higher than would be considered appropriate using good
engineering practice design standards. An interesting corollary of the tall stack issue is
the potential for such sources to enhance the production of particulate sulfate. When S0? is
emitted at higher levels in the atmosphere, the probability of its removal by dry deposition
is lowered, thus extending its lifetime in the atmosphere and subsequently enhancing the prob-
ability of its being transformed to particulate sulfate through chemical reaction.
Of particular importance to horizontal transport is the strength and time of formation of
the nocturnal inversion, a stable layer of air formed at the surface due to the differential
cooling of the earth's surface relative to the night air. This stable layer varies in thick-
ness from approximately 50 m to 500 m depending on meteorological conditions. At the onset of
the nocturnal inversion, all pollutants present in the well-mixed layer from the day's
emission are cut off from the surface by this stable layer of air: There are no major
mechanisms for transport through this layer, so dry deposition processes virtually stop,
leaving the pollutant reservoir aloft free to travel long distances with negligible losses.
In addition, horizontal mean windspeeds are typically higher in layers aloft, due to the re-
duction in frictional drag at the earth's surface as a result of the presence o^ Jfhe stable
nocturnal layer.
Associated with this overall process is a phenomenon of particular importance to night-
time transport, the nocturnal jet. According to Blackadar (1957), the nocturnal jet forms as
a result of decoupling of winds previously restrained by frictional forces at the surface.
These winds are now free to accelerate in response to existing pressure gradients. As a
6-28
-------
RESIDENCE
TIME.hr
103
10^ *+
101 -i
10° -I
10'1 *t
m*3 ^
HORIZONTAL
LENGTH
SCALE
10 000 km '"—
2 000 km
200 km -- i
20km
CLIMATOLOGICAL
SCALE
CH4
SYNOPTIC AND
PLANETARY
SCALE
0,1— 1.0 jum
PARTICLES
MESO
SCALE
so2
r
MICRO-SCALE
io2
PARTICLES
Figure 6-6. Estimated residence times for select pollutant species and their associated hori-
zontal transport scale.
6-29
-------
result, some overshooting in windspeed occurs as the flow attempts to establish a new balance
with inertia! forces. Bonner (1968) examined 2 years of upper-level wind data from the
National Weather Service's rawinsonde network to determine the frequency and geographical dis-
tribution of the low-level jet. Recently, high-resolution measurements of wind profiles col-
lected over central Illinois (Sisterson and Frenzen, 1978) showed that nocturnal, low-level
wind maxima occur more frequently than indicated in Bonner's analysis. In these studies, made
during the summers of 1975 and 1976, low-level wind maxima were observed on 24 out of 30
nights for which meteorological field experiments were conducted. Typical average windspeed
profiles observed under the decoupled conditions showed windspeeds of the order of 1 to 2 m
sec near the surface, increasing to maximum values of 8 m sec at 100 to 200 meters above
the surface.
In summary, it appears that the nocturnal jet and nighttime flows are significant factors
in the transport of pollutants over long distances.
Definitive studies on the long-range transport of atmospheric tracers have been primarily
associated with radioactive debris (Islitzer and Slade, 1968) and in many instances at heights
not of particular interest for air pollution-related work. Some analyses of ambient data have
been performed to provide qualitative indications of the long-range transport and dispersion
of certain pollutant species (Altshuller, 1976; Lyons and Husar, 1976; Rodhe et al., 1972;
Brosset and Akerstrom, 1972); but very few quantitative studies exist, primarily because of a
lack of appropriate experimental data. Recent monitoring and field studies of long-range air
pollutant transport phenomena should alleviate this problem somewhat (Perhac, 1978; MacCracken,
1978; Schiermeier et al., 1979).
6.5 AIR QUALITY SIMULATION MODELING
The air quality simulation model (AQSM) primarily describes the quantitative relationship
between the distribution of emissions and ambient air quality in time and space. Air quality
simulation models are intended to improve understanding of the physical and chemical processes
operating in polluted atmospheres and provide a basis for sound, credibly based scientific
decisions on the nature and extent of emission control required to meet specified ambient air
quality standards.
The AQSM was first developed in the early 1930's, when Sutton (1932) introduced his basic
theory on diffusion in the atmosphere. Sutton's theory established the foundation for the
Gaussian equations used in describing the dispersion of effluents in the atmosphere. Since
then, the evolution of AQSM's has continued, treating increasingly complex air pollution prob-
lems and using advanced theoretical approaches to describe the details of physical and chemi-
cal atmospheric processes. The use of mathematical models for air quality impact analysis
associated with SO,.and total suspended particulate matter, both criteria pollutants, had
become a standard practice, though discretionary, prior to the passage of the Clean Air Act
Amendments in 1977.
With the passage of the 1977 amendments, EPA was required to take certain regulatory
steps related to the use of air quality simulation models. The workhorse of operational air
6-30
-------
quality simulation modeling has been the single and multiple source Gaussian plume models.
These models have been used primarily to predict ground-level concentrations ranging from the
immediate vicinity to several kilometers downwind of the effluent source. Many reviews on
dispersion modeling are available (e.g., Gifford, 1968; Strom, 1976; and Turner, 1979).
Section 6.5.1 provides a brief discussion of the status of the Gaussian modeling techniques,
while section 6.5.2 discusses the scientific basis and current status of air quality simula-
tion modeling over long distances, and the extent to which available modeling techniques have
furthered our understanding of the physical and chemical processes affecting the fate of S0?
and particulate matter in the air environment. A discussion of model evaluation and data bases
is provided in Section 6.5.3.
6.5.1 Gaussian Plume Modeling Techniques
The Gaussian diffusion formulation is used in a variety of air quality simulation model-
ing approaches. The formulation is a result of the Gaussian or normal distribution function
being a fundamental solution to the Fickian diffusion equation. Strictly speaking, the
Gaussian distribution applies only in the limit of large diffusion time and for homogeneous,
stationary conditions.
The Gaussian diffusion formulation for a continuous point source emitting pollutants at
height h and calculated receptor concentrations at ground level is given by
x (x,y) =
noy
-------
The Gaussian plume model formulas, however, have not been without criticism. Aside from
fundamental disadvantages, such as inability to treat spatially varying meteorological param-
eters, discontinuity under calm wind conditions, and inability to treat nonlinear reactive
pollutants, the major criticisms relate to the improper application of the models through the
use of inadequate or inappropriate input data. Turner (1979) pointed out that the ease of
using the Pasquill-Gifford dispersion estimates rather than collecting recommended onsite
meteorological data for developing inputs to the model has led to applications beyond the
scope intended for the dispersion schemes. Turner (1979) indicates that a significant improve-
ment in dispersion modeling estimates would Be achieved through the collection and use of on-
site measurements, including: hourly averaged windspeed and direction at a height of 10 m;
standard deviation of horizontal wind fluctuations; bulk Richardson number (a quantitative
measure of atmospheric stability) as determined from temperature and 2 and 10 m temperature
differences; height of the top of the boundary layer under unstable conditions; and the top of
the inversion under stable atmospheric conditions. Additional improvements will grow as
existing theories and experimental data bases are drawn together in a unified scheme for
estimating dispersion parameters as a function of stability, effluent release height, and
surface roughness.
A variety of operational Gaussian air quality dispersion models, used for most S0? and
total suspended particulate matter regulatory applications, are available through EPA's User's
Network for Applied Modeling of Air Pollution (UNAMAP). Turner (1979) briefly describes these
models.
A final aspect of dispersion modeling calculations is the prediction of the effective
height of the effluent release, the so-called "plume rise," which strongly affects the pre-
dicted ground-level concentration of pollutants. Research into the processes affecting plume
rise and its prediction has been underway for the past 20 years. Briggs (1975) reviewed the
physics of plume rise and its prediction and presented basic formulations for calculating the
height to which plumes rise as a function of atmospheric stability and several standard stack
parameters. He indicated that further investigation is needed in the area of plume rise
limited by ambient turbulence under convective atmospheric conditions.
6.5.2 Long-RangeAir Pollution Modeling
The recent growing interest in long-range transport of air pollutants has resulted from
extensive reviews on the subject by Bass (1980), Eliassen (1980), Pack et al. (1978), and
Smith and Hunt (1978). (Long range is defined in this document as horizontal scales of the
order of 1000 km resulting from transport times of the order of several days.) Typical model
spatial resolutions on this scale range from 20 to 100 km. Bass (1980) concluded that most
long-range transport models are Lagrangian based. In the Lagrangian approach, an emitting
source element is represented by a series of discrete pollutant parcels that are advected and
diffused by a time- and space-dependent wind field. In principle, calculation on the indi-
vidual pollutant parcels can treat time-dependent chemical transformation, dry deposition, and
precipitation scavenging processes. Fixed space-time averages of pollutants are generated by
6-32
-------
superimposing all elements that pass a specified point over the averaging time of interest.
The Lagrangian models, although all based on the same theory, have evolved individual nuances.
Little consensus of opinion exists for standard treatments of: (1) wind field analysis;
(2) choice of wind height level for trajectory; (3) mixing height variations; and (4) dry and
wet removal and chemical transformation rates. Even the basic generation of discrete air par-
cels is viewed from four different approaches: puff,. superposition, segmented plume, square
puff, and statistical. In Figure 6-7, three of the approaches are contrasted to the idealized
continuous plume they are attempting to represent (Bass, 1980).
Eulerian or grid-based approaches are less prevalent in long-range transport air pollu-
tion modeling. This may result from problems associated with the numerical integration of the
advection equation that give rise to pseudodiffusion effects. A more likely reason is the in-
creased complexity and enhanced data base and computation requirements of the Eulerian models.
Table 6-7 provides a representative sampling of long-range transport models discussed in
the available literature and for each referenced model, presents a brief description of the
modeling approach, including characteristic averaging times; approaches to dry and wet removal
and chemical transformation; and pollutant species modeled.
A review of the models presented in Table 6-7 indicates that S0? and sulfates have re-
ceived the,most attention. This is because many of the models were developed specifically to
study the acid rain phenomenon. These sulfur species have been identified as major contribu-
tors to the acidification of precipitation. Though none of the models consider primary
emitted particulate matter, its inclusion would be reasonably straightforward given the avail-
ability of appropriate emission inventories. Considering the gas-to-particle forming pro-
cesses of nitrates and organic species, aerosol dynamic processes and size distributions need
further research to be understood. Until these basic processes, as well as gas-liquid phase
transfer and solution phase chemistry of rain droplets, are treated adequately within the
models, significant skepticism toward the scientific credability and usefulness of the mo'dels
will remain.
Similarly, regional visibility impairment, which results from the physical interaction of
sunlight with light-absorbing gases and particles, and light-scattering aerosols, requires the
consideration of many of the processes described above. Though several empirical analysis
techniques have been developed that provide a qualitative understanding of the scope and
general meteorological characteristics of the visibility impairment problem, no adequate quan-
titative relationships are available for emission control strategy assessments.
Although verification studies of long-range transport models are limited, it has been
recognized for some time that errors in observed wind direction (Pack et al. 1978) and the
specification of wind fields in general (Sykes and Hatton, 1976; Smith and Hunt, 1978;
Draxler, 1979) can result in drastic errors in spatial predictions over long-range travel
distances.
6-33
-------
CONTINUOUS PLUME MODEL
SEGMENTED PLUME MODEL
PUFFSUPERIMPOSmON MODEL
'SQUARE PUFF MODEL
Rgure 6-7. Trajectory modeling approaches are shown.
Source: Bass (1980).
-------
TABLE 6-7. SUMMARY OF SELECT LONG RANGE TRANSPORT AIR POLLUTION MODELS
Model
ARL-ATAD1
EURMAP-1,-22
ENAMAP-1
MESOPUFF3
MESOPLUME4
ASTRAP5
l^IRSOX
MESOGRIO6
Type
Segmented plume
trajectory
Puff supenmposition
trajectory
Puff super-imposition
trajectory
Segmented plume
trajectory
Statistical
trajectory
Puff superimposition
trajectory/vertical
finite difference
Squence puff grid
Description
averaging time
daily to yearly
daily to yearly
daily to yearly
daily to yearly
monthly to yearly
daily to yearly
daily to monthly
Removal process
First order wet
and dry removal
First order wet
and dry removal
First order dry
removal
First order dry
removal
Diurnal and sea-
sonal dry removal/
first order wet
removal
First order wet
and dry removal
First order wet
and dry removal
Chemical process
None
SO- first order decay
SO- first order decay
SO. first order decay
Diurnal and seasonal
dependent SO. first
order decay
SO, first order decay
Typically SO, first
order decay
Pollutant species
Inert substances
S02 and SO*"
SOj and SO^
SO. and SO*"
S0? and S0*~
2-
SO^ and SO^
SOZ and SO*"
Reference
Heffter et al. (1975)
Heffter (1980)
Johnson et al. (1978)
Mancuso et al (1979)
Bbumralkar (1980)
Benkley and Bass (1979a)
Benkley and Bass (1979b)
Shannon (1979)
Shieh (1977)
Meyers et al. (1979)
Morris et al (1979)
see also model by Start and Wendell (1974)
see also models by Eliassen (1978), Nordo (1976)
and Eliassen and Saltnones (1975)
see also models by Draxler (1977, 1979)
see also models by McNaughton (1980), Hales et al (1977), Pendergast (1979),
and Henmi (1980)
see also models by Bolin and Persson (1975); Fisher (1975, 1978); and
and Scriven and Fisher (1975)
see also models by Lui and Durran (1977), Rao et al (1976), Lavery et al. (1980);
and Cantnchael and Peters (1979)
-------
The sparse temporal and spatial resolution of upper-level wind information leads to
uncertainties in predicting transport residues. The National Weather Service's rawinsonde
network provides vertical profiles for windspeed and direction, temperature, and moisture
every 12 hours at 70 sites across the continental United States. This provides upper-level
winds at a horizontal resolution of the order of 400 kilometers, considerably less resolved
than the 20 to 80 kilometer grid spacing required in air quality simulation modeling techni-
ques. Increased temporal resolution in upper-level winds should also diminish transport pre-
diction uncertainties.
6.5.3 Model Evaluation and DataBases
Almost no long-range transport models have been verified through comparison of model pre-
dictions with actual observation. The major deterrent has been insufficient or inappropriate
monitoring data for the spatial scales of interest. Long-range transport models predict
ambient pollutant concentrations that represent horizontal spatial averages of the order of
10 square kilometers. Standard monitoring networks, established for local high concentration
measurement within the vicinity of the emission sources, do not provide representative data
for long-range models. Some routine monitoring data for S02 and S0f~ from EPA's Storage and
Retrieval of Aerometric Data (SAROAD) system have been useful in model testing (Bhumralkar,
1980} but in general, the data have proven less than adequate. The Electric Power Research
Institute (EPRI)-sponsored Sulfate Regional Experiment (SURE) air quality network, (Perhac,
1978} which operated from August 1977 to October 1978, has provided the most extensive S00 and
2~ "•
SQ, data base to date for long-range transport model evaluation. But even these data col-
lected over this limited period are only sufficient in providing spatially resolved S00 and
2- <;
SO. regional concentration fields during the intensive study periods—August 1977, October
1977, mid-January to mid-February 1978, April 1978, July 1978, and October 1978—when the
extended 54-site monitoring network was activated. During the SURE study period, data also
exist from the Department of Energy-funded Multi-State Atmospheric Power Production Pollution
Study (MAP3S) precipitation chemistry network, (MacCracken, 1978) which had at least four
sites operating during the program. No dry deposition data are available for the study
period. Since the data have only recently become available, they have had limited use, .but
future long-range transport model evaluations are certain to consider them.
Another limitation in model evaluation studies is the quality of the emission inventory.
Until recently, there was no national gridded emission inventory. Clark (1980) prepared an
annual gridded emissions inventory for the United States and Southern Canada east of the Rocky
Mountains, using data compiled by EPA, the Ontario Ministry of the Environment, and Environ-
ment Canada. In preparing the gridded inventory, Clark found significant errors in many of
the United States point source records, which had to be corrected. Models, like chains, are
only as strong as their weakest links.
Although long-range transport model evaluations are extremely limited, two recent
studies should be noted. Mancuso et al. (1979) evaluated a trajectory puff model using monthly
6-36
-------
averages from the Organisation for Economic Cooperation and Development (OECD) monitoring pro-
gram. They generated two sets of evaluation results. The first considered model predictions
versus observations using parameters originally specified for the model. This resulted in
root mean square (RMS) differences of predicted versus observed monthly averaged concentration
•* 3 ?—
at all receptors of 12.9 (5 ppb) and 4.8 pg/m for S0?* anS' SO. ", respectively. In the second
evaluation, half of the data were used to optimize model parameters through a regression anal-
ysis technique and the other half of the data were used to evaluate the model. This resulted
o ^_
in RMS differences of 7.7 (3 ppb) and 2.9 ug/m for S09 and SO. and correlation coefficients
2- £4
of 0,72 both for S0« and SO, , a marked improvement in the model's performance. None of the
optimized parameters assumed values that were physically unrealistic. Lavery et al. (1980)
evaluated a grid model using data from the SURE monitoring network and based on 24-hour aver-
aged concentrations. Four days were selected for parameter "adjustments" and "fine tuning"
2-
and an additional three days were used for the model evaluation. RMS differences for SO,
3 3
ranged from 6.9 to 23.4 ug/m for the 3 days, with a mean value of 14.1 ug/m . RMS dif-
2~ 3 3
ferences for SO. ranged from 5.2 to 14.4 ug/m with a mean value of 9.3 ug/m . Mean concen-
32-
tration for the 3 days were 26.1 (10 ppb) and 13.9 ug/m for SO, and SO, , respectively. Mean
2-
correlation coefficients for S0? and SO, for the three days tested were 0.31 and 0.53,
respectively. Based on results of the 3-day comparison between observations and model predic-
tions, Lavery et al. (1980) concluded that the model typically overpredicts observed 24-hour
average sulfate by 20 to 30 percent and observed 24-hour average S02 by a factor of 2 to 3.
Overall, the results from both models are encouraging. Evaluation of a trajectory puff
model for the United States (Bhumralkar, 1980) using the SURE data base, is also showing
promising results.
6.5.4 Atmospheric Budgets
Atmospheric budgets have proven a convenient technique for quantitatively evaluating the
overall source and sink contributions of specified pollutant species within a selected region
of interest. The budget is formed by estimating the various input and output processes asso-
ciated with the region, such as anthropogenic and natural emissions, pollutant concentration
inflow and outflow, and wet and dry removal. Budget analyses provide a general long-term
indication of the significant factors contributing to the pollutant burden in a given region.
Sulfur budgets have been of interest both in Europe (Rodhe, 1972, 1978; Garland, 1978) and in
North America (Galloway and Whelpdale, 1980) because of the association of sulfur with acid
precipitation phenomena. Conclusions drawn from the eastern North American sulfur budget by
Galloway and Whelpdale (1980) were that manmade emissions exceed natural ones by a factor of
10; wet and dry deposition over the region is approximately equivalent; and at least one-
quarter of the emissions leave the region via the atmosphere to the east. As with Western
Europe, the North American budget showed that human activities dominate the regional
atmospheric sulfur cycle.
6-37
-------
6.6 SUMMARY
The processes governing transport and diffusion, chemical transformation, and wet and dry
removal of SO, and PM in the atmosphere are extremely complex and are not completely under-
stood. The oxidation rates of S0? observed in industrial plumes and urban atmospheres range
from 0 to 15 percent/h and would seem to be only partially accounted for through homogeneous
gas-phase reactions. Liquid-phase catalytic oxidation reactions involving Mn2 and carbon, as
well as solution-phase oxidation by H202 are possible contributors to the observed oxidation
rates, but further research is required to determine the rates and detailed mechanisms of
these processes under typical atmospheric conditions.
Dry deposition of S0« is fairly well understood as a result of extensive measurements
over various surfaces. Particle dry deposition studies have focused more on the physical
aspects of the deposition process, and have generated very few supporting data on particles
with compositions typical of those found in the polluted atmosphere.
Our understanding of the wet removal of S0? has progressed considerably in recent years,
including consideration of solution-phase chemistry within rain droplets. Particle removal,
like gas removal, depends on the physical characteristics of the precipitation events, which
may, in many instances, be the determining factors in accurate wet removal prediction.
Characterization of the dynamics of the planetary boundary layer is essential to an ade-
quate understanding of pollutant transport and diffusion over all spatial scales. Though con-
siderable advances have been made in this area, our ability to predict mean transport and dif-
fusion over long distances is inadequate. No doubt this is partly due to the sparse spatial
and temporal resolution of the data from the upper air wind observation network used to gener-
ate models of the transport winds.
Present generation long-range air pollutant transport models use simple parameterization
for chemical transformation and wet and dry removal, and varying degrees of sophistication in
the treatment of transport and diffusion. None of the models adequately treat the dynamics of
the planetary boundary layer. Evaluations of long-range transport models, though limited
because of lack of data bases, have shown that with further research and development these
models should be adequate tools in addressing air pollution issues associated with the mean
movement of pollutants over long distances.
6-38
-------
6.7 REFERENCES
Acres Consulting Services Ltd. and Applied earth Science Consultants Inc. Atmospheric loading
of the upper Great Lakes, vol. 2, Report for Canada Centre for Inland Waters, 1974.
Alkezweeny, A. J., and D. C. Powell. Estimation of transformation rate of S0? and SO. from
atmospheric concentration data. Atmos. Environ. 11:179-182, 1977.
Altshuller, A. P. Regional transport and transformation of sulfur dioxide to sulfates in the
United States. J. Air Pollut. Control Assoc. 26:318-324, 1976.
Bakulin, V. N., E. E. Senko, B. G. Starikov, and V. A. Trufakin. Investigation of turbulent
exchange and washout by measurement of natural radioactivity in surface air. J. Geophys.
Res. 75:3669-3674, 1970.
Banerji, P., and D. D. Chatterjee. Radon content of rainwater. Nature (London)
204:1185-1186, 1964.
Barrie, L. A. An improved model of reversible, S0~~washout by rain. Atmos. Environ.
12:407-412, 1978. *
Bass, A. Modeling long range transport and diffusion. In: Second Conference on Meteorology,
American Meteorological Society and Air Pollution Control Association, Second Joint
Conference on Applications of Air Pollution Meteorology, New Orleans, Louisianna, March
24-28, 1980. American Meteorological Society, Boston, MA, 1980. pp. 193-215.
Batchelor, G. K. The Theory of Homogeneous Turbulence. Cambridge University Press,
Cambridge, England 1953.
Beilke, S. Laboratory investigations on washout of trace gases. In: Precipitation
Scavenging (1970), Proceedings of a Symposium. Battelle Pacific Northwest Laboratory and
U.S. Atomic Energy Commission, Richland, Washington, June 1-5, 1970. R. J. Engelmann and
W. G. N. Slinn, eds., AEC Symp. Ser. 22, U.S. Atomic Energy Commission, Oak Ridge, TN,
December 1970. pp. 261-269.
Beilke, S., D. Lamb, and J. Muller. On the uncatalyzed oxidation of atmospheric SOp by oxygen
in aqueous system. Atmos. Environ. 9:1083-1090, 1975.
Belot, Y., and D. Gauthier. Transport of micronic particles from atmosphere to foliar
surfaces. In: Heat and Mass Transfer in the Biosphere. Part I. Transfer Process in
Plant Environment. D. A. deVries and N. H. Afgan, eds., John Wiley and Sons, Inc., New
York, NY, 1975. pp. 583-591.
Benkley, C. W., and A. Bass. Development of mesoscale air quality simulation models. Vol. 3.
User's guide to MESOPUFF (mesoscale puff) model, September 1979. EPA-60Q/7-80-056, U.S.
Environmental Protection Agency, Research Triangle Park, NCS 1979a.
Benkley, C. W., and A. Bass. Development of mesoscale air quality simulation models. Vol. 2.
User's guide to MESOPLUME (mesoscale plume segment) model, September, 1979.
EPA-600/7-80-057, U.S. Environmental Protection Agency, Research Triangle Park, NC,
1979b.
Berlyand, M. E. Modern Problems in Atmospheric Diffusion and Atmospheric Pollution. Hydro-
meteorological Publishing House, Leningrad, USSR, 1974.
Bhumralkar, C. M., R. L. Mancuso, D. E. Wolf, R. A. Thuillier, and K. C. Nitz. ENAMAP-1
Long-Term Air Pollution Model: Adaptation and Application to Eastern North America.
EPA-6QO/4-80-039, U.S. Environmental Protection Agency, Research Triangle Park, NC, July
1980.
6-39
-------
Blackadar, A. K. Boundary layer wind maxima and their significance for the growth of
nocturnal inversions. Bull. Am. Meteorol. Soc. 38:283-290, 1957.
Bolin, B., and C. Persson. Regional dispersion and deposition of atmospheric pollutants with
particular application to sulfur pollution over western Europe. Tellus 27:281-310, 1975.
Bonner, W. D. Climatology of the low leva! jet. Mon. Weather Rev. 96:833-850, 1968.
Boussinesq, J. Essai sur la theorie des eaux courantes. [Essay on the theory of circulating
currents.] Mem. Acad. Sci. Inst. Fr. 23:1-61, 1877.
90
Bowen, V. T., and T. T, Sugihara. Strontium in the "mixed-layer" of the Atlantic Ocean.
Nature (London) 186:71-72, 1960.
Briggs, G. A. Plume rise predictions. In: Lectures on Air Pollution and Environmental
Impact Analysis. D. Haugen, ed., American Meteorological Society, Boston, MA, 1975.
Brimblecombe, P., and D. J. Spedding. The catalytic oxidation of micromolar aqueous sulfur
dioxide. Atmos. Environ 8:937-945, 1974.
o o
Brosset, C., and A. Akerstrom. Long distance transport of air pollutants - measurements of
black airborne particulate matter (soot) and particle-borne sulphur in Sweden during the
period of September-December 1969. Atmos. Environ. 6:661-672, 1972.
32
Burtsev, I. I., L. V. Burtseva, and S. G. Malakhou. Washout characteristics of a P aerosol
injected into a cloud. In: Atmospheric Scavenging of Radio-isotopes, Papers from a
conference, Lithuanian SSR Academy of Sciences, Palanga, Lithuania SSR, June 7-9, 1966.
B, Styra, C. Garbaliauskas and V. Lujanas, eds., Israel Program for Scientific
Translations, Ltd.; Jerusalem, Israel, 1970. pp. 242-250.
Cantrell, B. K., and K. T. Whitby. Aerosol size distributions and aerosol volume formation
for a coal-fired power plant plume. Atmos. Environ. 12:323-334, 1978.
Carmichael, G. R., and L. K. Peters, Numerical simulation of the regional transport of SO*
and sulfate in the eastern United States. In: Preprints - Fourth Symposium on
Turbulence, Diffusion and Air Pollution, American Meteorological Society, Reno, Nevada,
January 15-18, 1979. American Meteorological Society, Boston, MA, 1978, pp. 337-344.
Cawse, P. A. A survey of atmospheric trace elements in the U.K. (1972-73). Environmental &
Medical Science Division. AERE Harwell, U.K. Report R 7669, 1974.
Chamberlain, A. C. Aspects of travel and deposition of aerosol and vapour clouds. AERE HP/R
1261, 1953.
Chamberlain, A. C. Aspects of the deposition of radioactive and other gases and particles.
Int. J. Air Pollut. 3:63-88, 1960.
Chamberlain, A. C. Transport of Lycopodium spores and other small particles to rough
surfaces. Proc. R. Soc. London Ser. A 296:45-70, 1967.
Chamberlain, A. C. Deposition of S02 by gaseous diffusion. Unpublished manuscript, 1973.
Chamberlain, A. C. Movement of particles in plant communities. j_n: Vegetation in the
Atmosphere. J. L. Monteith, ed., Academic Press Inc., New York, NY, 1975. pp. 155-201.
6-40
-------
Chamberlain, A. C. Dry deposition of sulfur dioxide. In: Atmospheric Sulfur Deposition,
Environmental Impact and Health Effects, Proceedings of the 2nd Life Sciences Symposium,
U.S. Department of Energy, U.S. Environmental Protection Agency, and Tennessee Valley
Authority, Gatlinburg, Tennessee, October 14-18, 1979. D. S. Shriner, C. R. Richmond,
and S. E. Lindberg, eds., Ann Arbor Science Publishers Inc., Ann Arbor, MI, 1980. pp.
185-197.
Chamberlain, A, C., and R. C, Chpdwick., Deposition of airborne radio-iodine vapor. Nucleonics
8:22-25, 1953.
Chamberlain, A. C., and R. C. Chadwick. Deposition of spores and other particles on vegeta-
tion and soil. Ann. Appl. Biol. 72:141-158, 1972.
Chang, T. Y. Estimate of the conversion rate of S0« to SOA from the Da Vinei Flight Data.
Atmos. Environ. 13:1663-1664, 1979. , 4
Clark, T. L. Annual anthropogenic pollutant emissions in the United States and southern
Canada east of the Rocky Mountains. Atmos. Environ. 14:961-970, 1980.
Clough, W. S. Transport of particles to a surface. J. Aerosol Sci. 4:227-234, 1973.
Cobourn, W. G., R. B. Husar, and J. D. Husar. .Continuous in situ monitoring of ambient
particulate sulfur using flame photometry and thermal analysis. Atmos. Environ. 12:89-
98, 1978.
Craig, D. K. , B. L. Klepper, and R. L. Buschbom. Deposition of various piutoniurn-compound
aerosols onto plant foliage at very low wfnd velocities. Atmosphere-Surface Exchange of
Particulate and Gaseous Pollutants (1974), Proceedings of a symposium, Battelle Pacific
Northwest Laboratories and U.S. Atomic Energy Commission, Richland, Washington, September
4-6, 1974. R. J. Engelmann and G, A. Sehmel, coord., ERDA symposium series 38, Energy
Reserach and Development Administration, Oak Ridge, TN, January 1976. pp. 244-263.
Csanady, G. T. Turbulent Diffusion in the Environment. D. Reidel Publishing Co., Dordrecht,
Holland, 1973.
Dana, M. T. Scavenging of soluble dye particles by rain. In: Precipitation Scavenging,
(1970), Proceedings of a Symposium, Battelle Pacific Northwest Laboratory and U.S. Atomic
Energy Commission, Richland, Washington, June 1-5, 1970. R. J. Engelmann and W. G: N.
Slinn, eds., AEC symposium series 22, U.S. Atomic Energy Commission, Oak Ridge, TN,
December 1970. pp. 137-147.
Dana, M. T. S09 versus sulfate wet deposition in the eastern United States. J. Geophys. Res.
85:4475-4480, 1980.
Dana, M. T. , and J. M. Hales. Statistical aspects of the washout of polydisperse aerosols.
Atmos. Environ. 10:45-50, 1976.
Dana, M. T., J. M. Hales, and M. A, Wolf, Rain scavenging of S09 and sulfate from power plant
plumes. J. Geophys. Res. 80:4119-4129, 1975. £
Davidson, C. I., and S. K. Friedlander. A filteration model for aerosol dry deposition:
Application to trace metal deposition from the atmosphere. J. Geophys. Res. 83:2343,
1978.
Dennis, R. , C. E. Billings, J. A. Record, P. Warneck, P. Arin, and M. L. Arin. Measurements
of sulfur dioxide losses from stack plumes. Presented at the 62nd Annual Meeting, Air
Pollution Control Association, New York, NY, June 22-26, 1969.
6-41
-------
Dittenhoefer, A. C., and R. G. de Pena. A study of production and growth of sulfate particles
in plumes from a coal-fired power plant, Atmos. Environ. 12:297-306, 1978.
Drake, R. L., and S. M. Barrager. Mathematical Models for Atmospheric Pollutants. EA-1131,
Electric Power Research Institute, Palo Alto, CA, August 1979.
Oraxler, R. R. Determination of atmospheric diffusion parameters. Atmos. Environ. 10:99-106,
1976.
Oraxler, R. R. _A Mesoscale Transport and Diffusion Model. NOAA technical memorandum ERL-ARL-
64, U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Air
Resources Laboratories, Silver Spring, MD, June 1977.
Draxler, R. R. Modeling the results of two recent mesoscale dispersion experiments. Atmos.
Environ. 13:1523-1533, 1979.
Eliassen, A. The OECD study of long range transport of air pollutants: Long range transport
modeling. Atmos. Environ. 12:479-488, 1978.
Eliassen, A. A review of long-range transport modeling. J. Appl. Meteorol. 19:231-240, 1980.
Eliassen, A., and J. Saltbones. Decay and transformation rates for S0?, as estimated from
emission data, trajectories and measured air concentrations. Atmos. Environ. 9:425-429,
1975.
Engelmann, R. J. The Calculation of Precipitation Scavenging. BNWL-77, Battelle Memorial
Institute, Pacific Northwest Laboratory, Richland, WA, July 1965.
Engelmann, R. J. The calculation of precipitation scavenging. In; Meteorology and Atomic
Energy - 1968. D. H. Slade, ed. , U.S. Atomic Energy Commission, Division of Technical
Information, Oak Ridge, TN, July 1968. pp. 208-221.
Eriksson, E. The yearly circulation of chloride and sulfur in nature; meteorological, geo-
chemical and pedological implications. Part I. Tell us 11:375-403, 1959.
Esmen, H. A., and M. Corn. Residence time of particles in urban air. Atmos. Environ.
5:571-578, 1971.
Esmen, N. A. Particle retention efficiency of scavenging rainfall. In: 163rd National
Meeting, American Chemical Society, Division of Water, Air and Waste Chemistry, Boston,
Massachusetts, April 9-14, 1972. Prepr, Pap. Natl. Meet. Div. Water Air Waste Chem. Am.
Chem. Soc. 12: 12-18, 1972.
Pick, A, Uber diffusion [On diffusion.] Ann. Phys. Chem. 2:59-86, 1855.
Fisher, B. E. A. The long range transport of sulfur dioxide. Atmos. Environ. 9:1063-1070,
1975.
Fisher, B. E. A. The calculation of long term sulfur deposition in Europe. Atmos. Environ.
12:489-501, 1978.
Forrest, J. , and L. Newman. Further studies on the oxidation of sulfur dioxide in coal-fired
power plant plumes. Atmos. Environ. 11:465-474, 1977.
Forrest, J., and L. Newman. Oxidation of sulfur dioxide in the Sudbury smelter plume. Atmos.
Environ. 11:517-520, 19775.
Galloway, J. N., and D. M. Whelpdale. An atmospheric sulfur budget for eastern North America.
Atmos. Environ. 14:409-417, 1980.
6-42
-------
Garland, J. A. Dry and wet removal of sulfur from the atmosphere. Atmos. Environ. 12:349-
362, 1978.
Garte11, F, E. , F. W. Thomas, and S. B. Carpenter. Atmos. oxidation of S0« in coal-burning
power plant plumes. Am. Ind. Hyg. J. 24:113-120, 1963.
Gatz, D. F. A review of chemical tracer experiments on precipitation systems. Atmos.
Environ. 11:945-954, 1977.
Georgii, H. W. On the effect of rainfall on the sulphur dioxide concentration in the
atmosphere. Air Water Pollut. 7:1057-1059, 1963.
Gifford, F. A. An outline of theories of diffusion in the lower layers of the atmosphere.
In: Meteorology and Atomic Energy—1968. 0. H. Slade, ed., U.S. Atomic Energy
Commission, Division of Technical Information, Oak Ridge, TN, July 1968. pp. 65-116.
Gifford, F. A. Atmospheric dispersion models for environmental pollution applications. In:
Lectures on Air Pollution and Environmental Impact Analysis. D. Haugen, ed., American
Meteorological Society, Boston, MA, 1975.
Gifford, F. A. Tropospheric Relative Diffusion Observations. Unpublished manuscript, File
No. 76/10, National Oceanic and Atmospheric Administration, Oak Ridge, TN, 1976a,
Gifford, F. A. Turbulent diffusion typing schemes: a review. Nucl. Saf. 17:68-86, 1976b.
Gillani, N. V., R. Husar, D. E. Patterson, and W. E. Wilson. Project MISTT: kinetics of
particulate sulfur formation in a power plant out to 300 km. Atmos. Environ. 13:589-598,
1978.
Colder, D. Relations among stability parameters in the surface layer. Boundary Layer
Meteorol. 3:47-58, 1972.
Graedel, T. E. , and J. P. Franey. Field measurements of submicron aerosol washout by snow.
Geophys. Res. Lett. 2:325-328, 1975.
Graedel, T. E. , and J. P. Franey. Field measurements of submicron aerosol washout by rain.
In: Precipitation Scavenging (1974), Proceedings of a Symposium, Illinois State Water
Survey and U.S. Atomic Energy Commission, Champaign, Illinois, October 14-18, 1974. R.
G. Semonin and R. W. Beadle, coord. , ERDA symposium series 41, Energy Research and
Development Administration, Oak Ridge, TN, June 1979. pp. 503-523.
Gravenhorst, G. , T. H. Janssen-Schmidt, D. H. Ehalt, and E. P. Roth. The influence of clouds
and rain on the vertical distributions of sulfur dioxide in a one-dimensional steady-
state model. Atmos. Environ. 12:691-698, 1978.
Grover, S. N. , H. R. Pruppacher, and A. E. Hamielec. A numerical determination of the
efficiency with which spherical aerosol particles collide with spherical water drops due
to inertial impaction and phoretic and electrical forces. J. Atmos. Sci. 34:1655-1663,
1977.
Hales, J. M. Fundamentals of the theory of gas scavenging by rain. Atmos. Environ. 6:635-
659, 1972.
Hales, J. M. Wet removal of sulfur compounds from the atmosphere. Atmos. Environ. 12:389-
399, 1978.
Hales, J. M. , D. C. Powell, and T. D, Fox. STRAM - An air pollution model incorporating non-
linear chemistry, variable trajectories, and plume segment diffusion. EPA-450/ 3-77-012,
U.S. Environmental Protection Agency, Research Triangle Park, NC, 1977.
6-43
-------
Hales, J. H., J. M. Thorp, and M. A. Wolf. Field Investigation of Sulfur Dioxide Washout from
the Plume of a Large Coal-fired Power Plant by Natural Precipitation. APTD-0739, U.S.
Environmental Protection Agency, Research Triangle Park, NC, March 1971.
Hart, G. E,, and D. R. Parent. Chemistry of throughfall under Douglas Fir and Rocky Mountain
Juniper. Am. Midi. Nat. 92:191-201, 1974.
Haugen, D. A, ed. Workshop on micrometeorology, Boston, 1972. American Meteorological
Society, Boston, MA, 1973.
Heffter, J. L. Air Resources Laboratories Atmospheric Transport and Dispersion Model
(ARL-ATAD). NOAA technical memo ERL-.ARL-81, U.S. Department of Commerce, National
Oceanic and Atmospheric Administration, Air Resources Laboratories, Silver Spring, MD,
1980.
Heffter, J. L., A. D. Taylor, and G. J. Ferber. A Regional-Continental Scale Transport,
Diffusion and Deposition Model. NOAA technical memo ERL-ARL-50, U.S. Department of
Commerce, National Oceanic and Atmospheric Administration, Air Resources Laboratories,
Silver Spring, MD, June 1975.
Hegg, D, A. , and P. V. Hobbs. Measurements of gas-to-particle conversion in the plumes from
five coal-fired electric power plants. Atmos. Environ. 14:99-116, 1980.
Henmi, T. Long-range transport model of SO- and sulfate and its application to the eastern
United States. J. Geophys. Res. 85:443^-4442, 1980.
Hicks, B. B. An evaluation of precipitation scavenging rates of background aerosol. In:
Argonne National Laboratory Radiological and Environmental Research Division Annual
Report: Atmospheric Physics: January through December, 1976. ANL-76-88 Part IV, Energy
Research and Development Administration,'Argonne National Laboratory, Argonne, IL, 1976.
pp. 65-75.
Hicks, B. B., M. L. Wesely, and J. L. Durham. Critique of Methods to Measure Dry Deposition.
EPA-60Q/9-8Q-05Q, U.S. Environmental Protection Agency, Environmental Sciences Research
Laboratory, Research Triangle Park, NC, September 1980.
Hill, F. B., and R. ,F. Adamowicz. A model for rain composition and the washout of sulfur
dioxide. Atmos. Environ. 11:917-927, 1977.
Hobbs, P. V., D. A. Hegg, M. W. Eltgroth, and L. F. Radke. Evolution of particles in the
plumes of coal-fired power plants. I. Deductions from field measurements. Atmos.
Environ. 13: 935-951, 1979.
Huntzicker, J. J., R. S. Hoffman, and C. S. Ling. Continuous measurements and speciation of
sulfur-containing aerosols by flame photometry. Atmos. Environ. 12:83-88, 1978.
Husar, R. B., D. E. Patterson, J. D. Husar, N. V. Gillani, and W. E, Wilson. Sulfur budget of
a power plant plume. Atmos. Environ. 12:549-568, 1978.
Islitzer, N. F., and D. H. Slade. 'Diffusion and transport experiments. In: Meteorology and
Atomic Energy-1968. D. H. Slade, ed., U.S. Atomic Energy Commission, Division of
Technical Information, Oak Ridge, TN, July 1968. pp. 117-188.
Johnson, W. B., D. E. Wolf, and R. L. Mancuso. Long term regional patterns and transfrontier
exchanges of airborne sulfur pollution in Europe. Atmos. Environ. 12:511-527, 1978.
Junge, C. E. The cycle of atmospheric trace gases - natural and manmade. Q. J. R. Meteorol.
Soc. 98:711-729, 1972.
6-44
-------
Junge, C. E. Residence time and variability of tropospheric trace gases. Tellus 26:477-488,
1974. ~~
Kalkstein, M. I., et al. Natural aerosols and nuclear debris studies. AFCRL-TN-59-627, Air
Force Cambridge Research Laboratory, Hanscom Air Force Base, MA, 1959.
Klepper, B., and D. K. Craig. Deposition of ainbonne pa-rticulates onto plant leaves. J,
Environ. Qua!. 4:495-499, 1975.
Knutson, E. 0. , and J. D. Stockham. Aerosol scavenging by snow: comparison of single-flake
and entire-snowfall results. In: Precipitation Scavenging (1974), Proceedings of a
Symposium, Illinois State Water Survey and U.S. Atomic Energy Commission, Champaign,
Illinois, October 14-18, 1974. R. G. Semonin and R. W. Beadle, Coords., ERDA Symposium
Series 41, Energy Research and Development Administration, Oak Ridge, TN, June 1977. pp.
195-207.
Lavery, T. F. , R. L. Baskett, J. W, Thrasher, N. J. Lordi, A. C. Lloyd, and G. M. Hidy.
Development and validation of a regional model to simulate atmospheric concentrations of
sulfur dioxide and sulfate. In; Second Joint Conference on Applications of Air Pollu-
tion Meteorology, and Second Conference on Meteorology, American Meteorological Society
and Air Pollution Control Association, New Orleans, Louisiana, March 24-28, 1980.
American Meteorological Society, Boston, MA, 1980. pp. 236-247.
Little, P. Deposition of 2.75, 5.0 and 8.5 urn particles to plant and soil surfaces. Environ.
Pollut. 12:293-305, 1977.
Little, P., and R. D. Wiffen. Emission and deposition of petrol engine exhaust Pb—I Deposi-
tion of exhaust Pb to plant and soil surfaces. Atmos. Environ. 11:437-447, 1977.
Liu, M. K. and D. R. Durran. The Development of a Regional Air Pollution Model and its Appli-
cation to the Northern Great Plains. EPA-908/1-77-001, U.S. Environmental Protection
Agency, Denver, CO, July 1977.
Lusis, M. A., and H. A. Wiebe. The rate of oxidation of sulfur dioxide in the plume of a
nickel smelter stack. Atmos. Environ. 10:793-798, 1976.
Lusis, M. A., K. G. Anlauf, L. A. Barrie, and H. A. Wiebe. Plume chemistry studies at a
northern Alberta power plant. Atmos. Environ. 12:2429-2437, 1978.
Lyons, W. A., and R. B. Husar. SMS/GOES visible images detect a synoptic-scale air pollution
episode. Mon. Weather Rev. 104:1623-1626, 1976.
MacCracken, M. C. MAP3S: an investigation of atmospheric energy related pollutants in the
northeastern United States. Atmos. Environ. 12:649-659, 1978.
Makhon'ko, K. P. Determination of capture coefficient of radioactive dust by rain. Izv.
Akad. Nauk SSSR Ser. Geofiz. 11:1709, 1964.
Makhon'ko, K. P. Simplified theoretical motion of contaminant removal by precipitation from
the atmosphere. Tell us 19:467-476, 1967.
Makhon'ko, K. P., and G. V. Dmitrieva. Capability of various types of precipitation for wash-
out of fission products from the atmosphere and characteristics of washout. Izv. Akad.
Nauk SSSR Ser. Phys. Atm. I. Okeana 3:508, 1966.
Mamane, Y. , and R. F. Pueschel. Formation of sulfate particles in the plume of the Four
Corners Power Plant. J. Appl. Meteorol. 19:779-790, 1980.
6-45
-------
Mancuso, R. L., C. M. Bhumralkar, D. E. Wolf, and W. B. Johnson. The exchange of sulfur
pollution between the various countries of Europe based on the EURMAP model. In:
Preprints-Fourth Symposium on Atmospheric Turbulence, Diffusion and Air Pollution,
American Meteorological Society, Reno, Nevada, January 15-18, 1979. American Meteoro-
logical Society, Boston, MA, 1978. pp. 345-354.
McMahon, T. A., and P. J. Oenison. Review Paper - empirical atmospheric deposition parameters
- a survey. Atmos. Environ. 13:571-585, 1979.
McNaughton, D. J. Initial comparision of SURE/MAP3S sulfur oxide observations with long-term
regional model predictions. Atmos. Environ. 14:55-63, 1980.
Meagher, J. F., L. Stockburger, E. M. Bailey, and 0. Huff. The oxidation of sulfur dioxide to
sulfate aerosols in the plume of a coal-fired power plant. Atmos. Environ. 12:2197-2204,
1978.
Meyers, R. E., R, T. Cederwall, and W. D. Ohmstede. Modeling regional atmospheric transport
and diffusion: Some environmental applications. In: Advances in Environmental Science
and Engineering. J. R. Pfafflin and E. Ziegler, eds., Gordon and Breach, New York, NY,
1979.
HBller, U., and G. Schumann. Mechanisms of transport from the atmosphere to the earth's
surface. J. Geophys. Res. 75:3013-3019, 1970.
Honin, A. S., and A. M. Yaglom. Statistical Fluid Mechanics, vol. 1. MIT Press, Cambridge,
MA, 1971.
Neuberger, H., C. L. Hosier, and W. C. Kocmond. Vegetation as an aerosols filter. In: Bio-
meteorology. Volume 2. S.W. Tromp and W.H. Weihe, eds., Pergamon Press, Ltd., Oxford,
England, 1967. pp. 393-702.
Newman, L. Atmospheric oxidation of sulfur dioxide. In: Atmospheric Sulfur Deposition,
Environmental Impact and Health Effects, Proceedings of the 2nd Life Sciences Symposium,
U.S. Department of Energy, U.S. Environmental Protection Agency, and Tennessee Valley
Authority, Gatlinburg, Tennessee, October 14-18, 1979. D.S. Shriner, C.R. Richmond, and
S.E. Lindberg, eds., Ann Arbor Science Publishers Inc., Ann Arbor, MI, 1980. pp. 131-
143.
Newman, L., J. Forrest, and B. Manowitz. The application of an isotope ratio technique to a
study of the atmospheric oxidation of sulfur dioxide in the plume from a coal-fired power
plant. Atmos. Environ. 9:969-977, 1975.
NordiJ, J. Long-range transport of air pollutants in Europe and acid precipitation in Norway.
Water, Air, Soil Pollut. 6:199-217, 1976.
Obukhov, A. M. On the distribution of energy in the spectrum of turbulent flow. Izv. Akad.
Nauk SSSR Ser. Geogr. Geofiz. 5:453, 1941.
Pack, D. H«, G. J. Ferber, J. L. Heffter, K. Telegades, J. K. Angell, W. H. Hoecker, and L.
Machta. Meteorology of long-term transport. Atmos. Environ. 12:425-444, 1978.
Pasquill, F. The estimation of the dispersion of windborne material. Meteorol. Mag. 90:33-
49, 1961.
Pasquill, F. Atmospheric Diffusion. 2nd ed. John Wiley and Sons, Inc. New York, NY. 1974.
Pasquill, F. The dispersion of materials in the atmospheric boundary layer - the basis for
generalization. In; Lectures on Air Pollution and Environmental Impact Analysis. D.
Haugen, ed., American Meteorological Society, Boston MA, 1975.
6-46
-------
Pasquill, F. Atmospheric Dispersion Parameters in Gaussian Plume Modeling. Part 2: Possible
Requirements for Change in the Turner Workbook Values. EPA-600/4-76-030b, U.S. Environ-
mental Protection Agency, Environmental Sciences Research Laboratory, Research Triangle
Park, NC, June 1976.
Pendergast, M. M. Model evaluation for travel distances 30-140 km. In: Preprints-Fourth
Symposium on Turbulence, Diffusion, and Air Pollution, American Meteorological Society,
Reno, Nevada, January 15-18, 1979. American Meteorological Society, Boston, MA, 1978.
pp. 648-651.
Perhac, R. M. Sulfate regional experiment in northeastern United States: The "SURE" program.
Atmos. Environ. 12:641-647, 1978.
Perkins, R. W., C. W. Thomas, and J. A. Young. Application of short-lived cosmogenic radio-
nuclides as traces of in-cloud scavenging processes. J. Geophys. Res, 75:3076-3087,
1970. ~
Peterson, K. R., and T. Y. Crawford. Precipitation scavenging in a large-cloud diffusion
code. In: Precipitation Scavenging (1970), Proceedings of a Symposium, Richland,
Washington, June 1-5, 1970. R. J. Engelmann and W. G. N. Slinn, eds., Symp. Ser. 22,
Battelle Pacific Northwest Laboratory and U.S. 'Atomic Energy Commission, Oak Ridge, TN,
December 1970. Available as CONF-700601 from National Technical Information Service,
Springfield, VA. pp. 425-431.
Pierson, D. H., P. A. Cawse, L. Salmon, and R. S. Cambray. Trace elements in the atmospheric
environment. Nature, (London) 241:252-256, 1973.
Postma, A. K. Effect of solubilities of gases on their scavenging by raindrops. In:
Precipitation Scavenging (1970), Proceedings of a Symposium, Battelle Pacific Northwest
Laboratory and U.S. Atomic Energy Commision, Richland, Washington, June 1-5, 1970. R. J.
Engelmann and W. G. N, Slinn, eds. AEC symposium series 22, U.S. Atomic Energy
Commission, Oak Ridge, TN, December, 1970. pp. 247-259.
Pueschel R. V., and C. C. Van Valin. Cloud nucleus formation in a power plant plume. Atmos.
Environ. 12:307-312 1978.
Radke, L. F., E. E. Hindman, II, and P. V. Hobbs. A case study of plume scavenging by a rain
shower. In: Precipitation Scavenging (1974), Proceedings of a Symposium, Illinois State
Water Survey and U.S. Atomic Energy Commission, Champaign, Illinois, October 14-18, 1974.
R. G. Semonin and R. W. Beadle, coord.., ERDA symposium series 41, Energy Research and
Development Administration, Oak Ridge, TN, June 1977. pp. 425-436.
Radke, L. F., P. V. Hobbs, and M. W. Eltgroth. Scavenging of aerosol particles by precipita-
tion. J. Appl. Meteorol. 19:715-722, 1980.
Rao, K. S., J. S. Lague, and B. A. Egan. An air trajectory model for regional transport of
atmospheric sulfates. Preprint Volume-Third Symposium on Atmospheric Turbulence,
Diffusion and Air Quality, American Meteorological Society, Raleigh, North Carolina,
October 19-22, 1976. American Meteorological Society, Boston, MA, 1976. pp. 325-331.
Rasmussen, K. H., M. Taheri, and R. L. Kabel. Global emissions and natural processes for
removal of gaseous pollutants. Water Air Soil Pollut. 4:33-64, 1975.
Richardson, L. F. Atmospheric diffusion shown on a distance-neighbour graph. Proc. R. Soc.
London Ser. A 104:640-654, 1926.
Roberts, P. T., and S. K. Friedlander. Conversion of S02 to sulfur particulate in the Los
Angeles atmosphere. EHP Environ. Health Perspect. 10:103-108, 1975.
6-47
-------
Roberts, D. B. , and D. J. Williams. The kinetics of oxidation of sulphur dioxide within the
plume from a sulphide smelter in a remote region. Atmos. Environ. 13:1485-1499, 1979.
Robinson, E., and R. C. Robbins. Gaseous nitrogen compound pollutants from urban and natural
sources. 'J. Air Pollut. Control Assoc. £0:303-306, 1970.
Rodhe, H. Budgets and turn-over times of atmospheric sulfur compounds, Atmos. Environ.
12:671-680, 1978.
Rodhe, H., and J. Grandell. On the removal time of aerosol particles from the atmosphere by
precipitation scavenging. Tellus 24:442, 1972.
Schiermeier, F. A., W. E. Wilson, F. Pooler, J. K. S. Ching, and J. F. Clarke. Sulfur
Transport and Transformation in the Environment (STATE): A Major EPA Research Program.
Bull. Am. Meteorol. Soc. 60:1303-1312, 1979,
Scott, B. C. Parameterization of sulfate removal by precipitation. J. Appl. Meteorol.
17:1375-1389, 1978.
Scriven, R. A., and B. E. A. Fisher. The long range transport of airborne material and its
removal by deposition and washout. Atmos. Environ. 9:49-68, 1975.
Sehmel, G. A. Particle eddy diffusivities and deposition velocities for isothermal flow and
smooth surfaces. J. Aerosol Sci. 4:125-138, 1973.
Sehmel, G. A., and S. L. Sutter. Particle deposition rates on a water surface as a function
of particle diameter and air velocity. J. Rech. Atmos. 8:911-920, 1974.
Sehmel, G. A. Particle and gas dry deposition: a review. Atmos. Environ. 14:983-1011, 1980.
Shannon, J. D. The Advanced Statistical Trajectory Regional Air Pollution Model.
ANL/RER-79-1, U.S. Department of Energy Argonne National Laboratory, Argonne, IL, May
1979.
Sheih, C. M. Application for a statistical trajectory model to the simulation of sulfur
pollution over northeastern United States. Atmos. Environ. 11:173-178, 1977.
Sheih, C. M., M. L. Wesely, and B. B. Hicks. Estimated dry deposition velocities of sulfur
over the eastern United States and surrounding regions. Atmos. Environ. 13:1361-1368,
1979.
Shirvaikar, V, V. et al. A study of washout of radioactive fallout and particulate matter in
individual rain showers. AEET/AM/10 (India), 1960.
Sisterson, D. L., and P. Trenzen. Nocturnal boundary-layer wind maxima and the problem of
wind power assessment. Environ. Sci. Techno!. 12:218-221, 1978.
Slinn, W. G. N. Dry deposition and resuspension of aerosol particles —a new look at some old
problems. In: Atmosphere-Surface Exchange of Particulate and Gaseous Pollutants (1974),
Proceedings of a Symposium, Battelle Pacific Northwest Laboratories and U.S. Atomic
Energy Commission, Richland, Washington, September 4-6, 1974. R. J. Engelmann and G. A.
Sehmel, coord., ERDA symposium series 38, Energy Research and Development Administration,
Oak Ridge, TN, January 1976. pp. 1-40.
Slinn, W. G, N. Some approximations for the wet and dry removal of particles and gases from
the atmosphere. Water, Air, Soil Pollut. 7:513-543, 1977.
Slinn, W. G. N. Precipitation scavenging. In: Atmospheric Sciences and Power Production—
1981. D. Randerson, ed. , U.S. Department of Energy, Washington, DC. In press, 1981.
6-48
-------
Slinn, W. G. N., L. Hasse, B. B. Hicks, A. W. Hogan, D. Lai, P. S. Liss, K. 0. Munnich, G. A,
Sehmel, and 0. Vittorl. Some aspects of the transfer of atmospheric trace constituents
past the air-sea interface. Atmos. Environ. 12:2055-2087, 1978.
Small, S. H. Wet and dry deposition of fallout materials at Kjeller. Tellus 12:308-314,
1960. ~
Smith, F. B. A scheme for estimating the vertical <5-ispei%ion of a plume from a source near
ground level. In: Proceedings of the Third Meeting of the Expert Panel on Air Pollution
Modelling, NATO Committee on the Challenges of Modern Society, Paris, France, October
2-3, 1972. NATO-CCMS N. 14, North Atlantic Treaty Organization, Brussels, Belgium, 1972.
pp. XVII-1 - XVII-14.
Smith, F. B. , and R. D. Hunt. Meteorological aspects of the transport of pollution over long
distances. Atmos. Environ. 12:461-477, 1978.
Start, G. E. , and L. L. Wendell. Regional Effluent Dispersion Calculations Considering
Spatial and Temporal Meteorological Variations. NOAA technical memo ERL ARL-44, U.S.
Department of Commerce, National Oceanic and Atmospheric Administration, Idaho Falls, ID,
May 1974.
Stephens, W. T., and R. 0. McCaldin. Attenuation of power station plumes as determined, by^
instrumental aircraft. Environ. Sci. Techno!. 5:615-621 1971. ' "
Strom, G. H. Transport and diffusion of stack effluents. In: Air Pollution. Volume I: Air
Pollutants, Their Transformation and Transport. A. C. Stern, ed. , 3rd ed., Academic
Press, Inc., New York, NY, 1976. pp. 401-501.
Sutton, D. G. A theory of eddy diffusion in the atmosphere. Proc. R. Soc. London. Ser. A
135:143, 1932.
Sykes, R. I., and L. Hatton. Computation of horizontal trajectories based on the surface
geostrophic wind. Atmos. Environ. 10:925-934, 1976.
Taylor, G. I. Diffusion by continuous movements. Proc. London Math. Soc. 20:196-212, 1921.
Turner, D. B. Workbook of Atmospheric Dispersion Estimates. Office of Air Programs -Publica-
tion No. AP-26, U. S. Environmental Protection Agency, Research Triangle Park, NC, 1970.
Turner, D. B. Atmospheric dispersion modeling - a critical review. J. Air. Pollut. Control
Assoc. 29:502-519, 1979.
Ursenbach, W. 0., A. C. Hill, W. H. Edwards, and S. M. Kunen. Conversion rate of S02 to
submicron sulfate in the plumes of a coal-fired power plant in the western United StaT.es.
Presented at 70th Annual Meeting, Air Pollution Control Association, Toronto, Canada,
June 20-24, 1977.
Volchuk, H. L. , V. T. Bowen, T. R. Folsom, W. S. Broecker, E. A. Schuert, and G. S. Bien.
Oceanic distribution of radionuclides from nuclear explosions. In: Radioactivity in the
Marine Environment. National Academy of Sciences, Washington, DC, 1971.
Wang, P. K., S. N. Grover and H. R. Pruppacher. On the effect of electric charges on the
scavenging of aerosol particles by clouds and small raindrops. J. Atmos. Sci.
35:1735-1743, 1978.
Weber, E. Contribution to the residence time of sulfur dioxide in a polluted atmosphere. J.
Geophys. Res. 75:2909-2915, 1970.
6-49
-------
Weber, A. H. Atmospheric dispersion parameters of Gaussian Plume Modelling. Part I: Review
of Current Systems and Possible Future Developments. EPA-600/4-76-030a, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, 1976.
Wedding, J. B., R. W. Carlson, J. J. Stukel, and F. A. Bazzaz. Aerosol deposition on plant
leaves. In: Proceedings of the First International Symposium on Acid Precipitation and
the Forest Ecosystem, Ohio State University, Columbus, Ohio, May 12-15, 1975. L. S.
Dochinger and T. A. Seliga, eds., NEFES/77-1, U.S. Department of Agriculture, Forest
Service, Northeastern Forest Experiment Station, Upper Darby, PA, 1976. pp. 897-903.
White, E., and'F. Turner. A method of estimating income of nutrients in a catch of airborne
particles by a woodland canopy. J. Appl. Ecol. 7:441-461, 1970.
Wolf, M. A., and M. T. Dana. Experimental studies on precipitation scavenging. Battelle-
Northwest Annual Report. Report BNWL-1051 (Pt. 1), U.S. Atomic Energy Commission, 1969.
pp. 18-25.
Yaglora, A. M. Equations with time-dependent coefficients describing diffusion in a stationary
atmospheric surface layer. Izv. Acad. Sci. USSR Atmos. Oceanic Phys. (Engl. Transl.)
11:704-708, 1975.
6-50
-------
7. ACIDIC DEPOSITION
7.1 INTRODUCTION
Acidic precipitation has become a major environmental concern in many regions of the
world. Acidic precipitation (rain and snow) in the Adirondack Mountains of New York State,
eastern Precambrian Shield area of Canada, southern Norway, and southwest Sweden has been
associated with the acidification of waters in ponds, lakes, and streams and the subsequent
disappearance of animal and plant life. Acidic precipitation also is believed to have the
potential for leaching elements from sensitive soils and causing direct and indirect injury to
forests. It is also believed to play a role in damaging stone monuments and buildings, corrod-
ing metals and the deterioration of paint.
The story of acidic precipitation is ever-changing; new information concerning the
phenomenon is appearing continuously. This chapter explains how particulate matter and sulfur
*
oxides are involved in acidic deposition phenomena and associated ecological effects. The
information as presented reflects the understanding of the scientific community at the time
this chapter was written. A critical assessment document on acidic decomposition currently
being written under the direction of the Office of Research and Development of the Environ-
mental Protection Agency, will present a more detailed, up-to-date discussion of the many
facets of the acidic deposition problem.
The sections that follow emphasize the effects associated with the wet deposition of
sulfur and nitrogen oxides and their products on aquatic and terrestrial ecosystems. Dry
deposition also plays an important role, but the relative contribution of this process is
still unknown. Because sulfur and nitrogen oxides are so closely linked in the formation of
acidic precipitation, no attempt has been made to limit the discussion that follows to sulfur
oxides.
7.1.1 Overview of the Problem
The generally held hypothesis is that sulfur and nitrogen compounds are largely respon-
sible for the acidity of precipitation. The emissions of the sulfur and nitrogen compounds
involved in acidification are attributed chiefly to the combustion of fossil fuels. Emissions
may occur at ground level, as from automobile exhausts, or from stacks 300 meters (1000 feet)
or more in height. Emissions from natural sources are also involved; however, in highly in-
dustrialized areas, emissions from manmade sources far exceed those from natural sources. In
the eastern United States, the highest emissions of sulfur oxides are from electric power
generators burning coal, while on the West Coast, particularly around large cities, emissions
of nitrogen oxides, chiefly from automotive sources, predominate. (See Chapter 4.)
The fate of sulfur and nitrogen oxides, as well as other pollutants and gases emitted
from natural sources into the atmosphere, depends on their dispersion, transport, transforma-
tion and deposition. Sulfur and nitrogen oxides may be deposited locally or transported long
distances from the emission sources. Therefore, residence time in the atmosphere will be
7-1
-------
brief if the emissions are deposited locally or may extend to days or weeks if long range
transport occurs. The chemical form in which emissions ultimately reach the receptor is
determined by the complex chemical transformations that take place between the emission
sources and the receptor. Long range transport over distances of hundreds or thousands of
miles allows time for a greater number of chemical transformations to occur (see Chapter 6).
Sulfates and nitrates are among the products of the chemical transformations of sulfur
and nitrogen oxides. Ozone and other photochemical oxidants are believed to be involved in
the chemical processes that transform sulfur dioxide and nitrogen oxides in the atmosphere
into sulfuric and nitric acids. When these acids are brought to earth in rain and snow,
acidic precipitation occurs. Because of long range transport, acidic precipitation in a
particular state or region can be the result of emissions from sources in states or regions
many miles away, rather than from local sources. To date, however, the complex nature of the
chemical transformation processes has made it difficult to demonstrate a direct cause and
effect relationship between emissions of sulfur and nitrogen oxides and the acidity of pre-
cipitation.
Acidic precipitation is arbitrarily defined as precipitation with a pH less than 5.6.
This value has been selected because precipitation formed in an atmosphere relatively free of
natural or manmade emissions would have a pH of approximately 5.6 due to the combining of
carbon dioxide with water in the air to form carbonic acid.
Acidity of aqueous solutions is determined by the concentration of hydrogen ions (H )
present and is expressed in terms of pH units—the logarithm of the inverse activity of
hydrogen ions. The pH scale ranges from 0 to 14, with a value of 7 representing a neutral
solution. Solutions with values less than 7 are acidic, while values greater than 7 are
4
basic. Because pH is a logarithmic scale, a change of one unit represents a tenfold change in
acidity, hence pH 3 is ten times as acidic as pH 4. Currently the acidity of precipitation in
the Northeastern United States normally ranges from pH 3.9 to 5.0; in other regions of the
United States precipitation episodes with a pH as low as 3.0 have been reported. For
comparison, the pH of some familiar substances are: cow's milk, 6.6; tomato juice, 4.3; cola
(soft drink) 2.8; and lemon juice, 2.3.
The pH of precipitation can vary during an event, from event to event, from season to
season, and from geographical area to geographical area. Substances in the atmosphere can
cause the pH to shift by making it more acidic or more basic. Dust and debris swept up in
small amounts from the ground into the atmosphere may become components of precipitation. In
the West and Midwest soil particles tend to be basic, but in the eastern United States they
tend to be acidic. Industrial emissions of limestone particles and similar oxides and car-
bonates are basic. As gaseous ammonia" from decaying organic matter makes precipitation more
basic, ammonia influences the acidity of precipitation in areas where there are large stock-
yards or other sources of organic matter.
7-2
-------
In the Eastern United States sulfur oxide emissions are greater than nitrogen oxide
emissions; hence, sulfates are greater contributors in this region to the formation of acids
in precipitation. The ratio between the two emissions, however, has been decreasing. Sulfate
concentrations are greater in summer than in winter in the Eastern United States; however,
around some of the larger cities in California nitrate^, contribute more to the formation of
acidity in rainfall. In coastal areas sea spray strongly influences precipitation chemistry
by contributing sodium, calcium, potassium, chloride, and sulfates. In the final analysis,
the pH of precipitation reflects the overall contributions of all of these components.
The impact of acidic precipitation on lakes, streams, ponds, forests, fields and manmade
objects, therefore, is not the result of one, or even of several precipitation events, but of
continued additions of acids or acidifying substances over time. When did precipitation
become acidic? Some scientists state that it began with the Industrial Revolution and the
burning of large amounts of coal; others say that, in the United States, it began with the
introduction of tall stacks on power plants in the 1950's; other scientists disagree
completely and state that rain has always been acidic. In other words, no definitive answer to
the question exists at the present time, nor are there data to indicate with much confidence
trends of pH in precipitation because the pH of rain has not been continuously monitored in
the United States for any period of time. In Scandinavia, on the other hand, the pH of rain
has been continuously monitored for many years; it is more nearly possible, therefore, to
determine when precipitation began to become more acidic. Data from the European Chemistry
Network in the late 1950's revealed that the pH of the precipitation falling on southwestern
Scandinavia was < 4.7. At that time it was postulated that the acidification of freshwater
lakes and streams which had first been noted in southern Norway in 1911 and later during the
1920's had been caused by acidic precipitation.
Though acidic precipitation (wet deposition) is usually emphasized, it is not the only
process by which acids or acidifying substances are added to bodies of water or to the land.
Dry deposition also occurs. During wet deposition, substances such as sulfur and nitrogen
oxides are scavenged, by precipitation (rain and snow) and deposited on the surface of the
earth. Dry deposition processes include gravitational sedimentation of particles, impaction
of aerosols, adsorption and absorption of gases by objects at the earth's surface or by the
soil or water. Gases, solid and liquid aerosols can be removed by both wet and dry deposi-
tion. Dew, fog, and frost are also involved in the deposition processes but do not strictly
fall into the category of either wet or dry deposition. Dry deposition processes are not as
well understood as wet deposition at the present time; however, all of the deposition pro-
cesses contribute to the gradual accumulation of acidic or acidifying substances in the en-
vironment. In any event, precipitation in the Eastern United States at the present time is
acidic and has been associated with changes in ponds, lakes, and streams that are considered
by humans to be detrimental to their welfare.
7-3
-------
The most visible changes associated with both wet and dry acidic deposition are those
observed in the lakes and streams of the Adirondack Mountains in New York State, the
Precambrian Shield areas of Canada and in the Scandinavian countries. In these regions, the
pH of the fresh water bodies has decreased, causing changes in animal and plant populations.
The most readily observable effect, however, has been the decrease in fish populations.
The chemistry of fresh waters is determined primarily by the geological structure (soil
system and bedrock) of the lake or stream catchment basin, by the ground cover and by land
use. Near coastal areas, that is up to about 160 kilometers (100 miles) from the sea, marine
salts also may be important in determining the chemical composition of the stream, river, or
lake.
Sensitivity of a lake to acidification depends on the acidity of both wet and dry deposi-
tion plus the same factors—the soil system of the drainage basin, the canopy effects of the
ground cover, and the composition of the watershed bedrock—that determine the chemical com-
position of fresh water bodies. The capability of a lake and its drainage basin to neutralize
incoming acidic substances, however, is determined largely by the composition of the bedrocks.
Soft water lakes, those most sensitive to additions of acidic substances, are usually
found in areas with igneous bedrock which contributes few soluble solids to the surface waters,
whereas hard waters contain large concentrations of alkaline earths (chiefly bicarbonates of
calcium and sometimes magnesium and iron) derived from limestones and calcareous sandstones in
the drainage basin. Alkalinity is associated with the capacity of lakes to neutralize or
buffer the incoming acids. The quantity of acidic precipitation necessary to acidify a sensi-
tive lake system has yet to be determined.
The disappearance of fish populations from freshwater lakes and streams is usually one of
the most readily observable signs of lake acidification. Death of fish in acidified waters has
been attributed to the modification of a number of physiological processes by a change in pH.
Two patterns of pH change have been observed; the first involves a sudden short-term drop in
pH and the second, a gradual decrease in pH with time. Sudden short-term drops in pH often
result from a winter thaw or the melting of the snow pack in early spring and the release of
the acidic constituents of the snow into the water. These short-term changes in water
chemistry may have significant impacts on aquatic biota, especially if they occur at sensitive
times in the life cycle (e.g., during spawning or early stages of development).
A gradual decrease in pH, particularly below 5, can interfere with reproduction and spawn-
ing of fish until elimination of the population occurs. In some lakes, aluminum mobilization
in fresh waters at a pH below 5 has resulted in fish mortality .
Although the disappearance of and/or reductions in fish populations are usually empha-
sized as significant results of lake and stream acidification, changes of equal or greater im-
portance are the effects on other aquatic organisms ranging from waterfowl to bacteria. Or-
ganisms at all trophic (feeding) levels in the food web appear to be affected. Reduction in
7-4
-------
number and diversity of species may occur; biomass (total mass of living organisms in a given
volume of water) may be altered and processes, such as primary production and decomposition,
impaired.
Primary production and decomposition are the bases of the two major food webs (grazing
and detrital) within an ecosystem by which energy is passed along from one organism to another
through a series of steps of eating and being eaten. Green plants, through the process of
photosynthesis, are the primary energy producers in the grazing web, while bacteria initiate
the detrital food web by feeding on dead organic matter. Disruption of either of these two
food webs results in a decrease in the supply of nutrients, interferes with their cycling, and
also reduces energy flow within the affected ecosystems. Acidification of lakes and streams
affects both these processes when a slowing down in the rate of microbial decomposition causes
an alteration of the species composition and structure of the pondweed and algae plant communi-
ties.
At present, there are no documented observations or measurements of changes in natural
terrestrial ecosystems that can be directly attributed to acidic precipitation. The informa-
tion available is an accumulation of the results of a wide variety of controlled research
approaches largely in the laboratory, using in most instances some form of "simulated" acidic
rain, frequently dilute sulfuric acid.
Soils may become gradually acidified from an influx of hydrogen (H ) ions. Leaching of
the mobilizable forms of mineral nutrients may occur. The rate of leaching is determined by
the buffering capacity of the soil as well as the amount and composition -of precipitation.
Unless the buffering capacity of the soil is high and/or the salt content of precipitation is
high, leaching will, in time, result in acidification. At present, there are no studies show-
ing this process has occurred because of acidic precipitation.
Damage to stone monuments and buildings, corrosion of metals and deterioration of paint
can result from acidic precipitation. Because sulfur compounds are a dominant component of
acidic precipitation and are deposited during dry deposition also, the effects resulting from
the two processes cannot be distinguished. In addition, the deposition of sulfur compounds on
stone surfaces provides a medium for microbial growth that can result in deterioration. (See
Chapter 10 for a more detailed discussion of materials effects.)
Human health effects due to the acidification of lakes and rivers have been postulated.
Fish in acidified water may contain toxic metals mobilized due to the acidity of the water.
Drinking water may contain toxic metals or leach lead from the pipes bringing water into the
homes. Humans eating contaminated fish or drinking contaminated water could become ill. No
instances of these effects having occurred have been documented.
, Several aspects of the acidic precipitation problem remain subject to debate because
existing data are ambiguous or inadequate. Important unresolved issues include:
1. The rate at which rainfall is becoming more acidic and the rate at
which the phenomenon is becoming geographically widespread.
7-5
-------
2. The relative extent to which the acidity of rainfall in a region depends on local
emissions of nitrogen and sulfur oxides versus emissions transported from distant
sources.
3. The relative importance of changes in total mass emission rates compared to changes
in the nature of the emission patterns e.g., ground level versus tall stacks in con-
tributing to regional acidification of precipitation.
4. The relative contribution of wet and dry deposition to the acidification of lakes
and streams.
5. The geographic distribution of natural sources of NO and SO , and NH, and the
XX i5
significance and seasonality of their contributions.
6. The existence and significance of anthropogenic, non-combustion sources of SO , NO
and HC1.
7. The dry deposition rates for S02, NOp, sulfate, nitrate and HC1 over various
terrains and seasons of the year.
8. The existence and reliability of long-term pH measurements of lakes and headwater
streams.
9. The acceptability of current models for predicting long range transport of SO and
NO and of those for predicting the acid tolerance of lakes.
X
10. The feasibility and costs of using liming or other corrective procedures to prevent
or reverse damage from acidification.
11. The differential effects of SO and NO and hydrogen ion deposition on ecosystem
-------
fields, lakes, rivers, estuaries, and oceans) within the biosphere obtain energy from the sun,
nutrients from the earth's crust (the lithosphere), gases from the atmosphere, and water from
the hydrosphere. All of the living systems are interdependent. Energy and nutrients move
from one to another. The living systems, together with their physical environments, the
lithosphere, hydrosphere and atmosphere make up the ecosystem that is the planet Earth
(Billings, 1978; Boughey, 1971; Odum, 1971; Smith,, 1980).
Ecosystems are composed of biotic (living) and abiotic (nonliving) components. The bio-
tic component consists of: (1) producers, green plants that capture the energy of the sun;
(2) consumers that utilize as their energy source the food stored by the producers; and (3)
decomposers who obtain their energy by breaking down and converting dead^organic matter into
inorganic compounds. (See Table 7-1). The abiotic components are air, water, the soil
matrix, sediment, particulate matter, dissolved organic matter, and nutrients in aquatic
systems, and dead or inactive organic matter in terrestrial systems (see Table 7-1) (Billings,
1978; Boughey, 1971; Smith, 1980).
Ecosystems are basically energy processing systems "whose components have evolved to-
gether over a long period of time. The boundaries of the system are determined by the en-
vironment, that is, by what forms of life can be sustained by the environmental conditions of
a particular region. Plant and animal populations within the system represent the objects
through which the system functions." (Smith, 1980).
Ecosystems are open systems. They both receive from and contribute to the environment
that surrounds them.- The environment contributes gases, nutrients, and energy. Ecosystems
utilize these substances and, in turn, make their own contributions to the environment.
Energy flows through the system unidirectionally while water, gases and nutrients are usually
recycled and fed back into the system. The functioning of ecosystems is greatly influenced by
the extent to which the gases and nutrients are fed back into the system. When materials are
not returned to an ecosystem through recycling, they must be obtained in another-way. The
organismal populations are the structural elements of the ecosystem through which energy flows
and nutrients are cycled. (Smith, 1980; Billings, 1978; Odum, 1971).
Energy from the sun is the driving force in most ecosystems. Without it, virtually all
ecosystems would cease to function. The energy of the sun is captured by green plants through
the process of photosynthesis and is stored in plant tissues. This stored energy is passed
along through ecosystems by a series of feeding steps, known as food chains, in which
organisms eat and are eaten. Energy flows through ecosystems in two major food chains, the
grazing food chain and the detrital food chain. The amount of energy that passes through the
two food chains varies from community to community. The detrital food chain is dominant in
most terrestrial and shallow-water ecosystems. The grazing food chain may be dominant in deep-
water aquatic ecosystems (Smith, 1980). The two fundamental processes involved in these two
food chains are: (1) photosynthesis, the capture of energy from the sun by green plants, and
(2) decomposition, the final dissipation of energy and the reduction of organic matter into
inorganic nutrients.
7-7
-------
TABLE 7-1. COMPOSITION OF ECOSYSTEMS
Component
Description
Biotic (biological):
Individuals
Producers
Consumers
Decomposers
Populations
Communities
Plants, animals (man), and microorganisms.
These are either producers, consumers, or
decomposers.
Green plants.
Herbivores, carnivores.
Macroorganisms (mites, earthworms, millipedes,
and slugs) and microorganisms (bacteria
and fungi).
Groups of interbreeding organisms of the same
kind, producers, consumers or decomposers,
occupying a particular habitat.
Interacting populations linked together by
their responses to a common environment.
Abiotic (physical):
Atmosphere
Cosmic radiation, temperature, thermal
radiation, radioactivity including fallout.
Water vapor, cloud and precipitation.
Gases, pressure and wind,
Heat and temperature,
Fire and pollutants
Lithosphere
Rock and soil particles
Minerals, water,
Radioactivity, heat and temperature, gases
Topography
Earth mass
Gravity
Adapted from Billings (1978).
7-8
-------
In addition to the flow of energy, the existence of the living world depends upon the
circulation of nutrients through the ecosystems. Both energy and nutrients move through the
ecosystem as organic matter. It is not possible to separate one from the other. Both influ-
ence the abundance of organisms, metabolic rate at which they live, and the complexity and
structure of the ecosystem (Smith, 1980). Nutrients, unlike energy, after moving from the
living to the nonliving return to the living components of the ecosystem in a perpetual cycle.
It is through the cycling of nutrients that plants and animals obtain the minerals necessary
for their existence.
The gaseous and sedimentary cycles are the two basic types of nutrient or biogeochemical
cycles. The gaseous cycles involve carbon, oxygen, and nitrogen. Water, also, is sometimes
considered as belonging to the gaseous cycle. In the gaseous cycles, the main nutrient reser-
voir is the atmosphere and the ocean. In the sedimentary cycle, to which phosphorus belongs,
the reservoir is the soil and rocks of the earth's crust. The sulfur cycle, a combination of
the two cycles, has reservoirs in both the atmosphere and earth's crust.
Nitrogen, sulfur and water cycles are involved-in acidic deposition. Nitrogen, through
the agency of plants (chiefly legumes and blue green algae) and microorganisms, moves from the
atmosphere to the soil and back (see Figure 7-1). Human intrusion into the nitrogen cycle *
includes the addition of nitrogen oxides to the atmosphere, and ammonia and nitrates to
aquatic ecosystems. Sulfur enters the atmosphere from volcanic eruptions, weathering, the
surface of the ocean, gases released in the decomposition processes, and combustion of fossil
fuels (see Chapter 4 for details). There is no way to harness or use energy without creating
an environmental impact. Both the nitrogen and sulfur cycles have been overloaded in some
parts of the world by the combustion of fossil fuels by man. For these cycles to function, an
ecosystem must possess a number of structured relationships among its components. By changing
the amounts of nitrogen and sulfur moving through the cycles, humans have perturbed or upset
the structured relationships that have existed for millions of years and altered the movement
of the elements through the ecosystems. The pathways that the elements take through the
system depend upon the interaction of the populations and their relationships to each other.
Change is one of the basic characteristics of our environment. Weather changes from day
to day, temperatures rise and fall, rains come and go, soils erode, volcanoes erupt, and winds
blow across the land. These are natural phenomena. Significant environmental changes also
result when human beings clear forests, build cities and factories, and dam rivers. All of
these environmental changes influence the organisms that live in the ecosystems where the
changes are occurring (Moran et al., 1980).
Existing studies indicate that changes occurring within ecosystems, in response to pollu-
tion or other disturbances, follow definite patterns that are similar even in different eco-
systems. It is possible, therefore, to predict the basic biotic responses of an ecosystem to
disturbances caused by environmental stress (Woodwell, 1970; Woodwell, 1962; Odum, 1965).
These responses to disturbance are: (1) removal of sensitive organisms at the species and
7-9
-------
Jf-~*^, -^ minim
' \ f V sxcntiioN /
'• NMj' \ 06ATH 4 /
v_j/ I \ ORGANIC N
>=ZLj~~~±e—~—p-~~~'
I*' N IN SEDIMENTS r
•«; "^ SEE-AOi
Figure 7-1. Schematic representation of the nitrogen cycle, emphasizing human activities that affect fluxes of
nitrogen (National Research Council, 1978).
-------
subspecies level due to differential kill; (2) reduction in the number of plants and animals
(standing crop); (3) inhibition of growth or reduction in productivity; (4) disruption of food
chains; (5) return to a previous state of development; and (6) modification in the rates of
nutrient cycling.
Ecosystems can respond to environmental changes or perturbations only through the
response of the populations of organisms of which they are composed (Smith, 1980). The
species of organisms sensitive to environmental changes are removed. Therefore, the capacity
of an ecosystem to maintain internal stability is determined by the ability of individual
organisms to adjust their physiology or behavior to change. The success with which an
organism copes with environmental changes is determined by its ability to produce reproducing
offspring. The size and success of a population depends upon the collective ability of
organisms to reproduce and maintain their numbers in a particular environment. Those
organisms that adjust best contribute most to future generations because they have the
greatest number of progeny in the population (Smith, 1980; Billings, 1978; Woodwell, 1970;
Odum, 1971).
The capacity of organisms to withstand injury from weather extremes, pesticides, acidic
deposition or polluted air follows the principle of limiting factors (Billings, 1978; Odum,
1971; Moran et a!., 1980; Smith, 1980). According to this principle, for each physical factor
in the environment there exists for each organism a minimum and a maximum limit beyond which
no members of a particular species can survive. Either too much or too little of a factor
such as heat, light, water, or minerals (even though they are necessary for life) can jeopar-
dize the survival of an individual and, in extreme cases, a species (Billings, 1978; Smith,
1980; Boughey, 1971; Odum, 1971). The range of tolerance (see Figure 7-2) of an organism may
be broad for one factor and narrow for another. The tolerance limit for each species is
determined by its genetic makeup and varies from species to species for the same reason." The
range of tolerance also varies depending on the age, stage of growth, or growth form of an
organism. Limiting factors are, therefore, those which, when scarce or overabundant, limit
the growth, reproduction, and/or distribution of an organism (Billings, 1978; Smith, 1980;
Boughey, 1971; Odum, 1971; Moran et a!., 1980). The increasing acidity of water in lakes and
streams is such a factor.
Organisms can exist only within their range of tolerance. Some populations of organisms,
annual plants, insects, and mice, for example, respond rapidly to environmental change. They
increase in numbers under favorable conditions and decline rapidly when conditions are
unfavorable. Populations of other organisms, such as trees and wolves, fluctuate less in
response to favorable or unfavorable conditions by showing little variation in the rates of
reproduction and death. Adaptation is the ability of an organism to conform to its
environment. Ecosystem stability ultimately is based on the adaptability of the organisms
that compose it. Stability may be associated with the ability of a system to return to an
equilibrium state after a temporary disturbance (Holling, 1973; May, 1973). The less it
7-11
-------
ZONE OF
INTOLERANCE
tOWtR UNITS
OF TOLERANCE
ZONE OF
PHYSIOLOGICAL
STRESS
TOLERANCE RANGE
RANGE OF OPTIMUM
ZONE OF
PHYSIOLOGICAL
STRESS
ZONE OF
INTOLERANCE
ORGANISMS
INFREQUENT
ORGANISMS
ABSENT
GREATEST
ABUNDANCE
ORGANISMS
INFREQUENT
ORGANISMS
ABSENT
LOW4-
-BBAOIiNT-
-*HtGH
Figure 7-2. Law of tolerance.
Source: Adapted from Smith (1980).
7-12
-------
varies from and the faster it returns to its original state, the more stable the system
(Smith, 1980). Stability also involves persistence, the ability of the populations of an eco-
system to persist through time. Persistence involves resilience, the ability of an ecosystem
to absorb changes. Although individual populations within a system may fluctuate greatly in
response to environmental changes, the system may be highly resilient (Holling, 1973; Smith,
1980). Contrasted with resilience is resistance, the ability of a system, because of its
structure, to resist changes from disturbances. Typically, the most resistant ecosystems have
large living components, trees for example, and store nutrients and energy in the standing
biomass. Resistant systems such as forests, once highly disturbed, are very slow in returning
to their original state (Smith, 1980).
Aquatic ecosystems that lack components in which energy and nutrients may be stored for
long periods of time usually are not very resistant to environmental changes (Smith, 1980).
For example, an influx of pollutants, such as effluents from sewage disrupts the system be-
cause more nutrients enter the system than it can 'handle. However, since the nutrients are
not retained or recycled within the system it returns to its original state in a relatively
short time after the perturbation is removed.
No barriers exist between the various environmental factors, or between an organism or
biotic community and its environment. Because an ecosystem is a complex of interacting com-
ponents, if one factor is changed, almost all will change eventually. "The ecosystem reacts
as a whole. It is practically impossible to wall off a single factor or organism in nature
and control it at will without affecting the rest of the ecosystem. Any change no matter how
small is reflected in some way throughout the ecosystem: no 'walls' have yet been discovered
that prevent these interactions from taking place" (Billings, 1978).
Continued or severe perturbation of an ecosystem can overcome its resistance or prevent
its recovery, with the result that the original ecosystem will be replaced by a new system.
In the Adirondack Mountains of New York State, eastern Canada, and parts of Scandinavia, the
original aquatic ecosystems have been and are continuing to be replaced by ecosystems differ-
ent from the original due to acidification of the aquatic habitat. Forest ecosystems appear
to be more resistant thus far, because changes due to stress from acidifying substances have
not been detected.
7.2 CAUSES OF ACIDIC PRECIPITATION
7.2.1 Emissions of Sulfur and Nitrogen Oxides
The generally held hypothesis is that increased emissions of sulfur and nitrogen com-
pounds are largely responsible for the acidity of precipitation (Smith, 1872; Bolin et al.,
1972; Likens, 1976). The emissions of the sulfur and nitrogen compounds involved in the acidi-
fication are attributed chiefly to the combustion of fossil fuels. Emissions from natural
sources can also be involved; however, in highly industrialized areas emissions from manmade
sources usually exceed those from natural sources (see Chapter 4). (See Chapter 5 for environ-
mental concentrations of sulfur compounds.)
7-13
-------
After 1900 there was a nearly exponential increase in the consumption of gas and oil in
the United States (see Figure 7-3). Although the total consumption of coal has not increased
greatly since about 1925, the consumption of oil and gas has continued to rise precipitously,
thus overshadowing coal as the dominant fuel source during the past 50 years (Hubbert, 1976).
Within this overall increase in fossil-fuel use, there have been shifts in the pattern of con-
sumption. Whereas formerly a considerable proportion of coal was used for transportation and
heating, these functions have since been taken over by oil and gas. Coal is now predominantly
devoted to electric power generation (Figure 7-4). In fact, electric power generation is the
primary factor accounting for an absolute increase in coal consumption over the past two
decades. (The decline in the use of coal in the 1930s was due to the general economic depres-
sion, and the decline in the 1950s was due to the availability of relatively inexpensive oil
and gas.) Approximately 550 MM (million metric) tons (National Research Council, 1978b) were
used per year during 1918-1928 compared to 672 MM tons/year during 1979 (Hamilton, 1980).
There was, however, a seasonal shift in the pattern of coal consumption. Summer coal consump-
tion has increased since 1960 because of increased usage of electricity in air conditioning,
while winter consumption has decreased due to the use of alternative fuels.
These changes in the pattern of fuel use have been accompanied by changes in the pattern
of pollutant emissions. Figure 7-5A and 7-5B illustrate the rise since 1940 in emissions of
sulfur and nitrogen oxides, the primary gaseous pollutants resulting from the combustion of
fossil fuels (See also Chapter 4). Although there has been a net increase in both categories,
the more consistent rise has been in emissions of nitrogen oxides. Almost all (93 percent)
emissions of sulfur oxides in the United States arise from stationary point sources, princi-
pally industrial and power plant stacks. Nitrogen oxide pollutants, on the other hand, origi-
nate about 60 percent from transportation (mobile) sources and 30 percent from stationary
«
sources, which include not only industrial and power plants, but residential and institutional
heating equipment as well (U.S. Environmental Protection Agency, 1980). (see Chapter 5 of Air
Quality Criteria for Oxides of Nitrogen for a more detailed discussion.)
The geographic distributions of sources of the presumed gaseous precursors of acidic pre-
cipitation are depicted in Figures 7-6 and 7-7. Clearly, the main sources of sulfur oxides in
the United States are in the eastern half of the country, particularly the northeastern
quadrant. Major nitrogen oxide sources also show a tendency to be concentrated somewhat in
the Northeast. Chapter 4 should be consulted for a more detailed account of the sources and
emissions of sulfur oxides.
7.2.2 Transport of Nitrogen and Sulfur Oxides
Among the factors influencing the formation as well as the location where acidic deposi-
tion occurs is the long-range transport of nitrogen and sulfur oxides. Neither the gases nor
their transformation products always remain near the sources from which they have been emitted.
7-14
-------
I
10
o
o
o
cc
LU
50
40
30
20
10
1850
fOlL
1900
1950
YEAR
2000
2050
Figure 7-3. Historical patterns of fossil fuel consumption in the
United States
Source: Adapted from Hubbert (1976).
7-15
-------
o
oo
O
O
I
O
ce
800
700
600
BOO
400
300
200
100
1 I I I
I I
I I
TOTAL
RES. & COMM,
HEATING
RAILROADS Hi
I I I
- OTHER
-OVENCOKE
ELECTRIC
UTILITIES
I I
1900 10 20 30 40 50 60 70 80 90 2000
YEAR
Figure 7-4. Forms of coal usap in the United States, Electric
power generation is currently the primary user of coal. (Data
from U.'S. Bureau of Mines, Minerals Yearbooks 1933-1974).
Source: U,S. Bureau of Mines (1954, 1976).
7-16
-------
(a
O
S
tu
d*
tn
ta
O
35
30
2F
20 _
15
10
1940
35
30
.S 25
4=
09
§
20
15
10
0 I
1940
TRANSPORTATION
1950
1960
YEAR
1970
1980
T
TOTAL
TRANSPORTATION
I I
1950
1960
YEAR
1970
1980
Figure 7-5a. Trends in emissions of sulfur dioxides.
Figure 7-5b. Trends in emissions of nitrogen oxides.
Source: Office of Air Quality Planning and Standards
(1978).
7-17
-------
KEY
EMISSION DENSITY, tons/mi"2
20-50
>50
5'10
EH 10-20
Figure 7-6. Characterization of U.S. SOX emissions density by state.
{Roman numerals indicate EPA Regions.)
Source: U.S. Dept. of Energy (1981).
-------
0)
+*
R
S
0)
o
Jj
Q)
X «-.
i 1
°? "1?
3 (T
H« ^
O
o
i
a.
UJ
S "O
tS .5
2 «
co eg
1
O
Ul
e
Q
(0
3*
w i
* S »
^ c §
S I i
o> o o
E £ co
7-19
-------
They may be transported for long distances downwind (Altshuller and McBean, 1979; Pack el al.,
1978; Cogbill and Likens, 1974). (See Chapter 6 for a detailed discussion of transport and
transformation of sulfur oxides.)
The geographic picture of the problem of acidic precipitation in North America can be
better understood in the light of some information on prevailing wind patterns. Winds trans-
port the precursors of acidic precipitation from their points of origin to areas where the
acidified rain and snow eventually fall. Prevailing winds in the Eastern United States tend
to be from the'west and southwest. Atmospheric pollutants, therefore, are carried in a gener-
ally northeasterly direction. Thus, pollution originating in the Ohio River Val.ley can be
carried toward the New England States. Seasonal meteorological patterns, however, can modify
the direction of windflow, particularly in the summer. The Maritime Tropical air masses from
the Gulf of Mexico that occur in late summer have the greatest potential for the formation and
transport of high concentrations of sulfate into the Northeastern United States and into east-
ern Canada (Altshuller and McBean, 1979).
Cogbill and Likens (1974) associated acidic rainfall in central New York during 1972-73
with high altitude air masses transported into the region from the Midwest. They stated that
the NO and S02 that is involved in acidic rain formation may be transported distances of 300
to 1500 km. Reports by Miller et al. (1978), Wolff et al. (1979), and Galvin et al. (1978)
all support the concept that the trajectories of the air masses which come from the Midwest
carry sulfur and nitrogen compounds that acidify precipitation in New York State.
A significant, though disputed factor, in this transport picture is the height at which
the pollutants are emitted. Industrial and power plant smokestacks emit their effluents into
the atmosphere at higher elevations than do motor vehicles or most space heating equipment.
In fact, there has been a trend since the 1960s toward building higher stacks as a means of
dispersing pollutants and thereby reducing pollutant concentrations in the vicinity of power
plants, smelters, and similar sources (Grennard and Ross, 1974). The result has been that
sulfur and nitrogen oxides are carried by prevailing winds for long distances and allowed to
diffuse over greater areas through the atmosphere. Concomitantly, long-range transport allows
greater time for chemical reactions to convert these pollutant gases into particulate forms.
(Eliassen and Saltbones, 1975; Smith and Jeffrey, 1975; Prahm et al., 1976). Chapter 6 dis-
cusses the chemical transformations, wet and dry deposition, transport and diffusion of sulfur
oxides in the atmosphere. Sulfur dioxide and nitrogen oxides are oxidized and hydrolyzed in
the atmosphere to nitric (HNO~) and sulfuric (H-SQ,) acids. These acids are considered to be
the main components of acidic precipitation.
The mechanisms of these chemical reactions are quite complex and depend on a host of vari-
ables ranging from physical properties of the pollutants to weather conditions and the presence
of catalytic or interacting agents (Fisher, 1978). Although the atmospheric chemical processes
are not well understood, it does appear that the long-range transport of sulfur compounds can
cover 1000 to 2000 km over three to five days (Pack et al., 1978). Thus, the impact of sulfur
[
7-20
-------
pollutants in the form of acidic precipitation may be far removed from their points of origin.
It is not yet clear whether nitrogen oxide pollutants may be transported distances comparable
to that of sulfur compounds or are local in origin (Pack, 1978); in the Northeast, however,
nitrates are currently thought to contribute 15 to 30 percent of the acidity of polluted pre-
cipitation. This estimate has increased over the past few years and is expected to increase
still further in the future (Robinson et al., 1978).
Evidence from northern Europe (Oden, 1968) also supports the idea that acidic rainfall is
a large-scale regional problem involving long distances between emission sources and
deposition of acidic precipitation. The acid rains that have received intensive study in
southern Scandinavia have been shown to result primarily from emissions of nitrogen and sulfur
oxides in Great Britain and the industrial regions of continental Western Europe, e.g.,
Holland, Belgium, and West Germany (Brosset, 1973).
7.2.3 Formation
Precipitation is that portion of the global-water cycle by which water vapor from the
atmosphere is converted to rain or snow and then deposited on the earth surfaces (Smith,
1980). Water moves into the atmosphere by evaporation and transpiration (water vapor lost by
vegetation). Once it reaches the atmosphere, the water vapor is cooled, then condenses on
solid particles and soon reaches equilibrium with atmospheric gases. One of the gases is car-
bon dioxide. As carbon dioxode dissolves in water, carbonic acid (H^GOO is formed. Carbonic
•acid is a weak acid and in distilled water only dissociates slightly, yielding hydrogen ions
and bicarbonate ions (HCQ., ). When relatively pure water is in equilibrium with normal
atmospheric concentrations and pressures of carbon dioxide, the pH of rain and snow is
approximately 5.6 (Likens et al., 1979).
The pH of precipitation may vary and become more basic or more acidic depending on sub-
stances in the atmosphere. Dust and debris may be swept from the ground and into the
atmosphere in small amounts where it can become a component of rain. The amounts of the
various substances in the atmosphere originating from seawater, desert sands, volcanic
islands, or vegetated land influence the chemistry of natural precipitation. In regions with
calcareous soils, calcium and carbonate may enter precipitation as dust, subsequently
increasing the pH of rain or snow to 6.0 or above (Likens et al., 1979). Soil particles are
2+
usually slightly basic in distilled water and release positive ions, such as calcium (Ca ),
magnesium (Mg ), potassium (K ), and sodium (Na+) into solution. Bicarbonate usually is the
corresponding negative ion. Decaying organic matter adds gaseous ammonia to the atmosphere.
Ammonia gas, in rain or .snow forms ammonium ions (NH, ) and tends to increase the pH. In
coastal areas, sea spray pJ.ayS' a strong role in the chemistry of precipitation. The important
ions entering into precipitation—sodium, magnesium, calcium, potassium, and the anions
•» O«~
chloride (Cl ) and sulfate (SO. )—are also those most abundant in ocean water (Likens, 1976;
Likens et al., 1979).
7-21
-------
Gases, in addition to CQ^, which enter precipitation are ammonia, sulfur dioxide (S02)
and the nitrogen oxides (NO ). Sulfur gases originating from natural sources (e.g., volcanoes
and swamps) may also enter precipitation. In the wet atmosphere, S0« can be oxidized to sul-
furic acid (Likens, 1976; Likens et a!., 1979). Strong acids dissociate completely in dilute
aqueous solutions and may lower the pH of precipitation to less than 5.6. Rain or snow with a
pH below 5.6 has been arbitrarily considered acidic precipitation by many scientists (Galloway
and Cowling, 1978; Wood, 1975; Likens et al., 1979).
The most important natural sources of atmospheric sulfur are biologically reduced com-
pounds coming from lake and sea bottoms, marshes, swamps, polluted estuaries, streams, tidal
basins, and decaying vegetation. Such reduction occurs most readily under oxygen deficient
conditions when organic matter is present (Cullis and Hirschler, 1980). It is virtually
impossible to measure or calculate the quantities of volatile sulfur compounds emitted from
biogenic sources, mainly due to lack of quantitative information concerning their reactions in
the atmosphere (Cullis and Hirschler, 1980). Hydrogen sulfide, assumed initially to be the
predominant reduced sulfur compound, reacts very slowly with oxygen in the absence of
catalysts, but is oxidized photochemically (Eggleton and Cox, 1978). Estimates for the life
time of H,S near the Earth's surface range from 2 hours (Cadle and Ledford, 1966) to nearly 4
days (Friend, 1973). Dimethyl sulfide has been shown to be present in considerably higher
concentrations above a pond (Rasmussen, 1974); however, it is much less readily oxidized than
HqS (Cadle, 1976). Biogenic sulfur emissions are much higher over the oceans than land
(Cullis and Hirschler, 1980). (See also Chapter 4.)
7.2.3,1 Composition and pH of Precipitation—Sulfur and nitrogen compounds are chiefly re-
sponsible for the excess acidity of precipitation. Continuous measurement of pH in rain by
Likens et al. (197^) for the Hubbard Brook Experimental Forest in New Hampshire from 1964 to
1971 indicated the precipitation was acid with an annual weighted average pH range of 4.03 to
4.19. (A weighted average takes into account the amount of rain as well as its composition.)
Cogbill and Likens (1974), analyzing precipitation from the Ithaca area and Hubbard Brook,
which consistently had a pH of less than 4.4, reported that their analysis of precipitation
showed that 65 percent of the acidity was due to H,SO_ 30 percent to HN03, and less than 5
percent was due to HC1.
In 1976, Likens (1976) reported that the continued monitoring of precipitation at the
Hubbard Brook Forest through 1974 indicated the average annual weighted pH for the years
1964-1974 ranged from 4.03 to 4.21. There was a downward trend in annual pH values between
1964-65 and 1970-71 followed by an upward trend until 1973-74, but no statistically signi-
ficant trend was noted for the 1964-74 period; however, pH values of 2.1 and 3.0 were observed
for individual storms at various locations. There was an increase in nitric acid in the pre-
cipitation (rain and snow) falling there. This change in the composition of acidic precipita-
tion suggests that the sources of nitrogen oxide emissions increased while those for sulfur
oxides remained constant.
7-22
-------
The acidity of precipitation is a reflection of the hydrogen ions in precipitation. The
contribution of sulfate and nitrate anions has changed with time, and analysis indicates that
the nitrate anion makes up an ever-increasing fraction of the total negative ion equivalents.
+ 2-
Following the reasoning of Granat (1972), Likens et al, (1^,76) found [assuming 2H per SOt ion
as in hLSO. or one H ion per NO- as in (HNO,)] that the contribution of sulfate to acidity
declined from 83 to 66 percent of the total acidity between 1964 tp 1974 at Hubbard Brook, and
the contribution of nitrate increased from 15 percent to 30 percent of the total during the
same period. Furthermore, increased annual input of H was closely correlated with increased
input of nitrate, but there was little correlation between H input and sulfate input.
Data for nitrate, ammonium, and sulfate in rain at Ithaca and Geneva, New York, consti-
tute the longest record of precipitation chemistry in the United States (Likens, 1972). Data
are available from 1915 to the present, but long gaps exist in the measurements, especially at
the Geneva site. Figures 7-8 (A) to (C) show that marked changes in composition have occurred
at Ithaca: a gradual decline in ammonium, an increase in nitrate beginning around 1945, and a
marked decrease in sulfate starting between 1945 and 1950. Early data for Ithaca showed
higher concentrations of sulfate in winter than in summer, presumably because of greater local
burning of coal in winter. Data for 1971 showed the reverse trend, however, with nearly half
the annual sulfate input occurring during the months of June to August. Likens (1972)
concluded that, despite deficiencies in the historical data and questions concerning their
reliability, the trends are real and can be explained by changes in fuel consumption patterns,
i.e., natural gas began to replace coal for home heating near the time of the shifts in
precipitation chemistry. On the basis of United States Geological Survey data for nine
stations, Likens (1976) reported a sharp increase in nitrate concentrations in New York State
during the past decade (Figure 7-8 [D]).
Data for eastern North America (the U.S. east of the Mississippi River) indicate a
roughly three-fold increase in nitrate in rainfall since 1955, whereas sulfate in rain has
roughly doubled in this period. According to Nisbet (1975), sulfate/nitrate ratios in
rainfall averaged about 4 in the Eastern United States in 1955-1956, but the average ratio had
fallen to about 3 in 1972-1973. Nisbet calculated that the fraction of.H deposition attri-
butable to nitrate rose from 19 percent in 1955-1956 to 24 percent in 1972-1973, while the
deposition attributable to H«SO. decreased from 80 to 73 percent,
2~ +
Lindberg et al. (1979) noted that SO, and H were by far the dominant constituents of
precipitation at the Walker Branch Watershed, Tennessee. Comparison with the annual average
concentration of major elements in rain at the Walker Branch Watershed on an equivalent basis
indicated that H constitutes approximately 50 percent of the cationic strength and trace
elements account for only 0.2 percent. Sulfate constituted approximately 65 percent of the
anionic strength and on an equivalent basis was 3.5 times more concentrated than NQ3, the next
7-23
-------
Ul
u.
_l
I I I III I I I I
1920 1930 1940 19SO 1960
YEAR
1970
1920 1930 1940 1950 1960 1970
YEAR
8
z
01
o
D>
1920 1930 1940 1<«0 1960 1970
YEAR
06
05
04
iu 0.3
t-
K 02
Z
01
0
1968 1970 1972 1974
YEAR
Rgure 7-8. Trends in mean annual concentrations of sulfate, ammonia,
and nitrate in precipitation. (A), (B), and (C) present long-term data for
Ithaca, New York; (D) presents data for eight years averaged over eight
sites in New York and one in Pennsylvania. One point in (A), for 1946-7,
is believed to be an anomaly (see Likens, 1972, for discussion}.
Source: (A), (B), and (C) modified from Likens (1972); (D) modified from
Likens (1976).
7-24
-------
most abundant anion. The incident precipitation for the 2-year (1976-1977) period was des-
cribed as "a dilute mineral acid solution," primarily H,,S04, at a pH approximating 4.2 and
containing relatively minor amounts of various trace salts (Lindberg et a!., 1979). In
Florida, Hendry (1977) and Hendry and Brezonik (1980) found that the relative proportions of
sulfate, nitrate and chloride ions in rainfall at Gainesville, Florida, during 1976, were 69
percent, 23 percent and 8 percent, respectively.
2-
Based on most reports, sulfate (SO. ) appears to be the predominant anion in acidic pre-
cipitation in the Eastern United States. In California, however, nitrate (N03) seems to pre-
dominate. Liljestrand and Morgan (1978) reported that their analyses of acidic rainfall
collected from February 1976 to September 1977 in the Pasadena, CA area showed that the volume-
weighted mean pH was 4.0, with nitric acid being 32 percent more important as a source of
+ + + 2+ 2+
acidity than sulfuric acid. The major cations present were H , NH., K , Ca and Mg while
2-
the major amons were Cl , NO,, and SO. . McColl and Bush (1978) also noted the strong in-
fluence of nitrate on rain in the Berkeley, CA, region. However, they note that in bulk pre-
2-
cipitation (wet plus dry fall-out) sulfate (SO. ) constituted 50 percent of the total anions.
Nearly all of the nitrate in rainfall is formed in the atmosphere from NO . Little is
/\
derived from wind erosion of nitrate salts in soils. Similarly, nearly all of the sulfate in
rainfall is formed in the atmosphere from S0? (Galloway, 1978). Thus, all atmospherically
+
derived nitrate and sulfate can contribute to the acidification of precipitation, since H is
associated stoichiometrically with the formation of each. A second stoichiometric process
that affects the acidity of rain is the reaction of nitric and sulfuric acids with ammonia or
other alkaline substances (e.g., dust particles) in the atmosphere to form neutral nitrate and
sulfate aerosols. To the extent that such neutralization occurs, the acidity of precipitation
will be reduced (National Research Council, 1978). However, since much of the ammonium ion
reaching soil is converted to nitrate, these neutral salts can still have an acidifying effect
on the soil. (See Section 7.3.2.1)
Seasonal fluctuations in composition as well as pH of rainfall have been reported by many
workers. In addition, the pH and composition of rainfall fluctuate from event to event, with-
in an event, from locality to locality, and from storm to storm.
2- +
In general SO. and H concentrations in precipitation in the eastern United States are
higher in the summer than in the winter. Wolff et al. (1979) found this to be true for the
New York metropolitan area. Hornbeck et al. (1976) and Miller et al. (1978) both stated that
a summer maximum- for sulfate was associated with an increase in hydrogen ion concentration in
upstate New York, the Hubbard Brook Experimental Forest in New Hampshire, and in portions of
Pennsylvania. Pack (1978), using data (1977) from the four original MAP3S (Multistate
Atmospheric Power Production Pollution Study) precipitation chemistry stations, plotted the
7-25
-------
weighted monthly sulfate ion concentrations shown in the figure by Lindberg, (Figure 7-9).
Maximum sulfate concentrations occurred from June through August. Lindberg et al. (1979),
studying wetfall deposition of sulfate in the Walker Branch Watershed, also noted summer
2- +
maximum for SO* and H . Using the same MAP3S data as did Pack, they plotted weighted mean
concentrations of sulfate in rain collected from November 1976 through November 1977. The
peak summer concentrations at Walker Branch Watershed, Tennessee, are lower than all of the
stations except remote Whiteface Mountain, New York, The regional nature of the wet deposi-
tion of sulfate is apparent. Reasons for the existence of the high summer maxima of sulfate
for the Eastern United States are discussed in some detail in Chapter 5, Section 5.5.1.
Seasonal variations of nitrogen compounds and of pH in precipitation have been reported
by several workers, but no simple trends are apparent (see U.S. Environmental Protection
Agency Air Quality Criteria for Oxides of Nitrogen, 1982). Hoeft et al. (1972) found rela-
tively constant levels of nitrate in rain and snow collected in Wisconsin throughout the year,
but deposition of ammonia and organic nitrogen was lowest in winter and highest in spring, per-
haps because of the thawing of frozen animal wastes. Haines (1976) reported large random vari-
ations, but relatively small seasonal variations, for the various forms of nitrogen in wet-
only precipitation at Sapelo Island, Georgia; nitrogen concentrations were lowest during the
rainy months of July and September. The highest nitrogen loadings occurred during July and
were-asosciated with the lowest range in pH (4.2-4.8). Hendry (1977) and Hendry and Brezonik
(1980) found relatively smooth seasonal trends in ammonia and nitrate concentrations in both
wet-only and bulk collections (wet- and dryfall) at Gainesville, Florida, with lowest concen-
trations in winter (Figure 7-10). In addition, the pH of the bulk precipitation showed no
seasonal trend. Wet-only collections, however, showed the lowest pH value (4.0) during the
spring and summer. The historical record suggests*there has been an increase in the concen-
tration of inorganic nitrogen in florida over the past 20 years.
Scavenging by rainfall produces large changes in atmospheric contaminant concentrations
during a given rainfall event. The decline in constituent levels is usually rapid, at least
in localized convective showers; low, steady-state concentrations.are usually reached within
the first half hour of a rain event due to cleansing of the atmosphere by rain (Brezonik,
1975). Major ions [chloride (CT ) and sulfate (SQ.~)], inorganic forms of nitrogen [nitrate
*• -t-
(NO- ) and ammonium (NH. )], total phosphorus and pH were measured in rain collected in 5-
minute segments within three individual rainstorms. Initially, rapid decreases were observed
for nitrate and ammonium and total phosphorus. Therf was also a decrease,in pH from 4.65 to
4.4. Steady state concentrations were reached in 10 minutes. Two other storms sampled in the
same manner showed similar but less defined patterns (Hendry & Brezonik, 1980).
Wolff et al. (1979) examined spatial, meteorological and seasonal factors associated with
the pH of precipitation in the New York metropolitan area. Seventy-two events were studied
7-26
\
-------
Figure 7-9. Comparison of weighted mean monthly concentrations of
sulfate in incident precipitation collected in Walker Branch Water-
shed, Tenn. (WBW) and four MAP3S precipitation chemistry monitor-
ing stations in New York, Pennsylvania, and Virginia.
Source: Lindberg eit al. (1979).
7-27
-------
< s
p
gj 0>
o 5
o
4.80
4.60
4.40
4.20
0.40
0.30
0.20
0.10
M
M
I I
A S O
1976
Wl A
-1977 —
M
MONTH
Rgure 7-10. Seasonal variations in pH (A) and ammonium
and nitrate concentrations (B) in wet-only precipitation at
Galnaville, Florida. Values are monthly volume-weighted
averages of levels in rain from individual storms.
Source: Hendry (1977).
7-28
-------
from 1975 through 1977. There was some site-to-site variability in the hydrogen ion
concentration expressed as pH values among the eight sites they studied in the Manhattan area
(Table 7-2). The standard deviation for individual sites ranged between 0.20 and 0.37. They
also noted that the pH varied according to storm type (Table 7-3). Storms with a continental
origin have a slightly lower pH than storms originating over the ocean. The storms with tra-
jectories from the south and southwest had the lowest pH's, while those from the north and
east had the highest pH's (Wolff et al., 1979).
The mean pH of precipitation falling on the New York metropolitan area during a 2-year
(1975 to 1977) study was 4.28; however, a pronounced seasonal variation was observed (Figure
7-11). The minimum pH at all sites except Manhattan was recorded during July to September,
while the maximum occurred during October to December. The minimum pH in Manhattan, however,
was measured between January and March and then gradually increased through the year. The
lowest mean pH of 4.12 for the New York Metropolitan area occurred during the summer months
(Wolff et al., 1979). In general, the pH of rain is usually lower in the summer than in the
winter and is associated with the high summertime sulfate'concentrations. In addition, the
»
lowest pH's were associated with cold fronts and air mass type precipitation events. These
events occur more frequently during the summer months. The lower pH's also occurred on
westerly or southwesterly winds (Wolff et al., 1979).
Seasonal variations in pH measured at several sites in New York State 70 km (45 mi.)
apart demonstrated a significant difference between seasons (winter had an average pH of 4.2;
summer, 3.9.) but no significant difference between sites. In New Hampshire, however, six
summer storms sampled at 4 sites less than 3 km (2 mi.) apart showed a significant difference
(3.8 to 4.2) indicating considerable variation in pH may occur in the same storm (Cogbill
1976).
Stensland (1978, 1980) compared the precipitation chemistry for 1954 and 1977 at a site
in central Illinois. The pH for the 1954 samples had not been measured, but were calculated
from the data of Larson and Hettick (1956) and compared with those measured in 1977. The
calculated pH for 1954 was 6.05; the pH for 1977 was 4.1. The more basic pH in 1954, accord-
ing to the author, could have resulted from low levels of acidic ions (e.g. sul fate or
nitrate) or from high amounts of basic ions (evg. calcium and magnesium). Stensland suggests
I I -iil--|r
that the higher pH in 1954 was due to calcium (Ca ) and magnesium (Mg ) ions from the soil.
7.2.3.2 Geographic Extent, of Acidic Precipitation—Acidic precipitation has been a reality in
New York State for an undetermined period of time. Data collected by the United States Geo-
logical Survey (Harr and Coffey, 1975) are presented in Figure 7-12. The pH of precipitation
has remained nearly at the same average level during the entire ten-year period; therefore,
since data for the years prior to 1965 are lacking, it is difficult to determine when the pH
in precipitation first began to decrease (Harr and Coffey, 1975).
7-29
-------
TABLE 7-2. MEAN pH VALUES IN THE NEW YORK METROPOLITAN
AREA (1975-1977)
Site
Cal dwell, N.J.
Piscataway, N.J.
Cranford, N.J.
Bronx, N.Y.
Manhattan, N.Y.
High Point, N.J.
Queens, N.Y.
Port Chester, N.Y.
All sites
Mean pH
4.32
4.25
4.34
4.31
4.29
4.25
4.63
4.60
4.28
SD
0.26
0.36
0.34
0.37
0.25
0.30
0.35
0.19
0.32
No. obsd
50
64
48
57
39
25
20
21
72
Range
3.35-5.60
3.57-5.50
3.44-5.95
3.42-5.75
3.80-5.50
3.74-4.90
3.98-5.28
4.00-5.10
3.50-5.16
Source; Wolff et al. (1979.)
TABLE 7-3. STORM TYPE CLASSIFICATION
Type
Description of dominant storm
system
No.
obsd
Mean
PH
I
2
3
4
5
6
7
8
Closed low-pressure system which formed
over continental N. Amer.
Closed low-pressure system which formed in
Gulf of Mexico or over Atlantic Ocean
Closed low which passed to W or N of N.Y.C.
Closed low which passed to S or E of N.Y.C.
Cold front in absence of closed low
Air mass thunderstorm
Hurricane Belle
Unclassified
22
21
26
17
16
5
1
6
4.35
4.43
4.39
4.39
4.17
3.91
5.16
4.31
Source: Wolff et al. (1979).
7-30
-------
4.6
4.5
4.4
i. 43
4.2
4.1
4.0
JFM AMJ JAS OND
MONTHS OF THE YEAR (1975 THROUGH 1977)
Figure 7-11. Seasonal variation of precipitation pH in the New
York Metropolitan Area.
Source: Wolff et al. (1979).
7-31
-------
70
60
SO
40
30
«
00
TO
60
50
«0
ALBANY. NEW YORK
001
60
50
40
30
oo"
60
SO
ALLEGHENY STATE PARK. NEW YORK
< I I
p i i i. i . i i ii 11 r i i i i .1.1.1.1 i il
ATHENS. PENNSYLVANIA
40
CANTON NEW YORK
it,
00 1 . i i i t t i
1965 1966 1967 1MB %K9 1910 1971 1972
19J3
Figure 7-12. History of acidic precipitation at various sites In and adjacent to State of
New York.
Source: Harr and Coffey (1975).
7-32
-------
70
60
SO
40
00*
70
60
SO
40
oo":
HINCKLEV, NEW YORK
MAYS POINT. NEW YORK
70
60
50
40
OO-lim.Mm
60
SO
40
00
MINEOLA, NEW YORK
ROCK HILL NEW YORK
UPTON Nf W YORK
"
1965
1966
196S
1969
YEAR
1970
1973
Figure 7-12 (cont'd). History of acidic precipatation at various sites in and adjacent to
State of New York.
Source: Harr and Coffey (1975).
7-33
-------
Reports indicate that precipitation is acidic in parts of the country other than the
northeastern United States (See Figure 7-13.). Average pH values around 4.5 have been
reported as far south as northern Florida (Likens, 1976; Hendry and Brezom'k, 1980), from
Illinois (Irving, 1978), the Denver area of Colorado (Lewis and Grant, 1980), the San
Francisco Bay area of California (McColl and Bush, 1978; Williams, 1978), Pasadena, California
(Liljestrand and Morgan, 1978), the Puget Sound area of Washington (Larson et al., 1975), and
from eastern Canada (Glass et al., 1979; Dillon et al., 1978). Data from the San Francisco
Bay area indicate that precipitation has become more acidic in that region since 1957-1958
(McColl and Bush, 1978). The pH decreased from 5.9 during 1957-1958 to 5.0 in 1974, and seems
to be related to an increase in the NCs concentration (McColl and Bush, 1978). Another
report, using data from the California Air Resources Board (Williams, 1978), states that
acidic precipitation has been reported from such widespread areas as Pasadena, Palo Alto,
Davis, and Lake Tahoe.
Studies in the Great Smokey Mountain National Park (Herrmann and Baron, 1980) indicate a
downward trend in pH has occurred there over the past twenty years. Over a period of 20
years, there has been a drop in pH from a range of 5.3 to 5.6 in 1955 to 4.3 in 1979.
The absence of a continuous precipitation monitoring network throughout the United States
in the past makes determination of trends in pH extremely difficult and controversial. This
shortcoming has been rectified recently through the establishment of the National Atmospheric
Deposition Program. Under the program, monitoring stations collect precipitation samples,
determine their pH and then send the samples to a Central Analytical Laboratory in Illinois to
be analyzed. This long-term network plans to have 75 to 100 collection sites throughout the
United States; 74 are already operational.
7.2.4 Acidic Deposition
The previous sections of this chapter have discussed the formation, composition, and geo-
graphic distribution of acidic precipitation. Usually when the effects of acidic deposition
are discussed, emphasis is placed on the effects resulting from the scavenging of sulfur and
nitrogen compounds by precipitation. Dry deposition of gaseous and particulate forms of these
compounds also occurs and is beginning to receive more emphasis in research (Galloway and
Whelpdale, 1980; Schlesinger and Hasey, 1980; Stensland, 1980; Sehmel, 1980; Chamberlain,
1980). Gaseous compounds reach the surface of the earth by turbulent transfer, whereas
particulate sulfates and nitrates reach the earth's surface by gravitational sedimentation,
turbulent transfer, and impaction (Galloway and Whelpdale, 1980; Sehmel, 1980; Hicks and
Wesely, 1980). A comparison of the relative significance of wet and dry deposition is diffi-
cult (See Chapter 6). Dry deposition, however, is always removing pollutants from the
atmosphere, while removal by wet deposition is intermittent (Sehmel, 1980). Marenco and
Fontan (1974) suggest that dry deposition is more important than wet in removing air
pollutants.
7-34
-------
-a
i
5.0
* Wltto«(X>«for«m(NMJP)
a
-O- dWontelMMiltiafncMwKieir
A MlMMUUAMMflMltetaMr
Production Mftitfeo tuaxium)
A
0
e
IMxraNyglMnM
Figure 7-13. pH of rain sample as measured in the laboratory and used in combination with Jhe
reported amount of precipitation.
Source: Wisniewski and Keitz (1981).
-------
Lindberg et al. (1979) have calculated the annual mass transfer rates of sulfates to the
forest floor in Tennessee (Figure 7-14). Their calculations for SO/~ suggest wet deposition
by incident precipitation to be 27 percent of the total annual flux compared with a total dry
precipitation of 13 percent. The dry deposition and foliar absorption of S02, a very import-
ant component, is missing from this calculation. The wet and dry deposition percentages are
indicative only of the relative magnitude of the two processes. The percentages, however, do
point out that the effects of acidic deposition usually attributed to precipitation scavenging
alone are probably "a result of both wet and dry deposition. At the present time the accuracy
with which dry deposition can be measured is still under question.
The studies of McColl and Bush (1978), Hendry and Brezonik (1980), and Schlesinger and
Hasey (1980) also point out that both wet and dry deposition are important when considering
a. 9— *-
the effects of H , SO, , and N03 ions on aquatic and terrestrial receptors.
The effects of the dry deposition of SO, and particulate matter on vegetation and terres-
trial ecosystems is discussed in Chapter 8. The processes of wet and dry deposition of sulfur
oxides are discussed in Chapter 6 of this document; such processes for nitrogen oxides are
discussed in Chapter 6 of Air Quality Criteria for Oxides of Nitrogen (U.S. E.P.A., 1982).
7.3 EFFECTS OF ACIDIC DEPOSITION
Acidic precipitation has been implicated in the acidification of surface waters leading
to the degradation of aquatic ecosystems, in the erosion of stone buildings and monuments, and
as a potential source of harm to forests and other terrestrial ecosystems. The sections that
follow discuss these effects.
7.3.1 Aquatic Ecosystems
Acidification of surface waters is a major problem in regions of southern Scandinavia
(Oden, 1968; Aimer et al., 1974; Gjessing et al., 1976), Scotland (Wright et al. , 1980a),
eastern Canada (Beamish and Harvey, 1972; Dillon et al., 1978), and the Eastern United States
in the Adirondack Region of New York State (Schofield, 1976 a»b,c,; Pfeiffer and Festa, 1980),
Maine (Davis et al., 1978), and northern Florida (Crisman et al., 1980). Damage to fisheries
is the most obvious effect of acidification on freshwater life. The disappearance of fish
populations from acidified fresh water lakes and streams was first noted in southern Norway in
the 1920's. In 1959, Dannevig proposed that acidic deposition was the probable cause for
acidification and thus for the loss of fish populations (Leivestad et al., 1976). Subsequent
studies have attempted to verify this postulate. Declines in fish populations have been re-
lated to acidification of surface waters in southern Norway (Jensen and Snekvik, 1972; Wright
and Snekvik, 1978), southwestern Sweden (Aimer et al., 1974), southwestern Scotland (Wright et
al., 1980a), the Adirondack Region of New York State (Schofield, 1976a,b,), and the LaCloche
Mountain Region in southern Ontario (Beamish and Harvey, 1972). Acidification may also have
serious repercussions on other aquatic biota inhabiting these systems. Changes in the acidity
and chemistry of freshwater affect the communities of organisms living there. Pertinent de-
tails of these effects are described in the following sections.
7-36
-------
IN CLOUD
PRECIPITATION
SCAVENGING
25%
IBELOW CLOUD , '/.'< •
PRECIPITATION *.':."'.'.'.
SCAVENGING <[::^^
2% T TO GROUND
(DORMANT PERIOD)
2%
»,,! '«. J|t I,
INCIDENT'PRECIPITATION
• . TOTAL DRY
•..-:.• DEPOSITION
13%
/ \
TO LEAFY
CANOPY
10V.
•, ' i ' 27%" 1*1
'• jiWj
j
TO BRANCHES
(DORMANT PERIOD)
1%
INTERNAL EXTERNAL
100%
RELATIVE ANNUAL MASS TRANSFER RATES
OF SOj-S TO THE FOREST FLOOR
Figure 7-14. Annual mass transfer rates of sulfate expressed as a percentage of the estimated
total annual flux of the element to the forest floor beneath a representative chestnut oak stand.
Source: Lindberg et al. (1979).
7-37
-------
7.3.1.1 Acidification of lakes and streams—Precipitation enters lakes directly as rain or
snow or indirectly as runoff or seepage water from the surrounding watershed. The relative
magnitude of the influents from these two sources is dependent on the surface area and volume
of the lake, and the size of the watershed and its soil volume and type. In general, the
watershed plays a dominant role in determining the composition of water entering the lake. As
a result, the water will be strongly influenced by processes in the edaphic environment of the
watershed, such as weathering, ion exchange, uptake and release of ions by plants, carbon
dioxide production by vegetation, microbial respiration, and reduction and oxidation reactions
of sulfur and nitrogen compounds (Seip, 1980). Precipitation as a direct source of water to
the lake plays a relatively greater role when lake areas are large in comparison to the size
of the watershed.
Acidification of surface waters results when the sources of hydrogen ion exceed the abi-
lity of an ecosystem to neutralize the hydrogen ion. In general, the soils and crust of the
earth are composed principally of basic materials with large capacities to buffer acids. How-
ever, areas where bedrock is particularly resistant to weathering and soils are thin and
poorly developed have much less neutralizing ability. This inability to neutralize hydrogen
ions does not usually arise from a limited soil or mineral buffering capacity. Instead low
cation exchange capacity and slow mineral dissolution rates in relation to the relatively
short retention time of water within the soil system may result in incomplete neutralization
of soil waters and acidification of surface waters (Driscoll, 1980). Characteristics of
regions sensitive to surface water acidification are discussed in more detail in Section
7.4.1.
Sources of hydrogen ions to the edaphic-aquatic system include, besides acidic deposi-
tion, mechanisms for internal generation of hydrogen ion - oxidation reactions (e.g., pyrite
oxidation, nitrification), cation uptake by vegetation (e.g., uptake of NH4 or Ca ), or gene-
ration of organic acids from incomplete organic litter decomposition (Figure 7-15). The rela-
tive importance of the hydrogen ion content in acidic deposition to the overall hydrogen ion
budget of an ecosystem has been discussed by many researchers (Rosenqvist, 1976; SNSF Project,
1977).
The consensus is that changes in internal hydrogen ion generation related to land use or
other changes (e.g., Drabl^s and Sevaldrud, 1980) can not consistently account for the wide-
spread acidification of surface waters occurring in southern Scandinavia, the Adirondack
Region of New York, the LaCloche Mountain Area of Ontario, and elsewhere. Driscoll (1980)
developed a hydrogen ion budget for the Hubbard Brook Area in New Hampshire. Based on these
calculations, atmospheric hydrogen ion sources represent 48 percent of the total Hubbard Brook
ecosystem hydrogen ion sources.
As noted above, fresh water ecosystems sensitive to inputs of acids are generally found
in areas of poor soil development and underlain by highly siliceous types of bedrock resistant
to dissolution through weathering (Likens et a!., 1979). As a result, surface waters in such
7-38
-------
ALLOCHTHONOUS SOURCES OF HYDROGEN ION
PRECIPITATION,
DRY DEPOSITION,
, r DRAINAGE WATER
• ECOSYSTEM BOUNDARY
HYDROGEN ION
SOURCES
OXIDATION RXN
CATION UPTAKE
PYR1TE
OXIDATION
NH/ UPTAKE
HYDROGEN ION
SINKS
REDUCTION RXN
ANION UPTAKE
OXIDE
WEATHERING
STREAM EXPORTS
H*. HCOg, OH-LIGANDS,
ORGANIC ANIONS
Figure 7-1 i.
cycle.
Schematic representation of the hydrogen ion
Source: Driscol! (1980).
7-39
-------
areas typically contain very low concentrations of ions derived from weathering. The waters
are diluted with low levels of dissolved salts and inorganic carbon, and low in acid neutra-
lizing capacity. The chemical composition of acid lakes is summarized in Table 7-4 for lakes
in southern Norway (Gjessing, et a!., 1976), the west coast (Hornstrom et al., 1976) and west-
central regions of Sweden (Grahn, 1976), the LaCloche Mountains of southeastern Ontario
(Beamish, 1976), and the vicinity of Sudbury, Ontario (Scheider et al., 1976), as well as for
lakes not yet affected by acidification but in regions of similar geological substrata in west-
central Norway (Gjessing et al., 1976) and the experimental lakes area of northwestern Ontario
(Armstrong and Schindler, 1971). Basic cation concentrations (Ca, Mg, Na, K) are low (e.g.,
calcium levels of 18-450 jjeq/liter or 0.4 - 9 mg/liter) relative to world-wide averages (15
mg/liter calcium, Livingstone, 1963). Bicarbonate is the predominant anion in most fresh-
waters (Stumm and Morgan, 1970). However, in acid lakes in regions affected by acidic deposi-
tion, sulfate replaces bicarbonate as the dominant anion (Wright and Gjessing, 1976; Beamish,
1976). With a decreasing pH level in acid lakes, the importance of the hydrogen ion to the
total cation content increases.
Surveys to determine the extent of effects of acidic deposition on the chemistry of lakes
have been conducted in Norway (Wright and Snekvik, 1978; Wright and Henriksen, 1978), Sweden
(Aimer et al., 1974; Dickson, 1975), Scotland (Wright et al., 1980a), the LaCloche Mountain
area of Ontario (Beamish and Harvey, 1972), the Muskoka-Haliburton area of south-central
Ontario (Dillon et al., 1978), and the Adirondack Region of New York State (Schofield,
1976a,b), Maine (Davis et al., 1978), and Pennsylvania (Arnold et al., 1980). In regions of
similar geological substrata not receiving acidic deposition, lake pH levels average 5.6-6.7
(Armstrong and Schindler, 1971). Of 155 lakes systematically surveyed in southern Norway in
October 1974, over 70 percent had pH levels below 6.0, 56 percent below 5.5, and 24 percent
below 5.0 (Wright and Henriksen, 1978). Of 700 lakes in the SeJrlandet Region of southern
Norway surveyed in 1974 to 1975 (May-November), 65 percent had pH levels below 5.0 (Wright and
Snekvik, 1978). On the west coast of Sweden, of 321 lakes investigated during 1968-1970, 93
percent had a pH level 5.5 or lower. Fifty-three percent had pH levels between 4.0 and 4.5
(Dickson, 1975). In the LaCloche Mountain Region of Ontario, 47 percent of 150 lakes sampled
in 1971 had pH levels less than 5.5, and 22 percent had pH levels below 4.5 (Beamish and
Harvey, 1972). In the Adirondacks, 52 percent of the high elevation (> 610 m) lakes had pH
values below 5.0 (Schofield, 1976a,b). In each of these studies, the pH level of an indi-
vidual lake could be related to, in most cases, the intensity of the acidic deposition and the
geologic environment of the watershed. Atmospheric contributions of sea salts are also
important in coastal regions.
Several methods have been developed to assess the degree of acidification in a lake and
relate it to inputs of hydrogen ion or sulfate. Henriksen (1979) utilized alkalinity-calcium
and pH-calcium relationships in lakes to estimate the degree of acidification experienced by a
surface water. This technique is based on the premise that when carbonic acid weathering
7-40
-------
TABLE 7-4 CHEMICAL COMPOSITION (MEAN 1 STANDARD DEVIATION) OF ACID LAKES (pH <5) IN REGIONS RECEIVING HIGHLY
ACIDIC PRECIPITATION (pH <4 5), AND OF SOFT-WATER LAKES IN AREAS NOT SUBJECT TO HIGHLY ACIDIC PRECIPITATION
(pH M 8)
Region
I LAKES IM MID ARIAS
Scandinavia
Southernmost
Norway
Westcoast
Sweden
West-central
Sweden
North America
La Cloche Mtns,
Ontario
Sudbury,
Ontario
II. LAKES IN UNAFFECTED
Scandinavia
West-central
Norway
North America
Experimental
Lakes Area,
Ontario
No of
lakes
Measured. 26
Less s «*"
Measured 12
Less s **;
Measured! 6
Less s w*
Measured- *
Less % \f
Measured 4
Less s w*.
AREAS
Measured 13
Less s w*
Measured 40
Less s w*'
Specific
conductanceTt H (pH)
27*10 18*11 (4 76)
18
72"« 43" (4 37**)
43««
47*23 22*15 (4 66)
22
38±8 20*9 (4 7}
20
120*40 36*5 (4 5)
36
1313 612 (5 2)
6
19 02-2 (5 6-6 7)
0 2-2
Na
70+40
9
330
-50
1651120
20
26*4
9
100*30
50
50*20
9
40
4
K
5*3
4
20
13
1518
12
10i3
10
40110
40
»1
3
10
10
Ca
56*35
50
-
75*10
70
150*25
150
450*180
450
1849
16
80
80
Mg
41*16
25
*M»n
-
80+40
50
75*8
65
310*120
300
16±5
7
75
65
Meq/1
HC03
11±26
11
0
0
-
-
0
0
842
8
13*8
13
60
60
HC1
71± 45
0
440
0
170*90
0
2216
0
50*20t
0
46±21
0
40
0
so«
100*33
92
ZOO
155
200*70
180
290140
2iO
8004290
800
33*8
30
60
55
N03
442
4
8
8
19*4
19
-
-
-
"
5*2
5
<1 5
<1 5
I cations
189
106
673
-
360
175
280
255
940
880
93
41
200
160
I anions
186
107
648
-
390
200
310
290
850
800
97
48
160
120
Reference
G jess ing
et al , 1976
BornstrSn
et al , 1976
Grahn
et al , 1976
Beaiii sh ,
1976
Arnstrong,
1971
Gjessmg
et a) , 1976
Armstrong,
1971
*Less s w = Concontrations after subtracting the seawater contribution according to the procedure explained by
Wright and Gjessino, (1978)
"Data for 112 lakes
treasured after past liming of the lakes
ttyS/c«! at 20"C
-------
occurs one equivalent of alkalinity (acid neutralizing capacity) is released to the aquatic
environment for every equivalent of basic cation (Ca, Mg, K, or Na) dissolved. On the other
hand, if mineral acid weathering is occurring, for example as a result of acidic deposition,
one equivalent of hydrogen ion is comsumed for every equivalent of cation solubilized. There-
fore, for a given basic cation level, there is less aqueous acid neutralizing capacity in
lakes in systems experiencing strong acid weathering than in systems experiencing carbonic
acid weathering. When comparing alkalinity p,lots from two watersheds, one experiencing strong
acid contributions and the other undergoing largely carbonic acid weathering (assuming both
watersheds have similar edaphic environments), the difference in alkalinity between the two
plots for a given calcium level (the dominant basic cation) should be indicative of the amount
of strong acid the watershed receives and the degree of acidification of the surface water.
For waters with pH levels below 5.6, alkalinity is approximately numerically equal to the
hydrogen ion concentration with its sign changed. Therefore, pH level can be substituted for
alkalinity, and pH-calcium plots developed (Figure 7-16). In the figure most of the lakes in
northern and northwestern Norway fall below an empirically shown curve, whereas lakes in
southernmost and southeastern Norway lie above this curve. Data of this type for Norway in-
dicate that significant acidification of lakes has occurred in areas receiving precipitation
with volume-weighted average concentrations of H above 20-25 pec/liter (pH 4.7-4,6) and
sulfate concentrations above 1 mg/liter (20 peq/liter) (Henriksen, 1979).
Henriksen (1979) also utilized the concentration of excess sulfate in lake water (sulfate
in excess of that of marine origin) to estimate acidification. This suggests that bicarbonate
anions lost in acidified lakes have been replaced by an equivalent amount of sulfate. Aimer
et al. (1978) plot pH levels in Swedish lakes as a function of excess sulfur load (excess
sulfur in lake water multiplied by the yearly runoff) (Figure 7-17). Based on this relation-
ship, they estimate that the most sensitive lakes in Sweden may resist a load of only about
2 2
0.3 g/m of sulfur in lake water each year. At 1 g/m of sulfur, the pH level of the lake
will probably decrease below 5.0.
Elevated metal concentrations (e.g., aluminum, zinc, manganese, and/or iron) in surface
waters are often associated with acidification (Schofield and Trojnar, 1980; Hutchinson et al.,
1978; Wright and Gjessing, 1976; Beamish, 1976). Mobility of all these metals is increased at
low pH values (Stumm and Morgan, 1970). For example, an inverse correlation between aluminum
concentration and pH level has been identified for lakes in the Adirondack Region of New York
State, southern Norway, the west coast of Sweden, and Scotland (Wright, 1980b) (Figure 7-18).
Aluminum appears to be the primary element mobilized by strong acid inputs in precipitation
and dry deposition (Cronan, 1978). . .
Aluminum is the third most abundant element by weight in the earth's crust (Foster, 1971).
In general, aluminum is extremely insoluble and retained within the edaphic environment. How-
ever, with increased hydrogen ion inputs (via acidic deposition or other sources) into the
7-42
-------
I
a
200km
6.0
6.5
7.0
7.5
\
\
<^8o°0§ O °
°** s°°°
° SIR o %o
*o ^^ y
teo o o, - *
s»».
o "aJ-* "o O
ooocfgo o
o „
250
tCa]
Figure 7-16. pH and calcium concentrations in lakes in northern and northwestern Norway
sampled as part of the regional survey of 1975, in lakes in northwestern Norway sampled in
1977 (o) and in lakes in southernmost and southeastern Norway sampled in 1974 (•).
Southern Norway receives highly acid precipitation (pH 4.24.5) and a large number of lakes
have lost their fish populations due to high acidity. Inset shows areas in which these lakes
are located. Areas south of isoline receive precipitation more acid than pH 4.6.
Source: Henriksen (1979).
7-43
-------
D
O
I
I
CURVE 2
0123
EXCESS S IN LAKE WATER, g/m2/year
Figure 7-17. The pH value and sulfur loads In lake waters with extremely sensitive surroundings
(curve 1} and with slightly less sensitive surroundings (curve 2). (Load = concentration of
"excess" sulfur multiplied by the yearly runoff.)
Source; Aimer et al. (1978).
7-44
-------
1000
j£ 100
10
L
1000
1 100
<
10
4
— III-
SOUTH NORWAY 1974
— 154 LAKES —
*
* «
-•;•-:-••'•• .-
*,.* * * *
— % '•..** ... —
• *§ « * *
.*.>l **.-. « ~ .
*
* •
I i i
I 5 6 7 8
pH
\— I ! i —
WEST COAST SWEDEN
— * 37 LAKES
*
*
» *
- :'. .* .
*
* *
« *>* « * *
* *
I * i ** * I
t 6 7 f
pH
1000
OJ
5 100
10
t
1000
5 100
10
i i
— I 1 1 -
SCOTLAND
72 LAKES
*
•
**»*.. » * _
*'•."• . -' % •
9 •
~~ * » •* * . . "~~
* .
. * . ' * %*.
"" • * »» • '""='
* «
* *
1 I i '
15678
pH
* ..1 1 1
_ * * ADIRONDACKSUSA _
• 134 LAKES
».*«.* *
•Vw ;.*
..£* .*.,. ... . . .
• * a. " * »
(••WH ** * ^BBWH
** » '* .*** "
* * » * *
* *
* *
*
I I . I
I 5 6 7 8
pH
Figure 7-18, Total dissolved Al as a function of pH level in lakes in acidified areas in Europe and
North America.
Source: Wright et al. (I980b).
7-45
-------
edaphic environment, aluminum is rapidly mobilized. Cronan and Schofield (1979) suggest that
input of strong acids may inhibit the historical trend of aluminum accumulation in the B soil
horizon. Consequently, aluminum tends to be transported through the soil profile and into
streams and lakes. Evidence from field data (Schofield and Trojnar, 1980) and laboratory
experiments (Driscoll et al., 1979; Muniz and Leivestad, 1980) suggest that these elevated
aluminum levels may be toxic to fish. Concentration of aluminum may be as important or more
important than pH levels as a factor leading to declining fish populations in acidified lakes.
Aluminum toxicity to aquatic biota other than fish has not been assessed.
Surface water chemistry, particularly in streams and rivers, may be highly variable with
time. Since many of the neutralization reactions in soils are kinetically slow, the quality
of the leachate from the edaphic system into the aquatic system varies with the retention time
of water in the soil (Johnson et al., 1969). The longer the contact period of water with
lower soil strata, the greater the neutralization of acid contribution from precipitation and
dry deposition. Therefore, during periods of heavy rainfall or snowmelt and rapid water dis-
charge, pH levels in receiving waters may be relatively depressed.
Many of the regions currently affected by acidification experience freezing temperatures
during the winter and accumulation of a snowpack. In the Adirondack Region of New York
approximately 55 percent of the annual precipitation occurs during the winter months
(Schofield, 1976b). Much of the acid load deposited in winter accumulates in the snowpack and
may be released during a relatively short time period during snowmelt in the spring. In
addition, on melting, 50 to 80 percent of the pollutant load (including hydrogen ion and
sulfate) may be released in the first 30 percent of the meltwater (Johannessen and Henriksen,
1978). As a result, melting of the snowpack and ice cover can result in a large influx of
acidic pollutants into lakes and streams (Figure 7-19) (Gjessing et al., 1976; Schofield and
Trojnar, 1980; Hultberg, 1976). The rapid flux of this meltwater through the edaphic
environment, and its interaction with only upper soil horizons, limits neutralization of the
acid content. As a result, surface waters only moderately acidic during most of the year may
experience extreme drops in pH level during the spring thaw. Basic cation concentrations (Ca,
Mg, Na, K) may also be lower during this time period (Johannessen et al., 1980). Similar but
usually less drastic pH drops in surface waters (particularly streams) may occur during
extended periods of heavy rainfall (Driscoll, 1980). These short term changes in water
chemistry may have significant impacts on aquatic biota, especially if they occur at sensitive
times in the life cycle (e.g., during spawning or early stages of development).
7.3.1.2 Effects on decomposition. The processing of dead organic matter (detritus) plays a
central role in the energetics of lake and stream ecosystems (Wetzel, 1975). The organic
matter may have been generated either internally (autochthonous) via photosynthesis within the
aquatic ecosystem or produced outside the lake or stream (allochthonous) and later exported
to the aquatic system. Detritus is an important food source for bacteria, fungi, some proto-
zoa, and other animals. These organisms, through the utilization of detritus, release energy,
7-46
-------
F
1976/77
M
Figure 7-19, pH levels in Little Moose Lake, Adirondack region of New York State, at a depth of 3
meters and at the lake outlet.
Source: Adapted from Schofield and Trojnar (1980).
7-47
-------
minerals and other compounds stored in the organic matter back into the environment. Initial
processing of coarse participate detritus is often accomplished by benthic invertebrate fauna.
Among other things, the particles are physically broken down into smaller units, increasing
their surface area. Biochemical transformations of particulate and dissolved organic matter
occur via microbial metabolism and are fundamental to the dynamics of nutrient cycl-ing and
energy flux within the aquatic ecosystem.
In general, the growth and reproduction of microorganisms is greatly affected by hydrogen
ion concentration (Rheinheimer, 1971). Many bacteria can grow only within the pH range of 4
to 9; however, the optimum for most aquatic bacteria is between pH 6.5 and 8.5. There are
more acidophilic fungi than bacteria; consequently, in acid waters and sediments the proportion
of fungi in the microflora is greater than in waters or sediments with neutral or slightly
alkaline pH levels. Most aquatic fungi require free oxygen for growth (Rheinheimer, 1971).
Numerous studies have indicated that acidification of surface waters results in a shift
in microbial species and a reduction in microbial activity and decomposition rates. It should
be noted, however, that microorganisms in general are highly adaptive. With sufficient time,
a given species may adapt to acid conditions or an acid-tolerant species may invade and colo-
nize acidified surface waters. Therefore, some caution is necessary in interpreting
short-term experiments on the effects of acidification on microbial activity and
decomposition. On the other hand, increased accumulations of dead organic matter (as a result
of decreased decomposition rates) are commonly noted in acidic lakes and streams.
Abnormal accumulations of coarse organic matter have been observed on the bottoms of six
Swedish lakes. The pH levels in these lakes in July 1973 were approximately 4.4 to 5.4. Over
the last three to four decades, pH levels appear to have decreased 1.4 to 1.7 pH units (Grahn
et a!., 1974), In both Sweden and Canada, acidified lakes have been treated with alkaline
substances to raise pH levels. One result of this treatment has been an acceleration of
organic decomposition processes and elimination of excess accumulations of detritus (Andersson
et al., 1978; Scheider et al., 1975). Litterbags containing coarse particulate detrital mat-
ter have been used to monitor decomposition rates in acidified lakes and streams. In general,
the rates of weight loss were reduced in acidic waters when compared with more neutral waters
(Leivestad et al., 1976; Traaen, 1980; Petersen, 1980). Traaen (1980) found that after 12
months of incubation dried birch leaves or aspen sticks showed a weight loss of 50-80 percent
in waters with pH levels 6 to 7 as compared to only a 30-50 percent weight loss in waters with
pH 4 to 5. Petersen (1980) likewise found reduced weight loss of leaf packs incubated in an
acidic stream when compared to leaf packs in either a stream not affected by acidification or
a stream neutralized with addition of lime. Petersen, however, found no evidence of
differences in microbial respiration between the streams. The acidic stream did show a
reduction in the invertebrate functional group that specializes in processing large particles
(shredders). Gahnstrom et al. (1980) found no significant differences in oxygen consumption
by sediments from acidified and non-acidified lakes. Rates of glucose decomposition were also
7-48
-------
studied in lake sediment-water systems adapted to pH values from 3 to 9. Glucose
transformation increased at pH levels above 6. Lime treatment of acidic Lake Hogsjon in
Sweden also increased rates of glucose processing. However in a humic lake, the maximum rate
of glucose transformation occurred at the in situ value pH 5 (Gahnstrom et a!,, 1980)
Laboratory and field experiments involving decomposition rates have fairly consistently
found decreasing microbial activity with increasing acidity. Traaen (1980) found that litter
decomposition at pH level 5.2 was only 50 percent of that at pH 7.0 and at pH 3.5, only 30
percent that at pH 7.0. In addition, increasing acidity (pH 7.0 to 3.5) led to a shift from
bacterial to fungal dominance. Incubations of profundal lake sediments at pH 4, 5, and 6
indicated a significant reduction in community respiration with increasing acidity, as well as
a possible inhibition of nitrification and a lowering of sediment redox potentials. Bick and
Drews (1973) studied the decomposition of peptone -in the laboratory. With decreasing pH,
total bacterial cell counts and numbers of species of ciliated protozoans decreased,
decomposition and nitrification were reduced, and oxidation of ammonia ceased below pH 5. At
pH 4 and lower, the number of fungi increased.
Disruption of the detrital trophic structure and the resultant interference with nutrient
and energy cycling within the aquatic ecosystem may be one of the major consequences of acidi-
fication. Investigations into the effects of acidification on decomposition apparently have
produced somewhat inconsistent results. However, many of these apparent inconsistencies arise
only from a lack of complete understanding of the mechanisms relating acidity and rates of de-
composition. It is fairly clear that in acidic lakes and streams unusually large
accumulations of detritus occur, and these accumulations are related, directly or indirectly,
to the low pH level. The processing of organic matter has been reduced. In addition, this
accumulation of organic debris plus the development of extensive mats of filamentous algae on
lake bottoms (discussed in Section 7.3.1.3) may effectively seal off the mineral sediments
from interactions with the overlying water. As a result, regeneration of nutrient supplies to
the water column is decreased both by reduced processing and mineralization of dead organic
matter and by limiting sediment-water interactions. Primary productivity within the aquatic
system may be substantially decreased as a result of this process (Section 7.3.1.3). These
ideas have been formulated into the hypothesis of "self-accelerating oligotrophication" by
Grahn et al. (1974).
7.3.1.3 Effect on primary producers and primary productivity. Organisms obtain their food
(energy) directly or indirectly from solar energy. Sunlight, carbon dioxide, and water are
used by primary producers (phytoplankton, other algae, mosses, and macrophytes) in the process
of photosynthesis to form sugars which in turn are used by the plants or stored as starch.
The stored energy may be used by the plants or pass through the food chain or web. Energy in
any food chain or web passes through several trophic levels. Each link in the food chain is
termed a trophic level. The major trophic levels are the primary producers, herbivores,
carnivores, and the decomposers. Energy in an ecosystem moves primarily along two main
pathways: the grazing food chain (primary producers-herbivores-carnivores) and the detrital
7-49
-------
food chain (Smith, 1980; Billings, 1978; Odum, 1971). Interactions between these two food
chains are, however, extensive. Green plants convert solar energy to organic matter and, as
such, are the base for both food chains. The grazing food chain involves primarily living
organic matter; the detrital food chain, dead organic matter. Any changes as a result of
acidification in the green plants and primary production within the aquatic ecosystem may
therefore have a profound effect on all other organisms in the aquatic food web. As noted in
Section 7.3.1.2, a portion of the detrital food chain is supported by dead organic matter
imported into the aquatic system from external sources.
Extensive surveys of acidic lakes in Norway and Sweden (Leivestad et a!., 1976; Aimer et
al., 1974) have observed changes in species composition and reduced diversity of phytoplankton
correlated with decreasing lake pH level (Figure 7-20). Generally at normal pH values of 6 to
8, lakes in the west coast region of Sweden contain 30 to 80 species of phytoplankton per
100-ml sample in mid-August. Lakes with pH below 5 were found to have only about a dozen
species. In some very acid lakes (pH<4), only three species were noted. The greatest changes
in species composition occurred in the pH interval 5-6. The most striking change was the dis-
appearance of many diatoms and blue-green algae. The families Chlorophyceae (green algae) and
Chrysophyceae (golden-brown algae) also had greatly reduced numbers of species in acidic lakes
(Figure 7-21). Dinoflagellates constituted the bulk of the phytoplankton biomass in the most
acidic lakes (Aimer et al., 1978). Similar phenomena were observed in a regional survey of 55
lakes in southern Norway (Leivestad et al., 1976) and in a study of nine lakes in Ontario
(Stokes, 1980). Changes in species composition and reduced diversity have also been noted in
communities of attached algae (periphyton) (Leivestad et al., 1976; Aimer et al., 1978).
Hougeotia, a green alga, often proliferates on substrates in acidic streams and lakes.
Shifts in the types and numbers of species present may or may not affect the total levels
of primary productivity and algal biomass in acidic lakes. Species favored by acidic condi-
tions may or may not have comparable photosynthetic efficiencies or desirability as a prey
item for herbivores. On the other hand, decreased availability of nutrients in acidic water
as a result of reduced rates of decomposition (Section 7.3.1.2) may decrease primary produc-
tivity regardless of algal species involved. In field surveys and experiments, relationships
between pH level and total algal biomass and/or productivity were not as consistent as the
relationship between pH and species diversity.
Kwiatkowski and Roff (1976) identified a significant linear relationship of decreasing
chlorophyll a concentrations (indicative of algal biomass) with declining pH level in six
lakes near Sudbury, Ontario, with a pH range of 4.05 to 7.15. In addition, primary productiv-
ity was reduced in the two most acid lakes (pH 4-4.6). Stokes (1980) also reports a decrease
in total phytoplankton biomass with decreasing pH level for nine lakes in the same region of
Ontario. Crisman et al. (1980) reported a linear decrease in functional chlorophyll measure-
ments with declining pH for 11 lakes in northern Florida, pH range 4.5 to 6.9. On the
other hand, Aimer et al. (1978) note that in 58 nutrient-poor lakes in the Swedish west coast
7-50
-------
80
70
tn 60
01
o
£ 50
K
ui
CO
S
D
Z
40
30
20
10
T I I 1 I I I I 1 li I I I
PHYTOPLANICTON SPECIES IN 60 LAKES
ON THE SWEDISH WEST COAST
AUGUST 1976
pH 4,1 43 4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9 6.1 6.3 6.5 6.7 6.9 7.1
NUMBER 1 104324331210331002035054123101 1
OF LAKES
Figure 7-20, Numbers of phytoplankton species in 60 lakes having different pH values on the Swedish
West Coast, August 1976.
Source: Adapted from Aimer et al. (1978).
7-51
-------
pH 4.60-5.45
pH 6.25-7.70
BIOMASS
SPECIES
D1ATOMEAE
| | CHLOROPHYCEAE
CHRYSOPHYCEAE
CYANOPHYCEAE
PYRROPHYTA
SEPTEMBER 1972
Figure 7-21. Percentage distribution of phytoplankton species and their biomasses.
September 1972, west coast of Sweden. Biomass = living weight per unit area.
Source: Adapted from Aimer et al. (1978).
7-52
-------
region, the largest mean phytoplankton biomass occurred in the most acid lakes (pH <4.5). Van
and Stokes (1978) concluded that they have no evidence that the phytoplankton biomass in
Carlyle Lake, with a summer pH level about 5.1, is below that observed in circumneutral lakes
in the same region. In a continuing whole-lake acidification project (Schindler et a!,,
1980), a lowering of the epilimnion pH level from 6.7-7.0 in 1976 to 5.7-5.9 in 1978 resulted
in no significant change in the chlorophyll concentration or primary production. Both i_n situ
and experimental acidification have resulted in large increases in periphyton populations
(Muller, 1980; Hendrey, 1976; Hall et al. , 1980). Hendrey (1976) and Muller (1980) observed
carbon uptake by periphyton incubated i_n vitro. They found that, although the total rate of
photosynthesis increased with decreasing pH level due to the larger bioraass at the lower pH,
the photosynthesis per unit biomass decreased with pH.
From the above discussion it is obvious that not only is there no clear correlation
between pH level and algal biomass or productivity, but the effects of acidification appear
inconsistent between systems. Again, these apparent inconsistencies probably reflect a lack
of knowledge about exact mechanisms relating acidification and lake metabolism, and also the
complexity of these mechanisms and interactions. Changes in the algal community biomass and
productivity probably reflect the balance between a number of potentially opposing factors:
those that tend to decrease productivity and biomass versus those that tend to increase pro-
ductivity and/or biomass when acidity increases. Factors working to decrease productivity and
biomass with declining pH levels may include: (1) a shift in pH level below that optimal for
algal growth; (2) decreased nutrient availability as a result of decreased decomposition rates
and a sealing-off of the mineral sediments from the lake water; and (3) decreased nutrient
availability as a result of changes in aquatic chemistry with acidification. For example,
despite the fact that the optional pH range for growth of label 1 aria flocculosa is between 5.0
to 5.3 (Cholonky, 1968) or higher (Kallqvist et al., 1975), this species dominated experiment-
ally acidified stream communities at pH level 4 in three out of five replicates (Hendrey et
al., 1980a). As noted in Section 7.3.1.1, aluminum concentrations increase with decreasing pH
level in acidified lakes and streams. Aluminum is also a very effective precipitator of
phosphorus, particularly in the pH range 5 to 6 (Dickson, 1978; Stumm and Morgan, 1970). In
oligotrophic lakes, phosphorus is most commonly the limiting nutrient for primary productivity
(Wetzel, 1975; Schindler, 1975). Therefore, chemical interactions between aluminum and phos-
phorus may result in a decreasing availability of phosphorus with decreasing pH level and, as
a result, decreased primary production.
Factors working to increase productivity and/or biomass with acidification of a lake or
stream may include: (1) decreased loss of algal biomass to herbivores; (2) increased lake
transparency; and (3) increased nutrient availability resulting from nutrient enrichment of
precipitation. Decreased population of invertebrates (as discussed in Section 7.3.1.4),
particularly herbivorous invertebrates, may decrease grazing pressure on algae and result
in unusual accumulations of biomass. Hendrey (1976) and Hall et al. (1980) include this
7-53
-------
mechanism as one hypothesis to explain their observation of increased biomass of periphyton at
pH level 4 despite a decreased production rate per unit biomass.
Increases in lake transparency over time have been correlated with lake acidification in
Sweden (Aimer et a!., 1978) and the Adirondack Region of New York (Schofield, 1976c). In
addition, after the second year of experimental lake acidification (pH 6.7-7.0 to 5.7-5.9) in
northwestern Ontario (Schindler et a!., 1980), lake transparency increased by 1-2 m. These
increases in transparency have not been correlated with decreases in phytoplankton biomass.
Two mechanisms have been proposed. Aluminum acts as a very efficient precipitator for humic
substances. Dickson (1978) found that humic substances are readily precipitated in the pH
range 4,0 to 5,0. Dickson (1978) and Aimer et al. (1978) suggest that increases in aluminum
levels with lake acidification (Section 7.3.1.1) have resulted in increased precipitation of
humics from the water column and therefore increased lake transparency. Aimer at al. (1978)
provide data for one lake on the west coast of Sweden. The pH level declined from above 6 to
about 4.5 between 1940 and 1975. The secchi disc reading (Depth at which 9-inch disc is
visible to the naked eye when lowered into the water) increased from about 3m to about 10m
over the same period. Organic matter in the water (as estimated by KMnO* demand) decreased
from 24 to 8 mg/liter from 1958 to 1973. Schindler et al. (1980), on the other hand, found no
change in levels of dissolved organic carbon with acidification. Instead, changes in hydro-
lysis of organic matter with declining pH level may affect the light absorbancy characteris-
tics of the molecules. Levels of particulate organic carbon, and changes with pH level, were
not reported by Schindler et al. (1980).
Acidification of precipitation (and dry deposition) has been accompanied by increases in
levels of sulfate and nitrate. Both of these are nutrients required by plants. However, as
noted above, the primary nutrient limiting primary productivity in most oligotrophic lakes is
phosphorus. Aimer et al. (1978) report that atmospheric deposition rates of phosphorus have
also increased in recent years. The world-wide extent of the correlation between acidic
deposition and increased atmospheric phosphorus loading, however, is not known. It is expect-
ed that changes in atmospheric phosphorus loading would be much more localized than changes in
acidic deposition. It is possible that in some areas increased atmospheric loading of phos-
phorus has occurred in recent years coincidentally with increased acidic deposition.
Increased phosphorus nutrient loading into lakes may then increase primary production rates.
The effect of acidification on primary productivity and algal biomass of a particular
stream or lake system depends upon the balance of the above forces. Differences in the impor-
tance of these factors between systems may account for inconsistencies in the response of
different aquatic systems to acidic deposition. Acidification does, however, result in a
definite change in the nutrient and energy flux of the aquatic system, and this change may
eventually limit the total system biomass and productivity.
Acidification of lakes has also been correlated with changes in the macrophyte community.
Documentation for these changes comes mainly from lakes in Sweden. Grahn (1976) reported that
7-54
-------
in five of six lakes studied during the last three to five decades, the macrophyte communities
dominated by Lobe!ia and Isoetes have regressed whereas communities dominated by Sphagnum
mosses have expanded.- Acidity levels in these lakes apparently have increased, as indicated
by a drop of approximately 1,3 to 1.7 pH units since the 1930's to 1940's. In acid lakes
where conditions are suitable the Sphagnum peat moss may cpver more than 50 percent of the
bottom above the 4-m depth, and may also grow at much lower depths (Aimer et al., 1978). The
Sphagnum invasion may start at lake pH levels just below 6 (Aimer et al., 1978). Similar
growths of Sphagnum occur in Norwegian lakes (Galloway, 1978). Increases in Sphagnum as a
benthic macrophyte have been documented from one lake in the Adirondack Region of, New York
(Hendrey and Vertucci, 1980).
Under acid conditions the Sphagnum moss appears to simply outgrow flowering plant aquatic
macrophytes. In laboratory tests, the growth and productivity of the rooted macrophyte
Lobelia was reduced by 75 percent at a pH of 4, compared with the control (pH 4.3-5.5). The
period of flowering was delayed by ten days at the low pH (Laake, 1976). At low pH levels
(pH<5), essentially all the available inorganic carbon is in the form of carbon dioxide or
carbonic acid (Stumm and Morgan, 1970). As a result, conditions may be more favorable for
Sphagnum, an acidophile that is not able to utilize the carbonate ion.
Besides the shift in macrophyte species, the invasion of Sphagnum into acid lakes may
have four other impacts on the aquatic ecosystem. Sphagnum has a very high ion-exchange
capacity, withdrawing basic cations such as Ca from solution and releasing H (Anschiltz and
Gessner, 1954; Aimer et al., 1978). As a result, the presence of Sphagnum may intensify the
acidification of the system and decrease the availability of basic cations for other biota.
Second, dense growths of Sphagnum form a biotype that is an unsuitable substratum for many
benthic invertebrates (Grahn, 1976). Growths of Sphagnum in acidic lakes are also often
associated with felts of white mosses (benthic filamentous algae) and accumulations of
nondecomposed organic matter. In combination, these organisms and organic matter may form a
very effective seal. Interactions between the water column and the mineral sediments, and the
potential for recycling of nutrients from the sediments back into the water body, may be
reduced (Grahn, 1976; Grahn, et al., 1974). These soft bottoms may also be colonized by other
macrophytes. In Sweden, Aimer et al. (1978) report that growths of Juncus, Sparagam'um,
ytricularia, Nuphar, and/or Nymphaea, in addition to Sphagnum, may be extensive in acidic
lakes. Thus primary production by macrophytes in lakes with suitable bottoms may be very
large. Increased lake transparency may also increase benthic macrophyte and algal primary
productivity.
7.3.1.4 Effects on invertebrates—In regional surveys conducted in southern Norway (Hendrey
and Wright, 1976), the west coast of Sweden (Aimer et al., 1978), the LaCloche Mountain Region
of Canada (Sprules, 1975), and near Sudbury, Ontario (Roff and Kwiatkowski, 1977) numbers of
species of zooplankton were strongly correlated with pH level (Figure 7-22). Changes in
community structure were most noticeable at pH levels below 5. Certain species (e.g., of the
7-55
-------
25
20
CO
UJ
o
UJ
Q.
2 15
O
m
lio
<
ui
pH INTERVAL
NUMBER OF LAKES
Figure 7-22. The number of species of crustacean zooplankton observed in 57
lakes during a synoptic survey of lakes in southern Norway.
Source: Hendryetal(1976).
7-56
-------
genera Bosmina, Cyclops, Diaptomus,' as well as rotiferans, of the genera Polyarthra,
Keratella, and Kellicottia) apparently have a high tolerance of acidic conditions and were
commonly found in the pH interval 4.4 to 7.9. Others, such as cladocerans of the Daphnia
genus, apparently are more sensitive and were only rarely found at pH <6 (Aimer et al., 1978).
Similar studies of the relationship between pH level and biomass or productivity of zoo-
plankton are not available. Proposed mechanisms for interactions between lake acidification
and zooplankton populations are therefore largely hypothetical.
The species, population size, and productivity of zooplankton are affected both by
changes in the quality and quantity of the food supply and shifts in predator populations.
Changes in zooplankton species and production in response to changes in fish populations have
been clearly demonstrated (Brooks and Dodson, 1965; Walters and Vincent, 1973; Dodson, 1974).
Elimination of fish predators often results in dominance of the zooplankton community by
large-bodied species. Absence of invertebrate predators (e.g., large-bodied carnivorous zoo-
plankton) as a result of fish predation or other reasons often results in the prevalence of
>
small-bodied species (Lynch, 1979). Surveys of acidic lake waters often have shown the
dominance of small-bodied herbivores in the zooplankton community (Hendrey et al, 1980a). Fish
also often are absent at these pH levels (Section 7.3.1.5). Different zooplankton species may
have different physiological tolerances to depressed pH levels (e.g., Potts and Fryer, 1979).
Food supplies, feeding habits, and grazing of zooplankton may also be altered with
acidification as a consequence of changes in phytoplankton species composition and/or
decreases in biomass or productivity of phytoplankton. Zooplankton also rely on bacteria and
detrital organic matter for part of their food supply. Thus an inhibition of the microbiota
or a reduction in microbial decomposition (Section 7.3.1.2) may also affect zooplankton
populations. These alternate mechanisms postulated to underlie changes in community structure
and/or production of zooplankton communities probably play an important role in zooplankton
responses to acidification.
Synoptic and intensive studies of lakes and streams also have demonstrated that numbers
of species of benthic invertebrates are reduced along a gradient of decreasing pH level
(Sutcliffe and Carrick, 1973; Leivestad et al., 1976; Conroy et al., 1976; Aimer et al., 1978;
Roff and Kwiatkowski, 1977). In 1500 freshwater localities in Norway studied from 1953-73,
snails were generally present only in lakes with pH levels above 6 (Okland, J. , 1980). Like-
wise Gammarus Lacustris, a freshwater shrimp and an important element in the diet of trout in
the Norwegian lakes where it occurs, was not found at pH levels below 6.0 (0kland, K., 1980,
J. and K. 0kland, 1980). Experimental investigations have shown that adults of this species
cannot tolerate two days of exposure to pH 5.0 (Leivestad et al., 1976). Eggs were reared at
six different pH levels (4.0 to 6.8). At a pH of 4.5, a majority of the embryos died within
24 hours. Thus, the short-term acidification which often occurs during the spring melt of
snow could eliminate this species from small lakes (Leivestad et al., 1976). Fiance (1977)
concluded that ephemeropterans (mayflies) were particularly sensitive to low pH levels and
7-57
-------
their populations were reduced in headwater streams of the Hubbard Brook watershed in New
Hampshire. In laboratory studies, Bell (1971), Bell and Nebecker (1969), and Raddum (1978)
measured the tolerance of some stream macroinvertebrates to low pH levels. Tolerance seems to
be in the order caddisflies > (more tolerant than) stoneflies > mayflies (Hendrey et a!.,
1980).
2 2
Leivestad et al. (1976) reported on decreased standing crops (numbers/m and g/m ) of
benthic invertebrates in two lakes with pH levels near 4.5 as compared to five lakes with pH
near 6.0. Chironamids were the dominant group in all lakes. No fish were found in the acid
lakes. Lack of predation by fish should favor increases in benthic biomass, the opposite of
that observed. On the other hand, Hendrey et al. (1980a), based on data from eight Ontario
lakes (pH 4.3 - < 5.7), reported no reduction in abundance of benthos related to pH level.
Air-breathing aquatic insects (e.g., backswimmers, water boatmen, and water striders)
appear to be very tolerant of acidic environments. Population densities are often greater in
acidic lakes and in the most acid lakes than in circumneutral lakes. Abundance of these large
invertebrates may be related to reduced fish predation (Hendrey et al., 1980a).
Hall et al. (1980) experimentally acidified a stream to pH 4 and monitored reactions of
macroinvertebrate populations. Initially following acidification, there was a 13-fold
increase in downstream drift of insect larvae. Organisms in the collector and scraper
functional groups were affected more than predators. Benthic samples from the acidified zone
of Morris Brook contained 75 percent fewer individuals than those for reference areas. There
was also a 37 percent reduction in insect emergence; members of the collector group were most
affected. Insects seem to be particularly sensitive at emergence (Bell, 1971). Many species
of aquatic insects emerge early in the spring through cracks in the ice and snow cover. These
early-emerging insects therefore are exposed in many cases to the extremely acidic conditions
associated with srfowmelt (Hagen and Langeland, 1973).
Low pH also appeared to prevent permanent colonization by a number of invertebrate
species, primarily herbivores, in acidified reaches of River Dudden, England (Sutcliffe and
Carrick, 1973). Ephemeroptera, trichoptera, Ancylus (Gastropoda) and Gammarus were absent in
these reaches.
Damage to invertebrate communities may influence other components of the food chain.
Observations that herbivorous invertebrates are especially reduced in acidic streams, as
reported in Norris Brook and River Dudden, support the hypothesis (Hendrey, 1976; Hall et al.,
1980a) that changes in invertebrate populations may be responsible for increased periphytic
algal accumulations in acidic streams and benthic regions of acidic lakes (Hendrey et al.,
1980). Benthic invertebrates also assist with the essential function of processing dead
organic matter. Petersen (1980) noted that decomposition of coarse particulate organic matter
in leaf packs was lower in an acidic stream than in two streams with circumneutral pH levels.
The invertebrate community also showed a reduction in the invertebrate functional group that
specializes in processing large particles (shredders). In unstressed aquatic ecosystems, a
continuous emergence of different insect species is available to predators from spring to
7-58
-------
autumn. In acid-stressed lakes or streams, the variety and numbers of prey may be reduced.
Periods may be expected to occur in which the amount of prey available to fish (or other pre-
dators) is diminished.
7.3.1.5 Effects on fish—Acidification of surface waters has had its most obvious, and per-
haps the most severe, impact on fish populations. Increasing acidity has resulted not just in
changes in species composition or decreases in biomass>"but,'' in many cases, in total elimina-
tion of populations of fish from a given lake or stream. Extensive depletion of fish stocks
has occurred in large regions of Norway, Sweden, and parts of eastern North America. Both
commercial and sport fisheries have been affected in these areas. However, precise assess-
*.
ments of losses—in terms of population extinctions, reductions in yields, or economic and
social impacts—either have not been attempted or are still in the process of evaluation.
Potential damage to fish populations inhabiting other acid-sensitive aquatic ecosystems in New
England, the Appalachians, and parts of Southeastern, North central, and Northwestern United
States have not yet been assessed (Galloway, 1978).
Declines in fish populations have been related to acidification of surface waters in the
Adirondack Region of New York State (Schofield 1976c), southern Norway (Jensen and Snekvik,
1972; Wright and Snekvik, 1978), southwestern Sweden (Aimer et a!., 1974), the LaCloche Moun-
tain Region in southern Ontario (Beamish and Harvey, 1972), and southwestern Scotland (Wright
et al., 1980a). Schofield (1976c) estimated that in 1975 fish populations in 75 percent of
Adirondack lakes at high elevation (<610 m) had been adversely affected by acidification.
Fifty-one percent of the lakes had pH values less than 5, and 90 percent of these lakes were
devoid of fish life (Figure 7-23). Comparable data for the period 1929 to 1937 indicated that
during that time only about 4 percent of these lakes had pH values below 5 and were devoid of
fish (Figure 7-24), Therefore, entire fish communities consisting of brook trout (Salvelinus
fontinalis), lake trout ( Salvelinus namaycush), white sucker (Catostomus commersoni), .brown
bullhead (Icaturus nebulosus), and several cyprinid species were apparently eliminated over a
period of 40 years. This decrease in fish populations was associated with a decline in lake
pH level. A survey of more than 2000 lakes in southern Norway, begun in 1971, found that
about one third of these lakes had lost their fish population (primarily brown trout, Sal mo
trutta L.) since the 1940's (Wright and Snekvik, 1978). Fish population status was inversely
related to lake acidity (Leivestad et al., 1976). Declines in salmon populations in southern
Norwegian rivers were reported as early as the 1920's. The catch of Atlantic salmon (Salmo
salar, L.) in seven acidified southern Norwegian rivers is now virtually zero (Figure 7-25).
In northern and western rivers not affected by acidification, no distinct downward trend in
catch has occurred (Leivestad et al., 1976; Wright et al., 1976; Jensen and Snekvik, 1972).
Similar changes have been observed in Sweden (Aimer et al., 1974) where it is estimated that
10,000 lakes have been acidified to a pH less than 6.0 and 5,000 below a pH of 5.0 (Dickson,
1975). Populations of lake trout, lake herring (Coregonus artedii), white suckers, and other
species disappeared rapidly during the 1960's from a group of remote lakes in the LaCloche
Mountain Region of Ontario (Beamish et al., 1975).
7-59
-------
pH
Figure 7-23. Frequency distribution of pH and fish population status in Adirondack Mountain
lakes greater than 610 meters elevation. Fish population status determined by survey gill netting
during the summer of 1975.
Source; Schofield (1976b).
7-60
-------
20
10
60
LU
<
E
UJ
CQ
10
I I
1975
NO FISH PRESENT
FISH PRESENT
1930s
r-n
6
pH
Figure 7-24. Frequency distribution of pH and fish pop-
ulation status in 40 Adirondack lakes greater than 610 meters
elevation, surveyed during the period 1929-1937 and again in
1975.
Source: Schofield (1976 b).
7-61
-------
STAVANGER
RIVER TOVDAL
RIVER DALALVEN
300
250
200
150
1900
1920
1940
1960
1980
en
z
o
30
20
10
1900
1920
1940
1960
1980
Figure 7-25. Norwegian salmon fishery statistics for 68 unacidified and 7 acidified
rivers.
Source: Adapted from Aimer et a!, (1978).
7-62
-------
It is difficult to determine at what pH level fish species disappear from lakes. Disap-
pearance of the fish is usually not due to massive fish kills, but is the result of a gradual
depletion of the population following reproductive failures (Leivestad et a!., 1976). Field
surveys in Scandinavia and eastern North America (Wright and Snekvik, 1978; Aimer et a!., 1974;
Schofield, 1976c) suggest that many species do not occur in lakes with pH values below 5.0.
However, large spatial and temporal fluctuations,in ,pH, and the possibility for "refuge
areas" from acidic conditions during critical periods make it extremely difficult to genera-
lize about effects of acidification on fish populations based on grab samples or annual mean
pH levels. The pH levels identified in the literature as critical for reproduction of a
species or correlated with the absence of a species in lake surveys are summarized in Table
7-5. Values range from pH 4.4 to over 6.0, and are highly species dependent.
TABLE 7-5. pH LEVELS IDENTIFIED IN FIELD SURVEYS AS CRITICAL TO
LONG-TERM SURVIVAL OF FISH POPULATIONS
Family
Species
Critical pH
Reference
Salmonidae Brook trout (Salvelinus 5.0
fontinalis)
Lake trout (Salvelinus 5.1
namaycush) 5.2-5.5
Brown trout (Salmo trutta) 5.0
Arctic char (Salvelinus alpinus) 5.2
Percidae Perch (Perca fluviatilis) 4.4-4.9
Yellow perch (Perca flavescens) 4.5-4.7
Walleye (Stigostedion yitreum) 5.5-6.0+
Catostomidae White sucker (Catostomus 4.7-5.2
commersoni) 5.1
Ictajuridae Brown bullhead (Icaturus 4.7-5.2
nebulosus) 5.0
Cyprim'dae Minnow (Phoxinus phoxinus) 5.5
Roach (Rutilus rutilus)5.5
Lake chub (Couesius plumbeus) 4. 5-4. 7
Creekchub (Semotilus atromaculatus) 5.0
Commonshiner (Notropis cornutas) 5.5
Goldenshiner (Notemigonus 4.9
crysoleucas)
Centrarchidae Smallmouth bass (Micropterus 5.5-6.0+
dolomieui)
Rock bass (Ambloplites rupestris) 4.7-5.2
Esocidae
Pike (Esox jucius)
4.4-4.9
Schofield, 1976c
Schofield, 1976c
Beamish, 1976
Aimer et al., 1978
Aimer et al., 1978
Aimer et al., 1978
Beamish, 1976
Beamish, 1976
Beamish, 1976
Schofield, 1976c
Beamish, 1976
Schofield, 1976c
Aimer et al., 1978
Aimer et al., 1978
Beamish, 1976
Schofield, 1976c
Schofield, 1976c
Schofield, 1976c
Beamish, 1976
Beamish, 1976
Aimer et al., 1978
7-63
-------
Recent field and laboratory studies (Schofield and Trojnar, 1980; Dickson, 1978; Driscoll
et al., 1979; Baker and Schofield, 1980; Muniz and Leivestad, 1980) have indicated that
aluminum levels in acidic surface waters (Section 7.3.1.1, Figure 7-18) may be highly toxic to
fish (and perhaps other biota). Schofield and Trojnar (1980) analyzed survival of brook trout
stocked into 53 Adirondack lakes as a function of 12 water quality parameters. Levels of pH,
calcium, magnesium, and aluminum were significantly different between the two groups of lakes,
with and without trout survival. However, after accounting for the effects of aluminum con-
centrations on differences between the two groups of lakes, differences in calcium, magnesium,
and pH levels were no longer significant. Aluminum, therefore, appears to be the primary
chemical factor controlling survival of trout in these lakes. Likewise, in laboratory experi-
ments with natural Adirondack waters and synthetic acidified aluminum solutions, levels of
aluminum, and not the pH level per se, determined survival and growth of fry of brook trout
and white suckers (Baker and Schofield, 1980). In addition, speciation of aluminum had a sub-
stantial effect on aluminum toxicity. Complexation of aluminum with organic chelates elimin-
ated aluminum toxicity to fry (Baker and Schofield, 1980; Driscoll et al., 1979). As a
result, waters high in organic carbon, e.g., acidic bog lakes, may be less toxic to fish than
surface waters at similar pH levels but with lower levels of dissolved organic carbon.
Inorganic aluminum levels, and not low pH levels, may therefore be a primary factor lead-
ing to declining fish populations in acidified lakes and streams. However, many laboratory or
jjn situ field experiments have been conducted on the effects of pH on fish without taking into
account aluminum or other metal concentrations in naturally acidic waters. As a result, many
of the conclusions based on these experiments regarding pH levels critical for fish survival
are suspect. Therefore these experiments will not be reviewed here.
Sensitivity of fish and other biota to low pH levels has also been shown to depend on
aqueous calcium levels (Wright and Snekvik, 1978; Trojnar, 1977a,b; 8ua and Snekvik, 1972). In
southern Norway, the mean calcium level in lakes studied was approximately 1.1 mg/liter, as
compared to about 3 mg/liter in the LaCloche Mountain Region (Table 7-4) or 2.1 mg/liter in
the Adirondack Region (Schofield, 1976b). In Norwegian lakes, Wright and Snekvik (1978) iden-
tified pH and calcium levels as the two most important chemical parameters related to fish
status.
Decreased recruitment of young fish has been cited as the primary factor leading to the
gradual extinction of fish populations (Leivestad et al,, 1976; Rosseland et al., 1980; Wright
and Snekvik, 1978). Field observations (Jensen and Snekvik, 1972; Beamish, 1974; Schofield,
1976a; Aimer et al., 1974) indicate changes in population structure over time with acidifica-
tion. Declining fish populations consist primarily of older and larger fish with a decrease
in total population density. Recruitment failure may result from inhibition of adult fish
spawning and/or increased mortality of eggs and larvae. Effects on spawning and decreased egg
deposition may be associated with disrupted spawning behavior and/or effects of acidification
on reproductive physiology in maturing adults (Lockhart and Lutz, 1976). Field observations
7-64
-------
by Beamish et al. (1975) related reproductive failure in white suckers to an inability of fe-
males to release their eggs. On the other hand, Amundsen and Lunder (1974) observed total
mortality of naturally spawned trout eggs in an acid brook a few weeks after spawning. A sum-
mary of Norwegian studies (Leivestad et al., 1976) concluded that egg and fry mortality is the
main cause of fish reproduction failure. Spawning periods and early life history stages for
many fish species coincide with periods of extreme acidity, particularly during and immedi-
ately after snowmelt in the spring.
In some lakes, fish population decreases are associated with a lack of older fish
(Rosseland et al., 1980). In Lake Tveitvatn on the Tovdal River in southern Norway, brown
trout mortality apparently occurs primarily after the first spawning. Since 1976, no fish
past spawning age have been found, and population density has decreased steadily (Rosseland et
al., 1980). Fish kills of adult salmon in rivers in southern Norway were recorded as early as
1911 (Leivestad et al., 1976).
When evaluating the potential effects of acidification on fish or other biotic popula-
tions, it is very important to keep in mind the highly diversified nature of aquatic systems
spatially, seasonally, anc! year-to-year. As a result of this diversity, it is necessary to
evaluate each system independently in assessing the reaction of the population to acidifica-
tion. Survival of a fish population may depend more on the availability of refuge areas from
acid conditions during spring melt or of one tributary predominantly fed by baseflow (fed from
the bottom) and supplying an adequate area for spawning, than on mean annual pH, calcium, or
inorganic aluminum levels.
7.3.1.6 Effects on vertebrates otherthan fish. Certain species of amphibians may be the
vertebrate animals, other than fish, most immediately and directly affected by acidic deposi-
tion (Rough and Wilson, 1976). Their vulnerability is due to their reproductive habits. In
temperate regions, most species of frogs and toads, and approximately half of the terrestrial
salamanders, lay eggs in ponds. Many of these species breed in temporary pools formed each
year by accumulation of rain and melted snow. Approximately 50 percent of the species of
toads and frogs in the United States regularly breed in ephemeral pools; about one-third of
the salamander species that have aquatic eggs and larvae and terrestrial adults breed in
temporary pools. Most of these pools are small and collect drainage from a limited area. As
a result, the acidity of the eater in these pools is strongly influenced by the pH of the pre-
cipitation that fills them. Ephemeral pools are usually more acidic than adjacent permanent
bodies of water. Rough and Wilson (1976) report that in 1975, in the vicinity of Ithaca,
N.Y., the average pH of 12 temporary ponds was 4.5 (range 3.5 to 7.0), while the average pK of
six permanent ponds was 6.1 (range 5.5 to 7.0). Amphibian eggs and larvae in temporary pools
are exposed to these acidic conditions.
Rough and Wilson (1976) and Rough (1976) studied the effect of pH level on embryonic
development of two common species of salamanders, the spotted salamander (Ambystgma maculatum)
and the Jefferson salamander (A. Jeffersonianum). In laboratory experiments, embryos of the
7-65
-------
spotted salamander tolerated pH levels from 6 to 10 but had greatest hatching success at pH 7
to 9. The Jefferson salamander tolerated pH levels 4 to 8 and was most successful at 5 to 6.
Mortality of embryos rose abruptly beyond the tolerance limits. In a four-year study of a
large breeding pond (pH 5.0-6.5), 938 adult spotted salamanders produced 486 metamorphosed
juveniles (0.52 juveniles/adult), while 686 adult Jefferson salamanders produced 2157 juvenile
(3.14 juveniles/adult). Based on these findings, Rough and Wilson (1976) predict that con-
tinued acidic deposition may result in substantial shift in salamander and other amphibian
populations. Gosner and Black (1957) report that only acid-tolerant species of amphibians can
breed in the acid (pH 3.6 to 5.2) sphagnaceous bogs in the New Jersey Pine Barrens,
Frog populations in Tranevatten, a lake near Gothenberg, Sweden, acidified by acidic pre-
cipitation, have also been investigated (Hagstrom, 1977; Hendrey, 1978). The lake has pH
levels ranging from 4.0 to 4.5. All fish have disappeared, and frogs belonging to the species
Rana temporaria and Bufg bufo are being eliminated. At the time of the study (1977) only
adult frogs eight to ten years old were found. Many egg masses of Rana temporaria were
observed in 1974, but few were found in 1977, and the few larvae (tadpoles) observed at that
time died.
Frogs and salamanders are important predators on invertebrates, such as mosquitoes and
other pest species, in pools, puddles, and lakes. They also are themselves important prey for
higher trophic levels in an ecosystem. In some habitats, salamanders are the most abundant
vertebrates. In a New Hampshire forest, for example, salamanders were found to exceed birds
and mammals in both numbers and biomass (Hanken et a!., 1980).
The elimination of fish and vegetation from lakes by acidification may have an indirect
effect on a variety of vertebrates: species of fish-eating birds (e.g., the bald eagle, loon,
and osprey), fish-eating mammals (e.g., mink and otter), and dabbling ducks which feed on
aquatic vegetation. In fact, any animal that depends on aquatic organisms (plant or animal)
for a portion of its food may be affected.
Increasing acidity in freshwater habitats results in shifts in species, populations, and
communities. Virtually all trophic levels are affected. Summaries of the changes which are
likely to occur in aquatic biota with decreasing pH are listed in Tables 7-6 and 7-7.
7.3.2 Terrestrial Ecosystems
Determining the effects of acidic precipitation on terrestrial ecosystems is not an easy
task. In aquatic ecosystems, it has been possible to measure changes in pH that occur in
acidified waters and then observe the response of organisms living in aquatic ecosystems to
the shifts in pH. In the case of terrestrial ecosystems, the situation is more complicated
since no components of terrestrial ecosystems appear to be as sensitive to a change in pH as
organisms living in poorly buffered aquatic ecosystems. In addition, indirect effects may
only be expressed after long periods of time. Nonetheless, the possibility exists that soils
and vegetation may be affected, directly or indirectly, by acidic precipitation, albeit in
complex ways.
7-66
-------
TABLE 7-6. CHANGES IN AQUATIC BIOTA LIKELY TO OCCUR WITH INCREASING ACIDITY
1. Fish populations are reduced or eliminated.
2. Bacterial decomposition is reduced and fungi may dominate saprotrophic
communities. Organic debris accumulates rapidly, tying up nutrients and
limiting nutrient mineralization and cycling.
3. Species diversity and total numbers of species of aquatic plants and
animals are reduced. Acid-tolerant species dominate.
4. Phytoplankton productivity may be reduced due to changes in nutrient
cycling and nutrient limitations.
5. Biomass and total productivity of benthic macrophytes and algae may
increase due partially to increased lake transparency.
6. Numbers and biomass of herbivorous invertebrates decline. Tolerant
invertebrate species, e.g., air-breathing bugs (water-boatmen, back-
swimmers, water striders) may become abundant primarily due to reduced
fish predation.
7. Changes in community structure occur at all trophic levels.
7-67
-------
TABLE 7-7. SUMMARY OF EFFECTS ON AQUATIC ORGANISMS ASSOCIATED WITH A RANGE IN pH
CTI
00
8.0-6.0 • Long-term changes of less than 0.5 pH units in the range 8.0 to
6.0 are likely to alter the biotic composition of freshwaters to
some degree. The significance of these slight changes, however, is
not great.
• A decrease of 0.5 to 1.0 pH units in the range 8.0 to 6.0 may cause
detectable alterations in community composition. Productivity of
competing organisms will vary. Some species will be eliminated.
Phytoplankton plentiful and well distributed but numbers of species
begin to decrease as pH decreases.
6.0-5.5 • Decreasing pH from 6.0 to 5.5 will cause a reduction in species
numbers and, among remaining species, alterations in ability
to withstand stress, and change in species dominance.
Reproduction of some salamander species is impaired.
5.5-5.0 • Below ptf 5.5, numbers and diversity of species will be reduced.
Many species will be eliminated. Crustacean zooplankton, phy-
toplankton, molluscs, amphipods, most mayfly species, and many
stone fly species will begin to drop out. In contrast, several
pH-tolerant invertebrates will become abundant, especially the
air-breathing forms (e.g., Gyrinidae, Notonectidae, Corixidae),
those with tough cuticles which prevent ion losses (i.e.,
Sialis lutaria). and some forms which live within the sediments
(Oligochaeta, Chiromomidae, and Tubificidae). Overall, inver-
tebrate biomass may be reduced.
5.0-4.5 • Below pH 5.0, decomposition of organic detritus will be severely
impaired. Organic matter accumulates rapidly. Some fungal
species increase (Hyphomycetes, basidomycetes). Many fish
species are eliminated, (see Table 7-5 for fish species
eliminated.)
Aimer et al., 1974;
Leivestad et al., 1976;
Aimer et al., 1978
Aimer et al., 1974;
Leivestad et al., 1976;
Conroy et al., 1976;
Aimer et al., 1978
Aimer et al., 1974;
Kwiatkowski and Roff, 1976;
Aimer et al., 1978
Aimer et al., 1974;
Leivestad, 1976;
Kwaitkowski and Roff, 1976;
Aimer et al., 1978
Rough and Wilson, 1976
Aimer et al., 1974;
Leivestad et al., 1976;
Hendrey et al., 1976;
Grahn et al., 1974;
Grahn, 1976;
Kwiatkowski and Roff, 1976;
Hagen and Langeland, 1973;
Henriksen and Wright, 1977
Hultberg, 1976;
Aimer et al., 1978
Leivestad et al., 1976;
Schofield, 1976b;
Aimer et al., 1978
Hall et al., 1980.
-------
TABLE 7-7 (continued)
4.5 and
below
Macrophytes, such as Lobelia, are replaced by Sphagnum moss.
Number of algal species decreases. Acid-tolerant forms remain.
Below pH 4.5, all of the above changes will be greatly exacerbated,
and all fish will be eliminated. Lower limit for many algal
species.
I
en
vo
Leivestad et al., 1976
Hendrey et al., 1976;
Grahn et al., 1974;
Aimer et al., 1978
Leivestad et al., 1976;
Hendrey et al., 1976;
Grahn et al., 1974;
Aimer et al., 1978
Aimer et al., 1974
Leivestad et al., 1976;
Schofield, 1976a,b;
Wright et al., 1976
Beamish et al., 1975;
Menedex, 1976;
Trojnar, 1977a,b;
Schofield, 1979
Source: Modified from Hendrey (1978).
-------
7.3.2.1 Effects onsoils—Acidity Is a critical factor in the behavior of natural or
agricultural soils. Soil acidity influences the availability of plant nutrients and various
microbiological processes which are necessary for the functioning of terrestrial ecosystems;
therefore, there is concern that acidic precipitation over time could have an acidifying
effect on soils through the addition of hydrogen ions. As water containing hydrogen cations
(usually from weak acids) moves through the soil, some of the hydrogen ions replace adsorbed
exchangeable cations, such as Ca , Mg , K , and Na (see Figure 7-26). The removed cations
are then carried deep into the soil pr.ofile or into the ground water. In native soils
hydrogen ions are derived from the following sources (Wiklander, 1979):
1. nutrient uptake by plants—the roots adsorb cation nutrients and desorb H ;
2. COp produced by plant roots and micro-organisms;
3. oxidation of NH4+ and S, FeS2, and H2S to HN03 and H2S04;
4. very acid litter in coniferous forests, the main acidifying source for the A and B
horizons;
5. atmospheric deposition of H^SO. and some HNO-, NO , HC1 and NH. (after nitrifica-
tion to HN03).
In addition to the acidifying factors listed above, the use of ammonium fertilizers on culti-
vated lands increases the hydrogen cations in the water solution. Ammonium fertilizers are
oxidized by bacteria to form nitrate (NO., ) and hydrogen ions (H ) (Donahue et al., 1977).
Increased leaching causes soils to become lower in basic Ca , Mg , Na , and K cations
(Donahue et al., 1977). Sensitivity to leaching is according to the following sequence: Na
» K+ > Mg2+ > Ca2+ (Wiklander, 1979).
Norton (1977) cited the potential effects of acidic deposition on soils that are listed
in Table 7-8. Of those listed, only the increased mobility of cations and their accelerated
loss has been obse'rved in field experiments. Overrein (1972) observed an increase in calcium
++ -!_.{.
leaching under simulated acid rain conditions, and increased loss by leaching of Ca , Mg ,
+3
and Al were observed by Cronan (1980) when he treated New Hampshire soils with simulated
acid rain at a pH 4.4.
Wiklander (1979) notes that in humid areas leaching leads to a gradual decrease of plant
nutrients in available and mobilizable forms. The rate of nutrient decrease is determined by
the buffering capacity of the soil and the amount and composition of precipitation (pH and
salt content). Leaching sooner or later leads to soil acidification unless the buffering
capacity of the soil is great and/or the salt concentration of precipitation is high. Soil
acidification influences the amount of exchangeable nutrients and is also likely to affect
various biological processes in the soil.
o- -
Acidic precipitation increases the amounts of SO. and N03 entering the soils. Nitrate
is easily leached from soil; however,- because it is usually deficient in the soil for both
7-70
-------
ACID RAIN
1
SOIL PARTICLES
^flLtf* _^^^^
3INu ^
*
Mg2+
K*
^^
Na*^
NH4*
sof-_
SOIL SOLUTION
Ca2*
Mg2*
11+
K*
Na*
NOJ
so?-
V -rf
CAN BE LEACHEC
Figure 7-26. Showing the exchangeable ions
of a soil with pH 7, the soil solution com-
position, and the replacement of IMa+ by H+
from acid rain.
Source: Wiklander (1979).
7-71
-------
TABLE 7-8. POTENTIAL EFFECTS OF ACID PRECIPITATION ON SOILS
Effect
Comment
Increased mobility of
most elements
Increased loss of
existing clay minerals
A change in cation
exchange capacity
A general propor-
tionate increase in
the removal of all
cations from the soil
An increased flux in
nutrients through the
ecosystem below the
root zone
Mobility changes are essentially
in the order: monovalent,
divalent, trivalent cations.
Under certain circumstances may
be compensated for by production
of clay minerals which do not have
essential (stoichiometric) alkalies
or alkali earths.
Depending on conditions, this
may be an increase or a decrease.
In initially impoverished or
unbuffered soil, the removal
may be significant on a time
scale of 10 to 100 years.
Source: Norton (1977).
plants and soil microorganisms, it is rapidly taken up and retained within the soil-plant
system (Gjessing et al., 1976; Abrahamsen et a!., 1976; Abrahamsen and Dollard, 1979). The
fate of sulfate is determined by its mobility. Retention of sulfate in soils appears to
depend on the amount of hydrous oxides of iron and aluminum present. The amounts of these
compounds present varies with the soil type. Insignificant amounts of the hydrated oxides of
iron (Fe) and aluminum (Al) are found in organic soils; therefore, sulfate retention is low
(Abrahamsen and Dollard, 1979). The presence of hydrated oxides of iron and aluminum, however,
is only one of the factors associated with the capability of a soil to retain sulfur. The
capacity of soils to adsorb and retain anions increases as the pH decreases and with the salt
concentration. Polyvalent anions of soluble salts added experimentally to soils increase
adsorption and decrease leaching of salt cations. The effectiveness of the anions studied in
- - 2 -
preventing leaching increased in the following order: Cl ~ N03 < SO, < H^PO. (Wiklander,
1980). Additions of sulfuric acid to a soil will have no effect on cation leaching unless the
sulfate anion is mobile, as cations cannot leach without associated anions (Johnson et al.
1980; Johnson, 1980; Johnson and Cole, 1980).
7-72
-------
Leaching of soil nutrients is efficiently inhibited by vegetation growing on it. Plant
roots take up the nutrients frequently in larger amounts than required by the plants. Large
amounts of these nutrients will later be deposited on the soil surface as litter or as leach-
ate from the vegetation canopy (Abrahamsen and Dollard, 1979).
In lysimeter experiments in Norway, plots with vegetation cover were used. One plot had
a dense layer of the grass, Deschampsia flexuosa (L.) Trin. and the other a less dense cover.
O—
The soil retained approximately 50 percent of the SO, added to it. The greatest amount was
retained in the lysimeters covered with grass; the relative retention increased with increas-
ing additions of sulfate (Abrahamsen and Dollard, 1979). Leaching of cations from the soil
?— ?+ ?+
was reduced by the retention of the SO, ; however, leaching of Ca and Mg increased signi-
ficantly as the acidity of the simulated rain increased. In the most acid treatment leaching
of Al was highly significant. The behavior of K , NO, , and NH, was different in the two
lysimeter series. These ions were retained in the grass-covered lysimeters whereas there was a
net leaching of K and N03 in the other series. Statistically significant effects were
obtained only when the pH of the simulated rain was 3.0 or lower (Abrahamsen and Dollard,
1979).
The Scandinavian lysimeter experiments appear to demonstrate that the relative rate of
adsorption of sulfate increases as the amounts applied are increased. In the control lysi-
meters the output/input ratio was approximately one. These results are in agreement with
results of watershed studies which frequently appear to demonstrate that, on an annual basis,
sulfate outflow is equal to or greater than the amounts being added (Gjessing et al., 1976;
Abrahamsen and Dollard, 1979). Increased outflow may be attributed to dry deposition and the
weathering of sulfur-bearing rocks. The increased deposition of sulfate via acidic precipita-
tion appears to have increased the leaching of sulfate from the soil. Together with the reten-
tion of hydrogen ions in the soil this results in an increased leaching of the nutrient
+ 2+ 2 +
cations K , Ca , Mg , and Mn (Abrahamsen and Dollard, 1979). Shriner and Henderson (1978),
however, in their study of sulfur distribution and cycling in the Walker Branch Watershed in
eastern Tennessee noted the additions of sulfate sulfur by precipitation were greater than the
amount lost in stream flow. Analysis of the biomass and soil concentrations of sulfur indi-
cated that sulfur was being retained in the mineral soil horizon. It is suggested that leach-
ing from organic soil horizons may be the mechanism by which sulfur is transferred to the
mineral horizon. Indirect evidence suggests that vegetation scavenging of atmospheric sulfate
plays an important role by adding to the amounts of sulfur entering the forest system over wet
and dry deposition.
Studies of the nutrient cycling of sulfur in a number of forest ecosystems indicate that
some ecosystems accumulate (Johnson et al., 1980; Heinrichs and Mayer, 1977; Shriner and Hender-
son, 1978) while other ecosystems maintain a balance between the additions and losses of sulfur
or show a net loss (Cole & Johnson, 1977). Sulfur accumulation appears to be associated with
sulfate adsorption in subsoil horizons. Sulfate adsorption is strongly dependent on pH.
7-73
-------
Little adsorption occurs above pH 6-7 (Harward and Reisenaur, 1966). The amount of sulfate in
a soil is a function of a soil's adsorption properties and the amount of sulfate that has been
added to the soil, integrated over time. Soil properties may favor the adsorption of sulfate;
however, the fiet annual accumulation of sulfate at any specific time will be influenced by the
degree of soil saturation (Johnson et al., 1980),
Lowered soil pH also influences the availability and toxicity of metals to plants. In
general, potentially toxic metals become more available as pH decreases. Ulrich (1975)
reported that aluminum released by acidified soils could be phytotoxic if acid rain continued
for a long period. The degree of ion leaching increased with decreases in pH, but the amount
of cations leached was far less than the amount of acid added (Malmer, 1976). Baker et al.
(1977) found that sulfur dioxide in precipitation increased the extractable acidity and alumi-
num, and decreased the exchangeable bases, especially calcium and magnesium. Although dilute
sulfuric acid in sandy podsolic soils caused a significantly decreased pH of the leached
material, the amount of acid applied (not more than twice the yearly airborne supply over
southern Scandinavia) did not acidify soil as much as did nitrate fertilizer (Tamm et al.,
1977). Highly acidic rainfall, frequently with a pH less than 3.0, in combination with heavy
metal particulate fallout from smelters, has caused soils to become toxic to seedling survival
and establishment according to observations by Hutchinson and Whitby (1976). Very low soil
pH's are associated with mobility of toxic aluminum compounds in the soils. High acidity,
High sulfur, and heavy metals in the rainfall have caused fundamental changes in the structure
of soil organic matter. The sulfate and heavy metals were borne by air from the smelters in
the Sudbury area of Ontario and brought to earth by dry and wet deposition. Among the metals
deposited in rainfall and dustfall were nickel, copper, cobalt, iron, zinc, and lead. Most of
these metals are retained in the upper layers of soil, except in very acid or sandy soils.
The accumulation of metals in soils is mainly an exchange phenomenon. Organic components
of litter, humus, and soil may bind heavy metals as stable complexes (Tyler, 1972). The heavy
metals when bound may interfere with litter decay and nutrient cycling, and in this manner
interfere with ecosystem functioning (Tyler, 1972). Acidic precipitation, by altering the
equilibria of the metal complexes through mobilization, may decrease the residence time of the
heavy metals in soil and litter (Tyler, 1972, 1977).
Biological processes in the soil necessary for plant growth can be affected by soil
acidification. Nitrogen fixation, decomposition of organic material, and mineralization,
especially of nitrogen, phosphorus and sulfur, could be affected (Abrahamsen and Do!lard,
1979; Tamm et al., 1977; Malmer, 1976; Alexander, 1980). Nearly all of the nitrogen, most of
the phosphorus and sulfur, as well as other nutrient elements in the soil are bound in organic
combination. In this form, the elements are largely or entirely unavailable for utilization by
higher plants (Alexander, 1980). It is principally through the activity of heterotrophic
microorganisms that nitrogen, phosphorus, and sulfur are made available to the autotrophic
7.74
-------
higher plants. Thus, the microbial processes that lead to the conversion of the organic forms
of these elements to the inorganic state are crucial for maintaining plant life in natural or
agricultural ecosystems. The key role of these degradative processes is illustrated by the
fact that nitrogen is limiting for food production in much of the world and governs primary
productivity in many terrestrial habitats (Alexander, 1980).
Many, and probably most, microbial transformations in soil may be brought about by
several species. Therefore, the reduction or elimination of one population is not necessarily
detrimental since a second population, not affected by the stress, may fill the partially or
totally vacated niche. On the other hand, there are a few processes that are carried out, so
far as it is now known, by only a single species, and elimination of that species could have
serious consequences. Examples of this are nitrification in which ammonium is converted to
nitrate, and nodulation of leguminous plants, for which the bacteria are reasonably specific
according to the leguminous host (Alexander, 1980).
Nitrification is one of the best indicators of pH stress because the responsible
organisms, presumably largely autotrophic bacteria, are sensitive both in culture and in
nature to increasing acidity (Dancer et al., 1973). Although nitrification will sometimes
occur at pH values below 5.0, characteristically the rate decreases with increasing acidity
and often is undetectable much below pH 4.5. Limited data suggest that the process of sulfate
reduction to sulfide in soil is markedly inhibited below a pH of 6.0. (Connell and Patrick,
1968) and studies of the presumably responsible organisms in culture attest to the inhibition
linked with the acid conditions (Alexander, 1980).
It is difficult to make generalizations concerning the effects of soil acidification on
microorganisms. Many microbial processes that are important for plant growth are clearly
suppressed as the pH declines; however, the inhibition noted in one soil at a given pH may not
be noted at the same pH in another soil (Alexander, 1980). The capacity of some
microorganisms to become acclimated to changes in pH suggests the need to study this
phenomenon using environments that have been maintained at different pH values for some time.
Typically the studies have been done with soils maintained only for short periods at the
greater acidity (Alexander, 1980). The consequences of increased acidity in the subterranean
ecosystem are totally unclear, however; the pH of soil influences not only the microbial
community at large, but also those specialized populations that colonize the root surfaces
(Alexander, 1980).
The addition of nitrate and other forms of nitrogen from the atmosphere to ecosystems
through the activity of microorganisms is an integral function of the terrestrial nitrogen
cycle. The contribution of inorganic nitrogen in wet precipitation (rain plus snow) is
usually equivalent to only a few percent of the total nitrogen assimilated annually by plants
in terrestrial ecosystems; however, total nitrogen contributions, including organic nitrogen,
in bulk precipitation (rainfall plus dry fallout) can be significant, especially in unferti-
lized natural systems.
7-75
-------
Atmospheric contributions of nitrate can range from less than-0.1 kg N/ha/yr in the North-
west (Fredricksen, 1972) to 4.9 kg N/ha/yr in the Eastern United States (Likens et al., 1970;
Henderson and Harris, 1975). Inorganic nitrogen (ammonia-N plus nitrate-N) additions in wet
precipitation ranged from less than 0.5 kg/ha/yr to more than 3.5 kg/ha/yr in Junge's (1958)
study of rainfall over the United States. On the other hand, total nitrogen loads in bulk pre-
cipitation range from less than 5 kg/ha/yr in desert regions of the West to more than 30
kg/ha/yr near barnyards in the Midwest. Total contributions of nitrogen from the atmosphere
commonly range from about 10 to 20 kg N/ha/yr for most of the United States (National Research
'Council, 1978).
In comparison, rates of annual uptake by plants in ecosystems selected from several bio-
climatic zones; (deciduous forest, tundra, desert, western coniferous forest, grassland and
tropical forest) range from 11 to 125 kg N/ha/yr. (National Research Council, 1978). Since
the lowest additions are generally associated with desert areas where rates of uptake by
plants are low, and the highest additions usually occur in moist areas where plant uptake is
high, the contributions of ammonia and nitrate from rainfall to terrestrial ecosystems are
equivalent to about 1 to 10 percent of annual plant uptake. In eastern deciduous and western
coniferous forest ecosystems, contributions from bulk precipitation, on the other hand, repre-
sent from about 8 to 25 percent of the annual plant requirements. Although these comparisons
suggest that plant growth in terrestrial ecosystems depends to a significant extent on atmo-
spheric deposition, it is not yet possible to estimate the importance of these contributions
by comparing them with the biological fixation and mineralization of nitrogen in the soil. In
nutrient-impoverished ecosystems, such as badly eroded abandoned croplands or soils subjected
to prolonged leaching by acidic precipitation, nitrogen additions from atmospheric depositions
are certainly important to biological productivity. In largely unperturbed forests, recycled
nitrogen from the soil organic pool is the chief source of nitrogen for plants, but nitrogen
to support increased production must come either from biological fixation or from atmospheric
contributions. It seems possible, therefore, that man-generated contributions could play a
significant ecological role in a relatively large portion of the forested areas near industria-
lized regions (Galloway, 1978).
Sulfur, like nitrogen, is essential for optimal plant growth. Plants usually obtain
sulfur from the soil in the form of sulfate. The amount of mineral sulfur in soils is usually
low and its release from organic matter during microbial decomposition is a major source for
plants (Donahue et al., 1977). Another major source is the wet and dry deposition of atmos-
pheric sulfur (Donahue et al., 1977; Brady, 1974; Jones, 1975).
In agricultural soils crop residues, manure, irrigation water, fertilizers, and soil
amendments are important sources of sulfur. The amounts of sulfur entering the soil system
from atmospheric sources is dependent on proximity to industrial areas, the sea coast, and
marshlands. The prevailing winds and the amount of precipitation in a given region are also
important (Halstead and Rennie, 1977). Near fossil-fueled power plants and industrial sources,
7-76
-------
the amount of sulfur In precipitation may be as much as 150 pounds per acre (168 kg/ha) or
more (Jones, 1975). By contrast, in rural areas, based on the equal distribution of sulfur
oxide emissions over the coterminous states, the amount of sulfur in precipitation is gener-
ally well below the average 15 pounds per acre (17 kg/ha). Approximately 5 to 7 pounds per
acre (7 to 8 kg/ha) per year were reported for Oregon in 1966 (Jones, 1975). Shinn and Lynn
(1979) have estimated that in the Northeastern United States, the area where precipitation is
most acidic, approximately 5 x 10 tons of sulfate per year is removed by rain. Hoeft et al.
(1972) estimated the overall average sulfur as sulfate deposition at 26 pounds of sulfur per
acre per year (30 kg S/ha/year). Estimates for rural areas were 14 pounds of sulfur per acre
per year (16 kg/ha/yr). Approximately 40 to 50 percent of the sulfur additions occurred from
November to February. Tabatabai and Laflen (1976) found that SQ*-S deposition in Iowa was
greatest in fall and winter when precipitation was low.
Experimental data have shown that even though plants are supplied with adequate soil sul-
fate they can absorb 25 to 35 percent of their sulfur from the atmosphere (Brady, 1974).
Particularly if the soil sulfur is low and atmospheric sulfur is high, most of the sulfur re-
quired by the plant can come from the atmosphere (Brady, 1974). Atmospheric sulfur would be
of benefit chiefly to plants growing on lands with a low sulfur content (Brezonik, 1976).
Tree species vary in their ability to utilize sulfur. Nitrogen and sulfur are biochem-
ically associated in plant proteins, therefore, a close relationship exists between the two in
plants. Apparently, nitrogen is only taken up at the rate at which sulfur is available. Pro-
tein formation, therefore, is limited by the amount of sulfur available (Turner and Lambert,
1980). Conifers accumulate as sulfate any sulfur beyond the amount required to balance the
available nitrogen. Protein formation proceeds at the rate at which nitrogen becomes avail-
able. Trees are not injured when sulfur is applied as sulfate rather than S0? (Turner and
Lambert, 1980).
When discussing the effects of acidic precipitation, or the effects of sulfates or
nitrates on soils, a distinction should be made between managed and unmanaged soils. There
appears to be general agreement that managed agricultural soils are less susceptible to the
influences of acidic precipitation than are unmanaged forest or rangeland soils. On managed
soils more than adequate amounts of lime are used to counteract the acidifying effects of
fertilizers in agricultural soils. Ammonium fertilizers, usually as ammonium sulfate
2-
[(NH^nSOfl] or ammonium nitrate, (NH.NCU) are oxidized by bacteria to form sulfate (SO. )
and/or nitrate (NOl) and hydrogen ions (H ) (Donahue et al. 1977; Brady, 1974). The release
of hydrogen ions into the soil causes the soil to become acidified. Hydrogen ions are also
released into the soil when plants take up mineral nutrients. Hence, substances (notably
various complexes of ammonium and sulfate ions), although of neutral pH, or nearly so, are
acidifying in their effects when they are taken up by plants or animals. Thus, the concept of
"acidifying precipitation" must be added to the concept of "acid precipitation."
7-77
-------
The acidifying effects of fertilization or acidic precipitation is countered in managed
soils through the use of lime. Liming tends to raise the pH and thereby eliminate most major
problems associated with acidic soils (Donahue et al., 1977; Likens et al., 1977). Costs of
liming all natural soils sensitive to acidification would be prohibitive as well as extremely
difficult to carry out.
Precipitation adds many chemicals to terrestrial, aquatic, and agricultural ecosystems.
In addition to sulfur and nitrogen, phosphorus and potassium are biologically most important
because they often are in limited supply in the soil (Likens et al., 1977). Other chemicals
of varying biological importance and varying concentration found in precipitation over North
America are the following: chlorine, sodium, calcium, magnesium, iron, nickel, copper, zinc,
cadmium, lead, manganese (Beamish, 1976; Hutchinson and Whitby, 1976; Brezonik, 1976), mercury
(Brezonik, 1976), and cobalt (Hutchinson and Whitby, 1976). Rain over Britain and the
Netherlands, according to Gorham (1976), contained the following elements in addition to those
reported for North American precipitation: aluminum, arsenic, beryllium, cerium, chromium,
cesium, antimony, scandium, selenium, thorium, and vanadium. Again, it is obvious that many
of these elements will be found in precipitation in highly industrialized areas and will not
be of biological importance until they enter an ecosystem where they may come into contact
with some form of life, as in the case of heavy metals in the waters and soils near Sudbury,
Ontario. Of the chemical elements found in precipitation, magnesium, iron, copper, zinc, and
manganese are essential in small amounts for the growth of plants; however, at high
concentrations these elements, as well as the other heavy metals, can be toxic to plants and
animals. Furthermore, the acidity of precipitation can affect the solubility, mobility, and
toxicity of these elements to the foliage or roots of plants and to animals or microorganisms
that may ingest oh decompose these plants.
Wiklander (1979) has pointed out that based on the ion exchange theory, ion exchange
experiments, and the leaching of soil samples, the following conclusions can be drawn about
the acidifying effect on soils through the atmospheric deposition of mineral acids:
1. At a soil pH > 6.0, acids are fully neutralized by decomposition of CaCOo and other
unstable minerals and by cation exchange.
2. At soil pH < 5.5, the efficiency of the proton to decompose minerals and to replace
2+ 2+ + +
exchangeable Ca , Mg , K , and Na decreases with the soil pH. Consequently, the
acidifying effect of mineral acids on soils decreases, but the effect on the runoff
water increases in the very acid soils.
3, Salts of Ca +, Mg , K , and NH» in the precipitation counteract the absorption of
protons and, in that way, the decrease of the base saturation. A proportion of the
acids percolate through the-soil and acidify the runoff.
The sensitivity of various soils to acidic precipitation depends on the soil buffer
capacity and on the soil pH. Noncalcareous sandy soils with pH > 5 are the most sensitive to
acidification; however, acidic soils would be most likely to release aluminum.
7-78
-------
Very acid soils are less sensitive to further acidification because they are already
adjusted by soil formation to acidity and are therefore more stable. In these soils easily
weatherable minerals have disappeared, base saturation is low, and the pH of the soil may be
less than that of precipitation. The low nutrient level is a crucial factor which limits pro-
ductivity in these soils. Even a slight decrease in nutrient status by leaching may have a
detrimental effect on plant yield (Wiklander, 1979). Fertilization appears to be the only
preventive measure.
In properly managed cultivated soils, acidic precipitation should cause only a slight in-
crease in the lime requirement, with a cost compensated for by the supply of sulfur, nitrogen,
magnesium, potassium, and calcium made available to plants (Wiklander, 1979).
7.3.2.2 Effects on vegetation—The atmosphere, as well as the soil, is a source of nutrients
for plants. Chemical elements reach the plant surface via wet and dry deposition. Nitrates
and sulfates are not the only components of precipitation falling onto the plant surface.
Other chemical elements (cadmium, lead, zinc, and manganese), at least partially soluble in
water, are deposited on the surface of vegetation and may be assimilated by it, usually
through the leaves. (See Chapter 8 for discussion of particulate matter.) An average
raindrop deposited on trees in a typical forest washes over three tiers of foliage before it
reaches the soil. The effects of acidic precipitation may be beneficial or deleterious
depending on its chemical composition, the species of plant on which it is deposited, and the
physiological condition and maturity of the plant (Galloway and Cowling, 1978). Substances
accumulated on the leaf surfaces strongly influence the chemical composition of precipitation
not only at the leaf surface, but also when it reaches the forest floor. The chemistry of
precipitation reaching the forest floor is considerably different from that collected above
the forest canopy or a ground level where the canopy has no influence (Lindberg et al. 1979).
Except for the hydrogen ion (H ) the mean concentrations of all elements (lead, manganese,
zinc and cadmium) studied in the Walker Branch Watershed in Tennessee were found by Lindberg
et al. (1979) to be present in greater amounts in the throughfall than in incident rain.
'Their study indicated that the presence of trace elements was more variable than that of the
sulfate and hydrogen ions and that throughfall appeared to be a more dilute solution of
sulfuric acid than rain with a pH ~ 4.5 not influenced by the forest canopy. The solution was
found to contain a relatively higher concentration of alkaline earth salts of sulfate and
nitrate as well as a somewhat higher concentration of trace elements (Lindberg et al. 1979).
Lee and Weber (1980) studied the effects of sulfuric acid rain on two model hardwood
forests. The experiment, conducted under controlled field conditions, consisted of the appli-
cation of simulated sulfuric acid rain (pH values of 3.0, 3.5, and 4.0), and a control rain of
pH 5.6 to the two model forest ecosystems for a duration of 3 and 1/2 years. Rainfall appli-
cations_ approximated the annual amounts of areas in which sugar maple and red alder
communities normally occur.
7-79
-------
In evaluating the results of the study, the authors conclude that a well developed forest
canopy and litter layer can increase the pH and concentration of bases (i.e., calcium and
magnesium) in rainwater. Such conditions would tend to decrease the acidification rate of
forest soils by acid rain. However, as bases are continually leached from the soil column,
these cations could eventually be lost from the ecosystem and unavailable to influence the
acidification reactions. Changes in the ionic and pH balance of forest systems may impact the
productivity of forests through acidity-induced changes in the nutrient cycling process, de-
composition, reproduction, tree growth, and structure of forest systems (Alexander, 1980).
The additions of hydrogen, sulfate and nitrate ions to soil and plant systems have both
positive and negative effects. It has generally been assumed that the free hydrogen ion con-
centration in acidic precipitation is the component that is most likely to cause direct, harm-
ful effects on vegetation (Jacobson, 198Qa). Experimental studies support this assumption;
however, to date, there are no confirmed reports of exposure to ambient acidic precipitation
causing foliar symptoms on field grown vegetation in the continental United States (Jacobson,
198Qa) or Canada (Linzon personal communication 1980).
7.3.2,2.1 Direct effects on vegetation. Hydrogen ion concentrations equivalent to that
measured in more acidic rain events (5 pH 3.0) have been observed experimentally through the
use of simulated acid rain to cause tissue injury in the form of necrotic lesions to a wide
variety of plant species under greenhouse and laboratory conditions. This visible injury has
been reported as occurring between pH 3.0 and 3.6 (Shriner, 1980). The various types of direct
effects which have been reported are shown in Table 7-9. Such effects must be interpreted with
caution because the growth and morphology of leaves on plants grown in greenhouses frequently
are atypical of field conditions (Shriner, 1980). (See Chapter 8 for discussion of the
vegetational effects of SCL).
Small necrotic lesions, the most common form of direct injury, appear to be the result of
the collection and retention of water on plant surfaces and the subsequent evaporation of
these water droplets which concentrates the solution's constituents causes a lesion to occur.
The depression formed by the lesion further enhances the collection of water. A large
percentage of the leaf area may exhibit lesions after repeated exposures to simulated acid
rain at pH concentrations of 3.1, 2,7, 2.5 and 2.3 (Evans et al. 1977a, 1977b). In leaves
injured by simulated acidic rain, collapse and distortion of epidermal cells on the upper
surface is frequently followed by injury to the palisade cells and ultimately both leaf
surfaces are affected (Evans et al., 1977b). Evans et al. (1978), using six clones of Populus
spp. hybrids, found that leaves that had just reached full expansion were more sensitive to
simulated acid rain at pH 3.4, 3.1, 2.9, and 2.7 than those which were unexpanded or fully
expanded. On two of the clones, gall formation due to abnormal cell proliferation and
enlargement occurred. Other effects attributed to simulated acid rain include the
modification of the leaf surface, e.g. epicuticular waxes, and alteration of physiological
processes such as carbon fixation and allocation.
7-80
-------
TABLE 7-9. TYPES OF DIRECT, VISIBLE INJURY REPORTED IN RESPONSE TO SIMULATED ACIDIC WET DESPOSITION
Injury Type
Species
pH Range
Reference
Remarks
I
CO
Pitting, curl
shortening, death
1-mm necrotic lesions.
premature abscission
Cuticular erosion
Chlorosis
(A) small, shallow
circular depressions:
slight chlorosis
(B) larger lesions,
chlorosis always present
palisade collapse
(C) 1-ffim necrotic lesions
general distortion
(D) 2-iwii bifacial necrosis
sue to coalescence of
smaller lesions, total
tissue collapse.
Wrinkled leaves, excessive
adventitious budding, pre-
mature abscission
Yellow birch
Kidney bean,
soybean,
loblolly pine,
E. white pine,
willow oak
Willow oak
Sunflower,
bean
Sunflower,
bean
Sunflower,
bean
Sunflower,
bean
Sunflower,
bean
Bean
2.3-4.7 Wood and Bormann, 1974
3,2
Shriner et al., 1974
3.2
Shriner, 1978a,
Lang et al., 1978
2.3-5.7 Evans et al., 1977b
2.7 Evans et al., 1977b
2.7 Evans et al., 19J7b
2.7 Evans et al., 19775
2.7 Evans et al., 1977b
l.S-3.0 Ferenbaugh, 1976
More frequent near
veins. (A) - (D)
represent sequential
stages of lesion
development, through
time, up to 72 hrs (one
6-min rain event daily
for 3 days)
-------
TABLE 7-9 (continued)
Injury Type
Species
pH Range
Reference
Remarks
00
Incipient bronzed spot
Bifacial necrotic pitting
Necrotic lesions,
premature abscission
Marginal and tip necrosis
Galls, hypertrophy,
hyperplasia
Dead leaf cells
Bean
Bean
E. white pine,
scotch pine,
spinach,
sunflower,
bean
Bean, poplar,
soybean, ash
birch, corn,
wheat
Hybrid poplar
Soybean
2.0-3.0 Hindawi et al., 1980
2,0-3.0 Hindawi et al., 1980
2.6-3.4 Jacobson and
van Leuken, 1977
Submicron Lang et al., 1978
H2S04
aerosol
2.7-3.4 Evans et al., 1978
After first few hours
After 24 h (reported
pooling of drops =
more injury)
Injury associated with
droplet location
within 24-48 h.
3.1
Irving, 1979
Source: Shriner, 1980
-------
Lee et al. (1980) studied the effects of simulated acidic precipitation on crops. Dependa
ing on the crop studied, they reported beneficial, detrimental or no effects on yield when
crops were exposed to sulfuric acid rain at pH values of 3.0, 3.5, and 4.0 and were compared
to crops exposed to a control rain of pH 5.6. The yield of tomatoes, green peppers, straw-
berries, alfalfa, orchard gra.ss, and timothy were stiaulated. Yields of radishes, carrots,
mustard greens, and broccoli were inhibited. Potatoes were ambiguously affected except at pH
3.0 where their yield, as well as that of beets, was inhibited. Visible injury of tomatoes
might have decreased their market value. In sweet corn, stem and leaf production was stimu-
lated, but no statistically significant effects on yield were observed for 15 other crops.
Results suggest that the possibility of acid rain affecting yield depends on the portion of
the plant being commercially utilized as well as the species. Plants were regularly examined
for foliar injury associated with acid rain. Of the 35 cultivars examined, the foliage of 31
was injured at pH 3.0; 28 at pH 3.5; and 5 at pH 4,0. Foliar injury was not generally related
to effects on yield. However, foliar injury of swiss chard, mustard greens, and spinach was
severe enough to adversely affect marketability. These results are from a single growing
season and therefore considered to be preliminary.
Studies indicate that wet deposition of acidic or acidifying substances may result in a
range of direct or indirect effects on vegetation. Environmental conditions before, during
and after a precipitation event affect the responses of vegetation. Nutrient status of the
soil, plant nutrient requirements, plant sensitivity, growth stage and the total loading or
deposition of critical ions (e.g. H , N0« and SO. all play a role in determining vegeta-
tional response to acidic precipitation).
Wettability of leaves appears to be an important factor in the response of plants to acid
deposition. This has been demonstrated in the work of Evans and Curry (1977), Oacobson and van
Leuken (1977), and Shriner (1978a), who variously report a threshold of between pH 3.1 and 3.5
for development of foliar lesions on beans. The cultivars of Phaseolus vulgaris L. used in
the above studies are all relatively non-waxy and therefore fairly easily wettable. By com-
parison, studies with the very waxy leaves of citrus (Heagle et al., 1978) reported a thres-
hold for visible symptoms to be near pH 2.0. Waxy leaves apparently minimize the contact time
for the acid solutions, thus accounting for the <400X increase in H ion concentration re-
quired to induce visible injury. Table 7-10 summarizes the thresholds, species sensitivity,
concentration, and time for visible injury associated with experimental studies of wet deposi-
tion of acidic substances.
Leaching of chemical elements from exposed plant surfaces is an important effect that
rain, fog, mist, and dew have on vegetation. Substances leached include a great diversity of
materials. All of the essential minerals, ami no acids, carbohydrate growth regulators, free
sugars, pectic substances, organic acids, vitamins, alkaloids, and allelopathic substances are
among the materials which have been detected in plant leachates (Tukey, 1970). Many factors
7-83
-------
TABLE 7-10, THRESHOLDS FOR VISIBLE INJURY AND GROWTH EFFECTS ASSOCIATED WITH EXPERIMENTAL
STUDIES OF WET DEPOSITION OF ACIDIC SUBSTANCES (AFTER JACOBSQN, 1980a,b)
00
-fe
Effect
Species
Threshold
Reference
Remarks
Foliar lesions, decrease
in growth
Foliar aberrations,
decrease in growth
Foliar lesions
Foliar lesions
Foliar lesions
Foliar lesions
Foliar symptons, no
reduced growth
Increased growth,
i ncreased/decreased
nutrient content"
Reduced growth
Reduced yield
Reduced growth
Reduced yield
Yellow birch
Bean
Bean, sunflower
Bean
Hybrid poplar
Sunflower
Soybean
Lettuce
Pinto bean
Pinto bean
Soybean
Soybean
pH 3.1
pH 2.5
pH 3.1
pH 3.2
pH 3.4
pH 3.4
pH 3.0
pH 3.0, 3.2
pH 3.1
pH 2.7
pH 3.1
pH 2.5
Wood and Bormann (1974)
Ferenbaugh (1976)
Evans et al. (1977a)
Shriner (1978a)
Evans et al. (1978)
Jacobson and
van Leuken (1977)
Jacobson (1980b)
Jacobson (1980b)
Jacobson (1980b)
1/m
1/m
1/rn
greenhouse
greenhouse
greenhouse
greenhouse
greenhouse
greenhouse
greenhouse
greenhouse (varied
witft S04 & N03")
greenhouse
1/m
1/m
1/m
-------
TABLE 7-10 (continued)
"•J
I
co
en
Effect
Species
Threshold Reference
Remarks
Increased yield
Foliar symptoms
Reduced growth
Reduced yield
Reduced quality
No foliar symptoms, or
effects on growth
No foliar symptoms, but
a) decreased growth, yield
b) increased yield
No effect on growth, yield
Reduced quality
Soybean
Tomato
Tomato
Tomato
Tomato
Soybean
Soybean
Soybean
Soybean
Tomato
Tomato
pH 3.1
pH 3.0 Jacobson (1980b)
pH 3.0
pH 3.0
pH 3.0
pH 3.1 Irving (1979)
pH 2.8 Jacobson (1980b)
pH 2.8
pH 2.8
pH 3.0 Jacobson (1980b)
pH 3.0
greenhouse
field
field, low ozone
field, high ozone
field, low ozone
field
field
Highest pH to elicit a negative growth response, or lowest pH to elicit a positive growth response
Shriner, 1980.
-------
influence the quantity and quality of the substances leached from foliage. They include fac-
tors associated directly with the plant as well as those associated with the environment. Not
only are there differences among species with respect to leaching, but individual differences
also exist among individual leaves of the same crop and even the same plant, depending on the
physiological age of the leaf. Young, actively growing tissues are relatively immune to leach-
ing of mineral nutrients and carbohydrates, while mature tissue which is approaching senescence
is very susceptible. The stage of plant development, temperature, and rainwater falling on
foliage and running down plant stems or tree bark influences leaching. Rainwater, which
naturally has a pH of about 5.6, washing over vegetation may become enriched with substances
o
leached from the tissues (Nihlgard, 1970).
Leaching of organic and inorganic materials from vegetation to the soil is part of the
normal functioning of terrestrial ecosystems. The nutrient flow from one component of the
ecosystem to another is an important phase of nutrient cycling (Comerford and White, 1977;
Eaton et al., 1973). Plant leachates have an effect upon soil texture, aeration, permeability,
and exchange capacity. Leachates, by influencing the number and behavior of soil micro-
organisms, affect soil-forming processes, soil fertility, and susceptibility or immunity of
plants to soil pests and plant-chemical interactions (Tukey, 1970).
It has been demonstrated under experimental conditions that precipitation of increased
acidity can increase the leaching of various cations and organic carbon from the tree canopy
(ABrahamsen et a!., 1976; Wood and Bormann, 1975). Foliar losses of potassium, magnesium, and
calcium from bean plants and maple seedlings were found to increase as the acidity of an arti-
ficial mist was increased. Below a pH of 3.0 tissue damage occurred; however, significant
increases in leaching were measured at pH 3.3 and 4.0 with no observable tissue damage (Wood
and Bormann, 1975). Hindawi et al. (1980) also noted that, as the acidity of sulfuric acid
mist increased, so did the foliar leaching of nitrogen, calcium, phosphorous, and magnesium.
Potassium concentrations were not affected, while the concentration of sulfur increased.
Abrahamsen and Dollard (1979), in experiments using Norway spruce (Picea abies L. Karst),
observed that despite increased leaching under the most acid treatment, there was no evidence
of change in the foliar cation content. Wood and Bormann (1977), using Eastern white pine
(Pinus strobus L.), also noted no significant changes in calcium, magnesium or potassium con-
tent of needles. Tukey (1970) states that increased leaching of nutrients from foliage can
accelerate nutrient uptake by plants. No injury will occur to the plants as long as roots can
absorb nutrients to replace those being leached; however, injury could occur if nutrients are
in short supply. To date, the effects, if any, of the increased leaching of substances from
vegetation by acidic precipitation remain unclear.
Some experimental evidence suggests that acidic solutions affect the chlorophyll content
of leaves and the rate of photosynthesis. Sheridan and Rosenstreter (1973) reported marked
reduction of photosynthesis in a moss exposed to increasing H ion concentrations. Sheridan
7-86
-------
and Rosenstreter (1973), Ferenbaugh (1976), and Hindawi et al. (1980) reported reduced chloro-
phyll content as a result of tissue exposure to acid solutions. In the case of Ferenbaugh
(1976), however, the significant reductions in chlorophyll in the leaves of Phaseolus vulgaris
at pH 2.0 were associated with large areas of necrosis. A significant aspect of this study
was the loss of capacity by the plant to produce carbohydrates. The rate of respiration in
these plants showed only a slight, but significant increase, while the rate of photosynthesis
at pH 2.0 increased nearly fourfold as determined by oxygen evolution. Ferenbaugh concluded
that due to a reduction in biomass accumulation and sugar and starch concentrations, photo-
phosphorylation in the treated plants was in some way being uncoupled by the acidic solutions.
Irving (1979) reported a higher chlorophyll content and an increased rate of photo-
synthesis in field-grown soybeans exposed to simulated rain at pH 3.1. She attributed the in-
creases to improved nutrition due to the sulfur and nitrogen components of the simulated acid
rain overcoming any negative effects.
Vegetation is commonly exposed to gaseous phytotoxicants such as ozone and sulfur dioxide
at the same time as acidic precipitation'. Little information is available upon which to
evaluate the potential for determining the effects of the interaction of wet-and dry-deposited
pollutants on vegetation. Preliminary studies by Shriner (1978b), Irving (1979), and Jacobson
et al. (1980) suggest that interactions may occur. Irving (1979) found that simulated acid
precipitation at pH of 3.1 tended to limit the decrease in photosynthesis observed when field
3
-grown soybeans were exposed 17 times during the growing season to 500 ug/m (0.19 ppm) of
S09. Shriner (1978b), however, reported no significant interaction between multiple exposure
3
to simulated rain at pH 4.0 and four SO,, exposures (17860 M9/m . 3 ppm peak for 1 hr.) upon
3
the growth of bush beans. Shriner (1978b) also exposed plants to 290 ug/m (0.15 ppm) ozone
(4 3-hour exposures) in between 4 weekly exposures to rainfall of pH 4.0, and observed a sig-
nificant growth reduction at the time of harvest, Jacobson et al. (1980b), using ope"n-top
exposure chambers with field-grown soybeans, compared growth and yield between three pH levels
of simulated rain (pH 2.8, 3.4, and 4.0) and two levels of ozone (<60 and <240 |jg/m , <0.03
and £0.12 ppm). Results demonstrated that ozone not only depressed both growth and yield of
soybeans with all three rain treatments, but that the depression was greatest with the most
acidic rain. Ozone concentrations equal to or greater than those used in the studies are
common in most areas of the Northeastern United States where acidic deposition is a problem
(Jacobson et al. , 1980b); therefore, the potential for possible ozone-acidic deposition inter-
actions is great.
Shriner (1978a) studied the effect of acidic precipitation on host-parasite interactions.
Simulated acid rain with a pH of 3.2 inhibited the development of bean rust and production of
telia (a stage in the rust life cycle) by the oak-leaf rust fungus Cronartiurn fusiforme. It
also inhibited reproduction of root-knot nematodes and inhibited or stimulated development of
halo blight of bean seedlings depending on the time in the disease cycle during which the
7-87
-------
simulated acid rain was applied. The effects which inhibited disease development could result
in a net benefit to plant health. Shriner (1976, 1980) also observed that root nodulation by
Rhizobium on common Beans and soybeans was inhibited by the simulated acid rain, suggesting a
potential for reduced nitrogen fixation by legumes so affected.
Plants such as mosses and lichens are particularly sensitive to changes in precipitation
chemistry because many of their nutrient requirements are obtained directly through precipita-
tion. These plant forms are typically absent from regions with high chronic S0? air pollution
and may be affected by acidic precipitation (Nieboer et al., 1976; Denison et a!., 1977;
Sheridan and Rosenstreter, 1973). Gorham (1976) and Giddings and Galloway (1976) have written
reviews concerning this problem. Most investigations on the effects of air pollution on
epiphytes have dealt with gaseous pollutants. Very few studies have considered acidic precipi-
tation. Denison et al. (1977), however, did observe that the nitrogen-fixing ability of the
epiphytic lichen Lobaria oregana was reduced when treated with simulated rainfall with a pH of
4.0 and below. Investigations concerning the effects of acidic precipitation on epiphytic
microbial populations are very few (Abrahamsen and Do!lard, 1979).
Limited fertilization could occur in the bracken fern Pteridium aquilinum under condi-
tions of acidic precipitation (pH and sulfate concentrations) that prevail in the northeastern
United States. Evans and Bozzone (1977), using buffered solutions to simulate acidic precipi-
tation, observed that flagellar movement of sperm was reduced at pH levels below 5.8, Ferti-
lization was reduced after exposure to pH's below 4.2. Sporophyte production was also reduced
by 50 percent at pH levels below 4.2 when compared to 5.8. Addition of sulfate as sulfuric
acid (86 mM) to the buffered solutions decreased fertilization at least 50 percent at all pH
values observed. In another study, Evans and Bozzone (1978) observed that both sperm motility
and fertilization in gametophytes of Pteridium aquilinum were reduced when anions of sulfate,
nitrate, and chloride were added to buffered solutions.
Sulfur and nitrogen in precipitation have been shown to play an important role in vegeta-
tional response to acidic deposition. Jacobson et al. (1980b) investigated the impact of
simulated acidic rain on the growth of lettuce at acidities of pH 5.7 and 3.2. At pH 3.2,
solutions with NO.,:SO. mass ratios of 20:1, 2:1, and 1:7.5 were compared. For those growth
parameters (root dry weight and apical leaf dry weight) that responded to the treatments, the
results of the high nitrate concentrations applied at pH 3.2 could not be distinguished from
the control treatment at pH 5.7. The effects, however, were significantly less than those
obtained from the low nitrogen, high sulfur treatment. These observations suggest that sulfur
was possibly a limiting factor in the nutrition of these plants, with the result that the
plant response to sulfur overwhelmed the hydrogen ion effect. Other studies also have cited
the beneficial effects of simulated acidic precipitation. Irving and Miller (1978) observed
that an acidic simulant had a positive effect on productivity of field-grown soybeans as
reflected by seed weight. Increased growth was attributed to a fertilizing effect from sulfur
7-88
-------
and nitrogen delaying senescence. Irving and Miller (1978), in the same study, also exposed
soybeans to SO, and acidic precipitation. No visible injury was apparent in any of the plots;
however, a histological study revealed significant increases in the number of dead mesophyll
cells in all plots when compared to the control. The proportion of dead mesophyll cells of
plants exposed to acid rain and SCL combined was more than additive when compared to the
effects of each taken singly. Wood and Bormann (1977) reported an increase in needle length
and the weight of seedlings of Eastern white pine with increasing acidity of simulated precipi-
tation where sulfuric and nitric acid were used to acidify the mist. Increased growth was
attributed to increased NO, application. Abrahamson and Dollard (1979) also presented data
suggesting positive growth responses in forest tree species resulting from nitrogen and sulfur
in simulated rain. Simulated acidic precipitation was observed to increase the growth of Scots
pine saplings in experiments conducted in Norway. Saplings in plots watered with acid rain of
pH 3.0, 2.5, and 2.0 grew more than the control plots. The application of acid rain increased
the nitrogen and sulfur content of the needles. As the acidity of the artificial rain was ad-
justed using sulfuric acid only, the increased growth was probably due to increased nitrogen
mineralization and uptake. Turner and Lambert (1980) reported evidence indicating a positive
growth response in Monterey pine from the deposition of sulfur in ambient precipitation in
Australia.
Acidifying forest soils that are already acid by acidic precipitation or air pollutants
is a slow process. Growth effects probably could not be detected for a long time. To iden-
tify the possible effects of acidification on poor pine forests, Tamm et al. (1977) conducted
experiments using 50 kg and 100 kg of sulfur per hectare as dilute sulfuric acid (0.4 percent)
applied annually with and without NPK (nitrogen, phosphorous, potassium) fertilizer. Nitrogen
was found to be the limiting factor at both experimental sites. Acidification produced no
observable influence on tree growth. Lysimeter and soil incubation experiments conducted at
the same time as the experiments described above suggest that even moderate additions of sul-
furic acid or sulfur to soil affect soil biological processes, particularly nitrogen turnover.
The soil incubation studies indicated that additions of sulfuric acid increased the amount of
mineral nitrogen but lowered the amount of nitrate.
Soil fertility may increase as a result of acidic precipitation as nitrate and sulfate
ions, common components of chemical fertilizers, are deposited; however, the advantages of
such additions are possibly short-lived as depletion of nutrient cations through accelerated
leaching could eventually retard growth (Wood, 1975). Laboratory investigations by Overrein
(1972) have demonstrated that leaching of potassium, magnesium, and calcium, all important
plant nutrients, is accelerated by increased acidity of rain. Field studies in Sweden corre-
late decreases in soil pH with increased additions of acid (Oden et al., 1972).
Major uncertainty in estimating effects of acid rain on forest productivity is the capac-
ity of forest soils to buffer against leaching by hydrogen ions. Forest canopies have been
7-89
-------
found to filter 90 percent of the hydrogen ions from rain (pH 4.0) falling on the landscape
during the growing season (Eaton et al. , 1973). As a result, solutions reaching the forest
floor are less acidic (pH 5.0). Mayer and Ullrich (1977), however, point out that their
studies suggest that for most elements the addition by precipitation (wetfall plus dryfall) to
the soil beneath the tree canopy is considerably larger than that by precipitation to the
canopy surface as measured by rain gauges on a non-forested area. The leaching of metabo-
lites, mainly from leaf surfaces, and the washing out from leaves, branches, and stems of
airborne particles and atmospheric aerosols intercepted by trees from the atmosphere, are
suggested as the main reasons for the mineral increase.
Forest ecosystems are complicated biological organizations. Acidic precipitation will
cause some components within the ecosystem to respond even though it is not possible at pre-
sent to evaluate the changes that occur. The impact of the changes on the ecosystem can only
be determined with certainty after the passage of a long period of time.
7.3.2.3 Effects on Human Health—One effect of acidification that is potentially of concern
to human health is the possible contamination by toxic metals of edible fish and of water sup-
plies. Studies in Sweden (Landner and Larsson, 1975; Turk and Peters, 1977), Canada (Tomlin-
son, 1979; Brouzes et al. , 1977), and the United States (Tomlinson, 1979) have revealed high
mercury concentrations in fish from acidified regions. Methylation of mercury to monomethyl
mercury occurs at low pH while dimethyl mercury forms.at higher pH (Fagestrb'm and Jernelb'v,
1972). Monomethyl mercury in the water passing through the gills of fish reacts with thiol
groups in the hemoglobin of the blood and is then transferred to the muscle. As methyl mer-
cury is eliminated very slowly from fish, it accumulates with age.
Tomlinson (1979) reports that in the Bell River area of Canada precipitation is the
source of mercury. Both methyl mercury and inorganic mercury were found in precipitation.
< •
The source of mercury in snow and rain was not known at the time of the study.
Zinc, manganese, and aluminum concentrations also increase as the acidity of lakes in-
creases (Schofield, 1976b). The ingestion of fish contaminated by these metals is a distinct
possibility.
Another human health aspect is the possibility that, as drinking-water reservoirs acidify
owing to acidic precipitation, the increased concentrations of metals may exceed the public-
health limits. The increased metal concentrations in drinking water are caused by increased
watershed weathering and, possibly more importantly, increased leaching of metals from house-
hold plumbing. Indeed, in New York State, water from the Hinckley Reservoir has acidified to
such an extent that "lead concentrations in water in contact with household plumbing systems
exceed the maximum levels for human use recommended by the New York State Department of
Health" (Turk and Peters, 1977). The lead and copper concentrations in pipes which have stood
over night (U) and those in which the" water was used (F) are depicted in Table 7-11.
7-90
-------
TABLE 7-11. LEAD AND COPPER CONCENTRATION AND pH OF WATER FROM PIPES
CARRYING OUTFLOW FROM HINCKLEY BASIN AND HANNS AND STEELE CREEK BASIN,
NEAR AMSTERDAM, NEW YORK
Collection site
and date
Hinckley Dam
Nov. 21, 1974
Nov. 21, 1974
Nov. 7, 1974
Nov. 7, 1974
Oct. I, 1974
Oct. 1, 1974
Aug. 15, 1974
Aug. 15, 1974
Amsterdam
Jan. 6, 1975
Jan. 6, 1975
Pipe ,
condition
U
F
U
F
U
F
U
F
U
F
Copper
(Mi/D
600
20
460
37
570
30
760
40
2900
80
Lead
(M9/D
66
2
40
6
52
5
88
2
240
3
pH
7.4
6.3
6.3
6.8
7.1
6.3
6.3
4.5
5.0
U, unf lushed, (water stands
Source: Turk and Peters (1979).
pipes all night); F, flushed.
7-81
-------
7.3.2.4 Effects ofAcidic Precipitation on Materials—Acidic precipitation can damage the
abiotic as well as the biotic components of an ecosystem. Of particular concern in this sec-
tion are the deteriorative effects of acidic precipitation on materials and cultural artifacts
of manmade ecosystems. At present, in most areas, the dominant factor in the formation of
acidic precipitation is sulfur, usually as sulfur dioxide (Likens, 1976; Cowling and Dochinger,
1978). Because of this fact, it is difficult to isolate the effect of acidic precipitation
from changes induced by sulfur pollution in general. (The effects of sulfur oxides on mate-
rials are discussed in Chapter 10.) High acidity promotes corrosion because the hydrogen ions
act as a sink for the electrons liberated during the critical corrosion process (Nriagu, 1978).
Precipitation as rain affects corrosion by forming a layer of moisture on the surface of the
material and by adding hydrogen (H+) and sulfate (S02 ) ions as corrosion stimulators. Rain
also washes out the sulfates deposited during dry deposition and thus serves a useful function
by removing the sulfate and stopping corrosion (Kucera, 1976). Rain plays a critical role in
the corrosive process because in areas where dry deposition predominates, the washing effect is
greatest, while in areas where the dry and wet deposition processes are roughly equal, the
corrosive effect is greater (Kucera, 1976). The corrosion effect, particularly of certain
metals, in areas where the pH of precipitation is very low may be greatly enhanced by that
precipitation (Kucera, 1976). In a Swedish study, the sulfur content of precipitation,
o
expressed as meq/m per year, was found to correlate closely with the corrosion rate of steel.
The metals most likely to be corroded by precipitation with a low pH are those whose corrosion
resistance may be ascribed to a protective layer of basic carbonates, sulfates, or oxides, such
as those used on zinc or copper. The decrease in pH of rainwater to 4.0 or lower may accele-
rate the dissolution of the protective coatings (Kucera, 1976).
Materials reported to be affected by acidic precipitation, in addition to steel, are:
copper materials, linseed oil, alkyd paints on wood, antirust paints on steel, limestone, sand-
stone, concrete, and both cement-lime and lime plaster (Cowling and Dochinger, 1978).
Stone is one of the oldest building materials used by man and has traditionally been con-
sidered one of the most durable because structures such as the pyramids, which have survived
since antiquity, are made of stone. What is usually forgotten is that the structures built
with stone that was not durable have long since disappeared (Sereda, 1977).
Atmospheric sulfur compounds (mainly sulfur dioxide, with subsidiary amounts of sulfur
trioxide and ammonium sulfate) react with the carbonates in limestone and dolomites, calcar-
eous sandstone and mortars to form calcium sulfate (gypsum). The results of these reactions
are blistering, scaling, and loss of surface cohesion which, in turn, induces similar effects
in neighboring materials not in themselves susceptible to direct attack (Sereda, 1977).
Sulfates have been implicated by Winkler (1966) as very important in the disintegration
of stone. The surface flaking on the Egyptian granite obelisk (Cleopatra's Needle) in Central
Park, New York is cited as an example. The deterioration occurred within two years of its
erection in 1880.
7-92
-------
A classic example of the effects of the changing chemical climate on the stability of
stone is the deterioration of the Madonna at Herten Castle, near Recklinghausen, Westphalia in
Germany. The sculpture of porous Baumberg sandstone was erected in 1702. Pictures taken of
the Madonna in 1908 shows slight-to-moderate damage during the first 206 years. The features
of the Madonna—eyes, nose, mouth and hair—are readily discernable. In pictures taken in
1969 after 267 years, no features are visible (Cowling and Dochinger, 1978).
It is not certain in what form sulfur is absorbed into stone, as a gas (SO,) forming sul-
furous and/or sulfuric acid or whether it is deposited in rain. Rain and hoarfrost both con-
tain sulfur compounds. Schaffer (1932) compared the sulfate ion in both rain and hoarfrost at
Heading!ey, Leeds, England in 1932 (Table 7-12) and showed that the content of hoarfrost was
approximately 7 times greater than rain. Wet stone surfaces unquestionably increase the con-
densation or absorption of sulfates. Stonework kept dry and shielded from rain, condensing
dew, or hoarfrost will be damaged less by S0? pollution than stone surfaces which are exposed
(Sereda, 1977).
TABLE 7-12. COMPOSITION OF RAIN AND HOARFROST AT HEADINGLEY, LEEDS
Average rain Hoarfrost
parts per million parts per million
Suspended matter
Tar
Ash
Acidity
SO, as sulfur
SO- as sulfur
Total sulphur
Chlorine
Nitrogen as NH,
Nitrogen as N-Og
Nitrogen as albuminoid
115
15
28
1.9
22
5.7
27.7
7.3
1.98
0.196
0.434
4620
158
67
102.9
148
41.0
189.0
94.6
8.57
0.0
1.618
Source: Adapted from Schaffer (1932)
7-93
-------
Acid rain may leach ions from stonework just as acidic runoff and ground water leaches
ions from soils or bedrock; however, at the present time it is not possible to attribute the
deleterious effects of atmospheric sulfur pollution to specific compounds.
Hicrobial action (an indirect result of sulfur deposition) can also contribute to the
deterioration of stone surfaces. Tiano et al. (1975) isolated large numbers (250 to 20,000
cells per gram) of sulfate-reducing bacteria from the stones of two historical buildings of
Florence, Italy. The majority of the bacteria belonged to the genus Thiobacillus. This genus
of chemosynthetic aerobic microorganisms oxidize sulfide, elemental sulfur, and thiosulphate
to sulfate to obtain energy (Andersson, 1978). Limestone buildings, and particularly the
mortar used in the construction of brick and stone buildings, are susceptible to deterioration
when conditions permit Thiobacillus to convert reduced forms of sulfur to sulfuric acid.
Sulfate in acidic precipitation as well as other sulfur compounds deposited in dry deposition
permit the formation of sulfur compounds utilizable by microorganisms. (For more information
concerning the effects of sulfur oxides on materials, please consult Chapter 10.)
7.4 ASSESSMENT OF SENSITIVE AREAS
7.4.1 Aquatic Ecosystems
Why do some lakes become acidified by acidic precipitation and others not? What deter-
mines susceptibility? Are terrestrial ecosystems likely to be susceptible; if so, which ones?
The sensitivity of lakes to acidification is determined by: (1) the acidity of both wet
deposition (precipitation) and dry deposition; (2) the hydrology of the lake; (3) the soil
system, geology, and canopy effects; (4) the surface water. Given acidic precipitation, the
soil system and associated canopy effects are most important. The hydrology of lakes includes
the sources, amounts, and pathways of water entering and leaving a lake. The capability of a
lake and its drainage basin to neutralize acidic contributions as well as the mineral content
of its surface water is largely governed by the composition of the bedrock of the watershed.
The chemical weatheripg of the watershed strongly influences the salinity (ionic composition)
and the alkalinity of the surface water of a lake (Wetzel, 1975; Wright and Gjessing, 1976;
Wright and Henriksen, 1978). The cation exchange capacity and weathering rate of the water-
shed and the alkalinity of the surface water determine the ability of the system to neutralize
the acidity of precipitation.
Lakes vulnerable to acidic precipitation have been shown to have watersheds the geo-
logical compositions of which are highly resistant to chemical weathering (Wright and
Gjessing, 1976; Galloway and Cowling, 1978; Wright and Henriksen, 1978). In addition, the
watersheds of the vulnerable lakes usually have thin, poor soils and are poorly vegetated.
The cation exchange capacity of such soils is low and, therefore, their buffering capacity is
low (Schofield, 1979; Wright and Henriksen, 1978).
Wright and Henriksen (1978) point out that the chemistry of Norwegian lakes could be
accounted for primarily on the basis of bedrock geology. They examined 155 lakes and observed
7-94
-------
that 59 of them lay in granite or felsic gneiss basins. Water in these lakes was low in most
major ions and had low electrical conductivity. The fewer the minerals in water the lower its
conductivity (Wetzel, 1975). The waters in the lakes surveyed were "among the softest waters
in the world" (Wright and Henriksen, 1978). Sedimentary rocks generally weather readily,
whereas igneous rocks are highly resistant. The Adirondacks, as pointed out by Schofield
(1976a; 1979), have granite bedrock with much of the-araa covered with a mantle of mixed
gneisses. Shallow soils predominate in the area. Thus, these areas are susceptible to
acidification.
Limestone terrains, on the other hand, are capable of buffering intense concentrations of
acids. Glacially derived sediment has been found to be more important than bedrock in assimi-
lating acidic precipitation in the Canadian Shield area (Kramer, 1976). The detailed mineral-
ology of the unconsolidated post-glacial cover is the most important parameter in assessing
the H ion assimilation of acid precipitation in non-calcareous terrain. Knowledge of the
complete surface and subsurface hydrol,ogy is required as lower horizons may be calcareous,
whereas surface deposits may be non-calcareous (Kramer 1976). Generally, however, bedrock
geology is the best predictor of the sensitivity of aquatic ecosystems to acidic precipitation
(Hendrey et al., 1980b).
Areas with aquatic ecosystems that have the potential for being sensitive to acidic pre-
cipitation are shown in Figure 7-27. In Figure 7-27, the shaded areas on the map indicate
that the bedrock is composed of igneous or metamorphic rock while in the unshaded areas it is
of calcareous or sedimentary rock. Metamorphic and igneous bedrock weathers slowly; there-
fore, lakes in areas with this type of bedrock would be expected to be dilute and of low
alkalinity [<0.5 meq HC03/liter (Galloway and Cowling, 1978)]. Galloway and Cowling verified
this hypothesis by compiling alkalinity data. The lakes having low alkalinity existed in
regions having igneous and metamorphic rock (Galloway and Cowling, 1978). Hendrey et al.
(1980b) have developed new bedrock geology maps of the eastern United States for predicting
areas which might be impacted by acidic precipitation. The new maps permit much greater
resolution for detecting sensitivity than has been previously available for the region.
Henriksen (1979) has developed a lake acidification "indicator model" using pH-calcium
and calcium-alkalinity relationships as an indicator for determining increased surface water
acidification. The indicator is based on the observation that in pristine lake environments
(e.g., Northwest Norway or the Experimental Lakes area in northwest Ontario, Canada) calcium
is accompanied by a proportional amount of bicarbonate because carbonic acid is the primary
chemical weathering agent. "The pH-calcium relationship found for such regions is thus defined
as the reference level for unacidified lakes. Acidified lakes (e.g., Southeast Norway and the
Adirondack region) will exhibit lower pH or lower alkalinity than the reference lakes, at com-
parable calcium levels, due to the replacement of bicarbonate by strong acid anions. (See
Section 7.3.1.1).
7-95
-------
Figure 7-27. Regions in North America with lakes that are sensitive to acidification
by acid precipitation by virtue of their-underlying bedrock characteristics.
Source: Galloway and Cowling (1978).
7-96
-------
A report by Hendrey et al. (1980b) compared pre-1970 data with post-1975 data. A marked
decline in both alkalinity and pH was noted in waters of North Carolina and New Hampshire sen-
sitive to acidification. In the former, pH and alkalinity have decreased in 80 percent of the
streams tested and, in the latter, pH has decreased in 90 percent of the streams tested since
1949. These areas are predicted to be sensitive by geological mapping on the basis of their
earlier alkalinity values. Detailed county by county maps of other states in the eastern
United States suggest the sensitivity of these regions to acidic precipitation.
Though bedrock geology generally is a good predictor of the susceptibility of an area to
acidification due to acidic precipitation, other factors also have an influence. Florida, for
example, is underlaid by highly calcareous and phosphate rock, suggesting that acidification
of lakes and streams is highly unlikely. Many of the soils, however, (particularly in
northern Florida) are very mature and have been highly leached of calcium carbonate; as a
result, some lakes in which groundwater inflow is minimal have become acidified (Hendrey and
Brezonik, 1980). Conversely, there are areas in Maine with granitic bedrock where lakes have
not become acidified, despite receiving precipitation with an average pH of approximately 4.3,
because the drainage basins contain lime-bearing till and marine clay (Davis et al., 1978).
Small amounts of limestone in a drainage basin exert a strong influence on water quality in
terrain which would otherwise be vulnerable to acidification. Soils in Maine in the areas
where the pH of lakes has decreased due to acidic precipitation are immature, coarse, and
shallow, are derived largely from granitic material, and commonly have a low capacity for
assimilating hydrogen ions from leachate and surface runoff in lake watersheds (Davis et al.,
1978). The occurrence of limestone outcroppings in the Adirondack Mountains of New York State
is highly correlated with lake pH levels. The occurrence of limestone apparently counteracts
any effects of acidic precipitation. Consequently, when predicting vulnerability of a partic-
ular region to acidification, a careful classification of rock mixtures should be made. Rock
formations should be classified according to their potential buffering capacity, and the type
of soil overlying the formations should be noted. Local variations in bedrock and soils are
very important in explaining variations in acidification among lakes within an area.
7.4.2 Terrestrial Ecosystems
Predicting the sensitivity of terrestrial ecosystems to acidic precipitation is much more
difficult than for aquatic ecosystems. With aquatic ecosystems, it is possible to compare
affected and unaffected ecosystems and to note where the changes have occurred. With ter-
restrial ecosystems, comparisons are difficult to make because the effects of acidic precipi-
tation have been difficult to detect. Therefore, predictions regarding the sensitivity of
terrestrial ecosystems must, as much as possible, use the data which link the two ecosystems,
i.e., data on bedrock geology. Since, in most regions of the world, bedrock is not exposed
but is covered with soil, it is the sensitivity of different types of soil which must be
assessed. Therefore, the first step is to define "sensitivity" as it is used here in relation
to soils and acidic precipitation. Sensitivity of soils to acidification alone, though it may
7.97
-------
be the most important long-term effect, is too narrow a concept. Soils influence the quality
of waters in associated streams and lakes and may be-changed in ways other than simple pH-base
saturation relationships, e.g., microbiological populations of the surface layers or accele-
rated loss of aluminum by leaching. Therefore, criteria need to be used that would relate
soil "sensitivity" to any important change brought about in the local ecosystem by acidic
precipitation (McFee, 1980).
All soils are not equally susceptible to acidification. Sensitivity to leaching and to
loss of buffering capacity varies according to the type of parent material from which a soil
is derived. Buffering capacity is greatest in soils derived from sedimentary rocks, especi-
ally those containing carbonates, and least in soils derived from hard crystalline rocks such
as granites and quartzites (Gorham, 1958). Soil buffering capacity varies widely in different
regions of the country. Unfortunately, many of the areas now receiving the most acidic preci-
pitation are also those with relatively low natural buffering capacities.
The buffering capacity of soil depends on mineralogy, texture, structure, organic matter,
pH, base saturation, salt content, and soil permeability. Above a pH of 5.5 virtually all of
the H ions, irrespective of source, are retained by ion exchange and chemical weathering.
Below pH 5.5, the retention of the H ion decreases with the soil pH in a manner determined by
the composition of the soil (Donahue et a!., 1977). With a successive drop in the soil pH
below 5.0, an increasing proportion of hydrogen ions (H ) and deposited sulfuric acid will
pass through the soil and acidify runoff water (Donahue et a!., 1977). The sensitivity of
different soils based on pH, texture, and calcite content is summarized in Table 7-13.
Soils are the most stable component of a terrestrial ecosystem. Any changes which occur
in this component would probably have far-reaching effects. McFee (1980) has listed four
TABLE 7-1-3. THE SENSITIVITY TO ACID PRECIPITATION BASED ON: BUFFERING
CAPACITY AGAINST pH-CHANGE, RETENTION OF H , AND ADVERSE EFFECTS ON SOILS
Noncalcareous
Buffering
H retention
Adverse
effects
Calcareous
soils
Very high
Maximal
None
clays
pH > 6
High
Great
Moderate
sandy soils
pH > 6
Low
Great
Considerable
Cultivated
soils
pH > 5
High
Great
None -
slight
Acid
soils
pH < 5
Moderate
Slight
Slight
Source: Wiklander (1979).
7-98
-------
parameters which are of importance in estimating the sensitivity of soils to acidic precipita-
tion. They are:
1. the total buffering or cation exchange capacity, which is provided
primarily by clay and soil organic matter.
2. the base saturation of that exchange capacity, which can be esti-
mated from the pH of the soil.
3. the management system 'imposed' on the soil;»is«.>it eultivated and
amended with fertilizers or lime or renewed by flooding or by other
additions?
4. the presence or absence of carbonates in the soil profile.
McFee (1980) mapped soils of the Eastern United States taking into account several
factors, e.g., the sensitivity to acidic precipitation, effects of cultivation (Figure 7-28).
The areas containing most of the soils potentially sensitive to acidic precipitation are in
the upper Coastal Plain and Piedmont regions of the southeast, along the Appalachian
Highlands, through the east central and northeastern areas, and in the Adirondack Mountains of
New York (McFee, 1980). The current limited state of knowledge regarding the effects of
acidic precipitation on soils makes a more definitive judgment of the location of areas with
the most sensitive soils difficult at the present time.
The capacity of soils to absorb and retain am'ons, also important in determining whether
soils will become acidified, was not discussed by McFee (1980). The capacity for anion absorp-
tion is great in soils rich in hydrated oxides of aluminum (A!) and iron (Fe). Reduced leach-
ing of salt cations is of great significance not only in helping to prevent soil acidification
but in geochemical circulation of nutrients, fertilization in agriculture and preventing water
pollution (Wiklander, 1980; Johnson et al. 1980; Johnson, 1980). (See Section 7.3.2.1.) This
parameter, as well as those listed by Me Fee (1980), should be used in determining the sensi-
tivity of soils to acidification by both wet and dry deposition.
7,5 SUMMARY
Occurrence of acidic precipitation (rain and snow) in many regions of the United States,
Canada, and Scandanavia has been implicated in the disappearance or reduction of fish, other
animals, and plant life in ponds, lakes, and streams. In addition, acidic precipitation
appears to possess the potential for impoverishing sensitive soils, degrading natural areas,
injuring forests, and damaging stone monuments and buildings.
Sulfur and nitrogen oxides, emitted through the combustion of fossil fuels, have been
implicated as the chief contributors to the acidification of precipitation. The fate of sul-
fur and nitrogen oxides, as well as other pollutants emitted into the atmosphere, depends on
their dispersion, transport, transformation, and deposition. Emissions from automobiles occur
at ground level, those from electric power generators from smoke stacks 300 meters (1000 feet)
or more in height. Transport and transformation of the sulfur and nitrogen oxides are in part
associated with the height at which they are emitted. The greater the height, the greater the
likelihood of a longer residence time in the atmosphere and a greater opportunity for the
7-99
-------
REGIONS WITH SIGNIFICANT
AREAS OF SOILS THAT ARE
[~| NON SENSITIVE
SLIGHTLY SENSITIVE
SENSITIVE
WITHIN THE EASTERN U.S.
Figure 7-28. Soils of the eastern united states sensitive to acid rainfall.
Source: McFee (1980).
7-100
-------
chemical transformation of the oxides to sulfates, nitrates or other compounds. Ozone and
other photochemical oxidants are believed to be involved in the chemical transformations.
Because of long range transport, acidic precipitation in a particular state or region can be
the result of emissions from sources in states or regions hundreds of miles away rather than
local sources. To date the complex nature of the chemical transformation processes has not
made possible the demonstration of a direct cause-and-effect relationship between emissions of
sulfur and nitrogen oxides and the acidity of precipitation.
Natural emissions of sulfur and nitrogen compounds are also involved in the formation of
acidic precipitation; however, in industrialized regions anthropogenic emissions exceed
natural emissions.
Precipitation is arbitrarily defined as being acidic if its pH is less than 5.6.
Currently the acidity of precipitation in the Northeastern United States, the region most
severely impacted, ranges from pH 3.0 to 5.0. Precipitation episodes with a pH as low as 3.0
have been reported for other regions of the United States. The pH of precipitation can vary
from event to event, from season to season and from geographical area to geographical area.
The impact of acidic precipitation on aquatic and terristrial ecosystems is not the
result of a single or several precipitation events, but the result of continued additions of
acids or acidifying substances over time. Wet deposition of acidic substances on freshwater
lakes, streams, and natural land areas is only part of the problem. Acidic substances exist
in gases and particulate matter transferred into the lakes, streams, and land areas by dry
deposition. Therefore all the observed biological effects should not be attributed to acidic
precipitation alone.
Sensitivity of a lake to acidification depends on the acidity of both wet and dry deposi-
tion, the soil system of the drainage basin, canopy effects of ground cover and the composi-
tion of the watershed bedrock.
An extremely close mutual relationship exists between the chemistries of the environment
and of living organisms. There is a continuing exchange of nutrients and of energy. The two
are closely intertwined responses. There is no action without a reaction. Ecosystems can
respond to environmental changes or perturbations only through the response of the populations
of organisms of which they are composed. Species of organisms sensitive to specific environ-
mental changes are removed. Therefore, the capacity of an ecosystem to maintain internal
stability is determined by the ability of individual organisms to adjust their physiology or
behavior to environmental change. The success with which an organism copes with environmental
changes is determined by its ability to yield reproducing offspring. The size and success of
a population depends upon the collective ability of organisms to reproduce and maintain their
numbers in a particular environment. Those organisms that adjust best contribute most to
future generations because they have the greatest number of progeny in the population.
7-101
-------
The capacity of organisms to withstand injury from weather extremes, pesticides, acidic
deposition, or polluted air follows the principle of limiting factors. According to this
principle, for each physical factor in the environment there exists for each organism a
minimum and maximum limit beyond which no members of a particular species can survive.
Either too much or too little of a factor such as heat, light, water, or minerals (even though
they are necessary for life) can jeopardize the survival of an individual and in extreme cases
a species. When one limiting factor is removed another takes its place. The range of toler-
ance of an organism may be broad for one factor, narrow for another. The tolerance limit for
each species is determined by its genetic makeup and varies from species to species for the
same reason. The range of tolerance also varies depending on the age, stage of growth or
growth form of an organism. Limiting factors are, therefore, factors which, when scarce or
overabundant, limit the growth, reproduction and/or distribution of an organism. The increas-
ing acidity of water in lakes and streams appears to be such a factor. Significant changes
that have occurred in aquatic ecosystems with increasing acidity include the following:
1. Fish populations are reduced or eliminated.
2. Bacterial decomposition is reduced and fungi may dominate saprotrophic communi-
ties. Organic debris accumulates rapidly, tying up nutrients, and limiting
nutrient mineralization and cycling.
3. Species diversity and total numbers of species of aquatic plants and animals are
reduced. Acid-tolerant species dominate.
4. Phytoplankton productivity may be reduced due to changes in nutrient cycling and
nutrient limitations.
5. Biomass and total productivity of benthic macrophytes and algae may increase due
partially to increased lake transparency.
6. Number^ and biomass of herbivorous invertebrates decline. Tolerant invertebrate
species, e.g., air-breathing bugs (water-boatmen, back-swimmers, water striders)
may become abundant primarily due to reduced fish predation.
7. Changes in community structure occur at all trophic levels.
Studies indicate that pH levels between 6.0 and 5.0 inhibit reproduction of many species
of aquatic organisms. Fish populations become seriously affected at a pH lower than 5.0.
Disappearance of fish from lakes and streams follows two general patterns. One results
from sudden short-term shifts in pH, the other arises from a long-term decrease in the pH of
the water. A major injection of acids and other soluble substances occurs when polluted snow
melts during warm periods in winter or early spring. Fish kills are a dramatic consequence of
such episodic injections.
Long-term increases in acidity interfere with reproduction and spawning, producing a
decrease in population density and a shift in size and age of the population to one consisting
7-102
-------
primarily of larger and older fish. Effects on yield often are not recognizable until the
population is close to extinction; this is particularly true for late-maturing species with
long lives. Even relatively small increases (5 to 50 percent) in mortality of fish eggs and
fry can decrease yield and bring about extinction.
Aluminum is mobilized at low pH values. Aluminum may be as important or more important
than pH levels as factors leading to declining fish populations in acidified lakes. Certain
aluminum compounds in the water upset the osmoregulatory function of the blood in fish.
Aluminum toxicity to aquatic biota other than fish has not been assessed.
An indirect effect oF acidification potentially of concern to human health is possible
heavy metal contamination of edible fish and water supplies. Studies in Canada and Sweden
reveal high mercury concentrations in fish from acidified regions. Lead and copper have been
found in plumbing systems with acidified water, and persons drinking the water could suffer
from lead poisoning.
Acidic precipitation may indirectly influence terrestrial plant productivity by altering
the supply and availability of soil nutrients. Acidification increases leaching of plant
nutrients (such as calcium, magnesium, potassium, iron, and manganese), increases the rate of
weathering of most minerals, and also makes phosphorus less available to plants. Acidifica-
tion also decreases the rate of many soil microbiological processes such as nitrogen fixation
by Rhizobium bacteria on legumes and by the free-living Azotobacter, mineralization of
nitrogen from forest litter, nitrification of ammonium compounds, and overall decay rates of
forest floor litter.
Plants usually take up sulfur in the form of sulfate from the soil; however, they can
also take up S0« from the atmosphere through their leaves and utilize it as a sulfur source
for plant nutrition. If soil sulfur is low, plants may obtain most of their required sulfur
from the atmosphere. Though small amounts of S0? may be beneficial, large amounts and high
frequency of uncontrolled applications can be detrimental in the long term.
At present, there are no documented observations or measurements of changes in natural
terrestrial ecosystems that can be directly attributed to acidic precipitation. This does not
necessarily indicate that none are occurring. The information available on vegetational
effects is an accumulation of the results of a wide variety of controlled research approaches
largely in the laboratory, using in most instances some form of "simulated" acidic rain, fre-
quently dilute sulfuric acid and/or nitric acid. The simulated "acid rains" have deposited
hydrogen (H ), sulfate (SQ/~) and nitrate (NO,) ions on vegetation and have caused necrotic
lesions in a wide variety of plants species under greenhouse and laboratory conditions. Such
results must be interpreted with caution, however, because the growth and morphology of leaves
under greenhouse conditions are often atypical of field conditions. Based on laboratory
studies, sensitivity of plants to acidic depositions seems to be associated with the wettabi-
lity of leaf surfaces. The shorter the time of contact, the lower the resulting dose, and the
less likelihood of injury.
7T103
-------
Erosion of stone monuments and buildings and corrosion of metals can result from acidic
precipitation. Because sulfur compounds are a dominant component of acidic precipitation and
are deposited during dry deposition also, the effects resulting from the two processes cannot
be distinguished. In addition, the deposition of sulfur compounds on stone surfaces provides
a medium for microbial growth that can result in deterioration.
Certain aspects of the acidic deposition issue remain subject to debate because existing
data are ambiguous or inadequate. A comprehensive evaluation of scientific evidence bearing
on these issues is being prepared as part of a forthcoming EPA critical assessment document on
acidic deposition.
7-3.04
-------
7.6 REFERENCES
Abrahamsen, G. , and G. J, Dollard. Effects of acid precipitation on forest vegetation and
soil. In: Ecological Effects of Acid Precipitation, Report of a Workshop, Electric
Power Research Institute, Gatehouse-of-Fleet, Galloway, Scotland, September 4-7, 1978.
M. J. Wood, ed. , EPRI SQA77-403, Electric Power Research Institute, Palo Alto, CA, July
1979. section 4.2. 17 pp.
Abrahamsen, G., K. Bjor, R. Horntvedt, and B. Tveite. Effects of acid precipitation on coni-
ferous forest. In: Impact of Acid Precipitation on Forest and Freshwater Ecosystems in
Norway: Summary Report on the Research Results from the Phase I (1972-1975) of the
SNSF-Project. F. H. Braekke, ed. , Research Report FR 6/76, SNSF Project, Oslo-As,
Norway, March, 1976. pp. 37-63.
Alexander, M. Effects of acidity on microorganisms and microbial processes in soil. In:
Effects of Acid Precipitation on Terrestrial Ecosystems, North Atlantic Treaty Organiza-
tion, Toronto, Ontario, Canada, May 21-27, 1978. T. C. Hutchinson and M. Havas, eds. ,
Plenum Press, New York, NY, 1980. pp. 363-374.
Aimer, B., W. Dickson, C. Ekstrom, E. Hornstrom, and U. Miller.
Swedish lakes. Ambio 3:30-36, 1974.
Effects of acidification on
Aimer, B., W. Dickson, C. Ekstrom, and E. Hornstrom. Sulfur pollution and the aquatic eco-
system. In: Sulfur in the Environment, Part II: Ecological Impacts. J. 0. Nriagu,
ed,, John Wiley & Sons, Inc., New York, NY, 1978. pp. 271-311.
Altshuller, A. P., and G. A. McBean. The LRTAP Problem in North America: a preliminary over-
view. United States-Canada Research Consultation Group on the Long-Range transport of
Air Pollutants, Research Triangle Park, NC, and Downsview, Ontario, Canada, 1979. p. 48.
Amundsen, T., and K. Lunder. Report on fishery-biological surveys in Tjagevatn and Sonstevatn
in Grandsherod, Notodden in Telemark. Fiskerikonsulenten i Ost-Norge, Direktoratet for
vitt og ferskvannsfisk, Oslo, 1974.
Anderson, J. W. Oxidation of sulphur by sulphur by chemautotropic bacteria.
Biology. University Park Press, Baltimore, MD, 1978. pp. 17-19.
In: Sulphur in
Andersson, I., 0.0Grahn, H. Hultberg, and L. Landner. Jamforande undersokning av olika
tekniker fSr aterstallande av forsurade sjoar. [Research for developing different tech
niques for removing acid from lakes.] STU Report 73-3651. Stockholm: Institute for
Water and Air Research, 1975. Cited in: National Research Council, 1978b. pp. 63-75.
Anschiitz, J., and F. Gessner. Der lonehaustausch bei torfmoosen (Sphagnum). [Ion exchange by
peat moss.] Flora 141:178-236, 1954. Cited uj: Aimer et a!., 1978.
Armstrong, F. A. J., and D. W. Schindler. Preliminary chemical characterization of waters in
the experimental lakes area, Northwestern Ontario. J. Fish. Res. Board Can. 28: 171-187,
1971.
Arnold, D. E., R. W. Light, and V. J. Dymond. Probable Effects of Acid Precipitation on
Pennsylvania Waters. EPA 600/3-80-012, U.S. Environmental Protection Agency, Corvallis,
OR, January 1980.
Baker, J. P., and C. L. Schofield. Aluminum toxicity to fish as related to acid precipitation
and Adirondack surface water quality. In; Ecological Impact of Acid Precipitation,
an International Conference, SNSF Project, Sandefjord, Norway, March
Drablrfs and A. Tollan eds. SNSF Project, Oslo-As, Norway, October 1980.
Proceedings of
11-14, 1980. D.
pp. 292-293.
7-105
-------
Baker, J., D. Hocking, H. Nyborg. Acidity of open and intercepted precipitation in forests
and effects on forest soils in Alberta, Canada. In: First International Symposium on
Acid Precipitation and the Forest Ecosystem, U.S. Department of. Agriculture and Others,
Columbus, Ohio, May 12-15, 1975. Water Air Soil Pollut. 7:449-460, 1977.
Beamish, R. J. Growth and survival of white suckers (Catostomus commersoni) in an acidified
lake. J. Fish Res. Board Can. 31:49-54, 1974.
Beamish, R. J. Acidification of lakes in Canada by acid precipitation and the resulting
effects on fishes. In; First International Symposium on Acid Precipitation and the
Forest Ecosystem, U.S. Department of Agriculture and Others, Columbus, Ohio, May 12-15,
1975, Water Air Soil Pollut. 6:501-514, 1976.
Beamish, R. J., and H. H. Harvey. Acidification of the La Cloche Mountain Lakes, Ontario, and
resulting fish mortalities. J. Fish. Res. Board Can. 29:1131-1143, 1972.
Beamish, R. J., W. L. Lockhart, J. C. Van Loon, and H. H. Harvey. Long-term acidification of
a lake and resulting effects on fishes. Ambio 4:98-102, 1975.
Bell, H. L. Effect of low pH on the survival and emergence of aquatic insects. Water Res.
5:13-319, 1971.
Bell, H. L., and A. V. Nebecker. Preliminary studies on the tolerance of aquatic insects to
low pH. J. Kansas Entomological Soc. 42:230-236, 1969.
Bick, H., and E. F. Drews. Autopurification and ciliated protozoa in an acid environment
(model study.) Hydrobiologia 42:393-402, 1973.
Billings, W. D. Plants and the Ecosystem. 3rd Edition. Wadsworth Publishing Company, Inc.
Belmont, CA., 1978. pp. 1-62.
Bolin, B., L. Granat, L. Ingelstam, M. Johannesson, E. Mattsson, S. Oden, H. Rodhe, and C. 0.
Tanim. Air Pollution Across National Boundaries: The Impact on the Environment of Sulfur
in Air and Precipitation. Sweden's Case Study for the United Nations Conference on the
Human Environment, P. A. Norstedt and Sons, Stockholm, Sweden, 1972.
Boughey, A. S. Fundamental Ecology. Intext Educational Publishers, Scranton, PA, 1971. pp.
11-50.
Brady, N. C. The Nature and Property of Soils. 8th Edition. McMillan Publishing Co., Inc.,
New York, NY, 1974. pp. 462-472.
Braekke, F. H., Ed. Impact of Acid Precipitation on Forest and Freshwater Ecosystems in
Norway: Summary Report on the Research Results From the Phase I (1972-1975) of the
SNSF-Project. Research Report 6/76, SNSF Project, Oslo-fis, Norway, March, 1976.
Brezonik, P. L. Nutrients and other biologically active substances in atmospheric precipita-
tion. In: Atmospheric Contribution to the Chemistry of Lake Waters, First Specialty
Conference of the International Association for Great Lakes Research, Longford Mills,
Ontario, Canada, September 28 - October 1, 1975. J. Great Lakes .Res-. 2 (suppl. 1):
166-186, 1976.
Brock, T. D. Lower pH limit for the existence of blue-green algae: evolutionary and
ecological implications. Science (Washington, D.C.) 179:480-483, 1973.
Brooks, J. L., and S. I. Dodson. Predation, body size and composition of plankton. Science
150:28-35, 1965.
7-106
-------
Brosset, C. Air-borne acid. Ambio 2:2-9, 1973.
Brouzes, R. J. P., R. A. N. McLean, and G. H. Tomlinson. Mercury - the link between pH of
natural waters and the mercury content of fish. Presented at a meeting of the
Panel on Mercury of the Coordinating Committee for Scientific and Technical Assessments
of Environmental Pollutants, National Academy of Sciences, National Research Council,
Washington, DC, May 3, 1977. Domtar Research Center, Montreal, .Quebec, Canada, 1977.
Bua, B., and E. Snekvik. Hatching experiments with roe of salmonid fish 1966-1971. Effects
of acidity and salt content of hatchery water. Vann 7:86-93, 1972. Cited in: Wright and
Snekvik, 1978.
Cadle, R. 0. The photo-oxidation of hydrogen sulfide and dimethyl sulfide in air. Atmos.
Environ. 10:417, 1976.
Cadle, R. D. and M. Ledford. Reaction of ozone with hydrogen sulfide. Air Water
Pollut. 10:25-30, 1966.
Chamberlain, A. C. Dry deposition of sulfur dioxide. In: Atmospheric Sulfur Deposition:
Environmental Impact and Health Effects, Proceedings of the Second Life Sciences
Symposium, Oak Ridge National Laboratory and Others, Gatlinburg, Tennessee, October 14-18,
1979. D. E. Shriner, C. R. Richmond, and S. E. Lindberg, eds., Ann Arbor Science Pub-
lishers, Inc., Ann Arbor, MI, 1980. pp. 185-197.
Cholonky, B. J. Die Okologie der diatomeen in Binnengewasser. [Diatom ecology in inland
waters.] Cramer, Weinheim, 1968. Cited in: Hendrey et a!., 1980a.
Clapham, W. B., Jr. Natural Ecosystems. The Macmillan Co., New York, NY, 1973.
Cogbill, C. V. The history and character of acid precipitation in eastern North America.
In: First International Symposium on Acid Precipitation and the Forest Ecosystem,
U.S. Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air
Soil Pollut. 6:407-413, 1976.
Cogbill, C. V. The effect of acid precipitation on tree growth in eastern North America.
In: First International Symposium on Precipitation and the Forest Ecosystem, U.S.
Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil
Pollut. 8:89-93, 1977.
Cogbill, C. V., and G. E. Likens. Acid precipitation in the northeastern United States.
Water Resour. Res. 10:1133-1137, 1974.
Cole, D. W. , and D. W. Johnson. Atmospheric sulfate additions and cation leaching in a Doug-
las fir ecosystem. Water Resour. Res. 13:313-317, 1977.
Comerford, N. B, , and E. H. , White. Nutrient content of throughfall in paper birch and red
pine stands in northern Minnesota. Can. J. For. Res. 7:556-561, 1977.
Connell, W. E., and W. H. Patrick, Jr. Sulfate reduction in soil: effects of redox potential
and pH. Science (Washington, D.C.) 159:86-87, 1968.
Conroy, N. , K, Hawley, W. Keller, and C, Lafrance. Influences of the atmosphere on lakes in
the Sudbury area. In: Atmospheric Contribution to the Chemistry of Lake Waters, First
Specialty Conference of the International Association for Great Lakes Research, Langford
Mills, Ontario, Canada, September 28 - October 1, 1975. J. Great Lakes Res. 2
(Suppl. 1):146-165, 1976.
7-107
-------
Cowling, E. B., and L. S. Dochinger. The changing chemistry of precipitation and its effects
on vegetation and materials. In: Control and Dispersion of Air Pollutants: Emphasis on
NOV and Participate Emissions. AICE Symp. Ser. 74(175):134-142, 1978.
)\ ™T""™
Crisman, T. L., R. L. Schulze, P. L. Brezonik, and S. A, Bloom. Acid precipitation: the
biotic response in Florida lakes. _In: Ecological Impact of Acid Precipitation,
Proceedings of an International Conference, SNSF Project, Sandefjprd, Norway,* March
11-14, 1980, D. Drab1*5s and A. Tollan, eds., SNSF Project, Oslso-As, Norway, October
1980. pp. 296-298.
Cronan, C. S. Solution chemistry of a New Hampshire subalpine ecosystem: biochemical
patterns and processes. Ph.D. Thesis, Dartmouth College, Hanover, NH, 1978.
Cronan, C. S. Consequences of sulfuric acid inputs to a forest soil. In: Atmospheric Sulfur
Deposition: Environmental Impact and Health Effects, Proceedings of the Second Life
Sciences Symposium, Oak Ridge National Laboratory and Others, Gatlinburg, Tennessee,
October 14-18, 1979. D. S. Shriner, C. R. Richmond, and S. E. Lindberg, eds., Ann Arbor
Science Publishers, Inc., Ann Arbor, MI, 1980. pp. 335-343.
Cronan, C. S., and C. L. Schofield. Aluminum leeching response to acid precipitation:
effects on high-elevation watersheds in the Northeast. Science (Washington, D.C.)
204:304-305, 1979.
Cullis, C. F. and M. M. Hirschler. Atmospheric sulphur: natural and man-made sources.
Atmos. Environ. 14:1263-1278, 1980.
Dancer, W. S., L. A. Peterson, and G. Chesters. Ammonification and nitrification of N as
influenced by soil pH and previous N treatments. Soil Sci. Soc. Am. Proc. 37:67-69,
1973.
Dannevig, 6. Influence of precipitation on the acidity of water courses and on fish stocks.
Jeger og Fisker. 3:116-118, 1959. Cited in: Leivestad et al., 1976; Wright et a!.,
1976; Wright and Snekvik, 1978.
Davis, R. B., M. 0. Smith, J. H. Baily, and S. A. Norton. Acidification of Maine (U.S.A.)
lakes by acidic precipitation, Verh. Int. Ver. Theor. Angew Limnol. 20:532-537, 1978.
Denison, R., B. Caldwell, B. Bormann, L. Eldred, C. Swanberg, and S. Anderson. The effects of
acid rain on nitrogen fixation in western Washington coniferous forests. In: First
International Symposium on Acid Precipitation and the Forest Ecosystem, U.S. Department
of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil Pollut.
8:21-34, 1977.
Dickson, W. The acidification of Swedish lakes. Fish. Board Swed. Inst. Freshwater Res.
Drottingham Rep. 54:8-20, 1975.
Dickson, W, Some effects of the acidification of Swedish lakes. Verh. Int. Ver. Theor.
Angew. Limnol. 20:851-856, 1978.
Dillon, P. 0., D. S. Jefferies, W. Snyder, R. Reid, N. D. Van, D. Evans, J. Moss, and W. A.
Scheider. Acidic precipitation in south-central Ontario: recent observations. J. Fish.
Res. Board Can. 35:809-815, 1978.
Dodson, S. I. Zooplankton competition and predation: an experimental test of the size-
efficiency hypothesis. Ecology 55:605-613, 1974.
Donahue, R. L., R. W. Miller, and J. C. Shickluna. Soils: An Introduction to Soils and Plant
Growth. 4th Edition. Prentice-Hall, Inc., Englewood Cliffs, NJ, 1977. pp. 244-248.
7-108
-------
Drablj^s, D., and I. Sevaldrud. Lake acidification, fish damage, and utilization of outfields.
A comparative survey of six highland areas, southeastern Norway. In: Ecological Impact
of Acid Precipitation, Proceedings of an International Conference, SNSF Project,
Sandefjord, Norway, March 11-14, 1980, D. Drablsfe and A. Tollan, eds., SNSF Project,
Oslo-As, Norway, October, 1980. pp. 354-355.
Driscoll, C. T. , Jr. Chemical characterization of Some Dilute Acidified Lakes and Streams in
the Adirondack region of New York State. Ph.D. Thesis, Cornell University, Ithaca, NY,
1980.
Driscoll, C. T. , Jr., J. P. Baker, J. J. Bisogni, Jr., and C. L. Schofield. Effect of
aluminum speciation on fish in dilute acidified waters. Nature (London) 284:161-163,
1979.
Dvorak, A. J., B. G. Lewis, P. C. Ghee, J. D. Jastrow, J. C. Prioleau, E. H. Dettmann, F. C.
Kornegay, L. F. Soholt, R. F. Freeman, III, D. L. Mabes, W. S. Vinikour, R. M. Goldstein,
P. A. Merry, E. W. Walbridge, R. R. Hinchman, and E, D. Pentecost. Impacts of Coal-Fired
Power Plants on Fish, Wildlife, and Their Habitats. FWS/OBS-78/29, U.S. Department of
the Interior, Fish and Wildlife Service, Ann Arbor, MI, March 1978. pp. 64-70.
Eaggleton, A. E. J. and R. A. Cox. Homogeneous oxidation of sulfur compounds in the atmos-
phere. In: Sulfur in the Atmosphere, Proceedings of'the International Symposium, United
Nations Environment Program and Others, Dubrovnik, Yugoslavia, September 7-14, 1977.
Atmos. Environ. 12:227-230, 1978.
Eaton, J. S., G. E. Likens, and F. H. Bormann. Throughfall and Stemflow Chemistry in a
Northern Hardwood Forest. J. Ecol. 61:495-508, 1973.
Eliassen, A., and J. Saltbones. Decay and transformation rates of SO,, as estimated from
emission data, trajectories and measured air concentrations. Atmos Environ. 9_:425-429,
1975.
European Inland Fisheries Advisory Committee. Water quality criteria for European freshwater
fish - extreme pH values and inland fisheries. Water Res. 3:593-611, 1969.
Evans, L. S., N. F. Gmur, and J. J. Kelsch. Perturbations of upper leaf surface structures by
simulated acid rain. Environ. Exp. Bot. 17:145-149s 1977a.
Evans, L. S. , and 0. M. Bozzone. Effect of buffered solutions and sulfate on vegetative and
sexual development in gametophytes of Pteridium aquilinum. Am. J. Bot. 64:897-902, 1977.
Evans, L. S. , N. F. Gmur, and F. Da Costa. Leaf surface and histological perturbations of
leaves of Phaseolus vulgaris and He!ianthus annuus after exposure to simulated acid rain.
Am. J. Bot. 64:903-913, 1977b.
Evans, L. S., and D. M. Bozzone. Effect of buffered solutions and various anions on vegeta-
tive and sexual development in gametophytes of Pteridium aquilinum. Can. J. Bot.
56:779-785, 1978.
Evans, L. S., N. F. Gmur, and F. Da Costa. Foliar response of six clones of hybrid poplar to
simulated acid rain. Phytopathology 68:847-856, 1978.
Evans, L. S. , and T. M. Curry. Differential responses of plant foilage to simulated acid
rain. Am. J. Bot. 66:953-962, 1979.
Fagestrom, T., and A. Jernelov. Some aspects of the quantitative ecology of mercury. Water
Res. 6:1193-1202, 3972.
7-109
-------
Ferenbaugh, R. W. Effects of simulated acid rain on Phaseolus vulgaris L. (Fabaceae). Am. J.
Bot. 63:283-288, 1976.
Fiance, S. 'B. Distribution and biology of mayflies and stoneflies of Hubbard Brook, New
Hampshire. M.S. Thesis, Cornell University, Ithaca, NY, 1977.
Fisher, B. E. A, Long-range transport and deposition of sulfur oxides. In; Sulfur in the
Environment, Part I: The Atmospheric Cycle. J. 0. Nriagu, ed. , John Wiley and Sons,
Inc., New York, NY, 1978. pp. 243-295.
Foster, R. J. Geology. Charles E. Merrill Publishing Co., Columbus, OH, 1971. 162 pp.
Fredricksen, R. L. Nutrient budget of a Douglas-fir forest on an experimental watershed in
western Oregon. In; Research on Coniferous Forest Ecosystems: First Year Progress in
the Coniferous Forest Biome, US/IBP, Proceedings of a Symposium, Forty-Fifth Annual
Meeting, Northwest Scientific Association, Bellingham, Washington, March 23-24, 1972. J.
F. Franklin, L. J. Dempster, and R. H. Waring, eds., U.S. Department of Agriculture,
Forest Service, Portland, OR, 1972. pp. 115-131.
Friend, J. P. The global sulfur cycle. In: Chemistry of the lower atmosphere. S. I.
Rasool, ed. Plenum Press, NY, 1973.
Gahnstrom, G. , G. Andersson, and S. Fleischer. Decomposition and exchange processes in
acidified lake sediment. In: Ecological Impact of Acid Precipitation, Proceedings of an
International Conference, SNSF Project, Sandefj,ord, Norway, March 11-14, 1980. D.
Drabltfs and A. Tollan, eds., SNSF Project, Oslo-As, Norway, October 1980. pp. 306-307.
Gallo./ay, J. N. Effects of sulfur oxides on aquatic ecosystems. In; Sulfur oxides, National
Academy of Sciences, Washington, DC, 1978.
Galloway, J. N., and E. B. Cowling. The effects of precipitation on aquatic and terrestrial
ecosystems: a proposed precipitation network. J. Air Pollut. Control Assoc. 28:229-235,
1978.
Galloway, J. N., and D. M. Whelpdale. An atmospheric sulfur budget for eastern North America.
Atmos. Environ. 14:409-417, 1980,
Galvin, P. J. , P. J. Samson, P. E. Coffey, and D. Romano. Transport of sulfate to New York
State. Environ. Sci. Techno!. 12:580-584, 1978.
Galvin, P. J., and J. A. Cline. Measurement of anions in the snow cover of the Adirondack
Mountains. Atmos. Environ. 12:1163-1167, 1978.
Giddings, J. , and J. N. Galloway. The effects of acid precipitation on aquatic and
terrestrial ecosystems. In; Literature Reviews on Acid Precipitation. Cornell
University, Centers for Environmental Quality Management Water Resources and Marine
Sciences, Ithaca, NY, March 1976. 40 pp.
Gjessing, E. T., A. Henriksen, M. Johannessen, and R. F. Wright. Effects of acid precip-
itation on freshwater chemistry. In: Impact of Acid Precipitation on Forest and
Freshwater Ecosystems in Norway: Summary Report on the Research Results From the Phase I
(1972-1975) o£ the SNSF-Project. F. H. Braekke, ed. , Research Report FR 6/76, SNSF
Project, Oslo-As, Norway, March 1976. pp. 65-85.
Glass, N. R. , G. E. Glass, and P. J. Rennie. Effects of acid precipitation. Environ. Sci.
Techno!. 13:1350-1355, 1979.
7-110
-------
Gorham, E. The influence and importance of daily weather conditions in the supply of
chloride, sulfate and other ions to fresh waters from atmospheric precipitation. Philos.
Trans. R. Soc. London Ser. 8.241:147-178, 1958.
Gorham, E. Acid precipitation and its influence upon aquatic ecosystems—an overview. In;
First International Symposium ,on Acid Precipitation _and the Forest Ecosystem, U. S.
Department of Agriculture and Others, Columbus, tfhio," May 12-15, 1975. Water Air Soil
Pollut. 6:457-481, 1976.
Gosner, K. L., and I. H. Black, The effects of acidity on the development and hatching of New
Jersey frogs. Ecology 38:256-262, 1957.
Grahn, 0. Macrophyte succession in Swedish lakes caused by deposition of airborne acid sub-
stances. In: Proceedings of the First International Symposium on Acid Precipitation and
the Forest Ecosystem, May 12-15, 1975, Columbus, Ohio. L. S. Dochinger and T. A.
Seliga, eds., U. S. Forest Service General Technical Report NE-23, U. S. Department of
Agriculture, Forest Service, Northeastern Forest Experiment Station, Upper Darby, PA,
1976. pp. 425-458.
Grahn, 0., H. Hultberg, and L. Landner. Oligotrophication—a self-accelerating process in
lakes subjected to excessive supply of acid substances. Ambio 3_: 93-94, 1974.
Granat, L. On the relation between pH and the chemical composition in atmospheric preci-
pitation. Tell us 24:550-560, 1972.
Grennard, A., and F. Ross. Progress report on sulfur dioxide. Combustion 45:4-9, 1974.
Hagen, A., and A. Langeland. Polluted snow in southern Norway and the effect of the meltwater
on freshwater and aquatic organisms. Environ. Pollut. 5:45-57, 1973.
Hagstrom, T. Grodornas forsvinnande i an forsurad sjo. [Disappearence of frogs from an acid
lake.] Sver. Nat. 11/77:367-369, 1977. Cited in: Hendrey, 1978.
Haines, E. B. Nitrogen content and acidity of rain on the Georgia coast. Water Resour. Bull.
12:1223-1231, 1976.
Hall, R. J., G. E. Likens, S. B. Fiance, and G. R. Hendrey. Experimental acidification of a
stream in the Hubbard Brook Experimental Forest, New Hampshire. Ecology 61:976-989,
1980.
Halstead, R. L., and P. J. Rennie. Effects of sulphur on soils in Canada. In: Sulphur and
its Inorganic Derivatives in the Canadian Environment. NRCC N.C. 15015, National
Research Council of Canada, NRC Associate Committee on Scientific Criteria for
Environmental Quality, Ottawa, Canada, 1977. pp. 181-219.
Hamilton, W. C. Personal communication to Dr. J. H. B. Garner, U. S. Environmental
Protection Agency, Research Triangle Park, NC, January 31, 1980.
Hanken, I. , J. F. Lynch, and D. B. Wake. Salamander invasion of the tropics. Nat. Hist.
89:(12) December 47-53, 1980.
Harr, T. E., and P. E. Coffey. Acid Precipitation in New York State. Technical Paper No. 43,
New York State Department of Environmental Conservation, Albany, NY, July 1975. pp,
21-29.
Harward, M. E., and H. H. Reisenaur. Movement and reactions of inorganic soil sulfur. Soil
Sci. 101:326-335, 1966.
7-111
-------
Heagle, A. S., W. W, Heck, W. M. Knott, J. V. Johnston, E. P. Stahel, and E. B. Cowling.
Responses of citrus to acidic rain from simulated SRM fuel exhaust mixtures and exhaust
components. Interim report to NASA John F. Kennedy Space Center, Titusville, FL, 1978.
11 pp.
Heinrichs, H., and R. Mayer. Distribution and cycling of major and trace elements in two
central forest ecosystems. J. Environ. Qua!. 6:402-407, 1977.
Henderson, 6. S., and W. F. Harris. An ecosystem approach to characterization of the nitrogen
cycle in a deciduous forest watershed. Iji Forest Soils and Forest Soils Conference. B.
Bernier and C. H. Winget, eds., Les Presses de TUniversite Laval, Quebec, Canada, 1975.
pp. 179-193.
Hendrey, G. R. Effects of pH on the Growth of Periphytic Algae in Artificial Stream Channels.
Report IR 25/76, SNSF Project, Oslo-Ss, Norway, October 1976.
Hendrey, G. R., and R. F. Wright. Acid precipitation in Norway: Effects on aquatic fauna.
In: Atmospheric Contribution to the Chemistry of Lake Waters, First Specialty Conference
of the International Association for Great Lakes Research, Longford Mills, Ontario,
Canada, September 28-October 1, 1975. J. Great Lakes Res. 2(Suppl. 1):192-207, 1976.
Hendrey, G. R., and F. Vertucci. Benthic plant communities in acidic Lake Golden, New York:
sphagnum and the algal mat. In: Ecological Impact of Acid Precipitation, Proceedings
of an International Conference, SNSF Project, San^lefjord, Norway, March 11-14, 1980, D.
Drablife and A. Tollan, eds., SNSF-Project, Oslo-As, Norway, October 1980, pp. 314-315.
Hendrey, G. R. Aquatics Task Force on Environmental Assessment of the Atikokan Power Plant:
Effects on Aquatic Organisms. BNL 50932, U. S. Department of Energy, Brookhaven National
Laboratory, Upton, NY, November 1978.
Hendrey, G. R., K. Baalstrud, T. S. Traaen, M. Laake, and G. Raddum. Acid precipitation: some
hydrobiological changes. Ambio 5:224-227, 1976.
Hendrey, G. R. , J. N. Galloway, S. A. Norton, C. L. Schofield, P. W. Shaffer and D. A. Burns.
Geological and Hydrochemical Sensitivity of the Eastern United States to Acid Precipita-
tion. EPS-600/3-80-024, U.S. Environmental Protection Agency, Corvallis, OR, January
1980b.
Hendrey, G. R., N. D. Yan, and K. J. Baumgartner. Responses of freshwater plants and inverte-
brates to acidification. Presented at the EPA/OECD International Symposium for Inland
Waters and Lake Restoration, Portland, ME, September 8-12, 1980a.
Hendry, C. D., Jr. The Chemistry of Precipitation in North-Central Florida, M.S. Thesis M.S.
Thesis, University of Florida, Department of Environmental Engineering Sciences,
Gainesville, FL, 1977.
Hendry, C. D. , and P. L. Brezonik. Chemistry of precipitation at Gainesville, Florida.
Environ. Sci. 14:843-849, 1980.
Henriksen, A. A simple approach for identifying and measuring acidification of freshwater.
Nature (London) 278:542-545, 1979.
Henricksen, A. and R. F. Wright. Effects of acid precipitation on a small acid lake in
Southern Norway. Nord. Hydro!. 8:1-10, 1977.
Hermann, R., and J. Baron. Aluminum mobilization in acid stream environments, Great Smoky
Mountains National Park, U.S.A. In: Ecological Impact of Acid Precipitation, Proceedings
of an International Conference, SNSF Project, Sangiefjord, Norway, March 11-14, 1980. D,
Drablrfs and A Tollan, eds., SNSF Project, Oslo-As, Norway, October 1980, pp. 218-219.
7-112
-------
Hicks, B. B., and M. L. Wesely. Turbulent transfer processes to a surface and interaction
with vegetation. In: Atmospheric Sulfur Deposition: Environmental Impact and Health
Effects, Proceedings of the Second Life Sciences Symposium, Oak Ridge National Laboratory
and Others, Gatlinburg, Tennessee, October 14-18, 1979, D. S. Shriner, C. R, Richmond,
and S, E. Lindberg, eds., Ann Arbor Science Publishers, Inc., Ann Arbor, MI, 1980, pp.
199-207.
Hindawi, I. J., J. A. Rea, and W. L. Griffis. Response of bush bean exposed to acid mist.
Am. J. Bot. 67:168-172, 1980.
Hoeft, R. G., D. R. Keeney, and L. M. Walsh. Nitrogen and sulfur in precipitation and sulfur
dioxide in the atmosphere in Wisconsin. J. Environ. Qua!. 1:203-208, 1972.
Hornstrom, E., C, Ekstrom, U. Miller, W. Dickson. Effects of the Acidification on Lakes in
the Swedish West Coast Region. Statens Naturvardsverk, Slona, Sweden, Publikationer
1973: 7. Cited in: Wright and Gjessing, 1976'.
Hoi ling, C. C. Resilience and stability of ecological systems. Annu. Rev. Ecol. Syst.
4:1-23, 1973.
Hornbeck, J. W., G. E. Likens, J. S. Eaton. Seasonal patterns in acidity of precipitation and
their implications for forest stream ecosystems. 'In; Proceedings of the First Inter-
national Symposium on Acid Precipitation and the Forest Ecosystem, May 12-15, 1975,
Columbus, Ohio. L. S. Dochitiger and T. A. Seliga, eds. U. S. Forest Service, North-
eastern Forest Experimental Station, Upper Darby, PA, 1976. pp. 597-609.
Hubbert, M. K. Outlook for fuel reserves. In: McGraw-Hill Encyclopedia of Energy. D. N.
Lapedes, ed., McGraw-Hill Inc., New York, NY, 1976. pp 11-23.
Hultberg, H. Thermally stratified acid water in late winter - a key factor inducing self-
accelerating processes which increase acidification. In: Proceedings of the First
International Symposium on Acid precipitation and the Forest Ecosystem, May 12-15, 1975,
Columbus, Ohio. L. S. Dochinger and T. A. Seliga, eds. U. S. Forest Service, North-
eastern Forest Experimental Station, Upper Darby, PA, 1976. pp. 503-517.
Hutchinson, T. C. , and L. M. Whitby. The effects of acid rainfall and heavy metal parti-
culates on a boreal forest ecosystem near the Sudbury smelting region of Canada. In:
First International Symposium on Acid Precipitation and the Forest Ecosystem, U.S.
Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil
Pollut. 7:421-438, 1977.
Hutchinson, T. C. , W. Gizym, M. Havas, and V. Zoberns. Effects of long-term lignite burns on
Arctic ecosystems at the Smoking Hills, N.W.T. In: Proceedings of the First Inter-
national Symposfum on Acid Precipitation and the Forest Ecosystem, May 12-15, 1975,
Columbus, Ohio. L. S. Dochinger and T. A. Seliga, eds. U. S. Forest Service, North-
eastern Forest Experimental Station, Upper Darby, PA, 1976. pp. 745-746.
Irving, P. Rainfall acidity at argonne. In: Radiological and Environmental Research
Division Annual Report: Ecology, January through December 1978. ANL-78-65 (Pt. 3} U.S.
Department of Energy, Argonne National Laboratory, Argonne, IL, 1978, pp. 21-33.
Irving, P. M. , and J. E. Miller. The effects of acid precipitation alone and in combination
with sulfur dioxide on field-grown soybeans. In: Radiological and Environmental
Research Division Annual Report: Ecology, January through December 1978. ANL-78-65 (Pt.
3), U.S. Department of Energy, Argonne National Laboratory, Argonne, II, 1978, pp. 17-20.
Irving, P. M. Response of Field-Grown Soybeans to Acid Precipitation Alone and in Combina-
tion with Sulfur Dioxide. PhD. Thesis, University of Wisconsin, Milwaukee, WI, 1979.
-------
Jacobson, J.S. Experimental studies on the phytotoxiclty of acidic precipitation: the United
States experience. In: Effects of Acid Precipitation on Terrestrial Ecosystems, North
Atlantic Treaty Organization, Toronto, Ontario, Canada, May 21_27, 1978. T, C.
Hutchinson. and M. Havas, eds., Plenum Press, New York, NY, 1980a, pp. 151-160.
Jacobson, J.S. The influence of rainfall composition on the yield and quality of agricultural
crops. In: Ecological Impact of Acid Precipitation, Proceedings of, an International
Conference, SNSF Project, Sande|jord, Norway, March 11-14, 1980. D. DrabljJs and A.
Tollan, eds., SNSF Project, Oslo-As, Norway, October 1980b, pp. 41-46.
Jacobson, J.S. , and P. Van Leuken. Effects of acidic precipitation on vegetation. In:
Proceedings of the Fourth International Clean Air Congress, International Union of Air
Pollution Prevention Associations, Tokyo, Japan, May 16-20, 1977. S.,Kasuga, N. Suzuki,
T. Yamada, 6. Kimura, K. Inagaki, and K. Onoe, ets., The Japanese Union of Air Pollution
Prevention Associations, Tokyo Japan, 1977, pp. 124-127.
Jacobson, J.A., J. Troiano, L. J. Colavito, L. I. Heller, and D. C. McCune. Polluted rain and
plant growth. In: Polluted Rain, Twelfth International Rochester Conference on Environ-
mental Toxicity, University of Rochester, Rochester, New York, May 21-23, 1979. T. Y.
Toribara, M. W. Miller, and P. E. Morrow, eds., Environmental Science Research Volume 17,
Plenum Press, New York, NY, 1980, pp. 291-305.
Jensen, K. W., and E. Snekvik. Low pH levels wipe out salmon and trout populations in south-
ernmost Norway. Ambio I:223-225, 1972.
Johannessen, M., and A. Henriksen. Chemistry of snow meltwater: .changes in concentration
during melting. Water Resour. Res. 14:615-619, 1978.
Johannessen, M., A. Skartveit, and R. F. Wright. Streamwater chemistry before, during and
after snowmelt. In: Proceedings of an International Conference, SNSF Project,
Sandegjord, Norway, March 11-14, 1980. D. Drab¥s and A. Tollan, eds,, SNSF Project,
Oslo-As, Norway, October 1980, pp. 224-225.
Johnson, D. W. Site susceptibility to leaching by H^SC), in acid rainfall. In: Effects of
Acid Precipitation on Terrestrial Ecosystems, North Atlantic Treaty Organization,
Toronto, Ontarip, Canada, May 21-27, 1978. T. C. Hutchinson and M. Hawas, eds., Plenum
Press, New York, NY, 1980, pp. 525-535.
Johnson, D. W., and D. W. Cole. Anion mobility in sols: Relevance to nutrient transport from
forest ecosystems. Environ. Int. 3:79-90, 1980.
Johnson, D. W., J. W. Hornbeck, J. M. Kelly, W. T. Swank, and D. E. Todd. Regional patterns
of soil sulfate accumulation: relevance to ecosystem sulfur budgets. In: Atmospheric
Sulfur Deposition: Enfironmental Impact and Health Effects, Proceedings of the,Second
Life Sciences Symposium, Oak Ridge National Laboratory and Others, Gatlinburg, Tennessee,
October 14-18, 1979. D. S, Shriner, C. R. Richmond, and S. E. Lindberg, eds., Ann Arbor
Science Publishers, Inc., Ann Arbor, MI, 1980, pp. 507-520.
Johnson, N. M., G. E. Likens, F. H. Bormann, D. W. Fisher, and R. S. Pierce. A working model
for the variation in stream water chemistry at the Hubbard Brook Experimental Forest, New
Hampshire. Water Resources Res. 5:1353-1363, 1969.
Jones. M. B. Sulfur in agricultural lands. In: Sulfur in the Environment. Revised Edition.
Missouri Botanical Garden, St. Louis, Mo., July 1975, pp. 146-158.
Junge, C. E. The distribution of ammonia and nitrate in rainwater over the United States.
Trans. Am. Geophys. Union 39:241-248, 1958.
7-114
-------
Kallqvist, T., R. Romstad, and J. Kotai. Preliminary study of pH-effects on algal communities
in experimental channels. SNSF-Project IR 7/75, 1975. Cited in: Hendry et al., 1980.
Kramer, J. R. , Geochemical and lithological factors in acid precipitation. In: Proceedings
of the First International Symposium on Acid Precipitation and the Forest Ecosystem, U.S.
Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. L. S. Dochinger
and T. A. Seliga, eds. , NEFES/77-1", U.S. Department T»T Agriculture, Forest Service, Upper
Darby, PA, August 1976, pp. 611-618.
Kucera, V. Effects of sulfur dioxide and acid precipitation on metals and anti-rust painted
steel. Ambio 5:243-248, 1976.
Kwiatkowski, R. E., and J. C. Roff. Effects of acidity on the phytoplankton and primary pro-
ductivity of selected northern Ontario lakes. Can. J. Bot. 54:2546-2561, 1976.
Laake, M. Effekter av lav pH pa produksjon, nedbryting og stoffkretsltfp i littoralsonen.
[Effects of low pH on the production, breaking down of harmful substances in a coastal
region.] Research Project IR 29/76, 1432 Aas-NLH, Norway. SNSF-Projects Secretariat,
1976. 75 pp. Cited in: Galloway, 1978.
Landner, L., and P. 0. Larsson. Biological Effects of Mercury Fall-Out Into Lakes from the
Atmosphere, IUL Report BUS. Institute for Water and Air Research, Stockholm, Sweden,
1972. 18 pp. (in Swedish, translated by H. Altosaar, Domtar Research Centre, December
23, 1975)
Lang, D. S., S. V. Krupa, and D. S. Shriner. Injury to vegetation incited by sulfuric acid
aerosols and acidic rain. Presented at the 71st Annual Meeting, Air Pollution Control
Association, Houston, TX, June 25-30, 1978. Paper No. 78-7.3.
Larson, T. V., R. J. Charlson, E. J, Knudson, G. D. Christian, and H. Harrison. The influence
of a sulfur dioxide point source on the rain chemistry of a single storm in the Puget
Sound region. Water Air Soil Pollut. 4:319-328, 1975.
Larson, T. E. and J. Hettrick, Mineral composition of rainwater. Tellus 8: 196, 1956.
Lee, J. J. , and D. E. Weber. Effects of Sulfuric Acid Rain on Two Model Hardwood Forests:
Throughfall, Litter Leachate, and Soil Solution. EPA-600/3-80-014, U.S. Environmental
Protection Agency, Corvallis, OR, January 1980.
Lee, J. J., G. E. Neely, and S. C. Perrigan. Sulfuric Acid Rain Effects on Crop Yield and
Foliar Injury. EPA-600/3-80-016, U.S. Environmental Protection Agency, Corvallis, OR,
January 1980.
Leivestad, H., G. Hendrey, I. P. Muniz, and E. Snekvik. Effects of acid precipitation on
freshwater organisms. In: Impact of Acid Precipitation of Forest and Freshwater Eco-
systems in Norway: Summary Report on the Research Results from the Phase I (1972-75)0of
the SNSF-Project. F. H. Braekke, ed., Reserach Report FR 6/76, SNSF Project, Oslo-As,
Norway, March 1976, pp. 87-111.
Lewis, W. M., Jr., and M. C. Grant. Acid precipitation in the western United States. Science
(Washington, D.C.) 207:176-177, 1980.
Likens, G. E. The Chemistry of Precipitation in the Central Finger Lakes Region. Technical
Report 50, , Cornell University, Water Resources and Marine Science Center, Ithaca, NY,
October 1972.
7-115
-------
Likens, G. E. Acid precipitation. Chem. Eng. News 54:29-44, 1976.
Likens, G. E., and F. H. Bormann. Acid rain: a serious regional environmental problem.
Science (Washington, D.C.) 184:1176-1179, 1974.
Likens, G. E., G, H. Bormann, N. M. Johnson, D. W, Fisher, and R. S. Pierce. Effects of
cutting and herbicide treatment on nutrient budgets in the Hubbard Brook Watershed -
ecosystem. Ecol. Monogr. 40:23-47, 1970.
Likens, G. E., F. H. Bormann, J. S. Eaton, R. S. Pierce, and N. M. Johnson. Hydrogen ion
input to the Hubbard Brook Experimental Forest, New Hampshire, during the last decade.
Im First International Symposium on Acid Precipitation and the Forest Ecosystem, U.S.
Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil
Pollut. 6:435-445, 1976.
Likens, G. E., J. S. Eaton, and J. N. Galloway, precipitation as a source of nutrients for
terrestrial and aquatic ecosystems. In: Precipitation Scavenging (1974), Proceedings
of a Symposium, Illinois State Water Survey and Others, Champaign, Illinois, October
14-18, 1974, EROA Symposium Series 41, Energy Research and Development Administration,
Oak Ridge, TN, June 1977, pp. 552-570.
Likens, G. E., R. F. Wright, J. N. Galloway, and T. J. Butler. Acid rain. Sci. Am.
241:43-51, 1979.
Liljestrand, H. M., and J. J. Morgan. Chemical composition of acid precipitation in Pasadena,
Calif. Environ. Sci. Techno!. 12:1271-1273, 1978.
Lindberg, S. E., R. C. Harriss, R. R. Turner, D. S. Shriner, and D. D. Huff. Mechanisms and
Rates of Atmospheric Deposition of Selected Trace Elements and Sulfate to a Deciduous
Forest Watershed. ORNL/TM 6674, Environmental Sciences Division Pub. No. 1299, U.S.
Department of Energy, Oak Ridge National Laboratory, Oak Ridge, TN, June 1979.
Linzon, S. N. Statement made to workshop reviewers. Research Triangle Park, NC., November,
1980.
Livingstone, D. A. Chemical composition of rivers and lakes. In: Data of Geochemistry.
Sixth Edition - Chapter G. M. Fleischer, ed., Geological Survey Professional Paper 440-G,
U.S. Government Printing Office, Washington, DC, 1963.
Lockhart, W. L., and A. Lutz. Preliminary biochemical observations of fishes inhabiting an
acidified lake in Ontario, Canada. In: Proceedings of the First International Symposium
on Acid Precipitation and the Forest Ecosystem, May 12-15, 1975, Columbus, Ohio. L. S.
Dochinger and T. A. Seliga, eds. U. S. Forest Service General Technical Report NE-23, U.
S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station,
Upper Darby, PA, 1976. pp. 725-735.
Lynch, H. Predation, competition and zooplankton community structure: an experimental study.
Limnol. Oceanogr. 24:253-272, 1979.
Malraer, N. Acid precipitation: chemical changes in the soil., Ambio 5:231-234, 1976.
Marenco, A, and J. Fontan. Influence of dry deposition on the residence time of particulate
pollutants in the troposphere. In: Atmosphere-Surface Exchange of Particulate and
Gaseous Pollutants (1974), Proceedings of a Symposium, Bate'lle Pacific Northwest
Laboratories and Others, Richland, Washington, September 4-6, 1974. ERDA Symposium Seris
38, Energy Research and Development Administration, Oak Ridge, TN, January 1976, Pp.
54-61,
7-116
-------
May, R., Stability and Complexity in Model Ecosystems. Princeton University Press, Princeton,
N.J., 1973.
Mayer, R. , and B. Ulrich. Acidity of precipitation as influenced by the filtering of at-
mospheric sulphur and nitrogen compounds-its role in the element balance and effect on
soil. In: First International Symposium on Acid Precipitation and the Forest Ecosystem,
U.S. Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water, Air,
and Soil Pollut. 7:409-416, 1977.
McColl, J. G., and D. S. Bush. Precipitation and throughfall chemistry in the San Francisco
Bay Area. J. Environ. Qua! 7:352-357, 1978.
McFee, W. W., J. M. Kelly, and R. H. Beck. Acid precipitation effects on soil pH and base
saturation of exchange sites. In; First International Symposium on Acid Precipitation
and the Forest Ecosystem, U.S. Department of Agriculture and Others, Columbus, Ohio, May
12-15, 1975. Water Air Soil Pollut. 7:401-408, 1977.
McFee, W. W. Sensitivity of Soil Regions To Long Term Acid Precipitation. U.S. Environmental
Protection Agency, EPA-600/3-80-013, Corvallis, OR. January 1980,
Menendex, R. Chronic effects of reduced pH on brook trout- (salvelinus fontinalis). J. Fish.
Res. Board Can. 33:118-123, 1976.
Miller, J. M., J. N. Galloway, G. E. Likens. Origin of air masses producing acid precipita-
tion at Ithaca, New York, a preliminary report. Geophys. Res. Lett. 5:757-760, 1978.
Moran, J. M., M. D. Morgan, and J. H. Wiersma. Introduction to Environmental Science. W. H.
Freeman and Co., San Francisco, CA, 1980. pp. 7-75.
Muller, P. Effects of artificial acidification on the growth of periphyton. Can. J. Fish.
Aquat. Sci. 37:355-363, 1980.
Muniz, I. P., and H. Leivestad. Toxic effects of aluminum on brown trout, Salmo trutta, L.
In: Ecological Impact of Acid Precipitation, Proceedings of an International Conference,
SNSF Project, Sand%fjord, Norway, March 11-14, 1980. D. Drablrfs and A. Tollan, eds.,
SNSF Project, Oslo-As, Norway, October 1980, pp. 320-321.
National Research Council. Nitrates: An Environmental Assessment. National Academy of
Sciences, Washington, DC, 1978. pp. 275-317.
National Research Council, Sulfur Oxides. National Academy of Sciences, Washington, D.C.,
1978b.
Nieboer, E., D. H. S. Richardson, K. J. Puckett and F. D. Tomassini. The phytotoxicity of
sulfur dioxide in relation to measureable responses in lichens. In: Effects of Air
Pollutants on Plants, Society for Experimental Biology, Liverpool, England, April 10,
1975. T. A. Mansfield, ed. , Cambridge University Press, London, England, 1976*. pp.
61-85.
o
Nihlgard, B. Precipitation, its chemical composition and effect on soil water in a beech and
a spruce forest in south Sweden. Oikos, 21:208-217, 1970.
Nisbet, I. C. T. Sulfates and acidity in precipitation: their relationship to emissions and
regional transport of sulfur oxides. In; Air Quality and Stationary Source Emission
Control. A report by the Commission on Natural Resources, National Academy of Sciences,
National Academy of Engineering, National Research Council. Serial No. 94-4, U.S. Senate
Committee on Public Works, Washington, DC, March 1975. pp. 276-312.
7-117
-------
Norton, S. A. Changes in chemical processes in soils caused by acid precipitation. In: First
Inaternational Symposium on Acid Precipitation and the Forest Ecosystem, U.S. Department
of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil Pollut,
7:389-400,.1977.
NHagu, J. Deteriorative effects of sulfur pollution on materials. In: Sulfur in the
Environment. Part II: Ecological Impacts. J. 0. Nriagu, ed., John Wiley & Sons, Inc.,
New York, NY, 1978. pp. 1-59.
0kland, J. Environment and snails (Gastropoda): studies of 1000 lakes in Norway. In:
Ecological Impact of Acid Precipitation., Proceedings of an International Conference, SNSF
Project, Sandefjord, Norway, March 11-14, 1980. D. Drablrfs and A. Tollan, eds., SNSF
Project, Oslo-As, Norway, October 1980. pp. 322-323.
0kland, J., and K. A. 0kland. pH level and food organisms for fish: studies of 1000 Lakes in
Norway. In; Ecological Impact of Acid Precipitation, Proceedings of an International
Conference, SNSF Project, Sandegjord, Norway, March 11-14, 1980. D. Drablj^s and A.
Tollan, eds., SNSF Project, Oslo-As. Norway, October 1980. pp. 326-327.
0kland, K. A. Mussels and crustaceans: studies of 1,000 lakes in Norway. In: Ecological
Impact of Acid Precipitation, Proceedings of an International Conference, SNSF Project,
Sandeijord, Norway, March 11-14, 1980. D. Drabl^s and A. Tollan, eds., SNSF Project,
Oslo-As, Norway, October 1980. pp. 324-325.
Odin, S. The Acidification of Air and Precipitation and Its Consequences on the Natural
Environment. Ecology Committee Bulletin No. 1, State National Science Research Com-
mittee, Stockholm, Sweden, May 1968.
Odin, S. Acid precipitation: A world concern. In: Proceedings of the Conference on
Emerging Environmental Problems, Acid Precipitation, May 19-20, 1975, The Institute on
Man and Science, Rensselaerville, N.Y. U.S. Environmental Protection Agency, Region II.
Water Resources and Marine Sciences Center, Cornell University and Center for Environ-
mental Quality Management, Cornell University, EPA-902/9-75-001, November 1975. pp.
5-44.
Odin, S., R. Anderssan, and M. Barting. The long-term changes in the chemistry of soils in
Scandanavia due to acid precipitation. In: Supporting Studies to Air Pollution Across
National Boundaries. The impact on the environment of sulfur in air and precipitation,
B. Bolin, ed., Sweden's case study for the United Nations conference on the human en-
vironment, Royal Ministry of Foreign Affairs., Royal Ministry of Agriculture, Stockholm,
Sweden, 1972. p. 20.
Odum, E. P. Summary. In: Ecological Effects of Nuclear War, Proceedings of a Symposium,
Ecological Society of America, Amherst, Massachusetts, August 1963. G. M. Woodwell, ed.,
BNL 917(C-43), Brookhaven National Laboratory, Upton, NY, August 1965. pp. 69-72.
Odura, E. P. Fundamentals of Ecology. Third Edition. W. B. Saunders Co., Philadelphia, PA,
1971. p. 5, 8-139.
Office of Air Quality Planning and Standards. National Air Quality, Monitoring and Emissions
Trends Report, 1977. EPA-450/2-78-052, U.S. Environmental Protection Agency, Research
Triangle Park, NC, December 1978.
Overrein, L. N. Sulphur pollution patterns observed; leaching of calcium in forest soil
determined. Ambio. 1:145-147, 1972.
7-118
-------
Pack, D. H., G. J. Ferber, J. L. Heffter, K. Telegadas, J. K. Angell, W. H. Hoecker, and L.
Machta. Meterology of long-range transport. In: Sulfur in the Atmosphere, Proceedings
of the International Symposium, United Nations Environmental Program and Others,
Dubrovnik, Yugoslavia, September 7-14, 1977. Atmos. Environ. 12:425-444, 1978.
Pack, D. W. Sulfate betiavior in eastern U.S. precipitation.,.. Geophys. Res. Lett. 5:673-674,
1978.
Peterson, R. C. Dynamics of coarse particulate organic material processing in diversified pH
regimes. Presented at yie Internal. Conf. The Ecological Impact of Acid Precipitation,
March 11-14, 1980, Oslo-As, Norway.
Pfeiffer, M., and P. Festa. Acidity status of lakes in the Adirondack region of New York in
relation to fish resources. Publication FW-P168, New York State, Department of Environ-
mental Conservation, Albany, NY, August 1980.
Potts, D. T. W., and C. Fryer. The effect of pH and salt content on sodium balance in Daphnia
magnet and Acantholeberis curvirostris (Crustacea; Cladocera). J. Comp. Physio!. Series B
129:289-294, 1979.
Rough, F. H. Acid precipitation and embryonic mortality of spotted salamanders, Ambystoma
maculatum. Science (Washington, D.C.) 192:68-70, 1976.
Pough, F. H., and R. E. Wilson. Acid precipitation and reproductive success of Ambystoma
salamanders. In: Proceedings of the First International Symposium on Acid Precipitation
and the Forest Ecosystem, May 12-15, 1975, Columbus, Ohio. L. S. Dochinger and T. A.
Seliga, eds. L). S. Forest Service General Technical Report NE-23, U. S. Department of
Agriculture, Forest Service, Northeastern Forest Experiment Station, Upper Darby, PA, 1976.
Prahm, L. P. , U. Torp, and R. M. Stern. Deposition and transformation rates of sulphur oxides
during atmospheric transport over the Atlantic. Tellus 18:355-372, 1976.
Raddum, G. Invertebrates: quality and quantity as fish food. In: Limnological Aspects of
Acid Precipitation, Proceedings of the International Workshop, U.S. Environmental Pro-
tection Agency and Others, Sagamore, New York, September 25-28, 1978. G. R. Hendry, ed.,
BNL 51074, U.S. Department of Energy, Brookhaven National Laboratory Upton, NY, 1978.
pp. 17-24.
Rasmussen, R. A. Emission of biogenic hydrogen sulfide. Tellus 26:254-260, 1974.
Rheinheimer, G. Aquatic Microbiology. John Wiley & Sons, New York, 1971. 184 pp.
Robinson, E., R. B. Husar, and J. N. Galloway. Sulfur oxides in the atmosphere, jn: Sulfur
Oxides. National Academy of Sciences, Washington, DC, 1978. pp. 18-62.
Roff, J. R., and R. E. Kvnatkowski. Zooplankton and zoobenthos communities of selected nor-
therr, Ontario lakes of different acidities. Can. J. Zool. 55:899-911, 1977.
Rosenqvist, I. T. A contribution towards analysis of buffer properties of geological materials
against strong acids in precipitation water, report written for the Council for Research
in Natural Sciences, Norwegian General Research Council, 1976.
Rosseland, B. 0., I. Sevaldrud, D. Svalastog, and I. P. Muniz. Studies on freshwater fish
populations - effects of acidification or reproduction, population structure, growth and
food selection. In: Ecological Impact of Acid Precipitation, Proceedings of an Inter-
national Conference, SNSF Project, §andefjord, Norway, March 11-14, 1980, D. Drablris and
A. Tollan, eds., SNSF Project, Oslo-As, Norway, October 1980. pp. 336-337.
7-119
-------
Schaffer, R. J. The Weathering of Natural Building Stones, Spec. Ref. No. 18, London, HMSO,
1932« Cited in: Sereda, 1977.
Scheider, W., J. Adamski, and M. Paylor. Reclamation of Acidified Lakes Near Sudbury, Ontario
Ministry of the Environment, Rexdale, Ontario, Canada, June 1975.
Scheider, W., J. Adamski, and M. Paylor. Reclamation of Acidified Lakes Near Sudbury,
Ontario. Ontario Ministry of the Environment, Rexdale, Ontario, Canada, 1975. Cited in:
Wright and Gjessing, 1976.
Schindler, D. W. Whole lake eutrophication experiments with phosphorus, nitrogen and carbon.
Verh. Internet. Verein. Limnol. 19:577-582, 1975.
Schindler, D. W., R. Wagemann, R. B. Cook, T. Ruszczynski, and J. Prokopowich. Experimental
acidification of Lake 223, experimental lakes area: background data and the first three
years of acidification. Can. J. Fish. Aquat. Sci. 37:342-354, 1980.
Schlesinger, W. H., and M. M. Hasey. The nutrient content of precipitation, dry fallout, and
intercepted aerosols in the chaparral of southern California. Am. Midi. Nat.
103:114-122, 1980.
Schofield, C. L. Acid precipitation: effects on fish. Ambio. 5:228-230, 1976a.
Schofield, C. L. Dynamics and Management of Adirondack Fish Populations. I. Natural Waters
with Lethal Conditions for Fish. I-a. Acidification of Adirondack lakes by atmospheric
precipitation: Extent and Magnitude of the Problem, Final Report. Project No, F-28-R,
State of New York, Department of Environmental Conservation, Albany, NY, 1976b.
Schofield, C. L. Dynamics and Management of Adirondack Fish Populations: I. Natural Waters
with Lethal Conditions for Fish. I-b. Acidification of Adirondack Lakes by Atmospheric
Precipitation: Long Term and Seasonal Trends.Project No. F-28-R-4, State of New York,
Department of Environmental Conservation, Albany, NY, 1976c.
Schofield, C. L. The acid precipitation phenomenon and its impact in the Adirondack Mountains
of New York State. In: Scientific Paper from the Public Meeting on Acid Precipitation,
New York State Assembly Committee on Environmental Conservation and Others, Lake Placid,
New York, May 4-5, 1978. H. H. Icord and J. S. Jacobson, eds., New York State Assembly,
Science and Technology Staff, Albany, NY, March 1979. p. 86-91.
Schofield, C. L. Effects of acid rain on lakes. Presented at the Session on Acid Rain,
National Convention of the American Society of Civil Engineers, Boston, MA, April 2, 1979.
Schofield, C. L. and J. R. Trojnar. Aluminum toxicity to brook trout (Salvelinus fontinali§)
in acidified waters. In: Polluted Rain, Twelfth International Rochester Conference on
Environmental Toxicity, University of Rochester, Rochester, New York, May 21-23, 1979.
T. Y. Toribara, M. W. Miller, and P. E. Morrow, eds., Environmental Science Research,
Volume 17, Plenum Press, New York, NY, 1980. pp. 341-366.
Sehmel, G. A. Particle and gas dry deposition: a review. Atmos. Environ. 14:983-1011, 1980.
Seip, H. M. Acidification of freshwater-sources and mechanisms. In; Ecological Impact of
Acid Precipitation, Proceedings of an International Conference, SNSF Project, Sandefjojjd,
Norway, March 11-14, 1980, D. DrablizJs and A. Tollan, eds., SNSF Project, Report-As,
Norway, October 1980. pp. 358-366.
Sereda, P. J. Effects of sulphur on building materials. lYi: Sulphur and its Organic
Derivatives in the Canadian Environment. NRC No. 15015, National Research Council of
Canada. NRC Associate Committee on Scientific Criteria For Environmental Quality,
Ottawa, Canada, 1977. pp. 359-426.
7-120
-------
Sheridan, R. P., and R. Rosenstreter. The effect of hydrogen ion concentration in simulated
rain on the moss fortyla rural is (Hedw) Sm, Bryologist 76:168-173, 1973.
Shinn, J. H., and S. Lynn. Do man-made sources affect the sulfur cycle of northeastern
states? Environ. Sci. Techno!. 13:1062-1067, 1979.
Shriner, D. S. Effects of simulated rain acidified with sulfuric acid on host-parasite inter-
actions. In; Proceedings of the First International Symposium on Acid Precipitation and
the Forest Ecosystem, May 12-15, 1975, Columbus, Ohio. L. S. Dochinger and T. A. Seliga,
eds. U. S. Forest Service General Technical Report NE-23, U. S. Department of Agri-
culture, Forest Service, Northeastern Forest Experiment Station, Upper Darby, PA, 1976.
Pp. 919-925.
Shriner, D. S. Effects of simulated acidic rain on host-parasite interactions in plant
diseases. Phytopathology 68:213-218, 1978a.
Shriner, D. S. Interactions between acidic precipitation and S0? or 0.,: Effects on plant
response. Phytopathol. News, 12:153, 1978b.
Shriner, D. S. Terrestrial vegetation - air pollutant interactions: nongaseous pollutants,
w,et deposition. Presented at the International Conference on Air Pollutants and Their
Effects on Terrestrial Ecosystems, Banff, Alberta, Canada, May 10-17, 1980.
Shriner, D. S., and G. S. Henderson. Sulfur distribution and cycling in a deciduous forest
watershed. J. Environ. Qua!. 7:392-397, 1978.
Shriner, D. S., M. DeCot, and E. B. Cowling. Simulated acid rain caused direct injury to
vegetation. Proc. Am. Phytopath. Soc. 1:112, 1974.
Smith, F. B., and G. H. Jeffrey. Airborne transport of sulfur dioxide from the U.K. Atmos.
Environ. 9:643-659, 1975.
Smith, R. A. Air and Rain: The Beginnings of Chemical Climatology. Longmans, Green, London,
1872. 600 pp.
Smith, R. L. Ecology and Field Biology. 3rd Edition. Harper and Row, New York, NY, 1980.
pp. 11-199.
SNSF Project. Acid precipitation and some alternative sources as the cause of the acidifying
water courses. Norway's Agrarian Science Research Board, Norway's Technical Nature
Science Research Board, Dept. of Environmental Protection, ed., A. Tollan, 1977.
Sprules, G. W. Midsummer crustacean zooplankton communities in acid-stressed lakes. J. Fish.
Res. Board Can. 32:389-395, 1975.
Stensland, G. J. A comparison of precipitation chemistry data at a central Illinois site in
1954 and in 1977. Presented at 71st Annual Meeting, Air Pollution Control. Association,
Houston, TX, June 25-26, 1978. Paper No. 78-48.7.
Stensland, G. J. , Precipitation chemistry trends in the northeastern United States. In:
Polluted Rain, Twelfth International Rochester Conference on Environmental Toxicity,
University of Rochester, Rochester, New York, May 21-23, 1979. T. Y. Toribara, M. W.
Miller, and P. E. Morrow, eds., Environmental Science Research Volume 17, Plenum Press,
New York, NY, 1980. pp. 87-108.
Stokes, P. Phytoplankton of acidic lakes in Killarney Ontario: community structure related
to water chemistry abstracts of voluntary contributions. In: Proc. International Con-
ference on the Ecological Impact of Acid Precipitation, Sandefjord, Norway, March 11-14,
1980, SNSF-project report, Oslo, Norway, 1980.
7-121
-------
Stumm, W., and J. J. Morgan. Aquatic Chemistry, Wiley-Interscience, New York, NY, 1970. 583
pp.
Sutcliffe, D. W., and T. R. Carrick. Studies on mountain streams in the English lake district
I. pH, calcium, and the distribution of invertebrates in the River Duddon. Freshwater
Biol. 3:437-462, 1973.
Tabatabai, M. A., and J. M. Laflen. Nitrogen and sulfur content and pH of precipitation in
Iowa. J. Environ.' Qual. 5:108-112, 1976.
Tamm, C. 0., G. Wiklander, and B. Popovie. Effects of applications of sulphuric acid to poor
pine forests. In: First International Symposium on Acid Precipitation and the Forest
Ecosystem, U. S. Department of Agriculture and Others, Columbus, Ohio, May 12-15, 1975.
Water Air Soil Pollut. 8:75-87, 1977.
Tiano, P., R. Bianchi, G. Gargani, and S. Vannuci. Research on the presence of sulphur-cycle
bacteria in the stone of some historical buildings in Florence. Plant Soil 43:211-217,
1975.
Tomlinson, G. H. Acidic precipitation and mercury In Canadian lakes and fish. In:
Scientific Paper From the Public Meeting on Acid Precipitation, New York State Assembly
Committee on Environmental Conservation and Others, Lake Placid, New York, May 4-5, 1978.
H. H. Izard and J. S. Jacobson, eds., New York State Assembly, Science and Technology
Staff, Albany, NY, March 1979. pp. 104-118.
Traaen, T. Effects of acidity on decomposition or organic matter in aquatic environments. In:
Ecological Impact of Acid Precipitation, Proceedings of an International Conference, SNSF
Project, Sande/jord, Norway, March 11-14, 1980, D. Drablrfs and A. Tollan, eds., SNSF
project, Oslo-As, Norway. October 1980. pp. 340-341.
Trojnar, J. R. Egg and larval survival of white suckers (Catostomus commersoni) at low pH.
J. Fish. Res. Board Can. 34:262-266, 1977a.
Trojnar, J. R. Egg hatchability and tolerance of brook trout (Solvelinus fontinalis) fry at
low pH. J. Fish. Res. Board Can. 34:574-579, 1977b.
Tukey, H. B., Jr. «The leaching of substances from plants. Annu. Rev. Plant Physiol.
21:305-324, 1970.
Turk, J. T., and N. E. Peters. Acid-Rain Weathering of a Metasedimentary Rock Basin, Herkimer
County, New York. Open file Report 77-538, U. S. Department of the Interior, Geological
Survey, Washington, DC, July 1977.
Turk, J. T., and N. Peters. Acid-Rain Weathering of a Metasedimentary Rock Basin, Herkimer
County, New York. In: Scientific Paper From the Public Meeting on Acid Precipitation,
New York State Assembly Committee on Environmental Conservation and Others, Lake Placid,
New York, May 4-5, 1978. H. H. Izard and J. S. Jacobson, eds., New York State Assembly,
Science and Technology Staff, Albany, NY, March 1979. pp. 136-145.
Turner, J., and M. Lambert. Sulfur nutrition of forests. Jn: Atmospheric Sulfur Deposition
Environmental Impact and Health Effects. Proceedings of the Second Life Science Sym-
posium, Oak Ridge National Laboratory and Others, Gatlinburg, Tennessee, October 14-18,
1979. D. S. Shriner, C. R. Richmond, and S. E. Lindberg, eds. , Ann Arbor Science Pub-
lishers, Inc., Ann Arbor, MI, 1980. pp. 321-333.
7-122
-------
Tyler, G. Heavy metals pollute nature, may reduce productivity, Ambio, 1:52-59, 1972.
Tyler, G. Leaching rates of heavy metal ions in forest soil. Water Air Soil Pollut. 9_:
137-148, 1977.
Ulrich, B. Environmental influences ,on the nutrient cycle pf a beech" forest on acid soil.
Forstwiss. Centralbl. 94:280-287, 1975.
U.S. Bureau of Mines. Minerals Yearbook 1952. Volume II. Area Reports: Domestic. U.S.
Department of the Interior, Washington, DC, 1954. p. 117.
U.S. Bureau of Mines. Minerals Yearbook 1974. Volume II. Area Reports: ^Domestic. U.S.
Department of the Interior, Washington, DC, 1976. p. 392.
U.S. Department of Commerce. Climatic Atlas of the United States. National Climatic Center,
Asheville, NC, June 1968.
U.S. Department of Energy. Acid Rain Information Book. Final Report DOE/EP-0018, U. S.
Department of Energy, Washington, DC, May 1981. pp. 2-12; 2-13.
U.S. Environmental Protection Agency. Air Quality Criteria for Oxides of Nitrogen. Draft
Final. EPA-600/8-82-026, U.S. Environmental Protection Agency,.Research Triangle Park,
NC, September, 1982.
Walters, C. J., and R, E. Vincent. Potential productivity of an alpine lake as indicated by
removal and reintroduction of fish. Trans. Amer. Fish Soc. 102:675-697, 1973.
Wetzel, R. G. Limnology, p. 287-621, W. B. Saunders Co., Philadelphia, PA, 1975. 743 pp.
Whelpdale, D. M. Acidic Deposition. Ecological Effects of Acid Precipitation, Report of a
Workshop, Electric Power Research Institute, Gatehouse-of-Fleet, Galloway, Scotland, Sep-
tember 4-7, 1978. M. J. Wood, ed., EPRI SDA77-403, Electric Power Research Institute,
Palo Alto, CA, July 1979. Section 4.1. 22 pp.
Wiklander, L. Interaction between cations and anions influencing adsorption and leaching.
In: Effects of Acid Precipitation on Terrestrial Ecosystems, North Atlantic Treaty
Organization, Toronto, Ontario, Canada, May 21-27, 1978. T. C. Hutchinson and M. Havas,
eds., Plenum Press, New York, NY, 1980. pp. 239-254.
Wiklander, L. Leaching and Acidification of Soils. In: Ecological Effects of Acid Precipi-
tation. Report of a workshop, Electric Power Research Institute, Gatehouse-of-Fleet,
Galloway, Scotland, September 4-7, 1978. M. J. Wood, ed., EPRI SOA77-403, Electric Power
Research Institute, Palo Alto, CA, July 1979. Section 4.3. 24 pp.
Williams, T. Acid rain: the California context. Citizens for Better Environ.: 6-8, 10, May
1978.
Winkler, E. M. Important agents of weathering or building and monumental stone. Eng. Geol.
1:381-400, 1966.
Wisniewski, J. and E. L. Keitz. Acid rain deposition patterns in the continental United
States. Submitted to the Bull. Amer. Meteoro, Soc. 1981.
Wolff, G. T., P. J. Lioy, H. Golub, and J. S. Hawkins. Acid precipitation in the New York
Metropolitan Area: Its relationship to meteorological factors. Env. Sci. Techno!.
13:209-212, 1979.
7-123
-------
Wood, T. Acid Precipitation. In; Sulfur in the Environment. Revised Edition. Missouri
Botanical Garden, St. Louis, MO, July 1975. pp. 39-50.
Wood, T., and F. H. Bormann. The effects of an artificial acid mist upon the growth of Betula
alleghaniensis Britt. Environ. Pollut. 7=259-267, 1974.
Wood, T., and F. H. Bormann. Increases in foliar leaching by acidification of an artificial
mist. Ambio 4:169-171, 1975.
Wood, T., and F. H. Bormann. Short-term effects of a simulated acid rain upon the growth and
nutrient relations of Pinus strobus L. In: First International Symposium on Acid Pre-
cipitation and the Forest Ecosystem, U. S. Department of Agriculture and Others,
Columbus, Ohio, May 12-15, 1975. Water, Air Soil Pollut. 7:479-488, 1977.
Woodwell, G, M. Effects of ionizing radiation on terrestrial ecosystems. Science (Washington,
DC) 138:572-577, 1962.
Woodwell, G. M. Effects of pollution on the structure and physiology of ecosystems. Science
(Washington, DC) 168:429-433, 1970.
Wright, R. F., and E, T. Gjessing. Changes in the chemical composition of lakes. Ambio.
5:219-223, 1976.
Wright, R. F., and A. Henriksen. Chemistry of small Norwegian lakes, with special reference
to acid precipitation. Limnol, Oceanogr. 23:487-498, 1978.
Wright, R. F., and E. Snekvik. Acid precipitation: chemistry and fish populations in 700
lakes in southernmost Norway. Verh. Int. Ver. Theor. Angew. Limnol. 20:765-775, 1978.
Wright, R. F., N. Conroy, W. T. Dickson, R. Harriman, A. Henriksen, and C. L. Schofield.
Acidified lake districts of the world: a comparison of water chemistry of lakes in
southern Norway, southern Sweden, southwestern Scotland, the Adirondack Mountains of New
York, and southeastern Ontario. In: Ecological Impact of Acid Precipitation, Pro-
ceedings of an International Conference, SNSF Project, Sajidefjord, Norway, March 11-14,
1980, D. Drablrfs and A. Tollan, eds., SNSF project, Oslo-As, Norway, October 1980b. pp.
377-379.
Wright, R. F., T. Dale, E. T. Gjessing, G. R. Hendry, A. Henriksen, M. Johannessen, and I. P.
Huniz, Impact of acid precipitation on freshwater ecosystems in Norway. In: First
International Symposium on Acid Precipitation and the Forest Ecosystem, U. S. Department
of Agriculture and Others, Columbus, Ohio, May 12-15, 1975. Water Air Soil Pollut.
6:483-499, 1976.
Wright, R. F., R. Harriman, A. Henriksen, B. Morrison, and L. A. Caines. Acid lakes and
streams in the Galloway area, southwestern Scotland. In: Ecological Impact of Acid
Precipitation, Proceedings of an International Conference, SNSF Project, Sandefjord,
Norway, March 11-14, 1980, D. DrabljJs and A. Tollan, eds., SNSF project, Oslo-As, Norway,
October 1980a. pp. 248-249.
Van, N. D., and P. Stokes. Phytoplankton of an acidic lake, and its responses to experimental
alterations of pH. Environ. Conserv. 5:93-100, 1978. '
7-124
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing]
1 REPORT NO
EPA-600/8-82-029b
3 RECIPIENT'S ACCESSION NO
4 TITLE AND SUBTITLE
Air Quality Criteria for Participate Matter
and Sulfur Oxides. Volume II.
5 REPORT DATE
December 1982
6 PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
See list of Authors, Contributors, and Reviewers
8 PERFORMING ORGANIZATION REPORT NO
9 PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Environmental Criteria and Assessment Office
MD-52
Research Triangle Park, NC 27711
10 PROGRAM ELEMENT NO
11 CONTRACT/GRANT NO
12 SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
Office of Health and Environmental Assessment
401 M Street, SH, Washington, DC 20460
13 TYPE OF REPORT AND PERIOD COVERED
FINAL
14 SPONSORING AGENCY CODE
EPA/600/00
15. SUPPLEMENTARY NOTES
16 ABSTRACT
The document evaluates and assesses scientific information on the health and welfare
effects associated with exposure to various concentrations of sulfur oxides and
particulate matter in ambient air. The literature through 1980-81 has been reviewed
thoroughly for information relevant to air quality criteria, although the document
is not intended as a complete and detailed review of all literature pertaining to
sulfur oxides and particulate matter. An attempt has been made to identify the major
discrepancies in our current knowledge and understanding of the effects of these
pol1utants.
Although this document is principally concerned with the health and welfare effects of
sulfur oxides and particulate matter, other scientific data are presented and evalu-
ated in order to provide a better understanding of these pollutants in the environment
To this end, the document includes chapters that discuss the chemistry and physics
of the pollutants; analytical techniques; sources; and types of emissions; environ-
mental concentrations and exposure levels; atmospheric chemistry and dispersion
modeling; acidic deposition; effects on vegetation; effects on visibility, climate,
and materials; and respiratory, physiological, toxicological, clinical and
epidemiological aspects of human exposure.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
c cos ATI field/Group
18 DISTRIBUTION STATEMENT
RELEASE UNLIMITED
19 SECURITY CLASS /This Report)
UNCLASSIFIED
21 NO OF PAGES
625
20 'SFCUJRITY CLASS fThlSO&tei
UNCLASSIFIED
22. PRICE
iPA Farm 2220-1 (Rsv, 4-77) PREVIOUS EDITION is OBSOLETE
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Center for Environmental Research Information
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE. S3OO
AN EQUAL OPPORTUNITY EMPLOYER
If your address is incorrect, please change on the above label
tear off; and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HEREd; fear off label, and return it to the
above address.
Publication No. EPA-600/8-82-029b
-------
|