Air Quality Criteria for Particulate Matter and Sulfur Oxides, Volume 3


&EPA
              United States
              Environmental Protection
              Agency
              Environmental Criteria and
              Assessment Office
              Research Triangle Park NC 27711
EPA-60O/8-82-029c
December 1982
              Research and Development
Air Quality Criteria for
Paniculate Matter and
Sulfur Oxides
              Volume III

-------
                         EPA-600/8-82-029c

                             December 1982
  Air Quality Criteria
for Particulate Matter
  and Sulfur Oxides

        Volume
  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  III  of a  three-volume revision  of  Air Quality
Criteria for  Particulate Hatter 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 this volume (Volume III) of the document.   To assist
the  reader  in putting the effects  into  perspective with  the  real-world en-
vironment,   Chapters  2  through  7 in another volume (Volume II) of the document
have been prepared  and discuss:  physical and chemical  properties;  air moni-
toring  and  analytical measurement  techniques;  sources  and  emissions;  trans-
port,  transformation,  and  fate;  and observed  ambient concentrations  of the
pollutants..
     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.
                                      m

-------
                                  VOLUME III

                                   CONTENTS
8.  EFFECTS ON VEGETATION.	     8-1
    8.1  GENERAL INTRODUCTION AND APPROACH	     8-1
    8.2  REACTION OF PLANTS TO SULFUR DIOXIDE EXPOSURES.	     8-2
         8.2.1  Introduction	     8-2
         8.2.2  Wet and Dry Deposition of Sulfur Compounds on
                Leaf Surfaces	,.	,     8-3
         8.2.3  Routes and Methods of Entry Into the Plant	     8-3
         8.2.4  Cellular and Biochemical Changes	     8-5
         8.2.5  Beneficial  "Fertilizer" Effects	     8-7
         8.2.6  Acute Foliar Injury	    8-10
         8.2.7  Chronic Foliar Injury	    8-10
         8.2.8  Foliar Versus Whole Plant Responses	    8-11
         8.2.9  Classification of Plant Sensitivity to Sulfur Dioxide.    8-13
    8.3  EXPOSURE-RESPONSE RELATIONSHIPS - SULFUR DIOXIDE	    8-13
    8.4  EFFECTS OF MIXTURES OF SULFUR DIOXIDE AND OTHER POLLUTANTS...    8-28
         8.4.1  Sulfur Dioxide and Ozone	    8-28
         8.4.2  Sulfur Dioxide and Nitrogen Dioxide	    8-30
         8.4.3  Sulfur Dioxide and Hydrogen Fluoride	    8-31
         8.4.4  Sulfur Dioxide, Nitrogen Dioxide and Ozone	    8-31
         8.4.5  Summary	    8-31
    8.5  EFFECTS OF NON-POLLUTANT ENVIRONMENTAL FACTORS ON SULFUR
         DIOXIDE PLANT EFFECTS	    8-32
         8.5.1  Temperature	    8-32
         8.5.2  Relative Humidity	    8-32
         8.5.3  Light	    8-33
         8.5.4  Edaphic Factors	    8-33
         8.5.5  Sulfur Dioxide and Biotic Plant Pathogen Interactions.    8-34
    8.6  PLANT EXPOSURE TO PARTICULATE MATTER	    8-34
         8.6.1  Deposition Rates....	    8-34
         8.6.2  Routes and Methods of Entry Into Plants	    8-35
                8.6.2.1  Direct Entry Through Foliage	    8-35
                8.6.2.2  Indirect Entry Through Roots	    8-36
    8.7  REACTION OF PLANTS TO PARTICLE EXPOSURE	    8-36
         8.7.1  Symptomatology of Particle-Induced Injury	    8-36
         8,7,2  Classification of Plant Sensitivity—Particles	    8-41
    8.8  EXPOSURE-RESPONSE RELATIONSHIPS—PARTICLES..			    8-41
    8.9  INTERACTIVE EFFECTS ON PLANTS WITH THE ENVIRONMENT—
         PARTICULATE MATTER	    8-42
    8.10 EFFECTS OF SULFUR DIOXIDE AND PARTICULATE MATTER ON NATURAL
         ECOSYSTEMS	•	    8-43
         8.10.1 Sulfur Dioxide in Terrestrial Ecosystems.	    8-43
         8.10,2 Ecosystem Response to Sulfur Dioxide	    8-46
         8.10.3 Response of Natural Ecosystems to Particulate Matter..    8-54
    8.11 SUMMARY.	,	    8-56
    8.12 REFERENCES			    8-60

-------
                             CONTENTS (continued)



                                                                          Page

    APPENDIX 8-A	•	     8-78


 9.  EFFECTS ON VISIBILITY AND CLIMATE	      9-1
     9.1  INTRODUCTION	      9-1
     9.2  FUNDAMENTALS OF ATMOSPHERIC VISIBILITY	      9-2
          9.2.1  Physics of Light Extinction	     9-11
          9.2.2  Measurement Methods	     9-14
                 9.2.2.1  Human Observer (Total Extinction)	     9-16
                 9.2.2.2  Photography (Total Extinction)	     9-16
                 9.2.2.3  Telephotometry (Total Extinction)	     9-17
                 9.2.2.4  Long-path Extinction (Total Extinction)	     9-17
                 9.2.2.5  Nephelometer (Scattering)	     9-18
                 9.2.2.6  Light Absorption Coefficient	     9-18
          9.2.3  Role of Participate Matter in Visibility Impairment..     9-19
                 9.2.3.1  Ray!eigh Scattering	     9-19
                 9.2.3.2  Nitrogen Dioxide Absorption	     9-20
                 9.2.3.3  Particle Scattering	     9-20
                 9.2.3.4  Particle Absorption	     9-30
          9.2.4  Chemical Composition of Atmospheric Particles	....     9-31
                 9,2.4.1  Role of Water in Visibility Impairment	     9-34
                 9.2.4.2  Light Extinction Budgets	     9-38
          9.2.5  Considerations in Establishing a Quantitative
                 Relationship Between Fine-Particle Mass Concentration
                 and Visual Range	     9-40
     9.3  VISIBILITY AND PERCEPTION	     9-43
     9.4  HISTORICAL PATTERNS OF VISIBILITY	     9-49
          9.4.1  Natural Versus Manmade Causes	     9-63
     9.5  THE EVALUATION OF IMPAIRED VISIBILITY	     9-66
          9.5.1  Social Awareness and Aesthetic Considerations	     9-67
          9.5.2  Economic Considerations	,     9-68
          9.5.3  Transportation Operations	     9-71
     9.6  SOLAR RADIATION	     9-75
          9.6.1  Spectral and Directional Quality of Solar Radiation..     9-86
          9.6.2  Total Solar Radiation:   Local to Regional Scale......     9-92
          9.6.3  Radiative Climate:  Global Scale	     9-94
     9.7  CLOUDINESS AND PRECIPITATION	     9-95
     9.8  SUMMARY	     9-97
     9.9  REFERENCES	     9-100

10.  EFFECTS ON MATERIALS.	     10-1
     10.1 INTRODUCTION.	     10-1
     10.2 SULFUR OXIDES	/	     10-4
          10.2.1 Corrosion of Exposed Metals	     10-4
                 10.2.1.1 Physical and Chemical Considerations	     10-4
                 10.2.1.2 Effects of Sulfur Oxide Concentrations
                          on the Corrosion of Exposed Metals	     10-12
          10.2.2 Protective Coatings	     10-23
                 10.2.2.1 Zinc-Coated Materials	     10-23
                 10.2.2.2 Paint Technology and Mechanisms of Damage...     10-28
          10.2.3 Fabrics	     10-32
                                      VI

-------
                             CONTENTS (continued)
          10.2.4 Building Materials	    10-34
                 10.2.4.1 Stone	    10-34
                 10.2.4.2 Cement and Concrete	    10-35
          10.2.5 Electrical Equipment and Components...	    10-37
          10.2.6 Paper	    10-37
          10.2.7 Leather	    10-37
          10.2.8 Elastomers and Plastics		    10-38
          10.2.9 Works of Art	    10-38
         10.2.10 Review of Damage Functions Relating Sulfur Dioxide
                 to Material Damage	    10-39
     10.3 PARTICULATE MATTER	    10-41
          10.3.1 Corrosion and Erosion	    10-41
          10.3.2 Soiling and Discoloration	    10-42
                 10.3.2.1 Building Materials	    10-43
                 10.3.2.2 Fabrics	    10-45
                 10.3.2.3 Household and Industrial Paints	    10-45
     10.4 SUMMARY, PHYSICAL EFFECTS OF SULFUR OXIDtS AND PARTICULATE
          MATTER ON MATERIALS	    10-47
     10.5 ECONOMIC ESTIMATES.			    10-49
          10.5.1 Introduction	    10-49
          10.5.2 Economic Loss Associated with Materials Damage and
                 Soiling	    10-50
                 10.5.2.1 Metal Corrosion and Other Damage to
                          Materials Associated with Sulfur Oxides	    10-50
                 10.5.2.2 Soiling of Paint and Other Materials
                          Associated with Particulate Matter	    10-54
                 10.5.2.3 Combined Studies....	    10-64
          10.5.3 Estimating Benefits from Air Quality Improvement,
                 1970-1978	    10-70
          10.5.4 Summary of Economic Damage of Particulate Matter/
                 Sulfur Oxides to Materials.	    10-73
     10.6 SUMMARY AND CONCLUSIONS, EFFECTS ON MATERIALS	    10-74
     10. 7 REFERENCES	    10-75

11.  RESPIRATORY TRACT DEPOSITION AND FATE OF INHALED AEROSOLS AND
     SULFUR DIOXIDE	     11-1
     11.1 INTRODUCTION	     11-1
          11.1.1 General Considerations	     11-1
          11.1.2 Aerosol and Sulfur Dioxide Characteristics..	     11-2
          11.1.3 The Respiratory Tract	     11-4
          11.1.4 Respiration and Other Factors	     11-7
          11.1.5 Mechanisms of Particle Deposition	    11-12
     11.2 DEPOSITION IN MAN AND EXPERIMENTAL ANIMALS	    11-16
          11.2.1 Insoluble and Hydrophobic Solid Particles	    11-16
                 11.2.1.1 Total Deposition	    11-16
                 11.2.1.2 Extrathoracic Deposition	    11-20
                 11.2.1.3 Tracheobronchial Deposition	    11-23
                 11.2.1.4 Pulmonary Deposition	    11-27
                 11.2.1.5 Deposition in Experimental Animals	    11-29
          11.2.2 Soluble, Deliquiscent, and Hygroscopic Particles	    11-32
          11.2.3 Surface-Coated Particles.		    11-33
                                      VII

-------
                             CONTENTS (continued)
          11.2.4 Gas Deposition		    11-33
          11.2.5 Aerosol-Gas Mixtures	    11-37
     11.3 TRANSFORMATIONS AND CLEARANCE FROM THE RESPIRATORY TRACT....    11-38
          11.3.1 Deposited Particulate Material	    11-39
          11.3.2 Absorbed Sulfur Dioxide	    11-47
          11.3.3 Particles and Sulfur Dioxide Mixtures	    11-48
     11.4 AIR SAMPLING FOR HEALTH ASSESSMENT.	    11-48
     11.5 SUMMARY	    11-54
     11.6 REFERENCES	    11-57


12.  TOXICOLOGICAL STUDIES	     12-1
     12.1 INTRODUCTION	     12-1
     12.2 EFFECTS OF SULFUR DIOXIDE	     12-2
          12.2.1 Biochemistry of Sulfur Dioxide	     12-2
                 12.2.1.1 Chemical Reactions of Bisulfite with
                          Biological Molecules.		     12-3
                 12.2.1.2 Metabolism of Sulfur Dioxide	     12-5
                 12.2.1.3 Activation and Inhibition of Enzymes by
                          Bisulfite	     12-6
          12.2.2 Mortal ity	     12-7
          12.2.3 Morphological Alterations	     12-8
          12,2.4 Alterations in Pulmonary Function	    12-13
          12.2.5 Effects on Host Defenses	    12-19
     12.3 EFFECTS OF PARTICULATE MATTER	    12-22
          12.3.1 Mortal ity	    12-24
          12.3.2 Morphological Alterations	    12-24
          12.3.3 Alterations in Pulmonary Function	    12-28
                 12.3.3.1 Acute Exposure Effects		    12-28
                 12.3.3.2 Chronic Exposure Effects	    12-39
          12.3.4 Alteration in Host Defenses	    12-41
                 12.3.4.1 Mucociliary Clearance	    12-41
                 12.3.4.2 Alveolar Macrophages	    12-46
                 12.3.4.3 Interaction with Infectious Agents		    12-51
                 12.3.4.4 Immune Suppression.	    12-53
     12.4 INTERACTION OF SULFUR DIOXIDE AND OTHER POLLUTANTS		    12-54
          12.4.1 Sulfur Dioxide and Particulate Matter	    12-54
                 12.4.1.1 Acute Exposure Effects	"	    12-54
                 12.4.1.2 Chronic Exposure Effects	    12-56
          12,4.2 Interaction with Ozone		    12-63
     12.5 CARCINOGENESIS AND MUTAGENESIS OF SULFUR COMPOUNDS AND
          ATMOSPHERIC PARTICLES.	    12-66
          12.5.1 Airborne Particulate Matter	    12-68
                 12.5.1.1 Jji vitro Mutagenesis Assays of Particulate
                          Matter	    12-68
                 12.5.1.2 Tumorigenesis of Particulate Extracts	    12-70
          12.5.2 Potential Mutagenic Effects of Bisulfite and Sulfur
                 Dioxide	    12-72
          12.5.3 Tumorigenesis in. Animals Exposed to Sulfur Dioxide
                 or Sul fur Dioxide and Benzo(a)pyrene	    12-74
          12.5.4 Effects of Trace Metals Found in Atmospheric
                 Particles	    12-75
                                     vm

-------
                             CONTENTS (continued)
     12. 6 CONCLUSIONS	     12-75
          12.6.1 Sulfur Dioxide	     12-75
          12.6.2 Particulate Matter	     12-78
          12.6.3 Combinations of Gases and Particles	     12-81
     12. 7 REFERENCES	     12-83

     APPENDIX 12-A:   U.S.  EPA Analysis'of the Laskin et al.  and
         and Peacock and Spence Data (Memo from V.  Hasselblad
         to L. D. Grant)	    12-102

13.   CONTROLLED HUMAN STUDIES	      13-1
     13.1 INTRODUCTION	:	      13-1
     13. 2 SULFUR DIOXIDE	     13-2
          13.2.1 Subjective Reports	      13.2
          13.2.2 Sensory Effects	      13-3
                 13.2.2.1 Odor Perception Threshold	      13-3
                 13.2.2.2 Sensitivity of the Dark-Adapted Eye	      13-5
                 13.2.2.3 Interruption of Alpha Rhythm	      13-6
          13.2.3 Respiratory and Related Effects	      13-6
                 13.2.3.1 Respiratory Function	      13-6
                 13.2.3.2 Water Solubility	     13-12
                 13.2.3.3 Nasal Versus Oral  Exposure	     13-12
                 13.2.3.4 Subject Activity Level	     13-13
                 13.2.3.5 Temporal Parameters	     13-15
                 13.2.3.6 Mucociliary Transport	     13-17
                 13.2.3.7 Health Status	     13-19
     13.3 PARTICULATE MATTER	     13-23
          13.3.1 Sulfuric Acid and Sulfates	     13-23
                 13.3.1.1 Sensory Effects	     13-23
                 13.3.1.2 Respiratory and Related Effects	     13-24
          13.3.2 Insoluble and Other Non-sulfur Aerosols	     13-31
     13.4 PARTICULATE MATTER AND SULFUR DIOXIDE	     13-36
     13.5 SULFUR DIOXIDE, OZONE, AND NITROGEN DIOXIDE	     13-39
     13.6 SUMMARY AND CONCLUSIONS	     13-46
          13.6.1 Sulfur Dioxide Effects	     13-47
          13.6.2 Sulfuric Acid and Sulfate Effects	     13-52
          13.6.3 Effects of Other Particulate Matter Species	     13-53
     13. 7 REFERENCES	     13-55

     APPENDIX 13A	     13-62

14.   EPIDEMIOLOGICAL STUDIES ON THE EFFECTS OF PARTICULATE MATTER AND
     SULFUR OXIDES ON HUMAN HEALTH	      14-1
     14.1 INTRODUCTION	      14-1
          14.1.1 Methodological Considerations	      14-2
          14.1.2 Guidelines for Assessment of Epidemiological Studies.      14-5
     14. 2 AIR QUALITY MEASUREMENTS	      14-7
          14.2.1 Sulfur Oxides Measurements	      14-7
          14.2.2 Particulate Matter Measurements	      14-8
     14.3 ACUTE PARTICULATE MATTER/SULFUR OXIDE EXPOSURE EFFECTS	     14-11
          14. 3.1 Mortal ity	     14-11

-------
                        CONTENTS (continued)
            14.3.1.1 Acute Episode Studies	    14-11
            14.3.1.2 Mortality Associated with Non-episodic
                     Variations in Pollution	    14-15
            14.3.1.3 Morbidity	.,.	    14-26
14.4 CHRONIC PM/S02 EXPOSURE EFFECTS.	    14-35
     14.4.1 Mortal i ty	    14-35
     14.4.2 Morbidity	    14-44
            14.4.2.1 Respiratory Effects in Adults	    14-44
            14.4.2.2 Respiratory Effects in Children	    14-46
14.5 SUMMARY AND CONCLUSIONS	    14-49
     14.5.1 Health Effects Associated with Acute Exposures to
            Particulate Matter and Sulfur Oxides	    14-50
     14.5.2 Health Effects Associated with Chronic Exposures to
            Particulate Matter and Sulfur Oxides	    14-53
14.6 REFERENCES		....    14-56

APPENDIX 14-A:  Annotated Comments on Community Health Epidemio-
    logical Studies Not Discussed in Detail in Main Text of
    Chapter 14.	    14-73
APPENDIX 14-B:  Occupational Health Studies on Particulate Matter
    and Sulfur Oxides	   14-102

APPENDIX 14-C:  Summary of Examples of Sources and Magnitudes of
    Measurement Errors Associated with Aerometric Measurements
    of PM and S02 Used in British and American Epidemiologies!
    Studies	   14-107
APPENDIX 14-D:  EPA Reanalysis of Martin and Bradley (1960) Data
    on Mortal i ty During 1958-59 London Winter	   14-116

APPENDIX 14-E:  Summary of Unpublished Dawson and Brown (1981) Re-
    analysis of Martin and Bradley (1960) Data		   14-126

APPENDIX 14-F:  Summary of Unpublished Roth et al. (1981) Year-
    by-Year Analysis of London Mortality Data for Winters of
    1958-59 to 1971-72	   14-135
APPENDIX 14-G:  Summary of Mazumdar et al.  Year-by-Year
    Analysis of London Mortality Data for Winters of
    1958-59 to 1971-72	   14-138

-------
                             CONTENTS (continued)




                                   FIGURES


Figure                                                                     Page

8-1   Map of the United States indicating major areas of sulfur-
      deficient soils	       8-8
8-2   Conceptual model of the factors involved in air pollution
      effects (dose-response) on vegetation	      8-16
8-3 .  The sulfur cycle	      8-45


9-1   Map shows median yearly visual range (miles) and isopleths for
      suburban/nonurban areas, 1974-76	       9-3
9-2   Median summer visual range (miles) and isopleths for suburban/
      nonurban areas, 1974-76	.'	       9-3
9-3   (A) A schematic representation of atmospheric extinction,
      illustrates (i) transmitted, (ii) scattered, and (iii) absorbed
      light,  (B) A schematic representation of daytime visibility
      illustrates: (i) light from target reaching observer,
      (ii) light from target scattered out of observer's line^of
      sight, (iii) air light from intervening atmosphere, and
      (iv) air light constituting horizon sky	       9-4
9-4   The apparent contrast between object and horizon sky decreases
      with increasing distance from the target.   This is true for
      both bright and dark objects	       9-5
9-5   Mean contrast threshold of the human eye for 50% detection
      probability as a function of target angular diameter and adaption
      brightness (candles/m ) for targets brighter than their background.
      Daytime adaptation brightness is usually in the range 100 to '
      10,000 candles/m2	       9-8
9-6   Inverse proportionality between visual range and the scattering
      coefficient, a  , as measured at the point of observation	      9-10

9-7   Extinction efficiency factor (Q  .) of a single spherical

      particle as a function of diameter for a non-absorbing par-
      ticle of refractive index (1,5-O.Oi) and wavelength 0.55 urn......      9-12
9-8   Extinction efficiency factor (Qext) of a single spherical

      particle as a function of diameter for an absorbing particle
      of refractive index (2.0, -1.0) and wavelength 0.55 jjm	      9-12
9-9   For a light-scattering and absorbing particle, the scattering
      per volume concentration has a strong peak at particle
      diameter of 0.5 pm (m = 1.5-0.05i; wavelength = 0.55 pm).
      However, the absorption per aerosol volume is only weakly
      dependent on particle size.   Thus the light extinction by
      particles with diameter less than 0.1 |jm is primarily due to
      absorption.  Scattering for such particles is very low.  A
      black plume of soot from an oil burner is a practical
      example	     9-22
9-10  (A) Calculated scattering coefficient per unit mass
      concentration at a wavelength of 0.55 ym for absorbing and
      nonabsorbing materials is shown as a function of diameter for
      single-sized particles	 —	     9-23


                                      xi

-------
                             CONTENTS (continued)
Figure                                                                     Page

9-11  For a typical aerosol volume (mass) distribution, the calculated
      light-scattering coefficient is contributed almost entirely by
      the size range 0.1-1.0 pm.  The_total a   and total aerosol

      volume concentration are proportional to the area under the
      respective curves	.	    9-24
9-12  Scattering-to-volume concentration ratios are given for various
      size distributions.  The ratio values for accumulation (fine)
      and coarse modes are shown by dashed lines corresponding to
      average empirical size distributions reported by Whitby and
      Sverdrup (1980)	    9-26
9-13  Simultaneous in situ monitoring of a   and fine-particle mass

      concentration in St. Louis in April 1973 showed a high correla-
      tion coefficient of 0.96, indicating that cr   depends primarily
      on the fine-particle concentration	fv	    9-28
9-14  Aerosol mass distributions, normalized by the total mass, for
      New York aerosol at different levels of light-scattering
      coefficient show that at high background visibility, the fine-
      particle mass mode is small compared with the coarse-particle
      mode.  At the low visibility level, C, 60 percent of the mass
      is due to fine particles.	    9-29
9-15  Humidograms for a number of sites show the increase in a

      which can be expected at elevated humidities for specific sites
      or aerosol types (marine, Point Reyes, CA; sulfate, Tyson, MO)
      and the range observed for a variety of urban and rural sites
      (composite)	    9-35
9-16  Relative size growth as a function of relative humidity for an
      ammonium sulfate particle at 25°C	.	    9-37
9-17  Fine mass concentration (determined from equilibrated filter)
      corresponding to 4.8 km visual range, as a function of K and y,
      where K equals the Koschmieder constant (-log  e), and y equals
      a   + cr/fine mass concentration.	    9-44
9-18  Visual range as a function of fine mass concentration (deter-
      mined from equilibrated filter) and y, assuming K = 3.9	    9-45
9-19  Historical trends in hours of reduced visibility at Phoenix
      and Tucson are compared with trends in SO  emissions from
      Arizona copper smelters	    9-50
9-20  Seasonally adjusted changes in sulfate during the copper strike
      are compared with the geographical distribution of smelter SO
      emissions	    9-51
9-21  Seasonally adjusted percent changes in visibility during the
      copper strike are compared with the geographical  distribution of
      smelter SO  emissions	    9-52
9-22  The locations of sampling sites and smelters and the mean surface
      wind vectors at each sampling site from August 1979 through
      September 1980.	    9-55
9-23  Particle light extinction (a _ + cr  ) budget for the low visibil-
                                  sp    ap
      ity southern California incursion (June 30) and a clear day
      (July 10)	    9-56
                                      Xll

-------
                             CONTENTS (continued)
Figure                                                                     Page

9-24  Compared here are summer trends of U.S.  coal ^consumption
      and Eastern United'States extinction coefficient,	     9-58
9-25  In the 1950's, the seasonal coal  consumption peaked in the winter
      primarily because of increased residential  and railroad use.   By
      1974, the seasonal pattern of coal usage was determined by the
      winter and summer peak of utility coal usage.  The shift away
      from a winter peak toward a summer peak in coal consumption is
      consistent with a shift in extinction coefficient from a winter
      peak to a summer peak in Dayton,  OH, for 1948-52	     9-58
9-26  In 1974, the United States winter coal consumption was well
      below, while the summer consumption was above, the 1943 peak.
      Since 1960 the average growth rate of summer consumption was 5.8
      percent per year, while the winter consumption increased at only
      2.8 percent per year	     9-59
9-27  Trends in the light extinction coefficient (a  .) in the Eastern

      United States are shown by region and by quarters; 1 (winter), 2
      (spring), 3 (summer), 4 (fall).	     9-60
9-28  The spatial distribution of 5-year average extinction coeffi-
      cients shows the substantial increases of third-quarter extinc-
      tion coefficients in the Carolines, Ohio River Valley, and
      Tennessee-Kentucky area	     9-62
9-29  Average annual number of days with occurrence of dense fog.
      Coastal and mountainous regions are most susceptible to fog	..     9-65
9-30  Annual percent frequency of occurrence of wind-blown dust
      when prevailing visibility was 7 miles or less, 1940-70.
      Dust is a visibility problem in the Southern Great Plains
      and Western desert regions	     9-65
9-31  Percent of daily midday measurements (1971-75) in which
      visibilities were three miles or less in the absence of fog,
      precipitation, or blowing material		     9-76
9-32  Percent of daily midday measurements (1976-80) in which visibi-
      lities were three miles or less in the absence of fog,
      precipitation, or blowing material		     9-77
9-33  Solar radiation intensity spectrum at sea level in cloudless
      sky peaks in the visible window,  0.4-0.7 pm wavelength range,
      shows that in clean remote locations, direct solar radiation
      contributes 90 percent and the skylight 10 percent of the
      incident radiation on a horizontal surface	     9-85
9-34  Extinction of direct solar radiation by aerosols is depicted	     9-87
9-35t  On a cloudless but hazy day in Texas, the direct solar radiation
      intensity was measured to be half that on a clear day, but most
      of the lost direct radiation has reappeared as skylight	     9-88
9-36  To interpret these 1961-66 monthly average turbidity data in
      terms of aerosol effects on transmission of direct sunlight, use
                              •~R
      the expression I/I  = 10  , where 8 is turbidity and I/I  is the

      fraction transmitted.	     9-90
9-37  Seasonal turbidity patterns for 1961-66 and 1972-75 are shown for
      selected regions in the Eastern United States	     9-91
9-38  Analysis of the hours of solar radiation since the 1950's shows
      a decrease of summer solar radiation over the Eastern United
      States.  There may be several causes for this trend, including
      an increase of cloudiness; some of the change may also be due
      to haze.	     9-93

                                     xiii,

-------
                             CONTENTS (continued)



F1gure                                                                     Page

9-39  Numbers of smoke/haze days are plotted per 5 years at Chicago,
      with values plotted at end of 5-year period	     9-96
10-1  Relationship among emissions, air quality, damages and
      benefits, and policy decisions,,	     10-2
10-2  Steel corrosion behavior is shown as a function of average
      relative humidity at three average concentration levels of
      sulfur dioxide.	     10-6
10-3  Steel corrosion behavior is shown as a function of average
      sulfur dioxide concentration and average relative humidity	     10-7
10-4  Empirical relationship between average relative humidity and
      fraction of time relative humidity exceeded 90 percent (time
      of wetness) is shown for data from St. Louis International
      Airport	     10-9
10-5  Relationship between corrosion of mild steel and corresponding
      mean SQ% concentration is shown for seven Chicago sites.   (Corro-
      sion is expressed as weight loss of panel)	,	     10-17
10-6  Adsorption of sulfur dioxide on polished metal surfaces is
      shown at 90 percent RH	     10-22
10-7  Relationship between retained breaking strength of cotton fabrics
      and corresponding mean sulfation rate measured at selected sites
      in St.  Louis area		     10-33
10-8  Dust deposit patterns with corresponding coverage (% surface
      covered) are shown	     10-44
10-9  Representation of soiling of acrylic emulsion house paint
      as a function of exposure time and particle concentrations.	     10-48
10-10 Increases in particulate matter concentrations are plotted
      against reductions in outdoor cleaning task benefits (1978
      dollars).  The range of benefits increases progressively as
      pollution is reduced	     10-63
10-11 Improvement in U.S.  annual average SOg levels from 32 pg/m3
      in 1970 to 18 pg/m3 in 1978 has resulted in approximately $0.4
      billion in estimated economic benefit for 1978	     10-72

11-1  Features of the respiratory tract of man used in the description
      of inhaled particles and gases with insert showing parts of a
      silicon rubber cast of a human being showing some separated
      bronchioles to 3 mm diameter, some bronchioles from 3 mm
      diameter to terminal bronchioles, and some separated respiratory
      acinus bundles	   f 11-5
11-2  Representation of five major mechanisms of deposition of inhaled
      airborne particles in the respiratory tract	     11-13
11-3  Deposition of monodisperse aerosols in the total respiratory
      tract for nasal breathing in humans as a function of aerody-
      namic diameter, except below 0.5 (JOT, where deposition is plotted
      vs. physical diameter.	     11-18
11-4  Deposition of monodisperse aerosols in the total respiratory
      tract for mouth breathing as a function of aerodynamic diameter,
      except below 0.5 pm, where deposition is plotted vs. physical
      diameter	     11-19
11-5  Deposition of monodisperse aerosols in extrathoracic region for
      nasal breathing as a function of D2Q, where Q is the average
      inspiratory flowrate in 1 iters/min	     11-21


                                      xiv

-------
                             CONTENTS (continued)
Figure                                                                     Page

11-6  Deposition of monodisperse aerosols in extrathoracic region for
      mouth breathing in humans as a function of D2Q, where Q is the
      average inspiratory flowrate in liters/min	      11-22
11-7  Deposition of monodisperse aerosols in the tracheobronchial
      region for mouth breathing in humans in percent of the aerosols
      entering the trachea as a function of aerodynamic diameter,
      except below 0.5 urn, where deposition is plotted vs. physical
      diameter as cited by different investigators,	     11-24
11-8  Total and regional depositions of mono-disperse aerosols with
      mouth breathing as a function of the aerodynamic diameter for
      three individual subjects as cited by Stahlhofen et al.  (1980)...     11-26
11-9  Deposition of monodisperse aerosols in the pulmonary region for
      mouth breathing in humans as a function of aerodynamic diameter,
      except below 0.5 urn, where deposition is plotted vs. physical
      di ameter		     11-28
11-10 Deposition of inhaled polydisperse aerosols of lanthanum oxide
      (radio-labeled with 140La) in beagle dogs exposed in a nose-
      only exposure apparatus showing the deposition fraction of
      (A) total dog, (B) tracheobronchial regions (C) pulmonary
      alveolar region, and (D) extrathoracic region	     11-30
11-11 Deposition of inhaled monodisperse aerosols of ,fused alumino-
      silicate spheres in small rodents showing the deposition in the
      extrathoracic (ET) region, the tracheobronchial (TB) region,
      the pulmonary (P) region, and in the total respiratory tract.....     11-31
11-12 Single exponential model, fit by weighted least-squares, of the
      buildup (based on text equation 7) and retention (based on text
      equation 9) of zinc in rat lungs	     11-45
11-13 Comparison of sampler acceptance of BMRC and ACGIH conventions
      with the band for the experimental pulmonary deposition data of
      Figure 11-9	     11-51
11-14 Division of the thoracic fraction of deposited particles into
      pulmonary and tracheobronchial fractions for two sampling con-
      ventions (ACGIH and BMRC) as a function of aerodynamic diameter,
      except below 0.5 urn, where physical diameter is used (Interna-
      tional Standards Organization, 1981).  Also shown are bands for
      experimental pulmonary deposition data from Figure 11-9 and for
      tracheobronchial (TB) deposition as a percent of particles
      entering the month	      11-53

14-1  Martin and Bradley (1960) data as summarized by Ware et al.
      (1981) showing average deviations of daily mortality from 15-day
      moving average by concentration of smoke (BS) and S02 (London,
      November I, 1958 to January 1, 1959).	      14-17
14-2  Linear and quadratic dose-response curves plotted on the
      scattergram of mortality and smoke for London winters 1958-59
      to 1971-72	     14-22
14-3  Hypothetical dose-response curves derived from regressing
      mortality on smoke in London, England, during winters 1958-59
      to 1971-72	•	     14-23
14-4  History and clinical evidence of respiratory disease (percent) in
      5-year-olds, by pollution in area of residence	     14-48
                                       xv

-------
                             CONTENTS (continued)



                                    TABLES
                                                                 *
Tab!e                                                                       Page

 8-1   Relationship of biochemical response to visual symptoms of
       plant injury.	     8-6
 8-2   Sensitivity groupings of vegetation based on visible injury
       at different SOg exposures	    8-14
 8-3   Effects of exposure to S02 on plants under field
       conditions	,	.	.....*    8-20
 8-4   The degree of injury of eastern white pine observed at various
       distances from the Sudbury smelters for 1953-63..	    8-22
 8-5   Ambient exposures to sulfur dioxide that caused injury
       to vegetation	    8-23
 8-6   Summary of the effects resulting from the exposure of
       seedling tree species in the laboratory	    8-26
 8-7   Plants sensitive to heavy metals, arsenic, and boron as
       accumulated in soils and typical symptoms expressed  	    8-39

 9-1   Particle light scattering coefficient per unit fine-mass
       concentration	    9-27
 9-2   Median percent frequency of occurrence of selected RH classes
       for 54 stations in the contiguous U.S	.		    9-42
 9-3   Correlation/regression analysis between airport extinction
       and copper smelter SO  emissions	.-.-.	    9-53
 9-4   Seasonal average percent of time when midday visibility was
       3 miles (4.8 km) or less at U.S. airports from 1951 to 1980	    9-74
 9-5   Percent of visibility measurements at 3 miles (4.8 km) or less
       at 26 U.S. airports during the summer quarter	    9-78
 9-6   Some solar radiation measurements in the Los Angeles area	    9-92

10-1   Some empirical expressions for corrosion of exposed ferroalloys.   10-20
10-2   Critical humidities for various metals	   10-21
10-3   Experimental regression coefficients with estimated standard
       deviations for small zinc and galvanized steel specimens
       obtained from six exposure si tes	    10-25
10-4   Corrosion rates of zinc on galvanized steel products exposed to
       various environments prior to 1954	    10-26
10-5   Paint erosion rates and t-test probability data	    10-31
10-6   Mechanisms contributing to stone decay.  Principal atmospheric
       factors participating these mechanisms are denoted by solid
       circles:  secondary factors are indicated by solid triangles....    10-36
10-7   Selected physical damage functions related to 50% exposure	    10-40
10-8   Results of regression for soiling of building materials as a
       function of TSP exposure	    10-46
10-9   Summation of annual extra losses due to corrosion damage by air
       pol1ution to external metal structures for 1970.	    10-52
10-10  Selected characteristics of households in four air pollution
       zones	    10-56
10-11  27 cleaning and maintenance operations separated by sensitivity
       to air particulate levels in four pollution zones	    10-57
10-12  Annual welfare gain from achieving primary and secondary
       standards for TSP concentrati on	    10-62
                                      xvi

-------
                             CONTENTS (continued)
Table                                                                      Page

10-13  Economic loss, materials damage attributed to ambient exposure
       to SO  and PM, estimated by Salmon, 1970 (in billions of 1970
       dol lafs)	    10-65
10-14  Estimates of materials damage attributed to SO  and PM in 1970
       (in millions of 1970 dollars)	    10-66

12-1   Lethal effects of S02	     12-9
12-2   Effects of S02 on lung morphology	.	    12-10
12-3   Effects of S02 on pulmonary function	,	    12-20
12-4   Effects of S02 on host defenses		    12-23
12-5   Effects of H2S04 aerosols on lung morphology,.	...	    12-26
12-6   Respiratory response of guinea pigs exposed for 1 hour to
       particles in the Amdur et al. studies	    12-29
12-7   Effects of acute exposure to sulfate aerosols on pulmonary
       function	    12-40
12-8   Effects of chronic exposure to H2S04 aerosols on
       pulmonary function...		    12-42
12-9   Effects of H2SQ4 on mucocil iary clearance.	    12-45
12-10  Effects of metals and other particles on host defense mechanisms    12-47
12-11  Effects of acute exposure to S02 in combination with certain
       particles	,	    12-57
12-12  Pollutant concentrations for chronic exposure of dogs	    12-60
12-13  Effects of chronic exposure to SO  and some PM	.	    12-64
12-14  Effects of interaction of SO  and 03	    12-67
12-15  Potential mutagenic effects 8f S02/bisulfite	    12-73

13-1   Sensory effects of S02	     13-4
13-2   Respiratory effects of S02	     13-7
13-3   Pulmonary effects of sulfuric acid	    13-25
13-4   Pulmonary effects of aerosols	    13-33
13-5   Pulmonary effects of combined exposures to S02 and other gaseous
       air pollutants	    13-40

14-1   Excess deaths and pollutant concentrations during severe air
       pollution episodes in London (1948 to 1962)	    14-12
14-2   Summary of results, selected patients, 1964-65 and 1967-68	    14-29
14-3   Average deviation of respiratory and cardiac morbidity from 15-
       day moving average, by smoke level (BS)(London, 1958-60)	    14-31
14-4   Average deviation of respiratory and cardiac morbidity from 15-
       day moving average, by S02 level (London, 1958-60)	    14-31
14-5   Summary of key results regarding mortality-air pollution relation-
       ships in U.S. cities based on Lave and Seskin Model analyses for
       I960, 1969, and 1974 data.	    14-39
14-6   Summary of Lave and Seskin (1977) analysis of residuals from
       regression analysis for 1960 and 1969 U.S. .SMSA data....	    14-42
14-7   Summary of quantitative conclusions from epidemiological studies
       relating health effects to acute exposure to ambient air levels
       of S02 and PM	".	     14-52
14-8   Summary of quantitative conclusions from epidemiological studies
       relating health effects to chronic exposure to ambient air
       levels of S02 and PM	    14-54
14-9   Comparison of measured components of TSP in U.S. cities (1960-
       1965) and maximum 1-hour values in London (1955-1963)	    14-55

                                      xv ii

-------
Page Intentionally Blank

-------
8. EFFECTS ON VEGETATION

8.1 GENERAL INTRODUCTION AND APPROACH
The objective of this chapter Is to review available data relating atmospheric concentra-
tions of SOy and participate matter (PM) to effects on terrestrial vegetation. Many reviews
of the general effects of SO- (Daines, 1968; Guderian, 1977; Jacobson and Hill, 1970; Linzon,
1978; Mudd, 1975; National Academy of Sciences, 1978; Treshow, 1970; U.S. Environmental Pro-
tection Agency, 1973, 1978a; Van Haut and Stratmann, 1970) and, to a lesser extent, of parti-
culate matter (Guderian, 1980) exist in the literature. Additional reports of S0~ effects on
plants have also been prepared for use in diagnostic situations (Davis, 1972a,b; Lacasse and
Moroz, 1969; U.S. Environmental Protection Agency, 1976).
This chapter addresses factors that influence our ability to determine relationships
between pollutant concentration and plant responses. Principal focus is placed on specifying
concentrations of S0« and particles associated with vegetation responses ranging from the bio-
chemical level to that of plant populations. In this process, information of historical
interest has been kept to a minimum and emphasis has been placed on more recent studies that
have employed modern monitoring, experimental, and statistical techniques. Studies in which
SO, or particulate matter were determined to be the major cause of measured effects have been
emphasized.
As a backdrop against which to consider pollutant effects on plants, it is important to
recognize that the many factors that play a major role in determining whether a given quantity
of pollutant will produce a predictable level of effect vary tremendously in nature. These
factors include the type of exposure (acute or chronic), influences of stress from other
biotic (insects, disease) or abiotic (edaphic or climatic) factors, the type of response mea-
sured, and the species or population under study. These factors and associated terminology
have been addressed in Sections 8.2 and 8.4 for S0? and Sections 8.6 and 8.7 for particulate
matter.
While a broad variety of responses measured following exposure of vegetation to SO^ or
particulate matter are discussed, it should be noted that not all responses are detrimental,
and that not all short-term detrimental responses ultimately result in effects detrimental to
plant growth and development. These concepts are developed more fully in Sections 8.2.7 and
8.3 below.
The end-point of this presentation of concepts, components, and modifiers of pollutant
dose and plant response can be found in attempts to define exposure-response relationships for
SO- and PM effects on vegetation. This is done in Sections 8.3 and 8.8 for S0? and particu-
late matter, respectively.
The concluding section on ecosystem responses (Section 8.10) derives information from a
much more limited number of studies. Here reliance on the broadly based concepts of ecosystem
analyses forms the basis for strong inference rather than proof of effects on a more subtle

8-1

-------
scale. In this area definitive data to evaluate the degree or extent of ecosystem changes
over broad regions do not presently exist. ;
The following definitions have generally been acceptable to plant scientists working on
air pollutant-induced effects on plants (American Phytopathological Society, 1974) and are
relevant to information, discussed below in this chapter:
1. Injury - a change in "the appearance and/or'function of a plant that is deleterious
to the plant.
2. Acute Injury (effects) - injury, usually involving necrosis, which develops within
several hours to a few days after short-time exposure to a pollutant, and is
expressed as fleck, scorch, bifacial necrosis, etc.
3. Chronic Injury (effects) - injury which develops only after long-term or repeated
exposure to an air pollutant and is expressed as chlorosis, bronzing, premature
senescence, reduced growth, etc.; can include necrosis.
4. Damage - a measure of decrease in economic or aesthetic value resulting from plant
injury by pollutants.

8.2 REACTION OF PLANTS TO SULFUR DIOXIDE EXPOSURES
8.2.1 Introduction
The response of plants to SOp exposure is a complex process that involves not only the
pollutant concentration and duration of exposure but also the genetic composition of the plant
and the environmental factors under which exposure occurs. In simplistic overview, this pro-
cess involves entrance of SO,, into the plant through leaf openings called stomata, and contact
within the leaf with wet cellular membranes and subsequent liquid phase reactions resulting in
the formation of sulfite and sulfate compounds. The formation of these compounds can initiate
changes within the plant metabolic systems that will produce physiological dysfunctions. If
sufficient physiological modifications occur, plant homeostasis or equilibrium is disturbed
and visible symptoms of injury may or may not be manifested. Plant repair mechanisms" can
result in a return to homeostasis and recovery; however, should exposure occur again prior to
complete recovery, further injury may occur and plant recovery is less probable.
Several plant responses to exposure to SOp and related sulfur compounds are possible:
(1) fertilizer effects appearing as increased growth and yield; (2) no detectable responses;
(3) injury manifested as growth and yield reductions without visible symptom expressions on
the foliage or with only very mild foliar symptoms that are difficult to attribute to air pol-
lution without comparing them to a control set of plants grown in pollution-free conditions;
(4) injury exhibited as chronic or acute symptoms on foliage with or without associated reduc-
tions in growth and yield; and (5) death of plants and plant communities.
In some instances, the addition of low S0? concentrations to a plant's environment may
induce a fertilizer-like response, but relatively few studies of this phenomenon on agronomic
crops have been completed to date; none have shown beneficial effects on natural ecosystems.
In view of 'the scanty data on the subject, only limited consideration of potential beneficial
effects of S0« exposure can be undertaken at this time.
8-2

-------
Plant injury or death may result from continued exposure to high or low pollutant doses.
If such is the case, other mitigating factors may also be involved (e.g., abiotic agents or
biotic disease-inducing agents such as insects). Depending upon the plant species, exact con-
ditions of the seasonal stage of crop growth, pollutant dose, and environmental conditions,
many forms of injury may take place and their relative impact may vary. Symptoms of acute and
chronic injury may occur on a given plant simultaneously. Here, injury does not necessarily
imply damage (i.e., economic loss). Also, the timing of pollutant exposure in relation to the
physiological stage of crop development often determines the relationship of foliar injury to
subsequent yield losses.
Before discussing the different types of effects of S02 on vegetation, certain background
information is provided concerning deposition of S0« on plant surfaces, entrance into the
plant, and distribution and reaction within the plant after S0? entry into plant tissues.
8.2.2 Wet and Dry Deposition of Sulfur Compounds on Leaf Surfaces
Deposition processes limit the lifetime of sulfur compounds in the atmosphere, control
the distance traveled before deposition, and limit the atmospheric pollutant concentrations
(Garland, 1978), A fuller discussion of pollutant transport, transformation and dry deposition
is given in Chapter 6.
There have been several studies of the deposition of particulate material to natural sur-
faces (Chamberlain, 1975; Little and Wiffen, 1977; Sehmel and Hodgson, 1974). Very large
particles are chiefly deposited by sedimentation. Particles in the range of I to 100 urn are
also borne towards the surface by turbulence where sedimentation is supplemental to impaction
on rough surfaces. Submicrometer-sized particles (e.g., sulfuric acid aerosols) and gases,
including SCL, diffuse by Brownian motion through the thin laminar layers of air close to the
plant surface. This may be followed by active uptake by plants. The mean S02 deposition
velocities are surprisingly similar for a wide range of plant leaf surfaces (Garland, 1978).
(Wet deposition is discussed in Chapter 7.)
Dry deposition results in the removal of significant amounts of the larger particles from
the atmosphere within 2 or 3 days following emission, but several weeks are required to remove
the submicrometer fraction.
8.2.3 Routes and Methods of Entry Into the Plant
Stomata of leaves have been demonstrated to be the major avenues of SO, entrance into
plants. Although this is a widely accepted conclusion that has been presented in numerous re-
views (Guderian, 1977; Katz, 1949; Thomas and Hendricks, 1956; U.S. EPA, 1973), there is still
controversy as to the importance of stomatal movement relative to plant biochemistry in
determining plant sensitivity. Many factors that govern the mechanism of stomatal opening and
closing have been determined to be independent of S02 concentrations to which a plant is
exposed. Physical factors such as light, leaf surface moisture, relative humidity, and soil
moisture availability influence stomatal opening and closing and play a major role in plant
sensitivity by limiting passive entry of SO- into the leaf (Domes, 1971; Meidner and Mansfield,

8-3

-------
1968; Setterstrom and Zimmerman, 1939; Spedding, 1969; Mclaughlin and Taylor, ;1981).
These factors must therefore be considered when determining plant sensitivity or tolerance to
entry of SO,.
Internal resistances to flux of gases into plant leaves may also be substantial and may
exceed those imposed by stomata under some conditions. Barton et al. (1980) found that photo-
synthetic depression in kidney beans (Phaseolus vulgaris) during SCL exposure was explained
primarily by increases in mesophyll resistance, Stomatal resistances changed only slightly
and were a minor component of total leaf resistance to (XL at both high (71 percent) and low
(33 percent) relative humidity. Winner and Mooney (1980) also found that differences in both
stomatal and nonstomatal components of leaf resistance to S0? uptake were associated with dif-
ferences in resistance of deciduous and evergreen shrubs to S0?.
The rate of absorption of SO, by plants varies not only among species, but also is
influenced by previous exposures to SO-. Rates of S02 absorption and of translocation of
absorbed sulfur v/ere determined in sugar maple (Acer saccharum Marsh.), big toothed aspen
(Populus grandidentata Michx), white ash (Fraxinus amepicana L.) and yellow birch seedlings
(Betula alleghaniensis Britten Betula lutea Michx. f.). Bigtooth aspen, a species sensitive
to SO,, had the highest absorption rate with no prefumigation and sugar maple, a tolerant
species, had the lowest. The rate of sulfur absorption was reduced in all species except
q
white ash, a species of intermediate sensitivity, after prefumigation with SQ« at 1970 ug/m
(0.75 ppm) for 20 to 36 hrs. (Jensen and Kozlowski, 1975). The relationship between species
sensitivity and absorption rate thus changed with the prefumigation treatment. The sulfur
content of foliage in all species increased with SO, fumigation. Eight days after fumigation
35
with S0?, varying amounts of the labelled sulfur were translocated throughout the plants,
including roots (Jensen and Kozlowski, 1975). Subsequent effects were not indicated.
Sulfur dioxide has been shown both to increase and decrease stomatal resistance and thus
affect potential photosynthetic performance (Hallgren, 1978). Sulfur dioxide induced the clo-
sure of stomata in the Pelargonium hybrid, "Pelargonium x hortorum", especially when they had
been fully opened, and necrosis was not averted (Bonte et al., 1975). Kodata and Inoue (1972)
demonstrated that S02 entered leaves of Pinus resinosa (red pine) through stomata and accumu-
lated in the cells around stomata for some time before diffusing inward through the leaf;
i.e., internal diffusion was slower than diffusion into the leaves.
Once SO- has entered a leaf, it may induce stomata to remain open for longer periods of
3
time or to open wider than before fumigation. However, exposure of plants to SO- (1360 pg/m ,
0.5 ppm) at relative humidities above 40 percent caused an increase in stomatal opening
(Majernik and Mansfield, 1970; Mansfield and Majernik, 1970). A 3-minute fumigation with 6550
(jg/m (2.5 ppm) S0? increased carbon dioxide uptake and stomatal opening in white mustard
Brassica hirta Moench (Sinapsis alba B + B) plants. However, with the same concentration,
suppressed carbon dioxide uptake and stomatal closure have also been noted (Buron and Cornic,
1973).

8-4

-------
8,2.4 Cellular and Biochemical Changes
Based on the available literature, it is difficult to assess the relationship of SO^-in-
duced biochemical and/or physiological changes at the cellular level to subsequent effects on
photosynthetic activity or resultant growth and yield. Numerous studies have utilized
detached leaves and/or isolated chloroplasts in culture solutions for evaluation of biochemi-
cal or physiological effects, but their use for field estimations under ambient conditions re-
mains limited.
Recent studies have also shown a variety of SCL-induced biochemical effects: enzyme in-
hibition (Pahlich, 1971, 1973; Ziegler, 1972); interference with respiration (Haisman, 1974);
interference with energy transduction (Ballentyne, 1973); interference with lipid biosynthesis
(Malhotra and Kahn, 1978); alterations in ami no acid content and quality (Godzik and Linskens,
1974); and chlorophyll loss (Rao and LeBlanc, 1965). Pahlich (1975) has rationalized some of
this diverse list of effects in terms of sulfite and sulfate accumulation by exposed plant
tissues.
Vogl et al. (1965) attempted to integrate biochemical responses with the type and magni-
tude of resultant plant effects (Table 8-1). In Table 8-1 degrees of injury are classified
from A to E, i.e., from no visible injury to "catastrophic" in Vogls1 terminology. Develop-
ment of models is necessary to relate changes in physiology (biochemical responses) of parti-
cular plant species to altered growth and productivity, both qualitatively and quantitatively.
That is, the relationship between biochemical changes or responses and visible injury at vari-
ous S0~ exposure levels remains to be better defined.
Horsman and Wellburn (1976) have prepared an extensive listing of reported metabolic or
enzymatic effects of S09 on plants or plant tissues. In only one of eleven studies reviewed
14
was an increase in photosynthesis ( CO,, fixation) noted as a positive or beneficial effect in
response to exposure to S02 or its derivatives; the remaining effects observed were negative
or detrimental.
With SO-, which upon absorption is hydro!ized to SO- and then to SOl and subsequently in-
corporated into S-containing ami no acids and proteins, the rate of entry is particularly
important for determining toxicity. Plants have an inherent, and apparently species-dependent,
capacity to absorb, detoxify, and metabolically incorporate SO, and may absorb low concentra-
tions of SO, over long time periods without damage. Thomas et al. (1943), for example,
3
exposed alfalfa to SO, continuously at 520 ug/m (0.20 ppm) for 8 weeks without adverse
effects. Toxicity to S0? may occur during short episodes when the SOn to S07 conversion rate
is exceeded and the extremely toxic sulfite (SOI) form accumulates (Ziegler, 1975). During
longer exposures at lower S0? concentrations, SOT may accumulate as the capacity for metabolic
incorporation of SOT is exceeded, and chronic symptoms may appear. It is therefore reasonable
to expect that either no effects or beneficial effects may be associated with low-level SO,
exposures; but detrimental (even fatal) toxic effects may occur at increasingly higher S02 ex-
3-5

-------
                                        TABLE 8.1.   RELATIONSHIP OF BIOCHEMICAL RESPONSE TO VISUAL SYHPTQHS OF PLANT INJURY
Description of Injury
Degree
of
injury
A

8
C
Visible
symptoms
none

not detectable
loss of assimilation
capacity through:
Symptoms of biochemical
injury in leaf cells
stress on
buffer systems
photosynthesis
adversely affected,
diminished assimi-
lation rate
diminished activity
of enjymes
Injui-y
Assimilation
organs
none

temporary
impedance of
gaseous exchange
prolonged impe-
dance of gaseous
to;
Whole plant
none

not detectable
reduced growth
(deficiency
Ability
Assimilation
organs


very quick,
completely
slowly, com-
pletely
to recover in:
Whole plant



slowly, complete!;
for perennials
CO
 I
1) premature death
   of assimilation
   organs (leaves,
   needles)

2) diminished
   growth of new
   tissues (shorter
   needles, etc.)
                                              effect upon
                                              chlorophyll
D





E



necrosis of the
assimilating and
active plant
tissues


destruction of all
important assimila-
tory plant tissues

death of cells
through protein
and enzyiie degrada-
tion


death of organs



irreversible
injury: necro-
sis of some
assimilation
organs or parts
thereof
irreversible
injury to all
assimilation
organs
loss of assimi-
lation capability




destruction of
assimilation
capability

quick, not com-
pletely, some-
tlnes (for iso-
lated tissues)
not any more

not any more



slowly, completely
for perennials




sometimes (for
isolated
tissues)

            Vogl, et al., 1965.

-------
posure levels once the capacity for conversion of SO- to SO, and the transformation of SO, in
the leaf are exceeded.
8,2.5 Beneficial "Fertilizer" Effects
Under certain conditions, atmospheric S0? can have beneficial effects on agronomic vege-
tation (Noggle and Jones, 1979). Sulfur is one of the elements required for plant growth and
Coleman (1966) reported that crop deficiencies of sulfur have been occurring with increasing
frequency throughout the world. The sulfur required to maintain high crop production ranges
from 10 to 40 kg/ha per year. Figure 8-1 presents a map of sulfur-deficient soils of the
United States (The Sulphur Institute, 1979).
Cowling et al. (1973) found beneficial effects of S0?, such as increases in yield and
sulfur content, in perennial ryegrass that was grown with an inadequate supply of sulfur to
the roots. Faller (1970) conducted a series of experiments to determine effects of varying
atmospheric concentrations of S02 on sunflower, corn, and tobacco. In these studies, plants
were grown in nutrient media containing adequate supplies of all essential elements except
sulfur, which was low. Plants grown in the atmosphere without S0y developed sulfur-deficiency
symptoms within a few days. In other treatments, total plant yield increased to some extent
when increasing concentrations of S02 were added to the atmosphere during plant growth. For
tobacco, the total dry weight increased by up to 40 percent. Yields of leaves and stems alone
increased by 80 percent while dry weights of tobacco increased even at the highest S09 concen-
3
tration used (1490 |jg/m , 0.57 ppm); sunflower and corn had their highest biomass at S09 con-
33
centrations of 1050 pg/m (0.40 ppm) and 520 yg/m (0.20 ppm), respectively. Beyond these
35
concentrations, visible injury was observed. Additional studies by Faller (1970) with S
suggest that up to 90 percent of plant sulfur requirements may originate from the atmosphere
under the specific experimental conditions.
No monitoring or handling procedures for SO,, delivery were presented in a study by Thomas
et al. (1943); however, its results indicated that S00 could serve as a source of nutrient S
3
when alfalfa, grown in sulfur-deficient solutions, was fumigated with approximately 260 pg/m
(0.10 ppm) SQ~ for 6 to 7 hr/day, 6 days/week for the growing season, but had little effect
when the growth medium contained sufficient sulfur.
Noggle and Jones (1979) reported the results of a 2-year study comparing the contribu-
tions of soil and atmospheric sulfur to the sulfur requirements of cotton and fescue when the
plants were potted in soil and S was added as the S nutrient. Cotton was more efficient
than fescue in accumulating sulfur from the atmosphere. The amount of sulfur accumulated from
the atmosphere was apparently influenced by the amount of sulfur supply in the soil relative
to the sulfur requirements of the plant. A crop grown in a sulfur-deficient soil will accu-
mulate more sulfur from the atmosphere than the same crop grown in a soil that has an adequate
supply of sulfur. Noggle and Jones (1979) also showed that cotton grown in specifically de-
signed growth containers in the vicinity of certain coal-fired power plants accumulated signi-
ficant amounts of atmospheric sulfur (as S0?) and produced significantly more biomass than

8-7

-------
O3
1
co
               j     | STRONG TO
                       MODERATE NEED

                     LESSER NEED

                     NO DATA OR LITTLE NEED
               Figure 8-1. Map of the United States Indicating major areas of sulfur-deficient soils.


               Source: The Sulphur institute (1979).

-------
those grown at a location farther from the industrial source of sulfur. Thus, under appropri-
ate conditions, such as with sulfur-deficient soils, the atmosphere can be an important source
of sulfur for plant requirements. It should be noted, however, that SCL monitoring data were
not presented in the report.
Similar stimulatory responses of Oryzopsis hymenoides (a desert grass) were noted follow-
3 3
ing continuous exposure to SO,, concentrations of 80 $ig/m (0.03 ppm) and 160 M9/m (0.06 ppm)
for 6 weeks (Ferenbaugh, 1978), The noted increases in productivity expressed as mg dry wt/
plant were not statistically different from the control plants, but the trends were stable for
3 3
these doses. Exposure to 340 M9/m (0.13 ppm) and 660 |jg/m (0.25 ppm) S0« under identical
conditions however, resulted in foliar symptoms and decreased productivity.
Studies that have demonstrated beneficial effects, as well as those demonstrating detri-
mental effects, must be evaluated in terms of their single influence on one crop. There is
little doubt that direct application of sulfur as a nutrient to certain crops grown under bor-
derline or sulfur-deficient conditions may result in increased productivity for that crop.
Plants normally take up sulfur in the form of sulfate via, the roots; however, they can also
utilize SOp taken up through the leaf stomata. Faller (1971) has shown that tobacco plants
can utilize S0? as a sulfur source in deficient soils. Studies of this type (Faller, 1970)
have led some individuals to conclude that SO,, emitted from industrial sources is actually
beneficial as an aerial fertilizer for plants growing in sulfur deficient soils (Terman, 1978;
Noggle and Jones, 1979). The situation is not as simple as portrayed. A small amount of
atmospheric SO,, can be nutritionally beneficial to plants in the short term. However, in the
long term, large, uncontrolled, and frequent applications of sulfur compounds such as SO,, can
be detrimental. Not all plant species have the same nutritional requirements. In addition,
the rate at which plant species assimilate sulfur is influenced by such variables as the
physiological status of the plants, age and time during the growing season when the applica-
tion occurs, availability of soil nutrients, and light intensity. When more sulfur is avail-
able than can be assimilated by plants, it is accumulated in the leaf tissue (Ulrich et a!.,
1967; Legge et al., 1976; Cowling and Koziol, 1978; Thompson and Katz, 1978). Sulfur accumula-
tion in the leaf can reach toxic levels and adversely affect plant growth (Katz, 1949; Linzon
et al. , 1979). Growth rates of leaf tissues, available nutrient supplies, and environmental
factors affecting stomatal opening and closing should all be considered as influencing the
rate of sulfur accumulation in plant tissues (Bell and Clough, 1973).
The effects of nutrient forms of nitrogen being delivered to the plant along with S0? in
plumes under field conditions have not been investigated. However, Cowling and Lockyer (1978)
demonstrated that ryegrass (the S 23 cultivar), when grown under sulfur (S) deficiency and at
low nitrogen (N), did not respond to 50 M9/m (0.02 ppm) S0? for 85 days, whereas plants grown
at high N under the same exposure conditions responded with a 227 percent yield increase and
S-deficiency symptoms were alleviated. Data presented previously by Cowling and Jones (1971)
showed that at high levels of N and inadequate levels of S, nitrate-nitrogen accumulated in

8-9

-------
ryegrass, thus Indicating that protein synthesis was inhibited. Long-term natural ecosystem
studies showing the beneficial effects of SO, deposition are not available, wh'ereas those
showing detrimental effects are available (Legge et a!., 1981; Legge, 1980; Winner et al.,
1978; Winner and Bewley, 1978a). Since agronomic and natural ecosystems are often physically
proximal to one another, further research on the potential influence of sulfur compounds to
each singly and collectively is greatly warranted.
The Sulphur Institute (1971) published a review of crop response to sulfur; the reader is
referred to that review for more information on beneficial effects of sulfur on plants.
8.2.6 Acute Foliar Injury
This type of injury occurs following rapid absorption of a toxic dose of SO- and results
at first in marginal and intervene! areas having a dark green, watersoaked appearance. After
desiccation and bleaching of tissues, affected areas become light ivory to white in most broad-
leaf plants. Some species show darker colors (brown or red), but an exact line of demarcation
characteristically exists between symptomatic and asymptomatic portions of leaf tissues. Bi-
facial necrosis is common. In monocotyledons (corn, grasses), foliar injury occurs at the
tips and in lengthwise strips along parallel veins (U.S. EPA, 1976).
In conifers, acute injury on foliage usually appears as a bright orange red tip necrosis
on current-year needles, often with a sharp line of demarcation between the injured tips and
the normally green bases. Occasionally, injury may occur as bands at the tip, middle, or base
of the needles (Linzon, 1972). Recently incurred injury is light colored; but later, bright
orange or red colors are typical for the banded areas and tips. As needle tips die, they be-
come brittle and break, or whole needles drop from the tree. Pine needles are most sensitive
to SO, during the period of rapid needle elongation, but injury may also occur on mature
needles (Davis, 1972a).
8.2.7 Chronic Foliar Injury
Visible plant injury not involving collapse and necrosis of tissues is termed chronic in-
jury. This type of visible injury is usually the result of variable fumigations consisting of
either short-term, high-concentration or long-term, low-concentration exposures to S0?. It
has also been called "sulfate injury" since a slow accumulation of sulfate results from such
exposures (Daines, 1968). Within substomatal cavities, SO, reacts quickly with intercellular
water to form sulfite and bisulfite. These substances are slowly oxidized to sulfate which is
approximately one-thirtieth as toxic as sulfite and bisulfite (Thomas, 1951). If the capacity
of plant tissues to convert sulfite and bisulfite to sulfate is not exceeded or sulfate elimi-
nation processes are not overwhelmed, visible expression of symptoms will not occur. However,
as sulfite and bisulfite ions are formed and as sulfate accumulates to phytotoxic levels, then
chronic symptoms first appear as various forms of chlorotic (yellowing) patterns. As sulfite
and bisulfite ions continue to accumulate, destruction of individual chloroplast membranes or
reduction of chlorophyll production ensues, resulting in reddening or bleaching of cells with-
out necrosis (Thomas, 1951). Following such accumulations, a fine gradation occurs between

8-10

-------
chronic and acute symptom expressions. Sulfate levels in plants ambiently exposed to S0? have
been shown to be several times greater than those in control plants (Linzon, 1958).
In broadleaf plants chronic injury is often expressed in tissues located between veins,
with various forms of chlorosis predominating. Chlorotic spots or mottle may persist follow-
ing exposure or may subside and disappear after pollutant removal,or other changes in environ-
mental conditions (Jacobson and Hill, 1970). Chronic effects of SO, in conifers are generally
first expressed on older needles (Linzon, 1966). Chlorosis of tissues starting at the tips
progresses down the needle towards the base, i.e., symptoms progress from the oldest to
youngest tissues. Advanced symptoms may -follow, involving reddening of affected tissues
(Linzon, 1978).
8.2.8 Foliar Versus Whole Plant Responses
The presence of acute or chronic foliar injury is not necessarily associated with growth
or yield effects. Furthermore, when present, the degree of foliar injury may not always be a
reliable indicator of subsequent growth or yield effects. Yield effects in the absence of
foliar injury, for example, have been reported for soybeans in field fumigation experiments
(Sprugel et al., 1980; Heagle et al. 1974) and in chamber exposures (Reinert and Weber, 1980).
Sprugel et al. (1980), utilizing the Zonal Air Pollution System (ZAPS), reported significant
yield reductions in Wells soybeans exposed to mean S02 levels of 240, 260, 500, 660 or 940
ug/m3 (0.09, 0.10, 0.19, 0.25, or 0.36 ppm) for an average of 4.2 hr/day intermittently for 18
days during July and August 1978, but visible leaf injury was seen only at the highest SQ2
levels. The "ZAPS" permits significant variation in pollutant concentrations. For example,
in 1977, in experiments reporting a mean S02 concentration of 790 M9/m (0.30 ppm), the actual
concentration ranged from 0.00 to 3140 ug/m3 (0.00 to 1.20 ppm) S02 (Miller et al., 1980).
The "ZAPS" type of exposure system, however, tends to simulate ambient conditions near a point
source. Under the "ZAPS" system, S0? or other pollutants are delivered through a system of
one-inch pipes suspended about 75 cm above the soil surface in field plots of 0.5 hectares.
Pollutant samplers are located in the plots to measure the amount of pollutant delivered to
the vegetation. Though flow rates through the system are constant, the canopy-level concen-
trations vary depending upon the meteorological conditions, principally wind speed and tempera-
ture, so that concentrations are often higher near the pipes.
In a greenhouse chamber study of 0, + S02 interaction, Reinert and Weber (1980), exposed
Dare soybeans for 4 hr/day, 3 times per week, for 11 weeks. They reported significant growth
•j
reductions in the absence of visible injury at 660 ug/m (0.25 ppm) S02 when the treatment
sums of squares were partitioned. On the other hand, Heagle et al. (1974) also exposed Dare
o
soybeans but found that plants exposed to 260 |jg/m (0.10 ppm) SO, 6 hr/day for 133 days in
closed field chambers exhibited no significant yield reductions, even in the presence of
foliar injury.
8-11

-------
Additional examples of ambiguities are found in some of the literature dealing with gras-
ses. In a preliminary study, S 23 ryegrass exhibited significantly reduced growth when ex-
posed to 310 ng/m (0.12 ppm) S02 for 9 weeks or 174 ug/m3 (0.067 ppm) S02 for 26 weeks (Bell
and Clough, 1973); however, the only foliar injury noted was a slight chlorosis and an en-
hanced rate of leaf senescence. Ashenden.(1978) noted similar significant growth reductions
2
for cocksfoot grass when exposed to 290" ng/m (0.11 pp"m) VSO? for 4 weeks; reductions ranged
from 32 to 52 percent for various yield parameters while foliar necrosis was only 5 percent.
On the other hand, exposure of S 23 ryegrass to 50 or 370 ug/m (0.02 or 0.14 ppm) SCL in two
successive growth periods of 29 and 22 days was reported to result in foliar injury at the
high concentration, but no yield effects at either concentration (Cowling and Koziol, 3.978).
Neither net photosynthesis nor dark respiration was significantly affected.
Different plant species differ in tolerance to SO,, injury. Leaf injury and radial growth
were evaluated on Douglas fir and ponderosa pine growing in nursery plots exposed to various
doses of S0? in controlled fumigations to determine their tolerance to S0? exposure (Katz and
HcCallum, 1952). Slightly injured ponderosa pine (10 percent foliar symptoms) exhibited no
significant deviations in growth while slightly injured Douglas fir (10 percent foliar symp-
toms) showed definite growth retardation in comparison to control plants. The growth retarda-
tions were evident for 3 years after S0« exposure, followed by substantial recovery.
Sulfur dioxide concentrations have also been negatively correlated with annual ring width
in Norway spruce (Keller, 1980). Exposures to 130, 260 and 520 ug/m3 (0.05, 0.10, and
0.20 ppm) SO, were continuous for 10 weeks in the spring. Some injury was noted at 260 ug/m
3
(0.10 ppm) and 520 (J9/m (0.20 ppm), but a distinct decline of wood production was also found
in cases where no visible injury occurred. When dormant seedlings of beech were exposed to
the same S02 concentrations of 130, 260 and 520 ug/m3 (0.05, 0.10, and 0.20 ppm) for about 16
weeks, there was an increase in the number of terminal buds that failed to "break" in the
spring for the 260 (0.10 ppm) and 520 ug/m3 (0.20 ppm) treatments (Keller, 1978).
The literature is ambiguous concerning SO^-induced growth and yield effects and correla-
tions with visible foliar injury. No studies considered all of the potential variables that
could affect plant response. This is a virtual impossibility for a single study and is espe-
cially true for field studies (which are most relevant for present purposes) where many en-
vironmental variables cannot be controlled. Also, many of the studies that have demonstrated
adverse effects of S0? at low concentrations have utilized sensitive cultivars of plant
species (which may or may not be representative of plant populations as a whole) and main-
tained exposure conditions conducive to maximum plant sensitivity. However, from the data
available, we can conclude that growth and yield effects are not necessarily related to foliar
injury. Depending upon the plant affected, the environmental conditions, and the pollutant
exposure conditions, one may observe yield effects without injury, injury without yield
effects, or positive associations between injury and yield effects.
8-12

-------
8.2.9 Classification of Plant Sensitivity to Sulfur Dioxide
Limitations of space do not permit a listing here of all plants known to be sensitive to
various doses of SCL- Furthermore, in a listing of sensitive plants, the evidence collected
should also indicate environmental, genetic, and cultural considerations that may in fact de-
termine such sensitivities. In addition, general descriptions are difficult because plant re-
sponses to air pollutants vary at the genus, species, variety, and cultivar levels. Some
listings of plants according to relative sensitivities have been prepared, however, using as a
basis the expression of visible symptoms ,by any given plant. Injury expressed as losses in
growth or yield, however, has generally not been considered in preparing such lists.
Jacobson and Hill (1970) published a list of plants sensitive to major phytotoxic air
pollutants, and Linzon (1972) has classified 36 tree species as being either tolerant, inter-
mediate, or sensitive to SO,. These sensitivity lists do not identify the pollutant doses re-
quired to induce visible injury on indicator species. However, Jones et al. (1974) have pub-
lished such details based upon observations over a 20-year period of 120 species growing in
the vicinity of coal-fired power plants in the southeastern United States (Table 8-2).
Other compilations have also been presented, including that of Davis and Wilhour (1976)
which provides information on an international basis, and the report of Hill et al. (1974) for
vegetation native to the southwestern deserts of the United States.
Extensive efforts have been made to identify and develop certain sensitive plant species
as potential bioindicators of ambient air S0~ effects. Perhaps the most extensively examined
plants for this use are eastern white pine (Pinus strobus) and numerous species of lichens,
The literature on white pine has been reviewed by Gerhold (1977), and the most recent review
discussing lichens as bioindicators was prepared by LeBlanc and Rao (1975). Other reports
mention the use of various ornamentals (Daessler et al., 1972; Heggestad et al., 1973; Pelz,
1962); bluegrass cu-ltivars (Murray et al., 1975); scotch pine (Oemeritt et al., 1971); hybrid
poplar (Dochinger and Jensen, 1975); and trembling aspen (Karnosky, 1977).
Bioindicators for determining SO,, effects must be used with caution, however, because
other factors such as drought, nutrient imbalances, and other pollutants may induce injury
symptoms that mimic those of SO,,. Therefore, several bioindicators should be used if any
given geographic area is to be evaluated for possible SO, effects. In addition, individual
species and more complicated plant bioindicator systems are not as effective in detecting SO^
at low concentrations as are sophisticated instruments.
8.3 EXPOSURE-RESPONSE RELATIONSHIPS - SULFUR DIOXIDE
The primary focus of exposure-response studies should be to delineate relationships
between measurable indices of exposure and meaningful parameters of plant response. This
section will examine such relationships both from the perspective of defining empirically
demonstrated exposure-response (or exposure-effect) relationships based on our present know-
ledge and from the perspective of what generalizations might be reasonably made concerning
likely effects of S0? on plants under ambient conditions.

8-13

-------
 I
I—>
-p=>
               TABLE 8-2.  SENSITIVITY GROUPINGS OF VEGETATION BASED ON VISIBLE INJURY AT DIFFERENT S02 EXPOSURES'


Sensitivity
Grouping
SO 2 concentration, (jg/m (ppm),
h
Peak0 1-hr
and duration time, hr

3-hr


Plants
       Sensitive:
       Intermediate:
2620-3930
(1.0 - 1.5 ppm)
1310-2620 ng/nr
(0.5 - 1.0 ppm)
790-1570
(0.3 - 0.6 ppm)
3930-5240 jjg/m
(1.5 - 2.0 ppm)
2620-5240 jjg/m
(1.0 - 2.0 ppm)
1570-2100 |jg/nr
(0.6 - 0.8 ppm)
Ragweeds
Legumes
Blackberry
Southern pines
Red and black oaks
White ash
Sumacs

Maples
Locust
Sweetgum
Cherry
Elms
Tuliptree
Many crop and garden species
Resistant:
>5240
(> 2.
|jg/m
0 ppm)
>5240
(> 2.
|jg/m
0 ppm)
>2100
(> 0.
|jg/m
8 ppm)
White oaks
Potato
Upland cotton
Corn
Dogwood
Peach

       Based on observations over a 20-year period of visible  injury occurring  on  over  120  species growing  in  the
       vicinities of coal-fired power plants in the southeastern United States.
       Maximum 5 minute concentration.

      Source:  After Jones et al., 1974.

-------
The dose of SO, to which vegetation may be exposed is conventional\y designated as the product
of the  concentration  of S02 in the  plant's  environment times the duration  of exposure.   The
term exposure dose  refers  to the amount of pollutant at the external p.lant surface.   Response
may be characterized by, a measurable change in any plant function such as biochemical pathways,
gas exchange rates, photosynthetic rates, physiological reactions, degree of visibly recogniz-
able leaf injury, or subsequent growth and yield.  Plant responses may be beneficial  or detri-
mental (see Section 8.2.5).   They may also involve the expression of growth and yield effects
without foliar symptoms  (see Section 8.2.8) or  may  lead  to overt symptoms that seldom become
more  serious  than those  associated  with acute  injury (see Section 8.2.6)  or chronic injury
(Section 8.2.7).
     In interpreting exposure-response studies wherein a measured plant,response is correlated
with the length of exposure, it is important to realize that the relationship between exposure
and the  amount of  pollutant entering the  plant may be  influenced  significantly by environ-
mental  factors which  control rates of pollutant flux into plant leaves and  by plant factors
that determine the  metabolic fate of the pollutant  within leaf tissues as noted earlier (see
Figure 8-2).
     The role of short-term fluctuations in S02 may be particularly important where impacts of
point sources  are of  concern (Mclaughlin et  a!., 1976).  Here  concentrations may  fluctuate
widely  during  exposure, and damage  to  vegetation may be  most closely  associated with short-
term  averages  (1 hr)  or even brief  peak concentrations (The term "peak"  is  not precise,  but
depends on  instrumentation and the opinion  of  the scientist.   However, a peak is  usually of
only  a  few minutes duration.)   Consistent  with this, laboratory experiments  by Zahn (1961,
1970) have  demonstrated the greater relative toxicity of  short-term exposures at high concen-
trations of SO, than longer-term exposures with the same total treatment.  Also, more recently
Mclaughlin  et  al.  (1979)  studied the effects  of varying the peak  to  mean  SCL concentration
ratio on  kidney  beans  in  short-term (3  hr)  treatments with SO*.  They found that increasing
the peak to mean ratio from 1.0 (steady state treatments at 0.50 ppm for 3 hours) to 2.0 (3 hr
                                 3
treatments  with  peak = 2620 pg/m   or 1.0 ppm)  did  not alter post  fumigation photosynthetic
                                                                                          3
depression.  Further  increasing the ratio to 6.0 (1 hour exposure with peak = 5240 ng/m  or
2.0 ppm), however, tripled the postfumigation photosynthetic depression.  Total dose delivered
in the  three  exposures was  1.5,  1.8,  and 1.1 ppm hr,  respectively.   Clearly the quantity of
SO, to which the plants are exposed may have a very different effective potential as the para-
meters of the exposure are changed.
     Another important  aspect  of exposure is the frequency and duration of periods of low SO™
stress.   Zahn  (1970)  emphasized that periods of  low  SO,  concentration may be critical to the
recovery potential  of  plant systems following exposure to elevated levels of SO^.  Thus, con-
tinuous  exposure systems  probably overestimate  the  toxicity of the  delivered dose  in many
cases because physiological  recovery is not permitted.  Such recovery would be expected under
                                            8-15

-------
 NUMBER OF
 EXPOSURES"
 CLIMATIC FACTORS

 EDAPHIC FACTORS


 BIOTIC FACTORS
                                       POLLUTANT

                                    CONCENTRATION
PLANT RECEPTOR
                                 MECHANISM  OF ACTION
                                   DURATION OF
                                   "EACH EXPOSURE
                GENETIC MAKEUP



                 STAGE OF PLANT
                 DEVELOPMENT
                                        EFFECTS
                           ACUTE
   CHRONIC
  l r

SUBTLE
Figure 8-2. Conceptual model of the factors involved in air pollution effects (dose-response) on vegetation.


Source: Heck and Brandt, 1977.
                                          8-16

-------
most exposure regimes  In  the field where fluctuating  synoptic or local meteorological condi-
tions strongly Influence exposure patterns.
     Equally critical in defining the biologically significant features of exposure regimes is
a description of  the significant parameters of plant  response.   Sections 8.2.1 through 8.2.9
have emphasized many types  of responses that may  be elicited by exposure to  SCk-   In inter-
preting or predicting  plant response to SO,, it  is  important to keep in perspective the fact
that  plant  growth   and  development  represent  an  integration  of  cellular  and  biochemical
processes, just  as  community behavior  is  an  integration  of the  performance of  component
species.  The internal  allocation  of resources (carbon, water, and nutrients) to growth is an
integrative and,  in  many  cases,  resilient process that plays a major role in determining how
both individual plants  and  plant communities respond  to environmental  stress (Mclaughlin and
Shriner, 1980; see  also Section  8.10).  The fact  that a response is measured following expo-
sure to a given  concentration  of  S0«  may be  of  interest  in understanding  the mechanism of
action  and in  identifying the biologically significant features of exposure; however, it does
not  necessarily  mean that  an effect will be measured at a  subsequent  higher level  of plant
organization.   Responses  at  higher  levels  of  plant   organization,  however, must  be viewed
within  the  perspective of  the  increasingly complex biotic  and abiotic  factors that control
plant  response  (see Section  8.5)  and must  influence  our  attempts  to move  the focus of our
studies  from processes to  plants  and  from  plants  to communities.   These  factors  similarly
limit  our capacity  to extend  experimental  protocols beyond the  confines  of  our  carefully
controlled laboratory studies to more natural,  and more variable, field situations.
     Several  attempts have been made to characterize dose-response relationships in a mathema-
tical  sense  using monitored concentrations,  duration times,  and injury thresholds as modified
by  physical  and biotic factors  expressed  as constants (O'Gara, 1922; Thomas  and  Hill, 1935;
Zahn,  1963a,b).   However,  their consistency and  usefulness  are  limited by  numerous physi-
cal  and biotic  factors that must be  considered  in evaluating dose-response data.   Changes in
exposure conditions,  differences in exposure methodology and efficiency  of  monitoring equip-
ment,  and  consistency of measurements within a  study and between studies on the same plant
directly influence results.
     Data regarding SO^ effects on plant growth and yield in  most cases provide the most rele-
vant basis for studying dose-response relationships.   As a whole-plant measurement, plant pro-
ductivity is an  integrative parameter that considers  the  net effect of multiple factors over
time.   Productivity  data are  presently available for a wide range of  species  under a broad
range of experimental conditions.  However, results are neither necessarily closely comparable
across  sometimes  very divergent experimental  conditions, nor are all  necessarily useful for
criteria  development purposes.   Consequently,  rather  than  review all  such  studies  here in
detail, summaries of their data have been tabulated separately in Appendix 8A for:   controlled
field exposures (Table 8A-1); laboratory studies with agronomic and horticultural crops (Table
8A-2) and tree species (Table 8A-3); and a variety of studies with native plants (Table 8A-4).

                                            8-17

-------
Only the most important or salient points for present purposes will be discussed here, and
studies of SO- exposure effects on plants under field conditions will be emphasized.
Relatively few crops of economic importance have been studied under field conditions.
Results obtained for SOp exposure effects on yields for certain commercially important species
are summarized in Table 8-3 (statistical significances are noted in the table). Of the
results listed, those for soybeans, wheat and barley are economically the most important. For
soybeans, single 4.33 hr exposures to 3670, 4450, or 5240 pg/m (1.4, 1.7, or 2.0 ppm) S02
induced 4.5, 11, and 15 percent yield decreases, respectively. Also, 5.2 to. 15 percent yield
2
losses in soybean cv Wells were observed with intermittent exposures to 240 to 940 pg/m (0.09
to 0.36 ppm) SOp for an average fumigation period of 4.2 hours on 18 different days scattered
from July 19 through August 27 of the soybean growing season (Sprugel et al., 1980). A
greater yield loss (20.5 to 45 percent) resulted when soybeans were exposed for 4.7 hours to
S02 at 790, or 2070 pg/m (0.30 or 0.79 ppm) on 24 different days scattered through the soy-
bean growing season using the ZAPS exposure approach in an attempt to approximate ambient con-
ditions. The latter of the intermittently repeated exposures, therefore, appeared to have a
3
detrimental yield effect as great or greater than single exposures of 5240 ug/m (2.0 pprn) for
4 1/3 hours.
In contrast to effects reported by Sprugel et al. (1980) at 240 and 260 pg/m (0.09 and
0.10 ppm) S0_, Heagle et al. (1974) exposed the soybean cv. Dare in closed-top chambers in the
3
field for 6 hours per day at 260 pg/m (0.1 ppm) S02 for 133 days and noted little effect
other than defoliation. The contrasting results may be due to cultivar (Wells vs. Dare) dif-
ferences. Or, as Sprugel et al. (1980) suggest, the difference in results could be due to
differences in the exposure systems. The closed field chamber of Heagle et al. may have modi-
fied the macroenvironment of the plants being exposed, whereas with the open-air ZAPS system,
o
occasional high peak exposures may result. One such peak of 2100 jjg/ra (0.8 ppm) S02 for one
2-minute sample was measured in 1977. Thus, the reductions in yield seen in the Sprugel et
al. (1980) studies could have been due to the peaks rather than the general mean. Peaks of
this type, however, are not unusual but rather are typical of air pollution episodes, parti-
cularly near point sources. Another possible factor is that in the ambient air there is a
possibility of synergistic interactions. Such interactions with ozone may have increased the
effects of S00 in the Sprugel et al. (1980) study. In 1977, ozone concentrations in the field
3
were monitored during the first part of the season and averaged between 130 and 160 pg/m
(0.05 and 0.06 ppm) during the fumigations and, on at least two occasions, the concentrations
exceeded 260 pg/m (0.1 ppm) for one hour or more. Sprugel et al. (1980), however, point out
that the study of Heagle et al. (1974) produced no evidence for greater-than-additive effects
of SQ?/Q, on soybeans. Again, the conditions in the field under which the soybeans were ex-
posed using the ZAPS system probably more nearly simulate those under which field-grown crops
would normally be exposed, and the ZAPS results rather clearly indicate that repeated 4.7 hr
8-18

-------
exposures to X790 pg/m (>_0.30 ppm) S02 during the growing season markedly reduced (by >20 per-
cent) soybean yields under the conditions prevailing during the study.
As for other commercially important crop species, neither spring wheat or Durum wheat
exhibited significant yield loss effects when exposed to intermittent concentrations of SO,
3
from 660 (0.25 ppm) up to 3140 pg/m (1.2 ppm), 3 times or 7 times over the growing season.
o
When concentrations of 80 (0.03 ppm) to 390 jig/m (0.15 ppm) for 72 hours per week were used
for the growing season these intermittent exposures resulted in no significant yield loss
effects. Barley responded in the same manner,as Durum wheat, showing no significant response
at any of the .S0« concentrations or exposure times. Similar statements can be made regarding
3
the results of repeated 3 hr 660 (0.25 ppm) to 3140 ug/m (1.20 ppm) S0? exposures of alfalfa
noted in Table 8-3 (Wilhour et a!., 1978).
Table 8A-2 (Appendix 8A) presents a summary of studies that investigated effects of S02
on agronomic and horticultural crops grown and fumigated within artificial exposure-chamber or
growth-chamber systems rather than grown in typical field soils. It is difficult to determine
the significance of the results of such studies in relation to similar fumigations under field
conditions. Additionally, with the exception of a relatively few studies, the doses used for
exposure treatments would be considered in excess of those expected under ambient field condi-
tions.
A few major investigations of the effects of S02 on vegetational species growing under
natural conditions have been reported (Linzon, 1971, 198Q; Dreisinger, 1965; Dreisinger and
McGovern, 1970; Materna et al., 1969; Vins and Mrkva, 1973; Legge, 1980; Legge et a!., 1981).
Table 8-4 illustrates the degree of injury to eastern white pines (Pinus strobus) observed
over a 10-year period (1953-1963) in the sulfur-fume-effects area of smelters near Sudbury,
Ontario, Canada (Linzon, 1980). Linzon (1971, 1978) has indicated that a pollution (SO^)
gradient existed within the designated study area, and effects correlated well with this gradi-
ent. Chronic effects on forest growth were prominent where S09 air concentrations annually
3
averaged 45 or 115 ug/m (0.017 or 0.045 ppm) SO-, and such effects were not reported in areas
3
receiving 21 ug/m (0.008 ppm) S02 annually. Although monitoring of S02 was conducted in
these studies, neither concentrations of other air pollutants nor their potential effects were
evaluated by the authors. It is also not clear to what extent high, short-term, peak SO- expo-
sures may have contributed to the effects reported to be associated with these annual average
SOp concentrations.
Ambient air exposures to sulfur dioxide that caused injury to a variety of both sensitive
and non-sensitive species of vegetation are shown in Table 8-5. The data listed in the table
were derived from the work of Dreisinger and McGovern (1970) and has been previously reviewed
and cited in EPA's Revised Chapter 5 for Air Qua!ity Criteria for Sulfur Oxides (U.S. EPA,
1973) and the National Academy of Sciences report on sulfur oxides (NAS, 1978). In addition,
air quality data from the work have been analyzed and discussed by Linzon (1971). As pre-
sented and concluded by Dreisinger and McGovern in their work and cited in the N^S report,

8-19

-------
                                         TABLE 8-3.  EFFECTS OF EXPOSURE TO S02 ON PUNTS UNDER  FIELO CONOITIOHS*
00
po
o
Plant species
Soybeans
Soybeans
Soybeans
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv
Soybeans cv

Wheat
Wheat



Wells
Wells
Wells
Wells
Wells
Wells
Wells
Wells
Dare



Spring Wheat


Spring Wheat


Spring Wheat
Spring Wheat
Durw Wheat

Durum Wheat

Durum Wheat
OuruB Wheat






Type of
exposure
P
P
P
I
I
1
I
I
I
I
I
I

I
I
I

I

I
I
I

I

I
!
4
4
4
4.
4.
4.
4.
4.
4.
4.
4.
6
92
3
3
3

3

72
72
3

3

72
72
Title
1/3 hrs
1/3 hrs
1/3 hrs
2 hrs x 18 exposures
2 hrs x 18 exposures
2 hrs x 18 exposures
2 hrs x 18 exposures
2 hrs x 18 exposures
7 hrs x 24 exposures
7 hrs x 24 exposures
7 hrs x 24 exposures
hrs/day for 43-
and 133 days
hrs x 8 exposures/growing
hrs x 7 exposures/growing
hrs x 3 exposures/
growing season
hrs x 7 exposures/
growing season
hrs/wk/growing season
hrs/wk/growing season
hrs x 3 exposures/
growing season
hrs x 7 exposures/
growing season
hrs/wk/growing season
hrs/wk/growing season
Concentration
ug/n* (pp«)
3670
4450
5240
240
260
500
660
940
310
790
2070
260

season 100
season 1180
660-3140

660-3140

80- 260
390
660-3140

660-3140

80- 260
390
(1
(1
(2
(0
(0
(0
{0
(0
(0
(0
(0
(0

(0
(0
(0.

(0.

(0.
.4)
.7)
.0)
.09)
.10)
.19)
.25)
.36)
.12)
.30)
.79)
.10)

.04)
.45)
25-1.20)

25-1.20)

03-0.10)
(0.15)
(0.

(0.

(0.

25-1.20)

25-1.20)

03-0.10)
(0.15)
Effects
4.5% loss in yield (N.S.)
11% loss in yield**
15% loss in yield*
6.4% yield loss**
5.2X yield loss*
12. 2X yield loss**
19.2% yield loss**
15.9% yield loss**
12. 3% yield loss**
20.1% yield loss**
45.3% yield loss**
No effect until 92nd
day, 12X defoliation (H.S.;
No yield effect
No yield effect
No yield effect

No yield effect

No yield effect
No yield effect
Ho yield effect

No yield effect

No yield effect
42% yield loss (H.S.)
References
Killer et al. 1979


Sprugel et al. 1980 and
Hi Her et al. 1980






Heagle et al. 1974
J
Sij et al. 1974

Wilhour, et al. 1978





Wilhour, et al. 1978

-------
                                                           TABLE 8-3,  (continued)


co
i
ro
*-i
Plant species
Barley
Barley
Barley
Barley
Alfalfa
Alfalfa
yestern Wheat Grass
Prairie June Grass
Type of
exposure
I
I
I
I
I
I
c
c
Tine
',
3 hrs x 3 exposures/
growing season
3 hrs x 7 exposures/
growing season
72 hrs/wk/growi ng season
72 hrs/wk/growi ng season
3 hrs x 3 exposures/
growing season
3 hrs x 7 exposures/
growing season
growing season
Concentration
ug/m (ppm) Effects References
660-3140 (0.25-1.20) No yield effect Wilhour et
660-3140 (0,25-1.20) No yield effect
80- 260 (0.03-0.10) No yield effect
390 (0.15) 44% yield loss (N.S.)
660-3140 (0.25-1.20) No yield effect Wilhour "et
660-3140 (0.25-1.20) No yield effect
50-H60 (0.02-0.10)a Increasing S content Oodd et al.
with increasing SO-
50-260 (0.02-0.10)8 Significant decrease1'
in digestible protein

al. 1978
al. 1978
, , 1978
P = Peak (short-term, 8 hrs or less), I = intermittent and C = continuous exposure.'
*Statistically significant change at P <0.05.
"""Statistically significant change at P <0.01.
N.S. Results were not significant at P <0.05 probability level.
aValues are for a geometric mean.

-------
         TABLE 8-4.  THE DEGREE OF IHJURY OF EASTERN WHITE PIKE OBSERVED AT-VARIOUS DISTANCES FROM THE SUDBURY SHELTERS FOR 1953-63
Trees with
Current
Year's
Forest Saapling
Foliage
Trees with 1-Year-
Old (1962) Foliage
Injured


Station Injured in
(Distance and
Direction fron
Sudbury)
West Bay
(19 miles NE)
Portage Bay
(25 miles NE)

Grassy to Emerald Lake
(40-43 niles NE)
Lake Matinenda
(93 niles W)
Correlation
Coefficient (r)
August
1963
(%)
2.0

1.1


0.4

0.6

0.96

June
1963
(X)
38.0

21.5


2.5

0.3

0.96

August
1963
(X)
77.9

55.6


16.7

2.1

, 0.93**

Trees with 2- Year
Old Foliage,

Injured
in June
1963
(X)
96.0

77.0


$7.5

10.1

0.90**

Lacking
fn
August
1963
(X)
20.6

15.2


»9?.l

£9

0.94**

Net Annual
Average
Cain or
Loss in
Total
Volume,
1953-1963
(X)
-1.3

-0.5


*1,8~

-+2.1

0.90**

Average S0« b
Concentration
Annual
Average
Hortality
1953-1963
(X)
2,6

2.5


1.4

0.5 '
-
" 0.81




Degree of SO,
Damage
Acute and chronic
injury
Mostly chronic
and little
acute injury
Very little chronic
injury
Control: no S0«
injury


for Total
Measurement
Period 1954-
1963
(jg/ra3 (ppn)
115 (0.045)

45 (0.017).


21" (OlOOS)

3 (0.001)c :
(Sturgeon Falls)


 aLinzon (1971)  (Pollutants other than SO* were not measured and the monitoring was done several  niles  from the pine stands.)
 bDreisinpr (1961)
 C0ata for 5-nonth growing season-1971
 *p < 0.05
**p < 0.10
Derived fron Linzon,  1980.        .  .
                                                                                                                                               Is,

-------
        TABLE 8-5.   AMBIENT EXPOSURES TO SULFURDIOXIDE THAT CAUSED
                           INJURY TO VEGETATION3

Exposure where injured
SOo Concentration, M9/m3 (ppm
1 h
1070 (0.41)
1070 (0.41)
1100 (0.42)
1180 (0.45)
1180 (0.45)
1210 (0.46)
1210 (0.46)
1210 (0.46)
1360 (0.52)
1470 (0.56)
1650 (0.63)
1650 (0.63)
1680 (0.64)

1730 (0.66)
1730 (0.66)
1730 (0.66)
1830 (0.70)
1830 (0.74)
1990 (0.76)
2040 (0.78)
2150 (0.82)
2150 (0.82)
2280 (0.87)
2280 (0.87)
2310 (0.88)
2330 (0.89)
2460 (0.94)
2830 (1.08)
2990 (1.14)
3430 (1.31)
3510 (1.34)
2 h
1000 (0.38)
1000 (0.38)
1020 (0.39)
890 (0.34)
920 (0.35)
1000 (0.38)
1180 (0.45)
1130 (0.43)
1150 (0.44)
1020 (0.39)
1150 (0.44)
1550 (0.59)
1470 (0.56)

1130 (0.43)
1180 (0.45)
1410 (0.54)
1210 (0.46)
1650 (0.63)
1410 (0.54)
1810 (0.69)
1700 (0.65)
1700 (0.65)
1940 (0.74)
2070 (0.79)
1680 (0.64)
2150 (0.82)
2330 (0.89)
2070 (0.79)
1970 (0.75)
2020 (0.77)
2380 (0.91)
) for averaging
4 h
860 (0.33)
890 (0.34)
680 (0.26)
660 (0.25)
660 (0.25)
730 (0.28)
1130 (0.43)
1130 (0.43)
760 (0.29)
6fiO (0.26)
630 (0.24)
890 (0.34)
1130 (0.43)

970 (0.37)
1150 (0.44)
1050 (0.40)
710 (0.27)
1390 (0.53)
760 (0.29)
1150 (0.44)
1180 (0.45)
1620 (0.62)
1440 (0.55)
1830 (0.70)
1100 (0.42)
1600 (0.61)
1830 (0.70)
1310 (0.50)
1180 (0.45)
1180 (0.45)
1310 (0.50)
periods of:
8 h
790 (0.30)
680 (0.26)
340 (0.13)
550 (0.21)
550 (0.21)
550 (0.21)
550 (0.21)
550 (0.21)
520 (0.20)
390 (0.15)
310 (0.12)
450 (0.17)
1000 (0.38)

520 (0.20)
860 (0.33)
550 (0.21)
370 (0.14)
1020 (0.39)
370 (0.14)
790 (0.30)
680 (0.26)
1210 (0.46)
760 (0.29)
1310 (0.50)
710 (0.27)
1070 (0.41)
1180 (0.45)
660 (0.25)
600 (0.23)
600 (0.23)
890 (0.34)
Plant

Willow
Larch
Quaking aspen
Bracken fern
White pine
White birch
Bean
Alder
Jack pine
Buckwheat
Barley
Oats, peas, rhubarb
Lettuce, tomato,
potato
Large-toothed aspen
Austrian pine
Timothy
Red clover
Raspberry
Radish
Red pine
Balsam poplar
Sugar maple
Celery
White spruce
Swiss chard
Red oak
Cabbage
Carrot, cucumber
Witch hazel
Beet, turnip
Spinach

Derived from Dreisinger and McGovern, and reprinted from NAS Sulfur Oxides
Document, 1978, p.  6-29.
                                  8-23

-------
"one could be  reasonably certain of avoiding foliar injury to vegetation in the study ar,ea if
concentrations did  not  exceed 1830, 1050, 680 or 470 pg/m3 (0.70,. 0,40,,0.26, or 0.18 ppm) as

1-, 2-, 4-, or 8-h means, respectively."                          ,

     These conclusions  appear to be based  on the following discussion in  the  Drefeinger  and

HcGovern (1970) paper:
          "What levels  of  sulphur dioxide are considered potentially injurious to vegetation?
          It was shown in previous reports (Dreisinger, 1965, 1967)  that most of the injury to
          vegetation  that  occurred near  recorder stations in the Sudbury  area happ'ened when
          concentrations and  durations of  sulphur dioxide reached or  exceeded the following
          levels during the day time.

               0.95 ppm for 1 hour    )                                   .
            or 0.55 ppm for 2 hours   )
            or 0.35 ppm for 4 hours   )      Intensity factor of 100
            or 0.25 ppm for 8 hours   )            ,                                      •

          "It is not true that injury always occurs when these levels are reached or exceeded.
          Whether or  not damage  actually happens depends, to  a  large extent, on the environ-
          mental conditions prevailing  at the time of tile fumigation.   There have been occa-
          sions in  the  Sudbury area when levels  two, three, and even four times those listed
          above were  recorded and  no  injury occurred.  Those fumigations  simply happened at
          times when vegetation was resistant."

          "Conversely,  on  occasions,  some  very  sensitive species  were  injured with  levels
          approximately 25% less than the above.  In each case the injury occurred during June
          or July,  which,  for the Sudbury area,  are  the  months  when the most rapid growth of
          vegetation takes place.  Hot, humid weather prevailed in most cases, producing condi-
          tions extremely  favourable  for a  high photosynthetic rate  and  in  turn increased
          susceptibility to sulphur dioxide."

          "The occasions,  however,  when this happened, were quite  rare,  only nine times in a
          ten-year  period,  and the injury that did  result was not  extensive so  that  for  all
          intents and purposes, it is only when all S02 levels reach the 100-intensity mark or
          more that the potential for wide-spread injury is present."

     The general  conclusions  of  Dreisinger and  McGovern  (1970) have  essentially  been rein-

forced by Jones et al  (1974) and MeLaugh!in.(1981).
     Mclaughlin (1981), although recognizing that, especially in arid regions of the U.S.,  many

species of vegetation would not be visibly injured even in some cases by SO- concentrations of
11 ppm  for 2  hours as  suggested by  Hill et  al  (1974),  focused on  a  subset of vegetational

species in the more hot and humid Southeast which he believed to represent a spectrum of plant
response ranging from sensitive  to tolerant (Mclaughlin and Lee, 1974).  Mclaughlin concluded
that 20% of the selected species would show visible injury at peak (5 minutes),  1-,  and 3-hour

concentrations of 3140, 1700,  and 920 pg/m3 (1.20, 0.65,  and 0.35 ppm),  respectively.

     Jones et al.  (1974) also reported, as previously presented in  Table 8-2, a wide range of

SO* concentrations  which,  at peak  (maximum  of 5. minutes), 1-hour,  and 3-hour  durations,  in-
jured different species of plants that were classified as sensitive, intermediate, or resistant

to S02 exposures.   However, because of the many randomly occurring environmental factors affect-
ing injury, e.g.,  plant resistance, temperature, moisture, etc.,  Jones (1981) recommended that


                                            8-24

-------
an analysis of probability of injury occurrence should be applied prior to any generalizations
regarding the specific concentration of SO™ causing plant injury.
The work of Hill et al. (1974) also reinforces the widely accepted statement by Dreisin-
ger and McGovern (1970) that hot, humid weather produces conditions extremely favorable for
increased plant susceptibility to sulfur dioxide.
The studies of Vins and Mrkva (1973) and Haterna et al. (1969), although reporting foliar
and growth increment losses in forest trees as being due to SO^, were conducted in areas of
fluoride contamination or used only sporadic monitoring schedules, respectively. Pollution
gradients were evident and S0« exposure was probably involved, but conclusive proof of losses
attributed to SOp was not presented as was done for the Sudbury area.
Table 8A-3 (Appendix 8A) summarizes the results of tree studies that have utilized arti-
ficial exposure chamber systems under laboratory conditions. Two of the laboratory studies
listed used levels close to ambient concentrations. Houston (1974), for example, reported
significant effects on pine seedlings at levels as low as 390 (jg/m (0.15 ppm) SO™ for 6 hr;
however, the use of selected clones known to be sensitive to SO- hinders extrapolation of
these findings to ambient conditions. Most of the remainder of the studies presented in Table
8A-"3 have used doses above expected occasional exposures under field conditions.
The effects of exposing seedling tree species to SO, levels within or approaching the
3
ambient range are summarized in Table 8-6. Concentrations of greater than 660 (jg/m (0.25
ppm) SO, for 2 hours were required to induce slight injury to several pine species (Berry,
1971), but overall trends for increasing foliar injury do not follow increasing dose for coni-
fers per se. Smith and Davis (1978) exposed several conifers (pine, spruce, fir and Douglas
fir) to 2620 |jg/m3 (1.0 ppm) S02 for 4 hours or 5240 pg/m3 (2.0 ppm) S0« for 2 hours and only
the pines developed necrotic tips at the higher concentration. However, studies done with
tree seedlings under artificial conditions are difficult to extrapolate to expected yield-loss
effects under field conditions. Of the studies reviewed and accepted without caveats, none
demonstrated significant height or annual increment growth effects.
In an evaluation of SO, effects on photosynthesis, Keller (1977) used field chambers to
expose potted white fir, Norway spruce and Scotch pine to three concentrations of SO, (130,
260 and 520 |jg/m ; 0.05, 0.10, and 0.20 ppm) and a control of 0.0 ug/m for 10 weeks each
during the spring, summer, and fall, respectively. Several types of photosynthetic response
rates were obtained; however, trends of decreasing rate occurred as concentrations increased,
especially when administered during the 10-week spring period. Effects were less during the
3
summer and fall periods and spruce responded positively to S0« concentrations of 130 ug/m
(0.05 ppm) during the initial part of the fall period. Keller (1977) also reported that even
with the most severe depression of photosynthesis, there were no visible foliar symptoms in
evidence. Keller (1980) utilized a similar fumigation system and similar S0~ concentrations to
study effects on the annual ring width in two clones of Norway spruce. He reported signifi-
cantly depressed CO™ uptake with higher levels (260 and 520 ug/m ; 0.10 and 0.20 ppm S02, 10

8-25

-------
                           TABLE 8-6.  SUHHARY OF THE EFFECTS RESULTING FROM THE EXPOSURE OF SEEOLIHG TREE SPECIES IN THE LABORATORY
oa
Plant species
Scotch Pine

Eastern White Pine

Red Pine
Loblolly Pine
Shortleaf Pine
Virginia Pine
Slash Pine
Jack Pine
Jack Pine


Balsam Fir

Fraser Fir

White Fir

Blue Spruce

Douglas Fir

Austrian Pine

Ponderosa Pine

Ponderosa Pine
American Elm

American Elm

American Elm


Type of
exposure
P

P







C


P

P

P

P

P

P

P

P
P

P

C


Time
1 hr-6 hrs

2 hrs - 6 hrs

2 hrs
2 hrs
2 hrs
2 hrs
2 hrs
2 hrs
24 hrs


2, 4, 5 hrs

2, 4, 5 hrs

2, 4, 5 hrs

2, 4, 5 hrs

2, 4, 5 hrs

2, 4, 5 hrs

2, 4, 5 hrs

9 hrs/days x 8 wks
6 hrs

8 hrs

12 hrs


Concentration
pg/m3 (ppm)
1310-5240 (0.5-2.00)

6.6-1310 (0.025-0.5)

660 (0.25)
660 (0.25)
660 (0.25)
660 (0.25)
660 (0.25)
660 (0,25)
470-520 (0.18-0.2)


5240, 2620 and 1310
(2.0, 1.0 and 0.5)
5240, 2620 and 1310
(2.0, 1..0 and 0.5)
5240, 2620 and 1310
(2.0, 1.0 and 0.5)
5240, 2620 and 1310
(2.0, 1.0 and 0.5)
5240, 2620 and 1310
(2.0, 1.0 and 0.5)
5240, 2620 and 1310
(2.0, 1.0 and 0.5)
5240, 2620 and 1310
(2.0, 1.0 and 0.5)
1180 (0.45)
5240 (2.0)

2620 (1.0)

5240 (2.0)


Effects
Foliar injury - except
at 5 hrs at 0.5 ppm
Foliar injury even on
tolerant clones at 6 hrs
at 0.15 ppm

Foliar injury; most sensitive
period of plant 8-10 wks

Foliar injury
Inhibition of foliar lipid
synthesis; inhibition
reversable
Foliar injury only at
4 hrs, then less than 4%
Foliar injury only at
4hrs, then less than 4%
Foliar injury only at
4 hrs, then less than 4%
Foliar injury only at
4 hrs, then less than 4%
Foliar injury only at
4 hrs, then less than 4%
Foliar injury except at
4 hrs and 0.5 ppm
Foliar injury except at
5 hrs and 0.5 ppm
Tip necrosis
Severe foliar injury,
defoliation
Inhibited stomatal
closing
Stomatal closing induced,
increases content when
plants fumigated in light
References
Smith and Davis, 1978


Houston, 1974
Berry, 1972

Berry, 1972


Berry, 1974

Malhotra and Kahn, 1978







Smith and Davis, 1978







Evans and Miller, 1975
Constant! nidou and
Kozlowski, 1979a
Noland and Kozlowski,
1979
Temple, 1972

-------
                                                                    TABLE 8-6.   (continued)
TO
ro
Plant species
Chinese Elm
Chinese Elm
Gingko
Gingko
Trembling Aspen
Sugar Maple
Sugar Maple
Black Oak
Black Oak
Whfte Ash
White Ash
Norway Haplt
Pin Oak
Norway Spruce

Norway Spruce

Norway Spruce

Beech Spruce
Beech Spruce
Beech Spruce
Type of
exposure
P
C
P
C
P
C
I
C
I
C
I
C
C
C

C

C

C
C
C
Tine
6 hrs
24 hrs/day for 30 days
6 hrs
24 hrs/day for 30 days
3 hrs
24 hrs/day -for 1 wk
4 hrs/day for 3 wk
24 hrs/day for 1 wk
4 hrs/day for 3 wk
24 hrs/day for 1 wk
4 hrs/day for 3 wk
24 hrs/day for 30 days
24 hrs/day for 30 days
10 wks

10 wks

10 wks

16 wks
in winter

Concentration
Mg/m3 (ppm)
5240 (2.0)
1310 (0.5)
7900 (3.0)
1310 (0.5)
920; 1700 (0.35; 0.65)
1310 (0.5)
1310 (0.5)
1310 (0.5)
1310 (0.5)
1310 (0.50)
1310 (0.50)
1310 (0.50)
1310 (0.50)
1310 (0.05)

260 (0.10)

520 (0.20)

130 (0.05)
260 (0.10)
520 (0.20)
Effects
100% leaf necrosis
7 days for chlorosis
50% leaf necrosis
14 days for chlorosis
Foliar injury; 2% and 2.3%
54% reduction 1n photosynthesis
43% reduction in photosynthesis
48% reduction in photosynthesis
74% reduction in photosynthesis
20% reduction in photosynthesis
7% reduction in photosynthesis
12 days for chlorosis
30 days for chlorosis
No foliar injury, 25% <
volume
Foliar injury 38% <
volume
Foliar injury 53% <
volume
A few buds killed
Many buds killed
50% buds killed
References
Temple, 1972
Temple, 1972
Temple, 1972
Temple, 1972
Karnosky, 1976



Carlson, 1979


Temple, 1972
Temple, 1972
Keller, 1980

Keller, 1980

Keller, 1980

Keller, 1978
Keller, 1978
Keller, 1978
           *Data for this table come from Table 8A-3 in Appendix 8A.
           P, C, and I indicate peak (short-term average for 8 hrs or less),  continuous and intermittent exposures.

-------
weeks), but a trend was only noted for the 0.05 ppm SO, treatment; visible symptoms and trends
of reduced cambial growth occurred only at the higher concentration.
Based on the limited data available concerning the nonwoody components of native ecosys-
2
terns, there appear to be no adverse yield effects below 160 \ig/m (0.06 ppm) SO, for 6 weeks
(Table 8A-4, Appendix 8A). Most of the acceptable literature deals with long-term exposures
(several weeks) and results in SO, levels well above ambient atmosphere concentrations. There
3
is some indication of beneficial yield effects below 160 pg/m (0.06 ppm) SOp on one species
(i.e., Indian ricegrass). As stated in Section 8.2.8, low doses of SO, may increase the pro-
3
ductivity of certain crops. Concentrations of 80 to 160 ug/m (0.03 to 0.06 ppm) continuously
for 6 weeks of exposure increased the productivity of desert grass by 8 percent over the con-
trol plants which were grown in indigenous soils (Ferenbaugh, 1978). Other studies have also
demonstrated beneficial effects, but experimental conditions included.a sulfur-deficient grow-
ing medium.
More definitive dose-response studies, both with and without the addition of other pollut-
*
ants (see Section 8.4), are needed before the biologically significant features of typical re-
gional exposure regimes can be positively delineated. The National Crop Loss Assessment Net-
work (NCLAN), which has only recently begun its studies, is a step in this direction.
8.4 EFFECTS OF MIXTURES OF SULFUR DIOXIDE AND OTHER POLLUTANTS
Ambient atmospheres usually contain more than one pollutant. Atmospheric monitoring of
long-distance transport of photochemical oxidants and oxidant precursors (Husar et al., 1978;
U.S. EPA, 1971), and the presence of acidic precipitation over large areas of the eastern
United States (Cogbill, 1976) have documented the fact that emissions from specific sources
are mixed with ambient concentrations of one or more pollutants. Extrapolation from results
of single-pollutant effects on vegetation under ambient field conditions must be approached
with caution because reactions to pollutant combinations under controlled conditions, the in-
teraction of constantly changing environmental factors, and fluctuating pollutant exposures
must also be evaluated before a conclusive statement of the importance of such interactions
can be made. Reinert (1975) and Reinert et al. (1975) have prepared the most recent reviews
of this area of investigation. Some examples of the available literature follow.
8.4.1 Sulfur Dioxide andOzone
Table 8A-5 (Appendix 8A) summarizes studies on the effects of sulfur dioxide in combina-
tion with ozone.
A more-than-additive effect on vegetation was first reported for ozone and S09 (Menser
3 3
and Heggestad, 1966). Tobacco was severely injured by 80 pg/m (0.03 ppm) ozone and 630 pg/m
(0.24 ppm) SO, when the pollutants were combined for either 2 or 4 hours, whereas when used
alone neither pollutant produced foliar symptoms.
Since that first report, the effects of mixtures of ozone and SO, have been studied using
a variety of plant species. Radish and alfalfa plants showed more-than-additive foliar injury
after a 4-hr exposure to a mixture of 200 ug/m3 (0.10 ppm) 03 + 260 pg/m3 (0.10 ppm) S02

8-28

-------
(Tingey et al., 1973a), but less-than-additive growth reduction (top and root weights) from an
8 hr/day, 5 day/wk,  5-week exposure of radish (alfalfa total exposure time unknown) to a mix-
ture of  100 pg/m3 '(0.05 ppm)  03 + 130 pg/m3  (0.05 ppm)  S02  (Tingey et al., 1971a; Tingey and
Reinert, 1975).  Greater-than-additive foliar injury effects have also been reported for broc-
coli  and  tobacco,   while  additive  or  less-than-additive  effects  have  been  noted  for
cabbage, tomato,  lima bean,  bromegrass,  spinach,  onion, and  soybean  (Tingey et al., 1973a).
Soybean  has exhibited insignificant less-than-additive foliar injury  effects  (Tingey et al.,
1973a)  while  exhibiting  significantly greater-than-additive  growth  effects (Tingey  et al.,
1973b).
     Most researchers examining the effects of pollutant mixtures have utilized standard means
comparisons to express the  responses.  These  tests usually  do  not  adequately  evaluate  the
interaction:    the  failure of  one pollutant  to  be  consistent at  different concentrations of-
the second pollutant.   Reinert and Nelson (1980) used sums of squares partitioning by factor-
ial analysis to  examine the effects of 1310  (jg/m   (0.5 ppm) SOp and  490  pg/m   (0.25 ppm) 0-
(4-hr exposures, 4 times, 6 days apart) on Begonia.  A significantly less-than-additive effect
was found for  flower weight in 1  of  5 cultivars.   The same technique was utilized by Reinert
and Weber (1980) to  evaluate the  effects of  490 pg/m3 (0.25 ppm) 03 and 660 pg/m3 (0.25 ppm)
SO,  (exposed  4 h/day, 3  days/wk,  for 11  wks)  on  soybean  (Dare); an additive  effect of the
pollutant mixture was demonstrated.
                                                          3                                  3
     Field-grown soybeans  (cv Dare)  exposed  to 200 pg/m   (0.10 ppm) 0,  alone  or  200 pg/m
                        3
(0.10 ppm) 03 + 260 pg/m  (0.10 ppm) S02 for 6 hr/day for 133 days in field chambers exhibited
injury  and defoliation.   Due  to the mixture,  injury and yield were increased (9 percent) and
decreased  (19  percent),  respectively, when  compared to the  ozone-alone treatment,  but  the
differences were  not  significant  (Heagle  et  al.,  1974).  Two  cultivars  of bean  exposed to
sulfur dioxide and ozone showed interactive effects between these two gases, but the magnitude
and  direction  of the  effects depended on  the  cultivar  and on the  pollutant concentrations
(Jacobson and Colavito, 1976).
     Many studies have been conducted on the  effects of mixtures of sulfur dioxide and ozone
on  eastern white pine (Pinus  strobus  L.)  (Costonis, 1973; Oochinger and Heck, 1969; Houston,
1974; Houston  and Stairs,  1973).   Genetic control  of sulfur dioxide and ozone tolerance in
this species has  been demonstrated for low concentrations of S09 (66 pg/m  , 0.025 ppm) and 0,
          3
(100 pg/m  ;  0.05 ppm)  for  only 6 hr  with  consistent injury  to the  exposed sensitive clones
(Houston and Stairs, 1973).  Houston (1974) later used mixtures of sulfur dioxide and ozone at
concentrations simulating actual field conditions and reported that even the lowest concentra-
                       3                               3
tions of 03 (100 pg/m ;  0.05 ppm) and SO,  (130 pg/m ; 0.05 ppm) for 6  hr in  mixture caused
more serious damage than that  resulting from either pollutant alone at similar concentrations.
A  less-than-additive  effect  on foliar injury was noted when Scotch pine trees were exposed to
660 pg/m3  (0.25 ppm)  S02 and/or 270 pg/m3 (0.14 ppm) or 570 pg/m3 (0.29 ppm) 03, 6 hr/day for
varying  time periods  (Neilson  et al., 1977).

                                            8-29

-------
                                           3                              3
     Fumigation of aspen clones to 100 pg/m  (0.05 ppm) 03 and/or 520 ng/m  (0.20 ppm) S02 for
3 hr  resulted in  a number of  plants in the mix exhibiting  more-than-additive  foliar injury
(Karnosky, 1976).
     Oshima  (1978),  using enclosed,  round  chambers  in which the air was  constantly stirred,
conducted a  field experiment to  assess the effect of S00  on the yield of  red  kidney beans,
                                      *                       •          3
Phaseolus vulgaris, grown  in  pots,   the beans were expSsed Co 260 ug/m  (0.10 ppm) of S02 for
6 hr/day  for a total of 335  hours  and at the  same time  were exposed to a  gradient of ozone
doses.  The  ozone  concentrations were  those  that  resulted when  ambient air  at Riverside,
California,  was  passed directly  into the chamber or  when  25,  50,  75 and 100 percent of the
                                                                             \
ambient air  was  passed through  a  charcoal  filter.    The  temperature,  light,   and humidity
approximated that of ambient air.   Plants were in the  chambers for a total  of 78 days.  An in-
teraction between  ozone and  S0?  was  documented  in  the 50 percent  carbon  filtered treatment
(i.e., 50 percent  ambient  air)  (51 ppm-hrs) and produced a significant reduction in yield (37
percent) and plant  biomass.   The  data also suggested the possibility of an interaction in the
25 percent ambient  (75  percent filtered treatment) (28.22  ppm-hrs)  in  that yield was reduced
17 percent,  but  this  reduction  was not significant (p =0.20).  Exposure of beans to S09 alone
           3
at 260 ug/m   (0.10  ppra) did not produce detectable plant or yield responses nor was an inter-
active effect  observed  when  ozone doses exceeded 51 ppm-hrs; however, ozone alone at the same
concentrations did significantly (> 65%) reduce yield.
8.4.2  Sulfur Dioxide and Nitrogen Dioxide
     The occurrence  of S02 and nitrogen dioxide  (N02) together in the atmosphere has been
associated with power plant plumes as well as mobile sources.   However,  ambient concentrations
of NOp seldom  reach the injury threshold, and the literature for N0~ suggests that any injury
associated with N02 results from interactions with other pollutants  (Jacobson and Hill, 1970).
     No injury occurred to oats,  beans, soybeans, radish,  tomato,  or tobacco following, expo-
sure  for  4  hr  with up to 3760  yg/m  (2 ppm)  NOp  or 1310  ug/m   (0.50 ppm) SOp.   However,
following exposures  at 0.10 ppm  of  each gas  for 4 hr,  injury was noted on  all  species;  at
0.05 ppm of  each gas,  slight injury was noted on  all species except tomato  (Tingey et al.,
1971a).  A  greater-than-additiVe  suppression  of the   apparent  photosynthetic  rate of alfalfa
                                           3                                   3
was obvious  when treated  with  to 660 |jg/m   (0.25 ppm)  of  SO,, and/or 470 ug/m   N0? for 2 hr
(White et al., 1974).  At 0.15 ppm of each gas singly, there were no measurable effects, but a
7-percent suppression of apparent photosynthetic" rates was noted in  the mixture (White et al.,
1974).
     Field exposure of  seven  different species of plants  indigenous to the cold desert areas
of  the  southwestern  United  States  to  1310-28820  pg/m3  (0.50-11.0  ppm)  S02  singly,  or
1310-28820 pg/m3  (0.50-11.0 ppm)  S02 and 190-9400 ug/m3  (0.10-5.00  ppm) N02 combined in 2-hr
fumigations  resulted in no evidence of more-than-additive  foliar injury (Hill et al., 1974).
     More-than-additive  foliar  injury was  noted on  radish leaves   exposed  for 1 hr  to 1310
    •a                                o
ug/m   (0.50  ppm)  S02 and/or  940 pg/m  (0.50 ppm) N02-   No  interactive  effects were found for

                                            8-30

-------
other plants tested (oats, Swiss chard, and pea) (Bennett et al.,'1975). More-than-additive
2
effects have been associated with the enzyme activity in pea plants exposed to 520 ug/m
(0.20 ppm) S0? and 190 |jg/m (0.10 ppm) N0? for 6 days; peroxidase activity was increased and
ribulose-1,5 diphosphate carboxylase activity was decreased (Horsman and Wellborn, 1975).
Some selected SO, + NO- combination studies are shown in Table 8A-6 (Appendix 8A).
8.4.3 Sulfur Dioxide and Hydrogen Fluoride
Suppression of linear growth and leaf area in the absence of foliar injury of Koethen
o
orange plants exposed to SOp (2100 jjg/m ; 0,80 ppm) and/or hydrogen fluoride (2.3-13.0 ppb)
for 23 days was no greater than additive. Satsuma mandarin plants exposed to the same condi-
tions for 15 days exhibited only additive foliar injury and no growth suppression at all
(Matsushima and Brewer, 1972). Greater-than-additive foliar injury was exhibited-by barley
and sweet corn from to 160-210 |jg/m3 (0.06-0.08 ppm) SOp and/or 0.60-0.90 ppb hydrogen
fluoride for 27 days. Using higher concentrations of SO,, for only 7 days resulted in simply
additive foliar injury. Pinto beans were not injured in any of the treatments (Mandl et al. ,
1975).
8.4.4 Sulfur Dioxide, Nitrogen Dioxide and Ozone
Fujiwara et al. (1973) combined S02, N02, and 03 at concentrations ranging from 0 to
0.2 ppm in an artificially controlled environment and exposed peas and spinach for 5 hr.
Ozone was the most injurious, SO, was next, and N02 elicited only minor injury. More-than-
additive foliar injury followed treatment with SO, + 03, but only additive effects were
observed with S02 + N02 or N02 + 03- The addition of N02 to the 03 + S02 had little effect on
foliar injury. Reinert and Gray (1981) examined the effects of 0.2 or 0.4 ppm each of 03,
S0?, and NO/> (3- or 6-hr fumigations) on the growth of radish. The main effect of each pol-
lutant and the potential interactive effects of each mixture were examined through parti-
3
tioning. Sulfur dioxide depressed the root/shoot ratio at both 520 and 1050 ug/m (0.2 and
0.4 ppm); however, when N0» and SO, were both present there was a greater-than-additive
f. t. 3
depression of the root/shoot ratio at 1050 [jg/m (0.4 ppm).
8.4.5 Summary
As can be seen from the preceding analysis of research dealing with pollutant mixtures,
plant species vary in their response and the type of response (additive, less-than-additive,
greater-than-additive) may depend on the parameter measured. Understanding of how pollutant
combinations influence plant growth and development, and how environmental factors can modify
those responses is still fragmentary. Insufficient information exists to determine the influ-
ence of pollutant sequencing during combination exposures, meteorological influences,- the
effect of various cultural practices, and many other -variables in relation to vegetational
effects induced by SO, combined with other pollutants.
There is a need to determine the best technique for evaluating the effects of pollutant
mixtures. Only recently has the partitioning technique been utilized to elucidate the effects
of individual pollutants in pollutant mixtures>. This technique allows for separation of the

8-31

-------
main effects of each pollutant and also provides a statistical test of the significance of
potential interactions between pollutants.
Analysis of the data indicates that interactions can occur between pollutants and,
because the occurrence of pollutant mixtures is typical of the ambient atmosphere, knowledge
of interactive effects is very important. However, the nature of the effects of pollutant
mixtures is extremely complex. Most research studies have necessarily taken a rather
simplistic approach to this complex problem. It is, therefore, difficult to relate these
relatively few results to "real world" situations.
8.5 EFFECTS OF NON-POLLUTANT ENVIRONMENTAL FACTORS ON SULFUR DIOXIDE PLANT EFFECTS
The physical environment plays an extremely important role in determining response to SO^.
Most of the information accumulated at present deals with the factors leading to or inhibiting
the ingress of S0? into stomata and the immediate plant response as determined by the meta-
bolism of the plant at the time of exposure. The metabolic state of the plant is likewise
affected by the physical environment. As discussed in the following sections, the response of
any given plant species may also be quite different from any other given species grown under
identical physical conditions.
8.5.1 Temperature
Temperature plays an important part not only in determining the metabolic rate of the
plant, but also in determining (with moisture, fertility, and light) species diversity and
richness of a given ecosystem (NAS, 1978). The primary path of S02 entry into the leaf is
through stomata. Temperature exerts an effect on the guard cells that control the stomatal
opening and closing and thus the entry of SO^. Temperature regimes that increase the physio-
logical activity of the plant may also increase the plant response to S02 (Heck and Dunning,
1978). It is generally believed that plant sensitivity increases with temperature over a wide
range, from about 4° to 35°C (Guderian, 1977; Rist and Davis, 1979). Several studies suggest
that the greater resistance of conifers to SO- in the winter is attributable to lower rates of
physiological activity (NAS, 1978). However, according to Guderian and Stratmann (1968), in
areas with SO^ emissions, winter wheat and winter rye are more severely injured than the sum-
mer varieties. Guderian (1966) interpreted this to be due to gas exchange taking place
through stomata at temperatures as low as -2°C.
8.5.2 Relative Humidity
Relative humidity exerts an important control over plant sensitivity to SO* by affecting
stomatal opening and closing (Bonte et al., 1975; Majernik and Mansfield, 1970; Mansfield and
Majernik, 1970; Buron and Comic, 1973) and by affecting the internal leaf resistance to SO-
flux (McLaughlin and Taylor, 1981). Although plant sensitivity increases with relative
humidity, Setterstrom and Zimmerman (1939) found rather large changes (> 20 percent) were re-
quired to cause a change in plant sensitivity once RH levels become > 40 percent. McLaughlin
and Taylor (1981), in laboratory studies, found 2-to 3-fold increases in SO, uptake by kidney
3
beans over a range of S02 concentrations (420-1680 ng/m ; 0.16-0.64 ppin) as relative humidity

8-32

-------
was increased from 35 percent to 78 percent during exposure times of < 3 hours. According to
Zimmerman and Crocker (1934), although relative humidity Is Important in governing sensitivity
and, consequently, the sensitive plant population, 1t is not as important as the tissue turgidity
which may be- influenced by soil moisture as well as relative humidity. Based on the
water relations in certain trees and their sensitivity at relative humidities of over 75
percent, 50 to 75 percent, and below 50 percent, Halbwachs (1976) has rated plants as sensi-
tive, intermediate, and tolerant, respectively.
8.5.3 Light
Light controls stomatal opening and thus plant sensitivity. Plants are more tolerant
when fumigated in darkness with SO, or when held in the dark for several hours before exposure
(Zimmerman and Crocker, 1934). This relationship is complex, since injury is greater if the
night exposure follows a daylight exposure (HAS, 1978).
Setterstrom and Zimmerman (1939) observed that buckwheat grown at a light intensity of 35
percent or less of full sunlight was more sensitive to SO, than when grown under full sunlight.
Other investigators have found that injury was more severe when tomato stems and foliage were
fumigated on clear days than on cloudy days (HAS, 1978).
Plants seem to be more sensitive from midmorning to midafternoon, despite the high light
intensity that might continue after midafternoon (Rennie and Halstead, 1977; Thomas and Hen-
dricks, 1956). At the same time, plants may be more sensitive in the morning during good wea-
ther, but may become more sensitive if temperature and light increase in late afternoon in bad
weather (Van Haut and Stratmann, 1970).
8.5.4 Edaphic Factors
Soil factors influence directly and indirectly the responses of plants to SO,. Soil fer-
tility, moisture, and soil physics directly influence plant sensitivity to SO- (NAS 1978).
Adequate soil moisture and the resultant stomatal opening have been shown to increase the de-
gree of plant sensitivity, whereas wilting conditions confer tolerance (Setterstrom and Zim-
merman, 1939; Zahn, 1963a; Zimmerman and Crocker, 1934). As long as plants are grown with an
inadequate supply of water, they are less sensitive to S0? than are plants grown with an ade-
quate supply, even though the moisture content of the soil,is the same at the time of fuml
gation (Setterstrom and Zimmerman, 1939). Withholding water from some crops during periods of
•high pollution risk has been suggested as a technique to reduce injury (Brandt and Heck, 1968).
Soil fertility has a significant influence on plant response to S0_. Some plants become
more tolerant to SO, upon fertilization (Enderlein and Kastner, 1976; Zahn, 1963b). However,
with eastern white pine, increased nitrogen, phosphorus, and potassium concentrations in the
greenhouse raised tolerance (decreased needle necrosis) in sensitive clones but did not pre-
vent chlorotic banding in the field (Cotrufo and Berry, 1970). Nitrogen and sulfur deficien-
cies were correlated with increased tolerance to S0? 1n tobacco and tomato (Leone and Brennan,
1972). Conversely, nutrient deficiencies increased S02 sensitivity in alfalfa (Setterstrom
and Zimmerman, 1939). Fertilization of several dicotyledons with a complete fertilizer has

8-33

-------
been effective in decreasing their sensitivity to S02, but similar treatment of monocotyledons
like oats and barley has been ineffective (Van Haut and Stratmann, 1970; Zahn, 1963b).
8.5.5 Sulfur Dioxide and Biotic Plant Pathogen Interactions
Plant disease is caused by the interaction of a plant and a pathogen acting under suit-
able environmental conditions. The influence of S00 directly or indirectly on the interrela-
^»
tionships of a given plant and its possible biotic pathogens has been difficult to investigate.
Additionally, whenever the variables of the physical environment are considered within such
experimental sequences, the subject becomes even more difficult to examine. Heagle (1973) and
Laurence (1978) have provided the most recent reviews of the interaction between air pollut-
ants and plant parasites.
In seven of nine plant diseases reported in S0?-related studies as reviewed by Laurence
(1978), there was either no effect or a reduction in disease development demonstrated.
Disease increased only in needle cast of pine (Chiba and Tanaka, 1968), and increased virus
titer of southern bean mosaic virus has been reported by Laurence et al. (1981).
In a recent study by Laurence (1979) maize and wheat were exposed to 0.10 or 0.15
ppm SCL for either 2 or 10 days and innoculated at various times with He!minthospori urn maydis
or Puccinia graminis. The ability of these fungi to infect either corn or wheat, respective-
ly, was inhibited by SO,, exposures; greater inhibition occurred if plants were fumigated prior
to inoculation attempts.
Studies done under ambient conditions without monitoring of other pollutants have sug-
gested a decrease in disease incidence in areas of higher S02 pollution with the possible ex-
ception of tho'se pathogens that are better able to invade weakened plants. If SOp exposure
has resulted in an overall weakened condition, other agents, such as root-invading fungi, may
be able to gain entrance into an otherwise resistant host. Such is the suspected reason for
increased incidence and severity of attack by the root pathogen Armillaria jneljhea in trees
weakened by SO- (Donaubauer, 1968; Jancarik, 1961; Kudela and Novakova, 1962).
The effects of SO, on infection by organisms other than fungi have also been studied.
Abies concolor (white fir) and A. yeItchif were severely attacked by plant lice in an environ-
ment of high S0», but Pinus strobus (white pine) was attacked less and P. Griffithii
(Himalayan pine) and P. sylvestris (Scotch pine) were not attacked (Stewart et al., 1973).
The direct influence of S0? on plant pathogenic fungi has been demonstrated and a review
published by Saunders (1973); no direct effects of S02 on plant pathogenic bacteria have been
reported.
8.6 PLANT EXPOSURE TO PARTICULATE MATTER
8.6.1 Deposition Rates
Deposition of particles is strongly dependent on particle size. Most sulfates and ni-
trates are found in the size range of 0.1 to 1.0 |jm, and very little information is available
on the deposition rate for these particles (see Chapter 6). Shinn (1978) divided particulate
deposition into three categories based on particle size:

8-34

-------
CATEGORY 1. Particles more than 10 pm In diameter; includes dust
and spores.
CATEGORY 2. Particles between 1 |jm and 10 |jm in diameter where
the collection efficiency is highly dependent on the
particle diameter,
CATEGORY 3. Submicrometer particles between 0.1 and 1.0 |jm in
diameter, which have a nearly constant collection
efficiency.
Current experimental data suggest that collection efficiencies in Category 3 are less
than one-tenth of those in Category 2 (Shinn, 1978). According to Clough (1975), in the range
of wind speeds normally encountered, the larger particles in the atmosphere are much more
efficiently collected than the smaller fraction.
Little (1977) evaluated the effects of leaf surface texture on the deposition of mono-
disperse polystyrene aerosols on the leaf surfaces. Rough and hairy leaf discs collected
5.0 (jm particles up to seven times more efficiently than did smooth leaves. Very large dif-
ferences in particle deposition velocities were observed between the laminas, petioles, and
stems of each species. The velocity of deposition of particles to plant surfaces varies
according to both wind speed and particle size.
Further information on atmospheric transport, transformation and deposition of particulate
matter may be found in Chapter 6 of this document.
8.6.2 Routes and Methods of Entry Into Plants
8.6.2.1 Direct Entry Through Foliage—Foliage is continuously subjected to natural and man-
made coarse particles that are insoluble or sparingly soluble in water. Coarse particles in
general are too large to enter leaves through stomata. Based upon a review by Meidner and
Mansfield (1968) that presented stomatal data for 27 species of plants (e.g., pine, oak,
corn, soybean, and tobacco) the overall average pore (opening) width was 6 micrometers and
accounted for 0.15 to 2.0 percent of the total average stomatal area. In certain cases,
such as with cement kiln dusts (Lerman and Parley, 1975) and other types of aggregate par-
ticles (Smith, 1973), a limited amount of stomatal clogging can occur. This is apparently
dependent on the statistical probability of the particulate matter falling on the stomata,
the size of the particle, and the stomatal aperture. In many plants, the stomatal opening
is on the lower surface. Cement kiln dust forms a crust on leaves, twigs, and flowers.
According to Czaja (1961), crusts of this type form because some portion of the settling
dust consists of calcium aluminosilicates typical of the clinker from which cement is made.
Hydration of the dust on the leaf surface results in the -formation of a gelatinous calcium
aluminosilicate hydrate, which later crystallizes and solidifies to a hard crust.
When coarse particles are water soluble or have some water-soluble components, plant
uptake of ions from the leaf surface does occur. Because of analytical difficulties, the
exact magnitude of the uptake is difficult to measure. Since it is not possible to predict
the efficiency of any washing procedure used to remove particles from the leaf surface, it

8-35

-------
is difficult to delineate and separate the concentration of a given element on the leaf sur-
face from its concentration inside the tissue. In addition, leaching of elements from
within the tissue is known to occur during the washing (Little, 1973).
Smith (1973) evaluated metal contamination of urban woody plants by using a variety of
washing procedures. Indirect evidence from all these studies suggests that a variable
concentration of metals originating from coarse particles can accumulate in plant foliar tissue
through direct uptake.
At this time, one of the significant problems in deriving conclusions concerning the
magnitude of direct foliar deposition and uptake of atmospheric particulates is the lack of
coordinated data on size and frequency distribution of the particles and their chemistry,
rates of deposition, and dose in conjunction with changes in tissue concentrations over time
relative to background conditions.
8.6.2.2 Indirect Entry Through Roots—Many of the inorganic constituents of particulate air
pollutants occur naturally in the soil; others of equal or,greater importance do not. Deposi-
tion of these pollutants may increase the soil concentrations of the chemical species in ques-
tion. Some of the chronic effects caused by particulate air pollutants may result from
changes in soil physics and chemistry and from increased plant uptake of either the added
elements associated with the particles themselves or some other soil-borne elements made more
available by the influence of the deposited particles (Guderian, 1980).
It should be recognized that only a portion of the total elemental content of the soil is
available at a given time for plant absorption (Brady, 1974). As uptake of elements proceeds,
there may be a redistribution of nutrients or toxicants in the soil.
The availability of nutrients or other chemical elements from the soil is strongly
influenced by type, chemical composition, and acidity of the soils. Plant nutrients, when
present in optimal amounts, are usually available at a neutral pH; however, when the soil
becomes acidic, toxic elements such as aluminum become available.
8.7 REACTION OF PLANTS TO PARTICLE EXPOSURE
8.7.1 Symptomatology of Particle-Induced Injury
Particle-induced injury to plants has most often been associated with sustained accumula-
tion of particles such as dust or fly ash. Few investigations have dealt with direct or
indirect chemical interactions at the plant surface or subsequent effects. The toxicity of
accumulated heavy metals in soils has been established for several plant species.
The various forms of particles and their associated impacts on plants have been reviewed
(Darley, 1966; Lerman and Parley, 1975; U.S. Department of Health, Education, and Welfare,
1969; U.S. EPA, 1977). Krupa et al. (1976) and Linzon (1973) have also prepared extensive
reviews of heavy metal deposition and impact. Tolerance of plants For heavy metals and fine
particles and their bioenvironmental impacts have also been reviewed (U.S. EPA, 1975).
Dusts—Dust directly affects plants by coating exposed plant parts, including leaves,
stems, flowers, and fruits (Jennings, 1934; Katz, 1967; Linzon, 1973). Depending on the

8-36

-------
chemical nature of the particles and environmental conditions, deposits may accumulate as dry
dusts, as encrustations in the presence of free moisture, or as greasy films or tars. Encrus-
tations of dust particles on leaves result in reduced gas exchange, increased temperature, re-
duced photosynthesis, and eventual yellowing and tissue desiccation (Daessler et al., 1972;
Parish, 1910).
Terminal growth reduction and chlorosis of 2-year needles of hemlocks coated with heavy
limestone dust deposits have been reported. Manning (1971) also reported that fungal propa-
gules increased, but bacterial numbers decreased on such needles. Brandt and Rhoades (1972)
reported long-term changes in plant community structure and species composition, and later
indicated that radial growth rates were reduced in the tree species involved (Brandt and
Rhoades, 1973). On the exposed site they demonstrated a reduction in radial growth of red
maple (18 percent, chestnut oak (29 percent), and red oak (23 percent) but a 76-percent in-
crease in radial growth of tulip poplars as compared with representatives of these species
growing on a similar but nonexposed site. Reduction in growth of the dominant species (oak,
maple) probably gave a competitive advantage to tulip poplar and a greater than expected in-
crease may have occurred.
The primary result of the deposition of limestone dusts has been the raising of pH levels
of the environment, particularly tree bark. The substrate pH changes were followed by lichen
community changes in which replacement of acid-tolerant lichen communities by more alkaline-
tolerant species occurred. (Gilbert, 1976). A recovery of lichen diversity resulted in areas
where S0? pollution was of importance prior to limestone dust emission. No exact pollutant
dose of either limestone dusts or SO, was reported. Winter SO, levels were estimated to
3
average 65 ug/m .
Cement kiln^dusts collected from precipitators have been applied to vegetation. Visible
effects were demonstrated on beans following application of particles > 10 urn at rates
2 9
0.05 mg/cm /day to 0.38 mg/cm /day for 2-3 days. The lower dose induced a slight reduction in
carbon dioxide exchange, and the two higher doses reduced carbon dioxide uptake by 16-32 per-
cent (Darley, 1966).
The accumulation of dust caused increased reflection of solar radiation at wavelengths of
400 to 760 nm and has been demonstrated to reduce photosynthesis (Ricks and Williams, 1974).
Conversely, increased absorption of solar radiation by dusted leaves at wavelengths 750-1350 nm
has been demonstrated to lead to heat stresses within the leaf tissues (Spinka, 1971).
Growth and yield effects induced by the accumulation of dust have recently been reviewed
(U.S. EPA, 1977). Conflicting reports of yield increases and decreases from such accumulation
appear to be caused by variations in doses applied, substrate nutrient balances and pH, and
other specific physiological interferences with processes such as pollination of fruit trees
(Anderson, 1914).
Dusts, therefore, have only been considered of importance to vegetation growing near
emission sources. Accumulation of dusts has been demonstrated to reduce photosynthesis and
radial-increment growth of some forest tree species but has increased them in other species.
8-37

-------
The phytotoxicity of heavy metals, arsenic, and boron has been demonstrated after
accumulations in soils and subsequent uptake by various plants. Table 8-7 presents a
summation of toxic effects of individual elements (Krupa et a!,, 1976). Published reports of
direct effects on plants from specific sources are discussed in the following paragraphs.
Arsenic—Only one study was available to show foliar uptake of airborne arsenic
(Linzon, 1977). Phytotoxicological studies in the vicinity of gold smelters in Ontario,
Canada revealed the occurrence of several injuries to vegetation (primarily fireweed,
Epilobium angustifolium) as induced by airborne arsenic compounds in a sulfurous plume.
Chemical tissue analysis of affected leaves revealed arsenic at 200 ppm as compared to
1 ppm arsenic in leaves collected 50 miles from the source area. Linzon (1977) suggested
10 points of evidence leading to the conclusion of the airborne nature of the toxicant
including elevated concentrations on sides of plants closer to the source and correlation
of arsenic emissions with corresponding changes in tissue concentrations on an annual basis.
Arsenic sprays have been applied to the foliage of many plants to hasten fruit matura-
tion by causing premature defoliation and chemical changes in the fruit. For example, lead
arsenate sprayed on grapefruit trees caused a "fruit gumming" reminiscent of boron defi-
ciency (Liebig, 1966). Boertitz et al. (1976) reported that arsenic deposited at 22 mg/kg on
soil reduced the yield of wheat, rye, winter rape, and red clover by 25, 8, 0, and 6 per-
cent, respectively.
Cadmium—Host biologically active cadmium enters plants through root uptake (Jordan,
1975). Small oxide particles (0.01 to 0.03 pm) may enter leaves through stomata, but it
is thought that the oxides remain largely inert. Cadmium accumulated by apple leaves may
be translocated and incorporated into fruits as they develop (Yopp et al., 1974).
Copper—Wu and Bradshaw (1972) demonstrated a selection of individual plants of the grass
Agrostis stolonifera growing near metal smelters, thus indicating an indirect effect of
within-species simplification within a population through selection.
Lead—Davis and Barnes (1973) reported reduced growth of loblolly pine and red maple
-4 -3
seedlings in pots of two forest soils treated with 2 x 10 to 2 x 10 molar lead chloride.
Lead toxicity symptoms may include fewer and smaller leaves, reduced plant size, leaf
yellowing, and necrosis of elder, sugar beet, squash, and bush bean (Schoenbeck, 1973).
Plants growing in soil already high in these metals tended to be more sensitive to the
addition of metals by air pollution.
Nickel—Plants can absorb and translocate airborne nickel salts (NAS, 1975). Once
inside the plant, nickel affects photosynthesis and other processes such as stomatal
function (Bazzaz and Govindjee, 1974). In cases of incipient nickel toxicity to vegeta-
tion, no definite symptoms have been observed other than growth repression. In cases of
moderate or acute nickel toxicity, chlorosis-resembling symptoms of iron deficiency are
common (Anderson et al., 1973).
8-38

-------
                      TABLE 8-7.   PLANTS  SENSITIVE TO HEAVY METALS, ARSENIC, AND BORON AS ACCUMULATED IN SOILS AND TYPICAL SYMPTOMS EXPRESSED
      Metal
     Plant
                                                                           Symptoms
                                                     Reference
      Arsenic
      Boron
Snap bean, lima bean, onion, pea, cucumber,  alfalfa,  legumes,  sweet
corn, strawberry (on light and sandy soils).
Reduced germination of seeds, rotting of roots,      Liebig (1966)
leaf wilt, brown to red coloration of leaves,        Linzon (1977)
reduced yield in fruit trees, death.
Barley, var. Atlas 46; lima bean, var.  Henderson; kidney bean,  var.    Yellowing of leaf tips,  necrosis between lateral      Bradford (1966)
Navel; oats, var. Riverside; onion,  var.  Cabot;  pea,  var.  Alaska;      veins  and midrib of monocotyledons, marginal  leaf    Krupa and Kohut
peach, var. J.H. Hale; persimmon, var.  Kaki;  rose,  var.  Snow White;    scorch,  downward cupping of leaves, reduced flower-       (1976)
soybean, var. Wilson, var 0_'T°otan;  wheat,  var.  Opal; yellow zinnia,   ing, fruit lesions.                                   Yopp et al. (1974)
CO

ca
      Magne-
        sium

      Manga-
        nese
      Mercu
                Red oak,  birch,  trembling aspen, beet, carrot, celery, green pepper,  Reduced root elongation, general growth retardation. Yopp et al. (1974)
                lettuce,  radish, soybean-, Swiss chard, tomato, winter wheat.                                                               Jordan (1975)
                Bean,  citrus  fruits, corn, mustard.
                Bean dwarf French, var. Carters; beet, corn, fescue, lettuce,
                lupine,  loblolly pine, red maple.
Orange (only case known from published literature).
                                                                      Stunted  root development,  chlorotic leaves,  reduced  Reuther and Laban-
                                                                      vegetative  growth.                                      auskas (1966)
                                                                     Stunted  root growth,  shoot retardation,  increased
                                                                     leaf abscission,  reduced yields.
Leaf rolling, spiraling inhibition of leaf emer-
gence.  Narrow leaf development.
Alfalfa, broadbean, cabbage,  cauliflower,  cereals,  citrus,  clover,    Necrotic  spots  on  leaves,  necrosis of internal
lespedeza, pineapple,  potato, tobacco,  tung,  barley,  var. Atlas 46,   bark,marginal leaf yellowing,  incurling of
var. Herta; yellow birch,  cranberry,  peanut,  potato,  var. Kesweck;    leaf margins.
alfalfa, apple, apricot, barley,  bean,  brussels  sprout,  carrot,
clover, cotton, lettuce, medic, orange,  pea,  potato,  sugar  beet,
vetch, wheat.
Broadbean, oxalis, sunflower,  bean,  butterfly weed,  cinquefoil,
fern, Hydrangea. Mimosa,  Oxalis,  privet,  sunflower,  willow.
      Nickel     Citrus.fruit, alfalfa, oats, var. Victory; pear.
                                                                                     Possible reduced growth.
                                                                     Repression  of  vegetative  growth,  leaf chlorosis,
                                                                     white or  light yellow and green striping.
                                                     Yopp et al. (1974)
Embleton (1966)
                                                     Yopp et al. (1974)
                                                     Labanauskas (1966)
                                                     NAS (1973)
                                                     Yopp et al. (1974)
                                                     Lagerwerff,(1972)
                                                     Jacobson and
                                                          Hill (1970)

                                                     Vanselow (1966)
                                                     Yopp et al. (1974)

-------
                                                                       TABLE 8-7.  (continued)
       Metal
Plant
Symptoms
Reference
       Potas-
         sium     Orange,  tung  (only case known for published literature).
                                                                 Fruit coarseness  and leaf necrosis,  leaves  curl      Ulrlch  and
                                                                 downward,  marginal  leaf necrosis,  intervena!           Ohki  (1966)
                                                                 chlorosis, plant  dieback.
       Zinc      Oats,  orange, tung, barley, var. Trail; corn, var.
                 Whatley's  Prolific, var. Ida Hybrid 330; cowpea, var,
                 Suwannee;  wheat, var. Gaines; barley, citrus, oats,
                 sugarbeet.
                                                                "Uniform chlorosis,  reduced  terminal  growth,          Chapman  (1966)
                                                                 twig dieback,  chlorotic striping of  leaves,            Yopp et  al.
                                                                 stems stiff and erect,                                 (1974)
       Source:   Adapted  from Krupa et a*K (1976)
oo

-------
8.7.2 Classification of Plant Sensitivity--Particles
Coarse particles have not been shown to elicit responses in plants in such a manner as to
allow plants to be placed into sensitivity classes similar to those developed for gaseous pol-
lutants. Accumulations of particulate matter, such as roadside dusts, cement, quarry
particle emissions, or other forms of deposits, such as fly ash, are deposited on all
surfaces and induce responses discussed under the symptom portion of this chapter. How-
ever, heavy metals do elicit different responses in plants, and therefore it is possi-
ble to develop lists of particularly sensitive plants.
Heavy metals are constituents of many coarse particles emitted from various sources. To
our knowledge, there has not been an organized effort to establish the toxicity of specific
chemical constituents of particles in relation to sensitivity groups of vegetation under
field conditions. Table 8-7 lists plants that may be sensitive to heavy metals following
deposition and various symptoms as expressed following their respective accumulations.
8.8 EXPOSURE-RESPONSE RELATIONSHIPS—PARTICLES
Review of the published literature suggests that it is not possible at present to
define, even crudely, exposure-response relationships for the effects of particulate air
pollutants on plants. Many reports deal only with gross visible effects or tissue accumula-
tion of one or more constituents of the atmospheric particulate matter present. The emphasis
of research has been on settleable coarse particles. Since these are conglomerates of several
pollutants, their chemical compositions are frequently ill-defined, although their sources
have often been identified. Little information could be found on the effects of fine
particles on vegetation.
Where cause-and-effect relationships have been suggested, generally no information is
presented on the actual concentration, particle size, and frequency distributions.
Deposition rates ajid plant effects vary significantly with particle size. Few studies are
available where two independent scientists have evaluated the effects of particles on vege-
tation with closely comparable physical and chemical properties under reproducible condi-
tions.
Much of the literature refers to particles from point and line sources and their
accumulation in or on soils and vegetation. Tissue accumulation of a given element must
be considered as a plant response. Soil scientists have contributed most of the informa-
tion on plant toxicity symptoms that has been obtained under laboratory conditions.
Since many of the plant effects observed are due to the accumulation of elements up to toxic
concentrations, tissue concentrations prior to actual exposure will affect the amount and ele-
mental composition of particles that plants can tolerate. Exposure-response relationships are
determined by the concentrations of various elements in the soil where the plant is growing.
8-41

-------
Effects of surface accumulation of cement kiln dust on bean leaves have been Investigated
by Parley (1966). Doses of 0.6-3.8 g/m per day were applied for 2 or 3 days, and foliar in-
jury and reductions in carbon dioxide exchange were observed. Reductions in carbon dioxide
exchange of up to 33 percent were noted in the absence of visible foliar injury. Bean leaves
* " P
dusted with cement kiln dust at the rate of 4.7 g/m, pejr day for 2 days and then exposed to
dew developed leaf rolling and intervenal necrosis (Lerman and Darley, 1975). Leaves not
exposed to dew following the dust treatment remained asymptomatic.
Reduced yields and injury to leaves and flowers of several plant species were observed
when the plants were exposed once a week for 4 weeks to a dust containing cadmium, lead, cop-
per, and manganese (Krause and Kaiser, 1977). Yield reductions of up to 36 percent were
noted.
Plants accumulate different elements at differing rates. Tissue concentrations of some
elements are known to be significantly higher in the vicinity of a source for those elements
in comparison with background or baseline concentrations. This elevated tissue concentration
nay be due to direct foliar uptake or uptake from the pollutant accumulations in the soil. In
many cases, elevated tissue concentrations of a given metal or metalloid are not paralleled by
visible injury.
Demonstration of injury symptoms on vegetation under field conditions as a result of
accumulation of metals or metalloids is rare. The demonstrated cases are for strip mine
wastes. Predicted effects from atmospheric deposition include plant community changes, chron-
ic long-term" physiological changes, and indirect effects through a modification of the re-
sponse to other types of stress. Thus, the state of our knowledge concerning the effects of
particles on vegetation is inadequate at this time and does not allow the development of accu-
rate dose-response curves.
8.9 INTERACTIVE EFFECTS ON PLANTS WITH THE ENVIRONMENT—PARTICULATE MATTER
Few studies have examined the influence of dusts or heavy-metal-containing particles on
the interactions between organisms capable of causing disease and the predisposition of the
host plant to the disease process.
Infection due to Cercospora spp. increased on sugar beet leaves exposed to cement dust
containing 36 percent calcium oxide and 15 percent silica (Schoenbeck, 1960). Increased
occurrence of fungus-induced leaf spots on wild grape and sassafras has been observed near a
source of heavy emissions of limestone dust (Manning, 1971). After examining 40 leaves in each
of five locations exposed and not exposed to the dust accumulations, he found that disease
development was two to three and six to seven times greater, respectively, for the two leaf spot
fungi in the exposed areas.
Natural exposure to combustion nuclei from automobile exhaust, which supplied increased
levels of Aitken nuclei and atmospheric lead, reduced germination of uredospores of Puccinia
8-42

-------
striiformis (stripe rust of wheat); jm situ development of disease was prolonged about 4 days.
Similar studies at an unexposed site did not reveal reduced spore germination (Sharp, 1967,
1972).
8.10 EFFECTS OF SULFUR DIOXIDE AND PARTICULATE MATTER ON NATURAL ECOSYSTEMS
The previous sections of this chapter have discussed the effects of sulfur dioxide and
particulate matter on individual plants. This section discusses the effects of these sub-
stances on natural ecosystems since, because of their complexity, these systems respond to
environmental perturbations differently from individuals or monocultures of organisms.
Ecosystems are basically energy-processing systems, the components of which have evolved
together over a long period of time. They are composed of living organisms together with
their physical environment (see Chapter 7, Table 7-1). The boundaries of the system are
determined by the environmental conditions that determine the kinds of life forms that can
exist in a particular habitat or region. The plant and animal populations within the system
are the objects through which the system functions. Ecosystems respond to environmental
changes or perturbations only through the response of the organisms of which they are composed
(Smith, 1980).
Relationships among the various ecosystem components are structured, not haphazard. The
living (biotic) and the nonliving (abiotic) units are linked together by functional interde-
pendence. Processes necessary for the existence of all life, the flow of energy and the cy-
cling of nutrients, are based on the functional relationships that exist among the organisms
within the system (Odum, 1971; Smith, 1980; Billings, 1978). Because of these relationships,
unique attributes emerge when ecosystems are studied that are not observable when individuals,
populations or communities are studied. For a more detailed account of ecosystems, see Chapter
7, Section 7.1.2.
The discussion that follows emphasizes the response of terrestrial ecosystems to sulfur
dioxide and particulate matter. Natural ecosystems are seldom, if ever, exposed to a single
air pollutant. Therefore, the responses observed under ambient conditions cannot conclusively
be attributed to a single substance such as sulfur dioxide or particulate matter alone.
8.10.1 Sulfur Dioxide In Terrestrial Ecosystems
Sulfur is an element that is essential for the normal growth and development of plants
and animals. It is a basic constituent of protein and is required in large amounts by some
plants. Under normal circumstances sulfur in rainwater and in soil organic matter is suffi-
cient to meet plant requirements. Excessive sulfur in the form of sulfur dioxide, however, can
be toxic to plants. The phytotoxic forms of sulfur, routes of entry into plants, and the
symptomatology of S02 injury to plants have been discussed in the preceding sections.
Within any ecosystem, nutrient sources are to be found in the atmosphere, in living and
dead organisms, and in available and unavailable salts in the soil and rocks. The nutrients
are cycled from the living to the nonliving components and back again. Air pollution, how-
ever, can disrupt nutrient cycling by altering the amounts in the various compartments and the
rate of flow among them (Smith, 1980).
8-43

-------
There are two types of biogeochemical cycles, the sedimentary and the gaseous. The sul-
fur cycle includes both. Sulfur enters the atmosphere from the combustion of fossil fuels,
from volcanic eruptions, from the surface of oceans and in gases released by decomposition pro-
cesses (see Chapter 4). Anaerobic decomposition of organic matter releases hydrogen sulfide
(HpS) into the atmosphere where it is quickly oxidized into sulfur dioxide. Sulfur dioxide is
soluble in water and may be carried back to earth in rainwater as dilute sulfuric acid (FLSO.)
(Smith, 1980). Regardless of the source, sulfur in soluble form is taken up by vegetation and
is incorporated through a series of metabolic processes, including photosynthesis, into
sulfur-containing amino acids. Sulfur is transferred from the producers to the consumers in
food and through excretion and death back to the soil and to the sediments in the bottoms of
ponds, lakes, and seas where bacterial action releases it as hydrogen sulfide or as sulfate.
Sulfur in the long-term sedimentary phase is tied up in organic and inorganic deposits in the
soil, and sulfur is added to ecosystems through geological weathering and meteorological pro-
cesses, with the latter being the predominant source. Weathering and decomposition permit
*
sulfur to enter into solution and to be carried into aquatic and terrestrial ecosystems. In
its gaseous state, sulfur is circulated on a global scale (Figure 8-3),
Based on empirical watershed studies (Likens et a!., 1977; Shriner and Henderson, 1978)
and modeling (Coughenour, 1978), the soil is a major reservoir for atmospherically derived
sulfur within ecosystems. The majority of soil sulfur is unavailable to vegetation and is
organically bound in the humus (May and Downes, 1968). Microbial activity oxidizes organi-
cally bound sulfur to sulfates. Sulfates may be taken up by plants or leached from the soil.
The rate of sulfur released from the organic to the inorganic compartment is the major factor
controlling the movement of sulfur between the soil and vegetation (May et al., 1972; Moss,
1976). A distinction between natural and agroecosystems is that soils in agroecosystems
through continued cropping are depleted of their supply of organic sulfur, and it is not
renewable; therefore, sulfur must be added as fertilizer. Sulfur dioxide brought down in
precipitation is also added to the soil. The amount of sulfur added to soils through pre-
cipitation will depend on the industrial activity of the surrounding area (Kamprath, 1972).
The influence of anthropogenic sources of sulfur on the sulfur cycle is most pertinent
when addressed on a regional basis (Granat et al., 1976) as Shinn and Lynn (1979) have done
for the northeastern United States. Comparing global versus regional sulfur cycling,
atmospheric sulfur additions are not equally distributed over the global land areas, and the
northeastern United States experiences anthropogenic atmospheric additions of sulfur that are
28.4 times that expected if additions were distributed uniformly over the globe. The most
notable contrast between the global and regional sulfur cycle is the importance of atmospheric
sulfur sources. Globally, natural processes far exceed anthropogenic contributions, whereas
in the northeastern United States human activities generate 12.5 times the amount of sulfur
released by nature. A total of 27 x 10 tons of SO. enters the northeastern regional atmos-
fi
phere annually and 13 x 10 tons are deposited within this region by wet and dry deposition;
the remaining sulfur is exported to other areas of the globe.
8-44

-------
Oxidation
Direct Utilization
by Plants
t . _ , —f _ .

I ' ' /; 	 i Animals in the
j jr — ~~^i/ 	 : Grazing Food Cham
! \f \
{ Autotrophs I j

	 ( (-orestgj \
Grassland ">
	 '" 	 Mr ,s --!»• gQ2
j t
| "\ 1 Death & Wastes
j ~~-_^" Detritus Food Cham
' ! 	 	 	
f i
noryanic (* Oxidation
Sulfate N


l
i
^O7 Spontaneous
Bacterial , Sulfhydryl Oxidation
Reduction^ Sulfur in
to H2S N R-SH Atmosphere
x\
Oxidation

....._.__ Sulfur

hermg ' ' y J
ocks _ , _ ,. ^
Storage of Sulfur or -^ 	 '
Sulfur Compounds in
1
Volcanic _^_ ,, ,.
r- ^ — ^^ Hoo
Eruptions
Combustion
of
and Sedimentary Rocks Fuels
 Figure 8-3. The sulfur cycle (organic phase bounded by dashed line).
Source: Clapham (1973).

-------
Based on the above information,-man is clearly a major source of atmospheric sulfur with-
in the United States and, in the Northeast alone, anthropogenic sources exceed all others by a
o
factor of 12.5. Within this region, S02 levels annually average 16 |jg/m (0.006 ppm) (Shinn
and Lynn, 1979), which is several times that recorded in pristine areas. The immediate fate
of approximately 60 percent of this atmospheric sulfur is deposition on terrestrial and
aquatic ecosystems; however, the subsequent fate of sulfur»within the ecosystem is not fully
known. Experimental evidence from forested watersheds coupled with results from simulation
models indicates that sulfur in ecosystems is highly mobile. Although sulfur levels in the
soil and vegetation compartments in aggrading and mature forest ecosystems impacted by SOp
Increase with time, the majority of sulfur deposited annually is exported out of the system in
stream flow.
The conclusion that S02 emitted into the atmosphere through anthropogenic activity is
ultimately transferred to terrestrial and aquatic ecosystems is well documented (Meszaro et
a!., 1978). Unfortunately, the fate of sulfur in the ecosystem after deposition is not fully
resolved. The issue is critical because ecosystems subjected to excess nutrients or toxic
materials do not commonly distribute them uniformly throughout the system but, rather, prefer-
entially sequester them in specific pools or compartments. In addition, sulfur dioxide as a gas
can cause injury to the vegetative components of an ecosystem so that energy flow and the
cycling of other nutrients as well as sulfur may be disrupted.
8.10.2 Ecosystem Response, to Sulfur Dioxide
Pollutants often reduce community diversity when an ecosystem becomes simplified through
the removal of pollutant-sensitive plant species. Woodwell (1970) has described the effects
of pollution on the structure and physiology of ecosystems. The studies that follow discuss
the ecological changes in a forested ecosystem resulting from sulfur gas emissions.
The Kaybob I and II gas plants (Fox Creek, Alberta, Canada), which emit S02 during the
removal of hydrogen sulfide from natural gas, are located within the transition Montane-Boreal
forest that is dominated by a mixed assemblage of deciduous and coniferous trees; however,
white spruce (Picea glauca) is predominant in successional stands on well-drained soils.
(Winner and Bewley, 1978a,b; Winner et al., 1978). Because these white spruce forests have
less species variation than other sites, they were selected for analysis along a transect
showing decreasing SO, stress. The Kaybob facility began operation in 1968, and the field
study was conducted during the summers of 1975 and 1976. From 1973 to 1975, it is estimated
that the Kaybob facility emitted approximately 71,000 kg per day of SOp (Winner and Bewley,
1978b).
Relative species diversity showed no gradient pattern of response to SO,; however, the
percent of coverage by all understory plants, including vascular plant species and mosses,
showed a marked increase with distance from the source (Winner and Bewley, 1978b). White
spruce seedlings close to the refinery were reduced in number. Changes in moss communities
8-46

-------
were conspicuous and included decreasing values for moss canopy coverage, moss carpet depth,
dry weight, and capsule number and for the frequency of physiologically active versus inactive
moss plants. Close to the source, there were no mosses at all (Winner and Bewley, 1978a).
These results indicate that species diversity, particularly in the mosses, has changed as a
consequence of sulfur gas emissions. The changes in diversity observed during the summers of
1975 to 1976 probably were initiated soon after the facility began operation in 1968 and the
effects will be observed as long as the facility is operational. Atmospheric SO, concentra-
tions were not reported, but it is evident that operational upsets, particularly during the
early years of plant operation, resulted in high emission rates from the low elevation flare
sources causing high S02 concentrations near the ground in this area.
In a subsequent study, Winner et al. (1978) used sulfur isotope ratios to determine the
34
fate of sulfur emissions from the natural gas plants. The 8 S value of SO, emitted from gas
refineries is higher than that for nonindustrial sources in the study area. Through the use
of this tracer, it was determined that mosses absorb only airborne sulfur, whereas the sulfur
in conifer needles is derived from both air and soil. Krouse (1977) in his study of the Ram
River area of Alberta made a similar determination. Winner et al. (1978) conclude that
because mosses act as a sink and accumulate large quantities of sulfur, they are more
vulnerable than vascular plants to effluents such as S0?. Further, they state that mosses
participate in the process of soil formation, prevent erosion, and play a role in succession.
Removal of the moss carpet would ultimately be expected to influence the growth of the
vascular plants in the ecosystem because when mosses are not present, soils accumulate S0?
rapidly.
In another, more detailed study of the effects of low-level sulfur dioxide emissions-on a
forested ecosystem, it was observed that the main ecological process that was directly and in-
directly affected by the sulfur gas emissions was nutrient cycling (Legge et al., 1981, Legge
1980). The study concluded that the chronic exposure to sulfur gas emissions over time re-
sulted in an alteration of the mineral nutrient balances of the various ecosystem components
and modified the biological relationships among the components. Further, the four-year study,
begun in 1972 concluded that despite measurable deterioration of the forest ecosystem in the
West Whitecourt study area, it did not appear that the sulfur gas emissions caused irrever-
sible ecological degradation, particularly in view of the significant reductions in sulfur gas
emissions that have resulted since the inception of the West Whitecourt Gas Plant situated in
west central Alberta, Canada (Legge et al., 1981; Legge, 1980).
The region studied by Legge et al. (1981) was being subjected to the effluents from.the
West Whitecourt "sour gas" processing plant. "Sour gas", is natural gas (methane) that con-
tains hydrogen sulfide (H~ S). For commercial use, the hydrogen sulfide is removed from the
methane as elemental sulfur by a chemical process. Any H9S not converted to elemental sulfur
f- o
is incinerated in excess air and methane in a. high temperature (580 C) furnace and oxidized to
SO,, which is then vented to the atmosphere from a tall (122 m) incinerator stack. Smaller

8-47

-------
flare stacks are used to vent the effluents from the burning of small waste quantities of sul-
fur recovery gas plant process and compressor gases. The incinerator stack is the main source
of SO,, emissions, except when operational upsets occur within the gas plant. At these times,
the flare stacks may for short periods of time contribute more gas emissions to the atmosphere
daily than the incinerator stack. Since the gas plant began functioning in 1961-62, the S0?
output has dropped from 600 to 36 metric tons per day in 1976 because of improved operating
procedure.
The vegetation of the Whitecourt study area is within the predominantly forested subre-
gion of the Boreal Forest of Canada and is characterized as a transition forest area between
the Boreal and Subalpine Forest Regions. The tree populations in the area are hybrid repre-
sentatives of lodgepole pine x jack pine (Pinus contorta Loud, x P. banksiana Lamb.) and of
the alpine fir x balsam fir [Abies lasiocarpa (Hook.) Hutt. x A. balsamea (L.) Mill.)
To determine the effect of low-level sulfur dioxide emissions on the forested ecosystem,
an integrated ecological approach was used to study and explain the changes in ecosystem
structure and function. The four-year study, begun in 1972, divided the ecosystem into four
major components (air, vegetation, soil, and water) and developed a conceptual model to
illustrate the dynamic relationship between the sulfur dioxide "source" and the generalized
ecosystem "sink" (Legge et al., 1981; Legge, 1980).
Five experimental sites were located, sequentially, downwind in the main path of the sul-
fur dioxide emissions. The two intensive experimental sites were located 1.0 and 1.5 km east
of West Whitecourt Gas Plant. Analysis of the ambient air monitoring data from the two inten-
sive sites during the 1975 and 1976 growing seasons revealed that the air quality standard for
sulfur dioxide of 0.2 ppm per half hour per 24 hours as set by the Alberta Department of the
Environment was exceeded on only three occasions at the two sites in over 2500 hrs of monitor-
ing. Much higher concentrations existed at this sampling site from 1961-74, however, and
effects on soil pH, nutrient balance and tree growth would have also been influenced by these
higher concentrations. In addition, it was observed that the stable sulfur isotopic composi-
32 34
tion ( S/ S) of the sulfur dioxide emissions from West Whitecourt Gas Plant was different
from the natural background sulfur and so provided a tracer that indicated that the gas plant
was the major source of the sulfur gas detected in the forest ecosystem (Legge et al., 1981,
Legge 1980).
The gas exchange properties of field acclimated lodgepole x jack pine trees were measured
to determine the short- and long-term effects of chronic exposure to sulfur dioxide. Photo-
synthetic rates and leaf resistance of the trees had been modified. The extent of ecolo-gical
modification was dependent on the distance from the sulfur source. The rates of photosynthesis
were lower and leaf resistance rates were higher nearer the source of sulfur dioxide. The
reduced rate of photosynthesis was only partially attributable to increased leaf resistance;
therefore, ecological factors such as the mineral nutrient status of foliage and soil pH were
studied since these parameters were known to modify plant response (Legge et al., 1981),

8-48

-------
Analysis of the foliage of the lodgepole x jack pine trees indicated that the sulfate-
sulfur concentration decreased with increasing distance from the sulfur dioxide emission source.
The decrease in foliar sulfate concentration was also more pronounced with increased foliar
age. In addition, detailed analysis of foliar mineral nutrient concentrations revealed that
the mineral nutrient status of the trees had been altered. The concentration of P, K, Fe, Mg,
N, and Zn in the foliage tended to increase with distance from the West Whitecourt Gas Plant
while the concentration of Ca and Al tended to decrease. Much more than any of the other
minerals sampled, the Mn concentration in the foliage tended to decrease as the distance from
the sulfur source increased. Reduction iti foliar potassium and phosphorus may be associated
with reduction in the rate of photosynthesis since foliar potassium has been linked with
stomatal activity and phosphorus with phosphorylation. The alteration of mineral nutrients in
the foliage of the lodgepole x jack pine trees is, therefore, an important ecological factor
contributing to the modification of plant response by sulfur dioxide emissions (Legge et a!.,
1981 and Legge 1980).
Soil pH is known to affect the availability of mineral nutrients to plants; therefore,
soil pH profiles were measured. Soil pH and soil carbonate values increased with depth in 100
cm profiles and with distance away from the West Whitecourt Gas Plant. Total sulfur in the soil
likewise decreased with depth and distance from the emission source, suggesting that sulfur is
tied up in the organic matter and released very slowly. No relationship was found between soil
pH and 6 S values in the soil; however, the soil 6 S value was found to be an excellent
indicator of the presence and penetration into the soil profile of sulfur originating from
sulfur dioxide emissions, while soil pH indicated sulfur loading.
A direct relationship was found between lowered soil pH and the elevated levels of foliar
Mn in lodgepole x jack pine trees. This relationship suggests that the foliar Mn concentration
could be used as an indicator of mineral nutrient modification of the forest ecosystem by sulfur
dioxide emissions.
Biochemical changes in the trees were also noted. The most significant was a transient
metabolic effect involving adenosine triphosphate (ATP). ATP was found to decrease signifi-
cantly in the foliage of the trees when they were exposed to low concentrations of S02 for a
short duration. Foliar ATP concentrations returned to the pre-SO« fumigation levels when the
fumigation ceased. Trees grown in the laboratory in the absence of sulfur dioxide emissions
showed no fluctuation in foliar ATP concentrations and the ATP concentration was three times
the content of field-grown trees. The lower concentration of ATP in the foliage of field-
grown trees when compared to laboratory-grown trees suggests a partial explanation for the
lowered photosynthetic capacities of the lodgepole x jack pine trees growing in the Whitecourt
area. The short-term reversible change in foliar ATP concentration observed in the foliage of
the hybrid pine trees chronically exposed to sulfur gas emissions in the field is suggested as
a biochemical response to SOo toxicity.
8-49

-------
     It was assumed  that the observed ecological modifications  of  the forest ecosystem,  such
as reduced needle  biomass (premature needle drop), reduced biochemical energy, reduced photo-
synthetic rates,  reduced soil pH,  disruption  of mineral nutrient^cycling,  and  foliar sulfur
loading, when combined with a shortened growing season, should be measurable as a reduction in
forest  production.   There was a definite  reduction  in annual  basal  area  increments  in the
lodgepole x jack pine trees since 1962 .that was attributed to sulfur dioxide emissions from the
West Whitecourt Gas  Plant indicating that ecological modifications were initiated long before
the study began  in 1972.  The maximum reduction in basal area increments occurred nearest the
plant and the reduction fell to zero at 9.6 km.  The area affected by sulfur dioxide during the
past 14 years is estimated at 454 km2 (175 mi ) or 45,373 ha (112,130 acres).  The significant
reduction in  sulfur  dioxide emissions since 1970 suggests  areas being impacted by the sulfur
dioxide  will  decrease  in  the  future,  thus permitting  the  recovery of  the portion  of the
ecosystem no longer exposed to sulfur dioxide stress.
     In the near  future, the grasslands of  the  upper plains will be subject to SO,, emissions
from  new coal-burning  power facilities  that  are  being  constructed in  areas  rich  in  coal
reserves  (Durran  et  a!., 1979; Preston  and Lewis,  1978).   To address this  problem,  plots of
Montana  grasslands were  exposed to  SCL  during growing  seasons  of  successive  years.   The
                                                              33
monthly  median  exposure levels were  approximately  0,  52 |jg/m   (0.02), 106  ^ig/m  (0.04), and
         o
185 ug/m  (0.07 ppm)  S02 and were  delivered by  a zonal  air  pollution system or ZAPS  (Lee
et a!., 1978).   Field observations over four years verified that these concentrations were not
sufficient to elicit  any leaf lesions characteristic of acute SCL injury (Heitschmidt et al.,
1978).  Ambient pollutant concentrations were typically greater at night, and the concentration
decreased rapidly  from the interface of turbulent  air and grass canopy downward  to  the  soil
(Preston, 1979).
     The  most  prevalent producer  species  within  the  grassland  is  a perennial,  Agropyron
snnthii.  In  populations sampled over the growing season in*each of the exposure regimes, SCL
induced a variety  of changes in biochemical indices of plant performance.   Honthly samples of
tillers  and  leaves showed a positive correlation  of foliar sulfur with time  of exposure and
canopy-level  SCL   concentrations  (Lauenroth and Heasley 1980).   This relationship  was  most
conspicuous with  the two  higher exposure  regimens,  and total  foliar sulfur in  the highest
exposure plot was  three times greater than  that in vegetation sampled from control locations
(Lauenroth and  Heasley,  1980).   As the sulfur content of leaf tissue increased, the ratio of
nitrogen to sulfur decreased (Lauenroth and Heasley, 1980).
     These biochemical  changes in the major producer species were mirrored by other modifica-
tions in plant performance.  In A. smithii populations exposed to 52 ug/m  (0.02 ppm) SO^ over
the growing  season,  the  functional  leaf  life  (the period of active  photosynthesis)  was in-
creased by several weeks, while the same index of plant performance was shortened by two weeks
at 106 (0.04 ppm) and 185 ug/m  (0.07 ppm) SCL (Lauenroth and Heasley, 1980).  Parallel
                                            8-50

-------
increases and decreases In chlorophyll content at the low and high SO, levels, respectively,
were also recorded; however, the measurements of net primary production from harvested samples
over a period of 5 years revealed no significant treatment response. Finally, the single
significant crop change noted was a decrease in the carbon stored in the rhizomes of western
wheatgrass (Lauenroth and Heasley, 1980).
Dominant producers were not the only flora exhibiting sensitivity to SCL. In simulated
pollutant exposure using a bisulfite solution, Sheridan (1979) showed that nitrogenase activ-
ity in a major component of the lichen flora (Dollema tenex) was reduced. Although the appli-
cability of the data must be validated through field studies, the potential for such an effect
must be recognized, particularly in the light of the importance of soil lichens in regulating
nitrogen fixation in the grasslands (Sheridan, 1979).
Further evidence of SCL-associated effects on grasslands is recorded in both consumer and
decomposer populations. The density of grasshoppers, a major consumer of A._ smithii foliage,
decreased 25 percent in two successive seasons with increasing S0? stress (Lauenroth and
Heasley, 1980). Decomposition rates were apparently also altered, with less litter disappear-
ance in S0,~exposed plots. The mechanism involves a direct pollutant effect on decomposer
activity rather than an indirect effect, such as an increased sulfur concentration in the
litter (Lauenroth and Heasley, 1980).
Larger consumers also exhibited responses reflecting the presence of S0? in the atmos-
phere; however, the responses were not dose-dependent (Chilgren, 1978). Peromyscus
maniculatus, prairie deer mouse, is a common and active vertebrate in grassland communities.
Over one exposure season, the frequency of £_._ mam'culatus in control plots increased, imply-
ing an SOp-induced behavioral response (habitat preference) whereby individuals seek habi-
tats free of the pollutant.
2
In summary, at levels above 52 ug/m (0.02 ppm), SO- induced changes in the performance
of producers, consumer, and decomposers. Many of the responses are individually small, but
collectively, over time, they are gradually modifying the structure and function of the grass-
lands. The significance of these changes to the long-term persistence of the ecosystem
remains controversial (Preston, 1979). This is particularly true since plots were not
replicated in the Montana studies; on the other hand, the fact that the results are based on
data accumulated over several years tends to add to its credibility.
The results of these studies, particularly the West Whitecourt and Montana grasslands
studies, document the usefulness of addressing ecosystem-level responses to S0? from a multi-
disciplinary approach incorporating investigations of physiology, autecology, synecology, geo-
chemistry, meteorology, and modeling. The results confirm that producers are sensitive to
direct SQ? effects as evidenced by SQ?-associated changes in cell biochemistry, physiology,
growth, development, survival, fecundity, and community composition. Such responses are not
unexpected. An equally important point of agreement among the different research efforts is
8-51

-------
the potential for ecological modification resulting from either direct S02 effects on nonpro-
ducer species or direct changes in habitat parameters, which in turn affect an organism's per-
formance. Changes in biogeochemistry, particularly in the soil compartment, are notably re-
sponsive to chronic SO- exposure.
The influence of prolonged SCU exposures on plant communities is not well documented;
however, a theoretical basis is emerging from which to evaluate effects. This conceptual ef-
fort is exemplified by the development of a generalized forest growth model (Botkin et al.,
1972) that was designed to assess the consequences of the long-term interactions of air pollu-
tion stress and forest community dynamics (Botkin, 1976; Shugart and West, 1977). This model
has been applied to determine the response of a mixed-species deciduous forest in the south-
eastern United States to the differential levels of growth reduction (0, 10, and 20 percent)
occurring following simulated air pollution stress (West et al., 1980). Over several decades,
simulated pollution stress altered the biomass importance of the major tree species within the
forest; some species populations increased, while others decreased in importance. These re-
sults suggest that competitive interactions among species may significantly modify both the
level and direction of change in growth rate of individual species in response to air pollu-
tion stress. Stand age was also shown to influence strongly the role of competition in modi-.
fying responses of individual species within the forest community. Since community composi-
tion is determined in part by species interactions (e.g., competition, symbiosis), the eco-
logical importance of resistant species, the prominence of which in the community is deter-
mined by interaction with sensitive species, can be expected to be enhanced under stresses
such as air pollution that do not affect all species equally (West et al., 1980). An under-
standing of the governing role of species interactions is essential for predicting how eco-
systems will respond when exposed to low concentrations of pollution (Botkin, 1976). This is
also the justification for not freely extrapolating the results from intensely managed forest
and agroecosystems to predict how a mixed species community (e.g., natural forests or grass-
lands) will respond to a comparable perturbation (Miller and McBride, 1975; Kickert and
Hi Her, 1979).
The results from community-level studies in areas experiencing chronic levels of SO^ lend
credibility to the modeling effort. Using communities composed of only 2 or 3 different
species, Guderian (1967) analyzed community-level responses to S0? and their underlying
causes. Changes in community composition were a function of pollutant dose; the higher the
dose, the more rapidly the community changed. Altered community composition was attributed
both to direct SO, effects on populations of sensitive species and to indirect changes in
species interactions. Community biomass exhibited little quantitative change, but striking
differences in species composition. Similar conclusions have been reached in studies of
natural plant communities experiencing prolonged S0? exposure (Guderian and Stratmann, 1968;
Rosenberg et al., 1979). Rosenberg et al. (1979) assessed the species composition in 27
stands of a natural regrowth of a Northern hardwood forest dominated by oaks (Quercus spp.),
8-52

-------
white pine (P_. strobus), and hemlock (Tsuga canadensis). The stands, which varied in their
distance from a 25-year-old coal-consuming power plant, exhibited no obvious a priori composi-
tional differences. Atmospheric pollutant levels were not reported, although foliar symptoms
typical of SCL toxicity were recorded on several occasions. In both upwind and downwind
directions, the- number of vascular plants (canopy, understory, and ground) per unit area
(species richness) increased with distance from the source; however the differences were
greater downwind. A similar distance-dependent response was also recorded for species
diversity using the Shannor-Weiner index. For both species richness and species diversity, the
rates of increase were more gradual downwind from the power plant. In spite of these S00-
induced changes in community composition, an index of above-ground biomass (basal area of over-
story species) exhibited no variation among stands. Among the vascular plants, shrub and
ground vegetation was more sensitive (diversity and richness) than the overstory to S0« stress.
This susceptibility of the lower strata was attributed to the intense competition among indi-
viduals at an early phase in their life history when they are more sensitive, and to micro-
habitat factors that tend to increase SCL levels close to the ground.
Some of the most notable examples of SO- affecting plant communities are the responses of
cryptogamic flora (lichens and mosses), and several reviews are available (DeSloover and
LeBlanc, 1968; Hawksworth, 1971). A map of epiphytic lichen communities for England and Wales
has been devised that associates progressive shifts in species composition with S0? levels
(Hawksworth and Rose, 1970). In general, the higher ambient SOp levels were consistently
associated with fewer species and an increasing relative frequency of crustose versus foliose
or fruticose forms. The fidelity with which community composition changes in accordance with
S0? has led to the suggestion that analyses of lichen communities be used as a bioassay to
estimate ambient SO, levels.
Similar mapping efforts are reported for several regions of North America. In a rural
area of Ohio surrounding a coal-consuming power station (emitting 930 metric tons SO^ per
day), the distribution of two corticose lichens, Parmelia caperata and P. ruderta, was
markedly affected by high SCL concentrations (Showman, 1975). In regions experiencing an
3
annual S09 average exceeding 50 ug/m (0.020 ppm), both species were absent. The distribution
3
of more resistant lichens was not noticeably affected until SO- levels exceeded 65 M9/m
(0.025 ppm) (annual average). Somewhat lower levels were projected by LeBlanc and Rao (1973,
1975) to affect the ability of sensitive lichen species to survive and reproduce. Acute and
chronic symptoms of S09 toxicity in epiphytic lichens occurred when long range (May-October)
3 ' 3
average concentrations exceeded 80 ug/m (0.03 ppm) and were between 16-80 ug/m (0.006-0.03
ppm), respectively.
The susceptibility of cryptogamic flora to elevated levels of SO, may influence the move-
ment of materials within the ecosystem. In the northwestern coniferous forests, lichens fix
2-11 kg/ha of nitrogen annually, which represents 5-20 percent of the total nitrogen
requirement for the dominant producer, Douglas fir (Denison, 1973).

8-53

-------
Because the functioning of all ecosystems is due to a network of biotic and abiotic inter-
actions, it may be hypothesized that the effects of SQ~ on producers must have repercussions
at other trophic levels. Demonstration of such responses, however, is difficult experimental-
ly, and an accurate assessment of the specific importance of SOp in eliciting these responses
is complicated by the often complex rel-ationships among producers, consumers, and decomposers.
Consumers and decomposers may respond to SCL via a direct, adverse effect of the pollu-
tant. The presence of elevated atmospheric levels of S0? is particularly relevant to soil
organisms (Babich and Stotzky, 1974). This focus on soil-borne organisms takes on relevance
since the rhizosphere is not only biologically active but also the major site for sulfur accu-
mulation within the ecosystem (Legge et al., 1976). In a forested area experiencing atmos-
pheric S02 levels averaging 126 ng/m (0.048 ppm), the species composition of soil microflora
shifted toward a greater number and frequency of species capable of utilizing the soil sulfur
additions (Wainwright, 1979). Specifically, the levels of thiobacilli and sulfur-oxidizing
fungi were positively correlated with levels of S0? stress and soil depth.
The edaphic and climatic environments strongly influence the community of plants, animals
and microorganisms that develop at a given site. In natural ecosystems in sulfur-deficient
soils, communities have evolved within the constraints imposed by a limited supply of sulfur.
Although atmospherically derived sulfur may not be sufficient to cause injury, the prolonged
input of sulfur may relax the constraints of a limited sulfur supply, thereby inducing shifts
in species composition.
8.10.3 Response of Natural Ecosystems to Particulate Matter
Particulate matter originating from both natural and anthropogenic emission sources is a
common component of the atmosphere. As discussed in Section 8.7, the heterogeneous physical
and chemical nature of particulate matter presents problems in addressing the significance of
elevated atmospheric particulate levels for natural ecosystems and agroecosystems.
Wet and dry deposition are the two processes by which particles are transferred from the
atmosphere to terrestrial ecosystems. The fate of particles deposited on foliar surfaces
depends on the solubility of the constituents (chemicals and elements), the occurrence of pre-
cipitation, and the sorptive capacity of the leaf (Little, 1973). Furthermore, many elements
commonly associated with particles are essential for plant metabolism (e.g., zinc and phos-
phorus), and, as a consequence, absorption may be a means by which the plant can supplement its
nutrient supply (Kloke, 1974). Leaf surfaces may be only a transitory site for particulate
matter and its associated constituents. If not retained by the leaf, material is ultimately
transferred to the forest floor through washoff in rain events. The net effect of these pro-
cesses is to funnel leaf surface deposits to the litter-soil complex. This conclusion is
verified for many atmospherically derived heavy metals deposited in natural ecosystems (e.g.,
Coughtrey et al., 1979; Thompson et al., 1979).
8-54

-------
Given the regional character of participate emissions, particularly along the East Coast
(U.S. EPA, 1978b), the fate of participate matter in terrestrial ecosystems experiencing low
levels of participate pollution needs to be assessed. In a deciduous forest in the Southeast,
wet and dry deposition of aerosols, gases, precipitation and large particles were major sources
of trace element input to the forest floor, including 99 percent for lead, 44 percent for zinc,
42 percent for cadmium, 39 percent for sulfate and 14 percent for manganese (Lindberg et al.,
1979). These seemingly large percentages are typical for rural or remote areas even though
three major coal-consuming power facilities (total coal consumption of 8 x 10 metric tons per
year) were within 20 km of the forest studied.
Irrespective of the source of particles deposited in the forest, the atmosphere contri-
buted a major portion of the trace element inputs (Lindberg et al., 1979). Water solubility
was critical, since insoluble constituents associated with the particulate matter were not
readily mobilized within the forest. Any event promoting solubilization (e.g., aerosol forma-
tion, rainfall scavenging, moisture formation on leaves) enhanced an element's mobility.
The leaf surface is not the only accumulation site for particles and their associated
constituents within the ecosystem. Through precipitation scavenging of particles in air,
washoff of surface deposits, or litterfall, particles are transferred to the soil where they
are tightly bound to decaying organic matter. The upper soil horizon, including the decaying
organic material, is a region of intense biological activity as a result of the physical
degradation of litter, remineralization of the bound materials and root uptake of the plant"
available nutrients. Consequently, particulate emissions that interfere with microbial
activity can have delayed effects on primary production (Tyler, 1972) and soil consumer
species.
In summary, even though the impact of particulate matter on terrestrial ecosystems is
most apparent near large emission sources, ecosystems within the same geographic region may
be the site of deposition. Foliar surfaces are the most common site for initial dry and wet
deposition; however, most material is eventually transferred to the soil. Particulate matter
alone constitutes an ecological problem only where deposition rates are high. However,
concern for terrestrial ecosystems must also address elements and chemicals that may be asso-
ciated with the particulate matter. Solubility of these particulate constituents is a
critical factor since insolubility limits mobility within the ecosystem. One common behavior
of particles is their tendency to accumulate selectively within a given component of the land-
scape. Soils are long-term sites for the retention of many constituents found in particles.
While this accumulation in the soil-litter layer has had demonstrable adverse consequences for
ecological processes such as decomposition, mineralization, nutrient cycling and primary pro'-
duction around some point sources, the much lower levels chronically deposited over large
regions have not yet produced documented adverse impacts on natural ecosystems.
8-55

-------
8.11 SUMMARY
Sulfur dioxide and particulate sulfate are the main forms of sulfur in the atmosphere. Of
the two, sulfur dioxide is potentially more injurious, particularly when it is in combination
with other airborne pollutants. Plants may be exposed to these pollutants in several different
ways. Dry and wet deposition of gases and particles may bring sulfur compounds into contact
with plant surfaces and/or the soil substrate. The effects of such external exposures on
plants are more difficult to assess than those associated with the entry of SO,, through the
stomata. Plant response to pollutant uptake may be influenced by such dynamic physical
factors as light, leaf surface moisture, relative humidity and soil moisture. Such factors
influence internal physiological conditions in plants as well as stomatal opening and closing
and, therefore, play a major role in determining the sensitivities of the plant species or
cultivars.
Sulfur dioxide must enter a plant through leaf openings called stomata to cause injury.
Sulfur dioxide, after entering plant cells through the stomata, is converted to sulfite and
bisulfite, which may then be oxidized to sulfate. Sulfate is about one-thirtieth as toxic as
sulfite and bisulfite. As long as the absorption rate of SOp in plants does not exceed the
rate of conversion to sulfate, the only effects of exposure may be changes in opening or
closing of stomata or insignificant changes in the biochemical or physiological systems. Such
effects may abate if SO, concentrations are reduced. Both negative and positive influences on
crop productivity have been noted following exposures to low concentrations.
Symptoms of S0?-induced injury in higher plants may be quite variable since response is
governed by pollutant dose (concentration x duration of exposure), conditions of the exposure
(e.g., day yj» night, peak vs long-term), physiological status of the plant, phenological stage
of plant growth, environmental influences on the pollutant-plant interaction, and the environ-
mental influences on the metabolic status of the plant itself.
There are several possible plant responses to S0? and related sulfur compounds: (1) fer-
tilizer effects appearing as increased growth and yields; (2) no detectable response; (3)
injury manifested as growth and yield reductions, without visible symptom expressions on
foliage or with very mild foliar symptoms that would be difficult to perceive as being induced
by air pollution without the presence of a control set of plants grown in pollution-free con-
ditions; (4) injury exhibited as chronic or acute symptoms on foliage with or without asso-
ciated reduction in growth and yield; and (5) death of plants and plant communities.
A number of species of plants are sensitive to low concentrations of S0?. Some of these
plants may serve as bioindicators in the vicinity of major sources of S0?. Even these sensi-
tive species may be asymptomatic, however, depending on the environmental conditions before,
during, and after S0? exposure. Various species of lichens appear to be among the most sensi-
tive plants.
8-56

-------
As the exposure to SO, increases, plants may develop more predictable and more obvious
visible symptoms. Foliar symptoms progress from chlorosis, or other types of pigmentation
changes, to necrotic areas. The extent of necrosis increases with exposure. Studies of the
effects of S0~ on growth and yield have demonstrated a reduction in the dry weight of foliage,
shoots, roots, and seeds, as well as a reduction in the number of seeds. At still higher expo-
sures there are further reductions in growth and yield. Extensive mortality has been noted in
forests continuously exposed to SCL for many years.
The amount of sulfur accumulated from the atmosphere by leaf tissues is influenced by the
amount of sulfur in the soil relative to the sulfur requirement of the plant. After low-level
exposure to SO,,, plants grown in sulfur-deficient soils have exhibited increased productivity.
Plant growth and development represents an integration of cellular and biochemical processes.
The response of a given species or variety of plants to a specific air pollutant cannot be
predicted on the basis of the known response of related plants to the same pollutant. Neither
can the response be predicted by a given response of a plant to similar exposures to different
pollutants.
Each plant is a different individual genetically, and therefore its genetic susceptibi-
lity and the influence of the environment at the time of exposure must be considered for each
plant and each pollutant. The data presented in this chapter exemplify the fact that each
plant is a separate entity, and, because of the variation in response shown by the different
plant species and different cultivars of the same species, making generalizations is difficult.
With this in mind, this chapter concludes that in arid regions, some species of vegetation
would probably not show visible signs of SO, injury even at concentrations as high as 11 ppm
for 2 hours. On the other hand, in many nonarid regions where environmental conditions such
as high temperature, high humidity, and abundant sunlight enhance plant responsiveness to S0?
exposure, many species of sensitive and intermediately responsive vegetation would likely,
from time to time, show visible injury when exposed to peak (5 minutes), 1-hour, and 3-hour
S02 concentrations in the range of 2600-5200 ug/m3 (1-2 ppm), 1300-5200 ug/m3 (0.5-2 ppm), and
790-2100 ug/m3 (0.3-0.8 ppm) respectively.
In general, the studies cited in this chapter indicate that regardless of the type of
exposure and the plant species or variety, there is a critical SO- concentration and a criti-
cal time period after which plant injury will occur. This plant response appears to be asso-
ciated with the capability of a plant to transform within the leaf toxic SQy ar>d SOI into the
much less toxic S07 and ultimately to transfer or break down S07.
At present, data concerning the interactions of S0? with other pollutants indicate that,
on a regional scale, SO, occurs at least intermittently at concentrations high enough to pro-
duce significant interactions with other pollutants, principally 0^. A major weakness in the
approach to pollutant interactions, however, is the lack of in-depth analysis of existing
8-57

-------
regional air quality data sets for the three principal pollutants (SOp, 0_, and NO-), These
data should determine how frequently and at what concentrations the pollutants occur together
both spatially and temporally within regions of major concern. The relative significance of
simultaneous versus sequential occurrence of these pollutants to effects on vegetation is also
not well documented and is critical in evaluating the likelihood and extent of potential pol-
lutant interactions under field,conditions.
A few studies have reported that combinations of particulate matter and SO,, or particu-
late matter and other pollutants, increase foliar uptake of S02, increase foliar injury of
vegetation by heavy metals, and reduce growth and yield. Because of the complex nature of
particulate pollutants, conventional methods for assessing pollutant injury to vegetation,
such as dose-response relationships, are poorly developed. Studies have generally reported
vegetational responses relative to a given source and the physical size or chemical composi-
tion of the particles. For the most part, studies have not focused on effects associated with
specific ambient concentrations. Coarse particles such as dust directly deposited on the leaf
surfaces can result in reduced gas exchange, increased leaf surface temperature, reduced
photosynthesis, chlorosis, reduced growth, and leaf necrosis. Heavy metals deposited either
on leaf surfaces or on the soil and subsequently taken up by the plant can result in the accu-
mulation of toxic concentrations of the metals within the plant tissue.
Natural ecosystems are integral to the maintenance of the biosphere, and disturbances of
stable ecosystems may have long-range effects which are difficult to predict. Within the
United States, anthropogenic contributions to atmospheric sulfur exceed natural sources. In
the Northeast these contributions exceed natural sources by a factor of 12.5, and approxi-
mately 60 percent of the anthropogenic emissions into the atmosphere are deposited (wet and dry
deposition) on terrestrial and aquatic ecosystems. The subsequent fate and distribution of
sulfur in these systems is not well understood. Wet deposition of sulfur compounds is
discussed in Chapter 7.
Data relating ecosystem responses to specific doses of SO- and other pollutants are dif-
ficult to obtain and interpret because of the generally longer periods of time over which
these responses occur and because of the many biotic and abiotic factors that modify them.
Vegetation within terrestrial ecosystems is sensitive to SO,, toxicity, as evidenced by
changes in physiology, growth, development, survival, reproductive potential and community
composition. Indirect effects may occur as a result of habitat modification through in-
fluences on litter decomposition and nutrient cycling or through altered community structure.
At the community level chronic exposure to S0?, particularly in combination with other pol-
lutants such as 0~, may cause shifts in community composition as evidenced by elimination of
individuals or populations sensitive to the pollutant. Differential effects on individual
species within a community can occur through direct effects on sensitive species and through
alteration of the relative competitive potential of species within the plant community.
8-58

-------
     Participate emissions  have their  greatest  impact on  terrestrial  ecosystems  near  large
emission sources.  Participate  matter  in itself constitutes a problem only in those few areas
where deposition rates are very high.   Ecological modification may occur if the particles con-
tain  toxic  elements,  even  though deposition  rates  are  moderate.   Solubility of  particle
constituents is  critical,  since water-insoluble  elements are not mobile within the ecosystem.
Most of the material  deposited by wet and dry deposition on foliar surfaces in vegetated areas
is transferred to the soil where accumulation occurs in the litter layer.
                                            8-59

-------
8.12  REFERENCES

Adepipe,  N.  0.,  R.  E.  Barrett,  and  D.  P.  Ormrod.    Phototoxlcity and  growth response  of
     ornamental  bedding  plants  to  ozone  and  sulfur  dioxide.   J.  Am.  Soc.  Hortic.  Sci.
     97:341-345, 1972.

American  Phytopathologies!  Society.   Glossary of  air  pollution terms and selected  reference
     list.  Phytopathol. News 8:5-8,  1974.

Anderson,  A.  J.,  D.  R.  Meyer,  and  F.  K.  Mayer.    Heavy  metal  toxicities: levels of  nickel,
     cobalt, and chromium in the soil  and  plants associated with visual  symptoms  and  variation
     in growth of an oat corp.  Aust.  J. Agric. Res. 24:557-571,  1973.

Anderson,  P. J.   The effect of  dust from cement mills  on the setting  of  fruits.   Plant World
     17:57-68, 1914.

Ashenden,  T.  W.  Growth  reductions  in  cocksfoot  (Dactylis glomerata  L.)  as  a result  of S0?
     pollution.  Environ. Pollut. 15:161-166,  1978.

Ashenden,  T, W.   The effects of long-term exposures to  S0» and  N02 pollution  on  the  growth of
     Dactyli s gl ome rata L. and Poa pratensis  L. Environ. PQllut.  18:249-258, 1979.

Ashenden,  T. W. ,  and T.  A. Mansfield.   Influence of wind speed  on the  sensitivity  of ryegrass
     to SOg.  J. Exp. Bot. 28:729-735, 1977.

Babich,  H.,  and G.  Stotzky.   Air  pollution  and microbial ecology.   CRC  Crit. Rev. Environ.
     Control 4:353-420, 1974.

Ballentyne, D.  J.   Sulfite inhibition  of ATP  formation  in plant  mitrochondria.  Phytochemistry
     12:1207-1209, 1973.

iarton,  V.  R., S.  B.  Mclaughlin, and R.  K.  McConathy.  The effects of S02 on components  of
     least resistance to gas exchange.   Environ. Pollut. 21:255-265,  1980.

Bazzaz,  H.  B., and  Govindjee.   Effects of  lead chloride  on  chloroplast reactions.  Environ.
     Lett. 6:175-191, 1974.

Bell, J.  N.  B.,  and W. S. Clough.   Depression of yield  in ryegrass exposed to  sulfur dioxide.
     Nature (London) 241:47-49, 1973.

Bennett,  J.  H., and A, 'C. Hill.  Inhibition  of apparent phytosynthesis  by  air  pollutants.   J.
     Environ. Qua!.  2:526-530, 1973a.

Bennett,  J.  H., and A. C. Hill.  Absorption  of gaseous  air pollutants  by a standardized plant
     canopy.  J. Air Pollut. Control Assoc. 23:203-206,  1973b.

Bennett,  J.  H., A.  C.  Hill, A. Soleitnani, and W. H. Edwards.  Acute  effects of combination of
     sulfur dioxide and nitrogen dioxide on plants.  Environ. Pollut. 9:127-132,  1975.

BerigaM,  M. S.,  C.  F. Jordan, and  C. A.  Feickert.  The Effect  of Sulfur Dioxide on  Yield and
     Growth of  Kidney  Beans (Phaseglus  vulgaris L.).  Argonne Nat. Lab. Radio!.  Environ.  Res.
     Div. Annu. Rep. (Ecol.), January-December 1974.  pp. 51-61.

Berry, C. R.   Relative sensitivity  of red, jack, and white pine  seedlings  to ozone and  sulfur
     dioxide.  Phytopathology 61:231-232,  1972.
                                           8-60

-------
Berry, C.  R.   Age  of pine  seedlings  with  primary  needles affects  sensitivity to ozone  and
     sulfur dioxide.  Phytopathology 64:207-209, 1974.

Billings,  W.  D.    Plants  and  the  Ecosystem.   3rd  ed.   Wadsworth   Publishing   Company,
     Inc.  Belmont, PA., 1978.  pp. 1-62.

Black, V.  J.,  and  M.  H.  Unsworth.   Effects of low concentrations  of sulphur dioxide on  net
     photosynthesis and dark respiration of  Vicia faba.  J. Exp.  Bot.  30:473-483,  1979.

Boertitz, S. ,  H.  G.  Daessler, and E. Friedrich.  Effect of metallurgical  dust  on  agricultural
     plants.  Tech. Umweltschutz 15:247-254, 1976.

Bonte, J. ,  L.  DeCormis,  and P. Louguet.   Effect of sulfur dioxide pollution on the degree of
     opening   of   Pelargoni urn  x   hortorum  stomata.    C.R.   Seances   Acad.   Sci.   Ser.   D
     280:2377-2380, 1975.

Botkin,  D.  B.   The  role  of  species  interactions  in  the response  of a  forest  ecosystem  to
     environmental  perturbation.   In:    Systems  Analysis and  Simulation  in Ecology,   vol.  IV.
     B. C. Patten, ed., Academic Press, New  York, NY, 1976.  pp.  147-173.

Botkin,  D.  B. , J.  F.  Janak,  and  J.  R.  Wallis.  Some  ecological  consequences of a  computer
     model of forest growth.  J. Ecol. 60:849-872, 1972.

Bradford, 6. R.   Boron.   In:  Diagnostic  Criteria  for Plants and Soils.   H. G. Chapman,  ed.,
     University of California, Berkeley, CA, 1966.  pp.  33-61.

Brady, N. C.   The Nature  and Properties of  Soils, 8th edition.   MacMillan Publishing  Co.,  New
     York, NY,  1974.

Brandt,  C.  S. ,  and W. W.  Heck.  Effects of  air  pollutants on  vegetation.   In:  Air Pollution,
     vol. I:   Air Pollution and Its Effects.  A. C. Stern, ed. ,  Academic  Press, New York,  NY,
     1968.  pp. 401-443.

Brandt,  C.  J.,  and R. W.  Rhoades.   Effects  of limestone dust  accumulation on composition  of a
     forest community.  Environ. Pollut. 3:217-225, 1972.

Brandt, C. J. ,  and R'. W. Rhoades.   Effects of limestone  dust accumulation  on lateral growth of
     forest trees.  Environ. Pollut. 4:207-213,  1973.

Buron, A., and G.  Comic.  Effect of sulfur  dioxide on gas exchange  in white mustard (Sinapsis
     alba).   Bull. Soc. Vaudoise Sci. Nat. 71:451-461, 1973.

Carlson,  R.  W.   Reduction in the  photosynthetic rate  of  Acer  Quercus  and Fraxinus species
     caused by sulphur dioxide and ozone.  Environ. Pollut. 18:159-170, 1979.

Chamberlain, A. C.   The movement of particles  in plant communities.  In:  Vegetation and the
     Atmosphere,  vol.  1:   Principles.   J. L. Monteith,  ed.,   Academic Press,  London,  England,
     1975.  pp. 155-203.

Chapman,  H.  D.   Zinc.   In:   Diagnostic  Criteria for  Plants  and Soils.   H.  D.  Chapman,  ed.
     University of California, Berkeley, CA, 1966.  pp.  484-499.

Chiba,  0.,  and V.  Tanaka.  The effect  of sulphur dioxide on the  development of pine  needle
     blight  caused by  Rhizosphaera kalkhoffii  Bubak  (I).   J.  Jap.   For.  Soc.  50:135,  1968.

Chilgren, J. D.   The response of prairie  deer mice to a field SO,, gradient.   In:   Proceedings
     of  the 4th  Joint  Conference Sensing  of  Environmental  Pollutants.   American   Chemical
     Society, Washington,  DC, 1978. pp.  61-65.


                                           8-61

-------
Clapham, W. B. J.  Natural Ecosystems.  The Macmillan Company, New York, 1973.  p. 43.

dough,  W.  S.   The  deposition  of  particles  on  moss  and grass surfaces.   Atmos.  Environ.
     9:1113-1119, 1975.

Cogbill, C. V.  The history and character of  acid precipitation  in eastern  North America.   In:
     Proceedings  of the  First  International  Symposium  on Acid  Precipitation and the  Forest
     Ecosystem, U.S.  Department  of Agriculture and Ohio State University,  Columbus,  Ohio,  May
     12-15, 1975.   L. S.  Dochinger and R. A.  Seliga,  eds., USDA For.  Serv.  Gen. Tech.  Rep.
     NE-23, U.S. Department of Agriculture, Upper Darby, PA, 1976.  pp. 363-370.

Coleman, R.  The importance of sulfur as a plant nutrient  in world crop production.   Soil  Sci.
     101:230-239, 1966.

Constantinidou,  H.  A., and  T.  T.  Kozlowski.  Effects  of sulfur dioxide  and ozone  on Ulmus
     americana  seedlings.  I.  Visible  injury  and  growth. Can.  J.  Bot.  57:170-175,  1979a.

Constantinidou,  H.  A., and  T.  T.  Kozlowski.  Effects  of sulfur dioxide  and ozone  on Ulmus
     americana seedlings.  II.  Carbohydrates,  proteins, and lipids.   Can.  J.  Bot. 57:176-184,
     1979b.

Costonis,  A.  C.   Injury  to  eastern  white  pine  by  sulfur dioxide  and  ozone  alone and  in
     mixtures.  Eur. J. For. Pathol. 3_: 50-55, 1973.

Cotrufo,  C.,  and C.  R.   Berry.   Some effects  of  a soluble NPK fertilizer on sensitivity  of
     eastern white pine to injury from SO, air  pollution.  For. Sci. 16:72-73,  1970.

Coughenour, M.  B.  Grasslands  Sulfur Cycle  and  Ecosystem  Response  to Low-Level SO,,.  Ph.D.
     Thesis, Colorado State University, Fort  Collins, CO,  1978.

Coughtrey,  P.  J. ,  C.  H. Jones,  M. H.  Martin,  and S.  W.  Shales.   Litter  accumulation  in
     woodlands contaminated by Pb, Zn, Cd and Cu.  Oecologia 39:51-60, 1979.

Cowling,  D.  W.,  and  L.   H.  P.  Jones.  Sulphur deficiency on  two forage  plants  in  England.
     Sulphur Inst. J. 6:11, 1971.

Cowling,  D. W.,  and D. R. Lockyer.  The effect of SO- on  Lolium perenne  L, grown  at different
     levels of sulfur and nitrogen nutrition.   J. Expf Bot. 29:257-265, 1978.

Cowling,  D.  W.,  L.  H.  P.  Jones,  and D.  R.  Lockyer.   Increased yield through correction  of
     sulfur deficiency in  ryegrass  exposed  to sulfur  dioxide.   Nature (London)  243:479-480,
     1973.

Cowling,  D. W.,  and M. J. Koziol.   Growth of ryegrass (Lolium perenne L.)  exposed to SQ-.   I.
     Effects on photosynthesis and respiration.  J. Exp. Bot. 29:1029-1036, 1978.

Czaja, A.  T.  Zementstaubwirkungen auf pflanzen:  Die enstehung  der Zementkrusten  [The effects
     of  cement  dust  on  plants:   the  formation of cement crusts.]   Qua!.  Plant Mater.  Veg.
     8:201-238, 1961.

Daessler,  H.  G., J,  Ranft, and  K.  H.  Rehn.  The susceptibility of woody plants  exposed  to
     fluorine compounds and sulfur dioxide.   Flora (Jena)  161:289-302, 1972.

Daines, R. H.   Sulfur dioxide and plant response.  J. Occup. Med.  10:  516-534,  1968.

Darley,  E. F.   Studies  on the  effect of  cement-kiln dust  on vegetation.   J.  Air  Pollut.
     Control Assoc. 16:145-150, 1966.
                                           8-62

-------
Davis, D. D.  Sulfur Dioxide and Deciduous Plants.  Air Pollution and Plants Series. PA. Coop.
     Ext. Ser. U. Ed. 3-17, 1972a.

Davis, D. D.  Sulfur Dioxide and Evergreens.  Air Pollution and Plants Series.  PA. Coop. Ext.
     Serv. U. Ed. 3-526, 1972b.

Davis, D. D. ,  and R. G. Wilhour.  Susceptibility of woody plants to sulfur dioxide and photo-
     chemical oxidants:  a literature review.  EPA 600/3-76-102, U.S. Environmental Protection
     Agency, Corvallis, OR, September 1976.

Davis, J. B. ,  and R. L. Barnes.   Effects  of soil-applied fluoride and lead on growth of lob-
     lolly pine and red maple.  Environ. Pollut. 5:35-44, 1973.

Demeritt,  M.  E., W.  M.  Chang,  J.  D.  Murphy,  and  H.  D.   Gerhold.   Selection  system  for
     evaluating   resistance  of  Scotch   pine seedlings  to  ozone  and  sulfur dioxide.   In:
     Proceedings  of  the  19th  Northeastern For.  Tree Improvement Conference, 1971.  pp. 87-97.

Denison, W.  C.  Life in tall trees.  Sci. Am. 228:75-80, 1973.

DeSloover,  J. ,  and  F.  LeBlanc.   Mapping  of atmospheric  pollution  on  the  basis  of lichen
     sensitivity.   In:   Proceedings  of  the Symposium  Recent Advances Tropical  Ecology  R.
     Misra  and  B. Gopal  eds., The  International  Society for Tropical  Ecology,  Varanashi-5,
     India, 1968.  pp.42-56.

Dochinger,  L.  S., and  W.  W.   Heck.   An ozone-sulfur  dioxide synergism produces symptoms of
     chlorotic dwarf of eastern white pine.  Phytopathology ^9:399, 1969.

Dochinger,  L. S., and K. F. Jensen.  Effects  of chronic and  acute exposure to sulfur dioxide
     on the growth of hybrid poplar cuttings.  Environ. Pollut. 9:219-229, 1975.

Dodd, J.  L. ,  W.  K.  Lauenroth, G.  L.  Thor and Coughenour.  Effects  of Chronic Low Levels SOp
     Exposure  on  Producers  and  Litter Dynamics.   In:   The Bioenvironmental  Impact of  a
     Coal-Fired Power Plant.  Third Interim, Colstrip, Montana, December, 1977.   E. M. Preston
     and  T.  L.  Gullet,  eds.   EPA-600/3-79-044.  U.S.  Environmental  Protection Agency, 1978.
     pp. 384-493.

Domes, W.  Unterschiedlich COg-Abhangigkeit  des Gasaustaushes  beider Blattseiten  von Zea Mays.
     [Different C09~sensitivities  of  the gas exchanges of the two leaf surfaces  of Zea Mays.]
     Planta 98:186-189, 1971.

Donaubauer,  E.    Sekundlrschiden  in Osterreichischen  Rauchschadensgebieten.   Schwierigkeiten
     der  Diagnose  and  Bewertung.   [Secondary  damages  in  Austrian  smoke  damaged  areas.
     Difficulties of diagnosis and assessment.]  In:  Materialy VI Miedzynardowej Konferencji,
     Polska Academia Nauk, Katowice, Poland,  pp. 277-284.  1968.

Dreisinger,  B.  R.   Sulfur dioxide  levels  and the  effects  of  the  gas  on  vegetation near
     Sudbury,  Ontario.    Presented   at  the  58th  Annual  Meeting,  Air  Pollution  Control
     Association, Toronto, Ontario, 1965.  Paper no. 65-121.

Dreisinger, B. R.  The impact of sulphur dioxide pollution of  crops and forests,  Pollution and
     Our  Environment, Conference  Background Paper,  vol.  1, Montreal,  Canadian  Council  of
     Resource Ministers, Paper No. A4-2-1, 1967.  p. 1-7

Dreisinger,  B.  R. ,  and P.  C.  McGovern.   Monitoring atmospheric sulfur dioxide and correlating
     its  effects  on  crops  and forests  in  the  Sudbury area.    In:   Impact  of  Air  Pollution on
     Vegetation,  Air Pollution Control  Association Specialty  Conference,   S.  N.  Linzon, ed. ,
     Ontario Department of Energy and Resource Management, Toronto, 1970.  122 pp.
                                           8-63

-------
Durran, D.  R.,  M.  J. Meldgin, M.  Liu,  T.  Thoem, and D. Henderson.  A  study  of  long-range air
     pollution  problems  related  to  coal  development  in  the northern  Great Plains.  Atmos.
     Environ. 13: 1021-1037, 1979.

Embleton, T.  W.  Magnesium.   In:   Diagnostic Criteria for  Plants  and Soils.  H.D,  Champman,
     ed., University of  California  Division of  Agricultural Sciences,  Riverside,  CA, 1966.
     pp. 225-263.

Enderlein,  H.,  and  W.  Kastner,   What  effect has a nutrient deficiency on the resistance of
     one-year-old conifers to S02?  Arch. Forstwes. 16:431-435, 1976.

Evans,  L.  S. ,  and  P.  R. Miller.   Histologies!  comparison of single  and additive 0, and S0?
     injuries to elongating ponderosa pine needles.  Am. J. Bot. 62:416-421,  1975.

Faller, N.  Effects  of atmospheric S02 on plants.  Sulfur  Institute J.  6:5-7, 1970.

Faller, N.  Plant nutrient sulfur-SQ2 versus  SO..  Sulfur  Institute J.  7:5-6, 1971.

Fankhauser,  N.,  C.  Brunold,  and K. H. Erismann.  The influence of sublethal  concentrations of
     sulfur  dioxide  on  morphology,  growth and production  yield of the  duckweed  Lemna minor L.
     Oecologia  23:201-206, 1976.

Ferenbaugh,  R.  W.   Effects  of prolonged exposure of Oryzopsis hymenoides  to  SO,.  Water, Air,
     Soil Pollut. 10:27-31, 1978.                                              ^

Fujiwara, T., T. Umezawa, and H. Ishikawa.  Effects of  mixed  air pollutants on vegetation.  I.
     Effects of combinations of S0p, NO, and  03-  Cent. Inst.  of Elec.  Power. Res. Rep. 72007,
     Institute  of Agricultural Electricity, 1973.

Garland, J.  A.   Dry and wet removal of sulfur from the atmosphere. Atmos.  Eviron. 12:349-362,
     1978.                                                                         ~

Gilbert, 0.  L.   An  alkaline dust effect  on  epiphytic  lichens.  Lichenologist 8:173-178 1976.

Gerhold,  H.  D.   Effect  of  air  pollution  on  Pinus  strobus  L.   and  genetic  resistance.
     EPA-600/3-77-002,  U.S.   Environmental  Protection  Agency, Corvallis,  OR,  January 1977.

Godzik, S. ,  and H.  F.  Linskens.   Concentration  changes   of  free amino acids in primary bean
     leaves  after  continuous and  interrupted S0? fumigation and  recovery.  Environ. Pollut.
     7:25-38, 1974.                               ^

Granat, L.,  H.  Rehe,  and R. 0.  Hallberg.   The global sulfur cycle.   Ecol.  Bull. 22:89-134,
     1976.

Guderian,    R.     Reaktionen    von    Pflanzen-gemeinschaften   des    Feldfutterbaues   auf
     Schwefeldioxideinwirkungen.   [Responses  of plant  communities  to  air pollution.]  Essen:
     Girardet-Verlag,   Schriftenr.    Landesanst.    Immissions-Bodennutzungssch.   d.   Landes
     Nordrhein-Westfalen 4:80-100, 1966.

Guderian,  R.    Air  Pollution:   Phytotoxicity of  Acidic   Gases  and  its  Significance  in Air
     Pollution  Control.  Springer-Verlag, Berlin, W.  Germany,  1977.

Guderian, R.  Terrestrial  Vegetation—Air Pollutant Interactions:  Non-Gaseous Air Pollutants.
     Presented  at  the  International  Conference on  Air Pollutants  and Their Effects  on the
     Terrestrial Ecosystem.  Banff, Alberta,  Canada, May 10-17, 1980.
                                           8-64

-------
Guderian  R. ,  and H.  Stratmann.   Freilandversuche  zur Ermittlung von Schwefeldioxidwirkungen
     auf   die   Vegetation.     III.    Grenzwerte  schadlicher   SOp-Immissionen  fur   Obst-und
     Forstkulturen    sowie    fur    landwirtschaftliche    and   gartnerische    Pflanzenarten.
     Forschungber.   [Field  experiments  to  determine  the  effects  of   sulfur   dioxide   on
     vegetation.  Ill Threshold values of harmful SCL  pollution on fruit and  forest culture  as
     well  as  agricultural and  horticultural  species  of plants.]  Landes  Nordrhein-Westfalen
     Nr. 1920.   Westdeutscher Verlag, Cologne, 1968.   113 pp.

Haisman, D. R.   The effect of sulfur dioxide on oxidizing enzyme systems  in  plant  tissue.   J.
     Sci. Food Agric. 25:803-810, 1974.

Halbwachs, G.   Der Wasserhaushalt rauchgesch'digter Holzgewachse.  [Water conservation  in  smoke
     damaged woody plants.]  Allg. Forstz. 78:196-197, 1967.

Hallgren,  J-E.    Physiological  and  biochemical  effects of  sulfur  dioxide  on  plants.    In:
     Sulfur in  the  Environment,  vol. 2.  J.  0.  Nriagu,  ed. , John Wiley & Sons, New York,  NY,
     1978.  pp.  163-210.

Hawksworth,  D.   L.   Lichens  as   litmus  for  air pollution:   a  historical   review.   Int.   J.
     Environ.  Stud.  1:281-296, 1971.

Hawksworth, D.  L., and F. Rose.  Qualitative scale for estimating sulfur dioxide air pollution
     in England and Wales using epiphytic lichens.  Nature (London) 227:145-148, 1970.

Heagle,  A.  S.    Interactions  between  air  pollutants  and  plant  parasites.    Annu.  Rev.
     Phytopathol. 11:365-388, 1973.

Heagle, A.  S. ,  and  J.  W. Johnston.   Variable  responses  of soybeans  to  mixtures of ozone  and
     sulfur dioxide.  J. Air Pollut. Control Assoc. 29:729-732, 1979.

Heagle, A.  S. ,  D. E. Body, and G. E. Neely.  Injury and yield responses of soybean to chronic
     doses of ozone and sulfur dioxide in the field.   Phytopathology ^4:132-136, 1974.

Heck,  W.  W. , and C. S.  Brandt.    Effects  on vegetation:   native,  crops, forest.  _In:    Air
     Pollution.    Vol. II:   The Effects of Air Pollution.  A. C. Stern, ed.,  3rd ed.,  Academic
     Press, New York, NY, 1977.  pp. 157-229.

Heck, W. W. ,  and J. A.  Dunning.   Response  of oats to sulfur dioxide:  interactions of growth
     temperature  with  exposure   temperature  or humidity.   J.  Air  Pollut.  Control Assoc.
     28:241-246,  1978.

Heggestad,  H.   E. ,   K.   L.  Tuthill,  and  R.  N.  Stewart.   Differences  among  poinsettias   in
     tolerance to sulfur dioxide.  Horti. Sci. 8:337-338, 1973.

Heitschmidt, R.   K. ,  W.  K. Lauenroth, and  J.  L.  Dodd.    Effects of controlled levels of sulfur
     dioxide  on  western wheatgrass  in  a  southeastern  Montana grassland.   J.  Appl.  Ecol.
     14:859-868,  1978.

Hill, A.  C. ,  S.  Hill, C. Lamb, and T. W. Barrett.  Sensitivity of native  desert vegetation  to
     S02 and to S02 and N02 combined.  J. Air Pollut.  Control Assoc.  24:153-157, 1974.

Hofstra,  G. ,  and D. P.  Ormrod.   Ozone and  sulfur  dioxide  interaction in  white  bean  and
     soybean.   Can.  J. Plant Sci. 57:1193-1198,  1977.

Horsman,  D.  C. ,  and  A.  R.  Wellburn.   Synergistic effect  of S02  and N02 polluted  air upon
     enzyme activity in pea seedlings.  Environ. Pollut. 8:123-133, 1975.
                                           8-65

-------
Horsman, D.  C.,  and A. R. Wellburn.  Appendix  II.  Guide  to  metabolic  and  biochemical  effec
     of  air pollutants on  higher plants.   In:  Effects  of  Air  Pollutants on Plants.  T.
     Mansfield, ed. , Cambridge University Press, Cambridge, England  1976.   pp.  185-199.

Horsman, D.  C.,  T.  M.  Roberts,  and  A.  D. Bradshaw.  Evolution of sulfur dioxide  tolerance
     perennial ryegrass.  Nature .(London) 276:493-494, ,197,8.

Hou, L-Y.,  A.  C.  Hill, and A.  Soleimam*.   Influence of C09  on  the  effects of S09  and N09
     alfalfa.  Environ. Pollut.  12:7-16, 1977.             z                       z        ^

Houston, D. B.  Response of selected Pinus strobus  L. clones  to fumigation  with sulfur dioxi
     and ozone.  Can. J. For. Res. 4:65-68, 1974.

Houston, D.  C.,  and G. R, Stairs.   Genetic  control of sulfur dioxide  and ozone  tolerance
     eastern white pine.  For. Sci. 19:267-271, 1973.

Husar, R. B., D. E. Patterson, J. D. Husar, N.  V. Cillam and  W. E. Wilson.   Sulfur budget  of
     power plant plume.  Atmos.  Environ. 12:549-568, 1978.

Jacobson, J.  S.,  and L.  J. Colavito.  The combined effect of sulfur dioxide and ozone on  bee
     and tobacco plants.  Environ. Exp. Bot. 16:277-285, 1976.

Jacobson,  J. S.,  and A.  C.  Hill.   Recognition  of Air  Pollution  Injury  to Vegetation:
     Pictorial Atlas.  Air Pollution Control Association,  Pittsburgh, PA, 1970.

Jager, H. J.,  and H. Klein.  Biochemical and physiological detection of sulfur dioxide injur
     to pea plants (Pisum sativum).  J. Air Pollut. Control Assoc. 27:464-466,  1977.

Jancarik, V.  Vyskyt drevokaznych hub v hourem  poskozovani oblasti Krusnych hor.   [Occurrence
     of  wood-destroying  fungi   in  the  smoke-damaged  area  of  the  Erzgebirge.]   Lesnictv
     34(7):677-692, 1961.

Jensen,  K.   F.,  and T.  T.  Kozlowski.    Absorption and  translocation  of  sulfur dioxide b
     seedlings of four forest tree species.  J. Environ. Qua!. 4:379-382, 1975.

Jennings, 0. E.  Smoke injury to shade trees.   Natl. Shade Tree Conf.   10:44-48, 1934.

Jones, H. C.   Sulfur oxides and particulate  matter standards relative to  ecological  effects
     differing  viewpoints.   In:   Proceedings  of the  Proposed  SO   and Particulate  Standarc
     Atlanta, GA, Sept. 16-18, 1980, Air Pollution  Control Association, ed., 1981.   pp.  84-94

Jones, H.  C., D.  Weber,  and  D.  Balsillie.   Acceptable limits for  air pollution dosages an
     vegetation effects:  sulfur dioxide.  Presented at the 67th  Annual Meeting, Air Pollutic
     Control Association, Denver, CO, 1974.  Paper  no. 74-225.

Jordan,  M.   Effects of zinc smelter  emissions  and fire on a chestnut oak woodland.   Ecolog
     56:78-91, 1975.

Kamprath, E.  J.   Possible effects from sulfur  in  the atmosphere.   Combustion 44:16-17, 1972

Karnosky, D.  F.   Threshold levels for foliar  injury to Populus  tremuloides by sulfur dioxidt
     and ozone.  Can. J. For. Res. 6:166-169, 1976.

Karnosky,  D.  F.   Evidence for  genetic  control of response to  sulfur dioxide and ozone i
     Populus tremuloides.   Can. J. For. Res. 7:437-440, 1977.
                                           8-66

-------
Katz, M.  Sulfur  dioxide in the atmosphere  and its relation to  plant  life.   Ind. Eng. Cheml
     41:2450-2465, 1949.

Katz, M., and  A.  W,  McCallum.  The effect of sulfur dioxide on conifers.   In;   Proceedings  of
     the U.S.  Tech. Conf. Air Pollution, 1952.  pp. 84-96.

Katz, M.   Effect  of  contaminants  other  than  sulfur dioxide on  vegetation and animals.  In:
     Pollution and Our Environment Conf., Background Paper A4-2-2, Vol.  1,  Council  Res.
     Ministers, Montreal, Can.  1967.  pp. 1-18.

Keller,  T.  The  effect of long duration,  low  S0? concentrations upon photosynthesis  of coni-
     fers.  In:   Proceedings  of the 4th International Clean Air  Congress,  Tokyo,  Japan, 1977.
     pp. 81-83.

Keller,  T.   Wintertime  atmospheric  pollutants—do they  effect the  performance of deciduous
     trees in the ensuing growing season.  Environ. Pollut. 16:243-247,  1978.

Keller,  T.   The  effect of a continuous springtime  fumigation with  SO,,  and  CO,, uptake and
     structure of the annual ring in spruce.  Can. J.  For. Res.  10:1-6, T.980.

Kender,  W. J. , and F. H. F. G. Spierings.  Effects of sulfur dioxide, ozone, and  their  inter-
     actions  on   "golden delicious" apple  trees.  Neth.  J.  Plant  Pathol.  81:149-151, 1975.

Kickert, R. N.,  and  P.  R. Miller.   Responses of  ecological systems.  In:   Methodology for the
     Assessment of Air Pollution  Effects on Vegetations.   W.  W.  Heck,  S.  V. Krupa, and S.  N.
     Linzon, eds., Air Pollution Control Association, Pittsburgh, Pa., 1979.

Kloke, A.   Lead-zinc-cadmium enrichment in soils  and plants.  Staub-Rein.   Luft  34:20-24, 1974.

Kodata,  M. ,  and   T.  Inoue.   Invading path  of  sulfur dioxide into pine  leaves  as  revealed  by
     microradioautography.  J. For. Soc. _54:207-208, 1972.

Krause,  G. H.  M.,  and H. Kaiser.  Plant response  to heavy metals  and  sulfur dioxide.   Environ.
     Pollut. 12:63-71, 1977.

Krouse,  H.  R.  Sulphur  isotope abundance elucidate uptake of atmospheric sulphur  emissions  by
     vegetation.   Nature (London) 265:45-46, 1977.

Krupa,  S. V.,  B.  I.  Chevone, S. Fagerlie, F. Russo and D. F. Lang.   1976.  Impact of  air pol-
     lutants  on  terrestrial  vegetation  - a  literature  survey.  State  of Minnesota  Environ-
     mental Quality Council.

Krupa,  S. V.,  and R.  J. Kohut.  Impact  of stark emissions from  the  NSP-SHERCO  power  plant  on
     terrestrial   vegetation.  Annual Report, Northern States Power Co.,  Minneapolis, MN, 1976.

Kudela,   M.,  and   E.   Novakova.   Lesni  skudci  a skody  zerf v lesfch  poskozovanych  Kourum.
     [Insect  and  wild life damages in smoke-damaged forest stands.]  Lesnictvi  3_5(6): 493-502,
     1962.

Labanauskas, C. K.  Manganese.  In:  Diagnostic Criteria for, Plants and  Soils.   H.  D.  Chapman,
     ed., University of  California, Berkeley, CA, 1966.  pp. 264-285.

Lacasse,  N.  L.,   and  W.  J.  Moroz.   Handbook of  effects  assessment;  vegetation  damage.  CAES.
     Pennsylvania State  University, University  Park, PA, 1969.

Lagerwerff, J. V.  Lead, mand cadmium as  environmental  contaminants.   In:  Micronutrients  in
     Agriculture.   J.  J.  Mortuedt,  P.  M.  Giodana,  and  W.  L.  Lindsay,  eds., Soil  Science
     Society of America, Madison, WI, 1972.  pp.  593-636.


                                           8-67

-------
Lauenroth, W.  K.,  and  J.  E.  Heasley.   Impact of  atmospheric sulfur deposition on  grassland
     ecosystems.   In:   Atmospheric  Sulfur  Deposition:   Environmental  Impacts  and  Health
     Effects.   D.  S, Shriner,  C.  R.  Richmond,  and  S.  E.  Lindberg,  eds., Ann Arbor  Science
     Publishers, Ann Arbor, Mi., 1980. pp. 417-430.

Laurence, J,  A.  Effects of air pollutants on plant pathogen  interaction.   In:   Proceedings  of
     the 71st Annual Meeting, Air Pollution Control Association, Houston, TX,  1978.   pp.  3-24.

Laurence, J.  A.   Response of maize  and  wheat to sulfur dioxide.   Plant Dis.  Rep.  63:468-471,
     1979.                                                                          ~~

Laurence, J.  A., and L. G. Weinstein.   Effects of sulfur  dioxide on  southern  corn  leaf blight
     maize and stem rust of wheat.  Plant Dis. Reptr. 63:975-978, 1979.

Laurence, J.  A., A.  L.  Alusio, L. H. Weinstein, and  D. C. McCune.  Effects  of sulphur  dioxide
     on  southern  bean  mozaic  and maize dwarf moziac.   Environ.   Pollut.,  1981.   (in  press)

LeBlanc,  F., and  D. N.  Rao.   Effects  of  sulfur  dioxide   on  lichen  and moss transplants.
     Ecology 54:612-617, 1973.

LeBlanc,  F., and  D. N.  Rao.   Effects  of  air  pollutants   on  lichens  and bryophytes.   In:
     Responses  of  Plants to  Air Pollution.   J.  B.  Mudd 'and T.  T. Kozlowski, eds., Academic
     Press, New York, NY, 1975.  pp. 237-272.

Lee, J.  J.,  E.  M.  Preston, and  R.  A.  Lewis.  A system for  the experimental evaluation of the
     ecological effects of sulfur dioxide.  Proceedings of the 4th  Joint Conference on  Sensing
     of  Environmental  Pollutants,    American Chemical  Society,  Washington,   DC,  1978.   pp.
     49-53.

Legge, A. H.   Primary Productivity, Sulfur Dioxide and the Forest Ecosystem:   an  overview of a
     Symposium  on  Effects  of Air Pollutants on Mediterranean and Temperate  Forest  Ecosystems,
     June  22-27,  1980,  Riverside,  CA.,  P.  Miller,  ed.   Pacific  Southwest  Forest and  Range
     Experiment Station, Berkeley,CA, 1980.  pp. 51-62.

Legge,  A.  H.,  C.  W.  Harver,  P.  F.  Lester,  D.  R.  Jaques, H.   R.   Krouse,   J. Mayo, A.  P.
     Hartgerink, R.  G.  Amundson,  and R.  B.  Walker.   Quantitative  assessment  of  the  impact  of
     sulfur  gas  emissions  on a  forest  ecosystem.   Final  report  submitted to  Whitecourt
     Environmental  Study Group.   Environmental Sciences Centre,  Kananskis, The University  of
     Calgary, Alberta, 1976.

Legge, A.  H.,  D.  R.  Joques, G. W.  Harvey,  H, R. Krouse,  H.  M. Brown,  E. C. Rhodes,  M.  Nosal,
     H.  0, Schellhase,  J.  Mayo,  A.  P.  Hartgerink,  P.  F.  Lester,  R.   G.  Amudson, and  R.  B.
     Walker.   Sulphur gas emissions  in the  Boreal  Forest:    the  West Whitecourt  case study.
     Water, Air, and Soil Pollut. 15:77-85, 1981.

Leone, I.  A.,  and  E. Brennan.  Modification of sulfur dioxide injury to tobacco  and  tomato  by
     varying nitrogen and  sulfur nutrition.   J. Air  Pollut.  Control  Assoc. 22:544-547,  1972.

Lerman,  S. L. ,  and E. F. Darley.   Particles.   In:   Responses of Plants to  Air Pollution.   J.
     B.  Mudd and  T.  T.  Kozlowski,  eds.,  Academic  Press, New York,  NY,  1975.  pp. 141-158.

Liebig,  G, F.   Arsenic.   In:   Diagnostic Criteria  for Plants and  Soils.   H.  D.  Chapman,  ed.,
     University of California, 1966.  pp. 13-23.

Likens,  G. E.,  F.  H. Bormann, R. S.  Pierce, J. S. Eaton, and N. M. Johnson.   Biogeochemistry
     of a Forested Ecosystem.  Springer-Verlag, New York,  1977.
                                           8-68

-------
Lindberg, S.  E.,  A.  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.   Pub.  No.  1299, Oak  Ridge  National  Laboratory,  Oak Ridge,  TN, 1979.

Linzon,  S.  N.  The  influence of  smelter  fumes  on  the growth  of white  pine in the  Sudbury
     region.   Canadian  Department  of Agriculture  Publication,  Ontario  Dept.  Lands Forests,
     1958.

Linzon,  S.  N.   Damage  to  eastern white  pine by  sulfur  dioxide,  semimature tissue  needles
     blight and ozone.  J. Air Pollut. Control Assoc. 16:140-144, 1966.

Linzon,  S.  N.   Economic effects of sulphur  dioxide on forest growth.  J. Air Pollut.  Control
     Assoc.  21:81-86, 1971.

Linzon,  S.  N.   Some  effects of particulate  matter  on vegetation.   In: Ontario Proceedings of
     the 3rd  International Clean Air Congress.  Dusseldorf, West Germany,  1973,  pp. A118-A12Q.

Linzon, S. N.   Effects of sulfur oxides on vegetation.  For. Chron. 48:182-186, 1972.

Linzon,  S.  N.   Vegetation injury by airborne  arsenic and  sulphur dioxide emissions from gold
     smelters.  Proceedings of the Fourth International Clean Air Congress, Tokyo, Japan, 1977.

Linzon,  S.  N.   Effects  of airborne sulfur pollutants on plants.  In:  Sulfur in the Environ-
     ment:  Part  II,  Ecological  Impacts.  J. 0.  Nriagu, ed., John Wiley & Sons. New York, NY,
     1978.  pp. 109-162.

Linzon,  S.  N.   Acute  and  chronic  effects  of  sulfur  dioxide  on  natural  vegetation.   In:
     Proceedings  of  the  Specialty  Conference on the Proposed  SO   and Particulate Standard.
     Atlanta,   GA,  September  16-18,  1980.  Air  Pollution Control  (Association, ed. , 1981.  pp.
     50-61.

Linzon,  S.  N.  P. J.  Temple and  R.  G.  Pearson.  Sulfur  concentration  in plant foliage and
     related effects.  J.  Air Pollution Control Association 29:520-525, 1979.

Little,  P.    A  study  of  heavy  metal  contamination  of  leaf  surfaces.    Environ.   Pollut.
     5:159-172, 1973.

Little,  P.  Deposition  of 2.75,  5.0 and 8.5mm 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.
     Deposition of exhaust  Pb to plant  and  soil  surfaces.   Atmos. Environ.  11:437-447, 1977.

Majernik, 0.,  and T.  A.  Mansfield.   Direct effect of  SO, pollution on the degree of opening of
     stomata.   Nature (London) 227:377-378, 1970.      ^

Malhotra, S.  S. ,  and A. A.  Kahn.   Effects  of sulfur dioxide fumigation on lipid biosynthesis
     in pine needles.  Phytochemistry 17:241-244, 1978.

Mandl, R. H. ,   L. H, Weinstein, and M. Keveny.  Effects of hydrogen fluoride and sulfur  dioxide
     alone and in combination on several species  of plants.  Environ. Pollut.  £:133-143, 1975.

Manning,  W.  J.   Effects  of limestone  dust  on leaf  condition,  foliar  disease incidence, and
     leaf surface microflora of native plants.  Environ. Pollut. 2:69-76,  1971.

Mansfield.  T.   A.,  and 0.  Majernik.  Can  stomata  play a part in protecting plants against air
     pollution?  Environ.  Pollut. 1:149-154, 1970.
                                           8-69

-------
Markowski, A.,  S.  Grzesiak, and M. Schramel.   Indexes  of  susceptibility of various  species  of
     cultivated  plants to  sulfur  dioxide  action.   Bull.  Acad.  Pol.  des  Sci.  23(9):637-646,
     1975.                                                                       ~

Masaru,  N.,   F.  Syozo,  and K.  Saburo.   Effects of exposure  to various  injurious gases  on
     germination of lily pollen.   Environ.  Pollut.  11:181-187,  1976.

Materna, J.,  J.  Jirgle, and J. Kucera.  Vysledky merceni  koncentraci  lyclicniky siricitheo  v.
     lesich  krusnych  nor,    (Measurement results of  sulfur  dioxide  concentrations  in  the Ore
     Mountain Forests.)  Ochr. Ovzdusi.  6:84-93,  1969.

Hatsushima, J.,  and  R. F.   Brewer.  Influence  of  sulfur dioxide and  hydrogen ..fluoride as a mix
     or  reciprocal exposure on citrus  growth  and development.   J. Air  Pollut.  Control Assoc.
     22:710-713, 1972.

Hay, P.  F.,  and A. M. Downes,5 Nutrient cycling  In grazed pastures.   I.  A preliminary inves-
     tigation of the use of [  S]  gypsum.   Aust.  J. Agric. Res.  19:531-543, 1968.

Hay, P.  F. ,  A.  R.  Till, and M. J. Gumming.  Systems  analysis of  sulfur kinetics in pastures
     grazed by sheep.  J. Appl. Ecol. 9:25-49,  1972.

Mclaughlin, S.  B.   S0?, Vegetation effects and the air quality standard:   limits  of interpre-
     tation and  application.   In:   Proceedings of  the  Proposed SO  and  Participate Standard,
     Atlanta,  GA,  September 16-18,  1980,   Air  Pollution  Control  Association,  ed.,  1981, pp.
     62-83.

Mclaughlin, S. B., and N. T. Lee.  Botanical Studies  in the  Vicinity of  the Widows Creek Steam
     Plant.   Review  of Air Pollution Effects  Studies,  1952-1972,  and  Results of 1973 Surveys.
     Internal Report I-EB-74-1, TVA, 1974.

Mclaughlin,  S.  B. , and  D.   S.  Shriner.  Allocation  of resources  to defense and  repair.  In:
     Plant Disease:   An Advanced Treatise.  J. G.  Horsfall  and E.  B. Cowling,  eds., Academic
     Press, New York, NY, 1980.

Mclaughlin,  S.  B.,  and G.   E.  Taylor.   Relative humidity:   important  modifier  of  pollutant
     uptake by plants.  Science (Washington, DC)  211:167-169, 1981.

Mclaughlin, S.  B., V.  J. Schorn,  and H. C. Jones.  A programmable exposure system for kinetic
     dose-response studies  with air pollutant  J.  Air  Pollut. Control Assoc.  26:132-135.  1976.

Mclaughlin, S.  B. ,  D.  S. Shriner, R. K. M'Conathy, and L. K. Mann.  The  effects of  SQ  dosage
     kinetics  and exposure frequency  on   photosynthesis  and  transpiration of   kidney  beans
     (Phaseolus vulgaris L).  Environ.  Exp. Bot.  19:179-191, 1979.

Meidner, H.»  and T.  A. Mansfield.  Physiology of stomata.  McGraw-Hill, Great  Britian, 1968.

Menser,  H. A.,  and H.  E. Heggestad.   Ozone and  sulfur dioxide synergism:   injury  to  tobacco
     plants.   Science  (Washington, DC)  153:424-425, 1966.

Heszaros,  E.,  G. Varhelyi, and L.  Haszpra.    On  the  atmospheric sulfur budget  over  Europe.
     Atmos. Environ.  12:2273-2277, 1978.

Miller,  P.,  and J.  R. McBride.  Effects of air pollutants on forest.  In:   Response of Plants
     to  Air  Pollution.   J.  B.  Mudd and T.  T.   Kozlowski,  eds.,  Academic Press,  New  York, NY,
     1975.  pp. 195-235.
                                           8-70

-------
Miller, J.  E. ,  H.  J.  Smith,  P.  G.  Sprugel,  and P. B. Xerikos.  Yield  response  of  field-grown
     soybeans  to an  acute  S09  exposure.   Radiol.   Environ.  Res.  Div.  Annu.  Rep.,  Argonne
     National Laboratory, ANL-7B-65, Part  III, 1979.

Miller,  J.   E. ,  D.  G.  Sprugel,  R.  N.  Muller,  H.   J.  Smith,  and P.  B.  Xerikas.   Open-air
     fumigation  system  for  investigating sulfur dioxide effects on crops.   Phytopathology 70:
     1124-1128,  1980.

Moss,  M.   R.    Biogeochemical  cycles  as  integrative  and  spatial  modes   for  the  study  of
     environmental  pollution  (the  example  of  the   sulphur  cycle).  Int.  J.  Environ.  Stud.
     9:209-216,  1976.

Mudd, J. B.  Sulfur dioxide.  In:  Responses of  Plants to Air  Pollution.  J.  B.  Mudd  and  T.  T.
     Kozlowski,'eds.,   Physiological Ecol. Mono.  Series,  Academic Press,  Inc., New  York,  NY,
     1975.  pp.  9-22.

Murray, J.  J. ,  R.  K.  Howell,  and A.  C.  Wilton.   Differential  response of 17 Poa  pratensis
     cultivars  to ozone and  sulfur dioxide.  Plan  Dis. Rep. 59:852-854, 1975.

National  Academy of  Sciences.    Series  on  Medical   and  Biological  Effects of Environmental
     Pollutants.  National Academy of Sciences,  Washington, DC,  1973.

National  Academy  of   Sciences.    Principles   for  Evaluating  Chemicals  in  the Environment.
     National Academy of Sciences, Washington, DC, 1975.

National  Academy  of   Sciences.    Sulfur  oxides.   Board  on  Toxicology  and  Environ. Health
     Hazards.   U.S.  Environmental  Protection Agency Contract,  Report  No.  68-01-4655,  1978.

Neely, G. E., D. T. Tingey,  and  R. G. Wilhour.   Effects of ozone and sulfur dioxide singly and
     in  combination  on  yield,  quality  and  N-fixation  of  alfalfa.   Proceedings  of  the
     International   Conference   on   Photochemical   Oxidant   Pollution    and   Its   Control.
     EPA-600/3-77-001b, 1977.  pp. 663-673.

Nielsen,  D. G. ,  L.   E.  Terrell,  and T.  C.   Weidensaul.   Phytotoxicity of  ozone  and sulfur
     dioxide to  laboratory fumigated scotch pine.  Plant Dis.  Rep.  £1:699-703, 1977.

Noggle,  J.  C. ,  and  H.  C.  Jones.    Accumulation  of  Atmospheric Sulfur by Plants  and Sulfur-
     supplying  Capacity  of  Soils.   EPA-600/7-79-109,  U.S.  Environmental  Protection  Agency,
     1979.

Noland, T.  L. ,  and T.  T. Kozlowski.  Effect of  SO^ on stomatal  aperature and sulfur  uptake of
     woody  angiosperm seedlings.  Can. J.  For. Res. 9:57-62,  1979.

O'Connor,  J.  A., D.  G.  Parbery,  and  W.  Strauss.   The effect  of phytotoxic gases  on native
     Australian plant  species.    Part  I.  Acute effects  of  sulfur  dioxide.   Environ.  Pollut.
     7:7-23, 1974.

Odum,  E.  P.  Fundamentals of Ecology,   3rd  ed.   W.   B. Saunders,  Co.,  Philadelphia,  PA,  1971.

O'Gara,  P.  J.    Sulfur  dioxide  and  fume  problems and  their  solutions.  In:  14th Semiannual
     Meeting  of the  American Institute  of  Chemical  Engineers.   J. Ind.  Eng.  Chem.  14:744,
     1922.

Oshima,  R.  J.    The impact  of sulfur dioxide  on vegetation:   a sulfur dioxide-ozone response
     model.  Final  Report of the California  Air Resources Board  Agree.  No.  A6-162-30,  1978.
                                            8-71

-------
Pahlich, E.  Allosterische Regulation der Aktivitat der Glutamat-dehydrogenase aus Erbsenkeim-
     lingen durch  das  Substrat a-Ketoglutarsaure.  [Allosteric  regulation  of the activity of
     glutamate dehydrogenase from pea seedlings by the substrate alpha-ketoglutarate.]  Planta
     100:222-227, 1971,

Pahlich,  E.   Uber  den  Hemm-Mechanismus  mitochondriater  Glutamat-Qxalacetat-transaminase in
     SOp-begaster  Erbsen.   [Mechanism  of  inhibition of  mitochondria!  glutamate-oxalacetate-
     transaminase in sulfurdioxide fumigated peas.]  Planta 110:267-278, 1973.

Pahlich, E.  Effect of SO^-pollution on cellular  regulation.  A general concept of the mode of
     action of gaseous air contamination.   Atmos. Environ. 9:261-263, 1975.

Parish,  S.  B.   The effect of cement  dust  on citrus trees.   Plant World  13:288-291,  1910.

Pelz, E.  Intersuchungen ueber die individuelle Rauchhaerte von Frehten  [The individual smoke
     resistance of spruce.]  Wiss. Z. Tech Univ., Dresden  3J:3:595-600, 1962.

Preston,  E.  M.   The  ecological   implications  of chronic  sulfur dioxide exposure  for native
     grasslands.   Presented  at  the 72nd  Annual Meeting Air Pollution  Control  Association,
     Cincinnati, OH, 1979.

Preston,  E.  M., and  R.  A. Lewis.   The Bioenvironmental  Impact of  a  Coal  Fired Power Plant,
     Third  Interim   Report,   Colstrip,  Montana,  December,  1977.   EPA-600/3-78-021,  U.S.
     Environmental  Protection Agency, 1978.

Rajput, C. B. S.,  D.  P.  Ormrod,  and D. W. Evans.  The  resistance  of strawberry to ozone and
     sulfur dioxide.  Plant Dis.  Rep. 61:221-225, 1977.

Rao,  D.  N. ,  and  F.  LeBlanc.    Effects of  sulfur dioxide  on  the lichen alga  with special
     reference to chlorophyll.   Bryologist ^9:69-75, 1965.

Reinert,  R.   A,   Pollutant  interactions  and  their  effects  on  plants.    Environ.  Pollut.
     9:115-116,  1975.

Reinert, R. A.,  and T.  N.  Gray.  The  response of radish to nitrogen dioxide, sulfur dioxide,
     and ozone alone and in combination.  J.  Environ.  Qua!., 10:240-243, 1981.

Reinert, R. A.,  and P. V. Nelson.  Sensitivity and growth of five Elatior Begonia cultivars to
     S00 and 0-, alone and in combination.   J. Am. Soc.  Hort. Sci. 105:721-723, 1980.
       £      *3                                                    ""
Reinert, R. A.,  and D.  E.  Weber.  Ozone and sulfur dioxide-induced  changes in soybean growth.
     Phytopathology 70:914-916, 1980.

Reinert, R. A.,  A.  S. Heagle,  and W. W. Heck.  Plant Response to Pollutant Combinations.  In:
     Responses  of  Plants to Air Pollution.    J.  B.  Mudd and T.  T.  Koslowski,  eds. , Academic
     Press Inc., New York, NY,  1975.  pp.  159-177.

Rennie, P. J. ,  and R. L. Halstead.  The  effects of sulfur on  plants  in Canada.  In:  Sulfur
     and  its  Inorganic Derivatives  in the  Canadian  Environment.   National  Research Council
     Canada, NRCC No.  15015:69-179,  1977.

Reuther, W.,  and C.  K.  Labanauskas.  Copper.   Im   Diagnostic Criteria for  Plants and Soils.
     H. D. Chapman, ed., University  of  California, Berkeley, CA, 1966.  pp. 157-179.

Ricks,  G.  R. ,  and  J.  H.  Williams.   Effects of  atmospheric  pollution  on deciduous woodland.
     Part  2:   Effects of  particulate  matter upon stomatal diffusion  resistance in leaves of
     Quercus petraea (Mattuschka) Leiol. Environ. Pollut. §:87-109,  1974.
                                           8-72

-------
Rist, D. L. ,  and  D.  D,  Davis.   The Influence of exposure temperature and relative humidity on
     the response of pinto bean foliage to SCL.  Phytopathology 69:231-235, 1979.

Rosenberg,  C.  R. ,  R.  J.  Hutnik, and D. D.  Davis.  Forest composition of varying distances from
     a coal-burning power plant.   Environ. Pojlut. 14:307-317, 1979.

Saunders, P.  J. W.   Effects of atmospheric pollution of leaf surface microflora.  Pestic Sci.
     4:589-594, 1973.

Schoenbeck, H.  Beobachtungen  zur  Frage des Einflusses  von  industriellen Immissionen auf die
     Krankheitbereitschaft  der  Pflanze.   [The  effect of industrial pollution  on the disease
     susceptibility of plants.]  Berichte aus der Landestalt f. Bodennutzungsschutz des Landes
     Nordhein-Westfallen.  pp 89-98, 1960,

Schoenbeck, H.  Detection  of heavy metal-containing  air pollutants by selected plant indica-
     tors.   Ver.  Dtsch.  Ing. 203:75-85, 1973.

Sehmel,  G.  A., and  W.  A.  Hodgson.   Predicted  dry  deposition velocities.   In:   Atmosphere-
     surface exchange of particulate and gaseous pollutants symposium.   Conf-74092.  Richland,
     WA, 1974.  pp. 399-422.

Setterstrom,  C. ,  and P.  W. Zimmerman.  Factors influencing susceptibility of plants to sulfur
     dioxide injury.   Contrib.  Boyce Thompson Inst. 10:155-186, 1939.

Sharp,  E.   L.   Atmospheric ions  and  germination  of  uredospores of  Puccinia  striiformis.
    •Science (Washington, DC) 156:1359-1360, 1967.

Sharp,  E.  L.   Relation  of air  ions  to air pollution and  some  biological  effects.   Environ.
     Pollut. 3:227-239,  1972.

Sheridan,  R.  P.    Impact  of emissions  from  a  coal-fired electricity  generating facilties on
     N2-fixing in lichens.  Bryologist 82:54-58, 1979.

Shinn, J.  H.   A  critical survey of measurements of foliar deposition of airborne sulfates and
     nitrates.   Presented  at  the   71st  Annual  Meeting, Air  Pollution  Control  Association,
     Houston,  TX, June 25-30, 1978.  Paper No. 78-7.2.

Shinn, J. H.,  and S.  Lynn.  Do manmade sources affect the sulfur cycle of northeastern states?
     Environ.  Sci. Techno!. 13:1062-1067, 1979.

Showman, R.  E.   Lichens  as indicators of  air  quality  around a  coal-fired power generating
     plant.  Bryologist 78:1-6, 1975,

Shriner, D. S. ,  and G.   S.  Henderson.   Sulfur distribution and cycling  in  a deciduous forest
     watershed.  J. Environ. Qua!.  7:392-397, 1978.

Shugarti H.  H. ,   and D.  C.  West.   Development of an Appalachian deciduous forest succession
     model   and  its  application to  assessment  of  the  impact  of  the  Chestnut  Blight.   J.
     Environ.  Manag.   5:161-179, 1977.

Sij,  J.  W.,  E.  T.  Kanemasu,  and  S.  M.  Goltz.   Some preliminary results  of  sulfur dioxide
     effects  on  photosynthesis  and  yield  in  field-grown  wheat.  Trans  Kans, Acad.  Sci.
     76:199-207, 1974.

Smith, H.  J. , and D, D. Davis.  The  influence of needle age on sensitivity of Scotch pine to
     acute doses of S02-  Plant DIs. Rep. 61:870-874, 1977.
                                           8-73

-------
Smith,  H.  J.,  and D. D.  Davis.   Susceptibility of conifer  cotyledons  and primary needles  to
     acute doses of sulfur dioxide.  Hortic. Sci. 13:703-704, 1978.

Smith,  R.  L.   Ecology  and Field Biology,  3rd  ed.,  Harper and  Row,  New York, NY, 1980.  pp.
     11-199.

Smith,  W.  H.   Metal  contamination  of urban woody plants.   Environ.  Sci. Techno!. 7:631-636,
     1973.

Spedding,  D.  J.    Uptake  of  sulfur  dioxide  by  barley  leaves  at  low  sulfur   dioxide
     concentrations.  Nature (London) 224:1229-1231, 1969.

Spinka, J.  Effects of polluted air on fruit trees and  legumes.  Ziva 19:13-15, 1971.

Sprugel, D. G.,  J.  E. Miller, R.  N.  Muller,  H. J. Smith,  and  P.  B. Xerikos.  Sulfur dioxide
     effects on  yield and  seed quality  in field-grown soybeans.   Phytopathology  70:1129-33,
     1980.                                                                          ~~

Stewart, D., M.  Treshow,  and F. M. Harner.  Sulfur dioxide and  hydrogen fluoride  emissions  in
     regard to phytopathology.  Can. J. Bot. 51:983-988, 1973.

Sulphur  Institute.    Crop  Responses  to  Sulphur  in  North  America,  vol.   18.   The  Sulphur
     Institute, Washington, DC, 1971.

Sulphur  Institute.   Sulphur  in  Agriculture Series,  vol.  3.  J.  S.  Platu,  ed. ,   The  Sulphur
     Institute, Washington, DC, 1979.

Temple, P.  J.   Dose-response of urban trees to sulfur  dioxide.  J. Air Pollut. Control  Assoc.
     22:271-274, 1972.

Terman, G.  L.   Atmospheric sulfur  -  the  agronomic aspects.  Tech.  Bull.  no. 23.   The  Sulfur
     Institute.  Washington, DC, 1978.  15  p.

Thomas, M. D.   Gas damage to plants.  Ann.  Rev. Plant Physiol. 2:293-322,  1951.

Thomas, M.  D.,  and  R. H.  Hendricks.   Effect  of air pollution on  plants.  Iji:  Air Pollution
     Handbook.   P. L. Magill et al. eds., New York, NY,  1956.  pp. 9:1-44.

Thomas, M.  D. ,  and  G. R.  Hill,  Jr.   Absorption of sulfur dioxide by alfalfa and  its  relation
     to leaf injury.  Plant Physiol. 10:291-307, 1935.

Thomas, M.  D.,  R.  H. Hendricks, T.  R.  Collier, and G.  R.  Hill.  The  utilization  of  sulphate
     and  sulphur dioxide for  the  sulphur  nutrition  of alfalfa.   Plant  Physiol.  18:345-371,
     1943.

Thompson,  L.   K.,   S.  S.  Sidhu,  and  B.   A.   Roberts.   Fluoride accumulations   in  soil and
     vegetation  in  the vicinity  of a  phosphorus  plant.   Environ.   Pollut.   18:221-234,  1979.

Thompson, R. C. , and G. Katz.  Effects of continuous H?S fumigation on  crop  and forest plants.
     Environ.  Sci. and Tech. 12:550-553, 1978.        ^

Tingey, D.  T. ,  and  R. A.  Reinert.   Effect of ozone and sulfur  dioxide singly and in  combina-
     tion on plant growth.  Environ. Pollut. 9:117-125,  1975.

Tingey,  D.  T. , W.  W. Heck,  and R. A.  Reinert.   Effect  of low  concentrations  of ozone and
     sulfur  dioxide  on  foliage,   growth,  and  yield  of  radish.    J.   Am.   Soc.  Horti.  Sci.
     96:369-371, 1971b.
                                            8-74

-------
Tingey, D. T. , R. A. Reinert, J. A. Dunning, and W. W, Heck.  Vegetation  injury  from the  inter-
     action  of  nitrogen  dioxide  and  sulfur  dioxide.   Phytopathology  61:1506-1511,  1971a.

Tingey, D,  T. ,  R.  A.  Reinert,  J.  A. Dunning,  and W.  E.  Heck.   Foliar injury responses of
     eleven plant species to ozone/ sulfur dioxide mixtures. Atmos. Environ. 7:201-208, 1973a.

Tingey, D. T. ,  R.  A.  Reinert,  C.  Wickliff,  and W. W. Heck.   Chronic ozone or  sulfur dioxide
     exposures,  or  both, affect the  early  vegetative growth of  soybean.   Can.  J.  Plant  Sci.
     53:875-879, 1973b.

Treshow, M.  Environment and Plant Response.  McGraw-Hill Book Co., New York, NY, 1970.

Tyler, G.   Heavy metals pollute nature, may reduce productivity.  Ambio 1:53-59, 1972.

Ulrich, A.,  and K.  Qhki.  Potassium.   In:   Diagnostic Criteria  for  Plants  and Soils.   H. D.
     Chapman, ed., University of California, Berkeley, CA, 1966.  pp. 157-179.

Ulrich, A., M. A. Tabatabai, K. Ohki  and C. M. Johnson.  Sulfur content of alfalfa  in relation
     to growth in filtered and  unfiltered air.  Plant Soil 26:235-252, 1967.

Unzicker,   H.  J., H.  J.  Jager, and L.  Steubing.   Influence of S09 on  the vitamin content of
     plants.   Angew. Bot. 49:131-139, 1975.                       ^

U.S.  Department of  Health,  Education,  and  Welfare.    Air  Quality  Criteria  for  Particulate
     Matter.    Public  Health  Service,  EHS,  National  Air  Pollution  Control  Administration,
     Washington, DC 1969.

U.S. Environmental Protection Agency.   Mount Storm, West Virginia -Gorman, Maryland, and  Luke,
     Maryland  - Keyser,  West  Virginia  Air  Pollution Abatement Activity,  Research Triangle
     Park, N.C.  1971.

U.S. Environmental  Protection  Agency.  Effects of  Sulfur  Oxides in the  Atmosphere on Vegeta-
     tion;  revised chapter  5  for Air  Quality Criteria  for Sulfur Oxides.    EPA-R3-73-030,
     National  Environmental  Research Center,  U.S.  Environmental  Protection Agency, Raleigh,
     NC, 1973.

U.S.  Environmental  Protection  Agency.   The Bioenvironmental  Impact  of Fine Particulates:   A
     Critical Review and Summary.  USEPA/CERL Report, 1975.  Mimeo, 49 pp.

U.S. Environmental  Protection  Agency.  Diagnosing Vegetation  Injury  Caused by  Air Pollution.
     D, R.  Hicks, ed.,  EPA  Contract  Pub.  No.  68-02-1344,  Research  Triangle Park, N.C.  1976.

U.S.  Environmental  Protection  Agency.   Airborne  Particles.    Environmental  Health  Effects
     Series, Report No. EPA-600/1-77-053, 1977.

U.S. Environmental Protection Agency.   Office of Air Quality Planning and Standards.  National
     Air Quality, Monitoring and Emissions Trends Report, 1977, EPA-450/2-78-052, U.S. Environ-
     mental  Protection Agency,  Research Triangle Park, NC, 1978b,

U.S. Environmental Protection Agency.   Emission of Sulfur-Bearing Compounds  from Motor Vehicle
     and  Aircraft  Engines.   A  Report to Congress.   EPA-600/9-78-028, U.S.   Environmental  Pro-
     tection Agency, Washington, D.C.  August 1978a.

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.
                                           8-75

-------
Van  Haut,  H.,  and  H.  Stratmann.   Farbtafelatlas uber  Schwefeldioxidwirkugen and  Pflanzen.
     (Color-Plate  Atlas  of  the  Effects of  Sulfur  Dioxide  on Plants.)  Verlag W.  Glrardet,
     Essen, West Germany, 1970,

Vanselow, A,  P.   Nickel.   In:  Diagnostic  Criteria for Plants and  Soil.   H.  D. Chapman,  ed.,
     University of California, Berkeley,  CA, 1966.  pp. 302-309.
Vins, B., and R. Mrkva.   The diameter  increment  losses  of pine  stands  as  a  result of  injurious
     emissions.  Acta Univ. Ser. C Agric.-Brno 42:25-46, 1973.

Vogl, M.} S. Bortitz, and H. Roister.  Physiological  and biochemical contributions  to research
     of  fume  damage.  6th report.   Definitions  of degrees of  damage  and forms of resistance
     against the SO^ component of injurious fumes.  Biol. Zentralb.  84:763-777,  1965.

Wainwright, M.  Microbial S-oxidation in soils  exposed to  heavy atmospheric  pollution.   Soil
     Biol. Bioehem. 11:95-98, 1979.

West, D.  C.,  S.  B. Mclaughlin, and  H. H. Shugart.   Simulated  forest  response to  chronic  air
     pollution stress.  J. Environ. Qua!. 9:43-49,  1980.

White,  K.  L.,  A.   C.  Hill,  and J.  H.   Bennett.   Synergistic inhibition  of apparent photo-
     synthesis  rate  of  alfalfa  by   combinations  of  sulfur  dioxide and  nitrogen  dioxide.
     Environ. Sci. Techno!. 8:574-576, 1974.

Wilhour, R. G,, A. Neely, D. Weber and L. Grothaus.   The response of selected  small grains  and
     range grasses,  and  alfalfa  to  S0?.  In:    Bioenvironmental  Impact of a  Coal-fired Power
     Plant:  An Interim Report.   E,  M.  Preston  and T.  L. Gallett,  eds., EPA-600/3-79-044,  U.
     S.  Environmental   Protection  Agency, Corvallis  Environmental   Res.  Lab., Corvallis,   OR,
     December 1978, pp. 592-609.

Winner, W. E., and J. D.  Bewley.  Terrestrial mosses  as bioindicators  of  SO, pollution stress.
     Synecological analysis and the index of atmospheric purity.  Oecologia 35:221-230, 1978a.

Winner, W. E., and J. D.  Bewley.  Contrasts between bryophyte and vascular plant synecological
     responses  in an  S0»-stressed  white spruce  association  in central Alberta.    Oecologia
     33:311-325, 1978b.

Winner,  W.  E.  and  H.  A.  Mooney.   Ecology of  SO, resistance:   II  Photosynthetic changes  of
     shrubs in  relation to SO, absorption and stoHiatal behavior.  Oecologia 44:296-302, 1980.

Winner, W. E., J.  D. Bewley, H. R. Krouse, and H. M.  Brown.  Stable  sulfur  isotope  analysis  of
     SO, pollution impact on vegetation.  Oecologia _36:351-361, 1978.

Woodwell,  G.    Effects  of pollution  on  structure  and physiology  of  ecosystems.    Science
     (Washington,  DC) 168:429-433, 1970.

Wu,  L.,  and A, D.  Bradshaw.   Aerial  pollution and  the  rapid  evolution of copper tolerance.
     Nature (London) 238:167-169, 1972.

Yopp, J. H.,  W.  E. Schmid, and  R.  W.  Hoist.   Determination  of maximum permissible  levels  of
     selected chemicals that exert toxic  effects on plants of economic importance in  Illinois.
     Illinois Institute for Environmental Quality,  1974.

Zahn, R., Wirkungen  von Schwefeldioxyd  auf die  Vegetation, Ergebnisse aus Begasungsversuchen.
     [Effects of  sulfur-dioxide  on  vegetation:   results of  gas exposure experiments.]  Staub
     Reinholt Luft.  21:56-60, 1961.
                                           8-76

-------
Zahn, R, Uber Einfluss verschiedener Umweltfaktoren auf die Pflanzenempfindlichkeit
     gegenuberSchwefeldioxyd. [Effects  of various environmental factors  on  plant sensitivity
     to sulfur dioxide,]  Z. Pflanzenkr.  Pflanzenschutz 70:81-95,  1963a.

Zahn,   R.    Untersuchungen   uber  die  Bedeutung  Kontinuierlicher   and  interim'ttierender
     Schwefeldioxideinwirkung fur die  Pflanzenreaktion.   (Investigations  on plant reaction to
     continuous  and/or   intermittent  sulfur   dioxide   exposure.)    Staub   Reinholt  Luft,
     23:343-352, 1963b.

Zahn, R.  The  effect on plants of  a  combination of subacute and toxic sulfur dioxide doses.
     Staub 30:20-23, 1970.
                               2~
Ziegler,  I.   The  effect  of SOt  on  the ability  of ribulose-l,5-d1phosphate  carboxylase in
     insulated spinach chloroplasts.  Planta 103:155-163, 1972.

Ziegler,  I.   The  effect of air polluting  gases  on plant  metabolism.   Environ.  Qual.  Saf.
     2:182-208, 1975.

Zimmerman, P.  W. ,  and W. Crocker.  Toxicity of air containing  sulfur  dioxide gas.   Contrib.
     Boyce Thompson Inst. 6:455-470, 1934.
                                           8-77

-------
APPENDIX 8A
    8-78

-------
TABLE 8A-1.   SUMMARY OF STUDIES REPORTING  RESULTS OF SO, EXPOSURE UNDER FIELD CONDITIONS AND/OR CHAMBERS OVER PLANTS  FOR AGRONOMIC CROPS
Conc.a
M9/ma (PP«0
100 (0.04)
50 (0.02)
130 (0.05)
260 (0.10)
(Geon. means = 58;
100; 178 Mg/m3
or 0.02; 0.04;
0.07 ppm, re-
spectively)
80-260 (0.03-
0.10)
240 (0.09)
260 (0.10)
500 (0.19)
660 (0.25)
940 (0.36)
260 (0.10)
310 (0.12)
790 (0.30)
2070 (0.79)
Exposure Exposure
time condition
3 hr for 8 exp. F/CC
growing season
Growing season F-ZAP
72 hr/wk for
growing season F/CC
Mean cone. F-ZAP
4.2h/18 fumi-
gations from
19 July-27
August
6 hr/d F/CC
43 d
92 d
133 d
Mean cone. F-ZAP
4.7 hr/24 fumi-
gations 13 July
to 29 August
Plant
Wheat
Western wheat
grass, Prairie
June grass
Barley
Durum wheat
Spring wheat
Soybean
cv. Wells
Soybean
cv. Dare
Soybean
cv. Wells
Effects onc d
Foliage Yield Species effect Caveat
No effect on apparent photosynthesis,
no effect on the avg. head length or
no. of grains/head.
S content increased with increasing
SQ2 cone.; digestible dry matter was
decreased by 2 years of treatment;
crude protein content in winter
wheat decreased significantly.
No effect on yield
X X 6.4% yield reduction Sg of fumigation
X X 5.2% yield reduction cone, ranged 44-
X X 12.2% yield reduction 45% of x.
X X 19.2% yield reduction
X X 15.9% yield reduction
X No significant effect on foliar injury,
fresh wt. seeds/plant or wt. of seeds/
plant. 92nd day defoliation was 12% >
control; 135th day seed wt. only 1% <
control .
X X 12.3% yield reduction Sgf of fumigation
X X 20. 5X yield reduction cone, ranged 41-
X X 45.3% yield reduction 64% of x,
Reference
Sij et al.,
1974
Dodd et al, ,
1978
Wilhour et al. ,
1978
Sprugel et al. ,
1980 and
Miller et al.,
1980
Heagle et al. ,
1974
Sprugel et al. ,
1980 and
Miller et al. ,
1980

-------
                                                                        TABLE 8A-1.   (continued)
CO
o
Conc,a
ug/«i3 (ppn)
390 (0.15)
630 (0,25)
1050 (0.40)
2100 (0.80)
3140 (1.20)
1180 (0.45)
2100-5240
(0.80-2.00)
Exposure3 Exposure
time condition
72 hr/wk for F/CC
growing season
Once every week F/CC
(3 hr) to once
in 5 wks (3 hr)
3 hr for 7 exp. F/CC
growing season
4 hr 20 nin F-ZAP
Plant
Barley
Durum wheat
Spring wheat
Alfalfa,
Barley
Durum wheat
Spring wheat
Wheat
Soybean
cv. Wells
Effects onc
Foliage Yield
X
X


X
X
X
Species effect
44% lower yield in Barley (N.S.)
42% lower yield in Durum wheat (N.S,)
No effect on Spring wheat
No effect on yield
No effect on yield
No effect on yield
No effect on yield
No accumulative effect on yield, no
effect on avg. head length or no.
grains/head
4.5% lower yield at 3760 |ig/n3
(1.4 ppm)
11% lower yield at 4450 ug/m3
(1.7 ppm)
15% lower yield at 5240 pg/m3
(2.0 ppm)
Caveat6 Reference
Wilhour et al,
1978
Wilhour et al.,
1978
Sij et al.,
1974
Pollutant avg.; Miller et al.,
no est. on range 1979
of exposure doses
     Table arranged by increasing SO, concentration as first order and exposure time as second order divisions.   Treatments within a single study that did not induce
    specifically different effects are listed along with the lowest S0« concentration that induced said effect.
     F/CC = field, closed chambers; F/OT = field, open top chambers, F-ZAP = field, zonal  air pollution system.
    CX indicates study found foliage and/or yield effects.
     Host prominent or significant effect reported.
    eCaveats for consideration about proper study design and interpretations.
     Sg = Standard geometric deviation.
    N.S, = Results were not significant at 95 percent level of confidence.
    Note:  1 ppm SO  = 2620
                   2

-------
                                 TABLE 8A-2.  SUMMARY OF STUDIES REPORTING RESULTS OF S02 EXPOSURE UNDER LABORATORY CONDITIONS
                                                              FOR AGRONOMIC ANO HORTICULTURAL CROPS
oo
Conc.a
pg/rn3 (ppm)
92 (0.035)
465 (0.175)
130 (0.05)



130 (0.05)
260 (0.10)
660 (0.25)

130 (0.05)
520 (0.20)

178 (0.068)

180-1390
(0.07-0.53)







Exposure Exposure Effects onc
tine condition Plant Foliage Yield
8 hr EC/SD Broadbean

5 hr/d; 5 d/wk EC/SD Alfalfa X
for 4 wk X
Tobacco, Bel W3 X
Tobacco, Bur ley 21
4 hr EC/SD Oats
Radish
Soybean
Tobacco
8 hr/d EC/SO Soybean
5 d/wk
for 18 d
103.5 hr/wk GC Cocksfoot X
for 20 wks Meadowgrass X
24 hr/d GC Tobacco X
9-20 d

Sunflower X


Corn X


d e
Species effect Caveat
Depressed net photosynthesis

26% less foliage dry trt. at final harvest
49% less root dry wt. at final harvest
22% less leaf dry wt. at final harvest Sensitive plant
No effect
No foliar injury



No effect on top fresh or dry wt. , root
fresh or dry wt. ; plant height, shoot/root
fresh or dry wt. ratio
40X less total dry wt.
28% less total dry wt.
Increased dry wt. yield up to 1390 MS/w3 No monitoring
(0.53 pp») (44% > control) methods pre-
sented; Low S
Greatest dry wt. yield: 44X in soil medium
> control at 920 pg/m3 (0 35 ppm);
27X > control at 1390 ug/m3 (0.53 ppm).
Greatest dry wt. yield: 2A%
> control at 450 ug/n3 (0.17 ppm);
7X > control at 1390 ug/wa (0.53 ppm).

Reference
Black and
Unsworth,
Tingey and
Re inert,


Tingey et
1971a


Tingey et
1973b

Ashenden,




1979

1975


al.,



al.,


1979

Faller, 1970

-------
                                                                          TABLE 8A-2  (continued)
oo
 i
oo
ro
Cone. Exposure
ug/«a (ppn) tine
260 (0.10) 18 d
290 (0.11) 24 hr/day
4 wk
325-2620 1 and 3 hr
(0.125-1.0)
390 (0.15) 18 d
390-790 24 hr/d
(0,15-0.30) 7 d
465 (0.175) 2 hr
Exposure
condition Plant
GC Pea
GC Italian ryegrass
EC/SB Oats
Radish
Sweet pea
Swiss chard
GC Pea
EC/SD Barley
Bean
Corn
EC/SD Broadbean
Effects onc d
foliage Yield Species affect Caveat
X 3% less fresh wt. shoot
5% less dry wt, shoot
4% less tojal nitrogen
30% less H (buffer capacity)
10% more glutamate dehydrogenase activity
110% more inorganic sulfur content
X No difference from control at low wind;
17X-4QX less total dry wt. at high wind
No foliar injury at concentrations
< 0.50 ppm; 2% maximum foliar injury
experienced at all doses
X 3% less fresh wt. of shoot Water culture
8% less dry wt. of shoot
2% less total nitrogen
35% less H (buffer capacity)
32% more glutamate dehydrogenase activity
140% more inorganic sulfur content
X Severe foliar injury
X No injury
X Severe foliar injury
dec. photosynthetic rate, dec. stomatal
resistance if RH > 40%, inc. stomatal
resistance if RH < 40%
Reference
Jager and Klein,
1977
Ashenden and
Mansfield, 1977
Bennett et al.,
1975
Jager and Klein,
1977
Mandl et al.,
1975
Black and
Unsworth, 1979

-------
                                                                            TABLE  8A-2 (continued)
oo
CA>
Cone. a Exposure8
Mg/m3 (ppm) time
520-790 2 hr
(0.20-0.30)
520 (0.20) 30, 78, 100
hr

520 (0.20) 15 d



520 (0.20) Continuous
to maturation
660 (0.25) 4 hr










660 (0. 25) 18 d





Exposure
condition Plant
GC Alfalfa
Bar! ey
EC/SO Wheat


GC Tomato



EC/SD Kidney bean

EC/SD Broccoli
Tobacco, Bel B
Alfalfa
Onion
Soybean
Lima bean
Bromegrass
Cabbage
Radish
Spinach
Tomato
GC Pea





Effects onc ,
Foliage Yield Species effect Caveat
Threshold dose for inhibition of photo-
synthesis, reversible effect
X X Trend of increased dry wt. for 19 of 21 Trend, not
exposures; small amount foliar injury significant
from control
X Threshold dose for initial symptoi of
tissue death, dec. or no change in
vitamin BI, Be, and nicotinic acid
content
X < 15% of total yield; no change in
protein content
X 6% leaf injury
X 1% leaf injury
No effects
No effects
No effects
No effects
No effects
No effects
No effects
No effects
No effects
X 32% less fresh wt. of shoot Mater culture
26% less dry wt of shoot
24% less total nitrogen
42% less H (buffer capacity)
80% more glutamate dehydrogenase activity
150% more inorganic sulfur content

Reference
Bennett and
Hill, 1973a
Laurence, 1979


Unzicker et al. ,
1975


Berigari et al. ,
1974
Tingey et al. ,
1973a









Jager and Klein,
1977

-------
                                                                        TABLE 8A-2 (continued)
CO
Cone.8
(ig/m3 (ppn)
2620 (1,00)



2620 (1.00)






2620 (1.00)

2620 (1.00)


3930 (1.50)


3930 (1.50)

3930 (2.00)



Exposure3
time
2 hr



4 hr






6 hr/d
for 3 d
1.5 hr

3 hr
.75-3 hr


3 hr

2 hr



Exposure
condition Plant
EC Begonia
Petunia
Coleus
Snapdragon
EC/SO Broccoli
Bromegrass
Cabbage
Lima bean
Radish
Spinach
Tomato
EC/SO Strawberry

EC/SO Soybean


EC/SO Soybean


EC/SD Alfalfa

EC Begonia
Petunia
Coleus
Snapdragon
Effects
Foliage




X
X
X
X
X
X
X
X

X


X


X





onc
Yield

X
X
X









X

X
X




X
X
X
X
H
Species effect
No effect
30% less flower #'s; 19% less shoot wt.
27% less flower #'s; 19% less shoot wt.
14% less flower #'s; 16% less shoot wt.
38% leaf injury
65% leaf injury
70% leaf injury
25% leaf injury
46% leaf injury
49% leaf injury
33% leaf injury
No effect on growth and development;
necrotic lesions, lower leaf surface
9% less shoot fresh wt. , 4% leaf injury

21-29% less shoot fresh wt.
24-94% less shoot fresh wt. , 63-93%
foliar injury

Leaf necrosis at 315 ppiti COg was
2.5x that induced under 645 ppin CO^
14% fewer flowers; 22% less shoot wt.
32% fewer flowers; 24% less shoot wt.
30% fewer flowers; 20% less shoot wt.
15% fewer flowers; 15% less shoot wt.


Caveat Reference
Adedipe
1972


Tingey
1973a





Rajput
1977
Short- term Heagle
growth response ston,
only
Short-term Heagle
response ston,
only
Hou et

Adedpie
1972


et aL,



et al . ,






et al.,

and John-
1979

and John-
1979

al., 1977

et al . ,

-------
                                                                         TABLE 8A-2  (continued)
00
Ul

Conc.a
ug/m3 (ppm)
660 (0.25)
790 (0.30)
920 (0.35)
1050 (0.40)
1310 (0.50)
1570 (0.60)
1050 (0.40)
2620 (1 00)
2620 (1.00)
Exposure3
time
4 hr, 3
times/wk 11 wk
5 hr/d
6 d/wk
12 d
26 d
1 hr
4 hr
6 hr
Z hr
3 hr
Exposure
condition Plant
EC/SO Soybean
EC/SD Barley
Bean
Sunflower
Barley
Bean
Sunflower
EC/SD Alfalfa
EC/SD Tomato
EC/SD Apples
GC Barley
GC Poinsettia 8 cv. 's
Effects
Foliage

X
X
X
X
X
X

~
X
X
X
onc
Yield
X
X
X
X
X
X
X
~




Species effect
No foliar injury; significant dec plant
ht. at 5,7,9,11 wks; significant dec.
shoot dry wt. at 7,11 wks; significant
dec. root dry wt. at 9,11 wks; signifi-
cant dec total dry wt at 11 wks
11% foliar injury; 38% dec. dry wt. shoot
<1% foliar injury; 25% dec. dry wt. shoot
5% foliar injury, 41% dec. dry wt. shoot
21% foliar injury; 25% dec. dry wt. shoot
2% foliar injury; 15% dec. dry wt. shoot
16% foliar injury, 29% dec dry wt. shoot
8% decrease in apparent photosynthesis
increased accumulation of total
and soluble S content
2% leaf injury
Threshold dose for foliar necrosis, 30-60%
decrease in net photosynthesis
Foliar injury 1 cultivar
Caveat6
Grown in silica
sand with
Hoagland's
solution as
nutrient source
Monitoring system
explained in
unavailable
publication
—
—
—
—
--
Reference
Re inert and
Weber, 1980
Markowski
et al., 1975
White et al.,
1974
Bennett and
Hill, 1973a
Kender and
Spierings, 1975
Bennett and Hill ,
1973a
Heggestad
et al., 1973

-------
TABLE 8A-2 (continued)
Conc.a Exposure3
ug/m3 (ppin) time
5240 (2.00) 3 hr
6550 (2.50) 6 hr
7860 (3.00) 1 hr
2 hr
3 hr
10480 (4.00) 2 hr
CD
I
co
en
1050 (0.40) 30, 78,
100 hr
1310 (0.50) 1.5 hr
BIO (0.50) 100 hr
Exposure
condition
GC
EC/SD
GO
GC
GC
Plant
Poinsettia 8 cv's
Apples
Poinsettia 8 cv's
Poinsettia 9 cv's
Poinsettia 8 cv's
EC Begonia
Petunia
Coleus
Snapdragon
Marigolds
Celosia, Salvia
Impatiens
EC/SD
EC/SO
EC/SD
Wheat 7 cv's
Soybean
Corn
Effects onc
Foliage Yield
X
X X
X
X
X
X X
X X
X
X X
X
X
X
X
X X
—
Species effect
Foliar injury 2 cultivars
increased foliar injury; 62% more leaf
abscission; 191 less shoot growth
Foliar injury, 5 cv's
Foliar injury, 7 cv's
Foliar injury, 8 cv's
27% fewer flowers; 33% less shoot wt. ;
severe necrosis
42% fewer flowers; 32% less shoot wt. ;
slight injury
30% fewer flowers; 22% less shoot wt. ;
no foliar injury
20% fewer flowers; 19% less shoot wt. ;
slight injury
Slight injury
Slight injury
Slight injury
No effect on yield, small amount
foliar injury
7% decrease in shoot fresh wt. ;
trace foliar injury
Minimal foliar injury, no effect
on dry mass
Caveat8 Reference
Heggestad et al . ,
1973
Kender and
Spier ings, 1975
Heggestad et
al,, 1973
Adedipe et al . ,
1972
Laurence, 1979
Short-term Heagle and John-
growth ston, 1979
response only
Laurence, 1979

-------
                                                                    TABLE  8A-2  (continued)
Cone. Exposure Exposure
um/m3 (ppm) time condition Plant
1570 (0.60) 6 hr EC/SO
1570 (0.60) 30, 78, EC/SD
100 hr
1970 (0.75) 3 hr EC/SD
2100 (0.80) 2 hr GC
Apples
Wheat 7 cv's-
Alfalfa
Alfalfa
Effects onc ri
Foliage Yield Species effect0
X 7.3% more foliage injury; 5% more
leaf abscission
X X Trend of decreased dry wt. in 17
of the 21 exposures, small amount
of foliar injury
X No injury developed
X Threshold dose for foliar necrosis;
25-50% decrease in net photosynthesis
aTab1e arranged by increasing SQ2 concentration as first order and exposure time as second order division. Doses within
specifically different effects are listed along with the lowest SO. concentration that induced said effect.
co GC = Growth chambers, EC = Exposure chambers, EC/SD = Exposure chamber, special design.
Caveat8 Reference
Kender and
Spierings, 1975
Trend not sig- Laurence, 1979
nif leant from
control
Hou et al . , 1977
Bennett and Hill,
1973b
a single study that did not induce
 X indicates study found foliage and/or yield effects.
  Host prominent or significant effect reported
 Caveats for consideration about proper study design and interpretation.
 The symbol ">" means greater than; "<" means less than.
Note:   1 ppm SOZ = 2620 ug/n3.

-------
TABLE 8A-3.  SUWIARY OF STUDIES REPORTING RESULTS  OF SO, EXPOSURE UNDER LABORATORY CONDITIONS FOR VARIOUS TREE SPECIES
Cone. a Exposure3 Exposure
pg/n3 (ppn) tine condition
05 (0.025) 6 hr EC/SD
130-390 6 hr EC/SO
(0,05-0.15)
130 (0.05) 16 wk EC/SD
Winter
130 (0.05) 10 wk EC/SD
260 (0.10)
520 (0.20)
oo 470-520 24 hr EC/SO
oa (0.18-0.20)
00
660 (0.25) 2 hr EC/SD
660 (0.25) 2 hr EC/SD
920 (0.35) 3 hr EC/SD
1180 (0.45) S htr EC/SD

Effects onc
Plant Foliage Yield
E. white pine
E. white pine
Beech
Norway
spruce
Jack pine
E, white pine
Jack pine
Red pine
Loblolly pine
Shortleaf pine
Slash pine
Virginia pine
Trembling aspen
E. white pine
X
X
—
X X
X X
X X
__
X
X
X
X
X
X
X
X
X
d e
Species effect Caveat
Threshold dose for needle damage to most
sensitive clones only
60% of tolerant clones developed foliar Sensitive clones
injury
Number of dead buds in spring increased at
0.10 ppra and higher; 50% more killed at
0.20 ppm.
No foliar effect; 25% less vol. growth (avg. ) 2 clones only
Foliar injury; 38% less vol. growth (avg.)
Foliar injury; 53% less vol. growth (avg.)
Foliar lipid synthesis inhibition
(reversible); increasing dose =
increasing recovery time
6.5X foliar injury Plants naintained
4,5% foliar injury in sensitive
0.5% foliar injury condition
All equally sensitive; most sensitive Plants maintained
period 8-10 weeks of age or older in sensitive
condition
2% foliar injury
All tolerant clones developed foliar Injury — •
Reference
Houston, 1974
Houston, 1974
Keller, 1978
Keller, 1980
Malhotra and
Kahn, 1978
Berry, 1971
Berry, 1974
Karnosky, 1976
Houston, 1974

-------
                                                                         TABLE  8A-3 (continued)
00
 i
Conc.a
MS/*3 (ppm)
1180 (0.45)

1310 (0.50)


1310 (0.50)
1310 (0.50)






1310 (0.50)



1310 (0.50)


1310 (0.50)


1700 (0.65)
Exposure9 Exposure
Effects onc


time condition Plant Foliage Yield Species effect" CaveatK
9 hr/d EC
for 8 wk
2 hr EC/SO

,
3 hr EC/SO-
5 hr GC






24 hr/d GC
up to 30 d


24 hr/d EC/SC
1/wk

4 hr/d EC/SD
3 wk

3 hr EC/SD
Ponderosa pine X

E. white pine X
Jack pine X
Red pine X
Trembling aspen X
Austrian pine X
Ponderosa pine
Scotch pine
Balsam, Fraser fir
White fir
Blue, white spruce
Douglas fir
Chinese elm X
Gingko X
Norway maple X
Pin oak X
Sugar maple
Black oak
White ash
Sugar maple
Black oak
White ash
Trembling aspen X
Severe needle tip chlorosis and necrosis

12% foliar injury
11% foliar injury
2X foliar Injury
11% foliar injury
No injury






7 days to chlorosis
14 days to chlorosis
12 days to chlorosis
30 days to chlorosis
54% lower rate of photosynthesis, no symp.
48% lower rate of photosynthesis, no symp.
20% lower rate of photosynthesis, no symp.
43% lower rate of photosynthesis, no symp.
74% lower rate of photosynthesis, no symp.
7% lower rate of photosynthesis, no symp.
23% foliar injury

Reference
Evans and
Miller, 1975
Berry, 1971


Karnosky, 1976
Smith and
Davis, 1978





Temple, 1972



Carlson, 1979


Carlson, 1979


Karnosky i 1976

-------
                                                                      TABLE 8A-3 (continued)
03
 I
Cone.3 Exposure3
jig/m3 (ppm) tine
2620 (1.00) 4 hr
2620 (1.00) 8 hr
2620 (1.00) 1 hr
3 hr
5 hr
5240 (2.00) 2 hr
5240 (2.00) . 6 hr
b Effects onc
condition Plant Foliage Yield
GC Austrian pine X
Ponderosa pine
Scotch pine
Balsam, Fraser fir
White fir*
Blue, white spruce
Douglas fir
EC American eltn
EC Scotch pine X
X
X
GC Austrian pine X
Ponderosa pine
Scotch pine
Balsam, Fraser fir
White fir
Blue, white spruce
Douglas fir
GC American elm X X
Species effectd Caveat6
Less than 4X foliar injury all species
Inhibition of stoaatal closing
No injury, primary needles; slight injury,
secondary needles
14% maximum injury primary needles;
52% maximum injury secondary needles
37% maximum injury primary needles;
60% maximum injury secondary needles
Ho foliar injury on Douglas fir, firs,
spruce
Pine foliar injury threshold, necrotic tips
Induced severe foliar injury; defoliation
in older leaves; significant reduced
expansion of new leaves; no. of emerging
leaves and root dry wt. reduced
Reference
Smith and
Davis, 1978
Noland and
Kozlowski, 1979
Smith and
Davis, 1977
Smith and
Davis, 1978
Constantinidou
and Kozlowski,
1979a

-------
                                                                    TABLE 8A-3  (continued)
Conc.a
(ig/w3 (ppni)
5240 (2.00)
5240 (2.00)
5240 (2.00)
5240 (2.00)
oo 7860 (3.00)
i
<£>
Exposure8
time
6 hr
6 hr
6.5 hr
12 hr
6 hr
Exposure
condition
GC
GC
EC
GC
GC
Effects onc
Plant Foliage Yield
American elm
Chinese elm X
Gingko
American elm
Gingko X
Norway maple
d e
Species effect Caveat
No significant reduction in lipid content;
significant less in new leaf protein content;
significant less leaf, stem root carbohydrate
content
100% leaf necrosis
Water-stressed plant increased uptake
of S02
Induced stomatal closing; S content in-
creased in plants fumigated in light
50% leaf necrosis
Reference
Constantinidou
Temple, 1972
No land and
Kozlowski, 1979
Temple, 1972
Temple, 1972
aTable arranged by increasing S02 concentration  as  first  order and  exposure time  as  second  order divisions.   Doses within a single study that did not induce
  specifically different effects are listed along with the  lowest S02 concentration  that induced said effects.
 GC = growth chambers,  EC = exposure chambers, EC/SD  = exposure chamber, special  design.
CX indicates study found foliar and/or yield effects.
 Host prominent or significant effect reported.
 Caveats for consideration about proper study design  and  interpretation.
 The symbol ">" means greater than; "<" means less  than.
Note:  1 ppm S02 - 2620 ug/m3.

-------
                                     TABLE 8A-4.   DOSE-RESPONSE INFORMATION SUWWRIZEO FROM LITERATURE PERTAINING TO NATIVE PLANTS
                                               AS RELATED TO FOLIAR, YIELD, AND SPECIFIC EFFECTS BY INCREASING S02 DOSE
Cone.3
pg/m3 (ppra)
63 (0.024)
Exposure3
time
85 d
Exposure
condition
EC/SD
Plant
S23 Ryegrass
Effects onc
Foliage Yield
X
Species effect
Plants at high nitrogen; S02 exposure
alleviated S deficiency of plants not
Caveat Reference
Cowling and
Lockyer, 1978
CO
                                                                                     provided S0< and increased yield by 27%.
                                                                                     No effect on plants grown with adequate
                                                                                     S04

                                                                                   Plants at low nitrogen: no effect of S04
                                                                                     or S02
50 (0.02) 29 d and EC/SD
22 d
370 (0.14) later
80 (0.03) 6 wk EC/SO
160 (0.06)
176 (0.067) 26 wk EC/SO
210 (0.08) 13 hr/d EC
" for
28 d
S23 Ryegrass X No significant effects at 0.02 or 0.14 at
first harvest (29 days);
X At 0.14, significant reduction (16%) in
specific leaf area at second harvest
(22 days later). No significant effect
on dry wt. , tillers, dark respiration
or transpiration coefficients.
Indian X Insignificant increase in productivity (19%);
ricegrass Insignificant increase in productivity (21%);
significant decrease in chlorophyll content
(43%)
S23 ryegrass X Significant increase in number (88%) and dry
wt. (78%) of dead leaves; significant de-
crease in number of tillers (41%); leaf
area (SIX), dry wt. of stubble (55%), and
number (45%) and dry wt. (51%) of living
leaves; significant decrease in yield (52%)
Wild ryegrass X None
Foxtail grass X X Foliar injury as caused by heavy metals was
increased by S02 exposure; yield not sig-
nificantly affected

—
Preliminary
study mean
weekly cone
wintertime
exposure
Cowling and
Koziol, 1978
Ferenbaugh, 1978
Bell and
Clough, 1973
No technical Krause and
S02 monitor- Kaiser, 1977
ing information

-------
                                                                  TABLE 8A-4 (continued)
Cone. Exposure3 Exposure
pg/m3 (ppm) time condition
290 (0.11) 4 wk EC/SD
290 (0.11) 4 wk EC/SC
290 (0.11) ,103.5 hr/wk EC/SD
(weekly for 20 wk
mean/76 pg/ro3
0.067)
oo 310 (0.12) 9 wk EC/SD
i
U3
CO
340 (0. 13) 6 wk EC/SO
Effects onc
Plant Foliage Yield Species effectd Caveat6
Cocksfoot X X 5% foliar necrosis; significant (30%) de- Wind tunnel
crease In leaf area, dry wt. (45%), til- exposures
lers, green leaves, and root/shoot ratios
Ryegrass X Significant (20%) decease in leaf area, Wind tunnel
dry wt. (40%), root/sboot ratio at a exposures
windspeed of 25« «in (.93mph) ,
No effect at a windspeed of 10m min
(.37mph)
Smooth- stalked X Significant decrease in leaf area (28%), all
ineadowgrass dry wt. fractions (44%), leaves (37%), and
tillers (27%)
S23 rytgrass X Significant decrease In dry wt. (46%) and
number (34%) of living leaves, tillers
(42%), leaf area (44%), dry wt. of stubble
(47%); significant increase in number (46%)
and dry wt. (46%) of dead leaves; significant
decrease in yield
Wild ryegrass X None
Indian ryegrass X X Necrotic foliar lesions noted; Inslgnifi- Wind tunnel
cant decrease 1n productivity (6%), signi- exposures
Reference
Ashenden, 1978
Ashenden and
Mansfield, 1977
Ashenden, 1978
Bell and
Clough, 1973
Ferenbaugh,
1978
660 (0.25)
1310 (0.50)
2620 (1.00)
  fleant decrease in chlorophyll content (51%)
Necrotic foliar lesions noted; significant
  decrease in productivity (35%) and chloro-
  phyll content (61X)
Plants mostly dead after 4 weeks
Plants dead after 4 weeks

-------
TABLE 8A-4 (continued)
Cone. a Exposure3
jig/a3 (ppm) tine
390 (0.15) 6 wks
790 (0.30)
1570 (0.60)
520 (0.20) 2 hr
660 (0.25) 5 wk
710 (0.27) 8 wk
00
i
UD
** 1310-Z8820 2 hr
(0.50-11.00)
1860 (0.71) 1 hr
2 hr
5 hr
D;oo-urcb Effects or|C
condition Plant Foliage Yield Species effect
Duckweed X Decrease in diameter of fronds, no dry wt.
effects
Decrease in diameter of fronds
Decrease in starch content, no dry wt. effects
no irreversible damages up to 0.60 pp« S02
GC Kentucky bluegrass X Visible foliar injury, high degree of
variation among 17 cv's
EC/SO Ryegrass X Significant decrease in yield (17%), no
effect on number of tillers
Significant decrease in green wt. (38%),
total dry wt. (30%), no reduction in number
of tillers, senescence rate doubled
F/CC 87 desert species X Host plants required more than 2.0 ppm
S02 to produce foliar injury
EC/SD Lily — Significant pollen tube elongation,
inhibition at 1 and 2 hours
Caveat Reference
Ambient air +
S02 exposure
system not
, described
No SO, monitor-
ing Information
for plants pre-
viously exposed
to S02
Wind tunnel
exposure
Field plants
watered heavily
and exposed, to
ambient air
before and^after
fumigations
Pollen on agar,
relationship
these effects
have to ambient
conditions is
unknown
Fankhauser
et al., 1976
Murray et al. ,
1975
Horsuan
et al , 1978
Hill et al.,
1974
Masaru et al . ,
1976

-------
                                                                          TABLE 8A-4  (continued)
a r b Effects onc

fig/n3 (ppm) time condition Plant Foliage Yield Species effect
2620 (1.0) 6 hr EC/SD Eucalyptus X 40% more foliar necrosis
of Australian trees and

Caveat6 Reference
, 32 of 131 species — 0' Conner
shrubs were et al , 1974
rated as sensitive to acute (>1 ppm)
exposure to SOg

      aTable arranged by increasing SOj concentration as"first order  and  exposure time as  second  order divisions.   Doses within a single study that did not  induce
        specifically different effects are listed along with the lowest S02 concentration  that induced.said effect.
       F/CC = field, closed chambers;  GC ~ growth chambers;  EC = exposure chambers; EC/SD  -  exposure  chamber,  special design
      CX indicates study found foliar  and/or yield effects,
      Ttest prominent or significant effect reported.
       Caveats for consideration about proper study design and interpretations.
       The symbol  ll>" means greater than;  "<"  means less than.
"j0    Note:   1 ppm SO, = 2620 ug/m3.
>£>                   '
tn

-------
TABLE 8A-5,  EFFECTS OF HIXTURES OF S02 AW 03 OH PUWTS
Cone.3
ug/R3 (ppa)
66 S02 + 100 03
(0.025 S02 +
0.05 03)




130 S02 + 100 03
(0.05 S02 +
0.05 03)

130 S02 + 100 03
(0.05 S02 +
0.05 03)
130 S02 + 100 03
(0.05 S02 +
0.05 03)
00
to
en

160 S02 + 100 03
(0.06 S02 *
0.05 03)

197-1570 S02 +
290 03
(0.075-0.60 S02
+ 0.15 Os)
Exposure3 Exposure
tine condition
6 hr EC/SO






8 hr/d, 5 d/wk EC/SD
5 wk


8 hr/d, 5 d/wk EC/SO
18 d

8 hr/d, 5 d/wk EC/SD
4 wk

8 hr/d, 5 d/wk
until control
plants 40-45 cm
high
7 hr/d F/CC
68 d


5 or 10 d EC/SD



Effects onc
Plant Foliage Productivity Species effect
E. white X X No effect on needle elongation.
pine Foliar injury on sensitive clones
only; 10 of 10 trees with 75-100X of
needles with tip necrosis.
S02 alone caused tip necrosis on 75-
100% of the needles on 1 tree.
03 alone caused no injury.
Radish X X Plant weight reductions additive (leaf
fresh and dry weight) or significantly
less than additive (plant fresh wt. ,
root fresh and dry weight).
Soybean X X Additive foliar injury effects;
greater-than-additive root dry
weight
Tobacco X Additive growth reductions

i
Alfalfa '. X Less-than-additive growth reductions



Alfalfa X Ho significant alteration of plant re-
sponses (carbohydrate, protein, dry wt.)
compared to effects of single pollutants

White bean X X Less-than-additive growth reductions and
foliar injury

Soybean X Less-than-additive foliar injury
Caveat6
Host 03 meters
facto ry-
cal f brated
by sodium
thiosulfate
method

_-



—


—


—



Potted plants
set on soil
surface; grown
hydroponically
—


—
Reference
Houston, 1974






Tingey et al.,
1971b


Tingey et al. ,
1973b

Tingey and
Reinert, 1975





Neely et al . ,
1977


Hofstra and
Ormrod, 1977

-------
                                                                         TABLE 8A-5.   (continued)
CO
 i
Cone.8 Exposure3 Exposure
pg/m3 (ppm) time condition
260 S02 + 200 03 6 hr/d F/CC
(0.10 S02 + 133 d
0.10 03)
260-1310 S02 + 4 hr EC/SD
100-200 03
(0.10-0.50 S02 +
0.05-0.10 03)







520 S02 + 100 03 3 hr EC/SO
(0.20 S02 +
0.05 03)
920 S02 + 100 03 3 hr EC/SD
(0 35 S02 +
0,05 03)
630 S02 -f 530 03 2 hr 6C
(0.24 S02 *
0.27 03)

730 S02 * 550 03 4 hr 6C
(0.28 SOE +
0.28 03)

Effects onc
Plant Foliage Productivity
Soybean X X


Alfalfa. X
Broccoli
Cabbage
Radish
Tomato
Tobacco Bel-W,
a




Trembling X
Aspen
(5 clones)
Trembling X
Aspen
(5 clones)
Tobacco X
Bel-W3
Bel-B
Consolidation 402
Tobacco X
Bel-Wa
Bel-B
Consolidation 402
d e
Species effect Caveat
S02 alone and In the mix did not
significantly affect the yield
and Injury responses
Greater-than-additive foliar injury at 0.10
ppm of each gas for alfalfa, broccoli, and
radish. Less-than-additive effect for tomato.
At 0.25 ppm, S02 * 0.10 ppm 03 greater-than-
additive injury noted on alfalfa, radish, and
tobacco. At 0.50 ppm S02 and 0.05 ppm 03,
greater-than-additive injury on broccoli
and tobacco and less-than-additive injury on
alfalfa. At 0.50 ppm S02 and 0.10 ppm 03f
greater-than-additive effects on alfalfa,
cabbage, radish, tobacco.
Greater-than- additive injury to 3 clones
no injury due to S02 alone

Greater-than-additive injury to 4 clones


9-3fi% foliar Injury — no injury due to either
pollutant singly


23-76% foliar Injury—no Injury due to either
pollutant singly


Reference
Heagle at al.,
1974

Tingey et al. ,
1973a









Karnosky, 1976


Karnosky, 1976


Menser and
Heggestad,
1966

denser and
Heggestad,
1966

-------
                                                                         TABLE BA-5.   (continued)
U3
co
Conc.a
pg/m3 (ppm)
660 S02 + 100 03
(0.25 S02 +
0.05 03)


2620 S02 + 200 03
(1.00 S0j> +
0.10 03)

660 S02 + 490 03
(0.025 S02 +
0.25 03)
1180 S02 + 290
or 880 03 (0.45
S02 + 0.15 or
0.45 03)
660 S02 + 270 03
(0.25 S02 +
0.1403)
660 S02 + 570 Oa
(0.25 S02 +
0.29 03)
Exposure Exposure
time condition
4 hr




4 hr/d, 3
11 wk
4 hr
10 6hr/d
68 hr 5
154
164
10 6hr/d
68 hr 5
154
164
EC/SO
All species
exposed to
both concentra-.
tions of each
pollutant.


d/wk EC/SD
EC/SO
EC/SD
d/wk
EC/SD
d/wk
Effects onc
Plant Foliage
Alfalfa X
Onion
Soybean
Tobacco Bel-B
Tobacco
White gold
Tobacco Bel-W3
Lima bean
Broccol i
Bromegrass
Cabbage
Radish
Spinach
Toia to
Soybean X
Radish
Scotch Pine X
Scotch Pine X
d &
Productivity Species effect Caveat
Only tobacco Bel-W3 showed greater-than- Bel-W3 tobacco
additive foliar injury at 0.25 S02 •*• very sensitive
0.05 03


At 1.00 ppm S02 + 0.10 03 tobacco Bel-B
and Bel-W3 exhibited greater-than-additlve
efffects, and there were less-than-additlve
effects for bromegrass, cabbage, spinach,
and tomato.

X Additive growth effects
X Additive growth effects
Less-than-additive effects— no effects
due to 0, alone
Less-than-additive effects— no effect due
to 0, alone
•5
Reference
Tingey et al.,
1973a '


Ibid.

Re inert and
Weber, 1980
Tingey and
Re inert, 1975
Nielsen et al.
1977
Nielsen et al.
1977

-------
                                                                         TABLE 8A-5    (continued)
co
Conc.a
(jg/in3 (ppm)
1310 S02 + 490 03
(0.50 S02 +
0.25 03)
Exposure Exposure
time condition
4 hr/d EC/SO
4 times,
6 d apart
Plant
Begonia
(5 cv's)
Effects onc
Foliage Productivity Species effect
X X Less-than-additive effects for
of one cv, 0.50 ppm S02 alone
reduced flower production in
foliar injury for one cv.
Caveat6
flower weight
significantly
the absence of
Reference
Reinert and
Nelson,
1980
     Table arranged by increasing S02 soncentration as  first order and  exposure time  as  second  order divisions.   Doses within a single study that did not induce
      specifically different effects are listed along with the  lowest SOj  concentration  that induced said effect.
     F/CC = field, closed chambers;  EC/SD = exposure chamber, special design; GC = growth  chamber.
    c X indicates study found foliar and/or yield effects.
     Most prominent or significant effect reported,
Caveats for consideration about proper study design and  interpretation.
   :   1 ppm SO- = 2620 ug/n3.
      1 ppm 0, = 1960 ug/m .
    Note*

-------
TABLE 8A-6.  EFFECTS OF HIXTURES OF S02 AHO JW2 OH PLANTS












CO
1
f"-*
o
o













Cone,3 Exposure3 Exposure
Mi/a ppn time condition
130-660 S02 + 4 hr EC/SD
90-470 N02
(0.05-0.25 S02 +
0.05-0.25 N02)
390-660 S02 + 4 hr EC/SO
190-380 N02
(0.15-0.25 S02 +
0.10-0.20 N02)




290 S02 T 210 NQ2 103.5 hr/wk EC/SD
(0.11 S02 + 20 wk
0.11 N02)
325-2620 S02 + 1 hr or 3 hr EC/SO .
235-1880 N02
(0.125-1,0 S02 +
0.125-1.0 N02)

390-1310 S02 + 1 hr and 2 hr EC/SD
280-750 N02
(0.15-0.50 S02 +
0.15-0.40 N02)

520 S02 + 190 or 6 d EC/SO
1880 N02
(0.2 S02 + 0.1 or
1.0 H02)
Effects onc
Plant Foliage
Tobacco X
Pinto Bean X
Tomato X
Radish X
Oats X
Soybean X
Pinto Bean X
Tomato X
Radish X
Oats X




Cocksfoot
Meadow-grass

Radish X
Swiss chard X
Oats X
Sweet pea X

Alfalfa




Pea



Yield Species effectd
0-2% foliar injury at 0.05 ppm S02+
0.05 ppm N02. 1-35% foliar injury
at 0.10 ppm S02 and 0.10 or 0.15
ppm N02. Injury less at 0.20-0.25
ppra S02 than at 0.10 ppm SOZ. Thres-
holds~S02=0.50 ppin, N0=2,00 ppra
No foliar injury







X fireater-than-additive decreases in
X no. of tillers, no. leaves, and leaf
area
Greater-than-additive foliar injury
to radish at higher concentrations.
Thresholds for radish were 0.50 ppm
of each gas, and for" the other species
0.75 ppm of each gas.
Slight but significantly greater-than-
additive depression of photosynthesis
at S02 concentrations of 0.15 and 0.25.
No greater-than-additive effects at
0.35 or 0.50 ppm S02.
Significantly greater than additive
increase in peroxidase and RuDPC
enzyme activity at 0.20 ppm S02 and
and 0.10 ppm H02
Caveat6

Reel rail ating air
Experimental condi-
tions approximated
those of Tingey et al.
1971b, but used dif-
ferent cultivars of
tomato, radish, and
oats
Wintertime exposures


Recircul ating air




Reversible effects




Reversible effect



Reference
Tingey et al.,
1971a

Bennett et al . ,
1975
>




Ashenden,
1979

Bennett et al . ,
1975



White et al.,
1974



Horsman and
Well burn,
1975

-------
                                                                         TABLE 8A-6.  (continued)
Cone. Exposure Exposure
jjg/m3 (ppn) time condition
1310-28820 S02 + 2 hr F/CC
190-940 N02
(0.5-11.0 SQj, f-
0,1-5.0 N02
2100 S02 + 560 N02 2 hr EC/SD
(0.8 S02 +0.3 N02)
Effects onc
Plant Foliage Yield Species effect*1
87 desert *X No evidence of S02 + NOZ synergism at
species an N02/S02 ratio of 0.28. Host,
species required over 5240 ug/m
(2,00 ppn) S02 to cause injury.
Alfalfa X Apparent ohotosynthesis reduced at
567 mg/ffl (315 ppn) C02 but increased
at 1161 rag/in (645 ppn) C02
Caveat6
Plants exposed to
ambient air before
and after fumigation
Reversible effects
Reference
Hill et al.,
1974
Hou et al.,
1977
 Table arranged by increasing S02 concentration as  first order and  exposure time as  second order divisions.   Doses within a single study that did not induce
 specifically different effects are listed along with the lowest S02  concentration that induced said effect,

 EC/SO = exposure chamber,  special design;  F/CC = field,  closed chambers.

CX indicates study found foliar and/or yield effects.

 Most prominent or significant effect reported.

eCaveats for consideration about proper study design and interpretation.
                            3
Note:  1 ppB SO, = 2620 ug/iC,
       I ppm NOJ = 1885 ug/« .
       1 ppm C02 = 1.8 mg/m .

-------
Page Intentionally Blank

-------
9. EFFECTS ON VISIBILITY AND CLIMATE

9.1 INTRODUCTION
Pollutants can change the way we see the world and the climate in which we live. Because
these changes may have profound consequences for .the Duality of life, the relationships among
air quality, visibility, and climate must be understood. Although some questions about these
relationships remain unanswered, many have been resolved. To provide a framework for under-
standing effects of pollutants on visibility, this chapter discusses factors that affect vis-
ibility, ways to measure it, historical trends, and methods to determine its value to us.
Climatic effects are given similar but briefer treatment, since theoretical understanding in
this area is far less advanced.
Many complex physical and chemical atmospheric processes affect our ability to see
distant objects or to distinguish nearby objects clearly. For example, small particles, which
are invisible individually, collectively reduce the visual range. We easily understand the
visibility reduction caused by common phenomena such as rain, dust, or snow, but atmospheric
constituents that we cannot perceive directly without instruments also interfere with visual
range. These constituents are known as fine PM and do not include S0?. But since S0? forms
particulate sulfate in the atmosphere, it affects visibility indirectly.
Through both light scattering and absorption, airborne particles reduce visibility and
thereby affect transportation safety and aesthetic vistas. While human observers are remark-
ably sensitive to contrast, they are not ideal detectors. If pollution sufficiently reduces
the contrast between an object and its background, the only way a person can distinguish be-
tween them is to move closer. For the operator of a high-speed vehicle, this loss of visibi-
lity may be fatal, given the limits of human reaction time.
Scientists have made progress in evaluating the optical changes and perceptual con-
sequences due to increased air pollution, but it has proved more difficult to evaluate the
effects of pollution upon the aesthetic appeal of the environment. Economists have measured
the loss of aesthetic appeal in market simulation studies with only limited success. While
precise measurement of this factor remains problematic, reduced aesthetic appeal unquestion-
ably carries significant social and economic cost. This is true For geographic areas of both
good and poor visibility. More work is needed in evaluating the consequences of visibility
reduction.
The effects of light scattering and absorption by PM on climate may be threefold. First,
reduction of solar radiation at ground level makes less energy available for photosynthesis
and commercial exploitation of solar energy. Second, reductions in solar radiation may lead
to alterations in local or regional temperatures. Third, increased cloud formation may alter
precipitation patterns. These effects, however, cannot yet be quantitatively related to
pollutant emissions or concentrations.

9-1

-------
General visibility conditions in the United States can be understood by examining
available regional airport visibility data. The isopleths in Figures 9-1 and 9-2 (Trijonis
and Shapland, 1979) show median yearly and summer visibility, a statistic that is relatively
insensitive to the site-specific availability of markers as long as the farthest markers con-
sistently reported lie beyond the median visibility. The data represent midday visual ranges
for 1974-1976 from 100 suburban/nonurban locations. Visibility at 93 of the locations was
determined from airport observations, usually made by the National Weather Service. Instru-
mental visibility measurements from seven sites in the Southwest are also included. Although
some uncertainties arise from the use of airport visibility observations, there is reasonably
good consistency among airport observations within regions and between airport and instrumen-
tal results in the Southwest.
The best visibility (110+ kilometers, 70+ miles) occurs in the mountainous Southwest.
Visibility is also quite good (70 to 110 kilometers, 45 to 70 miles) north and south of that
region, but sharp gradients occur to the east and west. Most of the area east of the
Mississippi and south of the Great Lakes has a median visibility of less than 24 kilometers
(15 miles).
9.2 FUNDAMENTALS OF ATMOSPHERIC VISIBILITY
Figure 9-3(A) shows the simple case of a beam of light (e.g., from the sun or a search-
light) transmitted horizontally through the atmosphere. The intensity of the beam in the
direction of the observer, I, decreases with distance from the source as light is absorbed or
scattered out of the beam. Over a short interval, this decrease is proportional to the length
of the interval and to the intensity of the beam at that point:

-dl = a I dx (9-1)
where -dl is the decrease in intensity, cr . is the extinction coefficient (the propor-
tionality constant), I is the intensity of the beam at the beginning of the interval, and dx
is the interval length. The extinction coefficient has units of inverse length. The extinc-
tion coefficient is determined by the scattering and absorption of light by particles and
gases and varies with particle and gas concentration, particle size distribution and composi-
tion, and wavelength' of light.
Consider an observer looking at a distant target in the daytime (Figure 9-3(B)). Just as
a light beam is attenuated by the atmosphere, the light from the target reaching the observer
is diminished by absorption and scattering. In addition, the observer receives extraneous
light (often called air light) scattered into the line of sight by the intervening atmosphere.
The net effect, as Figure 9-4 shows, is that a target darker than the horizon appears brighter
than it actually is and a target brighter than the horizon appears darker than it actually is.
9-2

-------
          15
           25 45
P  Based on photographic
   photometry data
IM  Based on nephetometry data
*  Based on uncertain extrapolation of
   visibility frequency distribution
          Figure 9-1.  Map shows median yearly visual range (miles) and
          isopleths for suburban/nonurban areas, 1974-76.

          Source;  Trijonis and Shapland  (1979).
P- Based on photographic
   photometry data
N Based on nepholometry data

*  Based on uncertain extrapolation of
   visibility frequency distribution
                                                                        101
          Figure 9-2. Median summer visual range (miles) and isopleths for
          suburban/nonurban areas, 1974-76.

          Source: Trijonis and Shapland (1979).
                                        9-3

-------
      G
Rgure 9-3. (A) A schematic representation of atmospheric extinction,
illustrates (i) transmitted, (ii) scattered, and (iii) absorbed light. (B) A
schematic representation of daytime visibility illustrates: (i) light from
target reaching observer, (ii) light from target scattered out of
observer's line of sight, (iii) air light from intervening atmosphere, and
(iv) air light constituting horizon sky. (For simplicity, diffuse illumina-
tion from sky and surface is not shown). The extinction of transmitted
light attenuates  the "signal" from the target at the same time as the
scattering of air light is increasing the background "noise."

Source: Adapted from U.S. Environmental Protection Agency (1979).
                              9-4

-------
X
2
cc
OQ
                  BRIGHT OBJECT
        LIGHT INTENSITY OF HORIZON
                        BLACK OBJECT
                     OBJECT-OBSERVER DISTANCE


Figure 9-4. The apparent contrast between object and horizon sky
decreases with increasing distance from the target. This is true for
both bright and dark objects.

Source: Middleton, (1952),
                                 9-5

-------
At  increasing distances, the  apparent brightness  of  dark and  bright  targets  approaches the
horizon brightness.  At sufficient distance, the target and horizon are so close in brightness
that the target is indistinguishable; for a black target this distance is the visual range, or
visibility.
     The example  just  discussed can be restated  in terms  of a target's contrast, defined as:

               C    =    Ltarget  _ I                                      (9-2)
                          background
          where     L is luminance  (brightness).
A  target's  apparent contrast  begins  with  some  intrinsic  value at the  target  and approaches
zero  as  distance  from  the  target  increases.    When the  observer's  contrast  threshold  is
reached, the  target  cannot  be distinguished and  for black targets the visual range is known.
     In  a uniform  atmosphere,  the  apparent contrast  between  a  target  and the  horizon sky
decreases exponentially with observer-target distance x (Middleton, 1952):

             C = CQ exp(-o-ext x)                                           (9-3)

Where  C«  is   the  intrinsic  contrast at  x = 0.   The maximum distance at which a given large
target can be distinguished from the horizon sky is, therefore, inversely proportional to the
extinction coefficient:
                       1°ge   CQ  - loge £
               X,
                max ~                                                           (9-4)
                              ext
where e is the observer's contrast threshold.
     The numerator in Equation 9-4 depends on the target's intrinsic contrast with the horizon
and on  the  observer's  contrast threshold.  For a black target, CQ = -1, so that loge  CQ  = 0
and the numerator reduces  to  -log  s.   The visual  range  (V)  of a black target  in  a uniform
atmosphere is thus given by the Koschmieder (1924) formula:
                       -log   £         K
                 V  =  - § -   =  -JS -                               (9-5)
                                       CT
                                        ext
     Although  Equation 9-5  is  based  on  use of  a black target,  little error  is  introduced
into the  determination of  visual  range  by  using as  targets  such nonblack  features  as  dark
forests and  deep  shadows.   If a target's  intrinsic  brightness  is as much as 30 percent that
of the  horizon  sky,  then the limiting  distance  given by Equation 9-4 is within 10 percent of
the visual range.  On  the other hand, the intrinsic brightness of artificial  lights or sunlit
objects can  approach that of the horizon  sky,  in which case their use as targets can lead to

                                            9-6

-------
large underestimates of visual range. It must also be noted that nonblack targets can lead to
serious underestimates of visual range if a target at a distance significantly less than the
visual range is used and visual range is calculated or estimated from the apparent contrast of
the target. For example, for an intrinsic target brightness of 30 percent of the horizon
brightness, an error of 18 percent results from using a target at half the true visual range;
an error of 91 percent results if the target is at one-tenth the true visual range.
The observer's contrast threshold is not a constant (Figure 9-5), but varies with the
individual observer (Middleton, 1952); apparent target size (Blackwell, 1946); spatial fre-
quency (Campbell and Maffei, 1974); brightness adaptation of the eye (Blackwell, 1946);
whether the target is stationary or moving; viewing time; background uniformity and complex-
ity; and target shape (Taylor, 1964). Neither is the contrast threshold a sharp boundary
between detection and nondetection; rather, it is usually defined as the value of contrast
that results in a 50-percent detection probability. Applications that require high detection
probabilities (e.g., aircraft operation) often double the value (Taylor, 1964), The value of
0.02 has been widely used for many years, beginning with Koschmieder (Middleton, 1952). It is
clear from the above considerations, however, that no fixed value of contrast threshold is
generally applicable. Fortunately, as can be seen in Equation 9-4, uncertainties in e and C«
are reduced by the logarithmic function; thus, the relationship of visual range to a , is not
SXlu
directly proportional to variations of £ and €„.
Surprisingly, a target's chromaticity (hue and saturation) has only a marginal effect on
changing an observer's contrast threshold and thus the visual range (Middleton, 1950). This
is because the apparent color of a target near the visual range is gray, due to scattering out
of the beam of colored target light and scattering into the beam of essentially white air
light. Furthermore, the detection threshold of chromaticity difference increases greatly for
targets of apparent angular diameter less than half a degree, angles that most brightly
painted objects (e.g., radio towers) subtend. Thus, at the visual range colored targets
appear about the same as gray targets of the same luminance. This is not to say that target
chromaticity is unimportant in judging visibility degradation by observing targets closer than
the visual range. To the contrary, washed-out colors and details of relatively nearby objects
are probably the most noticeable facet of visibility degradation to observers of scenic vistas
(U.S. Environmental Protection Agency, 1979).
The most direct method of determining the proportionality constant, K, involves instru-
mental^ measuring the apparent contrast of a target that is marginally perceptible to an
unaided observer. This method has been used both in laboratory determinations (Blackwell,
1946) and in atmospheric studies (Middleton, 1952).
Another way to determine K is to measure the visual range and its associated a . and
9-7

-------
1
I
DC
Ui
i
5
5
K
(9

<
te
CD
OC
   1000
    100
     10
                      I     I     I
      0,001       0.01        0.1        1        10

                           CONTRAST THRESHOLD
                                                       100
1000
        Figure 9-5. Mean contrast threshold of the human eye for 50%
        detection probability as a function of target angular diameter and
        adaptation brightness (candles/m2) for targets brighter than their
        background. Daytime  adaptation  brightness is usually in the
        range of 100—10,000 candles/m2.

        Source:  Middleton (1952).
                                      9-8

-------
take their product. It is imperative that: (1) the intrinsic contrast of each target be
known and corrected for; (2) targets be sufficiently numerous and well-spaced so as to give
reasonable resolution of the visual range; (3) total extinction be measured; (4) the ex-
tinction measurement be representative of that along the sight paths to the targets; and (5)
aerosol sampling for the extinction measurement be conducted carefully and detailed measure-
ment conditions be recorded. Observations at small solar scattering angles should probably be
excluded unless the method for determining a . responds to the increased solar scattering as
GXt
well.
A third approach for obtaining K, multiplying visual range by the fine mass concentration
(FMC) and an assumed aext/FMC ratio, is much less desirable due to the additional uncertainty
of aeX4/FMC ratio discussed in Section 9.2.3.3.
Several studies have s"hown substantial deviation from the historical K value of 3.9,
which results from assuming large perfectly black targets and a contrast threshold of 0.02.
Studies by Horvath and Noll (1969), Samuels et al. (1973), and Griffing (1980) can be criti-
cized because they assumed absorption by particles was negligible. It is now known that par-
ticle absorption is rarely negligible, thus their estimates of K may be low by 10 to 50 per-
cent. Griffing's work can also be criticized because of the poor quality of markers available
at the airport he studied. The only study that reasonably adheres to the recommendations of
the earlier paragraph is that of Ferman et al. (1981), which reports a value of 3.5 for well-
mixed periods. This value may be corrected to 3.7 to account for the imperfect blackness of
the tree-covered mountain targets, assuming the trees to be 20 percent as bright as the
horizon sky (Middleton, 1952).
It is to be emphasized that no fixed value of K is generally applicable. As Middleton
(1952) reviewed, contrast thresholds from 0.0077 to 0.06 were measured over the years, corre-
sponding to K ranging from 4.9 to 2.8, respectively. Given the observed range of values, it
seems reasonable to retain the historical choice of K = 3.9 when a value must be chosen for
illustration. Users of "visual range" data derived from non-black targets must remember that
they do not have true visual ranges until they correct the data for non-zero intrinsic target
brightness.
The Koschmieder formula's neglect of both pollution gradients and the earth's curvature
and topography limit its applicability near sources of primary PM and in very clean air. The
formula also assumes equal illumination of all parts of the atmosphere in the horizontal
plane, which limits its usefulness under partly cloudy skies. Where visibility is restricted
by diffuse haze, however, Koschmieder1s formula performs well. Comparisons of daytime visual
range, measured by a human observer, and extinction from scattering, measured instrumentally
at a single point, show visual range to correlate with the reciprocal of extinction, as illu-
strated in Figure 9-6 (Horvath and Noll, 1969; Samuels et al. , 1973). The correlation co-
9-9

-------
     0.50

     0.40


     0.30
  *

  1
   _- 0.20
  UJ
  o
  u.
  Ul
  8 0.10
  O
   < 0.05
     0.03
            ii  i  i 11
                                            i  i  i  i
10       20

    VISUAL RANGE, km
                                           50
                                                    100
Figure 9-6. Inverse proportionality between visual range (V)
and the scattering coefficient )osp) as measured at the point
of observation. The straight line is derived from the Koseh-
mieder formula for visual range, assuming V -  3.9/osp and
nonabsorbing media (%xt = °sp)- The correlation coefficient
for V and oSn is —0.89.

Source:  Horvath and Noll (1969).
                              9-10

-------
efficients are commonly in the neighborhood of 0.9, which is quite good considering that the
point measurement of extinction is being extrapolated along a sight path several tens of kilo-
meters long. In summary, visibility is inversely proportional to the atmospheric extinction
coefficient.
9.2.1 Physics of Light Extinction
Since light is electromagnetic radiation, it is altered by the interaction of its elec-
tric and magnetic fields with all matter through which and near which it passes. It may be
altered by scat ten' ng (redirection) or by absorption. Two areas of theoretical investigation
have led to a thorough understanding of the physics of light extinction in the atmosphere.
The first, derived by Lord Rayleigh in the late 1800' s, pertains to the scattering of light by
the gaseous molecules comprising the atmosphere (Middleton, 1952; Kerker, 1969). Rayleigh
scattering is directly proportional to the molecular number density and thus decreases as
elevation increases. It is inversely proportional to the fourth power of wavelength (color);
thus, blue light is scattered by air molecules about five times more than is red light.
Rayleigh scattering is of limited interest in visibility work because it represents the
cleanest possible condition of the atmosphere. Almost all cases of visibility impairment are
caused by the presence of particles, the sole exception being discoloration caused by the
pollutant gas N0».
The second theory pertains to the extinction of light by homogeneous spheres of arbitrary
size and was advanced by several workers from 1890 to 1910; the name of Mie (1908) is most
commonly associated with this theory (Kerker, 1969; Middleton, 1952). This approach, based on
electromagnetic theory, is necessary because resonances occur in scattering by particles of
diameter within about an order of magnitude of the wavelength of the radiation. For visible
light, the size range requiring use of Mie theory extends from 0.05 (jm to 10 urn (and larger if
the particles are spherical), the sizes at which most long-lived atmospheric particles accumu-
late. Mie theory allows computation of the scattering in any or all directions and the
absorption of light by a single spherical particle. The wavelength of light, particle size,
and complex refractive index must be specified. For absorbing particles, the imaginary part
of the refractive index is nonzero. For a monodisperse aerosol of number concentration N, the
extinction coefficient becomes:
where r is the particle radius and Q . is the extinction efficiency factor calculated from
Mie theory. Examples of the behavior of Q . are shown in Figures 9-7 and 9-8. Note that
GX T»
after its large rise at a few tenths of a micrometer, Q , approaches 2 and the extinction by
a single particle becomes proportional to r2.
9-11

-------
3
s
a
2
II 1 I I I i HI I I I i III l| I I I I I III
HI
/
/
r i i i 11111-L^ i i i i 11111 » i i i i M IT
10'2 10'1 1 10

PARTICLE DIAMETER

Figure 9-7. Extinction efficiency factor (Qext) of a single
spherical particle as a function of diameter for a non-
absorbing particle of refractive index (1.5—O.Oi) and
wavelength 0.55 jum.
_ I I I I I Ill| I I I i i 111|
i r ifrm
X
0°

/
t—T"i I I I 111 i I I I I I 111 I III I | IT
i
-------
For polydisperse aerosols, integration over the range of sizes is necessary:

CText = "/ n(r) Vt(r) r" dr (9-7)

If the particles are of differing composition (and thus differing complex refractive index),
the number size distribution of each species must be known. Extinction coefficients for the
various species are then summed. The precise calculation of extinction by spherical particles
allowed by Mie theory demands detailed knowledge of the composition and size distribution of
the particles.
Because of the wavelength dependence of extinction, particles may cause discoloration of
scenes. The type of discoloration (e.g., a red hue or a blue hue) depends on the scattering
angle (i.e., the angle between lines of illumination and observation), the particle number
distribution, the particle refractive index distribution (i.e. composition as a function of
size), and the background sky and objects. The wavelength (A) dependence of scattering is
given by A a, where a ranges from 4 for extremely small particles (i.e., the Rayleigh limit),
to zero for very large particles (e.g., 10-pm diameter cloud droplets); for intermediate sizes
where most long-lived atmospheric aerosol particles accumulate, a ranges from 0.5 to 2.5, with
1.5 being typical (Charlson et al., 1972). The wavelength dependence of absorption is usually
found to approximate A for atmospheric aerosol, indicative of fine particles having a small
dispersion in the complex refractive index over the visible wavelengths (Weiss et al., 1978).
For particles larger than a few micrometers, however, absorption becomes wavelength
independent.
The intensity of scattering varies with the scattering angle (i.e., the angle between
initial and scattered directions). Particles of diameter much larger than the wavelength
scatter almost all light into the forward direction. At particle diameters comparable to
visible wavelengths (diameters at which particles of the most interest accumulate), a small
but significant fraction of light is scattered into wider angles. The exact values can be
calculated from Mie theory for a given size distribution, refractive index, and wavelength.
The practical significance of this angular dependence of scattering is obvious when viewing
objects toward the sun: contrast, colors, and details can appear extremely degraded compared
to objects in the opposite direction; plumes that appeared colored from one angle may appear
gray or white against the sun, their color washed out by the addition of large amounts of
essentially white light.
Mie theory is strictly applicable only to spherical particles. Fortunately, a majority
of scattering particles in the optically important size range of 0.1- to 2-um diameter are
spherical (many being droplets), or nearly spherical (Allen et al., 1979; Pueschel and
Wollman, 1978; Pueschel and Allee, 1980). Electromagnetic solutions have been achieved for

9-13
-------
several nonspherical shapes, such as spheroids (Latimer, 1980); disks (Weil and Chu, 1980);
cylinders; and for the important case of layered or coated spheres (Kerker, 1969; Sehuerman,
1980; Pinnick et a!., 1976; Fowler and Sung, 1979; Mugnai and Wiscombe, 1980). These exact
solutions can be applied if circumstances warrant. Absorbing.-particles are usually distinctly
nonspherical. Many are chain-aggregates such as flame soot. Janzen (1980) empirically showed
that Hie theory predicts measured absorption quite well for carbon black particles by assuming
the aggregates to be spheres of equal volume.
Multiple scattering of photons (i.e., more than one scattering event en route from object
to observer) is negligible for our application of Mie theory because of the relatively small
particle number concentrations and extinction cross sections involved in even the worst cases
of anthropogenic visibility impairment.
Light extinction by particles is determined by the number, size, shape, and type of
particles in a given volume and does not vary with fluctuations or differences of ambient
temperature or pressure It is recommended that mass concentration data collected at non-
standard temperature or pressure not be corrected to standard or normal conditions (as is
sometimes routinely done with air pollution data) for comparison with extinction values. Many
techniques for measuring size distribution (e.g., impaction, laser Doppler velocimetry)
measure aerodynamic diameter, not geometric diameter that is needed for Mie calculations.
These diameters are related by the square root of density for a given spherical particle, but
the usually unknown variation of density may cause errors of 20 percent or more in the
geometric size needed for Mie calculations.
Despite the relatively advanced state of knowledge of the physics of light extinction,
there are two remaining difficulties in visibility modeling;
1. How to model the transport and transformation of pollutants and precursors from
sources so that the concentration, size, and type of particles along a given sight
path can be predicted,
2. How to model the perception and evaluation of a scene by an observer.
The former problem is of considerable interest for reasons other than visibility and is the
subject of considerable research (see chapter 6). The latter problem encompasses the
disciplines of physiology, psychology, and economics, and requires an empirical approach.
9.2.2 Measurement Methods
The term "visibility" is often used to mean visual air quality, which takes into account
not only how far one can see, but also how well one can see nearby objects (i.e., their appar-
ent contrast or discoloration). These multiple meanings have confounded the choice of a
particular method for measuring visual air quality.
Human observer methods have been widely used, and their qualitative judgments are the
ultimate reference for any evaluation of visual air quality. Nonobserver or instrumental
9-14
-------
methods of visibility measurement, however, are very important in understanding visibility
impairment for a variety of reasons:
1. They may measure separate components of extinction (e.g., only scattering or only
absorption);
2. They may allow elimination of undesirable effects (e.g., increased contrast loss
because of looking toward the sun);
3. They may allow modification of an air sample to check for characteristic responses
of particular pollutants (e.g., humidifying a nephelometer sample to check for
deliquescence);
4. They may allow statistical analysis with particle composition data collected at a
site.
In many cases it is not the method that is deficient, but its application to the question at
hand. If reasonable correspondence can be demonstrated between certain instrumental methods
and human perception, it will probably be desirable to choose instruments for routine visi-
bility data collection, for we would then be inferring an average perception based upon the
accepted statistical link rather than the specific perception of a particular observer. In
addition, the response of observer methods, photography, and telephotometry to nonpollutant
conditions such as precipitation, snow on the ground, clouds, and the solar angle may limit
their use.
In general, correspondence between human observation and instrumental methods is good.
For specific comparisons, poor correlation is usually explainable as being due to humidity
effects or poor instrument techniques or application. Investigators have usually found
contrast methods (e.g., telephotometry and photography) to agree best with human observation
(Allard and Tombach, 1980; Cwalinski et a!., 1975). This finding is not surprising, since the
eye-brain system itself relies on contrast. The consistently superior correlation of contrast
methods with human observation is probably because of their common response to nonpollutant
conditions such as clouds and the solar angle, as noted above. These studies and another
(Horvath and Noll, 1969) have also shown that at RH below 60 to 70 percent, nephelometry
compares well with human observation. The fact that these investigators had to exclude high
humidity cases to achieve good agreement indicates that nephelometry is not inherently at
fault; the problem arises from failure to prevent heating or cooling of the aerosol during
aspiration.
Two other studies, one experimental (Hoffmann and Kuehnemann, 1979) and one theoretical
(Harrison, 1979), have shown that the nephelometer provides reasonable estimates of visual
range, at least for turbid and polluted atmospheres. Harrison (1979) calculated that in clear
atmospheres, nephelometers may underestimate visual range by as much as 35 percent. There
have been too few careful studies comparing transmissometry results and human observations to
establish a credible relationship.

9-15
-------
The extinction coefficient introduced in Section 9.2 is a summation of contributions froir
scattering and absorption by gases and by particles:

-------
the human eye); otherwise, the rendition may not be true, and both densitometry and
qualitative applications may be seriously affected.
9.2.2.3 Telephotometry (Total Extinction)--A telephotometer is a telescope that can measure
the apparent brightness o,f a faraway object (Middleton, 1952; Ellestad and Speer, 1981). By
measuring the brightness of an object' at distance x and the horizon sky around it, the
object's apparent contrast can be computed. This number may be used directly or, if
Koschmieder's restrictions are assumed (Middleton, 1952), may be converted to a t by the
equation:

CText = ^
Telephotometry is attractive for several reasons. It is a path measurement; thus, atmospheric
nonuniformities are averaged. The instrument's absolute calibration is unimportant; only its
linearity and short-term stability matter. It requires no sample aspiration, and therefore
avoids large particle losses and sample heating or cooling. Finally, it is perhaps the
closest instrumental approximation to human observation. The method has limitations when the
target's intrinsic contrast is unknown or assumed (a very common situation), when measuring
dark objects at close range (due to internal stray light errors), and when clouds cause uneven
illumination. The cloud problem makes telephotometry difficult to automate and thus the
method remains labor intensive.
9.2.2.4 Long-path Extinction (Total Extinction)—The most direct way to measure o . is to
measure the decrease in intensity of a.light beam over a known distance x,

CText = ^ lQ9e f (9-U>

where I and I are the final and initial intensities, respectively. The method is appealing
in that no assumptions are involved, it measures average extinction over the path, and it re-
™"|
quires no sample aspiration. Unfortunately, even for values of a . about 0.2 km (typical
in the eastern United States), the decrease over short paths (1 meter) is small (0.02 percent)
and cannot be measured accurately. An alternative is to increase the path length, but source
intensity fluctuation, mirror reflectivity changes for single-ended systems, detector sensi-
tivity drift, alinement, thermally induced scintillation, and the large background light of
daytime again make the measurement difficult.
Hall and Riley (1976) have measured extinction by observing an uncoilimated source at two
ranges. Any decrease in intensity with increasing range in excess of the inverse-square
decrease is due to extinction. This method avoids any need for absolute calibration, since
only the ratio of intensities at two ranges need be measured. This method as used by Hall and
Riley is labor intensive, but has been demonstrated in clean and urban environments during
night operation.

9-17
-------
9.2.2.5 Nephelometer (Scattering)—The integrating nephelometer measures only the scattering
coefficient of an aerosol (Beuttell and Brewer, 1949; Crosby and Koerber, 1963; Ruppersberg,
1964; and Charlson et al., 1967). By simple adjustment, Rayleigh scatter, cr , can be
excluded or included. The instrument consists of an enclosed volume painted black, a
sensitive light detector looking through the volume, and a light source at one side of the
volume. The only light reaching the detector is that scattered by gas molecules and particles
within the volume. The nephelometer is sensitive, easily calibrated and automated, enables
one to modify the sample if desired, and provides a point measurement for correlation analysis
with point measurements of mass concentration and chemical composition. Errors of application
of the nephelometer include siting where the aerosol is nonuniform (e.g., near aerosol sources
or in mountainous terrain) or siting where significant absorption may occur (e.g. urban
sites), unless a concurrent measurement of absorption is made. The operator must be careful
to avoid inadvertent heating or cooling of the sample and resulting modification of the
aerosol size distribution. The nephelometer has two sources of inherent errors: (1) angular
truncation (Ensor and Waggoner, 1970; Rabinoff and Herman, 1973), which results in underesti-
mation of scattering, especially when large particles are present; and (2) sample aspiration
(Heintzenberg and Quenzel, 1973), which results in the loss of large particles through
impaction on the ductwork. These inherent errors may result in depressed correlations between
scattering and total mass concentration when significant large particle concentrations occur.
Despite these limitations, nephelometry remains one of the most widely used visibility
measurement methods. Most of the data relating cause and effect (i.e., particle concentration
or composition and optical effect) have been acquired with the nephelometer. Agreement
between the nephelometer and the long-path transmission method has been demonstrated in
several cases (e.g., Waggoner and Charlson, 1976; Weiss et al., 1979).
9.2.2.6 light Absorption Coefficient—Were it not for the fact that elemental carbon (soot,
graphitic C, free C) is a prominent species in cities and industrial regions, 0 would be in-
consequential. Even a few percent of the submicrometer mass as soot, however, produces a
significant effect on a or a .. The methods that have so far been used to evaluate cr in-
ap GXT- «p
elude:
1. Determining the difference between cr . and a by using long-path transmissometry
and nephelometry (Weiss et al., 1979); p
®
2. Determining change of transmission of Nuclepore filters with scattered light
collected by an integrating plate of opal glass (Lin et al, 1973; Weiss et al.,
1979);
3. Determining change of transmission of Millipore® filters (Rosen et al., 1980);
9-18
-------
4, Determining the reflectivity of a white powder with aerosol mixed Into it, called
the Kubelka-Monk method (Lindberg and Laude, 1974);
5. Determining absorption of light by a sample of particles inside a white sphere (In-
tegrating sphere) (Elterman, 1970);
6. Estimating an imaginary refractive index from regular scattering or polarization and
size distribution (Eiden, 1971; Grams et a!., 1974);
7. Measuring the amount of graphitic C and its size distribution and then calculating
V
8. Detecting the acoustical pulse produced when energy is absorbed by particles as
light and is transformed to heat (spectrophone) (Truex and Anderson, 1979).
As yet, no single method is widely accepted, although filter methods (e.g., 2 and 3,
above) are simple, inexpensive, and show promise. More work is needed to learn how the
various filter methods relate to each other and, more importantly, how accurately each
@
measures the in situ a . The Nuclepore method has the distinct theoretical advantage of
— ap
collecting the particles in a monolayer and on an optically simple filter. This method may
overestimate a somewhat, however, due to backscatter losses and orientation of common
ap
chain-aggregates normal to the light beam, which maximizes their absorption cross section.
Filter methods employing depth filters produce greater overestimates of o , probably due to
ap
multiple scattering and the resulting multiple chances for each absorption site to attenuate
the light beam. For example, Sadler et al. (1981) report that a from quartz fiber and
P 8
membrane filters averages about a factor of 2.8 higher than values from Nuclepore filters.
Recent work on the BS method indicates that, as many researchers have speculated, this
method gives a rather poor indication of total or fine mass concentration and instead responds
principally to absorbing particles (Edwards, 1980; Pashel and Egner, 1981). (For further dis-
cussion see Chapter 3.) Edwards' work further demonstrates empirical conversions from BS and
coefficient of haze (CoH) data to 0 . If substantiated, these conversions will allow trend
ap
analysis of a , because BS or CoH data have been routinely monitored at many sites for
ap
several previous decades.
9-2.3 Role of ParticulateMatter in Visibility Impairment
As noted earlier, the extinction coefficient comprises contributions from gas and
particle scattering and absorption:
a . = a + a + a + a (9-12)
ext sg ag sp ap
This section discusses the relative magnitudes of these contributions.
9.2.3.1 Rayleigh Scattering--A particle-free atmosphere at sea level has an extinction
—1 *
coefficient of about 0.012 km for green light (wavelength 0.55 urn) (Penndorf, 1957),
limiting visual range to about 325 kilometers. The coefficient a decreases with altitude.

9-19
-------
In some areas of the Western United States, the extinction of the atmosphere is at times
essentially that of a particle-free atmosphere (Charlson et al., 1978a; Porch et al., 1970).
Rayleigh scattering thus amounts to a simply definable and measurable background level of
extinction with which other extinction components (such as those caused by manmade pollutants
or by natural sources of particles) can be compared. At 40-kilometer visual range, a better-
than-average value for the Eastern United States, Rayleigh scattering contributes only about
one-eighth of the total extinction (Tri-jonis and Shapland, 1979). Rayleigh scattering
decreases with the fourth power of wavelength, and contributes a strongly wavelength-dependent
component to extinction. When Rayleigh scattering dominates, dark objects viewed at distances
over several kilometers appear behind a blue haze of scattered light, and bright objects on
»
the horizon (such as snow, clouds, or the sun) appear reddened at distances greater than about
30 kilometers. Scattering by gaseous pollutant molecules is negligible because of their low
concentrations compared to N« and 0,,; thus, variations in pollutant gas concentrations have no
effect on Rayleigh scattering.
9.2.3.2 Nitrogen Dioxide Absorption—Of all common gaseous air pollutants, only NO,, has a
significant absorption band in the visible spectrum. Nitrogen dioxide strongly absorbs blue
light and can color plumes or urban atmospheres red, brown, or yellow if significant concen-
trations and path lengths are involved. The effects of NO,, on visibility are discussed more
fully in the NO criteria document and by Hodkinson (1966), White and Patterson (1981), and
)\
Charlson et al. (1972). Its contribution to total extinction is in general minor. For
example, in Denver, Colorado during November and December 1978, N02 levels contributed an
average of 6 percent of total light extinction (Groblicki et al., 1981).
9.2.3.3 Particle Scattering--In general, scattering by particles accounts for 50 to 95 per-
cent of extinction, 'depending on location, with urban sites in the 50- to 80-percent range and
nonurban sites in the 80- to 95-percent range. The measurements of Waggoner et al. (1981),
made with an integrating nephelometer and the integrating plate method, show that in urban-
industrial areas particle scattering accounts for 50 to 65 percent of extinction, in urban
residential areas 70 to 85 percent, and in remote areas 90 to 95 percent. A comparison re-
ported by Weiss et al. (1979), which employed a nephelometer and a long-path extinction de-
vice, found that scattering accounts for 55 to 65 percent of extinction in urban Phoenix,
Arizona, and about 95 percent on a plateau outside Flagstaff, Arizona. Wolff et al. (1980),
using a nephelometer and absorption values inferred from elemental carbon loadings, produced
data that show that scattering accounted for 60 to 85 percent of extinction at a variety of
sites.
Fine particles (i.e., those particles of diameter less than 1 to 3 jjm) usually dominate
light scattering. Coarse particles are occasionally important, particularly near roadways and
some industrial sites and during natural occurrences of fog and windblown dust. Charlson et
9-20
-------
al. (1978a) used Mie theory to calculate the light-scattering and absorption efficiency per
unit volume concentration of particles for a typical aerosol containing some light-absorbing
soot (Figure 9-9). As illustrated in the figure, ^par.ticle.s of 0.1 to 2 urn diameter are the
most efficient light scatterers. The remarkably high scattering efficiencies of these parti-
cles are illustrated by the following examples: a given mass of aerosol of 0.5-(jm diameter
scatters about a billion times more light than the same mass of air; a 1-mm-thick sheet of
transparent material, if dispersed as 0.5-(jm particles, would be sufficient td scatter 99 per-
cent of the incident light, that is, to obscure completely vision across such an aerosol
cloud. \
A more revealing explanation of the usual dominance of scattering by fine '^articles is
possible: particles smaller than 0.1 (jm, though sometimes present in high numbers, are
individually very inefficient at scattering light and thus contribute very little to visibi-
lity loss; particles larger than about 1 to 3 fjm, though individually efficient at scattering
light, usually exist in relatively small numbers and contribute only a small fraction of visib
ity loss.
Atmospheric particles are made up of a number of chemical compounds (see Chapter 2), All
these compounds exhibit a peak scattering efficiency in the same diameter range (0.1 to 1.0
|jm) calculated to be optically important in Figure 9-9. Because of differences in refractive
index, however, the values of the peak efficiency and the exact particle size at which it
occurs vary considerably among the compounds (Figure 9-10; Faxvog, 1975).
In Figure 9-10, note the high extinction efficiencies of C and water. As discussed in
Section 9.2,4, these compounds are often significant fine mass components and are therefore
often responsible for significant amounts of extinction. Figure 9-10 should not, however, be
taken to present invariable, precise extinction efficiencies of the various species. It was
produced with best estimates of the refractive indexes and for raonodisperse particles. In
reality, the species do not exist as monodispersions or in equal concentrations, and therefore
their relative roles in causing extinction may vary considerably.
Measured particle size distributions can be used in conjunction with Mie theory calcula-
tions to determine the contribution of different size classes to extinction. The results of
this kind of calculation are shown in Figure 9-11. The peak in scattering per unit volume
concentration is at 0.3 pm, so that the fine particles dominate extinction in most cases.
Because the peak and shape of the bimodal particle mass (or volume) distribution curve
can vary, the light-scattering characteristics of a given particle mass might also be expected
to vary. As noted by Charlson et al. (1978a), however, for the observed range of atmospheric
particle distributions, the calculated scattering coefficient per unit mass concentration is
relatively uniform. Latimer et al. (1978) have determined the scattering per unit volume con-
centration for aerosol size distributions having several geometric standard deviations, includ-
9-21
-------
10
§ i
^ w"
I E
^ a.
a. »
of™ O.
SCATTERINQ
10"
10"' 2

DIAMETER,
10
4.00
Figure 9-9. For a light-scattering and absorbing particle, the scatter-
ing per volume concentration has a strong peak at particle diameter
of 0.5 ;um (m = 1.5—0.05i; wavelength = 0.55 ,um). However,
the absorption per aerosol volume is only weakly dependent on
particle size. Thus the light extinction by particles with diameter less
than 0.1 .Aim is primarily due to absorption. Scattering for such
particles is very low. A black plume of soot from an oil burner is
a practical example.

Source: Charlson etal. (1978a).
9-22
-------
01 10
DIAMETER, fim
2 -
14

1 8

I6
£
° 4

0
X "0 55
001
0.1
10.0
DIAMETER,
Figure 9-10 (A) Calculated scattering coefficient per unit mass
concentration at a wavelength of 0 55 >im for absorbing and non-
absorbing materials is shown as a function of diameter for single-
sized particles The following refractive indices and densities (g/cmj)
were used carbon (m = 1 96-0.66i, d = 2 0), iron (m = 3.51-3 95i,
d = 7 86), silica (m = 1 55, d = 2 66), and water (m = 1 33, d = 1.0)
(B) Calculated absorption coefficient per unit mass concentration
at 0.55 yum for single-sized particles of carbon and iron (C) Calculated
extinction coefficient per unit mass concentration at 0 55 ^m for
single-sized particles of carbon, iron, silica, and water

Source (a) Faxvog (1975), (b and c) Faxvog and Roesslgr (1978).
9-23
-------
If J
s z z
[_ U) —
VOLUME
O
g^
15
C r-
v> o,
z 5
g3
35 m
< a

?l
3, ?»
•^>O
O
0.01
0.1
1.0
10
PARTICLE DIAMETER, jim
Figure 9-11. For a typical aerosol volume (mass) distribution, the
calculated light-scattering coefficient is contributed almost entirely
by the size range 0.1-1.0 /im. The total crsp and total aerosol
volume concentration are proportional to the area under the
respective curves.

Source: Charlson et al. (1978a).
9-24
-------
ing one that matches the Eastern-Western, urban-nonurban grand average of Whitby and Sverdrup
(1980). The calculated ratio for fine particles changes by no more than 40 percent as the MMD
ranges from 0.2 to 1.0 |jm (Figure 9-12). Figure 9-12 also demonstrates the relative in-
efficiency of coarse particles in degrading visibility.
The relative consistency of calculated light scattering per unit mass concentration over
a range of particle distributions and the dominant influence of fine particles suggest that a
reasonably good approximation of light-scattering coefficient can be obtained by measuring
fine-particle mass concentration. Indeed, agreement from simultaneous monitoring of the two
parameters at a wide variety of sites has been found by several investigators (Table 9-1).
The "normalized" column of Table 9-1 normalizes the reported a /FMC values to a uniform
nephelometer operating wavelength and the best nephelometer calibration values available (Ruby
and Waggoner, 1981). With each study having equal weighting, the mean normalized a /FMC is
2 SP
3.3 (±0.8) m /g; with each study weighted by the number of regression points, the mean normal-
2
ized 0 /FMC is 3.2 (±0.8) m /g. Correlations between the fine-particle mass concentration
and a are consistently high. Figure 9-13 (Macias and Husar, 1976) shows this relationship
for St. Louis. It should be noted that in determining mass concentration gravimetrically (or
by beta gauge), the filters are usually equilibrated at some reference RH before each weigh-
ing. Volatile components of the particles, such as water, may have desorbed (or adsorbed),
thus changing the apparent mass concentration from its true atmospheric value. To compensate
for this effect, many researchers heat the aerosol before nephelometry to reduce the RH and
simulate filter equilibration. Not heating the aerosol before nephelometry can be expected to
produce higher and more variable ratios of a /FMC. Heating the aerosol carries some risk of
distorting the results because volatile components other than water (e.g., ammonium nitrate or
organic particles) may also desorb. An unknown portion of the variation between studies in
Table 9-1 is due to differences in: the volatility of the fine mass; the RH of filter equili-
bration; care in maintaining that humidity throughout the weighing procedure; the mass measure-
ment method employed (gravimetric or beta gauge); the particle classifier employed (cutpoint
and efficiency characteristics); the statistical procedures used; and the detailed calibration
and operating conditions of the nephelometer used in the a determination. The remaining
variation is of course due to actual differences of particle composition and size.
The high correlations indicate that, at the sites studied, fine-particle mass dominates
particle scattering. This was documented in an experiment conducted by Patterson and Wagman
(1977), who monitored the ambient aerosol size distribution with a set of four cascade
impactors in the New York metropolitan area. One (background) impactor was operated only when
the light-scattering coefficient was between 0 and 0.15 km ; impactor A at 0.15 to 0.3;
impactor B at 0.3 to 0.45; and impactor C when values exceeded 0.45. The measured mass dis-
tributions (Figure 9-14; Patterson and Wagman, 1977) show that at good background visibility
9-25
-------
10

•I
0.1
a. -5
. I... I
0,1 1,0

MASS MEDIAN DIAMETER (fim)
10.0
Figure 9-12, Scattermg-to-volume concentration ratios are given
for various size distributions. The ratio values for accumulation
(fine) and coarse modes are shown by dashed lines corresponding
to average empirical size distributions reported by Whitby and
Sverdrup (1980).
Source: Latimer et al (1978).
9-26
-------
TABLE 9-1 PARTICLE LIGHT SCATTERING COEFFICIENT PER UNIT FINE-HASS CONCENTRATION
Location
Mesa Verde, CO
Seattle, WA (residential)
Seattle, WA (industrial)
Puget Island, WA
Portland, OR
New York, NY
St. Louis, HO
Los Angeles, CA
Oakland, CA
Sacramento, CA
Ohio (rural, winter)
Ohio (rural, summer)
Great Smoky Mtns, , NC
Shenandoah Valley, VA
Houston, TX
Denver, CO
Shenandoah Valley, VA
Reported
°sp/FMC
(n»Vg)
2 9
3,1
3,2
3,0
3,2
5 0
5 0
3.7
3 2
4,4
3 5
4.1
3.4
4.2
4.4
3.4
5 8
r
_
0.95
0.97
0.97
0 95
-
0 96
0 83
0 79
0.98
0 85
0 91
0.97
0.96
0.96
0.97
0.91
N
5
58
64
26
108
-
60
39
20
6
55
28
69
36
14
208
59
Normalized
o /FHC Effective Nephel
SP X
(mvg) Wavelength (ran)
2.5
2.9
3.0
2 8
3.0
3.7
4 4
2.8
2 4
3.3
3 1
3.6
3.0
4.4
4.6
2 5
5.0
525
550
550
550
550
475
525
475
475
475
525
525
525
538
538
475
475
Nephel.
Calib.*
old
old
old
old
old
old
old
old
old
old
old
old
old
new
new
old
new
Nephel.
Heated
yes
yes
yes
yes
yes
no
?
no
no
no
yes
yes
yes
yes
yes
yes
yes
Classifier
Cut
*
Point ((Jin)
2.5-3,0
2.5-3 0
2.5-3.0
2.5-3.0
2,5-3,0
1 5
3.0
~4.5
~4.5
~4 5
• 2 9
3.0
2 6
1 0
1.0
2.5
2.5
Reference
1
1
1
1
1
2
3
4
4
4
5
6
7
8
9
10
11,12

Normalized to 525 ran and new calibration values according to Ruby and Waggoner (1981), assuming o proportional to A
Stated in or inferred from reference.
References. 1, Waggoner and Weiss (1980); 2, derived by Charlson et al. (1978a) from Patterson and Wagman (1977); 3, Hacias and Husar (1976);
4, Samuels et al. (1973); 5, Nimnger et al (1981a), 6, Nininger et al. (1981b), 7, Ellenson et al. (1981); 8, Weiss et al. (1982), 9, Waggoner
et al. (1982); 10, Groblicki «t al. (1981), and 11, Ferman et al. (1981), as revised in 12, Wolff et al. (1982).
-------

60
40
§1S20
5 o
u.
,- 0.2
.a:
S- 0.1
I i i i
4/17
4/18
4/19 4/20

TIME, days
4/21
4/22
Figure 9-13, Simultaneous insitu monitoring of 0™ and fine-particle
mass concentration in St. Louis in April 1973 showed a high
correlation coefficient of 0.96, indicating that a™ depends primarily
on the fine-particle concentration.

Source: Macias and Husar (1976).
9-28
-------
Q
f

I
I
2.3-
2.0

1,5

1 0

0.5

1 1 i f 1 i 1 1 1 1 IIII1II I III
- BACKGROUND VISIBILITY f, -
Mtot = 44.3 pglm3
—

«»

_ ^^^ /
^xTr*^ ^ ff*?-
f
1
i
f
i
i
i

^•*l i iiiiin 'i t i in i

—
i
t

\
\
% «•
\
\ iii
0.1 0.2
0.5
10 20
SO 100
i.a
2.0
1 5

1.0

0.5

1 1 1 1 1 1 1 ! 1 I IIII1I I IT!
1 VISIBILITY LEViL
i Mtot = 78'5 ^9/fT|3

—
I
1
i
— nH
n

^ cg^f^^^

A _

•M
MINI i iiiiiii i i**rt"--,
0.1 0.2
0.5
10 20
SO 100
1.0 0.2 0.5
20
SO 100

2.0

1.5

1 0
OS

I i I I Dili 1 i I II 1 1 1 i i i I
VISIBILITY LEVEL C _
Mtnt - 212 jig/m3

_ /
^ -*
%
t
1

S ill)

—

•~~**S^«^rf^^^^™~™™l ****«•_
inn i limit i i 1'1«.
0.1 02 05
1 2 5
DIAMETER, M
10 20
50 100
Figure 9-14. Aerosol mass distributions, normalized by the total
mass, for New York aerosol at different levels of light-scattering
coefficient show that at high background visibility, the fine-particle
mass mode is small compared with the coarse-particle mode. At the
low visibility level, C, 60 percent of the mass is due to fine particles.
The solid lines represent histograms of the impactor data, while
dashed lines represent the best fit of a bimodal, log-normal
distribution.
Source: Patterson and Wagman (1977).
9-29
-------
levels, coarse particles constituted 70 percent of the mass concentration. At the low visibi-
lity level C, however, over 60 percent of the total mass was contributed by fine particles.
Thus, visibility in the New York metropolitan area was found to be lowest when the concentra-
tion of fine particles reached a maximum.
It is conceivable "that in the arid Southwestern United States the aerosol refractive
index and relative amounts of coarse and fine particles are so different that the a /FMC
ratios quoted above would not be applicable. Preliminary results from project VISTTA,
however, (Macias et al., 1980, 1981) suggest that 0 /FMC ratios in the Southwest are ~ 3 m /g
as measured elsewhere.
In areas where fine-particle concentrations are low, coarse particles may contribute
significantly to light extinction. However, coarse dust particles are much less efficient
scatterers per unit mass (Figure 9-10), so that much higher mass concentrations are required
to produce a given optical effect. In windblown dust, for example, Patterson and Gillette
(1977) reported values of the ratio of light scattering to mass concentration that were more
than an order of magnitude lower than those noted above for fine particles.
9.2.3.4 Particle Absorption—Particle absorption (0 ) appears to be on the order of 5 to 10
, ap
percent of particle extinction in remote areas such as a plateau outside Flagstaff, AZ (Weiss
et al,, 1979). Its contribution may rise to 50 percent of a . in urban areas, with values of
10 to 25 percent typical for suburban and rural sites (Weiss et al., 1979; Waggoner et al.,
1981; Groblicki et al., 1981).
The apparently large role of absorption in visibility in urban areas has prompted much
research on absorption in recent years. Weiss et al. (1978) fractionated aerosol at many
sites into coarse and fine portions (cut at 2.5 urn) and found that the fine particles ac-
counted for 80 to 98 percent of absorption. They also determined the wavelength- (color-) de-
pendence of a__ for the samples, which showed that at United States sites, 0__ is approxi-
dp op
mately proportional to the inverse of wavelength. Thus 0 is capable of causing brownish
ap
coloration of objects viewed through an aerosol.
The extinction per unit mass concentration for absorbing aerosols is a figure of some
importance because estimates show it to be significantly higher than for scattering-only
aerosols. Figure 9-10(c) shows the theoretical value for carbon particles to be higher than
for any other species considered (Faxvog and Roessler, 1978). Jennings and Pinnick (1980),
relying on the small size of combustion particles (diameters generally less than 0.3 |jm) and
the approximate linearity of the extinction efficiency factor in this region, predicted a
linear and size distribution-independent relation between extinction coefficient and mass
2
concentration of carbon particles, with a ratio of 9.5 m /g at 0.55 urn. Laboratory studies by
2 2
Roessler and Faxvog (1980) showed values of 9.8 m /g for acetylene smoke and 10.8 m /g for
diesel exhaust. They also summarized results from other investigators on various aerosols
9-30
-------
2
that showed values of 6.1 to 9.5 m /g. In developing a spectrophone, Truex and Anderson
2
(1979) measured an absorption of mass concentration ratio 17 m /g (at 0.417 (jm wavelength) for
aerosol from a propane-oxygen flame. As methods for measuring elemental carbon have improved,
Groblicki et al. (1981) have performed atmospheric measurements of fine absorption/fine
2
elemental carbon mass concentration in Denver and found an average of 11.8 m /g. While the
amount of absorption per unit mass concentration depends on chemical composition and particle
size distribution (Waggoner et al., 1973; Bergstrom 1973), the pattern emerging from these
empirical and theoretical studies is that absorbing particles have a more significant visibi-
lity impact than their mass would indicate, probably by a factor of 3 to 4, compared to
scattering-only particles.
Weiss and Waggoner (1981) calculated that, for constant mass concentration, changing 20
percent of each particle of a nonabsorbing aerosol to an equal volume of absorbing soot
reduces visual range (or increases a .) by about 35 percent. They pointed out the importance
of this concept as fuel conservation practices (e.g., use of diesel engines, wood burning)
lead to greater emissions of light-absorbing aerosol.
9.2.4 Chemical Composition of Atmospheric Particles
Given the dominance of particles in degrading visibility, it is natural that their compo-
sition should be studied. Such knowledge permits estimates of the roles various sources play
in visibility impairment. This section discusses the commonly observed particulate species in
the context of visibility impairment (see Chapter 5 for a more detailed discussion.) Before
discussing the aerosol components currently believed to be of significance, it is important to
consider some uncertainties in measurements and deductions.
Contrary to first impressions, visibility is an extremely complex subject, being deter-
mined by the sum of all atmospheric constituents, lighting conditions, the target, and the
observer. These factors are all extremely variable. Although it is generally accepted that
fine particles cause most visibility problems, the concentration and composition of these
particles can vary considerably at different times and sites. Furthermore, field studies that
characterize the aerosol must include diverse and sometimes elaborate instruments and
techniques to measure all relevant parameters simultaneously. There is evidence of volatile
aerosol (e.g., ammonium nitrate) that exists in the atmosphere and thus degrades visibility,
but cannot always be retained by conventional filtration for subsequent analysis. Some fine
particles (e.g., diesel exhaust) are distinctly nonspherical and hinder attempts to model
their optical influence theoretically or even report a size distribution. Until recently,
several fine mass components (e.g., ammonium nitrate and elemental C) have been measured
either inaccurately or not at all. Water is a fine mass component that is particularly
difficult to measure. Water degrades visibility only when in the liquid or solid phase, but
it is ubiquitous and causes particles of several common species to grow or shrink, which can
9-31
-------
affect visibility drastically. Thick hazes or fogs are often dismissed as being caused solely
by high humidity, whereas in some cases they might not have formed without the presence of
anthropogenic hygroscopic nuclei.
Fortunately, there are reasons to believe we can circumvent these uncertainties.
Improved analytical methods are evolving for nitrate measurement. Carbon measurement tech-
niques are being intercompared and applied to the atmosphere. Empirical work by Janzen (1980)
on highly irregular carbon black particles indicates that absorbing chain-aggregates (similar
to diesel exhaust and flame soot), despite their extreme nonsphericity, can be modeled
surprisingly well with Mie theory by assuming the aggregates to be spheres of equal volume.
Relatively constant scatter/fine mass concentration ratios are being reported for a variety of
sites (see Table 9-1). Extinction calculated from gross size distribution measurements is
usually within a factor of two of measured values (Ensor et al. , 1972; Patterson and Wagman,
1977). Regression analyses show consistent correlations between scattering or visibility and
sulfate concentration (White and Roberts, 1977; Cass, 1979; Trijonis, 1978a,b).
Current knowledge indicates that fine aerosol is composed of varying amounts of sulfate,
ammonium, and nitrate ions; elemental carbon and organic carbon compounds; water; and smaller
amounts of soil dust, lead compounds, and trace species. The following discussion separates
the components, although in reality they may exist as internal mixtures (i.e., coexist within
the same particle).
Sulfate occurs predominantly in the fine mass (Stevens et al,, 1978; Tanner et al., 1979;
Lewis and Macias, 1980; Ellestad, 1980). Sulfate ion generally constitutes 30 to 50 percent
of the fine mass at a wide variety of sites (Stevens et al., 1978; Pierson et al., 1980;
Stevens et al., 1980; Lewis and Macias, 1980; Ellestad, 1980; Macias et al., 1981), although
some urban sites have values of 10 to 20 percent, perhaps due to locally high values of other
fine mass constituents (Countess et al., 1980b; Cooper and Watson, 1979). Sulfate usually
occurs in combination with hydrogen and ammonium ions (Stevens, 1978; Pierson et al., 1980;
Charlson et al., 1978b; Stevens et al., 1980; Tanner et al., 1979) and to a lesser extent
calcium and magnesium. Indirect measurements of sulfate by examining the scattering response
of atmospheric aerosol to changes of relative humidity confirm the prevalence of H^SO^ and its
ammonium salts (Weiss eta!., 1977; Waggoner et al., 1981). Regression analyses by Cass
(1979), White and Roberts (1977), Trijonis and Yuan, (1978a,b), Grosjean et al. (1976),
Leaderer et al. (1978), and Heisler et al. (1980) show significant correlations between
sulfate concentrations and visibility or extinction. High correlations between variables do
not necessarily imply cause-and-effect; however, a lack of correlation would imply the absence
of cause-and-effect.
Ammonium ion is typically found to account for 5 to 15 percent of the fine mass (Lewis
and Macias, 1980; Patterson and Wagman, 1977; Countess et al., 1980b). It often correlates
well with sulfate levels (Tanner et al., 1979). Because of the possible reaction of ammonia

9-32
-------
with previously collected acidic particles, reported ammonium ion values may be higher than
actually exist in the atmosphere.
Recognized sampling problems prevent valid statements about ambient participate nitrate
concentrations at present (Appel et al., 1979; Spicer and Schumacher, 1977 and 1979). Simple
filtration (even with nonalkaline, nonreactive, high-purity filters) may not give true values,
due to the tendency of ammonium nitrate to seek equilibrium with ammonia and gaseous nitric
acid during sampling and storage.
Concentrations of apparent elemental carbon have been reported by Wolff et al. (1980) for
3
a variety of United States sites. Apparent elemental carbon was found to range from 1.1 (jg/m
3
at a remote site to an average of 5.9 |jg/m for urban sites, with about 80 percent in the fine
fraction. The word "apparent" is used because the value may be high due to charring of some
organic particles during the organic analysis. Shah et al. (1982), employing a combustion tech-
nique that included a quantitative correction for charring based on reflectance change, has
analyzed archived NASN hi-vol filters collected during 1975 at 46 urban and 20 rural United
3
States sites. The mean annual elemental carbon concentration was 3.8 ng/m at the urban sites
and 1.3 pg/m at the rural sites. Countess et al. (19805) determined that elemental carbon
accounted for 15 percent of Denver's fine mass. Considering the high extinction efficiency of
elemental C reviewed in Section 9.2.3.4, elemental C obviously has a significant effect on
Denver's visibility. Several investigators have concluded that elemental carbon is the only
significant light-absorbing species, including Rosen et al. (1978) who employed Raman spec-
troscopy, Allegrim (1980) who examined the spectral dispersion of the imaginary part of the
refractive index, and Pierson and Russell (1979) and Weiss et al. (1979) who employed various
solvent extraction schemes.
Determinations of particulate organic carbon concentrations are not discussed because of
uncertainties in their measurement from adsorption or volatilization. Also, most reported
organic particle concentration data are for unfractionated samples (i.e., all particle sizes
present). Therefore, even though some sampling and analytical techniques may have validity,
the interpretation of the concentrations for visibility purposes is hindered by the possible
contribution of coarse organic particles. Improved techniques for organic particulate
measurement are being developed.
Minor contributions to fine mass are made by soil-related elements, lead compounds
(especially in urban areas), and trace species (Stevens et al., 1978).
It is suspected that soil particles significantly impair visibility mostly in arid or
semiarid areas (Patterson, 1977) (in the United States, the Southwestern states). This obser-
vation may be due more to the relatively low fine-particle concentrations there, than to high
concentrations of soil particles. Macias et al. (1981a) indicated that coarse particles
accounted for 6 to 24 percent of extinction at Zilnez Mesa, Arizona, during the 3 days
reported from summer 1979 field work. What fraction of coarse particles is derived from

9-33
-------
natural sources has not been established. Hall (1981) points out that winds and dust devils
probably entrain much more dust when over anthropogenically disturbed soil surfaces (e.g.,
unpaved roads, off-road-vehicle trails) than when over undisturbed soils. Soil dust, whether
stirred up by winds or by anthropogenic turbulence, almost always forms coarse aerosol
(particle diameters greater than about 1 to 3 |jm), although '-a few of the softer species such
as carbonates can form submicrometer particles (Oraftz, 1974).
9.2.4.1 Role of Water in Visibility Impairment—Natural fluctuations of RH can greatly
influence the extinction of light by an aerosol. Since RH generally increases following
sunset due to declining temperature (assuming a constant dew point), particles usually grow
after sunset. After sunrise relative humidity usually declines as temperature rises, causing
particles to shrink as they release water to the vapor phase.
As mentioned earlier, water affects visibility only when in the liquid or solid phase.
Unfortunately, direct measurement of liquid water's contribution to mass is difficult due to
its rapid phase change and the fact that, except in fogs, typically less than 0.01 percent of
all water in a given volume exists in the liquid phase. For example, at 20°C, sea level, and
2
50 percent RH, the water vapor concentration is 8.65 g/m . By comparison, an aerosol composed
of 0.5-pm diameter water droplets of sufficient concentration that a = 0.2 km , has a
3
liquid water concentration of 42 pg/m , Ho et al. (1974) report a method employing microwaves
specific to unbound liquid water. Applying it to the Los Angeles Basin, they found that for
relative humidities between 40 and 70 percent water comprised 5 to 30 percent of the aerosol
mass. Data from Mainz presented by Hanel (1976) are in good agreement. It has been known for
many years that relative humidities above about 70 percent will often greatly reduce visibil-
ity due to the size growth of common aerosol species such as ammonium sulfate and sea salt
(Orr et al., 1958).
A straightforward demonstration of the effect of RH on light scattering is seen by
measuring the scattering coefficient of an atmospheric aerosol while increasing its RH (Covert
et al.» 1972). Figure 9-15 shows typical humidograms observed around the United States. The
method cycles the aerosol's RH in a short period, so the input aerosol can be assumed not to
have changed in composition or concentration while being measured.
Particles of certain inorganic salts commonly observed in the atmosphere (e.g., ammonium
sulfate and sodium chloride) exhibit the phenomenon of deliquescence (i.e., an abrupt trans-
formation from solid particle to liquid droplet, and growth at a RH specific to each com-
pound). Above the deliquescent RH, the droplets absorb water and grow smoothly as RH increases
(Orr et al.» 1958; Charlson et al., 1978b; Tang, 1980). As RH decreases, salt particles that
have already deliquesced do not crystallize until a RH well below the deliquescent RH is
achieved (Tang, 1980; Orr et al., 1958), a phenomenon generally referred to as hysteresis.
Until crystallization occurs, droplets become supersaturated, lose water gradually, and shrink
9-34
-------
T-r-r
(a)
TYSON, MO

1 I I , I , I ! I I I , 1,1,1,
o
£2
Q.
tn f\
b 0
£
Q.
m
to
TYSON, MO

, I ! I , I , I l I i I !
TYSON, MO,

,1,1,1,1,1
' I ' 1 ' I
(d)
TYSON, MO,
i I i I i I i I i I I I i
50% 100%

RELATIVE HUMIDITY
1 l ' i
(e)
i • i ' i • i
PT REYES.CA

I i I , I i I , I , I , I i I , I
O
M
SMOG
SANTA ANA CONDITIONS
PASADENA, CA,
, I , I i I i I i I , I , I , I , I ,
5 0
(g)
MONTANA DE ORO
STATE BEACH, CA
,1,1,1,1,1.1
DELIQUESCENT
RANGE
NON-DELIQUESCENT
RANGE
COMPOSITE
50% 100%

RELATIVE HUMIDITY
Figure 9-15. Humidograms for a number of sites show the increase in asp which can be expected at
elevated humidities for specific sites or aerosol types (marine, Point Reyes, CA; sulfate, Tyson, MO)
and the range observed for a variety of urban and rural sites (composite). Dashed lines in a-d
show the humidogram without injection of ammonia vapor, whereas the solid lines represent those
runs with ammonia injection. In f, the different lines differentiate between smog and Santa Ana
conditions, neither with any ammonia injection. In g and h, the range of humidograms is shown.

Source; Covert etal. (1980).
9-35
-------
gradually as RH declines. If crystallization has not occurred and RH increases, the growth of
droplets will follow the smoothly hygroscopic curve along which it just fell. Figure 9-16
depicts the behavior of an ammonium sulfate particle.
Hysteresis explains the persistence of some hazes at relative humidities below that at
which they formed (Orr et a!,, 1958). The RH of crystallization depends on the size of
insoluble nuclei present in each droplet. The dissipation of a water-enhanced haze may not
occur abruptly as RH falls (even though it may have formed abruptly) because the sizes of
nuclei within the droplets may vary, causing the droplets to crystallize at different relative
humidities.
The hygroscopic salts discussed above do not necessarily exist as external mixtures
(i.e., as pure and separate particles), but may be internally mixed (i.e., coexist within the
same particle). Tang et al. (1978) have shown that the phase diagram of a multi-component
system allows accurate prediction of its phase transformation and growth. In the important
sulfuric acid-ammonium sulfate system they found the molecular composition to be of consider-
able importance. For example, for ammonium sulfate'/total sulfate ratios above 0.95, deli-
quescence occurs at 80 percent RH, for ratios 0.75 to 0.95 at 69 percent, for ratios 0.5 to
0,75 at 39 percent, and for a ratio below 0.5 growth is not deliquescent but smooth. In some
cases (e.g., letovicite, (NH*)., H(SCL)/,) the phase diagram predicts two-stage growth, which
was confirmed by experiment. The phase diagram approach apparently leads to quite accurate
predictions of the growth of pure and mixed deliquescent aerosol particles.
To avoid the stringent requirement of knowing the detailed internal composition of the
particles, other investigators have pursued theoretical and empirical simplifications. The
most extensively developed model of this type is that of Hanel (1971, 1976). His model pre-
dicts that the extinction efficiency factor remains almost constant below relative humidities
of 90 percent due to the rise in size parameter from particle growth and simultaneous reduc-
tion of refractive index from dilution by water. Thus,
o-Sp (RH2)
r (RH2)
r (RHJ)
(9-13)
where r = particle radius.
Using a semiempirical formula for the size growth of particles, he arrives at:
CRHt)
- RH2
2e
(9-14)
where e = constant (~ 0.25.) This result is appealing in its simplicity and its independence
of size distribution and composition. Several restrictions, however, apply:
9-36
-------
o
•o
O
2.4
2.2
2.0
1.8
o
iu

55 •"
ui
_i

p 1-4

* 1.2

1.0
THEORETICAL

* ]• EXPERIMiRITAL
o J
20 30 40 50 60 70 80

% RELATIVE HUMIDITY
90
100
Figure 9-16. Relative size growth as a function of relative humidity

for an ammonium sulfate particle at 25° C.

Source: Tang (1980).
9-37
-------
1. Particle composition is the same for all sizes;

2. The particle'diameters of interest for visibility are 0.2 to 2. urn;

3. Applicable only for RHg < 95 percent and RH, of 40-80 percent, unless one assumes a
model aerosol and changes e;

4. Inapplicable to maritime aerosol that has not increased from very low RH.

The first restriction causes some concern. It is not clear that fine particles have uniform
composition. Though absolute nitrate concentrations are suspect, both Appel et al. (1978) and
Patterson and Wagman (1977) indicate that sulfate and nitrate may possess different HMD's.
Countess et al. (1980a) found that in Denver the mean daytime HMD's of sulfate and nitrate
were 0.36 and 0.30 jjm, respectively. Any success of Hanel's theory must be due to the fre-
quent atmospheric observation that ammonium sulfate is the predominant hygroscopic species.
The first and second restrictions may be severe in cities, where absorption by elemental
carbon is significant because elemental carbon resides mostly below 0.2 urn diameter and is
unlikely to be as hygroscopic as sulfate. The fourth restriction is necessary because of
hysteresis. Even for nonmaritime aerosol, ambient RH often does not become small enough to
cause crystallization of all major sulfate and nitrate species; thus, this restriction may
be serious. Atmospheric testing of Hanel's theory is difficult as suggested by Hanel (1976)
and demonstrated by Presle and Pirich (1980) due to continual advection and the resulting
potential changes of particle concentration, size, and type.
In summary, the role of water in visibility impairment can be significant and is
difficult to predict accurately to better than a factor of two, unless detailed and accurate
data are available about the hygroscopic particles from 0.1 to 2.0 pm diameter. Natural
fluctuations of RH will continue to occur. Their influence on visibility depends on the
presence of hygroscopic particles, both natural and anthropogenic.
9.2.4.2 Light Extinction Budgets—Light extinction budgets (LEBs) are an attempt to assign a
percentage of the total extinction to each chemical species. A common mistake in assessing
the compounds responsible for visibility impairment is to assume that a compound's contribu-
tion to mass or fine mass is proportional to its contribution to visibility impairment. Its
visibility contribution can be greater or lesser than its mass contribution, depending
primarily on its size distribution and refractive index. The great usefulness of LEB's is to
exonerate or implicate emission sources as primary causes of visibility impairment. If
impairment can be demonstrated to be due mostly to compound A, and source X emits no compound
A (or its precursors), source X probably has little to do with the visibility problem. The
complex issues of pollutant dispersion and transformation are thus avoided. Reasonably

9-38
-------
complete LEB's have been possible only recently because of a lack of good data for key species
(e.g., sampling artifacts for nitrates and organics, imperfect instrument performance, no
technique for elemental carbon). Another problem is their expense. Many diverse and some-
times elaborate techniques must be applied to measure all relevant parameters simultaneously
for a statistically significant period.
Two approaches have been used to arrive at light extinction budgets: (1) measurement of
each species' size distribution and calculation of a . by Mie theory; (2) statistical
analysis, usually multiple linear regression. The former method requires detailed size
distributions for each species, which is a problem because many significant species have HMD's
at the steepest part of the extinction efficiency curve (Figure 9-7). Thus, a small error in
the size distribution measurement may cause a large error in the calculated extinction.
Further, detection limits, artifact formation, volatilization, unknown particle density, or
imperfect instrument performance often lead to substantial uncertainties in the size distri-
bution. The advantages of the first approach are that it is not affected by interdependences
among pollutant concentrations and that it allows an independent estimation of a . , which can
then be compared to the observed a ...
Using multiple linear regression, one calculates a coefficient for each species that is
then multiplied by that species' concentration to yield its contribution. Calculation of the
coefficients requires use of the observed a ., so this method does not allow an independent
check of the results. Multiple linear regression may produce misleading results when the in-
dependent variables are highly correlated or when some variables are omitted (Snedecor and
Cochran, 1980). In applying this technique to determine a LEB it is important that only the
fine particles of the species be included in that species' concentration measurement; other-
wise the coarse portion (relatively unimportant in visibility) will degrade the accuracy of
the LEB.
Regrettably, few field studies have produced .reasonably complete LEB's. Most older
studies did not include key species or did not use only the fine component of species. Other
studies have been of very short term and may be of limited statistical significance. Two
studies will be presented herein, but it must be recognized that they may not be representa-
tive of other sites or even of the same sites at different times.
The most complete urban LEB_ to date is that for Denver in the winter (Groblicki et al.,
1981). Data were collected for 41 days during November and December 1978. Employing multiple
regression analysis, they found that, on the average, elemental carbon accounted for 38 per-
cent of the variable part of extinction (i.e., all extinction except Rayleigh scattering ex-
tinction), ammonium sulfate 20 percent, ammonium nitrate 17 percent, organic compounds 13 per-
cent, other PM 7 percent, and N0? 6 percent.
9-39
-------
The second example Is for Zilnez Mesa, Arizona, LEBs1 were calculated for 3 days in the
summer of 1979, using the size distribution-Mie approach (Macias et a!., 1981; Ouimette, 1980).
For these few days, a . caused by particles averaged 33 percent organic carbon, 18 percent
ammonium sulfate, 13 percent elemental C, 22 percent other fine species, and 15 percent coarse
particles.
No matter how accurately they may eventually be assessed, light extinction budgets are
predictions only for the present conditions of sources, meteorology, etc. Because of the
physical/chemical interactions among fine particles in the atmosphere, a good light extinction
budget may prove untenable as conditions change.
9-2.5 Considerations in Establishing a Quantitative Relationship Between Fine-Particle Mass
Concentration and VisualRange
For consistency, visual range must be defined in terms ,of a black target. Otherwise, the
same atmospheric conditions will allow different values of visual range depending on the in-
trinsic contrasts of different targets, a confusing and misleading situation that must be
avoided. For this reason, the use of airport visibility observations must be treated with
caution. Airport visibility markers are defined as "Dark or nearly dark objects viewed
against the horizon sky during the day," however, official guidelines for selecting the marker
set specify that "Insofar as possible, markers of the type described in paragraph 2.7 should
be used for determining visibility." (National Weather Service, 1979). There are no defined
quantitative selection criteria, nor even rules that the chosen markers be documented quanti-
tatively as to intrinsic brightness. Airport visibilities may be useful for visibility trend
analysis, presuming that careful checks are made for marker and observer changes, but com-
parisons among airports or with instrumental determinations may be fraught with uncertainties.
Beginning with Equation 9-5 and the (usually) reasonable assumptions made in deriving it,

V - ~ 1oge e = JL (9-15)
CText aext
where V = visual range
& - observer's contrast threshold
°V,«4. = extinction coefficient.
ext
Letting y = (a + a )/FHC, (FMC represents fine mass concentration),
sp ap
gsg + CTag
VV V

where a = scattering by gases (Rayleigh scattering) and a = absorption by gases.
9-40
-------
There are thus two unknowns, K and y, which must be established to relate FMC and V,
assuming that extinction by gases is small. It is, of course, possible to relate FMC and V
directly; however, studies have historically reported either K or y. Their separate deter-
mination distinguishes the effects of contrast threshold (on K) from those of RH (on y). We
thus have a better physical understanding of why variations exist at different sites and
times.
As discussed at the beginning of Section 9.2, no fixed value of K is generally applicable
because observer contrast threshold varies with many factors and because it may be desirable
to decrease K if greater than 50-percent detection probability is desired.
The choice of y requires attention to several factors, most notably the effect of RH. y
is usually determined empirically from a ratio of a measurement of o .to FMC determined from
filters equilibrated at a medium RH (typically 50 percent); however, the 0 . determination
often uses a nephelometer operated at an RH substantially different from ambient RH (either
intentionally or unintentionally). Thus, using these y values with FMC from equilibrated
filters results not in an estimate of in situ o +, but in an estimate of a at a different
— ex°c ext
RH. To correct for the RH effect, it is usually necessary to increase y by a factor of about
1.5 at 80-percent RH, about 2 to 3 at 90-percent RH, etc. (see Figure 9-15). The frequency of
occurrence of high RH is of obvious importance in choosing the factor. Table 9-2 addresses
this point. For example, in January at 6 a.m., half the stations report RH > 50%, 98 percent
of the time. By midafternoon, half of the stations report RH > 50 percent only 71 percent of
the time. Though high (:> 90 percent) daytime RH occurs relatively infrequently, the effect of
hysteresis must be remembered: until a RH substantially lower than the deliquescent RH is
reached, deliquescent particles (e.g., ammonium sulfate) remain as droplets and continue en-
hanced scattering due to water. The large daily variation of RH at a given site, hysteresis,
and sites where hygroscopic particles may not dominate fine mass complicate the choice of a
fixed factor for accounting for RH effects on y.
Other factors that influence y include the size and composition of fine particles. One
expects that at sites with high fine mass concentrations the fine particles will grow to some-
what larger sizes than at relatively cleaner sites, due to increased coagulation rates. Be-
cause of the steepness and monotonicity of the extinction efficiency curve in the usual fine-
particle diameter range, y will be enhanced in high loading areas relative to cleaner sites;
however, the enhancement factor is quite dependent on the size distributions involved. The
effect is no doubt of second order compared to RH effects on y.
Fine particle composition may have two effects on y. First, in areas with significant
absorption by particles (e.g., urban and industrial sites), the °"SD/FMC values reported in
Table 9-1 are underestimates of y, perhaps by 10 to 50 percent. Second, if the- volatile frac-
tion of fine mass changes from that present during each of the Table 9-1 studies, y should be
9-41
-------
TABLE 9-2. MEDIAN PERCENT FREQUENCY OF OCCURRENCE OF SELECTED RH
CLASSES FOR 54 STATIONS IN THE CONTIGUOUS U.S.

Relative Humidity
Time3
January,

April ,

July,

October,

Annual
average

6 a.m.
3 p.m.
5 a.m.
3 p.m.
4 a.m.
3 p.m.
5 a.m.
3 p.m.
a.m.
p.m.
= 50%
98
71
97
44
100
46
98
42
98
51
= 70%
82
35
77
18
95
12
84
14
85
20
= 90%"
29
10
27
4
36
2
37
4
32
5

Source: Derived from U.S. Department of Commerce (1968). Based mostly on the years
1949-1954.

aAverage for each midseasonal month. Hours (in local standard time) were
selected to conform most nearly to the average national daily maximum (early
morning) and minimum (3 p.m.) RH.
9-42
-------
increased if the volatile fraction increases and decreased if the volatile fraction decreases.
As noted earlier, the detailed sampling and storage conditions for y determination may in-
fluence the result significantly when volatile species are present. That atmospheric concen-
trations of volatile species may change can be predicted from projected emissions and atmos-
pheric chemistry. For example, current estimates of emission trends in the Ohio River Basin
from the mid-1970's to the year 2000 are that S02 emissions will increase only slightly, while
NO emissions will grow substantially (Glass, 1978; Stukel and Keenan, 1977). Because a
A
portion of NO emissions eventually form volatile nitrate particles (Orel and Seinfeld, 1977),
.X
there may be increases in the volatility of fine mass downwind of this region in the coming
years. Present developments in the measurement of volatile compounds will reduce our uncer-
tainty about this aspect of y as they come into use.
The net effect of all these factors (except a decrease in volatility) is to increase the
value of y from aSD/FMC values reported in Table 9-1. The points discussed above should be
considered in choosing a fixed value for y.
Figures 9-17 and 9-18 show the relationship of fine mass concentration to K, y, and
visual range, in one graph for a fixed visual range and in the other for a fixed K. Note that
these figures deal with visual range, which airport visibility observations always underesti-
mate to some extent. In both figures Rayleigh scattering is taken into account but 0,_ is
aa
assumed to be zero.
9.3 VISIBILITY AND PERCEPTION
The term "visibility" is used colloquially to refer to various characteristics of the
optical environment, such as the clarity with which distant details can be resolved and the
fidelity of their apparent coloration.
Traditionally, visibility has been defined in terms of visual range: the distance from an
object that corresponds to a minimum or threshold contrast between that object and some appro-
priate background. Threshold contrast refers to the smallest brightness difference between
two stimuli that the human eye can distinguish. The measurement of these quantities depends
on the nature of the observer, his or her physical health, and state of attention or dis-
traction, among other things.
Although visibility defined by visual range is a reasonably precise definition, visibi-
lity is more than being able to see a target at a distance for which the contrast is reduced
to the threshold value. Visibility also includes seeing vistas at shorter distances and being
able to appreciate the details of line, texture, color, and.form. Any definition of visibi-
lity and the selection of methods for monitoring visibility impairment should relate to these
9-43
-------
300
250
I 200
a-
Z
O
150
u

O
u

5
UJ
Z
EL
100
BO
Figure 9-17. Fine mass concentration (determined from equilibrated
filter) corresponding to 4.8 km visual range, as a function of K and y,
where K equals the Koschmieder constant {—loge c), and y equals (osp +
oap)/fine mass concentration.
9-44
-------
VISUAL RANGE = 3.9/oext
40
1 30
UJ
0

LTI
10
y = 3 m'/g
y = 6 m2/g
y = 10 m2/g
25
50 75

FINE MASS CONCENTRATION,
100
Figure 9-18. Visual range as a function of fine mass concentration (determined from equilibrated
filter) and 7, assuming K=3.9.
-------
aspects of perceiving distant objects. The selection of a good parameter (or parameters) to
characterize visibility depends on whether one's interest is in human perception of visual air
quality or in the cause of visibility degradation (that is, the air pollution itself).
Apparent target contrast relates well' to how a person perceives visual air quality and serves
as the fundamental measure of visibility impairment (U.S. Environmental Protection Agency,
1979). Color change is also a useful measure. Fine particle concentrations and scattering
coefficients are good measures when relating visibility impairment to pollutant sources.
Visual range cannot be measured directly by an instrument. Visual range, however, can be
derived from instrumental data.
Visual impairment resulting from air pollution can occur as layered haze or as uniform
haze (Malm et al., 1980a,b,c). Layered haze produces a visible spectral discontinuity between
itself and background (sky or landscape) while uniform haze exhibits reduced overall air
clarity. The classic example of a layered haze is a tight, vertically constrained, coherent
plume. As the atmosphere changes, however, from a stable to an unstable condition and a
plume mixes with the surrounding atmosphere, the diffused plume, though no longer visible as a
layer of haze, may reduce overall air clarity. The perceived effects of layered and diffused
haze are quite different, because the eye is much more sensitive to a sharp demarcation in
color or brightness than it is to a gradual change over time (Green, 1965; Patel, 1966).
Since a change in uniform haze takes place over hours or days, an evaluation of visual air
quality change resulting from a uniform haze requires remembering what the scene looked like
before a change in air pollution took place. A layered haze, on the other hand, is evaluated
by direct comparison with the background. Whether the pollution occurs as layered or uniform
haze, judgments of visual air quality as a function of air pollution might be altered by
variations in sun angle, cloud cover, and/or landscape features.
An assessment of visibility impairment produced by air pollution must consider the pro-
cess of human visual perception, as well as changes in the optical characteristics of the
atmosphere. The perception of brightness, contrast, and color is not determined simply by the
pattern and intensity of incoming radiation; rather, it is a dynamic searching for the best
interpretation of the visible scene. The relative brightness of an object may vary as a
function of its background, even though its absolute brightness remains constant. For
example, consider the difference between a candle in a brightly lit room and one in a dimly
lit room. The contrast between an object and its surroundings is fundamental to visibility.
As the contrast between object and background is reduced (for example, by fog or air
pollution), the object becomes less distinct. When the contrast becomes very low, the object
will no longer be visible. This liminal or threshold contrast has been the object of
considerable study. The threshold contrast is of particular interest for atmospheric
visibility, since it influences the maximum distances at which various components of a scene
can be discerned. Equally important to visibility is the smallest perceptible change in

9-46
-------
contrast in a viewed scene caused by an increase in pollution haze (U.S. Environmental
Protection Agency, 1979).
Malm et al. (1980a) investigated the relationship of contrast and color, and of changes
in these variables, to the perception of visual air quality. They determined that the various
demographic backgrounds of visitors to a national park did not influence perception, but that
changes in color contrast did influence the accuracy and consistency of perception of visual
air quality. In photographic slides of a mountain scene used as the test vehicle, color
contrast was determined by variables such as weather condition, time of day, and ground cover,
as well as by amount of air pollution. Although an incremental color contrast change was
perceived to be the same across air pollution levels, clean air environments appeared to be
more sensitive to contrast changes. The evidence indicates that a change in air pollution
level produces a larger contrast change in clean air than in relatively dirty air and is,
therefore, more perceptible.
For small objects, the size of the visual image on the retina of the eye also plays an
important role in the perception of contrast. As an object recedes from us and apparently
becomes smaller, details with low contrast become difficult to perceive. The reason for this
loss of contrast perception is not only that the relative brightness of adjacent areas
changes, but also that the visual system is less sensitive to contrast when the spacing of
contrasting areas decreases. If the contrast spacing is a regular pattern of light and dark
bands, (e.g., a picket fence), a "spatial frequency" can be readily described by the number of
pattern repetitions or "cycles" per degree of viewing angle. The human eye is much more
sensitive to contrast at certain spatial frequencies than at others (cited in U.S. Environ-
mental Protection Agency, 1979).
The relationship between perceived contrast threshold and target characteristics (size
and pattern) is important for visibility because a scenic vista usually contains a number of
targets of varying sizes and arrangement. The calculation of the perceptibility of all tar-
gets requires specification of their angular size distribution. The perception of "texture,"
consisting of contours of small angular size and high spatial frequency, is particularly
affected by this loss of threshold sensitivity.
Measurements of pattern perception thresholds by several researchers (Van Nes and Bouman,
1967; Schober and Hilz, 1965; Cornsweet, 1970) make it clear that the scattering and absorp-
tion of light by particles and gases added to the atmosphere can lead tp a dramatic loss of
visibility through contrast reduction. The operator of a motor vehicle or the pilot of an
airplane who must react quickly to minimal visual cues may be greatly disadvantaged by an
increase in atmospheric pollutants. This loss in contrast could make a target that is
normally visible at 100 meters (109 yards) visible only at 20 meters (22 yards), although a
reduction in visibility to levels this low rarely occurs.
9-47
-------
In many pristine areas, where viewing distances are 50 to 100 kilometers, the most
readily apparent effect of incremental pollution is a reduction in apparent contrast and
discoloration of nearby objects and sky. For large targets, calculation of contrast changes
accompanying increasing particle levels indicates that the maximum decrease in contrast occurs
for objects located at distances of about one-fourth of the visual range from the observer
(Malm, 1979a,b). Thus, in an initially clean atmosphere, an increase in fine particles
produces maximum contrast reduction for large objects 50 to 100 kilometers away. A reduction
in visual range of 5 percent results in a reduction in contrast of 0.02 for those objects.
Such a change may be just perceptible. The contrast detail (texture, small objects) and
coloration of closer objects may, however, be affected to a greater degree (Henry, 1979; Malm,
1979a,b).
The perceived color of objects and sky is also changed by the introduction of fine
particle aerosols. Because it is difficult to specify perceived color, only a qualitative
description is possible. In general, as distance from the observer increase, the apparent
color of a target fades toward the hue of the horizon sky. Without particles, scattered air
light is blue, and dark objects appear increasingly blue with distance. The addition of small
amounts (1 to 5 ug/tn ) of fine particles throughout the viewing distance tends to whiten the
horizon sky, making distant dark objects and the intervening air light (haze) appear more
gray. According to Charlson et al. (1978a), even though the visual range may be decreased
only slightly from the limit imposed by Rayleigh scattering, the change from blue to gray is
an easily perceived discoloration. The apparent color of white objects is less sensitive to
incremental fine particle loadings. Increments in particle levels produce a much greater
color shift in cleaner atmospheres (Malm, 1979a,b).
Aerosol haze can also degrade the view of the night sky. Star brightness is diminished
by light scattering and absorption. Perception of stars is also reduced by an increase in the
brightness of the night sky caused by scattering of available light. In or near urban areas,
night sky brightness is significantly increased by particle scattering of artificial light.
The combination of extinction of starlight and increased sky brightness markedly decreases the
2
number of stars visible in the night sky at fine particle concentrations of 10 to 30 ug/m
(Leonard et al., 1977).
Thus, aerosol haze reduces visual range and contrast, and changes color perception.
Visually, the objects are "washed out" and the aesthetic value of the vista is degraded, even
though the distances are small relative to the visual range.
Although natural sources of light scattering and light absorbing aerosols are undoubtedly
important in producing geographical and seasonal patterns of visibility impairment, analysis
of visibility trends and other information discussed in later sections suggests that manmade
air pollution is a significant influence. It is also important to note that changes in
pollution levels are most easily perceived in regions with the best visibility.

9-48
-------
9.4 HISTORICAL PATTERNS OF VISIBILITY
Records of visual range can be used to gain insight into the effects of changing emission
patterns on visibility. As an example, Marians and Trijonis (1979) have derived statistical
relationships between light extinction (computed from visibility data) and historical emission
trends. Yearly values of extinction from four Arizona airports were regressed against
statewide emissions of smelter SO , nonsmelter SO , NO , and RHC (reactive hydrocarbons).
/\ f\ f\
Smelter SO was found to be the most significant variable. Particularly close relationships
X . ,
between Arizona smelter SO and visibility at Tucson and Phoenix are shown in Figure 9-19.
J\
Table 9-3 summarizes the results of the correlation/regression analysis between yearly
airport extinction (visibility) data and Arizona smelter SO emissions. The correlation co-
*\
efficients and Student's t-statistics indicate significant statistical relationships at high
»1
confidence levels. The regression (extinction/emission) coefficients of 0.004 + 0.0005 km /
(1000 tons per day of SO ) are remarkably consistent from site to site and represent the
change in yearly median extinction associated with a given change in SO emissions; that is,
adding 1000 tons per day of SO tended to increase yearly median extinction by approximately
0.004 km"1.
Perhaps the best example of changed emission patterns is a strike that shut down the
copper industry for more than 9 months in 1967-1968. In the Southwest at this time, copper
production accounted for over 90 percent of the SO emissions, less than 1 percent of the NO
*\ )\
emissions, and less than 10 percent of the conventional particulate emissions (Marians and
Trijonis, 1979), and should therefore have affected visibility primarily through its contribu-
tion to sulfate loadings. Substantial decreases in sulfate occurred at five locations
(Tucson, Phoenix, Maricopa County, White Pine, and Salt Lake City) within 19 to 113 km (12 to
70 miles) of copper smelters as shown in Figure 9-20. More notably, sulfates dropped by about
60 percent at Grand Canyon and Mesa Verde; these remote sites are located 325 to 500 kilo-
meters (201 to 310 miles) from the main smelter area in southeast Arizona. Comparing measure-
ments taken during the strike with those taken during the surrounding 4 or 6 years, Trijonis
and Yuan (1978a) found a large decrease in Phoenix sulfate loadings, accompanied by a sub-
stantial improvement in visibility (Figure 9-21).
Visibility improved at almost all locations during the strike, with the largest improve-
ments occurring near and downwind (north) of the copper smelters in southeast Arizona and near
the copper smelters in Nevada and Utah. The nine locations showing statistically significant
improvements are all within 242 km (150 miles) of a copper smelter. Attributing the improve-
ment in visibility entirely to the drop in sulfate levels yields an estimated extinction
2
efficiency of 3.9 m /g» in agreement with data in Table 9-1.
Eldred et al. (1981) studied sulfate emissions and ambient air levels before, during, and
after a three-month strike beginning July 1, 1980, which shut down 9 of 11 copper smelters
9-49
-------
us
^* ***
SB 80
w w
EC iu
2= 60
u. U9 40

fe£ 20
ill C
cc m
S 0
I
TUCSON
6000

5000 z

4000

3000

2000
~
m
|-
~i
m
33
W
O
50
55
60 65

YEAR
70
75
80
s; in
5s
OT CO
K "J
S
u.
O
70
= 60
50
40
od
K CD
Si
PHOENIX
I
6000
_
N
O
5000
1
4000 co
s
3000 3

2000

33
50
55
60 65

YEAR
70
75
Figure 9-19. Historical trends in hours of reduced visibility at
Phoenix and Tucson are compared with trends in SOX emissions
from Arizona copper smelters. Data points represent yearly percent
of hours with reduced visibility. The Tucson observation site moved
in 1958; although this move did not produce a statistically significant
change in reported visibilities, open dots are used to distinguish data
prior to 1958. Lines represent yearly statewide SOX emissions from
Arizona copper smelters.

Source: Marians and Trijonis (1979).
9-50
-------
O
' 250 TONS/DAY SO,
URBAN SITE
NONURBAN SITE
100 200
I I
SCALE, miles
MESA VERDE
57%l NATIONAL PARK
Figure 9-20. Seasonally adjusted changes in sulfate during the copper strike are compared with the
geographical distribution of smelter SOX emissions.

Source: Trijonisand Yuan (1978a).
9-51
-------
O
' 250 TONS/DAY SO,
URBAN AIRPORTS
NONURBAN AIRPORTS
0 100 200
I I I

SCALE, miles
Figure 9-21. Seasonally adjusted percent changes in visibility during the copper strike are compared
with the geographical distribution of smelter SOX emissions.

Source: Trijonisand Yuan (1978a).
9-52
-------
TABLE 9-3. CORRELATION/REGRESSION ANALYSIS BETWEEN AIRPORT
EXTINCTION AND COPPER SMELTER SO EMISSIONS
1
tn
OJ

Data set
Tucson (1950-75)
Tucson (1959-75)
Phoenix (1959-75)
Winslow (1948-73)
Prescott (1948-75)
Prescott (1948-69)

Correlation
coefficient
0.91
0.88
0.81
0.68
0.70
0.70
Regression coefficient
extinction/emissions,
(104 m)"1/(1000 TPDa)
0.035
0.038
0.041
0.047
0.031
0.039
t-statistic
(t = 1.7 for 95% confidence)
(t s 2.5 for 99% confidence)
11.1
7.2
5.4
4.5
5.0
4.4

TPD = tons per day

Source: Marians and Trijonis (1979)
-------
near the southern borders of Arizona and New Mexico. At each sampling site, 72-hour concen-
trations were measured for fine and coarse PM, (i.e. those less or greater than 2.5 urn). The
location of samplers and smelters and mean surface wind vectors are given in Figure 9-22. The
investigators concluded that:
1. The largest sulfate concentrations at remote sites in Arizona and southern Utah from
August 1979 to June 1980 were accompanied by wind trajectories from copper
smelters;
2. During periods of southerly winds, the smelter SCL emissions produced significant
increase in sulfate levels throughout northern Arizolia and southern Utah.
Comparisons of one period prior to the strike with a similar one during the^trike
indicated that the smelters increased sulfate in the region from 0.5 \jg/m to 5
ug/m .
3. The sulfate impact from the smelters was increased in summer by higher conversion
rates of SOp to sulfate;
4. During the copper strike of the summer of 1980, mean sulfate concentrations
throughout Arizona decreased 50 percent to 90 percent from levels of the previous
summer.
Macias et al. (1981b) studied the contributions of major source types to air quality and
visibility in the desert Southwest United States. The study focused in the Utah-Arizona
border near the Grand Canyon and Canyonlands National Parks. In their emission inventory
estimates, southern Arizona copper smelters were the largest source of S emitted into the
atmosphere of this region, while southern California was the major source of NO and gaseous

-------
SEP 24-27,1979
SEP 12-14,1980
Figure 9-22. I: The locations of sampling sites (^} and smelters (o) and the mean surface wind
vectors ( $ ) at each sampling site from August 1979 through September 1980; Smelter B at center
of circle has area proportional to estimated 1975 SO2 emissions. Smelters A and B did not
stop production. The restart dates for the others are: 9/16 (C), 10/i {D,E,F,G}, 10/25 (H),
10/31 (I), 12/1 (J), 1/6 (K). The sampling sites are 1 (Tonto), 2 (Fort Bowie NHS), 3 (Organ
Pipe Cactus), 4 (Gila Cliff Dwelling), 5 (Montezuma Castle), 6 (Petrified Forest), 7 (Chaco Canyon),
8 (Grand Canyon), 9 (Bryce), 10 (Canyonlands), 11 (Fort Union), 12 (Gran Quivera), 13 (Carlsbad
Caverns). Sites 1-3,5-8, and 12-13 began in early August 1979, while the others began in late
September. The mean wind vectors at each sampling site are for the period from August 1979,
to September 1980. II: Sulfate concentrations and wind data for two comparison sample periods.
The area of each circle is proportional to the sulfate concentration, with the maximum area representing
6pg/m3. The arrows represent the 72-hour mean wind vectors at each site. The shaded area is the
envelope of 48-hour wind trajectories ending at Grand Canyon. The locations of operating smelters
are denoted by solid squares.

Source: Adapted from Eldred et al. (1981).
9-55
-------
LIGHT EXTINCTION DUE TO PARTICLES
Hizy (6/30)
Clear (7/10)
Organic Carbon

Soot
r.'J Fine Crystal
'• •'
I I Other Fine

1^3 Coarse
Figure 9-23. Particle light extinction (osp + aan) budget for the low
visibility southern California Incursion (June 30) and a clear day
{July 10). On the hazy day ammonium suifate and organic carbon
accounted for 63% of the particle light extinction, while on the clear
day these species accounted for only 31%.

Source: Macias et al. (1981 a and b).
9-56
-------
particle size range and chemical composition and visual ranges. They noted that episodes of
reduced visibility were associated with increased sulfur levels or increased levels of fine
soils in the 0.5 to 3.5 urn range.
Altshuller (1973) has noted an increase over the past decade in sulfate concentrations at
nonurban sites in the Eastern United States, which is not inconsistent with the decreasing
trend in nonurban site median visibilities noted by Trijonis and Yuan (1978b). Unfortunately,
the historical record of sulfate concentrations extends back only to the mid-I960's. Within
the Eastern United States, over 90 percent of the SO emissions are associated with the com-
bustion of coal and oil. One apparent conclusion is that visibility reduction is currently
due in large part to increases in sulfate aerosols, which are formed primarily from coal
combustion-related S0? emissions. Examination of the trends and changing spatial distri-
butions of coal use should be comparable with the change in the light-extinction coefficient.
Air pollutants emitted over the Eastern United States result mainly from the combustion
of fossil fuels: coal, oil products, and gas. The great spatial and seasonal variability of
haziness (inverse of visibility) prompted Husar et al. (1979) to examine the patterns of coal
consumption in the Eastern United States over the past few decades. For comparison with coal
consumption estimates, visibility data are expressed in terms of a light extinction coeffi-
cient, a ., via the Koschmieder formula: a . = 3.9/V.
Figure 9-24 illustrates the striking similarity "between summertime average haziness and
coal use within the Eastern United States over the past three decades. As shown in Figure
9-25, in 1951 the haziness was most pronounced in the winter, when the coal consumption was
highest. By 1974, there was a shift toward a summer peak, coincident with the increasing
summer use of coal. Such coincident behavior alone cannot establish cause-effect relation-
ships. Nevertheless, it is instructive to examine the more detailed spatial and temporal
patterns of coal use« and haziness (specifically, extinction).
Since 1940, the trend in coal consumption has been more pronounced in the summer than in
the winter (Figure 9-26; U.S. Bureau of Mines, 1933-74); since 1960, summer coal use has grown
by about 5.8 percent per year compared with 2.8 percent per year for winter coal demand.
Monthly coal combustion peaked in the winter in the early 1950's, but the seasonal pattern had
shifted to a summer peak by 1974 (Figure 9-25). The corresponding regional trends of haziness
in the Eastern United States (Figure 9-27; Husar et al. , 1979) exhibit changes similar to
those of coal combustion.
In the Ohio River Valley region, the winter (quarter 1) average extinction (oex*» km )
decreased slightly, whereas the spring (quarter 2) average increased. The summer (quarter 3)
extinction increased from roughly 0.25 in the 1950's (a visibility of 10 miles) to about 0.4
in the 1970's (a visibility of less than 6 miles). Fall (quarter 4) extinction remained
essentially unchanged. The summer avera-ge in New England increased from about 0.2 to 0.3,
9-57
-------
175

| 150
o

1 125
100
t

8
75
§ 50
o
25
COAL
\/«« A*
— ' mi*!
I I I I
0 40
0 35
0 30
0 25 ,_
0 20 t,
0.15
0.10
0 05
1940 1950 1960 1970 1980 1990 2000

YEAR

Figure 9-24. Compared here are summer trends of U.S. coal
consumption and Eastern United States extinction coefficient.

Source- Adapted from Husar and Patterson (1980).
0 7
I I I I 1 I I I I I
I
O
z
o
<
0
JFMAMJJASOND JFMAMJ J A S O N D

MONTH MONTH

Figure 9-25. In the 1950's the seasonal coal consumption peaked in the winter
primarily because of increased residential and railroad use (left figure). By
1974, the seasonal pattern of coal usage was determined by the winter and
summer peak of utility coal usage (right figure). The shift away from a winter
peak toward a summer peak in coal consumption is consistent with a shift in
extinction coefficient from a winter peak to a summer peak in Dayton, OH, for
1948-52.

Source: U.S. Bureau of Mines, Minerals Yearbooks 1933—1974.
9-58
-------
175
175
1940
1980
1940
1980
Figure 9-26. In 1974, the U.S. winter coal consumption was well below, while the summer consumption
was above, the 1943 peak. Since 1960 the average growth rate of summer consumption was 5.8 percent
per year, while the winter consumption increased at only 2.8 percent per year.

Source: U.S. Bureau of Mines, Minerals Yearbooks 1933-1974.
9-59
-------
OHIO RIVER
0.1'
_ »-

194'6~So"~60'"7'o~"80 90 194050 60 70 80

NE. MEGALOPOLIS
1940 SO 60 70 80 90 1940 50 60 70 80

EASTERN SUNBELT
1940 SO 60 70 80 90 1940 SO BO 70 80 90
YEAR
t _______ ____ _ __ ___ |..M,.:,,,i...,aii[

1940 SO 60 70 80 90 1940 50 60 '*70 80 90

SMOKY MOUNTAINS

„ 5 ^ ,., W!ffR, i,
0.2
01

1940^0 60^70 80 90 194050 60 70 80 90

MIDWEST
1940 50 60 70 80 90 1940 50 60 70 80 90
YEAR
Figure 9-27. Trends in the light extinction coefficient (oext) in the Eastern United States
are shown by region and by quarters; 1 (winter), 2 (spring), 3 (summer), 4 (fall).

Source; Husar et al. (1979).
9-60
-------
corresponding to a reduction in visibility from 20 to 13 kilometers (12 to 8 miles). The
Northeast megalopolis region shows a general decline in haziness during quarters 1 and 4,
whereas quarters 2 and 3 display a slight increase from a . = 0.27 (15 kilometers, 9 miles)
-i 6X t*
to 0.3 (13 kilometers, 8 miles) x km . The Smoky Mountain region displays a strong increase
in the average summer quarter extinction coefficient from about 0.16 to 0,4, corresponding to
visibility deterioration from 15 to 6 miles (24 km to 10 km). Smaller but still pronounced
increases are noted for quarters 2 and 4. Evidently the Smoky Mountains have become appre-
ciably "smokier" over the past 20 summers. The eastern Sunbelt region has an increased hazi-
ness for all quarters, most pronounced being the summer quarter, with an increased extinction
from 0.2 (12 miles; 19 km) to 0.35 (7 miles; 11 km). In the Midwest, extinction during the
first quarter fluctuated slightly, with no discernible trend. The spring and fall quarters
have increased appreciably, but summer values have nearly doubled, from 0.15 to 0.3 (16 to 8
miles; 26 to 13 km). The spatial shifts of Eastern United States haziness are displayed in
greater detail in Figure 9-28 (Husar et a!,, 1979).
Husar and Holloway (1981) and Sloane (1980) have used historical data to determine trends
in visibility in the Northeastern and Mideastern United States. Husar and Holloway (1981) an-
alyzed visibility observations at Blue Hill, Massachusetts, reported from 1889 to 1958. The
observations focused on the discernibility of three distant mountains. In their analyses, one
of the issues they hoped to resolve was whether visibility was ever as good anywhere in the
Eastern United States as current visibility in the Southwest, owing to the northeast's higher
humidities, greater vegetation densities conducive to secondary aerosol formation, and hygro-
scopic marine aerosols. They noted that two peaks of haziness corresponded roughly with two
peaks of combined coal and wood burning (1910-1920 and 1940-1950). The variability of the
data, however, precluded a conclusion about the relationship between the noted increase in
haziness and the burning of fossil fuels.
Sloane (1980) investigated yearly and seasonal patterns of visibility in the mideastern
United States from the 1948 to 1978. Cumulative percentile and ridit analyses of the data
showed that high-growth urban areas had declining visibility levels, while large metropolitan,
low-growth areas, and the Appalachian region had fluctuating but improved visibility. For
high-growth areas, significant declines occurred during the spring and summer quarter. Al-
though an overall improvement in visibility was shown generally for the Mideastern United
States during the summer quarter, the visibility was still below pre-1960 levels.
Wolff et al. (1981) studied three haze episodes associated with maritime tropical air
traveling northward from the Gulf of Mexico and reported that all three originated in the
Northeast and Midwest United States. The episodes were identified by traces of 0- and by the
temporal changes in light-scattering effects of particles, primarily sulfates. Air quality
data measured by a mobile laboratory and by satellite photography and wind trajectories were
used to~determine movements of the haze.

9-61
-------
1948-52
196O64
1970-74
EC
UJ
I-
" JIT" 0 288 /
I
O
§
a
ce
P
K
EXTINCTION
COEFFICIENT, km

VISIBILITY, miles
(km)
-1
>0.36

<6.6
03-036

86-8
(11-13)
024-030

8-10
(13-16)
0,18-0 24

10-13 3
(16-21)
IZH

<018
>133
Figure 9-28. The spatial distribution of 5-year average extinction
coefficients shows the substantial increases of third-quarter ex-
tinction coefficients in the Carolinas, Ohio River Valley, and
Tennessee-Kentucky area. In the summers of 1948-1952, a 1000-
km size multistate region around Atlanta, GA, had visibility greater
than 24 km (15 miles); visibility had declined to less than 13 km
(8 miles) by the 1970s. The spatial trend of winter (first quarter)
visibility shows improvements in the Northeast megalopolis region
and some worsening in the Sunbelt region. Both spring and fall
quarters exhibit moderate but detectable increases over the entire
Eastern United States.
Source: Husaret al. (1979).
9-62
-------
During the first and second episodes, August 5 to 9 and 14 to 19, 1980, respectively,
transport patterns were nearly identical. Hazy areas in the Midwest and Northeast followed
the 850 mbar (about 1570 meters) streamlines eastward and then southward in a clockwise
direction to the Gulf coast. This movement was followed by a return toward the southern
Midwest from the flow- of the surface high-pressure system above the 850 mbar level. During
the third episode, the hazy air mass travelled counterclockwise to the Gulf Coast, apparently
induced by Hurricane David. Subsequently, it was directed clockwise around a high-pressure
system that moved down from Canada into the Eastern United States and was transported back to
the Midwest.
9.4.1 Natural Versus Manmade Causes
Vision in the natural, unpolluted atmosphere is restricted by blue sky scattering, by
curvature of the earth's surface, and by suspended liquid or solid natural aerosols.
Important sources of natural aerosols include water (fog, rain, snow), windblown dust, forest
fires, volcanoes, sea spray, vegetative emissions, and decomposition processes. The
particle-free atmosphere scatters light and limits visual range to about 320 kilometers (200
miles) at sea level.
Dark objects, such as distant mountains, when viewed in daytime through a particle-free
atmosphere, appear bluish because blue light is scattered preferentially into the line of
sight. Bright snow-covered mountain tops or clouds on the horizon can appear yellow to pink
because the atmosphere scatters more of the blue light from bright "targets" out of the line
of sight leaving the longer wavelength colors. The actual visual range in the particle-free
atmosphere is also limited by the earth's curvature. Thus, Rayleigh scattering is seldom the
limiting factor in the detection of the most distant objects (i.e., the visual range).
Rayleigh scattering is, however, important in reduction of visual texture and in bluish colora-
tion of distant dark visual targets. Moreover, air scattering is solely responsible for the
blue color of the nonhorizon sky,
A recent study in the Shenandoah National Park (Ferman et a!., 1981) estimated that 14 to
22 percent of CT . was attributable to natural causes: sulfates and associated water 3 to 11
percent; organics, 5 percent; Rayleigh scattering, 5 percent; and crustal dust, 1 percent.
Thus, 78~86 percent of a . was estimated to be of anthropogenic origin. The sulfate contribu-
tion estimate was based on other studies in supposedly remote locations, while the organic and
crustal contribution estimates were inferred from actual Shenandoah measurements. The 11-
percent sulfate estimate was based on measurements in remote South Dakota; because of general
sulfate contamination of the northern hemisphere (Lawson and Winchester, 1979), this value may
be more of a background value than due to natural sources. The 3-percent figure (from measure-
ments in South America) is probably more realistic. The natural organic contribution was esti-
mated by assuming all elemental carbon to be anthropogenic and the anthropogenic organic/
elemental carbon emission ratio to be 1,5. Uncertainties in this ratio and carbon speciation

9-63
-------
analyses may have depressed the 5-percent estimate. In the worst case, assuming all measured
carbon was organic and from natural sources, the contribution would rise to 10 percent. The
sum of these slightly revised percentages (including a more realistic value of 3 percent for
Rayleigh scattering for visual observations) indicates that 12 to 17 percent of a t was
attributable to natural causes. Using Koschmieder's formula with K = 3.9, visual ranges of 60
to 80 kilometers in the absence of anthropogenic influence are estimated.
Fog is a natural phenomenon that can reduce the visual range to nearly zero. It is
o
characterized by high liquid water content, typically over 1000 ug/m , dispersed in droplets
with a mean diameter of several pm or more. In "natural" fogs all colors are scattered and
absorbed about equally, so the atmosphere appears white (Husar et a!., 1979).
The historical frequency of fogs in the Continental United States reveals considerable
geographic variability (Figure 9-29). Windward coastal areas experience the highest fre-
quency. Host inland portions of the United States west of the Appalachians can expect fewer
than 20 days of fog per year, with less than 5 days of fog annually in the arid West.
With the exception of coastal and mountainous regions, fogs are rare during the summer
months. Fogs tend to be localized events lasting a few hours at most, commonly during the
early morning hours. On an hourly basis, fogs exist less than 1 percent of the time
(Conway, 1963), Thus, the overall contribution of fog to the degradation of visual air
quality is small, and it is an insignificant cause of reduced visibility during the daylight
hours.
Thunderstorms, other rainfall, and snow can also reduce visibility. East of Nevada, most
of the United States experiences from 30 to 50 days of thunderstorm activity each year. Such
storms are most common on summer afternoons. Since thunderstorms are usually intense but
brief, they also contribute to visibility reduction less than 1 percent of the time on an annual
basis.
Snow is a major natural impediment to visibility. It is an important factor in many
regions of the North and in some mountainous areas, where blowing snow occurs from 1 to 12
percent of winter hours (Conway, 1963). During the winter months, snowstorms may account for
most of the hours of reduced visibility, and certainly may dominate the episodes of extremely
low visibility in winter months.
The natural contribution of fog, thunderstorms, snow, and other forms of precipitation
can thus cause severe degradation of visual air quality. With few exceptions, however, these
intense but infrequent events do not dominate the average visual range within the Continental
United States; typically, only a small percentage of the hours involve storms or fog.
In the arid West, the contribution of windblown dus't to degradation of visual air quality
is an important problem. Because human activities that disturb natural soil surfaces add sig-
nificantly to windblown dust, dust storms are only partially natural phenomena.
9-64
-------
DAYS

UNDER 5
5-10
10-20
20-40
OVER 40
Figure 9-29. Average annual number of days with occurrence of dense fog. Coastal
and mountainous regions are most susceptible to fog.

Source: Conway (1963).
Figure 9-30. Annual percent frequency of occurrence of wind-blown dust when
prevailing visibility was 7 miles or less, 1940-1970. Dust is a visibility problem
in the Southern Great Plains and Western desert regions.

Source: Adapted from Orgill and Sehmel (1976).

9-65
-------
The suspension of particles from the surface is determined by cohesiveness of the
particles to the underlying material, the force of the surface wind, and the topography of the
surface layer. The ideal situation leading to suspension of surface material is a dry,
crumbling, or disturbed crust in flat terrain without vegetation. Agitation of such surfaces
by strong winds and turbulence can transform a pristine arid atmosphere into a dust storm with
severely reduced visibility. Suspended crustal material in a dust storm usually consists of
coarse solid particles with volume mean diameters of tens of micrometers or more. Patterson
and Gillette (1977) found that the optically important fugitive dust particles include those
up to 40 pm in diameter, Orgill and Sehmel (1976) have analyzed the frequency of occurrence
of dust storms in the Continental United States in great detail, based on National Weather
Service observations of windblown dust and sand associated with visibility of 7 miles or less,
The peak hours for dust are noon to 8 p.m., during the period of maximum thermal turbulence.
Forested, coastal, and mountainous regions have few, if any, episodes. The Pacific coast has
high (>0.1 percent) incidence of dust only in the San Joaquin Valley and the Los Angeles Basin
Western desert areas in eastern Washington, western Nevada, Utah, New Mexico, and Arizona are
also prone to dust. The highest dust frequency is in the southern Great Plains, where
windblown dust is a serious problem up to 3 percent of the time (Figure 9-30).
9.5 THE EVALUATION OF IMPAIRED VISIBILITY
In the previous sections, fine particles have been established as a primary cause of
visibility degradation. Further, manmade emissions are the greatest contributors to levels of
fine particles. In this section, evaluation of visibility as an important component of public
welfare will be discussed. The evidence clearly shows that poor visibility lessens the
quality of life, that it presents a hazard to transportation, and that people are willing to
pay to improve it.
Although the Value of visibility may be intangible, it can be measured to some extent.
From the lack of consistent results for measuring aesthetic qualities such as visibility,
values can be categorized according to psychological, social/political, or economic criteria.
Psychological criteria relate to individual need and benefit associated with visibility. The
U.S. Environmental Protection Agency (1979) reported that certain psychological benefits, real
or perceived, are associated with a person's wish to preserve the option for a clear view of a
scenic area and with the knowledge that certain pristine areas exist, regardless of any intent
to visit those areas. The psychological benefits have been little studied and will not be
discussed further.
Social and political criteria for evaluating visibility relate to community opinions and
attitudes about visibility. These have typically been examined by surveys or questionnaires
to determine whether members of a community hold visibility to be a public good (a
nonexclusive asset, consumption of which by one individual does not preclude consumption by
another). The third category of criteria for evaluation is economic. Both surveys of

9-66
-------
willingness to pay for good or improved visibility and studies of property value differentials
as a measure of willingness to pay have been employed to determine what monetary value should
be assigned to good or improved visibility. The latter two criteria for evaluation have been
used in several studies and are discuss-ed later in thi>s section.
In addition to the aesthetic properties of good or improved visibility, inconveniences or
hazards to ground and air transportation have been associated with reduced visibility caused
in part by fine PM. These effects have both social and economic aspects and are also
discussed.
9.5.1 Social Awareness and Aesthetic Considerations
Reduced visibility is an aesthetic effect that may occur at pollution levels below those
that cause measurable health and other welfare effects. Although these aesthetic effects are
difficult to measure, several studies have shown that the public considers local air pollution
a nuisance and is willing to pay to reduce it. Although the importance of visibility relative
to other air pollution effects has not generally been examined in most of these studies, both
general expectations and some studies suggest that visibility has a major influence on the
social awareness of air pollution.
According to Barker (1976), for example, the proportion of the public bothered by air
pollution increases as the level of PM increases. Unlike most gaseous pollutants, which can be
perceived only at high concentrations, PM causes a variety of physical stimuli at lower
levels, such as direct smoke emission and visibility reductions, and therefore is a sensitive
indication of pollution.
The 1969 Air Quality Criteria Document (AQCD) for PM recognized the importance of public
concern over pollution by PM. Although research in this area was limited at that time, it
still pointed up a heightened social awareness about air pollution and a willingness "to act
to abate the nuisance." The 1969 AQCD for PM singled out a study in St. Louis, Missouri (U.S.
Department of Health, Education, and Welfare, 1965; Schusky, 1966; Williams and Bunyard, 1966),
to serve as an example of the association between public opinion and air pollution. Within
o
the approximate range of 50 to 200 ug/m PM (measured as TSP), an expression was derived to
relate the percentage of the St. Louis population concerned about air pollution (y) and the
annual geometric mean of PM levels (x)
y = 0.3x - 14 (9-17)
Accordingly, at levels of 80, 120, and 160 ug/m3 TSP, about 10, 20, and 33 percent,
respectively, of the public were bothered by the pollution. The same studies indicated that
the public became aware of pollution before it is regarded as a nuisance, with 30, 50, and 75
percent indicating awareness at the same PM levels noted. Booz, Allen and Hamilton (1970)
also confirmed this awareness by noting that there was a higher proportion of residents in
high TSP-level areas compared with residents of low pollution areas who believed their neigh-
borhoods were dirty.
9-67
-------
The degree of concern by the public varies according to several factors: the nature of
the pollutants, the extent of exposure, and perceiver characteristics (e.g., age, education,
occupation). Wall (1973) examined public perception of air pollution in three British com-
munities. These communities were similar in that they were coalfield localities, they were
sites of significant air pollution problems caused by domestic coal use, and were the focus of
a national effort to improve air quality. Findings were based on respondents' definition of
air pollution and air pollutants. Virtually all those included in the sample offered
definitions of air pollution and pollutants that indicated their awareness of the problem.
Wall noted that health effects of air pollution were rarely included in the definitions,
although other, more visible, effects were mentioned. Wall concluded that people were much
more aware of particulate pollution than gaseous pollution.
Wall also attempted to determine what behavioral adjustments people would make when faced
with a threatening air pollution episode. Although many respondents indicated they favored
"direct" action, (e.g., staying indoors or closing windows), 26.7 percent were unsure of what
a person could do or felt they could do nothing. One behavioral choice was complaint to a
third party. Although this was a common response, Wall noted that the actual ratio of
"complaint potential to complaint performance" was low. He explained that "many people do not
know who to approach and have doubts about the concern and efficiency of those in authority."
Flachsbart and Phillips (1980) offered a comprehensive analysis of human response to air
quality. One result of this empirical investigation was the development of an observer-based
air quality index (OBAQI), which incorporated three air quality measures demonstrated to
relate most significantly to reported observations of air quality. Visibility was one of
three highly significant parameters in Los Angeles County, the site of the study.
In testing the effectiveness of this perception-based index, the author found that people
accept a certain level of air pollution before defining it as "smoggy," the sixth value of a
seven-value scale ranging from extremely clean to extremely smoggy. They also found that
residents of areas with relatively clean air may be more sensitive to changes in air quality
than people accustomed to pollution.
9.5.2 Economic Considerations
The value of the "wilderness experience" and the importance of preserving the natural
heritage have long been recognized in the United States (U.S. Environmental Protection Agency,
1979). Long-range visibility, particularly in areas like the Southwest with its natural
vistas, may be what economists call a public good: an asset that is non-exclusive (U.S.
Environmental Protection Agency, 1979; Fox et a!., 1979). Ecbnomists have made some progress
toward quantifying the values of visibility using dollars as a measure.- According to Rowe and
Chestnut (1981) several economic methods are being applied and developed to estimate the value
of improvements in visual quality from air pollution control. These techniques, referred
to as visibility benefit analyses, can be used to estimate the dollar value of changes in

9-68
-------
visibility. Analyzing the economic effects of improved visibility requires an understanding
of the benefits and their monetary measures. The concepts of visibility as an economic good
and of monetary benefits from visibility improvements may be classified through application of
consumer demand theory. Consumer demand theory asserts that an individual derives well-being,
or what economists call utility, from the consumption of goods and services and that any
change affecting utility, such as improved visibility at a favorite park, has a value. The
monetary measure of benefits for an improvement in visibility indicates how much the increase
in utility is worth. Analyzing how individuals react to changes in visual quality may reveal
the value they place on these changes. A visibility benefit analysis attempts to determine
what individuals would be willing to pay for a change in visual quality if it were possible to
purchase it.
Two approaches have been used to measure the value of visibility: (1) asking people to
establish values directly through the use of hypothetical, or contingent, market situations;
and (2) using actual market and air quality data to determine the relationship between price
and visibility. The usual version of the contingent market approach is the bidding method, in
which participants are asked to indicate their maximum willingness to pay (WTP) or minimum
willingness to accept compensation (WTA) to obtain or prevent a change in air quality. In the
market approach to measuring the value of visibility, studies are based on the supposition
that if air quality varies across an area and if people are willing to pay more for a resi-
dence with better air quality, the amount they are willing to pay can be inferred from the
price differences between properties when all other influences except air quality are ac-
counted for.
The results of the dozen or so visibility benefit analyses that have been conducted, in-
cluding both contingent and actual market approaches, show that the value people place on good
or improved visibility is substantial. Although the circumstances of each study differ, the
results are broadly consistent, providing evidence that estimation procedures are valid (Rowe
and Chestnut, 1981).
Using the first approach, Brookshire (1979) explains the iterative bidding technique as
"a direct determination of economic values from data which represent responses of individuals
of contingencies posited to them via a survey instrument." The iterative bidding technique in
its current form was first developed and applied by Randall et al. (1974) in the Four Corners
region of the Southwest. Three contingencies were considered: (1) limited visibility reduc-
tions and a view of a power plant with limited visible emissions; (2) moderate emissions from
the plant, moderate visibility reductions, and moderate existence of unreclaimed spoil banks
and transmission lines; and (3) extensive emissions, visibility reductions, and unreclaimed
spoil banks and transmission lines. Unfortunately, this selection of scenarios prohibits dis-
aggregation of results into component values for visibility, power plant location, and unre-
claimed spoil banks and transmission lines. A mean reduction in sales tax of $85 per household

9-69
-------
was required to make bidders accept scenario 3 instead of scenario 1; and reduction of $50 was
required to accept 3 in place of 2 (U.S. Environmental Protection Agency, 1979}. Biases are
associated with these techniques; however, no bias tests were conducted in this experiment.
Because bids were requested without actual payment, there may have been incentives to misrep-
resent one's bids, or the questions may have distorted responses.
Three other studies employing the iterative bidding technique in the Southwest are
reviewed here: the Lake Powell study, by Brookshire et al. (1976); the Farmington study, by
Rowe et al. (1980a,b); and the South Coast Air Basin study, by Brookshire et al. (1979; 1980).
The Lake Powell study (Brookshire et al., 1976), conducted in 1975, considered the visual
impact of large power plants and visible smoke plumes in view from a predominantly recreational
site, Recreationists and a few local residents were asked what they would be willing to pay
to prevent construction of an additional plant if only the plant would be visible and if both
the plant and pollution would be visible. Average bids by users of the recreation area, in
terms of additional user fees per day, were between $0.87 and $2.11 to prevent an additional
plant and between $1.75 and $3.38 to prevent an additional plant and pollution. The lowest
average bids were made by residents, while the highest average bids were made by remote
campers. Aggregate bids indicate that the benefits of preventing visibility degradation at
that site alone (there are also several other recreation areas in the vicinity) were $400,000
to $700,000 yearly.
The Farmington study (Rowe et al., 1980a,b), conducted in 1977, was concerned with the
impact of coal-fired electric generation in the Southwest. The focus of this study was the
perception of reduced visibility by local residents in and near Farmington, New Mexico. Three
visibility levels were illustrated by photographs of long-distance landscape views from
Farmington showing (1) visual range of about 120 kilometers (75 miles), somewhat better than
current conditions, (2) visual range of about 80 kilometers (50 miles); and (3) visual range
of about 40 kilometers (25 miles). The average monthly bid by residents to prevent deteriora-
tion in visual range from 120 kilometers (75 miles) to 80 kilometers (50 miles) was $4.75,
and the bid to prevent deterioration from 120 kilometers to 40 kilometers (25 miles) was
$6.50. Nonresident recreationists in the area were willing to pay an average of $3.00 and
$4.00 in additional user fees per day for the same scenarios. Aggregate benefit estimates for
the study area over 35 years assuming a 10 percent discount rate, were $14.2 million for
preventing deterioration of condition 1 to 2 and $19.2 million for preventing deterioration
condition of 1 to 3. One of the purposes of this study was to test for biases in the bidding
process. Hypothetical, starting-point, and information biases were detected. Additionally,
it was found that, contradictory to theoretical expectations, WTA bids were much larger than
WTP bids. This outcome was attributed to differences in implied property rights in the two
questionnaire procedures used. Rowe et al. (1980a,b) also reviewed a number of iterative
bidding and property value differential studies as they might be applied to visibility values.

9-70
-------
The South Coast Air Basin study (Brookshire et al., 1979, 1980), conducted in 1978, was
an application of the bidding method and the property value approach to the same urban areas
to compare the results of these two techniques. The scenarios used in the bidding method were
illustrated with photographs of two views in Los Angele*. showing (1) poor air quality as a
visual range of about 2 miles, typical in much of the area; (2) fair air quality as a visual
range of about 12 miles, the predominant condition in the area; and (3) good air quality as a
visual range of about 28 miles. Residents in 12 different communities, each categorized as
having poor, fair, or good air quality, were interviewed. Average monthly bids" in communities
with poor air quality were between $11 and $22 to obtain fair conditions. In communities with
fair air quality, average monthly bids to obtain good conditions were between $5 and $28. In
communities with good air quality, residents offered average bids of $18 to $67 per month to
obtain good air quality in the entire region. The mean average household bid was approxi-
mately $30 per month for a 30-percent improvement in air quality. For all proposed changes,
aesthetic, acute health, and chronic health components each constituted about one third of the
bid. Brookshire et al. (1979) also reported the difference in property values based on
aesthetic consideration afforded by air pollution in six pairs of neighborhoods in the South
Coast Air Basin of Southern California. In this study, the WTP based on property value
differentials was about $40 per month per household for a 30-percent improvement in air
quality. This amount is comparable with the iterative bidding results of $30 per month. For
both survey and property value studies, the estimates ranged from $20 to $150 per month per
household. Additionally, 22 to 55 percent of the aggregate bids for all areas were for
aesthetic effects, the major component of visibility evaluation noted also by Flachsbart and
Phillips (1980) for the same region.
According to Rowe et al. (1980a,b), while much progress has been made in visibility bene-
fits valuation, much more work is needed. Early studies did not establish the link between
reduced visibility and actual physical parameters of visibility. In addition, other variables
such as health or other welfare effects also confound the valuation of visibility benefits.
(Some of these effects are discussed in Chapter 10.) Certain inherent variables may confound
estimating the value of visibility based solely on site value. For example, the perception of
visually dirty air may trigger an association with unhealthy air, thereby confounding aes-
thetic effects with health effects. Pollutants other than PM may also confound the benefits
analyses. Finally, studies of the economic value of visibility have only begun to address the
urban situation.
9.5.3 Transportation Operations
This section discusses effects on transportation safety and convenience. Both automobile
and aircraft safety and operations may be affected by pollution-related reductions in visual
range. Highway engineers use a concept called Recommended Sight Distances (RSD) to determine
visual range requirements for safe driving, passing, hill grading, etc. At recommended

9-71
-------
speeds, these distances provide adequate time for driver reaction and stopping. At 70 miles
per hour (113 km/hr) the RSD is about 400 yards (365 m). At 55 miles per hour (88 km/hr), it
is under 300 yards (274 m). Even with very high levels of humidity, which can greatly inten-
sify the effect of pollution upon visual range (U.S. Environmental Protection Agency, 1979),
very high pollutant concentrations are required to reduce visual range to less than one-fourth
mile (440 yards; 400 m). For example, concentrations exceeding 1150 ug/m of fine mode
particles (< 2.5 pm) would be required to reduce visual range to less than 440 yards (400 m)
at 90 percent RH for a highly scattering and absorbing aerosol, as depicted in Figure 9-18.
Improved visibility may result in air travel that is safer and involves fewer delays. Air
traffic is controlled using either Instrument Flight Rules (IFR) or Visual Flight Rules (VFR).
Generally when visual range (roughly measured as the average over three directions) falls to
below 3 (4.8 km) miles and there is less than a 1000-foot (300 m) ceiling, air travel is
judged unsafe in controlled airspace using VFR, and only IFR-rated pilots and planes may fly.
Commercial aircraft continue to operate, but certain general aviation flights are grounded.
Between 70 to 80 percent of general aviation aircraft are equipped to fly IFR (Federal
Aviation Administration, 1981a), but only 35 percent of certified general aviation pilots are
instrument rated (Federal Aviation Administration, 1981b). Out of 357,500 total certified
general aviation private airplane pilots, 36,500 (10 percent) are instrument rated (Federal
Aviation Administration, 1981b). General aviation (fixed-wing aircraft) logged approximately
40 million hours of flight time in 1979 distributed according to the following use (in
millions of hours): personal -v9.2; business ~8.7; instructional M5.4; executive -^4.7; rental
~4.1; air taxi <3.6; aerial application -v2.1; industrial M3.8; and other M3.8 (Federal
Aviation Administration, 1980b). Personal and rental amounted to 13.3 million hours, or 33
percent of the total general aviation hours. Since only 10 percent of the private pilots are
instrument rated, it follows that about 30 percent of general aviation flights operating in
controlled airspace would likely be grounded during visibilities less than 3 miles (4.8 km).
A U.S. Senate staff report (U.S. Congress, 1963) noted that air pollution was both
hazardous and costly to aircraft operations. Recognizing that fog slows air traffic, the
author noted that the addition of pollutants to the fog often reduced visibility enough to
ground aircraft not equipped with blindflying instruments and to delay traffic at busy
airports. Additionally, the report cited a review by the Civil Aeronautics Board of 1960
record cards, representing one-third of all U.S. aircraft accidents in 1962, and attributed
six accidents to obstruction of vision by smoke, haze, sand, or dust. Two of the six planes
were classified as large, and one was a commercial carri'er. The report noted that if the
sample study was representative, 15 to 20 plane crashes in 1962 could have been linked to poor
visibility. The staff report stated:
9-72
-------
Pollution contributes to fog formation and undoubtedly aggravated visibility
problems in which natural conditions play a major role. Where accidents are
attributed to impaired visibility due to weather, it is likely that in many
instances the risk had been increased by air pollution.

A 1967 report on air pollution abatement activity concluded that "interstate air pollu-
tion in the New York - New Jersey metropolitan area results in visibility restrictions en-
dangering the safety of persons in interstate travel, both by land and by air, and it causes
inconvenience and economic loss to the public and to transportation companies, due to disrup-
tion of traffic schedules" (National Center for Air Pollution Control, 1967).
The Federal Aviation Administration (1978) in the Airmen's Information Manual notes that
failure to see and avoid obstructions during flight is one of the 10 factors most frequently
associated with aircraft accidents. The manual notes that caution should be exercised when
visibilities are within the 3 to 4 mile (4.8 to 6.4 km) limit. The National Transportation
Safety Board (National Transportation and Safety Board, 1978b) also ranked weather as a major
factor related to total United States certified air carrier accidents from 1969 to 1978. In
1978, briefs of 928 total and 322 fatal accidents involving weather as a cause/factor list low
ceilings and visibilities as probable causes and factors associated with the fatal accidents
the majority of times (National Transportation and Safety Board, 1978b).
The Air Traffic Service of the Federal Aviation Administration has developed a perform-
ance measurement system for many major terminals (Federal Aviation Administration, 1980a). In
this system standards were set for individual airports according to runway configuration and
visibility levels. Under IFR and VFR conditions, the number of operations per hour (arrivals
and departures) are tabulated according to runway configuration. At most of the major
airports covered in the report, significant reductions in operations per hour occurred during
IFR conditions, i.e., when visual range is below 3 miles (4.8 km). In addition some airports
may require nonvisual approaches at visibilities between 3 and 5 miles, depending on the
minimum vectoring altitude (about 1000 to 1500 ft [300 to 460 m] ceiling). In this case, the
arrival and possibly the departure operations may be affected marginally at visibilities
greater than 3 miles (4.8 km).
To determine how often visibility is reduced to below 3 miles at airports throughout the
country, information from the National Weather Service historical data base was compiled and
plotted for 147 sites in the United States. Because only half mile increments are reported
for the 2- to 3-mile (3.2 m to 4.8 km) range, 3-mile (4.8 km) or less range is reported here.
The data consisted of local standard time midday readings (1200, 1300, or 1400 hours) for six
5-year periods: 1951-1955, 1956-1960, 1961-1965, 1966-1970, 1971-1975, and 1976-1980. Addi-
tionally, 3-month quarters were reported separately within the 5-year periods. Table 9-4
shows the frequency of poor midday visibility (i.e., 3 miles [4.8 km] or less) in the north-
eastern, southeastern, or midwestern regions of the United States. Except for the West Coast
and Texas, the frequency of poor midday visibilities of 3 miles (4.8 km) or less in most areas
9-73
-------
TABLE 9-4, SEASONAL AVERAGE PERCENT OF TIME WHEN MIDDAY VISIBILITY WA<
3 MILES (4.8 km) OR LESS AT U.S. AIRPORTS FROM 1951 to 1980e

Quarter 1
( Jan. -Mar. )
all good ,
Location weather weather

Northeast 16.2 2.4
Southeast 12.5 1.1
^ Midwest 19.8 4.9
.£.
Quarter 2
(Apr. -June)
all good .
weather weather
7.6 2.1
5.6 1.9
6.7 2.5
Quarter 3
(July-Sept. )
al 1 good .
weather weather
8.5 4.8
7.7 4.7
6.0 3.5
Quarter 4
(Oct. -Nov.)
all
weather
12.0
9.4
15.6
good .
weather
2.3
1.3
4.9
a Based on daily measurements over the entire 30 years. Data from the U.S. Weather Service.

Absence of fog, precipitation, or windblown material such as snow, sand, or dust.
-------
of the Western United States is low. Data are reported for poor visibility under all weather
conditions and under good weather conditions (i.e., in the absence of fog, precipitation,
blowing material). Under good weather conditions, poor visibility occurred most frequently in
the summer quarter, accounting for 60 percent of all instances of poor visibility in this
season. At other times of the year (Quarters 1, 2, and 4), poor visibility during good
weather represented only 20 percent of the total number of instances of poor visibility.
Figures 9-31 and 9-32 show the location of the sites and percent of times and occurrences
during quarter 3 (July-September) that visibility was 3 miles (4.8 km) or-less for the two
most recent 5-year periods, 1971-1975 and 1976-1980. Each quarter included approximately 460
measurements. In these figures, neither fog, precipitation, nor blowing material (e.g., sand
and dust) was the cause of the incidences of reduced visibilities. For all six 5-year periods
(1951-1980) Table 9-5 lists the occurrences of 3-mile (4.8 km) or less visibilities at midday
for all causes and for causes attributed primarily to pollution (i.e., during periods when no
fog, precipitation, or blowing material was present) for the summer quarters for 26 airport
sites in the United States. In most cases, the percent of occurrence of low visibility
reached a maximum from 1966 to 1975. Data from the National Weather Service also show that RH
greater than 90 percent occurred rarely during midday at most locations throughout the United
States. Examination of Table 9-5 leads to the conclusion that midday visibility reductions to
less than 3 miles (4.8 km) can occur in summer months more than 50 percent of the time from
causes other than fog, precipitation, or blowing material, and around 40 percent of the time
when RH is about 75 percent. Presumably, then, these instances of reduced visibility are
attributable to light-scattering or light-absorbing air pollutants (i.e., fine-mode PM).
From the limited information available, it was concluded that during periods of peak air
traffic fewer planes arrive and depart when visibility is below 3 miles (4.8 km) (IFR
conditions) than when visibility is greater than 3 miles (4.8 km) (VFR conditions). More than
half of the time that IFR conditions are in effect duririg the summer months, the weather is
good and the RH is below 90 percent. At such times, fine-mode PM is probably responsible for
the IFR conditions. The airport data listed in Table 9-5 indicate that poor midday visibility
(equivalent to IFR conditions) occurred an average of 7 percent of the time during the summer
quarter, from 1951-1980 in the Northeast, Southeast, and Midwest. It follows that more than
half of the time that visibility was poor enough to slow midday air traffic at some airports
in these regions, fine-mode PM was the cause. Thus, the overall frequency of poor airport
midday visibility related to fine-mode PM was about 4 percent. By the same line of reasoning,
fine-mode PM could have impeded midday air traffic only about 2 percent of the time annually.
9.6 SOLAR RADIATION
Incoming solar radiation is composed of the direct bean and the diffuse sky light arising
from the light-scattering atmosphere (Figure 9-33; Gates, 1966). The relative contribution of
the sky light is least at noon and greatest at sunrise and sunset. At sea level, and for a
clean atmosphere, sky light contributes at least 10 percent of the total radiation.
9-75
-------
Quarters 1971-75 Midday Observations
% of Times Vis O3 Miles
No Fog, Precipitation or Blowing Material
L Laguard
J JFK
N Newark
Quarters 1971-75 Midday Observations
% of Occurrences Vis < » 3 Miles
No Fog, Precipitation, or Blowing Material
Figure 9-31, Percent of daily midday measurements (1971-75) in which visibilities were three
miles or less in the absence of fog, precipitation or blowing material.

Source: Adapted from the Historical Data Base of the National Weather Service 1981.

9-76
-------
Quarter 3 1976-80 Midday Observations
% of Times Vis < = 3 MrI0s
No Fog, Precipitation, or Blowing Material
Quarter 3 1976-80 Midday Observations
% of Occurrences Vis < = 3 Miles
No Fog, Precipitation, or Blowing Material
N Newark
Figure 9-32. Percent of daily midday measurements (1976-80) in which visibilities were three
miles or less in the absence of fog, precipitation, or blowing material.

Source: Adapted from the Historical Data Base of the National Weather Service 1981,

9-77
-------
TABLE 9-5. PERCENT OF VISIBILITY MEASUREMENTS AT 3 MILES (4.8 km) OR LESS
AT 26 U.S. AIRPORTS DURING THE SUMMER QUARTER3
Location
to
1
•-J
oo
Period
Percent ofvisibilitymeasurements less than or equal to 3 miles
In the presence of
fog, precipitation, In the absence of fog,precipitation,
or blowing material ^ or blowing material
All DUtp"
AH
All RHV
RH < 75%
RH < 60%
Northeast
New York, NY (LaGuardia)

New York, NY (JFK)

Newark, NJ

Pittsburgh, PA (GP)

1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1975-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1975-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980

5.9
8.3
9.4
14.1
8.0
4.1
4.8
9,1
6.7
12.0
9.1
7.4
6.1
11.3
7.6
10.4
5.0
5.0
3.4
5.2
4.3
10.0
12.2
10.5

4.8
5.2
6.5
10.0
3.7
2.8
4.1
5.4
4.3
8.3
6.1
4.8
3.3
5.4
4.1
7.2
2.2
1.7
1.4
2.8
1.5
6.5
7.0
5.0

3.7
4.3
5.7
8.5
2.4
2.2
3.0
4.1
3.3
6.1
4.3
3.5
2.4
4.1
3.3
6.5
1.5
1.5
1.4
2.2
0.9
6.1
6.3
3.7

2.0
1.5
3.1
4.3
1.3
1.1
1.5
2.0
1.1
1.7
1.1
1.3
0.9
1.7
1.1
2.6
0.4
0.4
0.7
0.7
0.2
2.8
4.1
1.3
-------
TABLE 9-5. (continued)
10
I

Location
Albany, NY

Buffalo, NY

Providence, RI

Southeast
Montgomery, AL

Period
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980

1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
Percent of visibil
In the presence of
fog, precipitation,
or blowing material
All RH'sD
5.9
3.9
5.4
14.8
19.8
13.7
3.5
4.3
6.5
8.7
15.0
10.7
4.6
8.7
9.1
12.4
9.8
5.0

0.9
1.3
0.4
4.3
10.2
9.6
ity measurements less than or equal to 3 miles

In the absence of fog, precipitation,
All RH'sc
1.5
0.4
1.7
10.4
13.5
8.7
1.3
1.5
2.8
5.4
10.0
6.1
2. a
2.8
3.3
8.9
5.4
2.2

0.0
0.0
0.0
2.2
7.8
8.1
or blowing material
RH < 75%
1.3
0.2
1.7
9.8
13.0
7.2
0.9
1.1
2.4
4.6
9.3
4.1
2.4
1.3
1.3
6.7
3.7
1.3

0.0
0.0
0.0
2.0
7.4
7.6
RH < 60%
0.2
0.0
0.7
5.7
7.0
3.9
0.7
0.2
0.2
1.7
3.5
2.0
0.4
0.0
0.4
1.5
1.1
0.7

0.0
0.0
0.0
0.9
3.9
5.7
-------
TABLE 9-5. (continued)
I
oo
o

Location
Richmond, VA

Norfolk, VA

Greensboro, NC

Atlanta, GA

Period
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
Percent of visibil
In the presence of
fog, precipitation,
or blowing material
All RH's
3.9
5.4
7.4
18.3
28.0
22.0
2.8
4.6
7.4
15.4
20.4
12.6
2.4
4.6
4.1
6.3
16.5
9.2
0.2
3.1
2.2
3.0
5.4
4.1
ity measurements less than or equal to 3 miles

In the absence of fog, precipitation,
All RH'sc
0.9
2.4
2.0
13.7
21.3
16.1
0.2
1.3
3.7
11.7
15.2
9.8
0.4
0.7
2.2
3.9
12.0
6.1
0.0
0.7
0.4
1.3
3.9
1.1
or blowing material
RH < 75%
0.4
1.7
1.7
12.0
19.6
14.8
0.2
0.9
2.4
9.8
13.3
9.6
0.2
0.2
0.9
2.8
10.0
4.6
0.0
0.4
0.4
0.9
3.7
1.1
RH < 60%
0.0
0.4
0.9
6.7
10.7
8.7
0.2
0.0
0.2
4.6
3.9
6.1
0.0
0.0
0.2
1.1
6.1
2.8
0.0
0.2
0.0
0.4
1.5
0.9
-------
TABLE 9-5. (continued)
I
03

Location
Bristol, TN (Tri-City)

Memphis, TN

Baton Rouge, LA

Washington, DC (National)

Lexington, KY

Period
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
Percent of visibil
In the presence of
fog, precipitation,
or blowing material
All RH'sD
0.4
0.9
1.3
4.8
12.4
9.1
0.7
1.5
1.1
1.3
3.9
5.2
1.3
2.4
2.4
4.3
6.5
8.3
2.8
3.5
3.5
2.8
4.8
3.7
0.9
2.4
2.6
5.7
8.7
12.4
ity measurements less than or equal to 3 miles
In the absence of fog, precipitation,
All RH'
0.0
0.0
0.2
2.6
9.6
6.5
0.2
0.0
0.0
0.7
3.0
3.3
0.9
0.7
0.4
2.4
3.9
3.7
1.1
1.5
1.1
1.3
2.6
2.0
0.0
0.7
0.4
2.0
5.2
9.3
or blowing material
sc RH < 75%
0.0
0.0
0.0
2.4
8.0
6.3
0.2
0.0
0.0
0.2
3.0
3.1
0.7
0.4
0.4
2.2
3.5
3.5
0.4
0.9
0.9
1.1
2.6
2.0
0.0
0.7
0.2
1.7
4.1
0.0
RH < 60%
0.0
0.0
0.0
1.1
2.0
2.4
0.0
0.0
0.0
0.0
2.0
2.0
0.0
0.4
0.2
0.9
1.5
1.5
0.4
0.2
0.0
0.2
1.7
0.7
0.0
0.4
0.0
1.3
2.4
3.5
-------
TABLE 9-5. (continued)

Percent of visibility
measurements less than or equal to 3 mi
les
In the presence of
fog, precipitation, In the absence of fog, precipitation,
Location
or blowing material
Period All RH'sD All
or blowing material
RH's RH < 75% RH < 60%

Midwest

Detroit, MI (City)
CO
PO
Chicago, IL (Midway)
Chicago, IL (O'Hare)
St. Louis, MO
1951-
1956-
1961-
1966-
1971"
1976-
1955
1960
1965
1970
1975
1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980

1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-
1956-
1961-
1966-
1971-
1976-
1955
1960
1965
1970
1975
1980
3.9
7.8
8.0
12.8
8.3
7,4

2.6
5.9
7.6
5.7
6.5
8.2

4.9
5.9
4.3
5.9
3.9

2.2
1.7
4.3
8.9
6.3
6.1
2.0
4.8
4.1
9.3
5.0
4.1

1.5
4.1
3.7
4.1
4.6
5.4

4.3
2.0
3.0
2.8
2.0

0.9
0.7
2.0
3.9
3.5
3.0
1.3
3.7
2.8
8.0
3.9
3.7

1.1
3.3
2.8
2.8
3.0
3.0

3.3
1.5
2.2
2.0
0.9

0.7
0.4
1.3
3.5
2.8
2.8
1.3
1.5
1.1
5.2
1.5
2.2

0.2
1.5
1.1
2.0
1.7
2.2

1.6
0.9
0.9
0.9
0.4

0.4
0.0
0.7
2.2
1.3
2.0
-------
TABLE 9-5. (continued)
I
CD
CA!

Location
Southwest
Dallas, TX

Dallas/Ft, Worth, TX

West
Denver, CO

Los Angeles, CA (Int'l.)

Period
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1080
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980
1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1080
Percent of visibil
In the presence of
fog, precipitation,
or blowing material
All RH'sD
0.0
0.0
0.7
1.1
1.8
0.0
0.0
0.4
1.3
1.5
0.9
0.9
0.4
1.3
0.4
0.9
0.2
20.9
10.4
17.0
18.5
14.1
8.8
ity measurements less than or equal to 3 miles
In the absence of fog, precipitation,
All RH'sc
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.7
0.2
0.0
0.0
0.0
0.0
0.0
0.0
20.7
10.4
17.0
18.3
13.7
8.6
or blowing material
RH < 75%
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.4
0.2
0.0
0.0
0.0
0.0
0.0
0.0
18.9
9.1
15.2
15.2
11.7
5.5
RH < 60%
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.2
0.2
0.0
0.0
0.0
0.0
0.0
0.0
2.0
5.2
0.2
1.1
0.0
0.2
-------
TABLE 9-5. (continued)

Percent of visibility measurements less than or

ID
03

Location Period
Long Beach, CA 1951-1955
1956-1960
1961-1965
1966-1970
1971-1975
1976-1980

a Average number of observations during each
September).
In the presence of
fog, precipitation,
or blowing material
All RH'sD
21.3
12.0
17.4
18.7
10.4
6.5

5-year period was 460

In the
All RH's
20.9
12.0
17.2
18.7
10.4
6.1

equal to 3 miles

absence of fog, precipitation,
or blowing material
c RH < 75%
20.2
11.7
17.2
18.3
10.4
6.1

for the summer quarter (July,

RH < 60%
14.1
9.6
11.1
16.3
7.2
4.4

August, and

Blowing material refers to dust, sand, or snow.

Table 9-2 shows that midday RH greater than 90 percent occurs rarely; moreover, in the absence of
fog or precipitation, it occurs even more rarely.
-------
GLOBAL RADIATION

DIRECT SOLAR
\j v.' ^ i i \ i T •^ i i w*"^ i ~~ j ^m^ia^sai^- • i r-~^
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

WAVELENGTH, jU.m

Figure 9-33. Solar radiation intensity spectrum at sea level in
cloudless sky peaks in the visible window, 0.4-0.7 Mm wavelength
range, shows that in clean remote locations, direct solar radiation
contributes 90 percent and the skylight 10 percent of the incident
radiation on a horizontal surface. The airmass, m, is a measure of the
amount of air the direct solar beam has to pass through.

Source: Gates (1966).
9-85
-------
Aerosol layers in the atmosphere scatter and absorb solar radiation (Figure 9-34). Some
of the scattered radiation is directed upward and lost to space; some is directed downward to
the earth's surface. A fraction of the radiation may also be absorbed by aerosols, further
reducing the amount of radiation reaching the surface but at the same time heating the aerosol
layer itself from the earth's surface. Most of the solar radiation eventually reaches the
surface, but its spectral and directional composition (that is, the "quality" of the solar
radiation) may be changed by atmospheric haze (Weiss et a!., 1979).
Because aerosols are not uniformly distributed in the atmosphere, their effects are
spatially nonhomogeneous. First, the horizontal spatial scale encompassing aerosol source,
transport, and removal in the lower troposphere is variable but often about 1000 kilometers
620 miles). The vertical spatial scale of aerosols is also quite variable, but often the
particles are concentrated in a layer from 0.5 to 2 km (0.3 to 0.5 miles) meters deep at the
earth's surface. Hence, the aerosol effects should be concentrated in the lowest layers of
the atmosphere, especially in industrial regions.
Global-scale effects might also occur. If the effects in industrial regions are strong
enough, then the heat balance of the entire earth could be influenced. Additionally, effects
from long-lived aerosol, such as those in the stratosphere, might lead to direct physical
effects on a global scale.
9.6.1 Spectral and Directional Quality of Solar Radiation
Figure 9-35 (McCree and Keener, 1974) shows the spectral quality of solar radiation on a
clear day and on a hazy day in Texas. On the hazy day, the direct solar radiation is reduced
to about one-half of that on a clear day, but most of the energy has reappeared as diffuse sky
light. The net effect is that there is an overall loss of up to about 10 to 20 percent of the
radiation reaching the surface.
If we take the typical backscattered fraction for regional haze aerosols to be 10 percent
of scattering, and the absorption to be also about 10 percent, as suggested by the data of
Weiss et al. (1979), then we can estimate the amount of energy lost from the surface, the
amount lost to space, and the amount absorbed by the atmosphere. On a day with half of the
direct beam transmitted, we conclude that 10 percent of the other half, or 5 percent, is lost
to space, and the other 5 percent results in atmospheric heating. Together, these phenomena
lead to a loss of 10 percent of the radiation. Although it is not possible to calculate
accurately the influence this loss might have on surface temperature, rate of thawing of
frozen ground, growing season, or other climatological measures, it is highly probable that
this loss cools the ground and heats the hazy lower layers of the atmosphere. In turn, it
must increase atmospheric stability, decrease convective mixing, and therefore increase the
rate at which pollutants accumulate.
9-86
-------
BACKSCATTER

ABSORPTION

DIFFUSE
\.
DIRECT
«»/^. j L .ffin r * .frt. ' - * t* * . - 7? - jpnj _ * _ ' _ " _ t _ 7l"*T-^ _ _ u
^xSS^^^^
Figure 9-34. Extinction of direct solar radiation by aerosols is
depicted.
9-87
-------
u
u
z
oc
cc
1.0
0.5
«7 0
f 0.5
1.0
0.5
0.3
HAZY
CLEAR
0.4 0.5

WAVELENGTH,
0.6
SKY
SUN + SKY
0.7
Figure 9-35. On a cloudless but hazy day in Texas, the direct solar
radiation intensity was measured to be half that on a clear day, but
most of the lost direct radiation has reappeared as skylight. However,
there is about 20 percent of the solar radiation missing on the hazy
day, some absorbed, and some baclcscattered to space.
Source: McCree and Keener (1974).
9-88
-------
No detailed and routine measurements of the quality of solar radiation are available for
the United States, however, the total solar energy reaching the surface is monitored routinely
at many meteorological observation sites in the United States and worldwide. Unfortunately,
the large variability of such data does not allow manma4e aerosol effects to be distinguished
from the effects of natural sources,
A "data base that gives some information on the quality of solar radiation is the United
States turbidity network, operated at about 40 stations in the country since 1961 (Flowers et
al. , 1969). If there are no clouds between the observer and the sun, the intensity of direct
solar radiation for a given solar elevation depends on the variable amount of dust, haze, and
water vapor in the atmosphere. The extinction produced by these constituents is called
"atmospheric turbidity." Figure 9-36 (Flowers et al. , 1969) shows the seasonal pattern of
atmospheric turbidity in the United States at 29 sites for 1961-1966. At all 29 sites, the
highest turbidity occurs in the summertime and the lowest occurs in the winter, which is
consistent with the haziness pattern obtained from visibility observations (Section 9.4.2).
The turbidity of the atmosphere in the United States has a strong spatial dependence. In the
Southwest, with a mean annual turbidity coefficient of about 0.06, the incoming direct solar
radiation is attenuated by only 13 percent, compared with Midwest values of about 20 percent.
(In this and subsequent statements, only the reduction due to particles and, to a much lesser
extent, by NO,, is given.) The highest turbidity coefficients were observed in the Eastern
United States where winter values of 0.1 and summer values of 0.2 were typical, meaning that
about 20 to 35 percent of the direct solar beam is diverted to sky light, backscattered to
space, or absorbed. Thus, by earlier assumptions, 2 to 3.5 percent is backscattered to space
and another 2 to 3.5 percent is absorbed into the atmosphere.
Since the first report of Flowers et al. (1969), e.g., the turbidity data have been
reported yearly by the World Meteorological Organization (1977). Comparison of the seasonal
turbidity pattern for 1961-1966 and 1972-1975 is shown in Figure 9-37. Since the mid-I9601s
there has been a further increase of turbidity in the Eastern United States, particularly in
the summer season. Currently the summer average turbidity in the region including Memphis,
Tennessee, Oak Ridge, Tennessee, Greensboro, North Carolina, and Baltimore, Maryland, is about
0.3. This corresponds to a 50-percent attenuation of the direct solar beam on an average
summer day. During hazy episodes, turbidity coefficients of 0.6 to 1.0 are, often reported,
resulting in removal of 75 to 90 percent of the solar radiation from the direct beam, with 7.5
to 9 percent lost to space, and 7.5 to 9 percent lost as atmospheric heating. One of the
consequences of such hazy atmosphere is the disappearance of shadow contrasts. It is strongly
suspected, but has not yet been proved, that there are effects on agricultural productivity.
The spatial distribution and trends of regional-scale turbidity in the eastern United
States are consistent with the observed pattern of haziness obtained through visibility
9-89
-------
Figure 9-36. To Interpret these 1961-1966 monthly average turbidity data in terms of aersol effects on
transmission of direct sunlight, use the expression l/Io = 10"^, where B is turbidity and l/I0 is the fraction
transmitted.

Source: Flowers etal, (1969).
9-90
-------
TURBIDITY TREND
1961 «C ••••«
W72-7B —
ALBANY, NY
i.i.i.i.'.i.i.r.i.i.i.j.
t
S 04
c
3
0.3

0.1
[ I I"I" I * I" iTI ' T
MEMPHIS, TN
0 1 . 1 . 1 . 1 . 1 . 1 . 1 . 1 ...» . I . ! ! ,
OAK RIDGE. TN
ri'i ¥i*i "i r rrrrr
GREENSBORO, NC
,. 1 . 1 . 1. 1 .1 . 1 , 1 .1.1.1.
BALTIMORE, MD
I.I.1.1.1.1.1.1.1.1.1.
JFMAMJJASONO
MONTH
Figure 9-37. Seasonal turbidity patterns for 1961—1966 and 1972—1975 are
shown for selected regions in the Eastern United States.

Source; Flowers et al. (19ii), and WMO (1974 through 1977),
9-91
-------
observations. Both the turbidity and the visibility reduction by haze in the Eastern United
States can be attributed primarily to fine particles (Section 9.2.3.3). Bo!in and Charlson
(1976) suggest that many of these radiative effects are due to sulfates and conclude that the
magnitude of effects is comparable to that summarized here.
9.6.2 Total Solar Radiation: Local to Regional Scale
Changes in the total radiant energy have been observed within urban areas. Early
measurements in central city locations, primarily in Europe, showed levels typically 10 to 20
percent below surrounding rural areas. Robinson (1962) discussed some observations made in
London and Vienna. In London, the deficit was considerably reduced after the implementation
of a Clean Air Act. Measurements on 47 days in autumn 1973 in the Los Angeles area are sum-
marized in Table 9-6 (Peterson and Flowers, 1977). In the St. Louis area, however, smaller

TABLE 9-6. SOME SOLAR RADIATION MEASUREMENTS IN
THE LOS ANGELES AREA3

Measurement
Minimum
Average
Maximum
Total
4
11
20
uv
15
29
44

Values for the daily average percentage decrease of total
and UV solar radiation between El Monte (urban) and
Mt. Disappointment (rural).
Source: Peterson and Flowers (1977).

urban-rural differences were observed. On 12 cloudless days in summer 1972, the average solar
and U.V. fluxes at an urban site were only 3 and 8 percent, respectively, below those at a
rural site about 50 kilometers from the city. The difference between the St. Louis and Los
Angeles and European urban areas appears to involve both decreased urban and increased rural
attenuation, and it may be that neither the city of St. Louis/nor its surroundings over a wide
area modify solar radiation in a manner typical of other locations.
Angel 1 and Korshover (1975) analyzed the solar radiation duration data (hours of sun-
shine) for the eastern half of the United States. Data for 1950-1970 were obtained with
detectors that accumulate the time during which total illumination is above a set threshold.
These are believed to be more reliable for long-term trend analysis than data from recorders
of solar radiation intensity. Angel! and Korshover noted some marked trends: in the
southeast and south central United States, the solar radiation duration has decreased by about
4 to 6 percent; however, the north central area is increasing (Figure 9-38; Angel! and
Korshover, 1975). Although the authors do not attribute these trends to any specific cause,

9-92
-------
o
ec
10
D

65 -2
-4
1940
SOUTH CENTRAL
I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I
50
GO

YEAR
70
80
Figure 9-38. Analysis of the hours of solar radiation since the 1950's
shows a decrease of summer solar radiation over the Eastern United
States. There may be several causes for this trend, including an
increase of cloudiness; some of the change may also be due to haze.

Source: Angell and Korshover (1975).
9-93
-------
it should be noted that there has been an increase in haziness within that period in areas
with decreased solar radiation. It is conceivable, therefore, that increased haziness causes
sunshine-duration detectors to delay the turn-on time in the morning and advance the turn-off
time in the evening. It should also be stressed, however, that changes in the solar radiation
duration may be caused by other natural or manmade phenomena.
9.6.3 Radiative Climate: Global Scale
The attenuation of solar radiation from scattering and absorption by particles in the
atmosphere is probably an important factor in climatic change. The effect could arise from
redirecting the photon energy, from effects on cloud structure (perturbations of the colloidal
system), or from effects on optical properties, such as increased photon retention from absorp-
tion by soot particles. A report by Hobbs et al. (1974) argued that aerosols are most likely
the principal agents by which pollutants may affect weather and climate, by influencing the
structure and distribution of clouds. On a global scale, they noted that the effects of man-
made pollutants on climate are still a matter of debate. Others, such as Twomey (1974), have
pointed up a direct connection between pollution and the number of drops in a cloud and,
hence, an influence on optical thickness and reflectance of the clouds (cloud albedo); conse-
quently, climate is affected.
The importance of PM on climate may be overshadowed by that of COp. A doubling of the
concentration of COp could result in a 1.5° to 3° C warming of the lower atmosphere, according
to various predictions reviewed by the U.S. Department of Energy (U.S. Department of Energy,
1978). One set of calculations frequently cited in the scientific literature predicts a 2° to
3° C rise in the average temperature of the lower atmosphere in the middle latitudes with the
doubling of the COp content of the air and a three- to fourfold greater temperature increase
in the polar regions (Manabe and Wetherald, 1975). Current model estimates suggest that the
earth should have experienced a few tenths of a degree of warming since the late 1880s due to
the increase of COp concentrations from about 290 to the current 335 ppm. However, it appears
either that natural variations currently are large enough to mask the expected C02 effect on
temperature (U.S. Department of Energy, 1980; and Stuiver, 1980) or that models overpredict
the magnitude of the effect.
Effects of volcanic emissions (see Chapter 4) on weather have also been suggested.
Although spectacular sunsets have been associated with major eruptions (e.g., Krakatoa and
Katmai), the effects of scattering of solar radiation from volcanic dust are not clearly
understood. Whether a universal effect is created that can result in cooler weather is still
at issue. To date, surface temperature and rainfall changes are not conclusively related to
volcanic events. The large variability of weather and the self-preserving aspects of the
atmospheric system tend to obscure all but the most dramatic changes in climate.
On local scales associated with urban and industrial areas, any significant attenuation
of radiation by air pollution can, in addition to other well-recognized factors, result in

9-94
-------
changes in local weather (e.g., Landsberg, 1970 and 1981). It is possible that local- and
regional-scale changes in solar radiation caused by human activity may ultimately influence
the heat and water vapor content of the atmosphere on very large scales, but solar radiation
and aerosol levels measured at stations remote from pollutant sources have not as yet
displayed any trend that can be related to human causes (Fischer, 1967; Ellis and Pueschel,
1971; Hodge et al., 1972).
Unfortunately, there is little agreement about whether the net effect of increased air-
borne particle concentrations is the warming or cooling of the earth as a whole. Most models
can predict either an increase or a decrease in the effective albedo of the earth under cloud-
less skies, depending on which combination of surface albedo, sun angle, particle size distri-
bution, and particle complex refractive index is assumed. The effects of clouds are very im-
portant, and the contributions from infrared radiatiori must be considered to obtain a complete
energy budget (Wesely and Lipschutz, 1976).
9,7 CLOUDINESS AND PRECIPITATION
The global cloud cover plays a vital role in the earth's radiative budget in reflecting
energy back to space, in absorbing both solar and longwave (terrestrial) radiation, and in
emitting its energy downward and outward into space. Changes in cloud cover, therefore, alter
the global heat balance. Cloud- and precipitation-forming processes may be divided into two
broad classes: (1) macrophysical processes, which affect the rise and descent of air currents
and the amount of water vapor available for condensation; and (2) microphysical processes,
which affect the nature of cloud particles formed during condensation. The role of atmos-
pheric aerosols, primarily those that are strongly hygroscopic, is to influence the micro-
physics of cloud formation.
On a global or even regional scale, the very small amounts of moisture that man adds by
land practices or combustion of fossil fuels are negligible in comparison with global evapora-
tion. On a regional scale, only one form of increasing cloudiness suggests itself: the
formation of aircraft contrails (Machta and Telegadas, 1974). Aircraft contrail formation
results mainly from the injection of water vapor rather than of aerosols.
In urban areas, inadvertent changes of cloudiness as well as the quantity of precipita-
tion have been well established. Such urban impacts also include the frequencies of thunder-
storms and hail as well as total amounts of rain, Changnon (1968) has reported a notable
increase in days of precipitation, thunderstorms, and hail occurring since 1925 at La Porte,
Indiana. Since La Porte is 30 miles east of the Chicago urban-industrial complex, he proposed
that the increased precipitation results from inadvertent manmade modifications. Figure 9-39
(Changnon, 1968) shows the 5-year running totals of days with smoke- and haze-restricted
visibility in Chicago plotted with 5-year running total precipitation levels at three
stations. This measure of atmospheric pollution has a temporal distribution after 1930 rather
similar to the La Porte precipitation curve. A noticeable increase in smoke-haze days began

9-95
-------
300
OBSERVER ;
CHANGES AT •
LA PORTE /
1910 1920 1930 1940 1950 1960

ENDING YEAR OF 5-YEAR PERIOD

Figure 9-39. Numbers of smoke/haze days are plotted per 5 years at
Chicago, with values plotted at end of 5-year period.

Source: Changnon (1968).
9-96
-------
in 1935 and became more marked after 1940, when the La Porte precipitation curve began Its
sharp increase.
A recent report by Clark (1979) critically reexamined this anomaly historically and tech-
nically through the use^ of a budgetary hydrologic model. Clark proposed several explanations
of the anomaly. If the anomaly was real, it was, at most, a microscale phenomenon), and its
disappearance by the mid-1960s arose from the dissipation of the precipitation-producing
mechanism. He noted that although increases in local precipitation have been reported down-
wind of urban areas, no records exist of any alteration equaling the magnitude of the La Porte
precipitation variation relative to surrounding sites. Additionally, the 1929 to 1963 period
coincides roughly with the tenure of the sole observer. The accuracy of the gauge was also at
issue. Clark concluded that errors by the observer and/or gauge most likely explain the
anomaly.
As part of project Metromex, studies by the Illinois State Water Survey suggest increases
of about 30 percent in rain and 200 percent in thunderstorms and hail at single gauging sta-
tions downwind of St. Louis, with increases of about 10 percent over a two-county area
(Changnon et al., 1977; Ackerman et al., 1978). Here again, the physical causes of the maxima
are not well understood, but they do appear to be associated with perturbations of the plane-
tary boundary layer and enhanced cloudiness, possibly resulting from the addition of aerosols.
It is regrettable that the complex interactions of cloud- and precipitation-forming processes
obscure the specific role of manmade aerosols.
The incorporation of particles into rain and fog droplets can change the characteristics
of precipitation by changing its chemical composition. The most important impact on precipi-
tation water quality is probably that of "acid rain," discussed in more detail in Chapter 7.
9.8 SUMMARY
Traditionally, visibility has been defined in terms of the distance from an object that
is necessary to produce a minimum detectable contrast between that object and its background.
Although visibility is often defined by this "visual range," it includes not only being able
to see or not see a target, but also seeing targets at shorter distances and appreciating the
details of the target, including its colors. Visibility impairment can manifest itself in two
ways: (1) as a layer of haze (or a plume), which is visible because it has a visual discon-
tinuity between itself and its background, or (2) as a uniform haze which reduces atmospheric
clarity. The type and degree of impairment are determined by the distribution, concentrations,
and characteristics of atmospheric particles and gases, which scatter and absorb light trave-
ling through the atmosphere. Scattering and absorption determine light extinction.
On a regional scale, the extinction of light is generally dominated by particle scatter-
ing. In urban areas, absorption by particles becomes important and occasionally dominant.
Extinction by particles is usually dominated by particles of diameter 0.1 to 2 pm (fine par-
ticles). Extinction due to scattering is closely proportional to the fine-particle mass

9-97
-------
p
concentration, with extinction/fine mass concentration ratios in the range of about 3 m/g for
relative humidities below 50-70 percent. For higher humidities, the ratio increases, approxi-
mately doubling at 90 percent RH. Extinction/fine mass concentration ratios for absorbing
2
aerosols typically range from 7 to 12 m /g.
The currently available visibility monitoring methods measure different aspects of visi-
bility impairment. Generally, contrast-type measurements (such as photography, telephoto-
metry, and human eye observations) relate well to the perception of visual air quality, while
extinction or scattering measurements (such as transmissometry and nephelometry) relate to the
cause of visibility degradation. Each of the above measurement methods can be used to approx-
imate visual range. No single method is yet widely accepted for measuring light absorption.
Current knowledge indicates that fine particulate matter is composed of varying amounts
of sulfate, ammonium, and nitrate ions, elemental carbon, organic carbon compounds, water, and
smaller amounts of soil dust, lead compounds, and trace species. Sulfate often dominates the
fine mass and light scattering, while elemental C is sometimes the primary visibility-reducing
species in urban areas. Significant variations can occur at different times and sites. Our
knowledge of the roles of several possibly important species is hindered by the lack of data.
The 30-year record of the spatial and temporal trends of coal combustion and visibility
suggest that the increasing emissions of SO since the 1950's have been associated with
/\
similar increases in haziness. Nevertheless, the relationship between S02 emissions and
resulting sulfate concentrations is not as well known as the relationships between sulfate
concentrations and visibility reduction.
Studies performed over the last decade have shown that visibility is a sensitive para-
meter perceived by the public to indicate polluted air. Loss in the aesthetic value of
natural vistas has been ascribed to aerosols of fine PM. A number of studies have been
conducted to determine the economic benefit associated with good or improved visibility. The
results of these studies, including both contingent and actual market approaches, show that
the value people place on visibility is substantial. Although the circumstances of studies
differ, the results are broadly consistent, providing evidence that estimation procedures are
valid. And, finally, hazards to ground and air transportation have been associated with
greatly reduced visibility caused by high concentration of fine PM.
Although the effects on ground transportation from incidents of reduced visibility owing
to air pollution are not well documented, they are probably minimal. On the other hand, the
effects on aircraft operations are both well documented and significant. Historical records
from the National Weather Service indicate that of occasi-ons of visibility of 3 miles (5 km)
or less about half occur in the absence of fog, precipitation, or blowing material such as
sand or dust. At such low visibility, noninstrument-rated pilots or planes are grounded and
commercial air traffic operations at major airports may be significantly reduced.
9-98
-------
Pollutants released to the atmosphere alter the environment in ways other than visibility
reduction. They may lead to slow and subtle changes in the nature of the atmosphere and,
possibly, in climate. For example, a fraction of the solar radiation may be absorbed by
aerosols, further reducing the amount of radiation reaching the earth's surface and, at the
same time, heating the aerosol layer itself. On a hazy day, the direct solar radiation may be
reduced to about one-half of that on a clear day, but most of the energy reappears as diffuse
skylight. There is, however, an overall loss of up to about 10 to 20 percent of the radiation
reaching the surface.
If there are no clouds between the observer and the sun, the intensity of direct solar
radiation for a given solar elevation depends on the variable amount of dust, haze, and water
vapor in the atmosphere. The extinction produced by these constituents is called atmospheric
turbidity. During hazy episodes, turbidity coefficients as high as 0.6 to 1.0 have been
reported, translating into a 75- to 90-percent removal of solar radiation from the direct
beam, a 7.5- to 9-percent loss to space, and a 7.5- to 9-percent loss as atmospheric heating.
One of the consequences of such a hazy atmosphere is the disappearance of shadow contrast.
The attenuation of solar radiation from scattering and absorption by particles in the
atmosphere is probably an important factor in climatic change on all scales. But solar
radiation and aerosol levels measured at stations remote from pollutant sources have not, as
yet, displayed any trend that can be related to human causes.
Cloud- and precipitation-forming processes may be divided into two broad classes: macro-
physical and microphysical processes. Macrophysical processes involve the rise and descent of
air masses and the amount of water vapor available for condensation. Atmospheric aerosols,
primarily those that are strongly hygroscopic, influence the microphysics of cloud formation.
The incorporation of particles into rain and fog droplets can change the characteristics of
precipitation by changing its chemical composition. The complex interactions of cloud- and
precipitation-forming processes, however, obscure the specific role of manmade aerosols.
Accordingly, climatic effects cannot be related quantitatively to pollutant emissions.
9-99
-------
9.9 REFERENCES

Ackerman, B., S. A. Changnon, G. Dzurisin, D. F. Eatz, R. C. Grosh, S. P. Hilberg, F. A. Huff,
J. W. Hansel!, H. G. Ochs, M. E. Peden, P. T. Schickedanz, R. G, Semonin, J. L. Vogel.
Summary of Metromex, vol II: Causes of Precipitation Anomalies. Illinois State Water
Survey, Bulletin 63, 1978.

Allard, D., and I. Tombach. Evaluation of visibility measurement methods in the eastern
United States. Annual Meeting of the Air Pollution Control Association, Montreal,
Quebec, June 1980. Paper no. 80-29-3.

Allegrini, I. Optical absorption constant of suspended particulate matter. An air pollution
index. Environ. Sci. Tech. 14:1221-1227, 1980.

Allen, J. , R. B. Husar, and E. S. Macias. Aerosol size and the shape determination using a
laser light scattering spectrometer. In: Aerosol Measurement, a Workshop, University of
Florida, Gainesville, Florida, 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.312-320.

Altshuller, A. P. Atmospheric sulfur dioxide and sulfate, distribution of concentration at
urban and nonurban sites in the United States. Environ. Sci. Techno!. 7:709-713, 1973.

Angel 1, J. K., and J. Korshover. Variation In sunshine duration over the contiguous United
States between 1950 and 1972. J. Appl. Meteorol. 14: 1174-1181, 1975.

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 (ACHEX). Environ.
Sci. Tech. 12:418-425, 1978.

Appel, B. R., S. M. Wall, Y. Tokiwa, and M. Haik. Interference effects in sampling parti-
culate nitrate in ambient air. Atraos. Environ. 13:319-325, 1979.

Barker, M. L., Planning for environmental indices: observer appraisals of air quality. In:
Perceiving Environmental Quality: Research and Applications; Plenum Press, New York, NY,
1976, pp. 175J203.

Barnes, R. A., and D. 0. Lee. Visibility in London and the long distance transport of atmos-
pheric sulfur. Atmos. Environ. 12:791-794, 1978.

Bergstrom, R. W. Extinction and absorption coefficients of the atmospheric aerosol as a
function of particle size. Beitr. Phys. Atmos. 46:223-224, 1973.

Beuttell, R. G. and A. W. Brewer. Instruments for the measurement of the visual range. J.
Sci. Instrum. Phys. Ind. 26:357-359, 1949.

Blackwell, H. R. Contrast thresholds of the human eye. J. Opt. Soc. Am. 36:624-643, 1946.

Bolin, B., and R. J. Charlson. On the role of the tropospheric sulfur cycle in the shortwave
radiative climate of the earth. Ambio 5:47-54, 1976.

Booz, Allen and Hamilton, Inc. Study to determine residential soiling costs of particulate
air pollution. APTD-0715, U. S. Dept. of Health, Education and Welfare, National Air
Pollution Control Administration, Raleigh, NC, October, 1970.

Brookshire, D. S. Issues in valuing visibility: an overview. In: Proceedings of the Work-
shop in Visibility Values, Fort Collins, CO, January 28-February 1, 1979. D. Fox, R. J.
Loomis, and T. C. Green, eds., U.S. Department of Agriculture General Technical Report
WO-18, 1979.

9-100 ,
-------
Brookshire, D. S. , R. C. D'Arge, W. D. Schulze. Methods Development for Assessing Air Pollu-
tion Control Benefits, vol. II: Experiments on Valuing Non-Market Goods: A case Study of
Alternative Benefit Measures of Air Pollution Control in the South Coast Air Basin of
Southern California. EPA-600/5-79-001b, U. S. Environmental Protection Agency,
Washington, DC, 1979.

Brookshire, D. S. , R. C. D'Arge,' W. D. -Schulze. Valui-ng •Public Goods: A Comparison of Survey
and Hedonic Approaches. Resource and Environmental Economics Laboratory. University of
Wyoming, Laramie, WY, 1980.

Brookshire, D. S., B. C. Ives, and W. D. Schulze. The valuation of aesthetic preferences. J.
Environ. Manage. 3:325-346, 1976.

Cahill, T. A., B. H. Kusko, L. L. Ashbaugh, J. B. Barone, R. A. Eldred, and E. G. Walther.
Regional and local determinations of particulate matter and visibility in the south-
western United States during June and July, 1979. In: Proceedings of the Symposium on
Plumes and Visibility: Measurements and Model Components, Grand Canyon National Park,
AZ, November 10-14, 1980. Atmos. Environ. 15:2011-2016, 1981.

Campbell, F. W. and L. Maffei. Contrast and spatial frequency. Sci. Am. 231:106-114, 1974.

Cass, G. R. On the relationship between sulfate air quality and visibility with examples in
Los Angeles. Atmos. Environ. 13:1069-1084, 1979.

Changnon, S. A., Jr. The La Porte weather anomaly—fact or fiction? Bull. Am. Meteorol. Soc.
49:4-11, 1968.

Changnon, S. A., F. A. Huff, P. T. Schickedanz, J. L. Vogel. Summary of Metromex, Vol. I:
Weather Anomalies and Impacts. Illinois State Water Survey. Bulletin 62, 1977.

Charlson, R. J. , H. Horvath, and R. F. Pueschel. The direct measurement of atmospheric light
scattering coefficient for studies of visibility and pollution. Atmos. Environ.
1:469-478, 1967.

Charlson, R. J., A. P. Waggoner, J. F. Thielke. Visibility Protection for Class I Areas: The
Technical Basis. Council on Environmental Quality, Washington, DC, August 1978a.

Charlson, R. J. , D. S. Covert, T. V. Larson, and A. P. Waggoner. Chemical properties of
tropospheric sulfur aerosols. Atmos. Environ. 12:39-53, 1978b.

Charlson, R. J. , D. S. Covert, Y. Tokiwa, and P. K. Mueller. Multiwavelength nephelometer
measurements in Los Angeles, smog aerosol, III. Comparison to light extinction by N0?.
J. Colloid Interface Sci. 39:260-265, 1972.

Charlson, R. J. , .A. H. Vanderpol, D. S. Covert, A. P. Waggoner, and N. C. Ahlquist.
H2S04/(NH4)2S04 background aerosol: optical detection in St. Louis region. Atmos.
Environ. 8:1257-1268, 1974.

Clark, R. R. A hydrologic reanalysis of the La Porte anomaly. Bull. Am. Meteorol. Soc. 60:
1979.

Conway, H. M., ed. The Weather Handbook. Conway Research, Inc., Atlanta, GA, 1963.

Cooper, J. A. and J. G. Watson. Portland Aerosol Characterization Study. Presented at the
72d Annual Meeting, Air Pollution Control Association, Cincinnati, OH, June 1979. Paper
no. 79-29.4

Cornsweet, T. N. Visual Perception. Academic Press, Inc., New York, NY, 1970.

9-101
-------
Countess, R. J. , S. H. Cadle, P. J. Groblicki, and G. T. Wolff. Chemical analysis of size-
segregated samples of Denver's ambient particulate. GMR-3421, General Motors Research
Laboratories, Warren, MI, 1980a.

Countess, R. J,, G. T. Wolff, and S. H. Cadle. The Denver winter aerosol: a comprehensive
chemical characterization. J. Air Pollut. Control Assoc. 30:1194-1200, 1980b.

Covert, D. S., R. J. Charlson, and N. C. Ahlquist. A study of the relationship of chemical
composition and humidity to light scattering by aerosols. J. Appl. Meteorol. 11:968-976,
1972.

Covert, D. S. , A. P. Waggoner, R. E. Weiss, N. C. Ahlquist, and R. J. Charlson. Atmospheric
aerosols, humidity, and visibility. Adv. Environ. Sci. Techno!. 9:559-581, 1980.

Crosby, P., and B. W. Koerber. Scattering of light in the lower atmosphere. J. Opt. Soc. Am.
53:358-361, 1963.

Cwalinski, R., J. M. Lansinger, and W. G. Tank. Field testing and evaluation of methods for
measuring visibility. EPA-650/2-75-039, U.S. Environmental Protection Agency, Research
Triangle Park, N.C., 1975.

Draftz, R. G. Identification and sources of Denver aerosol. Presented at the 67th Annual
Meeting, Air Pollution Control Association, 1974. Paper no. 74-263.

Duntley, S. Q. Visibility, II. Summary. Appl. Opt. 3:551-556, 1964.

Edwards, J. D. A comparison of the British smoke shade method and the integrating plate
method. Master's Thesis, University of Washington, Seattle, WA, 1980.

Eiden, R. Determination of the complex index of refraction of spherical aerosol particles.
Appl. Opt. 10:749-754, 1971.

Eldred, R. A., L. L. Ashbaugh, T. A. Cahill, R. G. Flocchini, and M. L. Pitchford. The effect
of the 1980 smelter strike on air air quality in the southwest. Report #AQG 81-025, Air
Quality Group, Crocker Nuclear Laboratory, Univ. of California at Davis, Davis, CA, 1981.

Ellenson, W. D., R. C. Nininger, E. T. Myers, J. H. Overton, Jr., E. W. Corse, J. C. King, D.
B. Kirkton, and P. L. Smith. Data report - visibility study in the Great Smoky Mountains
region - September 1980. U.S. Environmental Protection Agency Contract 68-02-2566,
Northrop Services, Inc., Research Triangle Park, NC, 1981.

Ellestad, T. G. Aerosol composition of urban plumes passing over a rural monitoring site.
Ann. N. Y. Acad. Sci. 338:202-218, 1980.

Ellestad, T. G,, and R. E. Speer. Application of a telephotometer to visibility measurements
in the eastern United States. Atmos. Environ 15:2443-2449, 1981,

Ellis, A. T., and R. F. Pueschel. Absence of air pollution trends at Mauna Loa. Science
172:845-846, 1971.

Eltertnan, P. Integrating cavity spectroscopy. Appl. Opt. j):2140-2142, 1970.

Ensor, D., and A. P. Waggoner. Angular truncation error in the integrating nephelometer.
Atmos. Environ. 4:481-487, 1970.

Ensor, D. S., R. J. Charlson, N. C. Ahlquist, K. T. Whitby, R. B. Husar, and B. Y. H. Liu.
Multiwavelength nephelometer measurements in Los Angeles smog aerosol. I. Comparison of
calculated and measured light scattering. J. Colloid Interface Sci. 39:242-251, 1972.

9-102
-------
Faxvog, F. R. Optical scattering per unit mass of single particles. Appl. Opt. 14:269-270,
1975.

Faxvog, F. R. , and D. M. Roessler. Carbon aerosol visibility vs. particle size distribution.
Appl. Opt. 17:2612-2616, 1978.

Federal Aviation Administration. Airmen's information manual. Basic Flight Information - ATC
Procedures, January, 1978a.

Federal Aviation Administration. Air Traffic Service Performance measurement system for major
airports. Nov. 1975-1980. U.S. Dept. of Transportation, Washington, DC, 1980a.

Federal Aviation Administration. Census of U.S. Civil Aircraft. U.S. Dept. of Trans-
portation, Washington, DC, 1980b.

Federal Aviation Administration. General Aviation Avionics Statistics. U.S. Dept. of
Transportation, Washington, DC, April 1981a. p. 19.

Federal Aviation Administration. U.S. Civil Airmen Statistics. U.S. Dept. of Transportation,
Washington, DC, 1981b.

Ferman, M. A., G. T. Wolff, and N. A. Kelly. The nature and sources of haze in the Shenandoah
Valley/Blue Ridge Mountains area. J. Air Pollut. Control Assoc. 31:1074-82, 1981.

Fischer, W. H. Some atmospheric turbidity measurements in Antarctica. J. Appl. Meteorol.
6:958-959, 1967.

Flachsbart, P. G. , and S. Phillips. An index and model of human response to air quality. J.
Air Pollut. Control Assoc. 30:759-768, 1980.

Flowers, E. C., R. A. McCormick, and K. R. Kurfis. Atmospheric turbidity over the United
States, 1961-1966. J. Appl. Meteorol. 8:955-962, 1969.

Fowler, B. W. , and C. C. Sung. Scattering of an electromagnetic wave from dielectric bodies
of irregular shape. J. Opt. Soc. Am. 6_9:756-761, 1979.

Fox, D. , R. J. Loomis, and T. C. Green, eds. Proceedings of the workshop in Visibility
Values, Fort Collins, CO, January 28-February 1, 1979. U.S. Dept. of Agriculture General
Technical Report WO-18.

Gates, D. M. Spectral distribution of solar radiation at the earth's surface. Science
151:523-529, 1966.

Glass, N. R. (ed.) Environmental Effects of Increased Coal Utilization: Ecological Effects
of Gaseous Emissions from Coal Combustion. EPA-600/7-78-108, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1978.

Grams, G. W. , I. H. Blifford, Jr., D. A. Gillette, and P. B. Russell. Complex index of
refraction of airborne soil particles. J. Appl. Meteorol. 13:459-471, 1974.

Green, D. M. , and F. W. Campbell. Optical and retinal factors affecting visual resolution.
J. Physio!. 181:576-593, 1965.

Griffing, G. W. Relations between the prevailing visibility, nephelometer scattering coeffi-
cient, and sunphotometer turbidity coefficient. Atmos. Environ. 14:577-584, 1980.
9-103
-------
Groblicki, P. J. , G. T. Wolff, and R. J. Countess. Visibility reducing species in the Denver
brown cloud. Part 1: Relationships between extinction and chemical composition. _In:
Proceedings of the Symposium on Plumes and Visibility: Measurements and Model
Components, Grand Canyon National Park, AZ, November 10-14, 1980. Atmos. Environ. 15:
2473-2484, 1981.

Grosjean, 0., G. J. Doyle, T. M. Mischke, M. P. Poe, D. R. Fitz, J. P. Smith, and J. N. Pitts,
Jr. Concentration, size distribution, and modes of formulation of particulate nitrate,
sulfate and ammonium compounds in the eastern part of the Los Angeles air basin.
Presented at the 69th Annual Meeting, Air Pollution Control Association, 1976. Paper No.
76-20.3.

Hall, F. F. , Jr. Visibility reductions from soil dust in the western U.S. In: Proceedings
of the Symposium on Plumes and Visibility: Measurements and Model Components, Grand
Canyon National Park, AZ, November 10-14, 1980. Atmos. Environ. 15:1929-1933, 1981.

Hall, J. S., and L. A. Riley. Basic spectrophotometric measures of air quality over long
paths. In: Radiative Transfer and Thermal Control, vol. 49: Progress in Astronautics
and Aeronautics. A. M. Smith, ed. , American Institute of Aeronautics and Astronautics,
New York, NY, 1976. pp. 205-212.

Hanel, G. New results concerning the dependence of visibility on relative humidity and their
significance in a model for visibility forecast. Beitr. Phys. Atmos. 44:137-167, 1971.

Ha'nel, G, The properties of atmospheric aerosol particles as functions of the relative humid-
ity at thermodynamic equlibrium with the surrounding moist air. In: Advances in
Geophysics. Landsberg, H. E. and J. van Mieghen, eds. Academic Press, New York, NY,
1976.

Harrison, A. W. Nephelometer estimates of visual range. Atmos, Environ. 13:645-652, 1979.

Heintzenberg, J. , and Quenzel, H. On the effect of the loss of large particles on the deter-
mination of scattering coefficients with integrating nephelometers. Atmos. Environ. 7:
503-507, 1973.

Heisler, S. L., R. C. Henry, and J. G. Watson. The sources of the Denver Haze in November and
December of 1978. Presented at the 73d Annual Meeting, Air Pollution Control
Association, 1§80, Paper no. 80-58.6.

Henry, R. C. Psychophysics and visibility values. In: Proceedings of the Workshop in
Visibility Values, Fort Ceilings, CO, January 28-February 1, 1979. D. Fox, R. J. Loomis,
and T. C. Greene, U.S. Dept. of Agriculture, General Technical Report WO-18.

Ho, W. , Hidy, G. M. , and Govan, R. M. Microwave measurements of the liquid water content of
atmospheric aerosols. J. Appl. Meteorol. 13:871-879, 1974.

Hobbs, P. V., H. Harrison, and E. Robinson. Atmospheric effects of pollutants. Science
183:909-915, 1974.

Hodge, P. W., N. Laulainen, and R. J. Charlson. Astronomy and air pollution. Science
178:1123-1124, 1972.

Hodkinson, J. R. Calculations of colour and visibility in urban atmospheres polluted by
gaseous NOg. Air Water Pollut. 10:137-144, 1966.

Hoffmann, H. E., and Kuehnemann, W. Comparison of the results of two measuring methods
determining the horizontal standard visibility with the visual visibility range. Atmos.
Environ. 13:1629-1634, 1979.

9-104
-------
Horvath, H.» and K. E. Noll. The relationship between atmospheric light scattering
coefficient and visibility. Atmos. Environ. 3:543-550, 1969.

Husar, R. B. , and D. E. Patterson. Regional scale air pollution: sources and effects. Ann.
N.Y. Acad. Sci. 338:399-417, 1980.

Husar, R. B. , and J. M. Holloway. Visibility trend at Blue Hi-11, ME since 1889. Bull, of Am.
Meteorol. Soc. in press, 1981.

Husar, R. B. , D. E. Patterson, J. M. Holloway, W. E. Wilson, and T. G. Ellestad. Trends of
eastern U.S. haziness since 1948. |n: Proceedings of the Fourth Symposium on
Atmospheric Turbulence, Diffusion, and Air Pollution, American Meteorological Society,
Reno, NE, 1979. pp. 249-256.

Janzen, J. Extinction of light by highly non-sperical strongly absorbing colloidal particles:
spectrophotometric determination of volume distributions for carbon blacks. Appl. Opt.
19:2977-2985, 1980.

Jennings, S. G. , and Pinnick, R. G. Relationships between visible extinction, absorption and
mass concentration of carbonaceous smokes. Atmos. Environ. 14:1123-1129, 1980.

Kerker, M. The Scattering of Light and Other Electromagnetic Radiation. Academic Press,
Inc., New York, NY, 1969.

Koschmieder, H. Theorie der horizontalen Sichtweite. [theory of the horizontal visual
range.] Beitr. Phys. Freien Atmos. 12:33-53, 171-181, 1924.

Landsberg, H. E. Man-made climatic changes. Science 170:1265-1274, 1970.

Landsberg, H. E. The Urban Climate, International Geophysical Series, Vol. 28, Academic
Press, New York, 1981.

Latimer, D. A., R. W. Bergstrom, S. R. Hayes, M. K. Liu, J. H. Seinfeld, G. Z. Whitten, M. A.
Wojcik, and M. J. Hillyer. The Development of Mathematical Models for the Prediction of
Anthropogenic Visibility Impairment. EPA-450/3/78-110a, U.S. Environmental Protection
Agency, Research Triangle Park, NC, September 1978.

Latimer, P. Predicted scattering by spheroids: comparison of approximate and exact methods.
Appl. Opt. 19:3039-3041, 1980.

Lawson, D. R., and J. W. Winchester. Atmospheric sulfur aerosol concentrations and character-
istics from the South American Continent. Science 205:1267-1269, 1979.

Leaderer, B. P., D. M. Bernstein, J. M. Daisey, M. T. Kleinman, T. J. Kneip, E. 0. Knutson, M.
Lippmann, P. J. Lioy, K. A. Rahney, D. Sinclair, R. L. Tanner, G. T. Wolff. Summary of
the New York summer aerosol study (NYSAS). J. Air Pollut. Control Assoc. 28:321-327,
1978.

Leonard, E. M. , M. D. Williams, and J. P. Mutschlecner. The visibility issue in the Rocky
Mountain West. Prepared by Los Alamos Scientific Laboratory for the Dept. of Energy,
preliminary draft report, Sept. 30, 1977.

Lewis, C. W. , and E. S. Macias. Composition of size-fractioned aerosol in Charleston, West
Virginia. Atmos. Environ. 14:185-194, 1980.

Lin, C. I., M. Baker, and R. J. Charlson. Absorption coefficient of atmospheric aerosol: a
method for measurement. Appl. Opt. 12:1356-1363, 1973.
9-105
-------
Lindberg, J. D., and L. S. Laude. Measurement of the absorption coefficient of atmospheric
dust. Appl. Opt. 13:1923-1927, 1974.

Machta, L. , and K. Telegadas. Inadvertent large-scale weather modification. In; Weather and
Climate Modification. W. N. Hess, ed. , John Wiley & Sons, New York, NY, 1974. pp.
687-725.

Macias, E. S., and R. B. Husar. A review of atmospheric participate mass measurement via the
beta attenuation technique. In: Fine Particles, Aerosol Generation Measurement,
Sampling, and Analysis. B. Y. H. Liu, ed., Academic Press, New York, NY, 1976.

Macias, E. S,, D. L. Blumenthal, J. A. Anderson, and B. K. Cantrell. Size and composition of
visibility-reducing aerosols in southwestern plumes. Ann. N.Y. Acad. Sci. 338:233-257,
1980.

Macias, E. S., J. 0. Zwicker, J. R. Ouimette, S. V. Hering, S. K. Friedlander, T. A. Cahill,
G. A. Kuhlmey, and L. W. Richards. Regional haze case studies in the southwestern U.S.:
I. Aerosol chemical composition. In: Proceedings of the Symposium on Plumes and
Visibility: Measurements and Model Components, Grand Canyon National Park, AZ, November
10-14, 1980. Atmos. Environ. 15:1971-1986, 1981a.

Macias, E. S., J. 0. Zwicker, and W. H. White. Regional haze case studies in the southwestern
U.S. - II. Source contributions. In: Proceedings of the Symposium on Plumes and
Visibility: Measurements and Model Components, Grand Canyon National Park, AZ, November
10-14, 1980. Atmos. Environ. 15:1987-9997, 1981b.

Malm, W. Visibility: A physical perspective. In: Proceedings of the Workshop in Visibility
Values. Fort Collins, Colorado, January 28-February 1, 1979. D. Fox, R. J. Loomis, and
T. C. Greene, eds. U.S. Dept. of Agriculture General Technical Report WO-18, 1979a. pp.
56-68.

Malm, W. C. Considerations in the measurement of visibility. J. Air Pollut, Control Assoc.
29:1042-1052, 1979b.

Malm, W, C. K. K. Kelly, and J. V. Molenar. Human perception of visual air quality (Uniform
Haze). In: Proceedings of the Symposium on Plumes and Visibility: Measurements and
Model Components, Grand Canyon National Park, AZ, November 10-14, 1980. Atmos. Environ.
15:1875-1890, 1981a.

Malm, W. C., E. G. Walther, K. E. O'Dell and M. Kleine. Visibility in the Southwestern United
States from summer 1978 to spring 1979. In: Proceedings of the Symposium on Plumes and
Visibility: Measurements and Model Components, Grand Canyon National Park, AZ, November
10-14, 1980. Atmos. Environ. 15:2031-2042, 1981b.

Malm, W. , A. Pitchford, R. Tree, E. Walther, M. Pearson, and S. Archer. The visual air
quality predicted by conventional and scanning teleradiometers and an integrating
nephelometer. In: Proceedings of the Symposium on Plumes and Visibility: Measurements
and Model Components, Grand Canyon National Park, AZ, November 10-14, 1980. Atmos
Environ. 15:2547-2554, 1981c.

Manabe, F., and R. Wetherald. The effects of doubling the C02 concentration on the climate of
a general circulation model. J. Atmos. Sci. 3_2:3-15, 19/5.

Marians, M. , and J. Trijonis. Empirical Studies of the Relationship Between Emissions and
Visibility in the Southwest. Prepared under grant 802015 by Technology Service Corp.
U.S. Environmental Protection Agency, Cincinnati, OH, 1979.
9-106
-------
McCree, K. J. , and M. E. Keener. Effect of atmospheric turbidity on the photosynthetic rates
of leaves. Agric. Meteorol. 13:349-357, 1974.

Middleton, W. E. K. The colors of distant objects. J. Opt. Soc. Am. 40:373-376, 1950.

Middleton, W. E. K. Vision Through the Atmosphere. University of Toronto Press, Toronto,
Canada, 1952.
Mie, G. Optics of cloudy media, especially colloidal metal sols. Ann. Phys. 25:377-445,
1908. ~~

Mugnai, A., and W. J. Wiscombe. Scattering of radiation by moderately nonspherical particles.
J. Atmos, Sci. 37:1291-1307, 1980.

National Center for Air Pollution Control. New York - New Jersey Air Pollution Abatement
Activity. Particulate Matter. Phase II. Pre-Conference Investigations. U.S. Depart-
ment of Health, Education, and Welfare, Public Health Service, Cincinnati, OH, December,
1967. p.105.

National Park Service. My Eyes Need a Good Stretching; Seven Authorities Speak Out on Visi-
bility — Clean Air and Unique Natural Areas. U.S. Department of the Interior, Washing-
ton, DC, 1978.

National Transportation Safety Board. Annual review of aircraft accident data. NTSB-ARC-8Q-1
Washington, D. C. 1978a.

National Transportation Safety Board. Briefs of Fatal Accidents Involving Weather as a
Cause/Factor, U.S. General Aviation, 1978. NTSB-AMM-80-5, National Transportation Safety
Board, Washington, DD, 1978b.

National Weather Service, Federal Meteorological Handbook No. I, Surface Observations.
National Oceanographic and Atmospheric Administration, Washington, DC, 1979.

National Weather Service. Historical data base of visibility measurements for 147 U. S. sites
1951-1980. National Oceanographic and Atmospheric Administration, Washington, DC, 1981.

Nininger, R. C., J. H. Overton, Jr., M. J. Kaiserman, V. P. Aneja, E. W. Corse, D. B. Kirkton,
and E. T. Myers. Data report: visibility study in the Ohio River basin - February 1980.
U.S. EPA Contract 68-02-2566. Northrop Services, Inc. Research Triangle Park, NC, 1981a.

Nininger, R. C. , W. D. Ellenson, J. H. Overton, Jr., R. A. Carlson, W. S. Crews, J. W.
Faircloth, J. C. King, E. T. Myers, P. L. Smith, M, J. Turner, and M. L. Wheeler. Data
report: visibility study in the Ohio River basin - July 1980. U.S. EPA Contract
68-02-2566. Northrop Services, Inc. Research Triangle Park, NC, 1981b.

Orel, A. E, , and J. H. Seinfeld. Nitrate formation in atmospheric aerosols. Environ. Sci.
Tech. 11:1000-1007, 1977.

Orgill, M. M. , and G. A. Sehmel. Frequency and diurnal variation of dust storms in the con-
tiguous United States. Atmos. Environ. 10:813-825, 1976.

Orr, C. , F. K. Hurd, and W. J. Corbett. Aerosol size and relative humidity. J. Colloid
Interface Sci. 13:472-482, 1958.

Ouimette, J. R. Aerosol Chemical Species Contributions to the Extinction Coefficient. Ph.D.
Thesis, California Institute of Technology, Pasadena, CA, 1980.

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. 15:919-927, 1981.

9-107
-------
Pate!, A. S. Spatial resolution by the human visual system. The effect of mean retinal
illuminance. J. Opt. Soc. Am. 56:689-694, 1966.

Patterson, E. M. Atmospheric extinction between 0.55 mm and 10.6 mm due to soil-derived aero-
sols. Appl. Opt. 16:2414-2418, 1977.

Patterson, E. M. , and D. A. Gillette. Measurements of visibility vs. mass-concentration for
airborne soil particles. Atmos. Environ. 11:193-196, 1977.

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,

Penndorf, R. Tables of the refractive index for standard air and the Rayleigh scattering co-
efficient for the spectral region between 0.2 and 20.0 m and their application to atmos-
pheric optics. J. Opt. Soc. Am. 47:176-182, 1957.

Peterson, J. T. , and E. C. Flowers. Interactions between air pollution and solar radiation.
Sol. Energy 19:23-32, 1977.

Pierson, W. R., and P. B. Russell. Aerosol carbon in the Denver area in November 1973.
Atmos. Environ. 13:1623-1628, 1979.

Pierson, W. R., W. W. Brachaczek, T. J. Truex, J. W. Butler, and T. J. Korniski. Ambient
sulfate measurements on Allegheny Mountains and the question of atmospheric sulfate in
the northeastern United States. Ann. N.Y. Acad. Sci. 338:145-173, 1980.

Pinnick, R. G., D. E. Carroll, and D. J. Hofmann. Polarized light scattered from monodisperse
randomly oriented nonspherical aerosol particles: measurements. Appl. Opt. 15:384-393,
1976.

Porch, W. M. , R. J. Charlson, and L. F. Radke. Atmospheric aerosol: does a background level
exist? Science 170:315-317, 1970.

Presle, G. and R. Pirich. The relation between changes of the extinction coefficient and
relative humidity. J. Aerosol Sci. 11:255-257, 1980.

Pueschel, R. F. ahd P. A. Allee. Visibility, climatology, and air quality in east-central
Utah. Presented at the Symposium on Plumes and Visibility: Measurements and Model
Components, Grand Canyon National Park, AZ, November 10-14, 1980.

Pueschel, R. F. , and D. L. Wollman. On the nature of atmospheric background aerosol.
Presented at the 14th Conference on Agricultural and Forest Meteorology, Minneapolis, MN,
April 1978.

Rabinoff, R., and B. Herman. Effect of aerosol size distribution on the accuracy of the in-
tegrating nephelometer. J. Appl. Meteorol. 12:184-186, 1973.

Randall, A., B. Ives, C. Eastman. Bidding games for valuation of aesthethic environmental
improvements. J. Environ. Econ. Manage. 1:132-149, 1974.

Robinson, G. D. Absorption of solar radiation by atmospheric aerosol, as revealed by measure-
ments at the ground. Arch. Meteorol. Geophys,' Bioclimatol, Ser. B 12:19, 1962.

Roessler, D. M., and F. R. Faxvog. Optical properties of agglomerated acetylene smoke parti-
. cles at 0.5145 ym and 10.6 urn wavelengths. J. Opt. Soc. Am. 70:230-235, 1980.

Rosen, H., A. D. A. Hansen, R. L. Dod, and T. Novakov. Soot in urban atmospheres: deter-
mination by an optical absorption technique. Science 208:741-744, 1980.

9-108
-------
Rosen, H. , A. D. A., Hansen, L. , Grundel, and T. Novakov. Identification of the optically
absorbing component in urban aerosols. Appl. Opt. 17:3859-3861, 1978.

Rowe, R. D., and L. G. Chestnut. Visibility Benefits Assessment Guidebook. EPA-450/5-81-001,
U.S. Environmental Protection Agency, Research Triangle Park, NC, August 1981.

Rowe, R. D., R. C. d'Arge, and D. S. Brookshire. An experiment on the economic value of visi-
bility. J. Environ. Econ. Manage. 7:1-19, 198Qav

Rowe, R. D. , R. C. d'Arge, and D. S. Brookshire. Environmental Preference and Effluent
Charges. In: Progress in Resource Management and Environmental Planning. Timothy
O'Riordin and R. Kerry Turner, eds. London, England. John Wiley & Sons, Ltd. 1980b.
(Series in Environmental Economics)

Ruby, M. G. , and A. P. Waggoner. Intercomparison of integrating nephelometer measurements.
Environ. Sci. Tech. 15:109-113, 1981.

Ruppersberg, G. H. Registrierung der sichtweite mit dem streulichtschreiber. [Reading of the
visual range with a light scattering recorder.] Beitr. Phys. Atm. 37:252-263, 1964.

Sadler, M. , R. J. Charlson, H. Rosen, and T. Novakov. An intercomparison of the integrating
plate and the laser transmission methods for determination of aerosol absorption co-
efficients. Atmos. Environ. 15:1265-1268, 1981.

Samuels, H. J. , S. A. Twiss, E. W. Wong, H. Wong-Woo, P. K. Mueller, J. A. Maga, and G. S.
Taylor. Visibility, Light Scattering and Mass Concentration of Particulate Matter.
Report of the California Tri-City Aerosol Sampling Project. State of California Air
Resources Board, Sacramento, CA, July 1973.

Schober, H. A. W. , and R. Hilz. Contrast sensitivity of the human eye for square-wave grat-
ing. J. Opt. Soc. Am. 55:1086-1091, 1965.

Schusky, J. Public Awareness and Concern with Air Pollution in the St. Louis Metropolitan
Area, J. Air Pollut, Control Assoc., 16:72-76, 1966.

Schuerman, D. W. , ed. Light Scattering by Irregularly Shaped Particles. Plenum, New York,
NY, 1980.

Shah, J. J. , T. L. Johnson, E. K. Heyerdahl, and J. J. Huntzicker. Carbonaceous aerosol at
urban and rural sites in the United States. Presented at the 75th annual meeting of the
Air Pollution Control Association, New Orleans, June 1982. Paper 82-65.5.

Sloane, C. S. Visibility trends II. Mideastern United States 1948-1978, Report no. 10,
General Motors Research Laboratories, Warren, MI, 1980.

Snedecor, G. W. , and W. G. Cochran. Statistical methods. Iowa State University Press, Ames,
10, 1980.

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.

Steffens, C. Measurement of visibility by photographic photometry. Ind. Eng. Chem.
41:2396-2399, 1949.
9-109
-------
Stevens, R. K., T. G. Dzubay, G. Russworm, and D. Rickel. Sampling and analysis of atmos-
pheric sulfates and related species. Atmos. Environ. 12:55-68, 1978.

Stevens, R. K., T. G. Dzubay, R. W. Shaw, W. A. McClenny, C. W. Lewis, and W. E. Wilson.
Characterization of aerosol in the Great Smoky Mountains. Environ. Sci. Tech,
14:1491-1498, 1980.

Stuiver, M. Atmospheric carbon dioxide and carbon reservoir changes. Science 199:253-258,
1980.

Stukel, J. J., and B. R. Keenan. Ohio River Basin Energy Study Phase I: Interim Findings.
EPA-600/7-77-120. U.S. Environmental Protection Agency, Research Triangle Park, NC,
1977.
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, Ann Arbor, MI, 1980.

Tang, I. N., H. R. Munkelwitz, and J. G. Davis. Aerosol growth studies, IV. Phase trans-
formation of mixed salt aerosols in a moist atmosphere. J. Aerosol Sci. 9:505-511, 1978.

Tanner, R. L., R. Garber, W. Marlow, B. P. Leaderer.and M. A. Leyko. Chemical composition of
sulfate as a function of particle size in New York summer aerosol. Ann. N. Y. Acad. Sci.
322:99-113, 1979.

Taylor, J. H. Use of visual performance data in visibility prediction. Appl. Opt. 3:562-569,
1964.

Trijonis, J. , and R. Shapland. Existing Visibility Levels in the U.S.: Isopleth Maps of
Visibility in Suburban/Nonurban Areas During 1974-1976. EPA-450/5-79-010, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, 1979.

Trijonis, J., and K. Yuan. Visibility in the Southwest: An Exploration of the Historical
Data Base. EPA-600/3/78/039, U.S. Environmental Protection Agency, Research Triangle
Park, NC, April 1978a.

Trijonis, J., and K. Yuan. Visibility in the Northeast: Long Term Visibility Trends and
Visibility/Pollutant Relationships, EPA-600/3-78-075, U.S. Environmental Protection
Agency, Research Triangle Park, NC, August 1978b.

Truex, T. J., and J. E. Anderson. Mass monitoring of carbonaceous aerosols with a spectro-
phone. Atmos. Environ. 13:507-509, 1979.

Twomey, S. Pollution and the planetary albedo. Atmos. Environ. 8:1251, 1974.

U.S. Bureau of Mines. Minerals Yearbook, Annual Publication, 1933-74. U.S. Department of the
Interior, Washington, D.C.

U.S. Congress. Senate. Committee on Public Works. A Study of Pollution - Air. U.S. Govern-
ment Printing Office, Washington, D.C. September, 1963. p. 21.

U.S. Department of Commerce. Climatic Atlas of the United States. U.S. Dept. of Commerce,
Washington, D.C., 1968.

U.S. Department of Energy. Carbon Dioxide Effects, Research, and Assessment Program. A Com-
prehensive Plan for Carbon Dioxide Effects, Research, and Assessment, Part I: The Global
Carbon Cycle and Climatic Effects of Increasing Carbon Dioxide. DOE/EV-0094, U.S.
Department of Energy, Washington, DC, 1980.
9-110
-------
U.S. Department of Energy. Office of Carbon Dioxide Effects, Research, and Assessment. A
Comprehensive Plan for Carbon Dioxide Effects, Research, and Assessment. U.S. Department
of Energy, Washington, DC, 1978.

U.S. Dept. of Health, Education and Welfare. Public Awareness and Concern with Air Pollution
in the St. Louis Metropolitan Areas. U.S. Dept. of Health, Education and Welfare, Div.
of Air Pollution, Washington, DC, 1965.

U.S. Environmental Protection Agency. EPA Report to Congress: Protecting Visibility.
EPA-450/ 5-79-008, U. S. Environmental Protection Agency, Research Triangle Park, NC,
1979.

Van Nes, F. L., and M. A. Bouman. Spatial modulation transfer in the human eye. J. Opt. Soc.
Am. 57:401-406, 1967.

Veress, S. A. Extinction coefficient. Photo. Eng. 40:183-191, 1972.

Waggoner, A. P., M. B. Baker, and R. J. Charlson. Optical absorption by atmospheric aerosols.
Appl. Opt. 12:896, 1973.

Waggoner, A. P., and R. J. Charlson. Measurement of aerosol optical parameters. In: Fine
Particles: Aerosol Generation Measurement, Sampling, and Analysis. B. Y. H. Liu, ed.,
Academic Press, Inc., New York, NY, 1976.

Waggoner, A. P., and R. E. Weiss. Comparisons of fine particle mass concentration and light
scattering extinction in ambient aerosol. Atmos. Environ. 14:623-626, 1980.

Waggoner, A. P., R. E. Weiss, and T. V. Larson. _In situ, rapid response measurement of
HpSO./CNHOpSO. aerosols in urban Houston: a comparison with rural Virginia. Submitted
to Atmospheric Environment, August 1982.

Waggoner, A. P., R. E. Weiss, N. C. Ahlquist, D. S. Covert, S. Will, and R. J. Charlson.
Optical characteristics of atmospheric aerosols. Atmos. Environ. 15:1891-1909, 1981.

Wall, G. Public response to air pollution in South Yorkshire, England. Environ. Behav.
5:219-248, 1973.

Weil, H., and C. M. Chu. Scattering and absorption by thin flat aerosols. Appl. Opt.
19:2066-2071, 1980.

Weiss, R. E. The Optical Absorption Properties of Suspended Particles in the Lower Tropo-
sphere at visible wavelengths. Ph.D. Thesis, University of Washington, 1980.

Weiss, R. E. , and A. P. Waggoner. The importance of aerosol absorption and graphitic carbon
in visibility and atmospheric optics. Presented at the 74th Annual Meeting, Air Pol-
lution Control Association, Philadelphia, June 1981, Paper 81-54-6.

Weiss, R. E. , T. V. Larson, and A. P. Waggoner. In s_itu, rapid response measurement of
H2S04/(NH4)2 S04 aerosols in rural Virginia. Env. Sci. Tech. 16:525-532, 1982.

Weiss, R. E. , A. P. Waggoner, R. J. Charlson, and N. C. Ahlquist, Sulfate aerosol: its
geographical extent in the midwestern and southern United States. Science 195:979-981,
1977.

Weiss, R. E. , A. P. Waggoner, R. J. Charlson, D. L. Thorsell, J. S. Hall, and L. A. Riley.
Studies of the optical, physical, and chemical properties of light-absorbing aerosols.
In: Proceedings of the Conference on Carbonaceous Particles in the Atmosphere, March
20-22, 1978. LBL-9037, T. Novakov, Ed., Laurence Berkeley Laboratory, University of
California, 1979. p. 257-262.

9-111
-------
Wesely, M. L., and R. C. Lipschutz. An experimental study of the effects of aerosols on
diffuse and direct solar radiation received during the summer near Chicago. Atmos.
Environ. 10:981-987, 1976.

Whitby, K. T., and G. M. Sverdrup. California aerosols: their physical and chemical charac-
teristics. Adv. Environ. Sci. Tech. 9:477-517, 1980.

White, W. H., and D. E. Patterson. Nitrogen dioxide, particles, and the color of smoke
plumes. In: Proceedings of the Symposium on Plumes and Visibility: Measurements anc
Model Components, Grand Canyon National Park, AZ, November 10-14, 1980. Atmos. Environ.
15:2097-2104, 1981.

White, W. H., and P. T. Roberts. On the nature and origins of visibility reducing aerosols in
the Los Angeles Air Basin. Atmos. Environ. 11:803-812, 1977.

Williams, J. D. , and F. L. Bunyard. Interstate Air Pollution Study, Phase II Project report,
Vol. VII - Opinion Surveys and Air Quality Statistical Relationships. U.S. Dept. of
Health, Education and Welfare, Div. of Air Pollution, Cincinnati, OH, 1966.

WMO/EPA/NOAA/UNEP. Global Monitoring of the Environment for Selected Atmospheric Constituents
1972-1975. Environmental Data Service National Climatic Center, Asheville, NC,
1974-1977.

Wolff, G. T., M. A. Ferman, N. A. Kelly, D. P. Stroup, and M. S. Ruthkosky. The relationships
between the chemical composition of fine particles and visibility in the Detroit
metropolitan area. GMR-4072, ENV # 128, General Motors Research Laboratories, Warren,
MI, 1982.

Wolff, G. T., N. A. Kelly, M. A. Ferman. On the sources of summertime haze in the eastern
United States. Science 211:703-705, 1981.

Wolff, G. T., P. J. Groblicki, S. H. Cadle, and R. J. Countess. Particulate carbon at various
locations in the United States. Presented at the General Motors Symposium on Particulate
Carbon, Warren, MI, October 1980.
9-112
-------
10. EFFECTS ON MATERIALS

10.1 INTRODUCTION
Beginning with associative 17th century observations and continuing through modern ana-
lytical investigation, scientists have gathered evidence that air pollutants damage man-made
and natural materials. Pollutant-related damage may lead.-to measures such as increased main-
tenance, use of protective coatings, replacement of materials, or the control of emissions.
The decision to pursue one course rather than another is guided, ultimately, by weighing the
damage against the costs of prevention. Neither damage nor prevention cost is necessarily
measured monetarily, especially for works of art or items of historical significance. For
better or worse, however, society tends to assign a monetary value to value gained or lost.
This chapter presents and evaluates studies useful for estimating the dollar cost of material
damage caused by sulfur oxides and particulate matter.
Figure 10-1 depicts the relation between pollutant emissions and economic damage. As
shown, one may (1) proceed from ambient pollutant levels to economic damage estimates directly
or (2) estimate damage based on physical damage functions. The latter route, called the
damage function approach, has been the preferred method, although more recent studies have
employed the first route. The estimation of willingness to pay is common to both choices.
Economic damage (benefit) that results from increased (decreased) pollutant concen-
trations can be estimated by willingness-to-pay approaches. All willingness-to-pay approaches
try to estimate the aggregate monetary values that all affected individuals assign to the
effects of a change in pollutant concentration. These approaches can be divided into three
classes: damage function approaches, nonmarket approaches, and indirect market approaches.
The first step of the damage function approach uses the relationship of pollutant exposure to
physical damage. The second step links the physical damage to a dollar estimate of willing-
ness to pay. Most economic damage estimates using this approach have not considered substi-
tution possibilities for producers or consumers; however, proper consideration of these fac-
tors can yield good estimates of willingness to pay (via the damage function approach).
Nonmarket approaches generally use surveys which attempt to ascertain the monetary values
assigned to the effects. Indirect market approaches use information about the demand for
marketed goods to estimate the willingness to pay for nonmarketed environmental attributes
that are closely related to the marketed good (e.g., property value studies that estimate the
willingness to pay for a change in the level of pollutant concentration through analyses of
the changes in price of residential property) (Freeman, 1979a). Each of these three ap-
proaches has a different data requirement.
10-1
-------
POLLUTANT
EMISSIONS /!!«-. », „:>
m -,
-------
The physical damage function approaches are more extensively described here because they
have been the most widely used. Therefore, it is important to explain briefly the method and
some of its limitations. Nonmarket and indirect market studies are discussed in Section 10.5.
In the damage function approach, physical damage must be determined before economic
damage can be estimated by calculating the exposure of certain materials in specific loca-
tions. Exposure is estimated from ambient air levels over specific time intervals. Depending
on the pollutants and the kinds of material studied, both average short-term and long-term
exposure data may be needed. These data are used to develop a physical damage function, which
is a mathematical expression linking exposure to damage. The damage function is expressed in
terms appropriate to the interaction of the pollutant and material. For example, the corro-
sion of metal might be expressed in units of thickness lost, while the deterioration of paint
could be stated in units of reflectance or thickness lost. The pollutant level may be ex-
pressed in any concentration unit. Since material damage usually develops over time, average
long-term pollutant levels are often used in calculating damage. Damage studies are often
broken down by years, which helps economic comparisons and allows averaging of conditions in
changeable climates.
A major problem in establishing reliable damage functions has been to separate the influ-
ence of the target pollutant from those of meteorological parameters and other air pollutants.
For the corrosion of metals, RH is the most important variable. Rainfall, time of wetness,
sunlight, and windspeed and direction are available for many locations and, if relevant, can
be considered in damage calculations. The influence of atmospheric components should be
considered in the context of the material being studied; for example, in studies of SO -
)\
related steel corrosion, even low atmospheric chloride levels may significantly affect the
rate of damage.
A critical damage level should be determined before an economic value is placed on the
incremental damage caused by pollution. This level represents the point at which the service
life or functional utility of the material has ended or is severely impaired. When this point
is reached, replacement or repair is necessary, and cost is incurred. For example, if a
typical coat of paint is 60 nm thick, the critical damage level at which repainting is neces-
sary may occur when about 10 pm remain. By comparing the amount of surface erosion in a clean
environment with that in an area with a specific pollutant, the apparent damage from the
pollutant can be calculated and used in assessing physical damage. The value is determined
through economic damage functions, in which physical damage functions are coupled to the use
and value of the material. Exposure, replacement, protection, and other data are included in
the estimate. This approach cannot account for irreplaceable items, such as works of art,
where the only-, measurable cost is that of preservation. The social cost of losses of
historical artifacts cannot be quantified in monetary terms.
10-3
-------
As discussed above in the context of Figure 10-1, most estimates of economic damage to
materials have followed the damage function approach. This chapter first discusses the physi-
cal effects of SO and PM separately. Within these categories, laboratory studies are dis-
f\
cussed first,, then field studies. Results from field experiments are compared with labor-
atory-derived damage functions. When a damage function has been validated by field and
laboratory experimental results, it may be used as a basis for determining total damage to
materials caused by a given air pollutant.
In the damage function approach to cost estimation, physical damage is one component used
in assessing economic damage. Other components used are a mathematical expression of surface
area of materials exposed and cost factors associated with units of physical damage. Among
the problems with this approach are:
1. Valid physical damage functions do not exist for all pollutants and materials.
2. Estimates of the amount and type of material exposed have usually been based on some
surrogate such as production figures modified by service life data (field surveys of
exposed material in place have not been reported).
3. Cost factors may or may not accurately reflect cost associated with pollution as
opposed to other causes.
This latter consideration is especially a problem in estimates of soiling damage, where socio-
economic factors are heavily involved.
Past estimates of costs associated with PM, SO , or other pollutants must be considered
x>
1n light of the above discussion. The last section of this chapter discusses cost estimates
and their limitations for decisionmaking purposes.
10.2 SULFUR OXIDES
10.2.1 Corrosionof Exposed Metals
Sulfur oxides in the environment accelerate metal corrosion. Several factors other than
*
concentration of S0? are important.
10.2.1.1 Physical and Chemical Considerations—The atmospheric corrosion of most metals is a
diffusion-controlled electrochemical process. For electrochemical action to take place, the
following are necessary: (1) an electromotive force between points on the metal surface; (2)
a mechanism for charge transfer between the electronic conductors; and (3) a conduction path
between the cathode and anode reaction centers. Measurements of the rate of SO,-accelerated
rusting of iron vary greatly from site to site, despite careful monitoring of pollutant con-
centrations, a fact that has often puzzled researchers. Several factors might be responsible
for inconsistent results, including: (1) the deposition rate of gaseous or dissolved SOp and
particles; (2) the variability in the electrochemical actions that cause corrosion; (3) the
influence of rust on the rate of subsequent corrosion; (4) the interaction between the pollu-
tant effects and "wetness time", often indicated by RH, on surface electrolyte concentrations;
and (5) orientation of the metal surface.
10-4
-------
The actual mechanism for the oxidation of S02 (and its hydrated products) at metal-water
interfaces is little understood. Barton (1976) proposed the following schematic reaction:

S0£ + 02 + 2e -» SO*"

or
4HSO~ + 302 + 4e •* 4SO^~ + 2H20 (10-2)
The electrons are provided by the oxidation of the metal (M):
M •*• Mn+ + ne (l'D-3)
Duncan and Spedding (1974), using an electrophoretic method, found that the rates of sof~
formation on iron and zinc surfaces were similar; the pseudo-first-order half-life (equivalent
to corrosion product doubling time) was determined to be about 24 hours. Other workers
(Karraker 1963; Yoshihara et al., 1964) reported higher oxidation rates (half-life, 10 to 100
3+
minutes) in bulk solutions using Fe catalysts, as summarized in Nriagu (1978).
In his discussion of long-term corrosion rates, Barton (1976) noted that rusting occurs
first in localized areas and then spreads across the entire exposed surface. For iron, the
formation of rust increases the adsorption of S09 dramatically depending on RH and temper-
3
ature. For example, in an atmosphere of 10 ppm (26200 ug/m ) S0? at >96 percent RH, virtually
all the S0? that comes in contact with the corroded iron surface is adsorbed. On the other
hand, corrosion products in zinc, copper, and, particularly, aluminum lessen the rate of S0?
uptake. For the action of S00 on iron, Barton calculated that a surface uptake of 40 to 50 mg
2
SO^/m /day is required to accelerate corrosion, assuming critical RH and temperature. From
this uptake level, he calculated that corrosion rate increase will be linearly dependent on
2-
SO, concentration. From the correlation between SO. rates and ambient S0~ levels developed by
Guttman (1968), Barton calculated that the surface uptake level of 40 to 50 mg S09/m /day
3
corresponds to an average annual ambient air concentration of 0.015 ppm (40 ug/m ) S02.
10.2.1.1.1 Relative humidity and corrosion rate. According to Schwarz (1972), the corrosion
rate of a metal should increase by 20 percent for each increase of 1 percent in the RH above
the critical RH value. It is evident that RH has a considerable influence on the corrosion
rate, as established in laboratory trials by Haynie and Upham (1974) and Sydberger and
Ericsson (1976). Although these experimental results do not support the exact rate predicted
by Schwartz, it is apparent from Figures 10-2 and 10-3 (Haynie and Upham, 1974) that the
corrosion rate of steel increases with increasing RH as well as with increasing S02 concen-
tration.
The climate of an area is usually characterized by average RH rather than RH distribu-
tion. Since average RH is calculated from the distribution, there should be an empirical
relationship between average RH and the fraction of time some "critical humidity value"
(minimum concentration of water vapor required for corrosion to proceed) is exceeded, assuming
10-5
-------
!
a.
100

60
so
g
ta
° 40
cc
O
o
30
20
10
I I I
I I
SO, CONCENTRATION, fl g/m
I I I I I I
0 10 20 30 40 SO 60 70 80 90 100

AVERAGE RELATIVE HUMIDITY, %

Figure 10-2. Steel corrosion behavior is shown as a function of aver-
age relative humidity at three average concentration levels of sulfur
dioxide.

Source: Haynie and Upham {1974).
10-6
-------
I*
cc

o
35
O
IE
CC

6
D
100

10
55% RH
J_
100 200 300

AVERAGE SULFUR DIOXIDE CONCENTRATION,
400
Figure 1 0-3. Steel corrosion behavior is shown as a function of aver-

age sulfur dioxide concentration and average relative humidity (RH).

Source; Haynieand Upham (1974).
10-7
-------
a relatively constant standard deviation of RH (Mansfeld and Kenkel, 1976; Sereda, 1974). The
fraction of time that the surface is wet must be zero when the average RH is zero and unity
when the average RH is 100 percent. According to Haynie (1980), the following equation is the
simplest single-constant first-order curve that can be fitted to observed data:
f = a-*) RH ( .
T 100-k RH uu 4;

where f = fraction of time RH exceeds the critical value
RH = average relative humidity
and k = an empirical constant less than unity.
Ten quarter-year periods of RH data from St. Louis International Airport were analyzed and
fitted by the least-squares method to the above equation. The fraction of time the RH ex-
ceeded 90 percent gave a value of 0.86 for k. This fraction and the data points are plotted
in Figure 10-4 (Haynie, 1980).
When the temperature of a metal is below the ambient dewpoint, water condenses on the
metal surface. Whether or not the metal reaches the temperature at which condensation occurs
varies with heat transfer between ground and metal and between air and metal. Condensation
occurs when the RH adjacent to the surface exceeds a value in equilibrium with the vapor
pressure of a saturated solution of whatever salts are on the surface. The solution may
contain corrosion products, other hygroscopic contaminants, or both. Temperature, wind,
sunshine, and night sky cover then become factors in establishing corrosion rates, since they
determine whether there will be sufficient dew condensation.
Haynie (1980) reported on the relationship between diffusion theory and thermodynamics
for the observed effects of five variables: pollution level, RH, temperature, wind velocity,
and surface geome'try. He observed that metals must be wet to corrode electrochemically.
Surfaces are wet from condensation much more often than from precipitation. Further, while
precipitation dilutes electrolytes, dew concentrates them.
10.2.1.1.2 Influence of rainfall on corrosion. Steel surfaces shielded from the leaching
2-
effect of rain may corrode at a higher rate than those exposed to rain. The SO, content of
rust has been identified as a dominant factor in corrosion and is found at higher concen-
2-
trations on surfaces sheltered from rain than on exposed surfaces because soluble SO. is
leached from the rust. Sulfur deposition during rainfall, however, must also be considered.
Haagenrud and Ottar (1975) noted that the rate of corrosion of unalloyed steel and zinc cor-
O—
related wfth the amount of sulfur (SO- and SO, ) in air and in precipitation.
As Kucera's (1976) review indicates, the mode of deposition complicates the analysis of
the effects of acidic precipitation. For example, in areas where dry deposits of hydrogen and
2-
SO^, ions exceed deposits in wet precipitation, flat steel plates corrode more rapidly on their
10-8
-------
100

so
O 40
1U
S

I- 30

•20

0
0 10 20 30 40 50 60 70 80 90 100

AVERAGE RELATIVE HUMIDITY, %

Figure 10-4. Empirical relationship between average relative humid-

ity and fraction of time relative humidity exceeded 90 percent (time

of wetness) is shown for data from St. Louis International Airport.

Source: Haynie (1980).
10-9
-------
undersides than on their upper surfaces, suggesting that rainfall has more of a washing effect
than a corrosive action (vide supra). In other areas, where wet and dry deposition are about
equal, however, the upper sides of the plates corrode more quickly, suggesting that the corro-
sive effect of the rainfall predominates.
Matsushima et al. (1974), in studies of low-alloy weathering steels, considered the
impact of the washing action of rain, the ease with which water would drain off the surface,
and the drying effect of sunlight to determine the effect of these variables on the retention
of particles that influence the electrolytic corrosion mechanism and the time of wetness. The
authors hypothesized that the geometry of unpainted weathering steels may not favor the de-
velopment of a.protective oxide film of rust. The model structure used in the exposure trials
contained horizontal and inclined roofs, vertical wall panels, and window frames.
Two sites were chosen: an industrial location and a residential site in the Kawasaki
area, which has a cold, dry winter and a hot, humid summer. The results showed that the
successful use of weathering steel is related not only to the severity of pollution but also
to the specific interplay between shelter and the uniform washing action of rain. Thus, for
areas in which the structural factors are unfavorable, the optimal rust film forms slowly and
may deteriorate. Rust films develop and are then destroyed, and the surface never develops a
protective film. Generally, boldly exposed surfaces such as horizontal or inclined roofs show
the least corrosion.
Other variables, including amount and frequency of precipitation, and its pH level,
humidity, and temperature, also determine the impact of acidic precipitation (Kucera, 1976).
For a more thorough discussion of the role of acidic precipitation in corrosion, see
Chapter 7.
10.2.1.1.3 Influence of temperature on corrosion. Guttman (1968) and Haynie and Upham
(1974), using statistical techniques of multiple linear regression and nonlinear curve
fitting, found no significant correlation between corrosion and temperature. Other studies,
however, did find temperature to be a significant variable. Guttman and Sereda (1968) made
continuous measurements of SO , time of wetness, and temperature in their outdoor exposure
tests. The corrosion rate increased markedly with temperature. Barton (1976) found that the
effect of increased temperature was more pronounced when the rust contained little water and
2~
S0| . Haynie et al. (1976) found that temperature is a significant variable in chamber
studies of weathering steel.
The above results are in apparent conflict, but may be explained by Haynie1s (1980)
discussion of the effect of temperature on metal corrosion. He noted that if metal corrosion
were activation energy-controlled, then the logarithm of -the corrosion rate should be inverse-
ly proportional to the absolute temperature. In most cases, however, the rate of reaction is
diffusion-controlled. Whether this control occurs in the environment or in the corrosion
product film, the rate of corrosion is relatively insensitive to changes in ambient tempera-
ture. "A decrease in temperature," Haynie observed, "raises the relative humidity while it
10-10
-------
decreases diffusivity, thus normal temperature range effects on the overall corrosion rate
will most likely not be observed. Freezing should produce a step decrease in the corrosion
rate because diffusion is then through a solid rather than a liquid." (Haynie, 1980). This
last supposition is supported by Biefer (1981) in his report of the first atmospheric corro-
sivity study of the Canadian Arctic and sub-Arctic. Lowest average rates of penetration (2 to
5 |jm/yr) were recorded at mainland sites removed from the seacoast; the highest rates (21 to
34 |jm/yr) were those recorded at sites less than 1 km from the sea. Biefer (1981) attributed
the higher corrosion rates to "localized factors such as atmospheric sulfur dioxide, atmos-
pheric chlorides, and factors relating to the time-of-wetness of the corroding surface."
Sereda (1974) found that at -20°C metal corrosion is slowed, but not halted.
2-
10.2.1.1.4 Hygroscopiclty of metal sulfates. The SO, in rust stimulates further corrosion
by a mechanism related to the critical RH at which an electrolyte film is formed. The hygro-
scopicity of iron sulfates in the rust lowers the critical RH for corrosion; however, sulfates
are not the most deliquescent salts. For example, chloride and nitrate salts, which have
higher hygroscopicity than sulfates, make corrosion possible at lower humidities.
Surfaces contaminated by sea salt (mostly NaCl) can be expected to be wet when the RH
exceeds 75 percent. In contrast, calcium chloride keeps surfaces wet at an RH as low as 30
percent. A saturated solution of zinc sulfate at 20°C is in equilibrium at 90 percent RH.
Thus, zinc corroded by SOp is expected to be wet when the RH exceeds 90 percent (Haynie,
1980).
2-
10.2.1.1.5 Electrical conductivity ofrust. Barton (1976) postulated that SO, ions influ-
ence the anodic dissolution of iron as a function of their concentration at the steel-rust
interface. The corrosion rate of the rust layer is based in part on the high electronic
conductivity of rust, which allows the reduction of oxygen within the rust layer. The rate is
also influenced by the porosity of rust, which permits rapid diffusion of oxygen to the
cathode.
In the presence of S0?, FeSO, is formed before insoluble rust develops. The amount of
SOn required is small; each SO,, molecule can generate 20 to 30 molecules of rust. Once FeSO,
is formed, rusting can continue even though S0? is no longer present in gaseous form.
10.2.1.1.6 Electrical reduction of rust. Evans (1972) suggests that oxidative hydrolysis of
FeSO. occurs slowly, and would be important only in the initial stage of corrosion. He pro-
poses that there is a rate-controlling cathodic process. Thus, the corrosion products in the
ferric state would be converted to magnetite (Fe30.) by a reaction involving the reduction of
ferric oxyhydroxide (FeOOH):

Fe2+ + SFeOOH + 2e~ -* 3Fe0 + 4H0 (10-5)
10-11
-------
10.2.1.1.7 Corrosion-protective properties of sulfate In rust. The rust layer on steel is
somewhat protective against further corrosion, though far less so than the corrosion layer on
zinc and copper. The content of soluble compounds in rust limits its protection of steel.
Rust samples investigated by Chandler and Kilcullen (1968) and Stanners (1970) contained
2- 2-
2 to 2.5 percent soluble SO, and 3 to 6 percent total SO, . The outer rust layer contained a
2-
small amount (0.04 to 0.2 percent) of soluble SO. , compared with 2 percent in the inner rust
2- *
layer. The concentration of insoluble SO. was fairly uniform throughout the rust layers.
The emphasis on the composition of the rust layer has led to studies of the corrosion-
protective properties of rust as a function of exposure history (Nriagu, 1978; Sydberger,
1976). Steel samples initially exposed to low concentrations of SO and then moved to sites
X,
of higher SO concentrations corroded at a slower rate than did samples continuously exposed
to the higher concentrations. Exposure tests started in summer showed slower corrosion rates
during the first years of exposure than those started in winter.
The long-term corrosion rate of steel appears to depend on changes in the composition and
structure of the rust layer. During the initiation period, which varies with the S0? concen-
tration and other accelerating factors, the rate of corrosion increases with time (Barton,
1976). Because it is porous and nonadherent, the rust initially formed offers no protection;
in fact, it may accelerate corrosion by retaining hygroscopic sulfates and chlorides, thus
producing a microenvironment with a high moisture content. After the initiation stage, the
corrosion rate decreases as the protective properties of the rust layer improve. Satake and
Moroishi (1974) relate this slowing down to a decrease in the porosity of the rust layer.
2-
During a third and final stage, corrosion attains a constant rate and the amount of SO, in
rust is proportional to atmospheric SO concentrations.
*\
The quantitative determination and subsequent interpretation of corrosion rates becomes
difficult if it is' not known how long the metal has had a surface layer of electrolyte.
Variations in the "wet states" occur with RH, temperature, rain, dew, fog, evaporation, wind,
and surface orientation. The surface electrolyte layer may form on a metal surface as a
result of rain, dew, or adsorption of water from the atmosphere. Capillary condensation in
rust can be related to the minimum atmospheric moisture content that allows corrosion to occur
(i.e., critical RH). Centers of capillary condensation of moisture on metals can occur in
cracks, on dust particles on the metal surface, and in the pores of the rust (Tomashov 1966).
10.2,1.2 Effects of Sulfur Oxide Concentrations on the Corrosion of Exposed Metals—Most of
the laboratory studies reviewed in this section have measured corrosion rates related to
exposure to SO, alone or in combination with other compounds. In field exposure studies,
where SO almost invariably occur in combination with other airborne pollutants, an attempt is
X
made to assign separate values to SO and to describe pollutant interactive effects on corro-
sion. The discussion here unavoidably overlaps somewhat with a later section on particles,
since S0? contributes to formation of secondary sulfates. Here, the emphasis is on the direct
10-12
-------
role of SO in the corrosion process (e.g., the oxidation of SO, with moisture on a metal
/v c>
surface). In a later section, sulfates are discussed in terms of their indirect role (i.e.,
their ability to increase wetness time of a metal surface).
10.2.1.2.1 Ferrous metals. Ferrous metal products and structures are exposed widely to
ambient pollutant levels. Rusting of these metals is tne best documented form of metallic
corrosion affected by SO . This subsection reviews studies of rusting rates of ferrous
metals, such as iron, steel, and steel alloys.
A number of investigators reported data from 1959-1968 that showed that the addition of
0.05 to 0.5 percent copper to steel results in improved corrosion resistance (Larrabee, 1959;
Larrabee and Coburn, 1962; Brauns and Kail a, 1965; Schwenk and Ternes, 1968; and Barton,
1976).
Stainless steels contain more than 12 percent chromium and are widely used in outdoor
exposures. They are specified for use in many industrial processes involving corrosive
liquids that rapidly attack ordinary steels. The high corrosion resistance of stainless
steels that incorporate chromium, molybdenum, and nickel is attributed to the protective
properties of the oxide film formed on these alloys. In heavily polluted atmospheres,
however, this film is not completely protective. Particles in settled dust, including
sulfates and chlorides, can promote rupture of the oxide film and cause pitting corrosion,
which may be influenced by the surface finish (see Section 10.3 1),
The lowest alloyed stainless steels have little corrosion resistance. In particular, #13
Cr steel suffers pitting attack in industrial atmospheres. Ergang and Rockel (1975) report
that the austenitic steels of 18-percent Cr and 8-percent Ni are reasonably resistant in urban
atmospheres but have shown slight rusting in industrial areas. The rusting rate is decreased
when the steel surface is cleaned of atmospheric deposits.
10.2.1.2.2 Laboratory and field studies emphasizing ferrous metals. It is useful to consider
laboratory and field studies of corrosion effects separately because cause and effect is much
clearer in laboratory experiments; field studies are often beset by confounding variables.
Laboratory studies develop from controlled experiments, including as many variables as are
thought likely to influence damage to the material being studied. Data on materials damage
result from exposure to various concentrations of the air pollutant being studied. Analysis
of the data resulting from such laboratory studies is used to develop a mathematical expres-
sion of the relationship between the concentration of a pollutant and damage to materials.
Such a mathematical expression is called a damage function; that is, the quantitative expres-
sion of a relationship between exposure to specific pollutants and the type and extent of
damage to a target population. Factors that are shown to be significant influences in produc-
ing damage are included in the damage function.
10-13
-------
Information derived from laboratory studies is used to design field experiments, which
are performed to test the possibility of extrapolating laboratory results to ambient condi-
tions. The parameters that are measured in field studies are those found or suspected to be
important factors in the laboratory experiments. The results of laboratory experiments do not
easily translate to field situations, however, since ambient air pollution levels and other
influencing environmental factors vary widely both in time and space. Temporal variables
include fluctuations in temperature, wind moisture content, insulation, rainfall, and its
chemical characteristics (e.g., acidic rain), and atmospheric pollutant concentrations.
Spatial factors include differences such as aspect, altitude, electromagnetic fields, and
indigenous microorganisms. Initial conditions of the material being studied must also be
considered.
The results of field studies are compared with the laboratory-derived damage function.
In some cases the results are comparable, and the laboratory damage function is validated. In
other cases the data analysis may result in a markedly different damage function, with more or
fewer variables.
10.2.1.2.2.1 Laboratory studies. Spence and Haynie (1974) described the design of a
laboratory experiment to identify the effects of environmental pollutants on various materials
including ferrous metals. The environmental system consisted of five exposure chambers to
control temperature and humidity, and chill racks to simulate the formation of dew. Gaseous
pollutants included those usually monitored in field exposures: SCL, NCL, and CL. Experi-
ments were statistically designed for analysis of variance, and a system was selected to study
the interactive effects of pollutants and other variables. The effect of particles was not
included in the design. The chambers were equipped with a xenon arc light to simulate sun-
light. The system was designed to maintain air contact with the various materials at pre-
selected temperatures, RH's, flowrates, and pollutant concentrations. A dew-light cycle was
used; it produced faster deterioration than did conditions of constant humidity and temper-
ature.
Haynie et al. (1976 and 1978) exposed weathering steel in the chamber study described
above and measured concentrations of SCL, NOp, and CL in various combinations and at two
levels of pollutant concentration. Ozone was of interest since an earlier field experiment
(see Section 10.2.1.2.2, Field Studies) had indicated that the presence of oxidants inhibited
metallic corrosion. The corrosion rate was measured by loss in weight of the weathering
steel. Six panels each were exposed to 16 polluted-air and 4 clean-air conditions; measure-
ments were taken at 250, 500, and 1000 hours of exposure. The weight losses were converted to
equivalent thickness loss values. As expected, corrosion was most severe at high S02 concen-
trations and high humidity. Ozone neither inhibited nor accelerated corrosion. The authors
concluded that some other oxidant or unmeasured factor that was covariant with 0, caused the
inhibition effect. If the data from the sites with high oxidant concentrations in the field
exposure experiments were excluded, however, the damage from the laboratory study was an
10-14
-------
excellent predictor of the field results. The coefficient of determination for the field data
using the laboratory function was 0.986. The following physical damage functions were deve-
loped by Haynie et al. (1976) to relate SO, concentrations to weathering steel panels and
2
galvanized steel corrosion, respectively (R = 0.91).

corrosion = [5.64 VS5J + e (55'44 " 31>150/RT)] /j^ (10-5)
corrosion = (0.0187 S02 + e 41'85 ' 23,240/RT) ^ (1Q_7)
where:
corrosion is expressed in (jm thickness lost
SOp = sulfide dioxide concentration expressed in (jg/m
f = fractional time of panel wetness
w
t = time of wetness in years
W
R = the gas constant (1.9872 cal/gm mol/K)
T = geometric mean temperature of panels when wet, K.
Sydberger and Ericsson (1976) studied the corrosion of mild steel at 1, 10, and 100 ppm
(2620, 26,200, and 262,000 M9/m3) S02 across the range of cn'-Hcal humidities (80 to 96 per-
cent RH). The flowrate of the S02 atmosphere was varied, and some samples were sprayed with
water to simulate rain or condensation. The chemical composition of the corrosion products
was studied by X-ray diffraction, infrared spectrometry, and electron spectroscopy for chemi-
cal analyses (ESCA) techniques. The flowrates of the S0? atmospheres markedly influenced the
corrosion rates. It appears that corrosion rates are related not only to the SO, concen-
tration in the atmosphere, but also to the supply of S0? per unit surface area and time.
£. y
Spraying the samples with distilled water at intervals substantially increased corrosion.
Sydberger and Ericsson's (1976) based their analysis of the corrosion product (rust
2-
layer) on the concepts of Schwarz (1972) and Barton (1976) that SQ4 is the primary corrosion
stimulant in rust formation. Anodic activity is maintained by the concentration of ferrous
2
sulfate in the electrolyte. An S0? supply of 4 |jg/cm /hr at the lowest humidity initiated
corrosion at a low rate. A rise above 50-percent RH increased corrosion markedly. Of par-
3
ticular interest was the finding that different flow rates at 1 ppm (2620 (jg/m ) S0? with
96-percent RH gave significantly different corrosion rates. This study of the effect of rust
O —
on corrosion showed that even at high humidity and high SO, content, the corrosion rate
decreased to a low level when the SO, concentration was low.
10.2.1.2.2.2 Field studies. For outdoor exposures, eddy diffusion is the primary rate-
controlling factor in the delivery of pollutants to a surface. This flux is not constant and
is a function of the horizontal wind velocity gradient away from the surface. The transport
of a pollutant to a surface is usually expressed as a "deposition velocity", defined as the
flux to the surface divided by the ambient pollution level at some specific measuring height.
*
10-15
-------
Reported deposition velocities for gaseous pollutants have usually been within an order of
magnitude of 1 cm/sec. These values are consistent with calculated estimates based on an
analogy with momentum flux and measured wind velocity profiles (Sydberger, 1976).
The amount of S0? reaching a steel surface depends on wind direction, wind velocity, and
the orientation of the surface to the emission source. The concept that S02 deposition varies
with flow direction and velocity suggests that data on concentration alone cannot be used to
determine the supply of S0« to metal surfaces; therefore surface adsorption methods like the
lead candle method provide valuable information in relating supply of SO to metal surfaces
A,
(Sydberger, 1976). Upham's (1967) work indicated,"however, that corrosion of mild steel at
seven Chicago sites increased with time and with increasing mean SO- concentration (Figure
10-5). Difficulties in relating sulfation measurements to ambient SCL measurements are dis-
cussed in Chapter 3.
Haynie and Upham (1971) continuously monitored urban pollutants including SCL, N02, and
0, (oxidants) to determine whether previously unconsidered variables might affect steel cor-
rosion. They also considered temperature, RH, and TSP. Their 5-year program, begun in 1963,
involved sites in Chicago, Cincinnati, New Orleans, Philadelphia, San Francisco, Washington,
Detroit, and Los Angeles. They studied three types of steel expected to show different levels
of resistance to atmospheric corrosion: (1) a plain carbon steel containing some copper (0.1
percent copper); (2) a copper-bearing steel (0.22 percent copper); and (3) a low-alloy
weathering steel (0.4 percent copper with 0,058-percent phosphorus). The exposure periods
were 4, 8, 16, and 32 months. The same steels were exposed at rural sites as a control. The
rural sites proved to have higher-than-expected corrosion rates; however, S02 concentrations
were not measured at these sites. Multiple regression analysis established significant cor-
relations between average SQp concentrations and corrosion of all three steels at the urban
sites. Temperature was not a statistically significant variable. Average RH was insignifi-
cant because the ran'ge was only between 62 and 77 percent.
Inspection .of the monitored S02 and oxidant concentrations revealed wide variations from
site to site. Multiple-regression analysis showed that high concentrations of oxidants cor-
related with lowered metallic corrosion rates. The resulting physical damage functions
included terms for S0?, oxidants, and time of exposure. A more recent laboratory investiga-
tion by Haynie et al. (1976), however, (see preceding section) has shown that 0« is not a
significant corrosion controlling variable.
Hansfeld (1980) made observations at nine test sites in and around St. Louis for 30
months beginning in October 1974 as part of the Environmental Protection Agency's Regional Air
Pollution Study to determine the effect of airborne pollutants on galvanized steel, weathering
steel, stressed aluminum, marble, and house paint—essentially the same materials examined in
the chamber study reported by Haynie et al, (1976). During 1975 and 1976, atmospheric cor-
rosion monitors (ACM) of the type described by Mansfeld and Kenkel (1976) were installed at
10-16
-------
18.0
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18
0 (50) (100) (150) (210) (260) (310) (360) (415) (470)

MEAN SO2 CONCENTRATION, ppm (;ug/m3)

Figure 10-5, Relationship between corrosion of mild steel and cor-
responding mean SC>2 concentration is shown for seven Chicago sites.
(Corrosion is expressed as weight loss of panel.)

Source: Upham (1967).
10-17
-------
four sites to measure time of wetness. Each ACM consists of a copper-zinc or copper-steel
couple that registers current flow when an electrolytic path forms between"the two plates as a
result of deposition of water from the air, dew, or rain on corrosion products. The ACM
measures the time that the panel is wet enough for the electrochemical mechanism of corrosion
to occur. Parameters measured in Mansfeld's study included windspeed, wind direction, temper-
ature, 07 concentration, total hydrocarbon concentration, total sulfur and NO concentration,
•* 9- O
hydrogen sulfide concentration, S02 concentration, RH, SO. concentration, NO, concentration,
TSP concentration, and time of wetness.
Mansfeld determined weight losses for galvanized steel, weathering steel, house paint,
and marble (he removed aluminum tension samples after failure). His data show that damage to
a particular material does not necessarily occur at the same corrosion rate at each site.
Preliminary statistical analysis of the results failed to.show significant correlation between
corrosivity and pollutant concentrations. There was substantial error in the measurement of
RH, an extremely important corrosion variable. Relative humidity was, therefore, not included
in the regression analyses. Concentrations of S09 measured by Mansfeld (1980) were generally
" 3 \
an order of magnitude lower than the 130 ppb (340 (jg/m )1 concentrations reported by
(1967) at urban sites in St. Louis.
In another study Haynie and Upham (1974) exposed enameli
ib steel containing 0.019-percent
carbon and 0.028-percent copper at 57 site,s in National Air Surveillance Network (NASN). They
measured corrosion by weight loss and quantitatively determined other pollutants, including
2~
gaseous SOp, TSP, and the amount of SO. and NO, in the particles. (For a discussion of the
measurement methodology for NASN, see Chapter 3.) Sulfur /dioxide levels ranged from 9 to
7 \ 3 ' /
374 |jg/m , TSP from 11 to 182 (jg/m , and RH from 29 to 76 percent. The average temperature
/ '
remained within a fairly narrow range and was considered constant. The temperature, the
quantity of total particles, and the presence of NO, in /the particles did not significantly
/ O —
affect the corrosion rate of stee]. The concentration of S09 was significant only when SO,
' 2~
was not included in the regression analysis. At each/site, SO* content of the particles and
/ ^
SOp concentration were closely related. On the basis of this study, Haynie and Upham derived
the following empirical expression to obtain the best relationship between corrosion of ename-
ling steel and atmospheric SO^ content:

Corrosion = 325Vt e^0"00275 S0* ~ 163"2/RH> (10-8)
where: corrosion is expressed as (jm
o
SOp = sulfur dioxide expressed as (jg/m
RH = average relative humidity (percent)
t = time in years
Average RH is a substitute variable for the fraction of time the steel is wet.
10-18
-------
Considerable effort has gone Into isolating environmental variables that predict long-
term corrosion rates. Empirical expressions for corrosion of various steels exposed to the
atmosphere (see Table 10-1) have been developed by Chandler and Kilcullen (1968), and Haynie
and Upham (1974). These equations may be used to relate reduction in S02 and sulfates to
reduction in corrosion of metals, serving as a basis for a benefit appraisal.
All of the above-discussed damage functions are statistical models developed from regres-
sion analysis of the correlations of air chemistry and effects data resulting from laboratory
or field exposure experiments. Such functions vary in form, reflecting differences in param-
eters measured and/or measurement methodology. For instance, if the function includes a term
2
for SO, expressed as ug/cm /day, then deposition velocity is already accounted for. If,
3
alternatively, S0» is expressed as ug/m , then deposition velocity may be accounted for in the
equation. Barton (1976) proposed an equation for metal corrosion rate which combines both
kinetics of corrosion processes specific to the metal of interest and results of regression
analysis. Such an equation may be more universally applicable, although the statistical
damage functions discussed previously should be applicable across a range of environmental
conditions normally encountered in the United States. Barton's proposed equation is:

u = M • Tn • Sm (10-9)
IN
where
u = corrosion rate of metal in um/yr
K
M = constant involving specific corrosion kinetics of the metal
i = time of wetness, long-term average, expressed as hrs/day with RH > 80%
and temperature >^ 0°C
2
S = SOp expressed as mg/m /day
n and m = constants determined by regression analysis

10.2.1.2.3 Comparison of ferrousand nonferrous metals. Sydberger and Vannerberg (1972)
examined the influence of RH and rust on the adsorption of SO, on metal surfaces, using radio-
3
active sulfur. The concentration of SO, was 0.1 ppm (262 ug/m ), and RH varied between 50 and
98 percent. Polished and preexposed samples of iron, zinc, copper, and aluminum were compared
for their adsorption properties. Iron, zinc, and copper were preexposed to S0« concentrations
53
(100 ppm or 2.6 x 10 ug/m ) at 98 percent RH and 22°C for 3 hours. The aluminum samples were
preexposed for 30 hours. The principal corrosion product identified by X-ray diffractometry
was hydrated metal sulfate. Adsorbed SO- was measured at 30-minute intervals with a Geiger
counter. The corrosion rate at 90-percent RH was initially high for zinc and copper but quite
low for aluminum. Adsorption of S0? on preexposed iron samples was high. At 80-percent RH,
almost all of the SO,, was adsorbed. The high adsorption rate is perhaps explained by the
rapid oxidation of adsorbed S0~ caused by the catalytic effect of the rust.

10-19
-------
TABLE 10-1. SOME EMPIRICAL EXPRESSIONS FOR CORROSION
OF EXPOSED FERROALLOYS
Exposure Study
Material Site Duration Empirical Equation
Steel A* Sheffield 1 year y = 0.51 + O.Olx
England
Parameter Units
y = corrosion rate
in mils/year ,
x = SOg in ug/M
Source
Chandler and
Kilcullen
(1968)
Note/Coanents
Authors stated that S0? and smoke
has a major influence on the corrosion
rate of steel and accounted for about
50% of the variations found at the
different sites. Other factors, such
as tine of wetness, were found equally
important in determining the corrosion
rate of steel.
95% confidence limit + 0.75 mils for
any point on regression.
Steel B*
Enameling
steel
Sheffield 1 year
England
y = 0.82 + 0.006x
NASN
sites
1-2
years
corr =183.5
L"°-OM21sul-<163.2/RH)]
o
i
l\3
O
Enamel i ng
steel
NASN
sites
1-2
years
ff e[0.00275 SOZ-(163.2/RH)]
y - corrosion rate
in mils/year ,
x = SQ£ in \tg/m

corr = depth of
corrosion, \int
sul = average level
of sulfate in suspended
particulate, M§/B
RH = average relative
humidity, percent
t = time in years

parameter units
same as for previous
equation except
SO, = ug/ia SO,
Chandler and
Kilcullen
(1968)

Haynie and
Upman
(1974)
Haynie and
Upham
(1974)
95% confidence limit + 0.5 mils for
any point in regression.
strf. dev. 85.0; 95% conf. limits 13.4
lower, 353 upper
sid. dev. 0.00451; 95X conf. limits
0.0552 lower, 0.0732 upper
std. dev. 32.8; 95% conf. limits 97.6
Derived by substitution from above
equation, based on sufate and SO,
relationship defined as: sul =8.9
+ 0.0429 SO,, std. deviation ± 8.2.
Note: I ppm SOg = 2620 ug/m .

* composition of steels A and B are as follows:
El eitient

carbon
silicon
sulfur
phosphorous
manganese
tin
copper
nickel
aluminum
% Steel, A

0.058
trace
0.024
0.013
0.335
0.015
0.060
0.040
0.015
X Steel B

0.20
0.04
0.052
0.038
0.58

0.28
0.10
-------
The initial rate of adsorption on polished iron below 80 percent RH is related to the absence
of corrosion; however, at increased humidity corrosion is initiated and the adsorption rate
increases.
The observation that S02 adsorption can take place at humidities below the critical
humidity (Sydberger and Vannerberg, 1972) is of particular significance. This suggests that
SO, will be adsorbed on a rusty iron- surface during periods of low humidity and will affect
the corrosion rate when humidity rises. Table 10-2 shows the critical humidities for non-
ferrous metallic surfaces, as summarized from Nriagu (1978) and National Research Council (NRC,
1979). The corrosion products of copper and aluminum have an extremely low adsorption
capacity below 90-percent RH, confirming the lower sensitivity of these metaT's to corrosion by
SQ? (see Figure 10-6, from Sydberger and Vannerberg, 1972).

TABLE 10-2. CRITICAL HUMIDITIES FOR VARIOUS METALS
Metal Critical humidity, percent RH

Aluminum 75-80
Brass 60-65
Copper 65-70
Nickel 65-70
Zinc 70-75
Aluminum is generally considered to be corrosion resistant. It is quite resistant to SO
A.
when RH is less than 50 percent. At higher humidities aluminum can corrode rapidly in the
presence of very high S0? concentrations forming a hydrated aluminum sulfate [Alp(SO»)o '
ISHpO] surface deposit. At low concentrations of acid sulfate particles it forms a protective
film of aluminum oxide. When the film becomes contaminated with dirt and soot particles,
however, there is a change in surface appearance characterized by mottling and pitting.
Simpson and Horrobin (1970) reported that aluminum undergoing long exposure in industrial
areas displayed white areas of crystalline corrosion products. Aluminum surfaces exposed for
2- 3
more than 5 years to an SO, concentration of 550 |jg/m had pits as deep as 14 mils (0.36 mm).
This is, of course, some 50 times higher than typical ambient concentrations (see Chapter 5).
Fink et al. (1971) summarized measured corrosion rates and depth of pitting of aluminum
3 3
surfaces in rural, mild industrial (30 (jg/m or 0.01 ppm S0?), normal industrial (370 jjg/m or
0.14 ppm S0~ and 80-percent RH), and severe industrial areas. Their overall conclusion was
that, although some loss of thickness occurred in the first 2 years, structures composed of
aluminum and its alloys are resistant to air pollutants.
10-21
-------
T
o
O
tn
O
ui
CQ
K
O
CO
a
I I
COPPER
ALUMINUM
34567

EXPOSURE TIME, hr
10
Figure 10-6. Adsorption of sulfur dioxide on polished metal surfaces
is shown at 90 percent relative humidity.
Source: Adapted from Sydberger and Vannerberg (1972).
In another study, Haynie (1976) found that stressed aluminum specimens exposed to S02 at
concentrations of 79 (0.03 ppm) and 1310 MS/m3 (0.5 ppm) lost approximately 8.6 and 27.6
percent of their bending strength, respectively. He also noted that 7005-T53 high-strength
aluminum alloy tubing, which contains very little copper, is susceptible to stress-corrosion
cracking in industrial environments.
Abe et al. (1971) exposed copper and copper alloys for 2 years in marine, rural, highly
industrial, and urban areas in which there was great variation in' pollutant and salt content.
Analyses of the surface deposits showed the presence of basic sulfate, sulfide, and chloride
of copper. The green patina on copper was analyzed and determined to be basic copper sulfate
in urban areas and basic copper chloride in seacoast areas. These surfaces were protective
against further corrosion.
10-22
-------
According to Simpson and Horrobin (1970), the formation of these basic copper salts can
2~
take as long as 5 or more years and will vary with the concentration of SO, or chloride
particles, the humidity, and the temperature. They reported the corrosion rate of copper to
be 0.9 to 2.2 pm/year in industrial atmospheres, compared with 0.1 to 0.6 pm/year in rural
areas.
Sydberger (1976) attributed the high corrosion resistance of nickel and copper compared
2~
with unalloyed steel to the ability of these metals to form a layer of insoluble basic SO,
that protects the metal surface. Such layer formation does not occur on steel.
10.2.2 Protective Coatings
Susceptible materials are generally coated for protection against the effects of ex-
posure. The coatings provide either sacrificial protection or barrier protection. In galva-
nization, zinc is applied to ferrous metal for sacrificial protection. Thus, while the gal-
vanized surface may suffer corrosion damage, it helps to prevent rusting of steel products
such as gutters, cables, wire fencing, and building accessories. Barrier protection is pro-
vided by varnishes, lacquers, and paints by sealing the underlying surface material against
intrusion by moisture.
10.2.2.1 Zinc-Coated Materials—Zinc is generally exposed as a protective coating for steel
products since zinc coating is fairly resistant to atmospheric corrosion. Zinc is anodic with
respect to steel; when zinc and steel are in contact with an electrolyte, the electrolytic
cell provides current to protect the steel from corrosion with some oxidation of the zinc.
Guttman (1968) carried out a long-term exposure of zinc panels with measurement of the
atmospheric factors. He found that zinc is corroded by SO,, and that time of wetness and
concentration of S0? are the major factors that determine the rate of corrosion.
Fleetwood (1975) conducted 5-year exposure studies of zinc and iron in a number of loca-
tions ranging from dry tropical to industrial. He estimated the service life of galvanized
steel to be 15 to 20 years in an industrial area containing pollutants and 300 years in a dry
tropical unpolluted area. Kucera (1976) noted strong correlations between the corrosion rate
and (1) the adsorption of SO,, on zinc surfaces and (2) the concentrations of SO,.
Haynie and Upham (1970) exposed zinc panels in eight cities, continuously monitoring SOp
concentration, and collecting meteorological data, including temperature and RH, from the
nearest weather stations. They developed the following empirical equation, which correlates
3
corrosion rate with average SO, concentration (for the study range of 10 to 479 pg/m ) and RH:

Y = 0.001028 (RH - 48.8) S02> (10-10)
where
Y = zinc corrosion rate (pm/yr),
RH = average annual RH in percent, and
SO, = average SO, concentration (ns/m )•
10-23
-------
The regression intercept indicated that no corrosion would occur below an average RH of 48.8
percent. This expression gave a reasonably good linear fit with the experimental corrosion
results obtained by Haynie and Upham for S02 concentration and RH.
Haynie (1980) performed a regression on the products of the coefficients of the studies
listed in Table 10-3 and S02 levels. The results of the regressions yielded 1.087 with a
standard deviation of ± 0.283. Based on this analysis and the St Louis study results, Haynie
(1980) expressed the relationship of the corrosion of small specimens of galvanized steel as
follows:

C, = 2.32 t + 0.0134 v 0'781 - S00 - t ,, (10-11)
i W £ W
C = corrosion in pm
t = time of wetness in years,
v = wind velocity in m/s
so2 =
Because the equation was derived from the results of regression for several studies, not
2
an original data set, no R was calculated. A theoretical damage coefficient (for present
purposes, damage = any measurable adverse effect) for a pollutant can be calculated from the
stoichiometry of a reaction and the deposition velocity. For the reaction between S02 and
zinc to form zinc sulfate, the coefficient is 0.045, when the zinc corrosion rate is expressed
in micrometers per year, S02 in micrograms per cubic meter, and the deposition velocity (u) in
centimeters per second. For a small zinc or galvanized steel sheet specimen, the material
0 7R
damage coefficient for S02 is calculated to be 0.0123 v " , which agrees well with the above
determined empirical coefficient. At a wind velocity of 4 m/sec, the value is 0.0363 (nm/yr)/
2
(|jg/m ). For the same conditions, a similar calculation for marble yields a coefficient of
0.136 djm/yr)/0jg/m5).
Haynie (1980) restudied the results of six exposure investigations to relate the corro-
sion of zinc and galvanized steel to the concentration of S02. Each investigation was dif-
ferent and the data were evaluated differently, thus, no direct comparison of the results as
they were published was possible. In reevaluating the data from each study, however, Haynie
used techniques that permitted the comparison of the various data sets.
Table 10-3 compares the experimental regression coefficients obtained from all of these
studies. The S0? coefficient for the chamber study is low, whereas the analogous coefficients
for the Community Air Monitoring Program (CAMP) (Haynie and Upham, 1970) and Interstate
Surveillance Program (ISP, Cavender et al . , 1971) studies are high and agree with each other.
The remaining three S02 coefficients are generally in good agreement. The time-of-wetness
coefficients are all within a range of + 0.75 from a mean of 1.73 (jm/yr.
10-24
-------
TABLE 10-3. EXPERIMENTAL REGRESSION COEFFICIENTS WITH
ESTIMATED STANDARD DEVIATIONS FOR SMALL ZINC AND
GALVANIZED STEEL SPECIMENS OBTAINED FROM SIX
EXPOSURE SITES
Study
Time-of-wetness
coefficient,
S0? coefficient,
(lJi/yr)/(Mg/m )
Number of
data sets
CAMP (Haynie and
Uphara, 1970)
ISP (Cavender et al.,
1971)
Guttman, 1968
Guttman and Sereda,
1.15 + 0.60

1.05 + 0.96
1,79
0.081 + 0,005

0.073 + 0.007
0.024
37

173
>400
1968
Chamber study (Haynie
et al . , 1976)
St. Louis (Mansfeld,
1980)
2.47 +
1.53 +
2.36 +
0.86
0.39
0.13
0.027
0.018
0.022
+ 0.008
+ 0.002
+ 0.004
136
96
153

Note: 1 ppm S02 = 2620 ug/m .

The specified thickness of galvanized coating varies with intended use. Furthermore, the
thickness of a particular coating varies considerably from one point to another. Bird's
(1977) measurements revealed that 5 percent of coating thickness measurements varied from- the
mean by more than 46 percent. Haynie confirmed this variability with 475 thickness measure-
ments on a single galvanized steel sheet. The life of a coating is generally proportional to
its thickness; thus, rusting of the substrate steel will occur first at the thinnest spots and
last at the thickest spots. As reported in Haynie (1980), The American Society for Testing
and Materials (ASTM) has observed rusting at thin spots on galvanized steel wire, fencing, and
sheet exposed to various types of atmospheres over many years. Some of their exposures were
started in 1916 and continued until the test could reveal no additional information. In the
case of sheet, the product was completely rusted and showed perforations. In general, the
amount of corrosion at each site varied linearly with time. Corrosion rates at each site were
calculated on the basis of time to first rust and time to complete rust for various zinc
thicknesses, assuming +40 percent thickness variability.
10-25
-------
Originally, State College, Pennsylvania, was selected by ASTM as a control site represent
ing a "clean" rural environment; however, the corrosion was higher there than at five other
rural locations including the rural-marine environment of Santa Cruz, California, where high
RH is expected to accelerate corrosion. ASTM made no pollution measurements but recognized
the effects of "industrial" and "severe industrial" environments.
Zinc corroded nearly twice as fast on wire and fencing as it did on sheet, a finding that
is consistent with the theoretically predicted effects of surface configuration on S0? deposi-
tion velocity. One would expect a greater deposition velocity onto fencing than onto sheet
material. ASTM noted that fencing corrodes less near the ground than it does near the top be-
cause wind velocity increases with height, with a resultant increase in deposition velocity
and similar variation of S0? concentration. Another factor may be the decrease in concentra-
tion near the ground, since both soil and vegetation are sinks for S02 (see Chapters 6 and 8).
The average corrosion rates in the Haynie (1980) study shown in Table 10-4 correspond to
actual corrosion rates that are two to four times greater when the substrate is wet. Theore-
tical calculations indicate that the average S09 levels at the Pittsburgh site over the long
3
exposure period were between 350 and 700 jjg/m . The average at the Altoona site could have
o
been as high as 1000 ug/m .

TABLE 10-4. CORROSION RATES OF ZINC ON GALVANIZED STEEL PRODUCTS EXPOSED TO
VARIOUS ENVIRONMENTS PRIOR TO 1954

Mean corrosion rate and estimated standard
deviation, urn/year
Site Sheet Wire and fencing
Altoona, PA 7.57 + 0,54
Pittsburgh, PA 5.63 + 0.34 10.86 + 1.02
Sandy Hook, NJ 2.74 + 0.30 4.37 + 0.45
Bridgeport, CT — 4.25 + 0.44
Lafayette, IN — 2.94 + 0.34
Ithaca, NY — 2.68 + 0.42
State College, PA 1.27 + 0.29 2.48 + 0.24
Ames, IA — 1.68 + 0.19
College Station, TN — 1.22 + 0.43
Santa Cruz, CA -- 0.83 + 0.26
Manhattan, KS -- 0.79+0.27
Davis, CA — 0.76 + 0.42

Source: Haynie (1980).
10-26
-------
From the relationships between theoretical and experimental studies, Haynie (1980) con-
cluded the following:
1. Both short-term laboratory evidence and long-term exposure results for galvanized
steel are consistent with theoretical considerations.
2. Damage functions for some materials can be calculated from theoretical relationships
that consider factors controlling time of wetness and pollutant fluxes.
3. Wind speed and material geometry should be considered in evaluating atmospheric cor-
rosion effects.
Harker et al. (1980) examined the variables controlling the corrosion of zinc by SO- and
H,SO. using an aerosol flow reactor. Under steady-state conditions, they made the following
measurements:
Environmental measurements:
(1) Percent RH and temperature (at two points).
(2) Average flow velocity (Pitot tube).
(3) Flow velocity profile (recorded when a steady state had been estab-
lished).
Aerosol measurements:
(1) Aerosol size distribution and number concentration determined at
intervals during test by TSI 3050 analyzer.
(2) Two total-mass filter samples collected.
(3) Total-deposition sample collected on aluminum foil throughout each
experiment.
(4) TEM deposition grid samples collected continuously.
(5) X-ray photoelectron spectroscopy samples (both zinc plate and
aluminum foil) collected continuously during experiment.
Corrosion rate measurements were recorded continuously by an atmospheric corrosion monitor
(ACM), which had been pretreated with either 0.1 N H^SO. or ammonium sulfate. Experimental
conditions were selected from the following ranges:
Temperature, °C 12-20
Relative humidity, percent 65-100
Mean flow velocity, ra/sec 0.5-8
Sulfur dioxide concentration, ppb (volume) 46-216
Sulfate aerosol mass concentration, mg/m 1.2
Aerosol size distribution, urn diameter 0.1-1.0
In the Harker et al. (1980) study the factors controlling the rate of corrosion were RH,
pollutant flux, and chemical form of the pollutant. Corrosion occurred only at RH high enough
(more than 60 percent) to wet the surface; temperature did not appear to be a controlling
factor within the range 12 to 20°C. The results indicate that on initial exposure SOg-induced
corrosion of zinc proceeds at a rate approximately a factor of two greater than that for the
equivalent amount of deposited H^SQ* aerosol.
10-27
-------
2—
The investigators noted deposition velocities of 0.07 cm/sec for 0.1-1.0 urn S(n aerosols
and 0.93 cm/sec for S02 at a friction velocity of 35 cm/sec. These factors indicate that the
effects of SO, will dominate the effects of H?SO, in most urban areas.
10.2.2.2 Paint Technology and Mechanisms of Damage—Compared with other environmental factors
such as sun and precipitation, paint damage due to air pollutants is considered less im-
portant. There are at present no standard ASTM procedures for evaluating the effect of SCL,
NO-, and/or 0, on paints. Degradation by ultraviolet light has received the major emphasis;
outdoor test stations have been located where SCL levels are low.
Paint erosion can be measured by less of thickness of the paint layer, which can result
from the chemical action of S0? and the action of light and 0,. Film erosion rates are used
by paint manufacturers to determine the fail point for their formulations.
In paint formulas, the ratio of pigments to film formers is important to the overall
properties of gloss, hardness, and permeability to water. If the amount of film former is too
low, soiling is increased and the paint may lose the film flexibility needed for durability
and become brittle. Hay and Schurr (1971) reported on the permeability of paints to water.
High-permeability films are desirable for surfaces that must allow water to pass through, such
as wooden exterior walls behind poorly ventilated kitchens. Low-permeability coatings are
needed to protect surfaces that corrode when repeatedly moistened. The low permeability of
chlorinated rubber, as well as styrene-acrylic, is advantageous for use on concrete.
2-
Paint films permeable to water are also susceptible to penetration by S0« and SO*
aerosols. The absorption of S02 was observed by Holbrow (1962), who found sulfites and sul-
fates in paint, and by Walsh et al. (1977), who used radioactive SO- to determine rates and
saturation values for S02 absorption.
Concentrations of S0? found in fog or near industrial sites can increase the drying and
hardening times of certain kinds of paints. Holbrow (1962) found that the drying time of
linseed, tung, and certain castor oil paint films increased by 50 to 100 percent on exposure
to 2620 to 5240 MS/"" (1 to 2 ppm) SO^. The touch-dry and hard-dry times of alkyl and oleo-
resinous paints with titanium dioxide pigments were also reported to increase substantially;
however, the exposure time of the wet films was not reported. Analysis of the dried films in-
dicated that S0? had chemically reacted with the drying oils, altering the oxidation-poly-
merization process. No studies have been reported on the effects of SO- on the drying of
latex paints.
Holbrow (1962) also studied the effects of S0? on dried paint film. In these experi-
ments, paint films were allowed to dry, were refrigerated, and then exposed for 15 minutes to
an atmosphere containing 1.2 percent S02. The paint films with condensed moisture were
finally placed in an accelerated-weathering chamber. For all the paints except a pentaery-
thritol alkyd paint, the gloss decreased significantly after 1 day in the accelerated-weathering
10-28
-------
chamber. Without the accelerated weathering, the actions of SO, and moisture on the paint
films produced only a slight reduction in gloss. Holbrow concluded that the SO- had sen-
sitized the film, permitting water to be absorbed during the weathering cycle.
Bluing of green pigmented paint containing lead chromate has been observed during the
early life of the film. Holbrow (1962) reproduced this effect in the laboratory by exposing
the film to SO^ and moisture and then to warmth and^moisture. The bluing was probably caused
by conversion of the yellow lead chromate to colorless lead sulfate. Holbrow did not attempt
to correlate moisture, temperature, and pollutant concentration or to obtain dose-response
data. Although very high levels of SO, were used, this experiment indicates that condensation
and moisture evaporation are critical in concentrating the pollutant on the s-urface of exterior
paint films; under these conditions, the film deteriorates.
Svoboda et al. (1973) compared pigmented and unpigmented paint film for S02 permeability
and found that the rate of penetration of S0? into a paint film was related to the pigment
content. Zinc oxide and titanium dioxide pigments caused a 50- to 70-percent decrease in the
rate of penetration of S0? into the paint film.
Spence et al, (1975) carried out a chamber study of the effects of gaseous pollutants on
four classes of exterior exposure paints: oil-base house paint, vinyl-acrylic latex house
paint, and vinyl and acrylic coil coatings for metals. The house paints were applied to
aluminum panels by spraying. The coil coating panels were cut from commercially painted
stock. The oil-base paint film was 58 pm thick; the acrylic latex, 45 Mm> "the vinyl coil
coating, 27 urn; and the acrylic coil coating, 20 \M. The exposure chambers controlled temper-
ature, humidity, S0?, NQ~, and 0,. Each exposure chamber had a xenon arc lamp to provide
ultraviolet radiation. A dew/light cycle was included; light exposure time was followed by a
dark period during which coolant circulated through racks holding the specimens, thereby
forming dew on the panels. Each dew/light cycle lasted 40 minutes and consisted of 20 minutes
of darkness with formation of dew, followed by 20 minutes under the xenon arc. The'total
exposure time was 1000 hours. Damage was measured after 200 hours, 500 hours, and 1000 hours
by loss of both weight and of film thickness. In evaluating the data, loss of weight was
converted to equivalent loss of film thickness.
Visual examination of the panels coated with oil-base house paint revealed that all ex-
posure conditions caused considerable damage. The erosion rate varied from 28.3 to 79.1
(jm/year, with an average of 60 urn/year. The investigators concluded that SO,, and RH markedly
affected the rate of erosion of oil-base house paint. The presence of N0« increased the
weight of the paint film. A multiple linear regression on S0? concentration and RH yielded
the following relation:

E = 14.3 + 0.0151 SO, + 0.388 RH, - (10-12)
10-29
-------
where
E = erosion rate in |jm/year,
SO,, = concentration of S0? in |jg/m
RH = RH in percent.
The authors reported the 95-percent tolerance limits on 99 percent of the calculated rates to
be ± 44 urn/year.
Blisters formed on acrylic latex house paint at the high SO- levels. The blisters
resulted from severe pitting and buildup of aluminum corrosion products on the substrate. The
paint acted as a membrane retaining moisture under the surface and excluding oxygen that would
passivate the -aluminum. The vinyl coating and the acrylic coating are resistant to S0?. The
visual appearance of the vinyl coil coating showed no damage. The average erosion rate was
low, 3.29 urn/year. The average erosion rate for a clean air exposure was 1.29 |jm/year. The
acrylic coil coating showed an average erosion rate of 0.57 |jm/year.
A study of the effects of air pollutants on paint was conducted by Campbell et al.
(1974). The paints studied included oil and acrylic latex house paints, a coil coating,
automotive refinish, and an alkyd industrial maintenance coating. These coatings were exposed
to clean air, S02 at 262 and 2620 |jg/m3, and 03 at 196 and 1960 |jg/m3 (i.e. , equivalent to 0.1
and 1.0 ppm of each pollutant). Other controlled study variables included light, temperature,
and RH. In addition, one-half of the coatings were shaded during the laboratory exposures.
Similar panels (half facing north) were exposed at field sites in Leeds, North Dakota;
Valparaiso, Indiana; Research Center, Chicago, Illinois; and Los Angeles, California.
The laboratory exposure chamber operated on a 2-hour light-dew cycle (i.e., 1 hour of
xenon light at 70 percent RH and a temperature of 66°C followed by 1 hour of darkness at 100
percent RH and a temperature of 49°C). Coating erosion rates were calculated after exposure
periods of 400, 700, and 1000 hours. Table 10-5 summarizes the estimated erosion rates and
statistical characterizations of the results. Erosion rates at 03 or SO^ concentrations of
0.1 ppm were not significantly different from values for clean air exposures due to high
variability of the data. The erosion rates on the shaded specimens were significantly less
than the unshaded panel results shown in Table 10-5; panels facing north were also less
eroded. At 1 ppm pollutant concentrations, erosion rates were significantly greater than con-
trols, with oil-base house paint experiencing the largest erosion rate increases, latex and
coil coatings moderate increases, and the industrial maintenance coating and automotive
refinish the smallest increases (Yocom and Grappone 1976; Yocom and Upham 1977; and Campbell
et al., 1974). Coatings that contained extender pigments, particularly calcium carbonate,
showed the greatest erosion rates from the SO,, exposures. Results of field exposures also
support these conclusions (Campbell et al., 1974).
10-30
-------
TABLE 10-5. PAINT EROSION RATES AND T-TEST PROBABILITY DATA

PAINT EROSION RATES AND PROBABILITY DATA (T-TEST)
FOR CONTROLLED ENVIRONMENTAL LABORATORY EXPOSURES

Type of paint
House paint
oil
latex
Coil coating
Automotive refinish
Industrial maintenance
Mean erosion rate
limits)
Clean air
control
5.11 + 1.8
0.89 ± 0.38
3.02 ± .58
0.46 ± 0.02
4,72 ± 1.30
(nm/hour with 95-percent
for unshaded panels
S02
(1.0 ppm) (1
b35.8 ± 4.83 bll.
b2.82 ± 0.25 a2.
b8.66 ± 1.19 a3.
0.79 ± 0.66 bl.
5.69 ± 1.78 7.
confidence
03
. 0 ppm)
35 ± 2.67
16 ± 1.50
78 ± 0.64
30 ± 0.33
14 ± 3.56
PAINT EROSION RATES AND PROBABILITY DATA (T-TEST)
FOR FIELD EXPOSURES

Mean erosion rate (nm/month with 95-percent confidence
limits) for panels facing south
Type of paint
House paint
oil
latex
Coil coating
Automotive refinish
Industrial maintenance
Rural
(clean air)
109 ± 191
46 ± 13
53 ± 20
23 ± 28
91 ± 41
Suburban
b376 ±
b76±
b254 ±
a58 ±
a208 ±
124
18
48
18
361
Urban
(S02 dominant,
-60 Mg/m3)
a361 ± 124
a97 ± 8
b241 ± 20
41 ± 10
168 ± 99
Urban
(oxidant dominant,
~40 ug/m3)
b533 ± 157
165 ± 142
b223 ± 43
43 ± 10
b!98 ± 61

a Significantly different from control
b

at an

p < 0.05.
n s n AI

Source: Adapted from Yocom and Upham (1977).

Note: 1 ppm S02 = 2620 |jg/m3.
10-31
-------
10.2.3 Fabrics
Fibers that suffer destructive action upon exposure to acids derived from SO- include (1)
cellulosic fibers such as cotton and its close relative, viscose rayon, a regenerated cellu-
lose, and cellulose acetate; and (2) polyamide fibers such as nylon 6 and 66. Polyester,
acrylic, and polypropylene fibers are not damaged directly by S0?; however, S0« concentra-
tions can be a source of absorbed acid that can accelerate the fading of dyes and result in
fabric deterioration through chemical reactions. The possibility of higher acid content due
to oxidation of SO, to SO-, must be considered (Salvin, 1963).
Brysson et al. (1967) exposed cotton fabrics at six different environmental sites in St.
Louis, Missouri for 1 year (1963-64). The St. Louis metropolitan area sites represented
industrial, urban, and suburban environments. Suspended PM in the St, Louis area was measured
using periodic 24-hour hi-vol air samplers and monthly dustfall measurements. Sulfation
values were determined by the lead peroxide candle method. Two fabric types were exposed in
this study, a desized and scoured cotton print cloth and scoured cotton army duck. Study
results indicate that there is a significant relationship between air pollution and both
strength degradation and degree of fabric soiling. As shown in Figure 10-7, high pollutant
2 3
levels (mean sulfation 5 mg S03/100 cm /day and/or S0» concentrations of 0.2 ppm or 520 yg/m )
reduced effective fabric strength by one-sixth when compared with low pollution sites (0.5 mg
5 o
SQg/lQQ cm /day and/or 0.02 ppm or 60 (jg/m S02 concentrations). The relationship between
suspended PM and fabric strength degradation was not as good as that for S0». No correlation
between dustfall and strength degradation/effective life was demonstrated. Biological
deterioration did not appear to be a major factor in this study, since a plot of fluidity
values against breaking strength loss for exposed samples gave the essentially straight line
function indicative of nonbiological degradation (Brysson et al. 1967).
In a review of the Brysson et al. (1967) study, Upham and Salvin (1975) report a correla-
tion coefficient of 0.95 for breaking strength versus sulfation for cotton duck cloth. The
correlation coefficient for the thinner cotton print was 0.96. Of the pollutants measured,
S0? was most responsible for causing fabric damage (Upham and Salvin, 1975).
Zeronian (1970) carried out laboratory exposures in which cotton and rayon fabrics were
o
exposed for 7 days to clean air with and without 250 (jg/m (0.1 ppm) SO,, Both controlled
environments included continuous exposure to artificial light (xenon arc) and a water spray
turned on for 18 minutes every 2 hours. Loss in strength for all fabrics exposed to clean air
averaged 13 percent, while the fabrics exposed to S0? averaged 21 percent. Zeronian et al.
(1971) also exposed fabrics made from manmade fibers—nylon, polyester, and modacrylic—-to
3
controlled conditions similar to the cotton exposures, except that the S02 level was 486 yg/m
(0.2 ppm). They found that only the nylon fabrics were affected, losing 80 percent of their
strength when exposed to S02 and only 40 percent when exposed in clean air.
10-32
-------
•o
CM

O
5.0
4.0
E 3.0

O
oo
2.0
1.0
DUCK CLOTH: 12 months EXPOSURE
CORRELATION COEFFICIENT 0.95
SIGNIFICANT AT 0.5% LEVEL
O-DUCK CLOTH

A-PR1NT CLOTH
PRINT CLOTH: 5 months EXPOSURE
CORRELATION COEFFICIENT 0.96
SIGNIFICANT AT 0.5% LEVEL A>
I
50
0 10 20 30 40

BREAKING STRENGTH RETAINED, percent

Figure 10-7. Relationship between retained breaking strength of
cotton fabrics and corresponding mean sulfation rate measured at
selected sites in St. Louis area.

Source: Brysson et al. (1967), Upham and Salvin (1975).
10-33
-------
In a study designed to determine effects of air pollution on dye fading on fabrics, 67
dye-fabric combinations were tested. The test samples were exposed in the dark at 11 sites
representing climatic regimes and urban/rural conditions. High temperatures and humidities
appeared to increase fading in the presence of air pollutants. Urban sites produced sig-
nificantly higher fading than rural sites. Fading was highest at the site with the highest
SO.J and TSP concentrations (Beloin, 1972); however, a more rigorous test of the significance
of S0? to dye fading suggests that the only correlation between S09 and dye fading found in
the Beloin study may not have represented a cause-and-effect situation. Results of a two-
level factorial chamber experiment exposing various dyed fabrics to SO, at .03 ppm or .50 ppm
3
(79 and 1310 |jg7m , respectively) and two levels of N02 and 03 showed no significant fading
attributable to SO- (Upham et a!., 1975).
10.2.4 Building Materials
The deterioration of inorganic building materials occurs initially through surface weather-
ing. Moisture and salts are considered the most important factors in building material
damage. Many researchers believe that the mechanism of damage from air pollution involves the
formation of salts from reactions in the material; subsequently, these surface salts dissolve
in moist air and are washed away by rainfall. The components of inorganic building materials
can react with S02 and sulfates (Luckat, 1972; Winkler, 1975; Arnold et al. 1976). Luckat
(1977) reported good correlation of damage to stone with rate of SO,, uptake. Other research-
ers believe that the role of air pollutants in stone and concrete damage has frequently been
overestimated (Riederer, 1974; Niesel, 1979). These authors report that inorganic building
material damage is dominantly associated with RH >65 percent and freeze/thaw weathering. Some
researchers indicate that microorganisms must also be considered in order to quantify damage
to building materials due to ambient pollutant concentrations (Winkler, 1966; Riederer, 1974;
Krumbein and Lange, 1978; Eckhardt, 1978; Hansen, 1980). Sulfur chemoautotrophs are well
known for the damage they can cause to inorganic materials. These microorganisms (e.g.,
Thiobacillus) convert reduced forms of sulfur to H^SO* (Anderson, 1978). Presence of sulfur-
oxidizing bacteria on exposed monuments has been confirmed (Vero and Sila, 1976). The rela-
tive importance of biological, chemical, and physical mechanisms, however, have not been
systematically investigated. Thus, damage functions definitely quantifying the relationship
of pollutant concentrations to stone and concrete deterioration are not available in the
literature. Air pollution damage to glass is also not quantifiable at present (Newton, 1974).
10.2.4.1 Stone—Niesel (1979) completed a literature review on the weathering of building
stone in atmospheres containing SO , which includes references from 1700 to 1979. In brief,
he reports that weathering of porous building stone containing lime is generally characterized
by accumulation of calcium sulfate dihydrate in the near-surface region. The effect of at-
mospheric pollutants on the rate of weathering is believed to be dominantly controlled by the
10-34
-------
stone's permeability and moisture content. Migrating moisture serves primarily as a transport
medium. Sulfur dioxide is sorbed and thus can be translocated internally while being oxidized
to sulfates. Reacting components of the building stone are thus leached, the more soluble
compounds inward and the less soluble toward the surface, often forming a surface crust.
Sengupta and de Gast (1972) also report that SCL sorption causes physical changes in
stone involving changes in porosity and water retention. Removal of calcium carbonate changes
the physical nature of the stone surface. The hard, nonporous layer that forms as a result of
alternate freezing and thawing may blister, exfoliate, and separate from the surface. If the
stone contains some substances that are unaffected by SCL, the surface can deteriorate un-
evenly. The conversion of calcium carbonate to calcium sulfate results in a type of efflores-
cence called "crystallization spalling."
According to Gauri (1979) acidic precipitation also contributes to the weathering process
of building stone. Gauri reported that marble that is directly exposed to rainfall undergoes
nearly continuous erosion as the acid dissolves the calcium carbonate, allowing calcite
granules to break away and wash off.
At a recent MAS conference on conservation of historic stone monuments and buildings,
Tombach (1981) summarized mechanisms contributing to stone decay (see Table 10-6). While more
information on these mechanisms is becoming available, the relative contribution of ambient
S0? to deterioration of the various types of stone is not yet clear.
10.2.4.2 Cement and Concrete—Portland cement, the major active constituent of most concrete,
is manufactured by the high-temperature reaction of a mixture of limestone, alumina, sili-
cates, and iron salts found in clay. Cement, the binding agent in concrete, is an alkaline
material that reacts with SO^ and thus also suffers erosion and spalling effects.
The chemical action of SCL or sulfates on cement or concrete can be described as a dual
mechanism. Calcium hydroxide in cement and concrete can be converted to calcium sulfate,
which reacts to form calcium sulfate aluminum hydrate (ettringite), with a substantial in-
2-
crease in volume. Cement for dams and culverts requires special formulation for SO. re-
2-
sistance when exposed to SO. concentrations > 200 ppm in water (Nriagu, 1978).
Litvin (1968) examined concrete samples containing Portland cement and marble aggregate
with sand at an industrial site in Buffington, Indiana. Some changes were noted in the marble
aggregate, but a more observable change was found in the cement portion. Sealants were evalu-
ated as protective coatings; their use was accompanied in some cases by surface efflorescence.
Recently, specialized concretes in which sulfur replaces cement as the binding agent have
been successfully tested for resistance to acid and salt corrosive environments. These sulfur
10-35
-------
TABLE 10-6. MECHANISMS CONTRIBUTING TO STONE DECAY. PRINCIPAL ATMOSPHERIC FACTORS PARTICIPATING IN
THESE MECHANISMS ARE DENOTED BY SOLID CIRCLES; SECONDARY FACTORS ARE INDICATED BY SOLID TRIANGLES
co
en

A.
B.
C.
D.
t
Mechanism Rainfall Fog Huoldlty Temperature
External Abrasion
1. Erosion by windborne particles
2. Erosion by rainfall *
3. Erosion by surface ice • • •
Volume Change of Stone
1. Differential expansion of mineral grains
2. Differential bulk expansion due to uneven heating
3. Differential expansion of differing materials at joints
Volume Change of Material in Capillaries and Interstices
1. Freezing of water • •
2. Expansion of water when heated by sun • •
3. Trapping of water under pressure when • •
surface freezes
4. Swelling of water-imbibing Minerals by • • •
osmotic pressure
5. Hydration of efflorescences, internal Impurities, • 9
and stone constituents
6. Crystallization of salts 9 0
7. Oxidation of materials into more-voluminous forms • •
Dissolution of Stone or Change of Chemical Font
1. Dissolution in rain water • •
2. Dissolution by acids formed on stone by • • • •
atmospheric gases or particles and water
3. Reaction of stone with SO. to fora water- • • •
soluble material
4. Reaction of stone with acidic clay aerosol • • •
particles
Biological Activity
1. Chemical attack by chelating, nitrifying, sulfur- • •
reducing, or sulfur-oxidizing bacteria
2. Erosion by symbiotic assemblages and higher A A • •
plants which penetrate stone or produce
damaging excretions
Gaseous
Insolation Wind Pollutants Aerosol
• •
A
A A
A A
0 d A A
A
i i
•
-------
concretes, developed by the Federal Bureau of Mines, displayed good resistance to both acid
and salt attack and to damage by freeze-thaw cycling, (Sulphur Institute, 1979; McBee and
Sullivan, 1979; and Sullivan and McBee, 1976).
10.2.5 Electrical Equipment and Components
Robbins (1970) and ITT Electro-'Physics Labs (1971') studied the damaging effects of SCk
and particles on electronic components and estimated the cost of this damage. The ITT
Electro-Physics Labs report considered damage to 11 categories of electronic components for.
which a literature survey indicated that SO,, pollution would be mainly responsible. Infor-
mation gained directly from manufacturers, however, indicated that particles were the major
factor in degradation and failure of electronic components and equipment (see Section 10.3.1).
Reduction of S02 and particulate concentrations would have little effect on costs for the
prevention of corrosion; the same protective measure would still be required for low concen-
trations of pollutants. Since the cost of corrosion-resistant metals is far outweighed by the
expense of equipment failure, they are used even in environments where air pollution is mini-
mal .
10.2.6 Paper
Modern papers are manufactured from cellulose. On exposure to acids, paper is hydrolyzed
and loses strength. Analyses of paper from old books have shown H?SO- content on the edges up
to nearly 1.5 percent, with differences between the amounts in the edges and the center of a
page sometimes approaching 1 percent (Parker, 1955). Spedding et al. (1971), studying the
mechanism in work with radioactive labeling techniques, determined that S02 is readily
absorbed by paper and oxidized to H?SO. by the metallic impurities in the paper. The reaction
may also involve the lignins in the paper, resulting in the formation of lignosulfonic acids.
Walsh et al. (1977) showed that S0? is rapidly absorbed by uncoated wallpaper and less rapidly
absorbed by vinyl-coated paper. Although most paper is used in objects with a short service-
life, and steps have been taken to improve the acid resistance of books, the preservation of
documents is of concern in museums and archives. Coating paper with polymers impervious to
gases is an established process. It is estimated that 50 percent of the books printed between
1900 and 1940 are in need of conservation. The New York Public Library conserves old books by
microfilming, lamination, and electrostatic reproduction (Kingery, 1960). The library spent
$900,000 between 1952 and 1967 to microfilm books that had deteriorated (Waddell, 1974).
10.2.7 Leather
Leather has a high capacity for absorbing SOg. Spedding et al. (1971) reported that the
controlling factor in S0? uptake is the rate of S0» diffusion to the leather surface. The
formation of HpSCL in the presence of water is followed by hydrolysis of the protein (collagen)
of which leather is principally composed. This weakening of leather causes cracking and ulti-
mately results in reduction of the leather to a red-brown powder (Spedding et al., 1971; Yocom
and Grappone, 1976).
10-37
-------
The destruction of leather by absorption of SOp has long been known and was described in
detail by Prenderleith (1957). The buildup of H»SO. in aged leathers correlates with deteri-
f- 7j—
oration, which can be reduced by inactivating the SCL ion and by pH buffering. Deterioration
of leather is important in bookbinding and in leather upholstery. The use of artificial
leathers has reduced damage costs.
10.2.8 Elastomers and Plastics
The deterioration of natural rubber and synthetic elastomers under weathering conditions
has been studied extensively. Heat, light, oxygen, certain metallic ions, and particularly 0,
cause deterioration; but the literature does not mention SQp damage to rubber. Rubber, in
fact, is used as an acid-resistant coating. The problem of determining ambient air pollution
effects on rubber is complicated by the presence of 03, which attacks the double bonds in both
natural rubber and the butadiene-styrene and butadiene-acrylonitrile synthetics.
Haynie et al. (1976) conducted a chamber study on rubber to determine the effects of Q~,
S09, and N09 under controlled conditions of temperature, humidity, and light. Exposures were
3
made at concentrations of 0.1 and 1.0 ppm for each pollutant (in ug/m ; 262 and 2620 for SO,,
196 and 1960 for Q, and 188 and 1880 for NQp). As expected, 03 was responsible for acceler-
ated cracking of the rubber. Sulfur dioxide did not have any effect.
Verdu (1974) presented a theoretical study of the effect of air pollutants on the weath-
ering of plastics. He attributed a direct deteriorating effect on plastics to 0,, and sug-
gested that air pollutants such as S0? may form active compounds through photochemical re-
actions leading to oxidation chain reactions. In light-exposure trials, SO, increased the
rate of degradation of polystyrene.
10.2.9 Works of Art
Although works of art are composed of materials already discussed in earlier sections of
this chapter, they are briefly treated here as a separate category because the cost of the
materials involved does not represent the cost of the item.
The deteriorating effects of SOp and particles are well known to museum conservators
whose function is to preserve and restore works of art. The rate of pollutant-related deteri-
oration has increased markedly in the last 50 years. The damage is striking in Europe, where
ancient buildings, paintings, frescoes, stained glass windows, bronze sculptures, and marble
statuary have suffered deterioration.
Newton (1974) has investigated the cause of deterioration of medieval stained glass
windows. He found that the main cause of decay is the leaching of potassium ions from the
silicate glass by condensed water. Another cause is S0», which produces opaque white crusts
containing CaSO.-2H2Q and syngenite (K2Ca(SO.)2.H2Q). The poor durability of medieval glass is
due to its high content of alkaline earths such as lime and magnesia.
10-38
-------
Riederer (1974) conducted a study of the corrosion of bronze sculpture by air pollutants.
Sulfates were found in the corroded surfaces.
The sandstone of the cathedral in Cologne, located in a highly polluted urban area, has
suffered serious erosion due to the reaction of sulfur acids with calcium carbonate that form
calcium sulfate, which is leached out by rain (Luckat, 1972). Decay and deterioration of the
Taj Mahal in India has also been attributed to SCL and other pollutants emitted from a nearby
petrochemical complex (Gajendragakar, 1977).
Damage occurring in Venice, Florence, Rome, Athens, London, and Cologne has been attri-
buted to the effect of SCL from industrial areas in these cities (Yocum and Upham, 1977).- The
United States is also concerned about the deterioration of public buildings and monuments.
The National Bureau of Standards (NBS) was asked by the National Park Service to investigate
methods for preservation of stone after erosion was noted in the facade of the Lincoln
Memorial in Washington, D.C. (Sleater, 1977).
Sleater (1977) investigated damage to stone from the action of SCL, salt, sodium sulfate,
and light. Conservation materials including epoxy resins, fluorosilicates, and silicone
resins were evaluated. The conservation methods recommended to the National Park Service
varied with the exposure conditions.
The damage to the Acropolis caused by SCL and SCL has resulted in a massive interdisci-
plinary effort by the Greek government to protect the ancient buildings from further deterio-
ration (Yocom, 1979).
10.2.10 Review of Damage FunctionsRelating Sulfur Dioxide to MaterialDamage
Even the most reliable damage functions must be used with caution. As noted by Sereda
(1974), more data are required to take account of orientation, location, and design of mate-
rials in use. Those listed in Table 10-7 were selected on the basis of their treatment of
independent variables and their inclusion in major literature reviews. As discussed previous-
ly (see Section 10.2.1.2.2.2) damage functions vary in form, reflecting different parameters
measured and methods of measurement. Time-of-wetness (often expressed as RH above a critical
value) is the most important variable in these damage functions.
Functions for zinc or galvanized steel appear to show the best fit, followed by the func-
tions for oil-based house paint. The field studies by Haynie and Upham (1970) and Haynie
(1980) and chamber study by Haynie et al. (1976) incorporated critical variables and provided
relatively reliable damage functions for galvanized steel. The functions selected for weath-
ering and enameling steel and for oil-based paint also utilized these critical environmental
variables.
When these functions are used to estimate damage, other factors must be considered, such
as the amount of exposed (uncoated) metal, the percentage of buildings with oil-based (not
Latex) paint, and temperature and humidity variables (sites in the arid southwest compared
with sites in the relatively humid northeast).
10-39
-------
TABLE 10-7. SELECTED PHYSICAL DAMAGE FUNCTIONS RELATED TO SO EXPOSURE
Material
Reference
Exposure-Response relationships
Zinc
Haynie and Upham, 1970 Y = 0.001028 (RH - 48.8) S02
0.92
Galvanized steel Haynie et aL, 1976 corr = (0.0187 SO, + e 4L85 " 23«240/RT)t 0.91
c, W
Galvanized steel
Haynie, 1980
corr = 2.32 t + 0.0134v°'781S09t
W £, W
Not provided
by author.
o
1
-p.
o
Oil-base house paint Spence et al., 1975 Y = 14.3 + 0.0151 S02 + 0.388 RH

Enameling steel Haynie and Upham, 1974 corr = 325 Vt e^0'00275 S02 " <163-2/RH)J

Weathering steel Haynie et al.» 1976 corr = [5.64 ,/SC + e^55*44 " 31>150/RT)3
0.61
Not provided
by authors.
o.91
corr = depth of corrosion or erosion, urn
Y = corrosion/erosion rate, pm/yr
S02 = ug/m3 S02
R = gas constant (1.9872 cal/gm mol K)
RH = percent annual average relative humidity
w
fcw
V
T
t
= fractional time of panel wetness
= time of wetness in years
= wind velocity in m/s
= geometric mean temperature of panels when wet, K
= time exposure, years
Note: 1 ppm S02 = 2620 ug/m
-------
10.3 PARTICIPATE MATTER
A report by the National Research Council (1979) on airborne particles noted that
deposition of dust and soot on building materials not only significantly reduces their
aesthetic appeal, but also, either alone or in concert with other environmental factors,
results in direct chemical attack. Because of the paucity of data (see Chapter 5) regarding
TSP size distribution and composition, it is difficult to determine the specific types of
particles and chemical constituents that have damaged or soiled a particular structure. As
discussed below, chemical composition of PM is highly important to its corrosiveness.
Particle size may be important; research is under way on the role of particle size in soiling
of paint.
10.3.1 Corrosion and Erosion
Early studies indicated that suspended PM plays a significant role in metal corrosion.
Sanyal and Singhania (1956) wrote that the influence of suspended PM was "profound." They
ascribed the corrosive effects of particles to (1) electrolytic, hygroscopic, and/or acidic
properties and (2) their ability to sorb corrosive gases (e.g., S0?). Chandler and Kilcullen
(1968) pointed out that it is difficult to predict corrosion rates separately for S0? and PM
since they frequently coexist at high levels. Other field studies have not established a
conclusive correlation between total suspended PM and corrosion though analysis of data
continues (Mansfeld, 1980; Haynie and Upham, 1974; and Upham, 1967).
Moist air polluted with SO, and PM results in a more rapid corrosion rate than air
polluted with S09 alone (Yocom and Grappone, 1976; Johnson et al., 1977). Kottori (1980)
?~
observed that zinc and galvanized steel corrosion rates appeared to be related to the SO.
content of TSP. Chloride content of dust also may contribute to accelerated corrosion of
steels (Gibbons, 1970; Bresle, 1976).
Barton (1958) reported that dustfalls contribute to the initial stages of metal corrosion
but that their influence becomes less important as a layer of rust forms. Two classes of PM:
hygroscopic salts (including those of natural origin such as sodium chloride) and acid smut
appear to be definitely associated with corrosion.
A review of atmospheric factors affecting metal corrosion provides evidence of a relation-
ship between salinity and corrosion (Guttman and Sereda, 1968). Corrosion of metals can be
accelerated by deposition of particles that are hygroscopic and therefore increase surface wet-
ness time. The influence of hygroscopic substances on metal corrosion rates has been pre-
viously discussed in Section 10.2.1.1.4. As also noted in Section 10.2, particles can disrupt
the protective oxide films formed on metal surfaces such as nickel, copper, aluminum and
10-41
-------
stainless steel, resulting in pitting (Russell, 1976; NRC, 1979). Russell (1976) noted that
airborne particles often play an important role in the attack by SO on electrical contact
A
surfaces by acting as points for the concentration of active ionic species. Smoothly polished
or electroplated surfaces are less likely to retain solid deposits originating from airborne
particles (Larrabee, 1959).
Acid smut is a highly corrosive, sticky material formed in and emitted mainly from
furnaces, notably in power plants that burn liquid fuels containing sulfur (Ireland, 1968).
This material would not usually be considered suspended PM, as it occurs as agglomerates of
carbon, ash, and HpSO. up to 0.5 cm or more in diameter that fall close to the source (Rowden,
1968; Potter, 1971). According to Hoshizawa and Koyata (1970), acid smut mainly settles out
within 400 meters of the source under conditions of light wind.
Japanese investigators analyzed a large (>10 drums [sic]) sample of acid smut and found
the H«SO. content to be 30 percent (Oyama et al. , 1974). Damage to painted surfaces, auto-
motive finishes, and even agricultural crops can be substantial. As noted in a review of
residual oil firing problems, "public reaction can be quite severe" (Exley, 1970). A report
on the status of "public nuisances" in the electric power industry of Japan reported progress
in determining the cause of acid smut and in developing preventive techniques (Overseas Public
Nuisance Study Mission, 1965).
Finishes on automobiles parked near industrial sites have often been severely damaged.
Staining and even pitting of auto finishes have been traced to iron particles from nearby
plants. Cars parked near brick buildings being demolished have been damaged by alkali mortar
dust during humid weather. Damaged auto finishes had to be repainted because color changes
were not reversible by washing or polishing (Fochtman and Langer, 1957).
Parker (1955) reported that numerous black specks accumulated on freshly painted buildings
in industrial areas. The building exteriors became distinctly soiled and required cleaning or
repainting in 2 or 3 years, depending on PM concentrations in the air. When PM became
embedded in the paint film, the coating was both esthetically and physically damaged. Em-
bedding of particles provides nucleation sites at which other pollutants can concentrate.
Cowling and Roberts (1954) suggest that particles promote the chemical deterioration of paint
by acting as wicks to transfer the corrosive SO- solution to the underlying surface. Luckat
(1972) suggested that dusts containing heavy metals may accelerate stone erosion by catalytic
effects on the oxidation of ambient S0? to HpSO..
10.3.2 Soiling and Discoloration
Soiling is the accumulation of PM on the surface of a material. As defined by Faith
(1976), soiling arises from the deposition of particles less than 10 (jm by impingement on
surfaces which then mingle with settled dust. Soiling produces a change in reflectance from
opaque materials and reduces light transmission through transparent materials (Beloin and
Haynie, 1975; NRC, 1979). Soiling due to airborne particles from manmade sources results in
increased cleaning costs for building and other materials and in reduction in the useful life
of fabrics.

10-42
-------
Considerable uncertainty exists about the level of accumulated PM that leads to increased
cleaning, since not only are socioeconomic factors involved in a decision to clean, but the
accumulation of particles must first be perceived as soiling. Carey (1959), discussing house-
hold cleaning, observed that, although in a room with dusty air, particles descended continu-
ously onto paper, the paper remained apparently clean for a period of time and then, almost
suddenly, appeared dirty. The dingy appearance occurred when the paper surface was covered
with dust specks spaced about 10 to 20 diameters apart. The question of least perceptible
cover concerns the sensitivity of the eye and can be evaluated by use of patterns as shown in
Figure 10-8. If the patterns in this figure are viewed at a distance too great to perceive
the individual dots it will be found that when the contrast is strong, e.g., black on white,
it is possible to distinguish a clean circle from its surroundings when 0.2 percent of the
area is covered with dots; with a weaker color contrast 0.4 percent is the limiting coverage.
The effect is subjective, and it is not easy to judge between coverages that differ by a
factor of V2. The deposition of relatively large particles on flat, horizontal surfaces noted
in Carey's report applies more to household soiling than to the soiling of outside, vertical
structures.
Following the concept that dinginess or dustiness is a visually perceived phenomenon,
Hancock et al. (1976) conducted a quantitative experiment on dustiness. In support of Carey's
observations, these authors found that, with maximum contrast, a 0.2 percent effective area
coverage (EAC) by dust can be perceived against a clean background. The minimum perceivable
difference between varying gradations of shading was a change of about 0.45 percent EAC. The
results also revealed that a dust deposition level of 0.7 percent EAC was required before the
object covered in dust was deemed unfit to use. A telephone survey indicated that the minimum
tolerable interval between household dustings was every 4 days. Combination of the telephone
survey information with the level of objectionable dust coverage implied that a dustfall rate
of less than 0.17 percent EAC/day would be tolerable for the population at large. Based on
these evaluations and assuming an atmospheric particle density of 2 g/cm , an average pro-
jected area radius of 5 Mm f°r the settled particles, the total mass flux which corresponds to
2 2
0.17 percent EAC per day was estimated to be 1.36 mg/cm mo. (roughly 40 tons/mi mo).
Esmen (1973), found that exposed surfaces reached an EAC of 0.4 percent seven times
faster in an urban location than in a rural location. Ambient particle concentrations, how-
ever, were not reported.
10.3.2.1 Bui 1ding Materials—Under high wind conditions, large particles entrained in the
windstream tends to cause a slow erosion of surfaces similiar to sandblasting. Particles
also fill surface pores of many sandstones causing them to become uniformly darkened.
10-43
-------
01%
•i £'':- c> "^v"!^^iS^'^%
^'i ;':.>>:; ic^S-
W-r'-;''^; ^:?/-W
•/^v;?r:v;;....;, ..:• v-.^v^-v.i.

Vjr-r^:v:U.:. S'.^.:ff:0'
.. .-|.. .

•>SVXV':"'S'';*'«
0.2%
°-4%
08%
Figure 10-8. Dust deposit patterns with corresponding coverage {% sur-
face covered) are shown.
Source: Carey (1959).
10-44
-------
Particles can contribute to chemical decay of marble, limestone, and dolomite stone work
(see Table 10-6, Section 10.2.4), and concrete structures if they carry acids and soluble
salts (NRC, 1979). Scanning electron microscopy (SEM) examination of exposed marble from
several locations in northern Italy provides evidence of a major role for carbonaceous par-
ticles in marble deterioration. Del Monte et al. (1981) report that SEM morphological charac-
terization, together with particle analysis by X-ray'diffraction and X-ray energy dispersive
analyzer data, identifies the majority of the carbonaceous particles as products of oil fired
boiler/combustion. Particle median diameter was ~ 10 ^im. Del Monte et al. also presented
evidence strongly suggesting that the carbonaceous particles are very important in oxidizing
2-
SO, to SO* . If further research substantiates this finding, current understanding of the
relative roles of S02 and PM in marble deterioration might well be substantially modified.
Beloin and Haynie (1975) developed dose-response relationships for suspended PM and
soiling of various building materials. They compared the rates of soiling by different levels
of TSP on six different building materials over 2 years. The mean annual PM concentrations at
3 3
the five study sites ranged from 60 ug/m for a rural residential location to 250 ug/m for an
industrial residential environment. The exposed materials included painted cedar siding,
concrete block, brick, limestone, asphalt shingles, and window glass. Table 10-8 shows the
results of regression for soiling of building materials as a function of TSP exposure.
10.3.2.2 Fabrics—Although PM obviously soils fabrics, researchers have noted that it is only
damaging when the particles are highly abrasive and the fabrics are frequently flexed. Cur-
tains hanging in open windows, serving as filters in polluted areas, provide a good example.
Weakened as a result of such exposure, curtains often split in parallel lines along the folds
(Yocom and McCaldin, 1968.) The more tightly woven the cloth, the more resistant it is to
soiling (NRC, 1979).
•Because of soiling, fabrics may be washed more often. Excessive washing may reduce
fabric strength, leading to a poorer appearance and to shortened life. Sunlight, water vapor,
SO , NO and 0, concentrations, however, are believed to have a more significant effect on
A. X »3
the service life of fabrics. Insolation decoloration is considered to be the most important
service life reduction factor (NRC, 1979).
10.3.2.3 Household and Industrial Paints—Exterior paints can be soiled by liquids and solid
particles composed of soot, tarry acids, and various other constituents. Beloin and Haynie
(1975) determined by reflectance measurements that the degree of soiling of painted surfaces
was directly proportional to the square root of the PM dose, accounting for 74 to 90 percent
of the measured variability. As one example, the linear regression results of soiling of
acrylic emulsion house paint by exposure to suspended PM is shown below:

(720 samples, R2 = 0.902)
Y = B V(TSP x months of exposure) + A (10-13)
10-45
-------
TABLE 10-8. RESULTS OF REGRESSION FOR SOILING OF BUILDING MATERIALS AS A FUNCTION OF TSP EXPOSURE
Material
o
en
Oil base paint 400
Tint base paint 400
Sheltered acrylic 400
emulsion paint
Acrylic emulsion 720
paint
Shingles 48
Concrete 160
Coated limestone 80
Uncoated limestone 80
Coated red brick 80
Uncoated red brick 80
Coated yellow brick 80
Uncoated yellow brick 80
Glass 45
89.43
86.13
91.54

90.79

43.50
41.75
44.57
46.99
12.95
14.88
45.05
43.21
0.2806
-0.2768
-0.2618
-0.593

-0.4131

-0.199
-0.0458
+0.0779
-0.0503
-0.029i
-0.0374
-0.1133
-0.1133
+0. 0314
0.0641
0.0571
0.1156

0.0497

0.5771
0.1338
0.2464
0.1500
0.0223
0.0331
0.5337
0.2740
0.008077
0.000069
0.000061
0.000123

0.000026

0.000258
0. 000080
0.000164
0.000089
0.000013
0.000020
0.000317
0.000168
0.000007
7. 6510
6.8265
13.8143

8.3791

7.6992
7.5011
6.9046
4.2035
0.6255
0.9274
14.9533
7.6773
0.6851
0.745
0.738
0.880

0.902

0.769
0.143
0.347
0.266
0.459
0.477
0.342
0,503
0.340

Note; Equation used in this
except in the case of
regression analysis was
shingles, whei
N, = Number of data sets (dependent upon
re reflecta
the number
reflectance
nee = B(TSP

of control

= a
-------
Where: Y = Measured percent reflectance (Photvolt Model 625)
B = Slope of linear reflectance
3
TSP = Average TSP concentration (ug/m )
A = Intercept of linear regression
t = Exposure time (months)
Based on this equation, Figure 10-9 summarizes the soiling of acrylic emulsion house paint as
a function of exposure time and particle concentrations. Although it is recognized that
socioeconomic factors control ability and motivation to maintain clean surfaces (Beloin and
Haynie 1975), repainting frequencies can be estimated. Assuming an individual responds to a
defined change in the reflectance of house paint by repainting, that person will repaint a
3
house twice as often in an environment with a TSP concentration of 260 ng/m in comparison to
3
75 ug/m . Specifically, at a 35-percent change, this houseowner would repaint every 4 years
3 3
when TSP is 75 jjg/m and every 2 years when TSP is 260 ug/m .
10.4 SUMMARY, PHYSICAL EFFECTS OF SULFUR OXIDES AND PARTICULATE MATTER ON MATERIALS
Sections 10.2 and 10.3, indicate that the best data, base for evaluating the association
between SO and materials damage is that which concerns corrosion of metal. The parallel case
X
for PM is soiling. Of the damage functions developed for corrosion of metals by SO , the one
for zinc appears to show the best fit. Relationships between SO exposure and corrosion of
other metals are slightly less well established. There is eviuence that PM aggravates corro-
sion, especially when coexisting with SO . This is most likely due to the hygroscopic salt
content of the PM. No general mathematical expression of the relative contribution of PM to
corrosion of metal, however, has been established.
Both PM and SO damage paint. Damage functions for erosion of various paints by SO and
X /x
soiling of paint by PM have been developed. The varying properties of the several types of
paint used on exposed materials make it difficult to construct mathematical expressions for
effect of PM, S0?, or both on all paints.
Building materials also are eroded by SO and soiled by PM. That damage is attributable
X
to both pollutants is well established; quantitative general relationships between ambient
concentration and effect are lacking, however. For erosion of building materials, partic-
ularly stone and concrete, the contribution of sulfur oxides relative to other agents is not
clear. As for soiling, although some damage functions have been developed, the lack of under-
standing of the role of particle size and composition makes it difficult to generalize about
soiling effects of all suspended PM on all building materials.
(
10-47
-------
Ul
o
45
40
u.
Ul
tt 35

i 30
ui
B.

u 25
U
Z

p 20

2 15

9 10
o
5
J3

o
03
6 12 18 24 30

EXPOSURE TIME, months
36
Figure 10-9. Representation of soiling of acrylic

emulsion house paint as a function of exposure

time and particle concentrations.

Source: Abstracted from Beloin and Haynie (1975).
10-48
-------
Data on the effects of PM, SO , or both on other surfaces are even less well established,
A,
Some evidence shows damage to fabrics, leather, paper, glass, plastic, and works of art com-
posed of one or many materials, but this evidence is largely qualitative and sketchy. It is
difficult to develop reliable estimates of effects attributable to specific ambient pollutant
concentrations for these materials.
10.5 ECONOMIC ESTIMATES
10.5.1 Introduction
Several types of financial losses result from damage and soiling:
1. Reduced service life of a material.
2. Decreased utility of a material,
3. Use of suitable substitute materials.
4, Losses due to an inferior substitute.
5. Protection of susceptible materials.
6. Additional required maintenance, including cleaning.
The major losses of amenity, as defined by Maler and Wyzga (1976), are associated with
enduring and suffering soiled, damaged, or inferior products and materials because of pollu-
tion, in this case, PM, SO , or both. In addition, amenity losses are suffered when pollution
damage repair or maintenance procedures result in inconvenience or other delays in normal
operations. Some of these losses, such as effects on monuments and works of art, are espe-
cially difficult to specify (Miler and Wyzga, 1976).
The reduced value and attractiveness of property and the extra costs of cleaning and
maintenance resulting from air pollution levels must obviously be considered when assessing
financial losses. In addition to the diseconomy of property value losses, the consumer is
also burdened with less directly measurable psychological distresses encountered when experi-
encing pollution nuisances. Less tangible externalities such as these have an impact on
consumers as real as any directly measurable financial losses.
In calculating monetary damage, the approach selected depends on whether financial losses
or losses of amenity are emphasized, the type of damage being considered, and the availability
of cost information. The literature on pollutant effects on materials has been dominated by
calculations of financial loss based on the physical damage function approach. This approach
leads to error in financial loss estimates for the reasons listed below:
1. Reliable information on physical damage is not available for all economically impor-
tant materials.
2. Information is lacking on the spatial and temporal distribution of the materials
being used.
3. Techniques do not reflect the use of resistant materials that last longer and re-
quire less maintenance.
10-49
-------
4. Estimates assume that galvanized steels are left uncoated.
5. Many materials may wear out before they are significantly damaged by pollutants.
6. Many loss estimates assume that substitute materials cost more than the original
material, and that the cost differential is attributable to pollution.
Due to increased costs of labor, many materials are being sought that reduce maintenance and
also resist pollutant attack. Savings realized primarily from lower maintenance should be
deducted from costs charged to pollution (Glass, 1978).
The estimation of monetary damage associated with soiling is not dominated by the physi-
cal damage approach. In these studies, loss of amenity has been considered as well as direct
financial loss, since socioeconomic variables are heavily involved. The approaches reflect
the shift in emphasis: Nonmarket and indirect market approaches have been used to relate
changes in air quality to changes in the amount of money people are willing to pay for
improved air quality (reduced air pollution). This relationship can be documented through a
survey of affected individuals or development of relationships between environmental quality
and available data on price differentials. A major source of error is that these approaches
demand that all factors that affect cost other than air quality must be accounted for. It is
also difficult to distinguish among the effects of many different air pollutants.
In general, all approaches to estimating costs of air pollution effects on materials
are limited by the difficulty in quantifying the human response to damage based upon the
ability and the incentive to pay additional costs (Yocom and Grappone, 1976). The physical
damage approach requires specification of all relevant substitution possibilities and reliable
exposure estimates. The welfare economic approaches, on the other hand, require control of
several elements of survey bias, including "willingness to pay," and economic data that affect
or clearly reflect choices.
10.5,2 EconomicLoss Associated with Materials Damage and Soiling
Direct damage to materials is usually attributed to the corrosive action of SO , even
/C
though the interaction of SO and PM is recognized as an important determinant of the amount
of actual damage (see Sections 10-2 and 10-3). Soiling of materials is attributed to PM.
10.5.2.1 Metal Corrosion and Other Damage to Materials Associated with SulfurOxides—Realis-
tic estimates of the economic damage to metals from SO and PM must consider several factors,
including avoidance costs, such as the costs of specific protective treatment. For metals,
these costs include the use of anticorrosive primers, the practice of sandblasting before
painting, and the use of acid-resistant paints.
Gillette (1975) reported significant reductions in economic damage to materials from SO
/\
due to improved air quality levels throughout the United States. Comparing annual S02 concen-
trations from more than 200 monitoring sites with the estimated inventory of materials exposed
10-50
-------
near these sites, he estimated that nationwide material damage decreased from more than $900
million in 1968 to less than $100 million in 1972, for a total of $1.4 billion in realized
benefits over 5 years. These estimates were derived by distinguishing between physical and
economic losses and by attributing current estimates of losses to current exposure levels.
The distinction between the physical and economic damage to materials is based upon the
concept of normal or economic useful life of materials. Whereas physical deterioration of
materials may occur at relatively low exposure levels, economic losses will occur only if the
material requires early replacement or increased maintenance before its normal or economically
useful life is spent. Gillette (1975) used Internal Revenue Service guidelines for service
life assumptions, which may have introduced a downward bias to his damage estimates. Given
the prevailing ambient concentrations observed, Gillette reported that most materials were not
adversely affected economically except for metallic products that were subjected to corrosion
or paint damage. While material losses were much greater during the early 1960's, the later
losses were substantially lower and reflected the*considerable improvement in air quality
(Gillette, 1975).
A U.S. Department of Commerce report (Bennett et al., 1978) examined the cost of corro-
sion in the United States in 1975. Unfortunately, the developed damage costs were not
pollutant-specific and were not associated with ambient concentrations of pollutants. The
report estimated the total annual metallic corrosion cost at $82 billion with a model that
incorporated a broad range of cost items (e.g., materials, labor, energy, and technical capa-
bilities). About 40 percent of this cost, or $33 billion, was considered avoidable. Within
this avoidable cost is, of course, the cost of air pollution, a portion of which is in turn
the cost of metallic corrosion resulting from PM, SO , or both. The figure of $33 billion is
X
thus only useful as an upper limit for present purposes since the estimate reflects economic
damages from all pollutants as well as other avoidable costs.
Fink et al. (1971) estimated that corrosion of external metal structures caused by air
pollution costs $1.45 billion annually in the United States, as shown in Table 10-9. As with
the extensive Bennett et al. (1978) Department of Commerce report, these studies were not
specific to single pollutants, nor were the damage costs associated directly with ambient
pollutant concentrations. Furthermore, in some cases the cost estimates included material
damage resulting from causes other than air pollution (e.g., the Fink et al. study included
corrosion inside pipes of industrial systems). On the other hand, Haynie, (1974) noted that
within the estimate by Fink et al. of metal corrosion costs, damage to structural systems pri-
marily constructed of galvanized steel accounted for more than 90 percent of the cost. No
significant damage was assigned to corrosion of copper, aluminum, stainless steel, or lead.
Haynie reasoned that, based on the data of Fink et al.(1971), the accelerated corrosion of
zinc by S0? accounts for more than 90 percent of corrosion caused by air pollutants.
10-51
-------
TABLE 10-9. SUMMATION OF ANNUAL EXTRA LOSSES DUE TO CORROSION DAMAGE BY AIR POLLUTION
TO EXTERNAL METAL STRUCTURES FOR 1970
Steel system or structure
Basis for calculation
Annual
loss in $1000
o
i
en
r\>
Steel storage tanks
Highway and rail bridges
Power transformers
Street lighting fixtures
Outdoor metal work
Pole-line hardware
Chain-link fencing
Galvanized wire and rope
Transmission towers
Maintenance
Maintenance
Maintenance
Maintenance
Maintenance
Replacement
Maintenance and replacement
Replacement
Maintenance
$ 46,310
30,400
7,450
11,910
914,015
161,000
165,800
111,800
1,480
$1,450,165
Source: Fink et al. (1971).
-------
The economic damage from corrosion can also be estimated by determining the annual cost
of industrial paints used for corrosion control. This cost is estimated by summing the costs
of surface preparation, paint, and labor. Coatings applied at this combined cost would pro-
tect metal structures against corrosion for about 10 years, whereas "ordinary" coatings of
paint would provide protection for only 2 years., Moore §nd O'Leary (1975) investigated the
painting of structural steel in bridges, which involves sandblasting the steel to produce a
rust-free surface and to remove mill scale. Without such surface preparation, water is imme-
diately absorbed and sets up a corrosion system, rusting occurs, and the paint surface deteri-
orates in 2 to 3 years. The metal surface is protected by a primer that inhibits rust forma-
tion, and the primer coat is covered with two coats of SQp-resistant paint, such as vinyl
resin, which is substantially more expensive than household paint. Banov (1973), Michelson
and Tourin (1967), and others developed estimates to reflect these preparation and painting
costs for protection of metal structures; however, sound fractional allocation of these costs
to SO material damage is not available in the literature.
" •""
Stankunas et/al. (1981) developed a geographically distributed inventory of exposed
materials in the Boston metropolitan area, Isopleths for S0? were laid over a map reflecting
exposed materials. Damage functions were applied to determine an estimate of S0? effects on
materials for the entire area. Cost factors for repair were applied; the resulting annual
cost estimate for S0? damage to paint in the Boston area was $31 million. For bare galvanized
metal, the annual cost due to S0? was reported to be only $335,000. The authors noted that
the amount of exposed bare galvanized metal and of structural concrete was low when compared
to exposed painted surfaces. They also pointed out that the paint damage costs were derived
using a damage function for oil-based house paint. However, its applicability is questionable
due to the widespread shift to SO,,-resistant latex paints (see Table 10-5).
While this study represents a step toward developing and applying of realistic exposure
data, a national estimate cannot be reliably extrapolated because of wide variations in envi-
ronmental conditions (maritime versus inland locations, humidity levels) and material use
(commercial versus industrial cities). Similar studies of cities in various .regions may
provide a basis for a more reasonable composite national estimate.
Damage from pollutants during the manufacture of electrical components must be con-
trolled even in clean environments. To prevent such problems, parts are fabricated in "clean
rooms" with filtered air. Equipment malfunctions in plant or service areas lead to additional
maintenance costs for cleaning, repairing, or replacing defective equipment. ITT (1971)
estimated $15.5 million per year in added costs for clean rooms and maintenance. Robbins
(1970) conducted a survey of the effects of S02 and particles on electrical contacts such as
those in switches, relays, connectors, and computers. To reduce corrosion, contacts are
electroplated with corrosion-resistant metals (e.g., gold, platinum, palladium, and silver).
Less expensive metals are susceptible to corrosion failure, mostly from the action of S0? and
HS, Robbins estimated that 15 percent of the gold and platinum used in the United States for
10-53
-------
electrical contacts in 1970 was used specifically to combat SCL corrosion, with the remainder
for protection against other environmental pollutants. Protection against all deleterious
environmental conditions, however, is routinely provided in clean environments due to high
replacement costs associated with electrical system downtime. Thus, costs for protection of
electrical hardware cannot be allocated directly to ambient SO , PM, or both.
/\
10.5.2.2 Soiling of Paint and Other Materials Associated with Particulate Matter—Studies of
soiling have historically focused on cleaning costs, with the assumption that as particulate
pollution increases, so does the frequency of cleaning and maintenance. This was the thesis of
one of the first national estimates of soiling costs associated with air pollution. The
Beaver Report (1954) suggested an annual total for damage by all forms of air pollution in
Great Britain of 152 million pounds sterling in direct costs, of which 25 million was for
laundry, 30 million for painting and decorating, and 20 million for cleaning and depreciation
of buildings other than houses; thus, about half the total imputed cost of pollution was
attributed to soiling associated with particulate pollution.
In the United States, the same hypotheses underlay approaches to quantify soiling costs
associated with particulate pollution, or, as measured in the United States, TSP. Almost
without exception, these studies have focused on household cleaning and maintenance.
Michelson and Tourin (1966) compared cleaning and maintenance costs in Steubenville, Ohio,
where the average TSP level was 383 ug/m with costs in Uniontown, Pennsylvania, where TSP
averaged 115 (jg/m . The study was based on a 30-percent response to a questionnaire mailed to
2 to 6 percent of the households in these communities. Michelson and Tourin reported that per
capita costs for cleaning and maintenance were $84 higher in Steubenville than in Uniontown.
However, the results can be questioned on the following grounds: failure to account for socio-
economic factors, inadequate tests of statistical significance, and consumer attitudes.
In a second study using a similar approach, Michelson and Tourin (1967) sampled com-
munities in the Washington, D.C. area (Suitland and Rockville, Maryland and Fairfax, Virginia).
The TSP average levels for the three communities were 85, 68, and 60 ug/m , respectively.
Though the study did show an increase in frequency of cleaning with increased TSP levels,
errors associated with measurement techniques, averaging over a community, and influence of
socioeconomic factors could be of the order of the small differences reported.
Ridker (1967) undertook two studies attempting to identify and quantify the soiling costs
of air pollution. In 1965 Ridker divided Philadelphia into high, medium, and low pollution
zones, and administered a series of questions to residents of individual housing units se-
lected at random in each zone. Ridker diverged from previous questionnaires administered in
other studies by correcting for attitudes toward cleanliness, and asking willingness-to-pay
questions on controlling or mitigating air pollution soiling. The results of the study, how-
ever, were largely inconclusive.
Ridker (1967) also conducted a study on an air pollution episode in Syracuse, New York.
A survey instrument and selection of a sample, designed to collect data within a very short
time after the pollution incident, were used. The results of the household survey were better

10-54
-------
than the one conducted in Philadelphia. However, Ridker felt that the difference in response
to questions concerning clean-up after the pollution episode could be explained by variables
other than pollution. Respecification of the estimating equation to obtain a cost curve in-
dependent of the effect of the intervening variables (by combining questionnaire responses to
intervening variables with a pollution index) gave an upper and lower bound to the estimated
damages. Nevertheless, the results of the Syracuse study are limited to episode-type sit-
uations and cannot be generalized.
In 1968, the National Air Pollution Control Administration (NAPCA) became concerned that
the previous studies, individually and as a group, lacked statistical validity to form a basis
for assessing the benefits of air pollution control. To overcome the shortcomings of the
previous studies, NAPCA commissioned the Booz, Allen and Hamilton (1970) (BAH) study. The
objective of the BAH study was to determine residential soiling costs of particulate air pollu-
tion for the 11-county Philadelphia area, stretching approximately from Wilmington, Delaware,
to Trenton, New Jersey. Although the primary purpose of the study was to determine the resi-
dential soiling costs in the 11 county area, it also had the alternative objectives of pro-
viding methods of estimating residential soiling costs under various abatement strategies,
comparing its findings with previous work (Michelson and Tourin), and developing a sampling
methodology which could be applied in other metropolitan areas.
Important to note is that the BAH study was the first of its kind to combine a well-
designed sampling strategy and questionnaire. Although limited to the Philadelphia region, it
continues to serve as the only empirical basis for estimating soiling costs not only for the
Philadelphia region, but also the entire nation.
The BAH questionnaire and interview procedure consisted of eliciting information on the
following:
(1) Attitudes toward household cleaning and maintenance (e.g., neat and tidy, lacka-
daisical , etc.).

(2) Demographic and economic information, including but not limited to educational level
and occupation of household, number of children living at home, marital status,
total family income, national origin, etc. (See Table 10-10.)

(3) Maintenance and cleaning performance information on 27 indoor and outdoor tasks.
(See Table 10-11.)

(4) Cost of maintenance and cleaning operation estimates, including whether the task was
performed by "do it yourself" method or a contractor, cost of hired labor, cost of
materials, etc.
10-55
-------
TABLE 10-10. SELECTED CHARACTERISTICS OF HOUSEHOLDS IN FOUR AIR POLLUTION ZONES
(Booz, Allen and Hamilton, 1970)

All Households

Number *
Povr-ont -
Weighted
Total
inn%

1
471
inn%

2
421
-mn<£
Zone
3
299
inn%

4
251
inn^
Race
White 79 93 78 57 70
Non-White 21 7 22 43 30

Family Income of Household
Under $6,000 annually32 21 32 45 49
$6,000 - $9,000 annually 29 27 25 36 32
$10,000 or more annually 31 40 34 16 13

Marital Status of HouseholdHead
Harried with minor children at home 65 69 61 61 65
Married, no minor children at home 30 27 33 34 28
Not married 54 657

Tenure of Household
Own home 74 79 74 67 64
Rent home 26 21 26 33 36

Dwelling Units in Structure
Single-family structure51 62 42 44 43
Multifamily structure 43 33 51 52 47

Outside Wall Surface Areas
More than 10% painted wood 12 16 15 6 3
More than 10% aluminum siding 8 10 843
More than 75%.brick or masonry 45 31 40 59 79

Number of Rooms in Household**
36 27 33 45 59
More than 10 64 73 67 55 41

Mean Number of People Per Household 3.54 3.63 3.47 3.40 3.54

*
Some categories may not total to 100% because of nonresponses and/or multiple responses.

Closets and bathrooms/lavatories included in count.
10-56
-------
TABLE 10-11. 27 CLEANING AND MAINTENANCE OPERATIONS SEPARATED BY SENSITIVITY TO AIR PARTICULATE LEVELS IN FOUR POLLUTION ZONES
(Bool, Allen, and Hamilton, 1970)

ZONE 1 ZONE 2 ZONE 3 ZONE 4
OPERATION/TASKS Mean Annual
Frequency

(Number of Households Sampled)
% Households Mean Annual
Perform! ng Frequency
(471)
% Households
Perform! ng
(421)
Mean Annual
Frequency
% Households
Performi nq
Mean Manual
Frequency
(299)
% Households
Perf oral ng
(251)
Sensitive to Air Particulate Level
Inside
1.
2.
3.
4.
5.
Replace Air-Conditioner Filter
Wash Floor Surfaces
Wax Floor Surfaces
Wash Windows (inside)
Clean Venetian Blinds/Shades
0.36
40.55
13.76
10.06
4.04
20.4
96.0
65.2
97.2
34.0
0.50
42.06
14.36
11.78
6.17
26.
96.
55.
97.
48.
1
9
6
9
2
0.30
42.74
16.80
12.74
9.13
17.1
94.3
56.5
97.0
62.5
0.
45.
13.
18.
9,
98
17
01
95
21
21.9
97.2
51.8
98.0
59.4
Outside
1.
2.
3.
4.
S5.
i,6.
Clean/Repair Screens
Wash Windows (outside)
Clean/Repair Storm Windows
Clean Outdoor Furniture
Maintain Driveways, Walks
Clean Gutters
0.80
4.25
2.07
2.50
3.98
1.12
56.5
89.5
55.6
31.2
35.2
45.9
0.93
4.59
1.60
4.29
7.25
1.54
51.
89.
48.
29,
32,
41.
1
5
5
7
8
8
0.79
6.17
2.12
3.52
7.33
1,35
45.2
86.3
47.8
22.7
27.8
21,7
1.
10.
3.
1.
6.
2.
50
09
69
19
84
80
49.4
89.3
41.0
6.4
21.1
19.5
Not Sensitive to Air Particulate Level
Inside
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Painting Walls/Ceilings
Wallpapering
Washing Walls
Replacing Furnace Filter
Cleaning/Oiling Air-Conditioners
Cleaning Furnace
Dry-Cleaning Draperies
Dry-Cleaning Carpeting
Shampooing Carpeting
Shampooing Furniture
0.30
0.07
3.04
0.98
0.25
0.74
0.28
0.43
1.43
0.62
76.6
36.9
42.7
50 1
12,3
62.4
20.2
12.5
47.8
12.5
0.38
0.10
2,68
0.84
0.40
0.77
0.25
0.11
1.26
0.64
72.
45.
41.
38.
21.
51.
23.
13.
44.
14.
0
4
3
7
4
8
0
5
2
7
0.54
0.12
2.99
0.63
0.50
0.38
0.27
0.09
1.13
0.71
67. S
43.1
45.8
31.4
15.1
40.5
21.1
10.7
37.8
13.4
0.
0.
3.
0.
0.
0.
0.
0.
1.
0.
29
12
36
42
32
46
51
07
23
30
59.0
52.6
37.1
25.5
16.7
46.6
17.5
6.8
39.0
12.0
Outside
1.
2.
3.
4.
5.
6.
Painting Qutsfde Walls
Painting Outside Trim
Cleaning/Repairing Awnings
Washing Automobiles
Waxing Automobiles
Maintaining Shrubs, Flowers, etc.
0.11
0.22
0.05
19.38
3.47
15.47
38.6
66.5
4.0
80.9
44.4
71.3
0.10
0.29
0.05
16.63
3.49
13.16
28.
68.
4.
66.
42.
64.
5
9
3
0
5
6
0.04
0.20
0.04
14.94
2.70
10.15
11.4
54.5
4.3
50.8
29.8
38.8
0.
0.
0.
12,
1.
3.
03
20
31
59
75
27
10.4
55.0
5.2
49.8
28.3
14.3
-------
(5) Residence characteristics including length of tenure of occupants at residence, home
rented or owned, size and number of rooms, type of material used in inside and out-
side of structure, heating and air conditioning characteristics, fuel type, number
of windows, etc. (See Table 10-10.)

The BAH study covered four pollution zones in the Philadelphia area defined by ambient
o
air monitoring data. The four zones were identified as: (Zone 1) less than 75 ug/m , (Zone
2) 75 to 100 ug/m3, (Zone 3) 100 to 125 ug/m3, and (Zone 4) greater than 125 pg/m3. Frequency
rates of performing 27 cleaning and maintenance tasks (15 inside, 12 outside) were determined
for each pollution zone by various demographic and attitudinal subgroups (see Tables 10-11 and
10-10). The BAH study used appropriate statistical survey techniques (stratified random
sampling) and collected the best set of frequency of performance and cost data to date. The
overall finding by the BAH researchers was:

that over the 27 cleaning and maintenance operations studied
the range of annual air particulate levels experienced in the
Philadelphia area (approximately 50 to 150 micrograms per cubic
meter) has no measurable effect on out-of-pocket cleaning and
maintenance costs for the residence of the over 1,500,000 house
holds in the area.

A small number of contractors were interviewed in deriving the cost information. How-
ever, not only were no statistical sampling procedures used in their selection, but also the
estimates of costs failed to consider size of the project or types and conditions of materials
to be restored. It is not unexpected, therefore, that no significant relationship was found
between the prices for any given service and the location in the four pollution zones of the
places of business of the contractors (from which labor costs were estimated in BAH) within
thi region. All estimated costs were applied uniformly across the region.
The researchers stated other conclusions:
(1) Activities related to "looking outdoors" (e.g., washing windows) were more associ-
ated with air pollution levels. These tasks represented nonsignificant low "out-
of-pocket costs" because few households contracted out such types of work.

(2) Residents of high pollution zones believed the effect of pollution on cleaning costs
to be greater than the residents of low pollution zones.

The BAH findings for Philadelphia should not be entirely accepted without at least con-
sidering that the value of the direct personal labor or time of a "do-it-yourself" resident
was not included as a cost.
10-58
-------
Although the methodological approach employed by BAH is applicable to other sampling
sites, the specific findings of the Philadelphia study cannot be automatically assumed valid
for other metropolitan areas of the nation. Factors affecting cleaning costs, such as hourly
rates and unit costs of contractor work, may differ from one major region to another. Differ-
ent types of particulate matter found from region to region produce different kinds and
degrees of soiling. Also, Philadelphia, as well as many other urban regions of the Eastern
U.S., contains many dwellings constructed of brick or masonry. Cleaning estimates based upon
Philadelphia, which contains up to 80% brick or masonry structures in the most polluted zone,
cannot be extrapolated to other regions in the nation where use of other types of construction
materials may prevail. A need clearly exists for studies similar in approach to the BAH study
to be conducted in other regions of the nation.
In 1972, Spence and Haynie presented a survey and economic assessment of the deteriora-
tion and soiling of exterior paints caused by air pollution. Although the authors discussed
effects of S0? and other pollutants on paint, their economic assessment reflects only the ef-
fects of soiling by PH. Of an estimated total annual economic loss of $704 million in 1968
dollars, $540 million was for increased exterior household painting. No damage functions were
available for use in the estimate. Instead, the calculation was based on the assumption that
exterior household paint service life is reduced by half in an (urban) area averaging 110
pg/m TSP from a service life of 6 years in a (rural) area averaging 40 yg/m . This length of
service life was in turn based on frequency of painting. Spence and Haynie explained the
derivation of the service life assumption as follows: "The Michel son linear relationship...
predicts service life of 8 and 2 years, respectively, for paint exposed to 40 and 110 pg/m of
particulate matter. The Booz, Allen and Hamilton, Inc. study shows no difference in service
life for paint exposed over this concentration range, the average paint life being 3.14 years.
Because neither study was complete in its analysis, paint lives are estimated at 6 and 3
years, respectively, for rural and urban areas."
Narayanan and Lancaster (1973) conducted a questionnaire survey in a rural area and a
polluted area in New South Wales, Australia, to determine the difference in cost of household
upkeep. The cost of maintaining a house in the Mayfield area was about $90 per year higher
than in the relatively unpolluted Rotar area. The authors attributed this cost differential
to the higher levels of air pollution and airborne particulate matter in Mayfield. However,
the accuracy of the cost data was considered questionable, because like the earlier Michelson
and Tourin studies, socioeconomic factors, including respondents' attitude, were not dealt
with and could effect a bias in the estimates.
A study by Liu and Yu (1976) was designed to generated physical and economic damage func-
tions, by receptor, for S0? and TSP, and to establish cost/benefit relationships. The study
included the effects of air pollution damage on health, materials, vegetation, and household
soiling. These authors used the BAH results on cleaning frequency for nine household cleaning
tasks, and the data for TSP levels, in a Monte Carlo technique. The technique created a
10-59
-------
sample of data pairs (cleaning frequency and TSP level) for each cleaning task. The authors
regressed cleaning frequency on TSP levels without controlling for the social and economic
factors reported and included in the BAH report. Increased cleaning frequencies were trans-
formed into monetary units for each task and were used to calculate decreases in cleaning
2
costs as TSP is lowered to 45 yg/m . These costs were extrapolated to 148 SMSA's, then summed
for a national estimate of $5 billion household soiling cost attributed to TSP. Liu and Yu
have been criticized for their disregard of the socioeconomic factors in the BAH data base and
for ignoring evidence regarding the insensitivity of high-cost cleaning and maintenance tasks
to TSP levels.
Watson and Jaksch (1978) returned to the Booz, Allen and Hamilton (1970)(BAH) study for
their assessment of benefits associated with decreased TSP levels. They introduced a signifi-
cant extension of the earlier work by including in the benefits of pollution control the
psychological and other advantages of living in a cleaner environment. Watson and Jaksch
estimated the value of these psychological and health benefits by applying the standard
measure of net contribution to consumer welfare, namely the excess of the sum that consumers
would be willing to pay for the attained level of cleanliness over the actual cost to them of
attaining that level.
To estimate the cost of achieving a given level of cleanliness and the effect on that
cost of reducing the concentration of particulate matter, they relied largely on the work of
Beloin and Haynie (see Table 10-8). The formulas derived by Beloin and Haynie lead directly
to estimates of the cost of maintaining a given level of reflectance of different surface
materials, which, however, is not the same thing as the perceived level of cleanliness, though
related to it. "Cleanliness" is a vague concept, and in order to give it the precision nec-
essary for estimation, Watson and Jaksch posited that it is the reciprocal of the difference
between the actual reflectance of a surface and its maximum reflectance, raised to a power
that depends on the rate at which reflectance decreases over time. This 'enabled them to
express the marginal cost (or price to the consumer) of maintaining a given average level of
cleanliness, Q, in the form:

MC =aPaQ, (10-14)

where a and a are empirical constants and P denotes the ambient concentration of particu-
late matter. Beloin and Haynie's data for only one kind of surface indicate a value of a of
approximately 0.5, Less detailed observations by Esmen (1973) are consistent with a = 2.
This range of estimates was used in the subsequent analysis. It should be noted that this
formula depends on both the empirical studies of reflectance and the assumed relation between
reflectance and perceived cleanliness.
In the presence of a given ambient concentration of participates, consumers can be ex-
pected to choose the level of cleanliness at which the cost of further improvement, as re-
vealed by equation (1), is just equal to the amount they are willing to pay for it. In order

10-60
-------
to estimate the amounts that consumers are willing to pay for increases in cleanliness, Watson
and Jaksch derived a demand curve for cleanliness by a detailed reanalysis of the findings of
the Booz, Allen and Hamilton study of the frequency of household cleaning in Philadelphia. As
previously mentioned, that study indicated that the ambient concentration of pollutants had no
significant effect on the frequency with which cleaning tasks were performed. Thus, people
put up with lower levels of cleanliness in heavily polluted areas (where maintaining given
levels is expensive) than in less polluted areas. In fact, to the extent that the cost of
maintaining a given level is proportional to the frequency of cleaning required, households
adapted to levels that cost the same amount to attain at all levels of ambient pollution.
That observed behavior is consistent with a demand curve with constant elasticity of -1.
Accordingly, that type of demand curve was adopted. It implies that out-of-pocket costs of
home maintenance are not affected by ambient particulate levels, while psychological satis-
factions are affected.
Watson and Jaksch then brought the estimated cost and demand curves together in the usual
manner to estimate the changes in consumer welfare associated with changes in ambient concen-
tration, by calculating the level of cleanliness that would be chosen as a function of the
ambient concentration and then the area under the demand curve up to that level, less the cost
of attaining that level. The changes in consumer welfare corresponding to changes in particu-
late matter concentrations were estimated by comparing the results of that computation for
different ambient concentrations.
These estimates were made for the households in the BAH survey, then inflated to cover
the entire Philadelphia metropolitan area, and finally extended to 123 SMSA's in the United
States, with allowance for the differing prevailing TSP concentrations in the different
SMSA's. The results are summarized in Table 10-12. Converted to 1978 dollars, the nationwide
gains to consumers from attaining primary TSP standards in all SMSA's are valued at from $1.4
billion per year to $5.1 billion, depending on which value of alpha (0.56 or 2.0) is us'ed.
The corresponding estimates for achieving the secondary standard are $2.4 to $9.1 billion. No
allowance is made in this table for any sources of uncertainty or estimating error other than
the range of assumed values of alpha (see Figure 10-10).
The Watson and Jaksch (1982) study is conceptually sound. The data and statistical
methods on which the study rests, however, limit full acceptance of its generalized infer-
ences. Application of the method with better defined parameters and to data sets outside the
Philadelphia region is necessary. Nevertheless, despite these limitations, the study remains
a primary source of estimates of the benefits to consumers of the reductions in soiling
achieved by reducing ambient TSP concentrations.
A study by Hamilton (1979) used the framework developed by Watson and Jaksch to estimate
soiling benefits for six California SMSA's. Hamilton assumed a unit elasticity of demand for
cleanliness. He assumed a 25 percent reduction in total suspended particulate matter, which
produced estimated benefits of $40 per household, which is comparable to the estimates derived
10-61
-------
TABLE 10-12. ANNUAL WELFARE GAIN FROM ACHIEVING PRIMARY AND
SECONDARY STANDARDS FOR TSP CONCENTRATION

(1971 Prices)

Average gain per household,
Philadelphia sample, dollars
Aggregate gain, Philadelphia
$ million
Aggregate gain,
123 SMSA's, $ million
Primary
Standard

$16 - 57
23 - 85
860 - 3200
Secondary
Standard

$30 - 112
44 - 165
1500 - 5700

Source: Watson and Jaksch (1982).

Note: Gains calculated from reductions from TSP concentrations in 1970.
Primary-standard: 75 (jg/m , annual average. Secondary standard:
60 (jg/m . Ranges correspond to a = 0.56 (lower limit) to a = 2
(upper limit).
10-62
-------
t£
o
Q
o
V)

O
5 2
=2.0
a =0.86
20 40 60 80

TOTAL SUSPENDED PARTICULATE MATTER
100
Figure 10-10. Increases in particulate matter concentrations are plotted
against reductions in outdoor cleaning task benefits (1978 dollars). The
range of benefits increases progressively as pollution is reduced.

Source: Derived from Watson and Jaksch (1982).
10-63
-------
by Watson and Jaksch. Total benefits for a 25 percent reduction in PM for the six California
SMSA's (including Fresno, Los Angeles-Long Beach, Sacramento, San Bernadlno-Riverside-Ontario,
San Diego, and San Francisco-Oakland) are $223 million, in 1978 dollars.
10.5.2.3 Combined Studies—Salmon (1970) was one of the first to provide a cost estimate of
soiling and material damage associated with both ambient PM and SOp concentrations, shown in
Table 10-13. He calculated economic loss by determining the value of materials exposed to
pollution and then multiplying by the estimated difference in useful lifetime between clean-
rural and polluted-urban areas. The value of exposed materials was derived by multiplying
annual production by a product lifetime estimate and then applying a labor-factor estimate and
an exposure-factor estimate. Salmon cited economic damage from SO in the United States to
the following materials, listed in decreasing order of the extent of damage: metals, cotton,
finishes and other coatings, building stone, paints, paper, and leather. Deterioration of
paint was estimated at $1.2 billion, soiling of paints at $35 billion. Soiling costs attri-
butable to PM were calculated to be $99 billion. In discussing limitations to his study
results, Salmon pointed out that data were available to examine the economic value of mate-
rials loss for only six material categories: stainless steel, zinc, building stone, leather
and paper, cotton, and paint. He cautioned that his cost estimate indicated susceptibility to
economic loss or potential loss and should not be interpreted as actually incurred economic
loss. He claimed only that the study "may be a reasonably good approximation of reality."
The ranking of materials by decreasing order of potential economic loss may be reasonably used
for setting research priorities, which was the major purpose of the study. As Table 10-14
shows, however, the Salmon study has been used quantitatively to some extent in nearly every
national estimate for materials damage costs attributable to air pollution done since 1970.
In general, his figures are relied upon for costs not estimated elsewhere, reflected in the
"other" category of materials in Table 10-14. (Discussion of Waddell (1974), Yocom and
Grappone (1978), and Freeman (1979b) later in this section will show how each of these studies
used the Salmon estimates, directly or indirectly.) Nevertheless, no subsequent study has
arrived at costs nearly as large as Salmon's.
Waddell (1974) used Salmon's list of 32 economically important materials significantly
affected by air pollutants and associated damage costs, revising estimates for materials about
which more recent reports had become available. For those costs associated with paint, he
substituted the Spence and Haynie (1972) estimate of $704 million. Gillette's SO- damage
estimate as developed in an earlier 1973 version was used to replace the S02 damages assigned
to zinc, carbon and alloy steel, fibers, cement and concrete, building bricks and stone, plas-
tics, paper, leather, and wool. Gillette's estimate of damage to these materials for 1970
attributable to S02 was $400 million. It should be noted that the Gillette (1975) study
focused on damage to metals and paints and considered damages to other materials to be insig-
nificant. Waddell assigned a zero value to copper, aluminum, stainless steel, and lead, based
on the conclusions of Fink et al. (1971). Waddell then totaled the costs associated with
10-64
-------
Table 10-13 ECONOMIC LOSS, MATERIALS DAMAGE ATTRIBUTED TO
AMBIENT EXPOSURE TO SO AND PM,
ESTIMATED BY SALMONX(1970)
(IN BILLIONS OF 1970 DOLLARS)

Material
Pai nt
Zinc
aFibers
Cement and concrete
Nickel
Tin
Aluminum
Copper
Carbon steel
Building brick
Paper
Leather
Glass
Building stone
Wood
Brass and bronze
Magnesium
Alloy steel
Bituminous materials
Gray iron
Stainless steel
Clay pipe
Chromium
Malleable iron
Silver
Gold
Plastics
Lead
Molybdenum
Rubber
Refractory ceramics
Carbon and graphite
TOTALS5
S°x
1.195
0.778
0.358
0.316
0.260
0.144
0.114
0.110
0.054
0.024
0.023
0.021
<0.001
0.018
0.018
0.014
0.013
0.009
0.002
0.002
0.002
0.001
0.001
0.001
0.001
0.001
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
3.8
PM
35.0
24.0
0.5
5.4
1.0
0.1
4.9
0.2
o!i
1.1
2.5
19.0
0.1
0^2
<0.1
<0. 1
<0.1
<0. 1
<0. 1
<0. 1
0.2
<0. 1
<0.1
<0.1
4.7
0.1
O.I
-------
TABLE 10-14. ESTIMATES OF MATERIALS DAMAGE ATTRIBUTED TO SOV AND PM
IN 1970 (IN MILLIONS OF 1970 DOLLARS) *
Ol
Ol

ESTIMATES
WADDELL (1974)
MATERIAL
CATEGORY SO., PM TOTAL

PAINTS 100 100 200
TEXTILES
AND DYES
METAL
CORROSION 400 — 400
ELECTRICAL
SWITCHES AND
COMPONENTS included in
metal corrosion
OTHER0 100 200 300
TOTAL 600 300 900
YOCOM AND GRAPPONE (1976)
SOV PM SO/PM Total
A X
200 500 700
400 — 400
included in
metal corrosion
300 100 400
900 600 1,500
FREEMAN (1979b)
SOY/PMa
J\
704
636
400
80
400
2,200

SOURCES

Spence and Haynie, 1972
Salvin, 1970
Gillette, 1975b
Robbins, 1970
and ITT, 1971
Salmon, 1970
. .
No allocation of cost to PM or SO specifically,
h
Nickel, tin, brass, bronze, magnesium, gray iron, malleable iron, chromium, molybdenum, silver, gold,

clay pipe, glass, refractory ceramics, carbon, and graphite.

Sfaddell used earlier (1973) version of Gillette (1975) report.
-------
materials remaining on Salmon's list and assigned the resulting $400 million collectively to
the category "other." No evidence exists, however, for a specific association of SO or PM
s\
with effects on these materials. In allocating these costs specifically to PM or SO , Waddell
made an adjustment to the $704 million paint damage estimate. He considered that the portion
of the costs attributable to soiling of household paint by PM ($500 million) was included in
the soiling estimate. The remaining $200 million was evenly divided between PM and SO . The
X
"other" category (see Table 10-14), derived from Salmon (1970) was allocated in proportion to
criteria pollutant emissions, excluding carbon monoxide. On this basis, $100 million was
assigned to SO and $100 million to PM for effects on materials not addressed by Gillette
)\
(1975) or Spence-Haynie (1972) or excluded by Fink (1971).
In developing a national estimate for soiling damages attributable to particulate pollu-
tion, Waddell (1974) concluded that, on the basis of studies published up to 1973 designed to
determine household soiling costs attributable to PM, "the evidence to date indicates air
pollution does not have significant economic effects in terms of household maintenance and
cleaning operations," He declined to make an estimate,of economic damages associated with
soiling on the basis of these studies.
As an alternative, Waddell turned to studies relating residential property value differ-
entials to air pollution. Of the nine studies Waddell reviewed, six included PM and S0?,
while three related property value differentials only to S0~ (as measured by sulfation). Of
the six that included PM, three studies found significant negative correlations between pro-
perty value differentials and particulate pollution levels. The strongest associations re-
lated property value differentials to sulfation rates. From these studies, Waddell calculated
"marginal capitalized sulfation damage coefficients" for residential structures of $200-$500
2
for 0.1 mg/ S03/100 cm /day. Using these coefficients, he estimated the annual national cost
for 1970 of air pollution damage measured via the property differential method to be $3.4 to
$8.4 billion. He used the midpoint of this range, $5.9 billion, as a best annual estimate. In
estimating and allocating gross damage estimates between pollutants, Waddell adjusted the $5.9
billion estimate by subtracting $50 million (1970 dollars) for ornamental losses as determined
by Benedict, et al., (1971). By assumption, Waddell postulated the property value estimate
to measure aesthetic and soiling costs. He then assumed that the total cost of $5.8 billion
could be allocated by evenly dividing the damage between PM and SO , or $2.9 billion each.
Yocom and Grappone (1976) also estimated economic costs for air pollution damage to mate-
rials, in what was essentially a revision of Waddell's study. They concluded that the cost of
materials damage attributable to SO and PM in 1970 was $1.5 billion. As may be seen in Table
10-14, they departed from Waddell's approach in two categories: (1) The entire "other" cate-
gory derived from Salmon (1970) was assigned to PM and SO (both Salmon and Waddell had con-
sidered these costs attributable to all major air pollutants); and (2) The paint cost was not
adjusted to distinguish between industrial and exterior house paint. These two differences
10-67
-------
account for the $0.5 billion difference between the Waddell (1974) and Yocom and Grappone
(1976) estimates. The latter made no attempt to calculate a separate estimate for soiling
costs. Considering that the household paint soiling figure would have been assigned to a
soiling estimate had Yocom and Grappone developed one, the two studies are essentially in
agreement.
Freeman (1979b) estimated benefits from reduced materials damage and soiling associated
with improvements in air quality from 1970 to 1978. He assumed a 20 percent improvement in
levels of S0_ and in levels of TSP. To establish a basis for evaluation of the realized
benefits associated with materials damage, Freeman essentially revised Waddell's (1974)
estimate and updated the result to 1978 dollars. The significant features of the revision, as
shown in Table 10-14, are that Freeman included a $636 million figure for damage to textiles
and dyes, carried forward the entire $400 million "other" category derived from Salmon (1970),
did not assign the $500 million related to soiling of household paint to his separate soiling
estimate, and included a figure for damage to electrical contacts and switches, which Waddell
had assumed to be included in metal corrosion. Attribution of $636 million in textile and dye
damage to PM and SO was based on a study by Salvin (1970). Later evidence, as discussed in
Section 10.2.3, has shown effects of SO,, on dyes to be insignificant compared to the effects
of other pollutants, notably NO , although some evidence exists for S0? effects on fabric
strength. Similarly, the damage to electrical contacts and switches does not reflect recent
practice regarding use of corrosion-resistant metals. Freeman's estimate of damage to mate-
rials was not broken out for SO and PM separately; his estimate for damages attributable to
both pollutants in 1970 was $2.2 billion, equivalent to $3.7 (±1.9) billion in 1978 dollars.
Freeman then calculated the benefits for a 20-percent improvement in SO and TSP levels from
,X
1970 to 1978; realized benefits were estimated to be $0.7 (±0.4) billion per year.
In calculating benefits for reduced soiling and cleaning attributable to particulate
pollution, Freeman (1979b) also used the work of Watson and Jaksch (1978) to arrive at an
estimated range of $1.5 to 11.7 billion in 1978 dollars. Based upon 1970 and 1978 national
ambient particulate matter measurements, Freeman (1979b) inferred that between a third and a
half of the Watson and Jaksch (1978) estimate of the benefits of attaining the secondary
standard had been realized by 1978. Freeman's calculations showed this to be a range between
$0.5 and $5.0 billion per year, or a reasonable midpoint estimate of about 2.0 billion per
year. Incremental benefits of moving from 1970 particulate matter levels to achievement of
the secondary standard in 1978 are estimated at $2.6 billion per year. However, Watson &
Jaksch revised their paper subsequent to the publication of Freeman (1979b). The later Watson
& Jaksch (1982) paper, therefore, will result in Freeman's estimates being changed.
In a report prepared for the National Commission on Air Quality, SRI International (1981)
developed estimates of nonhealth benefits of meeting the secondary national ambient air qual-
ity standards. To estimate benefits resulting from reduced material damage, SRI used Salmon's
(1970) estimate of pollution damage for their base year of 1969. More recent physical damage
10-68
-------
functions were incorporated when available, and a national stock inventory for 1979 was devel-
oped. Salmon's assumptions regarding economic life of materials was also employed in the SRI
estimate. In calculating benefits associated with materials damage reduction from 1969 to
1979, SRI assumed no significant change for TSP and a 20 percent improvement in SCL levels.
The resulting estimate- of benefits associated with meeting the secondary standards was $3.9
billion in 1980 dollars. The authors stated that the estimate was similar to the conclusion
reached by Freeman (1979b), although the two studies were not directly comparable.
In considering the benefit of reduced soiling, SRI reviewed the state of the art of the
economics associated with soiling and concluded that: "(i) we are unable to quantify house-
hold soiling effects from alternative particulate levels; (ii) there is no basis at this time
for determining indoor-outdoor pollution concentration ratios and (iii) the soiling effects of
large versus small particulates [sic] are not well understood." To provide a rough estimate
of soiling-related benefits, SRI applied a $6.63 marginal benefit of soiling per household per
unit change in TSP concentrations to counties violating the secondary TSP standard to arrive
at an estimated $0.5 billion benefit for meeting the secondary standard. The marginal benefit
was derived from a damage function appearing in an unpublished EPA report. The very prelimi-
nary basis for the SRI soiling estimate diminishes its immediate utility; however, the identi-
fication by SRI of problems and research needs in the area of soiling is useful in evaluating
existing estimates of soiling benefits.
Freeman (1979b) noted that in the materials damage area, "the weakest link in most of
these (materials damage cost) studies is the estimation of quantities of material at risk and
exposure levels." While further application of the exposure inventory (Stankunas et al. 1981)
approach may provide information necessary for more reliable application of the physical
damage function approach, new approaches to estimating the cost of pollutant effects on wel-
fare, including materials damage, have been recently developed.
Manuel et al. (1981) developed an estimate of benefits to be realized from reductions in
ambient SO- and TSP concentrations. They drew heavily on econometric analysis to identify the
influence of changes in TSP and SO, levels on certain production and consumption decisions of
selected components of the agricultural, manufacturing, household, and electric utility
sectors of the economy. The approach allows consideration of substitution possibilities by
both buyers and suppliers. Indices of decision behavior of decision-making units within each
sector were chosen to reflect the effects of TSP and S02 on agricultural yield, materials
damage, and soiling. The theoretical approach and empirical methods employed appear to be
well based. However, additional analyses of the study results are needed to explain more
fully certain implied behavioral adjustments. In addition, more extensive air quality data
and more detailed economic data would have permitted not only greater coverage but might have
also served to reduce the degree of uncertainty associated with the estimated benefits.
10-69
-------
10.5.3 Estimating Benefits from Air Quality Improvement, 1970-1978
Although both the data base and the method used for estimating benefits associated with
reduced materials damage and soiling are fraught with shortcomings, it is possible to judge
the direction and rough order of magnitude of changes in these benefits as a function of
changes in air quality. Certain features of the data base discussed in Section 10.5.2 must be
carefully considered.
Between 1970 and 1978, almost without exception, aggregate estimates of materials damage
and soiling used 1970 as the base year for comparison with later years. Major studies of
material damage performed after 1970 relied principally on the physical damage function
approach and sought to revise or update relevant portions of cost estimates of 1970 damage.
As a result, the data base remained essentially unchanged. Largely because of this, estimates
of materials damage costs vary within a much narrower range than do estimates of soiling
costs, where a variety of approaches have been applied. These approaches were used to develop
estimates for a specific area or region of the country, then extrapolated for a national
estimate. Both differences in approach and extrapolation partially account for the wide range
in resulting estimates of aggregate soiling costs.
Host cost estimates are based on PM or SO levels over and above some base level. Because
,A
of changes in analytical techniques and in number and location of monitoring sites, determina-
tion of reduction in S0? levels between 1970 and 1978 is difficult and should be approached
with caution. Trends for TSP are less equivocal because, although monitoring sites were
changed, the methodology remained the same (see Chapters 3 and 5). Cost (benefit) estimates
of increasing (decreasing) pollutant levels must include any uncertainties of the air quality
trends data. For the purposes of this section, SO- and TSP trends discussed in Chapter 5
will be used.
Estimates of cost associated with SO are dominated by damage to materials, principally
/c
metals, exposed outdoors to ambient air. Very little material indoors is strongly affected by
SO,. In contrast, estimates of cost associated with soiling by PM have emphasized residential
household cleaning and maintenance, including a significant portion associated with indoor
exposure. As noted by SRI (1981) and discussed in Chapter 5, indoor TSP levels do not corre-
late well with outdoor TSP levels. Clearly, then, to the extent that soiling is attributable
to indoor PM levels, the relationship between soiling costs and outdoor PM levels (measured as
TSP) is tenuous. A further difficulty in relating soiling to TSP is the fact that physical
damage functions developed for the purpose have correlated loss of reflectance by the receptor
surface to variation in TSP concentrations. When the receptors vary in texture and color as
much as do the various building materials employed in developing damage functions for soiling,
the resulting poor correlation (see Table 10-8) is inevitable.
The range of estimated economic loss associated with SO as displayed in Table 10-14 was
yx
discussed in Section 10.5.2.1. There is general agreement that damage for 1970 was approxi-
mately $1 billion, plus or minus 50 percent. As noted earlier, there is evidence to suggest
that paint damage costs should exclude the $500 million developed from Spence and Haynie's

10-70
-------
association of TSP with exterior residential painting frequency. Similarly, exclusion of the
costs associated with damage to dyes, to electrical switches and contacts, and to a number of
materials included in the "other" category is indicated by physical evidence discussed pre-
viously. On the other hand, there is evidence of S0? effects on paint, on fabric strength,
and building stone, concrete, and masonry. The Stankunas et al. (1981) study reported $31
million damages (1981 dollars) in S0?-related paint erosion in the Boston area alone, although
this is probably an overestimate, since it was based on a damage function developed for oil-
based paint. Given no reliable basis for quantifying these effects, a reasonable course is to
retain the $400 million "other" category in Table 10-14 to reflect damage to materials that
has been shown to exist, but for which no national estimates of cost have been developed.
Adopting this course, one can sum the remaining costs by category: $100 minion for damage to
industrial paints based on the Spence and Haynie (1972) study and $400 million for metal corro-
sion, totaling $500 million. Adding this to the $400 million for damage to fabrics, stone,
concrete, and other paint results in a total estimate of $900 million for materials damage
associated with SO levels in 1970. Assuming 50 percent error, the range is from $0.45 to 1.4
billion in 1978 dollars, with a midpoint of $0.9 billion.
Figure 10-11 illustrates how an estimate of benefits associated with improvements in S0?
can be obtained. As shown in the figure, S0? levels have decreased by 44.5 percent during the
period 1970 to 1978. If one assumes that the improvement is reflected linearly in decreased
materials damage, the annual realized benefit for 1978 is approximately $0.4 billion.
Reported estimates of loss associated with soiling attributable to PM range up to $99
billion. Excluding the $99 billion estimated by Salmon (1970), who assumed that soiling of
any material carried a cost, estimates of soiling damages in 1970 dollars generally are about
$5 billion. Of the studies on household soiling attributable to airborne particles, three
form a continuous development of a common data base.
As previously mentioned, the Booz, Allen and Hamilton study (1970) is the most compre-
hensive collection of data for cost and frequency of performing household cleaning and main-
tenance tasks. The inclusion of socioeconomic variables and reliable TSP data allowed for
subsequent reanalysis by Watson and Jaksch (1978, 1982). Freeman's (1979b) estimate of $2.0
to 11.7 billion in 1978 dollars, equivalent to a range of $1.2 to 7.0 billion in 1970 dollars,
although requiring adjustment to the 1982 Watson and Jaksch work, is the best synthesis of the
data base developed by Booz, Allen and Hamilton (1970). In contrast, the Liu and Yu (1976)
study was superficial in its treatment of the Booz, Allen and Hamilton data, and the Waddell
estimate was based on the relationship of property value differentials to sulfation levels
rather than TSP.
Freeman's range included benefits attributed to decreased indoor soiling and maintenance
tasks. As SRI noted, and as is discussed in Chapter 5, the relationship between indoor
particulate pollution and outdoor TSP levels is tenuous. Levels of outdoor particulate matter
less than 2-3 urn in size, however, tend to be more strongly associated with levels of fine-
mode indoor particulate matter, although specific information on this subject is limited.

10-71
-------
50
40
"I
30
£ 20
10
i \ i r
i i i r
SO2,
I I
2.5
2.0
1.5
c/j
cc
o
Q
CO
o
1.0
1970 1971 1972 1973 1974 1975 1976 1977 1978

YEAR
Figure 10-11. Improvement in U.S. annual average SO2 levels from 3
in 1970 to 18 jug/m3 in 1978 has resulted in approximately $0.4 billion esti-
mated economic benefit for 1978. Area labeled '$ DAMAGES' refers to 1978
dollar value of damage to materials attributable to $ulfur oxides.
10-72
-------
To arrive at some lower bound estimate of benefit associated with reduced soiling asso-
ciated with decreased TSP levels, it seems prudent to confine such an estimate to outdoor
residential soiling and maintenance. From Watson and Jaksch (1978, 1982), the observed cost
outlays for cleaning and maintenance costs associated with outdoor pollution was 38 percent of
the total outlays for-both indoor and outdoor tasks. Applying the 38 percent factor to omit
indoor pollution from the Watson and Jaksch estimates reduces the amount of total benefits
from the range of $2.4 to $9.1 billion (Watson and Jaksch, 1982) to the range of $1.0 to $3.5
billion, or a best estimate of $2.0 billion in 1978 dollars for achieving a level of 60 ug/m
2
TSP. Considering an air quality improvement of 20 percent for annual average TSP (72 ug/m in
1970 to 58 ug/m in 1978), the range of benefits for improved TSP levels during the period
would be $0.2 to $0.7 billion, depending on which assumption of cleaning frequency is used.
One may, from the estimates of benefits resulting from improvement in TSP and S0~ over
the period 1970-1978, derive a rough notion of benefit per unit improvement in pollutant level
by dividing the 14 ug/m improvement in TSP or S09 by the associated annual benefit. This
3
exercise results in an, approximate figure of $30 million annual benefit for each ug/m re-
duction in S02; the range for TSP is $14 to $50 million. However, as noted at the outset of
this section, figures derived from a data base so replete with uncertainty can only be indic-
ative of the general direction and magnitude of the change in benefits associated with changes
in pollutant levels and by no means should be accepted as a reliable benefit estimate.
10-5.4 Summary of Economic Damage of Particulate Hatter/Sulfur Oxides to Naterials
The damaging and soiling of materials by airborne pollutants have an economic impact, but
this impact is difficult to measure. The accuracy of economic damage functions is limited by
several factors. One of the problems has been to separate costs related to SO and particles
from each other and those related to other pollutants, as well as from those related to normal
maintenance. Cost studies typically involve broad assumptions about the kinds of materials
that are exposed in a given area and then require complex statistical analysis to account for
a selected number of variables. Attitudes regarding maintenance may vary culturally, further
confounding the problem of quantifying economic impact.
Studies have used various approaches to determine pollutant-related costs for extra
cleaning, early replacement, more frequent painting, and protective coating of susceptible
materials, as well as other indicators of the adverse economic effects of pollutants. No
study has produced completely satisfactory results, and estimates of cost vary widely. In
1978 dollars, the estimated economic loss for 1970 SO damage was approximately $0.9 billion;
)\
for TSP exterior soiling of residential structures, $2 billion. Damage functions indicate
that reductions in pollutants will decrease physical and, therefore, economic damage. While
the data base and methodology for attribution of costs to SO and PM are incomplete at this
A
time, it is possible to estimate gains from improvements in air quality from 1970 to 1978. In
1978 dollars, annual gain from SO reduction is estimated to be $0.4 billion. Based on
/\
existing studies, the direct monetary benefits from further reduction in SO levels are likely
10-73
-------
to prove minor. The annual benefit of decreased soiling owing to declining levels of PM
between 1970 and 1978 is estimated to be $0.2 to $0.7 billion in 1978 dollars.

10.6 SUMMARY AND CONCLUSIONS, EFFECTS ON MATERIALS
The nature and extent of damage to materials by SO and PM have been investigated by
A
field and laboratory studies. Both physical and economic damage functions have been developed
for specific damage/effect parameters associated with exposure to these pollutants. To date,
only a few of these functions are relatively reliable in determining damage, while none has
been generally accepted for estimating costs.
The best documented and most significant damage from SO and PM is the acceleration of
A
metal corrosion and the erosion of paint. Erosion of building materials and stone due to SO
)\
is also established, but the importance of SO relative to other agents is not clear. Although
evidence of damage to fibers (cotton and nylon), paper, leather and electrical components has
been reported, reliable damage estimates have not.
Relatively accurate physical damage functions have been calculated for the effects of S0?
on the corrosion of galvanized steel. Determination of variables such as time of wetness and
surface configuration affect the applicability of the functions. Similar, but less accurate,
functions have also been developed for estimating erosion rates of oil-based paints from
exposure to SOy- The large-scale replacement of oil-based paint by much more SOp-resistant
latex paint, however, makes these estimates obsolete. The least reliable of the "significant"
damage functions are those for soiling from PM. The poorly understood deposition rates and
poorly characterized chemical and physical properties related to reflectance make general
application of the functions difficult, if not impossible.
The limitations of these and other physical damage functions hinder accurate estimates of
total material damage and soiling. Coupled with these limitations is the lack of material ex-
posure estimates. These problems presently preclude complete and accurate estimates of the
costs of damage based on a physical damage function approach. Studies based on this approach
estimated materials damage in 1970 attributable to SO as ranging from $0.6 to 1.2 billion, in
A.
1970 dollars. Estimates of soiling costs due to PM were based principally on other approaches
and ranged up to $99 billion. Best estimates of economic loss in 1970 attributable to S02 and
TSP, in 1978 dollars, are $0.9 billion in materials damage and $2 billion in soiling, respec-
tively. Improvements in S0? and TSP have resulted in estimated annual benefits of $0.4 for
S02 and $0.2 to $0.7 billion for TSP. These estimates are crude, but can serve to represent
the direction and magnitude of changes in benefits associated with improvement in air quality.
Approaches to cost estimation with data requirements different from those necessary for the
physical damage function approach have been attempted. Whether these approaches yield results
adequate for decisionmaking purposes is not clear at present.
10-74
-------
10.7 REFERENCES

Abe, H., Y. Ishii, and H. Kato. Evaluation of atmospheric factors by analyses of corrosion
products and surface deposits on copper plates. Rail. Tech. Res. Inst. 12:170-174, 1971.

Anderson, J. W. Sulphur in Biology. Studies in Biology 101, University Park Press,
Baltimore, MD, 1978. pp. 17-19.

Arnold, L., D.B. Honeyborne, and C.A. Price. Conservation of natural stone. Chem. Ind.
(London) 8:345-347, 1976.

Banov, A. Paintings and Coatings, Structures Publishing Co., Farmington, MI, 1973.

Barton, K. Protection Against Atmospheric Corrosion: Theories and Methods. John Wiley &
Sons, New York, NY, 1976.

Barton, K. The influence of dust on atmospheric corrosion of metals. Werkst. Korros.
8/9:547-549, 1958.

Beaver, H. Committee on Air Pollution: Report, HMSO, London, England, 1954.

Beloin, N.J. Fading of dyed fabrics by air pollution. , Text. Chem. Color. 4:77-82, 1972.

Beloin, N.J., and F.H. Haynie. Soiling of building materials. J. Air Pollut. Control Assoc.
25:399-403, 1975.

Benedict, H. L., C. J. Miller, and R. G. Olson. Economic impact of air pollutants on plants
in the United States, 1971.

Biefer, G.J. Atmospheric corrosion of steel in the Canadian Arctic. Mater. Perform.
20:16-19, 1981.

Bennett, L.H., J. Kruger, R.L. Parker, E. Passaglia, C. Reimann, A.W. Ruff, and H. Yakowitz.
Economic Effects of Metallic Corrosion in the United States. Part I: A Report to the
Congress by the National Bureau of Standards. NBS Special Publication 511-1, U. S.
Department of Commerce, National Bureau of Standards, Washington, DC, May 1978.

Bird, C.E. Corrosion behavior of galvanized sheet in relation to variation in coating
thickness. Mater. Prot. and Perform., 16, 14-16, 1977.

Booz, Allen and Hamilton, Inc. Study to Determine Residential Soiling Costs of Particulate
Air Pollution. APTD-0715, U.S. Department of Health, Education and Welfare, National Air
Pollution Control Administration, Raleigh, NC, October 1970.

Brauns, E., and U. Kalla. Corrosion rate of plain carbon steels in land, sea, and industrial
atmospheres. Stahl Eisen, 85:406-12, 1965.

Bresle, A. The corrosion of steel and the dangerous chlorides. Met. Finish. 74:23-25, 30,
1976.

Brysson, R.S., B.J. Trask, J.B. Upham, and S.G. Booras. The effects of air pollution on
exposed cotton fabrics. J. Air Pollut. Control Assoc. 17_:294~298» 1967.

Campbell, G.G. , G.G. Shurr, D.E. Slawikowski, and J.W. Spence. Assessing air pollution damage
to coatings. J. Paint Techno!. 46:59-71, 1974.

Carey, W. F. Atmospheric deposits in Britain—a study of dinginess. Int. J. Air. Pollut.
2:1-26, 1959.

10-75
-------
Cavender, J. H., W. M. Cox, M. Georgevitch, N. A. Huey, G. A. Jutze, and C. E. Zimmer.
Interstate Surveillance Project: Measurement of Air Pollution Using Static Monitors.
APTD-0666, U.S. Environmental Protection Agency, Research Triangle Park, 1C, May 1971.

Chandler, K.A., and M.B, Kilcullen. Survey of corrosion and atmospheric pollution in and
around Sheffield., Br. Corros. J. 3:80-84, 1968.

Cowling, J.E., and M.E. Roberts. Paints, varnishes, enamels, and lacquers. In: Deterioration
of Materials-Causes and Preventive Techniques. G. A. Greathouse and C. J. Wessel, eds.,
Reinhold Publishing Corp., New York, NY, 1954. pp. 596-645.

Crocker, T.D. Urban Air Pollution Damage Functions: Theory and Measurement. California
University, Riverside, CA, 1970.

Del Monte, M.C. Sabbioni, and 0. Vittori. Airborne carbon particles and marble deterioration.
Atmos. Environ. 15:645-652, 1981.

Duncan, J.R. , and D.J. Spedding. The mode of initial reaction of SO, at a metal surface.
Corros. Sci. 14:241-249, 1974. L

Eckhardt, F.G.W. Microorganisms and weathering of a sandstone monument. In: Environmental
Biogeochemistry and Geochemistry, vol 2: The Terrestrial Environment, Proceedings of the
Third International Symposium, Ministry of Sciences and Arts of Niedersachsen, Deutsche
Forschungsgemeinschaft, Wolfenb|itel, Germany, 1977. W. E. Krumbein, ed. , Ann Arbor
Science Publishers, Inc., Ann Arbor, MI, 1978. pp. 675-686.

Esmen, N.A. A direct measurement method for dustfall. J. Air Pollut. Control. Assoc,
23:34-36, 1973.

Evans, U.R. Mechanism of rusting under different conditions. Br. Corros. J. J:10-14, 1972.

Ergang, R., and M.B. Rockel. Corrosion resistance of stainless steels in the
atmosphere—evaluation of the results of weathering tests of up to 10 years duration.
Werkst. Korros. 26:36, 1975.

Exley, L.M. A practical review of residual oil firing problems and solutions. Combustion
41:16-23, 1970.

Faith, W.L. Effects of atmospheric pollutants. In: Encyclopedia of Environmental Science
and Engineering, J. R. Pfafflin and E. N. Ziegler, eds. Gordon and Breach Science
Publishers, New York, NY, 1976. pp. 219-223.

Fink, F.W., F.H. Buttner, and W. K. Boyd. Technical-Economic Evaluation of Air-Pollution
Corrosion Costs on Metals in the U.S. APTD-0654. U.S. Environmental Protection Agency,
Research Triangle Park, NC, February 1971.

Fleetwood, M.J. Corrosion—resisting metal. 1. Non-ferrous alloys for marine applications.
Corros. Prev. Control 22:6-12, 1975.

Fochtman, E.G., and G. Langer. Automobile paint damaged by airborne iron particles. J. Air
Pollut. Control Assoc. 6:243-247, 1957.

Freeman, A.M., III. The Benefits of Environmental Improvement, Theory and Practice. Johns
Hopkins University Press, Baltimore, MD, 1979a.

Freeman, A.M., III. The Benefits of Air and Water Pollution Control: A Review and Synthesis
of Recent Estimates. Report prepared for the Council on Environmental Quality. Bowdoin
College, Brunswick, ME, 1979b.

10-76
-------
Gajendragadkar, S.K. Effects of sulphur dioxide on materials. Chem. Age India 28:673-677,
1977.

Gauri, K. L. Effect of acid rain on structures. In: Acid Rain, Proceedings of the National
Convention Workshop Session, American Society of Civil Engineers, Boston, Massachusetts,
April 2-6, 1979. American Society of Civil Engineers, Boston, MA, 1979. pp. 55-75,
Reprint 3598.

Gibbons, E.V. The Corrosion Behaviour of the Major Architectural and Structural Metals in
Canadian Atmospheres. Summary of Ten-Year Results of Group I. NRCC 11630, National
Research Council of Canada, Ottawa, Canada, October 1970.

Gillette, D.G. S02 and material damage. J. Air Pollut. Control Assoc. 25:1238-1243, 1975.

Gillies, K.A., and D.J. Youston. Social effects of atmospheric emissions. Presented at the
69th Annual Meeting, Air Pollution Control Association, Portland, Oregon, June 27-July 1,
1976. Paper no. 76-19.4.

Glass, N. Environmental Effects of Increased Coal Utilization: Ecological Effects of Gaseous
Emissions from Coal Combustion. EPA-600/7-78-108, U.S. Environmental Protection Agency,
Corvallis, OR, June 1978.

Guttman, H. Effects of atmospheric factors on the corrosion of rolled zinc. In: Metal
Corrosion in the Atmosphere, a Symposium Presented at the Seventieth Annual Meeting,
American Society for Testing and Materials, Boston, Massachusetts, June 25-39, 1967.
ASTM Special Technical Publication no. 435, American Society for Testing and Materials,
Philadelphia, PA, 1968. pp. 223-239.

Guttman, H. , and P.J. Sereda. Measurement of atmospheric factors affecting the corrosion of
metals. In: Metal Corrosion in the Atmosphere: a Symposium Presented at the Seventieth
Annual Meeting, American Society for Testing and Materials, Boston, Massachusetts, June
25-30, 1967. ASTM Special Technical Publication no. 435, American Society for Testing
and Materials, Philadelphia, PA, 1968. pp. 326-359.

Haagenrud, S. , and B. Ottar. Long range transport of air pollution and corrosion effects.
In: Proceedings of the Seventh Scandinavian Corrosion Congress, Trondheim, Norway, 1975.
pp. 102-115.

Hamilton, J. D. A synthesis and critical review of methods for estimating the economic damage
of air pollution. In: Proceedings of the 72nd Annual Meeting of the Air Pollution
Control Association, June 1979.

Hancock, R.P., N.A. Esmen, and C.P. Furber. Visual response to dustiness. J. Air Pollut.
Control Assoc. 26:54-57, 1976.

Hansen, J. Ailing Treasures. Sci. 80 [Eighty] 1:58-61, 1980.

Barker, A.B., F.B. Mansfeld, D.R. Strauss, and D.D. Landis. Mechanism of SOp and H^SO.
Aerosol Zinc Corrosion. EPA-600/3-80-018, U.S. Environmental Protection Agency, Researcn
Triangle Park, NC, January 1980.

Hay, T.K., and G.G. Schurr. Moisture diffusion phenomena in practical paint systems. J.
Paint Technol. 43:63-72, 1971.

Haynie, F.H. The economics of clean air in perspective. Mater. Prot. Perf. 13:33-38, 1974.

Haynie, F.H. Air pollution effects on stress induced intergranular corrosion of 7005-T53
aluminum alloy. Jjn: Stress Corrosion - New Approaches. ASTM Special Technical
Publication 610, American Society for Testing and Materials, Philadelphia, PA, 1976.
pp. 32-43.

10-77
-------
Haynie, F.H. Theoretical air pollution and climate effects on materials confirmed by zinc
corrosion data. In: Durability of Building Materials and Components. P. J. Sereda and
G. G, Levitan, eds., ASTM Special Technical Publication 691, American Society for Testing
and Materials, Philadelphia, PA, 1980. pp. 157-175.

Haynie, F. H., J. W. Spence, and J. B. Upham. Effects of Gaseous Pollutants on Materials—A
Chamber Study. EPA-600/3-76-015, U.S. Environmental Protection Agency, Research Triangle
Park, NC, February 1976,

Haynie, F.H., J.W. Spence, and J.B. Upham. Effects of air pollutants on weathering steel and
galvanized steel: a chamber study. _In: Atmospheric Factors Affecting the Corrosion of
Engineering Metals. S. K. Coburn, ed. , ASTM Special Publication 646, American Society
for Testing and Materials, Philadelphia, PA, 1978. pp. 30-47.

Haynie, F.H., and J.B. Upham. Correlation between corrosion behavior of steel and atmospheric
pollution data. In: Corrosion in Natural Environments. ASTM Special Technical
Publication 558, American Society for Testing and Materials, Philadelphia, PA, 1974. pp.
33-51.

Haynie, F.H., and J.B. Upham. Effects of atmospheric pollutants on the corrosion behavior of
steels. Mater. Prot. Perform. 10:18-21, 1971.

Haynie, F.H., and J.B. Upham. Effects of atmospheric sulfur dioxide on the corrosion of zinc.
Mater. Prot. Perform. 9_: 35-40, 1970.

Hershaft, A. Air pollution damage functions. Environ. Sci. Techno!. 10:992-995, 1976.

Holbrow, G.L. Atmospheric pollution: its measurement and some effects on paint. J. Oil
Color Chem. Assoc. 45:701-718, 1962.

Hoshizawa, K., and K. Koyata. Formation and their prevention of acid smuts at oil-fired
burner. Nenryo Kyokaishi 49:656-665, 1970.

ITT Electro-Physics Laboratories, Inc. A Survey and Economic Assessment of the Effect of Air
Pollutants on Electrical Components, vol 1, Sections 1 through 9. APTD-0797, U.S.
Environmental Protection Agency, Research Triangle Park, NC, August 1971.

Ireland, F.E. Pollution by oxides of sulfur. Chem. Eng. (London) (221): CE 261-CE262, 1968.

Johnson, J.B., P. Elliott, M.A. Winterbottom, and G.C. Wood. Short-term atmospheric corrosion
of mild steel at two weather and pollution monitored sites. Corros. Sci. 17:691-700,
1977.

Karraker, D.G. The kinetics of the reaction between sulfurous acid and ferric iron. J. Phys.
Chem. 67:871-874, 1963.

Kingery, R. E. The extent of the paper problem in large research collections and the
comparative costs of available solutions. In: Summary of a Conference,
Permanent/Durable, American Library Association and Virginia State Library, Washington,
DC, September 16, 1960. Virginia State Library Publications no. 16, Virginia State
Library, Richmond, VA, 1960. pp. 36-41.

Kottori, M.S. The corrosion rates of metals exposed to pulp-mill .and smelter emissions.
Presented at the 73rd Annual Meeting, Air Pollution Control Association, Montreal,
Quebec, June 22-27, 1980. Paper no. 80-67.4.
10-78
-------
Krumbein, W.E,, and C. Lange. Decay of plaster, paintings and wall material of the interior
of buildings via microbial activity. In: Environmental Biochemistry and Geochemistry,
vol. 2: The Terrestrial Environment, Proceedings of the Third International Symposium,
Minister of Sciences and Arts of Niedersachsen and Deutsche Forschungsgemeinschaft,
Wolfbgttel, Germany, 1977. W. E. Krumbein, ed,, Ann Arbor Science Publishers, Inc., Ann
Arbor, MI, 1978. pp. 687-697.

Kucera, V. Effects of sulfur dioxide and acid precipitation on metals and anti-rust painted
steel. Ambio 5:243-248, 1976.

Larrabee, C.P. Mechanisms by which ferrous metals corrode in the atmosphere. Corrosion
(Houston) 15:526, 1959.

Larrabee, C.P., and S,K. Coburn. The atmospheric corrosion of steels as influenced by changes
in chemical composition. Jji: First International Congress on Metallic Corrosion,
London, April 10-15, 1961. Butterworth and Co., Ltd., London, England, 1962. pp.
276-285.

Litvin, A. Clear coatings for exposed architectural concrete. J. PCA Res. Dev. Lab.
10:49-57, 1968.

Liu, B., and E.S. Yu. Physical and Economic Damage Functions for Air Pollutants by Receptor.
EPA 600/5-76-011, U. S. Environmental Protection Agency, Con/all is, OR, September 1976.

Luckat, S. Investigations concerning the protection against air pollutants of objects of
natural stone. Staub Reinhalt. Luft 32: May 1972.

Luckat, S. Stone deterioration at the Cologne Cathedral and other monuments due to action of
air pollutants. In: Proceedings of the 4th International Clean Air Congress, Tokyo,
1976. Japanese Union Air Pollution Prevention Association, Tokyo, Japan, 1977. pp.
128-30.

Maler, K.G., and R.E. Wyzga. Economic Measurement of Environmental Damage: A Technical
Handbook. Organization for Economic Co-operation, Paris, France, 1976.

Mansfeld, F.B. Regional Air Pollution Study. Effects of Airborne Sulfur Pollutants on
Materials. EPA-600/4-80-Q07, U.S. Environmental Protection Agency, Research Triangle
Park, NC, January 1980.

Mansfeld, F. , and J.V. Kenkel. Electrochemical monitoring of atmospheric corrosion phenomena.
Corros. Sci. 16:111-122, 1976.

Manuel, Ernest H. , Jr., R.L. Horst, Jr., K.M. Brennan, W.N. Lanen, M.C. Duff, J.K. Tapiero.
Benefits Analysis of Alternative Secondary National Ambient Air Quality Standards for
Sulfur Dioxide and Total Suspended Particulates: vol. I-V. EPA Contract No. 68-02-3392,
Mathtech, Inc., Princeton, NJ, July 1981.

Matsushima, I., Y. Ishizu, T. Ueno, M. Kanazashi, and K. Horikawa. Effect of structural and
environmental factors on the practical use of low-alloy weathering steel. Boshoku
Gijutsu 23:177-182, 1974.

McBee, W.C. and T.A. Sullivan. Development of Specialized Sulfur Concretes. R.I. 8346, U.S.
Department of the Interior, Bureau of Mines, Washington, D.C. 1979.

McCarthy, E. F., A. R. Stankunas, and J. E. Yocom. Benefit model for pollution effects on
material TRC Environmental Consultants, Inc. 1981.

Michelson, I., and B. Tourin. Comparative methods for studying costs of air pollution.
Public Health Rep. 81(6):505-511, 1966.

10-79
-------
Miehelson, I., and B. Tourin. Report on Study of Validity of Extension of Economic Effects of
Air Pollution Damage from Upper Ohio River Valley to Washington, DC Area. Environmental
Health and Safety Research Associates, August 1967.

Hichelson, I., and B. Tourin. The household costs of living in polluted air versus the costs
of controlling air pollution in the twin Kansas Cities Metropolitan Area. Report to the
U.S. Public Health , Service Conference on Air Pollution Abatement in the twin Kansas
Cities Metropolitan Area, 1968. Environmental Health'and Safety Research Associates, New
Rochelle, NY, 1968a.

Hichelson, I., and B. Tourin. The household cost of air pollution in Connecticut. Report to
the Connecticut State Department of Health. Environmental Health and Safety Research
Associates, New Rochelle, NY, 1968b.

Moore, O.C., and O.R. O'Leary. Evaluation of Structural Steel Coatings in Relation to
Industrial Atmospheric Conditions, West Virginia Department of Highways, Research
Project 23, West Virginia Department of Highways, Charleston, WV, January 1975.

Narayanan, R.L., and B.W. Lancaster. Household maintenance costs and particulate air
pollution. Clean Air (Melbourne) ]_: 10-13, 1973.

National Research Council, Subcommittee on Air Particles. Airborne Particles. University
Park Press, Baltimore, MD, 1979.

Newton, R.G. Cathedral chemistry—conserving the stained glass. Chem. Br. 10:89-91, 1974.

Niesel, K, The weathering of building materials in atmospheres containing sulfur oxides~a
literature discussion. Fortsch. Mineral. 57:68-124, 1979.

Nriagu, J.O. Deteriorative effects of sulfur pollution on materials. In: Sulfur in the
Environment. Part II: Ecological Impacts. John Wiley & Sons, Inc., lew York, NY, 1978.
pp. 1-59.

Overseas Public Nuisance Study Mission. Present Status of Public Nuisance in Electric Power
Industry of Japan. Central Power Council, Tokyo, Japan, September, 1965.

Oyama, M., A. Moriyama, H. Miyazato, and H. Sakugawa. Examples on the acid-smut fallout.
Okinawa-ken Kogai Eisei Kenkyusho-Ho 8:32-33, 1974.

Parker A. The destructive effects of air pollution on materials. In: Proceedings of the
22nd Annual Conference, National Smoke Abatement Society, Bournemouth, England, September
28, 1955. National Smoke Abatement Society, Brighton, England, 1955. pp. 1-15.

Potter, P. Chimney corrosion and acid smut emission. N.Z. J. Dairy Sci, Techno!. 6:11-14,
1971.

Prenderleith, H.J. The Preservation of Leather Bookbindings. The British Museum, London,
England, 1957.

Ridker, Ronald G. Economic Costs of Air Pollution. Praeger, 1967.

Riederer, J. Pollution damage to works of art. In: New Concepts in Air Pollution Research.
Experientia Suppl. 20:73-85, 1974.

Robbins, R.C. Inquiry into the Economic Effects of Air Pollution on Electrical Contacts.
Stanford Research Institute, Menlo Park, CA, April 1970.
10-80
-------
Rowden, E. Acid corrosion and smut emission by kiln exhaust gases. Trans. Br, Ceram, Soc.
68:227-229, 1968.

Russell, C. A. How environmental pollutants diminish contact reliability. Insul. Circuits
22:43-44, 1976.

Salmon, R. L. Systems Analysis of the Effects of Air Pollution on Materials. APTD-0943,, U.S.
Department of Health, Education and Welfare, National Air Pollution Control
Administration, Raleigh, NC, January 1970.

Salvin, V.S. Effect of air pollutants on dyed fabrics. J. Air Pollut. Control Assoc.
13:416-422, 1963.

Salvin, V.S. Survey and economic assessment of the effects of air pollution on textile fibers
and dyes. Final Report, Contract no. PH-22-68-2, U.S. Department of Health, Education
and Welfare. National Air Pollution Control Administration. Raleigh, NC, June 1970.

Sanyal, B. , and G.K. Singhania. Atmospheric corrosion of metals: Part I. J. Sci. Ind. Res.
Sect. B. 15:248-255, 1956.

Satake, J., and T. Moroishi. Various factors affecting atmospheric corrosion" of steels. I_n:
Proceedings of the Fifth International Congress on Metal Corrosion, Tokyo, Japan, May
1972. National Association of Corrosion Engineers, Houston, TX., 1974. pp. 744-749.

Schwarz, H. On the effect on magnetite on atmospheric rust and on rust under a coat of paint.
Werkst. Korros. 23:648-663, 1972.

Schwenk, W., and H. Ternes. Rate of corrosion of plain carbon and low-alloy structural
steels. Stahl Eisen 88:318-321, 1968.

Sengupta, M., and A.A. de Gast. Environmental deterioration and evaluation for dimension
stone. Can. Min. Metal!. Bull. 65:54-58, 1972.

Sereda, P.L. Weather factors affecting corrosion of metals. In: Corrosion in Natural
Environments. ASTM Special Technical Publication 558, American Society for Testing and
Materials, Philadelphia, PA, 1974. pp. 7-22.

Simpson, J.W., and P.J. Horrobin, eds. The Weathering and Performance of Building Materials.
Wiley-Interscience, New York, NY, 1970. p. 286.

Sleater, G.A. Stone Preservatives: Methods of Laboratory Testing and Preliminary Performance
Criteria. NBS Technical Note 941, U.S. Department of Commerce, National Bureau of
Standards, Washington, DC, 1977.

Spedding, D.J., P.P. Rowlands, and J.E. Taylor. Sorption of sulphur dioxide by indoor
surfaces: leather. J. Appl. Chem. Biotechnol. 21:68-70, 1971.

Spence, J.W., F.H. Haynie, and J.B. Upham. Effects of gaseous pollutants on paints: A
chamber study. J. Paint Techno!. 47:57-63, 1975.

Spence, J.W., and F.H. Haynie. Design of a laboratory experiment to identify the effects of
environmental pollutants on materials. In; Corrosion in Natural Environments. STP 558.
American Society for Testing and Materials, Philadelphia, PA, 1974. pp. 279-291.

Spence, J.W., and F.H. Haynie. Paint Technology and Air Pollution: A Survey and Economic
Assessment. AP-103, U.S. Environmental Protection Agency, Research Triangle Park, NC,
February 1972.
10-81
-------
SRI International. An Estimate of the Nonhealth Benefits of Meeting the Secondary National
Ambient Air Quality Standards. Report prepared for the National Commission on Air
Quality. Menlo Park, CA, 1981.

Stankunas, A.R., D.F. Unites, and E.F. McCarthy. Air pollution damage to manmade materials.
Presented at the 74th Annual Meeting, Air Pollution Control Association, Philadelphia,
Pennsylvania, June 21-26, 1981. Paper no. 81-65.3.

Stanners, J.F. Use of environmental data in atmospheric corrosion studies. Br. Corros. J.
5:117-121, 1970.

Sullivan, T.A. and W.C. McBee. Development and Testing of Superior Sulfur Concretes. RI 8/60,
U.S. Department of the Interior, Bureau of Mines, Washington, D.C. 1976.

Sulphur Institute. Sulphur concrete. Sulphur Res. Dev. 2:2-14, 1979.

Svoboda, M., B. Knapek, and H, Kilcoua. The effect of pigments on the penetration of sulfur
dioxide into coatings. J. Oil Colour Chem. Assoc. 56:172174, 1973.

Sydberger, T. Influence of Sulphur Pollution on the Atmospheric Corrosion of Steel.
University o"f Gothenburg, Department of Inorganic Chemistry, Gothenburg, Sweden, 1976.
pp. 3-8.

Sydberger, T., and R. Ericsson. Laboratory testing of the atmospheric corrosion of steel.
Werkst. Korros. 28:154-158, 1976.

Sydberger, T., and N.G. Vannerberg. The influence of the relative humidity and corrosion
products on the adsorption of sulfur dioxide on metal surfaces. Corros. Sci. 12:775-784,
1972.

Tomashov, N.D. The Science of Corrosion. The Macmillan Co., New York, NY, 1966.

Tombach, I, Measurement of local climatological and air pollution factors affecting stone
decay. Presented at the Conservation of Historic Stone Buildings and Monuments
Conference, National Materials Advisory Board, National Research Council, National
Academy of Sciences, Washington, DC, February 2-4, 1981.

Upham, J.B. Atmospheric corrosion studies in two metropolitan areas. J. Air Pollut. Control
Assoc. 17:398-402, 1967.

Uphara, O.B., F.H. Haynie, and J.W. Spence. Fading of Selected Drapery Fabrics by Air
Pollutants. U.S. Environmental Protection Agency, Chemistry and Physics Laboratory,
Research Triangle Park, NC, 1975.

Upham, J.B., and V.S. Salvin. Effects of Air Pollutants on Textile Fibers and Dyes.
EPA-650/3-74-008, U.S. Environmental Protection Agency, Research Triangle Park, NC,
February 1975.

Verdu. Effect of air pollutants on aging of plastic materials. Trib, CEBEDEAU 27:360-370,
1974.

Vero, L.B., and M.M. Sila. Isolation of various sulphur-oxidizing bacteria from stone
monuments. In: The Conservation of Stone, Proceedings of the International Symposium,
Bologna, Italy, June 19-21, 1976.

Waddell, T.E. The Economic Damages of Air Pollution. EPA-600/5-74-012, U.S. Environmental
Protection Agency, Washington, DC, May 1974.
10-82
-------
Walsh, M., A. Black, A. Morgan, and G. Crashaw. Sorption of S0? by indoor surfaces including
carpets, wallpaper, and paint. Atmos. Environ. 11:1107-1111, 1977.

Watson, W.D., Jr., and J.A. Jaksch. Household cleaning costs and air pollution. Presented at
the 71st Annual Meeting, Air Pollution Control Association, Houston, Texas, June 25-30,
1978. Paper no. 78-52.3.

Watson, W.D., and J.A. Jaksch. Air pollution: household soiling and consumer welfare losses,
1982 (Paper to be published in the Journal of Environmental Economics and Management,
Volume 9, No. 3, September 1982).

Winkler, E.M. Stone: Properties, Durability in Man's Environment. 2nd rev. ed. ,
Springer-Verlag, New York, NY, 1975.

Winkler, E.M. Important agents of weathering for building and monumental stone. Eng. Geol.
(Amsterdam) 1:381-400, 1966.

Yocom, J.E. and Grappone. Effects of Power Plant Emissions on Materials. EPRI EC-139,
Electric Power Research Institute, Palo Alto, CA, July 1976.

Yocom, J.E. Air pollution damage to buildings on the Acropolis. J. Air Pollut. Control
Assoc. 29:333-338, 1979.

Yocom, J.E., and R.O. McCaldin, Effects of air pollution on materials and the economy. In:
Air Pollution, Vol. I, A. Stern, ed. Academic Press, New York, NY, 1968, 617-654.

Yocom, J.E., and J.B. Upham. Effects on economic materials and structures. In: Air
Pollution. 3rd ed. vol. II: The Effects of Air Pollution. A. C. Stern, ed., Academic
Press, Inc., New York, NY, 1977. pp. 65-116.

Yoshihara, K. , T.C. Huang, H. Ebihora, and H. Shibata. Chemical effects of the 35Cl(n,p)35 S
reaction. Radiochim. Acta 3:185-191, 1964.

Zeronian, S.H. Reactions of cellulosic fabrics to air contaminated with sulfur dioxide.
Text. Res. J. 40:695-698, 1970.

Zeronian, S.'H. , K.W. Alger, and S.T. Omaye. Reactions of fabrics made from synthetic fibers
to air contaminated with nitrogen dioxide, ozone, or sulfur dioxide. In: Proceedings of
the Second International Clean Air Congress, International Union of Air Pollution,
Washington, DC, December 6-11, 1970. H. M. Englund and W. T. Beery, eds., Academic
Press, New York, NY, 1971. pp. 468-476.
10-83
-------
-------
11. RESPIRATORY TRACT DEPOSITION AND FATE
OF INHALED AEROSOLS AND SULFUR DIOXIDE

11.1 INTRODUCTION
f >» **
11.1.1 General Considerations
The respiratory system is the major route of human exposure to airborne suspensions of
particles (aerosols) and sulfur dioxide (SO^) gas. During inhalation (and exhalation), a por-
tion of the inhaled aerosol and gas may be deposited by contact with airway surfaces, or it
may be transferred to unexhaled air; the remainder is exhaled. The portion transferred to un-
exhaled air may either be deposited by contact with airway surfaces or exhaled later. These
phenomena are complicated by interactions that may occur among the particles, the S0? gas,
other gases such as endogenous ammonia, and the water vapor present in the airways.
In inhalation toxicology, specific terminology is applied to these processes. The term
deposition refers specifically to the removal of inhaled particles or gases by the respiratory
tract and to the initial regional pattern of these deposited materials. The term clearance
refers to the subsequent trans!ocation (movement of material within the lung or to other
organs), transformation, and removal of deposited substances from the respiratory tract or
from the body. It can also refer to the removal of reaction products formed from S0? or par-
ticles. The temporal pattern of uncleared deposited particulate materials or gases and re-
action products is called retention.
The mechanisms involved in the deposition of inhaled aerosols and gases are affected by
physical and chemical properties, including aerosol particle size distribution, density, shape,
surface area, electrostatic charge, hygroscopicity or deliquescence, chemical composition, gas
diffusivity and solubility, and related reactions. The geometry of the respiratory airways
from nose and mouth to the lung parenchyma also influences aerosol deposition; the important
morphological parameters include the diameters, lengths, inclinations to vertical, and branch-
ing angles of airway segments. Physiological factors that affect deposition include breathing
patterns, airflow dynamics in the respiratory tract, and variations of relative humidity and
temperature within the airways. Clearance from the respiratory tract depends on many factors,
including site of deposition, chemical composition and properties of the deposited particles,
reaction products, mucociliary transport in the tracheobronchial tree, macrophage phagocytosis,
and pulmonary lymph and blood flow. An understanding of the regional deposition and clearance
of particles and S0? is essential to the interpretation of the results of health effects
studies described in Chapters 12-14.
Translocation of sulfur compounds or other materials from the lung to other organs is
also important, since the lung can be the portal of entry for toxic agents that affect other
organs of the body. Hence, multicompartment models of clearance from the respiratory tract to
other organs can provide predictive information about the potential for injury of those other
organs. Mathematical representations of lung retention and translocation require data on the
various factors that affect deposition and clearance.

11-1
-------
Since many conclusions concerning the deposition, clearance, and health impact of inhaled
aerosols and SOp are based on data obtained from animal experiments, care must be taken to
identify physiological and anatomical differences between human beings and animals which may
influence these phenomena. The following discussion will emphasize the regional deposition
and clearance that occur in human airways, but selected comparisons will be made with
other mammalian species to clarify differences that may affect health-impact analyses of
experimental data.
11.1.2 Aerosol and Sulfur Dioxide Characteristics
An aerosol may be defined as a relatively stable suspension of liquid or solid particles
in a gaseous medium (see Chapter 2 for a detailed discussion of the physicochemical properties
of aerosols and S02). Airborne particulate materials in the environment are aerosols with a
variety of physical and chemical properties. In particular, a given aerosol may include par-
ticles with a wide spectrum of physical sizes, even if all the particles have similar chemical
composition; the concentration of toxic components in particles may be different for different
sized particles (Natusch et al., 1974); or morphologically identical particles may have
totally different chemical compositions (Pawley and Fisher, 1977). Common assumptions that
particles in a given aerosol have a relatively homogeneous chemical composition, toxic poten-
tial, and physical density may be seriously misleading, especially when particles are found
in combination with SOp gas.
The relevant physical and chemical properties of aerosols and gases must be characterized
appropriately to evaluate the effect of their inhalation on health. These properties then can
provide predictive information concerning regional respiratory tract deposition and other
important dosimetric factors that need to be considered if biological responses described in
Chapters 12-14 are to be understood adequately.
If particles in an aerosol are smooth and spherical or nearly spherical, their physical
sizes can be conveniently described in terms of their respective geometric diameters. Even
unagglomerated aerosols of solids, however, rarely contain smooth, spherical particles. Vari-
ous conventions for describing physical diameters have been based on available methods of
observing and measuring particle size. For example, the size of a particle may be described
in terms of its projected area diameter (D ), defined as the diameter of a circle with an area
equal to the apparent cross-sectional area of the particle when lying on a collection surface
and viewed with an optical or electron microscope. Other conventions for describing physical
size are based on measurements of scattered light, surface area, electrical mobility, diffu-
sional mobility, or other physical or chemical phenomena (Mercer, 1973; Stockham and
Fochtman, 1979).
The aerodynamic properties of aerosol particles depend on a variety of physical para-
meters including size, shape, and physical density. Two important aerodynamic properties of
aerosol particles are the inertia! properties, which are most important for particles larger
than 0.5 urn in diameter and are related to the settling speed in air under the influence of
the" earth's gravity, and the diffusional properties, which are most important for particles

11-2
-------
smaller than 0.5 jjm in diameter and are related to the diffusion coefficient (Fuchs, 1964)
(see Section 11.2.1). When particles are inhaled, their aerodynamic properties, combined with
various anatomical and breathing characteristics, determine their fractional deposition in
various regions in the respiratory tract.
To avoid the complications associated with the effects of particle shape, size, and phys-
ical density on the inertia! properties of inhaled airborne particles, "aerodynamic diameters"
have been defined and used to describe particles with common inertia! properties with the same
"aerodynamic diameter." The aerodynamic diameter most generally used is the aerodynamic equi-
valent diameter (D ), defined by Hatch and Gross (1964) as "the diameter of a unit density
sphere having the same settling speed (under gravity) as the particle in question of whatever
shape and density." Raabe (1976) recommended the use of an aerodynamic resistance diameter
(Dgr), defined more directly with terms used in physics to describe the inertia! properties of
a particle. The difference between these two diameters is only 0.08 pm or less over all sizes
under normal conditions at sea level. Hence, the term aerodynamic diameter can be used to
refer to either or both of these two definitions.
Environmental aerosols have size distributions that are more complicated, reflecting the
production of particles by atmospheric processes, emission sources or other anthropogenic acti-
vities, and the particle dynamics. They may have several modes (Whitby, 1978). Photochemical
reactions and certain combustion processes create small particles that are generally smaller
than 0.1 (jm (the nuclei mode), whereas other combustion, condensation, and mechanical particle
generation processes yield larger particles. Another mode, between 0.1 and 2 urn, is known as
the accummulation mode and includes primary emissions plus aggregates and droplets formed by
coagulation of the primary nuclei mode particles and the materials that condense on them from
the vapor phase. Coarse particles, larger than about 2 urn, are formed primarily by mechanical
processes. The particle size distribution within each of the three modes (nuclei, accumula-
tion, and coarse) is generally lognormal, as defined below.
Since not all particles in an aerosol are of the same physical or aerodynamic size, the
distribution of sizes must be described. If either the physical diameter (D) or the aerody-
namic diameter is used to characterize particles, the distribution of particle sizes in a
mixed aerosol is most conveniently described as a probability density function. One such
generally useful function, the lognormal function, involves the geometric mean size (or
median) and the geometric standard deviation (o ) and refers to a normal distribution with
respect to the logarithm of particle diameter. Hence, if the particle number is being con-
sidered, the particle size may be reported as the count median (physical) diameter (CMD) and
a . Half of the number of particles in an aerosol has physical sizes less than the CMD and
half has larger. Since the mass of a material is usually more relevant to its potential
toxicity, the mass median (geometrical) diameter (MMD) or mass median aerodynamic diameter'
(MMAD) and o is usually preferred in describing aerosols in inhalation toxicology research.
Half the mass of particles in an aerosol is associated with particles smaller than the MMD and
11-3
-------
half with larger particles. Likewise, half the mass of particles Is associated with particles
whose aerodynamic diameters are smaller than the MMAD and half with particles with larger aero-
dynamic diameters. If an aerosol is radioactive or radiolabeled, mass measurements may be re-
placed by activity measurements. Interrelationships among these various ways to express the
diameter of the aerosol were examined for the lognormal distribution by Raabe (1971).
In addition to particle characteristics, conditions of the gas medium influence the pro-
perties of aerosol dispersions. Such environmental conditions as relative humidity, tempera-
ture, barometric pressure, and fluid flow conditions (e.g., wind velocity or state of turbu-
lence) affect the aerodynamics of aerosol particles.
The concentration of environmental aerosols or gases generally does not affect inhalation
3 3
deposition and particle dynamics. The mass concentration (mg/m or pg/m ) or concentration of
a specific potentially toxic species (mg of constituent/m ) provides information needed to
calculate inhalation exposure levels. For SO-, the concentration may be expressed in parts
3
per million (ppm) by volume or in mass concentrations (mg/m ); each I ppm of SO, equals 2.62
33
mg/m (2620 Mg/m ) at an air temperature of 25°C.
Sulfur dioxide gas is a rapidly diffusing reactive gas that is readily soluble in water
and body fluids (Aharonson, 1976). This property is responsible for the extensive removal of
SO, in the extrathoracic region and in the upper generations of the tracheobronchial tree.
Extraction of S0? during nose breathing is significantly greater than during mouth "breathing,
and over a 4- to 6-hour exposure to high levels of S0?, no saturation effect for absorption
can be seen (see Section 11.2.4). Through normal and catalyst-mediated oxidation processes in
air, SOp gas is slowly oxidized to sulfite (SO,) that rapidly hydrolyzes to form sulfuric acid
(HgSO,), leading to sulfate salts. Since ammonia (NH») is formed in natural biological pro-
cesses including endogenously in the airways, (NH.)pSO. and NH.HSO, are important products of
HpSO. neutralization.
11.1.3 The Respi ratory Tract
The respiratory tract (Figure 11-1) includes the passages of the nose, mouth, nasal phar-
ynx, oral pharynx, epiglottis, larynx, trachea, bronchi, bronchioles, and small ducts and
alveoli of the pulmonary acini. With respect to respiratory tract deposition and clearance of
inhaled aerosols, three regions can be considered: (1) extrathoracic (ET), the airways extend-
ing from the nares down to the epiglottis and larynx at the entrance to the trachea (the mouth
is included in this region during mouth breathing); (2) tracheobronchial (TB), the primary con-
ducting airways of the lung from the trachea to the terminal bronchioles (i.e., that portion
of the lung respiratory tract having a ciliated epithelium); and (3) pulmonary (P), the
parenchyma! airspaces of the lung, including the respiratory bronchioles, alveolar ducts,
alveolar sacs, atria, and alveoli (i.e., the gas-exchange region). The extrathoracic region,
as defined above, corresponds exactly to the nasopharynx, as defined by the International
Commission on Radiological Protection (ICRP) Task Group on Lung Dynamics (Morrow et al.,
1966).
11-4
-------
LEFT WALL OF NASAL CAVITY
AND TURBINATES
ORAL CAVITY
RIGHT MAIN BRONCHUS
UPPER LOBE BRONCHUS
MEDIAL LOBE
BRONCHUS
1
LOWER LOBE
BRONCHUS
LUNG PARENCHYMA
AND. ALVEOLI
UPPER LOBE BRONCHUS
LOWER LOBE BRONCHUS
Figure 11-1. Features of the respiratory tract of man used in the description of the deposition of
inhaled particles and gases with insert showing parts of a silicon rubber cast of a human lung show-
ing some separated bronchioles to 3 mm'diameter, some bronchioles from 3 mm diameter to term-
inal bronchioles, and some separated respiratory acinus bundles.
Source: Adapted from Hatch and Gross (1964) and Raabe (1979).
11-5
-------
The nose is a complex structure of cartilage and muscle supported by bone and lined with
mucosa (Holmes et al., 1950). The vestibule of the nares is unciliated but contains a low-
resistance filter consisting of small hairs. The nasal volume is separated into two cavities
by a 2~ to 7-mm thick septum. The inner nasal fossae and turbinates are ciliated, with mucus
flow in the direction of the pharynx. The turbinates are shelf-like projections of bone
covered by ciliated mucous membranes with a high surface-to-volume ratio that facilitate hurai-
dification of the incoming air. The larynx consists of two pairs of mucosal folds that narrow
the airway.
The trachea, an elastic tube supported by 16 to 20 cartilagenous rings that circle about
three-fourths of its circumference, is the first and largest of a series of branching airway
ducts (Tenney and Bartlett, 1967). The left and right lungs are entered by the two major
bronchi that branch off of the trachea (see Figure 11-1). The left lung consists of two
clearly separated upper and lower lobes; the right lung consists of the upper, middle, and
lower lobes. The conductive airways in each lobe of the lung consist of up to 18 to 20
dichotomous branches from the bronchi to the terminal bronchiole (Pump, 1964; Raabe et al.,
1976).
The pulmonary gas-exchange region of the lung begins with the partially alveolated res-
piratory bronchioles. Pulmonary branching proceeds through a few levels of respiratory bron-
chioles to completely alveolated ducts (Smith and Boyden, 1949; Whimster, 1970; Krahl, 1963)
and alveolar sacs (Tenney and Remmers, 1963; Pattle, 1961b; Machlin, 1950; Fraser and Pare,
1971). Alveoli are thin-walled polyhedral air pouches that cluster about the acinus through
connections with respiratory bronchioles, alveolar ducts, or alveolar sacs.
The airway spaces in the pulmonary region are coated with a complex aqueous liquid con-
taining several biochemically specialized substances (Green, 1974; Blank et al., 1969; Balis
et al., 1971; Pattle, 19615; Kott et al., 1974; Henderson et al., 1975; Kanapilly, 1977). An
understanding of the chemical composition and dynamic nature of the acellular layer at the air-
alveolar surface is needed to comprehend the general behavior of material deposited in the
pulmonary region. This acellular layer consists of a surfactant film < 0.01 (jm thick and a
hypophase about 0.1 to 0.2 |Jm thick (Clements and Tierney, 1965). The thickness of the com-
plex aqueous lining layer is not uniform because pools of surfactant even out pits, crevices,
folds and small surface irregularities, and thereby help to impart smoothness to the alveolar
surface (Gil et al., 1979). A mixture of phospholipids and neutral lipids is contained in the
surfactant film (Scarpelli, 1968; Pfleger and Thomas, 1971,; Pruitt et al., 1971; Reifenath,
1973). The major phospholipid is dipalmitoyl lecithin, and the major neutral lipids are
cholesterol and its esters. Protein content in lung surfactant is less than 20 percent by
weight (Pruitt et al., 1971; Klass, 1973). The composition of the hypophase is not well under-
stood, with lung surfactant materials, mucopolysaccharides, lipoproteins, and possibly serum
proteins such as albumin likely to be present (Scarpelli, 1968; Reifenath, 1973; Tuttle and
Westerberg, 1974). The pH of alveolar fluid may be similar to that of blood fluid. Various
11-6
-------
factors favoring and opposing transudation of fluid across the air-blood barrier result in the
cyclic movement of fluid in and out of alveoli, thereby helping to maintain the very thin layer
of alveolar fluid (Kanapilly, 1977). The concentrations of chelating and precipitating agents
in the alveolar fluid influence the retention and transport of particles deposited in the
pulmonary region; however, the concentrations of chelating agents are not sufficiently high to
prevent the transformation of polyvalent cations into an insoluble form (Kanapilly, 1977).
The parenchyma of the pulmonary region includes several types of tissue, circulating
blood, lymphatic drainage pathways, and lymph nodes. In humans, the weight of the lung, in-
cluding circulating blood, is about 1.4 percent of the total body weight. The weight of lung
blood is equal to about 0.7 percent of total body weight (10 percent of total blood volume)
(International Commission of Radiological Protection, 1975). Because a large portion of lung
o
is occupied by air, the average physical density of the parenchyma is about 0.26 g/cm (Fowler
and Young, 1959).
Models of the airways, which simplify the complex array of branching and dimensions into
workable mathematical functions, are useful in comparing theoretical predictions of deposition
with experimentally-obtained deposition data, thereby leading to more refined models and in-
creasing our understanding of the processes that affect respiratory tract deposition. An
early idealized model of the airways of the human lung was developed by Findeisen (1935) for
estimating the deposition of inhaled particles. Findeisen's model assumed branching symmetry
within the lung, with each generation consisting of airways of identical size. Landahl (1950),
Davies (1961), Weibel (1963), and Horsfield and Gumming (1968) proposed other models based on
a symmetry assumption. Asymmetric models that more closely approximate the human lung were
developed by Weibel (1963), Horsfield et al. (1971), and Horsfield and Gumming (1968). Yeh
and Schum (1980) proposed a typical pathway lung model and made particle deposition calcula-
tions for each lobe of the lungs. Although particle deposition models"currently available
ignore the dynamic nature of the airways, future models should consider this aspect.
11.1.4 Respiration and Other Factors
Both the humidity and temperature of inhaled aerosols and gases, as well as the subse-
quent changes that occur as the aerosol-gas mixture passes through various parts of the air-
ways, influence the inhalation deposition of airborne particles. Deposition of hygroscopic
aerosols will depend in part on the relative humidity in the airways, since the growth of such
particles will directly affect both the site and extent of inhalation deposition (see Section
11.2.2).
The complex anatomical structure of the nose is well suited for humidification, regula-
tion of temperature, and removal of many particles and gases. The relative humidity of in-
haled air probably reaches near saturation in the nose (Verzar et al., 1953). Since the human
nose is a short passageway, tranquil diffusion alone cannot account for rapid humidification.
Rather, convective mixing must play a role, suggesting a mechanism for enhancing S02 collec-
tion in the nose. The temperature of the inhaled air may not reach body temperature until
relatively deep in the lung. Deal et al. (1979a,b,c) measured retrocardiac and retrotracheal

11-7
-------
temperatures under different ambient temperatures and found airway cooling associated with
breathing cool air. Raabe et al. (1976) found that the temperature of the air at major bronchi
in a nose-breathing dog averaged 35°C, 4°C less than the body temperature.
The air-deflecting channels of the anterior nares cause impaction of large airborne par-
ticles and create turbulent airflow conditions. As the cross-sectional area expands beyond
the entrance, flow separation occurs resulting in turbulence and eddies, which continue as the
air traverses the passages around the turbinates. Proctor and Swift (1971) studied the flow
of water through a clear plastic model of the walls of the nasal passages and constructed
charts of the direction and linear velocity of airflow from this model. With a steady inspi-
ratory flow of 0.4 I/sec, the investigators found that the linear inspiratory velocity at the
nasal entrance reached at least 4.5 to 5 m/s and at most 10 to 12 m/s, values that are signi-
ficantly greater than the 2 m/s peak linear velocity in the tracheobronchial tree during quiet
breathing.
The caliber of the trachea and major bronchi and their cross-sectional geometry is about
15 percent larger during inspiration than during expiration (Marshall and Holden, 1963; Fraser
and Pare, 1971; Raabe et al., 1976), although the caliber of the smaller conductive bronchioles
may be up to 40 percent greater during inspiration (Marshall and Holden, 1963; Hughes et al.,
1972). Bronchial caliber correlates with body size (Thurlbeck and Raines, 1975).
Schroter and Sudlow (1969) studied a wide variety of flow patterns and flowrates in large-
scale symmetrical models of typical tracheobronchial tree junctions. For both inspiration and
expiration and irrespective of entry profile form, they observed secondary flows at all flow-
rates in their single bifurcation model. When a second bifurcation was added a short distance
downstream of the first, the entering flow profile was found to influence the resulting flow
patterns. Also, different results were obtained depending on the plane in which the second
bifurcation was located relative to the first bifurcation.
<<
Olson et al. (1973) studied convective airflow patterns in cast replicas of the human
respiratory tract during steady inspiration. They showed that the effect of the larynx is
such that flow patterns typical of smooth bifurcating tubes do not occur until the lobar
bronchi are reached. Small eddies were observed as far down as the sublobar bronchi with
flows of 200 ml/s in the trachea. In humans, the glottis of the larynx acts as a variable
orifice, since the position of the vocal cords changes. During inspiration, a jet of turbu-
lent air enters the trachea and is directed against its ventral wall, imparting additional tur-
bulence over that associated with the corrugated walls and length of the trachea.
In the tracheobronchial tree, with its many branches, changes in caliber, and irregular
wall surfaces, establishing exactly where flow is laminar, turbulent, or transitional is diffi-
cult. Viscous forces predominate in laminar flow, and streamlines persist for great distances;
with turbulent flow, there is rapid and random mixing downstream. As the flowrate increases,
unsteadiness develops and separation of the streamlines from the wall can occur, leading to the
11-8
-------
formation of local eddies. This type of flow is termed transitional. The Reynolds number,
the ratio of inertia! to viscous forces, is useful in describing whether flow is laminar or
turbulent. In smooth-walled tubes, values between approximately 2000 and 4000 are ascribed to
transitional flow, with smaller Reynolds numbers reflecting laminar flow and larger orfes tur-
bulent flow. Fully developed laminar flow probably only occurs in the very small airways;
flow is transitional in most of the tracheobronchial tree, although true turbulence may occur
in the trachea, especially during exercise when flow velocities are high (West, 1977).
Turbulence will gradually decay in any branch in which the Reynolds number is less than
3000 (Owen, 1969). Decays of 15, 16, and 10 percent are predicted to occur in the first three
generations of bifurcation, respectively, using the theory of Batchelor (1953) for the change
in turbulent energy at regions of rapid flow contraction. Although these decay calculations
neglect the possible effects of the strong secondary flows generated at the bifurcation, their
validity is supported by the data of Pedley et al.(1971) which show that the boundary layer
remains laminar in the daughter-tube for Reynolds numbers in the parent-tube up to at least
10,000. Hence, the turbulent eddies are localized in the center.
Flow oscillations in the segmental bronchi attributed to the beating of the heart are
only detectable during breathholding or during pauses between inspiration and expiration
(West, 1961). A peak oscillatory flowrate of 0.5 1/min was measured, which is about 20 per-
cent of the peak flowrate in the segmental bronchi during quiet breathing. Gas mixing is
improved by these oscillations.
Gas flow dynamics may be expected to be turbulent within the upper airways of humans and
dogs but laminar everywhere in the airways of small rodents (Dekker, 1961; Fry, 1968; Schroter
and Sudlow, 1969; Olson et a!., 1973; Martin and Jacob!, 1972; West, 1961). The larynx intro-
duces an important airflow disturbance that can influence trachea! deposition (Bartlett et a!.,
1973; Schlesinger and Lippmann, 1976). In the smaller human bronchi and bronchioles where
fluid flow is relatively tranquil, laminar flow prevails; but branching patterns, filling
patterns (Grant et a!., 1974), flow reversals with varying velocity profiles, and swirling
complicate a description of flow in the small airways (Silverman and Billings, 1961; Cinkotai,
1974). Because actual flow in the respiratory airways is difficult to describe, simplifying
assumptions, such as parabolic laminar or uniform velocity profiles, are usually incorporated
into analytic descriptions.
Inspiratory flowrate and depth of inhalation influence the deposition of inhaled par-
ticles. The air inspired in one breath is the tidal volume (TV). The average inspiratory
flowrate (Q) and TV (Bake et al., 1974; Clement et al., 1973) affect both inertia! and diffu-
sional deposition processes (Altshuler et al., 1967). The total air remaining in the lungs at
the end of normal expiration affects the relative mixing of inhaled particles and, when com-
pared with total lung capacity, is indicative of the extent of aerosol penetration into the
lung (West, 1974; Luft, 1958). Guyton (1947a,b) and Stahl (1967) developed interspecies rela-
tionships describing respiratory volumes and patterns.
11-9
-------
The inspiratory capacity, the maximum volume of air that can be inhaled after a given
normal expiration, is contrasted to the vital capacity, which is the maximum volume of air
that can be expelled from the lungs with effort after maximum forced inspiration. Air that
remains in the conductive airways (from nose or mouth to terminal bronchioles) at the end of
expiration is considered to occupy the anatomical dead space, since the conductive airways are
not involved in gas exchange.
Representative values for normal human respiratory parameters, which can be used for depo-
sition and dosimetric predictions, are available from various sources (Zenz, 1975; Higgs et
al., 1967; American Heart Association, 1973; Jones et a!., 1975; Intemountain Thoracic
Society, 1975; -International Commission on Radiological Protection, 1975). Considerable vari-
ability in respiratory parameters may occur among individuals in the population, particularly
when healthy adults are contrasted with children, the aged, and ill individuals. Average TV
has a reasonably fixed relationship with body weight of 7 to 10 ml/kg from birth to adulthood
(Doershuk et al., 1970, 1975). The gas-exchange area increases proportionally with age and
more or less with height, but not with body surface area. Average values for the gas-exchange
o
area are 6.5, 32, and 75 m at 3 mo, 8 y, and adulthood, respectively (Dunnill, 1962). Respi-
ratory frequency decreases from about 35 breaths per minute (BPM) at birth to 12 to 16 BPM with
normal respiration in adulthood (Polgar and Weng, 1979).
In some instances, the total and regional deposition data presented in Section 11.2 exhi-
bit considerable scatter. Some of this variability might be expected given the range of
breathing frequencies, TV's, and average inspiratory flowrates used in the various deposition
experiments. Deposition studies involving aerosol persistence during breath-holding led Lapp
and coworkers (1975) to conclude that marked differences exist in airway geometry among sub-
jects with similar heights and lung volumes. An interlaboratory comparison study of lung
deposition data by Heyder et al. (1978), besides identifying possible sources of errors con-
nected with the experimental technique, identified different deposition data in the subjects.
Further data on intersubject variability, including data on variability of regional deposi-
tion, were presented by Stahlhofen et al. (1981). Yu and coworkers (1979) used Monte Carlo
techniques to determine the total and regional deposition of inhaled particles in a population
of human lungs by taking into account variability in airway dimensions. Their results for
particle sizes ranging from 0.1 to 8 ym D suggest that observed subject deposition vari-
36
ability is caused primarily by differences in airway dimensions. When the total respiratory
tract deposition of particles between 0.3 and 1.5 ym 0 is studied, expressing the data as
B6
a function of the ratio of the relative expiratory reserve volume to the normal expiratory re-
serve volume greatly reduces intersubject variability (Tarroni et al., 1980).
The vast majority of studies on the deposition of particles in humans has been conducted
in young healthy adults. Consequently, there is a paucity of data on deposition (and clear-
ance) in other subpopulations, such as children, asthmatics, chronic bronchitics, etc. Signi-
ficant pathologic changes in airways and parenchyma can markedly alter the deposition of par-
ticles. For example, Lippmann et al. (1971) found substantially increased bronchial deposition

11-10
-------
in chronic bronchitic and asthmatic subjects. These increases may vary with different phases
of the disease (Goldberg and Lourenco, 1973). Tracheobronchial deposition appears to be en-
hanced at the expense of pulmonary deposition in most abnormal states. For example, the depo-
sition of 2 pro particles in patients with bronchiectasis is frequently more central than that
in normal subjects (Lourenco et al. , 1972). Partial or complete airway obstruction in bronchi-
tis, lung cancer, emphysema, fibrosis, and atelectasis may decrease or eliminate deposition of
particles in some,regions of the lungs (Taplin et al., 1970). The numerous and complex mech-
anisms responsible for alterations in the pattern of deposition in various disease states need
to be studied.
Currently available human deposition data have been collected from volunteers inhaling
aerosols through either mouthpieces or nose masks. Differences in mass burden of particles
and gases between these controlled inhalations and normal, spontaneous mouth breathing or nose
breathing are possible; however, Heyder et al. (1980a) found the same dependence on a deposi-
tion parameter incorporating minute volume, breathing frequency, and particle size for both
spontaneous and artificially controlled breathing patterns in three subjects. Any differences
are more likely to be seen at minute ventilations corresponding to heavy or maximal exercise,
since oral resistance to respiratory airflow is probably lowered when a subject breathes
through a mouthpiece. Studies in which the nose of the subject is completely occluded with a
clip do not simulate oronasal breathing because no air passes through the nose and the oral
airway is wider than usual. With partial nasal obstruction or in exercise, most human beings
resort to oronasal breathing. In studies designed to examine the switching point from nasal
to oronasal breathing (Niinimaa et al. , 1980) and to determine the oronasal distribution of
respiratory airflow in 30 healthy adult subjects (14 males, 16 females) (Niinimaa et al.,
1981), 20 of the subjects (67 percent) switched from nasal to orally augmented breathing with
exercise and 4 subjects (13 percent) breathed oronasally even at rest. In contrast, 5 sub-
jects (17 percent, all females) breathed solely through the nose both at rest and throughout
the exercise period, and one subject's nose/mouth breathing pattern was unpredictable and in-
consistent. The percentage of subjects Niinimaa et al. (1980) observed breathing oronasally
even at rest is in good agreement with the results of previous studies by Uddstromer (1940)
and Saibene et al. (1978), although the 17 percent incidence for nose breathing involving all
females is approximately double that found for males in studies by Uddstromer (1940) and
Saibene et al. (1978). The fact that Niinimaa et al. (1980) did not observe any male nose
breathers is probably a reflection of the small number of subjects studied.
With any of the common obstructive forms of nasal pathology such as allergic, viral, or
vasomotor rhinitis or septa! deviation, the proportion of ventilation passing through the
mouth is higher at rest and at any level of exercise. Healthy young adults without nasal
pathology, who breathe predominantly through the nose at rest, shift to breathing through the
nose and mouth when minute ventilation is approximately 35 1/min (Niinimaa et al., 1980;
Saibene et al. , 1978). Niinimaa et al. (1981) found that during exercise requiring a rate of
11-11
-------
ventilation of 35 1/min, 57 percent of the air passed nasally. The onset of oronasal breath-
ing did not show a significant sex difference and was quite consistent individually. Between
Individuals, however, the variation was considerable, as reflected by the large standard devia-
tion of 10.8 1/min (Niinimaa et al. , 1980). Camner and Bakke (1980) studied the proportion
between the air inhaled through the nose and mouth during two common activities, conversation
and reading a newspaper, where the breathing is calm. They found that while conversing casu-
ally all subjects inhaled less air via the nose than while reading with the mouth closed: on
the average less than half the volume. Studies on subjects breathing through a mouthpiece at
rest (minute ventilation of 6 to 8 1/min) provide conservative estimates of the mass burden of
particles, since the total quantity of ventilation passing through the mouth is significantly
less than that which would pass through the mouth in the same subjects breathing freely
through the nose and mouth while performing enough exercise to require a minute ventilation of
35 1/min. A minute ventilation of 35 1/min corresponds to anywhere from light to moderate
exercise according to various sources (Zenz, 1975; Higgs et al., 1967; American Heart Associa-
tion, 1973; Jones et al., 1975; Intermountain Thoracic Society, 1975; International Commission
on Radiological Protection, 1975).
11.1.5 Mechanisms of Particle Deposition
The behavior of inhaled airborne particles in the respiratory airways and their alterna-
tive fate of either deposition or exhalation depend on aerosol mechanics under the given
physiological and anatomical conditions (Yeh et al., 1976; DuBois and Rogers, 1968). This
behavior is usually described in terms of nonreactive stable spherical particles whose physical
properties do not vary during the breathing cycle. Behavior of hygroscopic and deliquescent
particles is more complex.
Figure 11-2 illustrates the five primary physical processes that lead to aerosol particle
contact with the wall of the airways. Contact of particles with moist airway walls results in
attachment and irreversible removal of the particle from the airstream. The contact process
can occur during inspiration or expiration of a single breath or subsequently if a particle
has been transferred to unexhaled lung air (Engel et al. , 1973; Davies, 1972; Altshuler, 1961).
Electrostatic attraction of particles to the walls of the respiratory airways is probably
a minor mechanism of deposition in most circumstances. Pavlik (1967) predicted that light air
ions (which would include some atmospheric aerosol nuclei) would be deposited by electrostatic
attraction in the mouth and throat and suggested that the tonsils were naturally charged for
this purpose. Fraser (1966) found that an average of 1000 electronic units of charge per
aerosol particle, a very large charge that does not normally occur, doubled the inhalation
deposition in experimental animals. Melandri et al. (1977) reported enhanced deposition of
inhaled monodisperse aerosols in humans when the particles were charged. Longley (1960) and
Longley and Berry (1961) found the charge of the subject to have an influence on deposition.
Similar observations have been made in in vitro studies (Chan et al., 1978). The airways are
covered by a relatively conductive electrolytic liquid that probably precludes the buildup of
11-12
-------
-I-1 ELECTROSTATIC

ATTRACTION
BROWNIAN DIFFUSION
GRAVITATIONAL SETTLING
Figure 11-2. Representation of five major mechanisms of deposition of inhaled airborne particles
in the respiratory tract.

Source: Raabe (1979),
11-13
-------
forceful electric fields. Charged particles are therefore collected primarily by image charg-
ing as they near the wall of an airway or by mutual repulsion from a unipolarly charged cloud
with a high concentration of particles (Yu, 1977). The role of this mechanism depends on par-
ticle source, concentration, age, and special electrical phenomena in the environment, as well
as the residence time of the aerosol in the airways. This mechanism can be expected to have a
small role, if any, in the deposition of atmospheric environmental aerosols.
Interception consists of noninertial incidental meeting of a particle and the lining of
the airway and thus depends on the physical size of the particle. It is important primarily
for particles with large aspect ratios, such as long fibrous particles of asbestos (Harris and
Fraser, 1976). Interception may be expected to play a minor role in the inhalation deposition
of most environmental aerosols.
Impaction dominates deposition of particles larger than 3 urn D in the nasopharyngeal
36
and tracheobronchial regions (Rattle, 1961a; Bohning et a!., 1975). In this process, changes
in airstream direction or magnitude of air velocity streamlines or eddy components are not
followed by airborne particles because of their inertia. For example, if air is directed
toward an airway surface (such as a branch carina) but the forward velocity is suddenly re-
duced because of change in flow direction, inertia! momentum may carry larger particles across
*r
the air streamlines and onto the surface of the airway. Impaction at an airway branch has
been likened to impaction at the bend of a tube, providing theoretical estimates of the impac-
tion probability (Johnston and Muir, 1973; Yeh, 1974; Cheng and Wang, 1975) and was studied in
a bifurcating tube model by Johnston and Schroter (1979). Aerodynamic separation of this type
is characterized satisfactorily in terms of the particle aerodynamic diameter. The airflow in
the trachea and major bronchi in humans is turbulent and disturbed by the larynx, so that tur-
bulent impaction plays a role in deposition in these larger airways (Schlesinger and Lippmann,
1976). Breathing patterns involving higher volumetric flowrates would tend to impact smaller
particles. In contrast, the passages of the nose contain smaller airways, and the convective
mixing spaces of the nasal turbinates would be expected to collect some particles as small as
1 or 2 pm D by impaction. Hence, impaction is an important process affecting the inhalation
ae
deposition in the human airways of environmental aerosol particles greater than 1 |jm in aero-
dynamic diameter.
Gravitational settling occurs because of the influence of the earth's gravity on airborne
particles. Deposition of particles by this mechanism can occur in all airways except those
very few that are vertical. The probability of gravitational deposition is usually estimated
with equations describing gravitational settling of particles in an inclined cylindrical tube
under laminar flow conditions (Wang, 1975; Heyder and Gebhart, 1977). This deposition depends
on the residence time and particle concentration distribution in the airway segments, the
angle of incline of the segment with respect to gravity, and the aerodynamic diameter of the
particle. Deposition by gravitational settling is therefore characterized in terms of the
11-14
-------
particle aerodynamic diameter. This mechanism has an important influence on the deposition of
particles larger than 0.5 urn D.. Settling has an important role in the deposition of environ-
ae
mental aerosols in the distal region of the bronchial airways and in the pulmonary region.
Deposition by diffusion results from the random (Brownian) motion of very small particles
caused by bombardment of the gas molecules in air. The magnitude of this motion can be des-
cribed by the diffusion coefficient for a given physical particle diameter. Since particles
larger than 0.5 pm D ^ have relatively small diffusional mobility compared with sedimentation
ctQ
or inertia, diffusion primarily affects deposition of particles with physical diameters smaller
3
than 0.5 urn D . For particles of 0.5 pm D with a physical density of about 1 g/cm , the in-
0© 36
fluences of inertia! properties and diffusional properties on lung deposition are about equal.
Accurate calculation of the diffusional deposition of aerosols in the airways requires infor-
mation concerning the three-dimensional velocity profile of airflow in each airway segment.
If the flow of a given segment is laminar and approximately Poiseuille, the probability of
deposition by diffusion might be approximated using the Gqrmley and Kennedy (1949) equation
for a cylindrical pipe. This assumes, however, the aerosol is mixed at the entrance of the
cylinder and that the flow is constant. It therefore overestimates deposition in lung seg-
ments where there is minimal mixing between tidal and residual air and reversible laminar flow
between segments.
The diffusivity and interception potential of a particle depend on its physical size,
whereas the inertial properties of settling and impaction depend on its aerodynamic diameter.
These two measures of size may be quite different, depending on particle shape and physical
density. Because the main mechanism of deposition is diffusion for particles whose physical
(geometric) size is less than 0.5 pm D and impaction and settling above 0.5 M™ D,_» 0.5 pm is
O©
convenient to use as the boundary. Although this convention may lead to confusion in the case
2
of very dense particles, most environmental aerosols have densities below 3 g/cm , and the
deposition probability tends to have a minimum plateau between 0.2 urn and 1 pm D".
ae
Information concerning breathing patterns and respiratory physiology, the anatomical and
geometrical characteristics of the airways, and the physical behavior of insoluble spherical
particles can be used to develop theoretical models of regional deposition (Findeisen, 1935;
Landahl et a!., 1951; Landahl, 1963; Beeckmans, 1965; Yu, 1978). In these models, deposition
of inhaled aerosols in a given region of the respiratory tract or in the entire tract is ex-
pressed as a fraction of inhaled particles. Deposition fraction is the ratio of the number or
mass of particles deposited in the respiratory tract to the number or mass of particles in-
haled. The undeposited fraction represents those particles that are exhaled after inhalation.
For example, pulmonary deposition is the ratio of the number or mass of particles deposited in
the unciliated small airways and gas exchange spaces of the parenchyma of the lung to the
number or mass of particles entering the nose or mouth. The fraction not deposited in the
pulmonary region is either deposited in some other region or exhaled. Similarly, deposition
fractions can be defined for the other regions of the respiratory tract.
11-15
-------
Host model calculations treat the various mechanisms of deposition as independently occur-
ring phenomena. Such processes as Brownian diffusion and gravitational settling, however,
will interfere with each other when their effects are of comparable magnitude, and that inter-
ference can reduce the combined deposition to less than the sum of the separate depositions
(Goldberg et a!., 1978). Taulbee and Yu (1975) developed a theoretical deposition model that
allows for the combined effects of the primary deposition mechanisms and features an imaginary
expanding tube model of the airway system (Weibel, 1963) based on cross-sectional areas and
airway lengths.
Historicallyj the most widely used models of regional deposition versus particle size
were developed by the ICRP under the chairmanship of P. E. Morrow (Morrow et al,,, 1966).
These models were developed to determine radiation exposure from inhaled radioactive aerosols.
Although the ICRP aerosol deposition and clearance models were not intended for broad applica-
tion to environmental aerosols, they have been so applied by some scientists. The ICRP Task
Group used the anatomical model and impaction and sedimentation equations of Findeisen (1935)
and the general methods of Landahl (1950, 1963) for calculating deposition in the tracheo-
bronchial and pulmonary regions. The Gormley and Kennedy (1949) equation for cylindrical
tubes was used for calculating diffusional deposition. For head deposition, inhalation
through the nose with a deposition efficiency given by the empirical equation of Pattle
(1961a) was used. Particles were assumed to be insoluble, stable, and spherical with physical
densities of 1 g/cm and log-normally distributed with a a as high as 4.5. When the results
were expressed in terms of HMD for these various sized distributions of unit density (equiva-
lent to the MMAD), the range of the expected regional deposition values was relatively narrow.
At the time the ICRP Task Group models were developed, the available human data were pri-
marily total deposition values for polydisperse and sometimes unstable aerosols (Landahl and
Herrmann, 1948; Pavies, 1964b; Van Wijk and Patterson, 1940; Brown et al., 1950; Dautrebande
and Walkenhurst, 1966; Morrow et al., 1958; Landahl and Black, 1947). Since then, the deposi-
tion in humans of monodisperse insoluble, stable aerosols of different sizes has been measured
under different breathing conditions. Extensive studies were conducted by Lippmann (1977),
Heyder et al. (1975, 1980a,b), Stahlhofen et al. (1980), Chan and Lippmann (1980), and
Giacomelli-Maltoni et al. (1972). Additional useful data were reported by Palmes and Wang
(1971), Shanty (1974), George and Breslin (1967), Altshuler et al. (1967), Hounam et al.
(1971a,b), Foord et al. (1976), Pavia et al. (1977), among others (Muir and Davies, 1967;
Taulbee et al., 1978; Hounam, 1971; Heyder, 1971; Heyder and Davies, 1971; Fry and Black,
1973),
11.2 DEPOSITION IN MAN AND EXPERIMENTAL ANIMALS
11.2.1 Insoluble and Hydrophobic Solid Particles
11,2.1.1 Total Deposition—The background information in Section 11.1 demonstrates that a
knowledge of where particles of different sizes deposit in the respiratory tract and the
extent of their deposition is necessary for understanding and interpreting the health effects
11-16
-------
associated with exposure to particles and S0?. As was seen, the respiratory tract can be
divided into regions on the basis of structure, size, and function. Insoluble particles
depositing in the various regions contact or affect different cell populations and have large
differences in retention times and clearance pathways (see Section 11.3).
If the quantity of aerosol exhaled is compared with that inhaled, the data can be ex-
pressed as total deposition, but regiona-1 involvement cannot be,.distinguished. By tagging the
test aerosols with radiolabels, investigators have been able to separate deposition by region,
beginning with either nasal and nasopharyngeal deposition for nose breathing or oral and
pharyngeal deposition for mouth breathing. The measurement pf clearance of the radiolabeled
aerosol from the thorax can be used to separate early clearance, indicative of tracheobronchial
(TB) deposition, from more slowly cleared pulmonary (P) deposition.
Total respiratory tract deposition with nose breathing is given in Figure 11-3, and total
deposition with mouth breathing is depicted in Figure 11-4. Analyses based on the difference
between concentrations of inhaled and exhaled particles, as well as those based on external j_n
vivo measurements of radiolabeled particles, are represented in the studies from which these
figures are taken. With nose breathing, complete deposition can be expected for particles
larger than about 4 urn D . Mouth breathing bypasses much of the filtration capabilities of
aG
the extrathoracic (ET) region, and a shift upward to around 10 urn D.^ occurs before there is
ae
complete deposition of the inhaled particles.
The various studies all appear to show the same trend. The particle size for minimum
deposition is less clear for nasal breathing than for mouth breathing, for which minimum depo-
sition is at about 0.5 urn D . Heyder and coworkers (1973a, 1973b, 1975) carefully matched
ac
breathing patterns in subjects in their studies of the deposition of 0.5 pm D particles on
ae
which there were no electrical charges; their data are the deposition minima in Figure 11-4.
Thus far, deposition of particles less than 0.1 urn diameter was studied in human subjects only
by Swift et al. (1977).
Heyder and coworkers (1975, 1980a,b) studied the effects of respiratory parameters on
aerosol deposition in systematic experiments comparing deposition of different sized monodis-
perse aerosols in human volunteers at different tidal volumes, Flowrates, and breathing fre-
quencies. For particles between 0.1 and 4.0 urn in diameter, Heyder et al. (1975) measured
total respiratory deposition during either nose or mouth breathing while sequentially main-
taining a given TV, breathing rate, or inspiratory flowrate and then varying the other two
parameters. They demonstrated several important features of aerosol deposition in the human
respiratory airways. Heyder et al. (1980a,b) extended these studies to particles as large as
9 urn D • in the mouth-breathing experiments they also determined alveolar deposition.
ac
With volumetric flowrate held at 15 1/min while the subject breathed through the mouth, the
particle size yielding the lowest deposition changed from 0.66 urn D,_ at TV 250 ml to 0.46 urn
ac
11-17
-------
I
h-«
00
1.0

0.9

0.8

0.7

0.6

| °-5

g 0.4
LU
Q

0.3

0.2

0.1
SOURCE

O GIACOMELLI-MALTONI et al. (1972)
O HEYDER et al. (1975)
A SHANTY (1974)
O GEORGE AND BRESLIN (1967)
-Z- HEYDER, et al. (1980a)
HEYDER, et al. (1980a)
TIDAL VOL,
ml
D

1 1 I I I I
I 1 1
0.1 0.2

PHYSICAL DIAMETER,
0.4 0.5 0.6 0.8 1.0
2.0

-AERODYNAMIC DIAMETER,/
4.0
6.0 8.0 10.0
Figure 11-3. Deposition of monodisperse aerosols in the total respiratory tract for nasal breathing in humans as a function of aero-
dynamic diameter, except below 0.5 urn, where deposition is plotted vs. physical diameter. The data are individual observations,
averages, and ranges as cited by the various investigators.
-------
SOURCE
O LANDAHL
D LANDAHL
TIDAL VOL, ml RES. RATE, breaths/min
eta!. (1951)
at al. (1952)
A ALTSHULER et al (1957)
V GEORGE AND BRESLIN (1967)
OGIACOMELLI-MALTONI it al. (1972)
• CHAN & LIPPMANN (1980)*
• FOORD etal. (1976)
A MARTENS & JACOBI (1973)
1.0
0:9
0.8
0.7
0.6

"y> 0,5
o
P
I0-4
0.
ill
Q
0.3
0,2
0.1
0
0.
500
1500
soo
760
1000
1000
1000
1000
15
15
15
11
12
14
15
14
SOURCE

T LEVER (1974)
• MUIR & DAVIES (1967)
O DAVIES etal.S1972)
B HEYDER etal (1975)
A SHANTY (1974)
V STAHLHOFEN et al. (1980)
O STAHLHOFEN et al. (1881)
V SWIFT etal. (1977)
•USED MMD FOH D <0 S jUm J^a H

—
_
—

"~ *
~~ ^
*

I I

*
V
* ^

01 0.02

i !
0.04 0.
PHVSIfi
"™~— rn ioi\*i

36
\L
II I

L. I
a M *
° JT *r
|i
.*
0.20

''
.«
•

;* -
» •
*, j' , x*""'
1

' <
i
. ™

0.4 0

EVDERetal. |
MM

; ?•
C3 w
j
i ii i
5 0.6 0.8

*
"l
\
I

73b)

<
>
G

a
>
o
v

TIDAL VOL, ml
600
500
600
1000
1140
1500
1000
500
500
! 1 1
• jyr
T'?,
• |0 "3*
> V
!*
^

i i i
.0 2.0 4.0
AERODYNAMIC DIAMETER, /ur
RES. RATE, breaths/

I la.
i
»•*
1

1
6.0
tl

16
15
16
15
18
15
7.5
15
15
IN
|i
i
•
—

"n™"™
—
—
1 1

8.0 10.0

Figure 11-4. Deposition of monodisperse aerosols in the total respiratory tract for mouth breathing in humans as a function of
aerodynamic diameter, except below 0.5 jum, where deposition is plotted vs. physical diameter The data are individual observa-
tions, averages, and ranges as cited by the various investigators.
-------
at TV 2000 ml. Breathing at TV 1000 ml changed this minimum deposition size from 0.58 [jm D
clG
at 30 BPM to 0.46 pm at 3.75 BPM. Hence, the particle size of minimum deposition was reduced
with increased residence time of particles in the lung and the net deposition for all par-
ticles was increased. In fact, as the breathing rate went from 3.75 to 30 BPM, the deposition
fraction at 1 pm D.,,, went from 0.08 to 0.4, an increase of a factor of 5. In contrast to
ae
mouth breathing, however, the particle size of minimum deposition with nose breathing was in-
dependent of the residence time of particles in the respiratory tract when 1000 ml of aerosol
at different flowrates was inhaled.
When Heyder et al. (1975) kept the breathing frequency constant while changing the flow-
rate and having the subjects breathe through the mouth, the deposition for particles smaller
than 1 pm D__ remained essentially unchanged, indicating that inertia! impaction was of little
cl6
importance in the deposition of submicrometer aerosols. On the other hand, the deposition of
particles larger than 1 |jm Q was enhanced at high flowrates, indicating the influence of
36
inertia! impaction on the deposition of larger particles.
Sedimentation and impaction are competing deposition mechanisms, being governed by mean
residence time and flowrate, respectively. Hence, impaction will be the dominant mechanism at
high flowrates and short residence times, and most particles will be deposited by sedimenta-
tion at low flowrates and long residence times. Heyder et al. (1980a) showed that for 1 pm
D.,Q particles, total deposition for mouth breathing at 1000 ml TV increased with increasing
etc
mean residence time, indicating these particles were mainly deposited by sedimentation. For 8
pm D particles, increasing flowrate increased deposition so that these particles were mainly
deposited by impaction. A transition region was observed for particles about 4 urn D : Heyder
06
and coworkers (1980a) noted the transition region was shifted towards smaller particles for
nose breathing.
11.2.1.2 Extrathoracic DeppsHIon—The fraction of inhaled aerosol depositing in the ET
region can be quite variable, depending on particle size, flowrate, and breathing frequency,
and whether breathing is through the nose or through the mouth. During exertion, the flow
resistance of the nasal passages cause a shift to mouth breathing in almost all individuals,
thereby bypassing much of the filtration capabilities of the head and leading to increased
deposition in the TB region. Extrathoracic deposition is shown for nose breathing in Figure
11-5 and for mouth breathing in Figure 11-6. Deposition in this region is usually plotted as
2
a function of D Q, since this is a convenient parameter for normalizing impaction-dominated
36
deposition data when the actual flowrates are not identical (Pattle, 1961a; Stahlhofen et al.,
1980; National Academy of Sciences, 1980; Hounam et al,, 1969, 1971a). For reference, a scale
showing aerodynamic diameter when Q = 30 1/min is also shown, since this flowrate approximates
the average flowrate for the studies cited in these figures.
11-20
-------
1.0

0.9

0,8

0.7

0.6
0
AERODYNAMIC DIAMETER (at 30 liters/mm), fim

2 345
8 9 10
I 1 1
SOURCE TIDAL VOL, mi R ESP. RATE, breaths/min
D HOUNAMetal. (1969) I
n -
O LIPPMANN (1970) 1000 14 M
D MARTENS & JACOBI (1973) 1000 14 j | ^
AGIACOMELLI-MALTONI etal (1972) 1000 12 AM
*. T V¥^
O RUDOLF &HEYDER (1974) ,. _ Jr

~ <

D T x
5JH ix^

>
1
r
" t
riX
oA.K

't^r 1 n
i I 1 I -
5 \r
JL. ,*•
<>Xb
^^^>
^ /l\
K^^o
X"II T
XT oln T
^° 1
T A V
1 T
-11 i i.

t
(
T
L
1

»
0

1
I 'P^ ' '
xf
T I

j. —

1)11
i 1 I Mill
Z
O
°-4
o.
u
Q
0.3

0.2

0.1
10
20
40 60 80 100
200 400 600 8001000

D2Q ^
2000
4000 6000 10,000
Figure 11-5. Deposition of monodisperse aerosols in the extrathoracic region for nasal breathing in humans as a function of
where Q is the average inspiratory flow rate in liters/min. The solid line is ICRP deposition model based on the data of Pattle (1961a).
Other data show the median and range of the observations as cited by the various investigators.
-------
I
1X3
(S3
1.0

0.9

0.8

0.7

0.6

O °'5

§0.4
u
Q
OJ

0.2

0.1
AERODYNAMIC DIAMETER (at 30 liters/min),pm
2 3456
8 9 10
SOURCE
O LIPPMANNf 19771
D STAHLHOFEN, at al. {1980S
^7 CHAN & LIPPMANN (1980)
O STAHLHOFEN, et al. (1980)
— CHAN & LIPPMAN (1980)
TIDAL VOL, ml
1000
1000
1000
1500
RESP. RATE, breaths/rain
14
7.5
14
15
REG. LINE OF ALL DATA
14
10
20
40
60 80 tOO
200
400
D2Q —
600 8001000
2000
4000 6000 800010,000
Figure 11-6. Deposition of monodisperse aerosols in extrathoracic region for mouth breathing in humans as a function of
where Q is the average inspiratory flow rate in liters/min. The data are the individual observations as cited by the various invest-
igators. The solid line is the overall regression derived by Chan and Lippmann (1980S.
-------
Particles larger than about 10 pm D entering the nose are effectively deposited in the
ET region (Figure 11-5), Also, deposition is slight (10 percent) for particles less than 1 pm
D . Similarly, for 10 and 1 |jm D particles under conditions of mouth breathing (Figure
90 3S
11-6), ET deposition is about 65 and 2 percent, respectively. The regression curve shown in
Figure 11-6 is from Chan and Lippmann (1980) who used their own data, as well as the data of
Lippmann (1977) and Stahlhofen et al. (1980) for Q = 45 liter/min, in their analysis. As in-
dicated by Chan and Lippmann (1980), some of the lower values of ET deposition may be due to
partial clearance to the stomach before the measurement of head deposition was obtained. Par-
ticles can be swallowed even when the subject consciously tries to avoid swallowing (Lippmann,
1977; Stahlhofen et al., 1980),
11.2.1.3 Tracheobronchial Deposition—As was seen earlier, when aerosols are inhaled through
the nose, relatively efficient ET filtration eliminates the passage of most particles larger
than about 10 pm D to the TB region. Mouth breathing markedly alters the deposition of in-
36
haled particles in humans, in that larger particles can,enter both the TB and P regions to a
greater extent (Morrow et al., 1966; Lippmann, 1977;*Heyder et al., 1980a,b; Stahlhofen et
al., 1980). For mouth breathing, TB deposition expressed as a fraction of the particles
entering the trachea is shown in Figure 11-7 plotted against particle size. Approximately 80
to 90 percent of 8 to 10 (jm D particles entering the trachea are deposited in the TB region,
ciS
as compared with less than 10 percent for particles less than 1 JJIR D . The increased penetra-
ae
tion of large particles deeper into the respiratory tract when a person breathes through the
mouth can be seen from the 20 to 30 percent experimental TB deposition data for particles 8 to
10 |jm D (Stahlhofen et al., 1980). The solid curve in Figure 11-7 is from Chan and Lippmann
9S
(1980), depicting the experimental TB deposition data from their investigations using the
average value of a new anatomic parameter, the bronchial deposition size, for the average Q
value measured in their study (Q = 39 1/min). This parameter enables the classification of
various individuals and populations according to their TB deposition efficiencies and is an
improvement over the characteristic airway dimension parameter developed earlier by Palmes and
Lippmann (1977).
Deposition in the TB region is influenced by both impaction and sedimentation, with the
relative contribution of these two mechanisms changing with particle size and airflow rate.
Impaction predominates for deposition of particles larger than about 3 ion D and flowrates
at;
greater than about 20 1/min; on the other hand, sedimentation deposition becomes a larger
fraction of a diminishing TB component for smaller particles and lower flows (Lippmann, 1977).
The importance of impaction for TB deposition is reflected by deposition often being plotted
2
against the inertia! parameter, D Q (Stahlhofen et al., 1980 and Lippmann, 1977).
36
For a given particle size, TB deposition with mouth breathing varies greatly from subject
to subject among nonsmokers, cigarette smokers, and patients with lung disease (Lippmann et al.t
11-23
-------
r
ro
•Pa
1.0

0,9

0.8

0.7

0.6

O 0.5
E
I 04
LU
Q
OJ

0.1
I I I I
SOURCE
O LIPPMANN & ALBERT (1§69!
D LIPPMANN (1977!
& STAHLHOFEN, et a\. (1i80)
O CHAN AND LIPPMANN (1980)*
CHAN AND LIPPMANN (1980!
—— ICRP MODEL FOR 1450 ml TV
Mill I
TIDAL VOLUME, ml
750
1500
1500
1000
REG. LINE FOR ALL DATA
I I I I I I J I
RESP. RATE, breaths/min '
0.01
•USED MMD FOR D <0,5 jllm
0.02
0.04 0.06 0.080.10
PHYSICAL DIAMETER, ju
0.20 0.400.500.600.80 1.0 2.0 4.0 6.0 8.0 10.0
_ A. AERODYNAMIC DIAMETER,/«n -
Figure 11-7. Deposition of monodisperse aerosols in the tracheobronchial region for mouth breathing in humans in percent of the
aerosols entering the trachea as a function of aerodynamic diameter, except below 0.5 jum, where deposition is plotted vs. physical
diameter as cited by different investigators. Dashed line is ICRP model for 1450 ml tidal volume. The solid line is the overall re-
gression derived by Chan and Lippmann (1980).
-------
1971). On the average, TB deposition is slightly elevated in smokers and greatly elevated in
patients with lung disease (Lippmann et a!., 1977; Cohen, 1977). Each subject, however, exhi-
bits a characteristic and reproducible relationship between particle size and deposition as
indicated by the data of Stahlhofen et al. (1980), depicted in Figure 11-8. For the two
breathing patterns shown, the steep increase of the ET deposition values with increasing par-
ticle size is accompanied by a corresponding decrease in TB deposition, so that TB deposition,
as a function of particle aerodynamic diameter, may be described by a bell-shaped curve with a
maximum (Stahlhofen et al., 1980). Although these investigators did not experimentally study
particles larger than 9 urn D , extension of their bell-shaped curves would support the con-
36
elusion of Miller et al. (1979) that about 10 percent of particles as large as 15 |jm D can
S6
enter the TB region during mouth breathing. Miller et al. (1979) used the TB deposition data
of Lippmann (1977) and aerodynamic diameters computed at a mean flowrate of 30 I/ min. This
flowrate is bracketed by the mean flowrates of 15 and 45 1/min used by Stahlhofen et al.
(1980).
The data of Stahlhofen et al. (1980) in Figure 11-8 on three subjects show lower values
and less scatter than the other data contained in the figure. Chan and Lippmann (1980) cited
two possible explanations for the differences. Stahlhofen and coworkers (1980) used constant
respiratory flowrates in comparison with the variable flowrates used by Chan and Lippmann
(1980). Also, the two laboratories used different bases to separate the initial thoracic
burden into TB and P components. Stahlhofen et al. (1980) extrapolated the thoracic retention
values measured during the week after the end of bronchial clearance back to the time of inha-
lation; they considered P deposition to be the intercept at that time, with the remainder of
the thoracic burden considered as TB deposition. This approach yields results similar to, but
not identical with, those obtained by treating TB deposition as equivalent to the particles
cleared within the first day.
Deposition calculations usually group lung regions without regard to nonuniformity of the
pattern of deposited particles within the regions. Schlesinger and Lippmann (1978) found that
nonuniform deposition in the trachea could be caused by the airflow disturbance of the larynx.
Bell and Friedlander (1973) and Bell (1978) observed and quantified particle deposition as it
occurs at a single airway bifurcation and found it to be highly nonuniform and heaviest around
the carinal arch. Raabe et al. (1977) observed that the relative lobar pulmonary deposition
of monodisperse aerosols was up to 60 percent higher in the right apical lobes than in others
of small rodents (corresponding to the human right upper lobe) and that the difference was
greater for 3.05 and 2.19 fjm D particles than for smaller particles. In addition, Raabe et al,
O.G
(1977) showed that these differences in relative lobar deposition were related to the geometric
mean number of airway bifurcations between trachea and terminal bronchioles in each lobe for
rats and hamsters. Since similar morphologic differences occur in human lungs, nonuniform
11-25
-------
TV * 1000 ml. BPM -7S/m!n
i
ro
1.0
0.8
0.6
0.4
0.2
O 0
2 i.o
LU
Q
0.8

0.6

0.4

0.2

0
= 1SOOmI, BPM = 15/mm
I SUBJ. 1I IT
1
689 2 4 6891

AERODYNAMIC PARTICLE DIAMETER, Mm
SUBJ. 4
8 9
Figure 11-8. Total and regional depositions of monodisperse aerosols with mouth breathing as a function of the aerodynamic diameter
for three individual subjects as cited by Stahlhofen et al. (1980). (T = Total, TB = Tracheobronchial, P = Pulmonary, ET = Extrathoracic,
TV = Tidal Volume, BPM = Breaths Per Minute.)
-------
lobar deposition should also occur. Sehlesinger and Lippman (1978) found nonuniform deposi-
tion in the lobar branches of a hollow model of the TB airways with enhanced carinal deposi-
tion and were able to demonstrate a correlation of higher lobar deposition and the reported
incidence of bronchogenic carcinoma in the different human lobar bronchi. Occupational lung
diseases, such as silicosis and asbestosis, also show distinctive distributional features
(Morgan and Seaton, 1975).
11.2.1.4 PulmonaryDeposition--Pulmonary deposition as a function of particle size is shown
in Figure 11-9. All of the experimental points plotted were obtained in mouth-breathing
studies on nonsmoking normal subjects who inhaled monodisperse aerosols.
The eye-fit band approximately encompasses the range of deposition values obtained in the
studies cited; a variety of TV's and breathing frequencies were used. Also shown in Figure
11-9 are the deposition curve from the predictive model of Yu (1978) and an estimate of the P
deposition that could be expected for nose breathing (Lippmann, 1977). Lippmann (1977)
derived the estimate by analysis of the difference in head retention during nose breathing and
mouth breathing.
The P deposition curve peaks at about 3.5 (jm D with the middle of the eye-fit band in
3.c
Figure 11-9 being located at about 50 percent deposition. However, the data of Stahlhofen et
al. (1980) for a tidal volume of 1000 ml and 7.5 BPM (reflective of breathing very slowly and
deeply) showed that P deposition of 3.5 pm D particles can be as high as 70 percent.
For nose breathing, the size associated with maximum deposition shifts downward to about
2.5 jjm D . Also, the deposition peak is much less pronounced (about 25 percent), with a
36
nearly constant P deposition of about 20 percent for all sizes between 0.1 urn and 4 urn D
3
Pulmonary and total deposition of Fe?0- (density 3.2 g/cm ) particles and di-2-ethylhexyl
sebacate droplets for mouth breathing was evaluated by Heyder et al. (1980a»b) as a function
of aerodynamic diameter for two breathing patterns. Some results with di-2-ethylhexyl
sebacate particles were reported by Heyder et al. (1980a) in terms of particle diameter. They
are presented here for uniformity in terms of aerodynamic diameter, since these particles were
close to unit density. Keeping the mean volumetric flowrate constant at 250 ml/s and allowing
the mean residence time to vary between 2 and 8 s, they observed that as the mean residence
time increased, the particle size having the greatest probability of deposition decreased.
With this mean flowrate, particles smaller than about 2.4 pm D were exclusively deposited in
36
the P region, indicating that their inertia was not sufficiently high for impaction loses.
When the mean flowrate was increased to 750 ml/s and the mean residence time was 2 s, particles
with an aerodynamic diameter smaller than about 1.5 ym were exclusively deposited in the P
region of the respiratory tract. The data of Heyder et al. (1980a,b) also showed that the
particle size associated with the peak of the deposition curve and the magnitude of the peak
decrease as the mean flowrate increases. In the above studies, maximum P deposition was at
3.5 and 3 urn D when Q was 15 and 45 1/min, respectively
616
11-27
-------
i
t\5
1.0

0.9

0.8

0.7
Z 0.6
O
H
m 05
O
Q.
LU
Q 0.4
0.3

0.2

0.1

0
TT
1 TT
1 T
SOURCE
TIDAL VOL, RES. RATE,
ml
O STAHLHOFENetal. (1980) 1500
A STAHLHOFENetDl. (1981). 1000
O ALTSHULERetal. (1967) 500
* GEORGE S. BRESLIN (1967) 760
* SHANTY (1974) 1140
• LIPPMAN& ALBERT (1969) 1400
* CHAN 8.LIPPMANS1980!« IOOO
— yy (1978!
— LiPPMAN (1977!
•USED MMD FOR D< 0.5 JUm
b/min
15
7.5
15
11
18
14
14
0.01
0.02 0.03 0.05
0.08 0.1 0.2 0.3

PHYSICAL DIAMETER, (a
0.5
0.8 1.0
2.0 3.0
5.0
8.0 10.0
- AERODYNAMIC DIAMETER, fm
Figure 11-94 Deposition of monodisperse aerosols in the pulmonary region for mouth breathing in humans as a function of aero-
dynamic diameter, except below 0.5 jum, where deposition is plotted vs. physical diameter. The eye-fit band envelops deposition
data cited by the different investigators. The dashed line is the theoretical deposition model of Yu (1978) and the broken line
is an estimate of pulmonary deposition for nose breathing derived by Lippmann (1977).
-------
11,2.1.5 Deposition In Experimental Animals—Since much information concerning inhalation
toxicology is collected with beagles or rodents, the comparative regional deposition in these
experimental animals must be considered to help interpret, from a dosimetric viewpoint, the
possible implications for humans of animal toxicologies! results.
The study by Holma (1967) on rabbits examined mucociliary clearance rates, but Lippmann
(1977) derived TB deposition information by further analyzing the data. Lung retention curves
indicated that the TB deposition of 6 |Jtn polystyrene spheres varied from 40 to 93 percent of
the total lung deposition, with a median of 60 percent. A median of 29 percent was found for
3 pro particles. The above values are remarkably close to the available data for humans.
Cuddihy et al. (1973) measured the regional deposition of polydisperse aerosols in beagles
with TV about 170 ml at about 15 BPM and expressed the results as mass deposition percentage
versus mass median aerodynamic resistance diameter (MMAD ) that ranged from 0.42 to 6.6 ym
3 r
with geometric standard deviation a = 1.8. These results are summarized in Figure 11-10 and,
compared with the ICRP Task Group Values for humans with TV 1450 ml, integrated to account for
a 0 = 1.8. In comparison with the TB deposition of large particles in rabbits exposed to
y
monodisperse aerosols for one test at 6.6 |jm D , the TB deposition in beagles was about 44
air
percent of the total lung deposition. With sizes between 2.5 and 3 pm D , the TB deposition
sr*
ranged from 5 to 39 percent, with a median deposition of 9 percent. The particle size for
minimum P deposition was approximately 0.6 pm D , with P deposition at this size ranging from
CU
about 12 to 35 percent and total deposition from about 18 to 55 percent.
Somewhat different results were obtained by Phalen and Morrow (1973) in dogs exposed to a
silver metal aerosol of 0.5 pm D_^ with a a = 1.5. Total deposition averaged 17 percent,
Of
with a range of 15 to 19 percent. In the Phalen and Morrow (1973) study, the dogs inhaled
through a trachea! tube so that there was no head deposition; in the study of Cuddihy et al.
(1973), head deposition varied from negligible to 5 percent for 0.5 ism D particles. In
3.6
experiments using donkeys (Albert el al., 1968, 1969; Spiegelman et al., 1968), eight animals
were tested periodically with monodisperse 3 to 3.5 |jm D Fe,0,, aerosol. Tracheobronchial
36 L. -3
deposition averaged 50 to 70 percent of the total lung deposition, with a median of 54 per-
cent.
Raabe et al. (1977) measured the regional deposition of 0.1 to 3.15 urn D monodisperse
36
aerosols in rats (TV about 2 ml, 70 BPM) and Syrian hamsters (TV about 0.8 ml at about 40
BPM). Their results are summarized in Figure 11-11. The P deposition of 1 to 3 p.m D par-
ticles is about 6 to 9 percent in rats and hamsters; deposition of these same size particles
in humans varies from 21 to 24 percent for nose breathing and from 20 to 50 percent for mouth
breathing. For particles smaller than 1 pm D , differences in P deposition between humans
36
and these animal species decrease. Tracheobronchial deposition of particles 5 |jm D is slight
(•v 5 percent) in rodents due to very efficient removal of these particles in the head. In
11-29
-------
1.0
I
—A. TOTAL
O 0.1

g
<
C
u.

3 1.0
CL.
Ul
a

050
0.20
0.10
0.05
0.03"—
0.1
. DOG NO.

O26GE

"D267B

. A 268E

V2G9A

O284A

L
TTPP
.0- -
l i i i i
0.10
0.05
l i i 1 1 1
_ B. TRACHEO-
— BRONCHIAL
0.2
0.5
1.0
10
a i 1 1 1 tt
I I I I ffj I I
_C. PULMONARY
e
0.02 —
0.01
0.2 0.5 1.0 2.0 5.0 10 0.1 0.2 0.5 1.0

ACTIVITY MEAN AERODYNAMIC DIAMETER, /urn
2.0
5.0 10,
Figure 11-10, Deposition of inhaled polydisperse aerosols of lanthanum oxide (radio-labeled with

in beagle dogs exposed in a nose-only exposure apparatus showing the deposition fraction (A) total dog,

(B) tracheobronchiai region, (C) pulmonary aveolar region, and (D) extrathoracic region (adapted from

Cuddihy et al. 1973). Dashed lines represent range of observed values.
11-30
-------
HAMSTER
A EXTRATHORACIC
• TRACHEOBRONCHIAL
• PULMONARY
0.2 0.3
PHYSICAL DIAMETER,
1.0 2.0 3.0
AERODYNAMIC DIAMETER (Dar),j
Figure 11-11. Deposition of inhaled monodisperse aerosols of fused aluminosilicate spheres in small
rodents showing the deposition in the extrathoracic (ET) region, the tracheobronchial (TB) region, the
pulmonary (P) region, and in the total respiratory tract based upon Raabe et al. (1977).
11-31
-------
contrast, as presented in Figure 11-7, 50 percent of 5 urn D particles inhaled via the mouth
36
deposit in the TB region of humans, showing that large differences can exist between humans
and rodents in the TB deposition of large particles. In rodents, the relative distribution
among the respiratory regions of particles less than 3 urn D during nose breathing follows a
36
pattern that is similar to human regional deposition .during nose breathing. Thus, for par-
ticles less than 3 urn D , the use of rodents or dogs in inhalation toxicology research for
aG
extrapolation to humans can be justified from the available data.
11.2.2 Soluble, Deliquescent, and Hygroscopic Particles
Most deposition studies and models tend to focus on insoluble and stable test aerosols
whose properties do not change during the course of inhalation and deposition. Environmental
aerosols, however, usually contain deliquescent or hygroscopic particles that may'grow in the
humid respiratory airways. That growth will affect deposition (Scherer et a!., 1979).
Although the ICRP Task Group on Lung Dynamics (Morrow et al., 1966) addressed this problem by
considering the equilibrium diameter for deliquescent materials at relative humidities near,
but less than, 100 percent, the residence times in the respiratory tract may be too short for
large particles to reach their equilibrium size (Nair and Vohra, 1975; Charlson et al., 1978).
Also, environmental aerosols may consist of a combination of components, including complex
mixtures, that may not behave like pure substances. Since the temperature of the inspired
aerosol will usually be less than that of the respiratory tract environment, supersaturation
of water vapor, with respect to the aerosol particles, may exist.
Perron (1977) described the factors affecting soluble particle growth in the airways dur-
ing breathing. His results suggest that particles 1 urn D_ will increase by a factor of 3 to
ae
4 in aerodynamic diameter during passage through the airways. Extrathoracic, TB, and P deposi-
tion of the enlarged particles would be greater than the deposition expected for the original
particle size. Submicrometer particles, including those as small as 0.05 pm, will grow by a
factor of 2 in physical diameter, with relatively little effect on deposition. The hygro-
scopic growth of particles in the diffusion size range (< 0.5 urn physical diameter), however,
may alter their deposition pattern substantially, as the diffusional displacement is related
to the actual size and not the aerodynamic diameter. Pulmonary deposition of particles
smaller than 0.3 urn may be reduced with growth because of reduced diffusivity.
Atmospheric sulfate aerosols can be described as H?SCL partially or completely
neutralized by NH~. Growth of these particles will occur in the respiratory airways during
respiration. This growth involves chemical dilution of the electrolyte or acid with absorbed
water. A particle growing a factor of 3 in physical diameter must absorb a volume of water
equal to 26 times its original particle volume. Also, the increased size will enhance losses
by inertial mechanisms, including impaction in the upper airways. A 1 urn D particle of H9SO,,
36 ^ H*
or (NH.)?SO. may grow to nearly 3 urn D in the nasal region, increasing both ET and TB deposi-
tion by a factor of 2 or more over the-deposition expected for a 1 urn Dgg particle, with the
11-32
-------
net result that P deposition is-reduced. Particle growth in the airways may in some cases be
protective, since the reduced electrolyte or acid concentration will probably reduce the level
of local toxicity.
11.2.3 Surface-coated Particles
Some environmental particles may consist of a relatively insoluble core coated with vari-
ous chemical species including metallic salts, (NH4)pSO., (NH.)HSO*, H-50^, organic compounds
including polynuclear aromatic hydrocarbons, and small* particles of other sparingly soluble
materials. Although some surface growth due to water adsorption may occur in the airways,
growth will be limited by the availability of deliquescent or hygroscopic components on the
particle surface. In general, the increase in aerodynamic diameter that may occur would be
much less for coated particles than for purer forms of insoluble materials.
Important examples of coated particles are the fly ash, soot, or other residual solid
particulate aerosols released into the environment by combustion of fossil fuels. The exact
chemical form of the core of these particles will vary from nearly pure fused aluminosilicate
particles produced during the combustion of coal to carbonaceous or metal oxide particles pro-
duced by internal combustion engines. Volatile trace metal compounds and organic compounds
condense on these particles during the cooling of the effluent stream in the powerplant smoke-
stack or engine exhaust line and during release to the atmosphere. Also, absorption or con-
densation of SO™ and other gaseous species from the atmosphere can produce a high surface con-
centration on particles that are already airborne. If these processes are diffusion-limited,
the condensation and coagulation will be quantitatively proportional to particle diameter for
particles larger than 0.5 pm D and to particle surface area for smaller particles. In
96
either case, the fractional mass of the surface-coating material will be greater on smaller
particles than on larger ones. Thus, surface deposition provides a layer of soluble material
present in high concentration and results in small-particle enrichment, leading to a shift of
the MMD's of the potentially toxic surface materials to smaller aerodynamic equivalent
diameters than that of the total particle mass (Natusch and Wallace, 1979). Consequently,
pulmonary deposition of surface-enriched material may be significantly enhanced. Important
elements such as Se, Cd, As, V, Zn, Sb, and Be have been found to exhibit this size dependence
in coal fly ash aerosols (Davison et al., 1974; Natusch et a!., 1974; Gladney et al., 1976).
(See also Chapters 3, 5, and 6.)
11.2,4 Gas Deposition
The major factors affecting the uptake of gases in the respiratory tract are the
morphology of the respiratory tract, the physicochemical properties of the mucous and
surfactant layers, the route of breathing and the depth and rate of airflow, physicochemical
properties of the gas, and the physical processes that govern gas transport. A brief
discussion of these factors serves to illustrate their general role in the deposition of gases
and convey some aspects specific to the uptake of SCL.
11-33
-------
The complex morphological structure of the human respiratory tract has been discussed in
Section 11.1.3. The nature and structure of the respiratory tract in humans and animals cri-
tically influence the deposition of gases, since the relative contribution of gas transport
processes varies as a result of this morphology. The human tracheobronchial tree is more
symmetric, with respect to diameter ratios and branching angles, than that of dogs, rats, or
hamsters but is closest to that of the dog (Phalen et al. , 1978). The structure of the
tracheobronchial tree is variable from species to species, from lobe to lobe within a given
lung, and from one depth to another in the lung.
Physicochemical properties of a gas relevant to respiratory tract deposition are its
solubility and diffusivity in mucus, surfactant, lipid, and water, and its reaction-rate con-
stants in mucus, surfactant, lipid, water, and tissue. Henry's law relates the gas phase and
liquid phase interfacial concentrations at equilibrium and is a function of temperature and
pressure. In general, the more soluble a gas is in biological fluids the higher it is removed
in the respiratory tract. Although the solubility of most gases in mucus and surfactant is
not known, Henry's law constant for many gases in water is known, the value for SO- being 59.7
mole fraction in air per mole fraction in water at 37°C and one atmosphere of pressure
(Washburn, 1928). The diffusivities of most gases in mucus, surfactant, tissue, and water are
also unknown, thereby complicating efforts to model gas uptake in the respiratory tract.
Diffusivity may be much smaller in a viscous mucous fluid than in water, but ciliary activity
induces turbulence, which effectively increases mass transfer. Generally, transport rates of
the gas across the mucus-tissue interface, tissue layer, and the tissue-blood interface are
needed to fully understand the absorption and desorption of gases in the respiratory tract.
Information on biochemical reactions, however, may enable one or more of these compartments to
be ignored for a given gas.
The major processes affecting gas transport involve convection, diffusion, and chemical
reactions. The bulk movement of inspired gas in the respiratory tract is induced by a pres-
sure gradient and is termed convection. Molecular diffusion due to local concentration gra-
dients is superimposed on this bulk flow at all times, with the transport of the gas being
accomplished by the coupling of these two mechanisms. Convection can be decomposed into the
processes of advection and eddy dispersion. Advection is the horizontal movement of a mass of
air that causes changes in temperature or in other physical properties, and eddy dispersion
occurs when air is mixed by turbulence so that individual fluid elements transport the gas and
generate the flux. Because of the morphology of the respiratory tract and respiratory airflow
patterns, the relative contribution of the various processes to transport and deposition is a
function of location and point in the breathing cycle.
During the respiratory cycle, the volumetric flow rate of air varies from zero up to a
maximum (dependent upon tidal volume, breathing frequency, and breathing pattern) and then
back to zero. Usually expiration is longer than inspiration, and intervening pauses may
occur. The net result of these variables is to impart complicated flow patterns and turbul-
ence in some portions of the respiratory tract (see Section 11.1.4).

11-34
-------
In the study of the nature of gas mixing in the tracheobronchial tree and its effects on
gas transport, numerous modeling efforts have used an approach in which all pathways from the
mouth or trachea to the alveoli are combined into one effective pathway whose cross-sectional
area is equal to the summed cross-sectional area of all bronchial tubes at a given distance
from the mouth or trachea (Davidson and Fitz-Gerald, 1974; Paiva, 1973; Pedley, 1970; Yu,
1975; Scherer et al. , 1972). In this formulation, the mechanical mixing imparted by tube bi-
furcations, turbulence, and secondary flows and the mixing due to molecular diffusion are
represented by the functional form of the effective axial diffusion coefficient (Scherer
et al., 1975). Thus, this coefficient of diffusion incorporates the effect of axial con-
vection. The effective axial diffusion coefficient is a constant equal to the molecular
diffusivity only in the P region, where gas velocity is very small. In other regions of the
TB tree, however, the local average gas velocity and the tube geometry will jointly determine
the value. Various functional forms were proposed in the studies cited above for an appro-
priate expression for the effective axial diffusion coefficient.
By constructing individual streamline pathways frqm the trachea to the alveoli, Yu (1975)
derived an expression for the effective axial diffusion coefficient which equalled the alge-
braic sum of the molecular diffusion coefficient and an apparent diffusion coefficient. The
apparent diffusion coefficient arises from two independent mechanisms: 1) the nonhomogeneous
ventilation distribution in the lung, and 2) the interaction of nonuniform velocity and con-
centration profiles due to Taylor's mechanism in individual airways. Using an average stand-
ard deviation of airway lengths based on the data of Weibel (1963) and various flow theory
limiting values, Yu (1975) demonstrated that Taylor diffusion is dominated everywhere in the
TB tree by the apparent diffusion due to nonhomogeneous distribution of ventilation, rather
than being a major mechanism for gas transport in some airways as claimed by Wilson and Lin
(1970).
In all of the studies described previously, the diffusivity expressions used assume fully
developed flow in straight pipes to describe gas mixing, a condition not truly applicable over
most of the TB tree. Since flow patterns at tube bifurcations are different for inspiration
and expiration (Schroter and Sudlow, 1969), the mixing process and hence the effective diffu-
sivities are different. To obtain diffusivities applicable to the TB tree, Scherer et al.
(1975) used airway lengths and diameters from Weibel (1963) and branching angles from
Horsfield and Gumming (1967) to construct a five-generation symmetrical branched tube model
and to determine experimentally the effective axial diffusivity for laminar flow of a gas as a
function of mean axial velocities up to 100 cm/s in the zeroth generation tube. The relation-
ship was approximately linear, and diffusivities for expiration were about one-third those for
inspiration. The values obtained by Scherer et al. (1975) for steady flow can be applied to
oscillating flow in the TB tree provided the oscillating flow can be considered quasi-steady,
i.e., steady at any instant of time. This condition should hold in the first 10 generations
whenever flowrates are approximately greater than 0.1 1/s (Jaffrin and Kesic, 1974).
11-35
-------
Additional experimental uptake data are needed to obtain a better understanding of the
effects of various factors on the transport and removal in the lung of gases such as S0?.
Also needed along with these experimental data are refined theoretical approaches, as well as
more flexible computational models, such as that of Pack et al. (1977). The amount of SOp re-
moved depends on solubility, the velocity and turbulence of the air, the diffusing capacity
across the air-tissue interface and through the tissue, the volume of tissue available for gas
storage, and the rate of fluid exchange between these tissues and the storage reservoirs in
the body for SCL (Aharonson, 1976). The rate-controlling factor in the deposition of S0~ is
probably the vapor pressure of dissolved SO^ in buffered body fluids.
2
The diffusion coefficient of S02 in air at body temperature is 0,144 cm /s at sea level
(Fish and Durham, 1971; Sherwood et al., 1975). The complicated flow patterns and turbulence
in the upper respiratory tract and upper generations of the TB tree, in combination with high
solubility in body fluids, are responsible for the large removal of SO- in these regions.
Frank et al, (1969) surgically isolated the upper respiratory tract of anesthetized dogs with
separate connections for the nose and mouth. Radio!abeled S0? (35S) was passed through this
isolated ET region for 5 min, and nearly complete removal was observed for concentrations of
2.62 to 131 mg/m (1 to 50 ppm) at a flowrate through the nose of 3.5 1/min. Uptake of the
3
mouth averaged more than 95 percent at 3.5 1/min with S02 levels of 2.62 and 26.2 mg/m (1 and
10 ppm). When flow was increased 10-fold to 35 1/min, however, uptake by the mouth fell to
under 50 percent. Since the plastic tube through which the gas was delivered was inserted
only 2 cm into the mouth, a small fraction of the total pathway through the oral cavity, the
use of a tube could not account for the lower uptake observed. Moreover, the results are in
agreement with the theory of Aharonson et al. (1974). Strandberg (1964), using a trachea!
cannula with two outlets that allowed sampling of inspired and expired air, studied the uptake
of S0y in the respiratory tract of rabbits. He observed 95-percent absorption in the respira-
3 3
tory tract at 524 mg S0?/m (200 ppm), but at 0.13 mg S0?/m (0.05 ppm) absorption was lowered
to about 40 percent during inspiration, demonstrating an apparent concentration effect.
Absorption of SO, at expiration was 98 percent in the 524 mg/m (200 ppm) studies, compared
3
with 80 percent for experiments using 0.13 mg/m (0.05 ppm). Dalhamn and Strandberg (1961)
found that rabbits exposed to 262 to 786 mg SQ2/m (100-300 ppm) absorbed 90 to 95 percent of
the SOp. They noted that absorption was to some extent dependent on the technique whereby
trachea! air samples were obtained.
Corn et al. (1976) studied the upper respiratory tract deposition of S02 in cats and
computed mass transfer coefficients that can be used with surface area data to calculate the
amount of S02 removed in various parts of the respiratory tract. Using a theoretical
approach, their own empirical data, and information available from the literature, Aharonson
et al. (1974) examined the effect of respiratory airflow rate on nasal removal of soluble
vapors. The only assumption made regarding factors affecting local uptake was that there was
no back pressure in the blood. Hence, whether the rate of uptake is limited by diffusion
11-36
-------
through the gas phase, diffusion through the tissue, chemical reactions in the tissue, or
local blood flow in the tissues, the analytical approach is valid, as long as the rate of up-
take is proportional to the gas phase pressure of the vapor. Their analysis for acetone,
ether, ozone, and SO,, showed that the uptake coefficient, which defines the average flux of
soluble vapors into the nasal mucosa per gas-phase unit partial pressure, increases with in-
creasing airflow rate.
In experiments described by Brain (19705), the amount of SO, present in the trachea of
dogs increased 32-fold when the airflow rate was increased 10-fold. However, had the uptake
coefficient not changed with the flowrate, Aharonson et al. (1974) pointed out that penetra-
tion would have increased 500-fold. If the uptake coefficient for SO- is concentration
dependent, as the data of Strandberg (1964) suggest, increasing airflow rate may increase up-
take because higher levels of SO,, are present along the center of the airstream for the same
inspired concentration.
The deposition and clearance of SO, has also been studied in i_n vitro model systems. In
a model of the TB airways lined with a simulated airway fluid (bovine serum albumin dissolved
in saline), SO, was absorbed primarily in the upper third of the simulated airway with only a
small fraction of the SO, reaching the simulated alveolar or bronchiolar regions (Kawecki,
1978).
Uptake and release of S0? in the nose of human subjects breathing 42.2 mg/m (16.1 ppm)
through a mask during a 30-minute exposure period was studied by Speizer and Frank (1966).
During inspiration, the concentration of SO- had dropped 14 percent at a distance 1 to 2 cm
within the nose and was too small to detect at the pharynx with the analytical method used.
Expired gas in the pharynx was also virtually free of S0«, but in its transit through the nose
the expired air acquired SO, from the nasal mucosa. The expired SO, concentration at the nose
3
was 5.2 mg/m (2,0 ppm), or about 12 percent of the original mask concentration. In most sub-
jects, the nasal mucosa continued to release small amounts of SO™ during the first 15 min
after the SO, exposure ended (see Section 11.3.2).
3
Melville (1970) exposed humans to S0? at levels ranging from 4 to 9 mg/m (1.5 to 3.4
ppm) for periods up to 10 min. Respiratory tract extraction of S0_ during nose breathing was
significantly greater (p < 0.01) than during mouth breathing (85 vs. 70 percent, respectively)
and was independent of the inspired concentration of SO,. Andersen et al. (1974) found that
3
at least 99 percent of 65.5 mg S0?/m (25.0 ppm) was absorbed in the nose of subjects during
inspiration. Values obtained after 1 to 3 h of exposure were the same as those obtained after
4 to 6 h of exposure, thereby indicating there was no saturation effect during this period of
time.
11.2.5 Aerosol-Gas Mixtures
Gases readily diffuse to the surface of particles and can participate in a variety of
surface interactions. Surface adsorption related to temperature and gaseous vapor pressure
occurs if adsorption sites for the gas molecules are present on the particles. Such physical
11-37
-------
adsorption can be described by the Langmuir isotherm or more complex isotherms (Gordieyeff,
1956). In addition, chemical adsorption can occur involving chemical transformations and
bonds that enhance transfer of gaseous materials to the particulate phase. Such transforma-
tions can include both inorganic and organic vapors. In addition, aerosols of liquid droplets
can collect and carry volatile species that are dissolved in the droplets. In these cases,
aerosols can serve as vectors carrying molecules of various substances deeper into the airways
than would occur if the substances were in their gaseous forms.
Sulfuric acid in the environment may be reduced in acidity by naturally occurring ammonia
(NHa) to form ammonium sulfate (NH4)n$CL and ammonium bisulfate (NH.HSO.). Larson et al.
(1977) made short-term measurements that suggest that endogenously generated NH, gas in the
human airways may rapidly and completely neutralize HUSO, aerosols at the concentrations that
are normally encountered in the ambient environment. Also, NH, is generated from food and
excreta in inhalation chambers used to expose experimental animals to H~S04, so that some
neutralization of HpSQ4 in these test atmospheres probably occurs.
Because S0« is found in the gas phase of various environmental aerosols, the reactions
that occur between SOp and aerosols, and the gas-to-particle conversions that may occur, can
influence greatly the regional deposition of biologically active chemical species. Since SO,
is highly soluble in water, droplet aerosols, including those formed by deliquescent particles,
will collect dissolved SOp and can carry some of the resulting sulfurous acid not neutralized
by NH, deep into the lung. The presence of certain sulfite species formed by such reactions
in environmental aerosols has been suggested (Eatough et al., 1978). Sulfur dioxide is also
known to be converted to sulfate by reactions catalyzed by some aerosols, including those con-
taining iron or manganese. The simple adsorption of SOp to aerosol surfaces by chemical re-
action may lead to the aerosol acting as a vector for transporting SOp to the P region.
The deposition of the aerosol and gaseous fractions of the sulfur species can be pre-
dicted from the properties of these fractions. Hence, the problem of estimating deposition
requires an understanding of the proportion of sulfur species associated with the aerosol
fraction and their chemical properties. Since these reactions are dynamic processes, the rate
and mechanics of the gas-particle chemical reactions, especially as they may occur in the air-
ways, must be understood in order to predict subsequent biological effects, such as the poten-
tiation of increased airway resistance in guinea pigs with SOp by some particles (Amdur and
Underbill, 1968).
11.3 TRANSFORMATIONS AND CLEARANCE FROM THE RESPIRATORY TRACT
Particulate material deposited in the respiratory tract may eventually be cleared by the
TB mucociliary conveyor or nasal mucous flow to the throat and is either expectorated or
swallowed. Other deposited material may be cleared by either the lymphatic system or transfer
to the blood. Sulfur dioxide reacts rapidly with biological constituents to produce S-
sulfonates (Gunnison and Benton, 1971, see Chapter 12, Section 12.2.1.2.1). The role of
11-38
-------
clearance as a protective mechanism for the respiratory tract depends on the physicochemical
characteristics of the particles (or gaseous species), the site of deposition, and respiratory
physiology. If the particles dissolve rapidly in body fluids, their deposition in the nasal
turbinates with subsequent absorption into the blood is important, and total deposition of
soluble particles may be more critical than regional deposition. For relatively inert and in-
soluble particles, deposition in the P region, where they may be tenaciously retained, may be
more hazardous, unless their action is mediated through ET and TB deposition. The deposition
by dissolution of S0? in the ET region may be protective, since it may involve less serious
biological effects than deposition in the TB or P airways. Mouth breathing would lessen the
ET absorption and increase the S0? levels entering the lung. If the particles or SCL chemi-
cally react with body fluids, transformations of the material can affect clearance. In all
respiratory regions, the dissolution of particles competes with other clearance processes.
Since respiratory tract clearance may begin immediately after the initial deposition, the
dynamics of retention can become quite complicated when additional deposition is superimposed
on clearance phenomena, especially if the deposited material affects clearance mechanisms.
Extended or chronic exposures are the general rule for environmental aerosols, and particulate
material may accumulate in some portions of the lung (Davies, 1963, 1964a; Walkenhorst, 1967;
Einbrodt, 1967).
11.3.1 Deposited Particulate Material
An understanding of regional deposition is requisite to an evaluation of respiratory
clearance and a description of the retention of deposited particulate materials. In addition,
significant differences may exist between the mechanisms of clearance in different mammalian
species. Particle deposition in the ET region is limited primarily to larger particles depo-
sited by inertia! impaction. Deposition of various aerosol particles may lead to specific
biological effects associated with this region. For particles that do not quickly dissolve or
do not react with body fluids, clearance from this region is mechanical. The anterior third
of the human nose (where most particles >5 pm may deposit) does not clear except by blowing,
wiping, sneezing, or other extrinsic means; and particles may not be removed until 1 or more
days after deposition (Proctor and Swift, 1971; Proctor et al., 1969, 1973; Proctor and Wagner,
1965, 1967).
The posterior portions of the human nose, including the nasal turbinates, have mucociliary
clearance averaging 4 to 6 mm/min, with considerable variation among individuals (Proctor and
Wagner, 1965, 1967; Ewert, 1965; van Ree and van Dishoeck, 1962). Particles are moved with
this mucus to the throat and are swallowed or expectorated. Various reactions can occur in
the gastrointestinal tract, and some assimilation into the blood is possible even for par-
ticles that were relatively insoluble in the nose. The ICRP Task Group (Morrow et al., 1966)
adopted a 4-min half-time for physical clearance from the human ET (nasopharyngeal) region
by mucociliary transport to the throat and subsequent swallowing.
11-39
-------
Soluble particles or droplets are-readily assimilated by the mucous membranes of the nose
directly into the blood. Solubility is graded from extremely insoluble to instantly soluble,
and the dissolution rate constant for an aerosol must be considered for each chemical species,
Since the TB region includes both very large and very small airways, particles of various
sizes can be deposited. The retention of deposited materials in this region can differ
markedly among individuals and can be affected by such factors as cigarette smoking, patho-
logical abnormalities, or responses to inhaled air pollutants. Clearly, the more rapid the
clearance, the less time available for untoward responses or latent injury at the site of
original deposition. In mouth breathing of aerosols, such as during smoking or under physical
exertion, the beneficial filtering of large particles in the nasal airways is lost, and a
greater fraction of these large particles can be deposited in the TB region.
An important characteristic of the TB region is that it is both ciliated and equipped
with mucus-secreting cells. Mucociliary clearance mechanisms were reviewed by Schlesinger
(1973). For relatively insoluble and inert particles, the primary clearance mechanism for the
TB region is mucociliary transport to the glottis, with subsequent swallowing and passage into
the gastrointestinal tract. Mucus flow influences the ciliary mucous conveyor (Van As and
Webster, 1972; Besarab and Litt, 1970; Dadaian et a!., 1971).
The rate of mucus movement is slowest in the finer, more distal airways and greatest in
the major bronchi and trachea. The pattern of mucus movement is complex, especially at airway
bifurcations, where whirlpools may be found that can slow the clearance of particles (Hilding,
1957). Coughing can accelerate TB clearance by the mucociliary conveyor. The size distribu-
tion of particles affects their distribution in the TB tree. The average clearance time for
small particles that preferentially deposit deep in the lung is longer than for larger par-
ticles, which tend to deposit in the larger airways (Albert et a!., 1967, 1973; Camner et a!.,
1971; Luchsinger eta!., 1968). Clearance rates for deposited materials vary considerably
among normal healthy adults (Yeates et al. 1975).
The clearance of material in the TB compartment cannot be described by a single rate.
Data from experimental studies imply that the larger airways clear with a half-time of about
0.5 h, intermediate airways with a half-time of 2.5 h, and finer airways with a half-time of 5
h (Morrow et al., 1967a; Morrow, 1973). There is also considerable variability among indi-
viduals (Camner et al., 1972, 1973a,b; Camner and Philipson, 1972; Albert et al., 1967).
Material with slow dissolution rates in the TB compartment will usually not persist longer
than about 24 h in healthy humans. Cigarette smoking has been reported under various condi-
tions to either increase, decrease, or have little effect on the efficiency and speed of TB
clearance (Camner and Philipson, 1972; LaBelle et al., 1966; Bohning et al., 1975; Albert et
al., 1974).
Particles smaller than about 10 pm D are deposited to some extent in the P region of
96
the lung on inhalation. Particles that deposit in the P region land on surfaces kept moist by
a complex liquid containing surfactants. Slowly dissolving materials that deposit in the
11-40
-------
human P region are usually retained for years. For example, McEuen and Abraham (1978) re-
ported that birefringent particle counts were significantly higher in 37 cases of pulmonary
alveolar proteinosis, both in regions of alveolar proteinosis and in perivascular and peri-
bronchiolar regions (dust retention areas), than in 13 control subjects. Out of 8619 par-
ticles, 4817 were < 1 |jm in physical diameter, 3771 were 1 to 10 urn in physical diameter, and
31 were > 10 pm in physical diameter, with 59 percent being round, 19 percent fibrous, and 22
percent irregular in shape. Although few particles larger than 10 urn D are deposited in the
ae
P region, a small number of large aeroallergen particles on the order of 25 urn D have been
ae
found in the deep lung parenchyma (Michel et at., 1977). Accumulation of pigment in the lungs
is reflective of exposure to particulate matter (Pratt and Kilburn, 1971; Sweet et a!., 1978).
Usually, relatively insoluble particles are rapidly phagocytized by pulmonary macrophages
(LaBelle and Brieger, 1961; Sanders and Adee, 1968; Green, 1971, 1974; Ferin, 1967, 1976,
1977; Camner et al., 1973a,b, 1974; Hibbs, et al., 1977; Ferin et al., 1965; Brain and
Corkery, 1977; Brain et al., 1977; Brain, 1970a). Some particles may enter the alveolar
interstitium by pinocytosis (Strecker, 1967). Some particles may be cytotoxic to alveolar
macrophages and thus influence this clearance mechanism (see Section 12.3.4.2). Migration and
grouping of macrophages laden with particles can lead to redistribution of evenly dispersed
particles into clumps and focal aggregations of particles in the deep lung. Such events have
been described in the sequence of pathological changes observed in experimentally-induced
silicosis (Heppleston, 1969). Silica particles ranging in size from less than 1 to 3 pm in
physical diameter have been found post mortem in fibrotic lesions associated with deposits of
crystalline silica (Craighead and Vallyathan, 1980). Sherwin and coworkers (1979) found an
abnormal number of birefringent particles in the lungs of seven patients in association with
early to late interstitial inflamation and fibrosis. Also, using scanning electron microscopy
and energy dispersive X-ray analysis of particles < 5 pm in physical diameter, they found
mostly silicates (especially aluminum, sodium, and potassium), with 5 to 10 percent silicon
dioxide.
Macrophages containing particles may enter the boundary region between the ciliated bron-
chioles and the respiratory ducts and then can be carried with the mucociliary flow of the TB
region. Some insoluble particles deposited in the lung are eventually trapped in the P inter- \
stitium (Strecker, 1967), impeding mechanical redistribution or removal (Felicetti et al.,
1975). Although protein molecules may pass across the air-blood barrier intact with a clear-
ance half-time of hours by pinocytotic vesicular transport (Bensch et al., 1967), there is
conflicting evidence at best on the passage of very small particles (< 10 nm in physical dia-
meter) across the air-blood barrier. For example, the data of Kanapilly and Diel (1980) on
239
the dissolution of ultrafine PuQ~ (plutonium dioxide) are in disagreement with the data and
interpretations of Raabe et al. (1978).
Another possible clearance route for migrating particles and particle-laden macrophages
is the pulmonary lymph drainage system, with translocation first to the TB lymph nodes (Thomas,
11-41
-------
1968; Lauweryns and Baert, 1977; Leeds et al., 1971). Little information is available about
the clearance rates for transfer from lung to lymph nodes in man. Estimates of half-times
range from 1 to 2 y for Pu02 in dogs and monkeys (Leach et al., 1970) to hours and days for
iron, cadmium, and lead in dogs (Oberdb'rster et al., 1978). The studies by Oberdb'rster and
coworkers (1978) indicate that both the chemical species and their physical states are
important in affecting alevolar permeation into the pulmonary lymph. Usually more than 10
percent of the material initially deposited in the P region can be recovered in the regional
lymph nodes; and since the amount of material passing through the lymph nodes is not known,
nodal values most likely underestimate the role of lymph transport in pulmonary removal
(Morrow, 1972). Because the mass of the P lymphoid tissue is only a few percent of the lung
weight, the average dose delivered to the lymphoid tissue may be orders of magnitude greater
than that delivered to lung tissue. Lymph node retention times usually exceed P retention
times, accentuating the disproportionality between lung and lymphoid dose (Morrow, 1972). In
Q
addition, Ferin and Feldstein (1978), using inhalation exposures of 15 and 100 mg Ti02/m ,
showed that the amount of material deposited in the lungs of rats can affect the fraction
cleared via the lymphatic system if the exposure level is sufficiently high enough to over-
whelm some components of host defenses. As in transfer to the TB region with clearance by the
mucociliary escalator, transfer to the lymph nodes may affect only a portion of the material
deposited in the lungs. For example, after intratracheal instillation in rats of a mixture of
3, 9, and 15 |jm physical diameter (calculated aerodynamic diameters of 3.4., 10.1, and 16.8
(Jra, respectively) polystyrene latex spheres, Snipes and Clem (1981) found that the 3 |jm
spheres translocated to the lung-associated lymph nodes, whereas the 9 and 15 |jm spheres did
not.
Waligora (1971) reported the P clearance of extremely insoluble and inert particles of
zirconium oxide radiolabeled with Nb. Although his results were not precise, the biological
clearance half-life in man was found to be about 1 y, a value about the same as for beagles.
By contrast, murine species have a more rapid pulmonary clearance (Morgan et al., 1977).
Leach et al. (1970, 1973) exposed experimental animals to insoluble PuO? (MMAD of about 3.5
urn) and observed lung retention half-times of 19.9 mo for dogs and 15.5 mo for monkeys.
Ramsden et al. (1970) measured the retention of accidentally inhaled, relatively insoluble
233pUQ -jn a man's lungs and found the clearance half-time to be about 240 to 290 days; some of
that material was dissolved into blood and excreted in the urine. Pulmonary clearance half-
times as long as 1000 days have been reported for particles of Pu02 in dogs (Raabe and Goldman,
1979). Cohen et al. (1979) reported an apparent half-time of about 100 days for nonsmokers
and about 1 y for smokers for P clearance of magnetite particles. Lung retention studies by
Snipes and Clem (1981) of 3, 9, and 15 |jm physical diameters polystyrene latex spheres in rats
after intratracheal instillation showed that half of the 3 urn spheres were retained with a
half-time of 69 days, and 24 and 76 percent of the 9 |jm spheres cleared with half-times of 17
and 580 days, respectively. In contrast, 14 percent of the 15 urn spheres cleared with a half-
time of 2.3 days and the remaining spheres were retained in the lung with a half-time that was

11-42
-------
not measurable over the course of the 106-day study. Their results indicate that large par-
ticles deposited in the P region could be retained indefinitely and yield unique dose patterns
in surrounding tissue.
Because of the slow clearance by the various mechanical pathways, dissolution and asso-
ciated physical and biochemical transformations are often the dominant mechanisms of clearance
from the P region (Morrow, 1973). The term "dissolution" is taken in its broadest context to
include whatever processes cause material in a discrete particle to be dispersed into the lung
fluids and the blood (Green, 1975), Many chemical compounds deposited in the lung in particu-
late form are mobilized faster than can be explained by known chemical properties at the
normal lung fluid pH of about 7.4 (Kanapilly, 1977). Raabe et al. (1978) suggested that the
apparent dissolution of highly insoluble PuO? actually may be due to fragmentation into par-
ticles small enough to move readily into the blood, rather than to true dissolution.
Mercer (1967) developed an analysis of P clearance based on particle dissolution under
nonequilibrium conditions. If the dissolution rate constant (k) is known for a material, the
time required to dissolve half the mass of (monodisperse) particles of initial physical
diameter (D ) is given by:

T1/2 =0.618 av pDQ/«sk (1)

with p the physical density of the particles and or and a the volume and surface shape fac-
tors, respectively (for spherical particles a /ct = 6).
The particles would be expected to be completely dissolved at a time, t~, given by:

tf = 3av p Oo/ask (2)

Mercer (1967) also calculated the expected dissolution half-time for polydisperse particles
when their mass median (physical) diameter in the lung is known'

T1/2 = 0.6 ay p(MMD)/ask (3)

Further, he showed that the resulting apparent lung retention function R(t) could be described
as the sum of two exponentials of the form:

R(t) = fie'AlP + f2

where f, = (l-f?), p = a kt/a p(MMD), and f,, fp, A,, and h^ are functions of the geometric
standard deviations as defined by Mercer (1967).
For dissolution-controlled P clearance, smaller particles will exhibit proportionately
shorter clearance half-times. When the dissolution half-times are much shorter than the half-
times associated with the translocations of particles to the TB region or to lymph nodes (i.e.,
much less than 1 y), dissolution will dominate retention characteristics. Materials usually
11-43
-------
thought to be relatively insoluble (such as glass) may have high dissolution rate constants
and short dissolution half-times for the small particles found in the lung; the dissolution
half-time for 1 pm D glass spheres is about 75 days (Raabe, 1979). Changes in structure or
36
chemical properties, such as by heat treatment of aerosols (Raabe, 1971), can lead to
important changes in dissolution rates and observed P retention.
Usually the retention time of material in the respiratory tract is measured (such as with
radiolabeled aerosols) rather than the clearance rates (Sanchis et al. , 1972; Camner et al.,
1971; Edmunds et al., 1970; Luchsinger et al., 1968; Aldas et al., 1971; Ferin, 1967; Barclay
et al,, 1938; Morrow et al., 1967a,b; Friberg and Hoi ma, 1961; Hoima, 1967; Kaufman and
Gamsus, 1974). The lung burden or respiratory tract burden can be represented by an 'appro-
priate retention function with time as the independent variable (Morrow, 1970a,b). For models
based on simple first-order kinetics, the lung burden, y, at a given time during exposure is
controlled by the instantaneous equation:
dt
= E - X-y (5)
where E is the instantaneous deposition rate of particulate material deposited in the lung per
unit time during an inhalation exposure and A, is the fraction of material in the lung cleared
from the lung per unit time (Raabe, 1967). For an exposure that lasts a time t , the lung
burden from the exposure is given by:
_ \ 4-
ye = (E - Ee ! e)A1 (6)

where E in this case, is the average exposure rate. After the exposure ends, the clearance is
governed by:

dy. _ _ A,y (7)
dt ~ A

and the lung burden is given by:

-A,t
y = ye e l (8)

where y is the lung burden at the end of the exposure period (t ). Hollinger et al. (1979)
used this simple model to describe the deposition and clearance of inhaled submicrometer ZnO
(zinc oxide) in rats (Figure 11-12) where the concentration of zinc (as Zn) in the lungs (as
described by equations 6 and 8) is superimposed on the natural background concentration of
zinc in lung tissue. The normally" insoluble zinc has only a 4,8-h dissolution half-time
11-44
-------
a.
LU

ts>
$2
H
I-

CD
EC
a
o

N
CD
1000:

800

600

400

300

200
I I I I
EXPOSURE PERIOD
I I I
CLEARANCE PERIOD
too
80

50
II I I I -
I
DATA VALUE ±SE
N=3
(SE)

~\
\
\Tl
\
\

4 = 4.8±0.3(SE), hr

I M I I
-20 -15 -10
0 5 10 15 20 25

POST EXPOSURE TIME, hr
30 35
40
45
50
Figure 11-12, Single exponential model, fit by weighted least-squares of the buildup (based on text
equation 7} and retention (based on text equation 9) of zinc in rat lungs.

Source: Hollingeretal. (1979).
11-45
-------
(A.-^ = 0.21 h ) for this aerosol. Of course, environmental aerosol exposures are likely to
continue so that a steady state lung burden may be expressed by:

yss = EA1 (9)
If several deposition and clearance regions, subregions, or special pools are involved, a
more complicated multicompartmental model may be required to describe lung or respiratory tract
buildup and retention of inhaled aerosols. If each compartment can be described by first order
kinetics, a general model can be specified by 1) subscripting E, A, and Y with the subscript i
whenever they appear on the right-hand side of equations 6, 8, and 9, and 2) performing a sum-
mation over i from one to the number of compartments. Each of the A. values translates to a
clearance rate for each of the compartments given by half-time T, ,„ = In 2/A..
For chronic exposures where the several pools are in complex arrays of change, a simple
power function may serve as a satisfactory model of P retention (Downs et al., 1967). In such
a model, the P region is treated as one complex, well-mixed pool into which material is added
and removed during exposure, as given by the instantaneous equation:

$ = E - Apy/t [y = 0 at t = 0] (10)

where y is the total lung burden at a given time, t, E is the average deposition rate of in-
haled particulate material in the lung, and A is the fraction of available lung burden being
cleared. Unlike the A. of the exponential retention models, A is dimensionless. The time
coordinate is not arbitrary; time is taken as zero only at the beginning of the inhalation
exposure, when the lung burden is nil. Thus, during an exposure lasting until time (t ), the
P burden (y ) is given by (Raabe, 1967):

ye = Ete/(Ap + 1) (11)

On this basis, no steady-state concentration is reached even though clearance is progressing
and the lung concentration continues to increase during chronic exposures to environmental
aerosols. This model is therefore not applicable to relatively soluble species. The lung
burden, y, after the exposure has ended for a time, t , is given by (Raabe, 1967):

y = ye te p t P = At p [t = te + tp]. (12)

Deposited particulate material cleared from the lung> is usually transformed chemically
and transferred to other tissues of the body. The injurious properties of a toxic material
translocated from the lung may therefore be expressed in other organs. Identification of the
potential hazards associated with inhalation exposures to toxicants is compounded when the
11-46
-------
respiratory tract is not the only target for injury but still serves as the portal of entry
into the body. The metabolic behavior and excretion of inhaled toxicants after deposition in
the lung may define the probable target organs and indicate potential pathogenesis of result-
ing disease.
Multicompartmental models that describe biological behavior can become extremely complex.
Each toxicant or component of aerosol particles deposited in the respiratory tract may need to
be described by a separate rate constant and pool or compartment. A general model of the
metabolic behavior of inhaled particles developed by Cuddihy (1969) identified 39 different
places where rate constants may need to be determined. In this general model, the P region of
the lung is visualized as consisting of three independent clearance compartments, and the par-
ticles are presumed to be converted from their original participate state to some other physi-
cochemical form or transformed state prior to clearance from the respiratory tract. Such a
transformed state can be used to describe, for example, the behavior of hydrolyzable aerosols
in the respiratory tract.
11.3,2 Absorbed SulfurDioxide
Sulfur dioxide coming in contact with the fluids lining the airways (pH 7,4) should dis-
2-
solve into the aqueous fluid and form some bisulfite (HSO-) and considerable sulfite (SO- )
anions. Because of the chemical reactivity of these anions, various reactions are possible,
leading to the oxidation of sulfite to sulfate (see Section 12.2.1).
Clearance of sulfite from the respiratory tract may involve several intermediate chemical
reactions and transformations (see Section 12.2.1.2). Gunnison and Benton (1971) identified
S-sulfonate in blood as a reaction product of inhaled S0«. The reaction rate is rapid, if not
nearly instantaneous, so that there is no long-term clearance to characterize. Intermediate
and potentially toxic products may be formed, however. These products may have residence
times that are long enough to demonstrate an elevation of the sulfur content of the lung.
Oesorption from the upper respiratory tract may be expected whenever the partial pressure
of SOp on mucosal surfaces exceeds that of the air flowing by. Desorption of SO- from mucosal
surfaces was still evident after 30 min of flushing with ambient air the airways of dogs that
had breathed 2,62 mg/m3 (1.0 ppm) for 5 min (Frank et a!., 1969). Frank et al. (1967)
reported S0» in the lungs of dogs that apparently was carried by the blood after nasal deposi-
3
tion. In human subjects breathing 42.2 mg/m (16.1 ppm) through a mask for 30 min, 12 percent
of the SOp taken up by the tissues in inspiration reentered the airstream in expiration and
another 3 percent was desorbed during the first 15 min after the end of S02 exposure (Speizer
and Frank, 1966). Thus, during expiration, SOp was desorbed from the nasal mucosa in quanti-
ties totaling approximately 15 percent of the amount originally inspired.
The effects of S02 on TB clearance in nine healthy, nonsmoking adults were studied by
Wolff et al. (1975) (see Section 13.2.3.5). Technetium (Tc) 99 m albumin aerosol (3 urn MMAD,
a = 1.6) was inhaled as a bolus under controlled conditions. A 3-h exposure to 13.1 mg
11-47
-------
SOp/m (5.0 ppm) had no significant effect on mucociliary clearance in resting subjects,
except for a small transient increase (p < 0.05) after 1 hour. A significant decrease in nasal
mucus flowrates during a 6-h exposure of 15 young men to 13.1 mg S00/m (5.0 ppm) and 65.5 mg
3 3
S02/m (25.0 ppm), but not 2.62 mg S02/m (1.0 ppm), was observed by Andersen et al. (1974).
Decreases were greatest in the anterior nose and in subjects with initially slow mucus flow
rates. Newhouse et al. (1978) assessed the effect of oral exposure to S02 on bronchial clear-
ance of a radioactive aerosol (3 |jm MMAD) in healthy nonsmoking males and females who exer-
cised periodically during exposure at an exertion level sufficient to keep the heart rate at
o
70 to 75 percent of the predicted maximum. After a 2-h exposure to 13.1 mg SCL/m (5.0 ppm),
clearance was increased.
11.3.3 Particles and Sulfur Dioxide Mixtures
The presence of adsorbed SO- or other sulfur compounds on aerosol surfaces may alter the
clearance processes of both. Chemical reactions involving sulfur compounds on particle sur-
faces may enhance the apparent solubility of the aerosol particles. These aerosol particles
may also undergo reaction with sulfite or other species on contact with body fluids.
The formation of sulfate anions by oxidation of S0? to S0_ may be catalyzed by manganese,
iron, or other aerosol components. The SO,, reacts immediately with water to form H-SO.
that can react with other materials, such as metal oxides on fly ash aerosols, to produce
sulfate compounds. Since sulfate is a normal constituent of body fluids (Kanapilly, 1977),
the clearance of sulfate anions probably involves simple dissolution into body fluids.
11.4 AIR SAMPLING FOR HEALTH ASSESSMENT
The objective of air sampling in relation to health assessment is to obtain data on the
nature and extent of potential health hazards resulting from the inhalation of airborne parti-
cles. To be effective, the techniques used in air samplers must be based on a recognition of
the size-selecting qharacteristies of the human respiratory tract (see Section 11.2). The
usual variables affecting the selection of methods, such as the physical limitations of the
collection process and the sensitivity and specificity of the analytical procedures, must
still be addressed.
An increasing recognition of the importance of the selective sampling of "respirable"
dusts has occurred in recent years. The commonly measured index of gross air concentration
provides a crude and sometimes misleading indication of health hazard. Since most aerosols
are polydisperse, with a a >2, the mass median size approaches the diameter of the largest
particles in the sample, resulting in a relatively few large particles strongly influencing
the value reported for the mass concentration. The measured total mass concentration then
will not relate to the inhalation hazard if these particles are not inhaled. Also, the true
total airborne mass concentration may be underestimated when the aerosol contains very large
particles, since every sampler has its own characteristic upper size cutoff. This cutoff is
dependent on its entry shape, dimensions, and flowrate.
11-48
-------
The best dose estimates for a substance whose toxicity is proportional to absorbed mass
are obtained from information on the mass concentrations within various size ranges. Lippmann
(1978) cited several ways such data can be obtained: (1) During the process of collection,
separate the aerosol into size fractions that correspond to anticipated regional deposition;
(2) analyze the size distribution of the airborne aerosol; and (3) analyze the size distribu-
tion of the collected sample. The most reliable and useful information is obtained using
methods of fractionation based on aerodynamic diameters similar to the way fractionation
occurs within the respiratory tract, thereby automatically compensating for differences in
particle shape and density.
Many recent advances have been noted in the technologies needed to develop samplers that
will separate particles during the process of collection into "respirable" and "nonrespirable"
fractions. The absence of uniform criteria for "respirable" mass concentrations has been a
major factor limiting the application of selective sampling concepts in the United States.
Regulations established on the basis of only gross concentration limits do not promote field
measurements of "respirable" mass.
The recommendations by Miller et al. (1979) on size considerations for establishing a
standard for "inhalable" particles indicate a possible future departure from the current
approach to the setting of a particulate standard in the United States. Also, the recent
report on respirable dust by the International Standards Organization ad hoc Group to TC 146
(1981) contains recommendations for size definitions for particle sampling for the healthy
normal segment of the population and high-risk subpopulations. A perspective on these recent
events can be obtained by examining the development of the field of respirable dust sampling.
In 1952, the British Medical Research Council (BMRC) adopted a definition of "respirable
dust" which essentially considered respirable dust to be that dust reaching the alveoli, there-
by making "respirable dusts" applicable to pneumoconiosis-producing dusts. The horizontal
elutriator was chosen as a particle size selector, and respirable dust was defined as that
dust passing an ideal horizontal elutriator. The elutriator cutoff was chosen to result in
the best agreement with experimental lung deposition data. The Johannesburg International
Conference on Pneumoconiosis in 1959 adopted the same standard (Orenstein, 1960).
In January 1961, at a meeting in Los Alamos sponsored by the Atomic Energy Commission
(AEC) Office of Health and Safety, a second standard was established, which defined "Respir-
able Dust" as that portion of the inhaled dust which penetrates to the nonciliated portions of
the lung (Hatch and Gross, 1964). This definition was not intended to be applicable to dusts
that are readily soluble in body fluids or are primarily chemical intoxicants, but rather only
for "insoluble" particles that exhibit prolonged retention in the lung. Criteria for respira-
bility were such that all 2 pm D particles were considered respirafale and particles 10 urn
ae
D were considered to be nonrespirable.
11-49
-------
Other groups, such as the American Conference of Governmental Industrial Hygienists
(ACGIH), have incorporated respirable dust sampling concepts in setting acceptable exposure
levels for other toxic dusts. Such applications are more complicated, since animal and human
exposure data, rather than predictive calculations, form the data base for standards. The
size-selector characteristic specified in the ACGIH standard for respirable dust (Threshold
Limits Committee, 1968) is almost identical to that of the AEC, differing only at 2 pm D ,
61 ©
where it allows for 90 percent passing the first-stage collector instead of 100 percent. The
difference between them appears to be a recognition of the properties of real particle separa-
tors, so that, for practical purposes, the two standards may be considered equivalent
(Lippmann, 1978).
The sampler acceptance criteria of the BMRC and of the ACGIH and the P deposition curves
from Figure 11-9 are shown in Figure 11-13. The cutoff characteristics of the precollectors
preceding respirable dust samplers are defined by these criteria. The two sampler acceptance
curves have similar, but not identical, characteristics, due mainly to the use of different
types of collectors. The BMRC curve was chosen to give the best fit between the calculated
characteristics of an ideal horizontal elutriator and available lung deposition data; on the
other hand, the design for the AEC curve was based primarily on the upper respiratory tract
deposition data of Brown et al. (1950). The separation characteristics of cyclone type
collectors simulate the AEC curve. Whenever the particle size distribution has a a >2,
samples collected with instruments meeting either criterion will be comparable (Lippmann,
1978). Various comparisons of samples collected on the basis of the two criteria are avail-
able (Knight and Lichti, 1970; Breuer, 1971; Maguire and Barker, 1969; Lynch, 1970; Coenen,
1971; Moss and Ettinger, 1970).
The various definitions of respirable dust are somewhat arbitrary, with the BMRC and AEC
definitions being based on the "insoluble" particles that reach the P region. Since part of
the aerosol that penetrates to the alveoli remains suspended in the exhaled air, respirable
dust samples are not intended to be a measure of P deposition but only a measure of aerosol
concentration for those particles that are the primary candidates for P deposition. Given
that the "respirable" dust standards were intended for "insoluble dusts," most of the samplers
developed to satisfy their criteria have been relatively simple two-stage instruments. In
addition to an overall size-mass distribution curve, multistage aerosol sampler data can pro-
vide estimates of the "respirable" fraction and deposition in other functional regions. Field
application of these samplers has been limited because of the increased number and cost of
sample analyses and the lack of suitable instrumentation. Many of the various samplers, along
with their limitations and deficiencies, were reviewed by Lippmann (1978).
After analyzing industrial health data, Morgan (1978) concluded that different sites of
deposition were sufficient to explain the lack of association between pneumoconiosis and bron-
chitis in coal workers. He found that particles less than 5 urn D_ were associated more with
ae
11-50
-------
SAMPLER ACCEPTANCE CRITERIA

— —— ACGIH

VIA NOSE

_. .— BMRC
0.1 02 0.4 0.50.6 0.8 1.0

PHYSICAL DIAMETER, jUin—*
2.0 4.0 6.0 8.0 10.0

AERODYNAMIC DIAMETER, ^m.
20.0
Figure 11-13. Comparison of sampler acceptance curves of BMRC and ACGIH conventions with the
band for the experimental pulmonary deposition data of Figure 11-9.
11-51
-------
P deposition, as compared to 5 to 10 urn D particles, which deposited primarily in areas
etc
above the gas-exchange region with nose breathing. In addition, many large particles between
10 and 20 urn D=0 vtsre deposited in the trachea and bronchi with mouth breathing. He suggested
oG
that when studies are designed for the purpose of relating biological and environmental
measurements, it would be wise to measure not only "respirable," but also total dust. Size
definitions for participate sampling which expand the area of concern beyond just "insoluble"
dust penetrating to the P region were advanced by Miller et al. (1979) and recently by the
International Standards Organization ad hoc Group report to TC 146 (1981). As knowledge of
the regional deposition of particles increases through experimental studies, such as those
discussed in Section 11.2, it is logical to envisage using samplers that broadly simulate the
relative collection efficiencies of the major regions of the respiratory tract. These devices
would first select from the total airborne material the inspirable fraction and then sequenti-
ally divide this fraction into ET, TB, and P fractions.
Such a division into these three regions (vide supra) was recommended in the Inter-
national Standards Organization ad hoc Group report to TC 146 (1981). Various options being
considered are the at risk population (healthy adults, children, and sick and infirm indi-
viduals) and the use of 10 or 15 urn D as the 50 percent cut-point for material penetrating
3S
to the TB region. Their division of the thoracic fraction into the P and TB fractions, where
the target population is healthy adults, is shown in Figure 11-14 using 15 Mm D ^ as the 50
* csG
percent cut-point for the total thoracic fraction. The 50 percent cut-point refers to the
aerodynamic diameter for which 50 percent of the particles that enter the mouth or nose are
considered to pass the larynx. Thus, the material not passing the larynx forms the ET frac-
tion, which includes the oral pharynx. Particles larger than 15 urn D can enter and be de-
36
posited in the ET region. If any of these larger particles are readily soluble, they will be
absorbed into the bloodstream just as quickly as smaller particles, with one 20 pm D=rt par-
36
tide contributing as. much to the systemic dose as a thousand 2 pm D_^ particles.
3B
Also shown in Figure 11-14 are bands for the experimental P deposition data of Figure
11-9 and for TB deposition as a percent of particles entering the mouth. The band for TB
deposition was derived using the overall regression line of Chan and Lippmann (1980) for ET
deposition with mouth breathing and their equation for TB deposition, which was evaluated at
bronchial deposition size values one standard deviation from the mean for an average inspi-
ratory flowrate of 30 1/min. The range of values demonstrates the variability among indi-
viduals for TB deposition and illustrates the uncertainties as to the particle size corre-
sponding to maximum deposition and the paucity of deposition data for large particle sizes.
The ad hoc group basically followed the BMRC and ACGIH conventions for P deposition in
healthy adults, although their definition of the P fraction differs slightly because it is
11-52
-------
r-r-JJ
1 MUM
ACGIH COIS1V,
——— SMRCCONV. —
—™— PULMONARY VIA
NOSE
PULMONARY VIA —
MOUTH
TRACHEOBRONCHIAL
PULMONARY FRACTION
TRACKED
BRONCHIAL
FRACTION
0
0.1 0.2 0.3

PHYSICAL DIAMETER, j
1JQ 2 3457

- AERODYNAWIFC DIAMETER, pm
20 30 40 50
Figure 11-14. Division of the thoracic fraction of deposited particles into pulmonary
and tracheobronchial fractions for two sampling conventions (ACGIH and BMRC) as
a function of aerodynamic diameter, except below O.B^m, where physical diameter is
used (International Standards Organization, 1981). Also shown are bands for ex-
perimental pulmonary deposition data from Figure 11-9 and for tracheobronchial (TB)
deposition as a percent of particles entering the mouth. The band for TB deposition
was derived using the overall regression line of Chan and Lippmann (1980) for ex-
trathoracic deposition with oral breathing and their equation for TB deposition,
which was evaluated at bronchial deposition size values one standard deviation
from the mean, given an average inspiratory flow rate of 30 liters per minute. The TB
band is shown up to about the largest particle size used by Chan and Lippmann
(1980J.
Sources: ACGIH (Threshold Limits Committee, 1968); BMRC (Orenstein, 1960);
Pulmonary via nose (Lippmann, 1977); Pulmonary via mouth (see Figure 11-9);
Tracheobronchial via mouth (Chan and Lippmann, 1980).
11-53
-------
defined as a fraction of the inspirable material rather than the total aerosol. A "high-risk"
selection curve for children or the sick and infirm used a 50 percent cut-point at 2.5 instead
of 3,5 |.im D in recognition of the fact that a similar shift is seen in lung deposition in
these groups (Lippmann, 1977), Taken to its ultimate, by using size selective samplers that
separate inspirable material into ET, TB, and P components, standards for airborne particles
could specify which of these regional fractions should be measured, taking into account the
biological effects of the material, and in the case of the TB and P fractions, the population
at risk.
11.5 SUMMARY
Besides being a target of inhaled particles and gases, the respiratory tract is also the
portal of entry by which other organs may be affected (see Section 11.3). An understanding of
the mechanisms and patterns of translocation to other organ systems is required for evaluation
of the potential for injury or response in those organs. When SO,, or aerosols are inhaled by
humans or experimental animals, different fractions of the inhaled materials deposit by a
variety of mechanisms in various locations in the respiratory tract. Particle size distribu-
tion, particle chemical properties, physicochemical properties of S0?, respiratory tract
anatomy, and airflow patterns all influence the deposition. The three functional regions
[extrathoracic (ET) or nasopharyngeal, tracheobronchial (TB), and pulmonary (P)] of the respi-
ratory tract can each be characterized by major mechanisms of deposition and clearance.
Impaction, gravitational settling, and diffusion predominate for the deposition of most
types of particles in the respiratory tract, with electrostatic attraction and interception
being of relatively minor importance. Diffusivity and interception potential of a particle
depend on its geometrical size, whereas the inertia! properties of settling and impaction
depend on its aerodynamic diameter. Gravitational settling accounts for the deposition of
particles in the TB and P regions, and impaction contributes to deposition in the ET and TB
regions. Diffusion primarily affects respiratory tract deposition of particles with physical
diameters smaller than 1 urn. The major processes affecting the transport of SCL in the
respiratory tract are convection, diffusion, and chemical reactions. The rapid diffusivity of
S0? in combination with its high solubility in body fluids is responsible for the large re-
moval of SO, in the ET region and upper generations of the TB tree.
After deposition, inhaled particles will be translocated by processes that depend on
their character and site of deposition. The anterior third of the human nose does not clear
except by blowing, wiping, sneezing, or other extrinsic means, and particles may not be re-
moved until 1 or more days after deposition. If the particles are quite soluble in body
fluids, they will readily enter the bloodstream. Relatively insoluble material that lands on
ciliated epithelium, either in the ET region or TB airways, will be translocated with mucus
flow to the throat and will be swallowed or expectorated. Depending on particle size, rela-
tively insoluble material that deposits on nonciliated surfaces in the P region may be
phagocytized, may enter the interstitium and remain in the lung for an extended period, or may
11-54
-------
be translocated by phagocytic cells, blood, or lymphatic drainage. Some material from the P
region may enter the TB region and then be cleared by the mucociliary conveyor. Dissolution can
contribute to the clearance of particles in all regions of the respiratory tract.
Nose breathing and mouth breathing provide somewhat contrasting deposition patterns for
some respiratory tract regions. With nose breathing, nearly complete respiratory tract deposi-
tion can be expected for particles larger than about 4 urn D . Since mouth breathing bypasses
30
much of the filtration capabilities of the ET region, there is a shift upward to about 10 pm
D before there is complete deposition of inhaled particles. Given the three general regions
ae i^
into which the respiratory tract can be divided on the basis of anatomical structure, function,
particulate retention times, and clearance pathways, however, regional deposition data for
particles of various aerodynamic diameters are more useful than total respiratory tract deposi-
tion information.
Particles about 10 urn D or larger are deposited in the ET region during nose breathing,
3G
as compared with about 65 percent deposition of 10 pm 0 particles under conditions of'mouth
36
breathing. On the-other hand, for both routes of breath.ing, ET deposition of particles
smaller than about 1 urn D is slight. The increased penetration of larger particles deeper
36
into the respiratory tract when a person breathes through the mouth is reflected by experi-
mental deposition data showing that TB deposition of 8 to 10 urn D particles is on the order
ae
of 20 to 30 percent. Also, about 10 percent of particles as large as 15 pm D are predicted
3G
to enter the TB region during mouth breathing.
For nose breathing, as compared with mouth breathing, the peak of the P deposition curve
shifts downward from 3.5 to about 2.5 pm D . Also, the peak is much less pronounced (about
36
25 percent compared with about 50 percent for mouth breathing), with a nearly constant
pulmonary deposition of about 20 percent for all sizes between 0.1 and 4 urn D .
36
The deposition data cited above are based on studies in which healthy young adult sub-
jects usually were used. Although children are usually considered to be a subpopulation more
susceptible to the effects of environmental pollutants, deposition data for children are not
currently available or likely to be obtained soon. The few data available on other subpopula-
tions, such as asthmatics and chronic bronchi tics, indicate that T6 deposition appears to be
enhanced at the expense of P deposition in most abnormal states. Partial or complete airway
obstruction in bronchitis, lung cancer, emphysema, fibrosis, and atelectasis may decrease or
eliminate the deposition of particles in some regions of the lungs.
Regional deposition studies of particles less than 3 urn D__ have been conducted using
36
dogs and some rodents. In these species, the relative distribution among the respiratory
regions of particles less than 3 pm D, during nose breathing follows a pattern that is
ae »
similar to regional deposition in humans during nose breathing. Thus, in this instance, the
use of rodents or dogs in toxicologies! research for extrapolation to humans entails differ-
ences in regional deposition of insoluble particles that can be reconciled from available
data.
11-55
-------
When breathing through the nose under resting conditions, SO, removal by nasal absorption
is nearly complete in both humans and laboratory animals. Expired air acquires SCL from nasal
mucosa, with small amounts of SO,, continuing to be released after cessation of exposure. Ex-
traction of S0? by the total respiratory tract during mouth breathing is significantly lower
than during nose breathing, although regional uptake has not been studied in humans during
mouth or oronasal breathing. On the other hand, studies in which SO, was passed through the
surgically-isolated ET airways of dogs showed that S0? absorption in the El region can be de-
creased to less than 50 percent by mouth breathing at elevated airflow rates. Sulfur dioxide
may also enter into a variety of gas-to-particle conversions or gas-particle chemical reac-
tions. As a consequence of these reactions with particles, S0~ can be carried deeper into the
respiratory tract, thereby increasing the potential for adverse effects.
Both deposition and retention play roles in determining the effects of inhaled particu-
late toxicants and SO,. Everyone is environmentally exposed to a variety of dusts, fumes,
sprays, mists, smoke, photochemical particles, and combustion aerosols, as well as SO^ and
other potentially toxic gases. The particle size distribution and chemical and physical com-
position of airborne paniculate material require special attention in toxicological evalua-
tions, since a wide variety of physicochemical properties may be encountered in both experi-
mental and ambient inhalation exposures. The need to characterize the aerosols to which indi-
viduals are exposed requires the development of appropriate air sampling techniques so that
potential health hazards can be identified. For insoluble dusts whose site of action is the P
region, inhalation hazard evaluations based on "respirable" mass are clearly superior to esti-
mates based on gross air concentrations. Appropriate selective sampling procedures can and
are being developed to provide more meaningful data on inhalation hazard potential for par-
ticles as a function of their regional deposition in the respiratory tract. Gross concentra-
tion sampling techniques are appropriate for highly soluble aerosols or where the particle
size distribution is relatively constant. They can also be used if the particle size distribu-
tion is relatively constant and a known fixed ratio between the gross concentration and the
concentration in the size range of interest exists.
11-56
-------
11.6 REFERENCES

Aharonson, E. F. Deposition and retention of inhaled gases and vapors. In: Air Pollution
and the Lung. E. F. Aharonson, A. Ben-David, and M. A. Klingberg, eds. , John Wiley and
Sons, New York, NY, 1976. pp. 13-24.

Aharonson, E. F. , H. Menkes, G. Gurtner, D. L. Swift, andvD. Fv Proctor. Effect of respiratory
airflow rate on removal of soluble vapors by the' nose. J. Appl. Physio!. ^7:654, 1974.

Albert, R. E. , M. Lippmann,, J. Spiegelman, C. Strehlow, W. Briscoe, P. Wolfson,' and N. Nelson.
The clearance of radioactive particles from the human lungs. In: Inhaled Particles and
Vapours II. C. N. Davies, ed., Pergamon Press, Oxford, England, 1967. p. 361.

Albert, R. E. , J. R. Spiegelman, M. Lippmann, and R. Bennett. The characteristics of bronchial
clearance in the miniature donkey. Arch. Environ. Health 17:50-58, 1968.

Albert, R. E. , J. R. Spiegelman, S. Shatsky, and M. Lippmann. The effect of acute exposure to
cigarette smoke on bronchial clearance in the miniature donkey. Arch. Environ. Health
18:30-41, 1969.

Albert, R. E. , M. Lippmann, H. T, Peterson, Jr., J. Berger, K. Sanborn, and D. Bohning.
Bronchial deposition and clearance of aerosols. Arch. Intern. Med. 131:115-127, 1973.

Albert, R. E. , J. Berger, K. Sanborn, and M. Lippmann. Effects of cigarette smoke components
on bronchial clearance in the donkey. Arch. Environ. Health 29:96-101, 1974.

Aldas, J. S. , M. Dolovich, R. Chalmers, and M. T. Newhouse. Regional aerosol clearance in
smokers and nonsmokers. Chest 59:25, 1971.

Altshuler, B. The role of the mixing of intrapulmonary gas flow in the deposition of aerosols.
In: Inhaled Particles and Vapours. C. N. Davies, ed. , Pergamon Press, Oxford, England,
1961. p. 47.

Altshuler, B. , L. Yarmus, E. Palmes, and N. Nelson. Aerosol deposition in the human respira-
tory tract. AMA Arch. Ind. Health 15:293, 1957.

Altshuler, B. , E. D. Palmes, and N. Nelson. Regional aerosol deposition in the human respira-
tory tract. In: Inhaled Particles and Vapours II. C. N. Davies, ed. , Pergamon Press,
Oxford, England, 1967. p. 323.

Amdur, M. 0., and D. Underbill. The effect of various aerosols on the responses of guinea
pigs to sulfur dioxide. Arch. Environ. Health 16:460-468, 1968.

American Heart Association, Committee on Exercise. Exercise testing: performance and inter-
pretation. Ind. Med. Surg. 42:20-28, 1973.

Andersen, I., G. R. Lundquist, P. L. Jensen, and D. F. Proctor. Human response to controlled
levels of sulfur dioxide. Arch. Environ. Health 28:31-39, 1974.

Bake, B. , L. Wood, B. Murphy, P. Macklem, and J. Milic-Emili. Effect of inspiratory flow rate
on regional distribution of inspired gas. J. Appl. Physio!. 37:8, 1974.

Balis, J. V., S. A. Shelley, M. J. McCue, and E. S. Rappaport. Mechanisms of damage to the
lung surfactant system. Exp. Molec. Pathol. 14:243, 1971.

Barclay, A. E., K. J. Franklin, and R. G. Macbeth. Roentgenographic studies of the excretion
of dusts from the lungs. Am. J. Roentgenol. Radium Tber. 3_9:673, 1938.
11-57
-------
Bartlett, D., J. E. Remmers, and H. Gautier. Laryngeal regulation of respiratory airflow.
Resp. Physiol. 18:194, 1973.

Batchelor, G. K. Symmetrical contraction on isotropic turbulence. In: The Theory of
Homogeneous Turbulence. Cambridge University Press, London, England, 1953. p. 74.

Beeckmans, J. B. The deposition of aerosols in the respiratory tract. Can. J. Physiol.
Pharmacol. 43:157, 1965.

Bell, K. A. Local particle deposition in respiratory airway models. In: Recent Development
in Aerosol Science. D. T. Shaw, ed. , John Wiley and Sons, New York, NY, 1978. pp. 97-
134.

Bell, K., and S. Friedlander. Aerosol deposition in models of a human lung bifurcation. Staub-
Reinhalt. Luft 33:183, 1973.

Bensch, K. G. , E. Dominguez, and A. A. Liebow. Absorption of intact protein molecules across
the pulmonary air-tissue barrier. Science 15_7:1204-1206, 1967.

Besarab, A., and M. Litt. Model studies on the adhesive properties of mucus and similar
polymer solution. Arch. Intern. Med. 126:504, 1970.

Blank, M. , A. B. Goldstein, and B. B. Lee. The surface properties of lung extract. J.
Colloid. Interface Sci. 29:148, 1969.'

Bohning, D. E. , R. E. Albert, M. Lippmann, and W. M. Foster. Tracheobronchial particle
deposition and clearance. A study of the effects of cigarette smoking in monozygotic
twins. Arch. Environ. Health 30:457, 1975.

Brain, J. D. Free cells in the lungs--some aspects of their role, quantisation and regulatory.
Arch. Intern. Med. 126:477-487, 1970a.

Brain, J. D. The uptake of inhaled gases by the nose. Ann. Otol. Rhinol. Laryngol. 79:529-539,
1970b.

Brain, J. D. , and G. C. Corkery. The effect of increased particles on the endocytosis of
radiocolloids by pulmonary macrophages _iri vivo: competitive and toxic effects. In:
Inhaled Particles IV, Part 2. W. H. Walton, ed., Pergamon Press, Oxford, England, 1977.
pp. 551-564.

Brain, J. D., J. J. Godleski, and S. P. Sorokin. Quantification, origin and fate of pulmonary
macrophages. In: Respiratory Defense Mechansims. J. D. Brain, D. F. Proctor and L. M.
Reid, eds., Marcel Dekker, New York, NY, 1977. pp. 849-892.

Breuer, H. Problems of gravimetric dust sampling. Ir\: Inhaled Particles III. W. H. Walton,
ed., Unwin Bros., London, England, 1971. pp. 1031-1042.

Brown, J. H. , K. M. Cook, F. G. Nex, and T. Hatch. Influence of particle size upon the
retention of particulate matter in the human lung. Am. J. Public Health 40:450, 1950.

Camner, P., and K. Philipson. Tracheobronchial clearance in smoking-discordant twins.
Arch. Environ. Health 25:60-63, 1972.

Camner, P. and B. Bakke. Nose or mouth breathing? Environ. Res. 21:394-398, 1980.

Camner, P., K. Philipson, L. Friberg, and B. Holma. Human tracheobronchial clearance studies.
Arch. Environ. Health 22:444, 1971.
11-58
-------
Camner, P., K. Philipson, and L. Friberg. Tracheobronchial clearance in twins. Arch.
Environ. Health 24:82, 1972.

Camner, P., P. Hellstrom, and M. Lundborg. Coating 5 mm particles with carbon and metals for
lung clearance studies. Arch. Environ. Health 27:331, 1973a.

Camner, P., P. Hellstrom, and K. Philipson. Carbon dust and mucociliary clearance. Arch.
Environ. Health 26:294, 1973b.

Camner, P., M. Lundborg, and P. Hellstrom. Alveolar macrophages and 5 mm particles coated
with different metals. Arch. Environ. Health 29:211, 1974.

Chan, T. L., and M. Lippmann. Experimental measurements and empirical modelling of the
regional deposition of inhaled particles in humans. Am. Ind. Hyg. Assoc. J. 41:399-409,
1980.

Chan, T. L. , M. Lippmann, V. R. Cohen, and R. B. Schlesinger. Effect of electrostatic charge
on particle deposition in a hollow cast of the human larynx-tracheobronchial tree. J.
Aerosol Sci. 9:463, 1978.

Charlson, R. J., D. S. Covert, T. V. Larson, and A. P.t Waggoner. Chemical properties of
tropospheric sulfur aerosols. Atmos. Environ. 12:39,' 1978.

Cheng, Y. S. , and C. S. Wang. Inertia! deposition of particles in a bend. J. Aerosol Sci.
6:139, 1975.

Cinkotai, F. F, Fluid flow in a model alveolar sac. J. Appl. Physiol 37:249, 1974.

Clement, J., M. Afschrift, J. Pardens, and K. Van de Woestline. Peak expiratory flow rate and
rate of change of pleural pressure. Resp. Physiol. 18:222, 1973.

Clements, J. A., and D. F. Tierney. Alveolar instability associated with altered surface
tension. In: Handbook of Physiology. W. D. Fenn and H. Rahn, eds. , Section 3, Respira-
tion, vol. II, chapter 69 (American Physiological Society), Washington, DC, 1965. pp.
15-84.

Coenen, W. Berechnung von Umrechnungsfaktorem ,fur verscheidene Feinstaubmessverfahren.
[Estimation of conversion factors for different fine dust measurement methods.] In:
Inhaled Particles III. W. H. Walton, ed. , Unwin Bros., London, England, 1971. pp.
1045-1050.

Cohen, D., S. F. Arai, and J. D. Brain. Smoking impairs long-term dust clearance from the
lung. Science, 204:514-516, 1979.

Cohen, V. R. The effects of glyceryl guaiacolate on bronchial clearance in patients with
chronic bronchitis. M.S. Thesis, New York University, New York, NY, 1977.

Corn, M., N. Kotsko, and D. Stanton. Mass-transfer coefficient for sulphur dioxide and
nitrogen dioxide removal in cat upper respiratory tract. Ann. Occup. Hyg. 19:1, 1976.

Craighead, F. E., and N. V. Vallyathan. Cryptic pulmonary lesions in workers occupationally
exposed to dust containing silica. J. Am. Med. Assoc. 244: 1939-1941, 1980.

Cuddihy, R. G. Analog simulation of the biological behavior of inhaled radionuclides. In:
Fission Product Inhalation Program Annual Report 1968-1969. LF-41, Lovelace Foundation,
Albuquerque, NM, 1969. p. 136.

Cuddihy, R. G. , D. G. Brownstein, 0. G. Raabe, and G. M. Kanapilly. Respiratory tract depo-
sition of inhaled polydisperse aerosols in beagle dogs. J. Aerosol Sci. 4:35, 1973.

11-59
-------
Dadaian, J. H., S. Yin., and G. A. Laurenzi. Studies of mucus flow in the mammalian respira-
tory tract. Am, Rev. Resp. Dis. 103:808, 1971.

Dalhamn, T., and L. Strandberg. Acute effect of sulfur dioxide on rate of ciliary beat in
trachea of rabbit _in vivo and jri vitro, with studies on absorptional capacity of nasal
cavity. Int. J. Air Water Pollut. 4:154, 1961.

Dautrebande, L. , and W. Walkenhurst. New studies on aerosols XXIV. Arch. Int. Pharmacodyn.
162:194, 1965.

Davidson, M. R., and J. M. Fitz-Gerald.' Transport of 02 along a model pathway through the
respiratory region of the lung. Bull. Math. Biol. 36:275-303, 1974.

Davies, C. N. A formalized anatomy of the human respiratory tract. In: Inhaled Particles and
Vapours. C. N. Davies, ed., Pergamon Press, Oxford, England, 1961, p. 82.

Davies, C. N. The handling of particles by the human lungs. Br. Med. Bull. 19:49, 1963.

Davies, C. N. A comparison between inhaled dust and the dust recovered from human lungs.
Health Phys. 10:1029, 1964a.

Davies, C. N. Deposition and retention of dust in the human respiratory tract. Ann. Occup.
Hyg. 7:169, 1964b.

Davies, C. N. An algebraical model for the deposition of aerosols in the human respiratory
tract during steady breathing. J, Aerosol Sci. .3:297, 1972.

Davies, C. N., J. Heyder, and M. C. Subba Ramu. Breathing of half-micron-aerosols. I.
Exp. J. Appl. Physio!. 32:592-600, 1972.

Davison, R, L., D. F. S. Natusch, J. R. Wallace, and C. A. Evans, Jr. Trace elements in fly
ash: Dependence of concentration on particle size. Environ. Sci. Technol. 8:1107-1113,
1974.

Deal, E. C., E. , R. McFadden, R. H. Ingram, R. H. Strauss, and J. J. Jaeger. Role of
respiratory heat exchange in production of exercise-induced asthma. J. Appl. Physio!:
Resp. Environ. Exercise Physiol. 46:467, 1979a.

Deal, E. C., E. R. McFadden, R. H. Ingram, and J. J. Jaeger. Hyperpnea and heat flux: initial
reaction sequence in exercise-induced asthma. J. Appl. Physiol: Resp. Environ.
Exercise Physiol. 46:476, 1979b.

Deal, E. C., E. R. McFadden, R. H. Ingram, and J. J. Jaeger. Esophageal temperature during
exercise in asthmatic and nonasthmatic subjects. J. Appl. Physio!: Resp. Environ.
Exercise Physiol. 46:484, 1979c.

Dekker, E. Transition between laminar and turbulent flow in human trachea. J. Appl. Physiol.
16:1060, 1961.

Doershuk, C. F., T. D. Downs, L. W. Matthews, and M. D. Lough. A method for ventilatory
measurements in subjects 1 month to 5 years of age: normal results and observations in
disease. Pediatr. Res. 4:165-174, 1970.

Doershuk, C. F., B. J. Fisher, and L.' W. Matthews. Pulmonary physiology of the young child.
Im Pulmonary Physiology of the Fetus, Newborn and Child. E. M. Scarpelli, ed., Mea and
Febiger, Philadelphia, PA, 1975.

Downs, W. L. , H. B. Wilson, G. 1. Sylvester, L. J. Leach, and E. A. Maynard. Excretion of
uranium by rats following inhalation of uranium dioxide. Health Phys. 13:445, 1967.

11-60
-------
DuBois, A. B., and R. M. Rogers. Respiratory factors determining the tissue concentrations of
inhaled toxic substances. Resp. Physio!. 5:34, 1968.

Dunm'11, M. S. Postnatal growth of the lung. Thorax 17:329, 1962.

Eatough, D. J., T. Major, J, Ryder, M. Hill, N. F. Mangelson, N. L. Eatough, L. D. Hansen, R,
6. Meisenheimer, and J. W. Fischer. The formation and stability of sulfite species in
aerosols. Atmos. Environ. 12:263-271, 1978.

Edmunds, L, H., P. D. Graf, S. S. Sagel, and R. H. Greenspan. Radiographlc observations of
clearance of tantalum and barium sulfate particles from airways. Invest. Radio!. 5:131,
1970.

Einbrodt, H. J. Experiments on the elimination of dust from human lungs. Ann. Occup. Hyg.
10:47, 1967.

Engel, L. A., L. D. Wood, G. Utz, and P. T. Macklem. Gas mixing during inspiration. J. Appl.
Physiol. 35:18, 1973.

Ewert, G. On the mucus flow rates in the human nose. Acta Otolaryngol; Suppl. 200, 1965.

Felicetti, S. A., S. A. Silbaugh, B. A. Muggenberg, and F. F. Hahn. Effect of time
post-exposure on the effectiveness of bronchopulmonary lavage in removing inhaled 114Ce
in fused clay from beagle dogs. Health Phys. 29:89, 1975.

Ferin, J. The mechanism of elimination of deposited particles from the lungs. Ann. Occup.
Hyg. 10:207, 1967.

Ferin, J. Lung clearance of particles. |n: Air Pollution and The Lung. E. F. Aharonson, A.
Ben-David, and M. A. Klingberg, eds., Halsted Press-John Wiley, Jerusalem, Israel, 1976.
pp. 64-78.

Ferin, J. Effect of particle content of lung on clearance pathways. In: Pulmonary
Macrophages and Epithelial Cells. Proceedings of the Sixteenth Annual Hanford Biology
Symposium, Energy Research and Development Administration and Battelle Memorial Institute,
Richland, Washington, September 27-29, 1976. C. L. Sanders, R. P. Dagle, and H. A. Ragan,
eds., ERDA Symposium Series 43, Energy Research and Development Administration, Oak Ridge,
TN, 1977. pp. 414-428.

Ferin, J. and M. L. Feldstein. Pulmonary clearance and hilar lymph node content in rats after
particle exposure. Environ. Res. 16:342-352, 1978.

Ferin, J., G. Urbankova, and A. Vickova. Pulmonary clearance and the clearance of macrophages.
Arch. Environ. Health 10:790, 1965.

Ferron, G. A. The size of soluble aerosol particles as a function of the humidity of the air.
Application to the human respiratory tract. J. Aerosol Sci. 8:251-267, 1977.

Findeisen, W. Uber das absetzen kleiner in der Luft suspendieter Teilchen in der
menschlichen Lunger bei der-Atmung. [On the settling of squall suspended in air particles
deposited in the human lung by respiration.] Pflugers Arch. J. d. Physiol. 236:367, 1935.

Fish, B. R. , and J. L. Durham. Diffusion coefficient of S0« in air. Environ. Lett. 2:13-21,
1971. L

Foord, N. , A. Black, and M. Walsh. Regional deposition of 2.5 - 7.5 mm diameter inhaled
particles in healthy male non-smokers. AERE Harwell, ML. 76:2892, 1976.
11-61
-------
Fowler, J, F., and A. E, Young. The average density of healthy lung. Am. J. Roentgenol.
Radium Ther. 81:312, 1959.

Frank, N. R., R. E. Yoder, E. Yokoyama, and F. E. Speizer. The diffusion of 35S02 from
tissue fluids into the lungs following exposure of dogs to 35S02. Health Phys. 13:31-
36, 1967.

Frank, N. R., R. E. Yoder, J. D. Brain, and E. Yokoyama. S02 (35S labeled) absorption by the
nose and mouth under conditions of varying concentration and flow. Arch. Environ. Health
18:315-322, 1969.

Fraser, D. A. The deposition of unipolar charged particles in the lungs of animals. Arch.
Environ.- Health 13:152, 1966.

Fraser, R. G., and J. A. P. Pare. Structure and Function of the Lung. W. B. Saunders Co. ,
Philadelphia, PA, 1971.

Friberg. L., and B. Holma. External measurement of lung clearance. Arch. Environ. Health
3:56, 1961.

Fry, D. A preliminary model for simulating the aerodynamics of the bronchial tree. Comp.
Biomed. Res. 2:111, 1968.

Fry, F. A., and A. Black. Regional deposition and clearance of particles in the human nose.
J. Aerosol Sci. 4:113, 1973.

Fuchs, N. A. The Mechanics of Aerosols. The Macmillan Company, New York, NY, 1964.

George, A. C. , and A. J. Breslin. Deposition of natural radon daughters in human subjects.
Health Phys. 13:375, 1967.

Giacomelli-Maltoni, G., C. Melandri, V. Prodi, and G. Tarroni. Deposition efficiency of mono-
disperse particles in human respiratory tract. Am. Ind. Hyg. Assoc. J. 33:603, 1972.

Gil, J., H. Bachofen, P. Gehr, and E. R. Weibel. Alveolar volume-surface area relation in air-
and saline-filled lungs fixed by vascular perfusion. J. Appl. Physiol.: Resp. Environ.
Excercise Physiol. 47:990-1001, 1979.

Gladney, E. S,, J. A. Small, G. E. Gordon, and W. H. Zoller. Composition and size distribution
of in-stack particulate material at a coal-fired power plant. Atmos. Environ. 10:1071,
1976.

Goldberg, I. S., and R. V. Lourenco. Deposition of aerosols in pulmonary disease. Arch.
Intern. Med. 131:88-91, 1973.

Goldberg, I. S., K. Y. Lam, B. Bernstein, and H. 0. Hutchens. Solution to the Fokker-Planck
equations governing simultaneous diffusion and gravitational settling of aerosol particles
from stationary gas in a horizontal tube. J. Aerosol Sci. 9:209, 1978.

Gordleyeff, V. A. The adsorption of gases and vapors on aerosol particulates. Am. Ind. Hyg.
Assoc. Q. 17:411, 1956.

Gormley, P. G.» and M. Kennedy. Diffusion from a stream flowing through a cylindrical tube.
Proc. R. Ir. Acad. A52:163, 1949.

Grant, B. J. , H. A. Jones, and J. M. Hughes. Sequence of regional filling during a tidal
breath in man. J. Appl. Physiol. 37:158-65, 1974.
11-62
-------
Green, G. M, Alveolobronchiolar transport observations and hypothesis of a pathway. Chest
59:15, 1971.

Green, G. M. In defense of the lung. Am. Lung Assoc. Bull. 60:4, 1974.

Green, J, F, The pulmonary circulation. Ijn: The Peripheral Circulations. R. Zelis, ed. ,
Grune and Stratton, New York, NY, 1975. p. 9.

Gunnison, A. F. , and A. W. Benton. Sulfur dioxide: sulfite. Interaction with mammalian
serum and plasma. Arch. Environ. Health 22:381-388, 1971.

Guyton, A. C. Measurement of the respiratory volumes of laboratory animals. Am. J. Physio!.
150:20, 1947a,

Guyton, A. C. Analysis of respiratory patterns in laboratory animals. Am. J. Physiol. 150:78,
1947b.

Harris, R. L. , and D. A. Fraser. A model for deposition of fibers in the human respiratory
system. Am. Ind. Hyg. Assoc. J. 37:73, 1976.

Hatch, T. E., and P. Gross. Pulmonary Deposition and* Retention of Inhaled Aerosols. Academic
Press, New York, NY, 1964.

Henderson, R. F., J. J. Waide, and R. C. Pfleger. Replacement time for alveolar lipid removed
by pulmonary lavage: effects of multiple lavage on lung lipids. Arch. Int. Physiol.
Biochim. 83:261, 1975.

Heppleston, A. G. The fibrogenic action of silica. Br. Med. Bull. 25:282, 1969.

Heyder, J. Conditions for the determination of aerosol particle deposition in the human
respiratory tract. Staub-Rein. Luft 31:11, 1971.

Heyder, J. , and C. N. Davies. The breathing of half micron aerosols - III dispersion of
particles in the respiratory tract. J. Aerosol Sci. 2:437, 1971.

Heyder, J. , and J. Gebhart. Gravitational deposition of particles from laminar aerosol flow
through inclined circular tubes. J. Aerosol Sci. 8:289, 1977.

Heyder, J. L., G. J. Armbruster, and W. Stahlhofen. Deposition of aerosol particles in the
human respiratory tract. In: Aerosol in Physik, Medizin and Technik, Proceedings of a
Conference, Bad Soden, October 17-18, 1973. V. Bohlan, ed. , Gesellschaft fuer Aerosol-
forschung, Bad Soden, West Germany, 1973a. pp. 122-125.

Heyder, J. , J. Gebhart, G. Heigiver, C. Roth and W. Stahlhofen. Experimental studies of the
total deposition of aerosol particles in the human respiratory tract. J. Aerosol. Sci.
4:191-208, 1973b.

Heyder, J., L. Armbruster, J. Gebhart, E. Grein, and W. Stahlhofen. Total deposition of aerosol
particles in the human respiratory tract for nose and mouth breathing. J. Aerosol Sci.
6:311, 1975.

Heyder, J. , J. Gebhart, C. Roth, W. Stahlhofen, B. Stuck, G. Tarroni, T. DeZaiacomo, M.
Formignani, C. Melandri, and V. Prodi. Intercomparison of lung deposition data for
aerosol particles. J. Aerosol. Sci. 9:147-155, 1978.

Heyder, J. , J. Gebhart, and W. Stahlhofen. Inhalation of aerosols. Particle deposition and
retention. In: Generation of Aerosols and Facilities for Exposure Experiments. K.
Willeke, ed., Ann Arbor Science, Ann Arbor, MI, 1980a. pp. 65-103.
11-63
-------
Heyder, J., J. Gebhart, G. Rudolf and W. Stahlhofen. Physical factors determining particle
deposition in the human respiratory tract. J. Aerosol Sci., 11:505-515, 1980b.

Hibbs, J. B., R. R. Taintor, H. A. Chapman, and J. B. Weinberg. Macrophage tumor killing:
influence of the local environment. Science 197:279-282, 1977.

Higgs, B. E. , M. Clode, G. J. R. McHardy, N. L. Jones, and E. J. M. Campbell. Changes in
ventilation gas exchange and circulation during exercise in normal subjects, din. Sci.
32:329-337, 1967.

Hilding, A. C. Ciliary streaming in the bronchial tree and the time element in
carcinogenesis. N. Engl. J. Med. 256:634-640, 1957.

Hollinger, M. A., 0. G. Raabe, S. N. Giri, M. Freywald, S. V. league, and B. Tarkington.
Effect of the inhalation of zinc and dietary zinc on paraquat toxicity in the rat.
Toxicol. Appl. Pharmacol. 49:53-61, 1979.

Holma, B. Lung clearance of mono- and di-disperse aerosols determined by profile scanning and
whole-body counting. Acta Medica Scand. Suppl. 473, 1967.

Holmes, T. H., H. Goodell, S. Wolf, and H. G. Wolff. The Nose. Charles C. Thomas,
Springfield, 111., 1950.

Horsfield, K. , and G. Cumming. Angles of branching and diameters of branches in the human
bronchial tree. Bull. Math. Biophys. 29:245, 1967.

Horsfield, K. , and G. Cumming. Morphology of the bronchial tree in man. J. Appl. Physiol.
24:373, 1968.

Horsfield, K. , G. Dart, D. E. Olson, G. F. Filley, and G. Cumming. Models of the human
bronchial tree. J. Appl. Physiol. 31:207-217, 1971.

Hounara, R. F. The deposition of atmospheric condensation nuclei in the nasopharyngeal region
of the human respiratory tract. Health Phys. 20:219, 1971.

Hounara, R. F., A. Black, and M. Walsh. Deposition of aerosol particles in the nasopharyngeal
region of the human respiratory tract. Nature 221:1254-1255, 1969.

Hounam, R. F. , A. Black, and M. Walsh. The deposition of aerosol particles in the nasopharyn-
geal region of the human respiratory tract. J. Aerosol Sci. 2:47, 1971a.

Hounam, R. F., A. Black, and M. Walsh. The deposition of aerosol particles in the nasopharyn-
geal region of the human respiratory tract. In: Inhaled Particles III. W. H. Walton,
ed., Unwin Brothers Limited, Surrey, England, 1971b. p. 71.

Hughes, J. M., F. G. Hoppin, and J. Mead. Effect of lung inflation on bronchial length and
diameter in excised lungs. J. Appl. Physiol. 32:25, 1972.

Intermountain Thoracic Society. Clinical Pulmonary Function Testing: A Manuel on Uniform
Laboratory Procedures for the Intermountain Area. R. E. Kanner and A. H. Morris, eds.,
Salt Lake City, UT, 1975. pp. VI-1 to 140.

International Commission on Radiological Protection. Report of the Task Group on Reference
Man, Pergamon Press, Oxford, England, 1975.

International Standards Organization. Size definitions for particulate sampling: recommen-
dations of ad hoc working group appointed by TC 146 of the International Standards
Organiation. Am. Ind, Hyg. Assoc. J. 42:A~64 to A-68, 1981.
11-64
-------
Jaffrin, M. Y. , and P. Kesic. Airway resistance: a fluid mechanical approach. J. Appl.
Physiol. 36:354-361, 1974.

Johnston, J. R. , and D. C, F, Muir. Inertia! deposition of particles in the lung. J. Aerosol
Sci. 4:269, 1973.

Johnston, J. R. , and R. C. Schroter. Deposition of particles in model airways. J. Appl.
Physiol.: Resp.' Environ. Exercise Physiol. 47,:947-953., 1979.

Jones, N. L. , E. J. M. Campbell, R. H. T. Edwards, and D. G. Robertson. Clinical Exercise
Testing. W. B. Saunders, Philadelphia, PA, 1975. p, 202.

Kanapilly, G. M. Alveolar microenvironment and its relationship to the retention and transport
into blood of aerosols deposited in the alveoli. Health Phys. 32:89, 1977.
Kanapilly, G. M. , and J, H. Diel. Ultrafine P 0~ aerosol generation, characterization and
short-term inhalation study in the rat. Health Phys. 39:505-519, 1980.

Kaufman, L. , and G. Gamsus. Fluorescent excitation in the measurement of clearance of heavy
metals from the lungs. I.E.E.E. Trans, on Nuclear Sci. NS- 21: 1721, 1974.

Kawecki , J. M. Emission of Sulfur-Bearing Compounds from Motor Vehicle and Aircraft Engines,
A Report to Congress. EPA-600/9-78-028, U.S. Environmental Protection Agency, Research
Triangle Park, NC, August 1978.

Klass, D. J. Immunochemical studies of the protein fraction of pulmonary surface active
material. Am. Rev. Resp. Dis. 107:784, 1973.

Knight, G. , and K. Lichti. Comparison of cyclone and horizontal elutriator size selectors.
Am. Ind. Hyg. Assoc. J. 31:437-441, 1970.

Kott, A. T. , J. W. Gardner, R. S. Schecter, and W. DeGroot. The elasticity of pulmonary lung
surfactants. J. Colloid Interface Sci. 47:265, 1974.

Krahl , V. Microstructure of the lung. Arch. Environ. Health 6:37, 1963.

LaBelle, C. W. , and H. Brieger. Patterns and mechanisms in the elimination of dust from the
lung. In: Inhaled Particles and Vapours. C. N. Davies, ed. , Pergamon Press, Oxford,
England, 1961. p. 356.

LaBelle, C. W. , M. A. Bevilacqua, and H, Brieger. The influence of cigarette smoke on lung
clearance. Arch. Environ. Health 12:588, 1966.

Landahl , H. D. On the removal of airborne droplets by the human respiratory tract. I. The
lung. Bull. Math. Biophys. 12:43, 1950.

Landahl, H. D. Particle remove! by the respiratory system. Bull. Math. Biophys. 25:29, 1963.

Landahl, H, , and S. Black. Penetration of air-borne particulates through the human nose. J.
Ind. Hyg. Toxicol. 29:269, 1947.

Landahl, H. , and R. Herrmann. On the retention of air-borne particulates in the human lung.
J. Ind. Hyg. Toxicol. 30:181, 1948.

Landahl, H. D. , T. N. Tracewell , and W. H. Lassen. On the retention of airborne particulates
in the human lung II. AMA Arch Ind. Health Occ. Med, 3:359, 1951.
11-65
-------
Landahl, H. D, , T. N. Tracewell, and W. H. Lassen. Retention of airborne participates in the
human lung III. AHA Arch. Ind. Hyg. Occup. Med. 6:508-511, 1952.

Lapp, N. L., J. L. Hankinson, H. Amandus and E. 0. Palmes, Variability in the size of air-
spaces in normal human beings as estimated by aerosols. Thorax .30:293-299, 1975.

Larson, T. V., D. S. Covert, R. Frank, and R. J. Charlson. Ammonia in the human airways,
neutralization of inspired acid sulfate aerosol. Science 1.97:161-163, 1977.

Lauweryns, J. M., and J. H. Baert. Alveolar clearance and the role of the pulmonary lymphatics.
Am. Rev. Resp. Dis. 115:625, 1977.

Leach, L. J., E. A. Maynard, H. C. Hodge, J. K. Scott, C. L. Yuile, G. E. Sylvester, and H. G.
Wilson. A five-year inhalation study with natural uranium-dioxide (U02) dust - I.
Retention .and biological effect in the monkey dog and rat. Health Phys. 18:599, 1970.

Leach, L. J. , C, L. Yuile, H. C. Hodge, G. E. Sylvester, and H. B. Wilson. A five-year inha-
lation study with natural uranium dioxide (U02) dust - II. Postexposure retention and
biological effects in the monkey, dog, and rat. Health Phys. 25:239, 1973.

Leeds, S. E., S. Reich, H. N. Uhley, J. J. Sampson, and M. Friedman. The pulmonary lymph flow
after irradiation of the lungs of dogs. Chest 59:203, 1971.

Lever, J,, cited in C. N. Davies. Deposition of aerosol in the human lung. In: Aerosole in
Physik Medizin und Technik, Bad-Soden, W. Germany, Gesellschaft fur Aerosolforschung, 1974
p. 90-99.

Lippmann, H. Deposition and clearance of inhaled particles in the human nose. Ann. Otol.
Rhino!. Laryngol. 79:519-528, 1970.

Lippmann, M. Regional deposition of particles in the human respiratory tract. In: Handbook
of Physiology, Section 9: Reactions to Environmental Agents. D. H. K. Lee, H. L. Falk,
and S. D. Murphy, eds., The American Physiological Society, Bethesda, MD, 1977. pp.
213-232.

Lippmann, M, "Respirable" dust sampling. In: Air Sampling Instruments for Evaluation of
Atmospheric Contaminants, 5th ed. American Conference of Governmental Industrial
Hygienists, Cincinnati, OH, 1978. pp. G-l, G-23.

Lippmann, M., and R. Albert. The effect of particle size on the regional deposition of
inhaled aerosols in the human respiratory tract. Am. Ind. Hyg. Assoc. J. 30:257, 1969.

Lippmann, M., R. E. Albert, and H. T. Peterson, Jr. The regional deposition of inhaled
aerosols in man. In: Inhaled Particles III. W. H. Walton, ed. , Unwin Brothers Limited,
Surrey, England, 1971. p. 105. ,

Lippmann, M., M. S. Mok, and K. Wasserman. Anaesthetic management for children with alveolar
proteinosis using extracorporeal circulation: a report of two cases. Br. J. Anaesth.
49:173-177, 1977.

Longley, M. Y. Pulmonary deposition of dust as affected by electric charges on the body. Am.
Ind. Hyg. Assoc. J. 21:187, 1960.

Longley, M. Y., and C. M. Berry. Pulmonary deposition of aerosols: effect of electrostratic
charging of the animal body and the aerosol. Arch. Environ. Health 2_:533, 1961.
11-66
-------
Lourenco, R. V., R. Loddenkemper, and R. W. Cargon. Patterns of distribution and clearance of
aerosols in patients with bronchiectasis. Am. Rev. Resp. Dis. 106:857-866, 1972.

Luchsinger, ,P. G. , B. LaGarde, and J. E. Kilfeather. Particle clearance from the human
tracheobronchial tree. Am. Rev. Resp. Dis. 97:1046, 1968.

Luft, U. C. Spirometric methods. In: Aviation Medicine - Selected Reviews. C, S. White, W.
R. Lovelace, F. G. Hirsch, eds., Pergamon Press, New York, NY, 1958. p. 168.

Lynch, J. R. Evaluation of size-selective presamplers: I. Theoretical cyclone and
elutriator relationships. Am. Ind. Hyg. Assoc. J. 31:548-551, 1970.

Machlin, C. C. The alveoli of the mammalian lung: an anatomical study with clinical corre-
lations. Proc. Inst. Med. 18:78, 1950.

Maguire, B. A., and D. Barker. A gravimetric dust sampling instrument (SIMPEDS): preliminary
underground trials. Ann. Occup. Hyg. 12:197-201, 1969.

Marshall, R. , and W. Holden. Changes in calibre of the smaller airways in man. Thorax 18:54,
1963. ~

Martens, A., and W. Jacobi. Die In-Vivo Bestimmung der Aerosolteilchen-deposition im
Atemtrakt bei Mund-Bzw. Nasenatmung. [In vivo' determination of aerosol particle
deposition in the total respiratory tract.] In: Aerosole in Physik, Medizin und Technik,
Gesellschaft fur Aerosolforschung, Bad Soden, W. Germany, 1973. pp. 117-121.

Martin, D. , and W. Jacobi. Diffusion deposition of small-sized particles in the bronchial
tree. Health Phys. 23i_:23-29, 1972.

McEuen, D. D. , and J. L. Abraham. Particulate concentrations in pulmonary alveolar protei-
nosis. Environ. Res. 17:334-339, 1978.

Melandri, C. , V. Prodi, G. Tarroni, M. Formignam", T. DeZaiacomo, G. R. Bompane, G. Maestri,
and G. G. Giacomellf-Maltoni. On the deposition of unipolarly charged particles in the
human respiratory tract. In: Inhaled Particles IV. W. H. Walton, ed., Pergamon Press,
New York, NY, 1977. p. 193.

Melville, G. N. Changes in specific airway conductance in healthy volunteers following nasal
and oral inhalation of SO,. W. I. Med. J. 19:231-235, 1970.
c. ——

Mercer, T. T. On the role of particle size in the dissolution of lung burdens. Health Phys.
13:1211, 1967.

Mercer, T. T. Aerosol Technology in Hazard Evaluation. Academic Press, New York, NY, 1973.
pp. 66-280.

Michel, F. B. , J. P. Marty, L. Quet and P. Cour. Penetration of inhaled pollen into the res-
piratory tract. Am. Rev. Resp. Dis. 115:609-616, 1977.

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.

Morgan, A,, J. C. Evans, and A. Holmes. Deposition and clearance of inhaled fibrous minerals
in the rat: studies using radioactive tracer techniques. In: Inhaled Particles IV. W.
H. Walton, ed., Pergamon Press, Oxford, England, 1977. pp. 259-274.

Morgan, W. K. C. Industrial bronchitis. Br. J. of Ind. Med. 35:285-291, 1978.
11-67
-------
Morgan, W. K. C, , and A. Seaton. Occupational lung diseases. W. B. Saunders, Philadelphia,
PA, 1975.

Morrow, P. E. Experimental studies of inhaled materials. Arch. Intern. Med. 126:466, 1970a.

Morrow, P. E. Models for the study of particle retention and elimination in the lung. In:
Inhalation Carcinogenesis. M. G. Hanna, P. Nettesheirn, J. R, Gilbert, eds. , U.S. Atomic
Energy Commission, Oak Ridge, TN, 1970b. p. 103.

Morrow, P. E. Lymphatic drainage of the lung in dust clearance. Ann, N.Y. Acad. Sci.
22:64-65, 1972.

Morrow, P. E. Alveolar clearance of aerosols. Arch. Intern. Med. 131:101, 1973.

Morrow, P. E., E. Mehrhof, L. Casarett, and D. Morken. An experimental study of aerosol
deposition in human subjects. AMA Arch. Ind. Health 18:292, 1958.

Morrow, P. E., D. V. Pates, B. R. Fish, T. F. Hatch, and T. T. Mercer. International
commission on radiological protection task group on lung dynamics, deposition and
retention models for internal dosimetry of the human respiratory tract. Health Phys.
12:173-207, 1966.

Morrow, P. E., F. R. Gibb, and K. M. Gazioglu. A study of particulate clearance from the
human lungs. Am. Rev. Resp. Dis. 96:1209, 1967a.

Morrow, P. E., F. R. Gibb, and K. M. Gazioglu. The clearance of dust from the lower
respiratory tract of man: an experimental study, Irt: Inhaled Particles and Vapours.
S. C. N. Davies, ed., Pergamon Press, Oxford, England, 1967b. p. 351.

Moss, 0. R., and H. J. Ettinger. Respirable dust characteristics of polydisperse aerosols.
Am. Ind. Hyg. Assoc. J. 31:546-547, 1970.

Muir, D. C., and C. N. Davies. The deposition of 0.5 m diameter aerosols in the lungs of man.
Ann. Occup. Hyg. 10:161, 1967.

Nair, P. V. N., and V. G. Vohra. Growth of aqueous sulphuric acid droplets as a function of
relative humidvty. J. Aerosol Sci. 6:265, 1975.

National Academy of Sciences. Measurement and Control of Respirable Dust in Mines. National
Academy of Sciences, Washington, DC, 1980. pp. 348-405.

Natusch, D. F. S. and J. R. Wallace. Urban aerosol toxicity: the influence of particle size.
Science 186:695-699, 1979.

Natusch, D. F. S. , J. R. Wallace, and C. A. Evans, Jr. Toxic trace elements: preferential
concentration in respirable particles. Science 183:202-204, 1974.

Newhouse, M. T., M. Dolovich, G. Obminski, and R. K. Wolff. Effect of TLV levels of S02 and
HpSO^ on bronchial clearance in exercising man. Arch. Environ. Health 33:24-32, 1978.

Niinimaa, V., P. Cole, S. Mintz and R. J. Shephard. The switching point from nasal to
oronasal breathing. Resp. Physio!. 42:61-71, 1980.

Niinimaa, V., P. Cole, S. Mintz and R. J. Shephard. Oronasal distribution of respiratory
airflow. Resp. Physio!. 43:69-75, 1981.

Oberdorster, G., F. R. Gibb, H. Beiter, S. T. Lu, P. E. Morrow. Studies of the lymphatic
drainage of dog lungs. J. Toxicol. Environ. Health 4:571-586, 1978.
11-68
-------
Olson, D. E. , M. F. Sudlow, K. Horsfield, and G. F. Filley. Convective patterns of flow
during inspiration. Arch. Intern. Med. 131:51-57, 1973.

Qrenstein, A. J. , ed. Proceedings of the Pneumonoconiosis Conference 1959, J & A Churchill,
Ltd., London, England, 1960.

Owen, P. R. Turbulent flow and particle deposition' in* the "trachea. In: Circulatory and
Respiratory Mass Transport. G. E. W. Wolstenholme and J. Knight, eds., A CIBA Foundation
Symposium. Little, Brown and Co., Boston, MA, 1969. pp. 236-252.

Pack, A. , M. B. Hooper, W. Nixon, and J. C. Taylor. A computational model of pulmonary gas
transport incorporating effective diffusion. Resp. Physio!. 29:101-124, 1977.

Paiva, M. Gas transport in the human lung. J. Appl. Physio!. 35:401, 1973.

Palmes, E. D. , and C. S. Wang. An aerosol inhalation apparatus for human single breath
deposition studies. Am. Ind. Hyg. Assoc. J. ^32:43, 1971.

Palmes, E. D. and M. Lippmann. Influence of respiratory airspaces dimensions on aerosol
deposition. Inhaled Particles IV. W. J. Walton, ed., Permagon Press, Oxford, England,
1977, pp. 127-135.

Rattle, R. E. The retention of gases and particles in the human nose. In: Inhaled Particles
and Vapours, C. N. Davies, ed., Pergamon Press, Oxford, England, 1961a. p. 302.

Pattle, R. E. The lining complex of the lung alveoli. In: Inhaled Particles and Vapours.
C. N. Davies, ed., Pergamon Press, Oxford, England, 1961b. p. 70.

Pavia, D. , M. Thomson, and H. S. Shannon. Aerosol inhalation and depth of deposition in the
human lung. Arch. Environ. Health 32:131, 1977.

Pavlik, I. The fate of light air ions in the respiratory pathways. Int. J. Biometeor.
11:175, 1967.

Pawley, J. B., and G. L. Fisher. Using simultaneous three colour X-ray mapping and
digital-scan-stop for rapid elemental characterization of coal combustion by-products.
J. Microsc. 110:87, 1977.

Pedley, T. J. A theory for gas mixing in a simple model of the lung. In: Fluid Dynamics of
Blood Circulation and Respiratory Flow, Proceedings of Meeting, Naples, Italy, May 4-6,
1970 AGARD Conference Proceedings No. 65. Advisory Group for Aerospace Research and
Development, Paris, France, 1970.

Pedley, T. J. , R. C. Schroter, and M. F. Sudlow. Flow and pressure drop in systems of
repeatedly branching tubes. J. Fluid Mech. 46:365-383, 1971.

Pfleger, R. C. , and H. G. Thomas. Beagle dog pulmonary surfactant lipids. Arch. Intern.
Med. 9:70, 1971.

Phalen, R. F., and P. E. Morrow. Experimental inhalation of metallic silver. Health Phys.
24:509-518, 1973.

Phalen, R. F. , H. C. Yeh, G. M. Schum, and 0. G. Raabe. Application of an idealized model to
morphometry of the mammalian tracheobronchial tree. Anat. Rec. 190:167-176, 1978.

Polgar, G. , and T. R. Weng. The functional development of the respiratory system from the
period of gestation to adulthood. Am. Rev. Resp. Dis. 120:625-695, 1979.
11-69
-------
Pratt, P. C. and K. H. Kilburn. Extent of pulmonary pigmentation as an indicator of par-
ticulate environmental air pollution. In: Inhaled Particles III. II. W. H. Walton, ed.,
Unwin Brothers Limited, The Gresham Press, Surrey, England, 1971, pp. 661-670.

Proctor, D. F. , and D. L. Swift. The nose - a defense against the atmospheric environment.
In: Inhaled Particles III. V. W. H. Walton, ed., Unwin Brothers, Limited, Surrey,
England, 1971. p. 59.

Proctor, D. F., and H. N. Wagner. Clearance of particles from the human nose. Arch. Environ.
Health 11:366, 1965.

Proctor, D. F. , and H. N. Wagner. Mucociliary clearance in the human nose. In: Inhaled
Particles and. Vapours II. C. N. Davies, ed. , Pergamon Press, Oxford, England, 1967.
p. 25.

Proctor, D. F. , D. L. Swift, M. Quinlan, S. Salman, Y. Takagi, and S. Evering. The nose and
man's atmospheric environment. Arch. Environ. Health 18:671, 1969.

Proctor, D. F. , I. Andersen, and G. Lundquist. Clearance of inhaled particles from the human
nose. Arch. Intern. Med. 131:132, 1973.

Pruitt, K. M., M. J. Cherny, and H. L. Spitzer. Physical and chemical characterization of pig
lung surfactant 1ippoprotein. Arch. Intern. Med. 9:6, 1971.

Pump, K. K. The morphology of the finer branches of the bronchial tree of the human lung.
Dis. Chest 46:379, 1964.

Raabe, 0. G. Some important consideration in use of power function to describe clearance
data. Health Phys. 13:293, 1967.

Raabe, 0. G. Particle size analysis utilizing grouped data and the log-normal distribution.
Aerosol Sci. 2:289, 1971.

Raabe, 0. G. Aerosol aerodynamic size conventions for inertial sampler calibration. J. Air
Pollut. Control Assoc. 26:856, 1976.

Raabe, 0. G. Deposition and Clearance of Inhaled Aerosols. U.S. Department of Energy.
National Technical Information Service, UCD-472-503, Springfield, VA, 1979.

Raabe, 0. G., and M. Goldman. A predictive model of early mortality following acute
inhalation of PuO™ aerosols. Radiat. Res. 78:264-277, 1979.

Raabe, 0. G. , H. C. Yeh, G. M. Schum, and R. F. Phalen. Tracheobronchial Geometry: Human,
Dog, Rat, Hamster, LF-53. Lovelace Foundation, Albuquerque, NM, 1976.

Raabe, 0. G. , H. C. Yeh, G. J. Newton, R. F. Phalen, and D. J. Velasquez. Deposition of
inhaled monodisperse aerosols in small rodents. In: Inhaled Particles IV. W. H.
Walton, ed. , Pergamon Press, New York, NY, 1977. p. 3-22.

Raabe, 0. G. , S. V. Teague, N. L. Richardson, and L. S. Nelson. Aerodynamic and dissolution
behavior of fume aerosols produced during the combustion of laser-ignited plutonium
droplets in air. Health Phys. 35:663, 1978.

Ramsden, D. , M. E. D. Bains, and D. C. Fraser. In vivo and bioassay results from two
contrasting cases of plutonium-239 inhalation. Health Phys. 19:9, 1970.

Reifenath, R. Chemical analysis of the lung alveolar surfactant obtained by alveolar micro-
puncture. Resp. Physio!. 19:35, 1973.
11-70
-------
Rudolf, G., and J. Heyder. Deposition of aerosol particles in the human nose. In; Aerosole
in Naturwissenschaft, Medizin und Techm'k. V. Bb'hlan, ed. Proceedings of a Conference
held in Bad Soden, October 16-19, 1974. Gesellschaft fur Aerosolforschung, Bad Soden,
West Germany, 1974.

Saibene, F. , P. Mognoni, C. L. LaFortuna, and R. Mostardi. Oronasal breathing during
exercise. Pflugers Arch. J. d. Physio!. 378:65-69, 1978.

Sanchis, J. , M. Dolovich, R. Chalmers, and M. Newhouse, Quantitation of regional aerosol
clearance in the normal human lung. J. Appl. Physio!. 3_3:757, 1972.

239
Sanders, C. L. , and R. R. Adee. Phagocytosis of inhaled plutonium oxide - Pu particles by
pulmonary macrophages. Science 162:918-920, 1968.

Scarpelli, E. M. The Surfactant System of the Lung. Lea and Febiger, Philadelphia, PA, 1968.

Scherer, P. W. , L. H. Shendalman, and N. M. Greene. Simultaneous diffusion and convection in
a single breath lung washout. Bull. Math. Biophys. 34:393-412, 1972.

Scherer, P. W. , L. H. Shendalman, N. M. Greene, and A. Bouhuys. Measurement of axial
diffusivities in a model of the bronchial airways, J. Appl. Physiol. 38:719-723, 1975.

Scherer, P. W. , F. R. Haselton, L. M. Hanna, and D. R. Stone. Growth of hygroscopic aerosols
in a model of bronchial airways. J. Appl. Physiol.: Resp. Environ. Exercise Physiol.
47:544-550, 1979.

Schlesinger, R. B. Mucociliary interaction in the tracheobronchial tree and environmental
pollution. Biol. Sci. 23:567, 1973.

Schlesinger, R. B. , and M. Lippmann. Particle deposition in the trachea: i_n vivo and in
hollow casts. Thorax 31:678, 1976.

Schlesinger, R. B., and M. Lippmann. Selective particle deposition and bronchogenic
carcinoma. Environ. Res. 15:424, 1978.

Schroter, R. C. , and M. F. Sudlow. Flow patterns in models of the human bronchial airways.
Resp. Physiol. 7:341, 1969.

Shanty, F. Deposition of Ultrafine Aerosols in the Respiratory Tract of Human Volunteers.
Ph.D. thesis, Johns Hopkins University, Baltimore, MD, 1974.

Sherwin, R. P., M. L. Barman, and J. L. Abraham. Silicate pneumoconiosis of farm workers.
Lab. Invest. 40:576-582, 1979.

Sherwood, T. K. , R. L. Pigford, and C. R. Wilke. Mass Transfer. McGraw-Hill, New York, NY,
1975. pp. 13, 677.

Silverman, L. , and C. E. Billings. Pattern of airflow in the respiratory tract. _In: Inhaled
Particles and Vapours. C. N. Davies, ed., Pergamon Press, Oxford, England, 1961. p. 9.

Smith, F. A. , and E. A. Boyden. An analysis of the segmental bronchi of the right lower lobe
of fifty injected lungs. J. Thoroc, Surg. 18:195, 1949,

Snipes, M. B. , and M. F. Clem. Retention of microspheres in the rat lung after intratracheal
instillation. Environ. Res. 24:33-41, 1981.

Speizer, F. E. , and N. R. Frank. The uptake and release of S0? by the human nose. Arch.
Environ. Health 12:725-728, 1966.
11-71
-------
Spiegelman, J. R., G. D. Hanson, A. Lazarus, R. J. Bennett, M. Lippmann, and R. E. Albert.
Effect of acute SOg exposure on bronchial clearance in the donkey. Arch. Environ.
Health 17:321-326, 1968.

Stahl, W. R. Scaling of respiratory variables in mammals. J. Appl. Physiol. 22:453-460,
1967. ~

Stahlhofen W., J. Gebhart, and J. Heyder. Experimental determination of the regional
deposition of aerosol particles in the human respiratory tract. Am. Ind. Hyg. Assoc.
J. 41:385-398a, 1980.

Stahlhofen, W., J. Gebhart and J. Heyder. Biological variability of regional deposition of
aerosol particles in the human respiratory tract. Am. Ind. Hyg. Assoc. J. 42:348-352,
1981. ' ~~

Stockham, J. D., and E. G. Fochtman, eds. Particle Size Analysis. Ann Arbor Science, Ann
Arbor, MI, 1979. pp. 140.

Strandberg, L, G, SO, absorption in the respiratory tract. Arch. Environ. Health 9:160-166,
1964. £

Strecker, F. J, Tissue reactions in rat lungs after dust inhalation with special regard to
bronchial dust elimination and to the penetration of dust into the lung interstices and
lymphatic nodes. In: Inhaled Particles and Vapours II. C. N. Davies, ed. , Pergamon
Press, Oxford, England. 1967. p. 141.

Sweet, D. V., W. E. Grouse, and J. V, Crable. Chemical and statistical studies of contaminants
in urban lungs. Am. Ind. Hyg. Assoc. J. 39: 515-526, 1978.

Swift, D. L., F. Slanty, and J. T. O'Neil. Human respiratory deposition patterns of fume-like
particles. Presented in part at the 1977 American Industrial Hygiene Conference, New
Orleans, LA, May 26, 1977.

Taplin, G. V., N. D. Poe, E. K. Dore, A. Greenberg, and T. Isawa. Radioaerosol inhalation
scanning. In: Pulmonary Investigation with Radionuclides. W. M. Smoals, ed. , Charles
C. Thomas, Springfield, IL, 1970. pp. 296-317,

Tarroni, G., C. Melandri, V. Prodi, T. deZaiacomo, M. Formignani, and P. Bassi. An indication
on the biological variability of aerosol total deposition in humans. Am. Ind. Hyg,
Assoc. J. 41:826-831, 1980.

Taulbee, D. 8., and C. P. Yu. A theory of aerosol deposition in the human respiratory tract.
J. Appl. Physiol. 38:77, 1975.

Taulbee, D. , C. Yu, and J, Heyder. Aerosol transport in the human lung from analysis of
single breaths. J. Appl. Physiol. 44:803, 1978.

Tenney, S. M., and J. E. Remmers. Comparative quantitative morphology of the mammalian lung:
diffusing area. Nature 197:54, 1963.

Tenney, S. M., and D. Bartlett. Comparative quantitative morphology of the mammalian lung:
trachea. Resp. Physiol. 3:130, 1967.

Thomas, R. G. Transport of relatively insoluble materials from lung to lymph nodes. Health
Phys. 14:111, 1968.

Threshold Limits Committee. Threshold Limit Values of Air Borne Contaminants for 1968.
American Conference of Governmental Industrial Hygienists, Cincinnati, OH, 1968.
11-72
-------
Thurlbeck, W. M, , and J. B. Haines. Bronchial dimensions and stature. Am. Rev. Resp. Dis.
112:142, 1975.

Tuttle, W. C. , and S. C. Westerberg. Alpha-1-globulin trypsin inhibitor in canine surfactant
protein. Proc. Soc. Exp. Biol. Med. 146:232, 1974.

Uddstromer, M. Nasal respiration. Acta Otolaryngol. Suppl. 42:3-146, 1940.

Van As, A., and I. Webster. The' organization of ciliary activity and mucus transport in
pulmonary airways. S. A. Med. J. 46:347, 1972.

Van Ree, J. H. L., and H. A. E. van Dishoeck. Some investigations on nasal ciliary activity.
Pract. Otorhinolaryng (Basel) 24:383, 1962.

Van Wijk, A. M. , and H. S. Patterson. The percentage of particles of different sizes removed
from dust-laden air by breathing. J. Ind. Hyg. Toxicol. 22:31, 1940.

Verzar, F. , J. Keith, and V. Parchet. Temperatur und Feuchtigkeit der Luft in'den Atemwegen.
[temperature and moisture content of the air in the respiratory tract.] Pfluegers Arch.
J. d. Physio!. 257:400, 1953.

Waligora, S. J. , Jr. Pulmonary retention of zirconium oxide ( Nb) in man and beagle dogs.
Health Phys. 20:89, 1971.

Walkenhorst, W. Untersuchungen an einem nach teilchengrossen geordneten Mischstaub im
atembaren korngrossenbereich. In: Inhaled Particles and Vapours II. C. N. Davies, ed.,
Pergamon Press, Oxford, England, 1967. p. 563.

Wang, C. S. Gravitational deposition from laminar flows in inclined channels. J. Aerosol
Sci. 6:19, 1975.

Washburn, E. W., ed. National Research Council of the U.S.A. International Critical Tables of
Numerical Data, Physics, Chemistry and Technology, vol. 3. McGraw-Hill, New York, NY,
1928.

Weibel, E. R. Morphometry of the Human Lung. Academic Press, New York, NY, 1963.

West, J. B. Observations on gas flow in the human bronchial tree. In: Inhaled Particles and
Vapours. C. N. Davies, ed. , Proceedings of an International Symposium organized by the
British Occupational Hygiene Society, 1960. Pergamon Press, New York, NY, 1961. pp. 3-7.

West, J. B. Respiratory Physiology: the Essentials. Williams and Wilkins, Baltimore, MD,
1974.

West, J. B. Respiratory Physiology: the Essentials. Williams and Wilkins, Baltimore, MD,
1977.

Whimster, W, F. The microanatomy of the alveolar duct system. Thorax 25:141, 1970.

Whitby, K. T. The physical characteristics of sulfur aerosols. Atmos. Environ. 12:135, 1978.

Wilson, T. A., and K. Lin. Convection and diffusion in the airways and the design of the
bronchial tree. In: Airway Dynamics; Physiology and Pharmacology. A. Bouhuys, ed. ,
Charles C. Thomas, Springfield, IL, 1970. pp. 5-19.

Wolff, R. K. , M. Dolovich, C. 'M. Rossman, and M. T. Newhouse. Sulfur dioxide and
tracheobronchial clearance in man. Arch. Environ. Health 30:521-527, 1975.
11-73
-------
Yeates, D. B., N. Aspin, H. Levinson, M. T. Jones, and A. C. Bryan. Mucociliary
trachael transport rates in mon. J. of Appl. Physio!. 39:487-495, 1975.

Veh, H. C. Use of a heat transfer analogy for a mathematical model of respiratory tract
deposition. Bull. Math. Biol. 36:105, 1974.

Yeh, H. C., and G. M. Schuin, Models of human lung airways and their application to inhaled
particle deposition. Bull. Math. Biol. 42:461-480, 1980.

Yeh, H. C., R. F. Phalen, and 0. G. Raabe. Factors influencing the deposition of inhaled
particles. Environ. Health Perspect. 15:147, 1976.

Yu, C. P. An equation of gas transport in the lung, Resp. Physiol. 23:257-266, 1975.

Yu, C. P. Precipitation of unipolarly charged particles in cylindrical and spherical vessels.
J. Aerosol S.ci. 8:237, 1977.

Yu, C. P. A two component theory of aerosol deposition in human lung airways. Bull. Math.
Biol. 40:693-706, 1978.

Yu, C. P. , P. Nicolaides, and T. T. Soong. Effect of random airway sizes on aerosol
deposition. Am. Ind. Hyg. Assoc. J. 40:999-1005, 1979.

Zenz, C., ed. Occupational Medicine: Principles and Practical Applications. Year Book
Medical Publishing, Inc., Chicago, IL, 1975. pp. 113-115.
11-74
-------
12. TOXICOLOGICAL STUDIES

12.1 INTRODUCTION
This chapter describes the toxicity of sulfur oxides (SO ) and partieulate matter (PM) in
animals. The health effects of SO and PM have also been reviewed by the National Academy of
Sciences National Research Council (1978, 1979). The toxic effects of SO and of atmospheric
X
aerosols overlap because the salts of sulfuric acid (ammonium sulfate, sodium sulfate, and re-
lated compounds) are significant components of atmospheric aerosols (see Chapter 2). The toxi-
cology of all forms of SO must be considered as a whole. For example, in the ambient air,
sulfur dioxide (SO,) may interact with aerosols, be absorbed on particles, or be dissolved in
liquid aerosols. To a lesser degree, similar interactions may occur in the air within the
respiratory tract. Sulfuric acid (FLSO.) aerosols may react with ammonia, forming ammonium
sulfate [(NH»)pS04] and ammonium bisulfate (NH.HSO*) in the ambient air, in the animal ex-
posure chamber atmosphere before inhalation, or, to a lesser degree, in the respiratory tract
simultaneously upon inhalation (see Section 12.3). Biological interactions can also occur,
resulting in a mixture of pollutants that has additive, synergistic, or antagonistic health
effects compared to the effects of the single pollutants.
Discussions of the deposition and clearance of SO are limited here (see Chapter 11 for
A
more details). The major toxic effects of sulfur compounds, whether caused by S0~, HLSQ*, or
some sulfate salts, include immediate irritation of the respiratory tract. Most measurements
of this irritation have been made through studies of the respiratory mechanics of the experi-
mental animal. Similar studies of respiratory mechanics have been done with human subjects
either experimentally or environmentally exposed. The general effects of S0? on the respira-
tory mechanics of animals and man are the same. The animal studies reviewed here present some
details of the metabolism of S0? and bisulfite, the effects of S0? on the biochemistry,
physiology, and morphology of the respiratory tract, and the potential effects on organs other
than the lung.
The physicochemical diversity of PM in the atmosphere is a major problem. When the local
concentration of any organic compound exceeds its vapor pressure, it will condense and be
found in the partieulate fraction when sampled. Some may be sorbed on the surfaces of in-
organic particles, which can have a wide chemical variety. Progress is being made toward a
better understanding of the toxicity of these materials associated with particles, but at
present inadequate data are available. A more complete treatment of the health effects of
polycyclic organic matter is found in a review by the Environmental Criteria and Assessment
Office of EPA (1978), which gives an overview also for some of the heavy metals present in
polluted air. More detail is found in documents dealing specifically with each heavy metal
(Environmental Criteria and Assessment Office, 1979; Office of Research and Development,
1977; Committee on Biologic Effects of Atmospheric Pollutants, Vanadium, 1974; Committee on
Medical and Biologic Effects of Environmental Pollutants, Nickel, 1975; Committee on Biologic

12-1
-------
Effects of Atmospheric Pollutants, Lead, 1972; Committee on Biologic Effects of Atmospheric
Pollutants, Chromium, 1974; Committee on Medical and Biologic Effects of Environmental Pollut-
ants, Arsenic, 1977; National Academy of Sciences, Iron, 1979; National Academy of Sciences,
Zinc, 1979).
Interactions between SO and other pollutants are reviewed briefly because of the spar-
sity of the data available. Some of these studies are controversial and have not been dupli-
cated, especially those dealing with the potential mutagenic effects of SCL and the inter-
action of S0? with known carcinogens.
Another difficulty with a review of the toxicology of SO is the sparsity of recent
studies. Since the early 1970s few studies have appeared and work in progress is not included
here. As a result, a certain degree of sophistication is lacking in some of the intrepreta-
tions, not through a lack of appreciation of the problem, but simply because insufficient in-
formation is available.
12.2 EFFECTS OF SULFUR DIOXIDE
12.2.1 Biochemistry of Sulfur Dioxide
This section largely covers i_n vitro experiments, in which the potential target (i.e.,
cells, enzymes, other molecules, etc.) is exposed to the toxicant outside the body. The
potential problems of such a system include the fact that some homeostatic or repair
mechanisms are absent or that pollutants sometimes act by indirect mechanisms. For example,
the pollutant affects target A that in turn alters target B. Thus, if only target B were pre-
sent," the effect would not be observed. In addition, the dosimetric relationships of ]_n vitro
studies to jrt vivo studies have not been defined. Therefore, effective concentrations cannot
be extrapolated directly from j_n vitro to j_n vivo studies. For the above reasons, there is
some controversy as to whether observed i_n vitro reactions can be extrapolated to ijn vivo
mechanisms of toxicity. Nonetheless, sound i_n vitro investigations can show whether a given
pollutant has the potential of affecting a given target. In vitro studies are best used to
provide guidance for i_n vivo investigations or when j_n vivo results have been observed. In
the latter case, the relatively simplified HI vitro systems can sometimes elucidate the
potential mechanisms of toxicity. To these ends, they can be useful.
Knowledge of the chemistry of sulfurous acid and SO, is necessary to understand the
physiological and toxicological properties of SO,. Sulfur dioxide is the gaseous anhydride of
sulfurous acid. It dissolves readily in water, and at physiological pH near neutrality,
hydrated SO- readily dissociates to form bisulfite and sulfite ions as illustrated by Equa-
tions 12-1 and 12-2. The rate of hydration of S02 is very rapid, with a rate constant (k-,) of
3.4 x 106 H"1 sec"1; the rate constant of the reverse reaction is 2 x 108 M*1 sec"1 at 20°C
(Equation 12-1) (Tartar and Garetson, 1941). The logarithms of the inverse of the dissocia-
tion constants (pK ) of sulfurous acid are 1.37 and 6.25 (in dilute salt solutions) (Tartar
3
and Garetson, 1941); consequently, at pH 7.4, the concentration of sulfite ions is about 14
times that of bisulfite, but in rapid equilibrium. Hence, S02 can be treated as bisulfite/
sulfite and conversely.
12-2
-------
kl
SO, + H90 -< H,SO, 12-1
C- £. i £_ O
pKa=1.37

H2S03 -* - H + HS03 -« - H + SO 3 12-2

12.2.1.1 Chemical Reactions of Bisulfite with Biological Molecules—Sulfur dioxide reacts
readily with all major classes of biomolecules. Reactions of SO, or bisulfite with nucleic
acids, proteins, lipids, and other biological components have been repeatedly demonstrated i_n
vitro. There are three important reactions of bisulfite with biological molecules: sulfona-
tion, production of free radicals by autooxidation, and addition to cytosine.
Sulfonation, or sulfitolysis, (Gilbert, 1965) results from the nucleophilic attack of bi-
sulfite on disulfides:
RSSR1 + HS03^ - >- RSS03 + R'SH - 12-3
The reaction produces S-sulfonates (RSS03 ) and thiols (R'SH). Gunnison and Benton (1971) and
Gunnison and Palmes (1973) provided direct evidence for the formation of plasma S-sulfonates
ID vivo- Any plasma protein containing a disulfide group could react to form an S-sulfonate.
Small molecular weight disulfides, such as oxidized glutathione, can also be reactants.
Generally, analyses of plasma S-sulfonates have been restricted to diff usable (dialyzable or
small molecular weight compounds) and nondiffusable (nondialyzable or protein) S-sulfonates.
The exact molecular species have not been determined, and the results of these analyses
represent pools of the two groups of compounds. S-sulfonates can react with thiols, either
reduced glutathione or protein thiol groups, to form sulfite and disulfide. Since this
reverse reaction is facile, it is hypothesized that S-sulfonates are transportable forms of
bisulfite within the body. Sulfitolysis therefore represents both a mechanism of toxicity and
means of detoxification and redistribution of a reactive molecule, bisulfite.
Similar reversible nucleophilic addition of bisulfite to a variety of biologically
important molecules has been reported, but the toxicological importance of these chemical
species is uncertain. It is not likely, for example, that the reactions of bisulfite with
pyrimidine nucleotides (NAD or NADP), reducing sugars, or thiamine are important to the
toxicity of bisulfite or SOp.
Autooxidation of bisulfite occurs through a multistep chain reaction (Hayon et a!., 1972;
Backstrom, 1927; Fridovich and Handler, 1958, 1960; Asada and Kiso, 1973; Reiser and Yang,
1977; Yip and Hadley, 1966; Rotilio et a!., 1970; Nakamura, 1970; Klebanoff, 1961; Yang, 1967;
McCord and Fridovich, 1969a,b). These reactions may be important because they produce
hydroxyl (-OH) and superoxide (-Op ) free radicals as well as singlet oxygen (*0n)- These
12-3
-------
chemical species of oxygen, which are highly reactive and also produced by ionizing radiation,
are theoretically responsible for the lethal effects of ionizing radiation. Autooxidation of
bisulfite could lead to increased concentrations of these reactive chemical species within the
cell and hypothetical1y could lead to similar adverse effects. The reactive forms of oxygen
can also initiate peroxidation of the lipid bilayer of cells. Peroxidation of cellular lipids,
especially plasma membrane lipids, is thought to be highly deleterious (Kaplan et a!., 1975).
No direct evidence has been presented, however, to support peroxidation of cellular lipids as
a mechanism of toxicity of S0?.
Bisulfite addition to cytosine results in the formation of uracil. The reaction of bi-
sulfite with nucleic acids are as follows (Shapiro and Weisgras, 1970; Shapiro et al.,
1970a,b; Hayatsu, 1976):
m
Changes introduced by bisulfite reaction at specific locations in the genome were shown to pro-
duce mutants in SV40 with the expected DNA sequence change after replication _in vivo (Shortle
and Nathans, 1978).
While the respiratory effects of S0« may be due to sulfonation (Alarie, et al, 1973d),
the fact that all disulfides in the respiratory tract will likely undergo this reaction means
that no single protein or small molecular weight compound can presently be identified as the
target or receptor for SO,, in this toxic lesion.
12.2.1.2 Metabolism of Sulfur Dioxide
12.2.1.2.1 Integrated metabolism. There are several studies of the metabolism of exogenously
supplied SO?J sulfite, or bisulfite. While quantitative differences exist between inhaled SO,
and ingested bisulfite with regard to the rate of clearance of plasma S-sulfonates, one of the
key intermediates in sulfite metabolism (Gunnison and Palmes, 1973), no qualitative differ-
ences exist in the metabolism of inhaled S0? and injected or ingested bisulfite or sulfite.
12-4
-------
The importance of the appearance of plasma S-sulfonates (RSSO, ) lies in their potential abi-
lity to serve as a circulating pool of sulfite molecules (Gunnison and Palmes, 1973) as evi-
denced by the presence of S from S09 in nonpulmonary tissues such as ovaries (Frank et al.,
3
1967). Continuous inhalation of 26.2 mg/m (10 ppm) S02 resulted in 38 ± 15 nmole of plasma
S-sulfonates/ml in rabbits after about 4 days (Gunnison and Palmes, 1973). The clearance of
plasma S-sulfonates generated by either inhalation of S0~ or ingestion of sulfite in the drink-
ing water was exponential, exhibiting only a single compartment in most rabbits. The half-
life was 4.1 days for S-sulfonates generated by inhalation vs. 1.3 days for those generated by
ingestion (Gunnison and Palmes, 1973). The mechanism for this quantitative difference in
clearance rates has not yet been found.
Inhaled S09 quickly penetrates the nasal mucosa and airways as shown by the rapid appear-
35 35
ance of S in the venous blood of dogs inhaling S07 (Yokoyama et al.t 1971). A significant
35
fraction of the blood S was probably in the form of plasma S-sulfonates. Most of the in-
haled SO,, is presumed to be detoxified by the sulfite oxidase pathway (predominantly in the
liver but also in other organs), forming sulfate, which is excreted in the urine. The domin-
ance of this reaction has been supported by studies of sulfite oxidase inhibition (Cohen, et
al. 1972) that are discussed below, and by the appearance of about 85 percent of the inhaled
S09 as urinary sulfate in dogs (Yokoyama et al., 1971). A small fraction (10 to 15 percent)
35
of the urinary S was in the form of HpSO, esters. Sulfate arising from the oxidation of
sulfite can enter the sulfate pool and be incorporated into sulfate macromolecules including
glycosaminoglycans and glycoproteins. These macromolecules are actively synthesized by the
respiratory mucosa and may account for the presence of radiolabeled sulfur in the respiratory
35 35
tract following inhalation of S0? (Yokoyama et al. , 1971). Most of the nondialyzable S
detected by Yokoyama et al. (1971) was bound to the a-globulin fraction of plasma. The
35 35
chemical form of the S was not determined. Yokoyama et al. (1971) speculated that the S
present in the a-globulin fraction was in the form of sulfonated carbohydrates. The problem
needs further clarification. According to Gunnison and Palmes (1973), plasma S-sulfonated pro-
35
teins may also have contained the S. They have suggested that the slow clearance of plasma
S-sulfonates is an important factor in determining toxicity, but have not reported intra-
cellular levels of S-sulfonates or sulfite.
12.2.1.2.2 Sulfite oxidase. The biochemistry of sulfite oxidase is important as a mechanism
of detoxification of sulfite. Sulfite oxidase is a metallo-hemo protein with molybdenum and
protoheme as the prosthetic groups (Cohen et al., 1972). It exists in animals (Cohen et al.,
1972,1974; Howell and Fridovich, 1968; Cohen and Fridovich, 1971a,b; Wattiaux-DeConinck and
Wattiaux, 1974), bacteria (Lyric and Suzuki, 1970), and plants (Tager and Rautanen, 1955;
Arrigoni, 1959; Fromageot et al., 1960). In both plants and animals, the enzyme is located in
the mitochondria. Purified sulfite oxidase can utilize either cytochrome c or oxygen as the
12-5
-------
electron acceptor (Cohen and Fridovich, 1971a). When coupled with cytochrome c to the mito-
chondria! respiratory chain, sulfite oxidase reduces molecular oxygen to water (Equation
12-4), whereas during oxygen reduction, the product formed is hydrogen peroxide (Equation
12-5),
SQl2 2 Cyt c (Fe3+) H,0
.; ( .) (
SO/ 2 Cyt c (Fe^ ) 1/2
12-4
S0~2 + H20 + 02 * S0^2+ H202 12-5

Direct reduction of molecular oxygen by sulfite oxidase is prevented in the presence of ferric
cytochrome c. In intact mitochondria, therefore, sulfite oxidation occurs through the inter-
action of sulfite oxidase with the respiratory chain of the mitochondria, producing 1 mole of
ATP/mole of sulfite oxidized.
In three reported cases in humans, a rare genetic defect in sulfite oxidase resulted in
severe neurological problems (Hudd et al., 1967; Irreverre et al., 1967; Shih et a!., 1977;
Ouran et al., 1979). Dietary factors can, however, alter the enzymatic activity. Because
sulfite oxidase requires molybdenum, Cohen et al, (1973) were able to deplete rats of sulfite
oxidase by feeding them a low molybdenum diet and treating them with 100 ppm of sodium
tungstate in drinking water. Tungsten competes with molybdenum and essentially abolishes the
activity of sulfite oxidase and xanthine oxidase, the two major molybdo-proteins of rat liver.
Similar decreases were observed in the lung and other organs. The LD,,0 for interperitoneally
injected bisulfite'was found to be 181 mg NaHSO^/kg in the sulfite oxidase-deficient rats com-
pared to 473 mg/kg in the nondeficient animals.
Attempts to induce higher levels of sulfite oxidase through pretreatment of the rats with
SQ?/bisulfite or phenobarbital failed (Cohen et al. , 1973). Since sulfite oxidase is a mito-
chondria! enzyme with a long half-life, it is not likely that phenobarbital or chronic
exposure to SO, would result in adaptation through induction of higher levels of sulfite oxi-
dase.
12.2.1.3 Activation and Inhibition of Enzymes by Bisulfite—Both inhibition and activation of
specific enzymes have been reported. This may be due to formation of S-sulfonates, since di-
sulfide bonds often stabilize the tertiary structure of proteins. Sulfite ions activated
several phosphatases including ATP-ase (Marunouchi and Mori, 1967) and 2,3-diphosphoglyceric
acid phosphatase (Harkness and Roth, 1969). The mechanism by which activation occurs is un-
known. Inhibition of several enzymes has also been reported, including aryl sulfatase (Harkness
12-6
-------
and Roth, 1969), choline sulfatase (Takebe, 1961), rhodanase (Lyric and Suzuki, 1970), and
hydroxyl amine reductase (Zucker and Nason, 1955). Malic dehydrogenase was inhibited by
micromolar concentrations of bisulfite (Wilson, 1968; Ziegler, 1974). Other dehydrogenases
(Oshino and Chance, 1975) and flavoprotein oxidases are inhibited by bisulfite.
Bisulfite effectively inhibits a number of other., enzymes, including potato and rabbit
muscle phosphorylase (Kamogawa and Fukui, 1973). Bisulfite inhibition was competitive with
respect to glucose-1-phosphate and inorganic phosphate, suggesting that the bisulfite inhibi-
tion was caused by competition of bisulfite with the phosphate binding site of phosphorylase.
Several important coenzymes (such as pyridoxy!phosphate, NAD, NADP, FMN, FAD, and folic acid)
may react with sulfite to form additional products as discussed above. As a result, these co-
enzymes could theoretically aid in inhibiting a wide variety of critical enzymic reactions.
Pyridine coenzyme-bisulfite adduct (Tuazon and Johnson, 1977) and flavoenzyme-bisulfite adduct
(Muller and Hassey, 1969; Massey et al., 1969), which have been studied in detail, have been
shown to be biologically inactive.
Despite all of the data obtained using j_n vitro systems on the inhibition of enzymes by
bisulfite/S07, no inhibition or activation has been determined in vivo with SCL exposure.
C. ™™ £
Such inhibition may occur, but there has been no concerted effort to search for inhibition of
specific enzymes during S0? exposure.
12.2.2 Mortality
The acute lethal effects of S0« have been examined mostly in the older literature and
have been reviewed in the previous Air Quality Criteria Document for Sulfur Oxides (National
Air Pollution Control Administration, 1970), In early studies, several different animal
species were examined for susceptibility to SO-. These data show that neither rats nor mice
3
died at exposures of 65.5 mg/m (25 ppm) for up to 45 days, a conclusion confirmed by Laskin
et al. (1970). Mortality could be associated with long-term exposure to S0? at 134 mg/m (51
ppm) or higher. The clinical signs of SO, intoxication appear to vary with the dose rate
3
(Cohen et al. , 1973). At concentrations below approximately 1,310 mg/m (500 ppm), mortality
is associated with respiratory insufficiency; above this concentration, mortality is ascribed
to central nervous disturbances producing seizures and paralysis of the extremities. These
clinical signs depend upon the presence and activity of sulfite oxidase, as Cohen et al.
(1973) observed shorter survival times and higher mortality rates in tungsten-treated animals.
Injections of histamine or adrenalectomy can increase the lethality of S0_ (Leong et al.,
1961).
Matsumura (1970a,b) examined the effect of a 30-min exposure to several air pollutants on
mortality consequent to the anaphylactic response of guinea pigs to protein antigens.
Sensitization to the antigen administered by aerosol was augmented by pretreatm,ent with 786
mg/m3 (300 ppm) S02, but not with 472 mg/m3 (180 ppm).
12-7
-------
On the basis of mortality due to acute exposure, SO, is far less toxic than ozone and is
similar in toxicity to nitrogen dioxide. Concentrations required to produce mortality from
SO, are far in excess of those that occur in the atmosphere due to pollution (Table 12-1).
12.2.3 Morphological Alterations
Because of the high solubility of S0? in water and biological fluids, morphological and
physiological effects occur mostly in the upper airways; however, changes- have also been de-
o
tected in the lower airways (Table 12-2). At the relatively high concentrations (>26.2 mg/m ;
10 ppm) used in most studies designed to detect morphological alterations, most of the inhaled
SO, is removed by the extrathoracic (nasopharyngeal) cavity. (See Chapter 11, Section 11.2.4,
for an expanded discussion of S0? absorption.) In rabbits, the concentration of inspired SO,
determines how much is removed in the ET cavity as opposed to the tracheobronchial (TB) and
pulmonary (P) regions of the lung (Strandberg, 1964). At SO, concentrations greater than 26.2
3
mg/m (10 ppm), 90 to 95 percent is removed in the ET cavity. A small part, 3 to 5 percent,
is removed by the TB-P region. Thus, in this range, most of the dose is delivered to the
nasal turbinates with only a small percentage going to the lung parenchyma. At lower concen-
o
trations of inspired S0?, [such as 0.13 mg/m (0.05 ppm)] which are closer to ambient levels,
only 40 percent of the dose is absorbed by the ET cavity upon inspiration, while another 40
percent is removed by the respiratory tract upon expiration. Thus, at lower concentrations,
the actual percentage of SO, removed in specific regions of the respiratory tract is not known
precisely. In the dog, over 95 percent is removed by the upper airways and nose at concentra-
tions between 2.62 and 131 mg/m3 (1 and 50 ppm) S02 (Frank et al., 1967, 1969). A more de-
tailed consideration of SOp extraction by airways is given in Section 12.2.4 below.
Giddens and Fairchild (1972) pointed out that these differences in removal of inspired
SO* could explain the apparent anomaly of little damage to the lower respiratory tract at high
SO, concentrations. They studied the effects of inhaled SO, on the nasal mucosa of mice. Two
groups of mice were used; one group that was free of specific upper respiratory pathogens, and
a conventional group that was presumed to be infected or to have a latent infection of upper
3
respiratory pathogens. The mice were exposed continuously to 26.2 mg/m (10 ppm) SO- for a
maximum of 72 hours. Pathological changes in the nasal mucosa appeared after 24 h of exposure
and increased in severity after 72 h of exposure. Mice free of upper respiratory pathogens
had fewer pathological findings than the conventionally raised animals. Giddens and Fairchild
(1972) concluded that resident or acquired pathogens exacerbated the morphological changes.
These alterations were, however, qualitatively identical in both groups of animals. Cilia
werfe lost from the nasal mucosa, vacuolization appeared, the mucosa decreased to about one
half the normal thickness, and a watery fluid accumulated. Desquamation of the respiratory
and olfactory epithelia was evident. Alveolar capillaries were slightly congested, but edema
and inflammatory cells were absent. Martin and Willoughby (1971) reported loss of cilia, dis-
appearance of goblet cells, and metaplasia of the epithelium of the nasal cavity of pigs exposed
12-8
-------
TABLE 12-1. LETHAL EFFECTS OF SO,
SQ2 Concentration
•g/ra3 pp«
26.2
134
275
(See Text)
1,598
2,392
3,086
5,175
9,165
13,236
5,782
6,571
7,205
786
10
51
105

610
913
1,178
1,975
3,498
5,052
2,207
2,508
2,750
300
Duration
6 h/day x 5 day/wk
x 113 day
113 days
22 day
5 min/day x 5 day/wk
> x lifetime
LT5Q 285 6 rain
74.5 niln
38.7 min
LT5Q 197,6 Dim
71 7
41.0
LT50 68.2 min
28.7
35 5
30 mm
Species
Rat
II
11
Mice
Mice
(Connaught Med.
Res. Lab. Strain)
H
Rat (Sprague-
Dawley)
H
H
Guinea Pig
it
M
Guinea Pig
Remarks
No mortality in excess of control
No mortality in excess of control
64X mortality (treated-control)
No increased mortality; tumor formation
found
IP injection of 200 to 300 mg histamine/mouse
increased toxicity
it
IP injection of 200 to 300 mg histamine/rat
or adrenalectomy increased toxicity
ir

Increased mortality
due to anaphylaxis
Reference
Laskin et al. ,
1970
it
H
Peacock and °
Spence, 1967
Leong et al.
1961
H
Leong et al.
1961
11
'<
Leong et al.
1961
It
11
Matsumura,
1970a,b
from antigen challenge
to sensitized animals
-------
TABLE 12-2. EFFECTS OF
OH LUNG MORPHOLOGY
Concentration
Duration
Species
Results
Reference
0.34, 2.65, or 15.0 mg/m3
(0.13, 1.01, or 5.72 ppm)
S02
1 yr. , continuous
0.37, 1.7 or 3.35 mg/nr (0.14, 78 wk, continuous
0.64 or 1.28 ppm) S02

12,3 mg/m3 (4.69 ppm) then 30 wk then 1 h then
between 524 and 2620 mg/m3 48 wk
(200-1000 ppm) then 0 mg/m3 S02

13.4 mg/m3 (5.12 ppm) SQ2 18 mo, continuous
13.4 ng/m3 (5.1 ppm) S02

26.2 mg/m3 (10 ppm)-SOz
21 h/day, 620 days

72 h, continuous
91.7 »g/ra3 (35 ppm) [rose on 1 to 6 wk
occasion to 262 mg/m3 (100 ppm)]
S02

131, 262, 542, 786 mg/n3 (50, 3 h/day, 5 day/wk,
100, 200, 300 ppm) S02 6 wk
1048 mg/m3 (400 ppm) S02
3 h/day, 5 day/wk,
3 wk
Guinea pig Lungs of 15.0 mg/ra3 (5.72 ppm) group
showed less spontaneous pulmonary disease
than controls, and 0.34 and 2.64 »g/m3
(0.13 and 1.01 ppm) animals. Tracheitis
present in all but 15,0 mg/ttft (5.72 pp«)
group. Survival greater in the high dose
group.

Cynomolgus No remarkable morphologic alterations in
monkey the lung

Cynomolgus Persistent changes in lung morphomology,
monkey including alterations in the respiratory
bronchioles, alveolar ducts, and alveolar sacs.
Cynomolgus Ho alterations in lung morphology
monkey

Dog No alterations in lung morphology
Mouse Pathological changes in the nasal nucosa.
No alterations of tracheae; slight con-
gestion of alveolar capillaries, but
no alveolar edema (mice free of upper
respiratory pathogens were significantly
less affected than conventionally raised
animals.)
Pig Loss of cilia in nasal cavity, disappear-
ance of goblet cells, metaplasia of the
epitheliun
Rat Trachael goblet cells increased in number
and size. Incorporation of 3SS042 into
mucus increased. Sialidase resistant
mucus secreting cells were found much
more distally. Chemical composition of
mucus altered.
Rat Increased mitosis of goblet cells. Alter-
ation not lost by 5 wk postexposure.
Alarie et al., 1970
Alarie et al., 1972, 1973c

Alarie et al., 1972, 1973c

Alarie et al., 1975

Lewis et al., 1973
Giddens & Fairchild, 1972
Martin & Willoughby, 1971
Reid, 1970
Lamb and Reid, 1968
1ppm S02=2.62 mg/m3.
-------
3
to 91.7 mg/m (35 ppm) SCL for 1 to 6 weeks. This study, however, was marred by difficulties
3
with the control of the SO, concentration, which rose on occasion to 262 mg/m (100 ppm), and
with high relative humidity (RH) occurring during cleaning of the pig pens.
Lamb and Reid (1968) and Reid (1970) attempted to use SQ?-exposed rats as a model of
human chronic bronchitis.. They presented arguments that SO,,-induced bronchial hyperplasia is
analogous to human chronic bronchitis. "Most of their Studies were carried out at high concen-
trations of S0? (1,048 mg/m or 400 ppm S0~ for 3 h/day, 5 days/wk) for up to 6 weeks. Under
these conditions, the trachea! goblet cells clearly increased in number and size. The goblet
cell density also increased in the proximal airways, main bronchi, trachea, and distal air-
ways, with proximal airways and main bronchi showing the largest changes. The incorporation
35
of S-sulfate into mucus by goblet cells also increased with exposure, reaching a plateau at
approximately 3 weeks. The effects of SO, were concentrated in the central airways, again
suggesting that the solubility of SO, in water limits its accessibility to the periphery.
Mitosis reached a maximum after 2 or 3 exposures and declined rapidly as injured cells were
replaced. On repeated exposure up to 6 wk, mitosis remained elevated in the proximal airways
compared to the distal airway, in which the mitotic index returned to the control level. The
3
magnitude was proportional to the SO, concentration up to 524 mg/m (200 ppm) but was less at
3 3
786 mg/m (300 ppm). The mitotic index rose at S0? concentrations as low as 131 mg/m (50
ppm) when given for 3 h/day, 5 days/week. Major changes in the goblet cell type or substance
produced by the goblet cells were also detected. Goblet cells, which produce mucus resistant
to digestion by sialidase, increased in number, and their distribution extended distally from
the upper bronchioles towards the respiratory bronchioles. Since each molecular type of mucin,
sialidase-resistant or -susceptible, could be produced by one type of goblet cell, or each
goblet cell could produce different mucins, these results can be interpreted in two ways.
The elaboration of a specific type of goblet cell could occur, or more goblet cells could be
produced but with a change in their biochemical function towards sialidase-resistant mucins.
Goldring et al. (1970) observed similar changes in morphology in Syrian hamsters exposed
to 1700 mg/m (650 ppm) S02 for 4 h/day over several weeks. These alterations included stimu-
lation of mucus cell secretion and "dysplastic" changes in the bronchial epithelium. While
these studies, along with those of Lamb and Reid, present an interesting means of studying
experimental bronchitis, they do not provide evidence that ambient SO, levels cause similar
changes.
Alarie et al. (1970) examined the tissues of guinea pigs exposed continuously to 0, 0.34,
2.65, or 15.0 mg/m3 (0, 0.13, 1.01, or 5.72 ppm) SO, for 1 year. The lungs of the guinea pigs
3
exposed to 15.0 mg/m (5.72 ppm) and killed after 13 or 52 wk of exposure showed less sponta-
neous pulmonary disease than the control group. The prevalence of pulmonary disease in the
control groups, which was not observed prior to exposure, suggests that they acquired pulmonary
12-11
-------
disease during the exposure period. This and other studies by Alarie and coworkers (1972,
1973c, 1975), were limited to light microscopic observations of conventional hematoxylin-eosin
stained paraffin sections. The results are of limited value compared to more recent
approaches using scanning electron microscopy of surfaces, transmission electron microscopy of
organelles, or morphometric techniques. The control group, as well as those exposed to 0.34
o
and 2.64 mg/m (0.13 and 1.01 ppm) S0?, had evidence of lung disease as shown by histocytic
infiltration of the alveolar walls. Tracheitis was also present in these three groups, but
3
not in the 15.0 mg/m (5.72 ppm) group. The latter group developed hepatocyte vacuolation,
but the pathological significance of this change needs further investigation. The authors did
not address this-issue. The survival was greater (p <0.05) in the 15.0 mg/m (5.72 ppm) group
than in the other groups, Including the air-exposed control group. The possible effects of
SCL in the Alarie et al. (1970) study, however, cannot be determined accurately because of the
disease in the control animals.
Alarie et al. (1972, 1973c) subsequently exposed cynomolgus monkeys continuously to 0.37,
1.7 or 3.35 mg/m (0.14, 0.64 or 1.28 ppm) S09 for 78 wk but found no remarkable morphological
o
alterations. Another group exposed to 12.3 mg/m (4.69 ppm) SO, for 30 wk was accidentally
3
exposed to concentrations of SO, not higher than 2,620 mg/m (1,000 ppm) or lower than 524
3
mg/m (200 ppm) for 1 h, after which they were placed in a clean air chamber and held for
48 more weeks. Persistent changes were noted in this group. Alterations in the respiratory
bronchioles, alveolar ducts, and alveolar sacs were found. Proteinaceous material was found
within the alveoli. The distribution of such lesions was focal, but was observed within all
lobes of the lung. Alveoli containing proteinaceous material were generally those that arose
directly from respiratory bronchioles. Alveolar walls were thicker and infiltrated with his-
tocytes and leukocytes. Macrophages were present within these foci. Moderate hyperplasia of
the epithelia of the respiratory bronchioles was found, and frequently the lumina of the respi-
ratory bronchioles were plugged with proteinaceous material, macrophages, and leukocytes.
Bronchiectasis and bronchiolectasis were present in 8 of 9 monkeys. Vacuolation of hepato-
cytes was also observed, as with the guinea pig group exposed to 15.0 mg/m (5.72 ppm) S02 in
the prior Alarie et al. (1970) study.
3
In a replication of this study, cynomolgus monkeys were exposed to 13.4 mg/m (5.12 ppm)
S02 continuously for 18 mo (Alarie et al., 1975). No alterations in lung morphology were re-
ported to be due to S0?. The morphological alterations reported in the control group included
lung mite infections and associated "slight subacute bronchiolitis, alveolitis, and bronchi-
tis." Pulmonary function measurements were made in the above mentioned studies (Alarie et
al., 1970, 1972, 1973c, 1975) and are described in Section 12.2.4.
The absence of SO^-induced morphological alterations as reported by Alarie et al. (1970,
1972, 1973c, 1975) and Lewis et al. (1973), who exposed dogs for 620 days (21 h/day) to 13.4
mg/m (5.1 ppm) S02, is not unexpected considering the transient bronchoconstriction induced
12-12
-------
by acute SO, exposure reported by Amdur (1973) at lower concentrations (see Section 12.2.4).
Alarie et al. (1970) pointed out, "As recent literature attests, there is also an obvious lack
of knowledge about the correlation between subtle microscopic alterations in the lung and con-
comitant changes in this physiological parameter (lung function)." Further, the transient
nature of the pulmonary function effects observed during short-term exposures would be diffi-
cult to detect morphologically unless the lungs were fixed during the time of exposure. Even
then, if the cause of the increased pulmonary resistance were a subtle alteration of smooth
muscle tone, as has been hypothesized, it might be morphologically undetectable.
Most of the studies in which the lungs of S0?-exposed animals have been examined center
around tracheitis, bronchitis, ulceration, and mucosal hyperplasia (Table 12-2; see also Reid,
1970). The lowest concentrations of S09 at which these alterations have been reported have
3
been in the rat at 131 mg/m (50 ppm) for 30 to 113 days (Reid, 1970). At higher concentra-
tions (1,048 mg/m or 400 ppm S0? for 3 h/day, 5 days/wk for 3 wk), recovery to normal morpho-
logy did not occur after 5 wk postexposure (Reid, 1970).- The possibility of recovery from
lower concentrations and shorter durations of exposure .is not known (Lamb and Reid, 1968;
Reid, 1970). Discounting their first study (Alarie, et al., 1970), where the control group of
guinea pigs had a higher level of pulmonary infection than the exposed groups, Alarie et al,
3
(1973c) reported no effect from SOp exposure up to 5 ppm (13.1 mg/m ). These observations
were, however, restricted to light microscopy and did not include scanning or transmission
electron microscopic observations. This group also reported no observable effects at 0.37,
1.7 or 3,35 mg/m3 (0.14, 0.64, and 1.28 ppm) S09 (Alarie et al., 1975). The group of monkeys
3
exposed to 12.3 mg/m (4,69 ppm) is difficult to evaluate due to the accidental exposure to
high levels of S0? (Alarie et al., 1972). No effects were reported for monkeys exposed to
13.5 or 13.7 mg/m3 (5.15 or 5.23 ppm) S02 and HpSO. aerosols at 0.10 mg/m3 or fly ash at 0.44
mg/m3 (Alarie et al., 1975).
Because of the transient bronchoconstrictive effects of SO/,, conventional light micro-
scopic morphological studies are not likely to be useful in evaluating the acute effects of
2
S02 exposure. Persistent alterations have not been noted at less than 131 mg/m (50 ppm).
12.2.4 Alterations in Pulmonary Function
Changes in breathing mechanics have been among the most sensitive parameters of SQ« toxi-
city. They have also been useful in studying the effects of aerosols alone or in combination
with S0? (Sections 12.3.3.1 and 12.4.1.1). A variety of methods have been used, some of which
have been applied to human exposures. A method for measuring increases in flow resistance due
to bronchoconstriction in guinea pigs has been developed by Amdur (Amdur and Mead, 1955, 1958).
Animals are not anesthetized and breathe spontaneously, allowing sensitive measurements of
pulmonary function. Another method by Alarie and coworkers (1973d) measures changes in respi-
ratory rate.
Several investigators (Nadel et al. 1965a; Corn et al. , 1972; Frank and Speizer, 1965;
Balchum et al., 1960; Nadel et al. , 1965b) found that bronchoconstriction resulted from both

12-13
-------
head-only and lung-only exposures in cats and dogs. When corrected for the amount of SCL
hypothesized to reach the lung, Amdur's study (1966) with guinea pigs showed that SCL is
highly effective in producing bronchoconstriction through direct exposure of the lung. Two
sets of receptors are involved in the response of animals to SO,. At high concentrations of
SOp or following lengthy exposure, the nasopharyngeal receptors fatigue or become unrespon-
sive, whereas the bronchial receptors- do not (Alarie, 1973).- Widdicombe (1954a) originally
described the receptors responsible for the S0?-initiated bronchoconstriction. The broncho-
constriction is initiated through the activation of bronchial epithelial chemoreceptors whose
efferent and afferent pathways are through the vagus nerves (Nadel et a!., 1965a,b; Grunstein
et a!., 1977; Toraori and Widdicombe, 1969). Chilling the vagus prevents conduction of nervous
impulses produced on inhalation of SO,. Other receptors located in the same regions of the
lung respond to mechanical stimulation and particles such as talc (Widdicombe, 1954b;
Widdicombe et a!., 1962). Intravenous injection of atropine blocks the efferent impulses, pre-
sumably at the cholinergic preganglionic synapse (Grunstein et al., 1977). Sulfur dioxide-ini-
tiated bronchoconstriction involves smooth muscle contraction, since p-adrenergic agonists
such as isoproterenol reverse the SQp-bronchoconstriction (Nadel et al., 1965a,b). Histamine
may be involved in this response, as implied by other studies of hyperreactive airways
(Boushey, et al. 1980), but definitive proof of histamine involvement is not available. Re-
lease of acetylcholine could also cause increased mucus secretion as noted during SO,, ex-
posure. Chronic exposure to SO^ could lead to mucus hypersecretion and altered airway caliber.
Cholinomimetic drugs and histamine applied as aerosols mimic the SO,-initiated bronchoconstric-
tion (Islam et al., 1972). Using anesthetized, intubated, spontaneously breathing dogs ex-
posed to 2,62, 5.24, 13.1, or 26.2 mg/m3 (1, 2, 5, or 10 ppm) S02 for I h, Islam et al.
(1972) found an increased bronchial reactivity to aerosols of acetylcholine, a potent broncho-
3 3
constrictive agent. The greatest response occurred at 5.24 mg/m (2 ppm), although 2.62 mg/m
(1 ppm) also caused an effect. The effect at 26.2 mg/m (10 ppm) was less than that at 2.62,
5.24, and 13.1 mg/m (1, 2, and 5 ppm). While these results could suggest that SO™ may modify
bronchial reactivity, the authors point out that the reactions were highly reversible and
occurred in ranges where alterations can also be produced by the inhalation of saline aerosols.
Cholinomimetic drugs act through either the same autonomic reflex arc or directly upon the
cholinergic receptors on smooth muscles and mucus secreting cells and glands. As discussed in
Chapter 13, SO, also produces bronchoconstriction in man through the same autonomic reflex arc.
Exposure to SO- increases resistance to air flow in guinea pigs that can be repeated by
numerous exposures over several hours and exhibits none of the tachyphylaxis found with other
species (Corn et al., 1972; Frank and Speizer, 1965). However, different techniques were used
for different species. In a review of her data, Amdur (1973), reported that for a 1-h ex-
3 ' 3
posure, a mean of 0.68 mg/m or 0.26 ppm (range of 0.08 to 1.57 mg/m or 0.03 to 0.6 ppm) was
12-14
-------
the lowest concentration of SOn that increased flow resistance in guinea pigs. The response,
a 12.8 percent increase (p < .001) at these low levels of SCL (Amdur, 1973), was the average
for 71 guinea pigs; the individual data points were reported in other publications (Amdur.and
Underfill!, 1968, 1970; Amdur, 1974). For a 1-h exposure, the lowest concentration these re-
o
searchers tested that caused an increase (p <0.01) in resistance was 0.42 mg/m (0.16 ppm) SO,,
(Amdur and Underbill, 1970). In a more recent study, Amdur et al. (1978) showed that a 1-h
exposure of guinea pigs to 0.84 mg/m (0.32 ppm) SO, caused a 12 percent increase in resist-
ance (p <0.02) and a nonstatistically significant decrease in compliance. Investigations of
o
the interaction of oil mists and SO^ showed that 2.62 mg/m (1 ppm) S0?, the lowest concentra-
tion used, significantly increased resistance (Costa and Amdur, 1979a,b). At concentrations
of SOp below 2.62 mg/m (1 ppm), the response of individual animals varied considerably
(Amdur, 1964, 1973, 1974). Of 1,028 guinea pigs, 135 were "susceptible," responding to low
concentrations of SOp (and a variety of irritants) with greater changes in resistance than the
predicted mean. Amdur cites comparative data for other species, including man, to suggest
that a certain fraction of all subjects may exhibit this phenomenon (Amdur, 1973, 1974;
Horvath and Folinsbee, 1977). It might also be suggested that some groups of animals by
chance may not have a "susceptible" individual. In one study (Amdur et al., 1978), 3 groups
o
of 10 animals each exposed to 0.52, 1.05, or 2.1 mg/m (0.2, 0.4, or 0.8 ppm) S0? had no sig-
nificant increase in airway resistance above the control values. It must be noted, however,
that these results might also reflect intraspecies differences in susceptibility.
Based on data from earlier work (Amdur and Underbill, 1968), Amdur concluded that 10 to
13 percent of the guinea pig population is very much more responsive than the average (Amdur,
1974). Cats (Corn et al., 1972) and dogs (Frank and Speizer, 1965), on the other hand, rarely
o
were found to be sensitive to short-term (< I h) exposure to 52.4 mg/m (20 ppm) S0? (cats) or
3
18.3 mg/m (7 ppm) SOp (dogs). Even with the relatively small sample sizes used, some cats
and dogs responded and others did not.
Some of the problem of "susceptible" vs. "nonsusceptible" members of the experimental pop-
ulations can be understood if one assumes that the response to a given toxicant, such as S02,
is the result of a number of different genes within the population and not just a single gene.
In that case, a single individual could have a number of recessive or dominant genes that
could contribute to either "susceptibility" or "nonsusceptibility." Since experimental ani-
mals and human subjects are selected as randomly as possible (in most experimental designs),
there is a maximal chance of getting some "susceptible" responders in each experiment. The
total number of "susceptible" responders will be small and variable because of the low incid-
ence of "susceptible" responders in the general animal population, but will tend to shift the
dose- or concentration-response curve toward lower concentrations and to decrease the slope of
the curve (e.g., when the data are expressed as the log-probit transformation). Such
phenomena have been studied in detail for "resistant" insects that have different genomes
responsible for increased detoxification mechanisms. In the case of SCL, the matter is further

12-15
-------
complicated by comparisons between groups of animals and different strains or species. Even
With guinea pigs, the total number of animals examined to date (about 1,000 to 2,000) is too
small to give more than a crude estimate of those animals that have a "sensitive" genome. The
incidence of "susceptibility" in the guinea pigs (about 13 percent) is too low to have been
detected clearly in the 100 or so cats and dogs used in SOp experiments. Here only 1 or 2
"susceptible" animals would have been encountered in each experiment. Further, the small
number of animals has been studied in different laboratories and at different times, and the
animals have come from different genetic stocks. It is fortuitous that Amdur's laboratory has
persisted in these studies with the same animal, the guinea pig, and the same general method,
so this low incidence of "susceptibility" could be detected. While the mechanism(s) respon-
sible for "susceptibility" is not known, the question of "susceptibility" is an important
aspect deserving further study. A similar incidence of "susceptible" individuals found to
exist in man would present a major health problem. Adverse reactions might conceivably occur
o
among these individuals at exposures less than 2.62 mg/m (1 ppm) (Amdur, 1964, 1973, 1974),
which are encountered in ambient air. However, the frequency of susceptibility to S0? in man,
as well as the physiological or biochemical basis of such susceptibility, is not known.
A broad dose-response curve has been noted also for histamine-initiated bronchoconstric-
tion in man (Habib et al., 1979), guinea pigs (Douglas et a!., 1973, 1977; Brink et al.,
1980), dogs (Loring et al., 1978; Snapper et al., 1978), and monkeys (Michoud, 1978). Among
12 normal human subjects, a 38-fold range of inhaled histamine was observed in both the thres-
hold and median doses causing bronchoconstriction (Habib et al. , 1979). The concentration re-
quired to produce a 50-percent change in dynamic lung compliance in 131 female guinea pigs
varied over a 100-fold range (Douglas et al., 1973). While the interindividual amount varied
considerably, the values were lognormally distributed, indicating a single population
(Douglas et al., 1973, 1977). Dogs showed a 40-fold variation in histamine concentration
needed to initiate changes in airway diameter (Snapper et al., 1978). These values were also
log-normally distributed, indicating a single population among the 102 mongrel dogs examined.
A wide interindividual variation for histamine- and methacholine-initiated bronchoconstriction
was found among 8 rhesus monkeys, some of which were sensitive to Ascaris suum allergen
(Michoud, 1978). The sensitivity to histamine or methaeholine was not associated with Ascaris
sensitivity, however. While genetic differences in histamine sensitivity have been found in
quinea pigs, naturally occurring or acquired allergic reactions are not likely to cause the
large interindividual differences in sensitivity in either guinea pigs (Takino et al., 1971)
or monkeys (Michoud et al., 1978). A further complicating factor is the age-dependence of
histamine- and other drug-initiated bronchoconstriction (Brink et al., 1980). Younger guinea
pigs are more sensitive to histamine than older animals. This decreasing bronchial reactivity
to histamine with age in the guinea pig has been suggested as a model of human juvenile asthma.
12-16
-------
Human bronchial hyperreactivity does not seem to decrease with age in the same manner,
however (Boushey et al., 1980). While large interindividual differences apparently occur with
a wide variety of chemical agents causing bronchial reactivity in both man and animals, the
response of the same individual is quite reproducible regardless of species. The variability
in the lowest dose of SOp needed to evoke a given bronchoconstriction (measured, for example,
as an increased resistance to flow by the studies of Amdur) is apparently an inherent part of
the bronchial response to a broad range of chemicals and is not an artifact of the method.
Similar variations in lowest effective doses for S02 are likely to occur in man, judging from
the variability of response to inhaled histamine. The general observation that asthmatic
patients appear to be hypersensitive to a broad range of chemical and physical agents initiat-
ing bronchoconstriction (Boushey et al., 1980; see also Sect. 13.2.3) supports the contention
that the most susceptible animal species might possibly be used experimentally as a surrogate
for man. A major difference in pharmacology may exist between the guinea pig and man, however.
For example, autonomic mediators interact with hi stamina-induced bronchial reactivity in
guinea pigs but not in man, and beta adrenergic blockade by propranolol causes no difference
in bronchial reactivity in man (Habib et al., 1979) but potentiates histamine reactivity in
the guinea pig (Douglas et al., 1973). Insufficient numbers of animals and subjects have been
examined to predict the general shape of the dose-response curve for the human population,
even excluding the hypersensitive asthmatic population. These variations in interindividual
dose needed to evoke a specific amount of increased resistance to flow in guinea pigs by S0«
likewise apply to the measurement of increased resistance to flow evoked by aerosols, as dis-
cussed below in Section 12.3.3.
Using Strandberg's (1964) data from the rabbit to correct for the concentration of S0?
hypothesized to reach the lung, Amdur (1966) was able to normalize the concentration-response
curve for 50,,-induced bronchoconstriction in the guinea pig resulting from nose-only exposures.
3
A break occurs in the concentration-response curve at about 52.4 mg/m (20 ppm) SO^, perhaps
due to the poorer extraction of gaseous S0« by the upper airways at low concentrations. It
should be recognized, however, that S0,-extraction data for rabbits (Strandberg, 1964) and
dogs (Frank et al., 1967; Balchum et al., 1960; Frank et al., 1969) are in some conflict and
that the data for rabbits are not clear with respect to the site of SO.-, removal. Thus, use of
the rabbit data for guinea pig studies can be done only hypothetically. Sulfur dioxide intro-
duced directly into the lung by a trachea! cannula was much more effective in producing
bronchial constriction. Amdur (1966) suggested that, at concentrations of 1.05 to 1.31 mg/m
(0.4 to 0.5 ppm), very little removal of SO, occurs in the upper airways. These data contrast
with the radiotracer studies in dogs (Frank et al., 1967, 1969; Balchum et al., 1960). Others
2
have required concentrations greater than 18.3 mg/m (7 ppm) to evoke increases in flow resist-
ance in anesthetized cats (Corn et al., 1972) and dogs (Frank and Speizer, 1965). Differences
in the sensitivity of the two models may lie in the use of anesthesia, in the use of different
species, or in a different incidence of "susceptible" individuals.

12-17
-------
Animals chronically exposed to S0? have also been examined for alterations in pulmonary
function. Guinea pigs exposed continuously to 0.34, 2.64, or 15 mg/m3 (0.13, 1.01, or 5.72
ppra) SOp for up to 1 yr showed no changes in pulmonary function; however, spontaneous pul-
monary disease was present in all animals (including controls) except those exposed to the
highest concentration (Alarie, et a!., 1970). Dogs exposed for 21 h/day to 13.4 mg/m (5.1
ppm) SOp for up to 225 days demonstrated increased pulmonary flow resistance and decreased
lung compliance (Lewis et al., 1969). After exposure for 620 days, the mean nitrogen washouts
of dogs increased (Lewis et al., 1973)., Alarie and coworkers (Alarie et al., 1972, 1973c,
1975) exposed cynomologus monkeys continuously to 0.37, 1.7, 3.4, or 13.4 mg/m (0.14, 0.64,
1.28, or 5.12 ppm) S02, The latter concentration was used in an 18-mo study, whereas the
others were used for 78-wk exposures. Pulmonary function remained unchanged in all of these
groups. After 30 wk of exposure to 12.3 mg/m (4.69 ppm) SO,, the monkeys were inadvertently
3
exposed to concentrations between 524 and 2,620 mg/m (200 and 1000 ppm) for 1 hour. This
treatment resulted in pulmonary function alterations that persisted for the remaining 48 wk of
the study, during which the animals were exposed to clean air. Morphological alterations were
also seen in this group (see Section 12.2.3).
The respiratory rate of mice has been used as an indication of sensory irritation by
Alarie et al. (1973d). Mice were exposed for 10 min to 0, 44.5, 83.8, 162, 233, 322, 519, or
3
781 mg/m (0, 17, 32, 62, 89, 123, 198, or 298 ppm) SO,. About a 12 percent decrease in res-
3
piratory rate was observed at 44.5 mg/m (17 ppm). The respiratory rate decreased inversely
to the logarithm of the concentration of inspired S0?. The decrease in respiratory rate was
transient, however, as complete recovery to control values occurred within 30 min following
all exposures to S0_. The time for maximum response was inversely related to the logarithm of
3
the concentration of S02, being shortest at highest concentrations. Mice exposed to 262 mg/m
(100 ppm) S0« for 10 min were allowed to recover in clean air prior to a subsequent 10 min ex-
posure to the same concentration. As the length of the recovery period was decreased (from 12
min to 3 min), the effect of the subsequent SO. exposure on respiratory rate was lessened.
"Desensitization" thus appeared to occur during the course of exposures. When another
irritant, aerosols of chlorobenzilidene malononitrile (CBM), was used during the refractory
period following SO, exposure, the respiratory rate decreased at a rate comparable to that
following exposure to CBM alone. Thus, the refractory period associated with SO- exposures
appeared specific to SO- and not to CBM. When 262 to 328 mg/m3 (100 to 125 ppm) SOg was pro-
vided repeatedly for 90 sec, with each exposure separated by a 60-sec recovery period, the re-
fractory period was cumulative. Ten such exposures eventually abolished all respiratory rate
responses to S0«- Breathing clean air for 60 min resulted in a return of the response to
initial levels. When mice were exposed to S0? by means of a trachea! cannula, no changes in
the respiratory rate were observed, indicating that the decrease in respiratory rate was
mediated by a reflex arc. This concept has been developed in considerable detail in an exten-
sive review by Alarie (1973), who suggests that stimulation and desensitization occur via

12-18
-------
cholinergic nerve endings of the afferent trigeminal and glossopharyngeal nerves, allowing
activation of receptors in the nose and upper airways. Alarie et al. (1973d) also suggest
that S0? is hydrated to bisulfite and sulfite that react with a receptor protein to form an
S-thiosulfate and a thiol, cleaving an existing disulfide bond. The receptor protein slowly
regenerates to its original disulfide configuration by the oxidation of S-thiosulfide and free
thiol moieties of the receptor protein to disulfide. No direct evidence for this hypothesis
has been presented, however.
In summary, decreases in respiratory rate or increased resistance to flow are reprodu-
cible end points. There are at least two sets of receptors responsible for these changes in
respiratory function in animals acutely exposed to SO,. Increased resistance to flow results
3
from S0« concentrations as low as 0.42 mg/m (0.16 ppm) in guinea pigs. Of the animals so
far examined, guinea pigs are the most sensitive to S0~. The reason for this is not known,
but potential factors include species, strains, and experimental technique used. Large inter-
individual differences in dose-response curves for changes in pulmonary resistance to airflow
exist in all species. The exact number of animals responding to a given dose depends on the
shape of the dose-response curve. The nature of the dose-response curve at low levels is
poorly understood and has not been investigated directly. While pulmonary function in guinea
pigs appears to be highly sensitive to acute S02 exposures, it has not been proven that
chronic SOp exposures have a similar effect. Chronic studies with guinea pigs are unclear
because of disease in the control group. In other chronic studies, pulmonary function of
monkeys was unchanged at S00 concentrations up to 13.4 mg/m (5.12 ppm); dogs were affected by
3
225, but not 620, days of exposure to 13.4 mg/m (5.1 ppm). High levels of S02 likely to ini-
tiate airway narrowing and hypersecretion of mucus do alter several parameters of pulmonary
function. These results are not contradictory in view of the physiology of SO^-initiated
bronchoconstriction. Sulfur dioxide appears to cause bronchoconstriction through action on
the smooth muscles surrounding the airways. Since smooth muscles fatigue or become adjusted
to altered tone over time, chronic exposure to S0« is not likely to cause a permanent altera-
tion in bronchial tone. Unfortunately, investigations of the reactions of the airways after
chronic exposure to S02 have not appeared. We do not know if chronic exposure to SO^ causes
an alteration in response to S0? itself, since only direct measurements of pulmonary function
were made on the animals after chronic exposure. It would be informative to learn if chronic-
ally-exposed monkeys, for example, were more or less sensitive to SO, (Table 12-3).
12.2.5 Effects on Host Defenses
Because alterations in the ability to remove particles from the lung could lead to in-
creased susceptibility to airborne microorganisms or increased residence times of other non-
viable particles, the effects of S0? on particle removal and engulfment, as well as on inte-
grated defenses against respiratory infection, have been studied. Cilia function does not
appear to be affected by exposure. No changes were observed in the cilia beat frequency o'r
12-19
-------
TABLE 12-3. EFFECTS OF S02 ON PULM0KARY FUKCTIOH
Concentration Duration Species Results Reference

0.37, 1.7, 3.4, or 13.4 mg/m3 72-78 wk, continuous Cynomologus No change Alarie et al., 1972,
(0.14, 0.64, 1.28, or 5.12 ppm) monkey 1973c, 1975
S02

0.42 or 0.84 mg/m3 (0.16 or 1 h Guinea pig Increase In airway resistance Amdur et al., 1970, 1978a
0.32 ppm) SOa

0.52, 1.04, or 2.1 mg/m3 (0,2, 1 h Guinea pig No significant increase in airway resistance Amdur et al., 1978c
0.4, or 0.8 ppm) S02

2.62, 5.24, 13.1, or 26.2 mg/nt3 1 h Dog Increased bronchial reactivity to aerosols of Islam et al., 1972
(1, 2, 5, or 10 ppm) S02 acetylcholine, a potent bronchoconstrictive agent

13.4 mg/m3 (5.1 ppm) S02 21 h/day, 225 and Dog Increased pulmonary flow resistance and decreased Lewis et al , 1969,1973
620 days lung compliance at 225 days; increased nitrogen
washout at 620 days

0, 44.5, 83.8, 162, 233, 322, 10 min House Decreased respiratory rate proportional to the log Alarie et al., 1973d
519, or 781 mg/m3 (0, 17, 32, of the concentration; complete recovery within 30
52, 89, 123, 198, or 298 ppm) min. The time for maximum response was inversely
S02 related to the log of the concentration.

1 ppni S02 = 2.62 mg/m3.
-------
the relative number of alveolar macrophages laden with particles in rats exposed to 2.62 or
3 3
7.86 mg/m (1 or 3 ppm) SO, and graphite dust (mean diameter 1.5 |jm, 1 mg/m ) for up to 119
consecutive days (Fraser et al., 1968). Donkeys (Spiegelman et al., 1968) were exposed by
nasal catheters to 68.1 to 1,868 mg/m (26 to 713 ppm) SO, for 30 min. Clearance was not
, z ,
affected below 786 mg/m (300 ppm), but at high concentrations (786 to 1,868 mg/m or 376 to
713 ppm), clearance was depressed.
Ferin and Leach (1973) exposed rats to 0.26, 2.62, and 52.4 mg/m3 (0.1, 1, or 20 ppm) S02
for 7 h/day, 5 days/wk, for a total of 10 to 15 days and then measured the clearance of an
aerosol of titanium oxide (TiOp). The aerosol was generated at about 15 mg/m (1.5 |jm MMAD,
0 =3.3). These investigators took the amount of TiO« retained at 10 to 25 days as a measure
of the "integrated alveolar clearance." Low concentrations of S09 (0.26 mg/m or 0.1 ppm)
3
accelerated clearance after 10 and 23 days, as did 2.62 mg/m (1 ppm) at 10 days. By 25 days,
however, clearance was decreased with 1 ppm. Hirsch et al. (1975) found that the tracheal
q
mucus flow was reduced in beagles exposed for 1 yr to 2.62 mg/m (1 ppm) S02 for 1.5 h/day, 5
days/week. No differences in pulmonary function wer"e reported. Confirmation of this study
and determination of the persistence of the decreased mucus flow at this low level of SO,
would be important in light of other data available.
Sulfur dioxide may have more of an effect on antiviral than on antibacterial defense
mechanisms. Bacterial clearance was not depressed or altered in guinea pigs exposed to 13.1
or 26.2 mg/m3 (5 or 10 ppm) S02 for 6 h/day for 20 days (Rylander, 1969; Rylander et al.,
1970). Using the infectivity model (see Section 12.3.4.3), Ehrlich et al. (1978) found that
3
short (3 h/ day for 1 to 15 days) or long (24 h/day for 1 to 3 mo) exposures to 13.1 mg/m (5
ppm) SOp did not increase mortality subsequent to a pulmonary streptococcal infection. Virus
infections, however, are augmented by simultaneous or subsequent SO, exposure. Mice were ex-
3
posed to concentrations varying from 0 to 52.4 mg/m (0 to 20 ppm) SO, continuously for 7 days
•3
(Fairchild et al., 1972). Nice breathing 18.3 to 26.2 mg/m (7 to 10 ppm) SO, had an increase
3
in pneumonia. Lung consolidation was significant at 65.5 mg/m (25 ppm), but not at 26.2 or
39.3 mg/m (10 or 15 ppm). The rate of growth of the virus within the lung was unaffected by
S09 exposure. Further analysis of the data (Lebowitz and Fairchild, 1973) indicated that SO,
3
and virus exposure produced weight loss at concentrations as low as 9.43 mg/m (3.6 ppm).
Exposure to SO,, whether alone or in combination with a viral agent, had more of an effect on
weight reduction than on pneumonia. Since Giddens and Fairchild (1972) showed that mice with
apparent respiratory infection were more susceptible to S02 (Section 12.2.3), a rebound effect
may be possible in which SO, and microbial agents each potentiate the effect of the other.
Several studies of the effects of SQ~ on alveolar macrophages have been conducted, since
these cells participate in clearance of viable and nonviable particles in the gaseous exchange
regions of the lung. Rats were exposed for 24 h to 2.62, 13.1, 26.2, and 52.4 mg/m3 (1, 5,
10, and 20 ppm) SO- and their alveolar macrophages investigated by Katz and Laskin (1976).
12-21
-------
Exposure to the two highest concentrations increased i_n vitro phagocytosis of latex spheres
o
for up to 4 days in culture. At 13.1 mg/m (5 ppm) S02, phagocytosis was increased after 3 or
4 days in culture, but not after 1 or 2 days. Histochemical studies of pulmonary macrophages
from rats which had been exposed to 786 mg/m (300 ppm) SO, for 6 h/day on 10 consecutive days
showed no changes in the lysosomal enzymes, p-glucurom'dase, p-galactosidase, and N-acetyl-p-
glucosaminidase (Barry and Mawdesley-Thomas, 1970). Acid phosphatase activity was markedly
increased. This is in agreement with Rylander's observation (Rylander, 1969) that suggests
2
that SOg exposure (26.2 mg/m (10 ppm) for 6 h/day, 5 days/wk for 4 wk) does not affect the
bactericidal activity of the lung (see Table 12-4).
12.3 EFFECTS OF PARTICIPATE MATTER
Sulfur dioxide is oxidized to sulfuric acid (HLSO.) in the atmosphere. Sulfuric acid can
react with atmospheric ammonia (NH-) to produce ammonium sulfate and bisulfate. Similar re-
actions can also occur in the animal exposure chamber and confound experiments. Ambient PM
usually contains some proportion of sulfur compounds and the definition of the effects of
ambient aerosols independent of sulfur compounds may be impossible. Sulfur dioxide is often
present in polluted atmospheres with complex mixtures of other compounds including heavy
metals, which may be present as oxides or as sulfate or nitrate salts. In addition, organic
compounds present in the atmosphere in the gaseous phase can be associated with the partieu-
late fraction or become adsorbed on particles either i_n situ or during collection. Inhalation
of particles with surface coatings of toxic elements, organic compounds, allergens or gases
(LaBelle et al. , 1955) may result in greater effects due to localized surface reaction with
lung tissue or macrophages (Camner et al., 1974). The diversity of these types of particles
precludes discussion of their toxicity at this time, since little or no inhalation data are
available. The details of the composition of atmospheric aerosols are covered in Chapter 2
and the deposition and transport of particles are discussed in Chapter 11.
Since very few studies have appeared on the toxicity of complex atmospheric particles
themselves, this section primarily covers the toxicology of those compounds which have been
identified as constituents of atmospheric particles. Therefore, these discussions, no matter
how sophisticated for a single component, are inherently simplistic. For aerosols other than
HgSO., (NH.)pSQ and NH.HSO., this information is integrated in the perspective of the poten-
tial biological effects of atmospheric particles.
As will be apparent from the discussion of the toxicity of sulfate aerosols in this sec-
tion, the chemical composition of its constituent particles determines the toxicity of an
atmospheric aerosol. Those particles that are biologically active may have direct toxic
effects in themselves, indirect toxic effects through interactions with other pollutants, and
chronic effects through cell transformation or chronic alteration in cell function. Direct
12-22
-------
TABLE 12-4. EFFECTS OF S02 ON HOST DEFENSES

0.26,
Concentration
2.62, or 52.4 mg/m3
Duration
7 h/day, 5 day wk
Species
Rat
Results
Low concentrations (0 26 mg/m3

or 0.1 ppm)
Reference
Ferin and Leach, 1973
(0.1, 1, or 20 ppm) S02
2.62, 13.1, 26.2, and 52.4 mg/m3 24 h
(1, 5, 10, and 20 ppm) S02
2.62 mg/m3 (1 ppm) S02
1.5 h/day, 5 day/wk
2.62 or 7.86 mg/oa (1 or 3 ppm) Up to 119 days
S02 + graphite dust (mean
diameter 1.5 pm, 1 mg/m3)
9.43 to 52.4 mg/m3 (3.6 to 20
ppm) S02
7 days continuous
13.1 or 26.2 mg/m3 (5 or 10 ppm) 6 h/day, 20 day
S02
13.1 mg/m3 (5 ppm) S02
Varying from 0 to 52.4 mg/m3
(0 to 20 ppm) S02

26.2img/m3 (10 ppm) S02

65. S/ to 1868 mg/m3 (25 to 713
pp«) S02
786 mg/m3 (300 ppm) S02
3 h/day, 1-15 days
and 24 h/day, 1-3 mo

7 days, continuous
6 h/day for 20 days

30 rain
6 h/day, 10 days
continuous
Rat

Dog

Rat

Mouse

Guinea pig

Mouse

Rat

Donkey

Rat
accelerated alveolar clearance after 10
and 23 days, as did 2.62 mg/m3 (1 ppw) at 10
days; at 25 days, 1 ppm decreased clearance.

Exposure to the two higher concentrations increased
lQ vitro phagocytosis of latex spheres for up to
4 days in culture. At 13.1 mg/m3 (5 ppm), phago-
cytosis was increased after 3 or 4 days in culture,
but not 1 or 2 days.

Tracheal mucous flow was reduced.

No changes in the cilia beat frequency or the
relative number of alveolar aacrophages laden
with particles.

Exposure to S02 and a virus produced weight loss
Bacterial clearance was not altered
Katz and Laskin, 1976
Hirsch et al., 1975

Fraser et al., 1968
Lebowitz and
Fairchild, 1973

Rylander, 1969, 1970
Did not increase mortality subsequent to a pulmonary Ehrlich, 1979
streptococcal infection
Increase in viral pneumonia at 18.3 to 26.2 mg/m3
(7 to 10 ppm). Rate of growth of virus unaffected.
Fairchild et al., 1972
Did not affect the bactericidal activity of the lung Rylander, 1969

Below 786 mg/ms (300 ppm), mucociliary clearance was Spiegelman et al., 1968
not affected, but at high concentrations (786 to 1868
mg/m3 or 376 to 713 ppm), clearance was depressed.
No changes in selected lysosomal enzymes
Barry et al., 1970
1 ppm S02=2.62mg/m3.
-------
toxic effects are best substantiated by cytotoxicity studies. Those reviewed here are for
some specific compounds that occur in the particulate fraction. The studies cited are by no
means complete and could be expanded by including a number of other investigations carried out
JH vitro or by exposures other than inhalation. The review was purposefully restricted to
those most applicable to the inhalation route of exposure. Major exceptions to this policy
have been made for silica and the limited data on compounds in the so-called "coarse-mode" par-
ticles fraction. Most of the effects through interaction with other pollutants have pre-
viously been discussed for S0?. Some additional data implicating interactions between S0? and
PM or ozone, and*between H_SQ* and ozone are included.
Almost all of the studies (and all of the inhalation studies) discussed in this section
involve the health effects of particles in the "fine mode" size range and composition. Within
this category, several investigators examined the influence of particle size for a given
chemical. For coarse mode particles, only a few HI vitro and intratracheal instillation
studies could be found. This work is discussed separately.
12.3.1 Mortality
The susceptibility of laboratory animals to H«S04 aerosols varies considerably. Amdur
(1971) reviewed the toxicity of HLSCL aerosols and pointed out that, of the commonly used
experimental animals, guinea pigs are the most sensitive and most similar to man in their
bronehoconstrictive response to HySQ.. The lethal concentration (LC) of H^SO,, depends on the
2 t- *r 3
age of the animal (18 mg/m for 8 h for 1 to 2 mo-old vs. 50 mg/m for 18 mo-old animals),
the particle size (those near 2 urn being more toxic), and the temperature (extreme cold in-
creasing toxicity). Wolff et al. (1979) found the LC5Q (the concentration at which 50 percent
of the animals die) in guinea pigs for an 0.8 pm (MMAD) aerosol to be 30 mg/m , whereas for a
0.4 urn (MMAO) aerosol it was above 109 mg/m . In determining acute toxicity, the concentra-
tion of the aerosol appears to be more important than the length of exposure (Amdur et al.,
1952). The animals that died did so within 4 hours. Chronic studies have only recently been
undertaken, and they support this conclusion that mortality rarely occurs at moderate concen-
trations of HpSO^,.
Sulfuric acid aerosol appears to have two actions. Laryngeal and/or bronchial spasm are
the predominant causes of death at high concentrations. When lower concentrations are used,
bronchostenosis and laryngeal spasm can still occur. Pathological lesions in the latter case
include capillary engorgement and hemor-rhage. Such findings are in accord with anoxia as the
primary cause of death.
12.3.2 HgrphoIogi ca1 A1teratIons
Alarie et al. (1973a) investigated the effects of chronic H-SO, exposure. Guinea pigs
3 3
were exposed continuously for 52 wk to 0.1 mg/m H~S04 (2.78 urn, MMD) or to 0.08 mg/m H2S04
(0.84 urn, MMD). Monkeys were exposed continuously for 78 wk to 4.79 mg/m (0.73 pm, MMD),
2.43 mg/m3 (3.6 pm, MMD), 0.48 mg/m3 (0.54 urn, MMD), or 0.38 mg/m3 H2S04, (2.15 urn, MMD).
12-24
-------
Sulfuric acid had no significant hematological effects in either species. No light micro-
scopic lung alterations resulting from HpSCL exposure were observed in guinea pigs after 12 or
52 wk of exposure in this study (Alarie et al., 1973a) or in a later study (Alarie et al.,
1975). Morphological changes were evident in the lungs of monkeys. At the two highest con-
centrations, there were changes (more prevalent in the 4.79 mg/m hLSQ, group) regardless of
the particle size. Major findings included bronchiolar epithelial hyperplasia and thickening
of the walls of the respiratory bronchioles. Alveolar walls were thickened in monkeys exposed
3
to 2.43 but not to 4.79 mg/m H-SO*. Particle size, however, had an impact at lower H2S(h
concentrations. No significant alterations were seen after exposure to 0.48 mg/m of the
smaller particle size (0.54 urn). However, bronchiolar epithelial hyperplasia and thicken-
ing of the walls of the respiratory bronchioles were seen after exposure to the larger size
(1.15 |jm) and lower concentration (0.38 mg/m ). Pulmonary function changes followed a slight-
ly different pattern (see Section 12.3.4.2). Dogs also appear to be relatively insensitive to
H?SOd alone, as judged by morphological changes. Lewis et al. (1973) found no morphological
3
changes after the dogs had been exposed for 21 h/day for 620 days to 0.89 mg/m FLSO, aerosol
(90 percent of the particles were <0.5 |jm in diameter).
Cockrell and Busey (1978) and Ketels et al. (1977) studied the morphological changes re-
sulting from HpSO* aerosols. Cockrell and Busey (1978) examined the effects of 25 mg/m H2S04
(1 |jm, MMD, 0 1.6) for 6 h/day for 2 days in guinea pigs. Segmented alveolar hemorrhage,
type 1 pneumocyte hyperplasia, and proliferation of pulmonary macrophages were reported.
3
Ketels et al. (1977) examined the response of mice to 100 mg/m i-LSO^; these exposures pro-
duced injury to the top and middle of the trachea, but none to the lower trachea and distal
airways. In an investigation of the dose-response relationship for H9SOA, mice received
3 3
either 5 daily 3 h-exposures to 200 mg/m , 10 daily exposures to 100 mg/m , 20 daily exposures
3 3
to 50 mg/m , or any one of these doses combined with 5 mg/m carbon particles. The damage was
judged to be proportional to the concentration (C) of HpSO,, but not to the integrated dose (C
x T) or to the time of exposure (T). (All of the exposures had the same C x T and therefore
their equivalence might have been hypothesized.)
A number of other studies of the morphological effects of H?SO* when combined with other
pollutants have been conducted. (See Section 12.4.1.2. and Table 12-5)
Inhalation of silicon dioxide (SO,) results in silicosis, which is characterized by mor-
phological changes in the lungs. Because the extensive information on silicosis has been re-
viewed elsewhere (Ziskind et al., 1976; NIOSH, 1975; Reiser and Last, 1979; Singh, 1978), it
is not discussed in detail here. Due to the toxicity of Si02> a Threshold Limit Value
(American Conference of Governmental Hygenists, 1979) has been set. Because of the involve-
ment of alveolar macrophages in its toxicity and its presence in ambient particles, however,
some of the effects of SiOp are summarized briefly here. All the information given below for
silicon is derived from reviews by Ziskind et al. (1976) and NIOSH (1975).
12-25
-------
TABLE 12-5. EFFECTS OF H2S04 AEROSOLS ON LUNG MORPHOLOGY
Concentration
Duration
Species
Results
Reference
52 wk, continuous
0.08 mg/m3 H2S04 (0.84 pm,
HMD), or 0.1 mg/m3 H2S04
(2.78 pm, HMD)

0.38 mg/m3 HzS04 (1.15 pm, HMD), 78 wk, continuous
0,48 mg/m3 H2S04 (0.54 pm, HMO),
2.43 mg/m3 H2S04 (3.6 \im, HMD), or
4.79 mg/ui3 H2S04 (0.73 pm, HMD)
0.89 mg/m3 (90% <0.5 pn in
diameter)
21 h/day, 620 days
25 mg/m3 H2S04 (1 pm, HMD,
6 h/day, 2 days
50 mg/m3 H2S04, or
100 mg/m3 H2S04, or
200 mg/m3 H2S04, or any of
of these doses combined with
5 mg/m3 carbon particles
(at all three duration schedules)
3 h/day, 20 days; or
3 h/day, 10 days; or
3 h/day, 5 days
Guinea pig No significant hematological effect or microscopic
lung alterations
Monkey No significant hematologfcal effect. Morphological
changes in the lungs at the two highest concentra-
tions, regardless of the particle size.
Alveolar walls were thickened with 2,43 mg/m3,
but not 4.79 mg/m3. Hyperplasia and bronchiole
thickening only with larger size (1.15 pm, 0.38
•g/m3).

Dog No morphological changes
Guinea pig Segmented alveolar hemorrhage, type 1 pneumocyte
hyperplasia, and proliferation of pulmonary
macrophages

Mouse Damage proportional to concentration, not
duration
Alarie et al.,
1973a, 1975
Alarie et al., 1973a
Lewis et al., 1973
Cockrell and Busey,
1978
Ketels et al., 1977
-------
Silicon is ubiquitous in the earth's crust. Silicon dioxide (SiCL) is found in three
crystalline forms (quartz, Cristobalite, and tridynrite), whose toxicity is ranked tridymite
>cristobalite > quartz. While these uncombined forms of SiO, are generally called "free
silica," SiOo combined with cations is called silicate(s). Several hypotheses of the etiology
of silicosis have been developed, but no single one has been proven definitively. According
to one widely accepted hypothesis developed from both animal and human studies, alveolar
macrophages ingest the particles, die, and release their intracellular contents, including
lysosomal enzymes and SiOp. This is followed by a recycling of particle ingestion by
macrophages and their death, slow accumulation of other macrophage cells, increased collagen
synthesis in response to macrophage lysosomal enzymes, hyalinization, and perhaps complicating
factors. Since the alveolar macrophage hypothesis does not explain completely the etiology
and pathogenesis of the disease, it is likely that additional factors contribute to the
disease. These might include autoimmunity, coexisting tuberculosis or other infections,
and/or alterations of lung lipid content and metabolism.
Many animal toxicological studies of SiCL exist. Unfortunately, comparisons are diffi-
cult because of the species and strain of animal, accidental infections, and the size and crys-
talline form of SiO? particle used.
Silicosis similar to that observed in man has been produced in animals exposed to high
concentrations of quartz and other SiO, dusts via intratracheal instillation (30-50 mg) or
chronic inhalation. Chronic exposures (2.5 yr) of dogs to diatomaceous earth containing 61
percent Cristobalite produced fibrotic nodules in hilar lymph nodes, but not the lungs.
Several studies of the hemolysis of red blood cells by particles have been reported.
This model may correlate with the ability of mineral dusts to cause lung fibrosis j_n vivo and
thus is used for screening. Ottery and Gormley (1978) studied the influence of particle size
of quartz (Min-u-sil) and other materials on red cell hemolysis. The particle size of the
quartz ranged from 2.7 to 6.8 [im (mean volume diameter, MVD). At lower concentrations (0.025
to about 0.15 mg/ml), there was a linear dose-response increase in hemolysis for quartz (1.35
and 3.55 urn, MVD), kaolin (4.7 pm, MVD), Cristobal ite (3.05 pm, MVD), and bentonite (5 pm
MVD). In this experiment, the effectiveness for increasing hemolysis was ranked as bentonite
>kaolin > quartz > Cristobalite. Even though this test typically is used to predict fibrotic
potential, it should be noted that cristobalite usually is more fibrogenic than quartz, which
is much more fibrogenic than the silicates. In another experiment, kaolin and quartz were
additive when mixed. Linear increases in hemolysis were also observed with increasing numbers
of particles. When the various sizes of Min-u-sil were directly compared, as particle size
decreased from 6.8 to 2.7 pm, a smaller concentration was required to produce 5 percent
hemolysis. For example, the largest size tested required 2.7 mg/ml, whereas only 0.21 mg/ml
was needed for the smallest size tested.
12-27
-------
Kysela et al. (1973) administered high concentrations (50 ing) of 9 sizes of quartz dust
(0.7 to 35 pm) to rats by intratracheal instillation. A variety of biochemical determinations
as well as a histological examination of the lungs were made 3 mo after dosing. As particle
size decreased, there was a trend towards increased wet weight of the lung, hydroxyproline con-
tent, total lipids, esterified fatty acids, and phospholipids. Cholesterol showed only a
slight increase. For hydroxyproline in total lung, the increase was in steps, with increments
at about 0.9, 5, 7, and 10 pm. Between 0.7 and 14 pm, the increase was significant (p < 0.05).
Lipid changes, per gram of tissue, exhibited a trend towards linearity with decreasing par-
ticle size. .The larger particles (14 to 35 pm) caused a stationary granulamatous response.
With the intermediate particles (5 to 10 pm), the lungs had cellular nodules with a few colla-
genous fibers, increased tissue cellularity, and endoalveolar foam cells. With the smaller
particles, the nodules were more numerous and collagenous.
Goldstein and Webster (1966) also investigated the effects of size-graded quartz parti-
cles in rats exposed by intratracheal instillation and examined 4 mo later. The sizes and
amounts used (<1 pm, 13.99 mg; 1 to 3 pm, 46.1 mg; and 2 to 5 urn, 92.7 mg) were such that the
rats were exposed to an equivalent surface area (600 sq. cm.) for each of the size ranges.
The < 1 pm particles caused more numerous nodules. The two other sile ranges produced an
equivalent number of lungs with nodules, but there were many more lungs with confluent
nodules, compared to the smallest size quartz. The degree of fibrosis was similar in the 1 to
3 pm and 2 to 5 pm groups and was more severe than that observed in the < 1 urn group. The
weight of collagen in lungs increased as particle size increased. It should be recalled, how-
ever, that the concentration (weight per exposure) of particles was increased as particle size
increased.
In vitro studies (Mossman et al., 1978) with small carbon particles (0.5 to 1 pm) suggest
that particles deposited in the TB region that become attached to the mucosal surfaces of the
airways may subsequently enter the submucosa and be taken up by mesenchymal cells of the
trachea.
12.3.3 Alterations in Pulmonary Function
12.3.3.1 Acute Exposure Effects—On short-term exposure, respiratory mechanics are very sen-
sitive to inhaled H^SO. and some other compounds. Amdur (1971) has cautioned that her method
for measuring airway resistance (Amdur and Mead, 1955, 1958) should not be used as an indica-
tion of chronic toxicity and should be considered only for short-term toxicity. The Mead-
Amdur method uses unanesthetized guinea pigs in which a transpleural catheter has been im-
planted. Amdur (1971) suggests that, if anything, this procedure increases rather than de-
creases the sensitivity of the guinea pigs to inhaled irritants.
Using this method, Amdur and coworkers (Amdur and Underhill, 1968, 1970; Amdur, 1954;
1958, 1959, 1961; Amdur and Corn, 1963; Amdur et al., 1978a,b,c,) have studied the effects of
aerosols alone (see Table 12-6) or in combination with SO, (see Section 12.4.1.1). In all of
their studies, exposures were for 1 hour. The method records resistance to air flow in and

12-28
-------
TABLE 12-6. RESPIRATORY RESPONSE OF GUINEA PIGS EXPOSED FOR 1 HOUR TO PARTICLES

IN THE AMDUR et al. STUDIES
ro
i

Compound
V°4
C, ™

(NH4)2SO

NH.HSO.
n n

Concentration
mg/m3
0.10

0.51
1.00
1.90

5.30
15.40
26.1
42.00
0.11
0.40
0.69
0.85
2.30
8.90
15.40
43.60
30.50
0.50
2.14
1.02

9.54
0.93
2.60
10.98
Particle
size, |jm, HMD
0.3

0.3
0.3
0.8

0.8
0.8
0.8
0.8
1.0
1.0
1.0
1.0
2.5
2.5
2.5
2.5
7.0
0.13
0.20
0.30

0.81
0.13
0.52
0.77
Resistance
cm H20/ml/sec
% difference
from control
+41a

+60a
+78a
+51a

+54a
+69a
+89a
+120a
-H4a
+30a
+47a
+60a
+39a
+61a
+96a
+317a
+42a
+23a
-4a
+29a

0
+15a
+28a
+23a
Compliance
ml /cm H20
% difference
from control
-27a

-33a
-40a
-35a

-40a
-24a
-38a
-26a
-13
-8
-25a
-28a
-16
-26a
-43a
-76a
-17
-27a
-133
-23a

-12a
-15a
-30a
-19a
Reference
Amdur et al . ,
Amdur, 1977
Amdur et al . ,
Amdur et al . ,
Amdur, 1969;
1958
Amdur, 1958
Amdur, 1958
Amdur, 1958
Amdur, 1958
Amdur et al . ,
Amdur et al . ,
Amdur et al. ,
Amdur et al . ,
Amdur, 1958
Amdur, 1958
Amdur, 1958
Amdur, 1958
Amdur, 1958
Amdur et al . ,
Amdur et al . ,

1978b;

1978b
1978b
Amdur,

1978b
1978b
1978b
1978b

1978a
1978a
Amdur and Corn, 1963;
Amdur et al .
Amdur, 1974
Amdur et al. ,
Amdur et al . ,
Amdur et al . ,
Amdur et al . ,
, 1978a;

1978a
1978a
1978a
1978a
-------
TABLE 12-6. (continued).
i
to
o

Concentration
Compound mg/m3
Na2S04
ZnSO
T"
ZnS04- (NH4)?S04
T" T" C. T"

CuSO.
4

NaVO,
0.90
0.91

0.25

0.50
1.10

1.80

1.50

2.48
1.40

1.10

3.60
0.43
2.05
2.41
0.70
Particle .
size, pm, MMDD
0.11
1.4

0.29

0.29
0.29

0.29

0.51

0.51
0.74

1.4

1.4
0.11
0.13
0.33
-
Resistance
cm H20/ml/sec
% difference
from control
+2
Hla

+22a

H0a
+81a

+129a

H3a

+68a
+29a

+32a
+9
+25a
+14a
+7
Compliance
ml /cm HaO
% difference
from control Reference
-7 Amdur et al . , 1978a
Amdur and Corn, 1963;
Amdur, 1974
Amdur and Corn, 1963;
Amdur, 1977
Amdur and Corn, 1963
Amdur and Corn, 1963;
Amdur, 1974
Amdur, 1969; Amdur and
Corn, 1963
Amdur and Corn, 1963;
Amdur, 1977
Amdur and Corn, 1963
Amdur, 1969; Amdur and
Corn, 1963; Amdur, 1977
Amdur and Corn, 1963;
Amdur, 1977
Amdur and Corn, 1963
-lla Amdur et al., 1978a
-15a Amdur et al., 1978a
-11 Amdur et al. , 1978a
Amdur and Underbill,
FeSO,
1.00
+2
1968

Amdur and Underbill,

1968
-------
TABLE 12-6. (continued)
OJ

Compound
Fe203 (2hr)
(Fumes)
MnCl2
Mn02
MnS04
Open hearth
dust
Activated
carbon
Spectographic
carbon
Concentration
mg/m3
11.70
21.00
1.00
9.70
4.00
0.16
7.00
8.70

2.00
8.00
Resistance
cm H20/ml/sec
Particle % difference
size, pm, MMD from control
0.076 (GMD) -9
0.076 (GMD) 0
+4
-6
-1
0.037 (GMD) +9
0.037 (GMD) +6
-3

+7
+17
Compliance
ml/cm H20
% difference
from control Reference
5 Amdur
1968
0 Amdur
1968
Amdur
1968
Amdur
1968
Amdur
0 Amdur
1968
-16 Amdur
1968
Amdur
1968
Amdur
1968
Amdur
1968
and Underhill
; 1970
and Underbill
; 1970
and Underbill
and Underbill
, 1974
and Underhill
; 1970
and Underbill
; 1970
and Underhill

and Underhill
and Underbill
*
»
s
'

'
5

J
J

p < 0.05
Diameters are provided as mass median diameter (MMD) unless specified as geometric median diameter by count (GMD).
-------
out of the lungs and airways, compliance (a measure of lung distensibility), tidal volume (the
volume of air moved during normal breathing), respiratory frequency, and minute volume. While
increased flow resistance is often the most striking feature of the response to aerosols, cal-
culations of the elastic, resistive, and total work of breathing can also be made. The method
is, therefore, nearly as elaborate and inclusive an evaluation of pulmonary mechanics as could
be made in small laboratory animals until Drazen's (1976) work.
The importance of particle size on the site of pulmonary deposition is described in
Chapter 11. Sulfuric acid aerosols ranging in concentration from 1.9 to 43.6 mg/m were gene-
rated in three particle sizes: 0.8 jjm (a = 1.32), 2.5 urn (tr = 1.38), or 7 urn (a = 2.03)
y 3 g g
HMD. Particles 'of the largest size (7 urn, at 30 mg/m ) produced a significant increase in
flow resistance, but no other detectable changes in respiration. At the lowest concentration
tested (1.9 mg/m ), the 0.8 urn particles increased the resistance to flow and the elastic,
resistive, and total work of breathing; but they produced a decrease in compliance. The 2.5
3
jjm particles also increased the resistance to flow at concentrations from 2.3 to 43.6 mg/m .
The relative efficacy of the 0.8 and 2.5 urn particles differed. At concentrations of 2 mg/m ,
the 0.8 urn particles were more effective than the 2.5 pm particles. The time course of the
response also varied with the particle size, since the 2.5 |jm particles did not produce their
major effects until the last 15 to 20 min of the 1-h exposure. These differences in response
were probably associated with the degree and site of constriction within the bronchi. The 2.5
pm particles produced constriction in the larger bronchi, whereas the 0.8 urn particles caused
narrowing of the smaller bronchi. In earlier work, Rattle et al. (1956) reported that, at
higher concentrations, 2.7 urn aerosols are more toxic than 0.8 urn as measured by mortality and
increased airway resistance. While the results of the experiments by Amdur and coworkers are
reported in a straightforward concentration-response curve, the physiological mechanisms pro-
ducing the measurable effects are obviously highly complex. Detailed understanding is lack-
ing.
Amdur et al. (1978b) exposed guinea pigs for 1 h to either 0.3 or 1 pro (HMD) H9SO.. in con-
3
centrations ranging from 0.1 to 1 mg/m . The concentration-response ratio for percent change
in resistance was linear for both particle sizes; however, the smaller particle caused a
3
greater response, particularly at 0.1 mg/m , where a 26 percent increase in airway resistance
was observed. All increases in resistance were statistically significant. The smaller par-
ticle size also decreased compliance at all concentrations tested; however, the lowest effec-
3
tive concentration tested for the I ym particle was 0.69 mg/m . For equivalent concentrations,
the 0.3 pm particle decreased compliance more than the 1 .urn particle. Animals were also
3
examined 30 min after exposure. After exposure to 0.1 mg/m I-LS04 (0.3 |jm), resistance was
still elevated above control in guinea pigs; but for the 1 \im particle, recovery had occurred.
These exposures caused no alterations of tidal volume, respiratory frequency, or minute
volume. In comparing these results to earlieir work with S02 (Amdur, 1966), Amdur et al. (1978b)
12-32
-------
describe how the same amount of sulfur, when given as I-LSCL, produces 6 to 8 times the re-
sponse observed than when given as SCL.
Silbaugh et al. (1981) exposed Hartley guinea pigs for 1 h to 1 urn (MMAD) H,SO,, aerosols
3
at concentrations and relative humidities (RH) of 0 mg/m (control group) (40 or '80 percent
RH), 1.2 mg/m3 (40 percent RH), 1.3 mg/m3 (80 percent RH), 14.6 mg/m3 (80 percent RH), 24.3
3 3
mg/m (80 percent RH), or 48.3 mg/m (80 percent RH). Ten animals were exposed at each con-
o
centration except for the 24.3 and 48.3 mg/m groups, which consisted of 9 and 8 animals, re-
spectively. Measurements of tidal volume, breathing frequency, minute volume, peak inspi-
ratory and expiratory flow, tidal transpulmonary pressure excursions, total pulmonary resist-
ance, and dynamic lung compliance were obtained every 15 min during (1) a 30-min baseline
period, (2) the 60-min exposure period, and (3) a 30-min recovery period. Pulmonary function
changes in H^SO.-exposed guinea pigs did not differ from controls, except for 1 animal exposed
3 3
to 14.6 mg/m3, 3 animals exposed to 24.3 mg/m , and 4 animals exposed to 48.3 mg/m . Pulmo-
nary function changes in these 8 responsive animals included marked increases in total pulmo-
nary resistance and marked decreases in dynamic compliance. Two animals in each of the latter '
groups died during exposure. The proportion of responsive to nonresponsive animals increased
with exposure concentration, but the magnitude of pulmonary function change was similar for
all responsive animals. Compared to nonresponders, responsive animals tended to have higher
preexposure values of total pulmonary resistance and lower preexposure values of dynamic com-
pliance. The authors suggested that guinea pigs react to acute H?SCL'exposure with an essenti-
ally all-or-none airway constrictive response. The finding that resistance and compliance
changes are important components of the guinea pig's airway response to H~SO. aerosols is con-
sistent with Amdur et al. (1978b). The presence of high preexposure pulmonary resistance
values in responsive animals is similar to the finding (Amdur, 1964) that guinea pigs with high
preexposure resistance values were those most severely affected during irritant aerosol expo-
sure. The lack of effects at lower concentrations and the essentially all-or-none airway con-
strictive response observed in these studies, however, differs markedly from the graded re-
sponse observed by Amdur et al. (1958, 1978b) during similar exposures. The reasons for the
differences in experimental results are unclear, but may be at least partially related to
differences in animal strains and techniques. These results indicate that changes in respira-
tory function do not occur at environmental concentrations of HpSO. in most animals, but
suggest that susceptible subpopulations might possibly exist in some species. (See discussion
above about susceptible individuals.)
It should be noted, however, that Sackner et al. (1978a,b) evaluated pulmonary function
in anesthetized dogs either immediately or 2 h after exposure to approximately 18 mg/m H?SO.
o
for 7.5 min or to 4 mg/m H2$04 for 4 hours. The MMAD was < 0.2 urn. There were no signifi-
cant changes in respiratory resistance, specific respiratory conductance, specific lung com-
pliance, or functional residual capacity. At the higher concentration, cardiovascular
parameters (e.g., blood pressure, cardiac output, heart rate, and stroke volume) and arterial

12-33
-------
blood gas tensions were also studied, but no significant changes were observed. In addition,
the pulmonary function (pulmonary resistance and dynamic compliance) of donkeys was not
affected by H-SO, exposure (1.51 mg/rn3, 0.3 to 0.6 HMAD, 1 h) (Schlesinger et al., 1978,
1979). The small size of the particles may be responsible for the lack of an effect. Larger
particles may be more potent, since resistance and compliance measurements reflect alterations
of the larger conducting airways in the TB region of deposition.
Studies of the irritant potential of sulfate salts have shown that these aerosols are not
innocuous, but evoke increased flow resistance similar to H?S04 aerosols. The influence of
particle size on the effects of zinc ammonium sulfate has also been investigated by Amdur and
Corn (1963). They showed in guinea pigs exposed for 1 h that zinc sulfate had about half the
potency of zinc ammonium sulfate, with ammonium sulfate being one-third to one-fourth as
potent as zinc ammonium sulfate. Zinc ammonium sulfate was chosen for study because it was
reported as a major component of the aerosol from the severe air pollution episode in Donora,
Pa, in 1948 (Hemeon, 1955). Zinc ammonium sulfate is not a species commonly found in urban
air. Four sizes of aerosols were administered: 0.29, 0.51, 0.74, and 1.4 urn (particle mean
3
size by weight). When the aerosol concentration was held constant at 1 mg/m , the smaller
particles produced greater increased resistance to flow. This response was thought to be the
result of the number of particles rather than of differential sites of deposition. The dose-
response curve also became steeper with decreasing particle size. These data should be
compared carefully with those from similar human exposures (Chapter 13, Sect. 13.5,2).
Amdur et al. (1978a) compared the effects of (NH4)2S04, NH4HS04, CuS04, and Na2S04.
Although particle sizes and concentrations were not matched precisely throughout the study,
statistical analyses for ranking were not applied, and the degree of response increased with
decreased size (size range, 0.1 to 0.8 urn, HMD), the authors suggest that the order of
irritant potency was (NH4)2S04 > NH4HS04 > CuS04- Sodium sulfate (0.11 mg/m3, 0.11 MMD)
caused no significant effects on either resistance or compliance. At the lowest concentra-
tions used, (NH4)2S04 (0.5 mg/m3, 0.13 (jm MMD), NH4HS04 (0.93 mg/m3, 0.13 [im HMD), and CuS04
(0.43 mg/m , 0.11 ym MMD) decreased compliance. These concentrations of (NH4)pS04 and NH4HS04
also increased resistance. For CuSO,,, the lowest concentration tested that caused an increase
0 '
in resistance was 2.05 mg/m (0.13 |jm MMD). All of these compounds are less potent than H2$04
in the Amdur studies.
Comparisons between H2S04 and sulfate salt aerosols are difficult because of the marked
dependence of the efficacy on the aerosol size. If the particles are of identical size, FLSO.
is more efficacious than zinc ammonium sulfate; but if the zinc ammonium sulfate were present
as a submicron aerosol and the H«S04 as a large aerosol, then zinc ammonium sulfate would be
more efficacious at the same concentration (Amdur, 1971). Regardless of the particle size,
the equivalent amount of sulfur present as S00 is much less efficacious than if it were
3
present as a sulfate salt containing ammonium or H-SO.. When present as S02, 2.62 mg/m (1
12-34
-------
ppm) SO,, is equivalent to 1.3 mg/m sulfur and produces a 15 percent increase in flow resist-
ance. If this amount of sulfur were present as a 0.7 pm aerosol of H?S(K, it would result in
a 60 percent increase in flow resistance or be about 4 times more efficacious. If the sulfur
were present as zinc ammonium sulfate in a 0.3 urn aerosol, the increase in flow resistance
would be about 300 percent (20-fold) in efficacy. Some sulfate salt aerosols are not irritat-
ing. For example, though ferrous sulfate and manganous sulfate do not cause an increase in
flow resistance, ferric sulfate does cause this response. A summary of irritant potency is
presented below.

Relative Irritant Potency of Sulfates In Guinea Pigs
Exposed for 1 Hour (Amdur et al., 1978a)

Sulfuric acid 100
Zinc ammonium sulfate 33
Ferric sulfate 26
Zinc sulfate 19
Ammonium sulfate 10
Ammonium bisulfate 3
Cupric sulfate 2
Ferrous sulfate 0.7
Sodium sulfate (at 0.1 urn) 0.7
Manganous sulfate -0.9
Data are for 0.3 Mm (MMD) particles. Increases in airway resistance were
related to sulfuric acid (0.41 percent increase in resistance per [ig of
sulfate per m3 as sulfuric acid) that was assigned a value of 100.

Nadel et al. (1967) found that zinc ammonium sulfate (no concentration given) and hista-
mine aerosols produced similar increases in resistance to flow and decreases in pulmonary com-
pliance in cats. Histamine was more potent than zinc ammonium sulfate. The increase in flow
resistance could not be blocked by intravenous administration of atropine sulfate, but was
blocked by either intravenous or inhaled isoproterenol. The study suggested that the in-
creased flow resistance was due to an increase in bronchial smooth muscle tone. Histamine
appears to be a likely mediator of the bronchoconstriction following inhalation of sulfate
salt aerosols. Charles and Menzel (1975a) investigated the release of histamine from guinea
pig lung fragments incubated with varying concentrations of sulfate salts. Almost complete
release of tissue histamine occurred with 100 mM ammonium sulfate. Intratracheal injection of
ammonium sulfate also released all of the histamine from perfused and ventilated rat lungs
(Charles et al., 1977a). The potency ranking of different sulfate salts in the release of
histamine from lung fragments (Charles and Menzel, 1975a; Charles et al. , 1977a) was equi-
valent to that causing increased resistance to flow (Amdur et al. , 1978a). Bronchoconstric-
tion of the perfused lung occurred on intratracheal injection of sulfate salts or histamine
(Charles et al,, 1977a). About 80 percent of the constriction could be blocked by prior treat-
ment of the isolated lungs with an H-l antihistamine. These experiments, as well as the

12-35
-------
original observations of Nadel et al. (1967) and Amdur et al. (1978a), support the concept
that an intermediary release of histamine or some other vasoactive hormone is involved in the
irritant response of sulfate aerosols. An ammonium sulfate particle is calculated to reach a
concentration of about 275 mM at equilibration with the 99.5 percent RH of the respiratory
tract (National Research Council, 1978). Thus, the concentration of the hydrated particle on
striking the mucosa would be within the range found to cause release of histamine in guinea
pig and rat lung fragments (Charles and Menzel, 1975a; Charles et al. , 1977a). A published
estimate of the.dose of inhaled ammonium sulfate needed to release histamine in the lung is in
error (National Research Council, 1978). Complete release of histamine (100 percent) occurred
with 1 umole of ammonium sulfate/lung, and not a 1 jjM solution for the entire lung (Charles et
al., 1977a). Further, total release of all histamine stores of a tissue rarely, if ever,
occurs under physiological conditions. Only about 10 percent of the total histamine is
released during degranulation reactions j_n vivo, producing anaphylactic shock and death
Therefore, even if the calculations were correct, only a small fraction of the ammonium sul-
fate dose would be required to produce the far less violent increases in flow resistance re-
ported by Amdur et al. (1978a) for ammonium bisulfate and ammonium sulfate. Assuming the cal-
culation of ammonium sulfate to be correct, a 4-h, not a 2-day, inhalation would produce a
marked increase in resistance to flow. Additionally, Charles et al. (1977b) found the rate of
35 -2
removal of SO. from the rat lung both in vivo and i_n vitro to be a function of the cation
associated with the salt and to follow the same order of potency as reported by Amdur and co-
workers (1978a) in the guinea pig irritancy test. Especially noteworthy is the fact that
manganous sulfate was removed at essentially the same rate as sodium sulfate, both of which
did not produce increased resistance to flow in the guinea pig.
Hackney (1978) has presented a preliminary summary of the effects of aerosols of HUSO,
and nitrate and sulfate salts on squirrel monkeys (Saimiri sciurens). Monkeys were exposed
3
(head only) to aerosols (2.5 mg/m ) of the respective salts or HUSO, (40 or 85 percent RH at
25°C). The exposure system was designed to reduce stress on the unanesthetized monkey. A
noninvasive method of pulmonary function measurement was used in which total respiratory
resistance was measured by the forced pressure oscillation technique at sine wave frequencies
of either 10 or 20 Hz. The measurement of pulmonary resistance included the resistance of the
chest wall, which was assumed to be irrelevant to pollutant response and to be constant
throughout the experiments. To correct for stress, control values were taken as those for a
given monkey exposed on the previous day to an aerosol of distilled water (for aerosol experi-
ments).
Hackney (1978) reports that the measurement of respiratory resistance was frequency-
dependent, with changes in resistance appearing greater in the 10 than the 20 Hz measurement
frequency. (The measurement frequency is not to be confused with the breathing frequency.)
The exposure period in the experiments was either 1 or 2 hours. Some aerosols were studied
12-36
-------
only at 40 percent RH. No attempt was made to define a dose-response curve for aerosols, and
all exposures were at or near 2.5 mg/m . At low RH (40 percent RH; MMAD 0.3 urn, o = 2.0),
there were no differences between (NH,),,$CL-exposed and control values, while at high relative
humidity (85 percent RH, MMAD 0.6 pm, a = 2.3), 3 of 5 monkeys had increased airway resist-
ance by 1 hour. Zinc ammonium sulfate aerosols produced increased resistance at low humidity
(40 percent RH; MMAD 0.3 urn, a = 2,5) but no consistent increases over control values at high
humidity (85 percent RH; MMAD 0.6 jjm, a = 1.6). Ammonium bisulfate (40 percent RH; MMAD 0.4
9 3
(jm, cr = 1.8) also produced increased resistance at 2.7 mg/m ,
Hackney's (1978) data from exposures to H?SO, and ammonium nitrate aerosols were analyzed
by computer and differed quantitatively from the data reported above for those exposures that
were reduced by hand. Differences were probably due to a systematic error in the hand-reduced
data that required a judgement in selection of raw data points. The biological interpretation
does not appear to be altered by these two approaches, but it does point out the experimental
difficulties in interpretation of pulmonary function data from experimental animals. While
HpSO. aerosols (40 percent RH; MMAD 0.4 pm, a =2.0) caused no statistically significant in-
creases, there was a trend toward increased resistance after 60 min that then tended to de-
cline. Ammonium nitrate exposures produced no changes.
Multiple contrast statistical analysis of the Hackney (1978) data revealed no significant
differences between baseline or control values for any exposure using data collected at the 20
Hz measurement frequency. At the 10 Hz measurement frequency, the data were more variable,
but significant differences indicative of increased airway resistance could be found for
animals exposed to ZnSG*, (NH-^pSO^ and H«S04 at 40 percent RH. Several procedural aspects
should be recognized. First, data were analyzed on a group mean basis, even though large
differences between individual monkeys existed in both variability and absolute magnitude.
Second, the time course of exposure to the aerosols illustrated a trend indicative of a transi-
ent response by the monkeys to sulfate, nitrate, or H2$04 aerosols. The use of group means
ended to reduce the magnitude of the response and flatten the response-time curve. This is
certainly true for the S0? exposures. Third, there were major differences in the response mea-
sured at either 10 or 20 Hz. Fourth, the response estimated by both manual and computer reduc-
tion differed by as much as 40 percent. Compared to the data reported for guinea pigs, however,
these experiments support the general trends originally proposed from the guinea pig data.
Larson and coworkers (1977, 1982) have proposed that breath ammonia is important in
neutralizing inhaled H^SO^. Ammonia is released in the breath from blood ammonia and bacter-
ial decay products in the buccal cavity. Ammonia in the breath could react with H^SO, to
produce ammonium bisulfate or ammonium sulfate, depending on the amount of ammonia and HgSO^
present in the aerosol droplet. Complete neutralization of H2S04 produces ammonium sulfate
((NH4)pS04). This theory has been discussed at some length (National Research Council, 1978).
Much of the data has not yet been published, so a critical review of the model given for the
12-37
-------
neutralization of FLSO^ aerosol droplets by gaseous ammonia is not available. However, it
does appear that neutralization is inversely proportional to particle size (Larson et al.,
1982).
The biological effects of HUSO, aerosols could be due to a combination of factors. First,
the pH of the particle could be very important. Larson et al. (1977, 1982) have calculated
the neutralization capacity of the breath ammonia. Once the neutralization capacity of the
ammonia present in the breath is exceeded., the pH of the aerosol reaching the lung may fall
rapidly. At low pH, physical properties of the mucous layer lining the upper airways may be
altered (Holma et al., 1977) or the permeability of the lung may be increased (Charles, 1976).
Second, the chemical composition of the sulfate aerosol, if other than H?SO», may also alter
the permeability of the lung to sulfate (Charles et al. , 1977b; Charles, 1976; Charles and Men-
zel, 1975b). Third, the cation associated with the sulfate compound may have pharmacological
properties in and of itself. The permeability of the lung to sulfate ion presented as various
sulfate salts (Charles et al., 1977b) is in the same relative order as the irritant potential
found for aerosols of the same sulfate salts (Amdur et al., 1978a).
It is likely that ammonia functions within pulmonary tissue as a source of protons that
increases the flux of sulfate to the site of action. Ammonia can diffuse readily across cell
membranes to react with protons, forming ammonium ion. Intracellular transport of negatively
charged sulfate would result in the concomitant accumulation of protons to preserve electro-
chemical neutrality. At physiological pH values, a significant fraction of ammonium salts is
present as ammonia. Ammonium salts could augment the local ammonia concentration and thus in-
crease the uptake of sulfate ions and result in release of histamine. Ammonia increased up-
take of sulfate by the lung (Charles et al., 1977a; Charles, 1976), possibly by this mechanism.
In relation td HUSO,, ammonium sulfate and bisulfate are less irritating to the lung be-
cause of their higher pH values once dissolved in the milieu of the lung. Thus, neutraliza-
tion of HpSO- aerosols by breath ammonia could be an important detoxification step. The con-
cept of neutralization by breath ammonia does not negate the histamine release hypothesis,
since ammonium sulfate is active in the release of histamine in guinea pig lung fragments
(Charles, 1976) and in rat lungs (Charles and Menzel, 1975b).
An important problem is the relationship of these observations to human health effects.
Unfortunately, histamine release by nonimmune-mediated reactions, such as the apparent ion ex-
change process due to sulfate interaction with mast cell granules (Charles, 1976), is poorly
understood. Metabolism of histamine by man and rodents could have important different impli-
cations. Also, not all of the pharmacological action of ammonium sulfate instilled intra-
tracheally in the perfused rat lung could be blocked by an H-l antihistamine (Charles et al.,
1977a). A number of other inflammatory hormones, aside from histamine, mediate bronchial tone
in man. Slow reacting substances of anaphylaxis (SRS-A or leukotrienes), prostaglandins, and
kinins would not be blocked by an H-l antihistamine. Thus, species differences are not unan-
ticipated, but should be clarified so that the potential applicability of these data to man is
understood.
12-38
-------
The biological effect of s ill fate compounds is highly dependent on the chemical composi-
tion of the compound. For example, H?SO, has much more potent effects on pulmonary function
than any sulfate salt, but the sulfate salts also have differing potency. The cations asso-
ciated with the sulfate. ion may promote its transport, thereby increasing the biological re-
sponse. The cation has biological effects by itself; as disc'ussed here. It is not possible,
then, to predict the potential toxicity of a sulfate aerosol based solely on the sulfate con-
tent. Clearly, the acidity of the aerosol .plays an important role in the toxicity, as do par-
ticle size and other physical properties.
An important experimental problem is raised by the ammonia neutralization "of H^SO,.
Ammonia is produced in all animal experimental exposure systems through the accumulation of
urine and feces (Malanchuk et al., 1980), This is particularly so in whole-body chronic ex-
posures. Few exposure systems provide a rapid turnover of the chamber air (e.g., 1 chamber
volume/min), and given the technological problems in monitoring ammonia, even this rate of air-
flow may be insufficient. The usual turnover rate is 10 to 15 chamber volumes of air/h or
less. Under these conditions, animals exposed to H«S(L aerosols may, in fact, be inhaling
ammonium sulfate and ammonium bisulfate aerosols as well. The high concentrations of I-LSCL
aerosols needed to produce significant pathological effects during chronic exposure may be due
to these chemical conversions. The level of ammonia in the breath of animals is also unknown
and is sure to vary with the diet of the animals. Some commercial animal diets are low in pro-
tein, while others are high. The blood ammonia will depend, in part, on the total amount of
protein and quality of the protein, as well as on the kidney function of the animal. What
effects, if any, the buccal flora have on the exhalation of ammonia in animals is totally un-
known. Certainly, the propensity of SO,- and HLSO.-exposed animals to develop nasal infec-
tions raises disturbing questions. The ability of the buccal flora of animals to produce
ammonia may be very different from humans. This technical problem of ammonia in the exposure
atmosphere should be addressed and solved in both human and animal exposures before further
reliance can be placed on these data for H^SQ, (Table 12-7).
12.3.3.2 Chronic Exposure Effects—The influence of chronic exposure to HUSO, on pulmonary
function was investigated by Alarie et al. (1973a, 1975). Guinea pigs exposed continuously to
either 0.9 mg/m3 (0.49 |jm, MMD) (Alarie et al., 1975), 0.1 mg/m3 (2.78 urn, HMD) (Alarie et
al., 1973a), or 0.08 mg/m3 (0.84 urn, MMD) for 52 wk (Alarie et al., 1973a) had no significant
changes of pulmonary mechanics (including measurements of flow resistance, respiratory rate,
lung volumes, and work of breathing) that could be attributed to H^SQ^. Cynomolgus monkeys
exposed continuously and tested periodically during 78 wk, however, were affected by some
3
treatment regimens (Alarie et al., 1973a). Monkeys exposed to 0.48 mg/m (0.54, \jtrn, MMD)
experienced an altered distribution of ventilation (increased nitrogen washout) early in the
exposure period, but recovery occurred during exposure. Animals exposed to a similar concen-
tration (0.38 mg/m ) but a larger particle size (2.15 urn, MMD) had no change in this para-
meter. Higher concentrations altered distribution of ventilation, with the lesser concentration

,12-39
-------
TABLE 12-7. EFFECTS OF ACUTE EXPOSURE TO SULFATE AEROSOLS ON PULMONARY FUNCTION
Concentration
Duration
Species
Results
Reference
0 mg/m3 (40 or 80% RH), 1.2 mg/m3
(40% RH), 1.3 mg/m3 (SDK RH),
14.S mg/m3 (80% RH), 24.3 mg/m3
(80% RH), and 48.3 mg/m3 (80% RH)
1 urn (MMAD) H2S04

0.8 - 1.51 mg/m3 H2S04
(0.3 - 0,6 urn, MMAD) or
0.4-2.1 mg/m3 (NH4)ZS04
(0,3 - 0.6 (jm, MMAD)

2.5 mg/«3 (NH4)2S04,
ZnS04,(NH4)2S04, H2S04,
and NH4N03, 2.7 mg/m3
NH4HS04
1 h
1 h
1 h
Guinea pig Pulmonary function changes observed in one animal
(out of 10) exposed to 14.6 mg/ma, three animals
(out of 9) exposed to 24.3 mg/m3, and four animals
(out of 8) exposed to 48.3 mg/m3
Donkey No significant alterations in pulmonary resistance
and dynamic compliance
Monkey Increased airway resistance at high relative hu-
midity for (NH4)2S04, and low relative humidity for
ZnS04, (NH4)2S04. NH4HS04 also increased resistance.
No significant effects with H2S04 or NH4N03
Silbaugh et al., 1981
Schlesinger et al.,
1978
Hackney, 1978
See Table 12-6 for the Andur et al. studies on pulmonai^ function effects in guinea pigs*.
-------
3
(2.43 mg/m ) and larger particle size (3.6 pm, HMD) causing an onset sooner (at 17 compared to
49 wk) than in monkeys exposed to 4.79 mg/m hLSO. (0.73 urn, HMD). Beginning approximately at
8 to 12 wk of exposure, 0.38 mg/m3 (2.15 urn, MMD), 2.43 mg/m3 (3.6 \im, MMD) arid 4.79 mg/m3
(0,73 (JIB, MMD) H?SO» increased respiratory rate. The only alteration in arterial partial
3
pressure of 0, was a decrease observed in monkeys exposed to 2.43 mg/m . Except for respira-
tory rate as described above, mechanical properties (including resistance, compliance, tidal
volume, minute volume, and work of breathing) were not altered significantly by the chronic
HUSO^ exposures. Morphological studies of these animals are described in Section 12.3.2.
Lewis et al. (1969, 1973) performed chronic studies of dogs. The animals were exposed
for 21 h/day for 225 or 620 days to 0.89 mg/m3 H2S04 (90 percent <0.5 urn in diameter) alone
and in combination with S0~ (see Section 12.4.1.2 for expanded discussion). After 225 days
(Lewis et al., 1969), the dogs receiving H?SO, had a significantly lower diffusing capacity
for CO than animals that did not receive HLSO.. After 620 days of exposure, CO diffusing
capacity was still decreased (p < 0.05) (Lewis et al., 1973). In addition, residual volume
and net lung volume (inflated) were decreased (p < 0.05), and total expiratory resistance was
increased (p < 0.05). Total lung capacity, inspiratory capacity, and functional residual
capacity were also decreased (p = 0.1). Other pulmonary function measurements were not sig-
nificantly affected (see Table 12-8).
12.3.4 Alteration in Host Defenses
To protect itself against inhaled microorganisms and inanimate particles, the host has
several defense mechanisms. Microbes reaching the gaseous exchange regions of the lung can be
phagocytized and killed by alveolar macrophages. Later these macrophages can move to the
ciliated airways where they are cleared from the lung, along with other particles that are de-
posited on the airways, by the mucociliary escalator. Inanimate particles can also be en-
gulfed and removed from the lung by this means (see Chapter 11). Mucociliary clearance is an
important defense against both microorganisms and inanimate particles. It is likely then that
an impairment of mucociliary clearance might be expressed as increased infections.
12.3.4.1 Mucociliary Clearance—Fairchild et al. (1975b) investigated the influence of a 1-
hour exposure to H?$0* on deposition of inhaled nonviable bacteria (Streptococcus pyogenes,
2.6 pm MMAD) in guinea pigs. All exposure regimens used caused no significant alterations of
3
breathing frequency, tidal volume, or minute ventilation After exposure to 3.02 mg/m (1.8 (jm
CMD), a 60-percent increase (p < 0.01) in total pulmonary bacterial deposition and a proximal
shift in the total deposition pattern to the nasopharynx were observed. No alteration in depo-
sition was observed in the trachea or lung. After exposure to 0.32 mg/m (0.6 (jm CMD), no
significant effect on total or regional deposition was seen. However, at a lower concentra-
tion and particle size (0.03 mg/m , 0.25 |jm CMD), the deposition pattern did shift (p < .05)
to the trachea, but without a significant change in total pulmonary deposition.
12-41
-------
TABLE 12-8.
EFFECTS OF .CHRONIC EXPOSURE TO H2S04 AEROSOLS ON PULMONARY FUNCTION
Concentration
0.08 mg/m3 H2S04 (0.84 pm,
Duration
52 wk, continuous
Species
Guinea pig
Results
No effects on pulmonary function
Reference
Alarie et al., 1975,
, . *
(2.78 (ini, HMD)

0.38 mg/m3 H2S04 (1.15 pm, HMD)
0.48 mg/M3 H2S04 (0.54 pm, HMD)
2.43 ng/m3 H2S04 (3.6 pm, HMD)
4.79 mg/m3 H2S0< (0.73 pm, HMD)

0.89 rag/m3 H2S04 (90%

-------
After studying effects of H^SO* on deposition of bacteria, these investigators (Fairchild
et al., 1975a) turned their attention to effects on clearance of bacteria. They showed that
4-h exposures to 15 mg/m H,,SO, (3.2 urn, CMD) after exposure to a nonviable radiolabeled strep-
tococcal aerosol reduced the rate of ciliary clearance of the bacteria from the lungs and
noses of mice. When mice received a 90-min exposure-tOv15 mtj/m H^SCL (3.2 urn, CMD) 4 days
prior to the bacterial aerosol, clearance of nonviable bacteria was reduced in the nose, but
not in the lungs. Neither regimen affected clearance of viable streptococci. No significant
effects were seen at concentrations of 1,5 mg/m H?S04 (0.6 urn, CMD).
Schlesinger et al. (1978) demonstrated that 1-h exposures to 0.3 to 0.6 [im HLSQ,, mist at
3
concentrations in the range of 0.19 to 1.36 mg/m produced transient slowing of bronchial
mucociliary particle clearance in 3 of 4 donkeys tested (only one of these demonstrated this
response at 0.19 mg/m ). In addition, 2 of the 4 donkeys developed persistently slowed clear-
ance after about 6 exposures. These exposures had no effects on regional particle deposition
or respiratory mechanics, and corresponding exposures to (NH4)?S04 up to 2 mg/m had no measur-
able effects. In subsequent experiments (Schlesinger et al., 1979), the 2 animals showing
only transient responses and 2 previously unexposed animals were given daily 1-h exposures, 5
days/wk, to H_SO, at 0.1 mg/m . Within the first few wk of exposure, all 4 donkeys developed
erratic clearance rates (i.e., rates that, on specific test days, were either substantially
slower than or faster than those in their preexposure period). The degree and the direction
of change in rate, however, differed to some extent in the different animals. These changes
may herald subsequent alterations and might possibly represent important low level signals at
the detection limit of the method. The 2 previously unexposed animals developed persistently
slowed bronchial clearance sometime during the second 3 mo of exposure and during 4 mo of
follow-up clearance measurements, while the 2 previously exposed animals had clearance times
that consistently fell within the normal range after the first few weeks of exposure. The
alterations of clearance observed in 2 initially healthy and previously unexposed animals may
be a significant observation, since alteration of normal mucociliary clearance over a period
might have important implications. Lippmann et al. (1980, 1981) and Leikauf et al. (1981)
have conducted similar experiments in human subjects (see Chapter 13).
Trachea! mucociliary transport rates have been measured in several other animal studies.
Sackner et al. (1978a) failed to find significant changes in tracheal mucus velocity following
short-term exposures to 14 mg/m (0.12 \im) H?SO. in sheep. Similarly, Schlesinger et al.
(1978) saw no effect on tracheal transport in donkeys after 1-h exposures to concentrations up
to 1.4 mg/m3 (0.3 to 0.6 ym MMAD) H2S04. On the other hand, Wolff et al. (1981a) reported a
depression in tracheal transport rate in anesthesized dogs exposed for 1 h to 1.0 mg/m (0.9
urn, MMAD, a =1.4) that persisted at 1 wk postexposure. Recovery had occurred when the
9 3
animals were examined again at 5 wk postexposure. Following a 1-h exposure to 0.5 mg/m
12-43
-------
H^SO,, there were slight increases (p >0.05) in trachea! mucous velocities both immediately
and 1 day after exposure. One wk after exposure, however, clearance was significantly
decreased. The latter results are similar to those observed in the bronchi of individual
humans in the Lippmann et al. (1980) study (see Chapter 13), although they recorded no
significant change in the mean trachea! mucociliary transport rates.
Clearly, the results of the donkey studies support the human experiments (Chapter 13)
that indicate that sub-micrometer H-SO* aerosols affect mucociliary clearance in the distal
conductive airways. Mucociliary clearance depends on both the physicochemical properties of
the mucus and the coordinated beat of the underlying cilia. Mucus is excreted into the airway
lumen in an alkaline form that is then acidified by CO, (Holma et al. , 1977). In vitro
<£,
studies have shown that mucus is a solution in high pH solutions, while at lower pH it becomes
_i_
viscous (Breuninger, 1964). The H supplied by the, H?S04 may stiffen the mucus and increase
the efficiency of removal. This is consistent with the increase in bronchial clearance rate
3
observed in humans following exposure to 0.1 mg/m H^SO.. Major changes in mucous viscosity
could also impair clearance by making the mucus so stiff that ciliary movement is not
possible. Other studies (Grose et al., 1980; Schiff et al., 1979) have shown that exposures
to 0.9 to 1.1 mg/m H-SO. can cause a depression of trachea! ciliary beat frequency in
hamsters that may lead to a depression in overall bronchial clearance. See Sections 12.4.1.1
and 12.4.2 for more details on these latter studies (Grose et al., 1980; Schiff et al., 1979)
that were conducted with pollutant mixtures.
Based on the results summarized above, it is possible that chronic H9SO» exposures at
3
concentrations of about 0.1 mg/m could produce persistent changes in mucociliary clearance
(see Table 12-9).
Cadmium and nickel chlorides also disrupt the activity of the ciliated epithelium (Adalis
et al., 1977, 1978). Trachea! rings have been isolated from hamsters and the beat frequency
and morphology of the ciliated epithelium have been observed. Concentrations of CdCl- as low
as 6 jjR jjj vitro resulted in decreased beat frequency and degradation of the ciliated
epithelium architecture (Adalis et al. , 1977). A prior 2-h exposure j_n vivo to 2 urn aerosols
of CdCl, at 0.05 to 1.42 mg/m caused a significant decrease in cilia beat frequency propor-
3 ++
tional to the aerosol concentration. When hamsters were exposed to 1.33 mg/m Cd for 2
h/day for 2 days, the beat frequency did not return to control values until 6 wk after ex-
posure. Nickel chloride aerosols or solutions had similar, but less marked, effects (Adalis
et al., 1978). The beat frequency descreased by 60 beats/min on exposure to 0.1 mg/m" Ni
for 2 hours. The decline in beat frequency was proporti'onal to the concentration of Ni
aerosol or solution. A single 2-h exposure to 0.1 mg/m Ni depressed cilia beat frequency
12-44
-------
TABLE 12-9, EFFECTS OF H2$04 ON HUCOCILIARY CLEARANCE
Concentration
0.1 mg/m3 H2S04
0,19 to 1.4 mg/m3 H2S04 (0.3
to 0.6 pin, MMAD)
0.5 mg/m3 H2S04
1.0 mg/m3 H2S04 (0.9 pm, HMAD,
or 1.4)
1.4 mg/m3 H2S04 (0.3 to 0.6 put,
MMAD)
1.5 mg/m3 H2S04 (0.6 pm, CMD)
14 mg/m3 H2S04 (0,12 pm MMAD)
Duration
1 h/day, 5 day wk,
six mo
1 h
1 h
1 h
1 h
90 mm
Short-term
Species
Donkey
Donkey
Dog
Dog
Donkey
Mouse
Sheep
Results
All four animals developed erractic bronchial
mucociliary clearance rates within the first few
weeks Those animals never preexposed before the
0.1 mg/m3 H2S04 had slowed clearance during the
second 3 mo of exposure.
Bronchial mucociliary clearance was slowed
Slight increases in tracheal mucociliary transport
velocities immediately and 1 day after exposure.
One wk later clearance was significantly decreased.
Depression in tracheal nucociliary transport rate
persisted at 1 wk postexposure
No effect on tracheal transport
No significant effects
No significant changes fn-tracheal mucociliary
Reference
Schlesinger
1979
Schlesinger
1978
Wolff et al
Wolff et al
Schlesinger
1978

et al.,
et al.,
. , 1979
. , 1979
et al . ,
Fairchild et al. , 1975a
Sackner et
al., 1978a
15 mg/m3 H2S04 (3.2 MM, CHD) 4 h
15 mg/m3 H2S04 (3.2 pm, CMD) 90 min
transport rate

Mouse Exposure to H2S04 after exposure to a nonviable
streptococcal aerosol reduced the rate of ciliary
clearance from the lungs and nose

Mouse Exposure to H2S04 4 days prior to bacterial aerosol.
Clearance of nonviable bacteria reduced in nose,
but not lungs.
Fairchild et al , 1975a
Fairchild et al., 1975a
-------
24 h after exposure, but the frequency returned to near normal values after 72 hours.
After exposure to 0,1 nig/m , Cd+ was about 20 percent more effective than Ni++ in slowing
cilia beat (Table 12-10).
12,3.4.2 Alveolar Macrophages--The cytotoxicity of components of atmospheric aerosols has
been studied with alveolar macrophages (AM). The physiological role of AM in the prevention
of infection and the defense of the lung through removal of inhaled particles has been amply
demonstrated (Green, 1970).
The viability of guinea pig alveolar macrophages was decreased by Min-u-sil silica (6.8,
4.5, and 2.7 pm MVD), with the effect increasing as particle size decreased (Ottery and
Gormley, 1978).
Aranyi et al. (1979) reported cytotoxic effects on AM with fly ash particles coated with
PbO, NiO, or MnO_. The percentage of metal adsorbed on the fly ash was fairly similar across
particle size for a given metal. The fly ash particles were of three size ranges: <2, 2-5,
or 5-8 pm in diameter. All of the particles, regardless of the coating or particle size,
decreased cell viability and were phagocytized by the AM. Within a given chemical series of
coated particles, the effects were both concentration- and size-related, with smaller
particles and greater concentrations producing greater effects. It was suggested that the
greater surface area of the smaller particles was responsible for the greater toxicity of the
small particles. Total cellular protein and lactate dehydrogenase (LDH) also decreased
after treatment, probably as a nonspecific result of the death of the cultured AM. For each
particle size, Pb-coated particles were most toxic, NiO- and MnOo~coated particles had inter-
mediate effects, and the untreated fly ash was least toxic. The toxicity did not appear
related to the solubility of the metal oxide coating, since no soluble metal could be found
using the AM themselves as a bioassay. The toxicity appeared to be associated with the uptake
of the intact particle. No changes were observed in the total lysosomal enzyme content, but
the latency or intactness of the lysosomal membrane was not examined. Toxicity could have re-
sulted from the disruption of the intracellular lysosomal membrane, which in turn could have
released intracellular lysosomal enzymes, Lysosomal enzyme release has been proposed as one
potential mechanism for the toxicity of asbestos and silica particles (Heppleston, 1962).
These results support the concept that the 'surface activity of particles determines the
toxicity of the particle (Allison and Morgan, 1979).
Camner et al. (1974) exposed rabbit alveolar macrophages i_n vitro to 5 urn Teflon
particles coated with Al, Be, C, Pb, Mn, Ag, and U. All particles were phagocytized by the
cells, but only Be caused a decrease in viability.
12-46
-------
TABLE 12-10. EFFECTS OF METALS AND OTHER PARTICLES ON HOST DEFENSE MECHANISMS

Concentration
0.01 or 0 15 mg/m3 Pb203
(0 18 urn, MMAO)
0.01 fflg/ms (0.17 pm, MMAD) PbC12
or 0.11 mg/m-1 (0.32 pm, MMAD)
NiCljj or 0.15 mg/mj (0.15 urn,
MMAD) Pb20j or 0.12 mg/«3 (0.17
pin, MMAD) NiO
Duration Species
3 mo Rat
12 h/day, 6 day/wk, Rat
2 mo with PbCl2,
NiCl2, or NiO; con-
tinuously for 2 no
with Pb203
Results
Decreased the number of alveolar macrophages/1 ung
Exposure to Pb2Qj resulted in a depression of the
number of alveolar macrophages (AM) for up to 3 mo
but returned to control levels within 3 days after
discontinuation. NiO produced a marked AH
elevation. NiCl2 resulted in marked increases in
mucus secretion and bronchial hyperplasia.
Reference
Bingham et al.
Bingham et al.

, 1968
, 1972
0.05 to 1.42 mg/m3 CdCl2 2 h

0.1 mg/mj NiClz 2 h

Graded concentrations: 2 h
0.075 to 1.94 mgCd/nr1 as CdCl2
0 1 to 0.67 mgNi/mJ as NiCl2, or
0.5 to 5 mgMn/m3 as Mnj04;
all aerosols (94-99%) <1.4 urn
in diameter

109 mg/m3 MnOjj (0,70 un, mean 3 h/day
diameter)
0.2 mg/m3 CdS04, 0.6 mg/m3 CuS04, 3 h
1.5 ing/m3 ZnS04, 2.2 mg/B3
A12(S04)3, or 3.6 mg/rn3 MgS04
Hamster Decreased ciliary beating frequency in trachea

Hamster Decreased ciliary beating frequency in trachea

Mouse The aerosols increased the mortality from the sub-
sequent standard airborne streptococcal infection:
CdCljj affect the response at 0.1 mg/m'' NiCljt at
0.5 mg/nr* and Hnj04 at 1.55 mg/mj.
Mouse Increased mortality after 3 or 4 days exposure when
mice received bacterial aerosol immediately after
exposure. When the bacteria were administered 5 hr
post pollutant exposure, a single 3 hour exposure
increased mortality. In mice exposed to aerosols
of virus i or 2 days prior to Mn02> there were also
increased mortality and pulmonary viral lesions.

Mouse Estimated concentrations that caused a 20% enhance-
ment of bacterial-induced mortality over controls
Adalis et al., 1977

Ada!is et al., 1978

Gardner et al., 1977b
Adkins et al., 1979,
1980c
Mafgetter
et al., 1976
Ehrlich et al., 1978
Ehrlich, 197S
-------
TABLE 12-10. (continued)
Concentration
Duration
Species
Results
Reference
Ammonium sulfate at 5.3 mg/m3 3 h
S04, NH4HS04, it 6.7 mg/m4 S04,
N02S04 at 4 mg/m3 S04» Fe2(SQ4)2
at 2.9 mg/m3 S04, or
Fe(NH4)2S04 at 2.5 mg/m3 S04

5.0 ig/iii3 carbon black or 2.5 2 h
mg/m2 iron oxide

0.19 mg/ms CdCla 2 h
0.25 mg/m3 NiCl2
House No significant alterations of host defense
mechanisms
House No significant increases in mortality resulted
upon subseauent exposure to airborne infection

House Decreased number of antibody-producing spleen cells
Ehrlich et al., 1978
Ehrlich, 1979
Gardner, 1981

Graham et al., 1978
-------
Although White and Kuhn (1980) did not consider particle size, they did conduct j_n vitro
AM studies with iron carbonyl, SiO_ (size not given), crocidolite and crysotile asbestos (size
not given), kaolinite (size not given), and polystyrene latex beads (1.1 pm diameter).
Although particle to cell ratios were roughly equivalent (10 to 15 particles/ cell), the par-
ticle concentration differed markedly for each chemical. Enzyme release was measured. Com-
pared to a no particle control group, iron carbonyl, SiCL, both forms of asbestos and latex
beads, but not kaolinite, increased (p < 0.02) the percent of extracellular p-glucuronidase (a
lysosomal enzyme). Similar results were obtained for percent extracellular LDH, except latex
beads had no significant effect. All particles, except crocidolite asbestos and latex beads,
increased elastase secretion.
Allison and Morgan (1979) have summarized the evidence that AM ingest both toxic and non-
toxic particles in the same manner. In the case of fibers, ingestion appears more dependent
upon the length of the fiber (Allison, 1973). Short fibers >5 pm are almost always ingested,
while fibers >30 pm are seldom ingested completely and remain in contact with the plasma mem-
brane as well as with the lysosomal surface. Intermediate sized particles (5 to 20 pm) are
sometimes completely ingested and sometimes not. Once ingested, particles have two effects.
An immediate cytotoxicity appears that is apparently due to the interaction of the particle
with the plasma membrane (Allison and Morgan, 1979). This interaction is similar to the
hemolytic effects described for silica particles. The second effect results in delayed cyto-
toxicity and occurs after the particle has been ingested into a primary phagocytic vacuole
that then combines with a primary lysosome to yield a secondary lysosome containing the par-
ticle (Allison and Morgan, 1979). Here toxic particles exert an effect upon the permeability
of the lysosomal membrane, resulting in the release of lysosomal enzymes into the cell and
into the external medium. These proteolytic enzymes can cause tissue damage.
Hatch et al. (1980) examined the influence of i_n vitro exposure to a variety of particles
on AM oxidant production (0, and H?0p) and found the response to be chemical specific. All
the particles studied stimulated chemiluminescence, with amphibole asbestos the most active.
Silica, chrysotile asbestos, and metal oxide (Pb, Ni, Mn)-coated fly ash had intermediate
activity. Fugitive dusts and fly ash had the lowest activity.
Waters et al. (1974) found that AM cultured with particulate forms of vanadium had de-
creased cell viability, indicating a direct cytotoxicity. Alveolar macrophages were cultured
in medium containing vanadium pentoxide (VpO,-), vanadium trioxide (VgOg), or vanadium dioxide
(V0?). Cytotoxicity was directly proportional to the solubility of the vanadium compound:
V?0[- > V?0, > VOp. The concentration of V required to produce a 50-percent decrease in viabi-
lity after 20 h of culture was -found to be: 13 pg V/ml as V205, 21 (jg V/ml as VpQg, and 33 pg
V/ml as VtL. When VJ)5 was dissolved in the medium prior to incubation with the AM, only
about 9 |jg V/ml were required to reduce viability by 50 percent, thus indicating that the
soluble V was responsible for toxicity. Phagocytosis, an essential function for the defense
12-49
-------
of the lung, was decreased by 50 percent with 6 |jg V/ml as dissolved VpCL. Acid phosphatase,
a lysosomal degradation enzyme necessary for digestion of phagocytized bacteria, was inhi-
bited by 1 |jg V/ml as VpCL, while the lysosomal enzymes, lysozyme and p-glucuronidase, were
not inhibited by concentrations as high as 50 |jg V/ml.
The effects of FfipOo on AM have also been investigated. Rabbits were exposed for 3 h to
186 to 222 mg/m3 Fe203 (0.17-0.31 urn, MMAD), and AM were removed 0, 12, 18, and 24 h later
(Grant et al., 1979). When selected lysosomal enzyme activites were determined for the first
three postexposure times, there were no significant differences from control. However, since
an increased (p < 0.02) number of cells were recovered after 12, 18, and 24 h postexposure,
the total amount of some of the lysosomal enzymes in the lung was increased. It appears that
the increased number of cells was due to the influx of smaller cells into the lung.
Alveolar macrophages exposed j_n vitro for 20 h to metallic salts were also studied by
Graham et al. (1975b) using a technique to determine phagocytosis of viable cells only. The
chlorides of Cd++ (2.2 x 10"5M), Cr++ (3.1 x 10~3M), Mn++ (1.8 x 10~3M), and Ni++ (5.1 x 10~4M)
significantly inhibited phagocytosis. Ammonium vanadate (6.9 x 10 M) had no effect on phago-
cytosis, but did lyse and kill cells. Nickel, which caused the greatest reduction in phago-
cytosis, had very little effect on viability or cell lysis. Antibody-mediated rosette forma-
tion of AM was also inhibited j_n vitro by low concentrations of CdCl9 (2.2 x 10 M) or NiCl9
-4 T+ ++
(10 M) (Hadley et al., 1977). Inhibition was proportional to the Ni or Cd concentration
and reached its maximum within 20 min. These studies showed that the antibody-dependent recog-
nition system of AM was inhibited by trace concentrations of Ni and Cd almost immediately
after contact with these metal ions. Such an effect implies that these metals may affect re-
ceptors for phagocytosis of opsonized bacteria. Depression of AM viability, phagocytosis, and
receptors for phagocytosis may be a mechanism by which these heavy metal salts increase the
susceptibility to airborne infections as discussed later (Section 12.3.4.3).
Bingham and coworkers (1968, 1972} examined the effects of Pb and Ni inhalation on
the number and type of AM present in the lungs of rats. In a preliminary report, Bingham et
3
al. (1968) showed that a 3-mo exposure to 0.01 or 0.15 mg/m Pb^On (0.18 pm, MMAD) decreased
the number of AM/lung. The specificity of this response was investigated in a subsequent
study (Bingham et al. 1972) using soluble PbCl2 (0.1 mg/m3, 0.17 urn MMD) and NiCl2 (0.11
mg/m3, 0.32 urn MMD) and insoluble Pb^ (0.15 mg/m3, 0.15 pm MMD) and NiO (0.12 mg/m3, 0.25 urn
MMD) aerosols. Rats were exposed for 12 h/day, 6 days/wk for 2 months. The only exceptions
were those exposed to PbpOn continuously. Exposure to PbpOn, but not PbClp, aerosols resulted
in a depression of the number of AM that persisted throughout the experiment. The number of
o
AM was depressed on inhalation of 0.35 mg/m Pb90., for up to 3 mo, but returned to control
++
levels within 3 days after discontinuing the exposure. The solubility of the Ni compound
also effected the biological response. Nickel oxide produced a marked elevation in the number
of AM/lung, while NiCl? did not. The most significant effects in NiClp-exposed rats were marked
12-50
-------
increases in mucus secretion and bronchial hyperplasia. No morphological alterations were
observed in those rats exposed to PbCl? or Pb,0~,. Isolated AM also varied in diameter with
the exposure, but the biological significance of this size variation is not presently known.
Perhaps different cell populations were recruited into the lung with the differing exposure
conditions.
Cadmium chloride aerosols also altered the number and kind of cells recoverable by lavage
following.exposure (Gardner, 1977b, 1981). The total number of AM isolated from exposed rats
decreased following exposure to 1.5 mg/m Cd (99 percent <3 urn in diameter) but returned to
normal values within 24 hours. The viability of the isolated cells decreased by 11.2 percent
immediately after exposure and was still depressed 24 h later. There was an influx of
polymorphonuclear leukocytes, especially 24 h postexposure, but no increase in lymphocytes.
These effects were not observed at 0.5 mg/m Cd , indicating that the minimum effective dose
may lie somewhere between these two concentrations.
Nickel chloride aerosols (Adkins et al., 1979; Gardner 1981) produced neither an effect
on the number of AM isolated by lavage of rats the day following a 2-h exposure to 0.65
mg/m Ni nor an influx of polymorphonuclear leukocytes. The phagocytic capacity of the
3
isolated AM was, however, depressed. A 2-h exposure of mice to 0.9 mg/m Mn_04 reduced the
number of AM that could be recovered by lavage, but did not result in an influx of other cell
types (Adkins et al., 1980a). The AM had a reduced concentration of ATP and total protein and
acid phosphatase activity. Viability and phagocytic activity of AM were normal.
The number, function, and kind of cells isolated from the lung by lavage are influenced by
the prior exposure to heavy-metal containing aerosols. Not all of these produced the same
effect, but those containing Cd , Ni , and Mn also enhanced the susceptibility of mice to
subsequent airborne infections (Gardner, 1981). The observations of two independent labora-
tories (Bingham et al., 1968, 1972; Adkins et al., 1979) on NiCl? aerosols are essentially in
agreement (Table 12-10).
12.3.4.3 Interaction with Infectious Agents—Gardner (1981) and Ehrlich (1979) have reviewed
their groups' studies and presented new data on the effects of aerosols on host defense mecha-
nisms against infectious pulmonary disease in mice. In all of the Gardner studies, 94 to 99
percent of the aerosols was less than 1.4 urn in diameter (Gardner et al., 1977b; Gardner,
1981). Animals were placed in a head-only exposure system for 2 h and were given graded
concentrations ranging from 0.075 to 1.94 mg/m CdCl, (Gardner et al., 1977b), from 0.1 to
0.67 mg/m3 NiCl2 (Adkins et al., 1979), or from 0.5 to 5 mg/m3 Mn304 (Adkins et al., 1980c).
In mice, these exposures resulted in the deposition of 0.002 to 0.026 mg Cd (Gardner et al.,
1977b), 0.001 to 0.012 mg Ni++ (Adkins et al., 1979), or 0.005 to 0.042 mg Mn++ (Adkins et al.,
.1980b) per g dry weight of lung, respectively. Nickel clearance (Graham et al., 1978) from the
lungs of mice had a half-life of 3.4 days, while Mn (Adkins et al., 1980b) clearance was
rapid, with a half-life of only 4.6 hours. None of the exposures appeared to be edemagenic, /,
*f
12-51
-------
as judged by the ratio of dry weight to wet weight of the lung. After exposure, mice were
challenged with an aerosol of Streptococcus pyogenes (S. pyogenes). The aerosols of CdCI,
™_ ^
(Gardner et al., 1977b), NiCl? (Adkins et a!., 1979), or 'MnC1? (Gardner, 1981) increased the
4-f
mortality from the subsequent standard airborne infection. Cadmium was more toxic than Ni ,
which was more toxic than Mn . Exposure to Cd and Mn resulted in a significant linear
concentration response. The lowest concentration tested at which a significant increase in
mortality was detected was 0.1 mg/m Cd or 0.5 mg/m Ni . Manganese, as Mn.,04 (Adkins et al.,
1980c), was statistically estimated to produce a 10-percent increase in mortality at 1.55
mg/m Mn , while MnCl? (Gardner, 1981) required a higher concentration to produce a measur-
able increase in mortality. Using a different infectivity model (Maigetter et al., 1976), 3
or 4 days (3 h/day) of exposure to 109 mg/m MnO? (0.70 \an, mean diameter) were required to
increase mortality consequent to Klebsiella pneumom'ae infection when the mice received the
bacterial aerosol immediately after exposure.
The toxicity of NiClo is complex (Adkins et al., 1979). Nickel exposure had no effect on
the S. pyogenes infection if the bacteria were given immediately after Ni aerosol exposure.
When the bacterial exposure was delayed by 24 h, Ni aerosols increased the mortality in a
concentration-related fashion. In contrast, effects of CdCl2 (Gardner et al., 1977b) and Mn
(Gardner, 1981) were observed when the bacterial challenge immediately followed exposure. The
concentration-response curve of Ni was very steep compared to those of Cd and Mn ex-
posures (Gardner, 1981). No explanation has been offered for the delay in effect of Ni
Perhaps the delayed effects represent either redistribution of Ni to the site of action or
some major change in the lung, such as death of a specific cell type. The delayed toxicity
raises the possibility of carry-over of effects from a single exposure to a second.
The influence of a variety of sulfate species on host defense mechanisms against infec-
tious respiratory disease has been investigated by Ehrlich (1979) and Ehrlich et al. (1978)
using the infectivity model with S. pyogenes. Mice were exposed for 3 hours. The estimated
concentrations of the compounds that caused a 20-percent enhancement of bacterial-induced mor-
tality over controls were 0.2 mg/m3 CdS04, 0.6 mg/m3 CuS04, 1.5 mg/m3 ZnS04, 2.2 mg/m3
A12(S04)3, 2.5 mg/m3 Zn(NH4)2(S04)2, and 3.6 mg/m3 MgS04- Ammonium sulfate at 5.3 nig/m0 S04,
NH4HS04 at 6.7 mg/m3 S04, Na2S04 at 4 mg/m3 S04, Fe2(S04)3 at 2.9 mg/m3 S04, and Fe(NH,)2$04
at 2.5 mg/m SO. did not cause significant alterations The nitrates of Pb , Ca , Na , K ,
3
and NH4 did not cause an effect at concentrations of 2 mg/m or higher; however, Zn(NO,,)2
caused effects similar to ZnSCL. From this research, it appears that the NH4 ion rendered the
compound less toxic, and that the toxicity is due primarily to the cation. With the infec-
tivity model, ZnSCL, Zn(NH4)2(S04)2, and CuS04 ranked differently than with airway resistance
experiments (Amdur et al., 1978a). This is not unexpected, as airway resistance primarily
detects alterations of the medium to large conducting airways, while the infectivity model
(Gardner and Graham, 1977) is hypothesized to reflect alveolar level changes.
12-52
-------
3 3
When mice were exposed for 2 h to 5.0 mg/m carbon black or 2.5 mg/m iron oxide, no
significant increases in mortality resulted on subsequent exposure to airborne infection
(Gardner, 1981).
Death from S. pyogenes exposure in this infectivity model is due to septicemia (Gardner
et al., 1977b). Septicemia occurs when the bacteria have grown to 10 organisms per lung.
Removal and killing of the inhaled organisms reduces the growth of the bacteria within the
host and prevents the occurrence of septicemia. For these reasons, the infectivity model is an
integrative assessment of toxicity for host defense systems against infectious pulmonary
disease. As reported above, the number, kind, function, and viability of the cells isolated
by lavage from the lungs of animals exposed to heavy-metal containing aerosols are different
from those of control animals. Studies of trachea! rings isolated from aerosol-exposed
hamsters also indicate depression of mucociliary clearance. Both mucocilary and AM clearance
of bacteria are depressed by aerosols of these heavy metals (Gardner, 1981) (Table 12-10).
12.3.4.4 Immune Suppression—Antibodies play a significant role in the ability of macrophages
to recognize and engulf pathogenic bacteria. The functioning of the immune system also inter-
locks with the macrophage system in other ways. In mice, intramuscular injections of NiCK
depressed the number of antibody-producing cells in the spleen (Graham et al., 1975a). Using
the Jerne plaque assay, a negative linear dose-response curve was found with injections rang-
ing from 0.093 to 0.123 mg Ni /g body weight. No effect was observed with a dose of 3.09 jjg
Ni /g body weight. Graham et al. (1978) calculated that exposure to an aerosol of 0.25
mg/m Ni for 2 h results in a maximum deposition of 0.98/mg Ni , assuming complete re-
tention and a minute volume of 1,45 ml/g body weight. This concentration was found to be the
lowest tested that produced a significant depression in the immune response. The lowest dose
found to produce a similar effect by injection was 0.21 mg Ni /mouse (Graham et al., 1975a).
The inhalation dose was, therefore, approximately 200 times more potent. Nickel was found to
follow first order removal kinetics from the lung, but measurable elevations remained in the
lung up to 4 days after exposure. Similar kinetics of removal have been found using the iso-
lated, ventilated, and perfused rat lung (Williams et al., 1980) and human, rat, and cat type
II pneumocytes in culture (Saito and Menzel, 1978).
Inhaled Cd also depresses the number of antibody-producing cells and is more potent
than intramuscularly injected Cd . The highest intramuscular dose of CdClo examined by
Graham et al. (1978) was 0,012 mg Cd /g body weight (about 0.266 mg Cd /mouse), and it pro-
duced no immunosuppression. When mice were exposed to 0.19 mg/m Cd for 2 h, a signi-
ficant suppression was observed. In both cases, the Cd was administered as CdCl™, a highly
soluble salt. The inhalation dose can be calculated, on the same basis as that given above
for Ni , to be at most 0.0007 mg CdCl?/mouse. The inhaled dose was, therefore, at least 350-
fold more potent. Inhalation also appeared to be more potent than ingestion or interperitoneal
injection (Exon et al. , 1975; toller et al., 1975). Keller et al. (1975) found that 0.150 mg
Cd given orally was required to produce immunosuppression.

12-53
-------
For comparative purposes, the lowest inhalation exposure of CdCl9 found to be immunosup-
o o
pressive was 0.19 mg/m ; 0.2 mg/m was the 1971 Threshold Limit Value (TLV). The current TLV
q
is 0.05 mg/m . Conservative estimates of the human intake from air and water are 0.0074 rag/
day and 0.160 rag/day, respectively (Schroeder, 1970). NiCl» was found to be immunosuppressive
3 3
at an inhalation exposure of 0.25 mg/m , while its TLV is 1 mg/m . The human exposure is
estimated to be 0.0024 mg/day from inhalation and 0.60 mg/day from ingestion (Schroeder, 1970).
Should the effectiveness of inhaled aerosols be equivalent in mice and men, then the inhaled
doses are biologically almost equivalent to those ingested.
Inhaled Cd or Ni aerosols impair the bacterial defenses of the lung through direct
cytotoxicity to AM, inhibition of antibody dependent aggregation reactions, and perhaps by de-
pression of antibody production." All of these mechanisms might help to explain the increased
susceptibility of mice to airborne pathogens following inhalation of Ni or Cd aerosols.
The rapidity of clearance of Ni and Cd from the lung may allow rapid recovery (see
Table 12-10).
Hitogen-induced transformation (a reflection of immune function) has been measured in
mouse splenic lymphocytes exposed jji vitro to 0.50 rag of various sizes of silica (Wirth et
al. , 1980). Four silica samples were tested unfractionated, or size-fractionated into two
categories (0.3 and 5.3 pm). All the unfractionated samples depressed the blastogenic re-
sponse to Conconavalin A (a measure of T cell function) and lipopolysaccharide (a measure of B
cell function). The T cell response was decreased by the 0,3 |jm size fraction of all samples.
The 5.3 urn particles of 2 samples increased the response, however, while one sample caused no
change and another caused a small decrease. When B cell function was examined, it was more
depressed by the 0.3 pm si lea than the 5.3 pm particles, although three of the four larger-
sized samples did cause a decrease.
12.4 INTERACTION OF SULFUR DIOXIDE AND OTHER POLLUTANTS
12.4.1 Sulfur Dioxide and Particulate Hatter
Although man breathes a multitude of chemicals in various mixtures at various dose-rates,
most animal toxicological and controlled human exposures are conducted with single chemicals.
This simplifies the research and permits an improved estimate of cause-effect relationships,
but it prohibits evaluation of the effects of pollutant mixtures that may be additive, syner-
gistic, or antagonistic with respect to the individual pollutants. Some interaction studies
that elucidate the complexity of toxicological interrelationships, however, have been con-
ducted. Some of this work used pollutant combinations that favored the conversion of the pri-
mary pollutant to a secondary pollutant (i.e., SO, altered to H,,SQ«, etc.). Other research
was directed at evaluating the influence of several pollutants when delivered in combination
or in sequence.
12.4.1.1 Acute Exposure Effects—The question of the possible effect of aerosols on the re-
sponse to S0? is important in air pollution toxicology (Amdur, 1977). The phenomenon has been
investigated in simple model systems of SO- alone or in combination with an aerosol of a single

12-54
-------
chemical. The typical bioassay system has been the comparison of the increase in pulmonary
flow resistance in guinea pigs produced by a given concentration of SO, alone with that pro-
duced in the presence of the aerosol. The aerosols used in many of these studies were "inert"
in the sense that they did not produce an alteration in flow resistance when they were given
alone.
The initial simple prototype aerosol used was sodium chloride (NaCl) at concentrations of
3 3
10 mg/m and 4 mg/m (Amdur, 1961). These experiments with guinea pigs indicated that the re-
3
sponse to a given concentration of SO, was potentiated by 10 mg/m NaCl. For example, a con~
3
centration of 5.24 mg/m (2 ppm) SO, alone produced an increase of 20 percent in pulmonary
flow resistance; when the NaCl was present, the increase was 55 percent. The potentiation did
not occur until the latter part of a 1 h exposure. When the concentration of sodium chlo-
3
ride was reduced to 4 mg/m , the potentiation was greatly reduced. Examination of postexpo-
sure data indicated that the response to the combination resembled the response to a more
irritant aerosol. The length of recovery was related to the concentration of SO,, and the
presence of the aerosol delayed recovery to control values. The chamber relative humidities
were below 70 percent; but on entering the high humidity of the respiratory tract, the NaCl
would absorb water to become a droplet capable of dissolving SO,, thus favoring the production
of HoSO*. Sodium chloride alone does not catalyze the oxidation of SO, to hLS04.
Experiments by McJilton et al. (1973, 1976) indicate the importance of ambient RH and the
solubility of SO, in the sodium chloride droplet. They examined the effect of 1 mg/m NaCl on
3
the response to 2.62 mg/m (1 ppm) SO, at low (<40 percent) and high (>80 percent) RH. An in-
crease in pulmonary flow resistance in guinea pigs was the response criterion. As would have
been predicted from the earlier work, no increase was observed with this NaCl concentration at
low RH. At high RH, the potentiation was marked and evident during both the early and
late parts of the 1-h exposure. The rapid onset indicates the formation of an irritant
aerosol in the exposure chamber under high humidity. Predictably, no conversion to sulfate
was found, but the droplets were acidic with an estimated pH of 4. Presumably, this was sul-
furous acid. (See the discussion of the effect of RH on sulfate and nitrate aerosols and the
interaction of NaCl above and on human exposure experiments in Chapter 13).
Using the Mead-Amdur method, Amdur and Underbill (1968) studied the effect of aerosols of
soluble salts of metals shown to convert SO, to I-LSO^. Manganous chloride, ferrous sulfate,
and sodium orthovanadate caused a three-fold increase in the resistance to flow over that of
3
2,62 mg/m (1 ppm) SO, alone. The potentiation was evident during the first 10 min as well as
during the remainder of the 1-h exposure. Chamber RH was 50 percent, indicating that high
humidity was not necessary for the formation of an irritant aerosol in the chamber when the
catalyzing metals were present. Analysis of the collected aerosol indicated the presence of
sulfate, presumably as H^SO, (Amdur, 1973). These analyses indicated that, at an S02 concen-
tration of 0.52 mg/m (0.2 ppm), about 0.08 mg H,S04 was formed. When this amount of H,SO* was
administered with 0.52 mg/m (0.2 ppm) SO,, the increase in flow resistance duplicated the

12-55
-------
increase observed with the iron and vanadium aerosols (Amdur, 1974). - This suggests that H-SCL
formation is a likely mechanism of potentiation for the aerosols of these metals, although the
formation of stable sulflte complexes in the air may also account for this effect (Hansen et
al., 1974; Schlesinger et a!., 1980).
Atndur et al. (1978a) have reported that a 1-h exposure to 0.4 mg/m copper sulfate also
potentiated the response to 0.94 mg/m (0.36 ppm) SO,,. It is not certain whether this is
caused through the formation of H9SO, or the formation of a sulfite complex. The increased
3
resistance to flow from exposure to 0.79 to 0.84 mg/m (0,3 to 0.32 ppm) S09 with ammonium
3 3 3
sulfate (0.9 mg/m ), ammonium bisulfate (0.9 mg/m ), or sodium sulfate (0.9 mg/m ) was purely
additive. It should be pointed out that these salts have not been tested under conditions of
high relative humidity.
Amdur and Underbill (1968) also examined the effect of a variety of solid aerosols that
do not catalyze the conversion of S0« to H?SO,. None of these potentiated the increased
resistance to flow when compared to S0? (Table 12-11).
12.4.1.2 Chronic ExposureEffects—Animals were exposed continuously to various combinations
of S02, H£S04 (0.5 to 3.4 Mm, HMD), and fly ash (3.5 to 5.9 (jm, HMD) (Alarie et al., 1975).
The fly ash had been collected downstream from electrostatic precipitators of coal-burning
electric generating plants. Monkeys were exposed for 18 mo and guinea pigs for 12 months.
The monkeys were exposed to S02, H2S04 + fly ash, S02 + H2S04, or S02 + H2S04 + fly ash.
Guinea pigs received either 0.9 mg/m3 H2S04 (0.49 urn HMD) or 0.08 mg/m3 H2S04 (0.54 or 2.23 |jm
HMD) + 0.45 mg/m fly ash (3.5 or 5.31 um MMD). In monkeys, a battery of hematological and
pulmonary function (tidal volume, respiratory rate, minute volume, dynamic compliance, pulmon-
ary flow resistance, work of breathing, distribution of ventilation, CO diffusing capacity,
and arterial blood gases) tests were applied at various times during exposure, but no signifi-
cant effects were attributed to the exposures. Similar methods (except for distribution of
ventilation and CO diffusing capacity) were used with guinea pigs, and again no significant
3 3
effects were observed. At the end of the exposure to 2.59 mg/m (0.99 ppm) S02 + 0.93 mg/m
HoSO^ (0.5 urn MMD, a = 1.5 to 3.8), the lungs of monkeys had morphological alterations in the
bronchial mucosa (focal goblet cell hypertrophy and occasional hyperplasia and focal squamous
metaplasia). Monkeys exposed to 2.65 mg/m3 (1.01 ppm) S02 + 0.88 mg/m3 H£S04 (0.54 urn MMD,
a = 1.5 to 3.8) + 0.41 mg/m3 fly ash (4.1 urn MMD, or = 1.8 to 2.8) had similar alterations.
5 83
Thus, fly ash did not enhance the effect. Monkeys that received 0.99 mg/m H~S04 (0.64 urn
MMD, a = 1.5 to 3.0) + 0.55 mg/m3 fly ash (5.34 urn MMD, a =1.8 to 2.2) had slight altera-
tions in the mucosa of the bronchi and respiratory bronchioles. Focal areas of erosion and
epithelial hypertrophy and hyperplasia were observed. The other groups of monkeys had no
remarkable morphological changes. No monkeys exposed to fly ash displayed morphological
alterations, although the presence of the fly ash was easily observed. Guinea pigs had no
morphological effects that could be attributed to pollutant exposure.
12-56
-------
TABLE 12-11. EFFECTS OF ACUTE EXPOSURE TO S02 IN COMBINATION WITH CERTAIN PARTICLES
Concentration
Duration
Species
Results
Reference
5 24 mg/m3 (2 ppm) S02, 10 mg/m3 1 h
and 4 wg/tn3 NaCl
2 62 mg/m3 (1 ppji) SOz, 1 mg/m3 1 h
NaCl at low (40%) and high (80%)
(RH)

2 62 rag/M3 (1 ppm) S02, an 1 h
aerosol of soluble salts
(nianganous chloride, ferrous
sulfate, and sodium orthovana-
date) 50% RH

0 94 mg/m3 (0.36 pptn) S02, 1 h
0.4 mg/m3 copper sulfate

0 79 to 0.84 mg/m3 (0.3 to 1 h
0 32 ppm) S02 and 0,9 ntg/m3
ammonium bisulfate, or 0 9
mg/m3 sodium sulfate
Guinea pig 5.24 mg/m3 SOZ alone produced an increase Amdur, 1961
of 20% in pulmonary flow resistance; with NaCl at
10 mg/M3 the increase was 55%.

Guinea pig No increase in pulmonary flow resistance at low RH. HcJilton et al., 1973
At high RH, the potentiation was marked.

Guinea pig Presence of soluble salt increased pulmonary flow Amdur and
resistance about 3-fold Underhill, 1968
Guinea pig Potentiated pulmonary flow resistance
Guinea pig The effect on pulmonary flow resistance was
additive
Amdur et al , 1978a

Amdur et al., 1978a
1 ppm S02 = 2.62 mg/m3.
-------
In a previous study, Alarie et al. (1973bc) found no effects on pulmonary function, hema-
3
tology, or morphology of monkeys or guinea pigs exposed to approximately 0.56 mg/m fly ash in
combination with 3 concentrations of SO, (0.28, 2.62, or 13.1 mg/m ; 0.11, 1, or 5 ppm).
Monkeys were exposed continuously for 78 wk and guinea pigs continuously for 52 weeks.
Lewis et al. (1969, 1973) investigated the effects of SO, and H2S04 in normal dogs and in
3
dogs that had been exposed previously for 191 days to 48.9 mg/m (26 ppm) N0?. Dogs identic-
ally treated with NO- had morphological changes in the lung, and one of the animals had strik-
ing bullous emphysema. Sulfur oxide exposures were for 21 h/day for a maximum of 620 days
3 3
to 13.4 mg/m (5.1 ppm) S02, to 0.89 mg/m HgSO, (90 percent < 0.5 urn in diameter), or to a
combination of the two. These concentrations were averaged over time, and when the animals
3
were examined at 225 days, the concentration of H^SO. was lower (0.76 mg/m H2SO, in the H-SO.
group and 0.84 mg/m H2SO, in the H2SO. + S02 group). After 225 days of exposure (Lewis et
al., 1969), dogs receiving H~S04 had a significantly lower diffusing capacity for CO than those
that did not receive H^SO.. In the S0?~exposed animals, pulmonary compliance was reduced (p <
0.05), and pulmonary resistance was increased (p < 0.05) compared to animals that did not re-
ceive S0?. Dogs not preexposed to N0? that received S0? + HpSO, had a smaller residual volume
(p < 0.01) than all other dogs.
These dogs were also examined after 620 days of exposure (Lewis et al., 1973). At 3, 7,
19, or 20.5 mo of exposure, SO did not markedly affect hematological indices. No morpho-
A
logical changes were clearly identified as resulting from SO exposure; however, pulmonary
function was altered. Generally, the animals preexposed to N0? were more resistant to the
SO . Sulfur dioxide did not produce any significant effects except for an increase in nitro-
gen washout. Sulfuric acid caused a significant (p < 0.05) decrease in diffusing capacity for
CO, residual volume, and net lung volume (inflated) with an increase in total expiratory re-
sistance. There was also a marginally significant (p = 0.1) decrease in total lung capacity,
inspiratory capacity, and functional residual capacity. Total lung weight and heart weight
also decreased. Other measurements (other lung volumes, dynamic and static compliance, and
nitrogen washout) were not significantly affected. These alterations of diffusing capacity
for CO and lung volumes are interpreted as a loss of functional parenchyma, and, along with
the increase in total pulmonary resistance, are in the direction expected for animals that
develop obstructive pulmonary effects. Although the standard histological techniques used did
not detect morphological effects, it is conceivable that the pulmonary function effects pre-
ceeded measurable structural alterations.
Female beagle dogs were exposed 16 h/day for 68 mo to raw or photochemically-reacted
auto exhaust, oxides of sulfur or nitrogen, or their combinations. Table 12-12 describes the
exposure groups. More than 90 percent of the particles were <0.5 urn in diameter. The dogs
were examined after 18 (Vaughn et al., 1969), 36 (Lewis et al., 1974), and 61 mo (Lewis et al.,
12-58
-------
1974) of exposure and 24 mo (Gillespie, 1980) or 32-36 mo (Hyde et al., 1978; Orthoefer et
al., 1976) after the 68 mo exposure ceased. A monograph describing the entire study and re-
sults is available (Stara et al. , 1980). Only those results pertaining to SO are described
X
here.
Although cardiovascular function was also assessed'after* 4 yr of exposure and 3 yr after
exposure ceased, no significant changes attributable to SO were found (Gillespie, 1980).
Typical hematological examinations (except for differential counts) were made approximately
every 6 mo (Orthoefer et al., 1976). The SO group (see Table 12-12 for abbreviations) had no
major differences from control. In the presence of auto exhaust (with or without irradia-
tion), however, SO did cause some significant elevations in hematocrit and hemoglobin concen-
X
tration. Clinical chemistries were unchanged during or approximately 1 1/2 yr after exposure
(Gillespie, 1980).
A variety of other parameters were examined during or immediately after exposure
(Gillespie, 1980). Sulfur oxides caused no significant effect on visual evoked brain
potentials.
After 18 (Vaughn et al., 1969) or 36 mo (Lewis et al., 1974) of exposure, no significant
changes in pulmonary function were observed. A variety of alterations were found using
analysis of variance after 61 mo (Lewis et al., 1974) of exposure. Residual volumes were in-
creased in dogs receiving R + SO compared to those receiving I + SO , SO , and clean air
X X X ry
(CA). Residual volumes of the SO group were lower than those of the CA group. When x
X
analyses were applied to the data of the number of dogs/group having alterations as judged by
clinical criteria, additional significant differences were found. More dogs of the I + SO
group had higher total expiratory resistance than their controls (CA and SO ). The ratio of
residual volume to total lung capacity was higher in animals exposed to R + SO , compared to
X
those receiving CA. This change was interpreted as pulmonary hyperinflation. Although other
lung volumes, compliance, resistance, diffusing capacity for CO, nitrogen washout, peak
expiratory flow, and maximum breathing capacity were also measured, SO had no effects.
X
Two years after exposure ceased, pulmonary function was remeasured (Gillespie, 1980).
These measurements were made in a different laboratory than those made during exposure, but
consistency among measurements of the control group and another set of dogs of similar age at
the new laboratory indicated that this difference did not have a major impact on the findings.
Animals in the R, R + SO , and I + SO groups had an increased arterial pressure of C0£
(PaC09) compared to controls (p < 0.05). These groups and the SO group had a greater dead
£- X
space volume compared to controls. Respiratory frequency was increased in the SO,, group.
Although the diffusing capacity of CO in the lung (DLCQ) was unchanged, the ratio of DL^g to
total lung capacity decreased in all pollutant-exposed dogs. Vital capacity did not change.
12-59
-------
TABLE 12-12. POLLUTANT CONCENTRATIONS FOR CHRONIC EXPOSURE OF DOGS9

3
Pollutant Concentration, mg/m

HC OX
Atmosphere CO (as CHJ N09 NO (as 0,) S09 H9SO,,
T- L. O £ £ T-

Control Air (CA)b - - -

Nonirradiated auto
exhaust (R) 112.1 18.0 0.09 1.78

Irradiated auto
exhaust (I) 108.6 15.6 1.77 0.23 0.39

S02 + H2S04(SOX)C - - - - - 1.10 0.09
t—•
[NJ
cr> Nonirradiated auto
° exhaust + S02 + 113.1 17.9 0.09 1.86 - 1.27 0.09
H£S04 (R + S0x)

Irradiated auto
exhaust + SO '+ H9SOA 109.0 15.6 1.68 0.23 0.39 1.10 0.11
(I + SV

Nitrogen oxides, 1
(N02 high) - - 1.21 0.31 -

Nitrogen oxides, 2
(NO high) - - 0.27 2.05 -

aHyde et a!., (1978).

Abbreviations in parentheses
c>90% of H2S04 particles were < 0.5 (jm in diameter (optical sizing)
-------
Compared to control values, total lung capacity and residual volume of the SO group were
significantly increased; there were no changes in functional residual volume and respiratory,
pulmonary, and chest wall resistance; and quasistatic chest wall compliance decreased. There
was a greater change in dynamic compliance with increasing breathing frequency in dogs exposed
to SO . When the pulmonary function values at the end of exposure were compared directly to
X
those values 2 yr after exposure, the SO group had increases in residual volume, total lung
capacity, vital capacity, inspiratory capacity, functional residual capacity, DLCQ, and the
ratio of DL~Q to total lung capacity. The magnitudes of these changes were greater than
changes in controls, in most cases. From evaluation of all the data, the authors indicated
that functional loss continues following termination of the exposure, and the damage caused by
SO is primarily to the parenchyma. They also stated that the combination of auto exhaust and
SO "did not appear to augment specific functional losses caused by single species of pol-
lutants."
Thirty-two to 36 mo (Hyde et al., 1978) after exposure, the lungs of the beagles were
examined using morphologic (light, scanning electron and transmission electron microscopy) and
morphometric techniques. In the SO group, lung weight, total lung capacity, and the
displaced volume of the processed right lung were significantly increased over the controls
(CA). In the most severely affected SO dogs, the air spaces enlarged and the number and size
A.
of interalveolar pores increased. Only the high-NO? group of dogs had a greater degree of air
space enlargement. The SO -group animals had a loss of cilia in the conducting airways with-
out squamous cell metaplasia, nonciliated bronchiolar cell hyperplasia, and loss of inter-
alveolar septa in alveolar ducts. When SO was combined with R, cilia were also lost, but
X
squamous cell metaplasia occurred. Exposure to R + SO and I + SO produced nonciliated
bronchiolar cell hyperplasia and an increase in interalveolar pores and alveolar air space
enlargement. The enlargement of the distal air spaces was centered on respiratory bronchioles
and .alveolar ducts and was associated with an apparent loss of interalveolar septa in all
animals receiving S0? and H^SO,. The authors consider these changes analogous to an incipient
stage of human proximal acinar (centrilobular) emphysema. The important observation from
these experiments is that mixtures of SO,, and HUSO., representing an interacting gas-aerosol
system similar to that in urban atmospheres, produced anatomic alterations at low concentra-
tions.
In a monograph describing all the dog studies (Stara et al., 1980), the morphological and
functional changes are compared. In the SO group, the changes in pulmonary function corre-
X
lated well with the morphological effects. Since the changes in pulmonary function were pro-
gressive over the postexposure period, it is likely that morphological changes were also pro-
gressive.
12-61
-------
Biochemical analyses were performed on these dogs at the time of sacrifice, 2.5 to 3 yr
after exposure, Hydroxyproline concentration (used as an index of collagen content) and prolyl
hydroxylase activity (the rate-limiting enzyme in collagen synthesis) were measured (Orthoefer
et al., 1976). No significant changes in hydroxyproline were found. The SO and I + SO
X X
groups had significantly elevated prolyl hydroxylase activity compared to the R, R + SO , and
A
CA groups. While it is remarkable that effects on prolyl hydroxylase remained 2.5 to 3 yr
after exposure, it is not possible to interpret these results further. No significant altera-
tions were observed in brain, heart, lung or liver lipids among the experimental groups
(Gillespie, 1980).
Zarkower (1972) reported mixed effects on the immune system of mice exposed to 5.24 mg/m
2
(2 ppm) SO, and 0.56 mg/m carbon (as carbon black, 1.8 to 2.2 pm, HMD), alone and in combi-
nation for 100 h/wk for up to 192 days. The animals were immunized with aerosols of bac-
teria (Escherichia coli) at various times during exposure. After 102 days of exposure, there
were no statistically significant changes. Sulfur dioxide exposure caused an increase (p <
0.05) in serum antibody titer at 135 days and a decrease (p < 0.01) at 192 days. Carbon and
SO- + carbon produced an equivalent decrease (p < 0.01) in antibody titer at 192 days (but not
at 135 days) that appeared to be a greater decrease than that found in the SOp-exposed mice.
In the spleen, exposure to SOp caused an increase (p < 0.01) in the number of antibody-produc-
ing cells at 135 days and a decrease (p < 0.01) in number at 192 days. In the mediastinal
lymph nodes (that drain the lung), SO, caused no such changes. Carbon + S0?, but not carbon
alone, caused an increase (p < 0.01) in the number of antibody-producing cells in the medias-
tinal lymph nodes and a decrease (p >0.05) in the spleen at 135 days. After 192 days of ex-
posure to carbon or carbon + S0?, the number of antibody producing spleen cells decreased (p
<0.01). The immunosuppression in these 2 groups was roughly equivalent and appeared to be
more severe than that in the SO,, alone group. In the mediastinal lymph nodes, only carbon +
S0« caused immunoenhancement (p < 0.05). Thus, for the pulmonary immune system, only exposure
to the combination of SO- and carbon caused significant effects. After 192 days, the systemic
immune system was affected in all 3 exposure groups. It appeared that carbon and carbon + SO™
caused equivalent effects and that both regimens were more effective than SQ2 alone.
Fenters et al. (1979) showed that exposure for 3 h/day, 5 days/wk for up to 20 wk to
3 3
a mixture of 1.4 mg/m H9SO, plus 1.5 mg/m carbon (as carbon black, 0.4 urn, mean particle
3
diameter) or to 1.5 mg/m carbon only (0.3 pm, mean particle diameter) also altered the immune
system of mice. Serum immunoglobulins (Ig) decreased, with the exception of IgM, which was in-
creased after 1 wk of exposure to either carbon or HpSO, + carbon. After 1 wk, some Ig
classes decreased in both exposure groups, but after 4 or 12 wk of exposure, alterations were
observed only in the H2SO. + carbon group. Results for Ig were mixed at 20 weeks. In the
carbon group, the number of specific antibody-producing spleen cells was increased at 4 wk,
unchanged at 12 wk, and decreased at 20 weeks. A similar trend was observed in the HgSO^ +
carbon group, but only the immunosuppression at 20 wk was significant. In examining other

12-62 !
-------
host defense systems, no alterations of AM viability or cell numbers were observed. After 4
and 12 wk of exposure, pulmonary bactericidal activity was increased in both exposure groups.
By 20 wk of exposure, values were not significantly different from controls. Using the in-
fectivity model with influenza A_/Taiwan virus, a 20-, but not a 4-, wk exposure to H2$0. +
carbon increased mortality.
Morphological changes were observed in these mice (Fenters et a"!., 1979) using scanning
electron microscopy after 12 wk of carbon exposure. In the external nares, there was excess
sloughing of squamous cells. In the trachea, the number of mucous cells appeared to increase;
dying cells were present, and microvilli were lost. No alterations of the bronchi were seen.
The alveoli had some areas of congestion with thickening, loss of interalveolar septa, and en-
larged pores. After 20 wk of exposure, damage was similar, but to a lesser degree. Mice ex-
posed to the mixture of H?SQ. and carbon showed equivalent effects, but the damage was some-
what more severe than that seen in the carbon only group.
The influence of H?SO, and carbon on the tracheas of hamsters was investigated by Schiff
3
et al. (1979). Animals were exposed for 3 h to 1.1 mg/m H,SO. (0.12 jjm, mean size) and/ or
3
1.5 mg/m carbon (as carbon black, 0.3 urn, mean size) and were examined either immediately, or
24, 48, or 72 h later. Carbon caused no change in ciliary beat frequency. Sulfuric acid ex-
posure, however, caused depression in this frequency at all times. The combination of H»SO*
and carbon produced similar effects, but recovery had occurred by 48 h postexposure. Using
light microscopy, the percentage of normal trachea! epithelium was determined. Up to 48 h
after exposure, the combination of H?SO. and carbon resulted in more tissue destruction than
either pollutant alone, although the single pollutants did cause some damage. Morphological
alterations of all pollutant exposure groups were observed using light and scanning electron
microscopy (see Table 12-13).
12.4.2 Interaction with Ozone
3
Cavender et al. (1977) exposed rats and guinea pigs to H9SOA aerosols (10 mg/m , 1 urn
3
MMD), 3.9 mg/m (2 ppm) ozone (0~), or a combination of the two for 6 h/day for 2 or 7 days;
they then measured the ratio of lung to body weight and examined the lungs histologically. No
synergism was observed between the 0_ and H^SO. treatments. The histological lesions were
those ascribed to 0~ alone. Th'is same group (Cavender, 1978) exposed rats and guinea pigs to
H9SO. aerosols (10 mg/m3, 0.83 |jm, MMAD, a = 1.66), 1.02 mg/m3 (0.52 ppm) 0,, or a combina-
£. 4- g 3
tion of the two for 6 h/day, 5 days/wk for 6 months. The histological alterations were those
due to 03 alone.
Last and Cross (1978) found synergistic effects of a continuous exposure of H9SO. aerosol
3 3
(1 mg/m ) and 0- (0,78 to 0.98 mg/m or 0.4 to 0.5 ppm) when administered simultaneously to
O
rats for 3 days. Glycoprotein synthesis was stimulated in trachea! ring explants measured ex
vivo. Ozone alone caused a decreased glycoprotein secretion; H^SO. was relatively inactive,
12-63
-------
TABLE 12-13.
EFFECTS OF CHRONIC EXPOSURE TO SOX AND SOME PH
Concentration
Duration
Species
Results
Reference
Various combinations of S02,
H2S04 (0.5 to 3.4 Mm, HMD),
and fly ash (3.5 to 5.9 urn,
HMD): S02, H2S04 + fly ash,
S02 + H2S04, S02 + H2S04 +
fly ash
0.9 mg/m3 H^SO,, (0.49 urn,
HMD), 0.08 mg/mj H;,S04
(0.54 or 2.23 urn, HMD) +
0.45 mg/mj fly ash (3.5
or 5.31 urn, HMD)

Approximately 0.56 nig/nr1 fly
ash in combination with SO,, at
0.28, 2.62, or 13.1 mg/mj (0.11,
1, or 5 ppm).

Approximately 0.56 mg/nr5 fly ash
in combination with SO^ at
0.28, 2 62, or 13.1 mg/m-5
(0.11, 1, or 5 ppm)

13.4 mg/nr1 (5.1 ppm) SO,,, or
0.89 mg/mj H^S04 (90% <0.5
urn in diameter), or to a
combination of the two
18 mo,
continuous
Monkey
12 no,
continuous
78 wk,
continuous
52 wk,
continuous
Guinea pig
Monkey
Guinea pig
21 h/day, 620 days Dog
No significant effects on henatology or pulmonary Alarie et al., 1975
function tests during exposure. At end of exposure
to S02 + H2S04 lungs had morphological alterations.
Exposure to S02 + H2S04 + fly ash had similar
alterations; thus fly ash did not enhance effect.
Exposure to H<,S04 + fly ash had slight alterations.
(See pp 12-56 and 58 for details on size and con-
centration. )
No significant effects on hematology, pulmonary
function, or morphology
No effects on pulmonary function, hematology,
or morphology
No effects on pulmonary function, hematology,
or morphology
Alarie et al., 1975
Alarie et al., 1973b
Alarie et al., 1973b
After 225 days, dogs receiving H<.S04 had a lower
diffusing capacity for CO than those that did not
receive H^S04. In the SO^-exposed group, pulmonary
compliance was reduced and pulmonary resistance was
increased compared to dogs that did not receive SO^
Dogs not pre-exposed to NO^ who received SO;. + H^S04
had a smaller residual volume than all other dogs
After 620 days, SO^ increased mean nitrogen wash-
out, but no hematological or morphological changes
occurred. H^S04 decreased: diffusing capacity
for CO, residual volume, net lung volume, total
lung capacity, inspiratory capacity functional
residual capacity, total lung weight and heart rate.
Total expiratory resistance increased.
Lewis et al., 1969, 1973
-------
TABLE 12-13 (continued)
Concentration
Duration
Species
Results
Reference
(see Table 12-12)
16 h/day, 68 mo Dog
5.24 mg/mj (2 ppm) S0^t or 0.56
mg/mj carbon (1.8 to 2.2 pm,
HMD), or in combination
1.4 flip/m4 H^S04 plus 1.5
ftig/mj carbon (0 4 ym, mean
particle diameter), or 1.5
mg/uij carbon only (0.3 uoi,
mean particle diameter)

1.1 ng/irr* HuS04 (0.12 UM, mean
size), or 1 5 mg/m3 carbon (0.3
MO, mean size), or in combination
100 h/wk, 192 days House
3 h/day, 5 day/wk, House
20 wk
3 h
Hamster
+ SO group had
Residual volumes increased in group receiving
R + SOX compared to CA, which showed increase

compared to SOX group. The I

higher total expiratory resistance. Thirty-two
to 36 mo after exposure ceased, the SO group had

increases in lung weight, total lung capacity, and
displaced voluine of the processed right lung and
loss of cilia in the conducting airways. SO + R
Lewis et al., 1969,
1973
had loss of cilia and squamous metaplasia.
to R + SO and I +
$QX produced nonciliated
x
Exposure
bronchiolar cell hyperplasia, increased inter-
alveolar pores, and alveolar air space enlargement,

For the pulmonary iimune system, only exposure to
the combination caused significant effects;
the systemic immune system was affected
in all 3 exposure groups. Carbon and carbon + SO*
were more effective than SOj., although SO* did
cause significant effects.

Altered the immune systen. Morphological changes
observed; more severe with carbon only exposure.
Zarkower, 1972
Fenters et al., 1979
Ciliary beat frequency depressed after HjjSO^, but
not carbon, exposure. The combination produced
similar effects, but recovery occurred by 48 h
post-exposure. H^S04 + carbon resulted in more
tissue destruction than either pollutant alone.
Schiff et al., 1979
1 ppm
= 2.62 uig/m3
-------
3
requiring concentrations in excess of 100 mg/m to produce changes in glycoprotein secretion.
The DNA, RNA, and protein content of the lung increased in the group exposed to 0, and HLSCL
aerosols, while the 0-j-exposed group had only a small increase and the Hr,SCL group had none.
Grose et al, (1980) investigated the interaction of H^SO, and 03 on ciliary beat fre-
quency in the trachea of hamsters. A 2-h exposure to 0.88 mg/m H^SO* (0.23 |jm, VMD) signifi-
cantly depressed ciliary beat frequency. By 72 h after exposure, recovery had occurred.
2
Hamsters exposed to 0.196 mg/m (0.1 ppm) 0., for 3 h were not significantly affected; however,
when animals were exposed in sequence, first to 0., and then to H~SO,, ciliary beat frequency
was decreased significantly, but to a lesser extent than that caused by HjSO. alone. Analysis
showed that antagonism (p < 0.05) occurred in this sequential exposure.
Gardner et al. (1977a) found that the sequence of exposure to HpSO* aerosols and CL
altered the response of mice to airborne infections. Mice were exposed alone or in sequence
to 0.196 mg/m3 (0.1 ppm) 03 for 3 h and to 0.9 mg/m3 H2$04 aerosol (0.23 urn, VMD, a = 2.4) for
2 hours. When given alone, neither pollutant caused a statistically significant increase in
the mortality to a subsequent infection with S. pyogenes. When the pollutants were given
sequentially, a significant increase in mortality occurred only when 0, was given immediately
before exposure to HpSO,, and the response was additive. The reverse procedure had no effect
on mortality due to S. pyogenes infections. Because photochemical oxidants and SO often co-
— ^
exist in polluted air, these studies are of practical importance. The question of the
temporal sequence has been poorly investigated. Simple mechanisms to predict this additive
response sequence are not apparent. The results are opposite those of the Grose et al. (1980)
study described above with the tracheal model that showed that sequential exposure to 03 and
HpSCL had an antagonistic effect. The reasons for this difference are not known. The
infectivity model, however, is thought to reflect alveolar level effects (Gardner and Graham,
1977), whereas the'ciliary beat frequency model is a measure of effects at the level of the
trachea. In addition, different animal species were used. These findings also indicate the
complexity of interaction effects and the need to exercise care in extrapolating the effects
of pollutants from one parameter to another and one species to another (see Table 12-14).
12.5 CARCINOGENESIS AND MUTAGENESIS OF SULFUR COMPOUNDS AND ATMOSPHERIC PARTICLES
Attempts have been made for several decades to correlate various indices of particulate
air pollution with the development of cancer in man. In some cases, a positive association
has been found between increased community air pollution and cancer of the lungs and/or
gastrointestinal tract (see Chapter 14). This has led to suspicions concerning the chemical
nature of that portion or portions of airborne PM that may be contributing to an excess of
human cancer. As a result, at least three classes of potential etiologic agents have been
studied in animals: organic matter (including polycyclic hydrocarbons) that is adsorbed to
suspended particles; sulfur oxides; and trace metals.
12-66
-------
TABLE 12-14. EFFECTS OF INTERACTION OF SSL AND 0,
X j
Concentration
10 mg/m3 (1 urn, HMD) H2S04
aerosol, or 3.9 mg/m3 (2 ppm)
Oj, or combination of the two
Duration
6 h/day, 2 or 7
days
Species
Rat and
Guinea pig
Results
No synergism in effect on ratio of lung to body
weight. Histological lesions were those ascribed
to OA alone
Reference
Cavender et al.

., 1977
10 ing/m3 0 83 urn HMAD, 0. =
1.66) H2S04 aerosol, or9!.02
mg/m3 (0.52 ppm) 03, or com-
bination of the two

1 mg/m3 H2S04 aerosol and
0 78 to 0.98 fflg/m3 (OTT"to
0.5 ppm) Oa

0.196 mg/m3 (0.1 ppm) 03;
0.9 mg/m3 H2S04 aerosol (0 23
MID, VMD, a = 2.4)
y
exposed alone or in sequence

0.196 mg/m3 (0.1 ppm) 03;
0.88 mg/m3 H2S04 aerosol (0 23
Mm, VMD) exposed alone or in
sequence.
6 h/day, 5 day/wk,
6 mo
3 days,
continuous
3 h, 03;
2 h, H2S04
3 h, 03;
2 h, H2S04
Rat and Morphological alterations due to Oj alone
Guinea pig
Rat Synergistie effects. Glycoprotein synthesis was
stimulated in tracheal ring explants, lung ONA,
RNA, and protein content increased

House Significant increase in mortality in response to
airborne infections only when Og was given
immediately before exposure to H2S04; the
response was additive.

Hamster H2S04 depressed ciliary beat frequency,
Recovery had occurred by 72 h after exposure.
Sequential 03 then H2S04 exposure decreased
ciliary beat frequency significantly but to
a lesser extent than that caused by H2S04
alone. 03 exposure had no effect.
Cavender et al., 1978

Last and Cross, 1978

Gardner et al., 1977a

Grose et al., 1980
-------
Test systems for the bioassay of potential mutagens and carcinogens are diverse, ranging
from the measurement of chemically-induced reverse mutations in bacteria to the frank produc-
tion of carcinomas by administration to mammals. It is commonly believed, however, that fun-
damental similarities exist between the molecular mechanisms of both mutagenesis and carcino-
genesis. This assumption is based on the theory that chemical interaction with DNA and/or
other critical cellular macromolecules initiates a mutagenic or carcinogenic transformation.
Because of the assumed relationship between molecular events involved in mutagenesis and
carcinogenesis (Miller, 1978), the demonstration of mutagenic activity for a substance is
generally taken as strong presumptive evidence for the existence of carcinogenic activity.
Therefore, it i's believed that an investigation of the mutagenicity of a substance may be pre-
dictive of its carcinogenic potential, and may serve as an early warning of a possible threat
to human health in cases where positive results are obtained.
12.5.1 Airborne Particulate Matter
12.5.1.1 In Vitro Mutagenesis Assays of Particulate Matter—Organic material associated with
airborne particles has been investigated to some extent for its mutagenic potential. In these
studies, PM is limited experimentally to that which is retained by the filter medium used,
e.g., glass fiber, paper. ..etc. (Clark et al., 1980; Lee et al., 1980; Pitts et a!., 1978).
The particles that have created most interest are those with a carbonaceous core. These par-
ticles, because of their large surface area, adsorb many organic compounds, some of which are
known to be mutagenic and carcinogenic, e.g., benzo(a)pyrene (B(a)P). Because of the small
size (0.2-0.3 urn mean diameter) of many of the particles, they can be deposited in the pul-
monary regions of the lung (Verrant and Kittelson, 1977; Wolff et al., 1981b) where the
adsorbed organic material can desorb into the alveolar fluid and enter the associated tissue.
The ability of serum proteins to leach mutagens off particles has been demonstrated using
horse serum and coal fly ash (Crisp et al., 1978; Brooks et al., 1979; King et al., 1981;
Clark and Virgil, 1980; Wang and Wei, 1980).
A number of studies using the Ames Salmonella mutagenicity assay and other jji vitro tests
have been conducted with fractionated extracts of PM from urban air to obtain information on
the chemical nature of the mutagens present (Dehnen et al., 1977; Teranishi et al., 1978;
Moller and Alefheim, 1980; Tokiwa et al., 1980; Huisingh, 1981; Kolber et al., 1981; Ohnishi
et al., 1980). Estimates have been made as to the relative mutagenicity of each extract;
however, due to the possible interaction among the many compounds present in any fraction of
the extracts, the only conclusion that can be drawn is that both the moderately polar and
neutral fractions contain significant portions of the total mutagenic activity. The former is
the more potent direct-acting mutagen, while chemical derivatives of polycyclic aromatic
hydrocarbon (PAH) compounds are contained in the neutral fraction (Daisey, 1980; Daisey et
al., 1980; Kotin et al., 1955; Goff et al., 1980; Falk and Steiner, 1952; Henderson et al.,
1981). At present, the identity of the compounds that act as direct mutagens is uncertain.
12-68
-------
In a similar manner, various fractions were also extracted from particles emitted from a
coal powered electric plant (Crisp et al., 1978; Kubitschek et al., 1979; Clark and Hobbs,
1980; Fisher et al., 1979), gasoline engines (Wang et al., 1978), and light- and heavy-duty
diesel engines (Huisingh et al., 1977). The extracts obtained from these sources were direct-
acting frame-shift mutagens. Both the heavy- and light-duty diesel engine study was fraction-
ation carried out on the crude extract (Wang et al., 1978; Dukovich et al., 1981; Li et al.,
1980). A review of diesel engine PM is available (Santodonato, 1978).
Tne Salmonella assay has also been used in an attempt to define air quality by measuring
the mutagenic potential of total airborne particulates. Tokiwa et al. (1977) compared the
number of revertants per |jg of PM collected in the industrial area of Ohmata with that"
collected in the residential area of Fukuoka, Japan. In a similar manner, Pitts et al. (1977,
1978) compared eight urban samples in the California South Coast Basin with one collected in a
rural area of the San Bernadino mountains. In both cases, the mutagenic activity was less in
the residential and rural areas compared to the urban areas. Mutagenic potential was also
determined in a quantitative manner for a variety of air samples collected in Chicago
(Commoner et al., 1978).
Caution must be exercised when comparing in a quantitative manner results of Ames assays
on complex environmental mixtures. Indirect mutagenesis is extremely difficult to quantitate,
since microsomal oxidation to nonreactive as well as reactive compounds occurs (Kaden et al.,
1979; Skopek et al., 1979). Mixtures of direct and indirect mutagens may not produce an addi-
tive result. For any valid comparison there has to be nearly complete separation of these two
types of mutagens (Commoner et al., 1978). Also, the effects on mutagenesis of synergism and
antagonism among compounds in complex mixtures has not been investigated adequately. In the
case of complex mixtures obtained from tar sand, for example, the mutagenic activity of the
known mutagen, 2-aminoanthracene, was greatly inhibited by interaction with the mixture
(Shahin and Fournier, 1978). For these reasons, a quantitative assessment of air quality is
not readily obtainable with the use of the Ames Salmonella mutagenicity assay.
The data obtained with mammalian cell transformation assays (Rudd, 1979; Li, 1981; Liber
et al., 1979) generally support the conclusions derived from the Ames Salmonella assays
(Heidelberger, 1978; Sivak, 1979). There appear to be a variety of biologically active agents
present in the extracts of airborne PM, and these agents are of both a polar and nonpolar
nature. The identity of these compounds is unknown; however, the activity present is greater
than that which could be accounted for only by the PAH present in the samples. Since it is
presently unclear how the process of transformation in virus-infected cells relates to the
process of chemical carcinogenesis, cell transformation assays should be considered in the
same way as Ames assays, i.e., as only an indicator of the presence of biologically active
compounds.
12-69
-------
The dominant lethal assay of Epstein et al. (1972) is the only short-term i_n vivo assay
performed on airborne participate extracts. The water-soluble and benzene-soluble fractions
produced no fetal deaths or preimplantation losses beyond control limits. On the other hand,
the oxygenated fraction showed significant fetal deaths and decreased total implants.
12.5.1.2, Tumorigenesisof Participate Extracts—As early as the 1930s it was realized that
increasing amounts of air pollution may correlate with the increasing rate of human lung
cancer. Some of the earliest jji v ivo experiments dealt with the repeated exposure of mice to
clouds of soot, followed by autopsy examination for tumors at the end of their natural life-
span. A number of different kinds of soot have been chosen for these studies due to their
significant contribution to airborne PM. Upon bioassay of soot from chimneys (Campbell, 1939;
Seelig and Benignus, 1938), motor exhaust (Campbell, 1939), and airborne PH collected in the
vicinity of a factory and roadway (McDonald and Woodhouse, 1942), a slight increase over con-
trol levels in the number of lung tumors was observed. Only in the case where road dust from
a freshly tarred road was used were there significant increases, with 57 percent of the experi-
mental and 8 percent of the control group having lung tumors (Campbell, 1934). However, 5 yrs
later dust from the same road, which had not been retarred, was again tested and only 8 per-
cent of the experimental group and 1.4 percent of the control group developed lung tumors
(Campbell, 1942). Although these studies have all attempted to demonstrate the potential of
airborne PM to cause lung tumors, the results obtained are ambiguous due to the low tumor
incidence and the small size of the animal groups.
Among the various compounds associated with airborne particles, PAH (polycyclic aromatic
hydrocarbons) have received the most attention as potentially carcinogenic. PAHs were the
first compounds shown to be associated with carcinogenesis. Carcinogenic PAH are still dis-
tinguished by several unique features:
1. several compounds of this class are among the most potent animal carcinogens known
to exist, producing tumors by single exposures to microgram quantities.
2. they act both at the site of application and at organs distant from the site of
absorption.
3. their effects have been demonstrated in nearly every tissue and species tested, re-
gardless of the route of administration.
The most widely studied PAH, benzo(a)pyrene (B(a)P), is ubiquitous in the environment and pro-
duces tumors in animals that closely resemble human carcinomas.
The production of lung tumors in animals has been difficult. PAHs such as B(a)P are
mildly carcinogenic in the respiratory tract using intratracheal instillations. Certain in-
organic materials (iron ore, carbon, asbestos) have been found to potentiate this effect of
B(a)P in hamsters (Pylev and Shabad, ]973; Stenback et al., 1973, 1976; Saffioti et al., 1968)
and produce bronchogenic carcinomas with a morphology similar to human lung cancer. Organic
extracts of airborne particulates readily cause tumors when injected subcutaneously. Sarcomas
12-70
-------
have been produced in mice using the benzene extracts of PM collected from an urban area
(Leiter and Shear, 1942; Leiter et al. 194,2). The tumor incidence was low in these initial
studies, with only 8 percent of the mice developing tumors by the end of the study; however,
none of the control mice had sarcomas. In one later study, when particles were collected in
the vicinity of a petrochemical plant, the tumor incidence was as high as 61 percent (Rigdon
and Neal, 1971). Even in this case of, high tumor production,.no increase in the incidence of
tumors over the spontaneous rate was observed ip any organ of the animal distant to the site
of injection. Only when neonatal mice were injected subcutaneously with particulate extracts
did tumors appear distant from the injection site (Epstein et al., 1966), with a very high
incidence of hepatomas (83 percent) and multiple pulmonary adenomas (67 percent). Remote
tumor formation after subcutaneous injection of neonatal mice was confirmed with both the
crude extract of particles collected in New York City and subfractions of this extract; the
predominant tumors were again hepatomas (Asahina et al., 1972).
The carcinogenic nature of extracts of PM has also been demonstrated by painting the skin
on the backs of mice. With repeated application (three times per wk for the life of the
animal) of the benzene extract of particulates collected in the Los Angeles area, papillomas
were formed that subsequently progressed to carcinomas (Kotin et al., 1954). Papillomas first
appeared after 465 days, and at the time the data were presented, 42 percent of the mice had
developed tumors. Although papillomas and carcinomas of the skin were the most commonly
observed tumors, lung tumors have also been noted after skin application (Clemo et al., 1955).
Among the different methods of administering particulate extracts to the mouse for bioassay,
skin painting yields the highest tumor incidence, with greater than 90 percent of the surviv-
ing animals in some cases developing tumors.
In subsequent studies, the phenomenon of two-stage tumorigenesis was used to characterize
further the biological activity in airborne PM. In two-stage tumorigenesis, an initiator is
an agent (usually a carcinogen) that, when applied in a single dose to the skin of a mouse,
does not produce tumors at the applied concentration, but predisposes the skin so that later
repeated application of a promoter (an agent that by itself will not produce tumors) will
cause the formation of tumors. A complete carcinogen is one that, if applied in sufficient
concentration, can produce tumors by itself. Extracts of airborne particles from Detroit were
fractionated, and the fractions examined for complete carcinogenicity and tumor initiating and
promoting activity (Stern, 1968; Wynder and Hoffman, 1962). When applied to the skin of Swiss
ICR female mice, only the whole extract and the aromatic fraction proved to be a complete car-
cinogen, while the insoluble, acidic, aliphatic and oxygenated fractions produced no tumors
(there was insufficient basic fraction to perform the assay).
To examine the aromatic fraction for initiating activity, it was applied to the backs of
mice in a subtumorigenic dose followed by repeated application of the known promoter: croton
oil. Tumor initiating activity corresponded in a general way to the B(a)P content of the
12-71
-------
fraction. (The other fractions of the participate extract were not tested for initiating
activity.) It should be noted that an initiator does not necessarily have to be a complete
carcinogen, although most if not all complete carcinogens will initiate if applied at a low
dose where their complete carcinogenic action is not apparent. For this reason, it is
possible that some of these fractions could have initiated activity even though they did not
act as complete carcinogens when first tested. In addition, both the acidic and neutral frac-
tions exhibited tumor promoting activity in the presence of the known initiator: 7, 12-
*•
dimethylbenz(a)anthracene (OMBA). The relevance of two-stage skin carcinogenesis to environ-
mentally-caused cancer in other organs, however, is not known.
Several contributing sources of airborne PM (e.g., gasoline and diesel engines and the
soot from coal and oil burning furnaces) have been examined individually and shown to produce
tumors. Extracts of PM from gasoline engines show carcinogenic activity when painted on the
backs of mice (Brune, 1977; Wynder and Hoffman, 1962) and injected subcutaneously (Pott et
al., 1977). Extracts from diesel engines have shown tumorigenic activity in some studies but
not in others; the same holds true for extracts of chimney soot where activity was shown in
some instances (Campbell, 1939) while not in others (Wittier and Nicholson, 1957). The dis-
crepancies among these results could be due to qualitative and/or quantitative differences in
the nature of the organic compounds adsorbed to the particles or difference in assay systems.
Differences may have existed in the operating parameters of the generating source, or varia-
tions in PM collection procedures. For example, the mode of operation of diesel engines (the
load under which the engine was run), the type of fuel and the temperature at which the par-
ticles were collected all affect the biological activity of the sample. With soot collected
from chimneys, temperature is an important consideration. The organic material on PM is
generated in the gaseous phase, while condensation on nuclei occurs at lower temperatures.
Unless particles are collected under similar conditions, disparities will exist in their
chemical composition and biological activity.
Taken together, it is apparent that all the major types of airborne PM may contain
adsorbed compounds that are mutagenic and/or carcinogenic to animals and may contribute in
some degree to the incidence of human cancer associated with exposure to urban air pollution.
12.5,2 Potential Mutagenic Effects of Bisulfite and Sulfur Dioxide
Bisulfite addition to cytosine can result in deamination to form uracil (Shapiro., 1977;
Fishbein, 1976). The result would be conversion of guanine-cytosine in DNA to adenine-thymine
sites. Transamination of cytosine can also occur through reaction of an amine with cytosine-
bisulfite adduct. Since the nucleus is rich in polyamines, transamination would appear to be
a likely event; however, these reactions of cytosine occur most readily in high (1 M) concen-
trations of bisulfite. Because of the nature of the reactivity of bisulfite with cytosine,
the potential mutagenic properties of bisulfite and SCL have been examined. Such experiments
have been reviewed by Shapiro (1977) and Fishbein (1976). Microbial experiments with high con-
centrations of bisulfite in acid solutions i_n vitro have produced mutations (see Table L2-15).

12-72
-------
TABLE 12-15. POTENTIAL HUTAGENIC EFFECTS OF StyBISULFITE
Concentration

1310 mg/m3 S02
(SOO ppm)
13.1 - 105 mg/m3 S02
(5 - 40 ppm x 3 min)
14.9 mg/ma S02
(5.7 ppm)

Bisulfite
0.9 M HSOl
pH 5 0 3

3 M HSOl
pH 5-6 *

1 M HSOl
pH 5.2 J
5 x 10" 3 M HSO"
pH 3.6 3
0.04 or 0.08 M

Organism
Phage T4-R11 System

Phage T4-R11
System

E. coll K12 &
K15
S. cerevisiae

D. melanogaster

Hela cells
(Human)
Mouse fibroblasts &
Peritoneal aiacrophages
Human lymphocytes

End Point
GC+AT or
deami nation of
cysocine
deami nation of
cytosine

GC-»AT or
deamination of cytosine
Point Mutation

Point Mutation

Cytotoxicity

Point Mutation
Chromosomal aberrations
Cytotoxicity
Response Comments Reference
+ - Summers and
Drake, 1971

± Poor dose Hayatsu and Miura,
response 1970,
Iida et al., 1974
+ - Mukai et al . ,
1970
+ - Dorange and
Dupuy, 1972
Kay not be Valencia et al,,
bioavailable 1973
+ - Thompson and Pace,
1962
Nulsen et al., 1974

-
Kikigawa and
+ - Iizuka, 1972
1 ppm S02 = 2.62 mg/m3.
-------
These conditions would be similar to those favoring deamination of cytosine. On the other
hand, experiments conducted at low concentrations (> 10 M bisulfite) and neutral pH (7-7.4)
have not provided clear-cut evidence of mutagenesis. The microbial assays were not done with
strains of Salmonella known to be sensitive to mutagens (Ames assays). Background mutation
rates, mechanisms of error-prone repair, and corrections for cytotoxicity were not studied.
Negative experiments have been reported when insects (Drosophila) (Valencia et al., 1973) and
mammals (mice) were exposed. Cytotoxicity, rather than mutagenicity, appears when cultured
animal and human cells (Thompson and Pace, 1962; Nulsen et al. , 1974; Kikigawa and lizuka,
1972; Schneider and Calkins, 1971; Timson, 1973) are exposed to sulfite. (See Table 12-15 for
summary.)
12.5.3 Tumorigenesis in Animals Exposed to Sulfur Dioxide or Sulfur Dioxide and Benzo(a)pyrene
Tumorigenesis after exposure to SO- alone or to SOp and an aerosol of B(a)P has been
examined. For example, Peacock and Spence (1967) examined the lungs and other organs of mice
exposed over their lifetimes (300 days) in a 180-liter chamber into which 500 ppm S0~ was intro
duced at a rate of 20 ml/min for 5 min, 5 days/week. They found an increase in primary pul-
monary neoplasia in the males (n=35) from 31 percent in the control group to 54 percent in the
SCL-exposed group, and in the females (n=30) from 17 to 43 percent. Furthermore, while S02
did not affect the incidence of malignant tumors in males, the incidence of primary lung carci-
noma in females increased from 0 to 18 percent in the exposed group.
Unfortunately, the concentration of S0? used cannot be determined accurately from the
paper and thus, no concentration-related effects can be deduced. In addition, the statistical
analyses reported are vague and the significance of the observed increases in lung carcinoma
is, therefore, questionable. In view of these shortcomings, EPA undertook a reanalysis of the
data reported in Peacock and Spence (1967). This reanalysis, which is described in the memo
attached as an appendix to this chapter, used a one-sided Fisher's exact test to determine
that SO, did indeed increase the incidence of primary lung carcinoma in females (p=0.056), as
well as the incidence of primary lung adenomas in males (p=0.065) and females (p=0.011).
Despite these findings, it would still appear that Peacock and Spence's conclusion that this
particular study does not "justify the classification of SOp as a chemical carcinogen as
generally understood" remains valid.
Lung tumors or other significant pathological effects were not observed in hamsters ex-
posed for 98 wk to 26.2 mg/m3 (10 ppm) S09 plus 10 mg/m3 B(a)P for 6 h/day, 5 days/wk for
\ 3
534 exposure days, or to 9.17 mg/m (3.5 ppm) S02 plus 10 mg/m B(a)P for 1 h/day, 5
days/wk for 494 exposure days, or to a combination of the two regimens (Laskin et al., 1970).
When rats were exposed in the same fashion (Laskin et al., 1970), however, lung squamous cell
carcinoma was found in 23.8 percent of the animals exposed to the combination of the two
regimens described above and in 9.5 percent of animals exposed to 10 mg/m B(a)P plus SOg for
1 h/day. Renal metastasis also occurred.
12-74
-------
This study was subsequently extended to lifetime (exact time not specified) exposures (5
•3
days/wk) of rats (Laskin et a!., 1976). Exposure to air alone (n=15) or to 26.2 mg/m (10
ppm) SQ9 (n=15) for 6 h/day caused no squamous cell carcinoma. A 1-h/day exposure to 10 mg/m3
3
B(a)P caused cancer in-1 of 30 (3.3 percent) rats. A 6-h/day exposure to 26.2 mg/m (10 ppm)
S0? plus a 1-h/day exposure to 10 mg/m B(a)P resulted in a cancer incidence of 6.7 percent (2
3 ^
of 30). When animals received a combination of 10 mg/m B(a)P and 10.5 mg/m (4 ppm) S0p» 4
of 45 (8.9 percent) of the rats had cancer. The highest incidence (9 of 46, 19.6 percent) was
3 3
found in animals exposed for 6 h/day to 26.2 mg/m (10 ppm) S0? plus a combination of 10 mg/m
B(a)P and 10.48 mg/m3 (4 ppm) S02 for 1 h/day.
The biological significance of these studies (Laskin et a!., 1970, 1976) is complex and
difficult to interpret, particularly since statistical analyses were not reported. In an
attempt to clarify the matter, EPA undertook a reanalysis of the latter study of Laskin et al.
(1976). The EPA statistical analysis is described in the memo attached as an appendix to this
chapter. Using a multiple probit approach, it was determined that, while the cancer incidence
increases in response to exposure to S02 alone or B(a)P alone were not statistically signifi-
cant (p=0.116 and p=0.113, respectively), the increase due to the combination of the two was
indeed significant (p=0.005). However, given the lack of experimental details, especially
with the experimental design used, it is very difficult to come to a definitive conclusion
about the carcinogenicity of SO, and B(a)P as administered in this protocol.
12.5.4 Effects of Trace Metals Found in Atmospheric Particles
Among the numerous trace metals found in the atmosphere, systemic toxicity has been demon-
strated following inhalation of lead (Office of Research and Development, 1977), mercury
(Hammond and Beliles, 1980), arsenic (HAS, 1977), asbestos (NAS, 1971), and cadmium (Hammond
and Beliles, 1980). In addition, certain compounds of some of these (e.g., arsenic,
beryllium, cobalt, and nickel) have been found to produce tumorigenic effects under specific,
nonrespiratory laboratory exposure conditions (Furst and Hard, 1969; IARC, 1972, 1973, 1976;
Lau et al,5 1972; Stoner, et al., 1976; Sunderman, 1978, 1979). Limited evidence also points
to compounds of molybdenum and manganese as possible tumorigens (Clemo and Miller, 1960).
Although trace metals are ubiquitous in the environment, their levels are generally so
low that it is difficult to predict the magnitude of carcinogenic risk in community settings.
This problem is compounded by the fact that clear dose-response relationships have not been
well defined for most carcinogenic metals. At present, therefore, it is not possible to pre-
dict confidently the quantitative relationship of trace metals to the production of cancer due
to low-level exposures to particulate air pollution.
12.6 CONCLUSIONS
12.6.1 Sulfur Dioxide
Once inhaled, SO- appears to be converted to its hydrated forms, sulfurous acid, bisul-
fite, and sulfite. The rate of absorption and removal of inhaled S0? varies with species, but
it is at least 80 percent of the inhaled amount at relatively high concentrations.

12-75
-------
The metabolism of SOp is predominantly to sulfate and is mediated by the enzyme sulfite
oxidase. Since sulfite oxidase is a molybdenum-containing enzyme, dietary factors could in-
fluence the function of the enzyme in man. No conclusive evidence has yet been reported. The
reaction of bisulfite with serum proteins to form S-sulfonates is rapid. The S-sulfonates are
remarkably long-lived (t 1/2 = 4.1 days in rabbits), supplying a circulating pool of bisulfite
that can reach all tissues. Since some circulating S-sulfonates decompose to SO,, that is ex-
haled, S-sulfonates can donate their bisulfite content to distal tissues.
An immediate effect of acute (5 1 h) S0? inhalation is either a decrease in respiratory
rate or an increase in resistance to flow within the lung. The decrease in respiratory rate
depends on afferent conduction through the Vth or IXth cranial nerve following activation of
receptors in the nose and upper airways. Nasal air flow is decreased. The response is
transient in nature and occurs at 44,5 mg/m (17 ppm) S0?. Lower concentrations were not
tested.
The increased resistance to flow on inhalation of SQp is mediated through receptors in
the bronchial tree and persists during continued exposure. With this physiological parameter,
lower concentrations of SOp have been observed to cause reproducible changes in respiration.
The increased resistance to flow of air in the lung following SOp inhalation represents the
activation of an autonomic reflex arc through the vagus nerves. The same reflex arc occurs in
man. The reflex is cholinergic since atropine blocks the reflex, presumably at pregangliom'c
synapses. The guinea pig is the most sensitive animal in which to measure airway resistance,
with significant changes in pulmonary resistance to airflow occurring on the inhalation of con-
Q
centrations as low as 0.42 mg/m (0.16 ppm) SOp for 1 hour. Chronic exposures have produced
alterations in pulmonary function in cynomolgus monkeys, but only at concentrations greater
3 3
than 13.1 mg/m (5 ppm). Dogs exposed to 13.4 mg/m (5.1 ppm) S02 for 21 h/day for 225
days had increased pulmonary flow resistance and decreased compliance. Lower concentrations
were not examined. It should be remembered that SOp appears to cause its immediate broncho-
constrictive effect through action on airway smooth muscles, as evidenced by the antagonism of
the S0?-initiated bronchoconstriction by isoproterenol in man and animals. Since smooth
muscles adapt or fatigue during long-term stimulation, chronic exposure to SO, is not likely
to evidence bronchoconstriction equivalent to that occurring on short-term exposure. Altera-
tions in pulmonary function after chronic exposure to SOp are likely to occur through other
mechanisms, such as morphological changes in the airways or hypersecretion of mucus, which
will result in narrowing the airway. Concentration, rather than duration of exposure, seems
to be the most important parameter in determining responses to SOp, whether the response is
measured as a histopathological lesion or as a permanent alteration in respiration. There is
no theoretical hypothesis available at present to integrate the short-term effects observed
with 1-h exposures and the effects of long-term exposures of several months.
In rats, histopathological effects of S0? alone are confined to the bronchial epithelium,
with most of the effects occurring on the mucus secreting goblet cells. Goblet cell

12-76
-------
hypertrophy occurs on chronic exposure of rats, leading to the suggestion that SQ~ produces a
chronic bronchitis similar in many respects to that in man. Repeated exposure to a critical
2
concentration of SO- (not less than 131 mg/m or 50 ppm) may be needed to produce the chronic
bronchitis. While 50,,-produced chronic bronchitis in rats is similar to that in man and is a
useful model for the study of bronchitis, no evidence exists that chronic bronchitis is pro-
duced in man from ambient concentrations of S0?.
The nasal mucosa of mice (particularly those with upper respiratory pathogens) was
2
altered by a 72-h exposure to 26.2 mg/m (10 ppm) S00, Continous exposure to 0.37 to 3.35
3
mg/m (0.14 to 1.28 ppm) SQ~ for 78 wk did not cause any significant morphological altera-
tions in the lungs of monkeys. The effects of near-ambient concentrations of S0? on the
morphology and function of the nasal mucosa are not known.
Some pulmonary host defense mechanisms are also affected by S09 exposure. After 10 and 23
3
days of exposure (7 h/day, 5 days/wk) to 0.26 mg/m (0.1 ppm)), clearance of particles from
the lower respiratory tract was accelerated in rats. At a higher concentration of S00 (2.62
3
mg/m , 1 ppm) there was an initial acceleration (at 10 days), followed by a slowing at 25
days. A 5-day (1.5 h/day) exposure to 2.62 mg/m (1 ppm) reduced trachea! mucus flow in
dogs, but a longer exposure to this concentration caused no changes in ciliary beat frequency
of rats. These aberrations in trachea! clearance and mucus flow in several species are con-
sistent with the profound effects of higher concentrations of S09 on mucus glands in rats.
3
Antiviral defenses were altered by a 7-day continous exposure to 18.3 to 26.2 mg/m (7 to
10 ppm) SO,,, as evidenced by an increase in viral pneumonia. The combined exposure to S09 and
3
virus produced weight loss on exposure to concentrations as low as 9.43 mg/m (3.6 ppm) S0_.
3
Mice exposed to 13.1 mg/m (5 ppm) S0? for 3 h/day for 1 to 15 days or for 24 h/day for
1 to 3 mo did not have increased susceptibility to bacterial lung disease. A variety of
o
changes in the humoral immune response of mice exposed for up to 196 days to 5.24 mg/m (2
ppm) have been reported.
Sulfur dioxide and bisulfite have been reported to be mutagenic in microbial test systems
(E. coli and yeast systems) at nonphysiological (acid) pH, but the relevancy of inhaled S0« as
a mutagen is not clear. A mechanism for the mutagenicity of S0? could be the deamination of
cytosine at high concentrations. Free radical reactions breaking glycosidic bonds in ONA may
be responsible at low concentrations. The potency of bisulfite in these j_n vitro systems is
moderate to weak when compared to agents such as nitrosamines or polycyclic aromatic com-
pounds. To date, experiments testing for mutagenicity of bisulfite in mammals have been
equivocal. On the basis of present evidence, one can not decide whether or not bisulfite, and
hence S0?, is a mutagen in mammals.
The potential carcinogem'city of SO,-, alone and in combination with B(a)P has also been
examined. Unfortunately, the two key studies have not been replicated and both lacked experi-
mental details and proper statistical analyses. While subsequent reanalyses of the data by
12-77
-------
EPA revealed statistically significant increases in cancer, it is very difficult at present to
come to a definitive conclusion about these studies. However, SCL must remain suspect as a
carcinogen or cocarcinogen in view of these reanalyses and the positive results of mutageni-
city assays.
12.6.2 Particulate Matter
The chemical and physical diversity of the PM in the atmosphere presents a severe limita-
tion on the scope of the conclusions presented here. Most of the evidence for adverse health
effects of inhaled particles presented in this chapter relates to compounds arising from SO
(e.g., H-SCL, ammonium sulfate, metal sulfates, and related compounds), A brief treatment is
presented for heavy metals and their compounds. A summary of data related to organic material
associated with particles is also presented, focusing primarily on polycyclic organic com-
pounds. Due to limitations of space, details dealt with in other criteria documents or recent
major reviews on specific elements and the broader aspects of polycyclic organic compounds are
presented by reference only. This chapter, then, should not be taken as a summary of all of
the data available on the health effects of atmospheric particles, but rather as a selected
summary related mostly to SO . The reader should refer to the more detailed reports before
X
attempting to integrate the present limited material into the generic problem of the effects
of atmospheric particles. Similarly, the subsequent section on the interactions between SO
X
and PM relates to a limited scope of the atmospheric particles.
All inhalation studies of particles available for review in this document were conducted
with sizes that could be expected to fall within the approximate size range of the alveolar
envelope of deposition (<8 pm MMAD, depending upon the species exposed). A few jjn vitro or
intratracheal instillation studies have been performed that compared the effects of a wide
range of particles including those that occur in the atmospheric coarse mode particle sizes
(>2.5 urn Dafi).
Reports disagree as to the potency of acute exposure to sulfate aerosols. Some investi-
gators contend that H»S(K is highly irritating, producing increases in pulmonary flow resist-
ance at low concentrations. The increased resistance to airflow in the lung was directly pro-
portional to the sulfate aerosol concentration inhaled. The bronchoconstriction produced by
zinc ammonium sulfate was similar in many properties to that produced by histamine aerosols.
Unlike S02-initiated bronchoconstriction, intravenous atropine had no effect. Inhaled or in-
travenous isoproterenol, however, blocked the zinc ammonium sulfate-induced bronchoconstric-
tion. These data suggest that the zinc ammonium sulfate aerosol receptor and presumably other
sulfate receptors are not identical to the S02 receptor The two agents accordingly could act
at separate sites in the lung. Histamine is implicated in the sulfate aerosol action more
clearly than in the bronchoconstricting action of S0?.
The lowest effective concentration of H,SQ, producing bronchoconstriction so far reported
3
was 0.1 mg/m (1 h) in the guinea pig. Particle size influenced the results in several
ways, but the smaller sizes were generally more effective. Another study has observed an "all

12-78
-------
or none" response (increased airway resistance) in guinea pigs exposed for 1 h to 14.6,
3 3
24,3, or 48.3 mg/m KLSO.. Exposure to lower concentrations (1.2 or 1.3 mg/m H2SCL) caused no
effects. Some of these conflicts may be due to differences in technique or in age or strain
of guinea pig. Large interindividual differences in dose-response curves are characteristic
for inhaled histamine. In man, dogs, cats, and guinea pigs, 100-fold differences in the bron-
choconstrictive response to histamine have been observed. The dose-response curves for an
individual subject are remarkably reproducible. In dogs and guinea pigs, the bronchoconstric-
tive response to histamine fell within a single log-normal distribution, despite the large
interindividual differences in the dose of histamine required to elicit a specific response.
These large interindividual differences could represent differences in "susceptibility" of
different individuals, suggesting a small fraction of "susceptible" individuals; or they could
represent a very flat dose-response curve for a single population. Currently, the data favor
a single population hypothesis for histamine. The dose-response relationship for sulfate and
H?S04 aerosols is not adequate to distinguish between these two hypotheses. It is clear that
large interindividual differences in response to inhaled aerosols are a characteristic of the
biological response as measured by increased resistance to flow, regardless of the species
used, and are not an artifact of the exposure or measurement system.
Age may also play in important part in this response, since young guinea pigs are more
susceptible than older ones. For histamine sensitivity, age-dependence has been suggested as
an analog of juvenile asthma, but human airway sensitivity does not seem to follow the same
developmental pattern. Further research is needed to settle the question of special suscepti-
bility of young animals and children.
For the effects of 1-h exposures of guinea pigs to SO from one laboratory, an
A,
apparent ranking of potency (for increased flow resistence) is as follows: H?S04 >
ZnS04(NH4)2S04 > Fe2(S04>3 > ZnS04 > (NH4)2S04 > NH4HS04, CuS04 > FeS04> Na^, MnS04- The
latter three caused no effects.
The toxicology of H?SO, is complicated by its partial concentration-dependent conversion
to (NH.);>SO, and NH.HSO. by ammonia in the breath or in the air of exposure chambers, due to
excrement or exhalation. However, the actual concentrations of (NH4)pS04 and NH.HSO. in the
airways or chambers have not been measured definitively. Thus, comparing results of different
hLSQ. studies is confounded, particularly since some neutralization would even be expected in
the atmosphere of human exposure chambers (Kleinman et a!., 1981). One theory for the irritat-
ing action of sulfates contends that sulfate salts can act to promote release of histamine or
other mediators of bronchoconstriction and is supported by biochemical and pharmacological
evidence in two species. Anionic release of histamine may play a role in the bronchial con-
striction as evidenced by the blockade with H-l antihistamines. The effects of adrenergic
agonists and antagonists suggest the involvement of trachea! smooth muscle. Certainly, the
clearance of sulfurous acid, bisulfite, sulfite, and sulfate from the lung is influenced by
12-79
-------
the cations present in the aerosols inhaled simultaneously. Since polluted air is such a
complex mixture of these aerosols, the question of the toxicity of ambient aerosols can not be
approached on a simplistic basis by estimating toxicity from the acidity or sulfate content
alone.
Chronic exposure to H0SOA also produces changes in pulmonary function. Monkeys exposed to
3
0.48 mg/m H?SO. continously for 78 wk had altered distribution of ventilation early in the
3
exposure period. Higher concentrations (2.43 and 4.79 mg/m H?SO.) changed the distribution
of ventilation and increased respiratory rate, but caused no effects on other pulmonary
3
function measurements. A lower concentration (0.38 mg/m H..SO,,) caused no effects. Morpho-
3
logical changes occurred at the lowest concentration tested (0.38 mg/m HLSO.). The effects
appeared to be related to size of the particle as well as to concentration. Major findings at
3
2.43 mg/m H?SO. included bronchiolar epithelial hyperplasia and thickening of the respiratory
bronchioles and alveolar walls. Guinea pigs exposed continously for 52 wk to 0.08 or 0.1
mg/m H?SO, had no effects on pulmonary function or morphology. Dogs that inhaled 0.89 mg/m
H^SO, for 620 days (21 h/day) also had no morphological alterations; however, CO diffusing
capacity, residual volume, and net lung volume were decreased. Several other changes were
noted, including an increase in total expiratory resistance.
Sulfuric acid also alters mucociliary clearance, which is responsible for clearing the
lung of viable or inanimate particles. These particles affect the ciliated airways during
inhalation or reach this region as a result of alveolar clearance. A 1-h exposure of dogs
to 0.5 mg/m FLSO. increased trachea! mucocilary transport, whereas 1 mg/m H2SO, depressed
this rate. A 2- to 3-h exposure to 0.9 to 1 mg/m H0SO. also decreased trachea! ciliary
3
beat frequency in hamsters. Lower concentrations (0.1 mg/m H-SQ^, 1 h/day, 5 days/wk)
caused erratic bronchial mucociliary clearance rates in donkeys after several wk of exposure.
Continued exposure of the donkeys that had not received preexposures caused a persistent slow-
ing of bronchia! clearance after about 3 mo of exposure. From these and other studies, it
appears that repeated exposures to low concentrations of H_S04 can slow mucociliary clearance.
This might imply increased lung residence times of materials that would ordinarily be cleared.
Other host defense parameters (e.g., resistance to bacterial infection) are not altered
by low concentrations of FLSO., but are affected by metal sulfates. The apparent relative
potency of various particles for increasing susceptibility to infectious (bacterial) respira-
tory disease has been determined in mice exposed for 3 h: CdSO^ > CuS04 > ZnN03>
A12(SO,)_ > Zn(NH4)2(S04)2. At concentrations > 2.5 mg/m , the following particles had no
significant effects in this model system: H2$Q4, (NH4)2$04, NH4HS04, Na2S04, Fe2(S04)3,
Fe(NH4)2SQ4, NaN03, KN03> and NH4N03.
The chemical composition of the sulfate aerosols determines their relative toxicities.
For pulmonary irritants, the potency of a sulfate salt aerosol can be correlated with the
permeability of the lung to that specific sulfate salt. However, it is evident that accurate
estimates of the toxicity of complex aerosols occurring in urban air based solely on their

12-80
-------
sulfate contents are Inappropriate, since the metallic ions often associated with them may
also be toxic. Since urban air contains HpSO., ammonium sulfate, and metallic sulfates in
varying proportions, it is not possible to extrapolate accurately to man as he exists in a
complex environment from the currently inadequate toxicological data derived from studies of
single compounds in animals.
No data are available on the toxicity of secondary or complex atmospheric aerosols, since
only a very few published reports of animal studies have appeared. The problem is highly com-
plex because of the variability of aerosols from different urban localities and the composi-
tional changes on collection. Evaluations of toxicity can be approached, at present, only
from estimates of composition and toxicity of individual components. Using i_n vitro tests,
metal oxide-coated fly ash has measurable toxicity that can be ascribed to the insoluble
oxides when alveolar macrophages are exposed. The effects of soluble salts of Ni and Cd
have major differences. Nickel and Cd are removed from the lung with relative rapidity, but
may be stored or bound to intracellular proteins to an extent that is sufficient for accumu-
lation on repeated short-term exposures. Two-hour exposures to both Ni (0,5 mg/m ) and Cd
(0.1 mg/m ) aerosols impair the antibacterial defenses of the lung, leading to an increased
sensitivity to airborne pathogens in mice. Ciliary beat frequency in the trachea can be de-
creased by Cd and Ni also. Humoral immunosuppression in mice has been reported after a
2-h exposure to 0.19 mg/m CdC1o or °'25 m9/m NiClp.
It is apparent that all major types of airborne PM may contain adsorbed compounds that
are mutagenic and/or carcinogenic to animals. These may contribute, to some degree,,, to the
incidence of human cancer associated with exposure to urban air pollution.
12.6.3 Combinations of Gases and Particles
Although man is exposed to a complex mixture of gases and particles, few animal studies
have been conducted with mixtures. The dissolution of SCL into liquid aerosols or the sorp-
tion onto solid aerosols tends to increase the potency of SO^. The exact mechanism by which
potentiation occurs is still controversial. Sodium chloride and soluble salts (manganous
chloride, ferrous sulfate, or sodium orthovanadate) potentiated the effect (increased flow
resistance) of a 1-h SO, exposure of guinea pigs. Hypothetically, these particles favored the
conversion of S0? to H?SO,, thus increasing the response.
The effects of chronic exposure to a variety of mixtures of SQ?, HpSCL, and fly ash were
examined in guinea pigs and monkeys. None of these studies showed effects on pulmonary
function. Morphological changes were observed in monkeys after an 18-mo continuous exposure
o 3
to 2.6 mg/m (0.99 ppm) S0? plus 0.88 mg/m H-SO.; but the addition of fly ash did not
potentiate the response.
3 3
When dogs were exposed to SCL (13-4 mg/m , 5.1 ppm) and HLSCL (0.89 mg/m ) alone and in
combination for 21 h/day for 620 days, no morphological changes were observed. Sulfur dioxide
Jid not cause any significant changes in pulmonary function except for an increase in nitrogen
12-81
-------
washout, but HpSO, caused a variety of changes that were interpreted as the development of
obstructive pulmonary disease.
In another series of studies, dogs were exposed for 16 h/day for 68 mo to raw or photo-
chemically-reacted auto exhaust, oxides of sulfur or nitrogen, or their combinations. The
animals were examined periodically during exposure and at 32 to 36 mo after exposure. After
18 or 36 mo of exposure, no significant changes in pulmonary function were observed. After 61
o
mo, a few functional alterations were observed in dogs exposed to SO (1.1 mg/m (0.42 ppm)
3
SOp, and 0.09 mg/m H?S04) alone and in combination with other pollutants. The animals were
placed in clean air after exposure ceased, at which time the SO group had a variety of
pulmonary functional (24 mo postexposure) and morphological alterations (32 to 36 mo post-
exposure). These structural changes included a loss of cilia without squamous cell
metaplasia, nonciliated bronchiolar hyperplasia, and a loss of interalveolar septa in alveolar
ducts. The authors hypothesized that these changes are analogous to an incipient stage of
human proximal acinar (centrilobular) emphysema. Since the pulmonary function changes were
progressive during the postexposure period and they were correlated with the pathology, it can
be hypothesized that the morphological alterations were also progressive.
Combinations of carbon and H^SO, or SO, were investigated also. In mice exposed for 3
3 3
h/day, 5 days/wk for up to 20 wk to a mixture of 1.4 mg/m H?SO* and 1.5 mg/m carbon or to
carbon only, morphological and immunological alterations were seen. In hamsters, a 3-h
3 3
exposure to 1.1 mg/m H?S04 + 1.5 mg/m carbon depressed ciliary beat frequency, as did HpSO.
alone. Alterations of both the pulmonary and systemic immune systems were found in mice at
3 3
various lengths of exposure (100 h/wk up to 192 days) to 5.2 mg/m (2 ppm) S0? and 0.56 ing/m
carbon, alone or in combination. Generally, carbon and carbon + SO, caused more extensive
effects than SO, a\one.
When the interaction of 0, and H0SOn was studied, the morphological effects of a 6-mo
3 3
intermittent exposure of rats and guinea pigs to the mixture (10 mg/m hLSO^ + 1.02 mg/m
(0.52 ppm) 0.,) were attributed to 0,. alone. However, combined exposure to 1 mg/m H^SQ,, and
0.78 to 0.98 (0.4 to 0.5 ppm) 03 resulted in synergistic effects on glycoprotein synthesis in
the trachea and certain indices of lung biochemistry. Acute sequential exposure to first
•3 3
0.196 mg/m (0.1 ppm) 0- and then 0.9 mg/m H~S04 caused additive effects on increased
susceptibility to infectious pulmonary disease and antagonistic effects on depression of
trachea! ciliary beat frequency. From these studies, the interaction of 0., and H2$0^ appears
quite complex and dependent on the sequence of exposure as well as on the parameter examined.
12-82
-------
12.7 REFERENCES

Ada!Is, A, D.E. Gardner, F. J. Miller, and D. I. Coffin. Toxic effects of cadmium on ciliary
activity using a trachea! ring model system. Environ. Res. 13:111-120, 1977.

Adalis, A., D. E. Gardner, and F. J. Miller. Cytbtoxic effects of nickel on ciliated
epithelium. Am. Rev. Resp. Dis. 118:347-354, 1978.

Adkins, B., Jr., J. H. Richards, and D. E. Gardner. Enhancement of experimental respiratory
infection following nickel inhalation. Environ. Res. 20:33-42, 1979.

Adkins, B,, Jr., G. H. Luginbuhl, and D. E. Gardner. Biochemical changes in pulmonary cells
following manganese oxide inhalation. J. Toxicol. Environ. Health 6:445-454, 1980a.

Adkins, B., Jr., G. H. Luginbuhl, and D. E. Gardner. Acute exposure of laboratory mice to
manganese oxide. J. Am. Ind. Hyg. Assoc. 41:494-500, 1980b.

Adkins, B. , Jr., G. H. Luginbuhl, F. J. Miller, and D. E. Gardner. Increased pulmonary
susceptibility to streptococcal infection following inhalation of manganese oxide.
Environ Res. 23:110-120, 1980c.

Alarie, Y. Sensory irritation by airborne chemicals. CRC Crit. Rev. Toxicol. 2:299-363,
1973.

Alarie, Y., C. E. Ulrica, W. M. Busey, H. E. Swann, Jr., and H. N. MacFarland. Long-term con-
tinuous exposure of guinea pigs to sulfur dioxide. Arch. Environ. Health 21:769-777,
1970.

Alarie, Y. , C. E. Ulrich, W. M. Busey, A. A. Krumm, and H. N. MacFarland. Long-term
continuous exposure to sulfur dioxide in cynomolgus monkeys. Arch. Environ. Health
24:115-128, 1972.

Alarie, Y. , W. M. Busey, A, A. Krumm, and C. E. Ulrich. Long-term continuous exposure to sul-
furic acid mist in cynomolgus monkeys and guinea pigs. Arch. Environ. Health 27:16-24,
1973a.

Alarie, Y. , R. J. Kantz II, C. E. Ulrich, A. A. Krumm, and W. M. Busey. Long-term continuous
exposure to sulfur dioxide and fly ash mixtures in cynomolgus monkeys and guinea pigs.
Arch. Environ. Health 27:251-253, 1973b.

Alarie, Y. , C. E. Ulrich, W. M. Busey, A. A. Krumm, and H. N. MacFarland. Long-term
Continuous Exposure to Sulfur Dioxide in Cynomolgus Monkeys. In: Air Pollution and the
Politics of Control. MSS Information Corporation, New York, 1973c. pp. 47-60.

Alarie, Y. , I. Wakisaka, and S. Oka. Sensory irritation by sulfur dioxide and chloroben
zilidene malononitrile. Environ. Physiol. Biochem. 3^:53-64, 1973d.

Alarie, Y. C., A. A. Krumm, W. M. Busey, C. E. Ulrich, and R. J. Kantz. Long-term exposure to
sulfur dioxide, sulfuric acid mist, fly ash, and their mixtures. Results of studies in
monkeys and guinea pigs. Arch. Environ. Health 30:254-262, 1975.

Allison, A. C. Experimental methods - cell and tissue culture: effects of asbestos particles
on macrophages, mesothelial cells and fibroblasts. In: Biological Effects of Asbestos.
P. J. Bogorshi, V. Timbrel 1, J. C. Gil son, and J. C. Wagner, eds., IARC Scientific Publi-
cations No. 8, International Agency for Research on Cancer, Lyon, France. 1973, pp. 89-
92.
12-83
-------
Allison, A. C. , and D. M. L. Morgan. Effects of silica, asbestos, and other particles on
macrophage and neutrophil lysosomes. In; Lysosomes in Biology and Pathology, vol. 6,
J. T. Dingle, P. 0. Jaques, and I. H, Shaw, eds., North Holland, New York, NY, 1979. pp.
149-159.

Amdur, M. 0, Effect of a combination of SO, and H9SOA on guinea pigs. Public Health. Rep.
69:503-506, 1954. * * 4

Amdur, M. 0. The respiratory response of guinea pigs to sulfuric acid mist. Arch. Ind. Health
18:407-414, 1958.

Amdur, M. 0. The physiological response of guinea pigs to atmospheric pollutants. Int. J.
Air Pollut.-1:170-183, 1959.

Amdur, M. 0. The effect of aerosols on the response to irritant gases. In: Inhaled Particles
and Vapors. C. N. Davies, ed. , Pergamon Press, Oxford, England, 1961. pp. 281-294.

Amdur, H. 0. The effect of high flow-resistance on the response of guinea pigs to irritants.
Am. Ind. Hyg. Assoc. J. 25:564-568, 1964.

Amdur, M. 0. Respiratory absorption data and SO- dose-response curves. Arch. Environ. Health
12:729-732, 1966. ^

Amdur, M. 0. Toxicologic appraisal of particulate matter, oxides of sulfur, and sulfuric acid.
J. Air Pollut. Control Assoc. 19:638-646, 1969.

Amdur, M. 0, Aerosols formed by oxidation of sulfur dioxide. Review of their toxicology.
Arch. Environ. Health 23:459-468, 1971.

Amdur, M. 0. Animal studies. In: Proceedings of the Conference on Health Effects of Air
Pollutants, Washington, D.C. October 3-5, 1973. A Report prepared for the Committee on
Public Works, United States Senate. Serial No. 93-15, U.S. Government Printing Office,
Washington, DC, 1973. pp. 175-205.

Amdur, M. 0. 1974 Cummings Memorial Lecture. The long road from Donora. Am. Ind. Hyg.
Assoc. J. 35:589-597, 1974.

Amdur, M. 0. Toxicological Guidelines for Research on Sulfur Oxides and Particulates. In:
Proceedings of the 4th Symposium on Statistics and the Environment, Washington, DC,
March 3-5, 1976. American Statistical Association, Washington, DC, 1977. 48-55.

Amdur, M. 0., and M. Corn. The irritant potency of zinc ammonium sulfate of different particle
sizes. Am. Ind. Hyg. Assoc. J. 24:326-333, 1963.

Amdur, M. 0., and J. Mead. A method for studying the mechanical properties of the lungs of
unanesthetized animals. In: Proceedings of the 3rd NatioRal Air Pollution Symposium.
National Air Pollution Symposium, Pasadena, CA. April, 1955. pp. 150-159.

Amdur, M. 0., and J, Mead. Mechanics of respiration in unanesthetized guinea pigs. Am. J.
Physio!. 192:364-368, 1958.

Amdur, M. 0., and D. Underhill. The effect of various aerosols on the response of guinea pigs
to sulfur dioxide. Arch. Environ. Health 16:460-468, 1968.

Amdur, M. 0., and D. W. Underhill. Response of guinea pigs to a combination of sulfur dioxide
and open hearth dust. J. Air Pollut. Control Assoc. 20:31-34, 1970.

Amdur, M. 0,, R. Z. Schulz, and P. Drinker. Toxicity of sulfuric acid mist to guinea pigs.
AMA Arch. Ind. Hyg. Occup. Med. 5:318-329, 1952.

12-84
-------
Amdur, M. 0., J. Bayles, V. Ugro, and D. W. Underbill. Comparative irritant potency of sulfate
salts. Environ, Res. 16:1-8, 1978a.

Amdur, M. 0., M. Oubriel, and D. A. Creasia. Respiratory response of guinea pigs to low levels
of sulfuric acid. Environ. Res. 15:418-423, 1978b.

Amdur, M. 0., V. Ugro., and D. W. Underbill, Respiratory response of guinea pigs to ozone alone
and with sulfur dioxide. Am. Ind. Hyg. Assoc. J. 39:958-961, 1978c.

American Conference of Governmental Industrial Hygenists. TLVs: Threshold Limit Values for
Chemical Substances in Workroom Air Adopted by ACGIH for 1979. ACGIH, Cincinnati, OH,
1979,

Aranyi, C. , F. J. Miller, S. Anders, R. Ehrlich, J. Fenters, D. E. Gardner, and M. D. Waters.
Gytotoxicity of alveolar macrophages of trace metals adsorbed on fly ash. Environ. Res.
20:14-23, 1979.

Arrigoni, 0. The enzymatic oxidation of sulphite in mitochondria! preparations of pea inter-
nodes. Ital. J. Biochem. 7:181-186, 1959.

Asada, K. , and K. Kiso. Initiation of aerobic oxidation of sulfite by illuminated spinach
chloroplasts. Eur. J. Biochem. ^3:253-257, 1973.

Asahina, S., J. Andrea, A. Carmel, E. Arnold, Y. Bishop, S. Joshi, D. 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,

Backstrom, H. L. J. The chain-reaction theory of negative catalysis. J. Am. Chem. Soc.
49:1460-1471, 1927.

Balchum, 0. J. , J. Dybicki, and G. R. Meneely. The dynamics of sulfur dioxide inhalation,
absorption, distribution, and retention. Arch. Ind. Health 21:564-569, 1960,

Barry, D. H., and L. E. Mawdesley-Thomas. Effect of sulphur dioxide on the enzyme activity of
the alveolar macrophage of rats. Thorax 25:612-614, 1970.

Bingnam, E. , E. A. Pfitzer, W. Barkley, and E. P. Radford. Alveolar macrophages: Reduced
number in rats after prolonged inhalation of lead sesquioxide. Science 162:1297-1299,
1968.

Bingham, E. , W. Barkley, M. Zerwas, K. Stemmer, and P. Taylor. Responses of alveolar macro-
phages to metals. I. Inhalation of lead and nickel. Arch. Environ. Health 25:406-414,
1972.

Boushey, H. A., M. J. Holtyman, J. R. Sheller, and J. A. Nadel. Bronchial hyperreactivity.
Am. Rev. Resp. Dis. 121:389-413, 1980.

Breuninger, H. Uber das physikalisch - chemische Verhalten des Nasenschleims. [Physical-
chemical behavior of nasal mucus.] Arch. Ohren Nasen Kehlkopfheilkd. 184:133-138, 1964.

Brink, C. , P. G. Duncan, M. Midzenski, and J. S. Douglas. Response and sensitivity of female
guinea pig respiratory tissues to agonists during ontogenesis. J. Pharmacol. Exp. Ther.
215:426-433, 1980.

Brooks, A. L. , R. K. Wolff, R. E. Royer, C. R. Clark, A. Sanchez, and R. 0. McClellan.
Biological availability and mutagic chemicals associated with diesel exhaust particles.
EPA International Symposium on the Health Effects Diesel Engine Emission, Cincinnati,
December, 1979. pp. 345-357.

12-85
-------
Brune, H. F. K. Experimental results with percutaneous applications of automobile exhaust
condensates in mice. Air Pollution and Cancer in Man. IARC Scientific Publications No.
16, 1977. pp. 41-48.

Camner, P., M. Lundborg, and P. Hellstrom. Alveolar macrophages and 5 urn particles coated
with different metals. Arch. Environ. Health 29:211-213, 1974.

Campbell, J. A. Cancer of skin and increase in incidence of primary tumours of lung in mice
exposed to dust obtained from tarred roads. Br. J. Exp. Pathol. 15:287-294, 1934.

Campbell, J. A. Carcinogenic agents present in the atmosphere and incidence of primary lung
tumours in mice. Br. J. Exp. Pathol. 20:122, 1939.

Campbell, J. A. Lung tumours in mice. Incidence as affected by inhalation of certain carcino-
genic agents and some dusts. Br. Med. J. 1:217-221, 1942.

Cavender, F. L. Effects in rats and guinea pigs of six-month exposures to sulfuric acid mist,
ozone, and their combination. J. Toxicol. Environ. Health 4:845-852, 1978.

Cavender, F. L. , W. H. Steinhagen, C. E. Ulrich, W. M. Busey, B. Y. Cockrell, J. K. Haseman,
M. D. Hogan, and R. T. Drew. Effects in rats and guinea pigs of short-term exposures to
sulfuric acid mist, ozone, and their combination. J. Toxicol. Environ. Health 21:521-533,
1977.

Charles, J. M^ A Mechanism for Inhaled Sulfate Initiated Bronchoconstriction. Ph.D.
Thesis, Duke University, Durham, NC, 1976.

Charles, J. M. , and D. B. Menzel. Ammonium and sulfate ion release of histamine from lung
fragments. Arch. Environ. Health 30:314-316, 1975a.

Charles, J. M. , and D. B. Menzel. Sulfate removal from the airways and histamine release in
the isolated perfused rat lung. Pharmacologist 11:213, 1975b.

Charles, J. M. , W. G. Anderson, and D. B. Menzel. Sulfate absorption from the airways of the
isolated perfused rat lung. Toxicol. Appl. Pharmacol. 41:91-99, 1977a.
<
Charles, J. M. , D. E. Gardner, D. L. Coffin, and D. B. Menzel. Augmentation of sulfate ion
absorption from the rat lung by heavy metals. Toxicol. Appl. Pharmacol. 42:531-538,
1977b.

Clark, C. R. , and Hobbs, C. H. Mutagenicity of effluents from an experimental fluidized bed
coal combustor. Environ. Mutagen. 2:101-109, 1980.

Clark, C. R., and Vigil, C. L. Influence of rat lung and liver homogenates on the
mutagenicity of diesel exhaust particulate extracts. Toxicol. Appl. Pharmacol. 56:110-
115, 1980.

Clemo, G. R. , and E. W. Miller. Tumour promotion by the neutral fraction of cigarette smoke.
Br. J. Cancer 14:651-656, 1960.

Clemo, G. R. , E. W. Miller, and F. C. Pybus. The carcinogenic action of city smoke. Br. J.
Cancer 9:137-141, 1955.

Cockrell, B. Y., and W. M. Busey. Respiratory tract lesions in guinea pigs exposed to sulfuric
acid mist. J. Toxicol. Environ. Health 4:835-844, 1978.

Cohen, H. J. , and I. Fridovich. Hepatic sulfite oxidase. Purification and properties. J.
Biol. Chem. 246:359-366, 1971a.

12-86
-------
Cohen, H. J., and I. Fridovich. Hepatic sulfite oxidase. The nature and function of the heme
prosthetic groups. J. Biol. Chem. 246:367-373, 1971b.

Cohen, H. J., S. Betcher-Lange, D. L. Kessler, and K, V. Rajagopalan. Hepatic sulfite oxidase.
Congruency in mitochondria of prosthetic groups and activity. J. Biol. Chem.
247:7759-7766, 1972.

Cohen, H. J. , R. T. Drew, J. L. Johnson, and K. V. Rajagopalan. Molecular basis of the bio-
logical function of molybdenum. The relationship between sulfite oxidase and the acute
toxicity of bisulfite and S02- Proc. Natl. Acad. Sci. U.S.A. 70:3655-3659, 1973.

Cohen, H. J. , I. Fridovich, and K. V. Rajagopalan. Hepatic sulfite oxidase. A functional
role for molybdenum. J. Biol. Chem. 246:374-382, 1974.

Committee on Biologic Effects of Atmospheric Pollutants. Lead. National Academy of Sciences,
Washington, DC, 1972,

Committee on Biologic Effects of Atmospheric Pollutants. Vanadium, National Academy of
Sciences, Washington, DC, 1974,

Committee on Biologic Effects of Atmospheric Pollutants. Chromium. National Academy of
Sciences, Washington, DC, 1974.

Committee on Medical and Biologic Effects of Environmental Pollutants. Nickel. National
Academy of Sciences, Washington, DC, 1975.

Committee on Medical and Biologic Effects of Environmental Pollutants. Arsenic. National
Academy of Sciences, Washington, DC, 1977.

Commoner, B. , P. Madyastha, A. Bronsdon, and A. J. Vithayathil. Environmental mutagens in
urban air particules. J. Toxicol. Environ. Health 4:59-77, 1978.

Corn, M. , N. Kotsko, D. Stanton, W. Bell, and A. P, Thomas. Response of rats to inhaled
mixture of S00 and S00 - NaCl aerosol in air. Arch, Environ. Health. 24:248-256, 1972.
tL L-
Costa, D. L. , and M, 0. Amdur. Effect of oil mists on the irritancy of sulfur dioxide. I.
Mineral oils and light lubricating oil. Am. Ind. Hyg. Assoc. J. 40:680-685, 1979a.

Costa, D. L. , and M. 0. Amdur. Effect of oil mists on the irritancy of sulfur dioxide. II.
Motor oil. Am. Ind. Hyg. Assoc. J. 40:809-815, 1979b.

Crisp, C. E. , G, L. Fisher, and J, E. Lammert. Mutagenicity of filtrates from respirable
coal fly ash. Science 199:73-75, 1978.

Daisey, J. M. Organic compounds in urban aerosols. In: Aerosols: Anthropogenic and Natural
Sources and Transport. T. J, Kneip and P. J. Luoy, eds. , Ann. N.Y. Acad. Sci. 338:50-
69, 1980.

Daisey, J. M. , T. J. Kneip, I. Hawryluk and F. Mukai. Seasonal variations in the bacterial
mutagenicity of airborne particulate organic matter in New York City. Environ. Sci.
Techno!. 14:1487-1490, 1980.

Dehnen, W. , N. Pitz, and R. Tomingas. The mutagenicity of airborne particulate pollutants.
Cancer Lett. 4:5-12, 1977.

Dorange, J. L., and P, Dupuy. Mise en evidence d'une action mutagene du sulfite de sodium sur
la levure. [Evidence of mutogenic action of sodium sulfite on yeast.] C. R. Seances
Acad. Sci., Ser. D. 274:2798-2800, 1972.

12-87
-------
Douglas, J. S. , M. W. Dennis, P. Ridgway, and A. Bouhuys. Airway constriction in guinea pigs.
Interaction of histamine and autonomic drugs. J. Phamacol. Exp. Ther. 184:169-179, 1973.

Douglas, J. S. , P. Ridgway, and C. Brink. Airway responses of the guinea pig j_n vivo and j_n
vitro. J. Pharmacol. Exp. Ther. 202:116-124, 1977.

Drazen, J. M. Physiologic basis and interpretation of common indices of respiratory mechanical
function. Environ. Health Perspect. 16:11-16, 1976.

Dukovich, M., R. E. Yasbin, S. S. Lestz, T. H. Risby, and R. B. Zweidinger. The mutagenic and
SOS-inducting potential of the soluble organic fraction collected from diesel particulate
emissions. Environ. Mutagen. 3:253-264, 1981.

Duran, M. ; J. Korteland; F. A. Beemer; C. van der Heiden; P. K. de Bree, M. Brink; and S. K.
Wadman. Variability of sulfituria: combined deficiency of sulfite oxidase and xanbhine
oxidase. In: International Symposium on Models for the Study of Inborn Errors of
Metabolism. F. A. Hommes (ed), Elsevier/North-Holland Biomedical Press, Amsterdam, The
Netherlands, 1979. p. 103.

Ehrlich, R. Interaction between environmental pollutants and respiratory infections. In:
Proceedings of the Symposium on Experimental Models for Pulmonary Research. D. E.
Gardner, E. P. C. Hu, and J. A. Graham, eds., EPA-600/9-79-022, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1979. pp. 145-163.

Ehrlich, R. , J. C. Findlay, and D. E. Gardner. Susceptibility to bacterial pneumonia in
animals exposed to sulfates. Toxicol. Lett. 1:325-330, 1978.

Environmental Criteria and Assessment Office. Health Assessment Document for Polycyclic
Organic Matter. External Review Draft No. 1, U.S. Environmental Protection Agency,
Office of Research and Development, Research Triangle Park, NC, May 1978.

Environmental Criteria and Assessment Office. Health Assessment Document for Cadmium.
Preprint. EPA-600/8-79-003, U.S. Environmental Protection Agency, Research Triangle
Park, NC, January 1979.

Epstein, S. S. , S. Joshi, J. Andrea, N. Mantel, E. Sawicki, T. Stanley, and E. C. Tabor.
Carcinogenicity of organic particulate pollutants in urban air after administration of
trace quantities to neonatal mice. Nature 2JL2:1305-1307, 1966.

Epstein, S. S. , E. Arnold, J. Andrea, W. Bass, and Y. Bishop. Detection of chemical mutagens
by the dominant lethal assay in the mouse. Toxicol. Appl. Pharmacol. 2:288-325, L972.

Exon, J. H. , N. M. Patton, and L. D. Roller. Hexamitiasis in cadmium-exposed mice. Arch.
Environ. Health 30:463-464, 1975.

Fairchild, G. A., J. Roan, and J. McCarroll. Atmospheric pollutants and the pathogenesis of
viral respiratory infection. Arch. Environ. Health 25:174-182, 1972.

Fairchild, G. A., P. Kane, B. Adams, and D. Coffin. Sulfuric acid and streptococci clearance
from respiratory tracts of mice. Arch. Environ. Health 30:538-545, 1975a.

Fairchild, G. A., S. Stultz, and D. C. Coffin. Sulfuric acid effect on the deposition of
radioactive aerosol in the respiratory tract of guinea pigs. Am. Indust. Hyg. Assoc. J.
36:584-594, 19755.

Falk, H. L. and P. E. Steiner. The identification of aromatic polycyclic hydrocarbons in
carbon black. Cancer Res. 12:30-39, 1952.
12-88
-------
Fenters, J. D, , J, N. Bradof, C. Aranyi, K. Ketels, R. Ehrlich, and D. E. Gardner. Health
effects of long-term inhalation of sulfuric acid mist - carbon particle mixtures.
Environ. Res. 19:244-257, 1979.

Ferin, J. , and L. J. Leach. The effect of S0~ on lung clearance of TiO« particles in rats.
Am. Ind. Hyg. Assoc. J. 34:260-263, 1973. ^
-------
Gilbert, E. E. Sulfonation and Related Reactions. Wiley Interscience, New York, NY, 1965.
p. 125.

Gillespie, J. R. Review of the cardiovascular and pulmonary function studies on beagles
exposed for 68 months to auto exhaust and other air pollutants. (In: Long-term Effects
of Air Pollutants in Canine Species.) J. F. Stara, D. L. Dungworth, J. C. Orthoefer, and
W. S. Tyler eds. EPA #600/8-80-014, U.S. Environmental Protection Agency, Cincinnati,
Ohio, 1980, pp. 115-154.

Goff, E. U. , J. R. Coombs, D. H. Fine, and T. M. Baines. Determination of n-nitrosamines from
diesel engine crankcase emissions. Anal. Chem. 52:1833, 1980.

Goldring, I. P., L. Greenburg, S. S. Park, and I. M. Ratner. Pulmonary effects of sulfur
dioxide exposure in the Syrian hamster. II. Combined with emphysema. Arch. Environmental
Health 21:32-37, 1970.

Goldstein, B. , and I. Webster. Intratracheal injection into rats of size-graded silica
particles. Br. J. Ind. Med. 23:71-74, 1966.

Graham, J. A., D. E. Gardner, F. J. Miller, M. J. Daniels, and D. L. Coffin. Effect of nickel
chloride on primary antibody production in the spleen. Environ. Health Perspect.
12:109-113, 1975a.

Graham, J. A., D. E. Gardner, M. D. Waters, and D. C. Coffin. Effect of trace metals on
phagocytosis by alveolar macrophages. Infect. Immun. 11:1278-1283, 1975b.

Graham, J. A., F. J. Miller, M. J. Daniels, E. A. Payne, and D. E. Gardner. Influence of
cadmium, nickel, and chromium on primary immunity in mice. Environ. Res. 16:77-87, 1978.

Grant, M. M., S. P. Sorokin, and J. D. Brain. Lysosomal enzyme activities in pulmonary
macrophages from rabbits breathing iron oxide. Am. Rev. Resp. Dis. 120:1003-1012, 1979.

Green, G. M. The J. Burns Amberson Lecture - In defense of the lung. Am. Rev. Resp. Dis.
102:691-703, 1970.

Grose, E. C., D. E.« Gardner, and F. J. Miller. Response of ciliated epithelium to ozone and
sulfuric acid. Environ. Res. 22:377-385, 1980.

Grunstein, M. M. , M. Hazucha, J. Sorli, and J. Milic-Emili. Effect of SOg on control of
breathing in anesthetized cats. J. Appl. Physio!, Resp. Environ. Exercise Physio!.
43:844-851, 1977.

Gunnison, A. F., and A. W. Benton. Sulfur dioxide: Sulfite. Interaction with mammalian
serum and plasma. Arch. Environ. Health 22:381-388, 1971.

Gunnison, A. F., and E. D. Palmes. Persistence of plasma S-sulfonates following exposure of
rabbits to sulfite and sulfur dioxide. Toxicol. Appl. Pharmacol. 24:266-278, 1973.

Habib, H. P., P. D. Pare, and L. A. Engel. Variability of airway responses to inhaled
nistamine in normal subjects. J. App!. Physio!.: Resp. Environ. Exercise Physiol.
47:51-58, 1979.

Hackney, J. D. Effects of sulfate aerosols upon cardiovascular function in squirrel monkeys.
Final Report. APRAC Project CAPM-2Q-74, Coordinating Research Council, Inc., New York,
Dec. 1, 1978.
12-90
-------
Hadley, J. 6., D. E. Gardner, D. L. Coffin, and D. B. Menzel. Inhibition of antibody mediated
rosette formation by alveolar macrophages: A sensitive assay for metal toxicity. RES J.
Reticuloendothel. Soc. 22:417-425, 1977.

Hammond, P. B. , and R. B. Beliles. Metals. In: Toxicology: The basic science of
poisons. Second edition, J. Doull, C. D. Kalassen and M. 0. Amdur, (eds.). MacMillan
Publishing Co., Inc. New York, NY, 1980. pp. 409-462,

Hansen, L. 0., D. J. Eatough, L. Whiting, C. H. Bartholomew, C. L. Cluff, R. M. Izatt, and J.
J. Christensen. Transition metal S032 complexes: A postulated mechanism for the
synergistic effects of aerosols and S02 on the respiratory track. In: Trace Substances
in Environmental Health Vol. VIII U, Montana Press, 1974. pp. 393-397.

Harkness, D. R., and S. Roth. Purification and properties of 2,3-diphosphoglyceric acid
phosphatase from human erythrocytes. Biochem. Biophys. Res. Commun. 34:849-856, 1969.

Hatch, G. E. , D. E. Gardner, and D. B. Menzel. Stimulation of oxidant production in alveolar
macrophages by pollutants and latex particles. Environ. Res. 23:121-136, 1980.

Hayatsu, H. Bisulfite modification of nucleic acids and their constituents. Prog. Nucl. Acid
Res. Mol. Biol. 16:75-124, 1976.

Hayatsu, H., and A. Miura. The mutagenic action of sodium bisulfite. Biochem. Biophys. Res.
Commun. 39:156-160, 1970.

Hayon, E. , A. Treim'n, and J. Wilf. Electronic spectra, photochemistry, and autoxidation
mechanism of the suitite-bisulfite-pyrosulfite systems. The S0?, SCL, SO,, and SOr
radicals. J. Am. Chem. Soc. 94:47-57, 1972.

Heidelberger, C. Oncogenic transformation of cell cultures by polycyclic aromatic
hydrocarbons and their derivatives. In: Polycyclic Aromatic Hydrocarbons and Cancer,
Vol. 2. H. V. Gilboin and P. 0. P. T'So, eds. Academic Press, New York, NY 1978.

Hemeon, W. C. L. The estimation of health hazards from air pollution. AMA Arch. Ind. Health
11:397-402, 1955.

Henderson, T. R. , A. P. Li, R. E. Royer, and C. R. Clark. Increased cytotoxicity and
mutagenicity of diesel fuel, after reaction with NO,,. Environ. Mutagen. 3:211-220, 1981.

Heppleston, A. G. The disposal of dust in the lungs of silicotic rats. Am. J. Path.
40:493-506, 1962.

Hirsch, J. A. , E. W. Swenson, and A. Wanner. Tracheal mucous transport in beagles after
long-term exposure to 1 ppm sulfur dioxide. Arch. Environ. Health 3^0:249-253, 1975,

Holma, B. , J. Lindegren, and J. M. Andersen. pH effects on ciliomotility and morphology of
respiratory mucosa. Arch. Environ. Health 32:216-226, 1977.

Horvath, S. M. , and L. J. Folinsbee. Interactions of Two Air Pollutants, Sulfur Dioxide and
Ozone, on Lung Functions. University of California, Institute of Environmental Stress,
Santa Barbara, CA, 1977.

Howell, L. G. , and I. Fridovich. Sulfite: Cytochrome c oxidoreductase. J. Biol. Chem.
243:5941-5947, 1968.
12-91
-------
Huisingh, J. L. Bioassay of particulate organic matter from ambient air. In: Application of
Short-Term Bioassays in the Fractionation and Analysis of Complex Environmental Mixtures,
Vol. II, Plenum Press, New York, NY, 1981.

Huisingh, J. , R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S. Tejada, J. Bumgarner, F.
Duffield, and M. Waters. Application of bioassay to the characterization of diesel
particle emissions. Characterization of light and heavy duty diesel particle emissions.
EPA-600/9-78-027, U.S. Environmental Protection Agency, Health Effects Research
Laboratory and Environmental Sciences Research Laboratory, Research Triangle Park, NC,
1977.

Hyde, D. , J. Orthoefer, D. Dungworth, W. Tyler, R. Carter, and H. Lum. Morphometric and
morphologic evaluation of pulmonary lesions in beagle dogs chronically exposed to high
ambient levels of air pollutants. Lab. Invest. 38:455-469, 1978.

lida, S. , M. Inoue, K. Kai, N. Kitamura, I. Kudo, M. Sono, T. Tsuruo, H. Hayatsu, A. Miura,
and Y. Wataya. Some properties of the damage of DNA and phage 2 induced by bisulfite.
Mutat. Res. 20:433-434, 1974.

International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcino-
genic Risk of Chemicals to Man. Vol. 1, 1972, pp. 40-50.

International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcino-
genic Risk of Chemicals to Man. Vol. 2, 1973, pp. 126-149.

International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcino-
genic Risk of Chemicals to Man. Vol. 12, 1976, pp. 75-112.

Irreverre, F., S. H. Mudd, W. D. Heizer, and L. Laster. Sulfite oxidase deficiency: studies
of a patient with mental retardation, dislocated ocular lenses, and abnormal urinary
excretion of S-sulfo-L-cysteine, sulfite, and thiosulfate. Biochem. Med. 1:187-217,
1967.

Islam, M. S. , E. Vastag, and W. T. Ulmer. Sulphur-dioxide induced bronchial hyperreactivity
against acetylcholine. Int. Arch. Arbeitsmed. 29:221-232, 1972.

Kaden, D. A., R. A. Kites, and W. G. Thilly. Mutagenicity of soot and associated polycyclic
aromatic hydrocarbons to Salmonella typhimurium. Cancer Res. 3_9:415214159, 1979.

Kamogawa, A., and T. Fukui. Inhibition of a-glycan phosphorylase by bisulfite competition at
the phosphate binding site. Biochim. Biophys. Acta 302:158-166, 1973.

Kaplan, D. , C. McJilton, and D. Luchtel. Bisulfite induced lipid oxidation. Arch. Environ.
Health 30:507-509, 1975.

Katz, G. V., and S. Laskin. Pulmonary macrophage response to irritant gases. I_n: Air
Pollution and the Lung. E. F. Aharonson, A. Ben-David, and M. A. Klingberg, eds. , John
Wiley and Sons, New York, NY, 1976. pp. 83-100.

Ketels, K. V., J. N. Bradof, J. D. Fenters, and R. Ehrlich. SEM studies of the respiratory
tract of mice exposed to sulfuric acid mist-carbon particle mixtures. Jji: Scanning
Electron Microscopy. Volume II, IIT Research Institute, Chicago, IL, 1977. pp. 519-526.

Kikigawa, K. , and K. lizuka. Inhibition of platelet aggregation by bisulfite-sulfite. J.
Pharm. Sci. 61:1904-1907, 1972.
12-92
-------
King, L. C., Kohan, M. J,, Austin, A. C. , Claxton, L D. , and Huisingh, J. L. Evaluation of
the release of mutagens from diesel particles in the presence of physiological fluids.
Environ. Mutagen. 3:109-121, 1981.

Klebanoff, S. J. The sulfite-activated oxidation of reduced pyridine nucleotides by peroxi-
dase. Biochim, Biophys. Acta 48:93-103, 1961.

Kleinman, M. T. , R. M. Bailey, Y. C. Chang, K. W. Clark, M. P. Jones, W. S. Linn, and J. D.
Hackney. Exposures of human volunteers to a controlled atmospheric mixture of ozone,
sulfur dioxide and sulfuric acid. Am, Ind. Hyg. Assoc. J. 42:61-69, 1981.

Kolber, A., T. Wolff, T. Hughes, E. Pellozzari, C. Sparacino, M. Waters, J. L. Huisingh, and
L. Claxton. Collection, chemical fractionation and mutagenicity bioassay of ambient air
particulate. In: Application of Short-Term Bioassays in the Fractionation and Analysis
of Complex Environmental Mixtures, Vol. II, Plenum Press, New York, NY, 1981.

Roller, L. D. , J. H. Exon, and J. G. Roan. Antibody suppression by cadmium. Arch. Environ.
Health 30:598-601, 1975.

Kotin, P., H. L. Falk, and M. Thomas. Aromatic hydrocarbons: III. Presence in particulate
phase of diesel-engine exhaust and carcinogenicity of exhaust extracts. AMA Arch.
Ind. Health 11:113-120, 1955.

Kotin, P., H. L. Falk, P. Mader, and M. Thomas, Aromatic hydrocarbons. I. Presence in the
Los Angeles atmosphere and the carcinogenicity of atmospheric extracts. AMA Arch. Ind.
Hyg. Occup. Med. 9:153, 1954.

Kubitschek, H. E., and L. Venta. Mutagenicity of coal fly ash from electric power plant
precipitators. Environ. Mutagen. 1:79-82, 1979.

Kysela, B., D. Jirakova, R. Holusa, and V. Skoda. The influence of the size of quartz dust
particles on the reaction of lung tissue. Ann. Occup. Hyg. 16:103-109, 1973.

LaBelle, C. W., J. E. Long, and E. E. Cristofano. Synergistic effects of aerosols: Particles
as carriers of toxic vapors. Arch. Ind. Health 11:297-304, 1955.

Lamb, D., and L. Reid. Mitotic rates, goblet cell increase, and histochemical changes in
mucus in rat bronchial epithelium during exposure to sulphur dioxide. J. Pathol.
Bacteriol. 96:97-111, 1968.

Larson, T. V., 0. S. Covert, R. Frank, and R. J. Charlson. Ammonia in the human airways:
Neutralization of inspired acid sulfate aerosols. Science 197:161-163, 1977.

Larson, T. V., R. Frank, D. S. Covert, D. Holub, and M S.Morgan. Measurements of
respiratory ammonia and the chemical neutralization of inhaled sulfuric acid
aerosol in anesthetized dogs. Am. Rev. Resp. Dis. (in press), 1982.

Laskin, S. , M. Kuschner, and R. T. Drew. Studies in pulmonary carcinogenesis. In:
Inhalation Carcinogenesis. AEC Symposium Series 18. M. G. Hanna, Jr. , P. Nettesheim,
and J. R. Gilbert, eds., U. S. Atomic Energy Commission, Oak Ridge, TN, 1970. pp.
321-351.

Laskin, S., M. Kuschner, A. Sellakumar, and G. V, Katz. Combined carcinogen-irritant animal
inhalation studies. I_n: Air Pollution and the Lung. E. F. Aharonson, A. Ben-David, and
M. A. Klingberg eds., John Wiley and Sons, New York, NY, 1976. pp. 190-213.

Last, J. A., and C. E. Cross. A new model for health effects of air pollutants: Evidence for
synergistic effects of mixtures of ozone and sulfuric acid aerosols on rat lungs. J.
Lab. Clin. Med. 91:328-339, 1978.

12-93
-------
Lau, T. J. , R. L. Hackett, and F. W. Sunderman. The carcinogen!city of intravenous nickel
carbonyl in rats. Cancer Res. 32:2253-2258, 1972.

Lebowitz, M. D. , and G. A. Fairchild. The effects of sulfur dioxide on A^ influenza virus on
pneumonia and weight reduction in mice: An analysis of stimulus-response relationships.
Chem. Biol. Interact. 7:317-326, 1973.

Lee, F. S. C. , W. R. Pierson, and J. Ezike. The problem of PAH degradation during filter
collection of airborne particulates - an evaluation of several commonly used filter
media. In: Polynuclear Aromatic Hydrocarbons: Chemical and Biological Effects. Pro-
ceedings of the Fourth International Symposium. Battelle Press, Columbus, OH, 1980.

Leikauf, G., D. B. Yeates, K. A. Wales, D. Spektor, R. E. Albert and M. Lippmann. Effects of
sulfuric acid aerosol on respiratory mechanics and mucociliary particle clearance in
healthy non-smoking adults. Am. Ind. Hyg. Assoc. J. 42:273-292, 1981.

Leiter, J., and M. J. Shear. Production of tumors in mice with tars from city air dusts. J.
Natl. Cancer Inst. 3:167, 1942.

Leiter, J., M. B. Shimkin, and M. J. Shear. Production of subcutaneous sarcomas in mice with
tars extracted from atmospheric dusts. J. Natl. Cancer Inst. 3:155-165, 1942.

Leong, K. J. , H. N. MacFarland, and E. A. Sellers. Acute sulfur dioxide toxicity. Effects of
histamine and histamine liberation. Arch. Environ. Health 3:66-73, 1961.

Lewis, T. R. , D. E. Campbell, and T. R. Vaught, Jr. Effects on canine pulmonary function via
induced N09 impairment, particulate interaction and subsequent SO . Arch. Environ.
Health 18:5^6-601, 1969. x

Lewis, T. R. , W. J. Moorman, W. F. Ludmann, and K. I. Campbell. Toxicity of long-term
exposure to oxides of sulfur. Arch. Environ. Health 26:16-21, 1973.

Lewis, T. R., W. J. Moorman, Y. Yang, and J. F. Stara. Long-term exposure to auto exhaust and
other pollutant mixtures. Effects on pulmonary function in the beagle. Arch. Environ.
Health 29:102-106, 1974.

Li, A. P. Antagonistic effects of animal sera, lung and liver cytosols, and sulfhydryl
compounds on the cytotoxicity of diesel exhaust particle extracts. Toxicol. Appl.
Pharmacol. 57:55-62, 1981.

Li, A. P., Brooks, A. L. , and Royer, R. E. Cytotoxicity and mutagenicity of diesel exhaust
soot-extracts. Proc. Annu. Meeting Environ. Mutagen Society LL:94 (Abstract), 1980.

Liber, H. L. , B. M. Andon, R. A. Hites, and W. G. Thilly. Diesel soot mutation measurements
in bacterial and human cells. Health Effects of Diesel Engine Emissions: Proceedings
of an International.Symposium, EPA-600/9-80-057a Vol. 1, pp. 402-412, 1979.

Lippmann, M. , R. E. Albert, D. B. Yeates, K. Wales, and G. Leikauf. Effect of sulfuric acid
mist on mucociliary bronchial clearance in healthy non-smoking humans. J. Aerosol Sci.
11:247, 1980 (abstract).

Lippmann, M. , R. B. Schlesinger, G. Leikauf, D. Spektor, and R. E. Albert. Effects of
sulfuric acid aerosols on respiratory tract airways. In: Inhaled Particles. V. W. H.
Walton, ed. , Univin Brothers Imt., Surrey, England (In Press), 1981

Loring, S. H. , J. M. Drazen, J. R. Snapper, and R. H. Ingram. Vagal and aerosol histamine
interactions on airway responses in dogs. J. Appl. Physiol.: Resp. Environ. Exercise
Physiol. 45:40-44, 1978.

12-94
-------
Lyric, R. M., and I. Suzuki. Enzymes involved in the metabolism of thiosulfate by
thiobaci'Tlus thioparus. Survey of enzymes and properties of sulfite: Cytochrome c
oxidoreductase. Can. J. Biochem. 48:334-343, 1970.

Maigetter, R. Z. , R. Ehrlich, J. D. Fenters, and D. E. Gardner. Potentiating effects of
manganese dioxide on experimental respiratory infections. Environ. Res. 11:386-391,
1976. ~~

Malanchuk, M. , N. P. Barkley, and G. L. Center. Interference of animal source ammonia with
exposure chamber atmospheres containing acid particulate from automobile exhaust. J.
Environ. Pathol. Toxicol. 4:256-276, 1980.

Martin, S. W. , and R. A. Willoughby. Effect of sulfur dioxide on the respiratory tract of
swine. J. Am. Vet. Med. Assoc. 159:1518-1522, 1971.

Marunouchi, T. , and T. Mori. Studies on the sulfite-dependent ATPase of a sulfur oxidizing
bacterium, thiobacillus thiooxidans. J. Biochem. 62:401-407, 1967.

Massey, V., F. Muller, R. Feldberg, M. Schuman, P. A. Sullivan, L. G. Howe11, S. G. Mayhew, R.
G. Matthews, and G. P. Foust. The reactivity of flavoproteins with sulfite. J. Biol.
Chem, 244:3999-4005, 1969.

Matsumura, Y. The effects of ozone, nitrogen dioxide, and sulfur dioxide on the
experimentally induced allergic respiratory disorder in guinea pigs. I. The effect on
sensitization with albumin through the airway. Am. Rev. Resp. Dis. 102:430-437, 1970a.

Matsumura, Y. The effects of ozone, nitrogen dioxide, and sulfur dioxide on the
experimentally induced allergic respiratory disorder in guinea pigs. III. The effect on
the occurrence of dyspneic attacks. Am. Rev. Resp. Dis. 102:444-447, 1970b.

McCord, J. M. , and I. Fridovich. Superoxide dismutase. J. Biol. Chem. 244:6049-6055, 1969a.

McCord, J. M., and I. Fridovich. The utility of superoxide dismutase in studying free radical
reactions. J. Biol. Chem. 244:6056-6063, 1969b.

McDonald, S. Jr., and D. L. Woodhouse. On the nature of mouse lung adenomata, with special
reference to the effects of atmospheric dust on the incidence of these tumours. J.
Pathol. Bacteriol. 54:1-12, 1942.

McJilton, C., R. Frank, and R. Charlson. Role of relative humidity in the synergistic effect
of a sulfur dioxide-aerosol mixture on the lung. Science 182:503-504, 1973.

McJilton, C. E. , R. Frank, and R. J. Charlson. Influence of relative humidity on functional
effects of an inhaled S0«-aerosol mixture. Am. Rev. Respir. Dis. 113:163-169, 1976.
£
Michoud, M. C. , P. D. Pare, R. Boucher, and J. C. Hogg. Airway responses to histamine and
methacholine in Ascaris suum-allergic rhesus monkeys. J. Appl. Physio"!,: Resp.
Environ Exercise Physio!. 45:846-851, 1978.

Miller, E. C. Some current perspectives on chemical carcinogenesis in humans and
experimental animals: presidential address. Cancer Res. 38:1479-1496, 1978.

Mittler, S. , and S. Nicholson. Carcinogenicity of atmospheric pollutants. Ind. Med. Surg.
26:135, 1957.

Meiller, M. , and I. Alefheim. Mutagenicity and PAH-analysis of airborne particulate matter.
Atmos. Environ. 14:83-88, 1980.
12-95
-------
Mossman, B. T., K. B. Adler, and J. E. Craighead. Interaction of carbon particles with
trachea! epithelium in organ culture. Environ. Res. 16:110-122, 1978.

Hudd, S. H,, F. Irreverre, and L. Laster. Sulfite oxidase deficiency in man: demonstration
of the enzymatic defect. Science 156:1599-1602, 1967.

Mukai, F. , I. Hawryluk, and R. Shapiro. The mutagenic specificity of sodium bisulfite.
Biochem. Biophys. Res. Commun. 39:983-988, 1970.
"V
Muller, F. , and V. Massey. Flavin-sulfite complexes and their structures. J. Biol. Chem.
244:1007-1016, 1969.

Nadel, J. A., H. Salem, B. Tamplin, and Y. Tokiwa. Mechanism of broncho-constriction during
inhalation of sulfur dioxide; reflex involving vagus nerves. Arch. Environ. Health
10:175-178, 1965a.

Nadel, J. A., H. Salem, B. Tamplin, and Y. Tokiwa. Mechanism of bronchoconstriction during
inhalation of sulfur dioxide. J. Appl. Physio!. 20:164-167, 1965b.

Nadel, J. A., M. Corn, S. Zwi, J. Flesch, and P. Graff. Location and mechanism of airway con-
striction after inhalation of histamine aerosol and inorganic sulfate aerosol. In:
Inhaled Particles and Vapours. Vol II. C. N. Davies, ed., Pergamon Press, Oxford,
England 1967. p. 55-67.

Nakamura, S, Initiation of sulfite oxidation by spinach ferredoxin-NADP reductase and ferre-
doxin system: A model experiment on the superoxide anion radical production by
metallo-flavoproteins. Biochem. Biophys. Res. Commun. 41:177-183, 1970.

National Academy of Sciences. Asbestos: The need for and feasibility of air pollution
controls. National Academy of Sciences, Washington, DC, 40 pp. , 1971.

National Academy of Science. Arsenic. National Academy of Sciences, Washington, DC, 1977.

National Academy of Sciences. Iron. University Park Press, Baltimore, MD, 1979a.

National Academy of Sciences. Zinc. University Park Press, Baltimore, MD, 1979b.

National Air Pollution Control Administration. Air Quality Criteria for Sulfur Oxides.
AP-50, U.S. Government Printing Office, Washington, DC, 1970.

National Institute for Occupational Safety and Health. Criteria for a Recommended Standard:
Occupational Exposure to Crystalline Silica. Ch. Ill, Biologic Effects of Exposure.
DAEW (NIOSH), Pub. No. 75-120, Cincinatti, OH, 1975. pp. 15 foil.

National Research Council. Sulfur Oxides. National Academy of Sciences, Washington, DC,
1978.

National Research Council. Airborne Particles. University Park Press, Baltimore, MD, 1979.

Nulsen, A., P. G. Holt, and D. Keast. Sulfur dioxide. Acute effects on cell metabolism jjn
vitro. IRCS Libr. Compend. 2:1464, 1974.

Office of Research and Development. Air Quality Criteria for Lead. EPA-600/8-77-OL7, U.S.
Environmental Protection Agency, Washington, DC, December 1977.

Ohnishi, Y., K. Kachi, K. Sato, I. Tohara, H. Takeyoshi, and H. Tokiwa. Detection of mutagenic
activity in automobile exhaust. Mutat. Res. 77:229-246, 1980.
12-96
-------
Orthoefer, J. G. , R. S. Bhatnagar, A. Rahman, Y. Yang, S. D. Lee, and J. F. Stara. Collagen
and proly'l hydroxylase levels in lungs of beagles exposed to air pollutants. Environ.
Res. JJ:299-305, 1976.

Oshino, N. , and B. Chance. The properties of sulfite oxidation in perfused rat liver; inter-
action of sulfite oxidase with the mitochondria! respiratory chain. Arch. Biochem.
Biophys. 170:514-528, 1975.

Ottery, J. , and I. P. Gormley. Some factors affecting the hemolytic activity of silicate
minerals. Ann. Occup. Hyg. 21-131-139, 1978.

Rattle, R. E. , F. Burgess, and H. Cullumbine. The effects of a cold environment and of
ammonia on the toxicity of sulfuric acid mist to guinea pigs. J. Pathol. Bacteriol.
72:219-232, 1956.

Peacock, P. R. , and J. B. Spence. Incidence of lung tumours in LX mice exposed to (1) free
radicals; (2) S02- Br. J. Cancer 21:606-618, 1967.

Peiser, G. D., and S. F. Yang. Chlorophyll destruction by bisulfite-oxygen system. Plant
Physio!. 60:277-281, 1977.

Pitts, J. N. , D. Grosjean, J. M. Mischke, V. F. Simmons, and D. Poole. Mutagenic activity of
airborne particulate organic pollutants. Toxicol. Lett. 1:65-70, 1977.

Pitts, J. N. , Jr., Van Cauwenberghe, K. A., Grosjean, D., Schmid, J. P., Fitz, D. R., Belser,
W. L. , Jr., Knudson, G. B. and Hynds, P. M. Atmospheric reactions of polycyclic aromatic
hydrocarbons: facile formation of mutagenic nitro derivatives. Science 202, p. 515-519,
1978.

Pott, F. , R. Tomingas, and J. Misfeld. Tumours in mice after subcutaneous injection of auto-
mobile exhaust condensates. In: Air Pollution and Cancer in Man. IARC Scientific
Publications No. 16, 1977. pp. 79-88.

Pylev, L. N. and L. M. Shabad. Some results of experimental studies in asbestos
carcinogenesis. In: Biological Effects of Asbestos. IARC Scientific Publication No.
8, 1973, 99-106.

Reid, L. Evaluation of model systems for study of airway epithelium, cilia, and mucus. Arch.
Intern. Med. 126:428-434, 1970.

Reiser, K. M. , and J. A. Last. Silicosis and fibrogenesis: fact and artifact. Toxicol.
13:51-72, 1979.

Rigdon, R. H. , and J. Neal. Tumors in mice induced by air particulate matter from a
petrochemical industrial area. Texas Rep. Biol. Med. 29:109-123, 1971.

Rotilio, G. , L. Calabrese, A. Finazzi Agro, and B. Mondovi. Indirect evidence for the
production of superoxide anion radicals by pig kidney diamine oxidase. Biochem. Biophys.
Acta 198:618-620, 1970.

Rudd, C. J. Diesel particulate extracts in cultured mammalian cells. Health Effects of
Diesel Engine Emissions: Proceedings of an International Symposium, EPA-600/9-80-057a
Vol. 1, pp. 385-403, 1979.

Rylander, R. Alterations of lung defense mechanisms against airborne bacteria. Arch
Environ. Health 18:551-555, 1969.
12-97
-------
Rylander, R., M. Ohrstrom, P. A. Hellstron, and R. Bergstrom. SOg and particles: synergistic
effects on guinea pig lungs. In: Inhaled Particles III. Vol. I. W. H. Walton, ed. ,
Unwin Bros., Ltd., Surrey, England, 1970. pp. 535-541.

Sackner, M. A., D, Ford, R. Fernandez, J. Cipley, D. Peroz, M. Kwoka, M. Reinhardt, E. 0.
Michael son, R. Schreck, and A. Wanner. Effects of sulfuric acid aerosol on
cardiopulmonary functions in dogs, sheep and humans. Am. Rev. Respir. Dis. 118:497-510,
1978a.

Sackner, M. A., D. Perez, M. Brito, and R. M. Schreck. Effect of moderate duration exposures
to sulfate and sulfuric acid aerosols on cardiopulmonary function of anesthetized dogs.
Am. Rev. Respir. Dis. 117:257, 1978b.

Saffioti, U. , F. Cefis and L. H. Kolb. A method for the experimental induction of
bronchogenic carcinoma. Cancer Res. 18:104-124, 1968.

Saito, K., and D. B. Menzel. Nickel uptake and efflux from cultured type II pneumocytes.
Pharmacologist 20:275, 1978.

Santodonato, J. , D. K. Basu, and P. H. Howard. Health effects associated with diesel exhaust
emissions: literature review and evaluation. EPA #600/1-78-063, U.S. Environmental
Protection Agency, 1978. 165 pp.

Schiff, L. J., M. M. Bryne, J. D. Fenters, J. A. Graham, and D. E. Gardner. Cytotoxic effects
of sulfuric acid mist, carbon particulates, and their mixtures on hamster trachea!
epithelium. Environ. Res. 19:339-354, 1979.

Schlesinger, R. B. , J. L. Gurman, and L. C. Chen. The production and characterization of a
transition metal [FE(III)]-S(IV) Aerosol. Atmos. Environ. 14:1279-1287, 1980.

Schlesinger, R. B. , M. Lippmann, and R. E. Albert. Effects of short-term exposures to
sulfuric acid and ammonium sulfate aerosols upon bronchial airways function in donkeys.
Am. Ind. Hyg. Assoc. J. 39:275-286, 1978.

Schlesinger, R. B. , M. Halpern, R. E. Albert, and H. Lippmann. Effect of chronic inhalation
of sulfuric aoid mist upon mucociliary clearance from the lungs of donkeys. J. Environ.
Pathol. Toxicol. 2:1351-1367, 1979.

Schneider, L. K., and C. A. Calkins. Sulfur dioxide-induced lymphocyte defects in human peri-
pheral blood cultures. Environ. Res. 3:473-484, 1971.

Schroeder, H. A. A sensible look at air pollution by metals. Arch. Environ. Health
21:798-806, 1970.

Seelig, M. G. , and E. L. Benignus, Coal smoke soot and tumors of the lung in mice. Am. J.
Cancer 28:96-111, 1938.

Shahin, M. M. , and F. Fournier. Suppression of mutation induction and failure to detect
mutagenic activity with athabasca tar sand fractions. Mutat. Res. 58:29-34, 1978.

Shapiro, R. Genetic effects of bisulfite (sulfur dioxide), Mutat. Res. 39:149-176, 1977.

Shapiro, R., and J. M. Weisgras. Bisulflte-catalyzed transamination of cytosine and cytidine.
Biochem. Biophys. Res. Commun. 40:839-843, 1970.

Shapiro, R., B. I. Cohen, and R. E. Servis. Specific deamination of RNA by sodium bisulphite.
Nature 227:1047-1048, 1970a.
12-98
-------
Shapiro, R., R. E. Servis, and M. Welcher. Reactions of uracil and cytosine derivatives with
sodium bisulfate. A specific deamination method. J. Am. Chem. Soc. 92:422-424, 1970b.

Shih, V. E. , I. F. Abroms, J. L. Johnson, M. Carney, R. Mandell, R. M. Robb, J. P. Cloherty,
and K. V. Rajagopalan. Sulfite oxidase deficiency. Biochemical and clinical
investigations of a hereditary metabolic disorder in sulfur metabolism. N. Engl. J. Med.
297:1022-1028, 1977.

Shortle, D. , and D. Nathans. Local mutagenesis: A method for generating viral mutants with
base substitutions in preselected regions of the viral genome. Proc. Natl, Acad. Sci.
U.S.A. 75:2170-2174, 1978.

Silbaugh, S. A., R. K. Wolff, W. K. Johnson, J. L. Mauderly and C. A. Macken. Effects of
sulfuric acid aerosols on pulmonary function of guinea pigs. J. Toxicol. and Environ.
Health 7:339-352, 1981.

Singh, J. Biochemistry of silicosis, J. Sci. Ind. Res. 37:328-333, 1978.

Sivak, A. Overview and status of in v itro transformation. J. Assoc. Off. Anal. Chem. 62:889,
1979.

Skopek, T. R. , Liber, H. L. , Kaden, D. A., Hites, R. A. and Thilly, W. G. Mutation of human
cells by kerosene soot. J. Natl. Cancer Inst. 63:309-312, 1979.

Snapper, J. R. , J. M. Drazen, S. H. Loring, W. Schneider, and R. H. Ingram, Jr. Distribution
of pulmonary responsiveness to aerosol histamine in dogs. J. Appl. Physio!.: Resp.
Environ. Exercise Physio!. 44:738-742, 1978.

Spiegelman, J. R. , G. 0. Hanson, A. Lazarus, R. J. Bennett, M. Lippmann, and R. D. Albert.
Effect of acute sulfur dioxide exposure on bronchial clearance in the donkey. Arch.
Environ. Health 17:321-326, 1968.

Stara, J. F. , D. L. Dungworth, J. G. Orthoefer and W. S. Tyler, eds. Long-term Effects of
Air Pollutants in Canine Species. EPA-600/8-80-Q14, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., 1980.

Stenback, F. , A. Ferrero, R. Montesano, and P. Shubik. Synergistic effect of ferric oxide on
dimethylnitrosamine carcinogenesis in the Syrian golden hamster. Z. Krebsforsch.
79:31-38, 1973.

Stenback, F. , J. Rowland, A. Sellabsumor. Careinogenicity of benzo(a)pyrene and dusts in
hamster lung (instilled intratracheally with titanium oxide, carbon and ferric oxide).
Oncology 33:29-34, 1976.

Stern, A. C. Air Pollution. Vol. II. Analysis, Monitoring, and Surveying. Academic Press,
New York, NY, London, England, 1968.

Stoner, G. D. , M. B. Shinkin, M. C. Troxel!, T. L. Thompson, and L. S. Terry. Test for
carcinogenicity of metallic compounds by the pulmonary response in strain A mice.
Cancer Res. 36:1744-1747, 1976.

Strandberg, L. G. S07 absorption in the respiratory tract. Arch. Environ. Health 9:160-166,
1964. z

Summers, G. A., and J. W. Drake. Bisulfite mutagenesis in bacteriophage T4. Genetics
68:603-607, 1971.

Sunderman, F. W. Carcinogenic effects of metals. Fed. Proceedings, _37(l):40-46, 1978.

12-99
-------
Sunderman, F. W. Metal carcinogenesis. Adv. Mod. Toxicol. 2:256-295, 1979.

Tager, J. M, , and N. Rautanen. Sulphite oxidation by a plant- mitochondria! system I.
Preliminary observations. Biochim, Biophys. Acta 18:111-121, 1955.

Takebe, I. Isolation and characterization of a new enzyme choline sulfatase. J. Biochem.
50:245-255, 1961.

Takino, Y., K, Sugahara, and I. Horino. Two lines of guinea pigs sensitive to chemical
mediators and anaphylaxis. J. Allergy 47:247, 1971.

Tartar, H. V., and H. H. Garetson. The thermodynamic ionization constants of sulfurous acid
at 25°. J. Am. Chem. Soc. 63:808, 1941.

Teranishi, K., K. Hamada, and H. Wantanabe. Mutagenicity in Salmonella typhimurium mutants of
the benzenesoluble organic matter derived from air-borne particulate matter and its five
fractions. Mutat. Res. 56:273-280, 1978.

Thompson, J. R., and D. M. Pace. Effects of S07 on established cell lines cultivated i_n
vitro. Can. J. Biochem. Physio!. 40:207-217, 1962.

Timson, J. Action of sodium sulphite on the mitosis of human lymphocytes. Chromosomes Today
4:211-214, 1973.

Tokiwa, H. , K. Morita, H. Takeyoshi, K. Takahashi, and Y. Ohnishi. Detection of mutagenic
activity in particulate air pollutants. Mutat. Res. 48:237-248, 1977.

Tokiwa, H., K. Shigeji, K. Takahashi, and Y. Ohnishi. Mutagenic and chemical assay of
extracts of airborne particulates. Mutat. Res. 77:99-108, 1980.

Tomori Z., and J. G. Widdicomble. Muscular, bronchomotor and cardiovascular reflexes elicited
by mechanical stimulation of the respiratous tract. J. Physio!. 200:25-49, 1969.

Tuazon, P. T., and S. L. Johnson. Free radical and ionic reaction of bisulfite with reduced
nicotinamide adenine dinucleotide and its analogues. Biochemistry 16:1183-1188, 1977.

Valencia, R. , S. Abrahamson, P. Wagoner, and L. Mansfield. Testing for food additive-induced
mutations in Orosophila melanogaster. Mutat. Res. 21:240-241, 1973.

Vaughan, T. R., Jr., L. F. Jennelle, and T. R. Lewis. Long-term exposure to low levels of air
pollutants: Effects on pulmonary function in the beagle. Arch. Environ. Health
19:45-50, 1969.

Verrant, J. A., and Kittelson, D. B. Sampling and characterization of diesel exhaust
aerosols. SEA Paper No. 770720, 1979.

Wang, Y. Y., and Wei, E. T. Detoxifying properties of liver homogenates, proteins and
glutathione on diesel exhaust particulates. Proc. EPA Symp. Appl. Short-Term Bioassays
Anal. Complex Environ. Mixtures, Vol. 2, 66 (Abstract) 1980.

Wang, Y. Y. , S. M. Rappaport, R. F. Swayer, R. E. Talcot't, and E. T. Wei. Direct-acting
mutogens in automobile exhaust. Cancer Lett. 5_:39-47, 1978.

Waters, M. D. , D. E. Gardner, and D. L. Coffin. Cytotoxic effects of vanadium on rabbit
alveolar macrophages in vitro. Toxcol. Appl. Pharmacol. 28:253-263, 1974.
12-100
-------
Wattiaux-DeConinck, S. , and R. Wattiaux. Subcellular distribution of sulfite cytochrome c
reductase in rat liver tissue. Eur. J. Biochem. 19:552-556, 1974.

White, R. , and C. Kuhn. Effects of phagocytosis of mineral dusts on elastase secretion by
alveolar and peritoneal exudative macrophages. Arch. Environ. Health 35:106-109, 1980.

Widdicombe, J. G. Respiratory reflexes from the trachea and bronchi of the cat. J. Physio!.
123:55-70, 1954a.

Widdicombe, J. G. Receptors in the trachea and bronchi of the cat. J. Physio!. 123:71-104,
1954b.

Widdicombe, J. G., D. C. Kent, and J. A. Nadel. Mechanism of bronchoconstriction during inha-
lation of dust. J. Appl. Physio!. 17:613-616, 1962.

Williams^ S. J., K. M. Holden, M. Sabransky, and D. B. Menzel. The distributional kinetics of
Ni ions in the rat lung. Toxicol. Appl. Pharmacol. 55_:85-93, 1980.

Wilson,, D. F. The inhibition of mitochondrial respiration by bisulfite ions. Fed. Proc. Fed.
Am. Soc. Exp. Biol. 27:830, 1968.

Wirth, J. J. , W. P. Carney, and E. F. Wheelock. The effect of particle size on the
immunodepressive properties of silica. J. Immunol. Methods ^2:357-373, 1980.

Wolff, R. K., S. A. Silbaugh, D. G. Brownstein, R. L. Carpenter, and J. L. Mauderly. Toxicity
of 0.4 and 0.8 urn sulfuric acid aerosols in the guinea pig. J. Toxicol. Environ. Health
5:1037-1047, 1979.

Wolff, R. K., B. A. Muggenburg, and S. A. Silbaugh. Effect of 0.3 and 0.9 urn sulfuric acid
aerosols on trachea! mucous clearance in beagle dogs. Am. Rev. Resp. Dis. 123: 291-294,
1981a.

Wolff, R. K. , G. M. Kanapilly, P. B. DeNee, and R. 0. McClellan. Deposition of 0.1 pm
chain aggregate aerosols in beagle dogs. J. Aerosol Sci. 12:119-29, 1981b.

Wynder, E. L. , and D. Hoffman. A study of air pollution carcinogenesis. III. Carcinogenic
activity of gasoline engine exhaust condensate. Cancer 15:103-108, 1962.

Yang, S. F. Biosynthesis of ethylene. Ethylene formation from methional by horseradish
peroxidase. Arch. Biochem. Biophys. 122:481-487, 1967.

Yip, C. C. , and L. D. Hadley. The iodination of tyrosine by myeloperoxidase and beef
thyroids. The possible involvement of free radicals. Biochim. Biophys. Acta
122:406-412, 1966.

Yokoyama, E. , R. E. Yoder, and N. R. Frank. Distribution of 35S in the blood and its
excretion in urine of dogs exposed to 35S02. Arch. Environ. Health 22:389-395, 1971.

Zarkower, A. Alterations in antibody response induced by chronic inhalation of SO- and
carbon. Arch. Environ. Health 25:45-50, 1972.

Ziegler, I. Action of sulfite on plant malate dehydrogenase. Phytochemistry 13:2411-2416,
1974.

Ziskind, M. , R. N. Jones, and H. Weill. Silicosis. Am. Rev. Resp. Dis. 113:643-665, 1976.

Zucker, M. , and A. Nason. Hydroxylamine reductase from neurospora crassa. Methods Enzymol.
2:415-419, 1955.

12-101
-------
APPENDIX 12-A
EPA Reanalysis of the Data of Peacock and Spence (1967) and Laskin et al. (1976).
12-102
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
DATE
SUBJECT
FROM
TO
October 24, 1980

Analysis of the Laskin, et al and Peacock and Spence Data for Chapter 12
of the SOX/PM Document.
Victor Hasselblad
Andrew Stead i/-^.
Biometry Division (MD-55)

Dr. Lester D. Grant
Director, ECAO (MO-52)

We analyzed the Laskin, et al data (p. 195) using multiple probit
analysis. The model used was:
y1 = PCPo
p3C)
.
where pc is the estimated background rate
Pi is the increase due to S02(S)
P2 is the increase due to BAP(B)
B3 is the increase due to the combination of BAP and S02(C)
e. is the binomial random variation attributed to each animal
P is the probit (cumulative normal) function
and y. is the response (1 means cancer, 0 means none).

The model was fitted using GLIM which is a British general linear model
fitting package. The chi-square tests are really asymptotic chi-squares
resulting from likelihood ratio tests computed by fitting reduced models.
Estimated
Coefficient Risk Chi-square P-value
Factor

Background (P0)

S°2 (Pi)

BAP (pa)

Combination (p3)

The estimated effect of S02 alone is very small, but there is a hint that
it increases the effect of the S02~BAP combination. BAP alone has a hint
of/an effect. The only truly significant effect is that of the S02-BAP
combination. There is no way to test for interactions since the model comes
close to "overpredicting" as it is. Compare the observed vs. predicted
cancers:
-5.068

.449
3.160
3.744
2.0X10 '
-K
1.9X10 b
.028
.093
-

2.468
2.511
7.943
-

.116
.113
.005
EPA Form 1320-6 (Rtv. 3-76)
Expose

A
I
A+C
I+C
A+CI
I+CI
Number of
Animals

15
15
30
30
45
46
Number of
Cancers
Predicted
Cancers
0
0
1
2
4
9
0.000000
0.000007
0.846
2.168
4.174
8.776
12-103
-------
Page 2 - Analysis of the Laskin, et al, and Peacock and Spence Data for
Chapter 12 of the SOX/PM Document.
A: air alone
I: S02 (10 ppm) alone
A+C: air plus benzo(a)pyrene (10 mg/m3)
I+C: S02 (10 ppm) plus b(a)p (10 mg/m3)
A+CI: air plus b(a)p (10 mg/m3) and SO^ (10 ppm)
I+CI: S02 (10 ppm) plus b(a)p (10 mg/m°) and S02 (4 ppm)

We also analyzed S02 data of Peacock and Spence (p. 616) using 2X2
contingency tables. The tests were made using a one-sided Fisher's
exact test which is appropriate for the small number of observed cases.
The following are the results:

One-sided
Response P value

Primary Carcinoma .604

Primary Carcinoma .056

Males Adenoma .065

Females Adenoma .011

These results can be summarized as marginally significant, with an in-
dication that females are more susceptible. I do not have a test of
the sex difference at this time, but I will try to forward a more
sophisticated analysis next week,

cc: Dr. Judy Graham
Or. Fred Miller
Dr. Daniel Menzel (Duke)
12-104
-------
13. CONTROLLED HUMAN STUDIES

13.1 INTRODUCTION
Controlled human exposure studies provide a necessary bridge between epidemiology and
animal toxicology data in attempting to characterize health effects induced by air pollution.
In general, such studies should utilize exposure conditions that realistically simulate expo-
sures experienced by human beings in their normal environment. However, the complexity and
variability of the ambient environment is such that most controlled human exposure studies
have typically been designed to evaluate initially the effects of exposure to single pollut-
ants and later to examine effects of more complex mixtures of pollutants analogous to those
present in the ambient environment.
The usually limited numbers of subjects that can be studied under controlled conditions
and the associated high costs make it imperative that such studies be conducted under strin-
gent conditions in order for their findings to be relevant to larger segments of the general
population and, therefore, maximally useful for criteria development purposes. Ideally, the
design of such controlled trials should include normal individuals of both sexes and all age
groups, subjects especially sensitive to some particular pollutant, and individuals from popu-
lations suspected to be more generally at special risk for air pollution effects. Considera-
tion should also be given to the activity levels of the subjects, ambient environmental con-
ditions prevailing prior to the subjects' testing, and exposure variables that realistically
simulate ambient conditions, including such factors as temperature, humidity, duration of ex-
posure, and mode of exposure. Controlled human exposure studies also require proper experi-
mental design, including such precautions as use of purified air and, ideally, double blind
exposure procedures, especially where psychological factors may be involved as in testing sug-
gestible asthmatics (double blind procedures, however, cannot always be complied with when
certain pollutants or levels of some pollutants are being evaluated in view of associated
medical risks). Furthermore, several concentrations of the" pollutant(s) being tested should be
employed in order to develop dose-response relationships, and comprehensive statistical treat-
ment of the data should be carried out as well. In addition, adequate (even duplicate) pollu-
tant monitoring equipment with documentation of quality control is needed. Proper attention
should also be accorded to monitoring for potentially interfering pollutants inadvertently
present or developing under certain exposure conditions. Ideally, in measuring dependent vari-
ables, 'not only should physiological and biochemical evidence be obtained, but subjective
symptoms and/or changes in performance capability should also be assessed. Since the respira-
tory tract is the initial target of many air pollutants, proper and sensitive respiratory
function measurements are a primary requirement and, since diurnal patterns are known to occur
in many physiological systems, the time of day during which experiments are conducted should
be controlled. Also, since various blood chemistry parameters and other non-respiratory
system biochemical processes may be affected if pollutants (or their reaction products or
13-1
-------
substances absorbed on particles) enter the circulatory system, such biochemical effects
should also be measured, as appropriate.
The above criteria should be applied to any evaluation of controlled human exposure
studies. However, because of particular restraints placed on investigators, virtually no
studies meet all of these ideal requirements. Nonetheless, certain useful information can be
derived from many existing studies, and this chapter assesses those controlled human exposure
studies most germane to understanding human health effects of sulfur dioxide, associated par-
ticulate matter, and nonsulfur particulate- compounds encountered in the ambient air. It
should be noted that laboratory studies utilizing man have generally been limited to the
evaluation of acute exposure effects; thus the potential for chronic health effects cannot be
fully evaluated based on such exposures.
13.2 SULFUR DIOXIDE
Exposure of man to sulfur dioxide has been shown to induce a number of physiological re-
sponses. Alterations in sensory system responses (such as irritation of eyes and nose,
changes in odor perception, and dark adaptation) have been reported. Various changes in
respiratory system pulmonary functions have also been reported. The following sections
attempt to characterize such physiological and pathophysiological changes, associated dose/
response or exposure/effect relationships, and interacting factors that may substantially
alter such relationships (e.g., alterations in respiratory tract S0»-deposition patterns are
dependent upon exercise levels or the health status of test subjects as noted in Chapter 11).
13.2.1 Subjective Reports
The perception of odor and the sensation of irritation in the eyes, nose, throat, or
other parts of the body are difficult to measure precisely. Thus subjects may be observed for
qualitative changes (coughing, rhinorrhea, lacrimation) or asked to report whether they detect
something in the air they are breathing. Several studies have used such subjective reports as
an indication of the effects of S0? on human subjects.
A number of early investigators exposed themselves to high concentrations of SO, (>500
3 '"
ppm [1310 mg/m ]) and experienced coughing, irritation of the eyes and nose, and difficulty
in breathing (e.g., Ogata, 1884; Yamada, 1905; Kisskalt, 1904). In the course of investiga-
ting the effects of S09 on industrial workers, Lehman (1893) and his associates experienced
3
nasal irritation during exposures of 10 to 15 minutes to 6.5 ppm (17 mg/m ) SOp. Holmes et
al. (1915 — cited by Greenwald, 1954) carried out an extensive study of 60 subjects, 28 of
whom were unaccustomed to breathing SO,,, and 32 of whom were familiar with it. All of the sub-
jects already famiMiar with the gas seemed to detect it (either as SO^, or as "something
3
foreign") at 3 ppm (7.9 mg/m ). But only 10 of 28 unaccustomed subjects detected something in
the air at 3 ppm (7.9 mg/m ) S0?. Few subjects found momentary whiffs of 5 ppm (13.1 mg/m )
disagreeable, although "long-continued breathing of air containing slightly more than 5 parts
per million would probably cause discomfort to most people..." (Holmes et al., 1915). Amdur
et al. (1953) noted that during exposure to 1 to 2 ppm (2.6 to 5.2 mg/m ) their subjects could

13-2
-------
not usually detect the odor of SO-; even at 5 ppm (13.1 mg/m ) most subjects could not smell
the gas, although they did complain of dryness in the throat. One subject, however, objected
2
so strongly to the odor of 5 ppm (13.1 mg/m ) S0~ that exposure was terminated. Above 5 ppm
3
(13.1 mg/m ) the odor was definitely detected by all subjects. In numerous more recent
studies, subjects have also been ask'ed to report " th&ir subjective experiences (e.g.,
Greenwald, 1954; Tomono, 1961; Frank et al., 1962; Toyama and Nakamura, 1964; Speizer and
Frank, 1966a,b; Melville, 1970; Weir and Bromberg, 1972, 1973; Lawther et al. , 1975; Horvath
and Folinsbee, 1977), and the results seem to be quite variable at exposures less than 5 ppm
so2.
The significance of such subjective reports is unclear at present. This is especially
true in light of findings by Frank et al. (1962) showing that subjective reports are in some
situations an unreliable indicator of physiological changes, since coughing and a sense of
throat irritation tended to subside in their subjects after a few minutes while other changes
in respiratory functions were still maximal.
13.2.2 Sensory Effects
Among the physiological functions affected by exposure to S0? are certain sensory pro-
t *-
cesses. Studies investigating sensory effects not only evaluated odor threshold for detection
of SQp but also S0? effects on the sensitivity of the dark-adapted eye and interruption of the
alpha (a) rhythm in electroencephalograms (see Table 13-1). Most of these investigations have
been summarized by Ryazanov (1962).
13.2.2.1 Odor Perception Threshold—In the Russian studies, odor threshold is typically deter-
mined in a well-ventilated chamber containing two orifices from which emerge two small streams
of gas, one being very pure air and the other other being a stream of the test gas. The sub-
ject sits in front of the apparatus, sniffs both orifices, and points out the odorous one.
This experiment is repeated with the same concentration of test gas over a period of several
days. The experiment is performed with sequentially reduced concentrations until the subject,
in the majority of instances, denies the presence of an odor or gives erroneous answers. The
threshold concentration for the most sensitive subject in a group of volunteers is defined as
the threshold for odor perception.
Using the two-orifice apparatus described above, Dubrovskaya (1957) conducted sulfur
dioxide odor perception threshold tests on 12 subjects. Sulfur dioxide concentrations of 0.5
3 3
mg/m to 13 mg/m (0.17 ppm to 4.6 ppm) were used in 530 threshold determinations. Six test
subjects sensed the odor of sulfur dioxide in the range 2.6 mg/m to 3.0 mg/m (1.0 to 1.1
3 3
ppm); four subjects sensed the odor in the range 1.6 mg/m to 2.0 mg/m (0.6 to 0.8 ppm); one
sensed the odor in the range 2.1 mg/m to 2.5 mg/m (0.8 to 1.0 ppm); and one sensed the odor
in the range 3.1 mg/m to 3.6 mg/m (1.2 to 1.4 ppm). Thus, the average sulfur dioxide odor
threshold concentration was 0.8 ppm to 1 ppm (~2.3 mg/m to ~2.9 mg/m ), and for the more
3 3
sensitive of these persons it was 0.5 ppm to 0.7 ppm (~1.5 mg/m to ~2.0 mg/m ). It should be-
noted, however, that most of the subjects were of an age at which odor perception was presumed
to be most sensitive.
13-3
-------
TABLE 13-1. SENSORY EFFECTS OF SO,
I
-fc,
Concentration
S02 (ppm)
400
6,5
140, 210, 240
210, 240
1, 2, 5
3, 5, 5 plus
0.17 - 4.6

0.34 - 6.9
0.23
0 2 - 1.7
1-10
Exposure
mins. Effects
120 Dyspnea
10 - 15 Nasal irritation
30 Harked nasal irritation, sneezing
30 Eye irritation, lacrimation
All subjects detect odor above 5 ppm
Discomfort to all subjects exposed to 5 plus.
Some noted disagreeable odor at 5 ppm.
Average SO, odor threshold was 0.8 - 1 0 ppm
Positive recognition of SO, was 0.47 ppm
15 Light sensitivity increased at 0.34 - 0 63 ppra
and above
Ocular sensitivity to light increased at SO,
levels of 0.23 ppm and above
0.33 Attenuation of B-waves at levels above 0.2 ppm
Organoleptic effects at levels 2 ppm and above
Reference
Ogata, 1884
Lehman, 1893
Yamada, 1905
Yamada, 1905
Amdur et al., 1953
Holmes, 1915 (see Green-
wald, 1954)
Dubrovskaya, 1957
Arthur D. Little, Inc., 1968
Dubrovskaya, 1957
Shalamberidze, 1967
Bushtueva, 1962
Greenwald, 1954
0.1 ppm S02 £ 262 yg/m*

0 5 ppm S02 s 1310 M9/raa
1,0 pp« s 2620 |jg/m3

5 0 ppm s 13,100 Mg/m
10 ppn s 26,200 (jg/m3

50 ppm £ 131,000 (jg/m3
-------
Studies of sulfur dioxide odor thresholds (1968) conducted for the Manufacturing Chemists'
Association in the United States gave somewhat lower values than those cited above (Arthur D.
Little, Inc., 1968). The concentration at which first one-half and then all of the panel
3
members could positively recognize the odor was reported to be 0.47 ppm (1.3 mg/m ). The de-
tails of the test procedure are thoroughly discussed in the report, but one important aspect
is reiterated as a reminder that odor thresholds usually represent values derived under
ideally suited conditions with trained individuals. That is, under ideal conditions, the
investigators (who were highly qualified to judge on the basis of substantial experience with
consumer evaluation of known flavor and odor situations) derived threshold values lower than
those which would be recognized by the majority of people under ordinary atmospheric condi-
tions. This does not mean that normal individuals exposed to sulfur dioxide under ideal test
o
conditions could not perceive the 0.47 ppm (1.2 mg/m ) level indicated. However, because of
background odor and lack of awareness or concern with ambient odor conditions, such indi-
viduals would probably be less responsive to this low concentration in everyday situations.
13.2.2.2 Sensitivity of the Dark-AdaptedEye--The sensitivity of the eye to light while a
subject is in darkness increases with time. Several investigations have been made of the
effects of inhalation of sulfur dioxide on this sensitivity. Typically, measurements of a
subject's normal sensitivity are taken in a dark, well-ventilated chamber in complete silence.
(Sudden stimuli, including noise, may affect the subject's response.) Each subject is tested
once daily following preliminary stimulation at a high light level. Light sensitivity is
measured at 5-minute or 10-minute intervals, and a curve of increasing sensitivity to light is
established from measurements taken over a period of 7 to 10 days.
Dubrovskaya (1957) studied the effect of inhaling sulfur dioxide in concentrations from
3 3
0.96 mg/m to 19.2 mg/m (0.4 to 7.3 ppm) for 15 minutes before measuring light sensitivity
during dark adaptation. She reported that light sensitivity was increased by sulfur dioxide
3 3
concentrations of 0.96 mg/m to 1.8 mg/m (0.34 ppm to 0.63 ppm), that the increase in sensi-
3 3
tivity reached a maximum at concentrations of 3.6 mg/m to 4.8 mg/m (1.3 "ppm to 1.7 ppm), and
that further increases in the sulfur dioxide concentration resulted in progressive lowering of
eye sensitivity to light unti
that of the unexposed subject.
In exposures during light adaptation, sulfur dioxide concentrations of 0.6 mg/m to 7.2
0
mg/m (0.23 ppm to 2.8 ppm) caused slight increases in eye sensitivity. Maximum sensitivity
was attained at 1.5 mg/m (0.52 ppm); at higher concentrations the increased sensitivity began
to abate. Two human subjects were used in these experiments. The odor threshold was between
33 33
2.5 mg/m and 3.0 mg/m (1.0 and 1.1 ppm) for one subject and between 3.0 mg/m and 3.6 mg/m
(1.1 and 1.4 ppm) for the other, so that changes in sensitivity to light during dark adapta-
tions were caused by sulfur dioxide concentrations below the odor threshold.
Shalamberidze (1967) investigated the effects of S0« and NO,, singly and in combination,
on visual light sensitivity as determined by measures of dark adaptation. According to this

13-5
eye sensitivity to light until, at 19.2 mg/m (7.3 ppm), the sensitivity was identical with
-------
report, SOy concentrations of 0.6 mg/m (0.23 ppm) and higher caused "a considerable increase
in the ocular sensitivity to light" (Shalamberidze, 1967). So few details on methods or re-
sults were presented, however, that this report cannot be accepted without reservations.
13.2.2.3 Interruption of Alpha Rhythm—The electroencephalogram is a composite record of the
electrical activity of the brain recorded as the difference in electrical potential between
two points on the head. In the resting adult, the electroencephalogram characteristically
shows a fairly uniform pattern of electrical frequency from 8 to 12 Hz (alpha-rhythms) in the
posterior head regions. Variations occur with age, the state of wakeful ness and attentiveness,
or as a result of incoming sensory stimuli from exteroceptive or interoceptive receptors. The
dominant frequency (a) is inhibited or attenuated by eye opening and by mental activity.
Subjects with well-defined crrhythms studied in a silent and electrically shielded cham-
ber show a temporary attenuation of the a-rhythm each time they are given a light signal.
When the light is excluded, the a-rhythm returns to normal. A concentration of test gas is
determined which is so low that by itself it does not cause attenuation of the
-------
13-2. RESPIRATORY EFFECTS OF SO,
CO
i
Concentration
SOZ (ppm)a
HEALTHY ADULT
10,15,25,50
9-60
5, 10
20
1-80
1-45
1-8
1(1-2), 5(4-7)
13(10-16)
1(1-2), 5(4-6)
15(14-17)
4-6
4-5
Duration of
exposure (fflins)
SUBJECTS AT REST
60
5
10
10
10 - 60
10
10
10 - 30
30
10
10
Oral (0) or
Number of nasal (N) Rest (R) or
subjects exposure exercise (E)*
1
10 Mask
18 0, N
6 0
8-12 Mask, chamber
N
46 Mask
14 Mask
11 0*
12 0*
7 0
5 0*
-
R
R
R
R
R
R
R
R
R
R
Effects'1
Mucociliary activity decreased
at higher cone. (>15 ppm SO.)
Airway resistance increased
Mo changes in pulse rate,
resp. rate or tidal vol.
(5, 10 ppm). Bronchospasm
in two subjects at 10 ppm
Br onchoconstH ct 1 on
above 5 ppm
Decreased peak flow,
decreased expiratory
capacity at fc 1.6 ppm
Pulse and respiratory rates
increased; tidal volume
rate decreased at £1.0 ppra
Pulmonary flow resistance
for groups increased 39%
at 5 ppm and 72% at 13 ppm.
At 1 ppm, one subject had 7%
increase in flow resistance,
another a 23% decrease
Increases in Rl (pulmonary
flow resistance) at §5 ppnt SOg
Airway conductance decreased
39%. Blocked by atropine
Increased respiratory and
Reference
Cralley,
Nakamura
Lawther,
Sim and
Tomono ,
Aidur et
1953
Frank et
Frank et
Nadel et
1942
, 1964
1955
Rattle, 1957
1961
al.,
al., 1962
al., 1964
al., 1965
Abe, 1967
inspiratory resistance
"Mouthpiece
-------
TABLE 13-2. (continued)
CO
I
00
Oral (0) or
Concentration Duration of Number of nasal (H) Rest (R) or
SO^ (ppn>3 exposure (mins) subjects exposure exercise (E)
5-30
1
3
Effects
Reference
15, 28 10 8 0,N R

1.0, 5.0 2-4 hr/d 15 Chamber (N) R
and 25.0
Pulmonary flow resistance
increased less with N
breathing
Significant decreases in
expiratory flow and FEV, „
Speizer and Frank,
1966a

Andersen et al. , 1

270 hr 16 controls Chamber (N)
16 exposed
120

10
60/DB
DB
2.5, 5.0, 10.0
0.5, 1.0, 5.0
1.1 - 3.6
0.50
Mouthpiece
DB = deep breaths
10
15
30
180
10 CO, stimulus (0)* R
13/12 Chamber N/0 R
17
0*
15
10
0*. N
0*
0, H*
0*
40 Chamber (0)
Nose clips
at 25 ppm. Decreased nasal
mucus flworate at 6 5 ppm.
Responses greater after 4 hr
than after 2 hr

50% decrease in nasal mucus
flowate but number of colds
similar in both groups

No effect on mucus transport

For group as whole (12 sub-
jects) small but significant
(14%) increase in SR follow-
ing 25 DB by air alone and 26X
increase after 25 DB SO™ at 1 ppw;
but no changes detected after
normal quiet breathing of 1-3 ppm
so2

Greater percentage decrease In
in SG with 0 breathing at
all concentrations
Andersen et a1., 1977
Wolff et al., 1975a

Lawther, 1975
Melville, 1970
Decreases in MEF
group were sig,
„. for
1 and 5
ppm ,; a ppm, ecreases
for N not sig. different from
0 breathing

Deep breathing produced no
effects

No pulmonary effects seen
Snell and Luchslnger,
1969
Burton et al., 1969

Jaeger et «1., 1979
-------
TABLE 13-2 (continued)
CO
I

Concentration Duration of Kumber of
S02 (ppm) exposure (rains) subjects
EXERCISING HEALTHY ADULTS
5 0 120
5.0 120
5.0 120
5.0 3
3.0 3
1.0 3
0.5 3
0.75 120
0.75 120
0. 50 120
0.40 120
0.40 120
0.37 120
0.37 120
RESPIRATORY DISEASE SUBJECTS
0.3, 1.0 and 96 - 120
3.0 hr
7.7 6 d
0.3-4 6-7 d
10
11
10
10
8+9
10+8
5
4
15 controls
16 exposed
24
9
11
8
4-12
12 (normal)
7 (COPD)
32 normals
27 subjects
w/obstrutive
Oral (0) or
nasal (N)
exposure
0
Chamber
Chamber (0)
0
0
0
0
Chamber
Chamber
Chamber
Chamber
Chamber
Chamber
Chamber
Chamber
Chamber (N)
Chamber (N)
Rest (R) or
exercise (E)
E
E
E
R
R + E
R + E
R
E
E
E
E
E
E
E
R
R
R
Effects
Increased tracheobronchial
clearance
Insignificant changes in
airway resistance and
arterial P02
HMFR decreased 8.5%; increased
tracheobronchial clearance
Light exercise potentiates
effect of SO, MEF4Q%
decreased at 3 ppm ana above
Decrease in MMFR, FVC, FEVj 0
(-8-10%) and 20% in MEFR5g%
Significantly elevated Raw
and trend toward decreased
FEF5Q and FEV/FVC after S02
exposure during heavy exercise
No pulmonary effects seen
with 0.50 ppm SO,, + 05 ppm
No pulmonary effects
No pulmonar/ effects seen
with 0.4 ppm S02 alone
No pulmonary effects
No pulmonary effects
No difference in response
between groups Slight
decrease in pulmonary
compliance but of question-
able significance
No significant changes in
airway resistance or other
effects in health subjects
Reference
Wolff et al , 1975b
von Neiding et al , 1979
Newhouse et al , 1978
Kreisman et al. , 1976
Bates and Hazucha, 1973
Stacy et al., 1981
Linn et al. , 1980
Horvath and Folinsbee, 1977
Bedi et al., 1979
Bedi et al., 1981
Bates and Hazucha, 1973;
Hazucha and Bates, 1975
Bell et al. , 1977b
Weir and Bromberg, 1972
Reichel, 1972
resp disease
or patients
-------
TABLE 13-2. (continued)
CO
I
o

Concentration Duration of
SO, (ppm) exposure (dins)
ASTHMATIC SUBJECTS
1, 3, 5 10

1.0 5
0.1, 0.25, 0.5 10

Number of
subjects

7 normals
7 atopies
7 asthmatics

6 asthmatics
7 asthmatics
Oral (0) or
nasal (H) Rest (R) or
exposure exercise (E)

0** R

0* £

Effects

SR increased significantly
at all cone for asthmatics;
only at 5 ppm for normals and
atopic subjects. Some asth-
matics exhibited marked
dyspnea requiring bronchodila-
tion therapy.
SR significantly increased
Tn the asthmatic group at

Reference

Sheppard et al . , 1980

Sheppard et al . , 1981

0.50

0.5

0 25, 0 5

0.30
180

120
40 Chamber (0) R
(asthmatics) Nose clips
5 asthmatics
24 asthmatics
19
(asthmatics)
0*
Chamber
Chamber
0.5 and 0.25 ppm S02 and at
0.1 ppm in the two most re-
sponsive subjects. At 0.5 ppm
three asthmatics developed
wheezing and shortness of
breath.
MMFR significantly decreased
2J%; recovery within 30 rain.

Specific airway resistance
(SR ) increases were ob-
served over exercise base-
line rates for 80% of the
subjects.

No statistically significant
changes in forced vital
capacity (FVC) or specific
airway resistance (SRa,J
3W

No pulmonary effects seen
with 0.3 ppn S0» and 0.5
ppm NO, exposure compared
to exercise basline
Jaeger et al., 1979
Linn et al., 1982
Linn et al., 1982
Linn et al., 1980
a0.1 ppm S02 s 262 pg/m3 1.0 ppm s 2620 (jg/m3 10 ppm s 26,200 |jg/m3
0.5 ppm S02 s 1310 ug/«3 5.0 ppn s 13,100 |jg/n3 50 ppn s 131,000 ug/m3
Significant increase or decreases noted here refer to "statistically significant" effects, independent of whether the
observed effects are "medically significant" or not
Chrome obstructed pulmonary disease
*Mouthpiece OB = deep breaths
-------
Sim and Rattle (1957) performed extensive clinical studies over a 10-month period on an
unspecified number (8 to 12) of "healthy males aged 18 to 45." Sulfur dioxide was adminis-
3
tered either by face mask at concentrations ranging from 1.34 to 80 ppm (3.51 to 210 mg/m )
for 10 minutes or in an inhalation chamber at concentrations of 1.0 to 23.1 ppm (2.6 to 60.5
mg/m ) for 60 minutes. Regardless of exposure route, the only notable effects of SO, were
said to be bronchoconstriction (increased resistance to air flow) at concentrations 5 5 ppm
3 3
(13.1 mg/m ) SO, and high-pitched chest rales at exposure > 49 ppm (128 mg/m ) SO,. They also
3
reported that when ammonia (no value given) was also present in the chamber (9.9 ppm [26 mg/m ]
SOp), the subjective impressions of bronchoconstriction disappeared.
Frank et al. (1962) examined the effects of acute (10 to 30 minute) exposures to S02 via
mouth in 11 healthy adult subjects. Each subject received approximately 1, 5, and. 13 ppm (2.6,
13 and 34 mg/m ) of the gas in separate exposures at least 1 month apart. The only statisti-
cally significant effects were a 39-percent increase (p <0.01) in pulmonary flow resistance at
3 3
5 ppm (13 mg/m ) and a 72 percent increase (p <0.001) at 13 ppm (34 mg/m ). Only one subject
2
showed a statistically significant increase with 1 ppm (2.6 mg/m ) SOp concentration; his con-
trol resistance was the highest encountered. The recovery of some subjects was complete with-
in a few minutes. As in Sim and Rattle's study (1957), other cardiovascular or pulmonary
measures did not show any statistically significant effects.
Tomono (1961) tested 46 men for the effects of SO, on their pulmonary physiology. The
3
subjects inhaled 1 to 45 ppm (2.6 to 118 mg/m ) SO, through face masks for 10 minutes. De-
creases in expiratory capacity and peak flowrate were proportional to the concentration of
2
SO,. Such effects were detected at a concentration as low as 1.6 ppm (4.2 mg/m ). Slight in-
creases in pulse and respiration rates were observed in about 10 percent of the subjects but
were not proportional to SO/, exposures. Nakamura (1964) exposed 10 adult subjects, each to a
3
different concentration of SO, (9 to 60 ppm [23.6 to 157 mg/m ]) for 5 minutes. Airway resist-
ance increased an average of 27 percent. Since each subject was exposed to only one concentra-
tion of SO, and there was considerable variability in response to the different concentrations,
the significance of those isolated findings may be questioned. No statistically significant
correlation between dosage and response was discovered. For example, one subject had a 17-
percent increase after exposure to 9 ppm (23.6 mg/m ), another 9 percent after exposure to 16
ppm (41.9 mg/m ), another 75 percent after exposure to 47 ppm (123 mg/m ), and another 22
q
percent after exposure to 57 ppm (149 mg/m ).
Snell and Luchsinger (1969) also found statistically significant decreases in pulmonary
function consequent to SO, exposure of healthy adults. Nine subjects inhaled SO, through a
3
mouthpiece at concentrations of 0.5, 1.0, and 5 ppm (1.3, 2.6 and 13.0 mg/m ) for 15 minutes
each, with 15-minute control periods interspersed. Maximum expiratory flow (MEFcno/ wr) was
Q DU/o VVj
significantly lower after exposure to 1 ppm (2.6 mg/m ) SO, (p <0.02) as well as 5 ppm (13.0
3
) (P <0.01). Reichel (1972) exposed 32 normal subjects in a chamber to 7.7 ppm (20.2
13-11
-------
mg/m ) SO, continuously for 6 days. Intrathoracic gas volume and intrabronchial flow resist-
ance were not altered consequent to any of the 6 days of exposure. Airway resistance was mea-
sured in 16 of these subjects after inhalation of 3 percent acetylcholine chloride solution.
The sensitivity response to this challenge was not altered as a consequence of the exposure to
S00. Jaeger et al. (1979) exposed 40 normal non-smoker subjects for 3 hours to 0.5 ppm (1.3
3
mg/m ) SOo. Oral inhalation was forced by having the subjects wear nose clips. These resting
subjects were also studied during exposure to ambient air having an average SO, content of
3
0.005 ppm (13.1 ug/m ). Three pulmonary function tests (VC, FEV, and MMFR) were performed at
intervals during the exposure, and a more intensive series of tests was made prior to and
after the exposure. No pulmonary changes were observed, although one subject (a probable
asthmatic) complained of wheezing during the night following the SCU exposure; these symptoms,
however, are not clearly attributable to the SO^ exposure experienced earlier in the day.
Nadel et al. (1965) helped to elucidate the mechanism of bronchoconstriction resulting
3
from SO, exposure. They exposed seven subjects to 4 to 6 ppm (10.5 to 15.7 mg/m ) SO, for 10
minutes via mouth in a closed plethysmograph. The mean decrease in specific airway conduct-
ance was 39 percent (p <0.001). After subcutaneous injection of atropine (1.2-1.8 mg), inha-
lation of SOg resulted in only a three-percent (p >0.02) decrement in specific airway conduct-
ance/thoracic gas volume. However, atropine did not affect the coughing or sensation of irri-
tation in the pharynx or substernal area. From this and other evidence, Nadel et al. con-
cluded that the bronchoconstriction induced by S0? depends on changes in smooth muscle tone
mediated by parasympathetic motor pathways. Thus, when sensory receptors in the tracheobron-
chial region are irritated by a substance such as S0?, a reflexive bronchospasm may be trig-
gered.
13.2.3.2 Water Solubility—An important point to note is that, because of its high solubility
in water, SO, is readily absorbed when it comes in contact with the moist surfaces of the nose
and upper respiratory passages (Frank et al., 1973). This has a number of important implica-
tions for the analysis of the effects of S02 on respiratory functions. These considerations
will be illustrated in the following sections (see also Chapter 11).
13.2.3.3 NasalVersus Oral Exposure—Several studies have demonstrated notable differences in
SO, concentrations necessary to elicit pulmonary effects depending upon the route (i.e , via
nose or mouth) of exposure to SO,. Speizer and Frank (1966a), for example, compared the
3
effects of 502 (10-minute exposures at 15 and 28 ppm [39.3 and 73.4 mg/m ]) in eight subjects
breathing the gas either by nose or by mouth. The subjects coughed less and reported less
irritation of the throat and chest when breathing through their noses. Also, pulmonary flow
resistance increased less during nasal exposure than during oral exposure.
A second study by the same investigators (Speizer and Frank, 1966b) refined their
analysis of these effects, using seven subjects and a specially designed face mask. Air was
13-12
-------
sampled at various points, including: (1) within the face mask before being inspired, (2)
within the subject's nose, and (3) within the subject's oropharynx. Exposures lasted 25 to 30
minutes. The average concentration of SCL within the mask was 16.1 ppm (42.2 mg/m ); within
the oropharynx the concentration was too low for the investigators' equipment to measure.
Thus, essentially all of the SCL (90 to 99 percent) in the inspired air was removed by the
nose. Similar results were obtained by Andersen et al. (1974) in a study described in detail
later.
Melville (1970) also compared oral and nasal routes of administration. He used 15 sub-
jects and exposed them (for 10 minutes) sequentially to 2.5, 5, and 10 ppm (6.6, 13.1, and 26.2
2
mg/m ) SO,. More SO, was removed per minute with nose breathing than with mouth breathing.
There was a clear dose-dependent response reflected in measures of the subjects' specific air-
way conductance (SG , not typical panting procedure). As the SO- concentration increased,
clW £.
SG decreased (p < 0.05). This was true regardless of administration route, but at 2.5 ppm
clW •-)
(6.6 mg/m ) SO- the average decrease under oral administration was greater (in 80 percent of
subjects) than the decrease under nasal administration (p <0.05). During a 1-hour exposure to
3
5 ppm (13.1 mg/m ) SO-, no statistically significant difference was observed (49 subjects) in
SG decreases after 5-minutes of mouth breathing in comparison to marked SG decreases after
nasal breathing and no further decrements in SG beyond the levels reached after 5 minutes
QW
exposure were observed with continuation of exposure up to 1 hour. During mouth exposures,
some subjects coughed at the onset of SO- exposure and complained of burning sensations in the
throat and substernal chest pains. Three returned a week later complaining of substernal
pains and bronchial infections.
Snell and Luehsinger (1969) also examined the differences between nasal and oral exposure
Q
using SO- at 5 ppm (13.1 mg/m ). Five subjects' average maximum expiratory flow (MEF50y „„)
was 10 percent lower following oral exposure than following nasal exposure. This difference,
however, was not statistically significant. See Chapter 11 for further discussion of S0~
deposition.
13.2,3.4 Subject Activity Level—One practical implication of the above findings is that
vigorous activity, such as heavy exercise or work, may significantly affect the actual dose
received by a person during exposure to SO-. Several studies indicate that at some level of
ventilation inhalation of air shifts from nasal to oronasal breathing and, also, that some in-
dividuals may always be oronasal breathers, even at rest. Saibene et al. (1978), for example,
studied 63 subjects while they exercised at increasing workloads. Incomplete information was
obtained on 13 subjects. Ten subjects were observed to breath through the mouth at all work-
loads, while five never opened their mouths. In the remaining 35 subjects, the highest minute
volume attained with nasal breathing was 40.2 liters per minute. Determination of the shift
from nasal to oronasal breathing was obtained by subjective observation. In a second study
using 10 subjects, ventilation was more precisely measured (but still not by a completely
13-13
-------
adequate technique), as determined by movements of the rib cage. The mean value of ventila-
tion at the point of shift to oronasal breathing was 44.2 liters/im'n.
Niinimaa et al. (1980, 1981) subsequently reported data indicating that: (1) some sub-
jects are essentially oral breathers even at rest; (2) a switch from nasal to oronasal breath-
ing occurs at Ve of 35.3 + 10,8 (mean ± standard deviation) liters per minute; and (3) after
the switch to oronasal breathing by persistent nasal breathers (at rest), the nasal portion of
Ve decreased to 57 percent of total Ve. With further increases in Ve, oral minute volume in-
creased rapidly. Other studies (D1 Alfonso, 1980) suggest that subjects who are nasal breathers
at rest move to oronasal breathing when minute volume is approximately 30 liters. It should
be noted that some air may continue to enter the lungs through the nose after the shift to
oronasal breathing, but the nasal volume is substantially reduced (Niinimaa et al . , 1980,
1981; D'Alfonso, 1980). Niinimaa et al., for example, reported that as much as 40 to 50
percent of inhaled air entered via the nose after the shift to oronasal breathing. The full
implications of these studies on oronasal breathing for understanding the distribution and
effects of inhaled pollutants under various activity conditions await further investigation,
but they appear to establish that some individuals utilize oronasal breathing even at rest and
wide variability exists in regard to Ve levels at which nasal breathers at rest shift over to
predominantly oral breathing under increased exercise conditions (i.e., across the range of Ve
~ 15 to 45 liters/minute).
In regard to the impact of increased activity levels on SOp effects, Kreisman et al.
(1975) reported that exercise may potentiate the effect of S02 on respiratory function. In
their study, subjects inhaled a mixture of S02 in air for 3 minutes while exercising on a
bicycle ergometer at a pace sufficient to double their resting minute ventilation rate. Eight
3 3
subjects received 1 ppm (2.6 mg/m ) S09 and nine subjects received 3 ppm (7.9 mg/m ). Those
3
receiving 3 ppm (7.9 mg/m ) showed a significant (p <0.05) decrease in maximal expiratory flow
/p-.) compared to a control (untreated air) exposure. However, it is not clear that
this change differed significantly from the change in MEF4f,n, .-„>. occurring in resting subjects.
Bates and Hazucha (1973) reported approximately a 20 percent decrease in maximal expiratory
flow rate C^EFR™^) and approximately an 8 to 10 percent decrease in other pulmonary function
measurements (FVC, FEV, „, MMFR) in 4 subjects (who exercised intermittently during the
exposure) exposed in a chamber containing 0.75 ppm S09; however, these differences were not
£. .,
statistically significant. Unlike these studies at 0.75 ppm (2.0 mg/m ) SO,,, subsequent
3
chamber studies by Hazucha and Bates (1975) at 0.37 ppm (0.97 mg/m ) S02 demonstrated no
decrements in any of the above pulmonary function parameters. Horvath and Folinsbee (1977)
and Bedi et al. (1979) exposed nine intermittently exercising subjects in a chamber to 0.4 ppm
(1.0 mg/m ) SO, and found no pulmonary function changes. Also, more recent studies of
exercising adults by Horvath et al . (1981) found neither significant pulmonary function
effects with 0.4 ppm S0? alone nor any enhancement of ozone effects by 0.4' ppm (1.0 mg/m )
13-14
-------
S09. Analogously, Linn et al. (1980) reported no pulmonary function changes in 24 exercising
33
healthy adults with exposure to 0.5 ppm (1.3 mg/m ) S0? plus 0.5 ppm (0.9 mg/m ) N0? in
comparison to exercising baseline control values obtained in an open chamber.
In a recent study by Stacy et al . (1981), sixteen young healthy males were exposed in a
3
chamber for 2 hours to 0.75 ppm (2.0 mg/m ) SCL. During the last 15 minutes of the first hour
of SO- exposure, they exercised at a rate sufficient -to 'increase their ventilatory volumes to
60 liters/minute and SQp effects on pulmonary function at the end of the first hour were com-
pared against nonexercise SO,, exposure baseline test resuts. Airway resistance was signifi-
cantly increased (mean = 14.6%; range = 2 to 55%) following the exercise period in 14 of the
16 subjects. Although this was the only statistically significant change among the 15 pulmo-
nary measurements made, there was also a trend toward decreased FEFj- and FEV,/FVC levels.
Half of the SOp-exposed subjects with one or more positive allergen skin tests appeared to be
significantly more reactive to S0?. Fifteen other subjects were utilized as controls (i.e.,
being exposed to filtered air and having their data compared to that obtained on the SCL-
exposed subjects). Subjects reacting positively to seven or more skin allergen or metacholine
tests were excluded from the study. The statistically significant changes in R and notable
trends observed by Stacy et al. (1981) in other pulmonary measures appear to imply a broncho-
constriction response occuring in healthy adults (especially atopic subjects without asthma)
at 0.75 ppm SOp exposure level under heavy exercise conditions.
Lawther et al . (1975) have demonstrated that simply instructing 12 subjects to take 25
deep breaths by mouth resulted in a statistically significant (p <0.001) increase in specific
airway resistance (SR ) during exposure to air alone (14%) and an overall increase of 26% for
aw~
SO- at 1 ppm (2.6 mg/m ). While sitting quietly in an inhalation chamber, the same subjects
had previously shown no such increase after breathing concentrations of 1 to 3 ppm (2.6 to 7.9
mg/m ) S09 for an hour. As part of a series of experiments in this study, 17 subjects also
3
received 3 ppm (7.9 mg/m ) SOp by a mouthpiece and were instructed to take 2, 4, 8, 16 and 32
deep breaths at 5-minute intervals. Increases in SR due to SO, were significantly greater
9W £.
after 16 (p <0.01) or 32 (p <0.001) deep breaths.
Burton et al . (1969), however, found no consistent effects in 10 subjects exposed to S09
3
at 1.1 to 3.6 ppm (2.9 to 9.4 mg/m ) for 30 minutes, regardless of whether the subjects
breathed normally or at a forced hyperventilation rate of up to 2.5 liters/second. One other
difference between these two studies was the duration of exposure. Burton et al. (1969) ex-
posed their subjects for 30 minutes, whereas, Lawther et al . (1975) maintained exposures for
an hour. This raises another important consideration in reviewing the effects of SOp on human
subjects, namely, temporal parameters.
13.2.3.5 Temporal Parameters — Early studies (e.g., Lehman, 1893) suggested that workers
chronically exposed to relatively high concentrations of SO, were less conscious of its pres-
ence in the atmosphere than persons not as familiar with the gas. However, the data of Holmes
13-15
-------
et al, (1915) indicated that subjects already accustomed to S02 could detect its odor at lower
concentrations than could persons unaccustomed to it. Nevertheless, it would seem plausible
that "self-selection" would tend to reduce the number of relatively sensitive persons among
the population of workers chronically in contact with supra-threshold S0? levels.
As previously noted, a study by Frank et al. (1962) indicated that subjective reports are
not a reliable indicator of physiological responses in any event. After 5 to 10 minutes of
exposure to either 5 or 13 ppm (13.1 or 34.1 mg/m ) SCL, changes in their subjects' pulmonary
resistance measures were just reaching their peaks, while subjective reports of an odor of S0?
had already subsided.
In a later study by Frank et al. (1964) the increase in pulmonary resistance induced by
SCL peaked at about 10 minutes and then gradually decreased over the next 15 minutes. This
finding corresponds closely to Sim and Rattle's (1957) report that, if lung resistance
increased at all in individual subjects, the increase occurred within the first 10 minutes.
Abe (1967) also investigated the temporal course of induction of S0«-exposure effects.
3
His five mouth-breathing subjects were given 4 to 5 ppm (11 to 13.1 mg/m ) SO • immediate sig-
nificant (p <0.05) increases in expiratory resistance (42 percent) and inspiratory resistance
(25 percent) were observed.
Similar short-term responses (within 5 to 10 minutes after the start of exposure) have
been recorded by other investigators. Melville (1970) found in 49 mouth-breathing subjects
that percentage decreases in specific airway conductance (SG ) were greatest during the first
clW r\
5 minutes of up to 60 minutes of exposure to S0?. At 5 ppm (13.1 mg/m ), for example, he
noted that SG decreased significantly (p <0.05) within 5 minutes of exposure but then stabi-
lized at slightly higher levels that, however, remained significantly lower (p <0.05) during
the rest of the one-hour exposure than levels recorded under control conditions of no S02.
Lawther et al. (1975) also noted that SR increased most during the first 5 minutes of
2W
exposure. However, if exposure was terminated after another 5 minutes, recovery to baseline
levels generally occurred in about 5 minutes, although 3 "S0--sensitive" subjects out of a
total of 14 took 10 to 65 minutes to recover from higher exposure levels (up to 30 ppm [78.6
2
mg/m ] S0?) after returning to a clean air environment.
Longer term effects (over a period of hours) have been reported by Andersen et al.
(1974), who investigated nasal mucus flow rates as well as airway resistance and subjective
responses. Nasal mucociliary flow was measured by placing a radioactive-labeled resin par-
ticle on the superior surface of the inferior turbinate and tracking its position with a slit-
coll imator detector. Fifteen subjects were exposed via an inhalation chamber to increasing
3
concentrations (1, 5, and 25 ppm [2.6, 13.1 and 65.5 mg/m ]') of S02 for approximately 6 hours
per day over 3 consecutive days. Baseline measurements were made under conditions of filtered
air on a day prior to experimental exposures. This study found a number of effects reaching
their maximum after 1 to 6 hours of exposure. Nasal cross-sectional airway area generally
decreased throughout the 6-hour daily trials, but the decreases were only significant (p <0.05)

13-16
-------
3 3
at 1 ppm (2.6 mg/m ) and 5 ppm (13.1 rng/m ), since there was an overall drop in this measure
(approaching a "floor level") by the time 25 ppm (65.5 mg/m ) was administered on the third
day of the study. Nasal airflow resistance increased significantly (p <0.05) with the 6-hour
exposure to each concentration (1, 5, 25, ppm [2.6, 13.1, 65.5 mg/m ] SCL). Significant (p
<0.05 or less) decreases in forced expiratory flow (FEF?C. 7,-,y) and forced expiratory volume
(FEV-, ~) also occurred both within daily exposures and across days (i.e., increasing concen-
trations), although the within-day decrease in FEV, n was only significant on day 3 at 25 ppm
3
[65.5 mg/m ] (see Andersen et al., 1974, Figure 7).
13,2.3.6 Mucociliary Transport—Cralley (1942) investigated mucociliary clearance before
sophisticated radioactive measurement techniques were available. A drop of red dye was placed
in the active ciliary region of the inferior meatus of a volunteer subject. The rate of mucus
clearance was reflected by the time between introduction of the dye and its appearance in ex-
q
pel led mucus. Exposure to SCL at 10 to 15 ppm (26.2 to 39.3 mg/m ) for 60 minutes produced
only a small decrease in the rate of mucus removal. A 30- to 60-minute exposure resulted in a
50 percent reduction in mucociliary transport at 25 ppm (65.5 mg/m ) SCL and a 65 to 70 per-
3
cent reduction at 50 ppm (131 mg/m ).
In another study (Anderson et al., 1974), mucostasis in the anterior region of the nose
was observed in 14 of 15 subjects after 4 to 5 hours of exposure on three successive days to
2
1, 5, and 25 ppm (2.6, 13.1 and 65.5 mg/m ) SCL, respectively. There appeared to be no carry-
over effect from the exposure of the previous day. Mucus flow rates on the first day of ex-
posure (to 1 ppm [2.6 mg/m ] S0?) tended to be lower but were not significantly lower than
those observed on the control day (0 ppm). Mucus flow rates were significantly lower on the
second day (5 ppm [13.1 mg/m ]) and were further decreased on the third day (25 ppm [65.5
mg/m ]) of exposure. The subjects noted discomfort only on the second and third day expo-
sures. At these concentrations some subjects also had sporadic mucostasis, although there
were pronounced individual differences in these measures even at baseline. Andersen et al.
(1974) calculated the cross-sectional area of the nasal airways. A statistically significant
decrease (0.02
-------
pace to yield heart rates 70 to 75 percent of estimated maximum values. Exposure in this
study lasted for 2.0 to 2.5 hours. The combination of exercise and exposure (via mouth) to 5
3
ppm (13.1 mg/m ) SO- resulted in a significantly (p <0.05) greater rate of tracheobronchial
mucociliary clearance. This result contrasts with the findings of Andersen et al. (1974) that
2
nasal clearance rates were reduced by exposure to 5 ppm (13.1 mg/m ) SOp. However, the differ-
ence between the two studies can probably be explained on the basis of dose. Dose to the lung
would be much lower than to the nose because of the absorption of S02 by upper airway mucosal
surfaces. Therefore, effects on the lung could be typical of lower concentrations and in-
creases might be anticipated as seen for low levels of H-SO*. Of course, the two studies
focused on different regions of the respiratory tract (tracheobronchial versus nasal), and
this in itself could account for these contrasting effects (Albert et al., 1969; Frances et
al, 1970). Both of these investigators replicated their findings in later studies (Andersen
et al., 1977; Newhouse et al., 1978).
These studies were extended by Newhouse et al. (1978), whose 10 subjects breathed either
3
S09 (5 ppm) or H,SO,, mist (1 mg/m ) delivered as an aerosol of 0.58 (jm MMAD, An aerosol con-
99m
taining a 0.025 percent solution of Tc-albumen was inhaled prior to pollutant exposure.
The bolus technique (exposure to short-term peak, concentrations) employed achieved deposition
of the aerosol, primarily in the large airways. One-half hour later the subjects were exposed
to the pollutants. They immediately exercised for the next 0.5 hour. A total of 20 minutes
of exercise at approximately 70 to 75 percent of predicted maximum heart rate was performed,
followed by an additional 1.5 hours of rest exposure. The subjects breathed through the mouth
to eliminate nasal ventilation and absorption of pollutants. Pulmonary function tests con-
ducted at the end of 2 hours exposure to SQp indicated no changes in FVC or FEV-, Q but maximum
midexpiratory flow rate (MMFR) decreased 8.5 percent, possibly due to a reflex bronchocon-
striction. No pulmonary changes were found consequent to the HpSQ* mist exposures. Tracheo-
bronchial clearance increased in both SO, (6 of 10 subjects) and H,SO. (5 of 10 subjects) expo-
sures. The investigators did not present their data in a manner which would provide informa-
tion as to the relationship between clearance rates and MMFR. It should be noted that these
data are in contrast to the replicated observations by Andersen et al. (1977), who showed a
3
slowing of nasal clearance on exposure to 5 ppm (13.1 mg/m ) S0?. It is possible that time-
dependent events (or subject selection) might explain the different results; or differences in
the respiratory tract regions studied (tracheobronchial versus nasal) may account for the ap-
parent inconsistency.
Mucociliary transport is a significant aspect of the respiratory system's defense against
airborne agents. A disturbance in this function might have important implications for a
number of health effects, such as susceptibility to cold-virus infections. Andersen et al.
(1977) evaluated this possibility by inoculating volunteers with a strain of rhino-
virus (RV3). The basic design of the study and reactions of the subjects are shown in a
table (Andersen et al., 1977). Although there was no difference in the number of colds that

13-18
-------
developed in the two groups of subjects (all nose breathers), cold symptoms were judged (under
a double-blind procedure) to be less severe (p <0.05) in the group exposed to SOp- It was un-
known, however, whether this result reflected a direct effect of S0? on the host, the rhino-
virus, or both. In addition, the average incubation period was somewhat shorter for the group
exposed to S0? (p <0.06). Virus shedding (a measure of infection determined from nasal wash-
ings) also seemed to be somewhat decreased in the SCL-exposed group but not significantly.
S0? may affect mucociliary clearance via direct effects on secretory or ciliated cells or
indirectly through S0? effects on neural reflexes involved in clearance control. Related to
the latter, Camner et al. (1974, 1976) evaluated the role of the autonomic nervous system in
the regulation of tracheobronchial clearance in man. They employed a test aerosol of 6 urn
Teflon particles tagged with technetium 99m. Their first study (1974) indicated the basic
mucociliary transport was not under parasympathetic control. However, activation of the vagus
did induce increases in transport velocity. Their second study (1976) suggested that the reg-
ulation of the mucociliary transport rate was under sympathetic control and blood concentra-
tions of catecholamines might be important in modulation of mucociliary transport. This is in
contrast to the likelihood that parasympathetic cholinergic neural mechanisms are of key impor-
tance in mediation of S0?-induced bronchoconstriction effects.
13.2.3.7 Health Status—Apart from fairly consistent bronchoconstriction effects, a common
element in the above and other reports of the effects of S0? has been the notable variability
among subjects in their responses to such exposures. In the study of Frank et al. (1962), for
example, 9 of 11 subjects showed no effects at 1 ppm, but 1 subject showed a significant (p
<0.01) decrease in pulmonary flow resistance, whereas the remaining subject showed a signifi-
cant (p <0.01) increase. Sim and Rattle (1957) reported that they themselves appeared to have
developed an increasing sensitivity to S0? encountered during the course of their research,
especially upon initial contact with the gas. The persistent and uncomfortable spells of
coughing and wheezing upon contact with the gas appeared to increase with repeated exposures.
Other investigators (e.g., Burton et al., 1969; Frank, 1964; Nadel et al, 1965; Lawther, 1955;
Lawther et al., 1975; Jaeger et al., 1979) have reported "hyperreactors" among their subjects.
Indeed, some investigators have suggested that about 10 percent of the total population is
made up of especially sensitive persons (Amdur, 1973, 1974; Horvath and Folinsbee, 1977). How-
ever, in at least one instance (Andersen et al. , 1974), a subject's response was exaggerated
even under control conditions, which raises the possibility of psychological factors contribut-
ing to this observed sensitivity.
Reports of the prevalence of asthma within the population vary with the criteria used to
identify asthmatics, differences in study populations, and other factors. Data from the 1970
Health Interview Survey indicate the prevalence of asthma to be about 3 percent of the popu-
lation. However, Dodge and Burrows (1980) reported that, in a sample of white (non-Mexican)
individuals living in Tucson, Arizona, some 6.6 percent had physician-diagnosed asthma and
13-19
-------
another 9 percent had spontaneous attacks of wheezing and shortness of breath. Another 10 per-
cent had wheezing without colds, and 10 percent had wheezing with colds. The generally cited
figure of 8-15 percent of the population having asthma may, therefore, be too low. It should
be recalled that most investigators of cinical studies have reported that 10 to 20 percent of
their "normal" subjects have shown unusual sensitivity to S0? exposure.
Some studies have considered the preexisting health status of subjects as a variable in
assessing the physiological effects of SO,. Weir and Bromberg, for example, conducted sepa-
rate studies on 12 healthy subjects (Weir and Bromberg, 1972) and on 1 smokers who showed
early signs of chronic obstructive pulmonary disease (Weir and Bromberg, 1973). The subjects
were exposed to 0," 0.3, 1, and 3 ppm (0, 0.8, 2.6 and 7.86 mg/m ) SO- in an inhalation chamber
for 96 or 120 hours (smokers or nonsmokers, respectively), with several days separating each
trial. In non-smoking, healthy subjects there was a significant, but reversible effect on the
pulmonary function of adult males at 3 ppm S0? exposure for 120 hours. The individual vari-
ablity among the smokers in their daily lung functions was so great that no effects could be
attributed to SQp exposure. Also, subjective complaints appeared to be randomly distributed
throughout the course of the study and could not be related to SQp exposure levels.
Gokenmeyer et al. (1973), however, reported that bronchitic patients exposed to 10 ppm (26.2
mg/m ) S0? had maximal changes in SR at the end of a 3-tninute period of inhalation.
Recovery to control levels required some 45 to 60 minutes when subjects were returned to a
clean-air environment.
Reichel (1972) exposed two groups of subjects with obstructive bronchial disease to vary-
ing concentrations of SO, in his chamber. Patients with minor obstructive disease were ex-
3
posed continuously for 4- days to 3.81 ppm S00 (10 mg/m , n= 8), for 4 days to 1.80 ppm S09
3 3
(4.7 mg/m , n=4) and for 6 days to 0.29 ppm S00 (0.75 mg/m , n = 5). Patients with serious
3
obstructive bronchial disease were exposed to 1.52 to 1.80 ppm SCL (4 to 4.7 mg/m ) for 4 days
3
and 1 to 1.52 ppm SQp (2.6 to 4.0 mg/m ) for 6 days. Airway resistance was not influenced by
such exposure. The details of the measuring procedures were not adequately presented in his
report.
Gunnison and Palmes (1974) compared 7 heavy smokers and 13 non-smokers with respect to
blood plasma levels of S-sulfonate after exposure to 0.3, 1.0, 3.0, 4.2, and 6.0 ppm (0.8,
2.6, 7.9, 11.0, and 15.7 mg/m ) SCL. Both groups showed highly significant correlations
(p <0.001) between S02 concentrations and S-sulfonate levels. But there was no significant
differentiation between the two groups of subjects in this regard.
Jaeger et al. (1979) exposed 40 asthmatics (mild to moderate but with no recent exacer-
bations) for 4 hours to 0.5 ppm (1.3 mg/m ) SQp. Oral inhalation was forced by having the
resting subjects wear nose clips. Control studies were made during exposure to ambient air
o
having an average SO, of 0.005 ppm (13.1 ug/m ). The only statistically significant effect (p
3
<0.04) observed was a 2.7-percent decrease in MMFR after 0.5 ppm (1.3 mg/m ) SOp exposure.
13-20
-------
This minimal change was stated to have little physiological significance. Two asthmatic sub-
jects exhibited discomfort and audible wheezing (requiring standard asthmatic medication) dur-
ing the night following SO, exposure; again, however, attribution of these "delayed" symptoms
to the earlier SO. exposure may be questionable.
Sheppard et al. (1980) exposed three groups of seven subjects (normal, atopic, and mild
asthmatic) for 10 minutes to 0, 1, 3, and 5 ppm (0, 2.6, 7.9 and 26.2 mg/m ) SO,,. The subjects
breathed these gases orally, via mouthpieces, while their specific airway resistance (SR )
was measured in a body plethysmograph. The intermittent exposures to SO- consequent to the
methods used may have influenced the results. Several statistical approaches were utilized,
one of which may be inappropriate, but the general conclusions appear to be valid. Despite
large inter- and intrasubject variability in these subjects breathing clean air, it was found
that in asthmatic subjects SR increased significantly (p <0.05 to 0.025) at all concentra-
tions of SO,, Normal and atopic (skin sensitive to common allergens) subjects had statistic-
3
ally significant increases in SR only while breathing 5 ppm (13.1 mg/m ) SO,. Some asthma-
tic subjects exhibited marked dyspnea requiring bronchodilator therapy. The increased SR
seen in either normal or mild asthmatic subjects was prevented by treatment with atropine,
confirming the involvement of parasympathetic pathways in this response.
Sheppard et al. (1981), using 13 non-smoking mildly asthmatic volunteers (10 men, 3 women,
20 to 30 years of age), demonstrated that moderate exercise (minute ventilation ~ 30 liters)
increased the bronchoconstriction effect of SO, administered by mouth at S09 concentrations of
3
< 1 ppm (2.6 mg/m ). The first set of studies evaluated the effect of exercise on S09-induced
£
bronchoconstriction in seven subjects (six men, one woman) at concentrations of 0.5, 0.25, and
0.1 ppm (1310, 660 and 260 |jg/m ). In these subjects, inhalation of 0.50 and 0.25 ppm (1310
and 660 (JQ/m ) °f S09 during the performance of moderate exercise significantly increased SRaw,
3
whereas neither inhalation of 0.50 ppm (1.3 mg/m ) of SO, at rest nor inhalation of humidi-
3
fied, filtered air during exercise had any effect on SRaw. Inhalation of 0.50 ppm (1.3 mg/m )
during exercise significantly increased SRaw in all seven asthmatic subjects (p < 0.05), and
three developed wheezing and shortness of breath. During the corresponding period of exercise
q
alone and during inhalation of 0.50 ppm (1.3 mg/m ) at rest, SRaw did not increase in any sub-
ject. After inhalation of 0.50 ppm (1.3 mg/m ) of SO, during exercise, SRaw was significantly
greater than after exercise alone or inhalation of 0.50 ppm of S09 at rest (p < 0.05). Inha-
3
lation of 0.25 ppm (660 ug/m ) during exercise significantly increased SRaw in three of the
seven subjects, and the increase in SRaw for the group was significant (p < 0.05). No subject
developed wheezing or shortness of breath at 0.25 ppm (660 ug/m ) SO, exposure. During the
corresponding period of exercise alone, SRaw did not increase in any subject. In the two most
responsive subjects at 0.5 (1310 ug/m ) and 0.25 (660 Mg/m3) SO,, inhalation of 0.1 ppm (260
3
ug/m ) SO, significantly increased SRaw, and there appeared to be a dose-response relationship
of successively greater increases in SR as a function of increasing SO, concentrations (0.1,
0.25, and 0.50 ppm).

13-21
-------
The second set of studies, involving all six subjects (four men, two women), compared the
3
bronchoconstriction produced by breathing SO- (1 ppm [2.6 mg/m )] during exercise and during
eucapnic hyperventilation. The magnitude of the increase in SR was the same when the sub-
jects inhaled SOy while they exercised or when they performed eucapnic hyperventilation at the
same minute ventilation. In every case, the increase in SRaw was accompanied by dyspnea and
audible wheezing.
o
The bronchoconstriction produced by inhalation of 0.50 ppm (1.31 mg/m ) SCL during exer-
cise was gradual in onset. Immediately after exercise, SRaw did not differ significantly from
baseline values. It then increased over the first 3.5 minutes, reached a plateau, and gradu-
ally returned to baseline values by 30 minutes after exposure. A similar time course was seen
in those subjects who developed bronchoconstriction after exposure to 0.25 and 0.10 ppm (660
2
and 260 pg/m ) SO,. In contrast, the bronchoconstriction produced by inhalation during eucap-
nic hyperventilation was rapid in onset, suggesting that the bronchoconstriction observed dur-
ing exercise was a function of the increases in minute ventilation. These investigators util-
ized two different statistical procedures to analyze their data. It is not clear which was
utilized for each of their conclusions.
Linn et al. (1982) confirmed qualitatively the results of Sheppard et al (1980). Five
3
asthmatic subjects were exposed to 0.5 ppm (1.3 mg/m ) SO, via mouthpiece for periods of 10
minutes during moderate exercise (~ 400 kg-m/min.). Compared to baseline conditions employing
clean air plus exercise, all subjects except one showed greater increases in specific airway
resistance while breathing 0.5 ppm SO, than while breathing clean air. Also in a larger scale
followup study employing open chamber exposures, Linn et al. (1982) measured pulmonary func-
tions (airway resistance, forced vital capacity) and recorded various symptoms (cough, sputum
production, wheezing, chest tightness, substernal irritation, dyspnea, throat irritation or
congestion, headache, eye irritation, and fatigue) in 24 young adult asthmatic subjects
exposed to sulfur dioxide (0.25 ppm and 0.5 ppm) for 1 hour under intermittent periods (10
minutes) of moderate exercise (Ve ~ 27 liters/minute). None of the measurements of pulmonary
function (SR and FVC) showed statistically significant variation attributable to sulfur
9W
dioxide, although small significant increases in resistance attributable to exercise were
found. Similar results were observed by Linn et al. (1980) in earlier chamber studies
employing 19 asthmatics exercising at workload levels sufficient to approximately double their
resting minute ventilation rates (normal Ve ~ 8 to 10 liters/minute). No changes in pulmonary
functions were observed over exercise baseline rates when the subjects were exposed to 0.3 ppm
3 3
(0.79 mg/m ) S02 in combination wtih 0.5 ppm (0.9 mg/m ) NO, while exercising in an open expo-
sure chamber.
The Linn et al. (1980, 1982) chamber study results are in contrast to results obtained in
the mouthpiece exposure studies conducted by Sheppard et al. (1981) and Linn et al. (1982).
These differences are most likely accounted for by greater doses of SQ~ reaching tracheo-
bronchial regions with mouthpiece exposure than with open chamber exposures at slightly lower

13-22
-------
exercise levels. Also, individual variation in bronchial reactivity to SCL among the subjects
tested may have contributed to the contrasting results.
Several other studies of SO/, (e.g., Snell and Luchsinger, 1969; Andersen et al., 1974;
Gokenmeijer et al., 1973; Burton et al., 1968, 1969) have included asthmatic patients or
smokers but have not -provided even qualitative ratings of their health status. This alone
makes it difficult to compare the results of these to other studies using "healthy" or
"impaired" subjects. Morever, the great individual variability among both normal and impaired
persons in these particular studies makes it difficult to reach any conclusions based on their
results regarding the relative importance of an individual's health status in determining his
physiological response to S0?.
13.3. PARTICULATE MATTER
As chemical interactants, particles such as aerosols of certain soluble salts (e.g.,
ferrous iron, manganese, vanadium) may act as catalysts to convert SO/, to H/,50*. Water from
atmospheric humidity or from physiological sources figures prominently in these reactions.
The following sections deal with common compounds of sulfur dioxides and point up the influence
of a number of variables that affect human physiological response to these compounds.
13.3.1 Sulfuric Acid and Sulfates
13.3.1.1 Sensory Effects--A number of studies have been directed toward determining threshold
concentrations of H?SO» for various sensory response. In a study with 10 test subjects,
Bushtueva (1957) found that the minimum concentration of sulfuric acid aerosol (particle size
not given) which was sensed by odor ranged from 0.6 mg/m to 0.85 mg/m (average 0.75 mg/m ).
3
In tests with five subjects (Bushtueva, 1961), a combination of sulfur dioxide at 1 mg/m
o
(0.35 ppm) and sulfuric acid mist at 0.4 mg/m was below the odor threshold. Amdur et al.
(1952) reported on 15 subjects (males and females) exposed for 5 to 15 minutes to various con-
centrations of sulfuric acid mist that the subjects breathed via face masks. It was found
that 1 mg/m was usually not detected, while 3 mg/m was detected by all subjects.
Bushtueva (1957) studied the effect of sulfuric acid mist on the light sensitivity of two
test subjects. Sensitivity was measured every 5 minutes during the first half-hour of each
test, then at 10-minute intervals thereafter. A control curve was established for each sub-
ject by seven repeated tests, and then sulfuric acid aerosol was administered for 4 minutes
and for 9 minutes at the 15th and 60th minutes, respectively. With sulfuric acid mist of un-
determined particle size at a concentration of 0.6 mg/m , a just detectable increase in light
sensitivity occurred with the first exposure but not with the second. Concentrations in the
3 3
range of 0.7 mg/m to 0.96 mg/m brought about a well-defined increase in light sensitivity.
3
With 2.4 mg/m , increased sensitivity to light was elicited by the exposures at both the 15th
and 60th minutes of the test; normal sensitivity was restored in 40 to 50 minutes.
Bushtueva (1961) studied the effect of sulfur dioxide, sulfuric acid mist and combina-
tions of the two on sensitivity of the eye to light in three subjects. The combination of
13-23
-------
3 3
sulfur dioxide at 0.65 mg/m (0.23 ppm) with sulfuric acid mist at 0.3 mg/m resulted in no
change in sensitivity of the eye to light. An increase of approximately 25 percent in light
3
sensitivity resulted from exposure to either sulfur dioxide at 3 mg/m (~1.0 ppm) or sulfuric
3 3
acid mist at 0.7 mg/m . The combination of sulfur dioxide at 3 mg/m (~1.0 ppm) with sulfuric
acid mist at 0.7 mg/m resulted in an increase of approximately 60 percent in light sensi-
tivity. Exposures lasted for 4^ minutes.
-j
Bushtueva (1962) demonstrated that combinations of sulfur dioxide at 0.50 mg/m" (0.17
3 3
ppm) with sulfuric acid mist at 0.15 mg/m or sulfur dioxide at 0.25 mg/m (0.087 pprn) with
sulfuric acid mist at 0.30 mg/m could produce electrocortical conditioned reflexes. There
are some uncertainties regarding this study.
Bushtueva (1961) studied the effects of different concentrations of sulfur dioxide, sul-
furic acid mist, and combinations of the two on the optical chronaxie of three subjects.
Optical chronaxie was determined in each test subject at 3-minute intervals as follows: at
the start and on the 3rd, 6th, 9th, 12th and 15th minutes. Between the 6th and 9th minutes
the subjects inhaled sulfur dioxide, sulfuric acid mist, or their combination for 2 minutes.
In each subject, the threshold concentrations of sulfur dioxide and sulfuric acid mist were
first determined independently, and then threshold concentrations for combinations of the two
were determined. Sulfuric acid mist (0.75 mg/m ) increased optical chronaxie.
13.3.1.2 Respiratory and Related Effects—Studies investigating respiratory and related
effects of human exposures to sulfuric acid under controlled conditions are summarized in
Table 13-3. Amdur et al. (1952) found respiratory changes in all subjects exposed for 15
3 3
minutes to HpSO, aerosol at concentrations of 0.35 mg/m to 5 mg/m . Vapors from an electric-
ally heated flask containing concentrated sulfuric acid were carried by compressed air into
the main air stream and then into a lucite mixing chamber, delivering a mist with a mass
median diameter of 1 urn. The subjects breathed through a pneumotachograph, permitting mea-
surement of inspiratory and expiratory flow rate. In 15 subjects, exposed to 0.35, 0.4, or
3
0.5 mg/m , the respiration rate increased about 35 percent above control values, while the
maximum inspiratory and expiratory flow rates decreased about 20 percent. Tidal volume
o
decreased about 28 percent in subjects exposed to 0.4 mg/m . These changes occurred within
the first 3 minutes of exposure and were maintained throughout the 15-minute exposure period.
Lung function returned rapidly to baseline levels after the exposure ended. The tidal volume
rose above control values during the first minute after termination of the exposure and then
returned to preexposure levels. Breathing through the same apparatus without the acid mist
was done as a control, and no such changes were observed. Some subjects showed a marked reac-
tion to 5 mg/m , a level of acid mist perceptible to all. Individual responses were much more
varied at this level, the main effect being a decrease in minute volume. The investigators
suggest that bronchoconstriction may have been the response to sulfuric acid.
The effect of breathing sulfuric acid mist at different relative humidities (RH) was
studied by Sim and Rattle (1957). Healthy males (variable number of subjects), 18 to 46 years

13-24
-------
13-3, PULMONARY EFFECTS OF SULFURIC AC10
w
i
ro
ui

Duration of
Concentration exposure (mins)
0.35 - 5.0 mg/m3 H.SO, 15
HMD 1 Mm

3-39 mg/m3 H9SO.. 10 - 60
MHO 1-1.5 pit) *
S0? (1-60 ppm) plus Variable
O, to form H.SO.
alrBsol i
CMO 1.8 and 4,6 pm
H,SO, mist - 120
flOOO Mg/m
MHO 0.5 Mm (erg = 2.59)
H,SO. aerosol , 10
10, 100, 1000 jjfl/m
HMD 0.1 Mm
H-SO. (75 M9/m3) 120
MMAO 0.48 - 0.81 \im
H.SO. (0, 100, ,300, 60
or 1,000 Mg/m
MMAO 0.5 MB
(o-g = 1.9)
H-SO. (0, 223, 418, 120
939 Mg/m3 MHO
0.90 - 0.93 Mm
(crg = 1.66 - 1.73)
Number of
subjects
15

Variable
24
10
6 normal
6 asthmatics
6 normal
6 asthmatics
10
11
(2 exsmokersj
6 allergies;
1 childhood
asthmatic; 2
normals)
Source
Mask (rest)

Mask (rest)
Chamber (rest)
Mask (rest)
Oral
Chamber
(exercise)
Oral
Oral*
Chamber
(exercise)
Nasal
(rest)
Chamber
(exercise)
Effects Reference
Respiratory rates increased, Amdur et al., 1952
max insp. and expiratory
flow rates and tidal volumes
decreased
Longer particles due to "wet Sim and Rattle, 1957
mist" resulted in increased
flow resistance, cough, rales
and bronchoconstriction
Airway resistance Toyaina and Nakamura,
increased especially 1964
with larger particles
No pulmonary function Newhouse et al., 1978
changes but increased
tracheobronchial clearance
No pulmonary function Sackner et al., 1978
changes, no alterations
in gas transport
No pulmonary effects Kleinman and Hackney,
in either group 1978; Avol et al., 1979
No pulmonary function Leikauf et al., 1981
effects. Bronchial
mucociliary clearance
t following 100 ug/m »but
* following 1000 \ig/m ;
mucociliary clearance distal
to trachea wore affected
Small statistically signi- Horvath et al., 1981
ficant change only in
FEV,j0 at 939 Mg/«3 but
physiological significance
questioned.
-------
TABLE 13-3. (continued)
Concentration
H2SO.
100 Mg/m3
MHO 0.14 Mm
og = 2.9
High cone, aerosol
(1 mg/m3 each)
Duration of
exposure (mins)
240
16
Number of
subjects
28 normals
16 normals
17 asthmatics
Source
Chamber
(exercise)
Oral*
(rest)
Effects
No pulmonary function effects
SG induced by carbachol
si g. potentiated in asth-
Reference
Kerr et al . , 1981
Utell et al., 1981
Low cone, aerosol
(0.1 mg/m3 each)
KMAD = 0.5-1.0 Mm
Aerosols included:
NaHSO.
NH.HSO.
(NH.),SO.
H2S94Z
matics at 1 mg/m3 H,SO. and
NH.HSO. each. FEV/0 tig.
decreased after H.SO. and NH.HSO..
No changes in SGJ: with all tul-
fates; but two most responsive
asthmatics to high H.SO. dose ex-
hibited potentiation effect on
carbahol-induced bronchocon-
striction at lower H,SO. level.
Z 4
GO
I
ro
CTl
*Mouthpiece
-------
3
of age, breathed 3 to 39 mg/m concentrations of hLSO^ at 62 percent RH either via mask or ex-
posure chamber. Subjects were also exposed in the chamber to 11.5 to 38 mg/m concentrations
at 91 percent RH. At the lower RH, particles were 1 (Jtn in size. The addition of water vapor
to raise RH increased the mean particle size to 1.5 Mi" and intensified irritant effects of ex-
3
posure. For example, the irritancy of wet mist at 20.& mg/m- was much more severe ("almost
intolerable at the onset") than that of the dry mist at 39.4 mg/m ("well tolerated by all").
Air flow resistance ranged from 43 to 150 percent above normal in response to the wet mist,
compared to increases ranging from 35.5 to 100 percent above normal in response to the dry
mist. Two subjects exposed to sulfuric acid mist developed bronchitic symptoms but may have
been previously exposed to other substances. Adding ammonia (quantity not given) to the acid
mist annulled its irritant properties. There was no consistent evidence that the acid mist
caused changes in respiratory functions or blood pressure, pulse rate, or other cardiovascular
functions.
Toyama and Nakamura (1964) investigated the synergistic effects of S0~ in combination
with hydrogen peroxide (H?0?) aerosol mixtures, the latter of which oxidizes SO,, to form SO-,,
which reacts with moisture (H«0) to form H9SO.. Sulfur dioxide concentrations ranged from 1
3 3
to 60 ppm (2.6 to 157 mg/m ); the HJ39 concentrations were 0.29 mg/m for particles of 4.6 |jm
o
CMD (Horvath estimated MMAD was 13) and 0.33 mg/m for particles of 1.8 pm CMD (Horvath esti-
mated MMAD was 5). Airway resistance increased significantly in the combination (H-O, + SO.)
exposure, particularly for the group of 15 subjects inhaling the larger particles (p <0.01).
Toyama and Nakamura (1964) exposed subjects to a mixture of S02 and H2S04 aerosols. They used
an inadequate method to measure airway resistance. They described the aerosols as having a
4.5 urn diameter. They found a strong constricting effect on the upper airways.
Sackner et al, (1978) studied normal resting young adults and seven asthmatic middle-aged
subjects who breathed, by mouth, either sodium chloride or sulfuric acid aerosols for 10
3
minutes at concentrations of 10, 100, and 1000 pg/m (0.1-0.2 MMAD). Measurements on these
individuals continued for up to 3 hours after exposure. The asthmatic patients represented a
wide range of clinical status and treatment. Neither normal nor asthmatic individuals showed
statistically significant alterations of lung volumes, distribution of ventilation, earoxime-
try, dynamic mechanics of breathing, oscillation mechanics of the chest-lung system, pulmonary
capillary blood flow, diffusing capacity, arterial oxygen saturation, oxygen uptake, or pulmo-
nary tissue volume. No delayed effects were observed during a follow-up period of a few weeks.
Kleinman and Hackney (1978) and Avol et al. (1979) reported on the pulmonary responses of
six normal subjects exposed in an ambient environment of 88°F dry bulb and 40 percent relative
3
humidity, and 94 M9/ro H~SO.. A sham exposure was followed by 2 consecutive days of acid ex-
3
posure. Sufficient excess acid aerosol to neutralize the NH3 present (about 56 M9/m ammonia
neutralization product) was added to the air to provide for the desired acid concentration (75
o
(jg/m ). The aerosol MMAD was approximately 0.48 to 0.81 |jm. The effective exposure time was
13-27
-------
2 hours, with the first 15 minutes of each half hour devoted to exercise that increased venti-
lation to twice the resting level. Only one subject was exposed at a time to minimize the
effects of ammonia neutralization. The normal subjects showed no exposure-related changes
Utell et al. (1981) exposed 16 normal subjects (all subjects non-smokers) to acidic aero-
sols, each at a concentration of 1 mg/m , (MMAD = 0.5-1.0 nm> cr 1.5-2.2) for periods of 16
y
minutes. Several aerosol exposures were given each day in a double-blind random pattern. 'At
the beginning of each study, an approximate dose-response curve to inhaled carbachol was
obtained. All aerosols [NaHS04; (NH4)2S04; NH4HS04; H2S04; NaCl] were given orally. Although
normals demonstrated no reduction in airway conductance, the bronchoconstrictor action of car-
bachol was significantly potentiated by the prior inhalation of sulfuric acid or ammonium bi-
sulfate. The bronchioconstriction action of carbachol was potentiated by the sulfate aerosols
in proportion to their acidity.
Lippmann and coworkers (Lippman et al., 1980; Leikauf et al., 1981) had 10 non-smokers
inhale via nasal mask 0.5 nm ( it was
significantly (p <0.03) retarded. Mucociliary transport in the airways distal to the trachea
was affected more by H?SO« exposure than was transport in the trachea. Out of ten subjects,
four did not respond. These four had the fastest clearance rates of the ten subjects in their
•a
control tests. They were retested at 1000 (jg/m , with the H2SQ4 exposure preceding the radio-
labelled aerosol exposure, in order to determine if the H?S04 effect had occurred too late in
the tagged particle clearance to have affected the measurement. In these followup tests,
three of the four responded with a slowing of mucociliary clearance comparable in magnitude to
that seen in the other subjects. Thus, there was only one subject in the ten whose clearance
was not markedly affected by exposure to H?S04. The alterations in bronchial clearance half-
time were all transient, which was consistent with the results seen earlier in similar inhala-
tion tests on donkeys (Schlesinger et al., 1978). However, when donkeys were repeatedly ex-
posed to sulfuric acid at comparable concentrations, four of six animals developed persist-
ently slowed clearance, which remained abnormal for at least several months (Schlesinger
13-28
-------
et al., 1978, 1979). Taken together, these results suggest that under chronic exposure con-
ditions at the concentrations employed, persistent changes could occur in tnucociliary clear-
ance in previously healthy individuals and exacerbate preexisting respiratory disease.
Kleinman and Hackney (1978) and Avol et al. (1979) presented in greater detail the pre-
liminary findings reported by Bell and Hackney (1977a). They evaluated the effects of various
sulfate compounds on normal subjects, ozone-sensitive subjects, and asthmatic subjects (re-
quiring medical treatment). The exposures were approximately 2.5 hours in duration, with the
subjects exercising the first 15 minutes of each half hour at a pace sufficient to double
their ventilation rates. Measurements of pulmonary functions, which included FVC, FEV,, MEFR,
FEFr^y, FEFycaj, TLC, RV, delta nitrogen (AN,), closing volume, and total respiratory resist-
ance (R.) were made before and 2 hours after the work-rest regimen began. The ambient condi-
tions were 88°F dry bulb and either 40 or 85 percent relative humidity. Host of the exposure
studies were made on five to seven subjects. Four to five sensitive subjects and six asthma-
tics completed the subject pool. Subjects were first exposed to a control (no pollutant) en-
vironment and then to 2 or 3 consecutive days of the pollutants. Nominal exposure concentra-
3 3
tions were 100 ug/m for ammonium bisulfate (NH.HSO.) and 85 ug/m f°r ammonium sulfate
[(NH.^SO.j. The sulfate aerosol size distribution was nominally 0.4 urn MMAD (a 2.5 to 3).
There was some ammonia (NH,) in the exposure chamber. Pulmonary functions were unaffected by
exposure to the two types of aerosol.
Kleinman et al. (1981) conducted studies in which 19 volunteers with normal pulmonary
function and no history of asthma were exposed on two separate days to clean air and to an
3
atmosphere mixture containing 03 (0.37 ppm), SCL (0.37 ppm), and H^SQ^ aerosol (100 ug/m ,
MMAD 0,5 JJDI; erg = 3.0). Chemical speciation data indicate that 93 percent of the sulfuric
acid aerosol had been partially neutralized to ammonium bisulfate. Additional data suggest
that the acidity of the aerosol in the chamber decreased as a function of time during exposure
so that at the beginning of the exposures, subjects were exposed to higher concentrations than
they were at the end of the exposures. During this 2-hour period, the subjects alternately
exercised for 15 minutes, at a level calibrated to double minute ventilation, and rested for
15 minutes. Statistical analysis of the group average data suggested that the mixture may
have been slightly more irritating to the subjects than 0_ alone. A large percentage (13 of
19) of the subjects exhibited small decrements in pulmonary function. The group average
FEV, „ on the exposure day was statistically significantly depressed (p <0.001) by 3.7 percent
of the control value. One might expect 0, alone to depress FEV, ~ by approximately 2.8 per-
cent under similar exposure conditions.
Kerr et al. (1981) investigated the respiratory effects associated with exposure to low
levels of sulfuric acid (hLSO,,) aerosol. Twenty-eight normal subjects were exposed (1 or 2
3
subjects/chamber) for 4 hours to 100 ug/m H^SO. aerosol of particle size 0.1 to 0.3 urn (HMD =
0.14 urn; og = 2.9) in an environmentallly controlled exposure chamber. Over this four-hour
13-29
-------
exposure period the acidity of the aerosol in the chamber was not monitored as a function of
time; however, based on the results of Kleinman et al. (1981), it appears likely that partial
neutralization of the chamber atmosphere did occur. The degree of this neutralization is de-
pendent, however, on the number of subjects in the chamber during the exposure and chamber
flow rates. At one and three hours into the study on each day, bicycle ergometer exercise was
performed at a workload at 100 watts at 60 RPM for 15 minutes. Of the 28 subjects, 14 were
nonsmokers and 14 were cigarette smokers. None of the subjects complained of symptoms attri-
butable to the exposure. Measurements of pulmonary function were obtained 2 hours into the
exposure, immediately following exposure and 2 and 24 hours postexposure. These measurements
were compared with control values obtained at comparable hours on the previous day when the
subject breathed only filtered, clean air in the chamber. No statistically significant dif-
ferences in pulmonary function were observed either during the exposure, immediately after ex-
posure or 2 and 24 hours post-exposure. Similar results, which are described below, were also
observed for asthmatic subjects.
Kleinman and Hackney (1978) and Avol et al. (1979) also studied the responses of six
asthmatics to sulfuric acid utilizing the same procedure as described on page 13-27. These
subjects had pulmonary function test results that ranged widely from normal to abnormal. The
3
2-hr exposures were in environments containing 75 ug/m hLSO.. The lung functions of the
asthmatics showed no statistically significant changes. Two asthmatics, the extent of their
disease state not given, exhibited increases in respiratory resistance on both exposure days.
Nonetheless, it was concluded that there were no convincing adverse short-term health effects
of sulfuric acid. However, the authors also noted the small size of their subject pool and
recommended additional studies.
Utell et al. (1981) also evaluated the responses of seventeen non-smoking asthmatics
following the protocol described on page 13-28. The exposures were to various sulfate aero-
sols. The data presented in the manuscript are incomplete, but those available suggest that
specific airway conductance (SG ) induced by carbacol was significantly potentiated (p <0.01)
3 3
in asthmatics breathing HpSO- and NH.HSO- (each 1 mg/m ). Low sulfate exposure (0.1 mg/m )
produced no changes in SG ; however, the two asthmatics most responsive to the high H?SO.
dose via inhalation exhibited a potentiation effect of sulfuric acid (at the lower concentra-
tion) on bronchoconstriction induced by carbacol. A more extensive presentation of the data
obtained by these investigators will be required before a clearer evaluation of the effects of
sulfates on asthmatics can be determined.
Kleinman and Hackney (1978) and Avol et al. (1979) also included six asthmatics in their
studies concerned with exposures to sulfates. The asthmatics were not studied under high
2
humidity conditions but were exposed to high concentrations (up to 372 ug/m ) of (NH^SO^.
Pulmonary functions were not modified by these exposures. However, an interesting side
observation was made on the asthmatics. On their first day of exposure to NH.HSO. aerosol,
13-30
-------
they exhibited worse lung functions in the preexposure measurements than they had on a control
day; and their lung functions improved following the aerosol exposure. Subsequent analysis of
local ambient conditions showed that these subjects arrived for the aerosol testing after a
3-day period of increased S00 and ozone levels during a "mild air pollution episode."
£. ' * >
Most recently Horvath et al. (1981), using eleven male subjects, investigated the effects"
on pulmonary function of breathing sulfuric acid aerosol- (223, 418 and 939 Mi/m ; 0.9 urn) for
120 minutes in a chamber under exercising conditions (Ve ~ 30 I/minutes). The only measure-
ment to show an interaction across time (from preexposure to postexposure) and H~SOA concen-
3
tration was FEV, „ which was significantly decreased with exposure to 939 [jg/m H?SQ« but not
at the other concentrations. Symptoms commonly reported at this latter concentration include
throat irritation and dryness and cough. Horvath et al. (1981) conclude that their results
support previous studies (Avol et al., 1979; Kerr et al., 1981; Sackner et al., 1978; Lippmann
et al., 1981; Leikauf et al., 1981), finding relatively minor or no pulmonary function changes
in humans exposed to sulfuric acid aerosols in the 0.3 to 0.5 |jm size range at similar concen-
3
trations (100-1000 (jg/rn ). However, they do not confirm pulmonary function and other effects
reported by Amdur et al. (1952) and Sim and Rattle (1957), whose subjects were likely exposed
to larger size H2SO» aerosols.
13.3.2 Insoluble and Other Non-SulfurAerosols
It is well-known that the ambient air contains many other particulate matter species be-
sides sulfates. The non-sulfate species include, for example, condensed organic vapors (POM),
lead, arsenic, selenium, hydrogen ions, ammonium salts, and carbon as soot. Health effects
associated with animal and/or human occupational studies have been addressed for POM, lead,
and arsenic in separate health assessment documents. Because of ethical considerations, con-
trolled human exposure studies on these inherently toxic and/or potentially carcinogenic com-
pounds do not exist. However, in addition to sulfate aerosols which have been extensively
studied using humans, a few investigators have conducted controlled human exposure studies
using carbon or other inert dust. These studies are described below and are summarized in
Table 13-4.
In a study designed to determine the efficiency of nasal dust filtration, Andersen et al.
(1979) exposed 16 resting subjects (12 men, 4 women) for 5 hours to three levels of plastic
dust impregnated with carbon black (2, 10, and 25 mg/m ). The number of these particles, ex-
pressed as a percent of the total number of particles was 36, 41, 14, 17, and 2, respectively,
for the aerodynamic size ranges < 1.8, 1.9 to 5.3, 5.4 to 8.9, 9.0 to 12.4, and > 12.5 urn.
Nasal mucus flows were measured at 5 different positions in the nose and varied markedly from
subject to subject and in relation to the three dust concentrations. The only statistically
significant effect observed was a small but statistically significant decrease in FEV, „. The
information reported was inadequate to support this conclusion and the approximate data pre-
sented in their Figure 3 fails to confirm their statements. These investigators also noted
13-31
-------
only slight discomfort regardless of the level of dust in the environment. Apparently some
further discomfort (primarily dryness in the nose and pharynx) was reported after the expo-
sures were terminated.
Other investigators, in addition to Andersen et al, (1979), have evaluated pulmonary
effects associated with breathing particles under controlled exposure conditions. Widdecombe
et al. (1962) had 9 subjects inhale 20 breaths of granulated charcoal (7 to 14 mesh). Airway
conductances were measured before and 2 minutes after oral inhalation of the dust. Airway
conductance and thoracic gas volume ratio decreased in every subject (increased resistance) by
a mean of 41 p'ercent (p <0.001). Coughing occurred in many of the subjects, and a few had
mild discomfort in the upper airways. Subcutaneous injection of atropine blocked this effect,
suggesting that it is produced via a vagal reflex. The response to dust inhalation was rapid
and reversible, but the effective concentration levels could not be determined from the
report.
Acute physiological effects of inhaling fine particulate matter were evaluated by DuBois
and Oautrebande (1958). Five normal subjects were given a large number of pulmonary function
tests before and after inhalation (for 1 to 3 minutes) of small (5 to 10 mg) quantities of
o
fine chemically inert dust particles consisting of CaCQ-, (~ 250 mg/m ),* coal dust (~ 500
3
mg/m ),* activated charcoal powder, aluminum powder and aerosolized India ink. The size of
the particles (< 0.5 |jm; mean 0.04 pm) and the amount inhaled varied depending upon the sub-
stances tested. The subjects usually showed an immediate marked increase in airway resistance
and pulmonary resistance. Other pulmonary functions were inconsistently altered. The effects
observed were transitory, usually disappearing within 30 minutes. Constantine et al. (1959)
followed up on this study, utilizing six normal subjects and seven patients with either emphy-
sema or asthma without acute symptoms. They used dust particles (< 0.5 jjm) from a colloidal
3 3
iron-water suspension (~ 16.5 mg/m )* and from Mclntyre aluminum powder (~ 100 mg/m ).* These
inhalations resulted in an increased airway resistance, confirming the previous observations
of DuBois and Dautrebande. However, in patients with chronic respiratory disease, a signifi-
cant reduction in vital capacity was also found. The constricting effects of these dusts per-
sisted for a slightly longer time than previously observed (more than 25 to 45 minutes).
These results are of interest in that patients with chronic respiratory disease, asthma and
emphysema exhibit reactions similar to those of normal subjects. These patients, already
characterized by an increase in airway resistance, exhibit reactions to constricting aerosols
more severe than normal subjects; yet the percentage of increase of the airway resistance in
the seven patients studied after constricting aerosols is not greater than that of the six
normal control subjects.
Similar controlled human exposure studies were carried out by Toyama (1964). Ten healthy
o
males (20 to 35 years) inhaled 20 deep breaths of concentrated (10 mg/m ) dust which was
^Approximate values calculated from references.

13-32
-------
13-4. PULMONARY EFFECTS OF AEROSOLS
Concentration
Duration of
exposure (mins)
Number of
subjects
Oral or nasal
exposure
Rest or
exercise
Effects
Reference
CJ
i
OJ
CJ
SO, (T.6 - 5 ppm)
NiCl(CHO = 0,22 Mm)

SO, (9-60 ppm)
NSC1 (CMO = 0.95 (ira)
SO, (0.5, 1,0 and 5.0 ppm) 15
NSC1 (CMD = 6-8 pm)
SO, (1.1 - 3.6 ppm) 30
N5C1 2.0 - 2.7 »g/n!s
HMD = 0.25 pm
SO, (1, 5, 15 ppm) 30
NaCl 10-30 mg/nr*
HMD 0.15 pm, ag = 2.3
SO, (1 ppm) 60
NaCl 1 mg/m3
MHO 0.9 pin
og = 2,0 pm

S02 (1 ppm) 30
NaCl 1 mg/m3
HMD = 0.9 fjm,
ag - 2.0 pm

Mixture of : SO, 120
(0 37 ppm); 0, i
(0.37 ppm) ana
H SO. (100 pg/nt3)
MTO 0.5 pm, og = 3.0

(NH,) SO,; 85 pg/m3 120
(NHT) H So,; 100 pg/Bs
(MMAD 0.4 pn; og = 2.5-
3.0 for both salts)
13
10
10
12
(asthmatics)
8
(asthmatics)
19
Mask
Mask
Oral*
Oral*
Oral*
Oral
Oral*
Chamber
5 normals Chamber
5 ozone sensitive
6 asthmatics
R & E
Synergistic increases in
airway resistance with
aerosol

Airway resistance greater
after exposure to aerosol
than to exposure to SOg
alone

Significantly decrease in
MEF50X VC only at 5 ppm;
however, magnitude not
different from SO- alone

Ho effect on pulmonary
functions

Changes in pulmonary
function similar to
changes due to SO-
alone not influenced
by aerosol
Significant decreases in

Vmax SOX and V«ax 75X
Vnax 50%' Vmax 753£'FEVl-°
and R, decreased signifi-
cantly after aerosol
Small but statistically
significant decrements in
in FEVi.o and slight
increases in the incidence
of clinical symptoms

No changes in pulmonary
function
Toyana, 1962

Nakamura, 1964
Snell and Luchsinger,
1969
Burton et al., 1969

Frank et al., 1964

Koenig et al., 1980

Koenig et al., 1981

Klelniaan et al., 1981
Bell and Hackney,
1977a; Kleinman
and Hackney, 1978
Avol et al., 1979
-------
TABLE 13-4. (continued)
CO
1
CJ
Duration of
Concentration exposure (mins)
Inert plastic dust 300
impregnated with
carbon black (Xerox
Toner 6R9000T) (2,10,
25 mg/m3)
Inert plastic dust ira- 300
pregnated with carbon
black (Xerox Toner 6R90005)
2, 10 mg/nt3 plus SO,
(1 and 5 ppm)
NaNO,; NaCl (control) 16
Both 7 mg/m3 HMAD = 0.49
t», ag = 1.7 RH ~ 25%;
crystalline solids
High cone, aerosol 16
(1 ng/a3 each)
Number of Oral or nasal Rest or
subjects exposure exercise Effects Reference
16
16
11 normals
(with acute
resp. influenza
Infection)
16 normals
17 asthmatics
Chamber
Nasal
Chamber
Nasal
Oral*
Nasal
Oral*
R No significant detrimen- Andersen
tal effect on airway or
nasal mucociliary clear-
ance. Small significant
decrease in FEV1-0
R Reductions in nasal mucus Andersen
flowrate, forced expira-
tory flow (F£F,,~ 7«r) and
discomfort reiaifd'pfinci-
pally to SO.. SO, and dust
effects were, at most, additive.
R Significant decrease In Utell et
SG (P <0.005) and V
40f*(P <0.05). Asympto-
matic airway obstruction
R SG induced by carbachol Utell et
fig. potentiated in
et «!., 1979
et il., 1981
al., 1980
al., 1981
Low cone, aerosol
(0.1 mg/n3 each)
HMAD = 0.5-1.0 pm
Aerosols included:
NaHSO.
NHHSS.
asthmatics at 1 mg/m3
H-SO. and NH.HSO, each.
FEVj o sig. decreased
after H-SO, and NH.HSO,..
No changes in SG . with all
sulfates; but two most re-
sponsive asthmatics to high
H.SO. dose exhibited potentia-
tfon effect on carbachol-induced
bronchoconstriction at lower
H2S04 level
a0.1 ppm S02 S 262 M9/«3
1-0 ppm = 2620

0.5 ppa S02 a 1310 pg/in3 5.0 pp» s 13,100

*Mouthpiece
10 ppm = 26,200

50 ppm s 131,000 ng/m3
-------
similar in composition to fly ash and coarse-mode fugitive dust (64 percent crustal material,
10 percent sulfate, 19 percent volatile). The experimental dust (0.5 to 10 [jm, 2.0 HMD) was
generated, using essentially the techniques of DuBois and Dantrebande, from dust particles col-
lected by deposition in Kawasaki City, an industrial area of Japan. Dust-induced changes in
airway resistance showed a wide range of response from a postexposure decrease of a minus 18
percent to a considerable increase of up to 73 per*cerft. The majority of subjects (8 out of
10), however, showed increses between 11 to 73 percent. The authors believe that the response
appeared to be initiated by mechanical irritation rather than chemical action of the particle,
since isoproterenol (a bronchodilator) abolished changes in airway resistance. Additional
experiments reported as being done on 8 subjects exposed sequentially to SO, and dust par-
ticles (and showing potentiating effects of the two pollutants) are difficult to interpret
because the experiments were not designed to account for potential carryover effects from one
exposure to another.
McKerrow (1964) discussed in more detail earlier findings reported by McOermott (1962).
•3
Six subjects were exposed to four concentrations of coal dust (9, 19, 33, and 50 mg/m ) for
four hours at a time on two occasions. Responses, as determined by changes in airway resist-
2
ance (Raw)> occurred early in the exposure to all but 9 mg/m concentration, and it was con-
cluded that "significant increases in airway resistance occurred and the response was corre-
lated to the weight of coarse particles between 3.6 and 1 urn."
Morris and Bishop (1966) exposed normals, asthmatics and bronchitics to calcium carbonate
dust. A "dust cloud highly aggregated and containing about 120,000 particles/ml (~ 21 to 170
mg/m ; 0.5 to 5.0 (jm)* and about 40,000 particles/ml (~ 0.45 to 7 mg/m , 0.2 to 0.5 urn)"* was
breathed through a mouthpiece. The volume or mass of dust inhaled was not precisely deter-
mined. Arterial blood samples were obtained so that A~aD~p (alveolar-arterial difference in
Qy partial pressure) could be measured. They used dust particles (< 0.5 |jm) from a colloidal
3 3
iron-water suspension (~ 16.5 mg/m )* and from Mclntyre aluminum powder (~ 100 mg/m ).* These
inhalations resulted in increased airway resistance, confirming the previous observations of
DuBois and Dautrebande. However, in patients with chronic respiratory disease, a significant
reduction in vital capacity was also found. The constricting effects of these dusts persisted
for a slightly longer time than previously observed (more than 25 to 45 minutes). These
results are of interest in that patients with chronic respiratory disease, asthma and
emphysema exhibit reactions similar to those of normal subjects. These patients, already char-
acterized by an increase in airway resistance, exhibit reactions to constricting aerosols more
severe than normal subjects; yet the percentage of increase of the airway resistance in the
seven patients studied after constricting aerosols is not greater than that of the six normal
control subjects.
^Approximate values calculated from references.
13-35
-------
Utell et al. (1980) studied the effects of exposure to nitrate on the airways of 11 sub-
jects following a respiratory infection [influenza A (H,N-,) infection]. Each subject, under a
double blind randomization protocol, orally (mouthpiece) breathed either sodium chloride (NaCl)
or sodium nitrate (NaNO~) for an initial period of 16 minutes and then the other aerosol for
16 minutes three hours later. The mass median aerodynamic diameter (MMAD) of the NaNO, aerosol
3
was 0,49 urn (concentration 7,000 (jg/m ). Exposure to the nitrate aerosol (compared to NaCl
aerosol) resulted in significant decreases in specific airway conductance (p <0.005) and par-
tial expiratory flows at 40 percent of TLC (p <0,05) at the first two test times (48 hours and
one week postinfection). They concluded that the constriction observed was a specific effect
of the sodium nitrate rather than a nonspecific response to the particles.
Some experimenters have attempted to determine physical or chemical characteristics of
atmospheric particles that may affect their patterns of respiratory tract deposition, inter-
actions with other pollutants such as S0?, and ultimately, their relative potential for con-
tributing to toxic effects under various conditions. For example, in a physical system, Corn
and Cheng (1972) found that insoluble particulate samples of CaCO,, VpOg, and fly ash from a
coalburning power plant were essentially inert to S0?; but Fe^O,, MnO?> activated carbon and
suspended particulate matter from urban Pittsburgh air sorbed S0?. They suggested that pul-
monary flow resistance (measured in their earlier guinea pig studies) consequent to SOp-aerosol
exposure is sensitive to the chemical reaction product of SO,, with specific particulate
compounds.
Stahlhofen et al. (1980) determined the deposition of particles in subjects breathing at
the slow rate of 7.5 breaths per minute. When the total volume was 1 liter, pulmonary deposi-
tion of 3.5 (jm particles was as high as 70 percent. The deposition of various other particles
(such as FeoOoi di-2-ethylhexyl sebacate) have also been studied, but these are not commonly
found in the environment. Additional basic information on respiratory system deposition of
respirable dust particles is presented in Chapter 11.
13.4 PARTICULATE MATTER AND SULFUR DIOXIDE
As suggested by Amdur (1969) and the above Corn and Cheng (1972) study findings, one of
the most significant factors influencing physiological responses to S0? may be the presence of
certain specific kinds of particulate matter in the atmosphere. Controlled human exposure
studies addressing this issue are also summarized in Table 13-4. Particulate matter appears
to interact with S0? in at least two distinct ways: as a carrier of SO,, and as a factor in
chemical reactions resulting in the conversion of SOp to other forms. In their carrier role,
particles may adsorb SO, and, depending on their size, solubility, and other characteristics,
transport it deep into the respiratory system (see Chapter 11, Section 11.2 for more detailed
discussion of deposition).
Respiratory function effects are illustrated by the results of studies by Nakamura (1964)
and Toyatna (1962), who reported that sodium chloride aerosol potentiated the response of human
subjects to S02- In Nakamura's (1964) study, 10 subjects were first exposed to NaCl aerosol
(CMD = 0.95 urn; Horvath's estimate MMAD = 5.6 t-im) alone for 5 minutes, allowed to recover for
13-36
-------
10 to 15 minutes, exposed to SOp alone at 9 to 60 ppm (23.6 to 157 mg/m ) for 5 minutes,
allowed 20 to 30 minutes to recover, and then exposed to SOp and the NaCl aerosol together for
5 minutes. Airway resistance was greater after the combination exposure than after exposure
to SOp alone (Nakamura, 1964). As noted, the combination condition always followed exposure
to SOp alone, thus raising the possibility that the effects of the latter exposure were con-
founded, However, on average, the subjects' airway resistance measures returned to only 4
percent above their pre-exposure control levels, thus making it more likely the reported
effects were independent of preceding conditions. Toyama (1962) also reported that 5 minute
exposures to SOp in combination with submicrometer (0.22 pm MMD; Horvath's estimae MMAD = 0.36
|jm) particles of NaCl aerosol produced synergistic increases in airway resistance in 13
3
subjects, even at levels as low as 1.6 ppm (4,2 mg/m ) SOp. There was also a linear relation-
ship between S0? concentration and percentage increase in airway resistance.
On the other hand, Burton et al. (1969) were unable to demonstrate comparable effects in
10 subjects exposed to S09 (1.1 to 3.6 ppm [2.9 to 9.4 mg/m ]) in combination with NaCl
3
aerosol (2.0 to 2.7 mg/m ; 0.25 pm MMD; Horvath's estimate MMAD 0.4 (jm). There was, however,
a great deal of variability within and between subjects in this study, including one or two
possible "hyper-reactors" who did show effects below 3 ppm. Frank et al . (1964) studied 12
subjects who were exposed to three conditions of SOp and NaCl aerosols. There were six
subjects in each group, but the same subjects were not evaluated under each of the three con-
ditions The purpose of this study was to determine whether acute changes in respiratory
dynamics Rl (pulmonary flow resistance) noted to occur during S0? exposure were intensified by
the presence of sodium chloride particles. The NaCl aerosols had a mean geometric diameter of
2
0.15 m (Horvath's estimate MMAD = 0.3 (jm) and a concentration of 10 to 30 mg/m ; SOp concen-
trations were 1 to 2, 4 to 6, and 14 to 17 ppm (2.6 to 5.2, 10.5 to 15.7, and 36.7 to 44.5
mg/m ). The subjects' response to the SOp exposures was as previously noted, in that Rl was
not affected by the lower levels of SOp and progressively increased at the higher levels. The
only statistically significant difference (p <0.05) between the effects of the gas alone and
the gas-aerosol mixture was a slightly greater average increase in pulmonary flow resistance
at 4 to 6 ppm (10.5 to 15.7 mg/m ) S02 than under the combination condition. Addition of the
NaCl aerosol resulted in changes similar to those observed for S0? alone.
Snell and Luchsinger (1969) also compared the effects SOp alone and in mixture with
aerosols of either NaCl or distilled water. Nine subjects inhaled SO, at 0.5, 1, and 5 ppm
3
(1.3, 2.6, and 13.1 mg/m ) alone and in combination with aerosols for 15-minute periods sepa-
rated by 15-minute control periods. For the SQ0-saline aerosol exposure, decreases in maximum
3
expiratory flowrate (MEFj-nv v/0 were significant (p <0.01) only at 5 ppm (13.1 mg/m ) SOp
whereas the SOp-distilled water aerosol exposure produced significant decreases (p <0.01) at
all SOp exposure levels, i.e., 0.5, 1, and 5 ppm (see Figures 3 and 4, Snell and Luchsinger,
1969). The authors noted that the size of the aerosol particles differed considerably, saline
particles averaging around 7 |jm in diameter and water aerosols averaging less than 0.3 pm in
diameter (Snell and Luchsinger, 1969; Ulmer, 1974).
13-37
-------
Koenig at al. (1980) exposed nine adolescent resting subjects (extrinsic asthmatics) for
3 3
60 minutes to either filtered air, 1 ppm (2.6 mg/m ) S0» and I mg/m of sodium chloride drop-
3
let aerosol or 1 mg/m of NaCl droplet aerosol (HMD 0.9 pm; a = 2.0) alone. Exposure to S0?
alone was not performed. Total respiratory resistance (R,-), maximal flow at 50 and 75 percent
of expired vital capacity (partial flow volume), FEV, Q> and functional residual capacity were
measured before, during (30 minutes), and after exposures via forced oral breathing. No sta-
tistically significant changes were found during exposures to filtered air or NaCl aerosol
alone. Significant decreases (p <0.025) were observed in V trw and V -,^j after combined
SOg'NaCl exposure, suggesting that broncho'constriction occurred in the peripheral airways. It
should be noted that V (-„„/ was significantly depressed (8 percent) only at the midpoint of
fflciX DU/o
exposure. However, since possible chemical reactions can occur between dissolved S0? and the
NaCl droplets (producing sulfite, bisulfite, and hydrogen ions), the pulmonary effects ob-
served cannot be directly attributed to either gaseous SOp or to the chemical substances
produced.
Koenig et al. (1981) exposed 8 adolescent extrinsic asthmatics to the same conditions as
in the previous study but had them also undergo a 10-minute period of moderate exercise during
the exposures. Maximum flow at vital capacity 50% and 25% above residual volume (V Kno/ VC
max 7%% ^^ decreased 44 and 50 percent respectively from the baseline mean after the
exercise. Statistically significant changes in forced expiratory volume in one sec. (FEV, n)
and Ry (airway resistance total) were also observed, suggesting that exercise and SOp-NaCl
exposure resulted in effects on both large as well as small airways. The functional changes
seen after exercise with exposure to filtered air or NaCl droplet aerosol alone were not
statistically significant. Although V „ rraf was depressed in resting subjects (extrinsic
max
asthmatics) 8 percent (t = 2.83 p < 0.025) and 6 percent (t = 0.38, p = N.S), respectively, in
the 1980 and 1981 studies by Koenig et al . , it should be mentioned that the latter change was
not significant after 30 minutes of exposure. In the 1980 study, all subjects (N = 9) dec-
reased in V_,v cno/; however, in the 1981 study some of the eight subjects increased and some
maX DU«3
decreased.
Recently, Andersen et al . (1981) exposed 16 subjects (male and female) for 5 hours in an
environmental chamber to several levels of SO, and dust. Subjects were exposed to clean air,
31 3
to combinations of S0« (2.6 or 13 mg/m and inert plastic dust (2 or 10 mg/m ) or to S09 (13
33
mg/m ) and dust (10 mg/m ) coated with vanadium. Nasal mucus flowrate, nasal airflow resis-
tance, forced vital capacity, and subjective discomfort were measured. Reductions in nasal
mucus flowrate, forced expiratory flow (FEF?E- t Tcy)» and discomfort were related principally
to SQy. Sulfur dioxide and dust effects were, at the most, additive without evidence of po-
tentiation effects. Measurements were made initially in clean air and after 2 to 3 and 4 to 5
hours of exposure to the various combinations employed. Exposures were made on five consecu-
tive days. There was no randomization of the exposures. This may have been responsible for
some of the variability in their accumulated data. They relate the data from the present

13-38
-------
study to an earlier report (Andersen et a]., 1974), where S0~ exposures induced a decrease (p
<0.01) in nasal mucus flowrate, more pronounced at the higher SO- concentration. This depres-
sion of nasal mucus flowrate caused by S0? exposure is apparently further increased by the
plastic dust in the SCL environment. No potentiation effects were observed. Some caution
should be expressed as 'to these conclusions, since different subjects were utilized and the
two studies were reported 1 years apart. FEV-, 0 was reduced during S0? exposures (p <0.05),
and no change occurred during dust-alone exposures. They also report that combined exposure
to dust and S0? produced an additive effect with respect to experienced discomfort.
13.5 SULFUR DIOXIDE, OZONE, AND NITROGEN DIOXIDE
Another important issue is whether or not S0? interacts synergistically with other gase-
ous air pollutants to produce greater-than-additive effects beyond those due to exposures to
S0? or the other pollutants alone. Several studies have addressed this issue in relation to
possible combined effects of SO, and other common gaseous air pollutants, as discussed below
and summarized in Table 13-5.
Bates and Hazucha (1973) and Hazucha and Bates (1975) exposed eight volunteer male sub-
3
jects to a mixture of 0.37 ppm 0, and 0.37 ppm (0.99 mg/m ) $(L for 2 hours. Temperature,
humidity, concentrations and particle sizes of ambient aerosols (if any) were not measured.
Sulfur dioxide alone had no detectable effect on lung function, while exposure to ozone alone
resulted in decrements in pulmonary function. The combination of gases resulted in more
severe respiratory symptoms and pulmonary function changes (10 to 20% decrements) than did
ozone alone. Using the maximal expiratory flow rate at 50 percent vital capacity as the most
sensitive indicator, it was evident that, after 2 hours exposure to 0.37 ppm (0.99 mg/m ) SO,
alone, no change occurred. However, during exposure to 0.37 ppm 0,,, a 13-percent reduction
3
was observed, while exposure to the mixture of 0.37 ppm 0, and 0.37 ppm (0.99 mg/m ) S0? re-
sulted in a reduction of 37 percent in this measure of pulmonary function. The effects
resulting from 0., and SO, in combination were apparent in 30 minutes, in contrast to a 2-hour
time lag for exposure to 0- alone.
Bell et al. (1977b) attempted to replicate these studies with four normal and four ozone-
sensitive subjects. They showed that the 0, + S0? mixture had a greater detrimental effect on
all pulmonary function measures than did 0, alone. However, only some of the lung function
parameters measured showed statistically significant decrements when compared to 0,. Four of
the Hazucha and Bates (1975) study subjects were aso studied by Bell et al. (1977b). Two of
these subjects had unusually large decrements in FVC (40 percent) and FEV, (44 percent) in the
first study (Bates and Hazucha, 1973), while the other two had small but statistically signi-
ficant decrements. None of the subjects responded in a similar manner in the Bell study. Re-
trospective sampling of the ambient air conditions utilizing particle samplers and chemical
analysis in the chamber showed that acid sulfate particles could have been 10- to 100-fold
higher in the Hazucha and Bates chamber and thus might have been responsible for the synergis-
tic effects observed, however, this hypothesis was not confirmed in recent studies conducted

13-39
-------
13-5.
PULMONARY EFFECTS OF COMBINED EXPOSURES TO S02 AND OTHER GASEOUS AIR POLLUTANTS

Concentration
S02 (0.15 ppm)
and
03 (0.15 ppm)
S02 (0.37 ppm)
and
03 (0.37 ppm)
00
o
S02 (0.37 ppm)
and
03 (0 37 ppm)
S02 (0 40 ppin)
and
03 (0 40 ppm
Duration of Number of
exposure (dins) subjects Source
120 6 Chamber
(exercise)
120 8 Chamber
(exercise)

120 4 (normal) Chamber
4 (ozone (exercise)
sensitive)
4 (from Bates)
120 9 Chamber
(exercise)
Effects
Decrease in SG after expo-
aw
sure to 03< Synergistic
potentiation reported for
S02 + Oj, but statistics
basis unclear
Decreased pulmonary functions
(synergistic effect of
S02/03 greater than 03
alone on FRC, FEV, -,
1 0
MMFR, MEFR 50%)
Unable to confirm S02/03
synergistic effects beyond
pulmonary decrement due
to 0_ alone
Unable to confirm S02/03
synergistic effects
beyond changes due to
Reference
Kagawa and Tsuru, 1979
Hazucha and
Bates, 1975
Bates and
Hazucha, 1973
Bell et al., 1977b

Horvath and Folinsbee
1977;
Bed! et al , 1979
O., alone
O
-------
TABLE 13-5, (continued)

Duration of Number of
Concentration exposure (mins) subjects
S02 (0 3 ppm) 120 19 asthmatics
+ N02 (0.5 ppm)
S02 (0.5 ppm) 120 24 normals
+ N02 (0 5 ppM)
S02 (5 ppra) 120 11
and
N02 (5 ppm)
S02 (5 ppm) 120 11
N02 (5 ppn)
and
03 (0 1 ppm)
S02 (0.12 ppm) 120 11
N02 (0 06 ppm)
and
03 (0.025 ppm)
Source
Chamber
(exercise)
Chamber
(exercise)
Chamber
(exercise)

Chaaber
(exercise)

Chamber
(exercise)

Effects
No pulmonary alterations
No pulmonary alterations
No changes in Pa02, PaC02>
pHa or TGV. Airway re-
sistance (Raw) increased
significantly.
No changes in Pa02> PaC02,
pHa or TGV. Airway re-
sistance (Raw) increased
significantly.
No changes in pulmonary
functions

Reference
Linn et al., 1980
Linn et al., 1980
von Nieding et al. , 1979

von Nieding et al, 1979

von Nieding et al., 1979

-------
TABLE 13-5. (continued)
i
.£»
ro

Concentration3
Mixture of:
SO 2 (5 ppm)
N02 (5 ppm)
03 (0.1 ppm)
Mixture of:
S02 (0.33 ppm)
NO- (0.16 ppm)
0, (0.075 ppm)
Duration of Number of
exposure (mins) subjects

120 24

8 hr/d for 15
4 successive
days
Source Effects

Chamber Data not adequately analyzed
(exercise) and could not be from the
data presented.

Chamber Statistical analysis of the
(rest) data not adequate
Reference

Islam and Ulner, 1979a

Islam and Ultnr, 1979b

a0.1 ppm SO,, 3 262 pg/n3 1.0 ppm = 2620 \>g/m

i S02 3 1310 |ig/«a 5.0 ppm = 13,100
0.5 ppm
10 ppm s 26,200 ug/m3

50 ppw S 131,000 (ig/*3
-------
by Kleinman et al. (1981) (see Section 13.3.1.2) involving identical concentrations of ozone
3
and SO,, and 100 |jg/m of sulfuric acid. In the Montreal chamber studies (Bell et al., 1977b),
concentrated streams of S09 and 0,, exited from tubes separated by 8 inches (20 cm) under a fan
3 3
which forced 167 ft /minute (4.7 m /minute) of air-conditioned laboratory air with SO- and 0,
through the chamber and out an exhaust, line on the opposite wall. The concentrated streams of
SCL and Oq could have reacted rapidly with each other and with ambient air impurities like
olefins, to form a large number of H^SO* nuclei which grew by homogenous condensation, coagu-
lation, and absorption of MR, during their 2-minute average residence time in the chamber.
Horvath's group (Horvath and Folinsbee, 1977; Bedi et al. , 1979) exposed nine young men
(18 to 27 years old) to 0.4 ppm 03 and 0.4 ppm (1.0 mg/m ) SO, singly and in combination for 2
hours in an inhalation chamber at 25°C and 45 percent relative humidity. The subjects exer-
cised intermittently for one-half of the exposure period. A large number of pulmonary func-
tion tests were conducted before, during, and after the exposure. Subjects exposed to fil-
tered air or to 0.4 ppm SO, showed no significant changes in pulmonary function. When exposed
to either 0^ or 0, plus SC^, the subjects showed statistically significant decreases in maxi-
mum expiratory flow, forced vital capacity, and inspiratory capacity. There were no signifi-
cant differences between the effects of 03 alone and the combination of 03 + SO,; thus, no
synergistic effects were discernible in their subjects. Although particulate matter was not
present in the inlet air, it is not known whether particles developed in the chamber at a
later point.
Chamber studies were also conducted by Kagawa and Tsuru (1979) who exposed six subjects
for 2 hours with intermittent exercise (50 watts; i.e. ventilation of 25 1/min) for periods of
15 minutes exercise separated by periods of 15 minutes rest. The exposures were performed
weekly in the following sequence: filtered air, 0.15 ppm 0,; filtered air, 0.15 ppm (390
3
ug/m ) SO,.,; and finally filtered air, 0.15 ppm 0~ - 0.15 ppm S0~. Pulmonary function measure-
ments were obtained prior to exposure after 1 hour in the chamber and after leaving the cham-
ber. Although a number of pulmonary function tests were performed, change in specific airway
conductance (SG ) was utilized as the most sensitive test of change in function. They
reported a significant decrease in five of the six subjects exposed to 0, alone. In three of
the six young male subjects, they reported a significantly enhanced decrease in SG after
clW
exposure to the combination of pollutants compared to the decrease in SG in these subjects
with 0., exposure alone. Two other subjects had similar decreases in either 0, or Q~ + SOp
exposure. Kagawa and Tsuru (1979) suggested that the combined effect of the two gases on SG
is more than simply additive. Questions concerning the statistical approach employed by them,
however, argue for caution in accepting their conclusions regarding greater-than-additive
effects. Also subjective symptoms of cough and bronchial irritation were reported to occur in
subjects exposed to 0-, or the 0,, + S0« combination. The question of potential synergistic in-
teraction between S0? and 0,, therefore, remains unresolved by this study.
13-43
-------
Recent studies by Bedi et al. (1981), under conditions of high temperature and humidity,
attempted to explain the conflict of opinion as to the cause of synergistic effects reported
by Hazucha and Bates (1975) and Kagawa and Tsuru (1979) for humans exposed to the combined
gases of 0^ and SOp. While exercising (Ve ~ 30 I/minutes), eight young nonsmoking adult males
were randomly exposed for 2 hours to filtered air, 0.4 ppm (1.1 mg/m ) SOp, 0.4 ppm 03 and 0 4
ppm SOp plus 0.4 ppm 03 at 35°C and 85% RH. No functional changes in FEV, Q occurred as a
result of exposure to filtered air or 0.4 ppm (1.1 mg/m ) SOp, but significant decreases
occurred following exposure to either 0.4 ppm 0, (6.9%) or the combination of/0.4 ppm 0, plus
«3 s5 I O
0.4 ppm (1.1 mg/m ) S02 (7.4%). However, no significant differences were found between the
ozone exposure and the ozone plus sulfur dioxide exposure. Observed alterations in pulmonary
functions in the SO, + 0~ exposure reflects the changes occurring during the 0,, exposure,
clearly indicating that no synergistic effects related to the additional presence of S0? were
evident in this study.
In another recent study, Linn et al. (1980) exposed 24 normal subjects to clean air and
3 3
to a combination of 0.5 ppm. (0.9 mg/m ) NO,, and 0.5 ppm (1.3 mg/m ) SOp for two hours. The
exposure was conducted at an ambient temperature of 31°C dry bulb and 40% relative humidity.
Fifteen-minute exercise periods requiring a ventilatory exchange of approximately 30 liters/
minute were alternated with 15-minute rest periods. Group mean lung function was essentially
similar in both exposure conditions with no changes occurring consequent to the exposures.
There was a small, statistically significant increase in symptoms during and after the ex-
posure. (Follow-up postexposure symptomatology reports were obtained by telephone.) Linn et
al. (1980) also exposed 19 asthmatic subjects (of widely different clinical status) to 0.5 ppm
3 3
(0.9 mg/m ) N02 and 0.3 ppm (0.8 mg/m ) S02 as well as to filtered air following a protocol
similar to that conducted in their concurrent studies of normal subjects. Again no pulmonary
alterations were observed, a finding similar to that obtained for normal subjects. Also, no
subjective symptoms were observed during the exposure, but small, statistically significant
increases in symptoms were reported during the telephone followup later in the day. These
changes were not necessarily attributable to the earlier test exposure. Overall, then, the
Linn et al. (1980) study does not provide evidence for synergistic interactions between SOp
and NO- at the concentrations studied, for either normal or asthmatic subjects.
Von Nieding et al. (1979) exposed 11 subjects to 0,,, N0« and SOp singly and in various
combinations. The subjects were exposed for 2 hours with 1 hour devoted to exercise that
doubled their ventilation. The work periods were of 15-minute duration interspersed with 15-
minute periods at rest. In the actual exposure experiments, no significant alterations were
observed for 0,, COp, and pH in arterialized pulmonary blood and thoracic gas volume (TGV).
Total airway resistance (RT) and arterial oxygen tension (PaOp) were altered in certain
studies. Arterial oxygen tension was decreased (7 to 8 torr) by exposure to 5.0 ppm (9.2
3 3
mg/m ) N02 but was not further decreased following exposures to 5.0 ppm (9.2 mg/m ) N02 and
13-44
-------
5.0 ppm (13.1 mg/m3) S02 or to 5.0 ppm (9.2 mg/m3) N02, 5.0 ppm (13.1 mg/m3) S02, and 0.1 ppm
03 or 5.0 ppm (9.2 mg/m3) NOp and 0.1 ppm 0,. Airway resistance increased significantly [0.5
to 1.5 cm HnO/(L/s)] in the combination experiments to the same extent as in the exposures to
NCL alone. In the 1-hour post exposure period of the NOp, S0?, and 0, experiment, RT con-
tinued to increase. Subjects were also exposed to a mixture of 0.06 ppm (110 ug/m ) NO,, 0.12
3
ppm (310 ug/m ) SQp, and 0,025 ppm 0,. No changes in any of the measured parameters were
observed. These same subjects were challenged with a 1, 2, and 3 percent solution of acetyl-
o
choline following control (filtered air) exposure and to the 5.0 ppm (9.2 mg/m ) NO,, 5.0 ppm
3 1
(13.1 mg/m ) S02, and 0.1 ppm 03 mixture, as well as after the 0.06 ppm (110 ug/m ) NCk, 0.12
ppm (310 ug/m ) SO,,, and 0.025 ppm 03 mixture exposures. Individual pollutant gases were not
evaluated separately. The expected rise in airway resistance was observed in the control
study. Specific airway resistance (R x TGV) was significantly greater than in the control
study following the combined pollutant exposures.
In another study of simultaneous exposure to S0?, N0? and 0,, three groups of eight sub-
jects, each of different ages (<30, >49 and between 30-40 years) were exposed for 2 hours in a
chamber on three successive days (Islam and Ulmer, 1979a). On the first day, subjects
breathed filtered air and exercised intermittently (levels not given); on the second day they
3 3
were exposed at rest to 5.0 ppm (13.1 mg/m ) SO,, 5.0 ppm (9.2 mg/m ) NO, and 0.1 ppm 0,; on
3 3
the third day the environment was again 5.0 ppm (13.1 mg/m ) SO,, 5.0 ppm (9.2 mg/m ) N0? and
0.1 ppm 0- but the subjects exercised intermittently during the exposure. Statistical eval-
uation of the data for 11 lung function test parameters and two blood parameters (PaOp and
PaCOp) were not adequately performed. These measurements were made before, immediately, and 3
hours post exposure. Individual variability was quite marked. The investigators concluded
that no synergistic effects occurred in their healthy subjects. However, since they did not
systematically expose these subjects to the individual components of their mixed pollutant
environment, the conclusion can only be justified in that they apparently saw no consistent
changes. There were some apparent changes in certain individuals related to exercise (unknown
level) and age, but the data were not adequately analyzed nor could they be, based on the in-
formation presented.
Islam and Ulmer (1979b) studied 15 young healthy males during chamber exposures to 0.9
33 3
mg/m S02, 0.3 mg/m NO, and 0 15 mg/m 0,. Ten subjects were exposed to 1 day of filtered
air and 4 succesive days to the above gas mixture. Another group of 5 subjects were exposed
for 4 days to the pollutant mixture followed by 1 day to filtered air. Each exposure was 8
hours in duration. Following each exposure, the subjects were challenged by an acety"Icholine
aerosol. Eight pulmonary function tests and four blood tests (PaOp, PaCQ,, Hb, and lactate
dehydrogenase) were performed before and after the exposure. No impairments of lung func-
tions, blood gases or blood chemistry were found, but the study suffers from a deficiency in
statistical analysis of the data. Also, some of the subjects were said to have exhibited
unusual responses.

13-45
-------
13.6 SUMMARY AND CONCLUSIONS
Unlike community epidemiological studies that investigate health responses of large popu-
lation cohorts under highly variable ambient exposure conditions, controlled human exposure
(clinical) studies typically evaluate much smaller numbers of subjects but under much better
defined and carefully controlled exposure conditions. In the latter type of studies, expo-
sures to either single pollutants or combinations of pollutants are usually carried out in
x
environmentally controlled chambers in which relative humidity, temperature, and pollutant con-
centrations are designed to approximate representative ambient air exposure conditions,
especially those thought to be associated, with the induction of acute effects.
Generally inherent in the design of controlled human exposure studies carried out in the
United States are limitations on the range or types of pollutant exposures and types of sub-
jects studied so as to assure (as approved by human rights and medical ethics committees) that
the experimental exposures to the pollutants being tested per se will not lead to serious mor-
bidity, irreversible illness, or death. Consequently, the types of pulmonary responses typi-
cally assessed in controlled exposure studies are typically "transient" and "reversible."
However, depending upon the population at risk, the method of exposure, and the level of sub-
ject activity, the so-called mild and reversible health effects measured in controlled human
exposure studies may be indicators of other more serious associated health effects likely to
occur if more prolonged or repeated ambient exposures to the same concentrations of pollutants
were encountered by study subjects; or the observed effects per se may be sufficient to inter-
fere with normal work or social activities of certain individuals under some ambient circum-
stances. For example, relatively small increases in airway resistance of no particular health
concern for healthy, nonsensitive adults may be of medical importance for asthmatics or other
sensitive groups with already compromised pulmonary functions, especially when accompanied by
symptoms associated with or indicative of the onset of more severe breathing difficulties for
them under ambient-conditions.
In general, the population groups at special risk to air pollution include the young, the
elderly, and individuals predisposed by some particular disease, including asthma, bronchitis,
cystic fibrosis, emphysema, and cardiovascular disease. In the normal population, there are
also nondiseased but hypersensitive individuals. Such nondiseased "hyperreactors" have been
found among at least three of the distinct population groups (normals, chronic bronchitics,
and asthmatics) that have been evaluated under controlled exposure conditions in regard to
their responses to SO, and particulate matter (Lawther, 1955; Frank et al., 1964; Nadel et
a!., 1965; Burton et al., 1969; Lawther et al.s 1975; Jaeger et al., 1979; Sheppard et al.,
1980, 1981; Stacy et al., 1981).
In evaluating responses of the above population groups, various investigators have
assessed the effects of varying the activity levels of the subjects, the mode of exposure
(e.g., nasal, oral, oronasal, or open chamber), and the duration of exposure. One purpose of
13-46
-------
increasing the activity level during exposure is to increase minute ventilation (Ve), so as to
simulate outdoor exposures during daily activities. A large majority of normal subjects at
rest breathe almost exclusively through the nose with a minute volume (Ve) of approximately 5
to 10 liters/minute. However, some healthy individuals may have abnormally obstructed nasal
passages, or, for other reasons, regularly breathe oronasally even at rest, as indicated by
recent studies (Niinimaa, 1980, 1981; D1Alfonso, 1980), and certain population groups at risk
(such as asthmatics) include some individuals who tend to breathe orally even at rest. At
some level of increased ventilation, individuals who normally breathe through the nose at rest
also shift over to oronasal breathing. In regard to ventilation levels at which that shift
has been observed to occur, Niinimaa et al. (1980, 1981) reported a switch from nasal to
oronasal breathing at a minute volume of 35.3 ± 10.8 (mean ± S. D.) liters per minute, and
after the switch to oronasal breathing by persistent nasal breathers (at rest), the nasal por-
tion of Ve decreased to 56 percent of total Ve. In addition to the study recently published
by Niinimaa et al. (1980, 1981), D1Alfonso (1980) also observed the shift to oronasal breath-
ing in response to increasing ventilation rate and found that subjects who are nasal breathers
at rest move to oronasal breathing at a mean minute volume (Ve) of 30 liters per minute. At
maximum exercise levels (90 liters/minute), subjects breathe, at most, 40% of the total minute
volume through the nose.
The results of such studies are extremely important in aiding our understanding of
results reviewed here as being derived from controlled human exposure studies of PM and S0«.
Sulfur dioxide, for example, is very soluble in water and, when inhaled nasally, is readily
(95 to 99 percent) absorbed on the moist surfaces of the nose and upper respiratory passages
(Frank et al., 1973). This, in fact, may protect individuals breathing nasally at rest from
even relatively high levels of SQp exposure. At some level of ventilation, however, breathing
shifts from nasal to oronasal, thereby increasing the dose of SO- reaching the tracheobron-
chial region of the lung and probably leading to enhanced S0? effects at ambient exposure
levels below those affecting the same individuals while breathing nasally at rest or at lower
activity levels. Forced oral breathing yields less nasopharyngeal absorption than either
nasal or oronasal breathing and would be expected to yield a more intense exposure-effect re-
lationship than is observed with either nasal or oronasal breathing. As summarized below,
these expected patterns can be discerned clearly when examining the results of available con-
trolled human exposure studies, especially in regard to SOp effects.
13.6.1 Sulfur Dioxide Effects
Sulfur dioxide has been found to affect a variety of physiological functions. These in-
clude sensory processes, subjective perceptions of irritative or painful SOp effects, and more
objectively measured changes in respiratory function parameters. Although the reliability of
subjective reports of perceived effects of S0~ has been questioned by some, certain statements
can be made with confidence concerning SOp effects on sensory processes. For example, exposure
to 5 ppm of SOp results uniformly in the detection of the odor of SOp, while odor detection

13-47
-------
below that level varies considerably. Other changes (e.g., alterations in electroencephalo-
gram alpha rhythms or an impact on response of the dark adapted eye to light) have been
reported to occur at SQ~-exposure levels as low as 0.20 to 0.23 ppm. However, the health sig-
nificance of such "sensory effects" is unclear at this time, but would appear to be of rela-
tively little concern unless any resulting discomfort or other outcome would markedly alter
normal activities of affected subjects.
Of much more concern are cardiovascular or respiratory effects found to be associated
with exposure to S09. For healthy subjects at rest, in general, such effects have not been
3
consistently observed except at exposure levels above 5 ppm (13.1 mg/m ). These include, for
example, observations by Frank et al. (1962) of marked pulmonary flow resistance increases
(mean = 39%) at 5 ppm (13.1 mg/m ) and consistent observations by numerous other investigators
listed in Table 13-2 of increased airway resistance or other bronchoconstrictive effects with
exposures of healthy adult subjects to S02 levels of 5 ppm (13.1 mg/m ) or higher. Only Amdur
et al. (1958) has reported observations of significant cardiorespiratory effects in healthy
o
adults at rest following S00 exposures below 5 ppm (13.1 mg/m ), including exposures as low as
3
1 ppm (2.6 mg/m ). Other investigators (e.g., Lawther, 1955; Frank et al., 1962) have not
observed similar results in attempting to replicate the findings of Amdur et al. (1958) at
o
levels below 5 ppm (13.1 mg/m ). Numerous explanations could be offered for this apparent
discrepancy in reported exposure-effect relationships for bronchoconstrictive effects in
healthy adults at rest, but no clear resolution of the issue is presently available. Never-
3
theless, available evidence points to 5.0 ppm (13.1 mg/m ) as being the most probable lowest
observed effect level for induction of bronchoconstriction effects in healthy adults exposed
to SOp while at rest.
Probably of more crucial importance are the findings of several investigators suggesting
potentiation of SO, airway effects in normal subjects as the result of increased oral inhala-
tion of SO,, due either to forced mouth breathing or increased exercise levels or both. As
3
indicated in Table 13-2, for example, deep breathing of S0? at 1 ppm (2.6 mg/m ) increased
SR significantly in comparison to breathing air alone (Lawther, 1975). Also, Melville
9W
(1970) reported greater decreases in SG with oral breathing than nasal breathing at 2.5 ppm
*j 3W
(6.6 mg/m ) S02; and Snell and Luchsinger (1969) found significant decreases in MEFgrw at 1
ppm (2.6 mg/m3) SOg with oral breathing at rest but not at 0.5 ppm (1.3 mg/m ) SOp. Simi-
larly, Jaeger et al. (1979) observed no pulmonary effects in resting normal subjects with
forced oral breathing at 0.5 ppm (1.3 mg/m ) SO,. These studies suggest possible bronchocon-
3
striction effects in healthy adults with oral breathing of 1.0 to 2.5 ppm (2.6 to 6.6 mg/m )
SOp, raising the possibility of such effects being seen at similar concentrations in healthy
adults exercising at sufficient workloads to induce a shift to oronasal breathing.
Examining the effects of exercise, Kreisman et al. (1976) found that light exercise poten-
tiated the effect of SO,,, with VIEF«Q0, being significantly decreased with exercise during oral
13-48
-------
o
exposure of normal subjects to 3 ppm (7.9 mg/m ) SOp or above. Another study, by Bates and
Hazucha (1973), reported a 20 percent (but not statistically significant) decrease in MEFR
with 0.75 ppm (2.0 mg/m ) exposure of exercising adults in an open chamber; and Stacy et al.
(1981) reported slight but statistically significant SR increases in healthy adults exposed
•^ ctW
to 0.75 ppm (2.0 mg/m ) SCL while exercising in a controlled exposure chamber. These effects
were the only significant ones found from among numerous pulmonary function tests even under
rather extreme exercise conditions employed in the Stacy et al. (1981) study. These results
(Bates and Hazucha, 1973; Stacy et al., 1981), therefore, provide only very weak evidence for
3
effects in exercising healthy adults at SCL levels <1.0 ppm (2.6 mg/m ). In other studies, no
pulmonary effects were observed with chamber exposures of exercising healthy adults at S0~
3
exposure levels of 0.50, 0.40, or 0.37 ppm (1.31, 1.05, or 0.97 mg/m ) (Horvath and Folinsbee,
1977; Bedi et al., 1979; Bates and Hazucha, 1973; Hazucha and Bates, 1975; Bell et a«!., 1977b;
Linn et al. , 1980; Bedi et al., 1981). The weight of available evidence, therefore, appears
to indicate that induction of pulmonary mechanical function effects may occur at > 1 to 3 ppm
3 ' ~ 3
(2.6 to 7.9 mg/m ) SO^ in exercising healthy adults but not at <_ 0.50 ppm (1.31 mg/m ) SO,
even with exercise or forced oral breathing.
In attempting to define populations at special risk for SOp effects, Weir and Bromberg
(1972) and Reichel (1972) exposed patients with obstructive pulmonary disease to S09 levels
3
across the range of 0.3 to 4.0 ppm (0.8 to 10.5 mg/m ) and observed no statistically signifi-
cant increase in airway resistance or other pulmonary function effects. The exposures were
carried out while the subjects were at rest in a controlled exposure chamber, but no assess-
ment was conducted regarding possible enhanced effects of increased oral inhalation due to
exercise or forced mouth breathing. Thus, although no evidence was obtained for increased
susceptibility of these patients at rest, possibly enhanced vulnerability to SOg effects of
such subjects at elevated activity levels cannot be ruled out based on the reported results.
A clearer picture of probable enhanced susceptability or special risk for SO^-pulmonary
function effects appears to be emerging now in regard to asthmatic subjects. For example,
Jaeger et al. (1979) reported observing small, statistically significant (mean 2.7%) decreases
in MMFR levels (which recovered in 30 minutes) following forced oral exposure (by use of nose
clips) to 0.5 ppm (1.3 mg/m ) SO/, of 40 asthmatic subjects at rest in a controlled exposure
chamber. Two subjects experienced delayed effects requiring medication that may have been due
to the SQy exposures. (Other uncontrolled factors, however, cannot be ruled out as possibly
having caused the delayed symptoms.) While the small pulmonary function decrements observed
by Jaeger et al. (1979) may be physiologically insignificant per se, they are suggestive of
possible S0? effects occurring in asthmatics at S0? levels below those affecting nonsensitive
healthy adults.
Consistent with this possibility, Sheppard et al. (1980) observed statistically signifi-
cant SR=l-/ increases in clinically defined mild asthmatics with oral exposures to 1, 3, or 5
aw
aw o
ppm (2.6, 7.7 or 13.1 mg/m ) SO™ via mouthpieces while at rest but observed significant SR
13-49
-------
« 3
increases in normal and atopic subjects only at 5 ppm (13.1 mg/m ). In further studies,
Sheppard et al. (1981) observed statistically significant increases in SR with oral exposure
•3 3W
of asthmatics to 0.25 and 0.5 ppm (0.7 and 1.3 mg/m ) S0? via forced mouth breathing while
exercising at moderately elevated level (Ve = 30 liters/minute). The two most responsive
subjects of six tested experienced increased SR , with oral exposure to levels as low as 0.10
•3 SW
ppm (260 |jg/m ) SO^. At 0.5 ppm three of the subjects experienced wheezing and shortness of
breath, and at 1.0 ppm all six subjects experienced such symptoms. Sheppard et al. (1980)
also employed pharmacologic tests, which indicated that the very rapid onset bronchoconstric-
tive effects seen in the asthmatics are under parasympathetic neural control, as was earlier
demonstrated (Nadel et al., 1965) to be the case for normal subjects experiencing bronchocon-
striction in response to exposure to SO^ at a higher level (i.e., 5 ppm) while at rest.
The Sheppard et al. (1980, 1981) results appear to demonstrate that some asthmatic sub-
jects may be approximately an order of magnitude more sensitive to SO, exposure than normal,
nonsensitive healthy adults. That is, whereas nonsensitive healthy adults display increased
bronchoconstriction at 5 to 10 ppm while at rest and at levels possibly as low as 1 ppm with
oral or oronasal breathing, clinically defined mild asthmatics appear to be sensitive, as a
group, down to 0.25 ppm SQp and the most sensitive (as individuals) possibly down to 0.1 ppm
under moderate exercise (Ve ~ 30 liters/minute) conditions. Most importantly, with brief 10
minute exposures to S0? concentrations encountered in U.S. cities (0.1 to 0.5 ppm), Sheppard
et al. (1981) demonstrated that moderate exercise increased the bronchoconstriction produced
by SO,, in subjects with mild asthma. The results were qualitatively confirmed by Linn et al.
(1982) using techniques similar to those employed by Sheppard et al. (1981). In this pilot
study five asthmatics were exposed, via mouthpiece, to 0.5 ppm SO, for a period of 10 minutes
while exercising (~ 400 kg-m/min). Similar results using face mask have been recently
described (see Appendix, Chapter 13, for an abstract summarizing pertinent research recently
completed and described in a manuscript submitted for peer-review and publication). However,
caution should be employed in regard to any attempted extrapolation of these observed quanti-
tative exposure-effect relationships to what might be expected under ambient conditions.
Additional research results from studies using open chamber oronasal breathing conditions more
analogous to those encountered in daily activities have recently been described by Linn et al.
(1982). In this large-scale chamber study employing 24 asthmatic subjects, no statistically
significant pulmonary function decrements were found with 0.5 ppm SOy exposures for 1 hour
under intermittent exercising conditions. These negative results are in contrast to the
findings of Sheppard et al. \1981) and Linn et al. (1982) obtained with 0.5 ppni SO,, exposure
via mouthpiece while exercising. These differences may be due to the delivery of a higher
proportion of inhaled S02 to the tracheobronchial and lung regions with mouthpiece exposure or
to individual variations in bronchial reactivity to SO- among subjects used in the different
studies.
13-50
-------
The health significance of pulmonary function changes and associated symptomatic effects
demonstrated to occur in response to SO, by the above human exposure studies is an important
issue for present air quality criteria development purposes. In contrast to the sensory
effects of SOp earlier described as probably being of little health significance, much more
concern is generally accorded to the potential health effects of pulmonary function changes
(such as increased bronchoconstriction)' and associated symptomatic effects (such as coughing,
wheezing and dyspnea or shortness of breath) observed with human exposures to SCL, especially
in sensitive population groups such as asthmatics. Temporary, small decrements in pulmonary
airway functions observed in some of the above studies for nonsensitive healthy adults at SCL
concentrations = 1 to 5 ppm are generally of less concern in terms of their implications re-
garding the potential health impact of ambient air SO,, exposures than are the pulmonary func-
tion and symptomatic effects observed in mild asthmatics at similar (1 to 5 ppm) or lower (< I
ppm) concentrations of SO,- Probably of most concern are marked increases (> 10 percent) in
airway resistance and symptomatic effects (wheezing, dyspnea) observed by Sheppard et al .
(1981) in a group of mild asthmatics with oral exposure via mouthpiece to 0.5 ppm (1.3 mg/m )
sulfur dioxide during exercise. A recent article (Fischl et al . , 1981) and accompanying
editorial (Franklin, 1981) in the medical literature discuss the inclusion of indices of air-
way obstruction and presenting symptoms such as wheezing and dyspnea among factors to be con-
sidered in attempting to predict the need for hospital ization of asthma patients following
initial emergency room treatment (e.g., bronchodilator therapy, etc.) for asthmatic attacks.
Particulate matter, especially hygroscopic salts, have been shown to be potentially
important in enhancing the pulmonary function effects of S0? exposure. Airway resistance in-
creased more after combined exposure to S0? and sodium chloride in several studies, although
others have failed to demonstrate the same effect. This difference in response to the
SCL-NaCl aerosol mixtures may be due principally to the relative humidity at the time of the
exposure. McJilton et al. (1976) have demonstrated in guinea pigs that changes in pulmonary
mechanical function were seen only when the mixture (SOp/Nacl) was administered at high
relative humidity (r.h. > 80%). The effect is ascribed to absorption of the highly soluble
SOp into the droplet before inhalation, whereas at a r.h. < 40% the aerosol was a crystal.
Significant reduction in MEF™™ VC (maximal expiratory flowrate) was observed for the
group-mean after oral exposure to a combination of saline aerosol and 13.3 mg/m (5 ppm) S09;
3
however, no effects were observed at S09 levels of 0.5 ppm and 1.0 ppm (1.3 mg/m and 2.6
3
mg/m ) (Snell and Luchsinger, 1969). The validity of this study has been questioned based on
the lack of an air sham control group and also based on the methodology used to measure
VC, More recently, studies have been reported showing pulmonary function changes in extrinsic
asthmatics both at rest (Koem'g et al., 1980) and during exercise (Koenig et al . , 1981) with
Q O
exposure to 2.62 mg/m (1 ppm) SO, and 1 mg/m NaCl. Statistically significant decreases in
V cno/ and V -,,-<*• were observed both at rest and during exercise for asthmatics but not
fflclX DUA> IH9X / *3%
13-51
-------
for all normals. Although NaCl alone produced no such effects, the lack of an "S0? .alone"
group and the difference in the number of subjects used with NaCl alone and in combination
with S0? make interpretation difficult.
In contrast to the apparent enhancement of S0?-induced pulmonary airway effects by com-
bined exposure with certain particulate matter aerosols, there is less evidence that supports
the hypothesis that synergistic interactions between S0? and other gaseous pollutants, such as
0_ or NO,, produce greater-than-additive effects cf each individually on pulmonary mechanical
functions.
Controlled human exposure study evidence regarding SQ« effects on respiratory defense
mechanisms, such as mucus clearance processes, is highly limited at present. For healthy
adults exposed to S00 while at rest, nasal mucus flowrate appeared to decrease markedly (by 50
3
percent) at 5.0 ppm (13.1 mg/m ) S02 (Andersen et al., 1977), but tracheobronchial mucociliary
clearance appeared to be unaffected by exposure at the same SO, level while at rest (Wolff et
al,, 1975a). These observed differences may be due to the much greater dose of SCL delivered
to nasal passages than to tracheobronchial regions by nasal breathing at rest. Oral exposure
of healthy adults to 5.0 ppm (13.1 mg/m ) S0? during exercise (which notably increases
tracheobronchial deposition of S0?), however, was observed in two studies (Wolff et al.,
1975b; Newhouse et al., 1978) to increase tracheobronchial clearance rates. No studies, to
date, have investigated whether or not repeated exposures to 5.0 ppm (13.1 mg/m ) S02 would
continue to induce increased nasal or tracheobronchial clearance or, possibly, cause eventual
slowing of mucus clearance. (Note that one early study by Cralley [1942] reported decreased
mucociliary activity in healthy adults exposed to high (> 15 ppm) S0? concentrations while at
rest.) Nor have any controlled exposure studies investigated the effects of SO, exposure on
mucus clearance activities in asthmatics or other potentially sensitive human population
groups, such as individuals with chronic obstructive pulmonary diseases. Thus, while S02
effects on nasal and tracheobronchial mucus clearance processes cannot be said to have been
demonstrated to occur in sensitive population groups at exposure levels below those affecting
healthy adults, such a possibility cannot be ruled out at this time.
13.6.2 Sulfuric Acid and Sulfate Effects
In addition to SO,, being absorbed by hygroscopic particles, whereby its effects may be
potentiated, sulfur dioxide is also transformed into sulfur trioxide during transport and (in
combination with moisture) sulfuric acid is formed. The latter may exist as a sulfuric acid
droplet or can be converted to sulfates in the presence of ammonia, which is found in the
ambient air and in expired human breath.
Sulfuric acid and other sulfates have been found to affect both sensory and respiratory
function in study subjects. The odor threshold for sulfuric acid has been estimated to be at
3 3
0.75 mg/m based on one study and at 3.0 mg/m based on another.
3
Respiratory effects from exposure to sulfuric acid mist (0.35 to 0.5 mg/m ) have been
reported to include increased respiratory rate and decreased maximal inspiratory and expiratory

13-52
-------
flowrates and tidal volume (Amdur et al., 1952). However, several other studies of pulmonary
function in nonsensitive healthy, adult subjects (Newhouse et al,, 1978; Sackner et al., 1978;
Kleinman et al, , 1978; Avol et al., 1979; Leikauf et al., 1981; Kerr et al., 1981; Horvath et
al., 1981) indicated that pulmonary mechanical function was little affected when subjects were
exposed at 0.1 to 1.0 mg/m sulfuric acid for 10 to 120 minutes, although in one study (Utell
et al., 1981) the bronchoconstrictor action of carbachol was potentiated by the sulfuric acid
and sulfate aerosol, more or less in relation to their acidity.
In regard to mucociliary clearance effects, tracheobronchial clearance was significantly
3 3
increased at 100 (jg/m HnSO., was not significantly altered at 300 ug/m , but significantly
3
decreased at 1000 M9/m (Leikauf et al., 1981). Although transiently depressed following a
3
single 60-minute exposure, the latter decreased clearance rates seen at 1000 ug/m raise the
possibility of more persistent or chronic depression of tracheobronchial clearance after
repeated exposures to the same concentrations of H^SO The possible occurrence of such an
effect in humans would be consistent with observations of persistently slowed clearance for
several months following repeated exposures of donkeys to comparable H?SO, concentrations
(Schlesinger et al., 1978, 1979).
In studies with asthmatic subjects, no changes in airway function have been demonstrated
2
after exposure to sulfuric acid and sulfate salts at concentrations less than 1000 pg/m . How-
•>
ever, at higher concentrations (1000 pg/m ), reduction in specific airway conductance (SG )
and forced expiratory volume (FEV, «) have been observed after sulfuric acid (H?SO^) and
ammonium bisulfate (NhLHSO,) exposures as reported by Utell et al. (1981). No studies, on the
other hand, have as yet evaluated the effects of sulfuric acid or other published sulfate salt
aerosols on nasal or tracheobronchial mucus clearance functions.
13.6.3 Effects of Other Particulate Matter Species
Water-soluble sulfates have been the most frequent ingredients of experimental aerosol
exposure atmospheres because ambient sulfate levels were earlier reported as likely being
associated with morbidity epidemiologically. However, in addition to sulfuric acid and sul-
fates, other nonsulfur particulate matter species exist in the ambient air. These include
POM, lead, arsenic, selenium, hydrogen ions, ammonium salts, and carbon as dust. Although con-
trolled human exposure to some of these inherently toxic compounds is forbidden for obvious
reasons, several investigators have conducted studies using carbon and other inert particles.
The relatively sparse results involving insoluble and other nonsulfur aerosols under con-
trolled human exposure conditions preclude drawing conclusions regarding quantitative exposure/
effect or dose/response relationships for the particulate chemical species studies. This is
due to the fact that extremely high aerosol concentrations were typically employed in such
studies. Nor can any clear conclusions be drawn, based on the available controlled human
exposure data, in regard to size ranges of insoluble and other nonsulfur aerosols that may be
associated with the induction of significant respiratory system effects at concentrations
13-53
-------
commonly found in the ambient air (although most of the controlled exposure studies generally
appear to have employed either fine mode-sized particles < 2.5 |jm diameter or inhalable par-
ticles < 10-15 |jm diameter). However, the effects in polydispersed aerosol studies cannot be
ascribed to fine particles alone. Only studies by McDermott (1962), Andersen et al. (1979),
and Toyoma (1964) have clearly studied the effects of larger particles but at highly elevated
levels of insoluble particulate matter not usually associated with ambient conditions.
13-54
-------
13.7 REFERENCES

Abe, M. Effects of mixed nitrogen dioxide-sulfur dioxide on human pulmonary functions. Bull.
Tokyo Med. Dent. Univ. 14:415-433, 1967.

Albert, R. E., J. R, Spiegelman, S. Shatsky, and M. Lippmann. The effect of acute exposure to
cigarette smoke on bronchial clearance in the miniature donkey. Arch. Environ. Health
18:30-41, 1969.

Amdur, M. 0. lexicological appraisal of particulate matter, oxides of sulfur and sulfuric
acid. J. Air Pollut. Control Assoc. 19:638-646, 1969.

Amdur, M. 0. Animal studies. |ni: Proceedings of the Conference on Health Effects of Air Pol-
lutants, National Academy of Sciences, Washington, DC, October 3-5, 1973. Serial No.
93-15, U.S. Senate, Committee on Public Works, Washington, DC, 1973. Washington, DC,
1973. pp. 175-205.

Amdur, M. 0. The long road from Donora. 1974 Cummings Memorial Lecture. Am. Ind. Hyg.
Assoc. J. 35:589-597, 1974.

Amdur, M. 0., W. W. Melvin, Jr., and P. Drinker. Effects of inhalation of sulfur dioxide by
man. Lancet 2:758-759, 1953.

Amdur, M. 0., L. Silverman, and P. Drinker. Inhalation of sulfuric acid mist by human sub-
jects. Arch. Ind. Hyg. Occup. Med. 6:305-313, 1952.

Andersen, I., L. Molhave, and D. F. Proctor. Human response to controlled levels of combina-
tions of sulfur dioxide and inert dust. Scan. J. Environ. Health 7:1-7, 1981.

Andersen, I., G. R. Lundquist, P. L. Jensen, and D. F. Proctor. Human response to controlled
levels of sulfur dioxide. Arch. Environ. Health 28:31-39, 1974.

Andersen, I., G. R. Lundquist, D. F. Proctor, and D. L. Swift. Human responses to controlled
levels of inert dust. Am. Rev. Resp. Dis. 119-:619-627, 1979.

Andersen, I. , P. L. Jensen, S. E. Reed, J. W. Craig, D. F. Proctor, and G. K. Adams. Induced
rhinovirus infection under controlled exposure to sulfur dioxide. Arch. Environ. Health
32:120-126, 1977.

Arthur D. Little Incorporated. Research on Chemical Odors. Part I. Determination of Odor
Thresholds for 53 Commercial Chemicals. The Manufacturing Chemists' Association, Wash-
ington, DC, January 1968.

Avol, E. L., M. P. Jones, R. M. Bailey, N. M-N. Chang, M. T. Kleinman, W. S. Linn, K. A. Bell,
and J. D. Hackney. Controlled exposures of human volunteers to sulfate aerosols. Health
effects and aerosol characterization. Am. Rev. Respir. Dis. 120:319-327, 1979.

Bates, D. V., and M. Hazucha. The short-term effects of ozone on the lung. In: Proceedings
of the Conference on Health Effects of Air Pollutants, National Academy of Sciences,
Washington, DC, October 35, 1973. Serial No. 93-15, U.S. Senate Committee on Public
Works, Washington, DC, 1973. pp. 507-540.

Bedi, J. F. , L. J. Folinsbee, S. M. Horvath, and R. S. Ebenstein. Human exposure to sulfur
dioxide and ozone: absence of a synergistic effect. Arch. Environ. Health 34:233-239,
1979.

6edi,,J. F. , S. M. Horvath, and L. J. Folinsbee. Human exposure to sulfur dioxide and ozone
in a high temperature-humidity environment. Am. Ind. Hygi. Assoc. J. , 1981 (in press).

13-55
-------
Bell, K. A., and J. D. Hackney. Effects of Sulfate Aerosols upon Human Pulmonary Function.
Coordinating Research Council, Inc. APRAC Project CA PM-27-75, 1977a.

Bell, K. A., W. S. Linn, M. Hazucha, J. D. Hackney, and D. V. Bates. Respiratory effects of
exposure to ozone plus sulfur dioxide in Southern Californians and Eastern Canadians.
Am. Ind. Hyg. Assoc. J. 38:696-706, 1977b.

Burton, G. G. , M. Corn, J. B. L. Gee, D. Vassallo, and A. Thomas. Absence of "synergistic
response" to inhaled low concentration gas-aerosol mixtures in healthy adult males. Pre-
sented at 9th Annual Air Pollution Medical Research Conference, Denver, Colorado, July
1968.

Burton, G. G., M. Corn, J. B. L. Gee, C. Vasallo, and A. P. Thomas. Response of healthy men
to inhaled low concentrations of gas-aerosol mixtures. Arch. Environ. Health 18:681-692,
1969.

Bushtueva, K. A. The determination of the limit of allowable concentration of sulfuric acid
in atmospheric air. In: Limits of Allowable Concentrations of Atmospheric Pollutants.
Book 3. B. S. Levine, translator, U.S. Department of Commerce, Office of Technical
Services, Washington, DC, 1957. pp. 20-36.

Bushtueva, K. A. Threshold reflex effect of S02 and sulfuric acid aerosol simultaneously
present in the air. In: Limits of Allowabre Concentrations of Atmospheric Pollutants.
Book 4. B. S. Levine, translator, U.S. Department of Commerce, Office of Technical
Services, Washington, DC, January 1961. pp. 72-79.

Bushtueva, K. A. New studies of the effect 'of sulfur dioxide and of sulfuric acid aerosol on
reflex activity of man. In: Limits of Allowable Concentrations of Atmospheric Pollut-
ants. Book 5. B. S. Levine, translator, U.S. Department of Commerce, Office of
Technical Services, Washington, DC, March 1962. pp. 86-92.

Camner, P., K. Strandberg, and K. Philipson. Increased mucociliary transport by colinergic
stimulation. Arch. Environ. Health 29:220-224, 1974.

Camner, P., K. Strandberg, and K. Philipson. Increased mucociliary transport by andenergic
stimulation. Arch, Environ. Health 31:79-82, 1976.

Constantine, H., L. Dautrebande, N. Kaltreider, F. W. Lovejoy Jr., P. Morrow, and P. Perkins.
Influence of carbachol, and of fine dust aerosols upon the breathing mechanics and the
lung volumes of normal subjects and of patients with chronic respiratory disease before
and after administering sympathominetic aerosols. Arch. Int. Pharmacodyn 123:239-251,
1959.

Corn, M. and R. T. Cheng. Interactions of sulfur dioxide with insoluble suspended particulate
matter. J. Air Pollution Control Assoc. 22:870-875, 1972.

Cralley, L. V. The effect of irritant gases upon the rate of ciliary activity. J. Ind. Hyg.
and Toxicol. 24:193-198, 1942.

D'Alfonso, D. A. The Limiting Factors of Nasal Respiration. Ph.D. Thesis, University of Cal-
ifornia, Santa Barbara, 1980.

Dodge, R. R., and B. Burrows. The prevalence and incidence of asthma and asthma-like symptoms
in a general population sample. Am. Rev. Respir. Dis. 122:567-575, 1980.

Dubois, A. B. , and L. Dautrebande. Acute effects of breathing inert dust particles and of
carbachol aerosol on the mechanical characteristics of the lungs in man. Changes in
response after inhaling sympathomimetic aerosols. J. of Clin. Invest. 37:1746-1754,
1958.

13-56
-------
Dubrovskaya, F. I. Hygienic evaluation of pollution of atmospheric air of a large city with
sulfur dioxide gas. In: Limits of Allowable Concentrations of Atmospheric Pollutants.
Book 3. B. S. Levine, translator, U.S. Department of Commerce, Office of Technical
Services, Washington, DC,, 1957. pp. 37-51.

Fishchl, M. A., A. Pitchenik, and L. B. Gardner. An index predicting relapse and need for
hospitalization in patients with acute bronchial asthma. N. Engl. J. Med. 305:783-789,
1981.

Frances, R., D. Alessandro, M. Lippmann, D. E. Proctor, and R. E. Albert. Effect of cigarette
smoke on particle transport on nasociliary mucosa of donkeys. Arch. Environ. Health 21:
25-31, 1970

Frank, N. R. Studies on the effects of acute exposure to sulfur dioxide in human subjects.
Proc. R. Soc. Med. 57:1029-1033, 1964.

Frank, N. R. , M. 0. Amdur, and J. L. Whittenberger. A comparison of the acute effects of S0?
administered alone or in combination with NaCl particles on the respiratory mechanics of
healthy adults. Int. J. Air Water Pollut. 8:125-133, 1964.

Frank, N. R., M. 0. Amdur, J. Worcester, and J. L Whittenberger. Effects of acute controlled
exposure to S0? on respiratory mechanics in hea.lthy male adults. J. Appl. Physio!.
17:252-258, 1962.

Frank, R. , C. E. McJilton, and R. J. Charlson. Sulfur oxides and particles; effects on pul-
monary physiology in man and animals. In: Proceedings of Conference on Health Effects
of Air Pollution. National Academy of Sciences, Washington, DC, October 3-5, 1973.
Serial No. 93-15, U.S. Senate, Committee on Public Works, Washington, DC, 1973. pp.
207-225.

Franklin, W. Asthma in the emergency room. Assessment and treatment. N. Engl. J. Med.
305:826-827, 1981.

Gokenmeijer, J. D. M., K. DeVries, and N. G. M. Orie. Response of the bronchial tree to
chemical stimuli. Rev. Inst. Hyg. Mines (Ha'sselt) 28:195-197, 1973.

Greenwald, I. Effects of inhalation of low concentrations of sulfur dioxide upon man and
other mammals. Arch. Ind. Hyg. Occup. Med. 10:455-475, 1954.

Gunnison, A. F. , and E. D. Palmes. S-Sulfonates in human plasma following inhalation of
sulfur dioxide. Am. Ind. Hyg. Assoc. J. 35:288-291, 1974.

Hazucha, M., and D. V. Bates. Combined effect of ozone and sulphur dioxide on human pulmonary
function. Nature (London) 257:50-51, 1975.

Holmes, J. A., E. C. Franklin, and R. A. Gould. Report of the Selby Smelter Commission.
Bureau of Mines Bulletin 98, U.S. Department of the Interior, Washington, DC, 1915.

Horvath, S. M. , and L. J. Folinsbee. Interactions of Two Air Pollutants, Sulfur Dioxide and
Ozone, on Lung Functions. Grant ARB-4-1266, California Air Resources Board, Sacramento,
CA, March 1977.

Horvath, S. M. , L. J. Folinsbee, and J. F. Bedi Effects of large (0.9 (jm) sulfuric acid
aerosols on human pulmonary function. 'Environ. Res., 1981 (in press).
13-57
-------
Islam, M. S., and W. T. Ulmer. The effects of long-time exposure (8 h per day on 4 successive
days) to a gas mixture of SCL + NCL + and CU in the threefold MIC range (maximum emission
concentration) on lung function and reactivity of the bronchial system of healthy
persons. Wissenschaft und Unwelt 4:186-190, 1979(b).

Islam, M. S. , and W. T. Ulmer. The influence of acute exposure against a combination of 5.0
ppm S0?, 5.0 ppm N0?, and 0.1 ppm 0-, on the lung function in the MAK (lower toxic limit)
area (Short-time test). Wissenschaft und Unwelt 3:131-137, 1979(a).

Jaeger, M. J., D. Tribble, and H. J. Wittig. Effect of 0.5 ppm sulfur dioxide on the respira-
tory function of normal and asthmatic subjects. Lung 156:119-127, 1979.

Kagawa, J., and K. Tsuru. Respiratory effect of 2-hour exposure with intermittent exercise to
ozone and sulfur dioxide alone and in combination in normal subjects. Jap. J. Hyg.
34:690-696, 1979.

Kerr, H. D. , T. J. Kulle, B. P. Parrel 1, L. R. Sauder, J. L. Young, D. L. Swift, and R. M.
Borushok. Effects of sulfuric acid aerosol on pulmonary function in human subjects.
Environmental Research, 26:42-50, 1981.

Kisskalt, K. Uber den Einfluss der inhalation schwelfiger Saure auf die Entevickelung der
Lungentuberculose: Eip Bietrag zum Studien der Gewerbekrankheiten. [Effects of inhaling
sulfuric acid on the development of tuberculoses of the lungs. A continuation to studies
of occupational diseases.] Z. Hyg. 48:269-279, 1904.

Kleinman, M. T. , and J. D. Hackney. Effects of sulfate aerosols upon human pulmonary
function. APRAE Project CAPM-27-75, Coordinating Research Council, Inc., New York, NY,
1978.

Kleinman, M. T., R. M. Bailey, Y. C. Chang, K. W. Clark, M. P. Jones, W. S. Linn, and J. D.
Hackney. Exposures of human volunteers to a controlled atmospheric mixture of ozone,
sulfur dioxide and sulfuric acid. Am. Indus. Hyg. Assoc. J. 42:61-69, 1981.

Koenig, J. Q. , W. E. Pierson, and R. Frank. Acute effects of inhaled S0? plus NaCl droplet
aerosol on pulmonary function in asthmatic adolescents. Environ. Res. 22:145-153, 1980.

Koenig, J. Q. , W. E,. Pierson, M. Horike, and R. Frank. Effects of S0? plus NaCl aerosol com-
bined with moderate exercise on pulmonary function in asthmatic adolescents. Environ.
Res. 25:340-348, 1981.

Kreisman, H. , C. A. Mitchell, H. R. Hosein, and A. Bouhuys. Effect of low concentrations of
sulfur dioxide on respiratory function in man. Lung 154:25-34, 1976.

Lawther, P. J. Effects of inhalation of sulfur dioxide on respiration and pulse rates in
normal subjects. Lancet 2:745-748, 1955.

Lawther, P. J. , A. J. MacFarlane, R. E. Waller, and A. G. F. Brooks. Pulmonary function and
sulphur dioxide, some preliminary findings. Environ. Res. 10:355-367, 1975.

Lehmann, K. B. Experimentelle Studien liber den Einfluss technisch und hygienisch wichtiger
Case und Dampfe auf den Organismus. VI. Schwefliger Saure. [Experimental studies on
the effects on the system of technically and hygienically important gases and fumes.. VI.
Sulfuric Acid.] Arch. Hyg. 18:180-191, 1893.

Leikauf, G., D. B. Yeates, K. A. Wales, D. Spedtor, R. E. Albert, and M. Lippmann. Effects of
sulfuric acid aerosol on respiratory mechanics and mucociliary particle clearance in
healthy nonsmoking adults. Am. Ind. Hyg. Assoc. J. 42:273-282, 1981.
13-58
-------
Linn, W. S. , R. M. Bailey, D. A. Medway, J. G. Venet, L. H. Wightman, and J. 0. Hackney.
Respiratory responses of young adult asthmatics to sulfur dioxide exposure under simu-
lated ambient conditions. Environ. Res. 1982 (in press).

Linn, W. S. , M. P. Jones, R. M. Bailey, M. T. Kleinman, C. E. Spier, R. D. Fischer, and J. D.
Hackney. Respiratory effects of mixed nitrogen dioxide and sulfur dioxide in human
volunteers under simulated ambient exposure conditions. Environ. Res. 22:431-438, 1980.

Lippman, M. , R. E. Albert, D. B. Yeats", K. Wales, and G.* LeiKauf. Effect of sulfuric acid
mist on mucociliary bronchial clearance in healthy non-smoking humans. In: Aerosols in
Science, Medicine and Technology - The Biomedical Influence of the Aerosol, Association
for Aerosol Research, Dusseldorf, German (FRG), October 3-5, 1979. W. Stober and R.
Jaenicke, eds. , Association for Aerosol Research, Mainz, Germany (FRG), 1980. pp.
157-162.

McDermott, M. Acute respiratory effects of the inhalation of coal-dust particles. J, Physio!.
162:53 p. 1962.

McJilton, C. E. , R. Frank, and R. J. Charlson. Influence of relative humidity on functional
effects of an inhaled SQ~-aerosol mixture. Am. Rev. Respir. Dis. 113:163-169, 1976.

McKerrow, C. B. Chronic respiratory disease in Great Britain. Arch. Environ. Health 8:
182-187, 1964.

Melville, G. N. Changes in specific airway conductance in healthy volunteers following nasal
and oral inhalation of S00. West Indian Med. J. 19:231-235, 1970.
C-
Nadel, J., H. Salem, B. Tamplin, and Y. Tokiwa. Mechanism of bronchoconstriction during inha-
lation of sulfur dioxide. J. Appl. Physio!. 20:164-167, 1965.

Nakamura, K. Response of pulmonary airway resistance by interaction of aerosols and gases in
different physical and chemical nature. Nippon Eiseigaku Zasshi 19:38-50, 1964.

Newhouse, H. T. , M. Dolovich, G. Obminski, and R. K. Wolff. Effect of TLV levels of SO, and
FLSQ, on bronchial clearance in exercising man. Arch. Environ, Health 33:24-32, 1978.

Niinimaa, V., P. Cole, S. Mintz, and R. J. Shephard. Oronasal distribution of respiratory
flow. Resp. Physiol. 43:69-75, 1981.

Niinimaa, V., P. Cole, S. Mintz, and R. J. Shephard. The switching point from nasal to oro-
nasal breathing. Resp. Physiol. 42:61-71, 1980,

Morris, R. M., and J. M. Bishop. The effect of calcium carbonate dust on ventilation and res-
piratory gas exchange in normal subjects and in patients with asthma and chronic
bronchitis. Clin. Sci. 30:103-115, 1966.

Ogata, M. Uber die Giftigkeit der schweffigen Saure. [The Toxicity of Sulfuric Acid.] Arch.
Hyg. 2:223-245, 1884.

Reichel, G. The effect of sulfur dioxide on the airway resistance of man. Annual Meeting of
the German Scoiety for Industrial Medicine, 1972.

Ryazanov, V. A. Sensory physiology as basis for air quality standards. Arch. Environ. Health
5:479-494, 1962.
13-59
-------
Sackner, M. A. , D, Ford, R. Fernandez, J. Cipley, D. Perez, H. Kwocka, M. Reinhart, E. D.
Michael son, R. Schreck, and A. Wanner. Effects of sulfuric acid aerosol on cardiopulmo-
nary function of dogs, sheep and humans. Am. Rev. Respir. Dis. 118:497-510, 1978.

Saibene, F., P. Magnoni, C. L. Lafortuna, and R. Mostardi. Oronasal breathing during
exercise. Pfluegers Arch. 378:65-69, 1978.

Schlesinger, R. B. , M. Lippmann, and R. E. Albert. Effects of short-term exposures to
sulfuric acid and ammonium sulfate aerosols upon bronchial airway function in the donkey.
Am. Ind. Hyg. Assoc. J. 39:275-286, 1978.

Schlesinger, R. B. , M. Halpern, R. E. Albert, and M. Lippmann. Effect of chronic inhalation
of sulfuric acid mist upon mucociliary clearance from the lungs of donkeys. J. Environ.
Pathol. Toxicol. 2:1351-1367, 1979.

Shalamberidze, 0. P. Reflex effects of mixtures of sulfur and nitrogen dioxides. Hyg. Sanit.
32:10-15, 1967.

Sheppard, 0., A. Saisho, J. A. Nadel, and H. A. Boushey. Exercise increases sulfur dioxide-
induced bronchoconstriction in asthmatic subjects. Am. Rev. Respir. Dis. 123:486-491,
1981.

Sheppard, D. , W. S. Wong, C. F. Uehara, J. A. Nadel, and H. A. Boushey. Lower threshold and
greater bronchomotor responsiveness of asthmatic subjects to sulfur dioxide. Am Rev.
Respir. Dis. 122:873-878, 1980.

Sim, V. M. , and R. E. Pattle. Effect of possible smog irritants on human subjects. J. Am.
Med. Assoc. 165:1908-1913, 1957.

Snell, R. E. , and P. C. Luchsinger. Effects of sulfur dioxide on expiratory flow rates and
total respiratory resistance in normal human subjects. Arch. Environ. Health 18:693-698,
1969.

Speizer, F. E., and N. R. Frank. A comparison of changes in pulmonary flow resistance in
health volunteers acutely exposed to S09 by mouth and by nose. Br. J. Ind. Med.
23:75-79, 1966a.

Speizer, F. E. , and N. R. Frank. The uptake and release of S09 by the human nose. Arch.
Environ. Health 12:725-728, 1966b. ^

Stacy, R. W. , D. E. House, M. Friedman, M. Hazucha, J. Green, L. Raggio, and L. J. Roger.
Effects of 0.75 ppm sulfur dioxide on pulmonary function parameters of normal human
subjects. Arch. Environ. Health, 36:172-178, 1981.

Stahlhofen, W. , J. Gebhart, and J. Heyden. Experimental determination of the regional deposi-
tion of aerosol particles in the human respiratory tract. Am. Ind. Hyg. Assoc. J.
41:385-398, 1980.

Tomono, Y. Effects of S0? on human pulmonary functions. Sangyo Igaku 3_:77-85, 1961.

Toyama, T. Studies on aerosols. Synergistic response of the pulmonary airway resistance of
inhaling sodium chloride aerosols and S0? in man. Sangyo Igaku 4:86-92, 1962.

Toyama, T. Air pollution and its helath effects in Japan. Arch. Environ. Health 8:153-173,
1964.
13-60
-------
Toyama, T., and K. Nakamura, Synergistic response to hydrogen perixide aerosols and sulfur
dioxide to pulmonary airway resistance. Ind. Health 2:34-45, 1964.

United States Department of Health, Education and Welfare. Prevalence of selected chronic
respiratory conditions in the United States - 1970. Publication No. (HRA) 74-1511, Sept.
1973 by Center for Health Statistics, Rockville, MD.

Ulmer, W. T, Inhalative noxen; schwefeldioxyd. Toxic Inhalation: Sulfur Dioxide.
Pneumonologie 150:83-96, 1974.

Utell, M. J. , A. T. Aquilina, W. J. Hall, D. M. Speers, R. G. Douglas, Jr., F. R. Gibb, P. E.
Harrow, and R. W. White. Development of airway reactivity to nitrates in subjects with
influenza. Am. Rev. Respir. Dis. 121:233-241, 1980.

Utell, M. J., P. E. Morrow, and R. W. Hyde. Inhaled Particles. V. Proceedings of the 5th
International Symposium, Cardiff, Wales, September 1980, Pergamon Press, London, England,
1981 (in press).

von Nieding, G. , H. M. Wagner, H. Krekeler, H. Lb'llgen, W. Fries, and A. Beuthan. Controlled
studies of human exposure to single and combined-action of N0?, 0, and S0?. Int. Arch.
Occup. Environ. Health 43:195-210, 1979. £ d

Weir, F. W., and P. A. Bromberg, Further investigation of the effects of sulfur dioxide on
human subjects. Annual Report Project No. CAWC S-15, American Petroleum Institute,
Washington, DC, 1972. pp. 1-74.

Weir, F. W. , and P. A. Bromberg. Effects of sulfur dioxide on human subjects exhibiting
peripheral airway impairment. Project No. CAWC S-15, American Petroleum Institute,
September 1973. pp. 1-40.

Widdicombe, J. G. , D. C. Kent, and J. A. Nadel. Mechanism of bronchoconstriction during
inhalation of dust. J. Appl. Physidogy 17:613-616, 1962,

Wolff, R. K. , M. Dolovich, C. M. Rossman, and M. T. Newhouse. Sulphur dioxide and
trachea-bronchial clearance in man. Arch. Environ. Health 30:521-527, 1975a.

Wolff, R. K. , M. Dolovich, G. Obminski, and M. T. Newhouse. In: Effect of sulfur dioxide on
trachiobronchial clearance at rest and during exercise. Inhaled Particles, Proceedings
of the 4th International Symposium, Edinburgh, Scotland, September 22-26, 1975. W, H.
Walton, ed. , Pergamon Press, London, England, 1975b. pp. 321-332.

Yamada, J. Untersuchunger uber die quantitative Absorption der Dampfe einiger Sauren durch
Tier und Menscn. [Investigations on the quantitative absorption of some acid fumes in
animals and man.] Ph.D. Thesis, Wurzburg, 1905. (See Lehmann, K. B., Arch. Hyg.
67:57-98, 1908.)
13-61
-------
APPENDIX 13A

The following is an abstract of a manuscript recently submitted for peer review and pub-
lication, but not yet available in the published literature. This abstract is included here
strictly for informational purposes and cannot presently be definitely analyzed for possible
consideration as part of this criteria document.*

EFFECT OF THE ORONASAL BREATHING ROUTE ON THE BRONCHOCONSTRICTOR RESPONSE TO SULFUR DIOXIDE IN
EXERCISING ASTHMATIC SUBJECTS. M. B. Kirkpatrick, D. Sheppard, J. A. Nadel, H. A. Boushey.
Cardiovascular Research Institute, University of California, San Francisco.
V/e studied how the oronasal breathing route during exercise affects the bronchocon-
strictor response to inhaled sulfur dioxide (SO.,) in asthmatic subjects. In six subjects, we
compared the changes in specific airway resistance (SR ) caused by breathing humidified air
through a mouthpiece during 5 min of exercise on a bicycle ergometer (550 kpm/min) to the
changes caused by breathing humidified air plus 0.5 ppm of S0? by mouthpiece, by facemask, and
by facemask with the mouth occluded (nose breathing) during exercise. Breathing humidified
air plus 0.5 ppm 50^ by both mouthpiece and facemask significantly increased SR in all 6
subjects; breathing S09 by nose significantly increased SR. in 5 of 6 subjects. Although the
^ uW
increase in SR caused by breathing SOp varied considerably among subjects, for the group,
breathing S09 by all 3 routes increased SR (mouthpiece, from 6.8 ± 4.0 to 16.4 ± 9.0 L x cm
£. aW
H20/L/S [mean ± S. D.], facemask, from 7.4 ± 3.6 to 12.4 ± 5.9, nose only, 6.4 ± 2.7 to 10.6 ±
5.2) significantly more than breathing humidified air without SOy through a mouthpiece (from
7.2 ± 6.2 to 8.3 ± 6.8) (p <0.05 for each route of breathing S02 compared to breathing air
without 502). Breathing SOy through a mouthpiece increased SR significantly more than
breathing S0? by facemask with the mouth occluded (p <0.05) but not significanlty more than
breathing S02 by facemask (p <0.05). These results indicate that, although nasal breathing
partially protects against S0?-induced bronchoconstriction, both oral and oronasal breathing
low concentrations of S02 during exercise can cause significant bronchoconstriction in people
with asthma.
*Note: Since finalization of the present criteria document in December 1981, this and certain
other controlled human exposure studies of S02 effects on asthmatics appeared in the published
literature and were then later evaluated rn an addendum to the present document. That
addendum is included in Volume I after Chapter 1 (Executive Summary and Conclusions) of this
document.
13-62
-------
14. EPIDEMIOLOGICAL STUDIES ON THE EFFECTS OF PARTICULATE MATTER
AND SULFUR OXIDES ON HUMAN HEALTH

14.1 INTRODUCTION
This chapter evaluates epidemiological literature concerning health effects associated
with ambient air exposures to sulfur oxides and participate matter. The main focus of the
chapter is on: (1) qualitative characterization of human health effects associated with
exposure to airborne sulfur dioxide (S02), related sulfur compounds, and other particulate
matter (PM); (2) quantitative delineation of exposure-effect and exposure-response relation-
ships for induction of such effects; and (3) identification of population groups at special
risk for experiencing the effects at ambient exposure levels.
The epidemiological data discussed here both complement and extend information presented
as part of health effects analyses in preceding chapters (11,12,13) of this document. Those
chapters focus on information from animal toxicology and controlled human exposure studies,
which offer the advantage of characterizing, under well-controlled laboratory conditions,
differential patterns of respiratory tract deposition and clearance of: SO-; sulfates (SO.)
and sulfuric acid (H^SO.); and other particulate matter of varying size and chemical composi-
tion. Also, the animal toxicological studies provide evidence for notable health effects
occurring in mammalian species as the result of respiratory tract deposition of S0? and PM,
including: transient alterations in pulmonary functions; altered mucociliary clearance and
other respiratory tract defense mechanisms; and increased susceptibility to infection and
morphological damage seen especially after high-level or prolonged exposures. However, while
such results from animal studies are highly suggestive of analogous effects possibly being
induced in human beings, caution must be exercised in directly extrapolating the findings or
associated dose-effect relationships to human health. More direct delineation of quantitative
dose-effect or dose-response relationships is possible through controlled human exposure
studies, but such studies also have important limitations. For example, whereas controlled
human exposure studies have demonstrated S0« or PM induction of transient pulmonary function
decrements, altered mucociliary clearance patterns, and symptomatic effects consistent with
animal toxicology study findings, observation of such effects has generally been confined to
conditions involving single or a few repeated short-term (<3 hours) exposures, rather than
prolonged chronic exposure conditions. Also left unanswered by controlled human exposure
studies are questions concerning whether or not more severe effects (e.g., increased vulner-
ability to respiratory diseases or marked morphological damage) are associated with either
short-term or prolonged ambient exposure conditions.
Epidemiological studies, in contrast, offer several advantages beyond those of animal
toxicology or controlled human exposure studies. Health effects of both short- and long-term
pollutant exposures (including complex mixtures of pollutants) can be studied and sensitive
members of populations at special risk for particular effects at ambient air concentrations
14-1
-------
identified. Also, epidemiologies! evaluations allow for investigation of both acute and
chronic disease effects and associated human mortality. Epidemiologies! studies, then,
together with the results of controlled animal and human exposure studies, can significantly
contribute to more complete understanding of health effects of sulfur oxides and particulate
matter, especially in helping to characterize human health effects associated with those
pollutants under ambient conditions. Despite such advantages, however, important limitations
do exist with regard to the conduct, analysis, interpretation, and use of many of the avail-
able epidemiological studies on the health impact of SO,, and PM, as discussed next.
14.1.1 Methodological Considerations
•As noted by Lowrence (1976), epidemiological and other types of studies employed in
generating information relevant to human risk assessment typically focus on one or more of
four areas of investigation: (1) defining exposure conditions; (2) identifying adverse
effects; (3) relating exposures to effects; and (4) estimating overall risk of specific
population groups experiencing particular effects under various exposure conditions.
In relation to accomplishing these goals, one important limitation of most epidemiological
studies reviewed here has been less-than-optimum characterization of community air quality
parameters used to estimate exposures of population groups to varying atmospheric concentra-
tions of SO, and PM. Such characterization of air quality has generally involved relatively
crude estimates of levels of pollutants present, often allowing for only limited qualitative
statements to be made regarding exposure conditions—e.g., whether a given site or time period
had comparatively higher or lower atmospheric levels of SC^ or PM than some other site or time
period. Only rarely have the epidemiological studies relied on measurement methods or their
practical field applications that permit reasonably precise determinations of variations in
ambient levels of the pollutants so as to adequately quantitate S0? or PM levels associated
with observed health effects. Even when reasonable quantification of community air quality
parameters was achieved, however, the use of such data in estimating actual population expo-
sures has typically been further constrained by factors such as siting of air sampling devices
in relation to study populations, frequency and duration of sampling periods, activity
patterns of study population members, and contributions of indoor air pollution to overall
exposures of study groups. These, and other limitations noted, arise in part from the fact
that most of the presently-reviewed epidemiological studies utilized air quality monitoring
data obtained from sampling networks originally established for purposes other than health-
related research and, therefore, not optimally designed to provide the specific types or
quality of aerometric data needed for epidemiological assessment of health effects related to
SQy and PM. Therefore, the aerometric data reported should generally be viewed as yielding,
at best, only approximate estimates of actual study population exposures.
Adequate characterization of health effects associated with various S02 and PM exposure
conditions has represented a second major problem for most of the epidemiological studies
evaluated here. Various health endpoint measurements (mortality, morbidity, and indirect
14-2
-------
measures of morbidity) have been employed in such studies and each has advantages and dis-
advantages, as discussed elsewhere (Hill, 1965; Speizer, 1969; Holland, 1970; Higgins, 1974;
Goldsmith and Friberg, 1977; American Thoracic Society, 1978). Some health outcome measure-
ments have involved direct observations of signs and symptoms of disease states or objective
indicators thought to be associated with the occurrence of illnesses, e.g., patient visits to
hospitals or clinics or absenteeism from school or work. Direct quantification of health
effects has also included measurement of biochemical or physiological changes in study
populations, as in recording of pulmonary function changes by spirometry methods. Indirect
measures or indices of health effects have also been used, e.g., in gathering information on
frequency and duration of respiratory illnesses by telephone interviews, written question-
naires, or self-reported entries in diaries. The validity of such indirect measurements of
health effects, however, is highly dependent on the ability and motivation of respondents to
recall and report accurately past or present health-related events; this can be influenced by
numerous extraneous factors such as age, cultural and educational background, instructions
from experimenters, sequencing of questions, and interviewer variability or bias. Confidence
in results obtained by either direct or indirect measurement methods is enhanced if potential
interfering or biasing factors have been appropriately controlled for and, especially for in-
direct health endpoint measurements, if results have been validated against corroborating
evidence.
Adequately relating observed health effects to specific parameters of ambient exposure
conditions is another objective not often achieved by epidemiological studies reviewed below,
such that few allow for confident qualitative or quantitative characterization of S0? or PM
exposure-health effect relationships. For example, competing risks, such as cigarette smoking
and occupational exposures, may contribute to observed health effects results and, therefore,
usually must be controlled for or taken into account in order for much confidence to be placed
in reported air pollution-health effects relationships; however, many studies on S0? or PM
effects have not adequately controlled for such factors. Similarly, possible effects of other
covarying or confounding factors such as socioeconomic status, race, and meteorological para-
meters have not always been adequately controlled for or evaluated. Also, further compli-
cating the evaluation of the epidemiological data is the fact that exposure parameters are not
subject to experimenter control; thus, ambient levels of a given pollutant often vary widely
over the course of a study. This makes it extremely difficult to determine whether mean con-
centrations, peak concentrations, rapid fluctuations in levels, or other air quality factors
were most important as determinants of reported health effects. Significant covariation among
concentrations of S0?, PM, and other pollutants has also often made it very difficult to dis-
tinguish among their relative contributions to observed health effects.
Estimation of overall risk by means of epidemiological studies requires still further
steps beyond delineation of exposure-effect relationships that define exposure conditions
14-3
-------
(levels, durations, etc.) associated with induction of specific health effects. That is, risk
estimation also requires: (1) identification of particular population groups likely to mani-
fest health effects under exposure conditions of concern; and (2) ideally, determination of
numbers or percentages of such individuals (responders) likely to be affected at various
exposure or dose levels. Delineation of the former, i.e., identification of population groups
at special risk at lower exposure levels of SCL and PM than other groups, has only started to
be accomplished via the epidemiological studies reviewed here. Also, epidemiological delinea-
tion of quantitative dose-response (or, more correctly, exposure-response) relationships,
defining percentages of population groups likely to manifest a given health effect at various
levels or durations of exposure to S0« and PM, is largely lacking at this time.
Another limitation of epidemiological information reviewed here concerns its usefulness
in demonstrating cause-effect relationships versus merely establishing associations (which may
be non-causal in nature) between various health effects and S0? or PM. The interpretation of
epidemiological data as an aid in inferring causal relationships between presumed causal
agents and associated effects has been previously discussed by several expert committees or
deliberative bodies faced with evaluation of controversial biomedical issues (U.S. Surgeon
General's Advisory Committee on Smoking and Health, 1964; U.S. Senate Committee on Public
Works, Subcommittee on Air and Water Pollution, 1968). Among criteria selected by each group
for determination of causality, the following were included: (1) the strength of the associa-
tion; (2) the consistency of the association, as evidenced by its repeated observation by
different persons, in different places, circumstances and time; (3) specificity of the associa-
tion; (4) the temporal relationship of the association; (5) the coherence of the association in
being consistent with other known facts; (6) the existence of a biological gradient, or dose-
reponse curve, as revealed by the association; and (7) the biological plausibility of the
association. In discussing the use of such criteria, Hill (1965) further noted that strong
support for likely causality suggested by an association may be derived from experimental or
semi-experimental evidence, where manipulation of the presumed causative agent (its presence
or absence, variability in intensity, etc.) also affects the frequency or intensity of the
associated effects. Importantly, both Hill (1965) and the deliberative bodies or expert
committees were careful to emphasize, regardless of the specific set of criteria selected by
each, that no one criterion is definitive by itself nor is it necessary that all be fulfilled
in order to support a determination of causality. Also, Hill (1965) and several of the expert
groups noted that statistical methods cannot establish proof of a causal relationship in an
association nor does lack of "statistical significance" of an association according to
arbitrarily selected probability criteria necessarily negate the possibility of a causal
relationship. That is, as stated by the U.S. Surgeon General's Advisory Committee on Smoking
and Health (1964): "The causal significance of an association is a matter of judgment which
goes beyond any statement of statistical probability."
14-4
-------
14.1.2 Guidelines for Assessment of Epidemiclogical Studies

Taking into account the above methodological limitations, the following set of guidelines
can be stated by which to judge the relative scientific quality of epidemiologies! studies and
their findings reviewed here:

1. Was the quality of the aerometric data used- sufficient to allow for meaningful
characterization of geographic or temporal differences in study population pollutant
exposures in the range(s) of pollutant concentrations evaluated?

2. Were the study populations well defined and adequately selected so as to allow for
meaningful comparisons between study groups or meaningful temporal analyses of
health effects results?

3. Were the health endpoint measurements meaningful and reliable, including clear defi-
nition of diagnostic criteria utilized and consistency in obtaining dependent vari-
able measurements?

4. Were the statistical analyses employed appropriate and properly performed and inter-
preted, including accurate data handling and transfer during analyses?

5. Were potentially confounding or covarying factors adequately controlled for or taken
into account in the study design and statistical analyses?

6. Are the reported findings internally consistent, biologically plausible, and coher-
ent in terms of consistency with other known facts?

Few, if any, epidemiological studies deal with all of the above points in a completely
ideal fashion; nevertheless, these guidelines provide benchmarks for judging the relative
quality of various studies and for selecting the best for detailed discussion here.
Detailed critical analysis of all epidemiological studies on health effects of SO, and PM,
especially in relation to all of the above questions, represents an undertaking beyond the
scope of the present document Of most importance for present purposes are those studies
which provide useful quantitative information on exposure-effect or exposure-response rela-
tionships for health effects associated with ambient air levels of SO,, and PM likely to be
encountered in the United States during the next 5 years. Accordingly, the following criteria
were employed in selecting studies for detailed discussion in the ensuing text:

1. Concentrations of both S02 and PM were reported, allowing for potential evaluation
of their separate or combined effects.

2, Study results provide information on quantitative relationships between health
effects and ambient air S0? and PM levels of current concern (i.e., generally <_
1000 pg/m3). ^

3. Important methodological considerations were adequately addressed, especially
(a) in controlling for likely potentially confounding factors and (b) in carrying
out data collection, analysis, and interpretation so as to minimize errors or poten-
tial biases which could be reasonably expected to affect the results.

4. The study results have been reported in the open literature or are in press, typi-
cally after having undergone peer review.
14-5
-------
In addition, some studies not meeting all of the above criteria-are briefly mentioned or
discussed in the main text below as appropriate to help elucidate particular points concerning
the health effects of SCL and/or PM. Additional studies, evaluated by the present authors but
found to be of very limited usefulness for present criteria development purposes are noted in
Appendix 14A, along with annotated comments on specific methodological aspects associated with
each that limit their results to qualitative findings only or make clear attribution of
reported health effects to S0? or PM questionable based on their reported results.
As a starting point in this assessment, key information from Chapters 2 and 3 is summar-
ized regarding physical and chemical properties of S0? and PM indexed by air quality measure-
ments used in epidemiological studies evaluated in this chapter. The ensuing discussion of
community health epidemiological studies is then subdivided into two subsections: Section
14.3 deals with studies of acute mortality and morbidity effects most germane to development
of health criteria for possible short-term (e.g., 24 hour) ambient air standards; and Section
14.4 discusses studies of mortality and morbidity effects associated with chronic exposures
most pertinent for development of health criteria for long-term (annual average) ambient air
standards. The last section (14.5) provides an integrative summary and interpretation of the
overall pattern of results evaluated in preceding sections.
The extensive epidemiological literature on the effects of occupational exposures to S02
and PM presently available is not reviewed here for several reasons:
1. Such literature generally deals with effects of exposures to S02 or PM chemical spe-
cies at levels many times higher than those encountered in the ambient air by the
general population.
2. Populations exposed occupationally mainly include healthy adults, self-selected to
some extent in terms of being better able to tolerate exposures to S02 or PM
substances than more susceptible workers seeking alternative employment or other
groups often at special risk among' the general public (e.g., the old, the
chronically ill, young children, and asthmatics).
3. Extrapolation of observed occupational exposure-health effects relationships (or
lack thereof) to the general public (especially population groups at special risk)
could, therefore, be potentially misleading in terms of demonstrating health effects
among healthy workers at higher exposure levels than would affect susceptible groups
in the general population.
The occupational literature does, however, demonstrate links between acute high level or
chronic lower level exposures to S0? or many different PM chemical species and a variety of
health effects, including: pulmonary function changes; respiratory tract diseases; morpholog-
ical damage to the respiratory system; and respiratory tract cancers. The reader is referred
to National Institute of Occupational Safety and Health (NIOSH) criteria documents and other
assessments listed in Appendix 14B for information on health effects associated with occupa-
tional airborne exposures to SOp and various PM species.
14-6
-------
14.2 AIR QUALITY MEASUREMENTS
Of key importance for evaluation of epidemiological studies reviewed here is a clear
understanding of the limitations of the analytical methods employed in determining ambrent air
aerometric data (PM and S0~ levels) utilized in those studies. These methods are discussed in
more detail in Chapter 3.
14.2.1 Sillfur Oxides Measurements
Three main measurement methods or variations thereof were used to generate data cited for
sulfur dioxide (SOp) levels in epidemiological studies discussed below: (1) sulfation rate
(lead dioxide); (2) hydrogen peroxide; and (3) the West-Gaeke (pararosaniline) methods. With
sulfation rate methods, airborne sulfur compounds react with lead dioxide in a paste spread
over an atmospherically exposed plate or cylinder. The sulfur compound reaction rate is
2
expressed in SOo/cm /day; but the reaction is not S0?-specific, and atmospheric concentrations
of S02 or other sulfur compounds cannot be accurately extrapolated from the results, which are
markedly affected by variations in temperature and humidity. Lead dioxide gauges were widely
used in the United Kingdom prior to 1960 and provided aerometric data reported in some British
epidemiological studies; sulfation rate methods were also used in some American studies.
Use of the hydrogen peroxide method was gradually expanded in the United Kingdom during
the 1950s (usually in tandem with apparatus for PM (smoke) monitoring) and the method was
adopted in the early 1960s as the standard SO, method used for the United Kingdom National
Survey of Air Pollution and, as an OECD-recommended method, elsewhere in Europe. The method
2
can yield reasonably accurate estimates of atmospheric SO,, concentrations expressed in pg/m ;
but the results can be affected by factors such as temperature, presence of atmospheric
ammonia and titration errors. Little quality assurance information exists on sources and
magnitudes of errors encountered in the use of the method to obtain SO, data reported in
specific British or European epidemiological studies, making it difficult to assess the
accuracy or precision of reported SO, values. Even the extensive quality assurance informa-
tion reported (Warren Spring Laboratory, 1961; 1962; 1966; 1967; 1975; 1977; OECD, 1965;
Ellison, 1968) for SO, measurements made by the method for the United Kingdom National Survey
is of very limited use in evaluating the quality of specific S0? results reported in various
British epidemiological studies.
The West-Gaeke (pararosaniline) method more widely used in the United States involves
absorption of SO, in potassium tetrachloromercurate solution, producing a chemical complex
reacted with pararosaniline to form a red-purple color measured colorimetrically. The method,
suitable for sampling up to 24 hours, is specific for S02 if properly implemented to minimize
interference by nitrogen or metal oxides, but results can be affected by factors such as tem-
perature variations and mishandling of reagents. Limited quality assurance information (U.S.
Congress, House of Representatives, 1976) has been reported for some American S0? measurements
by the West-Gaeke method but is generally lacking by which to evaluate the SO, data reported
in most published American epidemiological studies.
14-7
-------
Measurement approaches for suspended sulfates and sulfuric acid, used mainly in the
United States, include turbidimetric and methyl thymol blue methods. The former usually
involves collection of samples on sulfate-free glass fiber filters by high-volume PM samplers.
Sulfate is extracted and precipitated with barium chloride, and turbidity of the suspension is
determined spectrophotometrically or nephelometrically. However, the method does not differ-
entiate between sulfates and sulfuric acid, and secondary formation of such products from SCL
in air drawn through the filter can affect estimation of atmospheric sulfate levels. Similar
collection procedures and limitations apply for the methylthymol blue method. Thus, these two
methods are not specific measures of suspended sulfates, and their results can only serve as
rough indicators of atmospheric levels of SO -related PM.
14.2.2 Particulate Matter Measurements
To be of maximum value, epidemiclogical studies on PM effects must utilize aerometric
methods that provide meaningful data, not only regarding the mass or amount of atmospheric PM
but also quantitative information on the size and chemical composition of particles present.
In actual practice, most epidemiological studies on PM effects have relied on air quality data
from air monitoring instruments of questionable sampling accuracy and not specifically de-
signed for health-related research. The resulting data have thus typically only provided
limited information regarding mass, size or chemical properties of the PM sampled.
Three measurement approaches or variations were mainly used to obtain PM data cited in
epidemiological studies reviewed below: (1) the British Smoke light reflectance method or
variations used in the United Kingdom and elsewhere in Europe; (2) the American Society for
Testing and Materials (ASTM) filter-soiling light transmittance method used in the United
States; and (3) the high-volume sampling method widely employed in the United States.
The British Smoke (BS) method and its variations in routine use typically employed stand-
ard monitoring equipment with a D™ cutpoint of = 4.5 pm (McFarland et al., 1982). Thus,
whether or not larger atmospheric coarse-mode particles were present during the sampling
period, predominantly small particles were collected. The D™ of the instrument may, however,
shift at higher windspeeds. The BS method neither directly measures the mass nor determines
chemical composition of collected particles. Rather, it primarily measures reflectance of
light from a stain formed by particles collected on filter paper somewhat inefficient for
collecting very fine particles (Liu et al., 1978). Reflectance of light from the stain
depends both on density of the stain or amount of PM collected in a standard period of time
and optical properties of the collected materials. Smoke particles composed of elemental
carbon of the type found in incomplete fossil fuel combustion products typically make the
greatest contribution to stain darkness, especially in urban areas. Thus, the amount of
elemental carbon, but not organic carbon, present in the stain is most highly correlated with
BS reflectance readings. Other non-black, non-carbon particles also have optical properties
such that they can affect the reflectance readings (Pedace and Sansone, 1972).
Since highly variable relative proportions of atmospheric carbon and non-carbon PM can
exist from site to site or from one time to another at the same site, then the same absolute
-------
BS reflectance reading can be associated with markedly different amounts (or mass) of parti-
cles or, even, carbon collected. Site-specific calibrations of reflectance readings against
actual mass measurements from collocated gravimetric monitoring devices are therefore neces-
sary in order to obtain approximate estimates of atmospheric PM concentrations based on the BS
Q
method. A calibration curve relating mass or atmospheric. PM concentration (in pg/m ) "to BS
reflectance readings obtained at a given site may serve as a basis for crude estimates of PM
(mainly small particle) levels at that site over time, so long as the chemical composition and
relative proportions of elemental carbon and non-carbon PM do not markedly change.
For British National Survey and OECD work site-specific BS mass calibration curves were
determined in the 1960s for numerous urban areas in the United Kingdom and Europe, and such
curves were interrelated or normalized to define certain "standard" curves. Two standard
calibration curves were adopted: (1) a British standard smoke curve defining relationships
between PM mass and BS reflectance readings for London's atmosphere in 1963, used to yield BS
2
concentration estimates (in M9/m ) reported in many published British epidemiological studies;
and (2) an OECD international standard smoke curve, against which smoke reflectance measure-
ments made elsewhere in Europe were compared to yield smoke concentration estimates (in pg/m )
reported in European studies on PM effects. Of crucial importance in assessing such studies
is the fact that the actual PM mass or smoke concentration at a particular site may differ
markedly (e.g., by factors of two or more) from the corresponding mass or concentration (in
jjg/m ) associated with a given reflectance reading on either of the two standard curves; and
great care must be applied in interpreting what any reported BS value in ug/m means at all.
Further complicating interpretation of smoke data used in most epidemiological studies is the
lack of reporting of specific quality assurance information for cited aerometric measurements.
Such information has only been reported (Warren Spring Laboratory 1961, 1962, 1966, 1967,
1972, 1975; OECD 1964; Moulds, 1962; Ellison, 1968) in general terms for United Kingdom
National Survey data used in some British studies.
The ASTM or AISI light transmittance method is similar in approach to the British smoke
technique. The instrument has a D™ cutpoint of =5 jjm and utilizes an air flow intake appa-
ratus similar to that used for the BS method, depositing collected material on a filter paper
tape periodically advanced to allow accumulation of another stain over a standard time period.
Opacity of the stain is determined by transmittance of light through the deposited material
For this reason, smoke data reported in MS/1"3 based on either the British or OECD Standard
curve are generally most appropriately interpreted in terms of "nominal" pg/m3 smoke units and
cannot be accepted as accurate estimates of airborne PM mass unless corroborated by local
site-specific gravimetric calibrations. In other words, unless based on local site-specific
calibrations, smoke readings in |jg/m3 cannot yield quantitative estimates of atmospheric PM
concentrations. Otherwise, such readings only allow for rough qualitative (i.e. <; =; or >)
comparisons of amounts of PM present at a given time versus another time at the same site and
do not permit meaningful comparisons between PM levels at different geographic areas having
airborne PM of different chemical composition (especially in terms of relative proportions of
elemental carbon).

14-9
-------
and filter paper, with results expressed in terms of optical density or coefficient of haze
units (CoHs) per 1000 linear feet of air sampled rather than in terms of mass units. Thus,
CoHs readings roughly index the soiling capacity of PM in the air and, like BS readings, are
most strongly affected by fine-mode elemental carbon particles. Coefficient of haze readings,
however, are more markedly affected by non-carbon particles than BS measurements. The ASTM
method does not directly measure mass or determine chemical composition of the PM collected.
Attempts to ever relate CoHs to pg/m would require site-specific calibration of CoHs readings
against mass measurements determined by a collocated gravimetric device, but the accuracy of
such mass estimates would be subject to question. Only one attempt at calibration of CoHs
readings against corresponding particle mass levels has been reported (Ingram, 1969; Ingram
and Golden, 1973), for only one city (New York). The calibration results, however, are appli-
cable in only a limited fashion to New York City aerometric data.
The high-volume (hi-vol) sampler method more widely used in the United States to measure
total suspended particulate matter (TSP) collects particles on a glass-fiber filter by drawing
air through the filter at a flow rate of approximately 1.5 m /minute, thus sampling a higher
volume of air per unit of time than the above PM sampling methods. This permits collection of
sufficient PM in a 24-hour period to allow for direct weighing and chemical analysis of
sampled material. The high flow rate, the geometric shape of the shelter housing the sampler,
and other features of standard types of apparatus used result in sampling efficiency varying
with windspeed. The D™ cutpoint for the hi-vol sampler is typically around 25 to 50 jjm and
collection of larger particles tends to drop off rapidly above such outpoints (Wedding et.
al., 1977; McFarland and Ortiz, 1980). Thus, the hi-vol sampler, as typically employed,
collects both fine- and coarse-mode particles that may include windblown crustal material of
natural origin (especially in dry rural areas). Only rarely have cyclone samplers or other
variations of the hi-vol sampler with smaller size cutpoints been utilized in epidemiological
studies to limit collected particles to an inhalable range, but even then the cutpoints
achieved were not sharp or independent of windspeed. Numerous factors other than windspeed,
as discussed in Chapter 3, can affect PM measurements by hi-vol sampling techniques. However,
quality assurance information for TSP measurements reported in most American epidemiological
studies is largely lacking, except for information on U.S. Environmental Protection Agency
CHESS Program data (U.S. Congress, House of Representatives, 1976).
Among the consequences of the broader size range of particles sampled by the hi-vol
method versus the BS or ASTM methods are severe limitations on intercomparisons or conversions
of PM measurements by those methods to equivalent TSP units or vice versa. No consistent
relationship was found, for example, between BS and TSP measurements taken at various sites or
during various seasons at the same site (Commins and Waller, 1967; Lee et al. , 1972; Ball and
Hume, 1977; Holland et al., 1979). Some exceptions exist, e.g., during severe London air
pollution episodes, when low wind speed conditions resulted in settling out of larger coarse-
mode particles and marked increases in fine-mode particles to constitute most of the PM present.
14-10
-------
Then TSP and BS levels (in excess of = 500 (jg/tn ) tended to converge, as would be expected
when both methods are essentially sampling only fine-mode particles (Holland et al,, 1979).
Taking into account the foregoing information on S0? and PM measurement methods and fac-
tors affecting the quality of results obtained with routine field monitoring, aerometric data
cited in various epidemic!ogical studies must generally be viewed as providing at best only
very approximate estimates of atmospheric levels of sulfur dioxide, other sulfur compounds, or
other PM associated with reported health effects. Further, to the extent that the aerometric
data cited are derived from use of techniques with limited specificity for the substance(s)
purportedly measured or the relative contributions of sulfur oxides or PM to observed health
effects cannot be distinguished from each other or from the effects of other covarying pollut-
ants, then the aerometric data and associated health effects reported might be more appropri-
ately viewed as relatively nonspecific indicators of the effects of overall air pollutant mix-
tures containing sulfur oxides and PM.
14.3 ACUTE PARTICIPATE MATTER/SULFUR OXIDES EXPOSURE'EFFECTS
14.3.1 Mortality
14.3.1.1 Acute Episode Studies—Detailed study of human health effects associated with epi-
sodes of severe air pollution spans a period of less than 50 years. The earliest reliable
documentation of such episodes describes a 1930 incident in the Meuse Valley of Belgium.
Dense fog covered the valley from Liege to Buy (Firket, 1931, 1936) from December 1 to 5,
1930, accompanied by an anticyclonic high pressure area with low winds and large amounts of
PM. Approximately 6,000 residents in the valley became ill and 60 deaths associated with the
fog occurred on December 4-5. The people who died were sick for only a short time and the
on-set of acute illnesses abated rapidly when the fog dispersed. Although no other immediate
deaths occurred, several persons affected by the fog died much later from complications
associated with fog-induced injuries. The death rate in the area was 10.5 times normal.
A similar event later occurred in Donora, Pennsylvania (Shrenk et al., 1949). Donora was
blanketed by a dense fog during October 1948, which adversely affected 43 percent of the popu-
lation of approximately 10,000 people. Twenty persons, mostly adults with preexisting
cardiopulmonary diseases, died during or shortly after the fog due to cardiorespiratory
causes; and 10 percent of the population was classified as being severely affected. No pollu-
tion measurements were made during the incident, but during subsequent inversioa-periods in
the same area, presumably not as severe in pollutant elevations as the one in October 1948,
daily averages of SOp as high as 0.4 ppm (~1140 (jg/m ) were recorded. In a follow up study of
Donora, Ciocco and Thompson (1961) found increased mortality rates and morbidity effects
(e.g., heart disease and chronic bronchitis) among those residents who reported acute illness
during the 1948 episode in comparison to those reporting no acute illness. The Meuse Valley
and Donora incidents demonstrated that severe air pollution can cause death and serious
morbidity effects in exposed human populations and raised the possibility of PM and SO, being
among air pollutants contributing to the induction of such health effects.
14-11
-------
As shown in Table 14-1, a series of episodes was also documented in London between 1948
and 1962 (Ministry of Health, 1954; Scott, 1953; Logan, 1953; Wilkins, 1954a,b; Hewitt, 1956;
Gore and Shaddick, 1958; Burgess and Shaddick, 1959; Martin and Bradley, 1960; Clifton et al.,
1960; Lawther, 1963), Excess mortality reported during those episodes occurred mainly among
the elderly and chronically ill adults during periods of prolonged marked elevations in air
pollution lasting for several days. Various factors have been discussed which might help to
explain some of the excess mortality (Holland et al., 1979), including possible influences not
only of increased air pollution but also of high humidity (fog) and low temperatures. Regard-
less of the relative contributions of these different factors, there exists a clear consensus
that increases in mortality were associated with air pollution episodes when 24-hour concen-
trations of both SO, and BS exceeded 1000 ug/m3 in London (Rail, 1974; Higgins, 1974;
Goldsmith and Friberg, 1977; NRC/NAS, 1978a,b; Shy et al., 1978; Holland et al., 1979; WHO,
1979; Shy, 1979). The available data, however, do not allow for clear delineation of the
effects of specific pollutants acting alone or in combination.

TABLE 14-1. EXCESS DEATHS AND POLLUTANT CONCENTRATIONS DURING SEVERE
AIR POLLUTION EPISODES IN LONDON (1948 to 1962)
Date
Nov.
Dec.
Jan.
Dec.
Jan.
Dec.
1948
1952
1956
1957
1959
1962
Duration
(days)
6
4
4
4
6
5
Deviation from
X of total
excess deaths
750
4000
1000
750
250'
700
Maximum 24-hr pollutant
concentration, ug/m3
Smoke S02
(BS) (H202 titration)
2780h
4460°
2830
2417
1723
3144
2150
3830
1430
3335
1850
3834
Note that the numbers of excess deaths listed represent 15 to 350 percent
increases in normal London baseline death rates during the years listed.
Note.,that peak and 24-hr BS levels were likely much higher than 4460
ug/m due to rapid saturation of filter paper by collected PM.

Acute episodes of high air pollution have also occurred in the United States since the
1948 Donora episode, but no single event reached the magnitude of the London episodes. Some
published studies (Greenburg et al. 1962, 1967; McCarroll and Bradley, 1966; Glasser et al.
1967), for example, suggested that increases in mortality and morbidity may have occurred dur-
ing some New York City episodes in the 1950s and 1960s, when PM levels exceeded 5.0 to 8.0
3
CoHs and SO,, exceeded 1000 ug/m (0.40 ppm), as measured at a single monitoring station in
central Manhattan. The earliest episode (November, 1953) involved a prolonged (10 to 12-day)
anticyclonic temperature inversion, during which PM levels built up to peak 24-hour levels of
6.0 to 8.0 CoHs on the last 4 days and S09 levels (estimated by "total acidity") reached
3
hourly peaks of 0.47 to 0.80 ppm (1231 to 1800 ug/m ) on each of those days. Under these
conditions of simultaneous elevations of PM and S0?, average daily death rates were estimated

14-12
-------
to have increased by 18 (8 percent) over a control total mortality comparision level of
226/day distributed equally across all age groups. However, close inspection of the published
data does not convincingly reveal the reported total mortality increases; nor do the published
cause-specific mortality data suggest notable increases in specific death categories plausibly
affected by air pollution.
Later, in November-December, 1962, during an episode of intermittent daily peaks of PM
o
exceeding 5.0-6.0 CoHs and SOp peaking at 0.4-0.5 ppm (1050 to 1310 ug/m ) at the same central
New York City sampling station, no increased death rates were detected by Greenburg et al
(1963); nor were clinic visits at four New York City hospitals for cardiopulmonary conditions
increased. Only the daily number of upper respiratory complaints appeared to increase signif-
icantly (P < .01) among elderly residents in four old-age homes. In contrast, significant
increases in respiratory morbidity and mortality among older (45 to 65 and > 65 years) age
groups were reported (Greenburg, et al. 1967) to have occurred during January 29 to February
12, 1963, when PM and SO, intermittently peaked at daily levels in excess of 5.0 to 6.0 CoHs
3
and 0.4 to 0.8 ppm (1050 to 1800 ug/m ), respectively. During this period of high PM and SOp
pollution, coincident with the occurrence of the coldest New York City temperatures in decades
and an influenza epidemic, Greenburg et al. (1967) estimated that 200 to 400 excess deaths (4
to 20 percent increases) occurred in comparison to various control baseline daily mortality
values. However, increased death rates did not appear on or immediately after all of the days
of peak PM and S0? levels during the January-February, 1963 episode or on other scattered days
of comparably high or higher PM or SO^ peaks in the weeks immediately preceding or following
the episode period. The data reported, therefore, provide only a very weak basis upon which
to assert that excess mortality attributable to air pollution was superimposed upon New York
City death rates already elevated by cold weather and influenza in early 1963.
Somewhat more convincing are elevated mortality levels observed (especially in those 65
years or older) during the 1966 Thanksgiving Day weekend as reportedly occurred in the absence
of cold weather (temperatures = 34° min to 64°F max) or epidemic illnesses (Glasser et al.
1967). Daily mean PM levels remained in the range of 5.0 to 6.0 CoHs for 3 successive days
(November 23 to 25) during which time daily SO, levels averaged 0.4 to 0.5 ppm (1050 to 1310
3
ug/m ) and mean daily total mortality levels reportedly exceeded those (237/day) of comparison
control periods by 24 (10 percent). Other comparable increases in daily mortality appeared to
occur in association with upward excursions of PM to peak hourly levels > 5.0 CoHs during 2-3
successive days 2 weeks prior to the November 23 to 25 episode and on 2 days about 2 weeks
after the episode. (Note that daily mean and hourly peak SO, levels did not exceed 0.3 ppm
3
(790 ug/m ) during the high PM days in early November.) Also, cause-specific mortality for
certain cardiovascular diseases appeared to rise noticeably over control levels on or
immediately after these same days of elevated PM levels, as well as during the November 23 to
25 episode. These overall results tend to suggest that the elderly and individuals with
prexisting cardiovascular disease may have been adversely affected by closely occurring
14-13
-------
instances of prolonged periods of severe air pollution containing high concentrations of PM
and/or S02 in late 1966.
Independent evaluations of the same New York City data on mortality and air pollution
relationships initially led to a published report (HcCarroll and Bradley, 1966) confirming
apparent associations between increased mortality and acute episodes of high PM and SQ^
levels. However, later reexamination of the New York City data and the published analyses by
the Greenburg group and by McCarroll and Bradley (1966) led Cassell et al. (1968) to question
the validity of the earlier published conclusions, especially in view of difficulties in
separating air pollution episode effects on mortality from effects of competing factors such
as temperature and humidity extremes and epidemic illnesses, which appeared to exert much
larger effects on death rates than the air pollution episodes. Still further doubts regarding
the reported New York City air pollution episode-mortality associations are raised by
inconsistencies in the data, such as no evident mortality increases being associated with some
days of PM and/or S0? elevations as high or higher than those on other days reported to be
associated with excess mortality. Only in November-December, 1966, did there seem to exist
some possible hints of consistent associations between high pollutant elevations and excess
mortality over control comparison levels. Thus, the results of the New York City episode
studies, while not necessarily inconsistent with the hypothesis of increased mortality being
associated with episodic elevations of PM and SO,,, do not provide strong evidence in support
of the hypothesis either. At best, the studies might suggest possible increases in mortality
and respiratory morbidity (mainly among older members of the population and those with
preexisting cardiovascular or respiratory diseases) under unusual conditions of prolonged
multiday periods of severe air pollution containing simultaneously high concentrations of PM
and S02 likely in excess of 5.0 to 8.0 CoHs and 1000 ug/m , respectively (see comments in
Section 14,3.1.2 and Appendix 14A on limitations concerning quantitative statements that can
be made based on aerometric data from the single central Manhattan monitoring station used to
index PM and S0? concentrations for the entire New York City area).
When a marked increase in air pollution is associated with a sudden rise in the death
rate or illness rate that lasts for a few days and both return to normal shortly thereafter
(as documented in some of the above studies), a causal relationship is strongly suggested.
Sudden changes in weather, however, which may have caused the air pollution incidents, must
also be considered as another possible cause of the death rate increase. On the other hand,
the consistency of associations between S0« and particulate matter elevations in London and
increases in mortality make it extremely unlikely that weather changes alone provide an ade-
quate explanation for all such observations. This view is further reinforced by: (1) the
fact that at least some episodes were not accompanied by sharp falls in temperature; and (2)
other weather changes of similar magnitudes to those accompanying the above pollution episodes
are not usually associated with such dramatic increases in mortality in the absence of greatly
increased levels of SO?J particulate matter, or other pollutants. In summary, the above
London episode studies appear to provide clear evidence for substantial increases in excess

14-14
-------
mortality when the general population was-exposed over several successive days to air pollu-
0 Q
tion containing S02 concentrations > 1000 ug/m in the presence of PM levels over 1000 ug/m
(BS). Certain New York studies also tentatively suggest that small increases in excess mor-
tality may have occurred at simultaneous elevations of 1000 ug/m S02 and PM above 5.0 - 8.0
CoHs, but this is much less clearly established.
Comparision of the New York City episode data and those for the Meuse Valley, Donora, and
London episodes may reveal further important information. Perhaps most striking are the much
lower estimates of excess mortality reported for the New York episodes (at most 4 to 20%) com-
pared to the 15- to 350-percent death rate increases during the London episodes and even
larger mortality rate increases in Donora and the Meuse Valley. Numerous factors might be
cited to explain the striking differences, including likely variations in the specific chem-
ical composition of the mixes of pollutants present in the different areas and the much
greater peak levels of pollutants (including PM and/or SCL) that were probably present during
the non-New York episodes. Also of probable considerable significance are two other features
typifying the episodes in the Meuse Valley, Donora, and London: (1) the presence of extremely
dense fog together with accumulating air pollutants, possibly providing the basis for trans-
formation of pollutants to potentially more toxic forms (e.g. formation of sulfuric acid
aerosol or absorption of PM into water droplet particles) resulting in more deposition of
toxic substances in tracheobronchial regions of the respiratory tract (see Chapters 11 and
13); and (2) the generally much more prolonged, continuous exposures of the non-New York popu-
lations to marked elevations of the pollutants. Examination of the published New York City
episode reports reveals that during such episodes the contributing temperature inversion con-
ditions typically intensified during evening hours (accumulating air pollutants over night)
but dissipated during the morning hours, resulting in invariably much higher peaks in PM and
S0? in the morning than in the afternoons (when PM and S0? levels fell back to near normal
levels). This is in contrast to the continuously high pollutant levels that apparently per-
sisted for several (> 4) successive days during the Meuse Valley, Donora, and London episodes,
with the largest increases in mortality tending to occur on the later days of each episode.
Duration of exposure, even at the extremely high levels of pollutants present, and the pre-
sence of certain other interacting factors such as high humidity (fog), then, appear to be
important determinants affecting increases in mortality observed in acute episode incidents.
Such factors must, therefore, be taken into account as important limiting considerations in
attempting to generalize or extrapolate observed episodic air pollution-mortality dose-
response relationships from one geographic location (or time period) to another.
14.3.1.2 Mortality Associated with Non-Episodic Variations in Pollution—A number of reports
have investigated relationships between mortality and air pollution in England during periods
with no unusual air pollution episodes (Gore and Shaddick, 1958; Burgess, 1959; Clifton et al. ,
1960; Martin and Bradley, 1960; Scott, 1963; Lawther, 1963; Martin, 1964; Waller et al., 1974).
For most of these studies, 15-day moving averages were constructed and the effects of pollu-
tion were assessed in terms of daily deviations from these baselines. Lawther (1963) reported

14-15
-------
that increases in daily deaths during the winter of 1958-59 appeared to be associated with
3 3 '
24-hr concentrations of BS >750 ug/m and SO- >715 ug/m (0.25 ppm) during a long (59-day)
period of thick fog. Further, Lawther (1963) reported that increases in daily deaths gener-
ally did not seem to be associated with pollutants at lower concentrations during 1958-59; nor
did they occur at similar pollutant levels during the prior winter having only 8 days of fog,
suggesting again the possible importance of high humidity conditions as a key factor interact-
ing with air pollutants to increase mortality. Similar studies in Sheffield (Clifton et al. ,
1960) did not yield confirmatory results; that is, while increases in deaths appeared to be
possibly associated with very high concentrations of pollutants, random variations in the
number of deaths were so large that firm conclusions could not be drawn.
Probably the most important British studies of mortality associated with nonepisodic ex-
posure to sulfur oxides and PM are those of Martin and Bradley (1960) and Martin (1964). The
first of these studies related daily mortality from all causes and from bronchitis and pneu-
monia to the level of SCL and smoke in London during the winter of 1958-1959, as measured by a
multistation air monitoring network operated throughout London. The authors found a consider-
able number of coincident peaks in pollution level and daily mortality. The correlation of
mortality from all causes with pollutants measured on the log scale was 0.613 for smoke (BS)
and 0.520 for SCL. Martin and Bradley (1960) reported that neither temperature nor humidity
was significantly correlated with London mortality studied during the winter of 1958-59, not-
ing a very low correlation coefficient (r = -0.030) for temperature and deaths for the entire
1958-59 winter and the occurrence of several peaks in mortality during November and December,
1958, when temperatures were substantially over 38°F. They further noted that a range of 30-
38°F is characteristic of most winter fogs and temperatures consistently below 30°F (when
temperature effects on mortality can be expected) are the exception. Though the authors
emphasized the relationship between change in pollution level and number of deaths and lack of
meteorological effects, an influenza epidemic during part of the study may have influenced
some of the results.
Martin and Bradley (1960), however, reported number of deaths, smoke levels, and SOp
levels from November 1, 1958 to February 28, 1959 allowing independent analysis of the data.
One such independent analysis was performed by Ware et al. (1981) but excludes the month of
February, in which the epidemic of Type A influenza also significantly influenced daily mor-
tality. For the remaining 92 days, the deviations of daily mortality from the 15-day moving
average (truncated at each end of the series) were computed and appeared to show a consistent
and significant trend of increasing mortality with increasing BS and S0? levels. The results
of the Ware et al. (1981) analysis are graphically depicted in Figure 14-1. Mean deviations
3
from the 15-day moving average changed from negative to positive between 500-600 (jg/m BS and
2
300-400 pg/m SQ~, which are concentrations at which apparent marked increases in mortality
occurred in comparison to lesser increments seen at lower pollutant levels. However, the 500-
3 3
600 |jg/m BS and 300-400 jjg/m SO- levels cannot be accepted as clearcut thresholds for
14-16
-------
+40
+30
+20
+10
z
o
a

<
LU
-10
-20
-30
-40
US)
O (27)
MS)
iMIOJ
9(9)
0(22)
'(11)
M7)
• (19)
0<29}«M18)
MS)
• BS
O so2

NUMBERS IN PARENTHESES INDICATE NUMBER OF DAYS DURING
THE WINTER WITH CONCENTRATIONS IN THE RANGE SUHHOUN
DINQ THE POINT
1
I
200 400 800 800 1000

BS, SO, lM8/m3). 24 hr, AVERAGE
1200
1400
Figure 14-1. Martin and Bradley (1960) data as summarized by Ware et
a!. (1981) showing average deviations of daily mortality from 15-day
moving average by concentration of smoke (BS) and $07 (London,
November 1, 1958 to January 1, 1959). As daily pollutant levels
decrease, the positive mortality deviations also decline. At still lower
pollution levels, dally mortality deviations fall increasingly below the
15-day moving mean, suggesting the possibility of a continuum of
exposure-response over the range examined. Note that the highest
value depicted for mortality associated with J5O2 levels perhaps
should more appropriately be shifted to some point to the right as
Indicated above by-*-?.
14-17
-------
mortality, given that the dose-response curves plotted in Figure 14-1 are highly suggestive of
a possible monotonic relationship across the entire range of exposure values shown. In fact,
it is unclear at what level significant excess mortality first occurred, although graphical
presentation (Figure 14-1) of the Martin and Bradley (1960) data as tabularly aggregated by
Ware et al. (1981) does indicate that notable increases clearly occurred at BS and S09 levels
3
somewhere in the range of 500-1000 ug/m and small increases in mortality may have occurred at
levels below 500 ug/m of either pollutant.
Although the graphical plot of the Martin and Bradley (1960) data in Figure 14-1 suggests
a monotonic relationship across the entire range of exposure values, adequate evaluation of
this possibility-would require more thorough statistical analyses, including consideration of
possible contributions of temperature and humidity, autoregression tendencies, and other
sources of error to the observed mortality patterns. Thus, whereas temperature and humidity
were reported by Martin and Bradley (1960) not to be significantly correlated with daily mor-
tality, both pollution levels and daily mortality increased throughout the period of study,
such that the possibility of other extraneous seasonal variables contributing to the mortality
increases cannot be ruled out. Also the above analyses of dose-response relationships by
Martin and Bradley (1960) and Ware et al. (1981) were both conducted assuming no error in the
PM and SOp concentrations used to estimate population exposure levels. The air pollutant esti-
mates used by them represent average BS and S02 levels as monitored at seven air sampling
stations located at widely dispersed sites in various areas of London. Several types of
possible error or variation in BS and S02 data could be associated with the 7-site average
concentration estimates reported in ug/m by Martin and Bradley (1960). Of most crucial
importance are: (1) the accuracy and precision of site-specific calibrations by which esti-
3
mates of PM mass (in ug/m ) were derived from BS reflectance readings; (2) errors associated
with field applications of BS and SOp measurement methods at the monitoring sites; and (3) the
variation or range of specific PM or S0? values obtained at the different sites from which
daily mean concentration exposure estimates for all of London were derived.
Little information exists by which to judge the error associated with the first factor
listed above, other than to note that site-specific calibrations of PM mass (ug/m ) against BS
reflectance readings carried out in 1956 at a central London site, as described by Waller
2
(1964), appear to confirm reasonably well the BS mass (in ug/m ) to reflectance calibration
(D.S.I.R.) curve employed in estimating mass from reflectance readings at the above seven
London sites in 1958-59 and for several more years until 1963. By 1963, however, the mass to
reflectance relationship defined by calibrations at another nearby London location (Waller,
1964) appeared to have shifted somewhat from the 1956 curve (see Figure 14C-2; Appendix 14C).
This possibly reflected changes in chemical composition of the ambient airborne PM as the
proportion of elemental carbon present declined due to reductions in emissions of sooty
incomplete coal combustion products as "Clean Air Zones" were established following enactment
of the British 1956 Clean Air Act.
14-18
-------
In regard to the second factor listed above, little specific information is presently
accessible by which to ascertain with confidence the accuracy or precision of PM or SO-
measurements taken at each of the seven monitoring sites more than 20 years ago. Only limited
quality assurance information on sources of error related to measurements of the same type
taken as part of the British National Air Pollution Survey exists by which to speculate on
possible measurement errors associated with the London aerometric data reported by Martin and
Bradley (1960). Examples of such quality assurance information are summarized in Appendix
14C; but Warren Spring Laboratory has commented that such measurement errors generally tend to
cancel each other out, so that the precision of the PM and S02 measurements are generally
assumed to be within about 6 percent (based on several intercomparison tests; see British
Standards Institution, 1964) in the absence of evidence to the contrary suggesting systematic
measurement error or bias in specific sampling data.
In regard to assessing the third source of possible error in the London exposure
estimates, the original BS and SO, data for the seven monitoring stations used to derive daily
London-wide average BS and S02 levels reported by Martin and Bradley (1960) were obtained by
EPA. Calculation of pair-wise correlation coefficients revealed very high correlations (all >
0.50; most > 0.75) between: (1) daily BS values reported at any two individual sites, even
the two most widely separated geographically; (2) daily S02 levels at any two specific sites;
and (3) BS and S0? readings at the same site or between any two given sites. The absolute
amount of variation of any individual site BS or S0~ readings from the daily 7-station mean
increased as a function of increasing concentrations but, in general, the 7-station average
seems to be an excellent index of city-wide increases or decreases in airborne PM or S02
levels. The daily London mortality data points reported by Martin and Bradley (1960) for the
1958-59 winter (excluding the influence of the epidemic period data in January-February, 1959)
were then plotted in relationship to the seven-station averages of PM and SOp as recalculated
from the original seven-station data. These relationships are depicted in Appendix D (Figure
140-3 for BS values and Figure 14D-6 for SO, values), along with non-linear fitted curves and
two sets of 95 percent confidence intervals, one set assuming error only in the mortality
estimates and a wider set of upper and lower 95 percent confidence bounds reflecting
additional error or variation in the PM and S0? exposure estimates. Information on the
derivation of the fitted curves and confidence intervals is also presented in Appendix 14 D.
Examination of results depicted in Figures 14D-3 and 14D-6 reveals very small predicted incre-
ments in total mortality over estimated 1958-59 London winter background levels of about 280
3 3
deaths/day (at < 150 pg/m ) as a function of increasing BS or SO, levels up to 500 yg/m .
However, increasingly more marked increases in daily total mortality over background levels
o
become apparent as BS concentrations rise from 500 to 1000 ug/m ar>d clearly large incremental
increases are apparent at successively higher BS levels over 1000 ug/m . Given the relatively
wide confidence bounds associated with these particular data, however, much caution should
be employed in interpreting reported dose-response relationships as depicted in Appendix
14D or as derived by other analyses of these data; and such analyses should not be taken as

14-19
-------
demonstrating precise quantitative relationships for uncaveated use for risk assessment pur-
poses.3
An analysis similar to that used by Martin and Bradley (1960) was carried out by Martin
(1964) for the winter of 1959-60. This winter had fewer incidents of high pollution. The
significant positive correlation between mortality and pollution was, however, still present
although the coefficients were somewhat lower than in the previous year. The Martin (1964)
results were based on analyses combining high pollution days from 1958-59 and 1959-60, after
excluding days on which pollution had fallen from a previously higher level. The mean
deviation was positive for every BS level above 500-600 ug/m and S00 level above 400-499
3
ug/m , but again no clear threshold for significant increases in mortality could be clearly
2
delineated. However, the most marked increases occurred for BS levels over 1200 ug/m and S09
3
levels exceeding 900 yg/m . Considerable covariation in levels of the two pollutants,
however, preclude attribution of the apparent mortality effects to either one alone. Bronchi-
tis mortality was also significantly, though less strongly, correlated with pollution level,
but pneumonia mortality was not correlated with pollution.
The above discussions of the Martin and Bradley (1960), Martin (1964), and Ware et al.
(1981) analyses, it should be noted, concern a relatively small number of observations on air
pollution-mortality relationships obtained during just two London winters (1958-59; 1959-60).
thus, while the results of these analyses are highly suggestive of certain relationships
between London daily mortality rates and BS or S0? levels, analyses of more London mortality
data from additional years would be valuable in order to assess the strength and represent-
ativeness of associations detected for the 1958-59 and 1959-60 London winters.
Such analyses of London nonepisodic air pollution-mortality relationships over a much
longer time-period have recently been reported by Mazumdar et al. (1981) in the published pro-
ceedings of an American Air Pollution Control Association conference. Mazumdar et al. (1981)
employed two different types of statistical approaches in analyzing excess mortality over
15-day moving averages for London winters from 1958-59 to 1971-72. An analysis was initially
used whereby variations in nonepisodic excess mortality were analyzed in relation to pollutant
levels categorized according to quartiles (i.e. 4 quartiles of BS values versus 4 SCL
quartiles). This allowed for statistical comparison of mortality for cells corresponding to
highest BS quartile values versus lowest SOp quartile levels and so on, as well as other
combinations of BS and S0? quartile values including 4 quartiles of S0? levels "nested" within
each quartile of BS levels. Mazumdar et al. (1981) reported that, by these analyses, they
were able to separate out the relative contributions to mortality of BS and S0? (which closely
covaried in London during most years studied) and that significant relationships existed
alt should also be noted that the analyses presented in Appendix 14D were not adjusted for
temperature, humidity, temporal trends or possible autoregression factors. Another recent,
unpublished analysis of the data taking such factors into account is summarized in Appendix
14E for informational purposes and appears to show small incremental increases in mortality
being associated with BS levels in the range of 150-500 |jg/m3 during the 1958-59 Lond'on winter.
14-20
-------
between excess mortality and BS levels but not between mortality and S0? levels during the
1958-59 to 1971-72 London winters. Serious questions can be raised regarding specific details
concerning the quartile analyses used and the validity of reported conclusions regarding the
separation of BS from S0? effects. For example, the numbers of data points falling in the
highest BS-lowest SCL and lowest BS-highest SCL quartile cells are extremely small (being only
2 and 1 respectively), and not likely allowing for any reasonable statistical comparisons
against other quartile combinations. However, based on their conclusions derived from the
quartile analyses, Mazumdar et al. (1981) attempted to further define possible dose-response
relationships between excess mortality and BS concentrations by means of regression analyses.
Mazumdar et al. (1981) used both a linear and a non-linear (quadratic) model in carrying
out their regression analyses on the 1958-59 to 1971-72 London winter mortality data. Figure
14-2 depicts the dose-response relationships defined by each of the two models, plotted on the
scatter diagram for excess mortality data versus BS levels for all of the winters analyzed. A
clearer depiction of the hypothesized dose-response relationships obtained with the linear and
quadratic models is presented in Figure 14-3. Both analyses (linear and quadratic) indicate
that small increases in mortality were associated with London PM levels in the range of
2
150-500 M9/m BS and more marked mortality increases occurred as BS levels rose to 500-1000
o
ug/m or more. The findings of mortality being significantly associated with the lower range
of BS values (150-500 M9/m ) were further confirmed by analyses of mortality rates occurring
only on days when BS levels did not exceed 500 M9/m • Importantly, temporal factors as well
as temperature and humidity effects were taken into account as part of the regression analyses
employed. However, shifts in the specific calibration curves used to relate BS reflectance
readings to estimates of PM mass (in (jg/m ) were not taken into account in the analysis.
Prior to 1963, the D.S.I.R. curve alluded to earlier (page 14-18) was used in determining BS
mass estimates from reflectance readings in London; after 1963-64, the British National Air
Pollution Survey standard curve was used instead, representing calibrations between PM mass
((jg/m ) and BS reflectance readings in London in 1963. Because of differences between the
D.S.I.R. curve and the 1963 curve, as well as the likely decreasing applicability of the 1963
curve for each successive subsequent year, probably a more appropriate statistical approach
would be to analyze the London mortality-air pollution data on a year-by-year basis. This
would allow for better determination of the consistency of any significant dose-response
relationships for any given year(s) versus findings for other years.
Inspection of London winter mortality and pollution data tabulated by Mazumdar et al.
(1981) for separate years indicates both higher mortality rates as well as PM and S00 levels
; t-
for 1958-59 than for all subsequent winters other than 1962-63. It is not clear as to what
interacting factors may distinguish 1958-59 (and perhaps 1962-63) as winters having apparently
higher mortality rates associated with increased PM and/or SOp levels. Possibly, consistent
with hypotheses noted earlier, more frequent occurrences of high humidity (fog) days in con-
junction with elevated pollutant levels during those two winters may offer one plausible
14-21
-------
205
190
175
It
ra
01
E
o

*
QUADRATIC
MODEL
I
20 30 40 50
100
200 300
500
1000
2000
SMOKE (M9/m3)
Figure 14-2. Linear and quadratic dose-response curves plotted on the scatter-
gram of mortality and smoke for London winters 1958/59 to 1971/72.
Source: Mazumdaretal. (1981).
14-22
-------
60
S so
o
UJ
CO

til
oc
o

UJ
o
ec
UJ
B.
40
30
20
10
X QUADRATIC MODEL
600
1000 1600

SMOKE (jjg/m3)
2000
2600
Figure 14-3. Hypothetical dose-response curves derived from regressing
mortality on smoke in London, England during winters 1958/59 to 1971/72,
Results obtained with linear {—) and quadratic f- -J models are depicted for
comparison.
Source: Mazumdaratal. (1981).
14-23
-------
explanation; but more detailed analyses of the London data on a year-by-year basis would be
required in order to assess this possibility.3
It should be noted that, although the above analyses point to significant associations
between mortality and PM or SCL levels in London, the relative separate contributions 'of PM
(as BS) or SCL cannot be clearly determined based on the available data; nor can the
possibility be ruled out that the observed mortality associations are due to other, unmeasured
covarying specific air pollutants or the total pollutant mix for which BS or SCL may be serv-
ing as excellent indices or surrogates.
A number of investigators have also evaluated possible relationships in the United States
(especially in New York City) between mortality and daily variations in PM or SO, air pollu-
tion during non-episodic periods (Hodgson, 1970; Glasser and Greenburg, 1971; Schimmel and
Greenburg, 1972; Lebowitz, 1973a,b; Buechley eta!., 1973; Buechley, 1975; Schimmel and
Murawski, 1975, 1976; and Schimmel 1978). However, most of these studies mainly provide, at
best, only limited qualitative evidence bearing on possible mortality relationships to
nonepisodic PM and S0? air pollution (see Appendix 14A) and have, collectively, yielded
apparently confusing and conflicting results.
Glasser and Greenburg (1971), for example, reported results from regression and other
types of statistical analyses of nonepisodic PM and S02 air pollution relationships to New
York City mortality, based on the evaluation of both excess mortality deviations from
15-day moving averages during October-March of the years 1960-64 and deviations from "normal"
5-year mortality levels for the same months. Significant relationships between excess
mortality and both S0? and PM (CoHs) were reported (with the association being stronger for
SO,) based mainly on analysis of unadjusted excess mortality deviations from the 5-year normal
averages. However, inspection of the reported data broken down by individual years or by
other factors (e.g., day of week, month, etc.) reveal numerous internal inconsistencies in the
data, including a number of "reversals" in terms of markedly lower positive (or even negative)
mortality deviations at higher PM or S02 levels than positive deviations seen at lower pollut-
ant levels. Only in 1963 did there appear to be clear and consistent dose-response relation-
ships, but the pollutant levels during that year were badly confounded with the occurrence of
extremely cold temperatures and a major influenza epidemic. About the only other consistent
dose-response relationship readily apparent in the overall 5-year data was an apparent steady
increase in mortality levels as a function of jointly ascending PM and S02 values, suggesting
The results of recent, unpublished analyses (on a year-by-year basis) of associations between
daily London mortality rates and BS or S02 levels for each winter from 1958-59 to 1971-72 are
summarized in Appendix 14F. Those analyses, taking temperature, humidity and temporal trends
into account, indicate that significant (P<.05) positive associations between London mortality
rates and daily S02 levels occurred during 1958-59 and 1962-63 as apparently exceptional win-
ters with simultaneously high S02 and frequent fog days. However, significant associations
between mortality and BS levels were found for winters during additional years, including some
such as 1967-68 when 24-hour BS levels exceeded 250 ng/m3 only once during the entire winter.
14-24
-------
possible interactive effects of the two pollutants on mortality. However, the mortality devi-
ations by PM and S02 concentrations were not cross-tabulated for individual years; it was
thusly impossible to evaluate the internal consistency of such a possible interactive effect
on a yearly basis.
In another study of New York City mortality data in the 1960s, Hodgson (1970) employed
multiple regression analyses to examine the relationship between deaths and air pollution on a
monthly basis and reported significant correlations between excess deaths and both S0? and PM
(CoHs). In addition, Lebowitz (1973a,b), utilizing a stimulus-response model to study daily
air pollution exposures, meteorology, and mortality in New York (1962-1965), reported that
adverse temperature and humidity changes were found to be important but did not completely
account for all apparent air pollution-mortality associations, Schimmel and Greenburg (1972)
also reported significant associations between excess mortality and nonepisodic air pollution
(indexed by PM and S0? levels) in New York City during 1963-68, with PM measured as CoHs ap-
parently being more strongly implicated than S0?-
None of the above studies, regardless of their specific results, permitted significant
increases in mortality levels to be quantitatively related to specific PM or SO, levels in
New York City. This is due, in part, to the particular statistical analyses employed. It is
also due to the fact that a single central monitoring station was typically used as an index
of variations in PM and SO, levels for the entire New York City area, a questionnable practice,
especially in view of a report by Goldstein and Landowitz (1977a,b) indicating relatively low
(r < 0.40) pair-wise correlations between either S0? or CoHs readings from that station and
comparable readings from other New York City monitoring stations. The lack of such correla-
tion raises serious questions about the probable validity of even the qualitative associations
between city-wide mortality and nonepisodic PM or SO^ air pollution in New York based on the
above study results (the overall lack of correlation, in contrast, may be less important dur-
ing episodic events when very high air pollution (PM and S02) levels may have approached a
common "ceiling" throughout the city). Additional questions can be raised in view of the
large number of correlation coefficients often calculated in the different studies and the
probability of some apparently significant associations being found by chance alone.
Besides the above caveats, it should be noted that Buechley et al. (1973), Buechley
(1975), Schimmel and Murawski (1975, 1976), and Schimmel (1978) have published further
analyses of New York City mortality data, including data from additional years extending into
the 1970s, with interesting results in comparison to those of the above earlier studies. For
example, Buechley et al. (1973, 1977) evaluated the relationship of daily deaths in the New
York/New Jersey metropolitan area (1962 through 1972) to S0? measured at a single New York
City monitoring station and reported some statistically significant associations, suggesting a
relationship between residual mortality and S0? levels. However, Beuchley (1975) noted that
mortality rates did not decrease in association with a three-fold decline in S0? over the
10-year period studied and questioned whether S02 was perhaps only serving as a covarying
surrogate of other pollutants more directly linked to health effects. In an independent

14-25
-------
analysis, Schimmel and Murawski (1975, 1976), controlling for common seasonal trends in mor-
tality, temperature, and pollution in New York City found that the central Manhattan monitor-
ing station PM (CoHs) levels varied little over three time periods (1963-66; 1967-69; 1970-72)
analyzed, while average SCL concentrations declined markedly over that timespan. Based on
their analyses, the percentage of premature deaths due to air pollution in the respective
periods was estimated to be 2.78, 2.48, and 3.20, i.e., lower percentages attributed to air
pollution than those (=4.0 percent) suggested by the earlier analyses of Schimmel and Green-
burg, 1972. Further, the percentage attributed to S02 was 0.58, 1.22, and 0.62 for the three
different time periods (percentages not significantly different from zero), whereas the
remaining 80 percent or so appeared to be attributable to PM (CoHs). Also, because percent-
ages of excess deaths attributable to S02 varied very little over the three time periods as
3-fold decreases in SCL occurred, it was concluded that S0? was likely a surrogate for other
(possibly unmeasured) variables but not directly related to health effects (mortality) across
the range of ambient concentrations (annual means <0.2 ppm; daily peaks up to 0.4-0.6 ppm)
studied. Subsequently, time-series analyses by Schimmel (1978), eliminating seasonal and
other cyclical effects on New York City mortality, demonstrated that the regression of
mortality on SOy was not significant, negative correlations being obtained at times. Overall,
then, these later analyses by Schimmel and Murawski (1976) and Schimmel (1978) create serious
doubt regarding reported associations between mortality and nonepisodic S0? levels present in
New York City during the 1963 to 1972 period. The same analyses, however, were interpreted by
the authors as being indicative of weak but positive associations between nonepisodic
mortality and PM (annual means =2.0 to 2.5 CoHs; daily peaks up to 5.0 to 6.0 CoHs) levels in
New York City during the same time period. Unfortunately, the use of aerometric data from the
single central Manhattan monitoring station as an estimate of pollutant exposures for the
entire New York area precludes clear quantitative statements regarding possible effect or
no-effect levels based on these results.
14.3.1.3 Morbidity—Studies of morbidity effects associated with acute or short-term air
pollution exposures are much less common in the epidemiological literature than morbidity
studies of chronic or long-term air pollution exposures. This reflects dual complications of
the difficulty of having adequate estimates of individual's pollution exposures as well as the
statistical analytical problems of the health data being collected. The main focus here is on
studies providing information on quantitative relationships or associations between ambient
air concentrations of SCL or PM and acute exposure morbidity health effects.
Several British studies have been published on health effects associated with acute or
short-term exposures of adults to sulfur oxides and particulate matter which appear to provide
useful information on quantitative dose-effect relationships. Illness data were obtained in
many of the early severe pollution episodes discussed above; much of this morbidity informa-
tion did little more than support the mortality results reported in those studies in providing
evidence that increases in illness occurred along with increases in deaths.
14-26
-------
Certain other studies, however, provide valuable information useful in estimating quanti-
tative relationships between morbidity effects and acute exposures to elevated BS and SO,
levels in London during the 1950s and 1960s. For example, Waller and Lawther (1955) and
Lawther (1958) studied associations between daily variations in smoke and S0? pollution and
self-indicated health status in 29 British chronic bronchitis patients during the winters of
1954-55 and 1955-56. Patients maintained diaries on which their daily condition was indicated
in relation to their usual condition. The alternatives were "better," "same," "worse," and
"much worse." During the month of January 1955, an episode of relatively high pollution
resulted in a sharp increase in the .number of patients whose condition worsened as 24-hour
3 3
smoke (BS) increased from about 400 ug/m to 2000 ug/m and 24-hour S0? levels increased from
about 450 ug/m3 (0.15 ppm) to about 1300 M9/m3 (0.50 ppm). Waller and Lawther (1955) reported
that, on the worst day of the episode (January 19), when smoke (BS) concentrations in London
increased tenfold during the course of 2 hours, there was a deterioration in the clinical
condition of some patients with bronchitis. On this day, peak hourly smoke (BS) concentration
3
may have reached 6500 ug/m . Sulfur dioxide also increased to an hourly maximum of about 2860
Mg/m (1.0 ppm) but H?SO* apparently did not, on the basis of washings from impactor slides.
Most of the mass of PM was determined by microscopic studies to consist of particles less than
1 urn in diameter (Waller, 1963).
In the winter of 1955-56, the study was extended to include 180 patients in the London
area and "degree of illness" scores were quantified as follows: better = -1; same = 0; worse =
1; much worse = 2. Lawther (1958) reported that the exacerbation of preexisting illness
appeared to be related more closely to pollution than to temperature or humidity during the
1955 winter months, but the relationship disappeared when the levels of pollution decreased in
the spring. Actual numerical data are not tabulated in the Lawther (1958) report, but inspec-
tion of the published graphed results indicate that exacerbation of illness notably increased
when winter smoke (BS) levels exceeded about 300-350 ug/m and SO, levels about 500-600 ug/m .
However, in the spring of the year, pollution concentrations were no longer as clearly
associated with health status. That is, the association between pollution and illness
decreased when spring smoke (BS) levels fell to a fairly consistent 24-hour concentration of
less than 250 pg/m and the few higher peaks in 24-hour smoke (BS) seemed to have little
effect on illness status. Also, although S0« reached some intermittent peak concentrations as
high as those associated with increased illness during the winter, no marked exacerbation of
preexisting illness seemed to occur during the spring. Lawther (1958) noted that the hazard
to subjects' health may be better indicated by markedly higher short-term hourly peaks in
pollution rather than the 24-hour average levels noted above and that the results are not
necessarily indicative of causal relationships, but rather that the measurements of smoke (BS)
and S0? may only be indicators of some other causal agent.
A later report by Lawther et al. (1970) both reviewed the 1954-55 and 1955-56 results
and, also, provided results for further extension of these studies into the winters of
14-27
-------
1959-60, 1964-65, and 1967-68. The techniques used in the later studies were similar to the
earlier ones except that the patients now reported on health status in relation to the pre-
vious day rather than in relation to usual conditions. These studies confirmed results from
previous years in that the worsening of health status continued to be associated with in-
creases in air pollution. The responses of 1,071 patients were evaluated for the winter of
1959-60 in relation to variations in BS and S02 levels measured at the same seven London air
monitoring sites as those yielding aerometric data used by Martin and Bradley (1960) and
Martin (1964) to evaluate daily variations in London mortality. Similarly, responses from
1,037 patients were evaluated for 1964-65 in relation to pollution levels measured at the same
sites. Although there were fewer days of high pollution in 1964-65 than in 1959-60, and an
impression of a slightly reduced and less consistent response in 1964-65 compared to 1959-60,
clear positive associations still persisted between worsening of bronchitic patients' condi-
tions and the air pollution variables measured in 1964-65. Further, Lawther et al. (1970)
stated that, although exact relationships between the responses of patients and the concentra-
tions of smoke and S09 could not be determined, the minimum pollution leading to any signifi-
3 3
cant response was about 500 pg/m (0.17 ppm) SOp, together with about 250 pg/m smoke (BS).
Lawther et al. (1970) also speculated that these responses may reflect the effects of brief
exposures to maximum concentrations several times greater than the 24-hour average but pro-
vided no data analyses clearly substantiating this hypothesis. As in the earlier studies, the
exacerbation or worsening of health status appeared to relate more closely to pollution
indices during the first part of the winter, and in some instances there was little reponse to
higher concentrations of pollutants near the end of the winter. Although the concentrations
of smoke and S09 closely correlate, examination of the data again suggests that often higher
concentrations of S09 near the end of the winter, occurring with generally lower concentra-
tions of smoke, produced less response in the study subjects than did the same concentrations
of S0? earlier in the winter, when smoke was higher. However, there was some evidence of a
loss of interest by participants over time, which may also partially explain this pattern.
Follow-up studies compared responses for a selected group of patients (apparently among
the most sensitive to BS and S02 pollution) for 1964-65 and 1967-68 winters, as shown in Table
14-2. Examination of the data reported for the winters of 1964-65 and 1967-68 for the group
of selected patients shows statistically significant associations of morbidity effects with
pollution for both winters. This includes the winter of 1967-68 when, as Lawther et al.
(1970) noted, there were hardly any periods of high pollution. In fact, the 24-hour BS
levels, except on one occasion, never exceeded 250 (jg/m and S09 levels were consistently
3
below 500 pg/m . Lawther et al. (1970) thought it likely that these selected patients were
especially sensitive to pollution since no significant correlations were expected considering
the low pollution levels to which this group was exposed.
These studies among chronic bronchitis patients in London continued into the 1970s as the
frequency of periods of high pollution declined. Lawther et al. (1973, 1974a,b,c) reported acut
14-28
-------
TABLE 14-2. SUMMARY OF RESULTS, SELECTED PATIENTS, 1964-65
AND 1967-68

Mean Score
Smoke (pg/m

SO, (|jg/m )
H2S04 (ng/m )
Temp (°C)

Corr. coeff . ,
mean score and

Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.

Smoke
SO,
H2S04
Temp.
1964-65
1.98
0.10
129
95
264
149
7.3
4.8
6.4
4.1

0.39*
0.30*
0.51*
0.24*
1967-68
1.96
0.11
68
48
204
100
6.3
4.0
6.3
4.3

0.31*
0.28*
0.26*
0.17*

These results are for the whole winter period. October to
March.

^Significant at .05 level.

Source: Lawther et al. 1970.
14-29
-------
decrements in ventilatory function in four healthy adults and 2 bronchitics in London to be
associated with variations in air pollutant levels measured at their place of work or treat-
ment. After multiple regression analysis to remove time trend effects, S0? concentrations
explained the largest proportion of residual variance in peak flowrates, with clearest
associations shown after walking exercise in heavy pollution.
Martin (1964) examined applications for hospital admissions in London for the winter of
1958-59 and found for men ages 45 to 79 (after adjustment for day of the week and correction
for 15-day moving average) significant correlations for both cardiovascular and respiratory
conditions with smoke (r = 0.46) and S0? (r = 0.40). Analogous significant correlations were
found for the same male age group for such conditions in relation to both smoke (r =0.41) and
SO, (r = 0.43) for the winter of 1959-60. The average deviations associated with increasing
SCL and smoke levels during both winters are summarized in Tables 14-3 and 14-4. As seen in
those tables, whereas no clear threshold for the onset of mean positive deviations across the
exposure ranges can be distinguished, very marked increments in the positive mean deviations
can be discerned starting at 800-899 Mg/m for S02 and 1100-1510 ug/m3 for PM (BS). Presenta-
tion by the authors of their results separately in relation to BS and S0? (and the present
summarization in Tables 14-3 and 14-4) is not meant to imply that the relative individual
contributions of BS and S0? alone to the observed effects can be ascribed to the concentra-
tions listed, in view of considerable covariation in SO- and BS levels during the two winters
studied.
Essentially no other British or European epidemic!ogical studies on morbidity effects
associated with acute exposures to PM or S0? appear to yield pertinent data useful in attempt-
ing to quantify such relationships for present criteria development purposes. Few even pro-
vide good qualitative evidence characterizing acute PM and SO., exposure-morbidity relation-
ships. See Appendix 14A for brief annotated comments regarding such additional studies
evaluated by the present authors.
In addition to the quantitative British studies reviewed above, several American studies
appear to provide limited but useful qualitative evidence for the association of particular
types of morbidity effects with acute (24 hour) exposures to SOp and PM. For example,
Greenburg et al. (1962, 1963) reported that during the 1953 New York City air pollution
episode, statistically significant increases in emergency clinic visits for upper respiratory
infections and cardiac illnesses, respectively, occurred at 3 of 4 and 2 of 4 New York City
hospitals studied, but no significant increases occurred in asthma clinic visits. During a
1962 New York episode, however, no significant increases in hospital clinic visits were
detected (Greenburg et al., 1963) for respiratory infections or cardiac illnesses, although
significant increases in physician visits to treat respiratory complaints among elderly
residents at 4 old-age homes were observed. Greenburg et al., (1967) further reported
significant increases in emergency clinic visits for bronchitis and asthma at 3 of 7 New York
City hospitals on the third day of the Thanksgiving Day weekend episode in New York in 1966.
14-30
-------
TABLE 14-3. AVERAGE DEVIATION OF RESPIRATORY AND CARDIAC MORBIDITY
FROM 15-DAY MOVING AVERAGE, BY SMOKE LEVEL (BS) (LONDON, 1958-60)

Smoke level
((jg/m3, BS)
500-599
600-699
700-799
800-1099
1100-1510
Source: Martin (1964).
Number
of days
9
6
9
8
7

TABLE 14-4. AVERAGE DEVIATION OF RESPIRATORY
FROM 15-DAY MOVING AVERAGE, BY S02 LEVEL
Mean
deviation
3.2
-0.7
2.4
4.9
12.9

AND CARDIAC MORBIDITY
(LONDON, 1958-60)

S02 level
(|jg/m3)
400-499
500-599
600-799
800-899
900-1280
Number
of days
9
6
9
6
5
Mean
deviation
2.2
5.1
6.9
12.8
12.8
Source: Martin (1964).
14-31
-------
However, these reported increases in numbers of daily emergency visits may have in part been
due to the clinics having been closed during the preceding holiday weekend. Overall, these
results suggest that increased cardiac and upper respiratory tract morbidity may have signi-
ficantly increased, especially among the elderly, during episodic marked elevations of PM
(5.0-8.0 CoHs) or S02 (0.4-0.5 ppm; 1050-1310 ug/m3). However, the one isolated instance
(November 1966) of increased asthma visits being associated even with the episodic elevations
can be reasonably questioned in light of other observations by Greenburg et al. (1964) that
demonstrate regular seasonal increases in asthma attacks in New York in the fall months (even
in the absence of notable peaks in air pollution) and link increases in such attacks to rapid
shifts in climatic conditions from warm to cold temperatures.
In other studies focusing on possible morbidity effects associated with nonepisodic ex-
posures to air pollution, McCarroll (HcCarroll and Bradley, 1966; McCarroll et al., 1966) and
colleagues (Mountain et al., 1968; Thompson et al. 1970; Cassell et al., 1969, 1972; Lebowitz
et al. , 1972) studied frequency of cough and the "common cold" in New York City residents in
relation to S0? and PM measurements obtained from monitoring stations within 2000 feet of
their apartment building residences. These studies demonstrated significant multiple correla-
tions between acute respiratory symptoms and air pollution, controlling for season, weather,
3
and social class, when seasonal SO- means were in the range of 0.10-0.24 ppm (~28Q-70Q pg/m )
and seasonal smoke shade means were in the range of 1.56-3.15 CoHs. Initially, multiple re-
gression analyses showed conflicting findings in that the pollutants were occasionally absent
from or negative in their regressions. This led to a separation of the combined meteoro-
logical and air polluant conditions into categories of: (1) stormy weather (low temperatures,
occasional precipitation, high wind speed) when the pollutants were low; (2) stagnation
periods (low wind speed, moderate temperatures) when SO- and TSP were high; (3) periods of
change in pollutant levels during the fall through spring periods; and (4) high photoxidant
conditions in the summer. This analysis yielded significant correlations between the pollut-
ants and acute symptoms for 1800 individuals studied weekly in New York (1962-65) during stag-
nation periods and significant correlations of the same acute respiratory symptoms (pre-
dominately common colds) with meteorological conditions during stormy periods. They also
found a lag of one to three days in symptoms and corresponding increases in school absenteeism.
Some individuals (mostly under the age of 10) were found to have reacted consistently and
frequently to increases in the pollutants, and their respiratory symptoms were of greater
duration and severity than "nonsensitive" individuals (Lebowitz et al., 1972). Those who were
"sensitive" during the first part of the period under study were found to be "sensitive" later
on in the study. The attack rates per person year were about double for the "sensitive", and
occurred predominantly in the winter period. However, while suggesting that respiratory
symptom morbidity effects may be qualitatively associated (especially for some "sensitive"
children) with nonepisodic increases in air pollution these studies do not allow specific
levels of PH or S0~ or their individual contributions to be related to the observed effects.
14-32
-------
Further studies have attempted to relate increased morbidity effects to acute (24 hour)
increases in air pollution containing PM and SCL occurring in American cities during the late
1960s or 1970s. Only a few, however, were conducted methodologically in such a manner so as
to have employed community PM or SO, aerometric data that can be viewed as being somewhat rea-
sonably representative of ambient outdoor exposures of th,e particular populations studied.
Some of the more important findings from those studies having potentially important bearing on
present criteria development efforts are evaluated next (see Appendix 14D for comments on
other, less useful studies for present criteria development purposes).
Some studies have examined relationships between acute elevations in PM or S0~ air pollu-
tion in certain American cities and the occurrence of asthma attacks, yielding at best equiv-
ocal results which, taken together, tend to call into question whether any such relationships
existed at ambient air PM or SO, levels present in the cities studied. For example, Cofien
et al. (1972) reported a weak association to exist between air pollution and the frequency of
asthma attacks in a preliminary pilot study of 43 self-reported asthmatics living near a
co'al-fueled power plant in West Virginia. Temperature was reported to be most strongly
s
correlated with the frequency of attacks based on limited data analyzed for 20 subjects having
at least one attack during the study (note that several subjects also dropped out of the study
or returned health survey data too infrequently during the study to be included in data
analysis). Based on multiple regression analysis of the remaining data, both SOp and TSP were
reported to be significantly (P<.01) correlated with attack rate after adjusting for
temperature; also, attack rates were reported to be significantly higher on high air pollution
3 3
days (TSP > 150 ug/m ; S0? >_ 200 ug/m ) defined by aerometric data monitored within 1.5 mile
of the subjects' residences. However, these reported findings cannot be accepted as validly
indicating that asthmatic attacks were associated with the ambient PM or SO, concentrations
present at the time. This is due to: (1) the reported large subject dropout rate and dele-
tion of data due to too-infrequent subject reports of health status; (2) the likely resulting
bias introduced into the data ultimately analyzed, along with any biasing of results due to
deletion of data for subjects reporting no attacks; (3) the lack of information on changes in
panel membership over time; and (4) insufficient information about the specific data analyses
employed to assure that proper adjustments for seasonal or other time-related factors were
carried out.
In addition to the above caveats regarding the Cohen study, it should be noted that Gold-
stein and Dulberg (1981) reported finding no significant relationships between asthma "events"
(i.e., increased hospital emergency room visits for asthma attacks) and acute 24 hr changes in
PM or SO, air pollution levels in New York City during 1969-71 (when average daily PM levels
1
=0.9-1.5 CoHs and mean daily S02 =0.03-0.07 ppm or 80-180 ug/m ) as measured by a 40-station
sampling network. Unfortunately, the authors did not adjust for temperature change (as a key
variable affecting asthma attack occurrence), thus making it difficult to fully accept their
reported "no-effect" findings until more proper analyses including temperature adjustments are
14-33
-------
carried out. However, in another study that did adjust for temperature and other important
potentially confounding factors, Samet et al., (1981) found no increases in hospital visits
for all respiratory diseases (including asthma) in Stuebenville, Ohio, when 24-hr means for
3 3
TSP fell in the range of 14-696 ng/m and 24 hr SO- levels were 2-369 ng/m , as monitored at a
centrally located station proximal (within 300 meters) to the hospital studied. Many resi-
dences of potentially affected subjects, however, were located further away from the monitor-
ing site, making it unclear as to what extent the stated TSP and S02 values can be taken to
represent "no-observed effect" levels. Nevertheless, the weight of all of the above
considerations still points towards the conclusion that the studies reviewed here provide
essentially no evidence for asthma attacks being associated with acute exposures to low
ambient 24-hr PM or S0? levels encountered in several different American cities in the late
1960s or 1970s.
The results reported by Samet et al., (1981) concern some of the findings just beginning
to emerge from a long-term epidemiological study of morbidity-air pollution relationships in
several American cities. As part of this so-called "Six-Cities Study," Samet et al., (1981)
evaluated^additional indices of respiratory effects, besides those noted above for asthma at-
tacks, in heavily industrialized Steubenville, Ohio. Hospital emergency room visits for
various respiratory diseases during March, April, October, and November of 1974-1977 were
related to daily levels of TSP, S0?, N0?, CO, and 0,,. After adjusting for meteorological
variables, weekly, seasonal, and yearly cycles, possible day-of-week effects, and multi-
collinearity of the pollutants, deviations from average numbers of daily emergency visits for
all respiratory diseases did not vary significantly with pollution, although the largest
deviations occurred at the highest pollutant levels (see ranges stated above) for TSP and S0?.
For respiratory diseases, a linear regression model identified a significant effect of both
unlagged TSP and S02, but not N0?. However, the authors warn that this result should be
viewed with caution since, although the regression coefficients attained significance at
2
P<.05, the contribution of pollutant variables to the multiple R is only 0.01 - that is, only
1 percent of the variance of the respiratory disease index is explained by TSP or S09.
3 3
Furthermore, the days of highest TSP (>202 ng/m ) or S0? (XL21 ng/m ) pollution were not
significantly associated with increased numbers of emergency room visits for respiratory
treatments. Other (more ambiguous) preliminary results from the Six-City Study, as reported
by Dockery et al., (1981), are commented on in Appendix 14A.
In summary, the above studies on acute exposure effects tend to suggest that the elderly,
those with chronic cardiorespiratory diseases, and children .may constitute populations at risk
for manifesting morbidity effects in response to acute exposure to elevated atmospheric levels
of sulfur dioxide and particulate matter. Qualitatively, increases in the occurrence of car-
diac and upper respiratory tract disease symptoms, including exacerberation of preexisting
chronic bronchitis (but not asthma attacks) appear to be among the morbidity effects most
clearly associated with exposures to the ambient levels of PM and S02 evaluated in the studies
14-34
-------
assessed above and are most clearly seen at markedly elevated levels of the two pollutants.
For example, increased applications by adults aged 45-79 for admissions to London hospitals
for cardiac and respiratory morbidity most clearly occurred, based on the Martin (1964) study,
when 24-hour BS and S09 levels approached or exceeded 900-1000 ug/m ; but Martin's data also
3
suggest that such effects may have occurred at somewhat lower levels down to 500 MS/111 f°r
both SOp and BS. Using a probably more sensitive morbidity indicator, Lawther's studies in
London further appear to demonstrate that worsening of health status among bronchitic patients
3
is associated with acute 24-hour exposures to BS of 250-500 ug/m in the presence of SO,
3
levels in the range of 500-600 ug/m . In contrast, no effects on most bronchitics appeared to
be detectable at 24-hour BS levels below 250 ug/m in the presence of 24-hour SO, levels below
3
500 ug/m ; limited evidence suggests some effects on a selected group of highly sensitive
bronchitics at somewhat lower, but not precisely-defined BS and SO, levels. Similarly,
American studies by Greenburg's group appear to most clearly demonstrate increased cardiac and
upper respiratory morbidity, especially among the elderly, during air pollution episodes in
New York City when extremely high levels of PM (5.0-8.0 CoHs) and SOp (>1000 ug/m3) were
present. On the other hand, much less clearly demonstrated were morbidity effects related to
nonepisodic elevations in air pollution containing PM and SQ?. The findings of McCarroll's
group (especially as reported by Lebowitz et a!,, 1972) in New York City, for example, suggest
at most an increase in upper respiratory tract symptoms (e.g., coughs and colds) in certain
"sensitive" children at lower nonepisodic levels of PM or SCL. Insufficient epidemiological
information from the studies of McCarroll and coworkers exists, however, by which to determine
specific quantitative acute exposure levels at which the health of such "sensitive" children
might be adversely affected.
14.4 CHRONIC PM/SO, EXPOSURE EFFECTS
14.4.1 Mortality
Numerous studies have been performed to compare general or cause-specific mortality in
areas of lowest-to-highest pollution concentrations. However, virtually all of these studies:
(1) used aerometric data of questionable accuracy or representativeness of study population
exposures; and (2) did not adequately account for the potential effects on mortality rates of
such confounding factors as cigarette smoking, occupation, social status, or mobility differ-
ences between areas (see Appendix 14A). These methodological problems preclude accurate
characterization of any quantitative relationships between mortality and air pollution para-
meters. Therefore, essentially no epidemiological studies are presently well-accepted as
providing valid quantitative data relating respiratory disease or other types of mortality to
chronic (annual average) exposures to sulfur oxides or particulate matter. On the other hand,
the findings of certain published studies of chronic air pollution effects on mortality may
warrant assessment here in regard to their potential for establishing qualitative links
between mortality and chronic exposures to PM or sulfur oxides. Two types of general
approaches have been employed in such studies: (1) aggregation of mortality and other
information, e.g. smoking, or socioeconomic status data, in relation to specific individuals

14-35
-------
within the study population(s); and (2) aggregation of analogous data for entire populations
across large geographic areas, e.g. cities, counties, or standard metropolitan statistical
areas (SMSAs). Examples of results obtained with each type of approach are evaluated below.
Among the best known and most often cited examples of the first approach listed above are
the Winkelstein et al. (1967), Winkelstein and Kantor (1967), and Winkelstein and Gay (1971)
studies of total and cause-specific mortality in Buffalo and Erie County, New York, during
Ii59 to 1961. A network of 21 sampling stations provided data on TSP (hi-vol sampler) and
oxides of sulfur (non-specific sulfation methods) for the period July 1961 to June 1963; and
these aerometric data were used to categorize geographic areas as "low" to "high" air pollu-
tion areas. Chronic respiratory disease mortality for white males 50 to 69 years old was
reported to be about three times higher in the high-pollution areas than in the low-pollution
areas, across all economic groups (Winkelstein et al. 1967). Additional positive associations
in relation to TSP concentrations were reported for both stomach cancer (Winkelstein and
Kantor, 1967) and deaths from cirrhosis of the liver (Winkelstein and Gay, 1971).
However, numerous criticisms of these studies can be noted which raise serious doubts
regarding the validity of the reported findings, including the following most salient methodo-
logical problems: (1) the use of 1961-1963 TSP and sulfur oxides measurement data as a basis
for retrospectively classifying geographic areas according to presumed past air pollution
gradients against which to compare mortality among the elderly during 1959-1961; (2)
inadequate controls for possible age differences between study groups that may have covaried
with the air pollution gradient used; (3) lack of information on lifetime (including occupa-
tional) exposures to PM or SO,,; (4) failure to correct for smoking habits; and (!5) the
implausibility of some of the reported findings, e.g., air pollution increasing mortality due
to liver cirrhosis. In a later presentation, Winkelstein (1972) commented on several of these
points and attempted to correct for some of them, such as by looking at smoking patterns among
certain populations living in some of the same study areas included in the earlier analyses.
Still, the 1972 analyses do not adequately counter the different major concerns regarding the
Winkelstein Buffalo mortality findings. For example, the later finding in the follow-up
investigation (Winkelstein, 1972) of no significant differences in smoking patterns among the
different study areas for females does not adequately control for possible smoking effects in
different specific population cohorts evaluated in the earlier study analyses reported by
Winkelstein and coworkers. These particular studies, therefore, are of questionable validity
even in regard to providing credible qualitative evidence for links between PM air pollution
and mortality in Buffalo during 1959-1961.
Turning to examples of the second type of approach listed above as being used for evalu-
ating chronic air pollution effects on mortality, Lave and Seskin (1970) employed regression
analyses to evaluate relationships between mortality rates and indices of air pollution in
Britain such as PM levels (measured by deposit gauges and BS readings) and SO, levels, all as
reported by'Stocks (1959, 1960) and Ashley (1967). Significant positive associations were
14-36
-------
reported to exist between PM levels (as indicated by deposit gauge and BS measurements) taking
into account socioeconomic status (SES). This Lave and Seskin (1970) study has been criti-
cized (Holland et al., 1979) regarding, for example, difficulties in justifying inclusion of
SES and air pollution levels in the regression analyses as if they were completely independent
variables and the failure to make some direct allowance for smoking habits in the analyses.
However, perhaps even more basic difficulties with tne*analyses derive from: (1) the use of
imprecise qualitative estimates of PM pollution (e.g., deposit gauge data and BS aerometric
data expressed in terms of mass concentration estimates not appropriately obtained by means of
site-specific calibrations of reflectance readings against local gravimetric measurement
data); and (2) ambiguities regarding locations of sampling devices in relation to study popu-
lations, therefore raising questions about the representativeness of the aerometric data in
estimating PM exposure levels.
Other studies discussed by Lave and Seskin (1970) and in three further publications (Lave
and Seskin 1972, 1977; Chappie and Lave, 1981) extended the original U.K. analysis approach
(Lave and Seskin, 1970) to standard metropolitan statistical areas (SMSAs) in the United
States. Based on such analyses, analogous positive associations between mortality and air
pollution variables were reported for the United States. Many criticisms similar to those
indicated above for the earlier Lave and Seskin (1970) publication apply here. Of crucial
importance are basic difficulties associated with all of their analyses in terms of: (1) use
of aerometric data without regard to quality assurance considerations, most notably including
use of sulfate measurements known to be of questionable accuracy due to artifact formation
during air sampling (see Chapter 3); and (2) questions regarding the representativeness of the
air pollution data used in the analyses as estimates of actual exposures of individuals
included in their study populations. For example, in some instances data from a single
monitoring station were apparently used to estimate pollutant exposures for study populations
from surrounding large metropolitan areas. Clearly, then, no useful information on quantita-
tive relationships between specific concentrations of PM or sulfur oxides and mortality can be
derived from these published analyses. However, the question remains as to whether any
clearly consistent qualitative conclusions regarding PM or SO air pollution-mortality rela-
X.
tionships can be drawn based on these and analogous studies, as discussed next.
Lave and Seskin (1972, 1977) reported first on the results of linear regression analyses
of relationships between: (1) mortality (mainly total mortality rates; TMR) in more than 100
United States SMSAs; (2) various measures of air pollution, such as sulfur dioxide (SO,,);
suspended sulfates; total suspended participate matter; and others (CO, NOp, ozone, hydro-
carbons, etc.); and (3) variables serving as proxies for certain demographic factors such as
age, sex, race, population density, and socioeconomic status known to be important deter-
minants of total and cause-specific mortality rates. Using such indices, aggregated by
geographic areas, Lave and Seskin (1977) reported results of regression analyses which
they interpreted as demonstrating consistent significant positive associations between air
14-37
-------
pollution in general or particular pollutants (e.g., sulfates or TSP) and TMR recorded in 1960
for 117 U.S. SMSAs. They went on to state that the implications of their findings meant, for
example, that for every 10 percent change (increase or decrease) in sulfate or TSP air pol-
lution levels, there would be a resulting change in TMR of 0.50 percent (for sulfates) and
0.44 percent (for TSP); in other words, decreases of 4.60 and 3.99 in deaths per 100,000 could
be expected for every 10 percent reduction in ambient air sulfates or TSP concentrations,
respectively.
Lave and Seskin (1977) reported further results obtained with similar regression analyses
of mortality, air pollution and demographic data for additional years (1961-69) for most of
the same SMSAs.. This included presentation of comparisons among results obtained with: (1)
analyses of data for all SMSAs or SMSA subsets for different years, e.g. results for 1960
versus 1969, (2) analysis of data for different individual SMSAs for the same year(s), e.g.,
data for Chicago versus St. Louis, Philadelphia, and certain other SMSAs for 1962-69; and (3)
use of various models beside the basic linear regression model used for earlier analyses.
Chappie and Lave (1981) have since reported additional analyses for previously unavailable
1974 data for 104 SMSAs. This latter paper compared (1) results obtained with the linear
model for 1974 data versus earlier (1960, 1969) results; (2) results obtained for 1974 data
with various models in addition to the basic Lave and Seskin linear model; and (3) results
obtained for the 1974 data as a consequence of adding or deleting various potential explan-
atory variables or proxy variables from the models. Table 14-5 summarizes comparison of some
salient findings from the 1960, 1969, and 1974 analyses that led to key conclusions stated by
Chappie and Lave (1981). Those conclusions include: (1) that a strong, consistent and statis-
tically significant association between sulfates and mortality persists across all years
studied (and that the association is little changed by adding variables for smoking and alco-
hol consumption, by using city or county instead of SMSA data or by adding a medical care
variable and going to a simultaneous equation framework); (2) that the association between
particulate matter and mortality, while previously consistently positive and significant in
1960 and 1969, was no longer significant or consistently positive in 1974; and (3) that these
results support and strengthen the earlier conclusions of Lave and Seskin (1977) that "strin-
gent abatement of sulfur oxides and particulates would produce social benefits (based on
health effects alone) that greatly exceed social costs."
Lave and Seskin (1977) stated the above conclusions and implications despite: (1) their
own expressed misgivings concerning the quality, accuracy and representativeness of the aero-
metric data used as population exposure estimates and other data used as inputs to serve as
proxies for key parameters or components of their regression model(s); (2) ambiguities and
concerns, again recognized by the authors themselves, regarding the appropriateness of the
main linear (OLS) model they used to represent the underlying relationships they were attempt-
ing to model; and (3) numerous internal inconsistencies in their reported results which do not
support or which contradict their main conclusions. A few examples illustrating the third
type of problem will suffice here.

14-38
-------
TABLE 14-5. SUMMARY OF KEY RESULTS REGARDING MORTALITY-AIR
POLLUTION RELATIONSHIPS IN UNITED STATES CITIES BASED ON
LAVE AND SESKIN MODEL ANALYSES FOR 1960, 1969, AND 1974 DATA
Dependent variables
1960 TMRC
1969 TMRL
1974 TMR1"
Air Pollution Variables:
Min S
Mean S
Max S
Sum S El as =
Min P
Mean P
Max P
Sum P El as =
Socioeconomic Variables:
% >65
% Nonwhite
% Poor
Density
Log Popn
Other Terms:
R2
N
Constant
4.733
1.67
1.726
0.53
0.279
0.25
0.50
0.199
0.32
0.303
0.71
-0.018
-0.19
0.44
68.802
16.63
3.960
3.82
0.384
0.26
8.285
1.54
-27.566
-1.38
0.831
117
343.230
2.87
-0.384
-0.07
6.329
1.81
-0.527
-0.67
0.59
0.434
0.69
0.055
0.13
0.130
1.83
0.56
64.030
17.11
2.037
2.24
5.113
2.13
0.013
2.51
-42.774
-2.22
0.817
112
387.011
3.36
0.294
0.04
16.915
3.09
-1.809
-2.09
1.32
2.366
1.32
-1.386
-1.51
0.294
1.80
0.06
64.265
17.59
2.000
1.96
5.148
2.15
9.687
1.92
-44.594
-2.80
0.861
104
313.342
3.19

Regression (1960 TMR) is from Lave and Seskin (1977; p. 31 - Regression 3.1-1).
Degression (1969 TMR) is from Lave and Seskin (1977; p. 121 - Regression 7.1-3).

Degression (1974 TMR) is from Chappie and Lave (1981; Table 1 - Regression 1-1).

Abbreviations: TMR=Total Mortality Rates; Min S, Mean S, Max S = average minimum
sulfate levels, etc.; Min P, Mean P, Max P = average minimum TSP levels, and so on.
Numbers beneath the regression coefficients for dependent variables are t-statistics.
The sums of elasticities represent the estimated percent change in the dependent
variable resulting from a ten percent change in the three air pollution variables.
14-39
-------
For instance, in discussing the 1960 results shown in Table 14-5, Lave and Seskin (1977)
noted not only that "the results are encouraging in that more than 83 percent of the total
mortality rate across the 117 SMSAs was accounted for by the eleven dependent variables", but
also that "the estimated coefficients for the air pollution variables were disappointing in
that none had a statistically significant coefficient, and one coefficient was negative; how-
ever the six variables made a statistically significant contribution as a group." The latter
statement is based on an F-test comparison of the regression results obtained without the six
pollutant variables present versus with them included in the regression analysis, which
rejected strongly the hypothesis that the apparent effects of the air quality variables could
be attributed to random sampling variability (P < 1/350). Tests of the other two regressions
in Table 14-5 rejected that hypothesis even more decisively. Thus the air quality measure-
ments appeared to be related to the inter-SMSA differentials in mortality rates, but in view
of the non-significance of the t-statistics, the causal source of the significance remained
obscured.
In order to clarify the relationship, Lave and Seskin (1970) experimented with various
alternative specifications but with indecisive results. As part of such experimentation, they
dropped "superfluous variables in order to derive an equation with predicted signs, plausible
magnitudes, and statistical significance" and retained only those air pollution variables
"whose coefficients were positive and exceeded their standard errors, with the further con-
straint that at least one sulfate and one particulate measurement were retained." They con-
cluded that: "In reestimating the relationship, we often found that the retained air pollu-
tion variables were now significant"; and "sometimes the retained air pollution variable con-
tributed little to the significance of the regression. Such variables were eliminated,
subject to the restriction that at least one air pollution variable be retained in the final
equation. This technique was used throughout our analyses of other mortality rates."
Several examples can be cited to illustrate problems associated with directly extra-
polating data generated by this selective technique to the general population when drawing
conclusions about the contribution of air pollution to mortality. For example, Lave and
Seskin (1977) reported that analyses of SO^-mortality relationships in Chicago during 1962-63
showed no significant association between daily mortality and S0?, using a regression analysis
(including SCU, climatic variables, and day-of-week variables) which explained only 27 percent
of the variance, implying that important factors affecting mortality were missing from the
regression. Regressing daily deaths lagged against high levels of S0? occurring 1, 2, 4 or 5
days prior to the observed deaths yielded some positive associations, but only that for the
5-day lag was significant. Carrying out similar analyses for 1963-64 data produced three of
six significant positive associations for S0? and daily mortality. However, when similar
analyses were carried out for 50,,-daily mortality relationships in Denver and St. Louis, no
significant associations were found; nor did any of several other air pollution variables
, CO, hydrocarbons, etc.) make a significant contribution to analogous analyses for these
« other cities (except for Philadelphia; nitric oxide). In fact, the sums of elasticities

14-40
-------
of the air pollution variables were said to be "generally small and often negative," Lave
and Seskin (1977) offered several possible explanations for the differences in results
obtained between Chicago and other cities, noting, for example, that Chicago's mean SO- levels
were almost 4 times as high as those in Denver and 10 times higher than those in St. Louis and
that at levels of air pollution substantially below those in Chicago acute effects may not be
important. Nevertheless, despite these clear indications against extrapolation to other
cities of the Chicago results for SCL, Lave and Ses,kin, (1977) cite the results for Chicago as
if they are generally applicable to all U.S. cities across all levels of air pollution concen-
trations.
Similarly, Lave and Seskin (1977) appear to make overly broad generalizations concerning:
(1) results obtained with their earlier 1960 cross-sectional regression (summarized in Table
14-5) and interpreted as indicating a 50 percent reduction in air pollution (as measured by
sulfates and TSP) was associated with a 4.7 percent decrease in unadjusted TMR; (2) results
from their cross-sectional analysis of a 1969 subset of SMSA data interpreted as indicating a
50 percent reduction in S0? to result in a 2.7 percent decrease in TMR; and (3) results of
cross-sectional time-series analysis of a larger, multiple SMSA data set leadiog to esti-
mations of a 50 percent reduction in SO, being associated with a 1.2 percent reduction in TMR.
In making the above reported generalizations, Lave and Seskin (1977) do not mention inconsis-
tencies of the above types for the 1962-63 and 1963-64 data analysis results or other numerous
inconsistencies and contradictory outcomes present among results from the 1960, 1969, and
cross-sectional time-series analyses. Nor did they highlight the fact that their analysis of
residuals, as reported by Lave and Seskin (1977), indicated that their regression equation(s)
overpredicted unadjusted and age-sex-race adjusted TMR for many of their study SMSAs geographi-
cally widely dispersed around the United States. See Table 14-6, which is reproduced from
Lave and Seskin (1977) and shows many examples of overprediction of the contribution of air
pollution to mortality as indicated by negative residuals for various SMSAs listed.
Numerous other inconsistencies and problems with the Lave and Seskin (1970, 1972, 1977)
analyses have been noted by others (Crocker et al. 1979; Lipfert, 1980; Gerking and Schultze,
1981; Ware et al. 1981); and the Chappie and Lave (1981) analyses attempt to correct for some
of the more serious criticisms advanced, with some success. However, some of the major im-
provements made in carrying out certain new analyses reported in the Chappie and Lave (1981)
paper have not been employed in reanalysis of the 1960-1969 data earlier reported by Lave and
Seskin (1977) and one is left, at this time, with a confusing array of often internally incon-
sistent and conflicting results derived from the series of analyses reported by Lave and
Seskin (1970, 1972, 1977) and Chappie and Lave (1981).
Further difficulties in discerning consistent patterns of association between mortality
and air pollution variables are encountered when the results of Lave and coworkers are com-
pared with those obtained by others using analogous "macro-epidemiological" approaches. For
example, compare the findings reported by Lave and coworkers of (1) consistently significant
14-41
-------
TABLE 14-5. SUMMARY OF LAVE AND SESKIN (1977)
ANALYSIS OF RESIDUALS FROM REGRESSION ANALYSES FOR 1950 AND 1969 U.S. SMSA DATA

Ten lariat SMSAi
Nc. York N Y
Qlicato, 111
L<*9 Anteles Calif
Vtailadcfphta Pa
Motion Mau
Detroit Mich
San Franaieo, Calif

Saint Louli Mo
OevtUnd. Ohio
Southwestern SMSA]
Albuquerque, N Mex
Denver, Colo
Lai Vegas. Ne«
Lot Afitelei Cauf

Salt Lake Cn> Utah
San Dtegc Calif
San Joe. Calif
Ten largest 1960 reaiduals
rmi*
IfiO
-7 06
35 J3
-41 11
11 2<
61 13
—26 77
Ji 12
39 J2
0 25
-29 37

-3 30
12 95
92 12
—41 18
-62 9}
-77 14
—31 20
-Tl 74

(unadjusted total mortality rates)
Tampa Fla -246 00
Witae»-Barre. Pa
Scramon Pa
Austin. Tex
Savannah Ga
New Orteani La
Canton, Ohio
Orlando, Fla
Terre Hauie Ind
Siotix Falli, S Dak
Ten larieit 1960 renduaU (ate-w
race-adjuited total mortallt}
Scranton Pa
Wlliev-Bane. Pa
Aom FalU 4 Dak
Auittn Tex
New Orleans La
Tampa. Fla
Canton, Ohio
Brockton Mas*
Savannah, Gi
Fall River. Mau
Tan largest 1969 residuals
22J 74
212 09
—14] 40
!26 41
125 22
-122 6'
-119 10
113 26
-108 00

ratea)
212 09
225 74
-108 00
-143 40
125 22
-24t 09
-122 61
68 n
126 41
73 »4

(unadjusted total mortal*!} rates)
Buluth, Minn 63 3<
Tampa Fla
Honolulu Hawitt
Witiej-Barre ft
Fareo, N Dak
Montiorncrj ALj
San Bernardino CaW
Miami Fla
Toledo Ohm
Alban>, K Y
Ten large*! 1969 residual!
-26 50
54 76
-1 16
14 29
15 19
9 31
42 74
mi.,..
-27 50
-3 25

-S2 51
-12 fait
ifte
-ai 29
43 43
-11 t!
3 Wi
12 10
9 14
12 30
M.
-22 12
-17 46

-49 99
-76 14
41 00
-11 63
-30 73
-75 52
—6! 65
-•9 74

-19 13
122 90
«—
—
)4 40
13 94
51 39
«—
—^
—

122 90
— *
13 94
-«9 13
31 39
H_
X 40
29 77

16! 62
-19 13
-U5 48
-122 90
-152 18
65 6J
-106 0!
-17 56
97 ««
76 00

161 &2
—133 It
-135 48
-112 90
-106 01
97 J9
97 2'
95 97
—94 0}
-19 74
t*eir The second eoluirm thowg nauduiin from the 196© unadjunwJ total mwtalMj r»te eQtation (f*ff*saion
7 1-1) b*fed on 117 SMSAi the third column §ho*i roucluili from the co*rrt{>ondm| 1969 rcpJ*c*uon (re
•nouuoci 7 1*3} b«ie
A neeiitw: residual indicates that ihe K$ft**iQei etjtaman overestimated th* monaliij rate for the pamcular
SMSA
From: Lave and Seskln (1977), page 133.
14-42
-------
positive associations between suspended sulfates and mortality in 1960, 1969, and 1974 and (2)
significant positive associations for TSP in 1960 and 1969, but not 1974 against the following
results reported by others:

(1) Results reported by Mendelsohn and Orcutt (1979), based on regression analyses of
associations between 1970 mortality rates (for 404 country groups throughout the
United States) and air pollution exposures retrospectively estimated on the basis of
1970 and 1974 annual average pollutant data from air monitoring sites in the same or
nearby counties, which suggested fairly consistent (though variable) associations
between mortality for some age groups (increasingly more positive with age) and
sulfate levels but much less consistent and sometimes negative associations with TSP
or other pollutant levels.

(2) Results obtained by Thibodeau et al. (1980), interpreted as not disagreeing with the
conclusions of Lave and Seskin concerning mortality being associated with chronic
exposure to PM, S0? or sulfates, but also accompanied by indications that the re-
gression coefficients for air pollution variables are quite unstable and should be
used with great care when interpreting their meaning.

(3) Results reported by Lipfert (1980), derived from an analysis taking into account a
smoking index based on state tax receipts, which Lipfert interpreted as showing sul-
fates to be least harmful of seven air pollutants (including S0? and TSP), although
no adjustments for urban-rural differences in study population residences were used.

(4) Analyses of 1970 United States mortality data by Crocker et al. (1979), taking into
account retrospectively estimated nutritional variables and a smoking index, but
indicating no significant relationships between air pollution and total mortality.

(5) Results of Gerking and Schultz (1981), using the same data base, that indicated a
significant positive relationship between TSP and total mortality when using an OLS
model similar to that of Lave and Seskin (1977) but finding negative, though not
significant, air pollution coefficients after adding smoking, nutrition, exposure-
to-cold, and medical-care variables to a two-equation model.

Various criticisms of each of the above studies could be presented, as cited in one or more of
the other respective studies or Chappie and Lave (1981), but to little avail at this time in
trying to ascertain which findings may be "more valid" than others. Thus, although many of
the studies qualitatively suggest some positive associations between mortality and chronic
exposure to certain air pollutants (e.g., sulfates in the case of several studies) in the
United States, many issues remain unresolved concerning any such associations. These include
questions regarding the strength and stability of the reported air pollution-mortality
relationships across geographic areas, time periods, ranges of pollutant exposure levels, and
variations in other variables or combinations of variables included in the underlying
analyses. Furthermore, to the extent that the analyses generally do not adqeuately rule out
the possibility of reported associations being due to potentially important unaccounted-for
variables that may covary with the air pollution gradients, then it is not possible to discern
whether the reported associations are either merely fortuituous on the one hand or, perhaps,
causal on the other hand.
14-43
-------
Other epidemiologies] studies have more specifically attempted to relate lung cancer mor-
tality to chronic exposures to sulfur oxides, PM undifferentiated by chemical composition, or
specific PM chemical species. However, little or no clear epidemiological evidence has been
advanced to date to substantiate hypothesized links between S0? or other sulfur oxides and
cancer; nor does there presently exist credible epidemiological evidence linking increased
cancer rates to elevations in PM as a class, i.e., undifferentiated as to chemical content.
See reviews by Higgins (1976) and Doll (1978) evaluating associations between cancer mortality
and environmental exposures to various air pollutants. Other epidemiological studies (e.g.,
of occupationally-exposed workers) also provide some evidence of increased cancer risk associ-
ated with exposure to certain types of PM, e.g., certain organic compounds or metals, often
found in the fine- and coarse-mode particulate fractions of many urban aerosols (see Appendix
14B). However, no well-accepted basis currently exists for quantitatively defining any
consistent relationships concerning relative contributions or levels of such PM components to
possible carcinogenic effects of PM pollution as a whole.
14.4.2 Morbidity
14.4.2.1 Respiratory Effects in Adults--Impairment of pulmonary function is likely to be one
of the effects of exposures to air pollution, since the pulmonary system includes tissues that
receive the initial impact when toxic materials are inhaled. Acute and chronic changes in
function may be significant biological responses to air pollution exposure. A number of
studies have been conducted in an effort to relate pulmonary function changes to the presence
of air pollutants in European, Japanese, and American communities. However, few provide more
than qualitative evidence relating pulmonary function changes to elevations in S0? or PM (see
Appendix 14A).
A series of studies, reported on from the early 1960s to the mid-1970s were conducted by
Ferris, Anderson, and others (Ferris and Andersen, 1962; Kenline, 1962; Andersen et a!., 1964;
Ferris et al., 1967, 1971, 1976). The initial study involved comparison of three areas within
a pulp-mill town (Berlin, New Hampshire). Kenline (1962) reported average 24-hour SO^ levels
during a limited sampling period in the summer (August-September 1960) to be only 16 ppb and
3
average 24-hour TSP levels for the two-month period to be 183 (jg/m . In the original preva-
lence study (Ferris and Anderson, 1962; Anderson et al., 1964), no association was found
between questionnaire-determined symptoms and lung function tests assessed in the winter and
spring of 1961 in the three areas with differing pollution levels after standardizing for
cigarette smoking. The authors discuss why residence is a limited indicator for exposure
(Anderson et al. , 1964). The study was later extended to compare Berlin, New Hampshire, with
the cleaner city of Chilliwack, British Columbia in Canada (Anderson and Ferris, 1965).
Sulfation rates (lead candle method) and dustfall rates were higher in Berlin than in
Chilliwack. The prevalence of chronic respiratory disease was greater in Berlin, but the
authors concluded that this difference was due to the interaction between age and smoking
14-44
-------
habits within the respective populations. Higher levels of respiratory function in some
cigarette-smoking groups in the cleaner area were observed, but this difference could be due
to socioeconomic and ethnic differences as well as air pollution (Higgins, 1974).
The Berlin, New Hampshire, population was followed up in 1967 and again in 1973 (Ferris
et al. , 1971, 1976). 'During the period between 1961 and 1967, all measured indicators of air
3 3
pollution fell, e.g. TSP from approximately 180 ug/m in 1961 to 131 ug/m in 1967. In the
1973 follow-up, sulfation rates nearly doubled from the 1967 level (0.469 to 0.901 mg S0~/100
9 ~\ '
cm /day) while TSP values fell from 131 to 80 ug/m . According to WHO (1979) only limited
data on SOp was available (the mean of a series of 8-hour samples for selected weeks). During
the 1961 to 1967 period, standardized respiratory symptom rates decreased and there was an in-
dication that lung function also improved. Between the period 1967 to 1973, age-sex standar-
dized respiratory symptom rates and age-sex-height standardized pulmonary function levels were
unchanged. Although some of the testing was done during the spring versus the summer in the
different comparison years, Ferris and coworkers attempted to rule out likely seasonal effects
by retesting some subjects in both seasons during one year and found no significant differ-
ences in test results. Given that the same set of investigators, using the same standardized
procedures, conducted the symptom surveys and pulmonary function tests over the entire course
of these studies, it is unlikely that the observed health endpoint improvements in the Berlin
study population were due to variations in measurement procedures, but rather appear to be
associated with decreases in TSP levels from 180 to 131 ug/m . The relatively small changes
observed and limited aerometric data available, however, argue for caution in placing much
weight on these findings as quantitative indicators of "effect" or "no-effect" levels for
health changes in adults associated with chronic exposures to PM measured as TSP.
One other American study was found to provide potentially useful qualitative or quantita-
tive information regarding association of morbidity effects in adults with ambient exposures
to SOp or particulate matter. A cross-sectional study was conducted by Bouhuys (1978) in two
Connecticut communities in which differences in respiratory and pulmonary function were
examined in 3056 subjects (adults and children). Hosein (1977a) reported on aerometric data
used in the study, which were obtained at three sites in Ansonia (urban) and four sites in
Lebanon (rural) near the residences of study subjects. The TSP levels during the period of
the study in Lebanon and Ansonia were 39.5 and 63.1 |jg/m and S09 levels were 10.9 and 13.5
3
ug/m , respectively. Site-to-site variations on the same day were frequently significant in
Ansonia and also occurred in Lebanon. During the years 1966-72, annual average TSP levels in
Ansonia ranged from 88 to 152 ug/m . No historical data for S0? or TSP in Lebanon were
provided. Size fractionation (Hosein, 1977b) of a limited number of TSP samples in Ansonia
showed that 81 percent of the TSP sample was 9.4 pm or less in diameter. Binder et al. (1976)
obtained for 20 subjects in Ansonia one 24-hour measure of personal air pollution exposure
for: particles (< 7 urn diameter), SOp, and NOp. Subjects with smokers in the home were
14-45
-------
exposed to significantly higher levels than those without such exposure. Personal exposure
and outdoor exposures were also significantly different. The mean personal respiratory level
was 114 (jg/m as compared to the outdoor TSP level of 58.4 (jg/m .
An extended version of the MRC Questionnaire was administered via a computer data-
acquisition terminal (Mitchell, 1976) between October 1972 and January 1973 in Lebanon and
from mid-April through July 1973 in Ansonia. For children (7-14 yrs) the response rate varied
from 91-96 percent for boys and girls. For adults (25-64 years) the response rate was 56
percent in Ansonia and 80 percent in Lebanon. After analysis of non-responder versus
responder differences, the responders were considered to be representative of the total
population, although some significant differences were noted between responders and
non-responders for some symptom reporting and current smoking in some age groups.
Bouhuys (1978) found no differences between Ansonia and Lebanon for chronic bronchitis
prevalence rates but did note that a history of bronchial asthma was highly significant for
male residents of Lebanon (the cleaner town) as compared to Ansonia (the higher pollution
area). No differences were observed between the communities for pulmonary function tests
adjusted for sex, age, height and smoking habits. However, three out of five symptoms (cough,
phlegm, and plus one dyspnea) prevalences were significantly higher for adult non-smokers in
Ansonia (P <.001). The inconsistencies apparent in terms of both positive and negative health
effect results obtained in this cross-sectional study make it difficult to interpret,
Although it appears to be a generally negative study, the significantly increased symptom
rates raise questions as to whether some impact on health might have occurred. A follow-up
longitudinal examination could have determined whether the effects persisted. Also, it is not
clear whether the reported effects relate to current or historical pollutant levels, and
occupational exposure was not examined. The seasonal difference in conduct of the two surveys
might have also had some effect on the acquired health data. A later study (Bouhuys et al.
1979) compared results obtained in the Connecticut communities with results obtained by
analogous testing procedures in Winnsboro, South Carolina, and found no signigicant dif-
ferences between results obtained in Ansonia versus those obtained in Winnsboro where air
pollution was markedly lower. However, because of differences in racial mixes between the two
communities and other demographic differences not properly controlled for, these results
cannot be accepted as adequately demonstrating no differences in health effects between the
towns.
14.4.2.2 Respiratory Effects inChildren—Numerous epidemiological studies have attempted to
evaluate possible relationships between chronic exposure to air pollution containing PM and
SO- and the occurence of health-related changes in children. However, very few studies
provide adequate evidence qualitatively establishing such relationships; and still fewer
provide dose-response (or exposure-response) data of a type or of sufficient quality by which
to estimate even approximate chronic PM or SQ^ exposure levels associated with changes in
health status in children.
14-46
-------
An apparent quantitative relationship between air pollution and lower respiratory tract
illness in children was demonstrated by Lunn et al. (1967). These investigators studied
respiratory illness in 5- and 6-year-old schoolchildren living in four areas of Sheffield,
England. Air pollution concentrations showed a gradient in 1964 across four study areas for
3 3
mean 24-hour smoke (BS) concentrations from 97 ug/m to 301 ug/m and the same gradient for
3 3
annual mean 24-hour S0« concentrations from 123 ug/m to 275 ug/m . The following year, the
annual concentrations of smoke were about 20 percent lower and S0« about 10 percent higher,
but the gradient was preserved for each pollutant. In high-pollution areas, individual
3
24-hour mean smoke concentrations exceeded 500 ug/m 30 to 45 times in 1964 and 0 to 15 times
in 1965 for the lowest and highest pollution areas, respectively. Sulfur dioxide exceeded 500
ug/m 11 to 32 times in 1964 and 0 to 23 times in 1965 for the lowest and highest pollution
areas, respectively. Information on respiratory symptoms and illness was obtained by
questionnaires completed by the parents, by physical examination, and by tests of pulmonary
function (FEVg 75 and FVC). Socioeconomic factors (SES) were considered in the analyses, but
home-heating systems were not. Although certain differences in SES between areas were noted,
the gradients between areas existed even when the groups were divided by social class, number
of children in house, and so on. Positive associations were found between air pollution
concentrations and both upper and lower respiratory illness. Lower respiratory illness was 33
to 56 percent more frequent in the higher pollution areas than in the low-pollution area
(p <0.005). Also, decrements in lung function as measured by spirometry tests were closely
associated with the occurrence of respiratory disease symptoms. The main respiratory symptom
results observed by Lunn et al. (1967) are illustrated in Figure 14-4.
The authors of the study (Lunn et al., 1967) highlighted the following points in
discussing their results;
"The respiratory measurement findings showed no association with area, social
class, children in the house, and sharing of bedrooms, although Attercliffe, the
area of highest pollution, showed reduced F.E.V.Q -,,- and F.V.C, ratios. On the
other hand, very clear evidence of reduced F.E.V.~'75 ratios emerged where there
was a past history of pneumonia and bronchitis, persistent or frequent cough, or
colds going to the chest. It must be stressed that these findings relate to first
year infant schoolchildren and that measurements were made during the summer term
when pollution levels were low and acute respiratory infections few and far
between. In other words, a pattern of respiratory disability had appeared at an
early age and was sufficiently established to persist although the factors of
pollution and infection were temporarily absent or at a low level."
In a second report, Lunn et al. (1970) gave results for 11-year-old children studied in
1963-64 that were similar to those provided earlier for the younger group. Upper and lower
respiratory illness occurred more frequently in children exposed to annual average 24-hour
o
mean smoke (BS) concentrations of 230 to 301 ug/m and 24-hour mean SOp concentrations of
181-275 ug/m3 than in children exposed to smoke (BS) at 97 ug/m and SO, at 123 ug/m . This
report also provided additional information obtained in 1968 on 68 percent of the children who
14-47
-------
60 r-
50
I 40
EC
UJ
U
z
UJ
111
EC
30
20
10
CHiST COLDS
PERSISTENT COUGH
100
200
300
BS (
Figure 14-4. History and clinical evidence of respira-
tory disease (percent) in 5-year-olds, by pollution in
area of residence. BS (/ug/m^) levels indicated above
must be taken as only very crude approximations of
actual PM mass present due to ambiguities regarding
use of site-specific calibrations in deriving the mass
estimates.

Source: Lunn et al. (1967).
14-48
-------
were 5 and 6 years old in 1963-64. By 1968, the reported concentrations of smoke (BS)* were
only about one-half those measured in 1964, -SO, concentrations were about 10 to 15 percent
below those measured in 1964, and the pollution gradient no longer existed, so the combined
three higher pollution areas were compared with the single original low-pollution area. Lower
respiratory illness prevalence measured as "colds going to chest" was 27.9 percent in the low-
pollution area and 33.3 percent in the combined high-p'ollut'ion areas, but the difference was
not statistically significant (p >0.05). Ventilatory function results were similar. Also,
the 9-year-old children had less respiratory illness than the 11-year-old group seen pre-
viously. Since 11-year-old children generally have less respiratory illness than do 9-year-
olds, this represented an anomaly that the authors suggested may have been the result of
improved air quality.
It should be noted that these Lunn et al. (1967, 1970) findings have been widely accepted
(Rail, 1974; Higgins, 1974; Holland et al., 1979; National Research Council, 1978a,b; Ferris,
1978; WHO, 1979) as being valid; and, on the basis of the results reported, it appears that
increased frequency of lower respiratory symptoms and decreased lung function in children may
3
occur with long-term exposures to annual BS levels in the range of 230-301 ug/m and S09
3
levels of 181-275 M9/ra • However, these must be taken only as very approximate "observed
effect" levels due to uncertainties associated with estimating PM mass based on BS readings.
Also, one cannot conclude based on the 1968 follow-up study results, that "no-effect" levels
were demonstrated for BS levels in the range of 48-169 yg/m and S0» levels in the range of
3
94-253 (jg/m due to: (1) the likely insufficient power of the study to have detected small
changes given the size of the population cohorts studied; and (2) the lack of site-specific
calibration of the BS mass readings at the time of the later (1968) study. The lack of
observed effects in this follow-up study suggests, however, that the retested children may
have recovered from air pollution-induced decrements in lung function detected by the earlier
study (Lunn et al., 1967).
In summary, the one above study by Lunn et al. (1967) in Sheffield, England, provides the
clearest evidence for association of significant pulmonary function decrements and increased
respiratory disease illnesses in children with chronic exposure to specific levels of 50^ and
PM in the ambient air.
14.5 SUMMARY AND CONCLUSIONS
Some epidemiological studies reviewed above appear to provide meaningful quantitative
information on health effects associated with ambient air exposures to PM and SO,. Others,
3
^However, in contrast to the reported smoke concentrations (in jjg/m ) for 1963 being based on
site-specific calibrations carried out in Sheffield in 1963, the later smoke estimates
reported for 1968 were not derived from any subsequent recalibration of the PM mass to BS
reflectance relationship. Such recalibration would be necessary to adjust for changes in
atmospheric PM chemical composition (especially the proportion of elemental carbon) likely to
have resulted from decreased emissions of partially combusted fossil fuel products from
open-hearth burning of coal, which markedly decreased in Sheffield from 1963 to 1968.

14-49
-------
however, either do not fully meet the various objectives discussed earlier under Section
14.1.2 or ambiguity exists regarding clear interpretation of their reported results. Only
some study results can, therefore, be accepted with a relatively high degree of confidence,
whereas others may provide, at best, only suggestive evidence for the reported associations
between air pollution and health effects. The main focus of this section is on summarizing
and interpreting results from those selected key studies yielding results that can now be
accepted with the greatest degree of certainty. However, even those findings must be viewed
as providing only very approximate estimates of ambient air levels of PM and SCL likely to be
associated with the health effects indicated; and the individual contributions of PM and SCL
cannot be clearly separated from the effects of each other or other covarying air pollutants.
In general, the epidemiological studies reviewed here provide strong evidence for induc-
tion by marked elevations of atmospheric levels of PM and SCL of severe health effects, such
as mortality and respiratory diesease, in certain populations at special risk. Populations at
special risk for such effects appear to include, mainly, the elderly and adults with chronic
preexisting cardiac or respiratory diseases (e.g., bronchitics). Increased respiratory tract
illnesses and more transient effects, e.g., decrements in pulmonary function, also appear to
be associated for children with lower chronic exposures to S0? and PM.
14.5.1 Health Effects Associated with Acute Exposures to Particulate Matter and Sulfur Oxides
As noted earlier in the present chapter, it is widely accepted that increases in
mortality occur when either S00 or PM (as BS) levels increase beyond 24-hour levels of 1000
3
pg/m . Such increased mortality, mainly in the elderly or chronically ill, might logically be
attributed in part to even brief exposures to very high short-term peak (hourly) levels in the
o
pollutants, which at times increased to several thousand pg/m during certain major pollution
episodes. However, none of the available epidemiological data have been collected or analyzed
in a manner so as to either credibly substantiate or refute this possibility. Much more
clearly established are marked increases in mortality and morbidity being associated with
o
prolonged episodic elevations of PM and SCL which average out to daily levels of 1000 jjg/m ,
especially in the presence of high humidity (fog) conditions, but which include continuous
exposures to high pollutant (PM and S0?) concentrations for several days without intermittent
relief or return to near normal levels at points between short-term pollution peaks. Thus,
although 24-hour concentrations of PM and S0? ^.1000 M9/m can be stated as levels at which
mortality has notably increased, great care must be exercised in generalizing these observa-
tions in attempting to predict likely effects associated with comparable elevations at other
times and locations. In particular, the prolonged or continuous nature of the high pollutant
exposures and other interacting factors present, e.g., high humidity levels, must be taken
into account as additional important determinants of mortality increases observed so far
during major air pollution episodes; and marked increases in mortality should not be expected
to occur regularly as a function of short-term peak excursions of 24-hour PM or SCL levels
3
above 1000 jjg/m . Consistent with this are numerous examples in the epidemiological litera-
14-50
-------
ture evaluated above where no detectable increases in mortality were found to occur on various
2
scattered days when PM and/or SO^ levels reached comparably high (>1000 ug/m ) 24-hour levels
as on other days (or sets of successive days) when mortality was clearly increased.
Even more difficult to establish are to what extent smaller but significant increases in
mortality and morbidity are associated with nonepisodic 24-hour average exposures to S09
3
and/or levels below 1000 ug/m . Concisely summarized in Table 14-7 are findings from several
key studies reviewed above which appear to demonstrate with a reasonably high degree of
certainty mortality and morbidity effects associated with acute exposures (24 hrs) to these
pollutants. The first two studies cited, by Martin and Bradley (1960) and Martin (1964), deal
with a relatively small body of data from London in the late 1950s. No clear "threshold"
levels were revealed by their analyses regarding S02 or BS levels at which significantly
increased mortality began to occur. However, based on their findings, and reanalyses of the
Martin and Bradley data by Ware et al (1981), mortality in the elderly and chronically ill
was clearly elevated in association with exposure to ambient air containing simultaneous SCL
3
and BS levels above 1000 jjg/m ; and some indications exist from these analyses that slight
increases in mortality may have been associated with nonepisodic BS and PM levels in the range
of 500-1000 pg/m (with greatest certainty demonstrated for levels in excess of 750 ug/m ).
Much less certainty is attached to suggestions of mortality increases at lower levels,
possibly based on the Ware et al. (1981) or other reanalyses (Appendix 14E) of the Martin and
Bradley data, especially in view of wide 95 percent confidence intervals demonstrated by one
analysis (Appendix 140) to be associated with estimation of dose-response relationships
between BS or SOp and mortality using the Martin and Bradley (1960) data. Analyses by
Mazumdar et al. (1981) for 1958-59 to 1971-72 are generally consistent with the above findings
but seem to suggest that the 1958-59 London winter may represent something of a worst-case
situation in comparison to most later winters. Still the Mazumdar et al. (1981) and certain
other analyses (Appendix 14F) of 1958-59 to 1971-72 London winter mortality data are strongly
indicative of small, but significant increases in mortality occurring at BS levels below 500
3 3
ug/m and, possibly, as low as 150 to 200 ug/m .
Only very limited data exist by which to attempt to delineate any specific physical and
chemical properties of PM associated with the observed increases in mortality. Based on
information noted earlier (Section 14.2), it would seem that marked increases in small parti-
3
cles to levels above 500-1000 ug/m appear to be most cleanly associated with increased
mortality, based on the BS aerometric measurements reported, although contributions from
larger coarse-mode particles cannot be completely ruled out. Nor is it possible to state with
certainty specific PM chemical species associated with the increases in mortality. We do know
that large amounts of pollutants (e.g., elemental carbon, tarry organic matter, etc.) from
incomplete combustion of coal were present in the air and mortality levels appeared to
decrease as PM concentrations declined over the years; but no single component or combinations
of particulate pollutants can clearly be implicated. Neither can the relative contributions
14-51
-------
TABLE 14-7. SUMMARY OF QUANTITATIVE CONCLUSIONS FROM EPIDEHIOLOGICAL STUDIES RELATING HEALTH
EFFECTS TO ACUTE EXPOSURE TO AMBIENT AIR LEVELS OF S02 AND PH
Type of study
Effects observed
24-hr average pollutant level (ug/m)
BS SO.
Reference
Mortality
i
en
ro
Morbidity
Clear increases in daily total
mortality or excess mortality above
a 15-day moving average among the
elderly and persons with preexisting
respiratory or cardiac disease during
the London winter of 1958-59.

Analogous increases in daily
mortality in London during
1958-59 to 1971-72 winters.
>10DO
>1000 Martin and Bradley
(1960); Martin (1964)
Mazumdar et al, (1981)
Some indications of likely increases
in daily total mortality during the
1958-59 London winter, with greatest
certainty (95% confidence) of increases
occurring at 8S and SO,, levels above
750 ug/m .
Analogous indications of increased
mortality during 1958-59 to 1971-72
London winters, again with greatest
certainty at BS and SO- levels above
750 pg/n> but indications of small
increases at BS levels <500 (jg/n and
possibly as low as 150-200 ug/m .
500-1000
500-1000 Martin and Bradley (1960)
Mazumdar et al, (1981)
Worsening of health status among
a group of chronic bronchitis
patients in London during
winters from 1955 to 1960
>25Q-500*
>500-600
Lawther (1958); Lawther
et al. (1970)
No detectable effects in most
bronchitics; but positive
associations between worsening
of health status among a
selected group of highly
sensitive chronic bronchitis
patients and London BS and SO-
levels during 1967-68 winter.
<250*
<500
Lawther et al.
(1970)
*Note that the 250-500 jjg/n BS levels stated here may represent somewhat higher PM concentrations than those actuajly
associated with the observed effects reported by Lawther (1970). This is due to the estimates of PM mass (in |jg/m BS)
used by Lawther being based on the 0 S I R. calibration curve found by Waller (1964) to approximate closely a site-specific
calibration curve developed by Waller in central London in 1956, but yielding somewhat higher mass estimates than another
site-specific calibration developed by Waller a short distance away in 1963. However, the precise relationship between
estimated 8S mass values based on the D.S.I.R curve versus the 1963 Waller curve cannot be clearly determined due to several
factors, including the non-linearity of the two curves and their convergence at low BS reflectance values
-------
of SOg or PM be clearly separated based on these study results nor can possible interactive
effects with increases in humidity (fog) be completely ruled out. Temperature change,
however, does not appear to be a key determinant in explaining mortality effects demonstrated
by the above analyses to be associated with atmospheric elevations of PM or SCL.
Studies conducted in New York City by Greenburg and coworkers (and other groups of
investigators) appear 'to suggest, with less confidence, that slight mortality increases may
have been associated with episodic increases in 24-hour S0? levels above 1000 yg/m and CoHs
levels of 5.0-7.0, but little credible evidence exists for mortality having occurred with
lower acute (24-hour) exposures. Again, specific particulate chemical species cannot be
clearly implicated nor the relative contributions of SO, and particulate matter separated in
regard to any induction of mortality by episodic pollution. It should be noted that, whatever
the causal agents, only very small increases in mortality may have been detected at the above
pollutant levels in New York City; and associated morbidity effects appear to have been
restricted mainly to increased respiratory or cardiac complaints among the elderly.
Similar analysis of the Lawther morbidity studies listed in Table 14-7 suggests that
3
acute exposure to elevated 24-hour PM levels in the range of 250-500 pg/m (BS) in association
with 24-hour SO^ levels of 500-600 pg/m were most clearly associated with the induction
of respiratory disease symptoms among large (>1000) populations of chronically ill London
bronchitis patients. A smaller population (~80) of selected, highly sensitive London
bronchitic patients appeared to be affected at somewhat lower BS and S0? levels, but specific
exposure-effect levels could not be determined on the basis of the reported data. Again,
however, little can be said in terms of specifying physical or chemical properties of PM
associated with these observed morbidity effects beyond the comments noted above in relation
to Martin's studies on mortality.
14.5.2 Health Effects AssociatedI with Chronic Exposures to PM and SO
/\
In regard to chronic exposure effects of S0? and particulate matter, the best pertinent
epidemiological health studies are summarized in Table 14-8. The studies by Ferris et al.
(1973, 1976) suggest that lung function decrements may occur in adults at TSP levels in excess
of 180 (jg/m in the presence of relatively low estimated S09 levels, whereas no effects were
3
observed by the same investigators at TSP levels below 130 jjg/m . Other studies listed in
Table 14-8 suggest that significant respiratory effects occur in children in association with
3
long-term (annual average) PM levels in the approximate range of 230-301 p.g/m (BS) in associ-
ation with S02 levels of 181-275 [jg/m3.
No specific particulate matter chemical species can clearly be implicated as causal
gents associated with the effects observed in the studies listed in Table 14-8. Nor can
'Otential contributions of relatively large inhalable coarse mode particles be ruled out based
m these study results. It should be remembered that various occupational studies listed in
•ppendix 14B at least qualitatively suggest that such sized particles of many different types
f chemical composition can be associated with significant pulmonary decrements, respiratory
ract pathology, and morphological damage.

14-53
-------
TABLE 14-8. SUMMARY OF QUANTITATIVE CONCLUSIONS FROM EPIOEMIOLOGICAL STUDIES
RELATING HEALTH EFFECTS TO CHRONIC EXPOSURE TO AMBIENT AIR LEVELS OF S02 AND PM
Annual average pollutant levels (ug/m )
particulate matter

Cross-sectional
(4 areas)
Longitudinal
and cross-
sectional
Longitudinal
and cross-
sectional
Likely increased frequency
of lower respiratory symp-
toms and decreased lung
function in children in
Sheffield, England
Apparent improvement in
lung function of adults
in association with decreased
PM pollution in Berlin, N.H.
Apparent lack of effects
and symptoms, and no apparent
decrease in lung function in
adults in Berlin, N.H.
230-301* - 181-275 Lunn et al.
(1967)
180 ** Ferris et al.
(1973, 1976)
80-131 ** Ferris et al.
(1973, 1976)

*Note that BS levels- stated here in ug/m must be viewed as only crude estimates of the approximate PM (BS) mass levels
. associated with the observed health effects, given ambiguities regarding the use or non-use of site-specific calibrations
in Sheffield to derive the reported BS levels in ug/m .

**Note that sulfation rate methods indicated low atmospheric sulfur levels in Berlin, N.H. during the time of these studies.
Crude estimation of S0,> levels from that data suggest that <25-50 ug/m SO, levels were generally present in Berlin, N.H.,
and did not likely contribute to observed health effects.
-------
Only very limited information has been published (Commins and Waller, 1967) on the chemi-
cal composition of participate matter present in London air during the period of some of the

above epidemiological studies demonstrating associations between mortality or morbidity

effects and elevations in PM levels, as summarized in Table 14-9. Such data may provide im-
portant clues as to possible causative agents involved in the etiology of health effects ob-
served in London during the 1950s and early 1960s. For the sake of comparison, information on
measured chemical components of TSP matter in U.S. cities during the early 1960s is also pro-
vided in Table 14-9. It must be noted, however, that likely substantial differences in speci-
fic components of the PM present in London air of the 1950's and 1960's versus the chemical
composition of PM currently present in urban aerosols over American cities argue for much cau-

tion in extrapolating results of London epidemiological studies for present criteria develop-
ment purposes.

TABLE 14-9. COMPARISON OF MEASURED COMPONENTS OF TSP IN U.S. CITIES (1960-1965)
AND MAXIMUM 1-HOUR VALUES IN LONDON (1955-1963)
UNITED. STATES'
LONDON1
Pollutant
Concentration |jg/m3
Number of Arith. Maximum
stations average 24-hour
Maximum
1-hour
Suspended Particles 291
Fractions;
Benzene-soluble organics 218
Chloride (water soluble)
Nitrates 96
Sulfates"I"""~""""I 96
Sulfuric acid
Ammonium 56
Antimony 35
Arsenic ~133
Beryl!ium~__I 100
Bismuth__~__~~~~~~~"~~~~~_ 35
Cadmium 35
Calcium
Chromium _ _ 103
Cobalt__m™_IIIIIIII_III 35
Copper ~ 103
jron ~_ ~~~ io4
LeadIIII_HI"im2III_I_III 104
Manganese 103
Molybdenum, _ 35
Nickel I 103
Tin „ 85
Titanium"*""' ~ 104
Vanadium ~_ 99
Zinc IIIIIIIIIHII.IIIII 99
Gross beta radioactivity 323
105

6.8

2.6
10.6

1.3
0.001
0.02
<0.0005
<0.0005
0.002

0.015
<0.0005
0.09
1.58
0.79
0.10
<0.005
0.034
0.02
0.04
0.050
0.67
1254 (TSP)
39.7
101.2

75.5
0.160

0.010
0.064
0.420

0.330
0.060
10.00
22.00
8.60
9.98
0,78
0.460
0.50
1.10
2.200
58.00
9700 (Smoke)
410
5
666
680
1
32
2
<1
2
25
22
5
<1
1
2
1
2
24
(0.8 pCi/rn ) (12.4 pCi/m )
"U.S. Dept. Health, Education, & Welfare, 1970.
Commins and Waller, 1967.
"Obtained from one London site.
14-55
-------
14,6 REFERENCES*

American Thoracic Society. Epidemiology Standardization Project. Am. Rev. Res. Dis.
118(6,pt.2), 1978.

Anderson, D. 0., and A. A. Larsen. The incidence of illness among young children in two
communities of different air quality: A pilot study. Can. Med. Assoc. J. 95:893, 1966.

Anderson, D. 0., and B. G. Ferris, Jr. Air pollution levels and chronic respiratory disease.
Arch. Environ. Health 10:307-311, 1965.

Anderson, D. 0., B. G. Ferris, Jr.,- and R. Zinkmantel. Levels of air pollution and
respiratory disease in Berlin, New Hampshire. Am. Rev. Respir. Dis. 90:877-887, 1964.

Angel, J, H., C. M. Fletcher, I. D. Hill, and C. M. Pinker. Respiratory illness in factory
and office workers. Br. J. Dis. Chest 59: 66-80, 1965.

Ashley, D. J. B. The distribution of lung cancer and bronchitis in England and Wales. Brit,
J. Cancer 21: 243, 1967.

Aubrey, F., G. W. Gibbs, and M. R. Becklake. Air Pollution and Health in three urban
communities. Arch. Environ. Health 34:360-368, Sept/Oct 1979.

Ball, D. J. , and R. Hume. The relative importance of vehicular and domestic emissions of dark
smoke in Greater London in the mid-19701s, the significance of smoke shade measurements,
and an explanation of the relationship of smoke shade to gravimetric measurements of
participate. Atmos. Environ. 11:1065-1073, 1977.

Bates, D. V. Air pollution and chronic bronchitis. Arch. Environ. Health 14.220, 1967.

Bates, D. V. The fate of the chronic bronchitic: A report of the ten-year followup in the
Canadian Department of Veterans Affairs Coordinated Study of Chronic Bronchitis. Am.
Rev. Res. Dis. 108:1043, 1973.

Bates, D. V., C. A. Gordon, G. I. Paul, R. E. G. Place, D. P. Snidal, and C. R. Woplf. (with
special sections contributed by M. Katz, R. G. Fraser, and B. B. Hale) Chrome bronchi-
tis. Report on the third and fourth stages of the Coordinated Study of Chronic
Bronchitis in the Department of Veterans Affairs. Canada. Med. Serv. J. Can. 22:5,
1966.

Bates, D. V., C. R. Woolf, and G. I. Paul. Chrome bronchitis: A report on the first two
stages of the Coordinated Study of Chronic Bronchitis in the Department of Veterans
Affairs. Canada. Med. Serv. J. Can. 18:211, 1962.

Becklake, M. R., J. Soucie, G. W. Gibbs, and H. Ghezzo. Respiratory health status of children
in three Quebec urban communities. An Epidemiologic Study. Bull. Europ. Physiopath.
Resp. 14:205-221, 1978.

Becker, W. H, , F. J. Schilling, M. P. Verma. The effect on health of the 196.6 Eastern
Seaboard air pollution episode. Arch. Environ. Health 16:414-419, 1968. !-

Bennett, A. E., W. W. Holland, T. Halil, and A. Elliot. Lung function and air pollution.
Chronic inflammation of the bronchi. Prog. Respir. Res. 6:78-89, 1971.
*Note that references are included here for studies discussed in accompanying Chapter 14
appendix materials, as well as for studies cited in the main text of Chapter.

14-56
-------
Biersteker, K. Polluted Air Causes, Epidemiological Significance, and Prevention of
Atmospheric Pollution. Assen, Netherlands, Van Gorcum and Co., pp. 21-23 (in Dutch),
1966.

Biersteker, K. , and P: van Leeuwen. Air pollution and peak flow rates of schoolchildren in
two districts of Rotterdam. Arch. Environ. Hea'lth- 2^:382-384, 1970.

Biersteker, K. , and P. van Leeuwen. Air pollution, bronchitis prevalence and peak flow rates
of schoolchildren in two districts of Rotterdam (Netherlands). In: 2nd Int. Clean Air
Cong. Proc. Washington, D.C., December 1970. H. M. Englund and W. T. Berry (ed.). New
York, Academic Press, p. 209-212.

Binder, R. E. , C. A. Mitchell, H. R. Hosein, and A. Bouhuys. Importance of the Indoor
Environment in Air Pollution Exposure, Arch, of Em/iron. Health. Nov./Dec. 277-279, 1976.

Bouhuys, A., G. J. Beck, and J. B. Schoenberg. Do present levels of air pollution outdoors
affect respiratory health? Nature 276:466-471, 1978.

Bouhuys, A., G. J. Beck, and J. B. Schoenberg. Priorities in prevention of chronic lung
diseases. Lung, 156:129-148, 1979.

British Standards Institution. Methods for the measurements of air pollution, Part 2.
Determination of Concentration of Suspended Matter. British Standard 1747: Part 2,
1964

Buck, S. F. , and A. D. Brown. Mortality from Lung Cancer and Bronchitis in Relation to Smoke
and Sulfur Dioxide Concentration, Population Density, and Social Index. Research Paper
No. 7. London, Tobacco Research Council. 1964.

Buechley, R. W. S02 levels, 1967-1972 and perturbation in mortality. A further study in the
New York-New Jersey Metropolis. Presented to the Health Research Council of the City of
New York Meeting, The Environmental Pollution Impact of Stationary Source Fossil Fuel Air
Pollution Control Measures in New York City, March 6, 1975.

Buechley, R. W. S02 Levels, 1967-1972 and Perturbations in Mortality. Contract No.
ES-5-2101. Report, available from National Institute of Environmental Health Sciences,
Research Triangle Park, N.C., 1977.

Buechley, R. W., W. B. Riggan, W. Hasselblad, and J. B. Van Bruggen. S0? levels and perturba-
tions in mortality. A study in New York-New Jersey metropolis. Arch. Environ.
Health 27:134-137, 1973.

Burgess, S. E. , and C. W. Shaddick. Bronchitis and air pollution. R. Soc. Health J.
79-10-24, 1959.

Burn, J. L. , and J. Pemberton. Air pollution bronchitis and lung cancer in Salford. Int. J.
Air Water Pollut. 7:5-16, 1963.

Burrows, B., A. L Kellogg, and J. Bushey. Relationship of symptoms of chronic bronchitis and
emphysema to weather and air pollution. Arch. Environ. Health 16:406-413, 1968.

Came, S. Air Pollution Study. Proc. Royal Soc. Medicine 57:30-34, 1964.

Carnow, B W. Sulfur oxides and particles. Effects on health. Proceedings of the National
Academy of Science Conference on Health Effects of Air Pollution. U.S. Government Print-
ing Office, Washington, DC. Stock No. 5270-02105, 1973. pp. 263-291.
14-57
-------
Carnow, B. W. , M. H. Lepper, R, B. Shekelle, and J. Stamler. Chicago air pollution study.
Arch. Environ. Health 18:768-776, 1969.

Cassell, E. J., M. D. Lebowitz, and J. R. McCarroll. The Relationship Between Air Pollution,
Weather, and Symptoms In an Urban Population. Am. Rev. Res. Dis. 106:677-683, 1972.

Cassell, E. J. , M. D. Lebowitz, I. M. Mountain, H. T. Lee, D. J. Thompson, D. W. Wolter, and
J. R. McCarroll. Air Pollution, Weather, and Illness in a New York Population. Arch.
Environ. Health 18:523-530, 1969.

Cassell, E. J., D. W. Walter, J. D. Mountain, J. R. Diamond, I. M. Mountain, and J. R.
McCarroll II. Reconsiderations of mortality as a useful index of the relationship of
environmental factors to health. Amer. J. Pub. Health 58:1653-1557, 1968.

Cederlof, R. Urban factor and prevalence of respiratory symptoms and "angina pectoris."
Arch. Environ. Health 13:743-748, 1966.

Chappie, M. , and L. Lave. The health effects of air pollution. A Reanalysis. J. Urban
Economics, 1981.

Chapman, R. S. , C. M. Shy, J. F. Finklea, D. E. House, H. E. Goldberg, and C. G. Hayes.
Chronic Respiratory Disease in Military Inductees and Parents of School Children. Arch.
Environ. Health ^7:138, 1973.

Chanska, M. Changes in peak expiratory flow rate (PEFR) in children during three year
observations. Ph.D. Dissertation, State Inst. Hygiene, Warsaw, 1980.

Chiaramonte, L. T., J. R. Bongiorno, R. Brown, and M. E. Laano. Air pollution and obstructive
respiratory disease in children. NY State J. Med. 70:394, 1970.

Christainsen, G. B. and C. G. Degen. Air pollution and Mortality Rates: A Note on Lave and
Seskin's Pooling of Cross-Section and Time-Series Data. J. of Env. Economics and Mang.
7:149-155, 1980.

Ciocco, C. A., and D. J, Thompson. A Follow-up at Donora Ten Years After: Methodology and
Findings. Am. J. Publ. Health 51:155-164, 1961.

Clifton, M., D. Kerridge, J. Pemberton, W. Moulds, and J. K, Donoghue. Morbidity and mortal-
ity from bronchitis in Sheffield in four periods of severe pollution. In: Proc. 1959
Int. Clean Air Conf, London, National Society for Clean Air. 1960. p. 189.

Cohen, A. A., S. Bromberg, R. W. Buechley, L. T. Heiderescheit, and C. M. Shy. Asthma and air
pollution from a coal fueled power plant. Am. Rev. J. Pub. Health. 62, 1972. p. 1181.

Colley, J. R, T., and D. D. Reid. Urban and social origins of childhood bronchitis in England
and Wales. Br. Med. J. 2:213-217, 1970.

Colley, J. R. T. , and L. J. Brasser. Chronic respiratory diseases in children in relation to
air pollution. Report of a WHO study. EVRO Reports and Studies 28, Regional Office for
Europe, Copenhagen, 1980.

Colley, J. R, T., and W. W. Holland. Social and Environmental Factors in Respiratory Disease.
Arch Environ. Hlth. 14:157, 1967.

Collins, J. J. , H. S. Kasap, and W. W. Holland. Environmental factors in child mortality in
England and Wales. Am. J. Epidemiol. 93:10, 1971.

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.

14-58
-------
Comstock, G. W. , R. W. Stone, Y. Sakai, T. Matsuya, and J. A. Tonascia. Respiratory findings
and urban living. Arch. Environ. Health 27:143, 1973.

Cowan, D. W., H. J. Thompson, H. J. Paulus; and P. W. Mielke. Bronchial asthma associated with
air pollutants from the grain industry. J. Air Poll. Contr. Assoc. 13:546, 1963.

Crocker, T. G., W. Schulze, S. Ben-David, and A. V. Kneese. Methods development for assessing
air pollution control benefits, Volume I: Experiments in the economics of air pollution
epidemiology. EPA-600/5-79-001a, Environmental Protection Agency, Research Triangle
Park, N.C., 1979.

Deane, M. , J. R. Goldsmith, and D. Tuma. Respiratory conditions in outside workers. Report
on outside plant telephone workers in San Francisco and Los Angeles. Arch. Environ.
Health 10:323, 1965.

Derrick, E. H. A comparison between the density of smoke in the Brisbane air and the preva-
lence of asthma. Med. J. Aust. 11:670-675, 1970.

Detels, R. , J. W. Sayre, A. H, Coulson, S. N, Rokaw, F. J. Massey, Jr., D. P. Tashking and
M. Wu. Respiratory Effect of Long Term Exposure to Two Mixes of Air Pollutants in
Los Angeles County. Chest 80:27S-29S, 1981

Dockery, D. W. , N. R. Cook, B. G. Ferris, F. E. Speizer, J. D. Spengler, and J. H. Ware.
Changes in pulmonary function in children with air pollution episodes. Proceedings 74th
Annual APCA, Philadelphia, Pa., June 21-26, 1981.

Dodge, R. The respiratory health of school children in smelter communities. Amer, J. Ind.
Med. 1-359-364, 1980.

Dohan, F. C. Air pollutants and incidence of respiratory disease. Arch. Environ Health
3:387-395, 1961.

Dohan, F. C. , and E. W, Taylor. Air Pollution and Respiratory Disease, A Preliminary Report.
Am. J. Med. Sci 240:337, 1960.

Doll, R. Atmospheric pollution and lung cancer. Environ. Health Perspect. 22:23-31, 1978.

Douglas, J. W. B. , and R. W. Waller. Air pollution and respiratory infection in children.
Br J. Prev. Soc. Med. 20:1-8, 1966

Ellison, J. The estimation of particulate air pollution from the soiling of filter paper.
Staub Reinhalt, Luft 28:28, 1968.

Emerson, P. A. Air pollution, atmospheric conditions and chronic airway obstructions. J.
Occup. Med. 15:635-638, 1973.

Fairbairn, A. S. , and D. D. Reid. Air pollution and other local factors in respiratory
disease. Br. J. Prev. Soc. Med. 12:94, 1958.

Ferris, B. G. , Jr. Health Effects of Exposures to Low Levels of Regulated Air Pollutants. A
Critical Review. JAPCA 28:482-497, 1978.

Ferris, B. G. , Jr., and D. 0. Anderson. The prevalence of chronic respiratory disease in a
New Hampshire town Am. Rev. Respir. Dis. 86:165-177, 1962.

Ferris, B. G. , Jr., Burgess, W. A., and J. Worchester, J. Prevalence of chronic respiratory
disease in a pulp mill and a paper mill in the United States. Br. J. Ind. Med., 24:26-37,
1967
14-59
-------
Fern's, B. G., Jr., H. Chen, S. Puleo, and R. L. H. Murphy, Jr. Chronic non-specific respira-
tory disease in Berlin, New Hampshire, 1967-1973. A further follow-up study. Am. Rev.
Respir. Dis. 113: 475-485, 1976.

Ferris, B. G., Jr., I. T. T. Higgins, M, W. Higgins, J. M. Peters, W. F. Van Guase, and M. D.
Goldman. Chronic non-specific respiratory disease, Berlin, New Hampshire, 196L-67: A
cross section study. Am. Rev. Respir. Dis. 104:232-244, 1971.

Ferris, B. G., Jr., I. T. T. Higgins, M. W. Higgins, and J. M. Peters. Chronic non-specific
respiratory disease in Berlin, New Hampshire, 1961-67. A follow-up study. Am. Rev.
Respir. Dis. 107:110-122, 1973.

Ferris, B. G., Jr., I. T. T. Higgins, M. .W. Higgins, and J. M. Peters. Sulfur oxides and sus-
pended particulates, possible effects of chronic exposure. Arch. Environ. Health
27:179-182, 1973.

Ferris, B. G., Jr., J. R. Mahoney, R. M. Patterson, and M. W. First. Air quality, Berlin, New
Hampshire, March 1966 to December 1967. Am. Rev. Respir. Dis. 108:77-84, 1973.

Ferris, B. G. , Jr., F. E. Speizer, Y. M. M. Bishop, J. D. Spengler, and J. H. Ware. The
six-city study: A progress report. In: Atmospheric Sulfur Deposition: Environmental
Impact and Health Effects. Ann Arbor Sicence, D. S. Shriner, C. R. Richmond, and S. E.
Lindberg, eds., Ann Arbor Science, Ann Arbor, Michigan, 1980. pp. 99-108.

Ferris, B. G., Jr., R. E. Speizer, J. D. Spengler, D. Dockery, Y. M. M. Bishop, M Wolfson,
and C. Humble. Effects of sulfur oxides and respirable particles on human health
methodology and demography of populations in study. Am. Rev. Respir. Dis. 120:767-779,
1979.

Firket, J. Sur les causes des accidents survenus dans la valee de la Meuse, Tors des brouil-
lards de Decembre, 1930. Bull. Acad. R. Med. Belg. 11.683-741, 1931.

Firket, J. Fog along the Meuse Valley Trans Fraday Soc. 32:1191-1194, 1936.

Fletcher, C. M. , R. Peto, C. M. Tinker, and F. E. Speizer. The Natural History of Chronic
Bronchitis and Emphysema ( an eight year study of early chronic obstructive lung disease
in working men in London). Oxford University Press, 1976.

French, J. G. , G. Lowrimore, W. C. Nelson, J. F. Finklea, T English, and M. Hertz. The
effect of sulfur dioxide and suspended sulfate on acute respiratory disease. Arch. Envirc
Health 27:129-133, 1973.

Fujita, S., T. Motoichi, K. Shoji, Y. Ichiro, F. Takashi, S. Seigo, K. Tatsuo, and M. Michiko
Studies on chronic bronchitis epidemiological survey (2nd report). Teishin Igaku 21:13,
1969.

Gervois, M., G. Dubois, S. Gervois, J. M. Queta, A. Muller, and C. Vorsin. Atmospheric pollu-
tion and acute respiratory disease. Denoin and Quavrechoin epidemiological study. Rev.
Epidemic!. Sante Publique 25:195-207, 1977.

Gerking, S. , and W. Schultze. What do we know about benefits of reduced mortality from air
pollution control? Amer. Economic Review 71:228-234, 1981.

Girsn, L. S., E. Shubin, C. Dick, and F. A. Schulaner. A Study on the Epidemiology of Asthma
in Children in Philadelphia. J. Allergy 39:347-357, 1967.

Glasser, M. , and L. Greenburg. Air pollution and mortality and weather, New York City,
1960-64. Arch. Environ. Health 22:334-343, 1971.
14-60
-------
Glasser, M. , L. Greenburg, and F. Field. Mortality and Morbidity During a Period of High
Levels of Air Pollution, New York, November 23-25, 1966, Arch. Environ. Health
15:684-694, 1967.

Goldsmith, J. R., and L. T. Friberg. Effects of Air Pollution on Human Health. In: Air Pol-
lution . II. A. C. Stern, ed., Academic Press, New York, 1977. pp. 458-610.

Goldstein, I. F. , and G. Block. Asthma and air poThttton in"two inner city areas in New York
City. J. Air Pollut. Control Assoc. 24:665-670, 1974.

Goldstein, I. F. and E. M. Dulberg. Asthma and air pollution in two inner city areas in New
York City. J. Air Pollut. Contro. Assoc. 24:370-376, 1981.

Goldstein, I,, and L. Landowitz (Letter to editor). J, Air Pollut. Control Assoc. 25:1195,
1975.

Goldstein, I. F. and L. Landovitz. Analysis of air pollution patterns in New York City--I.
Can one station represent the large metropolitan area? Atmos. Environ. 11:47-52, 1977a.

Goldstein, I. F. and L. Landovitz. Analysis of air pollution patterns in New York City—11.
Can one aerometric station represent the area surrounding it? Atmos. Environ. 11:53-57,
1977b.

Gore, A, T, , and C. W. Shaddick. Atmospheric pollution and mortality in the County of London.
Br. J. Prev. Soc. Med. 12:104-113, 1958.

Gorham, E. Bronchitis and the acidity of urban precipitation. Lancet 2:691, 1958.

Gorham, E. Pneumonia and atmospheric sulphate deposit. Lancet 2:287, 1959.

Greenberg, L., M. Jacobs, B. Drolette, F. Field, and M. Braverman. Report of an air pollution
incident in New York City, November, 1953. Pub Health Rep. 77:7-16, 1962.

Greenburg, L. , C. Erhardt, F. Field, J. I. Reid, and N. S. Seriff. Intermittent air pollution
episode in New York City, 1962. Public Health Rep. 78:1061-1064, 1963.

Greenburg, L., F. F. Field, J. I. Reed, and C. L. Erhardt. Air pollution and morbidity in New
York City. J. Am. Med. Assoc. 182:161-±64, 1962.

Greenburg, L. , F. Field, J. I. Reed, and C. L. Erhardt. Asthma and temperature change. An
epidemiological study of emergency clinic visits for asthma in three large New York
Hospitals. Arch. Environ. Health 8:642, 1964.

Greenburg, L. , F. Field, C. Erhardt, M. Glasser, and J. Reed. Air pollution, influenza, and
mortality in New York City. Arch. Environ. Health 15:430, 1967.

Gregory, J. The influence of climate and atmospheric pollution on exacerbations of chronic
bronchitis. Atmos. Environ. 4:453-468, 1970.

Hagstrom, R. M, , H. A. Sprague, and E. Landau. The Nashville air pollution study. VII. Mor-
tality from cancer in relation to air pollution. Arch. Environ. Health 15:237-248, 1967.

Hammer, D. I., F. J. Miller, A. G. Stead, and C. G. Hayes. Air Pollution and Childhood Lower
Respiratory Disease, I. Exposure to Sulfur Oxides and Particulate Matter in New York,
1972. In; Clinical Implications of Air Pollution Research. A. J. Finkel and W. C.
Duel, ed. , Publishing Sciences Group, Inc., Acton, MA, 1976. pp. 321-337.

Hammer, D. I. Frequency of lower respiratory disease in children: Retrospective survey of two
southeastern communities, 1968-1971. Ph.D. Dissertation, Harvard, Univ., 1976.

14-61
-------
Heimann, H. Episodic air pollution in metropolitan Boston. A trial epidemiclogic study.
Arch. Environ. Health 20:230-251, 1970.

Hewitt, D. Mortality in the London boroughs, 1950-52, with special reference to respiratory
disease. Br. J. Prev. Soc. Med. 10:45, 1956.

Higgins, I. T. T. Epidemiology of Chronic Respiratory Disease: A Literature Review.
EPA-650/1-74-007, U.S. Environmental Protection Agency, DC, 1974.

Higgins, I. T. T, Trends in respiratory cancer mortality in the U.S. and in England and
Wales. Arch. Environ. Health. 28:121-129, 1976.

Hill, A, 8. The Environment and Diseases: Associations and Causation. In. Proceedings of
the Royal Society of Medicine (Occ. Med.) 58:272, 1965.

Hodgson, A,, JK Short-term effects of air pollution on mortality in New York City. Environ.
Sci. Techno!. 4:589-597, 1970.

Holland, W. W. , A. E. Bennett, I. R. Cameron, C. du V. Florey, S. R. Leeder, R. S. F
Schilling, A. V. Swan, and R. E. Waller. Health Effects of Particulate Pollution:
Re-appraising the Evidence. Am. J Epidemic!. 110(5):525-659, 1979.

Holland, W, W. , and D. D. Reid. The Urban Factor in Chronic Bronchitis. Lancet 1:445-448,
1965.

Holland, W. W. , and R. W. Stone. Respiratory disorders in United States East Coast telephone
men. Am. J. Epidemic!. 82:92-101, 1965.

Holland, W. W. , ed. Data Handling in Epidemiology. Oxford University Press, London, 1970.

Holland, W. W., T. Halil, A. E. Bennett, and A. Elliot. Indications for measures to be taken
in childhood to prevent chronic respiratory disease. Milbank Mem Fund Q 47:215-227,
1969a.

Holland, W. W. , T. Halil, A. E. Bennett, and A. Elliot. Factors influencing the onset of
chronic respiratory disease. Br. Med. J. 2:205-208, 1969b.

Holland, W. W. , H. S. Kasap, J. R. T. Colley, and W. Cormack. Respiratory symptoms and ven-
tilatory function: A family study. Br. J. Prev. Soc. Med. 23:77-84, 1969c.

Hosein, H. R., C. A. Mitchell, and A. Bouhuys. Evaluation of outdoor air quality in rural and
urban communities. Archives of Environ. Health. Jan./Feb. 4-13, 1977a.

Hosein, H. R. , Mitchell, C. A., and A. Bouhuys. Daily variation in air quality Arch of
Environ. Health Jan./Feb. 14-21, 1977b.

Hrubec, Z. , R. Cederlof, L. Freberg, R. Horton, and 6. Ozolins. Respiratory symptoms in
twins. Arch. Environ. Health 27:189-195, 1973.

Ingram, W. Smoke Curve Calibration. PHS Contract PH-86-68-66, New York University, New York,
NY, 19S9.

Ingram, W. T., and J. Golden. Smoke curve calibration. J.'Air Pollut. Control Assoc. 23:110,
1973,

Ipsen, J., M. Deane, and F. E. Ingenito. Relationship of acute respiratory disease to atmos-
pheric pollution and meteorological condition. Arch. Environ. Health 18:462-472, 1969.
14-62
-------
Irwig, L. , D. G. Altman, R. J. W. Gibson, and C. Du V. Florey. Air Pollution: Methods to
study its Relationship to Respiratory Disease in British Schoolchildren. Proceedings of
the Intermath Symp on Recent Advances with Asses, of the Health Effects of Environ. Pol.,
Luxembourg: Commission of the European Communities, Vol 1, 1975 pp. 289-300.

Ishikawa, S. , D. H. Bowden, V. Fisher, and J. P. Wyatt. The "emphysema profile" in two mid-
western cities in North America. Arch. Environ. Health 18:660, 1969.

Jacobs, C. , and B. Langdoc. Cardiovascular deaths and air pollution in Charleston, South
Carolina. Health Services Reports 87:623-632, 1972.

Johnson, K. G., R. Gideon, and D. Loftsgaonden. Montana air pollution study. Synopsis of
pulmonary function studies. Proceedings of 74th Annual APCA Meeting, Philadelphia, Pa.,
June 21-26, 1981.

Kaaawa, J. , and T. Toyama. Photochemical Air Pollution. Arch. Environ. Health 30:117-122,
1975.

Kagawa, J. , T. Toyama, and M. Nakaza. Pulmonary function tests in children exposed to air
pollution. In: Clinical Implications of Air Pollution Research. A. J. Finkel and W. C.
Duel, ed. , Publishing Sciences Group, Inc., Acton, MA, 1976. pp. 305-320.

Kalpazanov, Y. , M. Stamenova, and G. Kurchatova. Air pollution and the 1974-1975 influenza
epidemic in Sofia. Environ. Res. ^2.1-8, 1976.

Kenline, P. In quest of clean air for Berlin, New Hampshire, U.S. Department of Health,
Education, and Welfare. R. A. Taft Sanitary Engineering Center, Cincinnati, Ohio, 1962.

Kenline, P. A. October 1963 New Orleans asthma study. Arch. Environ. Health 12:295-304,
1966.

Kevany, J. , M. Rooney, and J. Kennedy. Health effects of air pollution in Dublin. Ir. J.
Med. Sci. 144:102-115, 1975.

Kiernan, K. E. , J. R. T. Colley, J. W. B. Douglas, and D. D. Reid. Chronic cough in young
adults in relation to smoking habits, childhood environment and chest illness. Respira-
tion 33:236-244, 1976.

Lambert, P. M. , and D. D. Reid. Smoking, air pollution and bronchitis in Britain. Lancet
1:853-857, 1970.

Lave, L. B., and B. P. Seskin. Air pollution and human health. The quantitative effect, with
an estimate of the dollar benefit of pollution abatement is considered. Science
169:723-733, 1970.

Lave, L. B. , and B. P. Seskin. Air pollution, climate, and home heating: Their effects on
U.S. mortality rate. Am. J. Public Health 62:909-916, 1972.

Lave, L. B. , and B. P Seskin. Air Pollution and Human Health. Baltimore, The Johns Hopkins
University Press. 1977.

Lawther, P. J. Climate, air pollution and chronic bronchitis Proc. R. Soc. Med. 51:262-264,
1958.

Lawther, P. J. Compliance with the Clean Air Act: Medical aspects. J. Inst. Fuel 36:341,
1963.
14-63
-------
Lawther, P. J., A. G. F. Brooks, P. W. Lord, and R. E. Waller. Day-to-day changes in ventila-
tory function in relation to the environment. Part I. Spirometric values. Environ.
Res. 7:24-40, 1974a.

Lawther, P. J. , A. G. F. Brooks, P. W. Lord, and R. E. Waller. Day-to-day changes in ventila-
tory function in relation to the environment. Part II. Peak expiratory flow values.
Environ, Res. 7:41-53, 1974b.

Lawther, P. J., A. G. F. Brooks, P. W. Lord, and R. E. Weller. Day-to-day changes in ventila-
tory function in relation to the environment. Part III. Frequent measurements of peak
flow. Environ. Res. 8:119-130, 1974c.

Lawther, P. J. , P. W. Lord, A. G. F. Brooks, and R. E. Waller. Air pollution and pulmonary
airway resistance: A pilot study. 'Environ. Res. j5:424-435, 1973.

Lawther, P. J. , R. E. Waller, and M. Henderson. Air pollution and exacerbations of bronchi-
tis. Thorax 25:525-539, 1970.

Lebowitz, M. D. A comparative analysis of the stimulus-response relationship between morta-
lity and air pollution weather. Environ. Res. 6:106-118, 1973a.

Lebowitz, M. D. , T. Toyama, and J. McCarrol. The relationship between air pollution and wea-
ther as stimuli and daily mortality as responses in Tokyo, Japan, with comparisons with
other cities. Environ. Res. 6:327-333, 1973b.

Lebowitz, M. D. , E. J. Cassell, and J. D. McCarroll. Health and the Urban Environment. XV.
Acute Respiratory Episodes as Reactions by Sensitive Individuals to Air Pollution and
V/eather. Environ. Research 5(2):135-141, 1972.

Lebowitz, M. , P. Bendheim, G. Cristea, D. Markovitz, J, Misiaszek, M Staniec, and D. Van
Wyck. The effect of air pollution and weather on lung function in exercising children
and adolescents. Am. Rev. Respir. Dis. 109:262-273, 1974.

Lee, R. E. Jr., J. S. Caldwell, and G. B. Morgan. The evaluation of methods for measuring
suspended particulates in air. Atmos. Environ. 6:593-622, 1972.

Lepper, M. H. ,• N. Shioura, B. Carnow, S. Andelman, and L. Lehrer. Respiratory disease in an
urban environment. Arch. Indust. Med. 38:36, 1969.

Levy, D., M. Gent, and M. T. Newhouse. Relationship between acute respiratory illness and air
pollution levels in an industrial city. Am. Rev. Respir. Dis. 116-167-175, 1977.

Lindeberg, W. Correlations between air pollutant concentrations and death rates in Oslo. In:
Air Pollution in Norway. III. Oslo, Norway, Smoke Damage Council, 1968.

Linn, W. S, , J. D. Hackney, E. E. Pederseri, P. Breisacher, J. V. Patterson, C. A Mulry, and
J. F. Coyle. Respiratory function and symptoms in urban office workers in relation to
oxidant air pollution exposure. Am. Rev. Res. Dis. 114.477, 1976.

Lipfert, F, W. Sulfur oxides, particulates, and human mortality: Synopsis of statistical
correlations. J. Air Pollut. Control Assoc. 30-366-371, 1980.

Logan, W. Mortality in the London fog incident. Lancet 1:336-338, 1953.

Lowrence, W. W. Of Acceptable Risk. Science and Determination of Safety. Los Altos, William
Kaufman, 1976.
14-64
-------
Liu, B. Y. H., D. Y. H. PUT, K. L. Rubow, and G. A. Kuhlmey. Progress Report—Research on Air
Sampling Filter Media. Grant Report R804600, University of Minnesota, Minneapolis, MN,
May 1978.

Lunn, J. E. , J, Knowelden, and A. J. Handyside. Patterns of respiratory illness in Sheffield
infant schoolchildren. Br. J. Prev. Soc. Med. 21:7-16, 1967.

Lunn, J. E., J. Knowelden, and J. W. Roe. Patterns of respiratory illness in Sheffield junior
schoolchildren. A follow-up study. Br. J. Prev". S*oc. Med. 24:223-228, 1970.

Manfreda, J. , N. Nelson, and R. M. Cherniack. Prevalence of respiratory abnormalities in a
rural and an urban community. Am. Rev. Respir. Dis. 117:215-226, 1978.

Martin, A. E. Mortality and morbidity statistics and air pollution. Proc. R. Soc. Med.
57:969-975, 1964.

Martin, A. E. , and W. H. Bradley. Mortality, fog and atmospheric pollution—An investigation
during the winter of 1958-59. Mon. Bull. Minist. Health Public Health Lab. Serv.
19:56-72, 1960.

Mazumdar, S. , and N. Sussman. Relationships of air pollution to health: Results from the
Pittsburgh Study. Proceedings of 74th Annual APCA Meeting, Philadelphia, Pa. , June
21-26, 1981,

Mazumdar, S. , H. Schimmel , and I. Higgins. Daily mortality, smoke, and SOp in London,
England, 1959-1972. Proceedings of APCA Conference on the Proposed S0? and Particulate
Standard, Sept. 16-18, 1980, Pittsburgh, Pa., 1981. pp. 219-239. ^

McCarroll, J. R. , and W. H. Bradley. Excess mortality as an indicator of health effects of
air pollution. Am. J. Pub. Health 56:1933, 1966.

McCarroll, J. R., E. J. Cassell, W. T. Ingram, and D. Welter. I. Health and the Urban Envir-
onment. Am. J. Public Health 56.266-275, 1966.

McCarroll, J., E. J. Cassell, D. W. Woeter, J. D. Mountain, J. R. Diamond, and I. M. Mountain.
Health and the Urban Environment. Arch. Environ Health 14:178-1967.

McFarland, A. R. , C. A. Ortiz, and C. E Rodes. Wind tunnel evaluation of the British smoke
shade sampler, Atmos. Environ. 16:325-328, 1982.

McFarland, A. R. , and C. A. Ortiz. Aerosol characterization of ambient particulate samplers
used in environmental monitoring. Progress Report, EPA Contract 68-02-2720, Air Quality
Laboratory, Texas A & M University, College Station, Texas, 1980.

Mendelsohn, R. and G. Orcutt. An Empirical Analysis of Air Pollution Dose-Response Curves. J.
of Env. Economics and Mang. 6:85-106, 1979.

Ministry of Health. Mortality and Morbidity During the London Fog of December 1952. London,
Her Majesty's Stationery Office. 1954.

.Ministry of Pensions and National Insurance. Report on an Enquiry into the Incidence of Inca-
pacity for Work. II. Incidence of Incapacity for Work in Different Areas and Occupa-
tions. London, Her Majesty's Stationery Office, 1965.

Mitchell, C. A., R. S. F. Schilling, and A. Bouhuys. Community studies of lung disease in
Connecticut: organization and methods. Am. J. of Epi. 103:213-224, 1976.

Mork, T. A comparative study of respiratory disease in England , Wales, and Norway. Norwe-
gian University Press, Oslo, 1962.
&

14-65
-------
Morris, S. C. , M. A. Shapiro, and J. H. Waller. Adult mortality in two communities with wide-
ly different air pollution levels. Arch. Environ. Health 3_1: 248-254, 1976.

Mostardi, R. A., D. L. Ely, N. R. Woebkenberg, and M. Conlon. Air Pollution and Health
Effects in Children Residing in Akron, Ohio. Chest 80:265-275, 1981.

Hostardi, R. A. and D. Leonard. Air Pollution and Cardiopulmonary Functions. Arch. Environ.
Health 29:325-328, 1974.

Mostardi, R. A. and R. Martell. The Effects of Air Pollution on Pulmonary Functions in
Adolescents. The Ohio J. of Science. 75:65-69, 1975.

Moulds, W. Some instrumental variations arising in routine air pollution measurements. Int.
J. Air Water Pollut. 6:201-203, 1962.

Mountain, I. M.-, E. J. Cassell, D. W. Wolter, and J. D. Mountain. Health and the Urban Envir-
onment. VII. Air Pollution and Disease Symptoms in a Normal Population. Arch. Environ.
Health 17:343-352, 1968.

MAS. Proceedings of the Conference on Health Effects of Air Pollutants, prepared for the Com-
mittee on Public Works, U.S. Senate, Committee Print, Serial no. 93-15, U.S. Government
Printing Office, Washington, DC, 1978.

National Research Council. Airborne Particles. National Academy of Sciences. Washington,
DC, 1978, Chapter 9, Epidemiological Studies on the Effects of Airborne Particles on
Human Health. I. T. T. Higgins and B. G. Ferris, Jr. pp. 243-288, 1978a.

National Research Council. Sulfur oxides. National Academy of Sciences. Washington, DC,
1978, Chapter 7. Epidemiological Studies of Health Effects. F. E. Speizer and B. G.
Ferris, Jr.. pp. 180-209, 1978b.

Neri, L. C., J. S. Mandel, D. Hewitt, and D. Jurkowski. Chronic obstructive pulmonary disease
in two cities of contrasting air quality. Can, Med. Assoc. J. 113:1043-1046, 1975.

Nobuhiro, T. , M. Yozo, T. Yoshizo, K. Kiroyuri, H. Masamichi, K. Tachachiro, H. Teruo, and H.
Ken'ichi. Concerning air pollution and chronic bronchitis in Ako City. Report of the
Environment Pollution Research Institute of Hyogo Prefecture, Japan. 1:25-35, 1970.

Organization for Economic Co-operation and Development. Methods of Measuring Air Pollution.
Paris, France, 1965.

Pedace, E. A., and E. B. Sansone. The relationship between "soiling index" and suspended par-
ticulate matter concentrations. J. Air Pollut. Control Assoc. 22:348-351, 1972.

Pemberton, J., and C. Goldberg. Air polution and bronchitis. Br. Med. J. 2:557, 1954.

Petrilli, F. L. , G. Agnese, and S. Kanitz. Epidemiologic studies of air pollution effects in
Genoa, Italy. Arch. Environ. Health 12:733-740, 1966.

Prindle, R. A., G. W. Wright, R. 0. McCaldin, S. C. Marcus, T. C. Lloyd, and W. E. Bye. Com-
parison of pulmonary function and other parameters in two communities with widely differ-
ent air pollution levels. Am. J. Public Health 53:200, 1963.

Rail, 0. P. A Review of the Health Effects of Sulfur Oxides. National Institute of Environ-
mental Health Sciences, NIH, Research Triangle Park, NC, 1973, Environ. Hlth. Perspect.
8:97-121, 1974.

Ramaciotti, D., M. Bahy, B. Voinier, and P. Rey. The S0? pollution level and the incidence of
bronchitis. Medicine sociale et preventive 22:189-190, 1977.

14-66
-------
Ramsey, J. M. The relationship of urban atmospheric variables to asthmatic bronchoconstric-
tion. Bull. Environ. Contain. Toxicol. 16:107-111, 1976.

Rao, M,, P. Steiner, Q. Qazi, R. Padre, J, E. Allen, and M. Steiner Relationship of air pol-
lution to attack rate of asthma in children. J. Asthma Res 11:23, 1973.

Reichel, G. Effect of air pollution on the prevalence of respiratory diseases in West Germany.
In: Proceedings of the Second International Clean Air Congress, Washington, DC, 1970.

Rudnik, J Epidemiological Study on Long-term Effects on Health of Air Pollution. Probl
Med Wieku Rozwojowego 7a(suppl)•1-159, 1978.

Samet, J. M. , Y. Bishop, F. E. Speizer, J. D. Spengler, and B. G. Ferris, Jr. The Relation-
ship Between Air Pollution and Emergency Room Visits in an Industrial Community. JAPCA
31: 236-240, 1981.

Saric, M. , M. Fugas, and 0. Hrustic. Effects of Urban Air Pollution on School-Age Children.
Arch, of Environ. Health, 36:101-108, 1981.

Sawicki, F. Air pollution and prevalence of non-specific chronic respiratory disease In:
Ecology of Chronic Non-Specific Respiratory Diseases. Z. Brzezinski, J. Kopczynski, and
F. Sawicki. ed., Warsaw, Panstwowy Zaklad Wydawnictw Lekarskich." 1972. p. 3-13.

Sawicki, F. Chronic non-specific respiratory disease in the city of Cracow. X Statistical
analysis of air pollution by suspended particulate matter and sulfur dioxide. Epidemic!.
Rev. 23-221, 1969.

Sawicki, F. Chronic non-specific respiratory disease in the city of Cracow. XL The cross-
section study. Epidemic!. Rev. 23:242, 1969.

Sawicki, F., and P, S. Lawrence, eds. Chronic Non-specific Respiratory Disease in the City of
Cracow—Report of a 5 year Follow-up Study Among Adult Inhabitants of the City of Cracow.
National Institute of Hygiene, Warsaw, Poland, 1977.

Schimmel, H. Evidence for possible acute health effects of ambient air pollution from time
series analysis-methodology questions and some new results based on New York City daily
mortality, 1963-1976. Bull, N.Y. Acad. 54:1052-1108, 1978.

Schimmel, H. , and T. J. Murawski. S02—Harmful pollutant or air quality indicator? J Air
Pollut Control Assoc. 25:739-740, 1975.

Schimmel, H., and T. J. Murawski. The relation of air pollution to mortality. J. Occup. Med.
18:316-333, 1976.

Schimmel, N. , and L. Greenburg. A study of the relationship of pollution to mortality, New
York City, 1963-1968. J. Air Pollut. Control Assoc. 22:607-616, 1972.

Schoettlin, C, E. , and E. Landau. Air pollution and asthmatic attacks in the Los Angeles
area. Public Health Reports 76:545, 1961.

Schrenk, H. H. , H. Heimann, G. D Clayton, W. Gafafer, and H. Wexler. Air Pollution in
Donora, Pennsylvania. Epidemiology of the Unusual Smog Episode of October 1948. Public
Health Bulletin 306, U.S.G.P.O. Washington, DC, 1949

Schwing, R. C. and G. G. McDonald. Measures of Association of Some Air Pollutants, Natural
Ionizing Radiation and Cigarette Smoking with Mortality Rates. The Science of the Total
Environment 5:139-169, 1976.

Scott, J. A. Fog and deaths in London, December 1952 Pub. Health Rep. 68:474-479, 1953.

14-67
-------
Scott, J. A. The London fog of December, 1962. Med. Off. 109: 250-252, 1963.

Shy, C. M. Epidemiologic Evidence and the United States Air Quality Standards. Am. J.
Epidemic"). 110:661-671, 1979.

Shy, C. M. , J. R. Goldsmith, J. D. Hackney, M. D. Lebowitz, and 0. B. Menzel. Health Effects
of Air Pollution. American Thoracic Society, Medical Section of American Lung Associa-
tion, 1978.

Shy, C. M., V. Hasselblad, R. M. Burton, C. J. Nelson, and A. Cohen. Air Pollution Effects on
Ventilatory Function of U.S. Schoolchildren. Results of Studies in Cincinnati, Chatta-
nooga, and New York. Arch. Environ: Health 27:124-128, 1973.

Smith, T. J., and H. J. Paulus. An Epidemiology Study of Atmospheric Pollution and Bronchial
Asthma Attacks. Presented at the 64th Annual Meeting of the Air Pollution Control
Assoc. , Atlantic City, June 27-July 1, 1971. 44 pp.

Speizer, F. E. An Epidemiologies"! Appraisal of the Effects of Ambient Air on Health: Partic-
ulates and Oxides of Sulfur. J. Air Pollut. Control Assoc. 19:647-655, 1969.

Speizer, F. E. , and B. G. Ferris, Jr. Exposure to automobile exhaust. I. Prevalence of
respiratory symptoms and disease. Arch. Environ. Health 26:313, 1973a.

Speizer, F. E. , and B. G. Ferris, Jr. Exposure to automobile exhaust. II. Pulmonary function
measurements. Arch. Environ. Health ^26:319, 1973b.

Sprague, H. A., and R. M. Hagstrom. The Nashville air pollution study: Mortality multiple
regression. Arch. Environ. Health 18:503-507, 1969.

Stebbings, J. H,, Jr., and D. G. Fogleman. Identifying a Susceptible Subgroup: Effects of
the Pittsburgh Air Pollution Episode Upon Schoolchildren. Am. J. Epidemic!. 110:27-40,
1979.

Stebbings, J,, and C. Hayes. Panel Studies of acute health effects of air pollution. I.
Cardiopulmonary symptoms in adults, New York, 1971-1972, Environ. Res. 11:89-111, 1976.

Sterling, J. D. ,' J. J. Phair, S. V. Pollack, D. A. Schumsky, and I. De Grout. Urban Morbidity
and Air Pollution, A First Report. Arch. Environ. Health 13:158-1966.

Sterling, J. D., S. V. Pollack, and J. J. Phair. Urban Hospital Morbidity and Air Pollution.
A Second Report. Arch Environ Health 15:352-1967.

Stocks, P. Air Pollution and Cancer Mortality in Liverpool Hospital Region and North Walls.
Inter. J. Air Pollut. 1:1-13, 1958.

Stocks, P. Cancer and bronchitis mortality in relation to atmospheric deposit and smoke. Br.
Med. J. 1:74, 1959.

Stocks, P. On the relations between atmospheric pollution in urban and rural localities and
mortality from cancer, bronchitis and pneumonia with particular reference to 3:4-benzo-
pyrene, beryllium, molybdenum, vanadium, and arsenic. Br. J. Cancer 14:397-418, 1960a.

Stocks, P., and R. I. Davies. Epidemiologies! evidence from chemical and spectrographic anal-
yses that soil is concerned in the causation of cancer. Br. J. Cancer 14:8-22, 1960b.

Sultz, H., J. Feldman, E. Schlesinger, and W. Mosher. An effect of continued exposure to air
pollution on the incidence of chronic childhood allergic disease. Am. J. Public Health
60:891-900, 1970.
14-68
-------
Suzuki, T. , N. Ishinishi, R. Yoshida, Y, Tsunetoshi, M. Hitosugi, S. Tominaga, K. Fukutomi,
and A. Nozoe. The Relationship Between Air Pollution and the Respiratory Symptoms and
Functions of Housewives. Japan Public Health Society Foundation, Tokyo, Japan, 1978.

Thibodeau, L. A., R. B. Reed, Y. M. M. Bishop, and L. A. Kammerman. Air Pollution and Human
Health: A Review and Reanalysis. Environmental Health Perspectives 34:165-183, 1980.

Thompson, D. J. , M. 0. Lebowitz, E. J. Cassell, D.'Wolter, and J. McCarroll. Health and the
Urban Environment. VIII. Air Pollution, Weather, and the Common Cold. Am. J. Public
Health 60(4):731-739, 1970.

Toyama, T. Air pollution and its health effects in Japan. Arch. Environ. Health 8:153-173,
1964.

Toyama, T. , H. Kanyo, K. Makamura, J. Kagawa, S. Yakura, S. Adachi, N. Yamoto, F. Iriyama, F.
Kumagaya, S. Osawa, and T. Nakamura, Study on the prevalence of respiratory symptoms in
a rural area (Kashima, Ibaragi Pref ) in Japan. J. Jpn. Soc. Air Pollut. 7:24-35 (in
Japanese), 1966.

Tsunetoshi, Y. , T. Shimizu, H. Takahashi, A. Schenosowa, M Ueda, N. Nakayama, Y. Yamagata,
and A. Ohshino. Epidemiologic study of chronic bronchitis with special reference to ef-
fects of air pollution. Int. Arch, Arbeitsmed. 29:1-27, 1971.

U.S. Congress. Committee on Public Works, U.S. Government Printing Office, Washington, DC,
1968. Air Quality Criteria Staff Report, 90th Congress, 2d Session, 1968.

U.S. Department of Health, Education and Welfare. Air Quality Criteria for Sulfur Oxides.
Washington, D.C., U.S. Government Printing Office. 1970. 178 p. National Air Pollution
Control Administration Publication No. AP-50.

U.S. Department of Health, Education, and Welfare. Air Quality Criteria for Particulate Mat-
ter. Washington, D.C., U.S. Government Printing Office, 1970. 211 p. National Air Pol-
lution Control Administration Publication No. AP-49.

U.S. Environmental Protection Agency. Health Consequences of Sulfur Oxides: A Report from
CHESS, 1970-71, EPA-650/1-74-004. May 1974.

U.S. Environmental Protection Agency. Scientific and Technical Issues Relating to Sulfates.
Ad Hoc Panel of the Science Advisory Board., Washington, DC, 1975.

U.S. House of Representatives. Committee on Science and Technology. The Environmental Pro-
tection Agency's Research Program with Primary Emphasis on the Community Health and
Environmental Surveillance System (CHESS): an Investigative Report. Government Printing
Office, Washington, DC, November 1976.

U.S. Surgeon General's Advisory Comm. on Smoking and Health, 1964.

Ulmer, W. T., G. Reichel, A. Czeike, and A. Leuschner. Regional incidence of nonspecific
respiratory diseases. IV. Communication, Int. Arch. Arbeitsmed. 27:73, 1970.

Van der Lende, R. Epidemiology of Chronic Non-Specific Lung Disease (Chronic Bronchitis).
Assen, Royal Van Gorcum, and Springfield, 111., Charles C. Thomas. 1969.

Van der Lende, R. , C. Huygen, E J. Jansen-Koster, S. Knijpstra, R. Peset, B. F. Visser, E. H.
E. Wolfs, and N. G. M. Orie. A temporary decrease in ventilatory function of an urban
population during an acute increase in air pollution. Bull. Physiopathol. Respir.
11:31-43, 1975.
14-69
-------
Van der Lende, R., G. J. Temmeling, B, F. Visser, K. de Vries, J. Wever-Hess, and N. G. M.
On'e. Epidemiologies! investigations in the Netherlands into the influence of smoking
and atmospheric pollution on respiratory symptoms and lung function disturbances. Pneu-
monologie 149:119-126, 1973.

Van der Lende, R., J. P. M. de Kroon, G. J. Tammeling, B. F. Visser, K. de Vries, J. Wever-
Hess, and N. G. M. Orie, Prevalence of chronic non-specific lung disease in a non-
polluted and an air polluted area of the Netherlands. In: Ecology of Chrome Non-
Specific Respiratory Diseases. Z. Brzezinski, J. Kopczynski, and F. Sawicki, ed.,
Warsaw, Panstwowy Zaklad Wydownictw Lekarskick. 1972. p. 27-33.

Van der Lende, R., T. J, Kok, R. P. Reig, J. P. Schouten, N. G. M. Orie, and P. H. Quanjer.
The relationship between air pollution and diminished lung function values. Results of a
longitudinal epidemiological study in the Netherlands. Proceedings of 74th Annual APCA
Meeting, Philadelphia, Pa., June 21-26, 1981,

Verma, M. P., F. J. Schilling and W. H. Becker. Epidemiological Study of Illness Absences in
Relation to Air Pollution. Arch Environ Health 18:536-543, 1969.

Waller, R. E. Experiments on the calibration of smoke filters. J. Air Pollution Control
Association, 14:323-335, 1964.

Waller, R, E. , A. G. F. Brooks, and J. Cartwright. An electron microscope study of particles
in town air. Int. J. Water Pollut. 7:779-786, 1963.

Waller R. E. , A. G. F, Brooks, and H. W. Adler, Respiratory Symptoms and ventilatory capacity
in a cohort of Londoners born in 1952-53. Proceedings of the International Symposium on
recent advances in the assessment of the health effects of environmental pollution,
Paris, June 1974.

Waller, R. E, Health Effects of Exposure to Low Levels of Regulated Air Pollutants. J. Air
Pollut. Control Assoc. 28:884-887, 1978.

Waller, R. E. , and P. J. Lawther. Some observations on London fog. Br. Med. J. 1:1356-1358,
1955.

Ware, J. H., L. A. Thibodeau, F. E. Speizer, S. Colome, and B. G. Ferris, Jr. Assessment of
the Health Effects of Atmospheric Sulfur Oxides and Particulate Matter: Evidence from
Observational Studies. Environ. Health Perspect. 41: 255-276, 1981.

Warren Spring Laboratory. Accuracy and representativeness of the National Survey data. In:
National Survey of Air Pollution, 1961-1971. Volume 5. Scotland, Northern Ireland,
Accuracy of data, Index. Warren Spring Labotory, Stevenage, England, January 1975. pp.
111-119.

Warren Spring Laboratory. Measurement of Atmospheric Smoke and Sulphur Dioxide: Reproducibil-
ity of Results. RR/AP/70, Warren Spring Laboratory, Stevenage, England, August 1962.

Warren Spring Laboratory. National Survey of Smoke and Sulphur Dioxide: Instruction Manual.
Warren Spring Laboratory, Stevenage, England, 1966.

Warren Spring Laboratory. The Investigation of Atmospheric Pollution 1958-1966. Thirty-
second Report. Her Majesty's Stationary Office, London, England, 1967.

Warren Spring Laboratory. The National Survey of Air Pollution. The Use of the Daily Instru-
ment for Measuring Smoke and Sulphur Dioxide. Warren Spring Laboratory, Stevenage,
England, December 1961.
14-70
-------
Warren Spring Laboratory. The National Survey of Smoke and Sulphur Dioxide-Quality Control
Tests on Analyses of Samples, October 1975 to February 1977. Warren Spring Laboratory,
Stevenage, England, 1977.

Watanabe, H. Air pollution and its health effects in Osaka Presented at the 58th Annual
Meeting of Air Pollution Control Association, Toronto, Canada, June 20-24, 1965.

Watanabe, H. Health effects of air pollution in Osaka City. J. Osaka Life Hyg, Assoc.
10:147-157(in Japanese), 1966.

Watanabe, H. , and F. Kaneko. Excess death study of air pollution. In: Proceedings of the
Second International Clean Air Congress. H. M. Englund and W. T. Beery, ed. , Academic
Press, New York, 1971. pp. 199-200.

Wedding, J. B. , A. R. McFarland, and J. E. Cermak. Large particle collection characteristics
of ambient aerosol samplers. Environ. Sci. and Technol. 11:389-390, 1977.

Whittemore, A. S. , and E. L. Korn. Asthma and air pollution in the Los Angeles area. Am. J.
of Pub. Health 70:687-696, 1980.

WHO. Environmental Health Criteria (8): Sulfur Oxides and Suspended Particulate Matter.
World Health Organization, Geneva, 1979.

Wicken, A. J. , and S. F. Buck. Report on a study of environmental factors associated with
lung cancer and bronchitis deaths in areas of northeast England. Research Paper No. 8.
London, Tobacco Research Council. 1964

Wilkins, E. Air pollution and the London Fog of December, 1952. J. Roy. Sanit. Inst.
64:1-21, 1954a.

Wilkins, E. Air pollution aspects of the London Fog of December, 1952. Roy. Meterol. Soc. J.
80:267-271, 1954b.

Winkelstein, W. Utility or futility of ordinary mortality statistics in the study of air pol-
lution effects. In: Proceedings of the Sixth Berkeley Symposium on Mathematical Statis-
tics and Probability. L. LeCam, J. Newyman, and E. Scott, eds., University of California
Press, Berkeley, CA, 1972. pp. 539-554.

Winkelstein, W. , and M. Gay. Suspended particulate air pollution. Relationship to mortality
from cirrhosis of the liver. Arch. Environ. Health 22:174-177, 1971.

Winkelstein, W. , and S. Kantor. Stomach cancer. Arch. Environ. Health 14:544-547, 1967.

Winkelstein, W., Jr., and S. Kantor. Respiratory symptoms and air pollution in an urban popu-
lation of northeastern United States. Arch. Environ. Health 18:760, 1969.

Winkelstein, W. , S. Kantor, E. Davis, C. Maneri, and W. Mosher. The relationship of air pol-
lution and economic status to total mortality and selected respiratory system mortality
in men. I. Suspended particulates. Arch. Environ. Health 14:162-170, 1967

Winkelstein, W. , S. Kantor, E. Davis, C. Maneri, and W. Mosher. The relationship of air pol-
lution and economic status to total mortality and selected respiratory system mortality
in man. II. Oxides of Sulfur. Arch. Environ. Health. 15:401-405, 1968.

Yoshida, R., K. Motomiya, H. Saito, and S. Funabashi. Clinical and epidemiologies! studies on
childhood asthma in air polluted areas in Japan. In: Clinical Implications of Air Pol-
lution Research. Acton, Massachusetts, Publishing Sciences Group, Inc., 1976.
14-71
-------
Yoshii, M, , J. Nonoyama, H. Oshima, H. Yamagiwa, and S. Taked. Chronic pharyngitis in
air-polluted districts of Yo KKAICHI in Japan. Mie Med. J, 19:17-27, 1969.

Zapletal, A., J. Jech, T. Paul, and M. Samanek. Pulmonary function studies in children living
in an air polluted area. Am. Rev. Respir. Dis. ]J07:400-409, 1973.

Zagraniski, R. T., B. P. Leaderer, and J. A. J. Stolwijk. Ambient sulfates, photochemical
oxidants, and acute adverse health effects: an epidemiological study. Environ. Res.
19:306-320, 1979.

Zeidberg, L, D., R. A. Prindle, and E. Landau. The Nashville air pollution study. I. Sulfur
dioxide and bronchial asthma. A preliminary report. Am. Rev. Res. Dis. 84:489, 1961.

Zeidberg, L. D., R. J. M. Horton, and E. Landau. The Nashville air pollution study. Mortali-
ty and diseases of the respiratory system in relation to air pollution. Arch. Environ
Health 15:214-224, 1967.
14-72
-------
APPENDIX 14A

ANNOTATED COMMENTS ON COMMUNITY HEALTH
EPIDEMIOLOGICAL STUDIES NOT DISCUSSED IN
DETAIL IN MAIN TEXT OF CHAPTER 14
14-73
-------
APPENDIX 14A

Many community health epidemiologies! studies have been cited during the past 10 to 20
years as providing quantitative evidence for particular atmospheric levels of sulfur oxides
and/or particulate matter being associated with mortality or morbidity effects. In the course
of the present assessment, close examination of such studies and published evaluations or re-
interpretations of their findings have led to the conclusion that methodological considera-
tions or published results reported for many of them substantially limit or preclude their
usefulness in helping to define qualitative or quantitative air pollution-health effects
relationships as part of present criteria development purposes focusing mainly on effects of
S0« or PM at levels below 1000 pg/m . Based on this, many studies were excluded from detailed
discussion or consideration in the main text of Chapter 14 or mentioned only briefly in
support of certain points made in the chapter. Provided below are annotated comments addres-
sing limitations of various studies for qualitatively characterizing the health effects of PM
and SO™ or for quantifying exposure levels at which such effects occur.
A. STUDIES OF MORTALITY EFFECTS OF ACUTE EXPOSURES
1. British, European, and Japanese Studies
Numerous studies concern early (1950s-60s) severe air pollution episodes in England, when
atmospheric concentrations of particulate matter (BS) and sulfur dioxide were very markedly
elevated. These include studies by Logan (1953), Scott (1953), Ministry of Health (1954),
Wilkins (1954 a,b), Gore and Shaddick (1958), Burgess and Shaddick (1959), Clifton et al.
(1960), and Scott (1963). These studies are mainly useful in indicating mortality effects
occurring at BS or S0? levels well in excess of 1000 ug/m and are widely accepted as such,
regardless of particular methodological flaws or limitations associated with each.
Biersteker (1966) also reported on a study of a high pollution episode in Rotterdam in
December 1962. During the episode, 24-hour mean concentrations were recorded for particulate
matter and sulfur dioxide at about 500 pg/m and 1000 ug/m , respectively (OECD smoke/sulfur
dioxide methods), together with increased hospital admissions for the elderly (over 50 years
old) with cardiovascular diseases and weak indications of increased mortality. However, these
results were observed only once in Rotterdam and could have been due to other causes.
Further, they do not provide evidence of a strongly convincing statistical relationship
between observed mortality and hospital admissions and the air pollution levels reported. It
3
is not possible to determine precisely what the reported smoke levels (in (jg/m ) mean in terms
of actual particulate matter mass present in Rotterdam at the time (site-specific calibrations
for Rotterdam suggest that the actual smoke levels were about twice the level indicated by the
OECD standard curve or slOOO ug/m ). Also, the distance of air monitoring site(s) used from
study population residences or the admitting hospitals was not specified, making it impossible
to estimate how representative the aerometric data were of population exposures.
14-74
-------
A study of relationships between mortality and air pollution in Osaka Japan, was reported
by Watanabe (1966). Increases in mortality (about 20 percent) appeared to occur when daily
o
concentrations of PM (as measured by a light scattering method) exceeded 1000 ug/m (4-day
o
average) in association with SO- (probably sulfation method) levels of 250 ng/m . Low tem-
peratures may have contributed to the effects, and it is not possible to assess with
confidence the statistical relationship between observed mortality and the reported pollutant
levels which were apparently based on data from a single monitoring station. Nor is
sufficient information available by which to interpret the meaning or precision of the
reported suspended PM measurement results.
2. American Studies
A long series of studies, mainly employing regression analysis techniques, attempted to
define relationships between daily mortality and variations in particulate matter and SO, in
New York City during periods of the 1950s, 1960s, and 1970s (Greenburg et a!,, 1962; Glasser
et a!., 1967; Hodgson, 1970; Glasser and Greenburg, 1971; Schimmel and Greenburg, 1972;
Buechley et al., 1973; Schimmel and Murawski, 1976).
Among the disadvantages of these studies is the fact that only data from a single air
pollution station in Manhattan were used to correlate changes in air pollution with mortality
in the city, raising questions regarding how representative those aerometric measurements are
of exposures for the study population drawn from the entire New York City area. Goldstein and
Landowitz (1975, 1977a,b) found that most correlations between pollutant levels recorded at
any two New York City monitoring stations were <0.40 and concluded that use of aerometric data
from any single station generally did not adequately represent pollutant levels for the entire
city. A possible exception to this might occur during severe air pollutant episodes when very
markedly increased pollutant levels at all stations may tend to approach a common ceiling
elevation but no specific evidence supporting this hypothesis has been reported. Lastly, many
criticisms have been advanced regarding specifics of the statistical approach employed (e.g.,
some correlations likely to be significant by chance alone from among a very large number run)
and some of the same and other investigators have since reinterpreted these studies as gener-
ally not providing evidence of any association between mortality and SO^ levels (Buechley,
1975; Schimmel, 1978) and only very weak associations with PM levels (Schimmel, 1978).
Generally then, little or only weak evidence can be derived from these studies for excess
mortality having been associated with either elevated SCL or particulate matter levels in
New York City, with the possible exception of observations during severe pollution episodes
which suggest effects mainly on the elderly and those with chronic cardiovascular or
respiratory system diseases.
B. MORBIDITY EFFECTS AND ACUTE EXPOSURES
1- British,European, and Japanese Studies
Carne (1964) reported a study of general practitioners records of illness during
the winter of 1962-63 in London which suggests evidence of morbidity effects among elderly

14-75
-------
patients during the December, 1962, pollution episode when levels of smoke and SO, markedly
3
exceeded 1000 (jg/m . However, the health effects results cannot be reliably linked quantita-
tively to specific pollutant levels.
In another British study (Angel et al., 1965), acute respiratory illness attack and pre-
valence rates in a working population of London appeared to correlate with weekly peaks of
pollution measured at several nearby sites. No clear conclusions were stated by the authors
regarding the pollutant levels associated with notable illness increases. However, based on
the reported data only some apparent increases seemed to occur, generally when smoke and SO,
•3 f-
levels exceeded 1000 pg/m.
Emerson (1973) carried out weekly spirometry tests on 18 bronchitic patients in London
during 1969-71 and found no correlation of effects with air pollution levels. However, the
author noted that pollutant levels were generally too low to expect any notable effects, and
the small number of subjects studied would tend to preclude demonstration of statistically
significant differences.
Kevany et al, (1975) studied cardio-respiratory hospital admissions in Dublin during
1972-73 and found low but significant correlations between cardiovascular admissions and SO,
or BS levels in winters. However, the distances of the air monitors used from study
population residences or admitting hospitals were not reported, making it impossible to
determine the representativeness of the exposure estimates employed.
Derrick (1970) conducted a study of nighttime emergency room visits for asthma in
Brisbane, Australia and reported negative correlations between asthma visits with degrees of
smoke shade. However, day-of-week effects were not analyzed, and information on methods was
insufficient to determine location of air monitors in relation to study population residences
or hospitals included in the study.
Kalpazanov et al. (1976) studied daily incidence of influenza in Sofia, Bulgaria during
1972 and 1974-75 and reported the incidence to be significantly correlated with SO, and dust
levels. However, the pollutant levels were determined from a single station in Sofia; SES
status and crowding were not controlled for in the analysis,
Another report (Gervois et al., 1977), in French, gives little methodological information
regarding the pertinent aerometric or health measurements used, and it is impossible to
estimate where the air monitors were located in relation to residences of the study
populations. A positive association was reported between pollution concentration and acute
respiratory disease within one town (Quavrechoin) but not another (Denoin), although there
were no clear differences in pollutant levels between the towns.
Studies in the Netherlands reported by van der Lende and colleagues (van der Lende et
al., 1969, 1972, 1973, 1975) compared lung function in a large population group in 1969 and
again in 1972. In a more polluted area, age-, health-, and smoking-adjusted FEV-^ g values in
men increased from the first to the second survey rather than decreasing with age as expected.
This was reported to be associated with a concurrent decrease in air pollution concentrations,

14-76
-------
although no information was provided on proximity of monitoring sites to study population
residences or workplaces. The investigators considered other possible causes of the improved
pulmonary function but concluded that the most plausible was the effect of reduced air pollu-
tion. However, little hard evidence was advanced to support this hypothesis. In fact, the
changes in pulmonary function test (PFT) results observed over time are as likely due to
different experimenters' performing the PFT tests from the earlier to later years.
Van der Lende et al. (1981) recently reported further results for the cohort studied
longitudinally for more recent years, when pollutant levels continued to decrease in the Dutch
communities of Vlaardingen and Vlagtwedde. They demonstrate that after removing the effects
of age, smoking and sex, the residents of the more polluted city (Vlaardingen) exhibited a
statistically significant greater decline in VC and FEV, in the past nine years than did the
residents of the clean community (Vlagtwedde). They state that the association between this
more rapid decline in lung function capacity with age and exposure to air pollutants can be
seen when slopes of -individual regressions for VC and FEV, against increasing age are based
on several PFT measurements repeated over a long time, even when this effect is not seen in
cross sectional studies. The authors also noted that the PFT data were gathered by a large
team of trained collaborators over a short time period in the same month each time and
suggested that effects observed represent adverse health effects of PM and S02 levels present
during the nine years of the study in Vlaardingen. The annual average arithmetic means and
maxima for smoke (OECD method) during those years were reported to range from 19 to 26 ug/m
2
and 47 to 130 |jg/m , respectively; comparable annual average arithmetic means and maxima for
3 3
SOp were 76 to 106 pg/m and 136 to 564 pg/m , respectively. However, these concentrations
were recorded at the Vlaardingen town hall, an unspecified distance from the residences of
study cohort members and the smoke levels were based on the OECD standard curve rather than
site-specific calibrations in Vlaardingen per se and, therefore, cannot be used to quantita-
tively estimate effect or no-effect levels for PM.
Kagawa and Toyama (1975) and Kagawa et al, (1976) studied respiratory functions in 20
Tokyo schoolchildren and observed, after controlling for temperature, significant correlations
of acute respiratory decrements with various pollutants measured at a monitoring site on the
third-story roof of their school house. However, only 2 of the 20 children appeared to respond
to SO, changes and 1 of 20 to suspended PM changes. It is not possible to determine the
3
physical meaning of the PM measurements in terms of pg/m concentrations based on the light
scattering methodology information provided.
2. AMERICAN AND CANADIAN STUDIES
Greenburg et al. (1962, 1963, 1964) reported on studies of air pollution-morbidity
relationships in New York City. Peak pollution levels during these studies were often much
o
higher than 1000 pg/m , especially when 1-hour values are considered. Also, methods of
analyses preclude clear quantitative statements on pollutant/health effect relationships,
14-77
-------
especially since a single monitoring station was used to generate aerometric data employed in
the study analyses for all of New York City.
Glasser et al. (1967) investigated emergency room visits in seven New York City hospitals
during the November 1966 air pollution episode. They reported that increased emergency room
visits for asthma occurred in three of seven hospitals studied. The authors stated that the
results were inconclusive, and only a single monitoring site was used for aerometric data
employed in the study to estimate population exposures for the entire New York City area.
Chiaramonte et al. (1970) studied emergency room visits at a Brooklyn hospital during the
November 1966 air pollution episode. They reported a statistically significant increase in
emergency room visits for respiratory disease symptoms, continuing to 3 days after the peak
air pollution concentrations. However, three important S0? data points were missing from the
analysis.
Reports by McCarroll et al. (1966), Mountain et al. (1968), Cassell et al. (1969) and
Thompson et al. (1970) on the Cornell study in New York City gave results for correlations of
CoHs and SO,, levels with numerous health end-point measurements. Many of the effects
measured, such as tearing, are of modest health significance. Also, numerous analyses are
reported, predominantly multiple regression analysis; and coefficients frequently change sign
and magnitude as variables are added and deleted, making it difficult or impossible to quan-
tify health effects/air pollution relationships from most of these studies.
Both Burrows et al. (1968) and Carnow et al. (1969) studied chronic respiratory disease
patients in Chicago in relation to SCL and PM. Burrows et al. (1968) obtained aerometric data
from a Chicago Loop area station and compared this to symptom responses reported by a changing
subject pool and concluded that in Chicago "air contaminants do not appear to play a role in
producing these symptomatic exacerbations." But he felt guarded about this conclusion consid-
ering the several limitations noted in the study. Carnow et al. (1969) developed a personal
air pollution exposure index using a network of stations and found that positive associations
between SCL levels and health effects were restricted to seriously ill elderly patients over
age 55 with severe bronchitis. The representativeness of the exposure assessment is uncertain.
Heimann (1970) studied acute morbidity and mortality during air stagnation periods in
Boston episodes for 1965 and 1966. Respiratory patient visits to hospitals (but not mortal-
ity) were found to increase during the episodes. Even though a monitoring network was used,
the representativeness of the exposure data is difficult to determine. TSP levels were below
the 24-hour standard.
Lebowitz et al. (1974) conducted a study in an Arizona town to analyze the relationship
between air pollutants and health effects observed in exercising elementary students and other
groups. The monitoring site for the main study was a quarter of a mile from the school
exercising site in Tuscon. Appropriate pulmonary function testing of the subjects was done,
and the statistical analyses appeared to be appropriate. However, a pollution index was
14-78
-------
utilized in characterizing high and low pollution levels that included temperature, relative
humidity, TSP and sulfates, and this did not allow for any quantitative attribution of
observed effects to specific air pollutants or levels, although TSP levels ranged from 103 to
3
117 |jg/m • Also, only a limited number of data points were obtained and the exercise levels
of the subjects were not characterized.
Dodge (1980) reported the results of a cross sectional respiratory health study of over
700 school children living in or near nonurban smelter communities in Arizona. When the
subjects were grouped by area of residence, the children living in areas with relatively high
particulate levels had significantly lower pulmonary function levels than children living in
areas with low particulate levels. The two high TSP level communities had annual averages of
3 3
72 and 76 (jg/m with peak 24-hour levels of 119 and 550 |jg/m . The low TSP level communities
had annual averages ranging from 37 to 54 (jg/m and peak levels ranging from 84 to 125 ug/m .
Sulfur dioxide 24-hour peak levels ranged from 700 to 3658 |jg/m , with the highest S02 levels
occurring in the low TSP communities. A standardized questionnaire and pulmonary function
testing method were utilized. Ethic and social class analysis were conducted and the effects
of parental smoking habits were examined. However, the author noted several concerns that
include: families with members that develop respiratory problems tend to leave such
industrial communities so that the sample may be from a population with abnormally high
pulmonary function; the low TSP areas have high SCL levels, so that it might be stated that
the high SCL areas have significantly higher pulmonary function results; the study results may
not generalize to the U.S. population since the smelter and local fugitive dust may make these
study areas unique; and longitudinal studies are needed to determine the importance of the
findings. Additionally, neither an adequate discussion of aerometric methods nor a discussion
of the representativeness of the exposure estimates were provided in this report.
Ramsey (1976) tested pulmonary functions daily in seven male nonsmoking asthmatic (ages
19-21) students at the University of Dayton, Ohio, over three months. Multiple regression
analyses showed significant correlations for some tests in 5 subjects with weather variables,
but not with TSP or ozone monitored at a site on the University campus. However, no details
are provided regarding the air pollution measurement methods and the small number of students
studied tends to preclude finding of statistically significant effects unless they were rather
marked ones. Nor was information provided on any control for possible medication effects on
the study results.
Whittemore and Korn (1980) reported on newly developed statistical methods, the use of
which were illustrated by use of data collected as part of the EPA CHESS program concerning
panels of asthmatics living in the Los Angeles area. The analysis examines daily asthma
attacks among 443 asthmatics during 34 week periods during the years 1972-1975. Data for
24-hour average concentrations of TSP, respirable suspended particulates, suspended sulfates
and nitrates, S02, N02 and photochemical oxidants were utilized. Neither the quality nor
the representativeness of this aerometric data is certain. A separate multiple logistic

14-79
-------
regression was used for each panelist's asthma attack data. The presence of an attack on the
preceding day was the most significant predictor of attack. Most attacks occurred on days
with high oxidant and participate pollution, cool days, and on Saturdays (the last day of the
weekly reporting period). Because TSP, RSP, NO and SO were highly correlated, TSP was used
)\ )\
as a surrogate for this group of pollutants. The analysis did not use data on medication use
nor pollen counts. The most serious limitation is the absence of information on individual
pollutant exposure. Problems (including an inadequate exposure estimate, use of TSP as a
surrogate for pollutants, and high attack rate on last reporting day of the reporting period)
limit the usefulness of the study for present purposes; however, the statistical analysis
approach utilized (looking across time within an individual) is an important development that
may prove beneficial in future studies.
Measurements of air pollution levels for an air pollution episode in Pittsburgh, Pa.,
reported on by Stebbings et al. (1976, 1979), were compared to health endpoints measured
(mainly pulmonary function, e.g. FEV tests) after the highest levels of pollution had
occurred. Method of subject selection, lack of clear association of health results to speci-
fic participate matter (TSP) or SO, levels, and lack of baseline health data preclude quantita-
tive conclusions regarding health effects-air pollution relationships.
Becker et al. (1968) studied the relationship of air pollution changes to responses of
2,052 white collar workers on a questionnaire administered within 48 hours of the November 23-
27, 1966 episode in New York. S0?, CO, and smoke shade (CoHs) were measured at various loca-
tions with a peak for S02 of 0.55 ppm. How representative the aerometric measurements are for
the group studied is uncertain, but the greatest number of positive responses for six of eight
respiratory symptoms occurred on peak episode days.
Dockery et al. (1981) presented preliminary results of a study of pulmonary function
changes in children associated with acute exposure to TSP and SO^ during air pollution
episodes in Steubenville, Ohio. Initial results of four (when 24-hour TSP levels rose from
3
150-200 to 400 (jg/m ) study periods are suggestive of some decrements in pulmonary function
occurring in children during episodes and persisting for some time (3-4 weeks) thereafter.
However, the results thus far are somewhat ambiguous and both the study and analyses are con-
tinuing as part of the previously described "Six-City Study," which appears to have reasonably
representative exposure estimates; unfortunately, this preliminary brief report does not allow
for clear conclusions to be drawn yet.
A Canadian study (Levy et al., 1977) related hospital admissions records for acute
respiratory conditions to changes in average weekly levels of an air pollution index
(combining CoH and S0? measurements from a single monitoring site in Hamilton) and found
significant correlations after effects of temperature were taken into account. However, no
clear quantitative associations between specific CoH or S0? levels and increased acute
' respiratory conditions could be deduced from the reported results.
14-80
-------
C. MORTALITY AND CHRONIC EXPOSURES
1. British, European, and Japanese Studies
Pemberton and Goldberg (1954) studied 1950-1952 bronchitis mortality rates in men 45
years of age and older in county boroughs of England and Wales. They reported that sulfur
oxide concentrations, (sulfation rates) were consistently correlated with bronchitis death
rates in the 35 county boroughs analyzed. However, -smoktng and occupational exposures were
not controlled for in the analysis.
Gorham (1958, 1959) studied 1950-1954 deaths in 53 counties of England, Scotland, and
Wales and reported that bronchitis mortality was strongly correlated with acidity of winter
precipitation. However, no correlations with more direct measurements of SO,, or PM were
provided; nor is it possible to evaluate the quality or precision of acidic precipitation
measurements used in the study.
Studies by Stocks (1958; 1959; and 1960a,b) investigated associations in the 1950s
between standardized mortality ratios for bronchitis, lung cancer, and other cancers and
particulate matter (BS) levels in 101 urban and rural areas of Wales and England, but they
provide no way to determine clearly quantitative relationships between BS and mortality
effects. Interpretation of the BS aerometric data alone is clouded by ambiguities regarding
the actual mass of BS (in jjg/m ) indexed by measurements reported for various areas of England
•3
and Wales (i.e. were site-specific calibrations used to make the ufl/m estimates?). Also,
perhaps more importantly, differences in smoking history were not assessed as possibly
accounting for reported urban-rural differences.
Burn and Pemberton (1963) studied 1950-1959 deaths from all causes, bronchitis, and lung
cancer in three polluted areas of Sal ford, United Kingdom. The gradient of mortality from all
causes and from bronchitis and lung cancer followed the pollution gradient. However, no
direct correlation was provided specifically with measurements of SO- or PM in the study
areas, nor was smoking or SES taken into account in the analysis of study results.
Also, Buck and Brown (1964) attempted to relate standardized mortality ratios from 214
areas in the United Kingdom (1955-59) to daily smoke (BS) and SO, levels for March, 1962.
Several factors make it difficult to interpret or accept the results of this study including:
(1) pollution levels for 1962 do not provide an adequate basis for quantitatively estimating
what were probably much higher BS and SO- levels possibly contributing to mortality occurring
3
in 1955-59; (2) it is not clear what the reported 1962 BS data in (jg/m mean in terms of
actual mass indexed from the various U.K. areas, including most for which no site-specific
calibrations were carried out; (3) the 1962 BS levels calculated were likely affected to an
unknown extent by a computer error reported by Warren Spring Laboratory for British National
Survey BS data during 1961-64; and (4) potential effects of differences in occupational
exposure were not taken into account.
In another study (Wicken and Buck, 1964) of cancer and bronchitis in 12 areas of England,
no actual measurements of particulate matter (BS) or S0« were available except for two areas

14-81
-------
(North and South Easton) during 1963-1964 and an effort was made to relate these levels to
mortality during 1952-1962. Similar objections to retrospective linking of later aerometric
data to earlier mortality information apply here as were stated above for the Buck and Brown
(1964) study. Also, mortality differences for the two Easton areas may have been due as
easily to differences in study population age levels and smoking patterns as to any air
pollution gradient.
Lave and Seskin (1970) attempted to demonstrate, by mathematical analyses mainly in-
volving regression analyses, relationships between BS measurements in the United Kingdom and
bronchitis mortality data once the effects of other factors, such as smoking and socioeconomic
status (SES-) are removed. This Lave and Seskin (1970) study has been extensively critiqued in
detail by others (Holland et al., 1979) who have noted, for example, difficulties in justify-
ing inclusion of smoking, SES, and air pollution levels in the Lave and Seskin analyses as if
they were completely independent variables and the failure to make some direct allowance for
smoking habits in the actual analyses. Perhaps even more basic difficulties with the analyses
derive from: (1) the use of qualitative BS aerometric data not appropriately translated into
o
mass concentration estimates (in ug/m ) for smoke by means of site-specific calibrations of
reflectance readings against local gravimetric measurement data; and (2) ambiguities regarding
the location of sampling devices in relation to study population residences or work places
and, therefore, the representativeness of the aerometric data in estimating population PM
exposures.
Collins et al. (1971) studied death rates in children 0-14 years of ages, 1958-1964, in
relation to social and air pollution indices in 83 county boroughs of England and Wales.
Partial correlation analysis suggested that indices of domestic and industrial pollution
account for differences in mortality from bronchopneumonia and all respiratory diseases among
children 0-1 year of age. However, a visibility index was used as the basis for development
of the air pollution index and not BS or SO,, measurement data per se, making it impossible to
link the observed mortality findings quantitatively to those air pollutants.
Kevany et al. (1975) studied 1970-1973 deaths from various causes in Dublin, Ireland.
Partial correlation analysis yielded significant associations between air pollutants and some
specific causes of death. Although a multistation air sampling network was reportedly used to
derive exposure estimates, it is unclear as to whether site-specific calibrations were
employed for BS data or how representative the monitoring sites were in relation to study
population residences or workplaces.
Lindeberg (1968) studied deaths during Oslo winters in relation to air pollution indices.
Average deaths per week, in the 1958-1965 winters, were reported to be correlated with air
pollution. However, the reported information neither allows for judgments to be made regard-
ing the representativeness of the aerometric data in relation to estimating study population
exposures, nor adequately takes into account certain other factors, such as socioeconomic
status, in the data analysis.

14-82
-------
Toyama (1964) studied mortality rates in districts of Tokyo and reported that bronchitis
mortality (but not cardiovascular, pneumonia or cancer mortality) was associated with dustfall
levels. Insufficient information was provided by which to judge how representative even the
crude dustfall measurements of PM levels were in relation to study population residences or
workplaces.
Watanabe and Kaneko (1971) studied 1965-1966 mortality by cause in three areas of Osaka,
Japan. They reported a stepwise increase in mortality to be related to air pollution
independent of temperature. However, the PM measurement data were simply reported to have
been obtained for study areas from Institute of Hygiene records, and insufficient information
was provided by which to estimate the representativeness of the data in relation to study
population residences or workplaces.
2. Americanand Canadian Studies
Among published American reports on mortality associations with chronic PM and sulfur
oxides exposures are several regarding results from the Nashville, Tennessee, air pollution
study (Zeidberg et al. , 1961; 1967; Hagstrom et al. , 1967). One purpose of the Nashville
study was to study relationships between air pollution levels and mortality (total and
respiratory disease-specific) in areas of the Nashville, SMSA. Particulate matter and sulfur
oxides measurements obtained during 1958-1959 were related to deaths occurring during
1949-1960, opening this study to criticisms regarding retrospective use of later aerometric
data to look for links with earlier mortality. Also, data regarding smoking habits and
occupational exposures were not taken into account in these studies,
Lepper et al. (1969) studied 1964-1965 mortality rates in Chicago census tracts
stratified by socioeconomic class and S02 concentration. Increased respiratory disease rates
were seen in areas of intermediate and high S02 concentration, within a socioeconomic status,
without a consistent mortality gradient between the areas of intermediate and high SO-
concentration.
Another set of American mortality studies was conducted in Buffalo, N.Y., by Winkelstein
and associates (Winkelstein and Kantor, 1967; Winkelstein et al., 1967, 1968; Winkelstein and
Gay, 1971). Numerous criticisms of these studies bave been discussed by Holland et al. (1979)
and Ware et al. (1981). Among the more salient problems noted were: (1) the use of 1961-1963
particulate matter and sulfur oxides measurement data in trying to retrospectively relate air
pollution to mortality among the elderly during 1959-1961; (2) inadequate controls for possi-
ble age differences between study groups that may have covaried with the air pollution
gradient used; (3) lack of information on lifetime, including occupational, exposures; and (4)
failure to correct for smoking habits. In a later presentation, Winkelstein (1972) commented
on several of these points and attempted to correct for some of them, such as by looking at
smoking patterns among certain populations living in some of the same study areas included in
the earlier analyses. Still, this 1972 discussion does not lay to rest many of the different
major concerns regarding mortality findings reported by Winkelstein for Buffalo. For example,

14-83
-------
the finding of no significant differences in smoking patterns among study areas in the
follow-up investigation (Winkelstein, 1972) does not adequately control for possible smoking
effects among differing population cohorts included in the earlier study analyses reported by
Winkelstein and coworkers.
Jacobs and Landoc (1972) studied 1968-1970 mortality rates in industrial versus nonindus-
trial areas of Charleston, S.C. Higher total and heart disease mortality rates were found in
industrial areas, but socioeconomic differences were not adequately controlled for in the
analyses. Nor were TSP measurements used likely representative of study population exposures.
Morris et al. (1976) studied 1960-1972 mortality rates compared to 1959-1960 air pollu-
tion levels in Seward and New Florence, Pa. They reported that mortality could be related for
both smokers and nonsmokers to air pollution exposures. However, socioeconomic status and
occupational exposures were not adequately controlled for, despite a much larger percentage of
the study population being coal miners or steelworkers in New Florence, the more polluted
community.
Two further publications by Lave and Seskin (1972, 1977) attempted to extend their
original U.K. analysis approach (Lave and Seskin, 1970) to metropolitan statistical areas in
the United States. Many similar comments as indicated above for the earlier Lave and Seskin
(1970) publication apply here. Of crucial importance Is the basic difficulty encountered in
trying to determine how representative the air pollution data used in the analyses were of the
actual exposures of individuals included in their study populations. For example, in some
instances data from a single monitoring station were apparently used to estimate pollutant
exposures for study populations from surrounding large metropolitan areas. No clear informa-
tion on quantitative relationships between particulate matter or sulfur oxides levels and
mortality can be, derived from these published analyses. Similar criticisms and others noted
earlier in the main text apply to the recent Chappie and Lave (1981) report.
Thibodeau et al. (1980) do not disagree with the conclusions of Lave and Seskin regarding
mortality being associated with chronic exposure to PM, SCL, or sulfates; however, Thibodeau
et al. (1980) express reservations about the methods of estimating the magnitude of the
association and see the Lave and Seskin work as a prioritizing step leading to further
research that will more directly examine a possible cause-and-effect relationship. The
Thibodeau et al. (1980) reanalysis of U.S. mortality data demonstrated that the regression
coefficients for air pollution variables were quite unstable and should be used with care when
interpreting their meanings. Christainsen and Degen (1980) also examined the work of Lave and
Seskin and felt that the hypothesis of a constant relationship between levels of air pollution
and the explanatory variables should be rejected, which they feel weakens the validity of the
Lave and Seskin results; however, they state that their results still support the contention
that air pollution bears a significant relationship to mortality rates.
Lipfert (1980) presented three mortality analyses for: (1) air pollution episodes; (2)
time-series analysis; and (3) cross sectional analysis over a large number of U.S. cities.
)
14-84
-------
Based on those analyses, Lipfert concluded that no reliable statistical association existed
between air pollution and excessive mortality. However, the method and representativeness of
the exposure estimates are not clear and occupational exposure is not examined. Mendelsohn
and Orcutt (1979) similarly examined a large death certificate data base to explore the rela-
tionship of pollutants and mortality. However, they compared 1974 air quality data to 1970
mortality data and did not examine smoking habits, significantly weakening the meaning of any
results reported by them.
Schwin and McDonald (1976) also studied a pollutant-mortality data base and, utilizing
ridge regression and sign constrained least squares analysis, concluded that increased
concentrations of sulfur compounds are associated with a general increase in the total white
population mortality rate. A total of 23 independent variables were used, including:
socio-economic status, climate, pollution and cigarette smoking. Cigarette smoking was
indexed by state per capita sales, a very crude proxy for actual smoking data for individuals.
The method and representativeness of the exposure estimates are also not clear, lessening
confidence in the reported findings.
Mazumdar and Sussman (1981) studied mortality and morbidity in relation to air pollutants
in Pittsburgh, Pennsylvania, and reported significant associations between PM (CoHs) but not
SOg and mortality in Allegheny County. A time-series analysis of abstracted hospital
discharge records for the time period 1972-1977 was also carried out to assess morbidity
effects, with significant associations being found between both PM and SQ2 and increased rates
of respiratory, heart, and other circulatory diseases. Three monitoring stations were used to
yield exposure estimates, a clear improvement over the fewer sites used in many other studies.
Also, temperature was controlled for and a 15-day moving average was used to remove cyclic
components of periods greater than fifteen days, making for a generally credible study
overall. However, the rough terrain in Pittsburgh and the large geographical size of
Allegheny County raise concern over the representativeness of the exposure estimates based on
data from three monitoring sites. The authors also note questions concerning the relationship
between data from discharge records and the patients' hospital records which might affect the
classification according to disease type and indicate that further verification of disease
classification is being carried out now. Once accomplished, this study may provide some
useful information for criteria development purposes.
D. MORBIDITY AND CHRONIC EXPOSURES
1. British, European andJapanese Studies
Several British studies are often cited as demonstrating morbidity effects to be
associated with chronic exposures of particulate matter (BS) and sulfur oxides. For example,
Fairbairn and Reid (1958) carried out comparisons of respiratory illness among British postmen
living in areas of heavy and light pollution. Sick leave, premature retirement, and death due
to bronchitis or pneumonia were closely related to a pollution index based on visibility. The
14-85
-------
morbidity findings, however, were not analyzed in relation to more direct measures of SO- or
PM; thus, no means were provided to relate pollution indices data to population exposures to
PM or S02.
Work (1962) employed questionnaire and ventilatory function tests in studying male
transport workers 40-59 years of age in the Norwegian city of Bergen, and London, England.
Greater frequency of symptoms and lower peak flowrates were found in London. These
differences were not explained by smoking habits cr socioeconomic factors and were therefore
hypothesized to result from differences in air pollution levels. However, no direct
associations were demonstrated between morbidity effects and monitored BS and SO levels. Nor
A
could one determine the representativeness of any aerometric data used in relation to the
study populations.
Another study by Burn and Pemberton (1963) demonstrated increases in sickness absences
for bronchitis in association with increases in BS levels. However, this study failed to test
for effects of temperature decreases, which covaried with the occurrence of pollution in-
creases and did not adequately control for smoking effects.
In another study (Ministry of Pensions and National Insurance, 1965), rates of illness,
and absences for diseases such as bronchitis were related to smoke (BS) and S0? measurements
from six areas of England, Scotland and Wales in 1961-1962, yielding apparent correlations
between bronchitis and pollutant levels in some areas but not in others. However, socio-
economic and several other possible confounding factors were not taken into account. Also, BS
aerometric measurements were likely influenced by a computer error affecting 1961-1964 British
National Survey BS data, and one cannot determine what the reported BS levels actually mean in
terms of pg/m mass estimates in the absence of information on site-specific mass calibrations
for the widespread localities studied.
Douglas and Waller (1966) and Kiernan et al. (1976) reported on a long-term prospective
longitudinal study of the association of respiratory disease and symptoms in a cohort of
children in the United Kingdom in relation to air pollution. This widely cited prospective
study examined a group of over 3,000 children with adequate health effect endpoint measure-
ments and appropriate statistical analysis. However, no direct measurements of concurrent air
pollutant levels were used; rather only a very crude estimate of likely smoke or S02 levels
based on a coal consumption index was employed and retrospectively partially confirmed by
later aerometric data in some study areas. The crudeness of the exposure assessment does not
allow for quantitative determination of the levels of either PM or SQ^ (or any other
pollutant) possibly associated with the reported health effects, and even the classification
of study areas from "low" to "high" pollution areas can be seriously questioned in the
abscence of any direct confirmatory data reported for the years studied.
Holland et al. (1969a,b) and Bennet et al. (1971) studied the prevalence of respiratory
symptoms, ventilatory function, and past histories of respiratory illness in more than 10,000
schoolchildren 5-16 years of age residing in four different areas of northwest London (Kent),

14-86
-------
1964-1965. Childhood smoking habits and degree of air pollution were found to have the
greatest influence on respiratory symptom prevalence and ventilatory function. Social class,
family size, and past history of respiratory disease were also found to contribute to the
observed health effect. All factors operated independently and exerted their effects collec-
tively. However, aerometric data obtained 1-3 years after the health data were employed
retrospectively to estimate study population exposures, precluding quantitative conclusions
regarding morbidity/air pollution relationships.
Another study by Holland et al. (1969c) examined respiratory disease and pulmonary func-
tion in families of Harrow, England (a suburb of London) during 1962-1965. During this
period, mean winter smoke levels declined in ug/m from 108 (1962-1963) to 72 (1964-1965) in
two "clean" areas and from 175 (1962-1963) to 73 (1964-1965) in two "dirty" areas, but S02
levels for the same areas, respectively, were higher in 1963-1964 and 1964-1965 for the
"dirty" areas than the "clean" ones. The study was generally well conducted and controlled
for many potentially confounding factors. However, the observed differences in respiratory
symptoms may have been quantitatively related to earlier higher pollutant levels and the lack
of site-specific BS calibrations precludes any quantitative conclusions regarding effective BS
3
levels expressed in terms of ug/m mass concentrations.
Col ley and Reid (1970) also studied respiratory disease prevalence in more than 10,000
children 6-10 years old in England and Wales during 1966. A definite gradient for past
bronchitis and current cough was found from lowest rates in rural areas to highest rates in
the most heavily polluted urban areas. Differences were clearer in children of semiskilled
and unskilled workers. No effect on upper respiratory illness rates was seen. However, the
air pollution levels were qualitatively estimated from aerometric data that is not adequately
described. Other problems are also posed regarding exposure estimates, e.g., two of five
rural areas used had no air monitoring stations.
Lambert and Reid (1970) analyzed data from a respiratory symptom questionnaire mailed to
about 10,000 Britains in relation to pollutant levels of S0~ and smoke. Analysis of the
health data was appropriate and considered major confounding factors. However, study areas
for only 30 percent of the sample had smoke (BS) and S0? data derived from the National Air
Pollution Survey conducted in 1975, and the rest were based upon estimates from the Douglas
and Waller coal consumption index. In addition, the study did not adequately assess possible
confounding effects due to occupational exposures.
Gregory (1970) studied sickness absenteeism for Sheffield, United Kingdom steelworkers in
the 1950's and reported correlation of weekly absences with SO and PM levels. However,
X
smoking was not controlled for, and the author states that the number of aerometric data
points available during the winter studied were too few to allow firm conclusions to be drawn.
Waller et al. (1974) studied ventilatory function and respiratory symptom prevalence
among 18-year-olds born in London just before and after the smog episode of 1952. No
differences were found between the two groups. Both were exposed later to a high degree of

14-87
-------
pollution during the 1950s. History of lower respiratory illness in childhood did have a
major influence on later symptoms and ventilatory function, but effects were not attributable
clearly to either SOp or BS.
A study by Fletcher et al. (1976) concerned relationships between smoke (BS) levels and
morbidity effects in working men in London. However, although certain apparent relationships
were detected, the authors (Fletcher et al., 1976) noted several factors which complicate
interpretation of their findings in such terms, and it is difficult to link observed effects
to quantitative levels of BS.
Preliminary reports from a study not yet completed of children in many areas of the
United Kingdom have been presented by Irwig et al. (1975). The results reported, however, are
not based on final analyses of data and have not been subjected to peer review. Also,
aerometric data for BS measurements employed appear to be based on mass estimates derived from
calibration of reflectance readings for the British National Survey standard curve (for London
0
air in 1963) not necessarily accurately reflecting actual BS mass levels in yg/m existing in
non-London study areas included in the analyses. The study, then, cannot be expected to yield
any meaningful information even when completed .
Biersteker and van Leeuwen (1970a,b) studied respiratory symptoms and (pulmonary)
functions in schoolchildren in relation to pollutant levels in Rotterdam. The data from
approximately 1,000 children were analyzed by appropriate statistical methods. However,
insufficient details on the aerometric measurements were presented to determine the
representativeness of the exposure assessment, and socioeconomic status was not studied.
A series of studies from Poland by Sawicki (1969a,b; 1972) and (Sawicki and Lawrence,
1977) reported higher prevalence rates of chronic bronchitis in males (all smoking categories)
and females (smokers and nonsmokers but not exsmokers) in a high-pollution community.
However, many of'the reported differences by air pollution gradient disappeared when rates
were adjusted for age, sex, and smoking habits. Also, no consistent relationship was found
between the chronic bronchitis prevalence rate and length of residence in the high-pollution
community. Several reviewers (e.g., Holland et al., 1979) have taken this as evidence
indicating that Sawicki's findings do not show a relationship between air pollution and
bronchitis. In a repetition of this study in 1973, Sawicki and Lawrence (1977) found some
further evidence suggesting a relationship between the prevalence of chronic bronchitis and
air pollution levels. By 1973, annual smoke concentrations in the high-pollution area were
•J O
reported to average 190 pg/m (OECD) compared with 86 yg/m (OECD) for the low-pollution area.
However, these are essentially meaningless estimates of PM mass concentrations present, given
that the OECD standard curve was used rather than local site-specific calibrations in generat-
ing the smoke estimates. More accurate estimates of PM mass based on the latter could differ
substantially from those derived from the OECD curve. Sulfur dioxide average annual con-
centrations were 114 and 46 pg/m , respectively, for the high and low pollution areas. Both
14-88
-------
chronic bronchitis and asthma were more prevalent in the high pollution area in males and
females aged 31 to 50 and in smokers. Chronic bronchitis was also more prevalent in female
nonsmokers in the high pollution area in both 1968 and 1973, The investigators demonstrated
an interaction between air pollution and smoking. However, the authors concluded that air
pollution, in comparison to other factors (such as smoking) exerted a relatively minor effect
on the health of their study populations. Chanska (1980) dfscusses additional data from the
Sawicki study series, including methods used in the studies to take socioeconomic status into
account.
Rudnick (1978), as part of a generally well-designed and methodologically-sound study,
collected data by a self-administered questionnaire on respiratory symptoms and disease, as
well as by PEFR tests, in 3805 children 8 to 10 years old living in three communities in
Poland with differing air pollution concentrations. The questionnaire sought information on
respiratory symptoms and asthma symptoms during the previous 12 months. Mean pollutant
q
concentrations in the higher pollution area for the years 1974 and 1975 were 108 to 148 ug/m
for SO, (OECD) and 150 to 227 pg/m3 for smoke (OECD). The low-pollution areas had SO, concen-
. 3 3
trations of 42 to 67 M9/m and smoke concentrations of 53 to 82 (jg/m . However, the smoke
2
readings in (jg/m were not obtained by site-specific calibration and, therefore, cannot be
used for more than a basis for very rough qualitative comparisons of results from different
geographic areas investigated in the study. Most symptoms of respiratory illness in both boys
and girls occurred more frequently in the high-pollution area but the differences were, in
general, nonsignificant. Adjusted PEFR values were lowest (non-significantly) in children
living in Cracow, the high-pollution area, but it is unclear as to how variation due to
instrumentation drift with the Wright Peak Flow meters used was minimized. There was a higher
prevalence of breathlessness, sinusitis, and asthma attacks in boys living in the high-pollu-
tion area, but only "runny nose in the last 12 months" occurred more frequently in girls in
the same area. There were no significant differences between the frequencies of nonchronic
cough, attacks of breathlessness, shortness of breath, or multiple cases of pneumonia assoc-
iated with the different pollution levels. These results, however, again cannot be taken as
indicative of no-effect levels at the above SO, and smoke concentrations, due to ambiguities
concerning the measurement methods used (e.g., the lack of site-specific calibrations of the
smoke levels) and given that the author concluded that "the results of the main cross-
sectional study pointed to dominating effects of living in urban highly polluted environment
on the frequency of respiratory symptoms and past respiratory illnesses."
Col ley and Brasser (1980) reviewed the results from eight European studies (of which the
Rudnik study is one) conducted under the guidance of a WHO working group. Standardization of
design and aerometric measurement was sought but not totally attained. The results were not
consistent, relationships between air pollution and respiratory diseases or pulmonary function
declines being found in some countries but not in others. Nevertheless, Colley and Brasser
14-89
-------
did data analyses for results from all of the studies pooled together, despite noting numerous
caveats and concerns about the appropriateness of such pooled analyses, and concluded that
"this study has clearly shown a strong association between air pollution and various
respiratory indices in children." They stressed that because of the small number of
observations, varying pollutant levels, different prevalence levels in different countries,
and problems in pooling the data that "caution must be exercised in drawing any firm
conclusions on the precise relationship between levels of air pollution and prevalence of
respiratory indices." However, it is questionable whether any meaningful comparisons can be
made at all between the different study results in relation to PM effects due to smoke "levels
in \ng/m being" determined based on the OECD standard curve rather than site-specific
calibrations of gravimetric mass measurements versus reflectance readings in each of the
communities studied.
During the winter season of 1977-78, Saric et al. (1981) studied pulmonary function and
respiratory disease in two groups each of more than 70 second grade students, one group living
in an industrialized community with elevated air pollution and the other in a cleaner-air
community in Yugoslavia. The authors concluded that adverse effects on pulmonary function and
respiratory symptoms were associated with exposure to annual average S0«, smoke, and suspended
particulate matter (SPM) levels of 70-80 pg/m3, 60-80 M9/m3 and 130-200 Mi/m3, respectively,
with frequent exposure to high peak episodes during the heating season. Pollutants were
measured both indoor and outdoors at the schools, and activity patterns were considered in
evaluating the exposure. However, the smoke data reported were not based on site-specific
calibrations and the SPM measurement may not be directly related to TSP measurements, although
the SPM measurements were reported to have been obtained by hi-volume samplers. Additionally,
not all major pollutants (e.g., NCL) were measured. Pulmonary function tests were admin-
istered each week by the same physician. Symptoms reported on provided postcards were
followed by a visit by a nurse to collect details of the respiratory disease. SES status was
evaluated. The confounding factors of smoking parents and number of family members were found
not to affect the results. Also, no epidemic of any respiratory disease was observed during
the study period, reducing the possibility of this confounding the study. In addition,
although the respiratory function measurements were taken by the same experimenters in both
towns, "blind" procedures were not utilized that would have enhanced confidence in the
reported findings concerning the pulmonary function tests. However, the reported decrements
in pulmonary function tests are consistent with the other findings of increased incidences in
respiratory disease symptoms, the reporting of which was not likely affected by the particular
survey procedures employed. This cross sectional study does not allow us to separate the
possible long-term effects of exposure to pollutants from exposure around the day of measure-
ment. Besides the limitations mentioned above, the paper does not provide adequate informa-
tion to consider it in other than qualitative terms since one cannot determine the adequacy of
their analysis method, nor the representativeness of their exposure estimates.

14-90
-------
Petrilli et al. (1966) studied chronic respiratory illness, rhinitis, influenza, and
bronchopneumonia in several areas of Genoa, Italy, in relation to air pollution concentra-
tions, between 1954 and 1964. Respiratory illness rates in non-smoking women over age 64 with
a long residential history and no industrial exposure history were strongly correlated with
SOp concentrations. These investigators found that all illness rates were higher in
industrial districts with higher annual mean pollution concentrations. However, differences
in socioeconomic status between study areas were not adequately controlled for, and
ambiguities exist regarding methodology and interpretation of reported aerometric measure-
ments.
Watanabe (1965) studied peak flow rates in Japanese school children residing in Osaka and
found lower peak flow rates in children from more polluted communities. Improved peak flow
rates occurred when air pollution levels decreased. However, it is not clear how Watanabe
divided study areas into three levels of varying pollution. Nor was socioeconomic status
(SES) controlled for in the study.
Toyama et al. (1966) evaluated respiratory symptoms and spirometry in an agricultural
area of Japan during 1965 in subjects ages 40-65, categorized by smoking and sex. Much lower
prevalence rates and higher lung function levels were found than elsewhere in Japan based on
other study results. No direct associations were demonstrated between S0? or PM measures and
the effects observed; nor were SES factors taken into account.
Tsunetoshi (1968) conducted a bronchitis survey of seven areas of Osaka, Japan, in 1966
among 36,000 adults 40 years of age and over. Bronchitis rates, standardized for sex, age,
and smoking were greater among men and women in the more polluted areas and appeared to follow
the air pollution gradient. However, it is not clear how the air pollution gradient was
determined or how representative the aerometric data used were of study population exposures.
Fujita et al. (1969) conducted a prevalence survey (Medical Research Council question-
naire) of post office employees in Tokyo and adjacent areas in 1962 and resurveyed the same
areas in 1967. A two-fold increase over time in prevalence of cough and sputum production was
seen in the same persons, irrespective of smoking habits. This change was attributed to in-
creasing degrees of air pollution, but no direct evidence for this hypothesized relationship
was provided. Dustfall estimates for 1964 and 1967 alone were used to estimate population
exposures, but how representative these data were of study population exposures was not clear
based on the data provided.
Yoshii et al. (1969) investigated chronic pharyngitis and histopathological changes
in 6th grade children in three areas of Yokkaichi, Japan. Correlation of both effects with
sulfation rates were seen. Although 18 monitoring sites were used to obtain the sulfation
rate data, the relationship of monitoring sites to population residences is not clear. Also,
14-91
-------
of 5331 clinic visits only 287 cases were followed up and, of these, 71 were thought to be due
to air pollution because no other apparent etiology was identified.
Nobuhiro et al. (1970) studied chronic respiratory symptoms via a survey of high- and
low-exposure areas of Osaka and Ako City, Japan. Higher prevalences of chronic respiratory
symptoms were seen in more polluted areas based on SQy measured by the PbO, method and dust-
fall measured by deposit gauge at 4-6 and 6-19 sites, respectively. However, little con-
tinuous data were available during the study periods and SES differences were not controlled
for in the study.
Tsunetoshi et al. (1971) conducted a prevalence survey of chronic bronchitis in nine
areas of Osaka and Hyogo Prefecture, Japan, in residents aged 40 or older. Multiple
regression analysis indicated increasing bronchitis prevalence, adjusted for age, sex and
smoking, corresponding to the area gradient of air pollution based on dustfall and S02
measurements obtained over a 3-year period prior to obtaining the health data. Insufficient
information was provided to allow for estimation of the representativeness of the exposure
data in relation to study population exposures.
Yoshida et al. (1976) studied the prevalence of asthma in schoolchildren in areas of
Japan, Increased prevalence rates were found in areas with higher sulfation rates, but it was
unclear how well the aerometric data related to population exposures.
Suzuki et al. (1978) conducted a prevalence survey of respiratory symptoms in housewives
ages 30 or over in Japan. They found that prevalence rates correlated with various pol-
lutants, especially in older smokers. However, the study did not control for SES factors and
the authors themselves expressed concern over their poor exposure estimates based on limited,
qualitative aerometric data.
Biersteker and van Leeuwen (1970a,b) evaluated peak flow rates and respiratory symptoms
in 935 schoolchildren living in two districts of Rotterdam, one relatively affluent and having
3 3
good air quality (40 |jg of smoke per m and 120 |jg of S02 per m ) and the other less affluent
and having 50 percent higher concentrations of smoke and SO^. No significant area differences
were found in peak flow, adjusted for height and weight. Significantly more childhood
bronchitis occurred in the more polluted district, but differences were judged to be due to
poor living conditions, because the low-pollution area consisted of higher class residences.
Details of the pollutant measurements were not provided; nor could one determine how repre-
sentative they were of population exposures.
Reichel (1970) and Ulmer et al. (1970) studied respiratory morbidity prevalence via
surveys of random samples of populations in three areas of West Germany with different degrees
of air pollution. No differences in respiratory morbidity (standardized for age, sex, smoking
habits, and social conditions) were found between populations living in the different areas.
Unfortunately, no monitoring data were presented and the monitoring methods used to classify
14-92
-------
pollution areas were unspecified; therefore, no credence can be placed in the results as
presently reported by Reichel (1970) or Ulmer et al. (1970).
Zaplatel et al. (1973) studied pulmonary function in 19 schoolchildren in Most,
Czechoslovakia and reported that some children living in areas of high air pollution had func-
tional abnormalities. However, no details were givea on aerometric monitoring methods or
information by which to judge how well such data represented study population exposures to SO-
or PM.
Ramaciotti et al. (1977) studied bronchitis symptoms and peak flow rates in 1182 men in
Geneva in relation to SOp, smoking, and age in 1972-1976. Regression analyses showed an
independent effect of 862 after controlling for smoking and age. It is not clear how the air
pollution exposure indices were derived or how representative the data were of study
population exposures.
2. Americanand Canadian Studies
Numerous American and Canadian studies are often cited as showing associations between
morbidity and chronic exposures to paniculate matter or sulfur oxides. Comments on various
American studies are as follows,
Dohan and Taylor (I960), Dohan (1961), and Ipsen et al. (1969) studied industrial
absenteeism rates and dispensary visits in women RCA workers in several Philadelphia
localities and other cities in 1957-1963. Correlations with sulfation rates were explained by
temperature and season. However, the authors employed annual mean estimates of an air
pollution index developed for an entire city based on TSP, CoHs and suspended sulfates data
from an unspecified number of monitoring sites not necessarily in close proximity to either
workers' residences or work places. SES was not controlled for in the analysis but may not
have been crucial given the use of females generally engaged in similar work capacities for
RCA.
Zeidberg et al. (1961) conducted a 1-year study of 49 adults and 34 children with asthma
in Nashville, Tennessee, and reported a doubling of asthma attack rates for persons living in
more S0~-polluted neighborhoods. Monitoring data for S0~ and TSP were obtained from
monitoring sites located within one-half mile of study population residences. However, no
adjustments for demographic or social factors were made.
Cowan et al. (1963) and Smith and Paulis (1971) evaluated history of asthma and allergen
skin tests for University of Minnesota students in relation to exposure to dust from a nearby
grain elevator. A significant association was seen between asthma attacks and grain-dust PM
exposure as measured by usmoke spot" monitoring on campus, which sampled PM smaller than 5 pm
MMAD. This observed association remained significant after effects of temperature, humidity
and other possible confounding factors were taken into account in the regression analysis.
The study did not allow, however, for clear delineation of effective PM exposure levels
inducing increased asthma attacks in the presence of very low levels of SO (40 (jg/m ).
14-93
-------
Prindle et al. (1963) carried out comparisons of respiratory disease and lung function in
residents of Seward and New Florence, Pennsylvania, and found increased airway resistance in
inhabitants of the more polluted community (New Florence). However, differences in occupa-
tion, smoking, and socioeconomic level could account for these observed effects. For example,
many more of the residents of New Florence were employed as coal miners or steelworkers, but
occupation was not controlled for in the study.
Deane et al. (1965) and Holland and Reid (1965) conducted a questionnaire and ventilatory
function survey of outdoor telephone workers 40-59 years of age on the west coast of the
United States. No differences in symptom prevalence between San Francisco and Los Angeles
workers were found, although particulate concentrations were approximately twice as high in
Los Angeles. Also increased prevalence of respiratory symptoms, adjusted for smoking and age,
a larger volume of morning sputum, and a lower average ventilatory function were found in
London workers compared with American workers. However, the aerometric data used in the
studies were simply reported for American cities unspecified as to number or location of
monitoring sites; and no information was provided regarding place of residence or work for the
study population in relation to the monitors used to estimate PM exposures. In fact, some
study subjects may have lived or worked outside the limits of the city for which the monitor-
ing data was reported.
Kenline (1966) analyzed daily visits to the emergency clinic of Charity Hospital, New
Orleans, for emergency treatment of asthma in relation to pollutant data from a six-site air
sampling network, of which one station had monitoring for particles in the respirable range
and for pollen. Specified air pollutant levels, including those for TSP, showed no signifi-
cant geographic or temperature variation with respect to asthma outbreaks with the exception
of small particles detected with special sampling that appeared to be significantly positively
correlated with asthma admissions. However, few of the more obvious potentially confounding
factors were evaluated or controlled for in the analysis.
Sterling et al. (19S6, 1967) studied hospital admissions for "relevant" physiological
causes in relation to measures of various air pollutants in Los Angeles. Significant
decreases in respiratory symptom admissions were seen with decreasing S0~, though SO, was low
and varied little during the study. Other pollutants and weather may have been more
important. How representative the pollutant data were of population exposures could not be
determined based on scant information regarding monitor locations. Also, indications of
considerable "adjustment" of raw data for health and pollution indices raise concerns
regarding the validity of the reported findings.
A study in Philadelphia conducted by Girsh (196?) concluded that stagnant air conditions
seemed to correlate with peak incidences of respiratory attacks. It is not clear how
particles were measured or how representative the exposure data are for the subjects studied.
Controls for day-of-week effects and medication usage were not discussed, but the effect of
temperature was considered.

14-94
-------
Cederlof (1966) and Hrubec et al. (1973) studied chronic respiratory symptom prevalence
in large panels of twins in Sweden and in the United States. An index of air pollution was
used based on estimated residential and occupational exposures to S02, PM, and CO. Increased
prevalence of respiratory symptoms in twins related to smoking, alcohol consumption, socio-
economic characteristics, and urban residence, but not to indices of air pollution.
Ambiguities exist, however, regarding how representative the pollution index used would be of
study population exposures and there was no clear delineation of symptom prevalence
associations with SCL or PM levels specifically.
Verma et al. (1969) retrospectively studied absenteeism among New York City insurance
workers over a 3-year period. They reported respiratory disease absenteeism to be correlated
with SO, (controlled for temperature and season) but not significantly so after adjusting for
time trends. No basis was provided by which to evaluate how representative the monitoring
sites used were in relation to study population residences.
Winkelstein and Kantor (1969) conducted a survey of respiratory symptoms in a random
sample of white women in Buffalo, New York. Among nonsmokers 45 years of age and over and
smokers who did not change residence, respiratory symptoms were correlated with particulate
matter concentrations obtained in the neighborhood of residence using a 21-station monitoring
network. No association of symptom prevalence was seen with S0? concentrations. The authors
expressed concern regarding the interviewing instrument used to collect health data, noting
that it had not been standardized and had only a moderate degree of reproducibility.
Ishikawa et al. (1969) carried out comparisons of lungs obtained at autopsy from
•residents of St. Louis, Missouri and Winnipeg, Canada. Autopsy sets, matched for age, sex and
race, showed more emphysema in the more polluted city (St. Louis). However, the autopsied
groups may not accurately reflect the prevalence of disease in the general population, and
familial and genetic factors were not taken into account. Nor were any means provided by
which to credibly associate the health observations specifically with SO, or PM exposures.
Sulz et al. (1970) investigated hospitalizations for asthma (1956-1961) and for eczema
(1951-1961) in Erie County, New York. Attack rates of patients were stratified into air
pollution and social class categories. Area gradients in asthma and eczema hospitalization
rates, adjusted for social class differences, were reported to correspond to the air pollution
gradient. Meteorological differences between areas were not analyzed, and insufficient infor-
mation was provided on pollution measurement methods.
Rao (1973) studied pediatric emergency room visits for asthma at Kings County Hospital,
Brooklyn, N.Y., from October 1970 to March 1971, and reported a negative correlation of asthma
visits with smoke shade (CoHs) levels. Lack of temperature adjustments and control for other
confounding factors plus insufficient information concerning the representativeness of the
exposure assessment, however, preclude acceptance of these findings.
Goldstein and Block (1974) similarly studied emergency room visits for asthma at a
hospital in Harlem and one in Brooklyn, during September-December 1970 and September-December

14-95
-------
1971. Temperature-adjusted asthma rates were positively correlated with S0« values in
Brooklyn but not in Harlem. In the 1971 period, a 50-90 percent increase in asthma visits
occurred on the 12 days of heaviest pollution. The size of the possible subject pool varied
markedly between Brooklyn and Harlem, and the limited SOo data presented were below the
24-hour S02 standard.
Comstock et al. (1973) carried out a repeat survey in 1968-1969 of east coast telephone
workers in New York, Philadelphia, and Baltimore and of telephone workers in Tokyo. After
adjustment for age and smoking, no significant association of respiratory symptom prevalence
was found with place of residence. However, no basis was provided by which to judge whether
study population members either resided or lived within the limits of the city for which
aerometric data were reported.
Speizer and Ferris (1973a,b) also conducted comparisons of respiratory symptoms and
ventilatory function in central city and suburban Boston traffic policemen. These
investigators found only slight, insignificant increases in symptom prevalence among
nonsmokers and smokers, bqt not exsmokers, from the group in the more polluted central city
area. No group differences in ventilatory function were found, however. The authors
expressed caution regarding the findings, noting the small sample size, relatively limited ex-
posure estimate data, and possibility of self-selection factors operating among the study
population in choice of occupation and location of duty assignment.
Shy et al. (1973) examined the relationship between pulmonary function test (PFT) results
for children in areas with different levels of pollutants in Cincinnati, Ohio. Appropriate
pulmonary function data was gathered from over 300 students; and the TSP-and S02-exposure,
assessments are based upon data from a station within three blocks of the students' school.
Host major potential confounding factors were also evaluated. Since acute effects of
pollutant levels 2'4 hours prior to the PFT did not influence the performance of the children,
the slight differences over the study period are difficult to ascribe to the small differences
in then current TSP levels present at the time of the study. However, the potential
contribution of previous pollutant exposures to the reported findings was not fully examined.
Mostardi and Hartell (1975) evaluated pulmonary functions in junior high school students
from a rural school district in comparison to a more polluted urban school district adjacent
to Akron, Ohio. Pulmonary function levels were found to be lower among the urban students
than the rural study group students, adjusted for height. Similar results were obtained by
Mostardi and Leonard (1974) in a followup study of cohorts of the same students when in high
school. Problems arguing for caution in attributing the observed differences to air pollution
effects include self-selection of study populations via volunteering to participate, higher
percentages of athletic participants likely having better lung function among the rural study
population, and lack of more-than-cursory determination of SES factors. Also, the air
pollution data reported (TSP and sulfation rate levels) were obtained from air monitoring
14-96
-------
sites located at the urban high school, up to 2-3 miles distant from student residences in the
school district, and at a county site about 5 miles from the rural school complex and up to
10-13 miles from student residences in the rural school district.
Mostardi et al. (1981) similarly studied pulmonary function and respiratory disease in a
group of grade school children in Akron, Ohio. They concluded that SQ?, in conjunction with
NO,, may be responsible for the observed pulmonary function decrements and symptoms. However,
not enough information or discussion was provided to assess the adequacy of the aerometric
data used, and no aerometric data were presented in this brief article.
Linn et al. (1976) studied respiratory symptoms and function in an office working popula-
tion in Los Angeles and San Francisco in 1973, They found no significant difference in
chronic respiratory symptom prevalence between cities; women in the more polluted community
(Los Angeles) more often reported nonpersistent (< 2 years) production of cough and sputum.
However, while the authors stated that they assumed the workers studied lived within close
proximity to the single air monitor used in each city to yield aerometric data for the study,
no information was provided by which to substantiate how representative the monitoring sites
actually were in relation to population residences, and the study cannot be accepted as demon-
strating clear no-effect levels at the TSP or other pollutant concentrations reported in the
study.
A more recent study by Manfreda et al. (1978) was a qualitative study of urban-rural
differences in two communities with similar air pollution concentrations and provides no clear
quantitative information on the health effects of either S0? or PM.
Zagraniski et al. (1979) reported a cross-sectional investigation of two study groups in-
structed to report daily respiratory symptoms via postcard diaries in New Haven, Connecticut
for 10 weeks in the summer of 1976. The two groups consisted of 192 telephone employees and
82 allergy and asthma clinic patients. The symptom rates were related to daily 8- and 24-hour
ambient pollutant data recorded at two monitoring sites, 0.8 km from where the subjects were
recruited. Suspended sulfate levels and TSP concentrations present in the ambient air at the
time were found not to be statistically associated with acute adverse health effects, while
oxidants appeared to be associated. Most major confounding factors, such as smoking and
pollen levels, were examined. Several sources of bias, subject dropout, misuse of diaries,
allergy sensitivity, and lack of indoor or personal monitoring were considered by the authors
as unlikely to have biased their analysis. However, the study cannot be interpreted as demon-
strating no-effects levels for sulfates or TSP due to the small study population and con-
sequent low power for detecting small changes in health status likely to be occurring, if at
all, at the air pollution levels reported.
Johnson et al. (1981) studied pulmonary function and symptoms in relation to pollutant
levels in several Montana communities. Schoolchildren and chronic obstructive pulmonary
disease (COPD) patients were administered pulmonary function tests and standardized question-
naires. Exposure estimates were based upon two to six monitoring sites in each community of

14-97
-------
the several studied. The adequacy of control for socioeconomic status or other possible
confounding factors is not clear. The study does not show significance by some covariance or
regression analyses but does by a sign test, making it difficult to determine the meaningful-
ness of the reported results at this time.
Detels et al. (1981) studied over 8,000 Los Angeles County residents' pulmonary function
and respiratory symptoms in relation to ambient air pollutant levels. The results suggested
that the lowest pulmonary function and greater number of symptoms occur most frequently in
the area with high levels of photochemical oxidants and particles. The authors caution that
these results are based on cross sectional data and note the limitation of the exposure
assessment.
Ferris et al. (1979) reported on methodology being used in prospectively studying
cohorts of 1500 to 1800 persons 25 to 75 years of age in each of six United States communities
in regard to reported respiratory symptoms and measurements of lung function. Aerometric data
(SO,, NCL, Qg, TSP, and mass respirable particles) are being gathered at centrally located
stations and satellite stations. Additionally, some indoor/outdoor and personal monitoring
for selected pollutants is being done. Ferris et al. (1980) reported preliminary cross-
sectional data from this longitudinal study of the effects of pollutants on human health,
which indicate that differences exist between geographic sites for the incidence of various
respiratory symptoms studied and for age- and height-adjusted values for FEV. The data suggest
trends toward positive associations of the observed effects with levels of pollution
(including S0? and TSP) present in the different cities. However, Ferris et al. (1979)
caution that any difference found between the cities in the cross-sectional study of adults
might be attributable to past exposures or confounding factors such as differences in socio-
economic status. Further analysis of prospective data, still being collected, is therefore
needed before definitive statements about the effects of exposure to specific levels of
pollutants can be made based on these initial results of the Six Cities Study.
A series of studies was conducted by the U. S. Environmental Protection Agency under the
Community Health and Environmental Surveillance System (CHESS) program, an integrated set of
epidemiological studies performed between 1969 and 1975. In those studies, the health status'
of volunteer participants was either ascertained during single contacts or followed for time
periods of up to nine months. Attempts were made to coordinate these health measures with
air pollution observations from the residential neighborhoods of the study participants
in an effort to derive information on quantatitive relationships between morbidity effects
and both acute and chronic exposures to sulfur oxides, particulate matter and other air
pollutants.
The results of approximately 10 CHESS studies were published in summary or review form
in the early 1970s (Chapman et al., 1973; Hammer et al., 1976; Shy et al., 1973; French et al.,
14-98
-------
1973) and were later presented in more detail in a 1974 EPA monograph entitled Health Conse-

quences of Sulfur Oxides: A Report from CHESS 1970-1971, U.S. EPA document No

EPA-65Q/1-74-Q04 (May, 1974). The manner in which the CHESS results were reported and

interpreted in the 1974 monograph raised questions regarding inconsistencies in data

collection and analyses, as well as possible mis- or overinterpretation of results for CHESS

data sets discussed in the early published reports or the 1974 monograph.3 Of particular

concern with regard to many of the studies was the adequacy of aerometric data or other

estimates of air quality parameters, as well as the collection and analysis of health endpoint

measurement data, upon which numerous key conclusions were based regarding possible health

effects-air pollution relationships. Many questions regarding the validity of most CHESS

study findings published in the early 1970s still remain to be fully resolved, including the

accuracy of data entries onto computer data tapes utilized in generating the analyses

contained in the reports published in the early 1970s. In view of this, the potential

usefulness of CHESS studies in yielding information on quantitative relationships between

health effects and well-defined air concentrations of sulfur oxides and particulate matter is

extremely limited at this time; such CHESS Program study results are, therefore, not included

in the discussion in the main chapter text concerning morbidity effects.
The matter of misinterpretation or overinterpretation of data or results of analyses of data
collected as part of the CHESS Program contributed to considerable controversy regarding the
validity and accuracy of results of early CHESS studies, as interpreted and reported in a 1974
EPA monograph entitled Health Consequences of Sulfur Oxides: A Report from CHESS 1970-71,
U.S. EPA Document No. EPA-655071-74-004 IFlay 197457" The controversy eventually led to the
1974 "CHESS Monograph" becoming the subject of U.S. Congressional oversight hearings in 1976.
Subcommittees of the U.S. House of Representatives Committee on Science and Technology
produced a report on the Monograph, other aspects of the CHESS Program, and EPA's air
pollution research programs generally—a report entitled The Environmental Protection Agency's
Research Program with Primary Emphasis cm the Community Health and Environmental Surveillance
System (CHESS): A_n Investigative Report. Of primary importance for the present discussion,
that report, widely referred to either as the "Brown Committee Report" or the "Investigative
Report" (IR), contained various comments regarding sources of error in CHESS Program air
quality and health effects data and quality control problems associated with such data
collection and analysis. The IR also contained various recommendations to be implemented by
the Administrator of EPA pursuant to Section 10 of the Environmental Research, Development,
and Demonstration Authorization Act of 1978 ("ERDDAA," P.L. 95-155, 91 Stat. 1275, November 8,
1977). ERDDAA also requires that EPA and the Agency's Science Advisory Board report to
Congress on the implementation of the IR recommendations.
One recommendation of the IR was that an addendum to the 1974 sulfur oxides monograph be
published, to be used in part to qualify the usefulness of the CHESS studies, and to apprise
the public of the controversy surrounding CHESS. An addendum has been published, and is
available from EPA, as announced in the Federal Register of April 2, 1980, 45 F.R. 21702. The
addendum is incorporated by reference in this document in partial qualification of the CHESS
studies cited herein and is part of the public file (or docket) established for revision of
this criteria document. The Addendum contains the full text of the IR, reports to Congress by
EPA on its implementation of the IR recommendations, and a report to Congress by EPA's Science
Advisory Board on the same subject. The reader is directed to the Addendum and IR for
detailed critiques of individual CHESS studies.
14-99
-------
Several Canadian studies have investigated relationships between morbidity and chronic
exposure to PM or SO,, as described below.
Anderson and Larsen (1966) studied peak flowrates and school absence rates in children
6~7 years of age from three towns in British Columbia, Canada. Significant decreases in peak
flowrates for children in two towns were reported to be associated with kraft pulp mill emis-
sions. No effects on school absences were observed; however, ethnic differences were not
studied.
Bates et al. (1962,1966) and Bates (1967) compared symptom prevalence, work absences, and
ventilatory function in Canadian veterans residing in four Canadian cities. Lower prevalence
of symptoms and work absences and better ventilatory function were found for veterans living
in the least polluted city. However, it is not clear how representative the aerometric data
were in relation to exposures of the study populations, based on lack of information regarding
distance of the air monitoring sites from residences or work places of the subjects studied.
Bates (1973) also conducted a 10-year follow-up study of the Canadian veterans initially
evaluated in 1960 and followed at yearly intervals with pulmonary function tests and clinical
evaluations. The least decline in pulmonary function with age was seen for veterans from the
least polluted city. Again, however, it is difficult to ascertain how representative any of
the reported air pollution data were as estimates of study population exposures.
Another Canadian study (Neri et al., 1975) compared the prevalence of symptomatic disease
and the level of pulmonary function in Sudbury, a mining community, and in Ottawa. Although
chronic bronchitis was more prevalent in Sudbury men, 58 percent of Sudbury men had an occupa-
tional history suggesting high pollution exposure. Lung function levels were lower for both
men and women in Sudbury. Very high periodic peak S0? exposure levels (exceeding 1000 |jg/m )
likely account more for any pollutant effects than long-term chronic exposures to relatively
low annual average levels of S0? or annual mean particulate levels (which did not vary by much
between Sudbury and Ottawa).
Becklake et al. (1978) compared the health status of children (9-11 years old) in three
urban communities of Montreal, Canada with differing levels of air pollution. In the three
3
areas studied, ambient SO, was reported to be 15, 123, and 59 pg/m , and annual mean high-
o
volume TSP values were 84, 99, and 131 pg/m , respectively, for the low-, intermediate- and
high-pollution areas, but there was a large overlap between areas. No significant differences
were found in symptom prevalence, nor in overall pulmonary function measured by five lung
function tests after controlling for socioeconomic factors. However, the analysis showed that
one measurement, the closing volume (CC/TLC%), an index of small airway function, was con-
sistently higher in children living in the two more polluted cites (one with high SOg, the
other with high TSP). In addition, there was a trend towards reduction in peak flowrates in
the two more polluted cities compared to the third city. The authors speculated that these
subtle differences might reflect the early stages of an irreversible obstructive airway
disease, although follow-up studies would be necessary to test this.

14-100
-------
A more recent study of the three communities in the Montreal area was marked by rela-
tively small air pollution differences between cities (Aubrey et al., 1979). Cough and phlegm
were weakly associated with air pollution concentration, but lung function was not. However,
after accounting for intercity differences in age and smoking, none of the associations were
statistically significant. Little meaningful quantitative information can be extracted from
the report, and it is difficult to evaluate the representativeness of the aerometric data
reported in relation to study population exposures.
14-101
-------
APPENDIX 14B

OCCUPATIONAL HEALTH STUDIES ON PARTICIPATE
MATTER AND SULFUR OXIDES
14-102
-------
Armstrong, B, K., J. C. McNulty, L. J. Levitt, K. A. Williams, and M. S. T. Hoffs. Mortality
in Gold and Coal Miners in Western Australia with Special Reference to Lung Cancer.
Brit. J. Ind. Med. 36:199-205, 1979.

Axelson, 0., E. Dahlgren, C. D. Jansson, and S. 0. Rehnlund. Arsenic Exposure and Mortality:
A Case-Referent Study from a Swedish Copper Smelter. Brit. J. Ind. Med. 35:8-15, 1978.

BarZiv, J. and G. M. Goldberg. Simple Siliceous Pneumoconiosis in Negev Bedouins. Arch.
Environ. Health 29:121-126, 1974.

Bennett, J. G., J. A. Dick, Y. S. Kaplan, P. A. Shand, D. H. Shennan, D. J. Thomas, and J. S.
Washington. The relationship between coal rank and the prevalence of pneumoconiosis.
Brit. J. Ind. Med. 36:206-210, 1979.

Bernard, T. E., E. Kamon, and R. L. Stein. Respiratory Responses of Coal Miners for Use with
Mechanical Simulators. AIHAJ 39:425-429, 1978.

Bonnevie, A. Silicosis and Individual Susceptibility-Fact or Myth? Ann. of Occup. Hyg.
20:101-108, 1977.

Brambilla, C., J. Abraham, E. Brambilla, K. Benirschke, and C. Bloor. Comparative Pathology
of Silicate Pneumoconiosis. Amer. J. Path. 96:149-163, 1979.

Carlson, M. L., and G. R. Peterson. Mortality of California Agricultural Workers. J.O.M.
20:30-32, 1978.

Chan-Yeung, M., M. Schulzer, L. MacLean, E. Dorken, F. Tan, D. Enarson, and S. Grzybowski. A
follow-up Study of the Grain Elevator Workers in the Port of Vancouver. Amer. Rev. Resp.
Dis. 121:228, 1980. (Supplement).

Chan-Yeung, M. , M. Schulzer, L. MacLean, E. Dorken, and S. Grzybowski, Epidemiologic Health
Survey of Grain Elevator Workers in British Columbia. Amer Review Resp. Disease
121:329-338, 1980.

Corey, P. , I. Broder, and M. Hutcheon. Relationship Between Dust Exposure of Grain Elevator
Workers and Both Baseline Pulmonary Function and Acute Work-Related Changes in Status.
Amer. Rev. Resp. Dis. 121:228, 1980 (Supplement).

Corn, M. , Y. Hammad, D. Whittier, and N. Kotski. Employee Exposure to Airborne Fiber and
Total Particulate Matter in Two Mineral Wool Facilities. Environ. Res. 12:59-74, 1976.

Craighead, J. E. , and N. V. Vallyathan. Cryptic pulmonary lesions in workers occupationally
exposed to dust containing silica. JAMA 244:1939-1941, 1980,

Crosbie, W. A., R. A. F. Cox, J. V. Leblanc, and D. Cooper. Survey of Respiratory Disease in
Carbon Black Workers in the U.K. and U.S.A. Amer. Rev. Resp. Dis. 119, 1979
(Supplement).

Decoufle, P., and D. J. Wood. Mortality Patterns Among Workers in a Gray Iron Foundry. Amer.
J. Epidemic!. 109:667, 1979.

Dosman, J. A., D. J. Cotton, B. L. Graham, K. Y. Robert, F. Froh, and G. D. Barnett. Chronic
Bronchitis and Decreased Forced Expiratory Flow Rates in Lifetime Nonsmoking Grain
Workers. Amer. Review Resp. Disease 121:11-16, 1980.

Dutkiewicz, J, Exposure to Dust-Borne Bacteria in Agriculture. I. Environmental Studies.
Arch. Envr. Health. 33:250-259, 1978.
14-103
-------
Dutkiewica, J, Exposure to Dust-Borne Bacteria in Agriculture. II. Immunologies! Survey.
Arch. Envr. Health. 33:260-270, 1978.

Farant, J., and C. F. Moore. Dust Exposures in the Canadian Grain Industry. AIHAJ
39:177-194, 1978.

Federspiel, C. F., J. T. Layne, C. Auer, and J. Bruce. Lung Function Among Employees of a
Copper Mine Smelter: Lack of Effect of Chronic Sulfur Dioxide Exposure. J.O.M.
22:438-444, 1980.

Ferris, B. G., S. Puleo, and H. Y. Chen. Mortality and morbidity in a pulp and a paper
mill in the United States: A ten-year follow-up. Brit. J. Ind. Med. 36:127-134, 1979.

Gibson, E. S., R. H. Martin, and J. N. Lockington. Lung Cancer Mortality in a Steel Foundry.
J.O.M., 19:807-12, 1977.

Glover, J. R., C. Sevan, J. E. Cotes, P. C. Elwood, N. G. Hodges, R. L. Kell, C. R Lowe, M.
McDermott, and P. D. Oldham. Effects of Exposure to Slate Dust in North Wales, Brit. J.
Ind. Med. 37:152-162, 1980.

Gold, A., W. A. Burgess, and E. V. Clougherty. Exposure of Firefighters to Toxic Air
Contaminants. AIHAJ 39:534-539, 1978.

Jedrychowski, W. A Consideration of Risk Factors and Development of Chronic Bronchitis in a
Five-Year Follow-up Study of an Industrial Population. J. Epidemic, and Comm. Health
33:210-214, 1979.

Karol, M. H., H. H. loset, and Y. C. Alarie. Effect of Coal Dust Inhalation on Pulmonary
Immunologic Responses. AIHAJ 40:284-290, 1979.

Klosterkotter, W. New Aspects on Dust and Pneumoconiosis Research. AIHAJ 36:659-668, 1975.

Kung, V. A. Morphological Investigations of Fibrogenic Action of Estonian Oil Shale Dust.
Environ. Health Perspect. 30:153-156, 1979.

Lapp, N. L. , J. L. Hankinson, D. B. Burgess, and R. O'Brien. Changes in Ventilatory Function
in Coal Miners After a Work Shift. Arch. Environ. Health. 24:204-208, 1972.

Lowe, C. R., H. Campbell, and T. Kosha, Bronchitis in Two Integrated Steel Works. III.
Respiratory Symptoms and Ventilatory Capacity Related to Atmospheric Pollution. Br. J.
Industr. Med. 27:121-29, 1970.

Lowe, C. R. , P. L. Pelmear, H. Campbell, R. A. N. Hitchens, T. Khosla, and T, C. King.
Bronchitis in Two Integrated Steel Works. I. Ventilatory Capacity, Age, and Physique of
Non-Bronchitic Men. Br. J. Prev. Soc. Med., 22:1-11, 1968.

Milham, S. Mortality in Aluminum Reduction Plant Workers. J. Occup. Med., 21:475-480, 1979.

Morgan, W. K. C. Industrial Bronchitis. Brit. J. Ind. Med. 35:285-291, 1978.

Morgan,;,W. K. C. Magnetite Pneumoconiosis. J.O.M. 20:762-763, 1978.

Musk, A. W., J. M. Peters, D. H. Wegman, and L. J. Fine. Pulmonary Function in Granite Oust
Exposure: A Four-Year Follow-up. Amer. Review Resp. Disease 115:769-776, 1977.
14-104
-------
National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Beryllium. OHHS (NIOSH) 72-10268, U, S. Department of Health
and Human Services, Washington, DC, 1972.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Coke Oven Emissions. DHHS (NIOSH) 73-11016, U. S. Department of
Health and Human Services, Washington, DC, 1973*. '

National Institute for Occupational Safety and Health, Criteria for a recommended standard:
Occupational Exposure to Inorganic Mercury. DHHS (NIOSH) 73-11024, U. S. Department of
Health and Human Services, Washington, DC, 1973.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Sulfuric Acid. DHHS (NIOSH) 74-128, U. S Department of Health
and Human Services, Washington, DC, 1974.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Ammonia. DHHS (NIOSH) 74-136, U. S, Department of Health and
Human Services, Washington, DC, 1974.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Cotton Dust. DHHS (NIOSH) 75-118, U. S. Department of Health
and Human Services, Washington, DC, 1975.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Inorganic Arsenic. (Revised). DHHS (NIOSH) 75-149, U. S.
Department of Health and Human Services, Washington, DC, 1975.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Zinc Oxide. DHHS (NIOSH) 76-104, U. S. Department of Health and
Human Services, Washington, DC, 1974.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Chromium VI. DHHS (NIOSH) 76-129, U. S. Department of Health
and Human Services, Washington, DC, 1976.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Nitric Acid. DHHS (NIOSH) 76-141, U. S. Department of Health
and Human Services, Washington, DC, 1976.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Cadmium. DHHS (NIOSH) 76-192, U. S. Department of Health and
Human Services, Washington, DC, 1976.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Hydrogen Cyanide and Cyanide Salts. DHHS (NIOSH) 77-108, U. S.
Department of Health and Human Services, Washington, DC, 1977.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Organotin Compounds. DHHS (NIOSH) 77-115, U. S. Department of
Health and Human Services, Washington, DC, 1977.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Inorganic Nickel. DHHS (NIOSH) 77-164, U. S. Department of
Health and Human Services, Washington, DC, 1977.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Asbestos. DHHS (NIOSH) 77-169, U, S*. Department of Health and
Human Services, Washington, DC, 1977.

14-105
-------
National Institute for Occupational Safety and Health. Criteria for a recommended standard.
Occupational Exposure to Vanadium. DHHS (NIOSH) 77-222, U. S. Department of Health and
Human Services, Washington, DC, 1977.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Tungsten and Cemented Tungsten Carbide. DHHS (NIOSH) 77-227, U.
S. Department of Health and Human Services, Washington, DC, 1977.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Coal Tar Products. DHHS (NIOSH) 78-107, U. S. Department of
Health and Human Services, Washington, DC, 1978.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Inorganic Lead. DHHS (NIOSH) 78-158, U. S, Department of Health
and Human Services, Washington, DC, 1978.

National Institute for Occupational Safety and Health. Criteria for a recommended standard:
Occupational Exposure to Carbon Black, DHHS (NIOSH) 78-204, U. S. Department of Health
and Human Services, Washington, DC, 1978.

Rencher, A. C., M. W. Carter, and D. W. McKee. A Retrospective Epidemiological Study of
Mortality at a Large Western Copper Smelter. J.O.M. 19:754-58, 1977.

Robertson, J. McD, and T. H. Ingalls. A Mortality Study of Carbon Black Workers in the United
States from 1935 to 1974. Arch. Environ. Health 35:181-186, 1980.

Rockette, H, E. Cause Specific Mortality of Coal Miners. J.O.M. 19:795-801, 1977.

Saric, M., I. Kalacic, and A. Holetic. Follow-up of Ventilatory Lung Function in a Group of
Cement Workers, Brit. J. Inc. Med. 33:18-24, 1976.

Sherwin, R. P., M. L. Barman, and J. L. Abraham. Silicate Pneumoconiosis of Farm Workers
Lab. Invest. 40:576-582, 1979.

Sweet, p. V., W. E. Crouse, J. V. Crable, J. R. Carlberg, and W. S. Lainhart. The Relation-
ship of Total Dust, Free Silica, and Trace Metal Concentrations to the Occupational
Respiratory Disease of Bituminous Coal Miners. AIHAJ 35:479-488, 1974.

Valic, F. , D. Beritic-Stahuljak, and B. Mark. A Follow-Up Study of Functional and Radio-
logical Lung Changes in Carbon-Black Exposure. Int. Arch. Occup. , 34:51-63, 1975.

Warren, C. P. W. Lung Disease in Farmers. Cand. Med. Assn. J. 116:391-394, 1977.

Whiting, W. B, Occupational Illnesses and Injuries of California Agricultural Workers.
J.O.M. 17:177-181, 1975.
14-106
-------
APPENDIX 14C

SUMMARY OF EXAMPLES OF SOURCES AND MAGNITUDES
OF MEASUREMENT ERRORS ASSOCIATED WITH
AEROMETRIC MEASUREMENTS OF PM AND S02 USED
IN BRITISH AND AMERICAN EPIDEMIOLOGICAL STUDIES
14-107
-------
TABIE MC-1. SWHtt* OF EWIUAIIQH OF SfflWCtS, HWHTUOES, W» DIBtCTIOKM. BIASES Of KWiS
ASSOCIATED W11H BRITISH SOj HCASWEWNIS
TIM
period
Keasurtaent
•elhod
Reported joyrce
of error
Direction and Magnitude of
reported error
Likely general Ixpact on
British S02 data
Pre-1961 Lead Dioxide HtnidUy (RH)
Tenptrature (T)

Wind speed (MS)
Reaction rate Increases with RH.
Reaction ratt increases 2%
per 5* rise.
Reaction rate Increases
with MS.
-ps
1
o
co
(Overall errors)

1961-1980 Hydrogen Siting of Sample Line
(British National Peroxfde Intake:
Air Pol. Survey) a. too near boiler ehinneys
b. too near vegetation

Sample Line Adsorption:
a. Good care & cleaning
b. Average care
e. Poor care (insects, dirt)

Flow Heter Problems.
a. Daily normal conditions

b. 8-port unit with only
one weekly flow reading

Allowable Filter Cla«p
Leakage
Poor Clamp Care & Technique

Grade 8 Glassware Usage
Improper Alkalinity Buffering
CO- in Bennneralized in
SO - 100 uf/ra overestlnatlon.
SO - 70 percent underestimation.
10 pj/m underestimation, ,
20-25 pg/ii low from SO MS/IS .
Probable greater underestimation
Atmospheric Ammonia
Titration Error-
a. Normal-sharp color
change of indicator at
PH 4 5
b. Gradual color change of
indicator at pH 4.5
c. Rounding off to 0 1 ml
of alkali volume added

Evaporation of reagent:
Teiisperature and Pressure-
a Corrections - normal
b Large 6P at filter
i 3 percent variation

i 5 percent variation.

1-2 percent underestiaation

5-10 peretnt underestimation.

2-5 ug/m -underestimation
5-10 pg/m underestimation.
40 pq/« low from 50 (ig/»
monthly mean

25 pg/m underestimation on lot of
summer samples in urban areas.
<80 pg/m low 1nd. days X 40

pg/m low monthly summer mean in
country areas
i5 MG,/m error In determination
Variable positive bias, especially in swmr.
Variable positive bias, especially In sumer.

Variable positive bias, under high wind cond.
Can be up to t ISO pg/n (2o).
Occasional (prob. rare) positive bias.
Occasional (prob, rare) negative bias.

Possible general 10 iig/m negative bias.
Occasional 40-50% negative bias
Likely rare 50-90X negative bias.

Negligible inpact. Presumed £3X precision
of data
-5% negative bias on high SO.-BS days
*S% positive bias on low SO.-BS days.

Negligible impact. Presumed 12X precision
of data.
Likely occasional 5-10X negative bias.

Negligible impact ,
Occasional 5-10 (jg/n negative bias h
Occasional negative bias of up to SOX
>Z5 (ig/in neg bias on 101. of sunnier
"samples in urban areas
Occasional neg bias in country ireas-

up to 80 Mg/nr daily data i up to 100X
monthly mean in summer.
Presumed t 5 |jg/ii precision of data.
±10 iig/"3 error in determination Actual * 10 Mi/* precision level
i5 pg/m error in determination
<15 (ig/m ovtr estimation,
"especially in summer nonths
5% underestimation
10* underestimation
Added + 5 pg/m precision error

15-100X pos bias for SO. data <100 pg/n ,
7 5-15* pos bias for SO, of 100-200 M9/« 3
3 25-7 5% pos bias for SO. of 200-400 ug/m
<3 25% pos bias for S0j data >400 pg/n

General SI neg bias in SO. data
Occasional - £10% negative bias in SO. data
Data from 1965-1968 most clearly impacted

Data from 1966-1967 most clearly impacted.

cAt <50 pg/m3 uncertainty due to these two errors Is - 7 tig/n3 or 14X That 1s, 68* of the data are within 1« and 5X are >28X in error
-------
TABLE 14C-2. SUMMARY OF EVALUATION OF SOURCES, MAGNITUDES, AND DIRECTIONAL BIASES OF ERRORS
ASSOCIATED WITH BRITISH SMOKE (PARTICULATE) MEASUREMENTS

Time
period
Measurement
method
Reported source
of error
Direction and magnitude
of reported error
Likely general impact
on published BS data
1944-1950s
Pre-1961
1961-1964
1964-1980
Smoke filter
Leakage at clamp.
Weights used to make the
seal.
Highly variable under-
estimation of BS levels.
Depending upon observer
and value of R.
Comparing reflectance to
photographs of painted
standard stains.
Reflectance (R) below 25%, 50-100% underestimation.
stain too dark with use
of Clark-Owens DSIR curve.
Computer not following
proper calibration curve
Clamp correction factor
for other than 1-inch
clamp.

Flow rate - normal 1 day.
Flow by 8-port with 1
reading per week.
Variability of reading
reflectance.
Averaging reflectance
instead~of averaging
mass/cm .
Use of coarse side of
filter facing upstream
<80% underestimation at low
~ R if not corrected by WSL
(See Moulds,1961) and
discussion of clamp size
correction factor.

Uncertain; derivation
cannot be verified
Possible *20%.

+3% variation.
-10% underestimation.
+10% overestimation.

+2 units of R

Highly variable under-
estimation due to non-
linearity of R.
6-15% underestimation.
Probable widespread highly
variable negative bias.
Probable widespread relatively
small negative bias

Occasional 50-100% negative
bias in some data sets
Negligible for BS <~100 \ig/aT.
Increasing negative bias up to 80%
. as BS values increase over 100
Possible underestimate for 2-inch
and 4-inch clamps
Possible overestimate for 1/2-inch
and 10 cm clamps
Presumed ± 3% precision level.
10% negative bias on high BS days.
10% positive bias on low BS days.

Error increases with BS level fronutlOX
at 50 (jg/m up to ±20% at 400 jig/m .
Probable small negative bias at low
BS levels, could be large at high BS.

Occasional negative bias of 6-15%.
-------
TABLE 14C-2 (continued).
Time
period
Measurement
method
Reported source
of error
Direction and magnitude
of reported error
Likely general impact
on published BS data
Reading of wrong side of
stained filter.

Leakage at filter clamp
a. Normal, with good care
b. With inadequate care,
c. Careless loading where
uneven stains are
produced.

Use of wrong clamp size
a. Stain too light R>90%.

b. Stain too dark R<25%.
50-75% underestimation.
1-2% underestimation.
2-8% underestimation.
10-20% underestimation.
Highly variable over-
estimation.
Highly variable under-
estimation.
Occasional negative bias
of 50-75%.
General 1-2% negative bias.
Occasional 2-8% negative bias.
Occasional 10-20% negative bias
Data usage not recommended.

Data usage not recommended.
-------
TABLE 14C-3 SUMMARY OF EVALUATION OF SOURCES, MAGNITUDES, AND DIRECTIONAL BIASES OF ERRORS
ASSOCIATED WITH AMERICAN S02 MEASUREMENTS
Time
period
1944-1968
Measurement
method
Lead dioxide.
Reported source
of error
Humidity (RH).
Temperature (T).
Windspeed (WS).
Direction and magnitude
of reported error
Reaction
Reaction
rise.
Reaction
rate increases
rate increases
rate increases
with RH.
2% per 5°
with WS.
Likely general impact
Variable
Variable
Variable
positive
positive
positive
on
bias
bias
bias
American SO.,
, especially
, especially
, especially
data
in summer,
in summer
in summer.
1969-1975
(EPA CHESS
PROGRAM)
west-Gaeke
Pararosanaline.
Saturation of Reagent
(sulfation plate mainly)
{Overall Errors)
Spillage of reagent
during shipment.

Time delay for reagent"
SO. complex
Concentration dependence
of sampling method
Low flow correction.

Bubbler train leakage

(Overall errors).
Variable underestimation beyond
pt. where 15% of PbO, on plate
reacted
18% of total volume 50% of time,
occasional total loss

SO. losses of 1.0, 5, 25, and
75% at 20, 30, 40, and 50°C,
respectively.

Underestimation of unspecified
magnitude at daily SO,, >200
|jg/m .

±10% to 50% variable error.
Small underestimation air of
unspecified magnitude
Possible large negative bias, especially for 30-
day samples for summer monthly readings.
Generally wide ± error band associated with data.
Possible negative bias up to >100%, mainly in
summer, with 30-day reading.
Half of SO
mean of
0, data likely negatively biased by
17%, some up to 100%
Usually small (<5%) negative bias, but consistent
negative^summer bias up to 25% at 40°C temp,
extreme
Probable general negative biaSuin daily,
monthly, and yearly SO- data

Usually error of < +10%; occasionally up to
± 50% in daijy, but dampened statistically in
annual mean,

Slight negative bias suspected.
From Nov , 1970, to Dec., 1971, data biased
low by 50-100% c From Nov . 1971, to
conclusion of CHESS Program in 1975, fall-
winter data appear valid but summer data biased
low by maximum of 60-80%. From 1972 to 1975
annual average data approximately 15-20% low
Daily data highly random, not useful
?November, 1970, to April, 1973, CHESS Progran data impacted before error corrected
Applies to CHESS Program SO™ data from all years 1970-1975
cAs summarized by Congressional Investigative Report (IR, 1976).
-------
TABLE 14C-4.
SUMMARY OF EVALUATION OF SOURCES, MAGNITUDES, AND DIRECTIONAL BIASES OF ERRORS
ASSOCIATED WITH AMERICAN TOTAL SUSPENDED PARTICULATE (TSP) MEASUREMENTS
Time
period
Measurement
method
Reported source
of error
Direction and magnitude
of reported error
Likely general impact
on published TSP data
1954-1980 Stap!ex Hi Vol TSP
Time Off (Due to power
failure).
Weighing error.
Flow measurement (with
control).
Flow measurement (without
control).
a. Constant TSP—Average
of flows.
1. Low TSP level.
2. High TSP level.
b. Rising TSP-Average of
flows
c. Falling TSP-Average of
flows.
Aerosol evaporation on
standing.
Condensation of water vapor.
Foreign bodies on filter
(Insects)
Windblown dust into filter
during off-mode.

Wind speed effect on pene-
tration of dust into the
Hi-Vol shelter.
Wind direction effect due to
Hi-Vol Asymmetry.
Artifact formation, NO,
SO ~ 4
Variable underestimation.

±2% random variation,

iZ% random variation.
2% underestimation.
5-10% underestimation,
10-20% underestimation.

10-20% overestimation.

1-2% underestimation.

5% overestimation.
Generally small over-
estimation

Generally small over-
estimation.

Less penetration at high
windspeed.

Higher penetration when
normal tq sides
5-10 yg/m overestination.
Negligible impact, rare negative bias.

Negligible impact,
Negligible impact.
Negligible impact
Possible 5-10% negative bias
Possible 10-20% negative bias.

Possible 10-20% possible bias

Probable negligible impact

Possible 5% positive bias.
Possible 5% positive bias.

Occasional (rare) positive bias.

Occasional (rare) negative bias.

Probable increase in random (±) error.

Occasional positive bias.
-------
TABLE 14C-4 (continued)

Time
period
Measurement
method
Reported source
of error
Direction and magnitude
of reported error
Likely general impact
on published TSP data
1969-1975
(EPA CHESS
Program).
Fed. Reference
Method Standard
Hi-Vol Sampler
Loss of sampling material
in field.
Loss of sampling material
in mailing
Evaporation of organic sub-
stances
Windflow velocity and
asymmetry.
(Overall errors).
No specific estimate of
magnitude of error; but
would be underestimation

Reported 4-25% apparent
loss, max likely due to
crustal (sand, etc )
fall-off from selected
Utah sampling sites

No specific estimate of error
magnitude, but not likely to
exceed 5% underestimation.

No specific estimate of error
magnitude, but most likely to
increase random variation of
small underestimation.
Probable slight negative bias
in Utah winter data No known impact
on other CHESS TSP data

Probable general small <10% negative bias;
occasional 25% negative bias
Probable slight negative bias
of <5% for TSP data from urban/
industrial areas
Negligible impact or slight
negative bias.
Generally <10% negative bias,
occasional 10 to 30% negative bias.
-------
A • BRITISH STANDARD CURVE
B • DSIR INTERIM CURVE
C - DSIR-CLARK-OWENS CURVE
D -1961 TO 1964 NATIONAL SURVEY CURVE
30 40 SO 60

DARKNESS INDEX
Figure 14C-1. Comparison of smoke calibration curves for
Eel reflectometer, Whatman No. 1 paper, and a 1-in.-diameter
filter. The computer followed curve D during 1961-64 instead
of the correct curvets) 6 and C. All British epidemiology
studies using BS (jug/m3) data computed by WSL from reflec-
tance readings for United Kingdom communities during
1961-64 period are, therefore, open to criticism on the basis
of utilizing and reporting erroneous BS aerometric data. Due
to the non-linear relationships involved, no simple correction
factor can be used by which to estimate the 'correct' BS
values in ^g/m3 from values reported in such studies.
Source: Warren Spring Laboratory (1967).
14-114
-------
80
60
40
LU
o
<
a
LU
LU
cc
20
10
1956 EXPERIMENT
1963 EXPERIMENT
D S. I. R CURVE
50
100 150
SMOKE WEIGHT, uglcm2
200
250
Figure 14C-2. Relationship between reflectance (log scale) and weight of smoke per unit area.

Source: Waller (1964).
14-115
-------
APPENDIX 14D

EPA REANALYSIS OF MARTIN AND BRADLEY (1960)
DATA ON MORTALITY DURING 1958-59 LONDON WINTER
14-116
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
DATE September 25, 1981

BJECT Reanalysis of the Martin and Bradley Mortality Data

FROM Victor Hasselblad "fa
Andrew G. Stead
Biometry Division (MD-55)"

Lester Grant
Director, ECAO (MD-52)

We attempted to reanalyze the Martin and Bradley data in a manner
that would not only describe the relationship of mortality to smoke, but
would also give some estimates of the variability involved. We decided to
fit a nonlinear curve to the data, using the form

Y = a + exp(b+c logX).

This form was chosen because it allows for a lower asymptote (a), and the
curve can rise sharply if necessary. The shape of the curve is similar
to the one found by Schimmel. Since deaths are counts we tried the classical
square root transformation, and found that this did improve the fit. We
did not use the 15 day moving average, nor did we use any other seasonal
or trend adjustment. We eliminated the month of February in order to
minimize the effect of influenza.

The data along with the fitted curve are in Figure 1. The estimates
of a, b, and c were 16,73, .169, and 1.58 respectively. The estimated
asymptotic background rate is (16.73)2 or 280 deaths per day.

The 95 percent confidence intervals can be estimated using the
formula (Kendall and Stuart, Vol. 1, p. 247) for the variance of a function:
Var{g(i)) = (||)' V (||), where

V is the estimated asymptotic covariance matrix of parameter estimates, and
1SU
36
in our case.
1*3.
3a

li
3b

la
sc

These estimated 95 percent confidence limits are also in Figure 1.

All of the above calculations are based on the assumption the dose
variable is measured exactly for each person, and without error. Since
14-117
Form 1320-6 (Rev 3-76)
-------
Page 2 - Reanalysls of the Martin and Bradley Mortality Data

there 1s both, variation end measurement error 1n the dose (smoke) variable,
we attempted to get some very crude estimates of this. By making the
assumption that each of the 7 monitoring stations were located »t "randorr,"
points throughout the city of London, we calculated the variances and
standard deviations for each day. We found that the standard deviations
Increased approximately linearly with the mean, as shown 1n Figure 2.
Because of this, and because of other work with smoke data, we based our
calculations on the logarithms of the smoke data. The average variance
of the logarithms was .0802, giving a geometric standard deviation of 1.33.
This means that an approximate 95 percent confidence Interval for a single
day's average of 7 readings would be given by multiplying and dividing the
geometric mean by

exp(1.96(.OB02/7)1/2) «= 1.233.

The next step 1n the calculations 1s extremely crude. We used this
range of dose values to recalculate the 95 percent confidence Interval for
the curve in Figure 1. The lower limits came from using X/1.233 instead
of X in the formula for the confidence interval. The upper limits came
from using 1.233 X 1n the formula. The resulting broader band about the
curve is shown in Figure 3.

Several important caveats must be mentioned. All of the calculations
are based on approximations and assumptions of unknown accuracy. Thus the
resulting curve may be little better than a best guess. Many contributing
factors should be considered and additional analysis techniques should be
investigated before these calculations could be considered scientifically
sound. The calculations represent the best estimate we can make in a very
limited amount of time.
14-118
-------
550
500
500
1,000

BRITISH SMOKE,
1,500
2,000
Figure 14D-1 Fitted curve and 95 percent confidence interval for data of Martin and Bradley (vertical axis
changed), assuming error in mortality but not in smoke data.
-------
ISS
O
250
500
750 1000 1250

MEAN SMOKE, mg/100m3
1500
1750
2000
Figure 14D-2 Relationship of standard deviation to mean for smoke data of Martin and Bradley.
-------
I
(_l
l\s
550
500
450
400
< 350
D
300
250
200
150
BOO
1,000

BRITISH SMOKE,pg/m3
1,500
2,000
Figure 14D-3 Fitted curve ( ) depicting dose-response relationship between mortality (number of daily deaths)
and atmospheric particulate matter concentrations (expressed in jug/nv* nominal British Smoke) during London win-
ter of 1958/59, as determined by EPA reanalysis of Martin and Bradley (1960) data. Also shown are: 95% confidence
intervals (• ) assuming error in mortality data but not smoke data; and 95% condidence intervals (——) taking
into account variation or error in both mortality and smoke data.
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

DATE October 2, 1981

Continued Reanalysis of the Martin and Bradley Mortality Data

Victor Hasselblad . xi
Andrew G. Stead *^*&'-*-\j
Biometry Division (MD-55)

TO Lester Grant, Director
Environmental Criteria and Assessment Office (MD-52)

We performed the same analyses on the S02 data of Hartin and
Bradley as we had with the smoke data. As with the smoke data, we found
that the log transformation of the S02 levels appeared to be the most
appropriate. The average variance of the logarithms of the SOa data was
.1725. Thus the 95tpercent confidence Intervals for SQ2 city wide means
are given by multiplying ar>d dividing the geometric means by:

exp(1.96(.1725/7)1/2) » 1.360

The graphs for the fitted mortality data are in figures 1, 1A, 3, and
3A. The model in this case became

(Y)1/2 « 16.94 + exp(1.73 + 1.47 log(S02))

We would like to again state that these calculations are crude.
Many factors have not been considered. For example, we did not use 15-
day moving averages as did riartin and Bradley. We do believe that
figures give some feeling for the inherent variability in the data.

One minor correction should be made in the smoke graphs. The
curves should not be extended beyond the range of the data.
14-122
Ff. Fetm 1320.6 (Re* 3.76!
-------
450
400
350
UJ
Q
300
250
200
* • •'
•
•
* * *
9 9 « *
* * %J*
V •-...._. i.t.»r-
*•" ". *• v ^
. .
9 »
* *
0 100 200 300 400 500 600 700 800 900 1000

" SULFUR DIOXIDE, uglm3

Figure 14D-4. Fitted curve and 95 percent confidence interval for Martin and Bradley
data, assuming error in mortality but not in sulfur dioxide data,
-------
0.20
^
I
h—
f\>
q.15 —
o

LU
O
Q
Ct
<
O
z
<

tn
010 —
005 —
1,000

SULFUR DIOXIDE, (jg/m3
1,500
2,000
Figure 14D-5. Relationship of standard deviation to mean for SO2data of Martin and Bradley.
-------
450
400
350
I
Ul
O
•f*
i
300
250
200
*

100
200
300
400
500
600
700
800
900
1000
Figure 14D-6 Fitted curve { ) depicting dose-response relationship between mortality (number of daily deaths)
and atmospheric sulfur dioxide concentrations (expressed in pg/m^} during London winter of 1958/59, as determined
by EPA reanalysis of Martin and Bradley (1960) data. Also shown are: 95% confidence intervals ( ) assuming
error in mortality but not SO2 data; and 95% confidence intervals ( ) taking into account variation or error in
both mortality and SO2 data
-------
APPENDIX 14E

SUMMARY OF UNPUBLISHED DAWSON AND BROWN (1981)
REANALYSIS OF MARTIN AND BRADLEY ,(1960) DATA
14-126
-------
A recently completed, unpublished reanalysis by Dawson and Brown of Martin and Bradley
(1960) London mortality data for winter 1958-59 was presented at the November 15-17, 1981,
meeting of EPA's Clean Air Scientific Advisory Committee (CASAC) as part of public participa-
tion and input concerning CASAC review of the penultimate draft version of the present cri-
teria document. The analysis, prepared by Stanley V. Dawson, Sc.D., and Scott Brown, M.S., on
behalf of the California Air Resources Board, helps to elucidate certain important points for
present criteria development purposes and a summary of its key findings is included here in
the Appendices of Chapter 14 for informational purposes.
The main purpose of the Dawson and Brown (1981) analysis was to reanalyze the Martin and
Bradley (1960) data to evaluate possible significant relationships between London daily
mortality rates and British Smokeshade (BS) measurements during a non-episodic period of four
months in 1958-59, with additional emphasis on assessment of associations between mortality
o
and 24-hour BS levels below 500 jjg/m . Three variants of same-day models were examined:
straight line, piecewise-linear, and a model log-linear for pollutants; in addition, time-
series models were developed. Comparative fits of the models were principally determined by
relative R values, and mean squared error (MSE) was also included as a measure of error
variance. The Durbin-Watson (DW) statistic was used to assess auto correlation of the error
terms to determine the need for a time-series model. The mortality data were also detrended
for seasonal effects, and regression analyses carried out together with determination of 95
percent confidence bands for the regression lines.
Graphs of total deaths, detrended total deaths (XTD), bronchitis deaths, detrended bron-
chitis deaths (XBD) and British Smokeshade (black suspended matter; BSM) levels are shown in
Figure 1; note association between peaks in BSM and peaks in total death rates and bronchitis
death rates. The high correlation between BSM and S0? levels in the data were found to pre-
clude isolation of individual effects on mortality and, so, BSM alone was used to represent
total pollution burden for a given day. Linear regression analyses were performed on de-
trended mortality and all levels of BSM, with the results confirming a highly significant
(P<.0001) association between BSM and total mortality (XTD) and a smaller, but still highly
significant association between BSM and bronchitis deaths (XBD). Also, the t-values were so
large for the regressions XTD and XDB that a small amount of autocorrelation in errors
detected by the Durbin-Watson statistic did not render the t-values insignificant. Piecewise
linear models of the threshold type (hockey sticks) and of the saturation type were fitted to
the data. For XTD and XBD, none of the threshold-type models gave a better fit (in terms of
2
improved R -value), whereas saturation models produced improvement in fit over simple linear
regressions for both XTD and XBD. Employing a logarithmic transformation of BSM, used by
14-127
-------
Martin and Bradley (1960), and a log-linear regression model Improved the fit even further
over the linear model (Figure 2, 3, and 6), although one model could not be rejected for the
2
other due to the closeness of the R values for both models. Time-series models, aimed at
resolving problems due to autocorrelation of the data, were found to be the best of the models
considered, but coefficients of BSM and lagged BSM for such models were close to corresponding
values in the linear and log-linear models without consideration of lags. Similar models were
employed to evaluate the same data set, but excluding days when BSM levels were > 500 (jg/m .
Estimates obtained were remarkably similar to those found for the whole range of pollutant
o
levels, and models with BSM restricted to <500 ug/m were still statistically significant
(P<.05). Group means and confidence intervals are plotted in Figure 6, along with the simple
linear regression line and its 95 percent confidence interval; similar means and 95 percent
confidence bounds are depicted in Figure 7 for the log-linear regression plot.
The present analyses were interpreted by Dawson and Brown as being indicative of strong,
statistically significant positive relationships between BSM and mortality across the whole
range of BSM measurements included in the Martin and Bradley (1960) data set. Thus, although
not as significant statistically due to lower signal to noise ratios at lower BSM values,
there appears to be strong indications that the relationship between BSM and mortality holds
Q
for values of BSM substantially below 500 ug/m and probably down to values at least as low as
200 ug/m . Dawson and Brown noted further that, based on the log-linear model results
(providing a better fit than the simple linear model), mortality appears to increase more
rapidly at lower levels of pollution for the same absolute incremental increase in BSM values.
That is, an increase of 100 ug/m from 200 ug/m BSM appears to be more potent in causing
further mortality than an increase of 100 ug/m from 500 ug/m , and so on. It was noted that
this deminishing effect of an absolute increase is common in toxicological experience,
offering motivation for customary lo'garithmic transformtions of dose.
14-128
-------
FIGURE 1
IS3
NOV. NOV. NOV
1 10 20
400-J '
NOV.
30
J.
DEC. DEC. DEC.
10 20 30
J l__ I
JAN.
9
i_
JAN.
19
J_
JAN.
29
400
BRONCHITIS DEATHS
-------
FIGURE 2
CO
O
x
UI
Q
H
O
H
O.
O
z
O
NHI

§
>
u
Q
Q

X
80 -i
60-
40-
20-
0
-20-
-40-
-60-
ESTIMATED LINEAR
ESTIMATED LOG-LINEAR \
. REGRESSION LINE
100 300 500 700 900 1100 1300 1500
1700 1900
BSM - BLACK SMOKE (ug/m1)
-------
FIGURE 3
W
0
H
I
U
2
O
QC
CO
U.
O
z
O
^x
H
LLl
Q
Q
00
X
20-i
15-
10
5-
0-
-5-
-10-
-15-
ESTIMATED LINEAR
ESTIMATED LOG-LINEAR
REGRESSION LINE
Q
Z
<
to
Ch
0
TiriiiiiiiiitiiTiTi
200 400 600 800 1000 1200 1400 1600 1800
BSM-BLACK SMOKE (ug/m*)
-------
FIGURE 4
LEGEND: A = 1 OBS, B = 2 OBS, C = 3 OBS
XTDt.2 + 25,39 log BSMt
Error from XTDt= -155.69 + .074 XTDt_j
60 -|

30-

-30-

A
AA
A
A A A A A
A A A A
A AA A A
A A A A A BA
A A A C 8 A
A BA B A
A A AA AA A A
AA A
i i i i
4.8 5.2 5.6 6.0
BSM =500

A
A A
A A A A
& AAB A
A A A AA
A A AAA
A A A
AA AA AA
A AA
A
1 1 1 1
6.4 6.8 7.2 7.6
loge BSM
FIGURE 5
LEGEND: A = 1 OBS, B •= 2 OBS, C = 3 OBS
Error from XBDf =-30.24 •»• .200 XBD^j + 4.933 log BSMt
20-i

-20-

-40-

A
A
A A A AA A AA 8 CAB
A A BA AAA 8 8A8 B
A 8A AB AB AAA
A 8 A
A

1 I I I
4.8 5.2 5.6 6.0

AA A
A A A
A A AAA
B AAAA B AB A A
AA A B A A A
A A A

BSM = 500

i i i i
6.4 6.8 7.2 7.6
loge BSM
14-132
-------
FIGURE 6
to
U)
U
Q
u-
O

O
MM
H
Q
H
X
80-i
60
40
20
-40-
-60
95% SIMULTANEOUS CONFIDENCE
INTERVAL FOR GROUPED MEANS
( )= # OF OBSERVATIONS
(19)
600-699
500-599
TOO-799
200-299
100-199
: (5J
1200+

9556 CONFIDENCE BAND
FOR REGRESSION LINE
200
400 600
800
T r \ r r i T i i i
1000 1200 1400 1600 1800
BSM - BLACK SMOKE (ug/ms)
-------
FIGURE 7
c*
X
H
<
01
O
uu
O

z
O
Q
H
X
80-i
40-
20-
0
-20
-40
•60
95% SIMULTANEOUS CONFIDENCE
INTERVAL FOR GROUPED MEANS

( )« * OF OBSERVATIONS
95X CONFIDENCE BAND
FOR REGRESSION LINE
1
200
I I i I
400
600
I I I i I i I ' I I I I i
800 1000 1200 1400 1600 1800
BSM-BLACK SMOKE (ug/ms)
-------
APPENDIX 14F

SUMMARY OF UNPUBLISHED ROTH ET AL. (1981)
YEAR-BY-YEAR ANALYSIS OF LONDON MORTALITY DATA FOR WINTERS
OF 1958-59 TO 1971-72
14-135
-------
A recently completed, unpublished analysis by Roth (1981) of London mortality data in
relation to British Smokeshade (BS) and SCU levels during the winters of 1958-59 to 1971-72
(using the same data set evaluated by Mazumdar et al., 1981) was presented at the November
15-17, 1981, meeting of EPA's Clean Air Scientific Advisory Committee (CASAC) as part of
public participation and input concerning CASAC review of the penultimate draft of the present
criteria document. The analysis, prepared by H. Daniel Roth, Ph.D. (of Roth Associates, Inc.)
on behalf of the Utilities Air Regulatory Group (UARG), helps to elucidate certain important
points for present criteria development purposes and a summary of key findings of the Roth
analysis is included here in the Appendices of Chapter 14 for informational purposes.
The main purpose of the Roth analysis was to evaluate, on a year-by-year basis, relation-
ships between daily mortality and BS or S0? levels occurring in London during the winters of
1958-59 to 1971-72. Multiple regression analyses were employed, taking into account temper-
ature, humidity, high correlations between BS and SCL levels, and certain temporal trends in
the data. The results of the analysis are concisely summarized in Table 14F-1. Statistically
significant (P < .05) associations were found between mortality derivations and either temper-
ature or humidity or both for several of the years evaluated by Roth. Based on this, Roth
indicated that reanalyses (discussed earlier in Chapter 14 or Appendix 14D) of Martin and
Bradley (1960) London mortality data by Ware et al. (1980) or by Hasselblad and Stead (see
Appendix 140) cannot be accepted as demonstrating quantitative exposure-response relationships
between deviations in London 1958-59 winter mortality rates and concurrent BS or SO, levels,
because the two reanalyses did not take into account the effects of temperature or humidity
(shown by Roth to be potentially important confounding variables) or other potentially impor-
tant factors (e.g. lag effects).
It should also be noted, however, that Roth's results also equally suggest that signifi-
cant associations may exist between mortality deviations in London and concurrent levels of BS
or SOo during various winters from 1958-59 to 1971-72. For example, significant positive
associations between mortality and SO, levels were found by Roth for the 1958-59 and 1962-63
winters, suggesting that positive mortality-SQ, associations detected in analyses of the
Martin and Bradley (1960) data set for the 1958-59 winter occurred, although they may be
somewhat unusual in comparison to other London winters. Roth's analysis also suggests that
statistically significant positive associations may exist between daily mortality and BS
levels for several additional winters, including the 1967-68 winter when 24-hour BS levels
rarely (if ever) exceeded 500 \.ig/m . EPA interprets the latter finding as tending to confirm
and reinforce indications based on the analyses of Mazumdar et al (1981) and others discussed
in Chapter 14 that positive associations exist between mortality and London BS values below
3 3
500 |ig/m (and possibly as low as 150-200 M9/ffl ). although none of the different individual
analyses can be said to conclusively delineate precise dose-response or exposure-response
relationships between London mortality and BS or SO, levels
14-136
-------
TABLE 14F-1
MULTIPLE REGRESSION ANALYSIS OF DEVIATIONS OF
MORTALITY VERSUS DEVIATIONS OF AIR QUALITY VARIABLES
TIME
PERIOD
Winter 1958-59
Winter 1959-60
Winter 1960-61
Winter 1961-62
Winter 1962-63
Winter 1963-64
Winter 1964-65
Winter 1965-66
Winter 1966-67
Winter 1967-68
Winter 1968-69
Winter 1969-70
Winter 1970-71
Winter 1971-72
STATISTICAL SIGNIFICANCE OF ASSOCIATION
(SIGN OF ASSOCIATION)
RELATIVE
TEMPERATURE HUMIDITY S02
S* S
-
-
S
S S S
S -**
-
S S
S
S
S
S
S
S

BRITISH
SMOKE

-
S
S

^Significance at the .05 level
**Negative association
14-137
-------
APPENDIX 14G

SUMMARY OF MAZUMDAR ET AL. YEAR-BY-YEAR
ANALYSIS OF LONDON MORTALITY DATA FOR WINTERS
OF 1958-59 TO 1971-72
14-138
-------
A third analysis of London mortality data in relation to PM and S0~ levels (in addition
to the two analyses summarized in Appendices 14E and 14F) was presented at the November 15-17,
1981, meeting of EPA's Clean Air Scientific Advisory Committee (CASAC) as part of public par-
ticipation and input concerning CASAC review of the penultimate draft of the present criteria
document. That third analysis, summarized here, was presented by Dr. Ian Higgins at the CASAC
meeting in order to provide an update regarding newly emerging results derived from the work
of Mazumdar et al. completed since the publication of the Hazumdar et al. (1981) report al-
luded to in the main text of Chapter 14.
Dr. Higgins described the results of a year-by-year analysis of the same 1958-59 to
1971-72 London winter mortality data previously analyzed by Hazumdar et al. (1981) but not on
a year-by-year basis. Certain allowances were incorporated into the analyses in order to
control for potentially confounding factors, e.g.: year to year variation in background mor-
tality rates; seasonal trends; temperature and humidity effects; lag effects; and day of week
or month of winter effects. Compared against 7-year averages of yearly coefficients for % mor-
3
tality per mg/m of each pollutant, notable increases (10-70%) in mortality over the 7-year
means were observed in relation to BS levels for most of the winters between 1958-59 and
1971-72, but smaller increases (10-20%) in mortality occurred in relation to S02 levels only
during about % of the winters. The results of regression analyses evaluating a change in %
mortality as a function of varying BS and SO,, levels over the entire 14 winters analyzed are
summarized in Figures 14G-1 and 14-G-2, which respectively show (1) variations in daily mor-
tality rates in relation to variations in particulate matter (BS) levels, holding concurrent
SOy levels constant and (2) variations in daily mortality rates in relation to SCL levels,
holding concurrent BS levels constant.
Dr. Higgins noted that the increase in mortality with increasing smoke concentrations
over different S09 concentrations is clear (as seen in Figure 14G-1). He further noted that
33
(as seen in Figure 14G-2) between about 600 and 800 nQ/m (0.6 and 0.8 mg/m ) S09 mortality is
33 3
flat for concentrations of smoke of 500 \jg/m (0.5 mg/m ) and above 500 ug/m mortality tends
to increase with increasing SO, concentrations (possibly due to interaction between smoke and
S0?). Taking this information into account and the results of the earlier regression analyses
reported by Mazumdar et al. (1981), as shown in Figures 14-2 and 14-3 of Chapter 14 of the
present document, Dr. Higgins concluded that: (1) the quadratic model dose/response curve of
o
Mazumdar et al. (1981) shows a possible threshold at 300 ug/m BS (though any material in-
crease in mortality appears to occur only at much higher levels); and (2) at daily pollutant
(BS and S0?) concentrations of less than 500 pg/m , any conclusions about dose-response rela-
tionships for mortality should be drawn with caution.
14-139
-------
Figure 14G-1.
15

„ 13
+*t
c
m
S 12
«
a
10
I
S 8
I 6
> 5

0
0.1
0 5 nig/m3
I / I I

1.1 mg/m3 SO2

1.0 mg/m3 SC>2

0.9 mg/m3 S02

0.8 mg/m3 SO2
0.2
0.3
0.4 O.S 0.6 0.7 0.8 Oi

BRITISH SMOKE (BS) LEVEL IN mg/m3
1.0
11
12
-------
Figure 14G-2.
15

1 "
o

I 10
fc
O
5
u.
O

5
0.7
1,1 mg/m3BS

1.0 mg/m3 BS

0,9 mg/m3 BS
?

0.8 mg/m3 BS —

0.7 mg/m3 BS —I
0.6 mg/m3 BS
0.5 mg/m3 BS
0 4 mg/m3 BS
0,1 0.2 0.3 0.4 0,5 0.6 07 0.8 0.9 1,0 1.1 1.2

SULFUR DIOXIDE (SO2) LEVEL IN mg/m3
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/8-82-029C
3 RECIPIENT'S ACCESSION NO
4. TITLE AND SUBTITLE
5 REPORT DATE
Air Quality Criteria for Participate Matter
and Sulfur Oxides. Volume III.
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, SW, Washington, DC 20460
13 TYPE OF REPORT AND PERIOD COVERED
FINAL
14 SPONSORING AGENCY CODE
EPA/600/00
IS, 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
participate 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
pollutants.

Although this document is principally concerned with the health and welfare effects of
sulfur oxides and particulate matter, other scientific data are presented and evaiju-
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; environr
mental concentrations and exposure levels; atmospheric chemistry ahd-dispersion
modeling; acidic deposition; effects on vegetation; effects on visibility, climate;,
and materials; and respiratory, physiological, toxicological, clinical and
epidemiological aspects of human exposure.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
COSATI I jfcld/Group
8. DISTRIBUTION STATEMENT

RELEASE UNLIMITED
19 SECURITY CLASS (1 hit Krportl

IJNCIASS i FLED
2O SfcCUP.TY CLASS {I/lit pane)
UNCLASSIFIED
21 NO Of PAGES

695
22 PRICE
EPA Form 2220—1 (Rev. 4— 77i PREVIOUS EDITION is OBSOLET E

.B. QOVSRNMiNT PRINTING QPFICS: 1986-646-116* it 0 6 3 7
-------