United St.i
Environ;:	t.il
Offn-t; ill
TOXK; Suhstai.i
Washington DC 20460
EPA 560 5 80 001
January 1981
  •
The Potential Atmospheric
Impact of Chemicals
Released to the
Environment

Proceedings of
Four Workshops

-------
                                        EPA 560/5-80-001
                                        January 1981
      THE POTENTIAL ATMOSPHERIC IMPACT OF
     CHEMICALS RELEASED TO THE ENVIRONMENT
         Proceedings of Four Workshops
                   Edited by
                John M. Miller
          Air Resources Laboratories
National Oceanic and Atmospheric Administration
           Silver Spring, MD  20910
                   98120AOTE
                Project Officer
                 William Wood
          Office of Toxic Substances
     U.S. Environmental Protection Agency
             Washington, DC  20460
     U.S. ENVIRONMENTAL PROTECTION AGENCY
   OFFICE OF PESTICIDES AND TOXIC SUBSTANCES
             WASHINGTON, DC  20460

-------
This document has been reviewed and approved for
publication by the Office of Toxic Substances,
U.S. Environmental Protection Agency.   Approval
does not signify that the contents necessarily
reflect the views and policies of the  Environ-
mental Protection Agency,  nor does the mention
of trade names or commercial products  constitute
endorsement or recommendation for use.
         The  report of a workshop on
  TOXIC SUBSTANCES  IN ATMOSPHERIC DEPOSITION
   has been distributed also  as  a report of
  the National Atmospheric  Deposition Program
                      11

-------
                           FOREWORD
     In the Toxic Substances Control Act of 1976, the Congress
expressed the finding that human beings and the environment
are being exposed each year to a large number of chemical
substances and mixtures, and that the effects of many on
health and environment are uncertain.  Congress decided
therefore that  (1)  adequate data should be developed with
respect to the effect of chemical substances and mixtures on
health and the environment and that such data should be the
responsibility of those who manufacture and those who process
such chemical substances and mixtures; (2)  adequate authority
should exist to regulate chemical substances and mixtures which
present an unreasonable risk of injury to health or the environ-
ment; and (3)  authority over substances and mixtures should be
exercised in such a manner as not to impede duly or create
unnecessary economic barriers to technological innovation.

     EPA is developing scientific risk assessment methodologies
which balance the primary purpose of TSCA, to assure that
chemical substances do not present an unreasonable risk, with
the admonition that the Agency not unduly impede innovation in
the chemical industry.  Development of these methodologies has
involved the identification of which health and environmental
effects EPA should consider to be important, the identification
of the information needed for a risk assessment, the organization
of this information into a logical assessment scheme, the
development of decision criteria to trigger testing, and the
identification and development of testing methods.

     As part of this activity EPA is soliciting the guidance of
the scientific community.  The workshops addressed in this
volume are a series sponsored by the Agency on the topic of
screening chemicals for atmospheric modification and its effects.
                        James J. Reisa, Ph.D.
                        Associate Deputy Assistant Administrator
                          for Toxic Substances
                        U.S. Environmental Protection Agency
                                 111

-------
                THE POTENTIAL ATMOSPHERIC IMPACT OF
               CHEMICALS RELEASED TO THE ENVIRONMENT
                             CONTENTS

                                                            Page

FOREWORD                                                     iii

ABSTRACT                                                      vi

Report of a workshop on TOXIC SUBSTANCES IN
  ATMOSPHERIC DEPOSITION
     Table of Contents                                         3

Report of a workshop on SCREENING CHEMICALS FOR
  INADVERTENT MODIFICATION OF THE STRATOSPHERE
     Table of Contents                                       149

Report of a workshop on THE IMPACT OF CHEMICALS
  ON THE RADIATIVE TRANSFER IMBALANCE
     Table of Contents                                       169

Report of a workshop on ANTHROPOGENIC CHEMICALS AS
  MODIFIERS OF CLOUDS AND PRECIPITATION
     Table of Contents                                       193
                               v

-------
                             ABSTRACT

     Four workshops are reported:  toxic substances in atmospher-
ic deposition, screening chemicals for inadvertent modification
of the stratosphere, the impact of chemicals on the radiative
transfer imbalance, and the impact of anthropogenic chemicals on
precipitation processes.
                              VI

-------
           Report of a Workshop on

TOXIC SUBSTANCES IN ATMOSPHERIC DEPOSITION:

           A REVIEW AND ASSESSMENT
           Jekyll Island, Georgia

                November 1979
                  Edited by

              James N. Galloway
            Steven J. Eisenreich
               Bryan C. Scott

-------
                   PARTICIPANTS AND CONTRIBUTORS
 Deborah A.  Banning
 Department of Environmental
  Sciences
 University of Virginia
 Charlottesville,  Virginia

 L. A.  Barrie
 Atmospheric Environment Service
 Downview, Ontario,  Canada

 T. F.  Bidleman
 Department of Chemistry
 University of South Carolina
 Columbia, South Carolina

 Anthony R.  Davis
 Water  Quality Branch
 Inland Waters Directorate
 Ottawa,  Ontario,  Canada

 Steven J. Eisenreich
 Environmental Engineering
 University of Minnesota
 Minneapolis,  Minnesota

 James  N.  Galloway
 Department of Environmental
  Sciences
 University of Virginia
 Charlottesville,  Virginia

 Donald F. Gatz
 Illinois  State Water Survey
 Urbana,  Illinois
C.  S.  Giam
Department of Oceanography
Texas  A  &  M
College  Station,  Texas

Ronald A.  N.  McLean
DOMTAR Research Center
Senmeville, Quebec, Canada

William Wood
Environmental Protection Agency
Washington, D.C.
John M. Miller
Air Resources Laboratories
NOAA/ERL
Silver Spring, Maryland
Michael D. Mullin
Large Lakes Research Station
Grosse lie, Michigan

Thomas J. Murphy
Chemistry Department
DePaul University
Chicago, Illinois

Stephen A. Norton
Department of Geological Sciences
University of Maine
Orono, Maine

Donald H. Pack
Certified Consulting
 Meteorologist
McLean, Virginia

Francis J. Priznar
Normandeau Associates
25  Nashua Road
Bedford, New Hampshire
 Bryan C.  Scott
 Battelle  Pacific Northwest
  Laboratories
 Richland, Washington

 David Thornton
 Environmental Engineering
 University of Minnesota
 Minneapolis, Minnesota

 Herbert L. Volchok
 Environmental Measurements
 Laboratory
U.S. Department of Energy
New York,  New York

-------
            TOXIC SUBSTANCES IN ATMOSPHERIC DEPOSITION


                             CONTENTS
                                                              Page

PARTICIPANTS AND CONTRIBUTORS                                    2

PREFACE                                                          9

ACKNOWLEDGMENTS                                                 10

INTRODUCTION                                                    11

EXECUTIVE SUMMARY                                               13

PART I:  TRACE METALS:  A REVIEW AND ASSESSMENT                 19

     1.   Introduction                                          19

     2.   Concentrations:  Naturally and Anthropogenically
          Controlled                                            20

          2.1. Comparison of Emission Rates                     20
          2.2. Comparison of Atmospheric Concentrations         22
          2.3. Determination of Historical Trends in
               Deposition                                       24
          2.4. Comparison of the Three Techniques               27

     3.   Concentration of Metals in the Atmosphere             27

     4.   Concentration of Metals in Wet Deposition             29

     5.   Deposition of Metals from the Atmosphere              39

          5.1. Deposition in Remote, Rural, and Urban
               Environments                                     39
          5.2. The Relative Importance of Dry versus
               Wet Deposition                                   39
          5.3. Physical Characteristics of Metals and
               Their Compounds Affecting Atmospheric
               Deposition                                       44

               5.3.1.   Vapor Pressure                           44
               5.3.2.   Particle Size                            46
               5.3.3.   Solubility                               48

          5.4. Measurement of Wet Deposition                    48
          5.5. Measurement of Dry Deposition                    50

-------
                                                              Page

     6.   Potential Effects of Increases of Concentration
          of Metals in Atmospheric Deposition                   51

          6.1. Recommended Upper Limits in Water                51
          6.2. Metal Speciation in Relation to Toxicity         51

     7.   Summary                                               54

     8.   Research Recommendations                              55

     9.   References                                            55

     APPENDIX A:  Concentrations of Metals in the
                  Atmosphere:  Selected Data                    64

     APPENDIX B:  Concentrations of Metals in Wet
                  Deposition                                    72

PART II:  TRACE ORGANICS: A REVIEW AND ASSESSMENT               83

     1.   Introduction                                          83

     2.   Vapor and Particle Distribution of Atmospheric
          Organics                                              85

     3.   Transfer of Gases Across Air/Water Interface          91

          3.1.   Resistance to Air/Water Interface               92
          3.2.   Problems                                        95

     4.    Wet Removal of Trace Organics from the Atmosphere     95

     5.    Dry Deposition of Particulates                        97

     6.    Sources of Trace Organics to the Atmosphere           98

          6.1.   Sources I                                       98
          6.2.   Sources II                                     100

     7.    Variability of Concentrations and Fluxes             100

     8.    Atmospheric Concentrations                           100

     9.    Total  Deposition in Urban,  Rural,  and Remote
          Environments                                          103

     10.   Applications  of Flux Calculations: Atmospheric
          PCB Input to  the Great Lakes                          105

          10.1. Dry Deposition                                105
          10,2. Wet Deposition                                106

-------
                                                              Page

     11.  Research Recommendations                             109

     12.  References                                           110

PART III:  GUIDE FOR ESTIMATING DEPOSITION RATES OF GASES
           AND AEROSOLS                                        114

     1.   Introduction                                         114

     2.   Wet Removal and Deposition of Aerosols and Gases     115

          2.1. Wet Removal of Aerosols                         115

               2.1.1.  Attachment                              115
               2.1.2.  Removal                                 118

          2.2. Wet Deposition of Aerosols                      120

               2.2.1.  Air Concentrations                      120
               2.2.2.  Wet-Deposition Rates                    120

          2.3. Wet Removal of Gases                            121
          2.4. Wet Deposition of Gases                         123

     3.   Dry Removal and Deposition of Aerosols and Gases     123

          3.1. Dry Removal of Aerosols                         124
          3.2. Dry Deposition of Aerosols                      125
          3.3. Dry Deposition of Gases                         125

     4.   Relative Removal by Wet and Dry Processes            127

     5.   Special Situations                                   128

          5.1. Unique Emission Configurations                  128
          5.2. Unique Receptor Configurations                  129
          5.3. Unique Meteorological Environments              130

               5.3.1.  Coastal Areas                           130
               5.3.2.  Precipitation Extremes                  130

     6.   Research Recommendations                             131

     7.   References                                           131

BIBLIOGRAPHY                                                   134

-------
                             FIGURES

                       Part I.   Trace Metals

Figure                                                        Page

 1. EF    ... values for atmospheric trace metals collected
      crust
     in the North Atlantic westerlies from R/V Trident,
     Bermuda, and at the South  Pole                             23

 2. Element concentrations versus depth in sediment of
     Woodhull Lake,  New York                                    25

 3. Pb concentration as a function of depth in the
     sediments of three lakes                                   26

 4. Airflow over the United States                              33

 5. Average deposition of Pb by precipitation                   34

 6. Average deposition of Zn by precipitation                   34

 7. Median concentrations of metals in precipitation
     in remote,  rural, and urban areas relative to
     organism-toxicity levels                                   52


                     Part II.  Trace Organics

 1.  Atmospheric  input of anthropogenic organics to
     natural waters                                              84

 2.  Vapor-aerosol distributions calculated from vapor
     pressure and the quantity  of atmospheric particulates       86


        Part III.   Guide for Estimating Deposition Rates

1.  Collision efficiency as a function of size of the
     collected particles                                      116

2.  Theoretical  growth curves for solution droplets  of
     sulfuric  acid and some inorganic salts of interest
     at 25°C                                                   116

3.  Growth of  aerosol  composed  of 50% ammonium sulfate         117

4. Variation  of washout  ratio,  W, with the mass median
    diameter  at the distance from the urban sources
    for St. Louis and Chilton,  United Kingdom                 llg

-------
                             TABLES

                      Part I.   Trace Metals

Table                                                         Page

 1. Global mobilization factors based on annual
     emission rates                                             21

 2. Three techniques for determining the influence
     of anthropogenic processes on the concentration
     of metals in atmospheric deposition                        28

 3. Ranges of metal concentrations in the atmosphere            30

 4. Approximate median concentrations of metals in
     the atmosphere                                             31

 5. Ratios of median concentrations of metals in
     the atmosphere                                             32

 6. Ranges of metal concentrations in wet deposition
     (rain, snow, and ice)                                      36

 7. Median concentrations of metals in wet deposition
     (rain, snow, and ice)                                      37

 8. Ratios of median concentrations of metals in wet
     deposition (rain,  snow, and ice)                           38

 9. Selected values of atmospheric-deposition rates
     for metals in remote, rural,  and urban areas               40

10. Means of data reported from all seasons for dry
     fraction of total deposition                               44

11. Vapor pressures and boiling points of metals
     and their common oxides                                    45

12. Mass median diameter of toxic metals in aerosols            47

13. Solubility of common oxides,  sulfates,  chlorides,
     and nitrates of metals at 20° C                            49

14. Recommended upper limits for metal concentrations
     in water                                                   53
                    Part II.   Trace Organics

 1.  Vapor pressures of trace-organic compounds                  87

-------
m wn                                                          Page
Table

 2. Apparent participate- and gas-phase distributions for       ^
     organics in urban air

 3. Ranges of gas-particulate distribution factors for
     organics, nitrates, and sulfates in Pasadena CA            by

 4. Particulate- and vapor-phase distributions for
     atmospheric PCBs

 5. Particle-size distributions for PAHs in ambient
     aerosols in urban, rural, and coastal areas                91

 6. Water solubilities of trace-organic compounds               93

 7. Henry's law constants                                       94

 8. Washout ratios for trace-organic compounds                  96

 9. Dry-deposition velocities                                   99

10. Atmospheric concentrations of trace-organic
     compounds in urban and rural environments                 101

11. Atmospheric concentrations of PAHs  in different
     environments                                              102

12. Atmospheric concentrations and fluxes of  PCBs
     in different environments                                 103

13. Atmospheric fluxes of trace-organic compounds
     to rural environments                                     104

14. Parameters for PCB flux to the Great Lakes                 107

15. Atmospheric flux of PCBs to the Great Lakes                108


        Part III.  Guide for Estimating Deposition Rates

 1. Washout Ratios                                             121

 2.  Particle-deposition velocities                             125

 3.  Ranges of Vd for a substance with a Jekyll number <10      12?
                               8

-------
                             PREFACE


     Our knowledge of the effect of human activities on atmos-
pheric deposition is essentially limited to three elements (S, N,
H) in the periodic table.  It is generally known that there are
other elements in atmospheric deposition and that their concen-
trations have been enhanced in recent years, but little specific
and systematic detail exists on their temporal and spatial trends
To establish the state of the art, a workshop was convened in
November 1979 to assess what is known and unknown about toxic
metal and organic compounds in atmospheric deposition.

     This assessment was made at the request of the Environmental
Protection Agency (EPA), the National Oceanic and Atmospheric
Administration (NOAA), and the National Atmospheric Deposition
Program (NADP) to determine the pathways of removal of toxic
substances from the atmosphere; the actual concentrations of
metal and organic compounds in atmospheric deposition; the degree
that human activities contribute to observed concentrations in
atmospheric deposition in urban, rural and remote areas; and the
possible threat to the environment.

     The workshop, held on Jekyll Island, Georgia, was composed
of three working groups.  The depositional processes group was
concerned with describing the specific processes and critical
parameters that remove toxic substances from the atmosphere.   The
groups on toxic metals and on organic compounds in atmospheric
deposition addressed two questions:

     (1)  What is known or can be predicted about the deposition-
          al processes of metal and organic compounds?
     (2)  What are the relative contributions of natural and
          anthropogenic processes to the concentration of metal
          and organic compounds in atmospheric deposition?

-------
                         ACKNOWLEDGMENTS


     As organizers, we thank EPA, NOAA, and the National At-
mospheric Deposition Program for giving us the opportunity to
have this workshop.  Very special thanks go to Tamara Gardner
for .providing all the logistical support for the workshop (under,
at times, arduous conditions) and for masterminding the prepara-
tion of this report and to Mary-Scott Marston for being creative
and accurate in her editing.  Finally we deeply appreciate the
long hours of preparation, participation,  and writing of the
attendees.

                                                  JNG
                                                  SJE
                                                  BCS
                              10

-------
                           INTRODUCTION
     The number of observations on atmospheric deposition and its
composition has substantially increased since the first observa-
tions of Noah1, Little2, Barrinchins3, and Smith4.  In the middle
1960s, several investigations began on the temporal and spatial
trends in the composition of atmospheric deposition.  As a result
of this work precipitation networks were established and research
on the effects of atmospheric deposition was accelerated.  Most
past studies focused primarily on three elements in the periodic
table, S, N, and H, with a secondary  focus on Na, K, Ca, Mg, and
Cl.  There has been a paucity of research on metal and organic
compounds in atmospheric deposition.

     Similarly, little  systematic attention has been paid to
concentrations in precipitation and atmospheric deposition rates
of metal and organic compounds in different regions of the world.
Therefore the assessment of our current state of knowledge was
easy (especially for organic compounds)--we know very little.
This ignorance is not healthy, literally or figuratively-  Our
deliberations have shown that for both metal and organic com-
pounds current concentrations in atmospheric deposition exceed
the limits established  to reduce the  effects on humans and other
organisms.  It is crucial that the present data base be systemat-
ically expanded so that the same questions are not still being
discussed ten years from now.

     The reports of the working groups on metal compounds and on
organic compounds in atmospheric deposition are in part litera-
ture reviews of existing information  and assessments of the
present state of our knowledge of the impact of human activities
on metal and organic compounds in atmospheric deposition.  Speci-
fic questions are addressed:

     (1)  What are the  toxic metal and organic compounds?
     (2)  What is the relative importance of natural versus
          anthropogenic processes acting as atmospheric sources
          for these substances?
     (3)  What are their concentrations in wet deposition in
          remote, rural, and urban areas?
     (4)  What are their depositional fluxes in remote, rural,
          and urban areas?
     (5)  What is the relative importance of dry versus wet
          deposition?
1Genesis 7:12; 2Little, C., in The Sky is Falling!  The Sky is
Falling! (date unknown) public domain; 3Cited in Hjarne, Acta et
Tentamina Chemica Holmiensia 2, 23 (1753) (Stockholm); 4Smith, R.
A., Air and Rain: The Beginnings of a Chemical Climatology-
London: Longmans, Green and Co. (1872) 617 pp.
                                11

-------
      (6)  What are the water-quality criteria for man and other
          organisms relative to the concentrations in precipita-
          tion?

      The report of the working group on depositional processes
describes the properties of the gases and aerosols that control
the relative importance of wet and dry deposition.  Simple rules
for estimating wet- and dry-depositional rates are presented in
terms of washout ratios and depositional velocities.

     Each report contains recommendations for research.  These
recommendations concentrate on those areas where the extent of
our ignorance is especially profound.  It was the consensus that
additional data are needed on more metal and organic compounds
before the impact of toxic metals and organic compounds in atmos-
pheric deposition can be assessed.
                              12

-------
                        EXECUTIVE SUMMARY
Part I.  TRACE METALS

     Information on nineteen metals in atmospheric deposition
potentially toxic to humans and other organisms was assimilated
to determine if metal concentrations are increasing in atmos-
pheric deposition and if these concentrations threaten human or
organism health.

     On the basis of rates of emission, atmospheric concentra-
tions, and known temporal trends in deposition, the greatest
increases in concentrations of metals in atmospheric deposition
due to human activity are expected for Ag, Cd, Cu, Pb, Sb, Se,
Zn, with smaller increases expected for Cr and V and with little
or no increases expected for Co, Mn, and Ni.  There were insuffi-
cient data to rank Mo, As, Be, Sn, Te, and Tl.

     Although actual data on these metals in atmospheric deposi-
tion are limited, the data available supported these expecta-
tions.  The metals Zn, Pb, Cu, Mn, Ag, As, and V had measured
concentrations 30 to 200 times higher in atmospheric concentra-
tion or deposition in rural, continental areas than at remote
areas such as the South Pole.  Other metals, Sb, Se, Cr, and Ni,
had concentrations that were 10 to 30 times greater in rural
areas than in remote areas.

     Metals can be deposited either wet or dry from the atmos-
phere.  An assessment of the relative importance of the two
processes revealed that, depending on the metal and the area, dry
deposition can be as great as or greater than wet deposition.

     From analyses of vapor pressures of metals and metal oxides,
only Hg, As, Se, and possibly Cd could be expected to have a
significant fraction of their atmospheric concentration in the
vapor phase.

     In regard to the effects of increased metal concentrations
in atmospheric deposition, only Pb and Hg are currently in pre-
cipitation in some areas at levels greater than the drinking-
water standard.  Cd, Cu, Hg, Pb, and Zn can be present in precip-
itation at levels higher than the standards for effects on other
organisms.
                               13

-------
Research Recommendations

     Assimilating the work on metals in atmospheric deposition
was less difficult than it might have been because so little has
been done in spite of the scope of the problem.  To avoid this in
the future, we propose the following research recommendations.

     (1)  More data must be acquired through more studies, in-
          cluding the analysis of more metals.
     (2)  The concentration of soluble versus insoluble metals in
          rain and melted snow should be determined under the
          assumption that soluble metals are probably more mobile
          in the environment.
     (3)  Sampling and analytical methods must be standardized
          throughout the scientific community for all metals,
          including those that take on different forms in the
          atmosphere (i.e., Hg, As).
     (4)  Since dry deposition can be just as important as, if
          not more important than, wet deposition, standardized
          collection procedures must be developed.
     (5)  A detailed study of the metal compounds that predomi-
          nate in rain and snow must be conducted.
     (6)  The size distribution of metals in urban, rural, and
          remote atmospheres needs to be determined.
     (7)  The fate of metals deposited in aquatic and terrestrial
          ecosystems must be determined.
     (8)  A national network to determine the temporal and spa-
          tial trends of metals in atmospheric deposition must be
          established.
Part II.  TRACE ORGANICS

     Atmospheric transport and deposition may be significant
contributors to the distribution and accumulation of trace organ-
ics in the biotic and abiotic compartments of the aquatic ecosys-
tem.  Trace organics, such as polychlorinated biphenyls (PCBs),
chlorinated hydrocarbon pesticides (CH), and polycyclic aromatic
hydrocarbons (PAHs), are emitted into the atmosphere from sani-
tary landfills and municipal incinerators (PCBs, CH), aerial
application to forests or crops (CH), and the high-temperature
combustion of fossil fuels (PAHs).  Airborne trace organics are
distributed between the gas and particle phases; the relative
importance of each depends on compound vapor pressure,  type and
quantity of absorptive surface area,  and emission-source strength.
Once in the atmosphere, airborne contaminants are removed by
either wet or dry removal processes.   PCBs and DDT remain in the
atmosphere for approximately 5 to 10 days.  Gas-phase organics
can be removed by precipitation scavenging and direct partition-
ing across the air/water interface,  both processes depending on
the magnitude of Henry's law constant.  Particle-phase trace
organics are usually associated with submicron-size particles and
                                14

-------
are removed by precipitation scavenging (W -\. 105 ) and impaction
on a water surface (V, ~ 0.1-1.0 cm s~M-

     Numerous reports and the literature published over the past
ten years were reviewed to assemble values for concentrations of
trace organics in the atmosphere and in precipitation.  At
present, a paucity of data on trace organics in the atmosphere is
readily evident.  Ranges and, where possible, medians of trace-
organic concentrations in the vapor and particle and in the rain
and snow phases in both urban and rural areas were compiled.
With simplified parameterizations of wet-  and dry-deposition
processes, fluxes of trace organics to water surfaces were calcu-
lated.  These should be interpreted as order-of-magnitude esti-
mates only.  Conservative estimates of trace-organic fluxes
suggest that dry deposition dominates wet deposition for all
compounds studied.

     The application of flux calculations  to atmospheric PCB
input to the Great Lakes has been included.  Atmospheric input to
aquatic systems is the major source of trace organics for lakes
lacking surface sources and having large surface-area/basin-area
ratios.  The Great Lakes and remote inland lakes,  therefore,
receive 50% to 100% of their trace-organic burden from the atmos-
phere .


Research Recommendations

     Implementing the following research recommendations is
necessary to relieve the paucity of data on trace organics in the
atmosphere and their deposition and effects.

     (1)  Methods must be developed to distinguish between vapor-
          and particulate-phase, high-molecular-weight organics
          in the atmosphere.
     (2)  The relationship between the mass median diameter, the
          deposition velocity,  and the receptor surface for
          atmospheric particles must be established.
     (3)  The size distribution of atmospheric particulates
          containing high-molecular-weight organics must be
          determined.
     (4)  Collection methods for dry deposition of particulate-
          and vapor-phase organics should be developed since this
          is an important fraction of total deposition.
     (5)  The collection efficiency of different receptor sur-
          faces for particulate- and vapor-phase organics must be
          determined.
     (6)  Accurate data on vapor pressures and water solubilities
          of slightly soluble,  high-molecular-weight organics
          need to be developed.
     (7)  The fugacities (true dissolved fraction) of trace
          organics in water need to be developed.
                               15

-------
     (8)  The trace-organic compositions of rain and snow and of
          the aerosol and gas phases in the atmosphere need to be
          determined before accurate deposition rates and ecolog-
          ical or health impacts can be evaluated.
Part III.  GUIDE FOR ESTIMATING DEPOSITION RATES OF GASES AND
           AEROSOLS

     This section provides procedures for predicting deposition
of gases and aerosols that are uniformly distributed in the
lowest several kilometers of the atmosphere.


Wet Deposition of Gases and Aerosols

     To compute the wet deposition of a material, it is necessary
to have information about the atmospheric state of the material
(gas or aerosol), the anticipated air concentration, and the
likely precipitation rates and durations.  The wet depositions of
aerosols and of gases are expressed in terms of the washout
ratio.  For the wet removal of aerosols, the washout ratio is
parameterized in terms of the particle size and the type of
rainfall (convective vs. continuous).  For the wet removal of
gases, the washout ratio is approximated from the inverse of
Henry's law constant.


Dry Deposition of Gases and Aerosols

     The dry depositions of gases and of aerosols are expressed
in terms of deposition velocity.  Aerosol-deposition velocities
are parameterized in terms of an anticipated mean particle size
and in terms of surface characteristics (land vs. water).  The
'dry deposition of gases is so complex and variable that no at-
tempt was made to generalize the deposition velocity; rather, a
first approach to making an order-of-magnitude estimate of gas
deposition velocity is outlined.

     Our primary conclusion was that the predicted deposition
rates should be within a factor of two or three of the true
values if the computations are carried out over time and space
scales on the order of 1 yr and 100 km.  Improved accuracy re-
quires a more sophisticated approach than could be provided here;
however, references to such approaches have been given.


Research Recommendations

     The research recommendations below are mandatory if the
uncertainty inherent in deposition calculations is to be allevi-
ated:
                                16

-------
(1)   Washout ratios for aerosols must be classified by storm
     type,  location,  season,  rainfall rate,  size,  and mass
     median diameter over many locations and seasons.
(2)   Gases  must be grouped by their chemical and physical
     properties to determine  if such categorization relates
     to the deposition properties.
(3)   How much gas is reemitted to the atmosphere after
     initial deposition must  be determined.   Many gases are
     transported to the surface but may be reemitted to the
     atmosphere after the precipitation is deposited on the
     ground.
(4)   Extensive measurements of the deposition velocity to
     various surfaces (snow and ice fields,  water surfaces,
     grasslands and,  especially, forests) are necessary to
     define dry deposition of aerosols more precisely.
(5)   The importance of dry deposition of a given (previously
     unknown) gas to total deposition should be established
     through further development, and eventual standardiza-
     tion and refinement, of  the use of the Jekyll number.
                          17

-------
           TOXIC SUBSTANCES IN ATMOSPHERIC DEPOSITION:
                   A REVIEW AND ASSESSMENT

                             Part I
             TRACE METALS:  A REVIEW AND ASSESSMENT
   J. N. Galloway (Co-chairman), H. L. Volchok (Co-chairman)
          D. Thornton, S. A. Norton and R. A. N.  McLean
1.   INTRODUCTION

     The passage of metals through the atmosphere is integral to
biogeochemical cycling.  Because of the dynamic nature of the
atmosphere, metals can be deposited in areas remote from the
initial source.  Historically, the rate of deposition has been
low because of the low volatility of most metals.  However, with
the advent of high-temperature anthropogenic processes (smelting
and fossil-fuel combustion), the rate of emission for some metals
has substantially increased.  With increased emissions have come
increases in metal concentrations in the atmosphere and in atmos-
pheric deposition.  Because of the known toxicity of some of
these metals to humans and other organisms, this workshop was
convened in part to assess the present knowledge of toxic metals
in atmospheric deposition.

     Metals classified as toxic or potentially toxic to humans
and other organisms (Ag,  As, Be, Cd, Co, Cr, Cu, Hg, Mn,  Mo, Ni,
Pb, Sb, Se, Sn, Te, Tl, V, and Zn; Wood 1974) have all been found
in atmospheric deposition.  Several critical questions need to be
asked to assess the potential risk of these metals to human
health and the welfare of organisms and ecosystems:

     (1)  What is the relative importance of natural versus
          anthropogenic processes acting as sources of metals in
          the atmosphere?
     (2)  What are the metal concentrations in the atmosphere in
          remote, rural,  and urban areas?
     (3)  What are the metal concentrations in precipitation in
          remote, rural,  and urban areas?
     (4)  What are the depositional fluxes of metals in remote,
          rural,  and urban areas?
     (5)  What is the relative importance of dry versus wet
          deposition?
     (6)  What are the water-quality criteria for humans and
          other organisms?
     (7)  What is the relationship between speciation and toxic-
          ity of the metals in atmospheric deposition?
     (8)  Where should future research be directed to fill in the
          gaps in our knowledge?


                                19

-------
2.   CONCENTRATIONS:  NATURALLY OR ANTHROPOGENICALLY CONTROLLED

     There are three ways to assess whether anthropogenic or
natural emission processes control current concentrations of
metals in precipitation:

     (1)  Compare the actual metal-emission rates of natural and
          anthropogenic processes;
     (2)  Compare the ratios of atmospheric concentrations of
          metals to the ratios of concentrations in source ma-
          terials;
     (3)  Determine temporal trends in the composition of metals
          in atmospheric deposition.

     These three techniques—emission to, concentration in, and
deposition from the atmosphere--are used to analyze the different
parts of the atmospheric cycle of metals to determine the impact
of anthropogenic processes on metal concentrations in atmospheric
deposition.


2.1  Comparison of Emission Rates

     On a global basis, the known natural sources of metals in
the atmosphere are injection of soil and volcanic dust and gas-
eous emanations.  The anthropogenic emissions are from industrial
gases and particulates and combustion of fossil fuels.  To esti-
mate the magnitude of the natural and anthropogenic fluxes on a
global basis, a recent analysis (Lantzy and Mackenzie 1979) has
been used to calculate a Mobilization Factor (MF) (Table 1-1)
where


          MF _ Emission rate from anthropogenic sources
               Emission rate from natural sources

An analysis of the calculated MF shows that the metals that would
be expected to be enriched in the atmosphere (and subsequently in
atmospheric deposition) by anthropogenic processes are Pb > Ag >
Sb, Mo > Zn,  Cd > Cu > Sn > V,  As, Se, Ni > Cr > Mn, Co, Hg.

     Using quite different methods, another study estimated the
global intensity of natural and anthropogenic emissions (Nriagu
1979) for some toxic metals.  Although the emission rates_ob-
tained were different from those in Table 1-1 (in 108 g y"1:
natural—Cd-8.3, Cu-185, Ni-260, Pb-245, and Zn-435; anthro-
pogenic—Cd-73,  Cu-560, Ni-470, Pb-4500, Zn-3140), the order of
mobilization factors was the same:  Pb > Cd > Zn > Cu > Ni.  The
absolute degree of enrichment of the metal is dependent on (1)
the spatial scale used, (2) the amount of metal that is deposited
close to the source, and (3) the accuracy of our estimate of the
natural and anthropogenic emission rates.
                                20

-------
          Table 1-1.  Global Mobilization Factors Based
                    on Annual Emission Rates
Emissions (108 g y""1 )

Ag
As
Cd
Co
Cr
Cu
Hg
Mn
Mo
Ni
Pb
Sb
Se
Sn
V
Zn
Natural
0.6
28 [210]
2.9
70
580
190
0.4 [250]
6,100
11
280
59
9.8
4.1 [30]
52
650
360
Anthropogenic
50
780
55
50
940
2,600
110
3,200
510
980
20,000
380
140
430
2,100
8,400
Mobilization
factor
83
3
19
0
1
13
0
0
45
3
340
39
4
8
3
23

.3

.71
.6

.44
.53

.5


.7
.3
.2

Source:  Adapted from Lantzy and Mackenzie 1979.

NOTE:  Natural emissions = Soil dust + volcanic dust and
volcanic-emanation fluxes.  For elements with known volatile
species in the atmosphere (As, Hg, and Se), vapor emissions
(in brackets) from land and sea were added to dust emissions
(Lantzy and Mackenzie 1979).  Anthropogenic emissions =
fossil-fuel and industrial-particulate fluxes.
                                21

-------
     Relative to the first point, the MF values in Table 1-1 are
calculated for the global scale; if this scale were reduced to
the area of the United States, the MF values would certainly
increase because of a greater density of anthropogenic processes
relative to the global average.  On the second point, the calcu-
lated MF values show that the atmospheric emissions of many
metals from anthropogenic processes are certainly greater than
those from natural processes.  However, this analysis assumes
that for all metals except As, Hg, and Se only soils and volca-
noes are the important natural sources.  However, as Lantzy and
Mackenzie (1979) pointed out, there is a possible net flux of
metals into the atmosphere from vegetation, oceans, and low-
temperature volatilization from soils of vapor-phase metals or
metal compounds.

     Therefore, on the scale of the United States, the MF values
will be larger  than those in Table 1-1, but if there are addi-
tional natural  sources these increases may be reduced.  Since the
primary focus of this report is to determine if metals are in
higher concentration in atmospheric deposition because of anthro-
pogenic activities, we will accept the limitations of the MF
values and compare the ranking of metals expected to be in in-
creased concentrations in atmospheric deposition with the results
from the other  techniques.


2.2  Comparison of Atmospheric Concentrations

     The second technique compares elemental ratios of atmos-
pheric concentrations with concentration ratios in the earth's
crust.  The function commonly used is the Enrichment Factor, EF:
                    EF = ^        air
where, if EF > 1, a metal, Me, is enriched in the atmosphere
relative to its concentration in the crust, which implies a
source other than the crust; where, if EF = 1, Me is not enriched
in the atmosphere, which implies a crustal source.

     Enrichment factors have been calculated for several metals
in different localities by Lantzy and Mackenzie (1979), Duce et
al.  (1975), Chester and Stoner (1974), and Zoller et al. (1974).
A summary of these studies shows that Se, Pb, Sb, Cd, Cu, and Zn
have greater concentrations in the atmosphere in remote areas
than would be expected from the contribution of crustal material;
Co, Mn, Cr, and V showed relatively little or no enrichment
(Figure 1-1).

     In remote areas (e.g., Antarctica, the South Atlantic Ocean)
there is some question as to the causes of EF > 1 (Duce et al.
1975).   Possible reasons, other than the long-range transport of
anthropogenic aerosols,  could be high-temperature volatilization


                                22

-------
  104
   10
 tfi
   10'
   101
| North Atlantic  Westerlies

Q South Pole
           _t_f_	
                           Element
Figure 1-1.  EFcrust values  for  atmospheric trace metals collected
in the North Atlantic westerlies from R/V Trident, Bermuda,  and
at the South Pole.   The vertical bars represent the geometric
standard deviation;  the horizontal dashes (crossing the bars),
the geometric mean-enrichment  factor (Duce et al. 1975, p. 60;
copyright 1975 by the American Association for the Advancement of
Science.)
                               23

-------
(volcanoes), low-temperature volatilization, or improper ref-
erence material.  For example, since Se is emitted into the
atmosphere in the vapor phase as well as by injection of crustal
material, the EF ratio may be artifically large.  The same may be
true of Hg and As, although their EF values have not been calcu-
lated.

     In densely populated areas of the United States, anthropo-
genic activity probably causes EF > 10.  For example, natural
emissions of sulfur are about 1.3 times greater than anthro-
pogenic emissions, on a global basis.  However, because of the
intensity of anthropogenic emissions in the eastern United States,
the ratio of anthropogenic to natural emissions is about 100
(Galloway and Whelpdale 1980).  Therefore, for the United States,
the relative order of the enrichment of metals in the atmosphere
by anthropogenic processes is estimated to be Se > Pb > Sb > Cd >
Cu, Zn > Cr > V > Co.


2.3. Determination of Historical Trends in Deposition

     The third  technique uses historical records on the composi-
tion of  atmospheric deposition found in glaciers (Herron et al.
1977; Weiss et  al. 1975) and in lake sediments from remote areas
(Davis and  Galloway 1980; Galloway and Likens 1979; Norton et al.
1978, 1980).

      In  the glacial records from Greenland, Zn and Pb have in-
creasing rates  of atmospheric deposition.  But, as Boutron (1980)
recently stated, the increasing rates are based on very few
reliable analytical data even for these elements.

     Most studies of mercury on the Greenland ice sheet can now
be discounted because of erroneous sampling and analytical de-
fects (McLean et al. 1980).  A more recent study based on excep-
tionally careful analytical techniques showed no increase in
mercury  deposition in the ice sheet between 1727 and 1971 (Appel-
quist et al. 1978).

     For some lake sediments in the eastern United States, large
increases (>10 times) in the rates of deposition in the sediment
are seen for Pb, Sb,  and Au, with smaller increases (>5 times)
for Ag, Cd, and Cu.  Cr, Zn, and V show only slight increases
(Figure  1-2).  These increases in sediment deposition are primar-
ily caused by increases in atmospheric deposition (Galloway and
Likens, 1979; Norton et al., 1978, 1980).  The magnitude of the
increases of the atmospheric deposition of metals depends par-
tially on the location of the lake relative to the source.  As an
example,  data on lead deposition from three lakes in the eastern
United States are presented in Figure 1-3.  The three lakes lie
on a line normal to the mean annual air flow (Figure 1-4) and
downwind from metropolitan areas that are sources for atmospheric
lead.
                                24

-------
MG / KG DRY WEIGHT
a
(
0
5
10
15
20
25
30
35
 Au (//G/KG)
01 2 3 4 5
•
.
- •
- '
- '
-.
-
•) Cu
3 6 12 W 24 30 3
.
-
*
.
t
- '
•
D v
1 30 36 42 48 54 61
•
•
-
-
>
*

(
B
0
3
e
9
12
15
•
21
24
(
«
O
5
10
15
2O
25
SO
35
4O
d
3 C
0
3
6
9
12
15
W
21
24
D Cd
0 1 2 3 4 5 «
•
-
*
-
-
- •
"•
D Pb
} 40 80 120 160 200 36
111 1 ' 1
•
•
•
..
-


) Zn
75 ISO 225 3OO 375 45C
.
-


*
- •
•
Figure 1-2.  Element concentrations versus depth in sediment of
Woodhull Lake, New York.  (Galloway and Likens 1979,  p.  429;
copyright 1979 by American Society of Limnology and Oceanography,
Inc. )
                                25

-------
        CL
        LU
0
5

10

15

20

25

30
in







•

•


• -




^___ 	

f
<




' 	 1 	 1 	
,,
x-^"
f
^/





SPECK POND, ME


i ' 	 1 	 1 	 1 	 L.
/
-

-

..


.

.

'
l»fO
1948

1918

1888


1858

1828

1798

O
LU
t
r-
(/)
O
Q.
LU
O

(T
LU
>•


                 20
40

0


5
"E
i 10


X
o. 15
LU
O
20
tn

^x-x. 	 .
	 | „ . I— ^» I
	 x — 	 	 * 	
X^~*
,/X
" X

/
/
'(
| WOODS LAKE, NY
-X
I | i i i i i i i i i i i
1978


1948 Q
LU
^_
1848 5;
O
i\
1748 Si
O
1648 en

1548 >
1448
                 20
40
O


4
E
o

X 8
1-
Q.
in
0 12
16



-
background




MOUNTAIN LAKE, VA
iii i i i i i i
l
/
X
./
«'"^

/
1

_

19/8

Q
1930 W


0
1880 Si
O

1830 <
LU
i7«n
                 20    40    60    80    100    120
                                    ,-1
                                Pb  9"
Figure 1-3.  Pb concentration  as  a  function of depth in the
sediments of three lakes.  Year deposited is approximate and was
determined by Pb210 and/or Cs137  dating.   Locations of the lakes
are shown in Figure 1-4.
                                26

-------
     These profiles of lead deposition (Figure 1-3) show that the
Maine and New York lakes have accumulated more lead, relative to
natural levels, than the Virginia lake.  This is consistent with
increases being due to atmospheric deposition; the lake in Vir-
ginia is less influenced by anthropogenic Pb sources because it
is farther from those sources.

     Although several studies have related the variation of
mercury concentration with its depth in sediment cores to varia-
tions in the deposition rate, Matsunaga (1978) reported that
diagenesis causes significant mercury movement in sediments.
Therefore, sediment cores may not give as valid a measure of
mercury deposition as they do of less mobile metals.


2.4  Comparison of the Three Techniques

     The results of the first two techniques (Mobilization and
Enrichment Factors) predict which metals in atmospheric deposi-
tion will be affected the most by anthropogenic processes.  The
third technique (historical trends in deposition) provides an
actual measurement (primarily for areas of the eastern United
States) of the metals that will be affected.  A comparison of the
relative rankings of all three techniques (Table 1-2) shows a
basic agreement on which metals should have the greatest degree
of enrichment.  For example, the predictions of the first two
techniques, that Ag, Cd, Cu, Pb, Sb, Se, and Zn should have
increased rates of atmospheric deposition, are supported by the
analyses of temporal trends in atmospheric deposition for the
metals Ag, Cd, Cu, Pb, Sb, V, and Zn.

     It is noteworthy that the three techniques agree as well as
they do since the predictions of the first two techniques rely on
the assumption that the primary natural source in the eastern
United States is the soil.  However, since they do agree, it
appears that, in the eastern United States at least, the rates of
atmospheric deposition of Ag, Cd, Cu, Pb, Sb, V, and Zn are
strongly influenced if not controlled by anthropogenic processes.

     Unfortunately, there are many metals for which we do not
have adequate data, i.e., As, Be, Co, Cr, Hg, Mn, Mo, Ni, Sn, Te,
and Tl.  To better understand the effects of anthropogenic proc-
esses on the cycling of these metals, their temporal trends of
atmospheric concentration and deposition must be determined.
3.   CONCENTRATION OF METALS IN THE ATMOSPHERE

     The amount of a metal in either wet or dry deposition is
directly related to the concentration of the metal in the atmos-
phere.  Most metals are associated with particulate matter in the
atmosphere and the concentrations are measured by collecting a
                                27

-------
 Table  1-2.  Three Techniques for Determining the  Influence  of
          Anthropogenic Processes on the Concentration
               of Metals in Atmospheric Deposition
 Techniques
Enrichment
   factor
Historical
   factor
    Low
                       Expected Enrichment
Moderate
   Large
  Co,  Mn,
  Ni
  Co,  Mn,
  Ni
 Cr, V
 Cr, Cd,
 V
Cd, Cu, Pb,
Sb, Se, Zn
Ag, Cu, Pb,
Sb, Zn
  No Data
Mobilization
factor*

Co, Mn,
Hg

As,
Ni,
V
Cr,
Se,

Ag,
Mo,
Se,
Cd,
Pb,
Sn,
Cu,
Sb,
Zn
Be, Te, Tl


Ag, As, Be,
Hg, Mo, Sn,
Te, Tl

As, Be, Hg,
Mo, Se, Sn
Te, Tl
NOTE: Low = <2 x enrichment; moderate = 2 to 5 x; and
large = >5 x.

*The MF is based on the comparison of global emission rates.  On
a reduced scale, as for the United States, the relative order
will change; for example, Hg, As, and Se would be expected to be
in categories of higher enrichment.
known volume of air, using among others, filters, impactors, etc.
For metals and their compounds existing in the gas phase, selec-
tive absorbents can be used for collection or direct determina-
tion can be made with atomic absorption following gas chroma-
tography.

     Many metal concentrations on atmospheric particulates have
been determined and are presented in Appendix A.  The data were
classified into the following regions:
     Remote -
     Rural   -


     Urban   -
Any areas of lowest concentration.  There was  a
large discrepancy between data  from the Antarctic
and the Arctic  (Norway, Greenland, Canada), with
the latter having somewhat higher concentrations.
Both sets of data have been included in the remote
range until the discrepancy can be resolved.
Any site not subject to the direct influence of
local anthropogenic sources but representative of
a regional background.
Any site (in a city or elsewhere) subject to a
local anthropogenic source.
                               28

-------
When available, the first set of data for each element in each
region was from Eisenreich et al. (1978).  Reported concentra-
tions for many elements varied widely, perhaps because of measur-
ing errors.  In addition, very few data have been reported for
many toxic metals.  For example, no remote data were available
for Be or Sn and only Dams and deJonge (1976) reported remote
values for Ag and Mo.

     Ranges of these concentrations are summarized in Table 1-3;
median concentrations are given in Table 1-4.  For the remote
sites, obvious outliers on the high end were omitted from the
ranges in Table 1-3.

     As measurement techniques improve, a more accurate set of
median remote concentrations for trace metals in the atmosphere
will be devised.  However, for the present, these median remote
concentrations can be used to assess the effect of emission
sources (natural or anthropogenic) on metal concentrations in the
atmosphere.

     Table 1-5 gives the ratios of the median concentrations in
the rural and urban areas to the median remote concentrations for
each metal.  This concentration factor for the rural areas com-
pared with the remote areas showed that anthropogenic activity
influences the concentration of metals in the atmosphere in the
order of Zn > Pb > Mn > Cu, Ag, As > Sb, Se, Cr > Cd > Hg, Ni, V
> Co.  This order, based on a very uncritical examination of
atmospheric metal concentrations, agrees with that in Tables 1-2
and I-10, except for Mn and Cd.  The anthropogenic contribution
to the atmosphere was clearly highest for Zn and Pb, elements
with widely dispersed emission sources.

     A number of the other elements have similar rural/remote
concentration ratios within a factor of two (Ag, As, Cr, Cu, Sb,
Se).  This could indicate common sources and similar atmospheric-
transport characteristics.  Although the low position of mercury
may seem rather surprising, its high volatility may have caused
it to be more widely dispersed.  However, the state of the sci-
ence is not adequate to distinguish anthropogenic from natural
emissions of volatile elements (Hg,  Se, As).
4.   CONCENTRATION OF METALS IN WET DEPOSITION

     The regional concentration of trace metals in atmospheric
deposition cannot at present be quantitatively assessed because
of the lack of a national atmospheric-deposition program for
metals.  However, some preliminary assessments can be made by
using data from historical networks and compiling the work of
several independent investigations of metals in atmospheric
deposition in remote, rural, and urban areas.
                               29

-------
  Table 1-3.  Ranges of Metal Concentrations in the Atmosphere
Metal
Ag


As


Be


Cd


Co


Cr


Cu


Hg


Mn


Site
Remote
Rural
Urban
Remote
Rural
Urban
Rural
Urban

Remote
Rural
Urban
Remote
Rural
Urban
Remote
Rural
Urban
Remote
Rural
Urban
Remote
Rural
Urban
Remote
Rural
Urban
Range
(ng m 3 )



0.17
0.31
0.3



0.003
0.07
0.07
0.0008
0.04
0.07
0.005
0.22
0.39
0.024
0.9
0.12
0.03
0.02
0.06
0.01
1.1
1.7
0.01
0.3
1.1
- 1.47
- 12.1
- 130
0.023
0.14

- 0.62
- 1.96
- 118
- 3
- 2.0
- 43
- 2
- 65
- 1100
- 10
- 72
- 4000
- 0.208
- 4
- 11.2
- 54
- 647
- 5400
Metal Site
Mo Remote
Urban

Ni Remote
Rural
Urban
Pb Remote
Rural
Urban
Sb Remote
Rural
Urban
Se Remote
Rural
Urban
Sn Rural
Urban

V Remote
Rural
Urban
Zn Remote
Rural
Urban



Range
(ng m~3 )
0.01 -
<1 -

0.35 -
1.3 -
0.2 -
0.1 -
0.6 -
0.8 -
0.05 -
0.18 -
12 -
0.042 -
0.09 -
4.
55
99

0.0015 -
0.37 -
2.8 -
0.03 -
3.8 -
0.3 -



54
3.4

3
50
1000
64
451
8300
0.64
7
43
0.40
3.3
7



14
15
600
31
1200
3120



NOTE:  For comparative purposes,  where ranges were not available
single values were used.
                               30

-------
        Table 1-4.   Approximate Median Concentrations of
                Metals in the Atmosphere (ng m~3)
Metal
Ag
As
Be
Cd
Co
Cr
Cu
Hg*
Mn
Mo
Ni
Pb
Sb
Se
V
Zn
Remote
0
0

0
0
0
0
0
0
0
0
1
0
0
1
0
.01
.2
--
.1
.05
.3
.2
.5
.4
.3
.36
.0
.2
.1
.0
.5
Rural
0.3
6
0.023
1.0
0.1
5.0
6.0
2.0
30.0
--
2
100
3
1.5
5
100
Urban
1.1
25
0.14
2.0
10.0
40.0
100
20
150
2
30
2000
30
4.7
50
1000
*Total Hg in atmospheric measurement.
                               31

-------
Table 1-5.   Ratios of Median Concentrations of
           Metals in the Atmosphere
Metal
Ag
As
Cd
Co
Cr
Cu
Hg
Mn
Mo
Ni
Pb
Sb
Se
V
Zn
Urban/Remote
110
125
20
200
133
500
40
375
7
83
2000
150
47
50
2000
Rural/Remote
30
30
10
2
17
30
4
75
--
6
100
15
15
5
200
                      32

-------
      Data from a 32-station national  network operating  during
 parts of 1966  and 1967 (Lazrus  et al.  1970)  show distinct maxima
 in Pb and Zn deposition  east of the Mississippi  River (Figures
 1-5  and  1-6).   The high  rate was caused by the large amount of
 fossil fuel used in  the  eastern United States and transport by
 the  prevailing easterly  winds (Figure  1-4).   It  also caused the
 increased deposition of  Pb  and  Zn found in sediments of remote
 lakes in the area (Figure 1-2).

      Because of the  lack of systematically collected data for
 other metals,  the results of many different  investigations of
 metal concentrations in  wet deposition have  been stratified into
 three categories in  Appendix B:   remote,  rural,  and urban areas.
 This compilation shows how  few  data are available for most metals
 and  how  few toxic metals have ever been studied.  No studies have
 measured Be, Sb,  Sn,  Te, and Tl and fewer than fifteen  studies
 have measured  Ag,  As,  Co, Cr, Se,  and  V-
Figure 1-4.  Airflow over the United States; X refers to loca-
tions of lakes in Figures 1-2 and 1-3; thickness of airflow lines
indicates degree of prevalence (data from Van Cleef 1908).
                                33

-------
                                                         100
                                                        100
                                                        50
                                                      10
Figure 1-5.  Average deposition (g ha"1 mo"1) of Pb by precipi-
tation from September 1966 to March 1967 as determined by Lazrus
et al. (1970).  X refers to locations of lakes in Figure 1-3
(Davis and Galloway 1980, with permission of the copyright holder,
Ann Arbor Science Publishers, Inc.).
                                                        50
Figure 1-6.  Average deposition (g ha"1 mo"1) of Zn by precipi-
tation from September 1966 to March 1967 as determined by Lazrus
et al. (1970).  X refers to locations of lakes in Figure 1-3
(Davis and Galloway 1980, with permission of the copyright holder,
Ann Arbor Science Publishers, Inc.).
                                34

-------
     The ranges of metal concentrations in precipitation varied
widely within these three categories—up to four orders of magni-
tude for some metals—reflecting the different methods of samp-
ling and analyses and the many different locations.  For example:

     (1)  Observation periods range from as short as one month to
          as long as two years.
     (2)  Collection frequencies range from event sampling to
          monthly and to even yearly sampling.
     (3)  Collector designs are different.
     (4)  Samples are treated differently (e.g., filtered vs.
          unfiltered).
     (5)  Analytical methods vary considerably-
     (6)  Distinctions between remote, rural, and urban sites are
          subjective, particularly between rural and urban sites.

Because of these differences in sampling techniques, the lack of
data, and the wide variations in concentrations, it is difficult
to assign a representative concentration value for any given
metal in any given area.

     With this realization, the data-analysis approach used
previously for metal concentrations in the atmosphere was also
used for metal concentrations in wet deposition.  Ranges of metal
concentrations were calculated for remote, rural, and urban
precipitation (Table 1-6).  For the urban and rural sites, all
available data were included.  It was assumed that within an
urban environment extremely high and low concentrations are
possible because of the concentration of sources.  Most remote
studies took place at either of the two poles and Greenland.
Concentrations of metals at these locations are now thought to be
lower than previously believed because of possible sample contam-
ination.  Therefore, older data were excluded when they were not
considered representative of remote environments.

     The median concentrations of metals in wet deposition are
consistently higher for urban and lower for remote sites (Table
1-7).  However,  because of the wide concentration ranges and lack
of data (Table 1-6), caution must be exercised if these data are
used as typical of an urban, rural, or remote area.
                                35

-------
       Table 1-6.  Ranges of Metal Concentrations  in  Wet
                Deposition (Rain, Snow, and  Ice)
Metal  Site
 Range    No.  of
(|jg S,'1)    Refs.
Metal  Site
  Range   No. of
 ((jg A"1)  Refs.
Ag


As

Cd


Co


Remote
Rural
Urban
Remote
Rural
Remote
Rural
Urban
Rural
Urban

0.006-0.01
0.023-0.48
3.2
0.01
<0.1-0.5
0.004-0.639
0.08-0.5
0.5-2.3
0.01-1.5
1.8

2
7
1
1
2
7
17
3
2
1

Mn Remote
Rural
Urban
Mo Urban

Ni Remote
Rural
Urban
Pb Remote
Rural
Urban
0.018-
0.1-4
1.9-8
0.20

<0.1
0.6-2
2.4-11
0.02-.
0.6-6
5.4-14
6
19
7
1

1
9
3
7
24
6
 Cr   Rural     0.1-<5       7
      Urban    0.51-1        3

 Cu   Remote  0.035-0.851    6
      Rural     0.4-3       22
      Urban     6.8-10       5
                    Sb
                    V
      Remote
      Remote
      Rural
      Urban
    0.034
0.016-0.3
 0.13-7
    68
3
3
1
Hg


Remote
Rural
Urban
0.011-0.429
0.01-0.2
0.002-4.0
4
7
6
Zn Remote
Rural
Urban
0.001-1
1-31
200-28
8
25
6
NOTE:  For comparative purposes, where ranges were not available,
single values were used.  No data were found in the literature
for Be, Se, Sn,  Te,  or Tl; Co,  remote; Mo, remote, rural; Sb,
rural, urban; Cr,  remote.  Only bulk values were found for Se
and for As, urban.
                               36

-------
       Table 1-7.  Median Concentrations of Metals_in Wet
            Deposition (Rain, Snow, and Ice) (pg a'1 )
Metal
Ag
As
Cd
Co
Cr
Cu
Hg
Mn
Mo
Ni
Pb
Sb
V
Zn
Remote
0
0
0


0
0
0

<0
0
0
0
0
.008
.019
.008
--
--
.055
.048
.22
--
.1
.14
.034
.022
.22
Rural
0
0
0
0
0
5
0
10
-
1
15
-
1
45
.040
.5
.6
.01
.27
.3
.11
.0
-
.5

-
.1

Urban
3.2
--
0.7
1.8
3.6
30
1.0
25
0.2
17
41
--
68
40
NOTE:  If only one concentration was available,  it was used.   If
the median value was reported as a range, the middle of the range
was used.  If the median fell between two numbers, the lowest was
used.  If the median value was reported as a "less than" value,
the next lowest was used.  Insufficient data existed for Be,  Se,
Sn, Te, and Tl.
                               37

-------
     To assess how human activities have affected precipitation,
the ratios of median metal concentrations in the urban and rural
studies to those in the remote studies were calculated where
possible (Table 1-8).  In rural areas, the toxic metals most af-
fected by human activities were Zn > Pb > Cu > Cd > V > Mn > As >
Ni > Ag > Hg.  Except for Ag, this agees well with the predictions
(Table 1-2).  And except for Cd and V, it also agrees well with the
analysis of the effect of anthropogenic activities on metal
concentrations in the atmosphere in rural relative to remote
areas.

     The low ratio for Hg was almost certainly due to its wide
distribution (see above) and the inefficient scavenging of the
predominantly vapor species in precipitation.  Insufficient data
existed for Be, Co, Cr, Mo, Sb, Se, Sn, Te and Tl to make any
calculations.
   Table 1-8.  Ratios of Median Concentrations of Metals in Wet
                 Deposition (Rain, Snow, and Ice)
     Metal
Urban/Remote
Rural/Remote
Ag
As
Cd
Cu
Hg
Mn
Ni
Pb
V
Zn
400
--
87
554
21
113
170
292
3090
181
5
26
75
96
2
45
15
110
50
200
NOTE:  Insufficient data existed for Be, Co, Cr, Mo, Sb, Se, Sn,
        Te,  and Tl.
                               38

-------
5.   DEPOSITION OF METALS FROM THE ATMOSPHERE
5.1  Deposition in Remote, Rural, and Urban Environments

     Our present knowledge of the deposition rate of toxic metals
is v.ery primitive; it is comparable with our knowledge of major
cations (H, Na, K, Mg, Ca) in precipitation twenty years ago.
This is due partially to the inability, until recently, to quant-
itatively analyze constituents in water at the microgram per
liter level or less and partially to an ignorance of the poten-
tial significance of this deposition.

     Table 1-9 presents data selected from the literature avail-
able on deposition rates of toxic metals.  These data vary in
quality and are not strictly comparable for the same reasons
stated for metal concentrations in wet deposition.  In addition,
because of the lack of data and the effect that annual precipita-
tion amounts have on total deposition, a relative ordering of
these metals most influenced by anthropogenic activities is not
possible.

     However, even given these constraints certain general con-
clusions can be drawn.

     (1)  Remote:  Values reported for metals deposited in
          Antarctica were generally an order of magnitude lower
          (Boutron 1979a,b) than those reported in Greenland
          (Boutron 1979c), which in turn were an order of magni-
          tude lower than those reported from North Atlantic
          precipitation (Duce et al. 1976).
     (2)  Rural:  Values reported for metals deposited over rural
          areas were from 10 to 102 times higher than those from
          North Atlantic precipitation.
     (3)  Urban:  Values reported for metals deposited over urban
          areas were from 102 times to 104 times higher than
          those from North Atlantic precipitation and up to 106
          times higher than those from Antarctica.


5.2  The Relative Importance of Dry versus Wet Deposition

     Most data in Table 1-9 are for bulk deposition, which is a
combination of dry and wet.  In terms of our understanding of
atmospheric processes and our assessment of the effects of toxic-
metal deposition,  it is important to know the relative contribu-
tions of each.

     Actual measurements of toxic metals in deposition suggest
that the fraction deposited dry is substantial.  Few cases in the
literature have indicated that the dry fraction is less than 0.1
of the total deposit.   For the most part,  the mean dry fraction
lies between 0.3 and 0.6.   Table I-10 summarizes data for several
                                39

-------
Table 1-9.   Selected Values  of Atmospheric-Deposition Rates
       for  Metals in Remote,  Rural,  and Urban Areas
Precipitation
Ag
Remote Bulk

Rural Bulk

As
Remote Bulk
Rural Dry
Wet
Bulk


Urban Dry
Wet
Bulk

Cd
Remote Bulk


Rural Dry
Wet
Bulk
<2



Urban Dry
Wet
Bulk



Co
Rural Bulk
Cr
Remote Bulk
Rural Dry
Wet
kg ha"1 yr"1
2.1 x 10~6
2.7 x 10~5
<1 x 10~2
3.4 x 10~4

3.1 x 10~4
<3.6 x 10~3
<8.0 x 10"1
2 x 10~2
0.01 - 0.50
<10.15
<3 x 10~3
<1.23
<1 . 14
2.5-6 x 10~2

5 x 10~5
2.3 x 10~6
3.7 x 10~5
<4.7 x 10"2
<9.4 x 10"2
<0 . 1
!.4-7.7 x 10~3
4.0 X 10~3

1.2 x 10~3
8.8 x 10"3
<9.4 x 10~2
<1.7 x 10"1
<1.3 x 10"1
8.3-36 x 10"3
7 x 10~3
2.6 x 10'1

<1.0

3 x 10~3
<2.0 x 10~2
<0.57
Reference
Boutron 1979b
Ibid.
Cawse 1977
Struempler 1976

Duce et al . 1976
U.S. DOE 1979
Ibid.
Dethier 1977
Cawse 1977
U.S. DOE 1979
Ibid.
Ibid.
Ibid.
Dethier 1977

Weiss et al . 1975
Boutron 1979b
Ibid.
U.S. DOE 1979
Ibid.
Cawse 1977
Peyton et al . 1976
Vermeulen 1977;
Semb 1978
Struempler 1976
Schlesinger et al . 1974
U.S. DOE 1979
Ibid.
Ibid.
Peyton et al. 1976
Vermeulen 1977
Peirson et al . 1973

Cawse 1977

Duce et al. 1976
U.S. DOE 1979
Ibid.
                           40

-------
Table 1-9.  (continued)
Precipitation
Cr

(continued)


Bulk
kg ha"]

0

.
10




Cu

















52






Mn











Urban



Remote



Rural









Urban



Remote

Rural


Urban

Remote
Rural





Urban



Dry
Wet
Bulk

Bulk



Bulk









Bulk



Dry
Bulk



Bulk

Bulk
Dry
Wet
Bulk



Dry
Wet
Bulk
0

<
.
0
0
10






0
0



0



0
0.



7
1
1
1
1
*
f
3


^

0
7
.

21
~2
36
.4
.7
.6

.0
.9
.2
.9
.8
23
12
.6


01
5
.1
.4
16


-





X
X
X
X
X


X


-
X

X
-
• yr"1


0.





10
10
10
10
10


10


0.
10

10
0.


5





~4
~3
~5
~4
~2


~2


5
~2

~2
32
168
0

>2







<0
<0
<0

0
0
<0
<0
<1

7
3

1

6
.
,
.
2
.
*
.
,
^
.6

.0
2
.3
.2
1
.5

.2
31
25
29
.1
25
01
83
99
00


X
X
X
X
X
X

X



X

-





10
10
10
10
10
10

10



10

0.





~2
~4
~4
~3
~2
~2

~3



~2

5



Reference

Ibid.
Cawse 1977
Andren and Lindberg
U.S. DOE 1979
Ibid.
Ibid.

Weiss et al. 1975
Duce et al. 1976
Boutron 1979b
Ibid.
Struempler 1976
Andren and Lindberg
Dethier 1977
Biggs 1979 (Personal
communication to S
Norton)
Cawse 1977
Vermeulen 1977
Lazrus et al. 1970
Semb 1978
Dethier 1977
Vermeulen 1977
Lazrus et al. 1970





1977









1977


.A.









Siegel and Siegel 1978
Weiss et al. 1975

Schlesinger et al. 1974
Andren and Lindberg
Cawse 1977
Lockeretz 1974

Duce et al. 1976
U.S. DOE 1979
Ibid.
Ibid.
Struempler 1976
Andren and Lindberg
Cawse 1977
U.S. DOE 1979
Ibid.
Ibid.
1977








1977




          41

-------
                     Table 1-9.   (continued)
           Precipitation  kg ha l  yr
                                  Reference
Mo
Ni
     Rural
     Rural
     Urban
Pb
     Remote
     Rural
 Bulk
 Dry
 Wet
 Bulk
 Dry
 Wet
 Bulk
Bulk
Dry
Wet
Bulk
    Urban
Dry
Wet
Bulk
 <10
  0.09
 <0.87
 <0.23
  0.01 - 0.5
  10'2

 <2.74
 <2.24
   0.1
   9.3 x 10~3
   1.2 x 10~5
  7.85 x 10~4
 <0.72

 <2.72
   1.9 x 10~2
  0.20
 0.246
  0.16
   6.2 x 10~2
  0.32
  0.01  -  0.5
  (filtered)
 0.048  -  0.216
 0.226
     6  x  10~2
  0.17
 <2.97
:26.5
 12.36
   0.1
  0.21
  0.70
  0.89  -  2.11
  2.15
  0.37
   1.2
 Cawse  1977
U.S.DOE  1979
Ibid.
Ibid.
Cawse  1977
Lazrus et al.
U.S.DOE  1979
Ibid.
Ibid.
Lazrus et al.
                                                        1970
                                                        1970
Duce et al. 1976
Boutron 1979b
Ibid.
U.S.DOE 1979
Ibid.
Ibid.
Struempler 1976
Schlesinger et al. 1974
Andren and Lindberg 1977
Dethier 1977
Biggs 1979 (Personal com-
  munication to S. A.
  Norton)
Siccama and Smith 1978
Cawse 1977

Peyton et al. 1976
Vermeulen 1977
Lazrus et al. 1970
Semb 1978
U.S.DOE 1979
Ibid.
Ibid.
Coello et al. 1974
Dethier 1977
Dethier 1977
Peyton et al. 1976
Page and Ganje 1970
Vermeulen 1977
Lazrus et al. 1970
                              42

-------
                     Table 1-9.  (continued)
           Precipitation  kg ha'1 yr~*
                                 Reference
Sb
Se
V
     Remote    Bulk
     Rural     Bulk
     Rural
Bulk
            1.6 x 10"4
          <10"2
Duce et al. 1976
Cawse 1977
Cawse 1977
Rural Dry
Wet
Bulk


Urban Dry
Wet
Bulk

Zn
Remote Bulk


Rural Dry
Wet
Bulk








Urban Dry
Wet
Bulk




<0.25
<4.8
12.4
0.01
3.6
< 0 . 3 5
<9.04
8.3
0.043

1.1
1.0
2.5
<2
<7.92
11
4
0.50
0.06
0,5
0.29
0.82
0.2
0.4
<5.94
<11 . 9
15.8
0.35
1.57
1.9
3.6



- 0.5
x 10"2





x 10~5
x 10~3
x 10"2
.02


x 10"2

- 0.17
- 10.0
-0.97







- 4.81


U.S. DOE 1979
Ibid.
Ibid.
Cawse 1977
Vermeulen 1977
U.S. DOE 1979
Ibid.
Ibid.
Vermeulen 1977

Boutron 1979b
Ibid.
Duce et al . 1976
U.S. DOE 1979
Ibid.
Ibid.
Struempler 1976
Andren and Lindberg 1977
Dethier 1977
Cawse 1977
Peyton et al. 1976
Vermeulen 1977
Lazrus et al . 1970
Semb 1978
U.S. DOE 1979
Ibid.
Ibid.
Dethier 1977
Peyton et al . 1976
Vermeulen 1977
Lazrus et al. 1970
NOTE:  No data available on the elements Be, Sn, Te, Tl.
                               43

-------
      Table 1-10.   Means of Data Reported from All Seasons
              for Dry Fraction of Total Deposition


  Metal                  Reported Measurement of Dry Fraction
                      Marine            Rural             Urban
As
Cd
Cu
Mn
Ni
Pb
V
Zn
	
0.4
0.5
0.5
0.6
0.6
0.4
0.7
	
0.4
	
0.5
0.5
0.3
	
0.4
0.2
0.6
	
0.5
0.5
0.2
	
0.5
SOURCE:  Values for the marine environment from Duce (1979); for
rural and urban areas from Feely and Larsen (1979); data for Ag,
Be, Co, Cr, Hg, Mo, Sb, Se, Sn, Te, and Tl were not included.
toxic metals in deposition from marine, rural, and urban areas.
No systematic trend between the different environments is evident;
the median value of all of the data is 0.5.


5.3  Physical Characteristics of Metals and Their Compounds
     Affecting Atmospheric Deposition

     The physical characteristics of a metal and its compounds
(particle size, vapor pressure, solubility, heats of solution,
etc.) largely determine the type of deposition process.  Eventual-
ly knowing what species of metal is present in the atmosphere and
its physical properties, particularly the vapor pressure, particle
size, and solubility.- will enable rates of metal deposition to be
predicted.

5.3.1.  Vapor Pressure

     In contrast to organic compounds  (Part II), only a few
metals in the atmosphere are in the vapor phase, as seen from an
examination of the vapor pressures of the elements and their
common oxides (Table 1-11).  Of all the elements, only mercury (Hg)
                               44

-------
      Table 1-11.  Vapor Pressures and Boiling  Points of Metals  and
                             Their Common Oxides
Element
Ag
As

Be
Cd
Co
Cr
Cu

Hg
Mn
Mo
Ni
Pb

Sb

Se


Sn

Te
Tl

V
Zn
Boiling
point
2240
613

2970
765
2900
2482
2595

356.58
2097
5560
2732
1744

1380

685


2270

999
1457

3000
907
Vapor pressure Vapor pressure
at 500 K at 1000 K
(mm Hg) (mm Hg)
4.56 x 10~6
6.27 x 10~4 >400

3.27 x 10~8
1.51 x 10~3 466
2.77 x 10~«
1.57 x 10"10
1.03 x 10~8

44.2
1.53 x 10~5
<1.9 x 10~10
9.36 x 10~8
< 3 x 10~9 1.11 x 10"2

1.56 x 10~10 7.46 x 10'1

7.8 x 10~3 >400


5.0 x 10~8

1.65 x 10~6 45.7
1.66 x 10~10 ,1.55 x 10'1

<1 x 10"10
2.61 x 10~5 85.6
Oxide
—
As203
As205
BeO
CdO
CoO
Cr203
Cu20
CuO
HgO
MnO
MoO3
NiO
PbO
Pb02
Sb203
Sb205
SeO
Se02
Se03
SnO
Sn02
Te02
T120
T12O3
V205
ZnO
Boiling Vapor pressure
point at 500 K
(°C) (mm Hg)
	
457
(d 315)
3900
1559
>2000
4000
1800
—
(d 500)
2600
1155(s)
>2000
>600
(d 290)
1550
(d-0.380)
180
(d 340-350)
(d 180)
(d 1080600)
1800
1245
1865
875
1750
(d 1975)
—
1.9 x 10~9
6.3 x 10~7
3.9 x 10~115
3.3 x 10"37
—
--
—
—
—
—
1.62 x 10~36
1.3 x 10~72
1.58 x 10~35
--
--
--
--
1.7 x 10~7
6.35 (400
--
--
1.1 x 10~32
—
--
--
~~
Vapor pressure
at 1000 K
(mm Hg)
--
--
—
3.9 x 10~26
9.0 x 10~4
—
--
--
1.53 x 10~4
—
—
0.22
6.46 x 10~14
4.5 x 10~3
--
—
--
--
--
K)
—
--
0.10
—
—
--
~~
NOTE:  d = decomposes; s = soluble

-------
has a significant vapor pressure at ambient temperature  (1.2 x
10"3 mm Hg at 20°C),  and therefore its major deposition process
will be dry deposition from the vapor phase.  Based on vapor-
pressure data at 500  K and 1000 K, the order of decreasing atmos-
pheric concentration  of the vapor phase of the elements is:  Hg >
Se > Cd > As > Zn > Te > Sb > Tl > Pb > Mn > Ag > Ni > Sn > Be >
Co > Cu > Cr > Mo > V.

     For the oxides,  the order of decreasing vapor pressure (or
increasing boiling point) is (Se03 > SeO2) > (Ag205 > As203) >
PbO > T1203 > Mo03 >  Te02 > CdO, Sb203 > V205 > Sn02 > T120 >
ZnO > Cu20 > CoO, NiO, MnO > BeO > Cr203.  Only the Se and As
oxides have sufficient vapor pressure to have a significant
proportion of the elements (if present as oxides) existing in the
vapor phase in "clean" rural or remote air.

     Therefore, from  this analysis, only the metals Hg, Se, As,
and perhaps Cd, could be expected to have a significant vapor-
phase concentration relative to the total atmospheric concentra-
tion.

5.3.2  Particle Size

     To estimate rates of both dry deposition and washout, or
rainout, of toxic metals in particulate matter, it is necessary
to have a knowledge of the particle-size distribution.

     Particles in the atmosphere are commonly characterized by
their "mass median diameter" (mmd):  the particle size for which
50% of the mass occurs on larger and 50% on smaller particles.
Some workers in the field of aerosol physics consider the mmd an
oversimplification that essentially overlooks the fine structure
of the mass versus particle size distribution.  The urban aerosol,
for instance, is often characterized by a bimodal particle-size
distribution (e.g., Whitby et al. 1972; Whitby 1973; Bernstein
and Rahn 1979).  Rahn (1976) discussed this issue rather thor-
oughly, and concluded

          ... it seems to be generally true that an
          element is  unimodally formed in a particular
          size range  of the aerosol, then remains there.
          Thus the mmd is a reasonable single-parameter
          description of the preferred particle size of
          an element  in an aerosol.

     Table 1-12 summarizes mmd's for several toxic metals.  The
data are extremely scanty, and clearly more effort is required to
determine size distribution of metals in different atmospheres.
                               46

-------
Table 1-12.  Mass Median Diameter of Toxic Metals in Aerosols (urn)
Metal Marine air
A'g
As
Cd 0.5
Co
Cr
Cu 0.8
Hg
Mn 0.4
Mo
Ni
Pb 0.4
Sb
Se
Sn
V 0.5
Zn
New York City
<2
<2
<2
-
<2
-
<2
<2
<2
<2
<2
<2
<2
-
<2
<2
.5
.5
.5
-
.5
-
.5
.5
.5
.5
.5
.5
.5
-
.5
.5
General
(rural to urban)
1
1
2
4
3
1
0
2
1
1
0
1
0
1
1
1





.8
.5

.2
.5
.4

.8
.2

.5
SOURCE:  For marine-air values, Duce et al.  (1979); for New York
City, Bernstein and Rahn (1979); and for general (rural to ur-
ban), Rahn (1976).

NOTE:  No data were found for Be, Te, or Tl.
                                47

-------
5.3.3.  Solubility

     If we consider a hypothetical case, that atmospheric parti-
cles are either pure compounds or aerosol solutions of these,
then the efficiency of absorption of the particle into the aero-
sol, and hence into cloud-water droplets, will increase with the
solubility of the species.  The efficiency of washout should
therefore increase with the solubility of the species.

     The solubilities of the four metal salts most likely to be
present in atmospheric particulates are given in Table 1-13.  In
most cases, because the solubility of the metal species is large
relative to the amounts found in wet deposition, these compounds
will dissolve in precipitation.  However, since data on metal
speciation in the atmosphere are limited, it is difficult to say
that all metals will be found in their dissolved phase.  Indeed,
experimental data indicate that for some metals significant
amounts are insoluble.

     Also, in considering the dry deposition of pure compounds,
deposition velocity varies with particle size (see Part III).
Because of the minimum in this relationship, the most insoluble
and the most soluble of the pure compounds have higher deposition
velocities than those of intermediate solubility.  The limits of
this intermediate solubility have not yet been established.

     Clearly to use the simplified deposition theory developed in
Part III more fully and to evaluate the deposition of toxic
metals, more detailed studies must determine the composition of
the atmospheric particulates with which these metals are associ-
ated.  Only then can rates of deposition be predicted.


5.4.  Measurement of Wet Deposition

     Collection techniques of samples for the analysis of major
constituents in precipitation have been developed and stand-
ardized over the past few years (Galloway and Likens 1976, 1978;
Berry et al. 1975; Granat 1976).  Basically, these techniques use
a collector that can be easily cleaned to eliminate contamina-
tion.

     However,  in the case of trace metals,  new sampling and
analytical problems are presented in their determination in dry
deposition, rain,  and snow.   First,  concentrations are much
lower, generally ranging from 0.01 to 10 (jg 2"1, thereby requir-
ing more sensitive analytical methods and greatly increasing the
potential for significant contamination.  In addition, because of
the low concentration and the properties of ionic metal species,
adsorption of the metals onto the collector surface is a problem.
Therefore,  in any procedure  for sampling precipitation for trace-
metal analyses,  the following precautions should be considered:
                               48

-------
         Table 1-13.  Solubility of Common Oxides,  Sulfates,
             Chlorides, and Nitrates of Metals at 20°C*
Element
Ag
As

Be
Cd
Co
Cr

Cu
Hg
Mn
Mo
Ni
Pb
Sb
Se

Sn

Te
Tl

V
Zn
Species
Ag2X
AS2 X3
As2X5
BeX
CdX
CoX
CrX
Cr2X3
CuX
HgX
MnX
MoX
NiX
PbX
Sb2X3
SeX2
SeX3
SnX
SnX2
TeX2
T12X3
T12X
V2X5
ZnX
Solubilityt
Oxide
0.013
37
150016
0.000230
i
i
i
i
i
0.053
i
	
i
0.017
	
38414
vs
i
i
d
i
	
8
0.001629
(g JT1)
Sulfate
5.7
	
	
i
755
362
123.5
1200
143
0.0316
5205
	
293
0.04325
	
	
	
33025
vs
	
i
48.7
	
96525
Species
AgX
AsX3

BeX2
CdX2
CoX2
CrX2
CrX3
CuX2
HgX2
MnX2
MoX2
NiX2
PbX2
SbX3
	
	
SnX2
SnX4
TeX2
T1X
T1X3
VX2
ZnX2
Solubilityt
Chloride
0.00089
d

vs
1680
450
vs
i
706
69
723
i
642
9.9
6016
	
	
839
s
d
2 g i s.6
vs
d
432025
(q A"1)
Nitrate
1220
	

vs
1090
	
	
s
1378
vs
4264
	
2385
377
d
	
	
	
	
	
95.5
s
	
327040
*Superscripts - solubility temperatures when not 20°C.

ti = insoluble, s = soluble,  vs = very soluble,  d = decomposes.
                                49

-------
     (1)  Neither the collection vessel nor the structure  should
          have metal surfaces.
     (2)  If the collector is automated, the motor assembly
          should be enclosed.
     (3)  The collection vessel should be washed in a clean room
          designed specifically for working with solutions with
          low trace-metal concentrations.
     (4)  The collection vessel should be acid washed and well
          rinsed; blanks of the washing and rinsing processes
          should be taken.
     (5)  After the precipitation has been collected, the sample
          should be acidified in the collection vessel to remove
          absorbed metals from the walls.  This, however, will
          dissolve or desorb metals on particulate matter suspended
          in the sample.
     (6)  For the collection of precipitation to determine vola-
          tile metal species (e.g., Hg), the scavenging acid must
          be in the collection vessel before collection to prevent
          reemission.  For Hg all vessels should be of glass and
          the scavenging acid be concentrated H2SO4 plus 5%
          K2Cr204 (McLean et al. 1980).

     Few interlaboratory, quality-control studies have been
 carried out on deposition samples collected for trace-metal
 analyses.  Further evaluation studies are urgently needed.


 5.5.  Measurement of Dry Deposition

     Several different devices have been designed, usually in the
 form of buckets or funnels (e.g., Volchok and Graveson 1976,
 Galloway and Likens 1976) to separate dry from wet deposition
 (historically to study fallout from nuclear-weapon tests).

     These collectors all rely on artificial collection surfaces
 (plastic or metal),  thus creating the basic question:  how closely
 do artificial surfaces approximate natural ones such as water,
 grass, or leaves?  Artificial surfaces are not completely satis-
 factory surrogates for "real-world" surfaces, particularly for
 studying the dry deposition of small particles and gases (e.g.,
 Lundgren 1977; Slinn et al.  1979).  However, Turekian et al.
 (1977), Pattenden (1977),  and Volchok (1977; 1979, personal
 communication to S.  Eisenreich) all found that quantitative data
 from their collections on artificial surfaces are at least par-
 tially accurate.

     Another approach for determining the dry-deposition rate of
metals is to measure atmospheric concentrations; then, using
 information on the phase of the metal,  its solubility, its vapor
pressure,  and the type of receiving surface, calculate a deposi-
tion velocity.   Both techniques have unique disadvantages.  For a
better understanding of trace-metal biogeochemical cycles, these
difficulties must be resolved.
                                50

-------
6.   POTENTIAL EFFECTS OF INCREASES OF CONCENTRATION OF METALS IN
     ATMOSPHERIC DEPOSITION

     High concentrations of metals deposited into lakes and
streams, a common source of drinking water, can affect humans and
other organisms by raising the levels of these metals above the
acknowledged safety limits.  The effect on other organisms is
evident in water bodies (lakes, streams) with high metal concen-
trations due to atmospheric deposition.


6.1  Recommended Upper Limits in Water

     Table 1-14 lists the recommended upper limits that concen-
trations of metals should not exceed if man and other organisms
are not to be affected.  These are the minimum values quoted in
the literature for reducing plant growth and for killing animals
and plants.

     Literature values, commonly site-specific, seldom evaluate
such factors as the amount of biologically available toxic sub-
stance.  For example, upper limits for Be range from 11 pg 2 {
(soft water) to 1100 ng £-1 (hard water) and for cadmium,  from
0.2 ug £-1 (soft water) to 3.0 |jg 2"1 (hard water).  These ranges
reflect the control of pH on greater complexing of various metals
of higher pH (commonly associated with higher levels of dissolved
anions).  Also, organic ligands may ameliorate the effect of
dissolved metals.

     Of the metals for which there are limits, only Pb and Hg in
precipitation approach the drinking-water standard as well as the
limits for biological effects; the levels of Cd, Cu, Hg, Pb, and
Zn in precipitation approach the limits for other biological
effects (Figure 1-7).


6.2  Metal Speciation in Relation to Toxicity

     Metals in the atmosphere may be present as several species,
with different levels of toxicity.  We must, therefore, establish
the speciation and how it relates to toxicity.

     Little information on which metal species are present in the
atmosphere and in precipitation is available.  Several recent
studies on the forms of mercury in the atmosphere have found that
elemental mercury (>80%) is present although varying proportions
of the more toxic methyl mercury species (0%-50%) have also been
reported.  (For reviews and evaluation of these studies, see
Barton et al. 1980;  Brouzes et al. 1980; McLean 1980; Matheson
1979.)  Very few studies have been done on the gaseous forms of
other metals (e.g.,  As and Se).
                                51

-------
     Cd

     Mg
                   REMOTE   RURAL  URBAN
Figure 1-7.  Median concentrations of metals in precipitation
in remote, rural, and urban  areas relative to organism-toxicity
levels. (	) denotes threshold of organism toxicity.  Median
values of metals in wet deposition are adapted from the data in
Appendix B.
                              52

-------
        Table 1-14.  Recommended Upper Limits for Metal
                Concentrations in Water (pg &~l)
Metal
As
Ag
Be
Cd
Co
Cr
Cu
Hg
Mn
Ni
Pb
Se
Sn
Tl
V
Zn
Potable water
50
50
no standard
10
no standard
50
1000
2
50
no standard
50
10
no standard
no standard
no standard
5*
Organism toxicity
1.1*
no standard
11.0
0.2
0.1*
30.0
20.0
0.05
1.0*
30.0
10.0
500.0
>40.0
50.0
500.0
100.0
SOURCE:  For potable water values U.S.EPA (1976) and Torrey
(1978); for organism toxicity values, Gough et al.  (1979).

NOTE:  No standard has yet been established for Mo, Sb, or Te.

*mg S,'1
                                53

-------
     For the metals in the atmosphere associated with particulate
matter, almost no speciation has been carried out.  However, most
metals are likely to be associated with oxide, sulfide, sulfate,
or nitrate.

     Even fewer studies have been done on the species of metals
present in precipitation than on those present in the atmosphere;
however, elevated rates of methyl mercury have been found in snow
close to an emission source (McLean 1980).  To evaluate the
effects of metals in atmospheric deposition, the data base on
metal speciation must be improved.
7.   SUMMARY

     Information on nineteen metals in atmospheric deposition
that are potentially toxic to humans and other organisms has been
assimilated to determine if metal concentrations are increasing
in atmospheric deposition and if these concentrations threaten
human or organism health.

     On the basis of rates of emission, atmospheric concentra-
tions, and known temporal trends in deposition, the greatest
increase in concentration of metals in atmospheric deposition due
to anthropogenic activity are expected for Ag, Cd, Cu, Pb, Sb,
Se, Zn, with smaller increases expected for Cr and V and with
little or no increases expected for Co, Mn, and Ni.  There were
insufficient data to rank Mo, As, Be, Hg, Sn, Te, and Tl.

     Although actual data on these metals in atmospheric deposi-
tion are limited, the data available supported these expectations.
The metals Zn, Pb, Cu, Mn, Ag, As, and V had measured concentra-
tions 30 to 200 times higher in atmospheric concentration or
deposition in rural, continental areas than in remote areas such
as the South Pole.  Other metals, Sb, Se, Cr, and Ni, had concen-
trations that were 10 to 30 times greater in rural areas than in
remote areas.

     Metals can be deposited either wet or dry from the atmos-
phere.  An assessment of the relative importance of these proc-
esses revealed that, depending on the metal and the area, dry
deposition can be as great or greater a process than wet deposi-
tion.

     On the basis of analyses of vapor pressures of metals and
metal oxides,  only Hg, As, Se, and, possibly, Cd could be ex-
pected to have a significant fraction of their atmospheric con-
centration in the vapor phase.

     In regard to the effects of increased metal concentrations
in atmospheric deposition, only Pb and Hg are currently in pre-
cipitation in some areas at levels greater than the drinking-
                                54

-------
water standard.  Cd, Cu, Hg, Pb, and Zn can be present in precip-
itation at levels greater than the standards  for effects on other
organisms..
8.   RESEARCH RECOMMENDATIONS

     Although it was difficult to assimilate the work on metals
in atmospheric deposition, the job was easier because so little
has been done in spite of the scope of the problem.  To avoid
this in the future, we propose the following research recommen-
dations :

     (1)  More data must be acquired through more studies,
          including the analysis of more metals.
     (2)  The concentration of soluble versus insoluble metals in
          rain and melted snow should be determined under the
          assumption that soluble metals are probably more mobile.
     (3)  Sampling and analytical methods must be standardized
          throughout the scientific community for all metals,
          including those that take on different forms in the
          atmosphere (i.e., Hg, As).
     (4)  Since dry deposition can be just as important as, if
          not more important than, wet deposition, standardized
          collection procedures must be developed.
     (5)  A detailed study of the metal compounds that predom-
          inate in rain and snow must be conducted.
     (6)  The size distribution of metals in urban, rural, and
          remote atmospheres needs to be determined.
     (7)  The fate of metals deposited into aquatic and terres-
          trial ecosystems must be determined.
     (8)  A national network to determine the temporal and spa-
          tial trends of metals in atmospheric deposition must be
          established.
9.   REFERENCES

Andren AW, Lindberg SE, Bate LC. 1975.  Atmospheric input and
geochemical cycling of selected trace elements in the Walker
Branch Watershed.  Environ. Sci. Div. Pub. No. 728, Oak Ridge
National Lab., Oak Ridge, TN.

Andren AW, Lindberg SE. 1977.  Atmospheric input and origin of
selected elements in Walker Branch Watershed, Oak Ridge, Tennes-
see.  Water Air Soil Pollut. 8:199.

Appelquist HK, Jenson O, Sevel T. 1978.  Mercury in the Greenland
ice sheet.  Nature 273:657.
                                55

-------
Barton SC, Johnson ND, Christinson J. 1980.  A study of  atmos-
pheric mercury deposition in Ontario.  Ontario Research  Founda-
tion Report AP-2699/G-04 for the Minister of the Environment,
Province of Ontario.

Beamish RJ, Van Loon JC. 1977.  Precipitation loading of acid and
heavy metals to a small acid lake near Sudbury, Ontario.  J. Fish
Res. Board Can. 34:649.

Bernstein DM, Rahn KA. 1979.  New York summer aerosol study:
Trace element concentrations as a function of particle size.  New
York Academy of Sciences Rep., NYA9-322-1-165, The New York
Summer Aerosol Study.

Berry RL, Whelpdale DM, Wiebe HA. 1975.  An evaluation of collec-
tors for precipitation chemistry sampling.  Rep. to WMO  Expert
Meeting on Wet and Dry Deposition, Downsview, Canada.

Biggs RB, Miller JC, Otley MJ, Shields CL. 1973.  Trace  metals in
several Delaware watersheds.  Rep. from Water Res. Center,
University of Delaware, Newark.

Bogen J. 1974.  Trace elements in precipitation and cloud water
in  the area of Heidelberg measured by instrumental neutron
activation analysis.  Atmos. Environ. 8:835.

Boutron C. 1979a.  Alkali and alkaline earth enrichments in
aerosols deposited in Antarctic snows.  Atmos. Environ.  13:919.

	. 1979b.  Past and present day tropospheric fallout fluxes
of  Pb, Cd, Cu, Zn, and Ag in Antarctica and Greenland.   Geophys.
Res. Lett. 6:159.

     . 1979c.  Trace element content of Greenland snow along an
east-west transect.  Geochim. Cosmochim. Acta 43:1253.

	• 1980.  Trace metals in remote Arctic snows:  Natural or
anthropogenic.  Nature 284:575.

Boutron C, Lorius C. 1979.  Trace metals in Antarctic snows since
1914.  Nature 277:551.

Brouzes RJP, McLean RAN,  Farbas MO, McGraw SE, Tomlinson GH.
1980.  Atmospheric input of mercury to natural waters.  Atmos-
pheric Input of Pollutants to Natural Waters, Eisenreich SJ, ed.,
Ann Arbor Science Pubs.,  Ann Arbor, MI (in press).

Carr RA,  Wilkniss PE. 1973.  Mercury in the Greenland ice sheet:
Further data.  Science 181:843.

Cawse PA. 1974.   A survey of atmospheric trace elements in the
United Kingdom (1972-73).  Environmental and Medical Science
Division, AERE Harwell Rep. R 7669, MMSO, London.
                                56

-------
Cawse PA. 1977.  Deposition of trace elements from the atmosphere
in the U.K., Abstracts from Conference on Inorganic Pollution and
Agriculture, Imperial College, London.

Chan KC, Cohen BL, Frohlinger JO, Shabason L. 1976.  Pittsburgh
rainwater analysis of P.I.X.E.  Tellus 27:24.

Chester R, Stoner JH. 1974.  The distribution of Mn, Fe, Cu, Ni,
Co, Ga, Cr, V, Ba, Sr, Sn, Zn, and Pb in some soil-sized partic-
ulates from the lower troposphere over the world ocean.  Mar.
Chem. 2:157.

Coello WF, Saleem ZA, Kahn MAQ. 1974.  Ecological effects of lead
in auto exhaust.  In:  Survival in Toxic Environments, Kahn MAQ,
Bederka JP, Jr., eds., Academic Press, New York.

Cooper HBH, Lopez JA. 1976.  Chemical composition of acid preci-
pitation in central Texas.  Water Air Soil Pollut. 6:351.

Dams R, de Jonge J. 1976.  Chemical composition of Swiss aerosols
from the Jungfraujoch.  Atmos. Environ. 10:1079.

Davis AO, Galloway JN. 1980.  Atmospheric Zn and Pb deposition
into lakes of the eastern United States.  In:  Atmospheric Input
of Pollutants to Natural Waters, Eisenreich SJ, ed., Ann Arbor
Science Pubs., Ann Arbor, MI (in press).

Dethier DP- 1977.  Precipitation chemistry at an isolated site in
the North Cascade range.  Am. Geophys. Union Trans. 58:386.

Dickson W. 1975.  The acidification of Swedish lakes.  Rep. 54,
Institute of Freshwater Research, Drottningholm, Sweden.

Drozdova VM, Makhonko EP. 1970.  Content of trace elements in
precipitation.  J. Geophys. Res. 75:3610.

Duce RA. 1979.  Pigeon key results.  SEAREX Newsletter 2(4).

Duce RA, Hoffman GL, Zoller WH. 1975.  Atmospheric trace metals
at remote northern and southern hemispheric sites:  Pollution or
natural.  Science 187:59-61 (10 January 1975).

Duce RA, Hoffman GL, Ray BJ, et al. 1976.  Trace metals in the
marine atmosphere:  Sources and fluxes.  In:  Marine Pollutant
Transfer, Windom H,  Duce R, eds., Heath & Co., Lexington, MA.

Eisenreich SJ. 1980.  Atmospheric input of trace metals to Lake
Michigan.  Water Air Soil Pollut. 13 (in press).

Eisenreich SJ, Hollod GL,  Langevin S. 1978.   Precipitation
chemistry and atmospheric deposition of trace elements in north-
eastern  Minnesota.   Rep.  to Minnesota Env.  Quality Council.
                                57

-------
Elgmork K, Hagen A,  Langeland A. 1973.  Polluted snow in southern
Norway during the winters 1968-1971.  Environ. Pollut. 4:41.

Evendijk JE. 1977.  Trace metal content of atmospheric dust in
Rotterdam.  Paper presented at Fourth Annual Clear Air Congress,
Tokyo.

Feely H, Larsen RJ.  1979.  The chemical composition of atmos-
pheric deposition.  Environmental Measurements Lab Quarterly EML-
363, Appendix.

Feely HW, Toonkel LE, Larsen RJ. 1979.  Radio nuclides and trace
metals in surface air.  Environmental Measurements Lab Quarterly
EML-367, Cl.

Forland EJ, Gjessing YT. 1975.  Snow contamination from washout/
rainout and dry deposition.  Atmos.  Environ. 9:339.

Galloway JN, Likens GE. 1976.  Calibration of collection proce-
dures for the determination of precipitation chemistry.  Water
Air Soil Pollut. 6:241.

	. 1978.  The collection of precipitation for chemical
analysis.  Tellus 30:71.

     . 1979.  Atmospheric enhancement of metal deposition in
Adirondack Lake sediment.  Limnol.  Oceanogr. 24:427.

Galloway JN, Whelpdale DM. 1980.   An atmospheric sulfur budget
for eastern North America.  Atmos.  Environ. 14:409.

Gatz DF. 1975.  Pollutant aerosol deposition into southern Lake
Michigan.  Water Air Soil Pollut. 5:239.

Gjessing YT, Gjessing ET. 1973.   Chemical composition of an
Antarctic snow profile.  Vatten 3:233.

Gough LP, Shacklette HT,  Case AA. 1979.  Element concentrations
toxic to plants, animals  and man.  U.S. Geol. Surv. Bull. 1466:80

Granat J. 1976.  Principles in network design for precipitation
chemistry measurements.  J. Great Lakes Res. 2:42.

Hanssen JE, Ramback JP, Semb A,  Steinues E. 1980.  Atmospheric
deposition of trace elements in Norway.  International Conference
of the Ecological Impact  of Acid Precipitation.  Sandefjord,
Norway,  March 11-14 (in press).

Harrison RM, Perry R,  Wellings RA.  1975.  Lead and cadmium in
precipitation:  Their contribution to pollution.  J. Air Pollut.
Control Assoc. 25:627.
                               58

-------
Henriksen A. 1972.  Quantitative chemical analysis of snow.
Vatten 5:409.

Herron MM, Langway CC, Weiss HV, Cragin J. 1977.  Atmospheric
trace metals and sulfate in the Greenland ice sheet.  Geochim.
Cosmochim. Acta 41:415.

Herron MM. Langway CC, Weiss HV, Huxey P, Kerr R, Cragin J. 1976.
Vanadium and other elements in Greenland ice cores.  CRREL Rep.
76-24.

Hovmand MF. 1976.  Deposition of metals from the atmosphere in
Denmark.  Institut for Okologisk Botanik, Copenhagen University.

Johannessen M, Henriksen A. 1978.  Chemistry of snow meltwater:
Changes in concentration during melting.  Water Resour. Res.
14:615.

Jonasson IR. 1973.  Migration of trace metals in snow.  Nature
241:447.

Kennedy VC, Zellweger GW, Avanzino RJ. 1979.  Variation of rain
chemistry during storms at two sites in northern California.
Water Resour. Res. 15:687.

King RB, Fordyce JS, Antoine AC, Leibecke HF, Neustadter HE,
Sidik SM. 1976.  Elemental composition of airborne particulates
and source identification—An extensive one-year survey.  J. Air
Pollut. Control Assoc. 26:1073.

Konasewich D, Traversy W, Zar H. 1978.  Organic and heavy metal
contaminants in the Lakes Erie, Michigan, Huron and Superior
basins.  Status Report, Great Lakes Water Quality Board.

Kramer JR. 1973.  Fate of atmospheric S02 (and related substances)
scavenged by precipitation. , MS thesis, Dept. of Geology,
McMaster University, Hamilton, Ontario.

Krupa SV, Coscio MR, Jr, Wood FA. 1976.  Evidence for multiple
hydrogen-ion donor systems in rain.  Water Air Soil Pollut.
6:415.

Lantzy RJ. Mackenzie FT. 1979.  Atmospheric trace metals:  Global
cycles and assessment of man's impact.  Geochim. Cosmochim. Acta
43:511.

Lazrus AL, Lorange E, Lodge JP, Jr. 1970.  Lead and other metal
ions in United States precipitation.  Environ. Sci. Technol.
4:55.

Likens GE. 1972.  The chemistry and precipitation in the central
Finger Lakes region.  Tech. Rep. 50, Water Resources and Marine
Sciences Center, Cornell University, Ithaca, NY.
                                59

-------
Liljestrand HM, Morgan JJ.' 1978.  Chemical composition  of  acid
precipitation in Pasadena, CA.  Environ. Sci. Technol.  12:1271.

Lockeretz W. 1974.  Deposition of airborne mercury near point
sources.  Water Air Soil Pollut. 3:179.

Lundgren DA 1977.  Remarks during Expert Meeting on Dry Deposi-
tion.  Gothenburg, Sweden.  (Available from WMO, Geneva.)

Martens CS, Harriss RC.  1973.  Chemistry of aerosols, cloud
droplets and rain in the Puerto Rican marine atmosphere.   J.
Geophys. Res. 78:949.

Matheson DH. 1979.  Mercury in the atmosphere and in precipita-
tion.  In:  Biogeochemistry of Mercury in the Environment, Nriagu
JO, ed., Elsevier-North Holland, NY.

Matsunaga K. 1978.  Mercury in sediments from the Okhotsk  Sea and
Funka Bay-  Geochem. J.  12:287.

Matsunaga K, Goto T. 1976.  Mercury in the air and precipitation.
Geochem. J. 10:107.

McCall JG, Bush DS. 1978.  Precipitation and throughfall chemis-
try in the San Francisco Bay area.  J. Environ. Qual. 7:352.

McLean RAN. 1980.  Methyl mercury transport in the environment:
The role of atmospheric precipitation.  Nature (in press).

McLean RAN, Findlay DM,  Farkas MO. 1980.  Mercury transport in
the environment—sampling and analysis problems.  In:   Polluted
Rain, Taribara TY, et al., eds., Plenum, NY.

Moore DI, Gosz JR, White CS.  1978.  Impact of a ski basin  on a
mountain watershed.  Water Air Soil Pollut. 10:81.

Murozumi M, Chow TJ, Patterson C. 1969.  Chemical concentrations
of pollutant lead aerosols, terrestrial dusts and sea salts in
Greenland and Antarctic snow strata.  Geochim. Cosmochim.  Acta
33:1247.

National Academy of Sciences. 1978.  An Assessment of Mercury in
the Environment.  Washington, D.C.

Norton SA, Dubiel RF, Sasseville DR, Davis RB. 1978.  Paleolimnol-
ogic evidence for increased Zn loading in lakes of New  England.
Verh. Int. Ver. Theor.  Angew Limnol. (Eng.) 20.

Norton SA, Hess CA, Davis RB. 1980.  Rates of accumulation of
heavy metals in pre- and post-European sediments in New England
lakes.   In:  Atmospheric Input of Pollutants to Natural Waters,
Eisenreich SJ,  ed., Ann Arbor Science Pubs., Ann Arbor, MI (in
press).
                                60

-------
Nriagu JO. 1979.  Global inventory of natural and anthropogenic
emissions of trace metals to the atmosphere.  Nature 279:409.

Page AL, Ganje TO. 1979.  Accumulations of lead in motor vehicle
traffic density-  Environ. Sci. Technol 4:140.

Pattenden NJ. 1974.  Atmospheric concentration and deposition
rates of some trace elements measured in the Swansea/Neath/ Port/
Talbot area.  AERE Harwell Rep. R-7729, Her Majesty's Stationery
Office, London.

	. 1977.  Remarks during Expert Meeting on Dry Deposition.
Gothenburg, Sweden.  (Available from WMO, Geneva).

Peirson DH, Cawse PA, Salmon L, Cambray RS. 1973.  Trace elements
in the atmospheric environment.  Nature 241:252.

Peyton T, Mclntosh A, Anderson V, Yost K. 1976.  Aerial input of
heavy metals into an aquatic ecosystem.  Water Air Soil Pollut.
5:443-451.

Rahn KA. 1976.  The chemical composition of the atmospheric
aerosol.  Technol. Rep., Graduate School of Oceanography, Univer-
sity of Rhode Island, Kingston.

Ronneau C, Cara J, Navarre JL, Priest P. 1978.  An automatic
sequential rain sample.  Water Air Soil Pollut. 9:171.

Ruppert H. 1975.  Geochemical investigations on atmospheric pre-
cipitation in a medium-sized city (Gottingen, FRG).  Water Air
Soil Pollut. 4:447.

Salmon L, Atkins DHF, Fisher EMR, Healy C, Law DV. 1977.  Retro-
spective trend analysis of the content of United Kingdom air
particulate material 1957-1974.  AERE Harwell Rep. 8680, Her
Majesty's Stationery Office, London.

Schlesinger WH, Reiners WA, Knopman DS. 1974.  Heavy metal con-
centration and deposition in bulk precipitation in mountain
ecosystems of New Hampshire, USA. Environ. Pollut. 6:39.

Schaefer VJ, Fuquay JJ. 1965.  The detection of silver iodide in
snow by neutron activation analysis.  J. Rech. Atmos. 2:49.

Semb A. 1978.  Deposition of trace elements from the atmosphere
in Norway.  Norwegian SNSF Project Res. Rep. FR-13178, Oslo.

Shiomi MT, Kuntz KW.  1973.  Great Lakes precipitation chemistry:
Part I, Lake Ontario basin.  Sixteenth Great Lakes Res. Confer-
ence .

Siccama TG, Smith WH. 1978.  Lead accumulation in a northern
hardwood forest.  Environ. Sci. Technol. 12:593.
                                61

-------
Siegel SM, Siegel BZ. 1978.  Mercury fallout in Hawaii.  Water
Air Soil Pollut. 9:113.

Slinn WGN, Katen PC,  Wolf MA, et al.  1979.  Wet and dry deposi-
tion and resuspension of AFCT/TFCT fuel processing radionuclides.
Report Corvallis Air Resources Center,  Oregon State University,
SR-0980-10.

Soderlund R. 1975.  Some preliminary views of atmospheric trans-
port of matter to the Baltic Sea.  Rep. AC-31 MISU-IMI, Arrhenius
Laboratory, Stockholm.

Struempler AW. 1976.   Trace metals in rain and snow during 1973
at Chadron, Nebraska.  Atmos. Environ.  10:33.

Sugawara K, Okabe S,  Tamaka M. 1961.   Geochemistry of molybdenum
in natural waters.  (Eng.) Chem. Abstr. 57:590.

Tanner TM, Rancitelli LA, Haller WA.  1972.  Multielement analysis
of natural waters by neutron activation, assays, chemical separa-
tion, and gamma ray spectrometric techniques.  Water Air Soil
Pollut. 1:132.

Thornton JD, Eisenreich SJ, Munger JW,  Gorham E. 1980.  Trace
metal and acid composition of rain and snow in northern Minnesota
In:  Atmospheric Input of Pollutants to Natural Waters, Eisen-
reich SJ, ed., Ann Arbor Science Pubs., Ann Arbor, MI (in press).

Torrey S, ed. 1978.  Trace contaminants from coal.  Noyes Data
Corporation, Park Ridge, NJ.

Turekian KK, Nozaki J, Benninger LK.  1977.  Geochemistry of
atmospheric radon and radon products, Volume 5 of Annu. Rev-
Earth and Planet. Sci.,  Donath RA, ed., Annual Reviews, Inc.,
Palo Alto, CA.

U.S.DOE. 1979.  U.S.  Department of Energy.  Environmental
Measurements Laboratory Quarterly (EML-363).

U.S.EPA. 1976.  U.S.  Environmental Protection Agency.  National
Interim Primary Drinking Water Regulations.  EPA 570/9-76-003.

Van Cleef J. 1908.  Is there a type of storm path.  Mon. Weather
Rev. 36:56.

Vermeulen AJ. 1977.  Immieeieonderzoek met behulp van regen-
vangers, opzet,  ervaringen en resultaten.   In:   Dienst voor de
Milieuhydiene, Prov.  Waterstaat van Noord-Holland, Haarlem.

Volchok HL. 1977.  Remarks during Expert Meeting on Dry Deposi-
tion.  Gothenburg,  Sweden.   (Available from WMO, Geneva).
                                62

-------
Volchok HL, Bogen J. 1973.  Atmospheric trace metal studies.  In:
Chemical Analysis of the Environment, Ahuja S, ed. ,  Plenum, New
York.

Volchok HL, Graveson RT. 1976.  Wet/dry fallout collection.
Second Federal Conference on the Great Lakes, ICMSE Federal
Council for Science and Technology, Great Lakes Commission.

Warburton JA, Young LG. 1972.  Determination of silver in precip-
itation down to 10 11m concentrations by ion exchange and neutron
activation analysis.  Anal. Chem. 44-2043.

Weiss H, Bertine K, Koide M, Goldberg EP. 1975.  The chemical
composition of a Greenland glacier.  Geochim. Cosmochim. Acta
39:1.

Weiss H, Herron MM, Langway CC. 1978.  Natural enrichment of
elements in snow.  Nature 274:352.

Whitby KT. 1973.  On the multimodal nature of atmospheric aerosol
size distributions.  Paper presented at Seventh International
Conference on Nucleation, Leningrad.

Whitby KT.   Husar RB, Liu BYH. 1972.  The aerosol size distribu-
tion of Los Angeles smog.  J. Colloid Interface Sci. 39:177.

Wiener JG. 1979.  Aerial inputs of Cd, Cu, Pb and Mn into a
freshwater pond in the vicinity of a coal-fired power plant.
Water Air Soil Pollut. 12:343.

Wisniewski J. 1975.  Variability of rainwater silver concentra-
tion in south Florida.  Florida Area Cumulus Experiment unpub-
lished observations, Environmental Research Laboratories, NOAA,
Boulder.

Wood JM. 1974.  Biological cycles for toxic elements in the
environment.  Science 183:1049.

Zitko V, Carson WV. 1971.  Heavy metals in the precipitation in
the vicinity of St. Andrews, New Brunswick.  Report Series 1129,
Biological Station, Fisheries Res. Board,  St. Andrews,  Canada.

Zoller WH, Gladney ES, Duce RA. 1974.  Atmospheric concentrations
and sources of trace metals at the South Pole.  Science 183:198.
                                63

-------
                       Part I - APPENDIX A

    Concentrations of Metals in the Atmosphere:  Selected Data


     The data presented here were classified into the following
regions:

     Remote -  Any areas of lowest concentration.  There is a
               large discrepancy between data from the Antarctic
               and the Arctic (Norway,  Greenland, Canada) with the
               latter having somewhat higher concentrations.
               Both sets of data have been included in the remote
               range until the discrepancy can be resolved.
     Rural  -  Any site not subject to the direct influence of
               local sources but representative of a regional
               background.
     Urban  -  Any site (in a city or elsewhere) subject to a
               local source.

     When available,  the first set of data for each element in
each region is from Eisenreich et al.  (1978).  Reported concentra-
tions for many elements varied widely,  perhaps because of measuring
errors.  In addition,  very few data have been reported for many
toxic metals.  For example,  no remote data were available for Be
or Sn, and only Dams and de Jonge (1976) reported remote values
for Ag and Mo.

     All  citations are included in the Reference list at the end
of Part I.  Ranges of concentration are summarized in Table 1-3;
median concentrations are in Table 1-4;  ratios of the median
concentrations in the rural and urban areas to the median remote
concentration for each metal are in Table 1-5.
                               64

-------
           Concentration of Metals in the Atmosphere
                          A Selection

Ag
Remote
Rural
Urban
As
Remote

Rural


Urban




Be
Rural
Urban
Cd
Remote
ng m~3

0.01
0.3
1.1

0.17-1.47
0.23
0.31-2.5
4-10
12.1
0.3-130

17.4
17.5-26
5.0

0.023
0.14

0.0059-0.28
Location

Switzerland
England
Ohio

USA, Europe
Switzerland
Europe, Canada
England
Ohio
Canada, Europe,
USA, N. Sea
Ohio
Norway
Netherlands

Ohio
Ohio

Europe, USSR,
Reference

Dams and de Jonge
Salmon et al. 1977
King et al. 1976

Eisenreich et al .
Dams and de Jonge
Eisenreich et al.
Salmon et al. 1977
King et al. 1976
Eisenreich et al .

King et al. 1976
Semb 1978
Evendijk 1977

King et al. 1976
Ibid.

Eisenreich et al.


1976



1978
1976
1978


1978








1978
Rural
Urban
             0.5
         0.003-0.62
0.124-1.96

    1.55
 0.07-0.33

 0.07-118
             3.9
             2.8
  Indian Ocean (S)
Switzerland
N. Atlantic,
  S. Pole

Europe, Indian
  Ocean (N)
Ohio
Norway

Europe, USSR, USA,
  Japan, Bermuda,
  N. Atlantic
Ohio
Netherlands
                                 Dams and de Jonge 1976
                                 Duce et al. 1975
Eisenreich et al.  1978

King et al.  1976
Semb 1978

Eisenreich et al.  1978
                                 King et al.  1976
                                 Evendijk 1977
                              65

-------
          Metal  Concentrations:   Atmosphere (continued)
             ng m
                           Location
                                       Reference
Co
 Remote
 Rural
 Urban
Cr
 Remote
 Rural
0.013-0.26

    0.00084
    0.045
0.006-0.09

0.042-0.061
  0.6-3


 0.04-0.245
  0.2-2.0
    0.75
 0.42-0.214

 0.07-43
              2.6
              3.5
             36
 0.11-1.29
              0.0053
              0.36
           0.07-1.1

           0.22-0.35
0.618-65
              5-10
              3.4
           0.22-1.30
Europe, S. Pole,
  Gulf of Guinea
S. Pole
Switzerland
N. Atlantic,
  S. Pole
Norway
Greenland, Easter
  Islands

Europe, Canada,
  Ivory Coast,
  Atlantic Coast,
England
Ohio
Norway

Europe, USA, Japan,
  Hawaii, Ivory
  Coast, N. Sea, N.
  Atlantic, Sudan
Ohio
Netherlands
IML sampling sites
  in the Americas
  (50°N-53°S)


Europe, Canada,
  S. Pole, Gulf
  of Guinea
S. Pole
Switzerland
N. Atlantic,
  S. Pole
Norway
Greenland, Easter
  Islands

Canada, Ivory
  Coast, Europe,
  USSR, Atlantic
  Ocean
England
Ohio
Norway
Eisenreich et al. 1978

Zoller et al. 1974
Dams and de Jonge 1976
Duce et al. 1975

Semb 1978
Feely et al. 1979


Eisenreich et al. 1978
Salmon et al. 1977
King et al. 1976
Semb 1978

Eisenreich et al. 1978
                                 King et al.  1976
                                 Evendijk 1977
                                 Feely et al. 1979
Eisenreich et al. 1978
Zoller et al. 1974
Dams and de Jonge 1976
Duce et al. 1975

Semb 1978
Feely et al. 1979


Eisenreich et al. 1978
                                 Salmon et al.  1977
                                 King et al.  1976
                                 Semb 1978
                               66

-------
          Metal Concentrations:  Atmosphere (continued)
             ng m
                Location
                          Reference
Cr (continued)
 Urban     0.39-1100
Cu
 Remote
 Rural
 Urban
Eg
 Remote
 Rural
 Urban
  19
  16


0.04-2.1
          0.029-0.036
              0.024
           0.12-10

            1.6-2.4
           <0.6-2
 0.9-12
  72
 2.2-13.9

0.12-4000
            130
            255
            343
   0.208
   0.03

0.02-0.428
 0.5-4
   0.223

0.06-11.2
USA, Europe, Japan, Eisenreich et al
  Bute, Hawaii, Ber-
  muda, N. Sea, N.
  Atlantic, Atlantic
  Ocean, Sudan
Ohio                King et al. 1976
Netherlands         Evendijk 1977
                                                  1978
Europe, Hawaii,     Eisenreich et al. 1978
  USSR, Indian Ocean
  (S), S. Pole
                    Zoller et al. 1974
                    Dams and de Jonge 1976
                    Duce et al. 1975
S. Pole
Switzerland
N. Atlantic,
  S. Pole
Norway
Greenland, Easter
  Islands

Europe, Canada,
  USSR, Indian
  Ocean (N)
Ohio
Norway
                    Semb 1978
                    Feely et al. 1979


                    Eisenreich et al. 1978
                                           King et al. 1976
                                           Semb 1978
USA, Europe, USSR,  Eisenreich et al. 1978
  Japan, Canada, Ber-
  muda, N. Atlantic,
  Atlantic Ocean,
  Tasmania, Sudan
Ohio                King et al. 1976
Netherlands         Evendijk 1977
  IML sampling      Feely et al. 1979
    sites in the
    Americas
    (50°N-53°S)
Gulf of Guinea
Switzerland
                    Eisenreich et al. 1978
                    Dams and de Jonge 1976
Europe, Ivory Coast Eisenreich et al. 1978
England             Salmon et al. 1977
Ohio                King et al. 1976

USA, Europe, Ivory  Eisenreich et al. 1978
  Coast, N. Atlantic
  Sudan
                                67

-------
          Metal Concentrations:  Atmosphere (continued)
             ng m
                           Location
                                     Reference
Mn
 Remote
 Rural
 Urban
Mo
 Remote

 Urban
Ni
 Remote
 Rural
0.2-4.5
              0.0103
              1.5
           0.05-54

            1.4-3.5
1.1-647
 15-60
 66.3
1.4-20.7

1.7-5400
            148
             80
            269
   0.3


   3.4


0.35-0.38


 0.6-3



 1.3-4.3

  10-50
Europe, Canada,     Eisenreich et al. 1978
  USSR, Indian Ocean
  (S), S. Pole,
  Atlantic Ocean
                    Zoller et al. 1974
                    Dams and de Jonge 1976
                    Duce et al. 1975
           S.  Pole
           Switzerland
           N.  Atlantic,
             S.  Pole
           Norway
           Greenland,  Easter
             Islands
                    Semb 1978
                    Feely et al. 1979
Canada, Europe,     Eisenreich et al. 1978
  Ivory Coast, USSR,
  Indian Ocean (N),
  Atlantic Ocean
England             Salmon et al. 1977
Ohio                King et al. 1976
Norway              Semb 1978
USA, Europe, USSR,
  Bute, Hawaii,
  Bermuda, N. Sea,
  Atlantic Ocean,
  Sudan
Ohio
Netherlands
IML sampling sites
  in the Americas
  (50°N-53°S)
           Switzerland

           Ohio
           Netherlands
                                Eisenreich et al. 1978
                               King et al.  1976
                               Evendijk 1977
                               Feely et al.  1979
                    Dams and de Jonge 1976

                    King et al. 1976
                    Evendijk 1977
           Europe,  USSR,        Eisenreich et al. 1978
             Indian Ocean (S),
             S.  Pole
           Greenland,           Feely et al. 1979
             Easter Islands

           Europe,  USSR,        Eisenreich et al. 1978
             Indian Ocean (N)
            England            Salmon et al. 1977
                               68

-------
          Metal Concentrations:  Atmosphere (continued)
             ng m
                Location
                          Reference
Ni (continued)
 Urban      0.2-1000
              0.3
             35
Pb
 Remote
 Rural
 Urban
Sb
 Remote
0.23-38
   0.20
   4.4
0.10-64

 0.6-10

 0.6-64
 100-200
 451
 9.4-23.6

 0.8-8300
            759
           4150
   0.2
0.05-0.64

0.18-0.35
            Europe, USSR, USA,  Eisenreich et al.  1978
              Japan, Bute,
              Hawaii, N. Sea,
            Netherlands         Evendijk 1977
            IML sampling sites  Feely et al.  1979
              (50°N-53°S)
Europe, Canada,
  Hawaii, USSR,
  Indian Ocean (S),
  S. Pole,
  Atlantic Ocean
S. Pole
Switzerland
N. Atlantic,
  S. Pole
Greenland,
  Easter Islands
Europe, USSR,
  Indian Ocean (N),
  Atlantic Ocean
England
Ohio
Norway
Eisenreich et al. 1978
Zoller et al. 1974
Dams and de Jonge 1976
Duce et al. 1975

Feely et al.  1979

Eisenreich et al. 1978
                                           Salmon et al. 1977
                                           King et al. 1976
                                           Semb 1978
USA, Europe, USSR,  Eisenreich et al. 1978
  Japan, Bute,
  Hawaii, Bermuda,
  N. Sea, Sudan
  Tasmania,
  Atlantic Ocean
Ohio                King et al. 1976
IML sampling sites  Feely et al. 1979
  in the Americas
  (50°N-53°S)
Switzerland
N. Atlantic,
  S. Pole
Norway
Dams and de Jonge 1976
Duce et al. 1975

Semb 1978
                                69

-------
Metal Concentrations:   Atmosphere (continued)
ng m~3
Sb (continued)
Rural 1.5-7
6.29
0.18-1.01
Urban 43
12
Se
Remote 0 . 042
0.09-0.40

0.09-0.17
Rural 1.2-2.2
3.3
0.09-0.66
Urban 4 . 7
Sn
Rural 55
Urban 99
V
Remote 0.0015-2.15

0.0015
0.29
0.06-14


-------
          Metal Concentrations:  Atmosphere (continued)
             ng m
                Location
                          Reference
Zn
 Remote
 Rural
 Urban
0.03-31
   0.030
   9.9
 0.3-27
 4.7-7.8
<0.8-3
 3.8-90.7
 100-1200
 264
 5.8-45

 0.3-3120
            413
            350
           2800
Europe, S. Pole
S. Pole
Switzerland
N. Atlantic
Norway
Greenland, Easter
  Islands

Europe, Canada
England
Ohio
Norway
Eisenreich et al.  1978
Zoller et al.  1974
Dams and de Jonge 1976
Duce et al. 1975
Semb 1978
Feely et al. 1979
Eisenreich et al.  1978
Salmon et al.  1977
King et al.  1976
Semb 1978
USA, Europe, Japan, Eisenreich et al.  1978
  Bute, Hawaii, Can-
  ada, Bermuda, N.
  Sea, N. Atlantic,
  Atlantic Ocean,
  Tasmania, Sudan
Ohio                King et al. 1976
Netherlands         Evendijk 1977
IML sampling sites  Feely et al.  1979
  in the Americas
  (50°N-53°S)
                                71

-------
                       Part I - APPENDIX B

           Concentrations of Metals in Wet Deposition

     Because of the lack of systematically collected data for
other metals, the results of many different investigations of the
metal concentrations in wet deposition have been divided into
categories for remote,  rural, and urban areas.  This compilation
shows how few data are available for most metals and how few
toxic metals have ever been studied.  No studies have measured Be,
Sn, Te, and Tl and less than fifteen studies have measured Ag, ,
As, Co, Cr, Sb, Se, and V-   All citations are included in the
reference list at the end of Part I.

     The ranges of metal concentrations in precipitation vary
widely within these three categories—up to four orders of magni-
tude for some metals--reflecting the different methods of sampling
and analyses and the many different locations.  For example:

     (1)  Observation periods range from as short as one month to
          as long as two years;
     (2)  Collection frequencies range from event sampling to
          monthly and to even yearly sampling;
     (3)  Collector designs are different;
     (4)  Samples are treated differently (e.g., filtered vs.
          unfiltered);
     (5)  Analytical methods vary considerably;
     (6)  Distinctions between remote,  rural,  and urban sites are
          subjective, particularly between rural and urban
          sites.

Because of these differences in sampling techniques, the lack of
data, and the wide variations in concentrations, it is difficult
to assign a representative  concentration value for any given
metal in any given area.

     Ranges of metal concentrations are summarized in Table 1-6;
the median concentrations are in Table 1-7;  the ratios of median
metal concentrations in the urban and rural areas to those in the
remote areas are in Table 1-8.
                               72

-------
           Concentrations of Metals in Wet Deposition
          Type
                 Precipitation
                    Location
                                        Reference
Remote  Snow
                  0.006
                  0.010
                  0.008
Rural
Rain
          0.48
          0.01
Rain/snow 0.084
Snow      0.023
          0.073
          0.040
          0.036
Urban   Rain
          3.2
                 Antarctica
                 Greenland
South Florida
Washington
Nebraska
Lake Erie
South Illinois
Colorado
Montana

Heidelberg
As

Remote
Ice/snow  0.019  Greenland
Rural   Bulk
        Rain
Urban   Bulk
          3.0    Lake Erie
          1.0    Lake Huron
          1.6    Uraymines (UK)
          9.6    Swansea (UK)
          3.9    Great Britain
          0.5    Washington
          :0.1    Washington
          0.5    N. Norway

          7.0    Seattle
                                 Boutron & Lorius
                                 Boutron 1979c
                                 Ibid.
                                 1979
                                         Wisniewski 1975
                                         Tanner et al.  1972
                                         Struempler 1976
                                         Warburton & Young 1972
                                         Gatz 1975
                                         Schaefer & Fuquay 1965
                                         Warburton & Young 1972

                                         Bogen 1974
                                 Weiss et al.  1975

                                 Konasewich et al.  1978
                                 Ibid.
                                 Peirson et al.  1973
                                 Pattenden 1974
                                 Cawse 1974
                                 Dethier 1977
                                 Tanner et al. 1972
                                 Hanssen et al.  1980

                                 Larson 1977
Cd

Remote
Rain
Snow
        Ice/snow
Rural   Bulk
          1.4
          0.004
          0.007
          0.005
          0.011
          0.008
          0.639
          0.5
          0.22
                  1
                  2
                  2
                  0.6
NWT (Canada)
Antarctica
                 Greenland

                 Greenland

                 South Carolina
                 N. Norway
                 Lake Ontario
                 Lake Erie
                 Lake Huron
                 Lake Michigan-S
                                         Kramer 1973
                                         Gjessing & Gjessing 1973
                                         Boutron & Lorius 1979
                                         Boutron 1979a
                                         Boutron 1979b
                                         Ibid.
                                         Herron et al.  1977
                                         Weiss et al. 1975
                                         Wiener 1979
                                         Hanssen et al.  1980
                                         Shiomi & Kuntz 1973
                                         Konasewich et al. 1978
                                         Ibid.
                                         Eisenreich 1980
                                73

-------
        Metal Concentrations:  Wet Deposition (continued)
          Type
                 Precipitation
                    Location
                                                Reference
 Cd  (continued)
                  0.4    Lake Michigan-N
                <17.7    Uraymines (UK)
                 18      Great Britain
                  0.6    New Hampshire
               0.18-0.30 NE. Minnesota
Urban
Rain     46.0
          1.0
         10.5
          4.0
          7.0
          0.3
          0.27
          0.20
          0.6
         <2.0
        0.4-80
          0.6
Rain/snow 0.31
          0.15
          0.18
          0.73
Snow      3 .4

          2.5
          0.08

Bulk     19.0
Rain      0.5
Rain/snow 2.3
Snow      0.7
Co
Rural
Bulk
Urban
Rain

Rain
7.0
2.0
                  0.25
                  3.6
1.5
0.01
1.8
                         Lake Erie
                         Lake Superior
                         Tennessee
                         Belgium
                         Sweden

                         Norway
                         Denmark
                         North Holland
                         Twin Cities
                         Delaware
                         New Brunswick
                         Nebraska
                         NE. Minnesota
                         N. Minnesota
                         SW.North Dakota
                         Norway
                         Alaska
                       Ibid.
                       Peirson et al. 1973
                       Cawse 1974
                       Schlesinger et al. 1974
                       Eisenreich et al. 1978
                       Konasewich et al. 1978
                       Eisenreich et al. 1978
                       Andren et al.  1975
                       Ronneau et al. 1978
                       Dickson 1975
                       Soderland 1975
                       Semb 1978
                       Hovmand 1976
                       Vermeulen 1977
                       Krupa et al. 1976
                       Biggs et al. 1973
                       Zitko & Carson 1971
                       Struempler 1976
                       Thornton et al. 1980
                       Ibid.
                       Ibid.
                       Johannessen &
                         Henriksen 1978
                       Henriksen 1972
                       Weiss et al. 1978
                                                         1975
       London          Harrison et al.
       Gottingen (FRG) Ruppert 1975
       Minneapolis     Thornton et al. 1980
       Ottawa          Jonasson 1973
Lake Erie
Lake Huron
Lake Michigan-S
Lake Michigan-N
Uraymines (UK)
Swansea (UK)
United Kingdom
NE.  Minnesota
Lake Superior
Washington
Heidelberg
Konasewich et al.  1978
Ibid.
Eisenreich 1980
Ibid.
Peirson et al.  1973
Pattenden 1974
Cawse 1974
Eisenreich et al.  1978
Ibid.
Tanner et al. 1972
Bogen 1974
                               74

-------
        Metal Concentrations:  Wet Deposition (continued)
  Type
                 Precipitation
                  jjg
           Location
                       Reference
Cr
Rural
Bulk      2.9
          8.3
          6.6
Rain      1.9
         30.0
         <5.0
          0.2

Rain/snow 0.24
          0.27
          0.44
Urban   Rain
         15.0
          3.6
Rain/snow 0.51
Cu

Remote
Rain
Snow
        Ice/snow
<0.5
16
 0.047
 0.035
 0.074
 0.055
 0.851
Rural   Bulk
        Rain
          6
          9
          7
          6
          3
         23
         57
         26
         21
        5.8-7
          2.8
          2.3

         13.0
         23.0
         14.4
         30.0
          9.0
          7.0
        Uraymines (UK)
        Swansea (UK)
        United Kingdom
        Tennessee
        Belgium
        Sweden

        Washington
        NE. Minnesota
        N. Minnesota
        SW.North Dakota

        New York City
        Heidelberg
        Minneapolis
NWT (Canada)
Antarctica
                         Greenland
        Lake Ontario
        Lake Erie
        Lake Huron
        Lake Michigan-S
        Lake Michigan-N
        Uraymines (UK)
        Swansea (UK)
        England
        United States
        NE.  Minnesota
        Washington
        South Carolina

        Lake Erie
        Lake Superior
        Tennessee
        Belgium
        Sweden
                Peirson et al. 1973
                Pattenden 1974
                Cawse 1974
                Andren & Lindberg 1977
                Ronneau et al. 1978
                Dickson 1975
                Soderland 1975
                Tanner et al.  1972
                Thornton et al. 1980
                Ibid.
                Ibid.

                Volchok & Bogen 1973
                Bogen 1974
                Thornton et al. 1980
Kramer 1973
Gjessing & Gjessing 1973
Boutron & Lorius 1979
Boutron 1979a,b
Boutron 1979b,c
Ibid.
Weiss et al. 1975

Shiomi & Kuntz 1973
Konasewich et al. 1978
Ibid.
Eisenreich 1980
Ibid.
Peirson et al. 1973
Pattenden 1974
Cawse 1974
Lazrus et al. 1970
Eisenreich et al. 1978
Dethier 1977
Wiener 1979

Konasewich et al. 1978
Eisenreich et al. 1978
Andren & Lindberg 1977
Ronneau et al. 1978
Dickson 1975
Soderland 1975
                               75

-------
        Metal Concentrations:   Wet Deposition (continued)
          Type
                 Precipitation
                            Location
                                        Reference
Cu (continued)
                  5.3
                  1.4
                  6.6
                <10.0
                  2.0
                0.3-60
                150
                  6.4
                  0.4
                  0.8
        Rain/snow 4.4
                  2.1
        Snow
Urban   Bulk
        Rain
          3.9
         <1.0
          3.0
         13.0

         33.0
          9.5

          6.4
         20.0
         67.0
          6.8
        120
Rain/snow 8.2
Snow     30
Hg

Remote
Ice
        Ice/snow
0.429
0.111
0.011
0.048
Rural   Bulk
        Rain
         <0.2
          0.4
          0.06
          0.18
          0.2
       <1.0-3.6
          0.11
       Norway
       Denmark
       N. Holland
       Twin Cities
       Central Texas
       Delaware
       Pennsylvania
       New Brunswick
       Washington
       Menlo Park, CA
       Nebraska
       NE. Minnesota
       N. Minnesota
       SW.North Dakota
       Manitoba
       Ontario
       Norway
       New Mexico

       San Francisco
       Seattle
       New York City
       Gottingen (FRG)
       Pittsburgh
       Minneapolis
       Ottawa
Greenland

Greenland
       Uraymines (UK)
       England
       New Hampshire
       Tennessee
       Sweden
       Central Texas
       New Brunswick
       Washington
                Semb 1978
                Hovmand 1976
                Vermeulen 1977
                Krupa et al. 1976
                Cooper & Lopez 1976
                Biggs et al. 1973
                Chan et al. 1976
                Zitko & Carson 1971
                Tanner et al. 1972
                Kennedy et al. 1979
                Struempler 1976
                Thornton et al. 1980
                Ibid.
                Ibid.
                Beamish & Van Loon 1977
                Jonasson 1973
                Johannessen & Henriksen
                  1978
                Henriksen 1972
                Moore et al. 1978

                McCall & Bush 1978
                Larson 1977. (Personal
                  communication to S. A.
                  Norton.)
                Volchok & Bogen 1973
                Ruppert 1975
                Chan et al. 1976
                Thornton et al. 1980
                Jonasson 1973
Herron et al. 1976
Carr & Wilkniss 1973
Appelquist et al. 1978
Weiss et al. 1975

Peirson et al. 1973
Cawse 1974
Schlesinger et al. 1974
Andren & Lindberg 1977
Soderland 1975
Cooper & Lopez 1976
Zitko & Carson 1971
Tanner et al. 1972
                               76

-------
        Metal Concentrations:  Wet Deposition (continued)
          Type
                 Precipitation
             Location
                       Reference
Hg (continued)
        Snow
Urban   Rain



        Snow


Mn
Remote  Rain


        Snow
        Ice
        Ice/snow
Rural   Bulk
  <0.01   East Ontario
  <0.005  Alaska
 .05-0.10 NW.  Quebec
   0.08   Quebec
   0.04
   1.3
 J.5-4.0
   0.002
   0.069
        Rain
   0.018
   0.32
   0.215
   0.139
   0.250

   2.2
   4
  37
  15
   8.1
  49
  20
  11
  12
  14.8
   5

  35
                0.1-6.0
                 10
                 84
                   5
                 11
                 11
                   0.19
                Jonasson 1973
                Weiss et al.  1978
                McLean et al. 1980
                Matheson 1979
Gottingen (FRG) Ruppert 1975
Heidelberg      Bogen 1974
Canada          NAS 1978
Japan           Matsunaga & Goto 1976
Ottawa          Jonasson 1973
0.05-2.0  Quebec
          South Pacific
Antarctica
Greenland
South Carolina
N. Norway
Lake Michigan-S
Lake Michigan-N
Uraymines (UK)
Swansea (UK)
England
USSR-E
Lake Superior
Tennessee
Central New
  York
New Jersey
          Menlo Park
          Oregon
          Belgium
          Sweden

          Norway
          Washington
                Matheson 1979
Volchok 1979. (Personal
  communication to S.
  Eisenreich.)
Boutron & Lorius 1979
Boutron 1979c
Herron et al. 1977
Ibid.
Weiss et al.  1975

Wiener 1979
Hanssen et al. 1980
Eisenreich 1980
Ibid.
Peirson et al. 1973
Pattenden 1974
Cawse 1974
Drozdova & Makhonko 1970
Eisenreich et al. 1978
Andren & Lindberg 1977
Likens 1972

Volchok 1979. (Personal
  communication to S.
  Eisenreich.)
Kennedy et al. 1979
Volchok 1979. (Personal
  communication to S.
  Eisenreich.)
Ronneau et al. 1978
Dickson 1975
SSderland 1975
Semb 1978
Tanner et al. 1972
                                77

-------
        Metal Concentrations:   Wet Deposition (continued)
                 Precipitation
          Type
          (jg &
                            Location
                                        Reference
Mn (continued)
        Rain/snow 5.2
                  3.2
                  5.6
                 22
        Snow
Urban
Bulk
Rain
Mo

Urban
        Rain/
          Snow
Snow
Ni

Remote  Rain

Rural   Bulk
        Rain
17

22
15
 0.9

 2.7
16
25
22
25
 1.9

80
26
 0.2
          4.0
         27
          2
        0.9-1.7
          4.0
          6
         35
         12
          3.8
         <3
          5.7
         48
          5
          0.6
                 Nebraska        Struempler 1976
                 NE.  Minnesota   Thornton et al. 1980
                 N. Minnesota    Ibid.
                 SW.North Dakota Ibid.
Manitoba
Norway

S.  Norway
E.  Ontario
Alaska

San Francisco
New York City
Gottingen (FRG)
Pittsburgh
Heidelberg
Los Angeles

New York City
Minneapolis
Beamish & Van Loon 1977
Johannessen & Henriksen
  1978
Elgmork et al. 1973
Jonasson 1973
Weiss et al. 1978

McCall & Bush 1978
Volchok & Bogen 1973
Ruppert 1975
Chan et al. 1976
Bogen 1974
Liljestrand & Morgan
  1978
Volchok 1979
Thornton et al.  1980
Nagoya, Japan   Sugawara et al.  1961
        NWT (Canada)

        Lake Ontario
        Lake Erie
        Lake Huron
        NE.  Minnesota
        United States
        Uraymines  (UK)
        Swansea (UK)
        England
        USSR-E
        Lake Superior
        Tennessee
        Sweden

        Denmark
                Kramer 1973

                Shiomi & Kuntz 1973
                Konasewich et al.  1978
                Ibid.
                Thornton et al. 1980
                Lazrus et al.  1970
                Peirson et al. 1973
                Pattenden 1974
                Cawse 1974
                Drozdova & Makhonko 1970
                Eisenreich et al.  1978
                Andren & Lindberg, 1977
                Dickson 1975
                Soderland 1975
                Hovmand 1976
                               78

-------
        Metal Concentrations:  Wet Deposition (continued)
                 Precipitation
          Type
                    Location
                       Reference
Ni (continued
        Rain/snow 1.5
                  1.5
                  3.6
        Snow     <2.0
                  8.2
Urban
Pb
Remote
Rain    114
Rain/snow 2.4
Snow     17
Rain
Snow
Ice/snow
                  1.6
                  0.040
                  0.033
                  0.02
                  0.41
                  0.229
                  0.2
                  0.14
Rural   Bulk
        Rain
         20
         11
         12
         39
         13
         13.4
        7.1-7.8
          3.2
         34
         39
         97
         40
          3.5
          4.7
          6.6
         17
         15.6
        <20
          4
          3-35
        0.6-60
         19
          2.8
         51
                 NE. Minnesota
                 N. Minnesota
                Thornton et al.
                Ibid.
                1980
                 SW.North Dakota Ibid.
                 Manitoba        Beamish and Van Loon 1977
                 New Mexico      Moore et al.  1978
New York City
Minneapolis
Ottawa
NWT (Canada)
Antarctica
                 Greenland
Lake Ontario
Lake Erie
Lake Huron
Lake Michigan-S
Lake Michigan-N
New Hampshire
NE. Minnesota
Washington
United States
Uraymines (UK)
Swansea (UK)
England
N.  Norway
USSR-E
South Carolina
Lake Erie
Tennessee
Twin Cities
Central Texas
Menlo Park
Delaware
Pennsylvania
New Brunswick
Belgium
Volchok & Bogen 1973
Thornton et al.  1980
Jonasson 1973
Kramer 1973
Gjessing & Gjessing 1973
Boutron & Lorius 1979
Boutron 1979b
Murozumi et al.  1969
Ibid.
Ibid.
Ibid.
Herron et al. 1977

Shiomi & Kuntz 1973
Konasewich et al. 1978
Ibid.
Eisenreich 1980
Ibid.
Schlesinger et al. 1974
Eisenreich et al. 1978
Dethier 1977
Lazrus et al. 1970
Peirson et al. 1973
Pattenden 1974
Cawse 1974
Hanssen et al. 1980
Drozdova & Makhonko 1970
Wiener 1979
Konasewich et al. 1978
Andren & Lindberg 1977
Krupa et al. 1976
Cooper & Lopez 1976
Kennedy et al. 1979
Biggs et al. 1973
Chan et al. 1976
Zitko & Carson 1971
Ronneau et al. 1978
                               79

-------
        Metal Concentrations:  Wet Deposition (continued)
          Type
                 Precipitation
          pg
                            Location
                                        Reference
Pb  (continued)
         64
         10
         13
          7.0
         28
Rain/snow 4.8
          7.1
          5.7
          8.0
Snow     <1.0
          7.5
         15.1
         30
Urban   Bulk

        Rain
        Rain/
          snow
        Snow
 15
 30
190
 20
147
  5.4
 41
 75

 31

 76
                         Sweden

                         Norway
                         Denmark
                         N.  Holland
                         Nebraska
                         NE .  Minnesota
                         N.  Minnesota
                         SW. North Dakota
                         Manitoba
                         E.  Ontario
                         New Mexico
                         Norway
                 S.  Norway
                 London
                 Seattle
                 New York City
                 Gottingen (FRG)
                 Pittsburgh
                 Los Angeles

                 Minneapolis

                 Ottawa
Sb
Remote  Ice/snow  0.034  Greenland
Se
Rural
Bulk      0.34   Uraymines  (UK)
          1.1    Swansea (UK)
          0.59   England
          1      N.  Norway
V

Remote
Rural   Bulk
Ice
Ice/snow
0.022
0.016
0.31
Greenland
N. Norway
  4.1    Uraymines  (UK)
 13      Swansea  (UK)
  8.9    England
Dickson 1975
Soderland 1975
Semb 1978
Hovmand 1976
Vermeulen 1977
Struempler 1976
Thornton et al. 1980
Ibid.
Ibid.
Beamish & Van Loon 1977
Jonasson 1973
Moore et al. 1978
Johannessen & Henriksen
  1978
Henriksen 1972
Elgmork et al. 1973
Harrison et al. 1975
Larson 1977
Volchok & Bogen 1973
Ruppert 1975
Chan et al.  1976
Liljestrand & Morgan
  1978
Thornton et al. 1980

Jonasson 1973
                                 Weiss et al.  1975
                                         Peirson et al.  1973
                                         Pattenden 1974
                                         Cawse 1974
                                         Hanssen et al.  1980
                                 Herron et al.  1976
                                 Herron et al.  1977
                                 Hanssen et al.  1980

                                 Peirson et al.  1973
                                 Pattenden 1974
                                 Cawse 1974
                               80

-------
        Metal Concentrations:  Wet Deposition (continued)
          Type
                 precipitation-
                    Location
                               Reference
V (continued)
        Rain
Urban
Snow
Rain
 4.7    N.  Holland
:20      Twin  Cities
 1.1    Puerto Rico
 0.13   Alaska
 68      New York  City
Zn
Remote









Rain
Snow




Ice
Ice/snow


<1.0
27
0.001
0.031
0.78
0.293
0.215
0.224
1.1

NWT (Canada)
Antarctica


Greenland

Greenland


Rural   Bulk
        Rain
         80      Lake Ontario
        140      Lake Erie
         53      Lake Huron
         57      Lake Michigan-S
         33      Lake Michigan-N
          0.5    Washington
        110      United States
         85      Uraymines (UK)
        160      Swansea (UK)
        120      England
          7.8    N. Norway
        130      Lake Erie
        176      Lake Superior
         34.9    Tennessee
         15      Twin Cities
          2-13   Menlo Park
          3      Central Texas
         82      Pennsylvania
          3      Washington
        251      Belgium
        311      Sweden
         60
         29      Norway
         15      Denmark
        187      N. Holland
         36      New Brunswick
Vermeulen 1977
Krupa et al.  1976
Martens & Harriss 1973
Weiss et al.  1978
Volchok & Bogen 1973
                        Kramer 1973
                        Gjessing & Gjessing 1973
                        Boutron & Lorius 1979
                        Boutron 1979a,b
                        Boutron 1979b,c
                        Ibid.
                        Herron et al. 1976
                        Herron et al. 1977
                        Weiss et al. 1975

                        Shiomi & Kuntz 1973
                        Konasewich et al. 1978
                        Ibid.
                        Eisenreich 1980
                        Ibid.
                        Dethier 1977
                        Lazrus et al. 1970
                        Peirson et al. 1973
                        Pattenden 1974
                        Cawse 1974
                        Hanssen et al. 1980
                        Konasewich et al. 1978
                        Eisenreich et al. 1978
                        Andren & Lindberg 1977
                        Krupa et al. 1976
                        Kennedy et al. 1979
                        Cooper & Lopez 1976
                        Chan et al.  1976
                        Tanner et al. 1972
                        Ronneau et al. 1978
                        Dickson 1975
                        Soderland 1975
                        Semb 1978
                        Hovmand 1976
                        Vermeulen 1977
                        Zitko & Carson 1971
                               81

-------
        Metal Concentrations:   Wet Deposition (continued)
          Type
                 Precipitation
  (jg £
   Location
                                                Reference
Zn (continued
        Rain/
          snow
        Snow
Urban
 10

 93
 99
147
 <1.0
 10
  1.7
 63

 56
 45
Bulk

Rain


Rain/
snow
Snow
16
6
280
88
48
40

20
Nebraska

NE. Minnesota
N. Minnesota
SW.North Dakota
Manitoba
E. Ontario
Alaska
Norway
S.  Norway

San Francisco
Seattle
New York City
Pittsburgh
Struempler 1976

Thornton et al. 1980
Ibid.
Ibid.
Beamish & Van Loon 1977
Jonasson 1973
Weiss et al. 1978
Johannessen & Henriksen
  1978
Henriksen 1972
Elgmork et al. 1973

McCall & Bush 1978
Larson 1977
Volchok & Bogen 1973
Chan et al.  1976
                        Gottingen  (FRG) Ruppert  1975
                 22
         Minneapolis

         Bergen,  Norway
         Ottawa
                Thornton et al.  1980

                Forland & Gjessing 1975
                Jonasson 1973
                               82

-------
        Part II - TRACE ORGANICS:  A REVIEW AND ASSESSMENT


    S. J. Eisenreich (Chairman), T. F. Bidleman, T. J. Murphy,
               A. R. Davis, D. A. Banning, C. S. Giam,
                 F. J. Priznar and M. D. Mullin
1.   INTRODUCTION

     In recent years, there has been a growing concern that
atmospheric transport and deposition are responsible for the
accumulation of trace organic constituents, such as DDT, PCBs,
and others, in areas of the world where local sources are absent
(Nisbet and Sarofim 1972; Risebrough et al. 1968a, b; Peel 1975).
Although_atmospheric concentrations often range from <10~9 to
10~8 g m~3 and thus are very low, atmospheric input into aquatic
ecosystems contributes to and in some cases is responsible for
the accumulation of trace organics in every level of the food
chain.  In some instances, a complete fishery may be eliminated
(as in Lake Michigan where PCB concentrations in fish from 2 to
19 mg kg'1 far exceed the FDA guidelines of 2 mg kg"1 for edible
fish).

     Anthropogenic chlorinated (CH) and nonchlorinated hydrocar-
bons are hydrophobic compounds of low water solubility that
partition rapidly into lipid layers of plankton and fish and
concentrate in the food chain.  While food chain bioconcentration
at lower trophic levels may not be a controlling factor in the
high concentrations observed in plankton (Clayton et al. 1972),
it is probably the source for the concentrations of these mater-
ials found in fish.  The bioconcentration of some CHs may be
controlled by equilibrium partitioning between the internal lipid
pools of the biota and ambient water.  Once CHs are incorporated,
they may alter species composition of mixed phytoplankton assem-
blages (Mosser et al. 1972), inhibit zooplankton reproduction
(Wildish 1972), and decrease phytoplankton biomass and natural
species size,  thereby altering and contaminating harvestable
fish.  Trace concentrations of some CHs in the 10~9 g i~l range,
derived mainly from atmospheric deposition, can support potenti-
ally toxic concentrations in numerous compartments of the aquatic
ecosystem.

     The atmospheric processes responsible for the removal of
trace organics from the atmosphere are wet deposition, dry parti-
cle deposition, and dry-vapor deposition (Figure II-l).  A know-
ledge of gas-particle-phase distributions of trace organics in
the atmosphere is critical to understanding the deposition proc-
esses.  Gases may be scavenged from the atmosphere by precipi-
tation; the extent of scavenging, or solubility coefficient (a),
                                83

-------
     WET              PARTICLE             VAPOR
                        (dry)                (dry)
                                            °°
          Rainout        "BV.» Settling          _..°o"  Partitioning
          Washout        V '• ' Impaction       °'\ °" °,
          Partitioning                          ,.f °'r             AIR
                                                           WATER
  Figure  II-l.  Atmospheric  input of anthropogenic organics to
  natural waters.
is dependent upon the magnitude of Henry's law constant H,  which
is the reciprocal of a.   Particle removal from the  atmosphere  by
wet deposition can also be estimated from the washout coefficient,
W (see Part III).  In both cases, the concentration of the  trace
organic in the atmosphere at a reference height must be known.

     Dry particle deposition can be estimated from  the deposition
velocity and the particulate trace-organic concentration at a
reference height.  The deposition velocities of trace organics
to different receptor surfaces are not well known.   The flux of
vapor-phase trace organics into aqueous systems depends on  the
overall mass-transfer rates across the air/water  interface  and
on the "dissolved" concentration of the trace organic in water
in equilibrium with the concentration in the atmosphere. Here
again, Henry's law constant H becomes important.  Therefore,
several parameterizations are required to estimate  the wet  and
dry flux of trace organics to water.

     Trace organics entering a water body from the  atmosphere
first come into contact with the surface film and must be trans-
ported through the film.  They are then subjected to the biologi-
cal uptake, aerosolization/volatilization, and selective chemical
or biological degradation that determine the mass and species
available to the water column.  Hydrophobic organics, such  as
DDT,  PCBs, and PAHs, exhibit low water solubilities and largely
associate with biotic and abiotic particulates.   However, even in
unusual cases when the particle:water distribution  coefficient is
high (104-105), a large fraction of the trace-organic concentra-
                                84

-------
tion may be dissolved because of low concentrations of suspended
particles  (i.e., low adsorptive surface area).  Ultimately, trace
organics that are biologically transformed or adsorbed to partic-
ulates accumulate in lake sediments and lipid compartments of the
water column.  The chemical and physical properties of trace
organics are important because deposition and emission processes
are, to some extent, influenced by aqueous cycling.

     This  section discusses the processes important in atmos-
pheric trace-organic fluxes to water, identifies the physical
properties and parameterizations necessary to estimate fluxes,
summarizes what is known about trace organics in the atmosphere,
and applies flux calculations to a specific system.
2.   VAPOR AND PARTICLE DISTRIBUTION OF ATMOSPHERIC ORGANICS

     High-molecular-weight organics in the atmosphere are present
in the vapor phase and are adsorbed on particulate matter.
Vapor-aerosol partitioning in the atmosphere depends on the vapor
pressure of the organic compound, the size and surface area of
the suspended particulates, and the organic content of the aero-
sol (Junge 1977).  Vapor-aerosol distributions calculated from
vapor pressure and the quantity of atmospheric particulates
(Figure II-2) show that the amount of an organic in the partic-
ulate phase in clean-air environments is rather small when the
saturation-vapor pressure (P ) is greater than 10 6 mm Hg.  Junge

(1977) suggested that most PCBs, DDT, and Hg in clean atmospheres
are in the vap_or phase and that compounds with saturation-vapor
pressures <10 7 mm Hg in clean air are in the particulate phase.
Considering clean-, rural-, and urban-air environments, organics
paving P  > 10~4 mm Hg should exist almost entirely in the vapor

phase, and those having P  < 10~8 mm Hg should exist almost en-

tirely in the particulate phase.  In reality, the high-molecular-

weight organic pollutants fall somewhere between these extremes,
and their distribution and atmospheric lifetimes depend largely
on the particulate content of air.  Table II-l gives a represent-
ative series of vapor pressures reported for selected trace
organics.

     Unfortunately, sampling techniques for airborne organics can-
not adequately distinguish vapor from particulate-phase species.
The most commonly used collection system employs a high-volume
sampler with particulates collected on a glass-fiber filter and
gases collected by a following adsorbent bed.  The adsorbents
used in the collection of atmospheric PCBs are polyurethane foam-
PUF (Simon and Bidleman 1979; Bidleman and Olney 1974b) and
XAD-2 macroreticular resin (Doskey and Andren 1979; Hollod GJ,
                                85

-------
  1.0
                         10
                                    10-
                                                10
                                                 -4
                                                           10-3
Figure 11-2.  Vapor-aerosol distributions of Hg calculated from
vapor pressure and the quantity of atmospheric particulates.
Ratio * = adsorbed on aerosol    compounds as a function of
      y   total concentration       ^
saturation vapor pressure P  and aerosol surface 6 (after Junge
1977).                     °

A: PQ = 10~8 mm Hg; B: PQ = 10~7;  C: PQ = 10~6; D: PQ = 10~5;

E: P  = 10~4 .
    o
Eisenreich SJ,  unpublished observations).  The high-volume col-
lection of airborne organics employing the filter/adsorbent
system may underestimate the actual proportion of particle:phase
organics because organics may be desorbed from the particles
collected on the filter and the resultant vapor retained by the
solid adsorbent.  Nevertheless, fractions of "filter-retained"
organics determined using high-volume samples consist of higher
molecular-weight and low-vapor-pressure species.   This was demon-
strated by Cautreels and vanCauwenberghe (1978) for a homologous
series of alkanes collected in the urban atmosphere.  As shown in
Table 11-2, a significant fraction in the particle phase in-
creases with increasing molecular weight and decreasing P  .

Also a fraction of other high-molecular-weight species, such as
PAH and phthalate esters, exist in the gas phase, even in urban
atmospheres (Giam et al.  1980).  Therefore, the distribution of
trace organics  between the glass-fiber filter and the solid
adsorbent cannot be taken as unequivocal evidence of vapor/parti-
cle partitioning, but rather the sum as the total atmospheric
                               86

-------
                           Table II-l.  Vapor Pressures of Trace Organic Compounds
00
Compound Molecular weight
Aroclor 1242
1248
1254
1260
2' ,3,4-PCB
p , p ' -DDT
-hexachlorocyclo-
hexane
Dieldrin
Aldrin
Chloroterpenes
(Strobane)
Chlorobenzene
Benzene
Toluene
Ethyl benzene
Naphthalene
Biphenyl
Acenaphthalene
Phenanthrene
Pyrene
Benzo(a)pyrene
258 (3-C1)
293 (4-C1)
328 (5-C1)
373 (6.3-C1)
258
356
291
377
363
412
113
78
92

128


202
228
4.
4.
7.
4.
1.

9.



12T.
95
28.

0.


6.
6.
mm Hg
06 x 10~4
94 x 10~4
71 x 10"5
05 x 10~5
2 x 10~4
1 x 10"7
4 x 10"6
1 x 10~7
6 x 10~6
3 x 10~7
9

4

09


8 x 10~6
6 x 10"7
Atm
5
6
1
5
1
2
1
2
7
8
1
1
3
1
1
2
5
5


.33
.50
.01
.33
.6
.8
.2
.8
.9
.4
.6
.25
.7
.25
.2
.86
.74
.07
9
9
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10"7
10~7
10"7
10~8
10"7
10"10
10~8
10'10
10~9
10~10
10'2
10"1
10'2
10~2
10"4
10'3
10'2
10~5
10'9
10'10
Reference
MacKay and Wolkoof (1973)
Ibid.
Ibid.
Ibid.
Westcott J, Bidleman TF,
unpublished observations
MacKay and Wolkoff (1973)
Ibid.
Ibid.
Ibid.
Melnikov (1971)
MacKay et al . (1979)
Ibid.
Ibid.
Ibid.
MacKay et al . (1979)
Calculated from data
of MacKay et al . (1979)

Pupp et al. (1974)
Ibid.
                 Diethylhexylphtha-
                  late
10
             10"10
Ibid.

-------
         Table  II-2.  Apparent Particulate-  and Gas-Phase
             Distributions  for Organics  in Urban Air
Compound
Hydrocarbons
n-C15
n-C17
n-C19
n-C22
n-C25
n-C29
PAHs
Phenanthrene
+ anthracene
Fluoranthene
Pyrene
Benzo ( a ) anthracene
+ chrysene
Benzo ( a ) f luoranthene
Benzo(a)-, benzo(e)
+ perylene
di-isobutylphthalate
di-n-butylphthalate
di-2-ethylhexyl
phthalate
Particulate
( ng m 3 )


0
2
9
15


1
2
1
12
23
20
1
101
54
^ ^
--
.8
.33
.50
.8


.21
.22
.64
.2
.1
.1
.73

.1
Gas_
( ng m 3 )
61
66.
15.
4.
5.
6.


44.
8.
3.
3.
2.
2.
32.
353
127

5
I
23
74
56


7
52
36
87
01
69
8


P/G
ratio


0
0
1
2


0
0
0
3
11
7
0
0
0
_ _
--
.053
.55
.66
.4


.027
.26
.488
.15
.5
.47
.053
.286
.426
SOURCE:   Cautreels  and vanCauwenberghe  (1978).
                               88

-------
 Table I1-3.  Ranges of Gas-Particle Distribution Factors for
 Organics (f  ), Nitrates  (f  ), and Sulfates  (f  ) in Pasadena, CA
            *—              n                  s
Day f f
2 c n
fs
 7/12/73    0.009-0.03        0.012-0.018          0.230-0.333

 7/25/73    0.009-0.062       0.019-0.068           0.13-0.39

10/17/73    0.007-0.024       0.007-0.032          0.171-0.466


SOURCE:  Adapted from Grosjean and Friedlander (1975).


NOTE:        OC                  NO,"                SOd~
     f  = - -  ,    f  = -  ,   f  =
                          n            ~      S
          HC  + OC            N0x + N03            S02 + S0
concentration.  This observation also points out the importance
of using an additional adsorbent as a back-up to the filter to
prevent the loss of the volatile fraction of the organics.

     Using similar methodologies, Grosjean and Friedlander (1975)
found that 97%-99.3% of organic carbon, 93%-99.3% of nitrogen,
and 53%-87% of sulfur in Pasadena air consisted of vapor-phase
components (Table II-3).  There is also a controversy about
whether atmospheric PCBs exist in the vapor or particle phase.
The data in Table II-4, summarizing some determinations in urban
and marine and rural environments, show that ^85%-100% of air-
borne PCBs are, at least operationally, in the vapor phase.
Junge (1977) presented vapor pressure and adsorption evidence
that -^90% of PCBs are in the vapor phase.  Thus, for many trace
organics, the vapor phase may contribute a sizeable fraction of
the atmospheric concentration.

     The particle size with which trace pollutants are associated
is largely unknown.  Doskey and Andren (1980) showed that PCBs
are probably associated with 0.1 to 1.0 pm particles.  Van Vaeck
and vanCauwenberghe (1978) showed that aerosol PAH in urban,
rural, and coastal environments have mass-median diameters (mmd)
of 0.7-1.6 pm with 34%-70% of the mass occurring in sizes less
than 1 (jm (Table 11-5).  Because of their higher surf ace: volume
ratio and higher organic content, submicron particles have higher
concentrations than do larger particles.  However, Van Vaeck and
vanCauwenberghe (1978) found that a significant fraction of the
                                89

-------
  Table II-4,
Particulate- and Vapor-Phase Distributions
    for Atmospheric PCBs
Percentages
Location
Marine and Rural
Atlantic
North Atlantic
Grand Banks,
Newf oundl and
Gulf of Mexico
Lake Superior
Lake Michigan
Bermuda

Urban
Toronto, ON
Hamilton, ON
Milwaukee, WI
Lake Michigan
(Chicago)
Sheridan Park, ON
Columbia, SC
Particulate
10
2
1
1
0
3
2

14-43
5-18
16
13
11
8
Vapor
90
98
99
99
100
97
98

57-86
82-95
84
87
89
92
Reference
Junge (1975)
Bidleman and Olney
(1974a)
Harvey and Stein-
hauer (1974)
Giam et al. (1980)
Eisenreich and
Hollod (1980)
Doskey (1978)
Bidleman and
Olney (1974b)
Gilbertson (1976)
Ibid.
Doskey (1978)
Murphy and
Rzeszutko (1977)
Gilbertson (1976)
Bidleman and
Chicago,  IL
                               Christensen (1979)

                       96     Murphy and
                               Rzeszutko (1977)
                            90

-------
  Table I1-5.  Particle-Size Distributions for PAHs in Ambient
           Aerosols in Urban, Rural, and Coastal Areas
Compound
                        Coastal
               < 1
                                     Rural
                Urban
mmd   < 1 |jm   mmd   < 1
(Mm)    (%)    (urn)
                     mmd
                     (Mm)
Phenanthrene
 and anthracene        49

Fluoranthene           46

Benzo(a)anthracene
 and chrysene          57

Benzo(b + k)
 fluoranthenes         64

Benzo(a + e)
 pyrenes               64

Dibenzo
 anthracenes           61
                          1.1

                          1.1


                          0.9


                          0.7


                          0.7


                          0.8
36

39


54


63


58


34
               1.5

               1.4


               0.9


               0.8


               0.9


               1.6
47

60


67


70


68


59
1.1

0.8


0.7


0.7


0.7


0.8
SOURCE:  Van Vaeck and vanCauwenberghe (1978).
high-molecular-weight n-alkanes, carboxylic acids, and PAHs were
associated with particles >1 |jm mmd.  Since these particles have
higher deposition velocities and washout ratios (Slinn et al.
1978), the flux of high-molecular-weight organics may be domi-
nated by the deposition of large particles.  That is, even though
most trace organics are on submicron particles, most of those
deposited may be due to large particles.
3.
TRANSFER OF GASES ACROSS AIR/WATER INTERFACE
     Transfer of gases across the air/water interface can be
predicted from a two-film diffusion model (Liss and Slater 1974).
In this model, the air and water reservoirs are assumed to be
well mixed except for thin, stagnant films of air and water at
the interface.  The rate of transfer is governed by the molecular
diffusion across these interfacial layers and driven by the
concentration gradients between equilibrium concentrations at the
interface and the concentrations in the bulk air and in water
reservoirs.  For the steady-state transfer across air and water
films, the flux (F) is given by (MacKay et al. 1979):
                               91

-------
               F = KnT (C - P/H)
and                 °L                                       (3-D
               1/K_T = 1/K  + RT/HKg
                  \JLj      J-J
                                                         tt

where F is the flux (mol m~2 h"1); KL and  Kg,  the  liquid and

gas-phase mass-transfer coefficients (m h'1);  KQL,  the overall

liquid-phase mass-transfer coefficient (m  h"1);  H,  the Henry's
law constant (atm m3 mol"1)(P/C); C, the solute  concentration in
the liquid phase (mol m~3); P, the solute  partial  pressure (atm);
T, the absolute temperature (K); and R, the  gas  constant (m3  atm
mol"1 K"1).

3.1  Resistance to Air/Water Interface

     Resistance to gas-phase transfer occurs  in  both  the liquid
and gas films.  For solutes of increasing  molecular weight,  both
Kr and Kg decrease although their ratio may  remain approximately
  J_i
the same. _MacKay et al. (1979) have shown that  for H >  5  x  10~3
atm m3 mol 1, resistance lies almost totally  in  the liquid phase;
the flux is then described by

                    F = KT (C - P/H).                        (3-2)
                         Li

If H < 5 x 10~6 atm m3 mol"1,  then resistance  lies  almost  totally
in the gas phase; the flux can be described by

                      =  KgH(C-P/H)
                             RT                              (3-3)

For H values between these extremes, resistance  to  mass  transfer
occurs in both the gas and liquid film; the  equation  for flux is

                         F = KQL (C-P/H).                    (3-4)

A knowledge of the Henry's law constant is essential  to  predict-
ing air/water exchange because one must be able  to  calculate  the
concentration of the organic compound in the water  in equilibrium
with the organic in the vapor phase.  If the water  is undersat-
urated with respect to equilibrium with the  atmosphere (negative
C-P/H in Eqs. 3-2,  3-3,  and 3-4), there will be  a  net transport
of vapor to the water.  If the water is oversaturated with re-
spect to the air (positive C-P/H),  there will be a  net transport
of vapor to the atmosphere.   The rate and  amount of material
transferred depends on whether the compound  is liquid- or  gas-
phase controlled.  This can be determined  from the  equations
above.

     Henry's law constant is calculated by dividing the  solute
vapor pressure at saturation (25°C) and the solute  solubility
                               92

-------
    Table I1-6.  Water Solubilities of Trace-Organic Compounds
   Compound
Solubility
(mol m~3)
       Reference
Aroclor 1242
        1248
        1254
        1260

p,p'-DDT

HCH

Dieldrin

Aldrin

Chloroterpenes
 (strobane)

Chlorobenzene

Benzene

Toluene

Naphthalene

Biphenyl
9.3 x 10~4
1.8 x 10~4
3.7 x 10~5
7.2 x 10~5

3.4 x 10~6

2.5 x 10~2

6.6 x 10"4

5.5 x 10~4


1.2 x 10~3

    4.2

   22.8

    5.7

    0.24

4.9 x 10~2
MacKay and Wolkoff (1973)
Ibid.
Ibid.
Ibid.

Ibid.

Ibid.

Ibid.

Ibid.


Ibid.

MacKay et al.  (1979)

Ibid.

Ibid.

Ibid.

Ibid.
 (25°C) in water.  Recently, direct measurements of H for relative-
 ly low-molecular-weight organics were made by stripping organic
 vapors from an aqueous solution using a gas stream (MacKay et al.
 1979).  The calculated values based on saturation vapor pressures
 and solubilities (Tables II-l and II-6) agree well with experi-
 mentally determined values (Table II-7).  Unfortunately, one (or
 both) of these physical constants is not accurately known for
 most high-molecular-weight organic pollutants because of the
 difficulties of working with slightly soluble organics.  Further-
 more, the true "dissolved" aqueous concentration in equilibrium
 with the atmospheric species needs to be known.

     Vapor pressures for PCBs have been determined at high tem-
peratures and estimated values at 25°C have been obtained by
 extrapolation using the Antoine equation (MacKay and Wolkoff
 1973).  However, these values are uncertain and possibly 101 to
                                93

-------
                Table II-7.   Henry's Law Constants
Compound
Henry's law constant Cont
( atm m3 mol 1 ) ph
Calculated Measured (gas/
Aroclor 1242
1248
1254
1260
2 ' ,3,4-PCB
p , p ' -DDT
HCH
Dieldrin
Aldrin
Chloroterpenes
Chlorobenzene
Benzene
Toluene
Naphthalene
Biphenyl
Acenaphthalene
Phenanthrene
5
3
2
7
5
8
4
4
1

3
5
6




.7
.6
.7
.4
.3
.2
.8
.2
.4
7
.8
.5
.5
5



X
X
X
X
X
X
X
X
X
X
X
X
X
X



10
10
10
10
10
10
10
10
10
10
10
10
10
10



"4 1.1 x 10~6 - 1.4 x 10~7*
~3
~3 2.9 x 10~7 - 6.9 x 10~8*
~4
~4
~5
~7
~7
~5
~7
~3 3.8 x 10~3
~3 5.6 X 10~3
~3 6.6 x 10~3
~4 4.8 X 10~4
4.1 x 10~4
1.5 x 10"4
3.9 x 10~5
:rolli
tase
'liqui
G
G

G,
G,
G
G
G,
G
L
L
L
G,
G,
G,
G,
•ng
-d)


L
L


L




L
L
L
L
SOURCE:  Experimental values of H (except as noted) from
         MacKay et al.  (1979).

NOTE:  For H <  5 X 10~6,  >  95% G; for H > 5 x 10~3, > 95% L.

*Air/water partition coefficients determined in natural lake
 water (Doskey and Andren 1980).
                               94

-------
103 too high.  Saturation vapor pressures of  individual PCB
isomers should be much higher  above commercial  liquid mixtures
than above a pure solid, which is the physical  form of individual
isomers.  Recently water solubilities were reported for pure PCB
isomers in distilled water  and seawater  (Dexter and Pavlou 1978),
and the vapor pressure of a trichlorobiphenyl isomer was measured
(We^tcott J, and Bidleman TF,  unpublished observations).


3.2  Problems

     Thus, vapor-phase trace organics partition into water and
can be described, in part,  by  the Henry's law constant (Table
II-7).  In theory, if H is  known, then the concentration of a
trace organic at equilibrium in air and water can be calculated.
However, in practice, the situation is more complex.  For exam-
ple, vapor-phase PCBs entering the water column may remain dis-
solved and unassociated, bind  with dissolved or colloidal organ-
ics, adsorb onto a particle surface, or be absorbed into an
organic particle.  Only the dissolved, unassociated form equili-
brates with the organic vapor  in the atmosphere.  To determine
whether the organics present as vapor in air are in equilibrium
with the organics in water, the concentration of the dissolved
species must be known.  Analytically, this measurement is, at
best, difficult to perform  at  the ambient concentrations observed
in the environment—1 to 10 x  10~9 g i l.  This problem might be
avoided by measuring the concentration on suspended particulate
matter and applying a particulate/water-distribution coefficient,
such as that determined by  Chen et al. (1980),  Karickoff et al.
(1979), and Dexter and Pavlou  (1978).
4.   WET REMOVAL OF TRACE ORGANICS FROM THE ATMOSPHERE

     Wet removal of airborne trace organics occurs by scavenging
of particles and vapor partitioning.  The relative importance of
these processes depends on the fraction of organic present as an
aerosol and the particle-size distribution of the particulates
and on H for vapor-phase partitioning.  A falling raindrop should
attain equilibrium with a trace-organic vapor in ~10 m (Slinn et
al. 1978).  Washout ratios calculated as the reciprocal of H are
given in Table I1-8 along with the field values calculated as

                         mass organic £  .   l
                    W =  	rain_   .
                         mass organic £  •  l
                                       O.11.


For PCBs and DDT, the field values are much higher than predicted
for values from vapor partitioning into rain.  This suggests that
wet removal of PCBs and DDT occurs predominantly by scavenging of
                                95

-------
                                        Table II-8.  Washout Ratios for Trace-Organic Compounds
CTl
Compound Vapor calculation
Aroclor 1242 43
2.2 x 104 - 1.7 x 105
1248 6.8
1254 9.0
8.4 x 104 - 3.5 x 105
1260 33
2',3,4-PCB 46
p,p'-DDT 298
p , p ' -DDE
-hexachloro-
cyclohexane 5.1 x 104
Dieldrin 5.8 x 104
Aldrin 1.7 x 103
Chloroterpenes 3.5 x 104
Chlorobenzene 6.4
Chlordane
Hexachloro-
benzene 370
Naphthalene 49
Phenanthrene 626
Diethylhexylphthalate
Field measurement
1.2 x 10"

8.6 x 104
8.6 x 104
7.8 x 104
1.9 x 104


5.4 x 104
7.4 x 104
1.9 x 10"
2.5 x 103 - 1.5 x 104
5 x 103 - 1 x 104
1.3 - 3.4 x 104
2 x 103
2 x 103 - 9 x 103

< 7 x 103 - 3 x 10s

7 x 103
4 x 103
>1.5 x 103


1.4 x 104 - 8.5 x 104
Reference
Murphy and Rzeszutko (1977
Doskey and Andren (1980)

Murphy and Rzeszutko (1977




)
Bidleman and Christensen (1979)
Giam et al. (1978)


Atkins and Eggleton (1974)
Bidleman and Christensen (
Giam et al. (1978)
Giam et al . (1980)
Atkins and Eggleton (1974)
Giam et al. (1978)
Atkins and Eggleton (1974)
Giam et al . (1980)

Bidleman and Christensen (



1979)




1979)

Bidleman and Christensen (1979)
Giam et al. (1980)
Giam et al. (1978)








-------
particulate matter.  For other high-molecular-weight organics,
vapor scavenging may be important.  This overall behavior is
supported by findings that early portions of rainfall are high in
particulates and PCBs (Murphy and Rzeszutko 1977; Strachan and
Huneault 1979).  However, as with air/water vapor exchange, the
accuracy of these predictions depends on the accuracy of H.
5.   DRY DEPOSITION OF PARTICULATES

     The dry deposition to a receptor surface of trace organics
associated with atmospheric particulates depends on the type of
surface, the resistance to mass transfer in the deposition layer,
and the particle size and concentration.  Ideally, all these
factors should be known to adequately describe the transfer of
atmospheric particulates to a surface.  For particles with radii
a in the range of 0.001 < a < 0.1 pm, the transfer velocity can
be given by (Slinn et al . 1978)
                                2/3—  '
                            (Sc)a//ci

where h is von Harmon's constant  (^0.4), u^ the friction velocity,
and Sc the Schmidt number.

     For large particles, transfer ratios increase, especially
above 1.0 urn.  In practical terms, the deposition of particles to
a surface can be described by

                         F = Vd(C)   ,                       (5-2)

where F is the flux, V, the deposition velocity, and C the con-

centration.  The V, depends primarily on the deposition surface

(i.e., water versus forest canopy) and particle size.  Deposition
velocities are minimal for particles in the range 0.1 < a < 1.0 pm
and increase for both larger and  smaller particles.  For parti-
cles having radii in the range of 2-10 (jm, Vd is ~l-5 cm s"1;

for particles having radii < 0.1  (jm, Vd is -^0. 02-0. 8 cm s"1;

and for particles in the range of 0.1-1.0 (jm, Vd is -^0.1-1 cm s'1

(see Part III).  Slinn and Slinn  (1980) proposed that deposition
of submicron particles to a water surface may be enhanced by
rapid particle growth from water  condensation in the deposition
layer.  Organic pollutants, especially those with low vapor
pressures,  that have high molecular weight and are hydrophobic in
nature are associated with particles above and below one micron,
with a slight preference for the  smaller particles (Table II-5).
Many organics condense on aerosol surfaces following emission
                                97

-------
from a high-temperature source.  Because of the nonpolar nature
of most trace organics, the organic carbon content of the parti-
cle is also important.

     Deposition velocities are given in Table  I 1-9 for various
airborne-organic species.  An average Vd for trace organics in
the atmosphere has been determined by measuring the flux to a
surface coated with glycerol, ethylene glycol, and mineral oil
or a plain, wet filter paper and simultaneously measuring the
atmospheric concentrations of the compound of  interest.  For
surfaces coated with mineral oil or another nonpolar substrate,
it is difficult to determine whether the organic flux is due only
to particle deposition or to a combination of  gas and particle
flux (Murphy et al .  1980).  The deposition velocities determined
in this way (V, = F/C  . ) are operational in nature but may
              Q
represent realistic estimates of actual total dry flux to a water
surface if polar, hygroscopic coatings are used.
6.   SOURCES OF TRACE ORGANICS TO THE ATMOSPHERE

6.1  Sources I

     The sources of many trace organics to the atmosphere are
obvious, but it may prove useful to briefly describe the import-
ant types of sources.  Most trace organics in the atmosphere
occur in three categories:  (1) compounds such as the volatile
freons with high vapor pressures, which upon use accumulate in
the air; (2) compounds such as pesticides and herbicides which
are often injected into the atmosphere by spraying and also
evaporate from plant surfaces; (3) compounds formed by combus-
tion, such as PAHs, which are discharged directly to the atmos-
phere.  The vapor pressures of high-molecular-weight PAHs pre-
clude their being in the atmosphere without originating from a
high-temperature source.

     Some compounds enter the atmosphere by a variety of sources.
Solvents, plasticizers,  etc.,  that are exposed to air enter the
atmosphere by vaporization.  Natural organic compounds produced
by vegetation,  such as terpenes,  evaporate from forest canopies.
Open burning,  such as forest and brush fires,  causes many organic
compounds to be emitted into the air, both as particulates and
vapors.   The incomplete combustion of municipal refuse, sewage
sludge,  and industrial products results in the emission of large
quantities of organics.

     Murphy and Rzeszutko (1977)  stated that high-temperature
sources  such as these may be responsible for much of the PCBs
"recently"  added to the environment.  PAHs either formed in or,
at least, emitted by the combustion of fossil fuels condense on
aerosols and can be transported great distances.
                                98

-------
                               Table I1-9.  Dry-Deposition Velocities
l£>
Compound
PCB (Aroclor 1242, 1254)
PCBs
PCBs, DDT
PCBs, DDT
PCSs (Total)
PCBs (Aroclor 1016)
PCBs (Aroclor 1254)
Chlordane
p , p ' -DDT
Toxaphene
Deposition
velocity
(cm s ~ 1 )
0
0
0
1
0
0
0
0
1
0
.5
.3-3
.19
.0
.14
.04
.43
.068
.3
.24
Comment
Estimated from Sehmel
and Sutter (1974)
Mineral-oil -coated
plates
Estimated gas phase
Estimated particulate
Glycerol-coated plates
Glycerin-water Al pans
(in Columbia, SC)
Glycerin-water Al pans
(in Columbia, SC)
Glycerin-water Al pans
(in Columbia, SC)
Glycerin-water Al pans
(in Columbia, SC)
Glycerin-water Al pans
(in Columbia, SC)
Reference
Doskey and Andren (1980)
McClure (1976)
Bidleman et al . (1976)
Ibid.
Eisenreich and Hollod
(1980)
Bidleman and Christensen
(1979)
Ibid.
Ibid.
Ibid.
Ibid.

-------
6.2  Sources II

     The manufacturing process and the sites of manufacturing,
distribution,  and use are other sources of trace organics in the
atmosphere, although mostly of local significance.  Much more
prevalent are dispersed sources,  such as landfills, dumps, and
other areas, where spent or spilled organic products are disposed
of.  Finally,  some manufacturing by-products and wastes are used
as an energy source and combusted alone or with other materials.
For example, the total quantity of PCBs present in waste oils,
although limited to -^50 mg H~l, is substantial when the total
amount of oil used in commerce is considered.
7.   VARIABILITY OF CONCENTRATIONS AND FLUXES

     One difficulty in determining concentrations of trace organ-
ics in air and precipitation and their flux to a receptor surface
is variability-  Concentrations and fluxes vary with airborne
concentration, macrometeorology and micrometeorology,- frequency,
intensity and duration of precipitation,  and organic speciation.
This results in field values varying by an order of magnitude or
more.  Thus, a statistically valid number of samples needs to be
taken to accurately estimate concentration.  Conclusions based on
a few samples or samples collected over short time periods and
from similar meteorological conditions must be applied with
caution.
8.   ATMOSPHERIC CONCENTRATIONS

     A paucity of data exists on the occurrence of trace-organic
contaminants in the atmosphere.  Only recently has the relative
impact of atmospheric deposition on water quality (e.g., acid
rain and PCBs in the Great Lakes) been recognized.  Polluted air
masses accumulate chemical components from local sources as well
as from sources hundreds or thousands of kilometers away.  Air-
mass circulation can transport airborne pollutants from urban or
industrialized centers,  depositing them in distant regions.  The
published and unpublished literature over the last ten to fifteen
years was reviewed to determine the nature and concentration of
airborne pollutants found in rain and snow and in vapor and
particle phases in urban and rural areas.  The data shown in
Tables 11-10 and 11-11 summarize our findings for a wide range of
trace-organic compounds.  In many instances, especially for PAHs,
airborne-contaminant data were available only for aerosol concen-
trations determined in urban areas.  Where possible, median
values are identified so that values outside the median range can
be more readily interpreted.
                               100

-------
Table 11-10.  Atmospheric  Concentrations of  Trace-Organic
         Compounds in  Urban and Rural Environments
Compound
Total PCBs

Total DDT
(IDDT, DDE, ODD)
Dieldrin

Aldrin
Toxaphene

a-BHC

Y-BHC
( Lindane )
Hexachlorobenzene
Chlordane
Heptachlor

Heptachlor
epoxide

Endrin
Methoxychlor

Trichloro-
ethylene

Tetrachloro-
ethylene
Mirex
a-endosulfan

p-endosulfan

Urban
V/P*
(ng m 3 )
0.5-30
(5-10)

0.01-1
0.5-10

1-10
0.02-10
(South)

2-10



0.1-0.5
1-20








100-1000


500-2500






R/S*
(ng £~l )
10-250
(50)

1-30
5-40

7
10-100
(South)
3-60

1-12
(6)
1-10















1-10

1-10
(3)
Rural
V/P* R/S*
(ng m 3) (ng S>~1 )
0.1-2 10-100
(1) (15-50)

0.01-0.05 1-10
0.01-0.1 0.5-30
(1-4)
0.1-1 0.5-3
0.02-2 1-10

0.25-0.4 1-35
(10-20)
0.2-4 1-15
(5)
0.1-0.3 1-4
0.01-0.5 1-5
0.2-5 1-9
(2)

0.5-5
(2)

1-20
(8)

100-1000
(Alaska)


<1
1-10
(2)
1-10
(2-5)
   Polychlorinated
    naphthalenes

   Dibutylphthalate

   Diethylhexylphtha-
    late
10-25
                       0.5-5


                       0.5-5
   NOTE:  Numbers in parentheses represent median values.

   *V/P = vapor/particulate,  R/S = rain/snow.
                                 101

-------
     Table 11-11.   Atmospheric Concentrations  of PAHs
                  in Different Environments
Compound
Benzene

Toluene

Anthracene

Phenanthrene

Fluor anthrene

Pyrene

Benz(a)-
anthracene
Chrysene

Benzo(k)-
fluoranthene
Benzo(a)pyrene

Benzo(e)pyrene

Indeno
[1,2,3-c-d]
pyrene
Perylene

1, 12-benzo-
perylene
Benzo(g,h,i)
perylene
Coronene

Total PAH
Urban
V/P* R/S*
(ng m"3) (ng A'1)
600-30,000
(1,000-2,000)
600-50,000
(1,000-10,000)
0.2-10
(0.5-4)
10-50

0.1-13 300
(1-5)
0.2-10
(1-5)
0.06-5
(0.3-3)
0.04-5
(1-4)
0.06-5
(0.1-1)
0.03-10
(0.3-3)
0.05-5
(0.5-2)
0.03-3
(0.3-2)

0.01-14
(0.1-1)
1-30 90
(1-10)
0.2-20
d-10)
0.2-20
(1-6)

Rural
V/P* R/S*
(ng m~3 ) (ng S.'1 )




0.01-1.0
(0.1-0.5)
0.02-6
(0.1-1)
0.2-7 60
(0.5-2)
0.1-10
(0.5-2)
0.4-5
(0.1-1)
0.1-12
(0.1-2)
0.08-4

0.01-5
(0.2-2)
0.02-3
(0.2-2)



0.01-2
(0.1-0.6)
0.04-3 10

0.06-1.9
(0.5-2)
0.02-0.2

50-300
NOTE:  Numbers in parentheses represent median values.

*V/P = vapor + particulate,  R/S = rain + snow.
                              102

-------
          Table  11-12.  Atmospheric  Concentrations  and  Fluxes
                  of PCBs  in Different  Environments
                 Air
Location
            Range   _Mean
               (ng m~3 )
  Precipitation
Range       Mean
    (ng jO)
                                                       Fluxes
m~2 yr"1
Urban 0.5-30 5-10
Rural 0.12 0.8
Remote 0.02-0.5 0.1
Marine 0.02-2 0.5
Great Lakes 0.4-3 1
10-250 50 100-700
1-50 20 50-500 (1.5)
1-30 5 0.2-20 (40)
0.5-10 1-5 0.1-20 (35)
10-150 20-50 20-150 (4)
NOTE:  Numbers in parentheses represent  approximate ratios of
       urban-to-area  fluxes.
9.   TOTAL DEPOSITION  IN URBAN, RURAL, AND REMOTE ENVIRONMENTS

     The total fluxes  (wet  and dry) of airborne trace organics to
urban and rural environments have been reported for remarkably
few compounds.  Table  11-12 shows the estimated aerosol and pre-
cipitation concentrations of total PCBs and their estimated
fluxes to the Great Lakes and to the urban, rural, marine, and
remote environments of the world.  Admittedly, some geographical
areas overlap but the  estimates are based on concentrations and,
in some cases, the fluxes were actually observed in the respective
areas.  It is useful to compare the fluxes estimated for urban
areas with those estimated  for the Great Lakes and the rural,
marine, and remote environments.  The ratio of urban PCB fluxes
to remote and marine fluxes is -^40, to the Great Lakes -v-4, and to
rural areas near urban centers -^1.5.  Waters in the respective
areas also show levels of contamination in the order of their
increasing input ratio.  On the basis of calculations made by
Bidleman and Olney (1974b), total DDT inputs ought to be ^lO1
less than inputs for PCBs but relative ratios should remain
approximately constant.

     Table 11-13 shows the fluxes of selected chlorinated hydro-
carbons estimated for  rural areas and based on the following
criteria:   (1) dry-deposition velocities for gases and submicron
particles are -\-0.1-1.0 cm s"1 (see Part III) and (2) wet fluxes
from rain and snow concentrations and precipitation intensities
are 70 or 100 cm yr'1.  These values are typical for the upper
                               103

-------
           Table 11-13.   Atmospheric Fluxes  of Trace-Organic  Compounds  to Rural Environments

Dry

Concentration Flux
Compound
Total PCB
Total DDT
Dieldrin
Aldrin
Toxaphene
a-BHC
(Lindane)
Hexachloro-
benzene
Chlordane
Heptachlor
Dibutyl-
phthlate
Diethylhexyl-
phthalate
Total PAHs
NOTE : For dry
for wet
Range Mean 0 . 1
( ng m 3 ) ( cm s : , p g
0.1-2 1 32
0.01-0.05 0.03 1
0.01-0.1 0.05 1.6
0.1-1 0.5 16
0.02-2 0.5 16
1-4 2 63
0.1-0.3 0.2 6.3
0.01-0.5 0.1 3.2
1-5 2 63
0.5-5 2 63
5-5 2 63
10-100 20 630


— _
flux, (jg m yr — (Cone.
m 2 yr
315
9. 5
16
160
160
630
63
32
630
630
630
6300
s"1) (3
ng H ~ l )

Wet
Concentration Flux
Range
1 ) (ng
15-50
1-10
1-4
0.5-3
1-10
1-15
1-4
1-5
1-9
4-40*
4-40*
50-300
.15 x 107 s
1Q3£
( - j (m yr
m3
Mean 70 _ _100
i 1) (cm yx~l,\ig m 2 yr '
25 18 25
5 3.5 5
21.4 2
21.4 2
53.5 5
8 5.6 8
21.4 2
21.4 2
3 2.1 3
10 7 10
10 7 10
100 70 1000
T 1 A "~ 3
-rrT-^ix / . r _ . w H9wri1~LT-t
yr j ( ) ( ) (con
102 cm ng
-i)(10^fl,.
ng
Ratio of
wet: dry flux
V =0.1 cm s l
) Po=70 cm yr"1
0.56
3.5
0.88
0.09
0.22
0.13
0.22
0.44
0.03
0.11
0.11
0.11
c. ng m"3 ) ,-

*Estimated from W = 10s for Concentration = 0.5-5 ng m

-------
Midwest at the western boundary of the Great Lakes (70 cm yr~M,
increasing eastward to the United States Atlantic coast (100 cm
yr 1).  In some cases trace-organic concentrations in precipita-
tion were estimated from washout ratios (W) of -^105 .  These
numbers represent the maximum input for Midwest and East Coast
areas in urban and industrial zones.  An adjustment may be re-
quired for other areas.

     The atmospheric fluxes of trace organics listed in
Table 11-13 are only estimates based on the available data base,
which is small.  They are probably accurate by an order of mag-
nitude only and should be construed to represent ranges for
researchers and governmental agencies to use.  One question often
posed is, what is the relationship between wet and dry fluxes?
Using conservative estimates for both dry (V, = 0.1 cm s"1 ) and
wet fluxes (P  = 70 cm yr"1), the values in the last column of
Table 11-13 suggest that dry deposition exceeds wet deposition by
1 to 5 times.  This is in contrast to atmospheric metal fluxes,
which are about evenly divided between wet and dry deposition
(Parts I and III).
10.  APPLICATIONS OF FLUX CALCULATIONS:  ATMOSPHERIC PCB INPUT TO
     THE GREAT LAKES

     The deposition of atmospheric organics to water bodies can
logically be separated into dry and wet deposition.  The proces-
ses affecting deposition and the various theoretical or empirical
approaches to estimating atmospheric fluxes have already been
discussed.  Thus, a relatively large body of information exists
on the processes controlling PCB deposition to water surfaces and
the environmental distribution of PCBs.  For these reasons, the
flux of PCBs from the atmosphere to the Great Lakes was estimated
as an example of a general approach to the problem.  The Great
Lakes were selected because they contain the greatest degree of
fish contamination not related to a specific source.  For exam-
ple, most game and commercial fishes in Lake Michigan, part of
which is in the industrialized region of northern Indiana and
Chicago, Illinois, show significant accumulations of PCBs
(>2 ppm).  Although Lake Superior is in a forest-dominated
watershed, PCBs in fish are commonplace.  This supports the
postulate that PCBs are distributed over long distances by
atmospheric transport and deposition.


10.1 Dry Deposition

     The dry deposition of particulate matter has been estimated
using the following equation:
                               105

-------
                         F = VCair part)'                 t10'1)


where F is the flux  (ng m~2 s"1); Vd, the  dry-deposition velocity

(m s"1); and C .      ,_,  the concentration  in  atmospheric partic-
              3.X2T pciiT...
ulate matter (ng m 3).  Available data suggest  that ^5%-15% of
the atmospheric PCBs in the particulate phase (Table 11-14) are
associated with submicron-size particles and  have  Vd values of

0.001-0.005 m s"1.  By using Cair part =0.13 and  0.15  ng m"3,

respectively, for Lakes Michigan (Doskey 1978)  and Superior
(Eisenreich and Hollod 1980) and Vd values of 0.005 (Doskey and

Andren 1980) or 0.0013 m s'1 (Hollod 1979), the  dry particle PCB

flux may be calculated as -^20-22 M9 ™~2 yr"1 .

     The dry deposition of vapor-phase PCBs directly to a water
surface is estimated from (MacKay et al. 1979)
                         F = K
                              og
                                   CH-P1
                                   	                     (10-2)
                                    RT J
where F is the flux (mol m 2 h"1); K   , the overall gas-phase

mass-transfer coefficient (m h"1); C,  the PCB concentration  in
dissolved phase in water (mol m 3); H, the Henry's law  constant
(atm m3 mol"1); P, the atmospheric partial pressure of  PCB
(atm); and RTi the gas constant times  absolute temperature
(m3 atm"1 mol"1 ) .

     By assuming gas-phase control for PCB transfer across the
air/water interface (Doskey and Andren 1980; MacKay et  al . 1979)
and applying the parameters in Table 11-14, the vapor-phase  flux
was estimated as ^2800 kg yr"1 (^48 m~2 yr"1 ) and 6000  kg yr"1
(^73 (jg m"2 yr"1 ) for Lake Michigan and Lake Superior,  respec-
tively (Table 11-15).   The higher PCB-vapor flux to Lake Superior
is primarily a consequence of the higher atmospheric concentra-
tion and higher relative percentage assumed in the gas  phase
(•v-90%).  Gas-phase control for PCB transfer across the  air/water
interface was based on the Henry's law constant (10~6-10~7 atm
m
 3
   mol M and arguments presented by Doskey and Andren  (1980).
10.2 Wet Deposition

     Wet deposition is calculated for gas scavenging by

                         F = a(PQ)Cb,                       (10-3)

where F is the flux (ng m"2 yr"1); a, the PCB-vapor washout
                               106

-------
               Table  11-14.   Parameters for PCB  Flux to the  Great  Lakes
                                 Lake Michigan
                            1242
                          1254
                                                 Lake Superior
                     1242
                                                                                         1254
Concentration in air
 Vapor (atm)
       (ng m~3)
 Particulate  (ngm~3)
5.6 x 10~14
Concentration in water
 (mol m~3)              7.7  x  10~9

Aerosol-deposition
 velocity (m s"1)

Air/water partition
 coefficient
 (atm m3 mol"1)          2.8  x  10~7

Overall gas-phase mass-
 transfer coefficient
 (Kog)(m h'1)              7.9

RT (25°C)

Surface area (m2)
               0.87
               0.13
                                      0.005
1.5 x 10~4
                                              2.6 x 10~8
                                              1.4 x 10"
                                                 7.0
                                   2.45 x 10~2

                                     5.9 x 1010
                                                               7.9 x 10
                          14
                                       3.8 x 10~9
                                       1.4 x 10
                               1.35
                               0.15
1.8 x 10
                                      3.1 x 10
                                                                     '14
                                                     0.0013
                                      6.9 x 10~8
                                          7.9                   7.0

                                                   2.45 x 10~2

                                                   8.21 x 1010
SOURCE:   For Lake Michigan  parameters, Doskey and Andren (1980);  for Lake  Superior parameters,
         Eisenreich and Hollod  (1980).

-------
             Table 11-15.  Atmospheric Flux_of PCBs  to
                     the Great Lakes (kg yr"1)

Dry deposition
Particulate
Vapor
Wet deposition
Total input
Lake Michigan
Calculated
1,200
2,800
5,000
9,000
Lake Superior
Calculated
70-2,900
6,000
2,800-3,600
8,900-12,500
Measured
5,100
1,500-3,000
6,600-8,300
SOURCE:  For Lake Michigan values, Doskey and Andren  (1980),
         and (wet deposition), Murphy and Rzeszutko (1977); for
         Lake Superior values, Eisenreich and Hollod  (1980).

NOTE:    Dry deposition:  flux = V,(C .  );
                                  Q  Cll JL.

         wet deposition:  flux = C  .  (P ).
                                        o
coefficient (H"1); P ,  the rainfall intensity (m yr"1); and C,,
the vapor phase concentration (ng m~3) at a reference height
above the water surface.  The washout of particles containing PCB
can be calculated from the following:
                         F ='W(PQ)Cb
                                                           (10-4)
where F is the flux (ng m~2 yr"1); W, the PCB-particle washout
coefficient; and Cb, the particle-phase concentration of PCB at a

reference height above the receptor surface (ng m~3).  The values
of a and W have been empirically derived or measured.  However,
Murphy and Rzeszutko (1977, 1978) and Eisenreich and Hollod
(1980) determined the wet flux from
                    Flux = C  .  (P )
                            rainv  o'
                                                           (10-5)
where
            is the total  volume -weighted PCB concentration in
            )  and PQ,  the rainfall or snowfall in
(m yr"1)  such  that flux has  units  of |jg m~2 yr"1 .
rain (|jg m~3) and PQ,  the rainfall or snowfall intensity
                               108

-------
     Eisenreich and Hollod also derived values for W and a from
data proyided_by Slinn et al. (1978).  Wet-flux rates varied from
•^85 ug m 2 yr"1 for Lake Michigan to -^27-34 pg m~2 yr"1 for Lake
Superior (Table 11-15).

     In addition, Eisenreich and Hollod (1980) have estimated
loadings to Lake Superior by experimentally determining a V,

value, calculating total dry flux, and estimating wet loadings
based on rain and snow concentrations of 25-50 |jg 2"1 .

     The relative importance of the atmospheric input of PCBs to
the Great Lakes compared with other sources is great.  Approxi-
mately 85%-90% of the total input to both lakes is deposited from
the atmosphere (Doskey and Andren 1980, Doskey 1978, Eisenreich
and Hollod 1980,  Hollod 1979).
11.  RESEARCH RECOMMENDATIONS

     The following recommended research is necessary to relieve
the paucity of data that exists on trace organics in the atmos-
phere and their deposition and effects.

     (1)  Methods must be developed to distinguish between vapor-
          and particulate-phase, high-molecular-weight organics
          in the atmosphere.
     (2)  The relationship between the mass median diameter,  the
          deposition velocity, and the receptor surface for
          atmospheric particles must be established.
     (3)  The size distribution of atmospheric particulates
          containing high-molecular-weight organics must be
          determined.
     (4)  Collection methods for dry deposition of particulate-
          and vapor-phase organics should be developed since this
          is an important fraction of total deposition.
     (5)  The collection efficiency of different receptor sur-
          faces for particulate- and vapor-phase organics must be
          determined.
     (6)  Accurate data on vapor pressures and water solubilities
          of slightly soluble, high-molecular-weight organics
          need to be developed.
     (7)  The fugacities (true dissolved fraction) of trace
          organics in water need to be developed.
     (8)  The trace-organic compositions of rain and snow and of
          the aerosol and gas phases in the atmosphere need to be
          determined before ,accurate deposition rates and eco-
          logical or health impacts can be evaluated.
                               109

-------
12.   REFERENCES

Atkins DHF,  Eggleton EEJ.  1974.  Studies of atmospheric washout
and deposition of -BHC,  dieldrin,  and p, p'-DDT using radiola-
belled pesticides.  In:   Nuclear Techniques in Environmental
Pollution, 521, International Atomic Energy Agency, Vienna.

Bidleman TF, Christensen EJ. 1979.  Atmospheric removal processes
for high molecular weight organochlorines.   J. Geophys. Res.
84:7857.

Bidleman TF, Olney CE. 1974a.  Chlorinated hydrocarbons in the
Sargasso Sea atmosphere and surface water.   Science 183:516.

	. I974b.  High-volume collection of atmospheric polychlor-
inated biphenyls.  Bull. Environ.  Contain. Toxicol. 11:442.

Bidleman TF, Rice CP, Olney CE. 1976.  High molecular weight
chlorinated hydrocarbons in the air and sea:  Rates and mecha-
nisms of air/sea transfer.  In:  Marine Pollutant Transfer,
Windom H, Duce RA, eds. , Heath & Co., Lexington, MA.

Cautreels W, vanCauwenberghe K. 1978.  Experiments on the distri-
bution of organic pollutants between airborne particulate matter
and the corresponding gas phase.  Atmos. Environ. 12:1133.

Chen PH, Shieh HH, Gaw JM. 1980.  Determination of PAH in air-
borne particulates at various locations in Taipei city by GC/MS
and glass capillary GC.   In:  Abstracts of American Chemical
Society Meeting, Houston,  Texas, 60.

Clayton JR,  Jr, Pavlou SP, Breitner NF. 1972.  Polychlorinated
biphenyls in coastal marine zooplankton:  Bioaccumulation by
equilibrium partitioning.   Environ. Sci. Technol. 11:676.

Dexter RN, Pavlou SP. 1978.  Mass solubility and aqueous activity
coefficients of stable organic chemicals in the marine environ-
ment:  PCBs.  Marine Chem. 6:41.

Doskey PV. 1978.  Transport of airborne PCBs to Lake Michigan.
Thesis, University of Wisconsin, Madison.

Doskey P,  Andren AW. 1979.  High-volume sampling of airborne
PCBs with Amberlite XAD-2.  Anal.  Chim. Acta, 110:129.

Doskey PV, Andren AW.  1980.  Modeling the flux of atmospheric
PCBs across  the air/water interface.  Environ. Sci. Technol.  (in
press).
                               110

-------
Eisenreich SJ, Hollod GJ. 1980.  Atmospheric input of PCBs to
the Great Lakes; impact on the Lake Superior ecosystem.  In:
Atmospheric Input of Pollutants to Natural Waters, Eisenreich SJ,
ed., Ann Arbor Science Pubs,  (in press).

Giam CS, Chan HS, Neff GS, Atlas E. 1978.  Phthalate ester
plasticizers:  A new class of marine pollutant.  Science 199:419.

Giam CS, Atlas E, Chan HS, Neff GS. 1980.  Phthalate esters, PCB
and DDT residues in the Gulf of Mexico atmosphere.  Atmos.
Environ. 14:65-69.

Gilbertson M. 1976.  Background to the regulation of PCBs in
Canada, Task Force on PCB.  Rep. to Environmental Contaminants
Committee of Environment  Canada and Health and Welfare, Canada.

Grosjean D, Friedlander SK. 1975.  Gas-particle distribution
factors for organic and other pollutants in the Los Angeles
atmosphere.  J. Air. Pollut. Control Assoc. 25:1038.

Harvey GR, Steinhauer WG. 1974.  Atmospheric transport of poly-
chlorinated biphenyls to  the North Atlantic.  Atmos. Environ.
8:777.

Hollod GJ. 1979.  PCBs in the Lake Superior ecosystem:
Atmospheric deposition and accumulation in the bottom sediments.
Dissertation, University  of Minnesota, Minneapolis.

Junge CE. 1975.  Transport mechanisms for pesticides in the
atmosphere.  Pure Appl. Chem. 42:95.

Junge CE. 1977.  Basic considerations about trace constituents
in the atmosphere as related to the fate of global pollutants.
In:  Fate of Pollutants in the Air and Water Environments,
Part 1.  (Advances in Environmental Science and Technology, Vol. 8)
Suffet IH, ed., Wiley-Interscience, New York.

Karickoff SW, Brown DS, Scott TA. 1979.  Sorption of hydrophobic
pollutants on natural sediments.  Water Res. 13:241.

Liss PS, Slater PG. 1974.  Flux of gases across the air-sea
interface.  Nature 247:181.

MacKay D, Wolkoff AW. 1973.  Rate of evaporation of low-
solubility contaminants from water bodies to the atmosphere.
Environ. Sci. Technol. 7:611.

MacKay D, Shiu WY,  Sutherland RP. 1979.  Determination of air-
water Henry's Law constants for hydrophobic pollutants.  Environ.
Sci. Technol. 13:333.
                               Ill

-------
McClure VE. 1976.   Transport of heavy chlorinated hydrocarbons in
the atmosphere.   Environ.  Sci.  Technol. 10:1223.

Melnikov NN. 1971.   Chemistry of Pesticides, Springer-Verlag,
New York.

Mosser JL, Fisher NS,  Wurster CF.  1972.  Polychlorinated biphe-
nyl's and DDT alter species composition in mixed cultures of
algae.  Science 176:533.

Murphy TJ, Rzeszutko CP.  1977.   Precipitation inputs of PCBs to
Lake Michigan.   J.  Great Lakes  Res.  3:305.

	. 1978.  PCBs in precipitation in the Lake Michigan
basin.  Rep. to US-EPA on Grant No.  803915, Environmental Re-
search Laboratory,  Duluth (EPA-60013-78-071).

Murphy TJ, Heeson TC,  Young DR,  McDermott-Ehrlich D. 1980.
Evaluation of a technique for measuring dry aerial deposition
rates of DDT and PCB residues.   Atmos.  Environ, (in press).

Nisbet ICT, Sarofim AF.  1972.  Rates and routes of transport of
PCBs in the environment.   Environ.  Health Perspect. 1:21.

Peel DA. 1975.   Organochlorine  residues in Antarctic snow.
Nature 254:324.

Pupp C, Lao RC,  Murray JJ, Pottie RF. 1974.  Equilibrium vapor
concentrations of some PAH,  As406,  and Se02 and the collection
efficiencies of these compounds.  Atmos. Environ.  8:915.

Risebrough RW,  Rieche P,  Peakall DB, Herman SG, Kirven NM.
1968a.  PCBs in the global ecosystems.   Nature 232:50.

Risebrough RW,  Griffin JJ, Huggett RJ,  Goldberg ED. 1968b.
Pesticides:  Trans-Atlantic movement in the Northeast Trades.
Science 159:1233.

Simon CG, Bidleman TF. 1979.  Sampling airborne PCB with poly-
urethane foam plugs.  Anal.  Chem.  51:1110.

Slinn WGN, Hasse L, Hicks BB et al.  1978.  Some aspects of the
transfer of atmospheric trace constituents past the air-sea
interface.  Atmos.  Environ.  12:2055.

Slinn SA, slinn WGN. 1980.  Modeling of atmospheric particulate
deposition to natural waters.  In:   Atmospheric Input of Pollut-
ants to Natural Waters,  SJ Eisenreich,  ed., Ann Arbor Science
Pubs, (in press).
                               112

-------
Strachan WMJ, Huneault H.-1979.  Polychlorinated biphenyls and
organochlorine pesticides in Great Lakes precipitation.   J.  Great
Lakes Res. 5:61.

Van Vaeck L, vanCauwenberghe K. 1978.  Cascade impactor measure-
ments of the size distribution of the major classes of organic
pollutants in atmospheric matter.  Atmos. Environ. 12:2229.

Wildish DJ. 1972.  Polychlorinated biphenyls (PCBs) in sea water
and their effect on reproduction of Gammarus Oceanicus.   Bull.
Environ. Cont. Toxicology 7:182-187.
                               113

-------
                             Part III

               GUIDE FOR ESTIMATING DEPOSITION RATES
                       OF GASES AND AEROSOLS

         B.  C.  Scott (Chairman), L. A. Barrie, D. F. Gatz,
                    J.  M. Miller and D. H. Pack
1.   INTRODUCTION

     The natures of depositional processes and atmospheric
transport are such that,  as space and time scales increase, the
ability to estimate deposition rates improves.  In this section,
no attempt has been made to describe depositional processes that
might occur over a few tens of kilometers or within a few hours.
Rather the removal parameterizations are for materials relatively
uniformly concentrated in the lowest several kilometers of the
atmosphere and for time and space scales on the order of 1 year
and 1,000 kilometers.  These scales permit atmospheric variabil-
ity in directional transport, storm paths, stability, etc., to be
smoothed or averaged.

     A guide for estimating deposition rates for gases and aero-
sols has been provided to be used with specific information about
the physical and chemical properties of the gas or aerosol in-
volved.  Simple rules for estimating both wet- and dry-deposition
rates are presented in terms of parameters called washout ratios
and deposition velocities.

     The following tables and formulas provide only deposition
rates.  The user must determine the time interval involved to
compute total deposition.  However, the procedure for determining
relative contributions of wet and dry deposition on an annual
basis is also given.  Air concentrations of the material involved
must also be known as well as how it is distributed, e.g., as a
gas, as an aerosol, or as both (see Parts I and II).

     A great deal of uncertainty can result from using the ap-
proaches outlined below.   The deposition rates obtained from the
simplified tables and formulas should be considered accurate only
to within a factor of 2 or 3 of the true values.  More accurate
estimates would require extensive investments of time and money.
If applications of the procedures presented here suggest that
particular materials are potentially damaging, then more accurate
procedures or even laboratory or field experiments should be used
to refine those estimates.  Indeed, one part has been devoted to
considerations that must be recognized when computing deposition
where the well mixed, annually averaged atmosphere is clearly not
appropriate.  There are also references for complex models and
for publications with additional details on deposition calcula-
tions .
                               114

-------
     Finally, implementation of the research recommendations as
listed at the end of this section is mandatory to the advancement
of the existing knowledge on estimating the deposition of pollu-
tants .
2.   WET REMOVAL AND DEPOSITION OF AEROSOLS AND GASES

2. 1  Wet Removal of Aerosols

     Regardless of how the wet removal of aerosols from the
atmosphere is described, the concentration of the pollutant x

associated with precipitation must first be considered.  Because
most aerosol pollutants associated with precipitation first
attach to suspended cloud droplets, both the collection of and
the chemistry of the tiny cloud droplets must be considered in
deriving x .  Aerosols not attached directly to cloud droplets

can also be scavenged by falling precipitation particles.

     What is being considered here is therefore a system with two
rate-limiting processes (listed below).  First, an aerosol atta-
ches to the condensed water; second, the condensed water falls.
The slower of the two rates determines the rate of deposition at
the surface.

     Rate Limiting Processes

     (1)  Attachment
          (a)  Collection efficiency
          (b)  Solubility
          (c)  Size
          (d)  Condensation, evaporation
          (e)  Age
     (2)  Removal
          (a)  Precipitation growth processes  (riming, accretion,
               etc. )
          (b)  Storm efficiency
          (c)  Seasonal variations

2.1.1  Attachment

     The collection and collision efficiencies are the most
important and the most difficult attachment processes  to esti-
mate.  Figure III-l illustrates how sensitive  collection is to
the size of the collected and collector particles.

     Particulate solubility is also important  in determining  the
size of the aerosol.  During precipitation, subcloud humidities
are typically between 90% and 100%.  At these  humidities,
initially dry, soluble aerosols will rapidly increase  their
dimensions by a factor of 4 or more.  Figure II1-2 shows the
                               115

-------
                    Radius of Unit Density Particles
                                          10n
Figure III-1.  Collision efficiency as  a  function of size
of the collected particles.   Radii of collector particles
are 0.1 mm  (dashed  line) and 1.0 mm (solid line) (after
Slinn 1977).
                0  10  20 30 40  50  60 70
                        % Relative Humidity
80  90 100
Figure II1-2.   Theoretical growth curves for solution
droplets of sulfuric acid and some inorganic salts at 25°C
(after Tang and Munkelwitz 1977).  A  =  NaCl; B = H2SO4;
C = NH4HSO4; D  =  (NH4)2SO4.
                             116

-------
 10.0 —
5
5
                                     Figure  II1-3.   Growth  of
                                     aerosol  composed  of  50%
                                     ammonium  sulfate.
                                     S  =  0.1%;  (NH4)2S04  =  50%.
relationship between the size of various soluble materials and
humidity-

     How long it takes a soluble aerosol to grow to cloud-droplet
size depends on what fraction of the particle is composed of
soluble material and on various cloud micro-physical properties
(such as, updraft velocity, temperature, and the number of aero-
sols competing for available water).  Figure II1-3 illustrates
how an aerosol composed of 50% ammonium sulfate responds after
being drawn into a cloud with water supersaturation of 0.1%.
Under these conditions, the growth of a submicron-size aerosol
to cloud-droplet size  (10-|jm diameter) takes 2 min or less.

     Aerosol growth in the humid-cloud environment also presents
a likely pathway for the removal of insoluble material from the
atmosphere.  Consider, for example, insoluble lead (Pb) aerosol,
thought to be attached to particles with a mass median diameter
(mmd) of ~0.5 (jm.  The tiny Pb-containing particles would ordinar-
ily be difficult to remove (see Figure III-l) because of their
                               117

-------
size.  However,  if the Pb is attached to a particle that is
partially soluble,  then,  upon entering a cloud, it would almost
instantly be incorporated into the cloud water and within minutes
would achieve a maximum collision efficiency of 1.  This rapid
growth of aerosol to cloud-droplet size is often called nuclea-
tion scavenging.  Thus,  the age of an aerosol is also important;
the older an aerosol,  the larger its mean diameter—because of
coagulation—and the greater the likelihood that it will attach
to a soluble aerosol.

2.1.2  Removal of Aerosol

     To properly discuss removal processes, distinctions must be
made between ice growth and water growth after cloud and precipi-
tation water have formed.  Ice-growth processes probably start
most precipitation in our latitudes.  However, during warm seasons
or at high freezing levels (~3 km),  the coalescence process invol-
ving water/water collisions becomes  the dominant precipitation-
growth mechanism.  Condensation onto drops always remains a minor
factor in the growth of precipitation.  Thus when coalescence is
important, a 1-mm raindrop is produced from about a million
collisions with cloud droplets.  Rather than being diluted, as
would occur if condensation were important, the concentration of
a pollutant is increased or decreased depending on the amount of
pollutant in the collected droplets.  That a raindrop acts as a
concentrating agent is clearly demonstrated by noting that one
drop takes 106 cloud droplets with a mean spacing of ~1 mm between
droplets and places them all into a  volume of about 1 mm3,  i.e.,
a concentration factor of 106.  This concentrating of pollutant
is the physical basis behind the washout ratio W.

     The washout ratio is an empirical relationship, valid over
one month or more.   Expected pollutant concentrations in rain can
be empirically related to known or assumed concentrations in the
air at ground level on the basis of  previous measurements of both
quantities.  When W is expressed as  a volume-weighted ratio (mass
of pollutant per volume of rain water divided by mass of pollut-
ant per volume of air),  the values are typically near 106,  imply-
ing that (1) the aerosol has been readily incorporated into the
cloud water, (2) the aerosol is probably soluble, and (3) precip-
itation is formed primarily through  a coalescence process.   Also
implied is that the concentration of a pollutant measured in air
near the ground is  representative of the concentration averaged
over approximately 3 km--from the ground through the cloud.

     Physical interpretation of the  scavenging mechanisms is
difficult if the precipitation is snow or is falling from a cold
cloud with low freezing levels—particularly if the washout ratio
is on the order of 105 .   Although precipitation from these cold
clouds may be predominantly snow, they can contain abundant
quantities of supercooled liquid water (average concentrations
                               118

-------
   1600
   1200
    800
    400
                             mmd
Figure I I 1-4.  Variation of the precipitation-weighted mean wash-
out ratio W (dimensionless) with the mass median diameter (mmd).
Cross-hatched area = Chilton, United Kingdom (after Cawse 1977);
straight line and • = 50 km northwest of St. Louis, MO (after
Gatz 1975); dashed line and o = 15 km northeast of St. Louis, MO
(after Gatz 1975).
near -^0.1 g m~3 ).  The collection of this supercooled water by
snow flakes (called riming) is the dominant mechanism for remov-
ing soluble aerosols from cold clouds.  However  in spite of this
water-mass contribution  from riming, one-half or more of the
precipitation mass from  cold clouds usually results  from vapor
deposition (the  ice analogy of liquid-phase condensation).  In
other words, in  cold clouds precipitation particles  grow primari-
ly because water vapor is depositing upon their  surfaces by
sublimation.  This direct deposition of vapor is only of minor
importance in warm clouds.  Thus, in cold clouds only 105 or
fewer collisions with cloud droplets are needed  to produce a  1-mm
raindrop at ground level—the additional mass being  obtained  by
vapor deposition.  In a  cold cloud, therefore, washout ratios for
soluble aerosols should  drop to 105 or less.  Thus,  a washout
ratio of £105 can indicate that efficient ice-phase  mechanisms
are operating in a cloud.  Equally likely, however,  is that such
low washout ratios imply that an aerosol is not  being efficiently
attached to the  cloud water.  Aerosols that are  insoluble or
relatively young, or have dimensions between 0.1 to  1.0 |jm are
not easily scavenged (Figure III-4).
                               119

-------
2.2  Wet Deposition of Aerosols

2.2.1  Air Concentrations

     Before leaving the discussion of washout ratios and aerosol
scavenging, it is important to show how the surface-deposition
rate D is related to the washout ratio.

     Surface-deposition rate D is given by

                        D = (JC)Q,                          (2-1)


where C is the pollutant concentration in precipitation (units of
mass of pollutant per mass of water) and J is the precipitation
rate (units of mass of water per unit area per unit of time).
Washout ratio W is defined as

                        W = -r^y-  '                        (2-2)
                             
-------
          Table III-l.  Washout Ratios (volume basis)
  Particle               Convective               Continuous
    mmd                   rainfall                 rainfall
                          5 x 10s                    106

                            10s                    2 x 10s
     The respective W values for convective and continuous pre-
cipitation reflect what is currently known about wet removal for
these two classes.  Within these classes, W may vary as a func-
tion of precipitation rate.  Further research is needed both to
verify the overall average values for each class and to gather
information on how W varies with precipitation rate within each
class.

     These calculations will yield only a gross estimate of the
deposition of a given material, accurate to within a factor of
perhaps 3.  If such an assessment shows even a marginal potential
for harmful effects from wet or dry deposition, laboratory or
field studies should measure the actual atmospheric-deposition
characteristics of the substance in question.


2.3. Wet Removal of Gases

     Unlike aerosol scavenging, gas scavenging is often revers-
ible.  Gases can be absorbed, desorbed, and chemically altered in
condensed water.  As with aerosols, however, the phases of the
condensed cloud water and of the collector particles are import-
ant in determining the deposition rate at the surface.

     In general, the flux of a gas into or out of a single rain-
drop can be related to the diffusivity of the gas D and to the
gradient of gas concentration C near the drop of radius rQ:


                         Flux = D -^-                      (2-4)


     The change of pollutant mass within a drop of radius rQ is
then given by


               Vt a£- = 4nroD(Ceq-C)(l+F),         (2-5)
                               121

-------
where V  is the fall speed of a drop, Ceg the equilibrium concen-

tration of a gas in water at an air concentration of XQ, and C

the actual concentration of a gas in water.  The term (1+F) is a
semi-empirical correction for the ventilation of a drop as it
falls through the environment.  From Eq.2-5, an e-fold equilib-
rium distance can be computed (i.e., the fall distance Z required
to reach ^63% of the equilibrium value):

                                r 2V
                         Z -^    ro Vt                       (2-6)
                               3D(1+F)

     Thus, we see that the bigger the drop and the smaller the
diffusivity, the longer it takes for a drop to come into equilib-
rium with the environment.  However, using reasonable values for
the parameters in Eq.2-6 establishes

                              Z < 1 m .                      (2-7)

The equilibrium time t is

                              t < 1 s.                      (2-8)

     Therefore, the gas concentration in liquid water should be
very close to the equilibrium value determined by the gas concen-
tration in the environment; i.e.,


                              C^ Ceq
where                                                       (2-9)

                              c-=  Is


and H is the Henry's law constant.

     For liquid precipitation, this equilibrium concentration is
strongly controlled by surface-air concentrations.  For snow,
however, this equilibrium concentration is determined at the al-
titude of impact between supercooled cloud droplets and collect-
ing snowflakes.  The collisions rapidly freeze the cloud droplets
and prevent any readjustment to near-equilibrium values at lower
altitudes.

     Thus to determine correct wet removal rates by snow, the air
concentrations of pollutants within a cloud must be estimated.
Unfortunately this cloud-level concentration is rarely known, and
the best one can do is to use surface-level concentrations in the
removal calculations.
                               122

-------
2.4  Wet Deposition of Gases

     Recalling the definition of the washout ratio (Eg.2-2), for
gases we have

                              W = -g   .                     (2-10)

The surface deposition rate (see Eg. 2-3) is then given by


                            D -    JXa -                    (2-11)
                            D ~   ~H~~

The values used for the solubility expression H must be appropri-
ate for typical atmospheric conditions.  That is, H must be
evaluated at temperatures between 0° and 25°C, at different air
concentrations x   and at different ionic concentrations in
rainwater (e.g., pH).

     To compute the deposition rate of gases, a washout ratio is
not needed; as noted in Eg.2-11, knowing the Henry's law constant
is sufficient.  Therefore, as a first approximation,  wet removal
of gas is independent of season and location and depends only
upon the Henry's law constant.

     Groups of materials, e.g., PCBs, DDT, etc., that have simi-
lar values of H should be identified.  A single value of H could
then be used for certain classes of molecules, thereby eliminat-
ing the need for tests for every material within a class.

     Eg.2-11 must be used with caution.  Various gases thought
to be harmless or proven to be inefficiently removed by different
depositional processes may combine with or transform to materials
(gases or aerosols) with entirely different properties.  Research
concerning the transformation products of every new gas is manda-
tory.

     Surfaces do not necessarily retain gases deposited on them.
For example, many pesticides and PCBs evaporate from the ground
guite readily (Junge 1977, Spenser and Cliath 1969) and Hg can
be reemitted at locations of previous deposition (Hogstrom et al.
1979).  In addition, the drying of a wet surface can result in
reemission caused by increased concentrations in the agueous
phase.
3.   DRY REMOVAL AND DEPOSITION OF AEROSOLS AND GASES

     Particles and gases are removed from the atmosphere not only
by precipitation but also by direct uptake at the earth's surface.
The rate at which this occurs is governed by several processes
acting simultaneously.  Each process has a transfer resistance
associated with
                               123

-------
     (1)  Turbulent transport through the atmospheric boundary
          layer ra;
     (2)  Inertial penetration and molecular diffusion through
          the near-surface air layer rb;
     (3)  Uptake at the surface rg.

The,total transfer resistance is then

                         r = ra + rb + rs  *                 (3'V

The inverse of r is the deposition velocity Vd used to calculate
deposition flux F:
                         F = VdC ,                           (3-2)

where C is the pollutant's atmospheric concentration at 1-10 m
above this surface.  When dry removal is computed, C and Vd can

be considered independent of height above 1 m from the surface.

     The relative importance of atmospheric transport (ra+rb)

and surface capture (r ) is primarily governed by the level of
                      5
turbulent mixing in the boundary layer and the reactivity of the
pollutant with the surface.  If surface resistance to pollutant
transfer is high, as for chlorofluoromethanes, then the surface
processes become rate limiting.  On the other hand, if a surface
is a perfect sink for a pollutant (viz., uptake of SO2 by the
Great Lakes), then the atmospheric-transport processes are rate
limiting.  Many pollutant/surface combinations fall between these
two extremes and, therefore, both atmospheric and surface resist-
ance to transfer must be taken into account.
3.1  Dry Removal of Aerosols

     Most particulates in the atmosphere have a diameter of 0.01-
10 |jm.  Because of the dynamics of particle production, growth,
and interaction, these size ranges are not uniformly distributed
but rather are concentrated in modes (Whitby 1978): a nucleus
mode, centered on particle diameters of 0.05 pm; an accumulation
mode, centered on particle diameters of 0.4 pm; and a coarse-
particle mode, centered on particle diameters of 5 jjm.  The
nucleus mode is maintained by a dynamic equilibrium between
particle production by condensation processes and particle remov-
al by adsorption onto the surfaces of accumulation-mode aerosols.
If production is stopped, particles in the nucleus mode expand to
the accumulation mode by coagulation within a few hours.  Parti-
cles in the accumulation mode are created from nucleus-mode
particles and primary emissions.  The coarse-particle mode com-
prises wind-blown-dust particles and coarse particles from pri-
mary emissions, such as flyash.  Most suspended particulates in
the atmosphere are in the accumulation or coarse-particle modes.
                               124

-------
3.2  Dry Deposition of Aerosols

     Particles are deposited onto surfaces by various processes.
Particles with diameters smaller than 0.3 pm are deposited by
Brownian diffusion, those with diameters ranging from 0.3 to 5 p
by inertial impaction-interception, and those with diameters
greater than 5 pm by gravitational sedimentation.  Because Brown
ian diffusion increases as particle size decreases below 0.3 pm
and inertial interception-impaction increases as particle size
increases above 0.5 pm, there is a minimum in the deposition
velocity in the size range of 0.3-0.5 pm.  The relationship
between deposition velocity and particle size resembles that for
particle-collection efficiency by raindrops (Figure III-l).

     In Table II1-2 particle-deposition velocities to land and
water surfaces are given for two ranges of particle size.  These
are only order-of-magnitude estimates for dry-deposition removal.
The V, values are uncertain by about a factor of 3.  Should a
substance be of serious environmental concern, more detailed
investigations should be undertaken.  Such studies must first
establish the substance's particle-size distribution in the
atmosphere and then measure actual deposition velocities to the
receptor surfaces under field conditions.


      Table III-2.  Particle-Deposition Velocities (cm s-1)


                                   Land           Water
Coarse-particle mode                 1              1
 (1 pm < D < 10 pm)

Accumulation mode                    0.2            1
 (0.1 pm < D < 1 pm)


   NOTE:  D is the mass median diameter; particle size refers
          to aerodynamic size.
3.3  Dry Deposition of Gases

     The quantitative prediction of the dry-deposition velocity
of a gas to the earth's surface is complicated by the multitude
of surfaces involved (see Garland 1978).  Since this assessment
was not meant for such detailed work, a first approach to making
an order-of-magnitude estimate of deposition velocity is outlined
below.  Should the gas involved be of environmental concern, more
detailed studies must be done under actual field conditions.
                               125

-------
     As mentioned before,  the dry-deposition rate of gases is
dependent on the resistance to atmospheric transport and on the
resistance to uptake at the surface.   For most environmental
surfaces, the surface resistance to S02 -uptake is of the same
order of magnitude as the  atmospheric-transport resistance;
consequently S02 deposition rates are controlled by both atmos-
pheric and surface processes.  However, if a gas is 10 times less
reactive with a surface than SO2 is,  its deposition is mainly
limited by surface uptake.   By making a simple laboratory compar-
ison of the relative reaction of a gas and sulfur dioxide with an
environmental surface,  an  es.timate can be made of that gaseous
deposition velocity.  Therefore, the  following procedure is
recommended.

     Through a laboratory  study, determine the Jekyll number Jfc
defined as
                         T  -   s' substance x               ,„ _
                            =                 '              '
The Jekyll number is the surface resistance of the substance
relative to that of sulfur dioxide.   Experimentally it is the
ratio of the half-life decay time,  Tx ,  of each substance in a
                                    •^
well-mixed chamber containing the natural surface of interest in
a condition similar to that found in nature (see Payrissat and
Beilke 1975).  Actual conditions must be simulated.  For instance,
if the uptake of chlorofluorome thanes by sand is being tested,
sunlight should be present.  Wall losses in the chamber must
always be taken into account.

     If J^ is 10 or greater,  the atmospheric dry-deposition

process is controlled by the surface uptake.  Surface uptake
rates are actually measured in the chamber experiment and the V,
in cm s"1 is given by

                         v   _ (100 h)ln2                   ,3_4v
                         vd        T,                        {   '
                                    \
where h is the height of the test chamber in meters and T,  is
                                                         %
the half-life of the gaseous substance of interest in seconds in
a 1-m2 chamber of height h whose bottom is covered with the
natural surface of interest and in which air is well mixed by
fans .

     Jf Jk is less than 10, the dry-deposition process is con-
trolled not only by surface uptake but also by transport through
the atmospheric boundary layer.  The latter is governed by the
time of year and the surface type.   Ranges expected for the V, in
these circumstances are listed in Table I I 1-3.
                               126

-------
       Table II1-3.  Ranges of V, for a Substance with

               a Jekyll Number < 10 (cm s"1)


                                      Season
Surface Type               Summer                Winter
Land
Water
0.5-2
0.05-0.2
0.05-0.2
0.5-2
     These kinds of calculations give only gross estimates of the
deposition of a given material, accurate to within a factor of
perhaps 3.  If such an assessment shows even a marginal potential
for harmful effects from wet or dry deposition, laboratory or
field studies should measure the actual atmospheric-deposition
characteristics of the substance in question.
4.   RELATIVE REMOVAL BY WET AND DRY PROCESSES

     To estimate the relative removal by wet and dry deposition,
the dry flux of a material is approximated as

                         F  = V x                           (4-1)

and the wet flux as

                         Fw = Wx J .                         (4-2)
                          \v     d

     Dry flux can occur throughout the year (denoted as At) but
wet flux occurs only during periods of precipitation (denoted as
At ).   The ratio R of wet to dry deposition is given by
  w
                                                            (4_3)

                      "   VdAt

Here, the product JAt  is the annual precipitation amount A.
Therefore,

                      „ _   AW                              (4-4)

                        "
For example, for a deposition velocity of 0.2 cm s x, a washout
ratio of 2 x 10s and an annual precipitation amount  (At = 3.15 x
107 s) of 90 cm, the ratio of wet to dry deposition would be
                               127

-------
about 3.  This value of 3 implies that 75% of the total deposi-
tion is due to wet removal.
5.   SPECIAL SITUATIONS

     The averagings given in this section may not be applicable
for

     (1)  Unique emission configurations—those that do not
          produce (relatively) uniform atmospheric concentra-
          tions;
     (2)  Unique receptor configurations—primarily single
          locations for which deposition values are required;
     (3)  Unique meteorological environments
          (a)  Coastal areas with frequent sea/land breezes;
          (b)  Arctic environments with persistent low-level
               stability (i.e., surface inversions) and snow
               cover;
          (c)  Precipitation extremes: arid or wet.

Other methodologies can be applied to estimate the deposition for
these situations.


5.1  Unique Emission Configurations

     If the emissions of a compound exceed about 10% of the area
or national total for that compound and if the source of emission
is relatively isolated from other significant releases (500-1000
km away), atmospheric transport and dilution should be carefully
analysed.

     The above value of 10% for a single source is based on an
analogy with sulfur emissions.  Here a single site in Sudbury,
Ontario, is the source of approximately 10% of the total sulfur
emitted in North America.  This emission rate is detectable
several hundreds of kilometers from the stack.

     The techniques for a detailed approach are well known
(Turner 1970, Pasquill 1974, 1978, Haugen 1975).  For short
periods of up to about one month, observations of wind direction,
wind speed,  and stability can be used to calculate the distribu-
tion of a three-dimensional plume concentration.  Dry deposition
can then be calculated by applying the appropriate deposition
velocity (vd as previously defined) with the necessary correc-

tions for the removal of a material from a plume (e.g., Horst
1977, Hogstrom 1979).

     Wet deposition is best estimated from observed precipitation
amounts and types (Hales 1972, 1975).  The techniques are reason-
ably accurate for the distance a plume travels in 10-20 h (~200-
                               128

-------
500 km).  Beyond these distances, vertical concentration becomes
uniform and the methods described previously for dry deposition
can be applied along the path the material travels.

     On the other hand, wet removal of a material from a discrete
source at long distances is much less certain since precipitation
indicates significant vertical motion.  These vertical motions
carry material aloft into layers moving in different directions
and at different speeds.  They further permit the incorporation
of compounds into clouds with the subsequent possibility for
aqueous chemistry modification.  Studies of storm characteristics
(e.g., Kreitzberg and Leach 1979) indicate the rapid incorpora-
tion of surface air into higher levels of storm systems, some-
times from a location 180° opposite to that of the surface
airflow.

     There are few validated wet-deposition data for these situa-
tions.  However, Smith and Hunt (1978) have shown that individual
storms produce large amounts of wet deposition even from area
sources.  Also nuclear-test debris has been removed by rain
thousands of kilometers from its source, further supporting both
the significance of the process and its explicit relation to
individual storms.

     Accurate quantitative evaluation of this phenomenon will
require research on the mesoscale flow patterns of precipitation
systems combined with measured data on vertical concentration
profiles and concentrations in collected precipitation (e.g.,
Gatz 1980).


5.2  Unique Receptor Configurations

     These are usually individual locations where deposition
levels are required, e.g., a lake, an urban complex, an agricul-
tural or forest plot.

     As a general approach, site-specific data should be used to
calculate removal.  For uniformly distributed material, the
methodology suggested for the regional approach can be modified
to use actual precipitation data (frequency, amounts, and rain-
fall rates) for wet deposition.  For dry deposition, the actual
or estimated atmospheric concentrations should be used in conjunc-
tion with deposition velocities that reflect the specific rough-
ness characteristics of the actual locations, which can vary with
the season.  For example, dry deposition on an open-water surface
will differ from that on a frozen lake; grassland deposition will
differ from that on a snow-covered surface; and removal by for-
ests in leaf will differ from removal during bare-branched winter
conditions.

     Unfortunately few data on dry-deposition velocity exist for
all these situations, even for the conventional pollutants.  For
                               129

-------
new compounds, surrogate data should be used, i.e.,  data based on
similarity of forms (gaseous or particulate) , size distributions,
and surface reactivities (see Section 3).  At the time  of  this
publication, the single most comprehensive collection of V,  data
is that tabulated by McMahon and Denison (1979), and studies of
this phenomenon are continuing (e.g., Barrie and Walmsley  1978).

     If removal calculations must be carried out for a  compound
that is not well mixed but that originates from a significant
source within, say 500 km,  "back-track" meteorological  trajector-
ies (e.g., Heffter et al .  1975) separated into precipitation/non-
precipitation occurrences will give more relevant statistics.


5.3  Unique Meteorological Environments

5.3.1  Coastal Areas

     These situations will be important, primarily,  to  discrete
sources or discrete receptor calculations of deposition.   For
example,  for a location with a daily onshore sea breeze (non-
source flow) and nocturnal land breeze  (potential source flow —
the sea), deposition calculations would be relevant  only for the
flow from the source, i.e., at night.

     Not  all special circumstances can be covered here.  Coastal,
mountainous, and ocean environments should be carefully consider-
ed before applying any generalized methods.

5.3.2  Precipitation Extremes

     Two  additional environmental circumstances — very arid or
desert areas and very rainy areas — should also be treated  as
special cases.  If the area has only infrequent and  sparse precip-
itation,  first calculate the ratio of wet to dry deposition,  R:

                              -  AW   .                      (5-1)
For annual precipitation amounts typical of deserts, dry deposi-
tion probably predominates.  For example, an annual rainfall of
only 12.7 cm would remove about 30% of the small particulates.
Attention could then be focused on details of the dry-removal
processes .

     Alternatively,  a detailed examination of a storm using
actual observations of individual precipitation events — rainfall
rates, storm type, etc. — would improve the estimates.

     The converse would be true for locations with copious
precipitation.   Annual precipitation values near 508 cm yr""1
(e.g., in portions of southeast Alaska) could remove more than
                               130

-------
90% of the fine particulates.  Under such circumstances, dry
deposition would be unimportant.

     Not all the special circumstances where generalized methods
could be applied are covered here.  Each new compound, its source
distribution, physical/chemical character, and potential for
environmental effects must be considered in deciding what
methods to use.
6.   RESEARCH RECOMMENDATIONS

     The research recommended below is mandatory if the uncer-
tainty inherent in deposition calculations is to be alleviated:

     (1)  Washout ratios for aerosols must be classified by storm
          type, location, season, rainfall rate, size, and mass
          median diameter over many locations and seasons.  The
          size distribution and the mass median diameter of each
          aerosol must also be reported.
     (2)  Gases must be grouped by their chemical and physical
          properties to determine if such categorization relates
          to the deposition properties.
     (3)  How much gas is reemitted to the atmosphere must be
          determined.  Many gases are transported to the surface
          but may be reemitted to the atmosphere after deposition
          on the ground.
     (4)  Extensive measurements of the deposition velocity to
          various surfaces (especially forests) are necessary to
          define the dry deposition of aerosols more precisely.
          Snow and ice fields, water surfaces, forests, grass-
          lands, and other surfaces influence how much aerosol is
          captured.
     (5)  Use of the Jekyll number should be developed and re-
          fined.  The dry deposition of a chemical is a difficult
          parameter to determine.  A known gas such as S02 can
          be compared with an unknown gas to define the Jekyll
          number.  This practical method could be used to stand-
          ardize procedures for evaluating how important the dry
          deposition of a given gas is to total deposition.
7.   REFERENCES

Barrie LA, Walmsley JL. 1978.  A study of sulfur dioxide deposi-
tion velocities to snow in Northern Canada.  Atmos. Environ.
12:2321.

Cawse PA. 1974.  A survey of atmospheric trace elements in the
United Kingdom.  AERE Rep. R7669, Harwell.
                               131

-------
Garland JA. 1978.  Dry and wet removal of sulfur  from the atmos-
phere.  Atmos. Environ. 12:349.

Gatz DF. 1975.  Scavenging ratio measurements in  METROMEX.
ERDA Symposium on Precipitation Scavenging 1974-Coordinators
RW Beadle, RG Semoin, NTIS-74-1014, pp. 71-87.

Gatz DF. 1980.  Association and mesoscale spatial relationships
among rain water constituents.  J. Geophys. Res.  (in  press).

Hales JM. 1972.  Fundamentals of the theory of gas scavenging by
rain.  Atmos. Environ. 6:635.

	. 1975.  Atmospheric transformations of pollutants.   In:
Lectures on Air Pollution and Environmental Impact Analyses, DA
Haugen, ed., American Meteorological Society, Boston,  MA.

Haugen DA, ed. 1975.  Lectures on Air Pollution and Environmental
Impact Analyses.  American Meteorological Society, Boston, MA.

Heffter JL, Ferber GJ, Taylor AD. 1975.  A regional-continental
scale transport, diffusion, and deposition model.  NOAA Tech.
Mem., ERL-ARL-50.  Air Resources Laboratories, Silver Spring, MD.

Hogstrom U. 1979.  Initial dry deposition and type of source in
relation to long distance transport of air pollutants.  Atmos.
Environ. 13:295.

Hogstrom U, Enger L, Svedung I. 1979.  A study of atmospheric
mercury dispersion.  Atmos. Environ. 13:465.

Horst TW. 1977.  A surface depletion model for deposition  from
a Gaussian plume.  Atmos.  Environ. 11:41.

Houze RH Jr., Locatelli JD, Hobbs PV. 1976.   Dynamics  and  cloud
microphysics of rainbands in,an occluded frontal system.   J.
Atmos. Sci. 33:1921.

Junge CE. 1977.  Basic considerations about trace constituents in
the atmosphere as related to the fate of global pollutants.  In:
Fate of Pollutants in the Air and Water Environment (Advancement
in Environmental Sciences and Technology series), Suffet  IH, ed. ,
Wiley-Interscience, New York.

Kreitzberg CW,  Leach MJ.  1979.   Diagnosis and prediction of
tropospheric projectories  and cleansing.   In:  Proceedings of
Advisory Workshop to Identify Research Needs on the Formation of
Acid Precipitation, Electric Power Research Institute, Palo Alto,
CA.

McMahon TA,  Denison PJ.  1979.   Empirical atmospheric  deposition
parameters--a survey.  Atmos.  Environ.  13:571.
                               132

-------
Pasquill F. 1974.  Atmospheric Dispersion.  Halstead Press, New
York.

	.  1978.  Atmospheric dispersion parameters in plume model-
ling.  EPA Rep. EPA-600 4-78-021.

Payrissat M, Beilke S. 1975.  Laboratory measurements of the
uptake of sulfur dioxide by European soils.  Atmos. Environ.
9:211.

Slinn WGN. 1977.  Some approximations for the wet and dry removal
of particles and gases from the atmosphere.  Water Air Soil Poll.
7:513.

Smith FB, Hunt RD. 1978.  Meteorological aspects of the transport
of pollution over long distances.  Atmos. Environ. 12:461.

Spencer WF, Cliath MM. 1969-  Vapor density of dieldrin.  Environ.
Sci. Technol. 31:670.

Tang IN, Munkelwitz HR. 1977.  Aerosol growth studies:  III.
Ammonia bisulfate aerosols in a moist atmosphere.  J. Aerosol
Sci. 8:321.

Turner DB. 1970.  Workbook of Atmospheric Dispersion Estimates.
EPA, Research Triangle Park, NC.

Whitby KT. 1978.  Physical characteristics of sulfur aerosols.
Atmos. Environ. 12:135.
                                133

-------
                           BIBLIOGRAPHY


     This section contains additional references on metal and
organic compounds in atmospheric deposition that the reader may
find useful.

Abbott DC, Harrison RB,  Tatton JO'G,  Thompson G. 1965.  Organo-
chlorine pesticides in the atmosphere.   Nature 208:1317.

Abdoh Y, Aghdaie N, Darvich MR,  Khorgami MH. 1972.  Detection of
some PAH's and determination of benzo(a)pyrene in Teheran atmos-
phere.  Atmos. Environ.  6:949.

Ahling B, Jensen S. 1970.   Revised liquid-liquid partition in
determination of PCBs and chlorinated pesticides in water.  Anal.
Chem. 42:1483.

Altshuller AP, Lonneman WA,  Sutterfield FD, Kopezynski ST. 1971.
Hydrocarbon composition of the atmosphere of the Los Angeles
basin, 1967.   Environ. Sci. Technol.  5:1009.

Andreae MO. 1980.  Arsenic in rain and the atmosphere mass bal-
ance of arsenic.  J. Geophys. Res. (in press).

Andren AW. 1973.  In:  Ecology and Analysis of Trace Contaminants,
Fulkerson W,  Schultz WD, Van Hook RI, eds., NSF-RANN Rep. ORNL/
NSF/EATC-6, Oak Ridge National Laboratory,- TN.

Antommaria P, Coon M, Demaio L. 1965.  Airborne particulates in
Pittsburgh associated with p,p' DDT.   Science 150:1476.

Appel ER, Hoffer EM, Kothney EL, Wall SM, Haik M. 1979.  Analysis
of carbonaceous material in southern California atmospheric
aerosols.  Environ. Sci. Technol. 13:98.

Arthur RD, Cain JD, Barrentine BF. 1976.  Atmospheric levels of
pesticides in the Mississippi Delta.   Bull. Environ. Contain.
Toxicol. 15:129.

	• 1977.  DDT residues in air in the Mississippi Delta.
1975.  Pestic. Monit. J. 10:168.

Atkins PR. 1969.  Lead particulate matter in the environment.
Paper presented at Air Pollution Control Association 62nd Annual
Meeting, New York.

Baldwin MK, Bennett D, Begnon KI. 1977.  The concentrations of
aldrin and dieldrin and their photoisomers in the atmosphere.
Pestic. Sci.  8:431.
                               134

-------
Belot Y, Diop B, Marini T. 1971.  Composition minerale de la
matiere particulaire en suspension dans 1'air dans des zones
urbaines.  In:  La Radioecologie Appliquee a la Protection de
I'Homme et de son Environment,  International Symposium of the
European Community, Rome.

Bengtson S-A, Sodergren A. 1974.  DDT and PCB residues in air-
borne fallout and animals in  Iceland.  Ambio 3:84.

Bertsch W, Chang RC, Zlatkis  A. 1974.  The determination of
organic volatiles in air.  J. Chrom. Sci. 12:175.

Bevenue A, Ogata JW, Hylin JW.  1972.  Organochlorine pesticides
in rainwater, Oahu, Hawaii 1971-1972.  Bull. Environ. Contain.
Toxicol. 8:238.

Bidleman TF, Olney CE. 1975.  Long-range transport of toxaphene
insecticide in the atmosphere of the western North Atlantic.
Nature 257:475.

Bjorseth A, Lunde G, Lindskog A. 1979.  Long range transport of
PAHs.  Atmos. Environ. 13:45.

Brar SS, Nelson DM, Kline JR, Gustafson PF, Kanabrocki EL, Moore
CE, Hattori DM. 1970.  Instrumental analysis for trace elements
present in Chicago area surface air.  J. Geophys. Res. 75:2939.

Bruland KW, Bertine K, Liode  M, Goldberg ED. 1974.  History of
metal pollution in southern California coastal zone.  Environ.
Sci. Technol. 8:425.

Bunn WW, Deane ER, Klein DW,  Kloepfer RD. 1975.  Sampling and
characterization of air for organic compounds.  Water Air and
Soil Poll. 4:367.

Burghardt E, Jeltes R. 1975.  Gas chromatographic determination
of aromatic hydrocarbons in air using a semi-automatic preconcen-
tration method.  Atmos. Environ. 9:935.

Butler LD, Burke MF. 1976.  Chromatographic characterization of
porous polymers for use of adsorbants in sampling columns.  J.
Chrom. Sci. 14:117.

Cambray RS, Jeffries DR, Topping G. 1975.  An Estimate of the
Input of Atmospheric Trace Elements into the North Sea and the
Clyde Sea (1972-73).  AERE Harwell Rep. R-7733, Her Majesty's
Stationery Office, London.

Chow TJ, Bruland KW, Bertine  K, Soutar A, Koide M, Goldberg ED.
1973.  Lead and pollution:  records in southern California
coastal sediments.  Science 181:551.
                               135

-------
Cohen JM  Pinkerton C.  1966.  Widespread translocation of pesti-
cides by air transport and rainout.  In:  Organic Pesticides  in
the Environment,  Gould RL, ed.,  American Chemical Society,
Washington, DC.

Cohen Y  Cocchio W, MacKay D. 1978.  Laboratory study of liquid-
phase controlled volatilization rates in presence of wind waves.
Environ. Sci. Technol.  12:553.

Colucci JM, Begeman CR. 1971.  Carcinogenic air pollutants in
relation to automotive traffic in New York.  Environ. Sci.
Technol. 5:145.

Commins BT, Hampton L.  1976.  Changing patterns in concentrations
of polycyclic aromatic hydrocarbons in the air of central London.
Atmos. Environ. 10:561.

Comerford NB, White EH. 1976.  Nutrient input by precipitation in
north Minnesota forest ecosystem.  J. Minn. Academy Sci. 42:22.

Crozat G, Domergue JL,  Bogui V,  Fontan J. 1973.  Etude de 1'aero-
sol atmospherique en Cote d'lvoire et dans le Gofe de Guinee.
Atmos. Environ. 7:1103.

Daisey JM, Leyko MA. 1979.  Thin-layer GC method for the deter-
mination of PAH and oliphatic hydrocarbons in airborne partic-
ulate matter.  Anal. Chem. 51:24.

Dams R, Robbins JA, Rahn KA, Winchester JW. 1971.  Quantitative
relationships among trace elements over industrialized northwest
Indiana.  In:  Nuclear Techniques in Environmental Pollution,
International Atomic Energy Agency, Vienna.

Davis AR. 1980.  The Canadian Network for Sampling Organic
Compounds in Precipitation.  Inland Waters Directorate, Ottawa.
DeBortoli M, Pecchio E. 1976.
compounds in air over Europe.
Measurements of some halogenated
Atmos. Environ. 10:921.
Delfino JJ. 1979.  Toxic substances in the Great Lakes.
Sci. Technol. 13:1462.
                          Environ.
Dewiest F. 1978.  Any factors influencing the dispersion and
transport of heavy hydrocarbons associated with airborne parti-
cles.  Atmos. Environ. 12:1705.

Dong L, Locke DC,  Ferrand E.  1976.  HPLC method for routine
analysis of major parent PAH in particulates.  Anal. Chem.
48:368.

Duce RA, Zoller WH,  Moyers JL. 1973.  Particulate and gaseous
halogens in the Antarctic atmosphere.  J. Geophys. Res. 78:7802.
                               136

-------
Dudey ND, Ross LE, Noshkin VE. 1969.  Application of activation
analysis and GE(Li) detection techniques for the determination of
stable elements in marine aerosols.  In:  Modern Trends in Acti-
vation Analysis, DeVoe JR, Lefleur PD, eds. ,  Nat. Bureau of Std.,
Special Pub. 312, Vol. I, Washington, DC.

Egorov W, Zhigalovskaya TN, Malakhov SG. 1970.  Microelement
content of surface air above the continent and the ocean.  J.
Geophys. Res. 75:3650.

Eichmann R, Nevling P, Ketseridis G, Hahn J,  Jaenicke R, Junge C.
1979.  N-alkane studies in the troposphere:  I. Gas and particu-
late concentrations in the North Atlantic air.  Atmos. Environ.
13:587.

Eisenreich SJ, ed. 1980.  Atmospheric Input of Pollutants to
Natural Waters, Proceedings of American Chemical Society Sympo-
sium on Metal Deposition of Lakes, Sept. 1979, Ann Arbor Science
Pubs,  (in press).

Eisenreich SJ, Emmling PJ, Beeton AM. 1977.  Atmospheric loading
of phosphorus and other chemicals to Lake Michigan.  J. Great
Lakes Res. 3:291.

Eisenreich SJ, Hollod GJ, Johnson TC. 1979.  Accumulation of PCBs
in surficial Lake Superior sediments:  atmospheric deposition.
Environ. Sci. Technol. 13:569.

Environmental Protection Agency. 1976.  National Interim Primary
Drinking Water Regulations.  EPA-570/9-76-003.

Erickson MD, Michael LC, Zweidinger RA, Pellizzai ED. 1978.
Development of methods for sampling and analysis of polychlori-
naphthalenes in ambient air.  Environ. Sci. Technol. 12:927.

Evered, DF. 1974.  Decline of PCB concentrations in North At-
lantic surface waters.  Nature 252:387.

Faoro RB. 1975.  Trends in concentrations of benzene soluble
suspended particulate fraction and benzo(a)pyrene.  J. Air
Pollut. Control Assoc. 25:638.

Farwell SO, Robinson E, Powell WJ, Adams DF.  1976.  Survey of
airborne 2,4-D in south-central Washington.  J. Air Pollut.
Control Assoc. 26:224.

Fine DH, Rounbehler DB, Sawicki E, Krost K. 1977.  Determination
of dimethylnitrosamine in air, water, and soil by thermal activa-
tion analysis.  Environ. Sci. Technol. 11:581.
                               137

-------
Foster JF, Beatty GH,  Howes JE,  Jr. 1974.  Interlaboratory
Cooperative Study of the Precision and Accuracy of the Meas-
urement of Lead in the Atmosphere using the Colorimetric Dithi-
zone Procedure.  American Society for Testing and Materials Data
Series Publication DS55-S5, Battelle Memorial Institute, Rich-
land, WA.

Fox MA, Staley SW. 1976.  Determination of PAH in atmospheric
particulate matter by HPLC coupled with fluorescence techniques.
Anal. Chem. 48:992.

Frankel LS, Black RF.  1976.  Automatic GC monitor for determina-
tion of ppb levels of bis(chloromethyl)ether.  Anal. Chem.
48:732.

Fraser PJB, Pearman GI.  1978.  Atmospheric halocarbons in the
Southern Hemisphere.  Atmos.  Environ. 12:839.

Giam CS, Chan HS, Neff GS. 1975.  Rapid and inexpensive method to
determine PCB and phthalates in air.  Anal. Chem. 47:2319.

	. 1976.  Concentration and fluxes of phthalates, DDT
and PCBs to the Gulf of Mexico.   In:  Marine Pollutant Transfer,
Windom H, Duce RA, eds., Lexington Books, Lexington, MA.

Giam CS, Chan HS, Neff GS, Atlas EL. 1978.  Phthalate esters
plasticizers:  a new class of marine pollutant.   Science 199:419.

Gordon RJ. 1976.  Distribution of PAH in Los Angeles.  Environ.
Sci. Technol. 10:370.

Gordon RJ, Bryan RJ. 1973.  Patterns in airborne PAH concentra-
tions at four Los Angeles sites.  Environ. Sci.  Technol. 7:1050.

Gray D, Kay DM, Eichor J, Bogt JR. 1972.  Determination of the
trace element levels in atmospheric pollutants by instrumental
neutron activation analysis.   IEEE Trans. Nucl.  Sci. NS19.

Grimsrud EP, Rasmussen RA. 1975.  The analysis of chlorofluoro-
carbons in the troposphere by gas chromatography-mass spectrom-
etry.  Atmos. Environ. 9:1010.

Grosjean D, vanCauwenberghe K, Schmidt JP, Kelley JM. 1978.
Identification of C3-C1C, aliphatic dicarboxylic acids in airborne
particulate matter.  Environ. Sci. Technol. 12:313.

Hague R, Schmedding DW.  1975.  A method of measuring the water
solubility of hydrophobic chemicals:  Solubility of five PCBs.
Bull. Environ. Contam. Toxicol.  14:13.

Hammerle RH, Pierson WR. 1975.  Sources and elemental composition
of aerosol in Pasadena,  California by energy-dispersive X-ray
fluorescence.  Environ.  Sci.  Technol. 9:1058.
                               138

-------
Hamilton WL, Miller JE. 1971.  High lead concentrations in
Columbus snow.  Ohio J. Sci. 7:313.

Harding LW, Philips JH, 1978.  PCBs:  Transfer from microparti-
culates to marine phytoplankton and effects on photosynthesis.
Science 202:1189.

Harrison PR, Rahn KA, Dams R, Robbins JA, Robbins JW, Winchester
JW, Brar SS, Nelson DM. 1971.  Areawide trace metal concentra-
tions measured by multielement neutron activation analysis.  J.
Air Pollut. Control Assoc. 21:563.

Harvey GR, Steinhauer WG, Teal JM. 1973.  Polychlorobiphenyls in
North Atlantic ocean water.  Science 180:643.

Hasan 0, Spyrou NM. 1972.  Air-pollution study of Guildford,
Surrey, using neutron activation analysis.  In:  Nuclear Activa-
tion Techniques in the Life Sciences, International Atomic Energy
Agency Symposium, Vienna.

Heesen TC, Young DR, McDermott ED. 1979.  Evaluation of a tech-
nique for measuring dry aerial deposition rates of DDT and PCB
residues.  Atmos. Environ. 13:1677.

Heindryckx R, Dams R. 1974.  Continental, marine, and anthro-
pogenic contributions to the inorganic composition of the aerosol
of an industrial zone.  J. Radioanal. Chem. 19:339.

Henry WM, Blosser ER. 1971.  Identification and estimation of
ions, molecules, and compounds in particulate matter collected
from ambient air.  Technical Report CPA-70-159, Battelle Labora-
tories, Columbus, WA.

Hester NE, Meyer RA. 1979.  A sensitive technique for measurement
of benzene and alkyl benzene in air.  Atmos. Environ. 13:107.

Hidym GM, Mueller PK, Wang HH, Karney J, Twiss S, Imada M,
Alcocer A. 1974.  Observations of aerosols over southern Califor-
nia coastal waters.  J. Appl. Meteorol. 13:96.

Hoffman GL, Duce RA, Hoffman EJ. 1972.  Trace metals in the
Hawaiian marine atmosphere.  J. Geophys. Res. 77:5322.

Hoffman GL, Duce RA, Zoller WH. 1969.  Vanadium, copper and
aluminum in the lower atmosphere between California and Hawaii.
Environ. Sci. Technol. 3:1207.

Huntzinger 0, Safe S, Zitko U. 1974.  The Chemistry of PCBs, CRC
Press, New York.

Ichikuni M. 1978.  Calcite as a source of excess calcium in
rainwater.  J. Geophys. Res. 83:6249.
                               139

-------
Jackson JW,  Chang DPY.  1977.   PCB recovery efficiency from stack
gas and ambient air.   J.  Air.  Pollut. Control Assoc. 27:1006.

Janssens MJ, Dams R.  1972.   Analyse en interpretatie van de
anorganische samenstelling der aerosolen in mechelen.  Rep. of
the Institute of Nuclear Sciences,  University of Ghent, Belgium.

John W, Kaifer R, Rahn K, Wesolowski JJ. 1973.  Trace element
concentration in aerosols from the San Francisco Bay area.
Atmos. Environ. 7:107.

Johnson ER,  Yu TC, Montgomery ML. 1977.  Trapping and analysis of
atmospheric 2,4-D residues.  Bull.  Environ. Contain. Toxicol.
17:369.

Jones PW, Glammer RD, Strup PE, Stanford TB. 1976.  Efficient
collection of polycyclic organics from effluents.  Environ. Sci.
Technol. 10:806.

Junge CE. 1974.  Residence time and variability of tropospheric
trace gases.  Tellus 26:477.

Katz J, Sakuma T, Ho A. 1978.   Chromatographic and spectral
analysis of PAH.  Environ.  Sci. Technol. 12:909.

Keane JR, Fisher EMR. 1968.  Analysis of trace elements in
airborne particulates by neutron activation and gamma-ray spec-
trometry.  Atmos. Environ.  2:601.

Keith LH, Telliard WA.  1979.   Priority pollutants.  Environ. Sci.
Technol. 13:416.

Kertesz-Saringer M, Morlin Z.  1975.  On the occurrence of poly-
cyclic aromatic hydrocarbons in the urban area of Budapest.
Atmos. Environ. 9:331.

Ketseridis G, Hahn J, Jaenicke R, Junge C. 1976.  Organic consti-
tuents of atmospheric particulate matter.  Atmos. Environ.
10:603.

Kneip TJ, Eisenbud M, Kleinman MT,  Bernstein DM, Lippman M,
Riddick R. 1974.  Trace Metals in Urban Aerosols.  Second Annual
Progress Report to the American Petroleum Institute and the
Edison Electric Institute.

Lee RE, Goranson SS,  Enrione RE,  Morgan GB. 1972.  National air
surveillance cascade impactor network:  II. Size distribution
measurements of trace metal components.  Environ. Sci. Technol.
6:1025.

Levin JO, Nilsson CA, Anderson K. 1977.  Sampling of organic
substances in workroom air with amberlite X+D-2 resin.  Chemo-
sphere 9:595.
                              140

-------
Likens GE, Borman FH, Johnson NM, Pierce RS. 1967.  The Ca, Mg,
K and Na budgets for a small forest.  Ecosystem Ecol. 5:772.

Louw CW, Richards JF, Faure PK. 1977.  The determination of vola-
tile organic compounds in city air by gas chromotography combined
with standard addition, selective subtraction, IR and MS.  Atmos.
Environ. 11:703.

Lovelock JE, Maggs RJ, Wade RJ. 1973.  Halogenated hydrocarbons
in and over the Atlantic.  Nature 241:194.

Lunde G, Bjorseth A. 1977.  PAH in long-range transported aero-
sols.  Nature 268:518.

Lunde G, Gether J, Gjos N, Lande M-BS. 1977.  Organic micro-
pollutants in precipitation in Norway.  Atmos. Environ. 11:1007.

MacKay D. 1978.  Volatilization of non-polar compounds from water
surfaces, presented at US-EPA Workshop on Transport and Fate of
PCBs in the Great Lakes, Grosse lie, Michigan, April 3-4.

MacKay D, Leinonen PJ. 1975.  Rate of evaporation of low-
solubility contaminants from water bodies to atmosphere.  Environ.
Sci. Technol. 9:1178.

MacKay D, Yurn ATK. 1980.  Transfer rates of gaseous pollutants
between the atmosphere and natural waters.  In:  Atmospheric
Input of Pollutants to Natural Waters, Eisenreich SJ, ed., Ann
Arbor Science Pubs, (in press).

Mamuro T, Matsuda Y, Mizohata A, Matsunami T. 1971.  Activation
Analysis of Airborne Dust.  Annual Rep. of the Radiation Center
of Osaka Prefecture 12:1.

Mann JE, Enos HF, Gonzales J, Thompson JF. 1974.  Development of
sampling and analytical procedure for determining HCB, HCl, and
3-butadiene in air.  Environ. Sci. Technol. 8:584.

Martin A, Barger FR. 1978.  Some observations of acidity and
sulphur in rainwater from rural sites in central England and
Wales.  Atmos. Environ. 12:1481.

McCall CS, Harriss RC. 1973.  Precipitation and throughfall
chemistry in the San Francisco Bay area.  J. Environ. Qual.
7:352.

McClure VE, Lagrange J. 1977.  Deposition of heavy chlorinated
hydrocarbons from the atmosphere.  Bull. Environ. Contain. Toxicol.
17:219.

Miguel AH, Friedlander SK. 1978.  Distribution B(a)P and coronene
with respect to particle size in Pasadena aerosols in the
submicron-size range.  Atmos. Environ. 12:2407.
                               141

-------
Miles JW, Fetzer LE,  Pearce GW.  1970.  Collection and determina-
tion of trace quantities of pesticides in air.  Environ. Sci.
Technol. 4:420.

Moller J, Alfheim I.  1980.   Mutagenicity and PAH analysis of
airborne participate  matter.  Atmos.  Environ. 14:83.

Morrow NL, Brief RS.  1971.   Elemental composition of suspended
particulate matter in metropolitan New York.  Environ. Sci.
Technol. 5:786.

Moyers JL, Zoller WH,  Duce  RA,  Hoffman GL. 1972.  Gaseous bromine
and particulate  lead,  vanadium,  and bromine in polluted atmos-
phere.  Environ. Sci.  Technol.  6:68.

Moyers JL et al. 1973.  Trace metal concentrations in the Tucson
area atmosphere.  Final Report to the Arizona Mining Association,
Atmospheric Analysis  Lab.,  Dept.  of Chemistry, University of
Arizona, Tucson.

Murphy TJ, Schinsky A, Paolucci G, Rzeszutko CP. 1980.  Inputs of
PCB from the atmosphere to  Lakes  Huron and Michigan.  In:
Atmospheric Input of  Pollutants to Natural Waters, Eisenreich SJ,
ed., Ann Arbor Science Pubs, (in press).

Murray KE. 1977.  Concentration of volatiles on chromosorb 105.
J. Chrom. Sci. 135:49.

National Research Council,  1979.   Polychlorinated biphenyls.
National Academy of Science, Washington, DC.

Nifong GD, Winchester JW. 1970.   Particle size distributions of
trace elements in pollution aerosols.  Tech. Rep. ORA 08903-8-T,
University of Michigan, Ann Arbor.

Nriagu JO, ed. 1979.   Biogeochemistry of Mercury in the Environ-
ment.  Elsevier-North Holland,  New York.

O'Connors HB, Wurster CF, Powers  CD,  Biggs DC, Rowland RG. 1978.
PCBs may alter marine trophic pathways by reducing phytoplankton
size and production.   Science 201:737.

Ohta T, Morita M, Mizoguchi I.  1976.   Local distribution of
chlorinated hydrocarbons in the ambient air in Tokyo.  Atmos.
Environ. 10:557.

Ohta T, Morita J, Mizoguchi I,  Tada T. 1977.  Washout effect and
diurnal variation for chlorinated hydrocarbons in ambient air.
Atmos. Environ.  11:985.

Olney EC, Bidleman TF. 1973.  Atmospheric pollutant transport and
deposition on the sea surface.   Rep.  to Department of Energy on
Grant No. GX33777, University of Rhode Island, Kingston.
                               142

-------
Orgill MM, Sehmel GA, and Peterson MR. 1976.  Some initial meas-
urements of airborne DDT over Pacific Northwest forests.  Atmos.
Environ. 10:827.

Peakall DB. 1976.  DDT in rainwater in New York following applica-
tion in the Pacific Northwest.  Atmos. Environ. 10:899.

Pellizzari ED, Bunch JE, Berkley RE, MacRae J. 1976.  Determina-
tion of trace hazardous organic vapor pollutants in ambient
atmosphere by GC/MS.  Anal. Chem. 48:803.

Penkett SA, Brice KA, Derwent RG, Eggleton AEJ. 1979.  Measure-
ment of CC13F and CC14 at Harwell over the period Jan. 1975-Nov.
1977.  Atmos. Environ. 13:1011.

Persiani C. 1971.  The detection of selected air pollutants using
a laboratory neutron generator.  Amer. Ind. Hyg. J. 32:573.

Petrenchuk OP, Selezneva ES. 1978.  Chemical composition of
precipitation in regions of the Soviet Union.  J. Geophs.  Res.
75:3629.

Pierce RC, Katz M. 1975.  Dependency of PAH content on size dis-
tribution of atmospheric aerosols.  Environ. Sci. Technol.
9:347.

Pilar S, Graydon WF. 1973.  Benzene and toluene distribution in
the Toronto atmosphere.  Environ. Sci. Technol. 7:628.

Pillay KKS, Thomas CC. 1971.  Determination of trace element
levels in atmospheric pollutants by neutron activation analysis.
J. Radioanal. Chem. 7:107.

Poltz K. 1970.  Inorganic chemical analysis of non-polluted
aerosols sampled at 1800 meters altitude.  J. Geophys. Res.
75:2347.

Rahn KA. 1972a.  Sources of trace elements in aerosols—an
approach to clean air.  Ph. D. Thesis, University of Michigan,
Ann Arbor (University Microfilms 72-4956).

Rahn KA. 1972b.  Study of the atmospheric aerosol on the
Jungfraujoch.  Progress Rep., Institut voor Nucleaire
Wetenschappen, Rijksuniversiteit, Ghent, Belgium.

Rao AMM, Vohra KG. 1975.  The concentrations of benzo(a)pyrene in
Bombay.  Atmos. Environ. 9:403.

Rappaport SM, Morales R. 1979.  Air sampling and analytical
method for 4,4' methylenebis (2-chloroaniline).  Anal. Chem.
51:19.
                               143

-------
Reiter R,  Saldkovic R,  and Potzl K.  1976.  Chemical components of
aerosol particles in the lower troposphere above central Europe
measured under pure air conditions.   Atmos. Environ. 10:841.

Robinson E,  Rasmussen RA,  Krasnec J,  Pierotti D, Jakubovic M.
1977.  Halocarbon measurements in the Alaskan troposphere and
lower stratosphere.  Atmos. Environ.  11:215.

Roberts RM.  1975.  A review of some  biological effects of lead
emissions from primary and secondary smelters, presented to
National Research Council  of Canada,  Ottawa, from Inst. for
Environmental Studies,  University of Toronto, Canada.

Sandalls FJ, Hatton DB. 1977.   Measurements of atmospheric
concentrations of trichlorofluoromethane, dichlorodifluoromethane
and carbon tetrachloride by aircraft sampling over the British
Isles.  Atmos. Environ. 11:321.

Sanderson M, LaValle PD. 1979.  Surface loading from pollutants
in precipitation in southern .Ontario:  Some climatic and statis-
tical aspects.  J. Great Lakes Res.  5:52.

Schrimpff E, Thomas W,  Hermann R. 1979.  Regional patterns of
contaminants in the snow of N. E. Bavaria.  Water Air and Soil
Poll. 11:481.

Seba DV, Prospero JM. 1971.  Pesticides in the lower atmosphere
of the northern equatorial Atlantic  Ocean.  Atmos. Environ.
5:1043.

Shriner DS.  1975.  Effects of simulated rain acidified with
sulfuric acid on host-parasite interactions.  In:  First Inter-
national Symposium on Acid Precipitation and the Forest Ecosystem,
Columbus,  OH (NTIS PB-258695).

Simmonds PG, Kerrin SL, Lovelock JE,  Shair FH. 1974.  Distribu-
tion of atmospheric halocarbons in the air over the Los Angeles
basin.  Atmos. Environ. 8:209.

Sodergren A. 1972.  Chlorinated hydrocarbon residues in airborne
fallout.  Nature 236:395.

Spencer WF,  Cliath MM.  1975.  Vaporization of chemicals, in
Environmental Dynamics of  Pesticides, Hague R, Freed VH, eds.,
Plenum, New York.

Stanley CW,  Barney JE,  Helton MR, Yobs AR. 1971.  Measurement of
atmospheric levels of pesticides. Environ. Sci. Technol. 5:430.

Stocks P,  Commins BT, Aubrey KV. 1961.  A study of polycyclic
hydrocarbons and trace elements in smoke in Merseyside and other
northern localities.  Water Air Soil Poll. 4:141.
                               144

-------
Strachan WMJ, Huneault H, Schertzer WM, Elder FC. 1980.  Organo-
chlorines in precipitation in the Great Lakes.  In:  Proceedings
of International Symposium on Hydrocarbons in the Environment,
Toronto, 1979.

Strand JW, Andren AW. 1978a.  Organic carbon content of aerosols
over Lake Michigan:  Concentration, particle size distribution,
and deposition.  Rep. of the Water Chemistry Program, University
of Wisconsin, Madison.

	. 1978b.  Polyaromatic hydrocarbons in aerosols over
Lake Michigan:  fluxes to the lake.  Presented in the Environment
Meeting, Battelle, Columbus, WA.

Sugimae A. 1975.  Sensitive emission spectrometric method for the
analysis of airborne particulate matter.  Anal. Chem. 47:1840.

Sugimae A, Hasegawa T. 1973.  Emission spectrographic determina-
tion of trace metals in airborne particulates collected on a
membrane filter.  Japan Analyst 22:3.

Suffet IH, ed. 1977.  Fate of Pollutants in the Air and Water
Environment.  Wiley-Interscience, New York.

Swain WR. 1978.  Chlorinated organic residues in fish,  water and
precipitation from the vicinity of Isle Royale, Lake Superior.
J. Great Lakes Res. 4:398.

Swank WT, Henderson GS. 1976.  Atmospheric input of some cations
and anions:  Forest ecosystems in North Carolina and Tennessee.
Water Resources Res. 12:541.

Tarrant KR, Tatton JO'G. 1966.  Organochlorine pesticides in rain
water in the British Isles.  Nature 219:725.

Tanner TM, Young JA, Cooper JA. 1974.  Multi-element analysis of
St. Louis aerosols by non-destructive techniques.  Chemosphere
5:211.

Thomas JF, Mukar M, Tebbens BD. 1968.  Fate of airborne benzo(a)-
pyrene.  Environ. Sci. Technol. 2:33.

Thrane KE. 1978.  Background levels in air of lead, cadmium,
mercury and some chlorinated hydrocarbons measured in south
Norway.  Atmos. Environ. 12:1155.

United Nations Scientific Committee on the Effects of Atomic
Radiation. 1977.  Sources and Effects of Ionizing Radiation.
UNSCEAR Pub. 40, 32nd Session (A/32/40).

Versino B, deGroot M, Geiss F. 1964.  Air pollution sampling by
adsorption columns.  Chromotographia 1:302.
                               145

-------
Volchok HL. 1980.   Atmospheric deposition of man-made radioactiv-
ity.  In:  Twelfth International Conference on Environmental
Toxicity-  Plenum, NY.

Warburton JA. 1973.  The distribution of silver in precipitation
from 2 seeded Alberta hailstorms.  J. Appl.  Meteorol. 12:677.

Weibel SR, Weidner RB, Cohen JM, Christiansen AG. 1966.  Pesti-
cides and other contaminants in rainfall and runoff.  J. Amer.
Water Works Assoc. 58:1075.

Welles DE, Johnstone SJ. 1978.  The occurrence of organochlorine
residues in rain.   Water Air Soil Poll. 9:271.

Wheatley GA, Hardman JA. 1965.  Indications of the presence of
organochlorine pesticides in rainwater in central England.
Nature 207:486.

Wilkniss PE, Swinnerton JW, Lamontagne RA, Bressan DJ. 1975.
Trichlorofluoromethane in the troposphere:  Distribution and
increase, 1971-1974.  Science 187:832.

Williams FW, Umsted ME. 1968.  Determination of trace contami-
nants in air by concentration on porous polymer beads.  Anal.
Chem. 40:2232.

Windom HL, Duce RA, eds. 1976. Marine Pollutant Transfer.
Lexington Books, Lexington, MA.

Wright RF. 1976.  The impact of forest fire on the nutrient
influxes to small  lakes in northeast Minnesota.  Ecology 57:649.

Yasuda SK. 1975.  Determination of MOCA in air.  J. Chrom. 104:
283.

Young DR, McDermott DJ, Heesen TC. 1976.  Aerial fallout of DDT
in southern California.  Bull. Environ Contain. Toxicol. 16:604.

Zafonte L, Hester  NE, Stephens ER, Taylor OC. 1975.  Background
and vertical atmospheric measurements of fluorocarbon-11 and 12
over southern California.  Atmos. Environ. 9:1007.

Zoller WH, Gordon  GE. 1970.  Instrumental neutron activation
analysis of atmospheric pollutants utilizing GE(Li) ray detec-
tors.  Anal. Chem. 42:257.
                               146

-------
             Report of a Workshop on
SCREENING CHEMICALS FOR INADVERTENT MODIFICATION
               OF THE STRATOSPHERE
                Boulder, Colorado
                  November 1979
                    Edited by
                 Daniel Albritton
                      147

-------
                   PARTICIPANTS AND CONTRIBUTORS
Daniel L. Albritton
U.S. Dept. of Commerce
NOAA/ERL, R443
325 Broadway
Boulder, CO  80303
303-497-5785
FTS-320-5785

Julius Chang
Lawrence Livermore Lab.
P.O. Box 808
Livermore, CA  94550
415-422-4081
FTS-532-4081
Dieter Kley
U.S. Dept. of Commerce
NOAA/ERL, R448
325 Broadway
Boulder, CO  80303
303-497-3355
FTS-320-3355
Jim Anderson
Harvard University
29 Oxford  St.
Cambridge, MA   02138
617-495-5922
Carleton J. Howard
U.S. Dept. of Commerce
NOAA/ERL, R443
325 Broadway
Boulder, CO  80303
303-497-5820
FTS-320-5820

Shaw Liu
U.S. Dept. of Commerce
NOAA/ERL, R446
325 Broadway
Boulder, CO  80303
303-497-3356
FTS-320-3356
Mack McFarland
U.S. Dept. of Commerce
NOAA/ERL, R448
325 Broadway
Boulder, CO  80303
303-497-3373
FTS-320-3373

Arthur L. Schmeltekopf
U.S. Dept. of Commerce
NOAA/ERL, R448
325 Broadway
Boulder, CO  80303
303-497-5340
FTS-320-5340
John M. Miller
NOAA/Air Resources Labs.
8060 13th Street
Silver Spring, MD  20910
301-427-7645
FTS-427-7645
William P. Wood
U.S. EPA
Office of Toxic Sub-
  stances (TS-792)
Washington,  DC
202-755-4860
FTS-426-0724
                               148

-------
                             CONTENTS

                                                             Page

PARTICIPANTS AND CONTRIBUTORS                                 148

PREFACE                                                       150

1.   INTRODUCTION                                             151

2.   METHODOLOGY                                              152

     2.1. Preliminary Analysis                                152
     2.2. Tropospheric Removal Processes                      152

          2.2.1.  Photoabsorption                             152
          2.2.2.  Destruction by Reactive Tropospheric
                  Species                                     154
          2.2.3.  Scavenging by Atmospheric Aerosols          155
          2.2.4.  Removal by Precipitation                    155
          2.2.5.  Dissolution Into the Oceans                 155
          2.2.6.  Surface Removal Processes                   156

     2.3. Stratospheric Interaction                           157

3.   DATA REQUIREMENTS                                        158

     3.1. Gas Phase UV and Visible Absorption Spectra         158
     3.2. Destruction by Reactive Tropospheric Species        158
     3.3. Scavenging by Atmospheric Aerosols                  159
     3.4. Removal by Precipitation                            159
     3.5. Dissolution Into the Oceans                         159
     3.6. Surface Removal Processes                           160

4.   DATA SOURCES                                             160

5.   REFERENCES                                               160

APPENDIX:  Sources of Physical and Chemical Data              163
                                149

-------
                              PREFACE

     In recent years an awareness has developed of the potential
for human activities to affect the stratosphere.  Activities over
which concern has been raised because of their potential impacts
on stratospheric ozone, one of the major constituents that absorb
near-UV solar radiation,  have included (a) the flights of high-
altitude aircraft, such as the supersonic transport, which re-
lease nitrogen oxides into the stratosphere, (b) the operation of
the Space Shuttle with attendant stratospheric emission of HCl
gas, and (c) the use of chlorofluorocarbons (CFCs), which reach
the stratosphere in significant quantities because of their
stability at the earth's surface and in the troposphere.  It has
been postulated, on the basis of a great deal of laboratory and
field data, that the stratospheric ozone layer may be modified by
a variety of catalytic reaction cycles involving oxides of nitro-
gen, oxides of hydrogen,  chlorine species, bromine species, and
possibly other yet-unidentified reactants.  The resulting modi-
fication of the stratospheric ozone layer by these catalytic
cycles could have significant biological and agricultural impacts
because of the extreme sensitivity of living organisms to the
altered UV irradiation that would accompany a change in the
stratospheric ozone shield.

     Perception of the long-range consequences for humans and the
biosphere has prompted a considerable expenditure of effort and
resources by industry,  academia, and several Federal agencies in
quantifying the risks associated with activities of potential
stratospheric significance.

     The methodology described herein provides a means by which
chemical substances may be evaluated to determine their potential
to be transported into and affect the stratosphere (particularly
the ozone in the stratosphere).  The essence of the methodology
is a step-by-step assessment of the capability of a number of
processes to remove a substance from the atmosphere.  The assess-
ment begins with the tropospheric removal processes thought to be
generally most significant, followed by those of somewhat lesser
significance.  In each case,  any products formed in the tropo-
sphere are also assessed for their potential to reach the strato-
sphere.
                               150

-------
         SCREENING CHEMICALS FOR INADVERTENT MODIFICATION
                        OF THE STRATOSPHERE
1.  INTRODUCTION

     The Environmental Protection Agency's initial involvement
with potential ozone modification began in the fall of 1976 when
the EPA announced the initiation of regulatory activities to
control emissions of chlorofluorocarbons (CFCs).  In March 1978
the EPA published regulations prohibiting the use of CFCs as
propellants in aerosol products.  The EPA derived its authority
for such action from the Toxic Substances Control Act (TSCA),
Public Law 94-469.  As the Federal Agency charged with the admin-
istration of TSCA, the EPA is responsible for determining whether
a new or existing substance presents or may present an unreason-
able risk of injury to health or to the environment.

     In addition, passage of the Clean Air Act Amendments of
1977, Public Law 95-95, assigned EPA specific aspects of the
responsibility for protecting the stratosphere, especially the
ozone in the stratosphere.  Part B of the amended Clean Air Act
provides that substances, practices, processes, and activities
that may affect the stratosphere, especially ozone in the strat-
ophere, should be investigated by the Administrator of the EPA to
give early warning of any potential problem and to develop the
basis for possible future regulatory action.  The Administrator
is required to conduct studies and research concerning the ef-
fects of such substances and activities, to report biennially to
Congress the results of these studies and research,  to report
annually to Congress recommendation for control, and to propose
regulations for the control of any substance or activity that may
reasonably be anticipated to affect the stratosphere.

     Because of the EPA's need to determine, often within a
relatively short time and with limited data, the potential strat-
ospheric importance of both new and existing chemicals, a method-
ology for the systematic examination of chemical substances is
required.  The methodology (1) must evaluate the potential of a
substance to enter the atmosphere; (2) must account systemati-
cally for each of the potential physical and chemical mechanisms
by which a substance may be removed from the atmosphere at the
earth's surface and in the troposphere, and must account for any
reaction products formed by these physical and chemical removal
processes; and, finally (3) must assess the potential for strato-
spheric modification.
                                151

-------
2.   METHODOLOGY

     The methodology described herein provides a means by which
chemical substances may be evaluated to determine their potential
to be transported into and affect the stratosphere (particularly
the ozone in the stratosphere).  The essence of the methodology
is a,step-by-step assessment of the capability of a number of
processes to remove a substance from the atmosphere (Figure 1).
The assessment begins with the tropospheric removal processes
thought to be generally most significant, followed by those of
somewhat lesser significance.  In each case, any products formed
in the troposphere are also assessed for their potential to reach
the stratosphere.


2.1.  Preliminary Analysis

     A prelude to the application of this methodology is the
determination that an expected significant transport of a sub-
stance into the atmosphere is based upon a consideration of
(1) materials balance data (i.e., production, use, and disposal
information), (2) environmental characterization data (i.e.,
meteorological, hydrological, and topographical information), and
(3) materials data (i.e., physical state and chemical properties).


2.2.  Tropospheric Removal Processes

     The following tropospheric removal processes should then be
considered:

2.2.1.  Photoabsorption

     It is first necessary to establish whether a substance   0
absorbs radiation in the spectral range between 2950 and 7000 A.
This constitutes the spectral region of concern because light at
shorter wavelengths is absorbed above the tropopause, and radia-
tion of longer wavelengths is not of sufficient energy to disso-
ciate chemical bonds of interest here.  If photodissociation is
predicted, that molecule may be eliminated from consideration,
provided that the quantitative criteria discussed below are met.
However, identical criteria must be applied to the potential
products of that dissociation.  Furthermore, if photodissociation
is predicted to lead to enhanced chemical reactivity of the
dissociation products, then further reactive cycles must be
considered as possible loss processes.

     The primary quantity of concern for a quantitative assess-
ment of photoabsorption is the product of the solar flux and the
absorption cross section at each wavelength and at each altitude
such that the photodissociation rate can be determined:
                               152

-------
            Partitioning
            Analysis
Absorption
Spectrum  0
(2950-7000 A)
Reaction with
Hydroxyl Radical
and Ozone
Vapor Pressure;
Solubility; Aqueous
Phase Reactivity;
Adsorption
Solubility; Vapor
Pressure;  Agueous
Phase Reactivity;
Adsorption
Solubility; Vapor
Pressure;  Aqueous
Phase Reactivity;
Adsorption
Absorption
Spectrum  0
(1750-2150 A)
Ul
u>
               Chemical Is Not
               of Stratospheric Importance
                                     Figure  1.    Schematic  of methodology for  identifying
                                     chemicals  of  stratospheric  importance.

-------
                      7000A
                             F(\)a(\) exp (-T(\,z)
                 2950A
                                          _  _  o _
where F(X) is the solar flux in photons cm~2s~1 A 1, a (A.) is the
absorption cross section in cm2 at wavelength A., and T(\,z) is the
atmospheric opacity above the altitude z.  The critical quantity
is the constituent mixing ratio, which, when combined with the
relevant production term P(z) provides the information needed to
calculate the tropospheric mixing ratio:


                      tropopause  /      tropopause
                     / P(z)dz   /  [M]   / J(z)dz,
where [X] is the released substance concentration and [M] is
the total concentration.

2.2.2.  Destruction by Reactive Tropospheric Species

     This destruction mechanism is limited to homogeneous gas
phase reactions that occur in the tropopause.  Although tropo-
spheric chemistry is not well understood, many reactions of
several important reactive species, such as O(3P), 0(1D), 03 , NO,
N02,  OH, and HO2,  have been studied.  There are other species
that undoubtedly play a very important role, but few of their
reaction rate constants are measured.  These include N03, alkoxy,
and alkylperoxy radicals.  At the present time only two of these
species, the hydroxyl radical OH and ozone 03 are known to be
significant scavengers of chemical substances in the atmosphere.
Chemical substances containing hydrogen are generally reactive
toward OH; others containing double bonds, specifically olefins,
are reactive toward both OH and 03.

     The atmospheric lifetime t of a chemical substance S toward
destruction by some species A is defined as the reciprocal of the
first-order rate constant for the reaction of S with A.  If that
reaction is bimolecular, then t = l/k[A], where k is the bimo-
lecular rate constant and [A] is the atmospheric concentration of

A.  If the reaction is termolecular, then k = k111 [M], where

k    is the termolecular rate constant and [M] is the effective
concentration of air.  Few reactions are purely termolecular in
the troposphere.

     If a chemical substance is found to be degraded by reaction
with some species, then the possibility of the formation of
persistent products must also be evaluated.


                                154

-------
2.2.3.  Scavenging by Atmospheric Aerosols

     Another possible tropospheric removal process is scavenging
by aerosols, which are subsequently removed from the atmosphere
by dry deposition or wet removal.

     The aerosol residence times in the troposphere are estimated
to be from 5 to 15 days (Junge 1977, Moore et al. 1973, Slinn et
al. 1978).  The gas kinetic collision frequency of molecules with
aerosols is in the range of 10"*2 to 1 per second.  If the stick-
ing coefficient (the fraction of the molecules that stays on the
aerosol following each collision) of a substance is, for example,
10~5, then its tropospheric lifetime is 10s to 107 s.

     The physical property of a substance that is of primary sig-
nificance in determining its sticking coefficient with aerosols
is its vapor pressure.  Junge (1977) has developed a model to
estimate the partial atmospheric removal rate of substances by
this process as a function of vapor pressure.  According to
Junge's model, substances with vapor pressures greater than
approximately 10~6 torr would have partial residence times of the
order of days and would be removed effectively by this process.
In addition, substances with a high solubility, or a high reac-
tivity in the dissolved phase, may also be scavenged effectively
by aerosols.

2.2.4.  Removal by Precipitation

     Three properties of a substance are important in evaluating
its removal by precipitation—solubility, dissociation in water,
and oxidation in water.  In mass transfer of a substance, a gas
to a cloud or rain droplet, the highly soluble gas is quickly
absorbed.  If further dissociation takes place in removal of a
substance, it reduces the chances of release back into the
atmosphere.  Aerosol removal also depends on the solubility of a
given aerosol and its ability to act as a condensation nucleus or
dissolve into a cloud or rain droplet.  A detailed description of
these processes is given in the workshop report on Toxic Sub-
stances in Atmospheric Deposition.

2.2.5.  Dissolution Into the Oceans

     Another heterogeneous removal process is the dissolution of
substances into the oceans.  The solubility of a substance is of
primary concern regarding this removal mechanism.  However,
because of the relatively slow mixing of the surface water above
the thermocline with the deep ocean (exchange time is approxi-
mately 15 years; Panel on Atmospheric Chemistry 1976), for disso-
lution into the oceans to be significant it must be followed by
the removal of the substance by reaction in the liquid phase
(e.g., oxidation, hydrolysis, dissociation, biological action,
etc.).
                                155

-------
     In an analysis of atmospheric removal processes, Slinn et
al. (1978) showed that the transfer rate of highly soluble (solu-
bility, a >102) or reactive substances from the atmosphere into
the oceans is controlled by the rate at which the substance is
transferred through the atmosphere to the ocean surface (gas
transfer rate).  The vertical gas transfer rate is approximately
1 cm s"1 for low molecular weight gases and is slightly slower
for high molecular weight gases.  If the height over which the
substance is distributed is 10 km, the corresponding tropospheric
lifetime of the substance is approximately 10 days.

     For substances of lower solubility (a <102), transfer of the
substance from the atmosphere to the ocean is controlled primar-
ily by the liquid phase transfer rate.  A lower limit to the
lifetime of these substances (a <102 ) can be estimated using a
relationship developed by Broecker and Peng (1974).  This estima-
ted lifetime is a lower limit because it is assumed in the deri-
vation of the relationship that the flux of the substance from
the ocean to atmosphere is zero; i.e., the effective concentra-
tion of the substance in the ocean is zero.  The relationship is

                               O TvlO8
                    Lifetime = ^   x   seconds,


where a is the solubility (volume of gas at STP absorbed by one
volume of water when the pressure of the gas above the water is
1 atm) of the substance.  Therefore, for the tropospheric life-
time to be less than a year, a must be greater than 7.  For gases
like 02, N2,  CH4 CO and Ar; a ~ 0.02.

     Factors such as chemical reaction of the dissolved substance
and solution pH can lead to an enhanced solubility coefficient
(Junge 1963) for the substance.  For example,  Hales and Sutter
(1973) have shown the effect that solution pH has on the dissolu-
tion of SO2.

     The substance can be removed from the dissolved phase by dis-
sociation, hydrolysis, or oxidation.  If any of these potential
reaction pathways for a substance is identified as being feasi-
ble, the reaction rates should be quantified so that the signif-
icance of that pathway for the removal of that substance from
the dissolved phase (and hence, the atmosphere) can be assessed.

2.2.6.  Surface Removal Processes

     There is very little information on surface removal proces-
ses.  Only one has had much attention.  The possibility that the
photoabsorption cross section of a substance will be modified
when it is physically adsorbed on a surface has been studied for
some of the chlorofluorocarbons adsorbed on sand (Ausloos et al.
1977).  if this enhanced photoabsorption leads to excitation of
dissociation,  the possibility exists for tropospheric removal by
this process.


                                156

-------
2.3.   Stratospheric Interaction

     If no significant tropospheric removal mechanism can be
identified for the substance of interest, then continued accumul-
ation of the substance in the troposphere will definitely lead to
significant transport into the stratosphere and beyond.  It is
reasonable to expect that the substance will begin to photodisso-
ciate above some altitude.  Furthermore, this substance and its
derivatives most likely will react with some radicals in the
stratosphere.  Consequently any impact analysis must be extensive
and draws upon all the available information on stratospheric
chemistry.

     If the tropospheric removal for the substance gives lifetime
of t,  then the maximum fraction of the substance dissociated in
the stratosphere can be determined by assuming that all of the
substance that is transported to the stratosphere gets dissociat-
ed there.  Thus, the exchange process between the troposphere and
the stratosphere will determine the maximum fraction.

     The average exchange time for stratospheric air with the
troposphere is about 15 years.  Since the troposphere contains
about four times as much air as the stratosphere, the lifetime of
the substance against the stratospheric sink is 6 years.  Conse-
quently, the lifetime of the substance against the tropospheric
and stratospheric removal is


                    1   - 6T  years.
                  1+1   T+6
                  T   6
Therefore, the maximum fraction of the substance dissociated in
the stratosphere is

                    6T  /6 =  T
                    T+6 '"   T+6 •


Since the amount of stratospheric ozone is maintained primarily
through the interaction of several photochemical catalytic cycles
the introduction of any substance that either directly or indir-
ectly interacts with chemical species involved in these cycles
will certainly affect the ozone layer.  Several well-known fami-
lies of hydrogen-, nitrogen-, and chlorine-containing molecules
are (H, HO, H02 ,  H2O2), (N, NO, N02, N03, N20s, HN03, HN04 ) and
(CH, C$,0, ECS,, EOCS,, C£ON02).  Therefore, a simple guideline  for
assessing potential impact on stratospheric ozone of any sub-
stance would be to study the reactivity of it or its derivative
substance with these molecules.  Any significant reaction that
may affect the balance of any family will lead to significant
impact on the ozone layer.  Furthermore, reactions with the so-
called source molecules such as CH4, N20, H20, CH3C£, CC£4,
CH3CC£3, etc., will also be of importance.  In addition to the
                                157

-------
chemical interaction,  one must also consider the potential impact
on the radiative and dynamic aspect of stratospheric climatology,
which further complicates the impact analysis.
3.  DATA REQUIREMENTS

     The proposed stratospheric impact assessment methodology
consists of a sequence of steps in which certain properties of a
substance are compared with evaluation criteria to determine the
significance of each of a series of potential removal mechanisms
or sinks.  A number of sources may be utilized in obtaining the
data and information required to assess the tropospheric removal
of chemical substances.  Data for existing substances may be
available from data banks of chemical properties, which are ac-
cessible in the literature; data for new and existing substances
may be obtained from industry through testing required under
TSCA; and physical or chemical properties of both new and exist-
ing of both new and existing substances may be estimated from
data that are available for structurally similar chemicals.  Test
methodologies and/or data sources are discussed here in terms of
the removal processes.


3.1  Gas Phase UV and Visible Absorption Spectra

     It is recommended to EPA that the gas phase ultraviolet and
visible absorption cross sections be measured for each constit-
uent throughout the 2950 to 7000 A wavelength interval.  A con-
tinuous scan survey of 15 A resolution is adequate for initial
appraisal.  Sample pressure, temperature, and relative humidity
should be varied over the range of conditions found in the
troposphere.  It must be recognized that in many cases such
measurements require an advanced level of experimental capability
and that generalized procedures may be misleading.  References
cited for consultation regarding the measurement techniques are
Chou et al. (1978), Tsubomura et al. (1964), Gordus and Bernstein
(1954), Lacher et al. (1950), and Calvert and Pitts (1977).
Extensive appraised cross sections appear in two recent NASA
publications edited by DeMore et al. (1979) and Hudson and
Reed (1980).


3.2  Destruction by Reactive Tropospheric Species

     There are, at present, no simple procedures for experimental
evaluation of the atmospheric reactivity of gaseous chemicals.
The information that is needed to evaluate the importance of
destruction of a substance by a particular tropospheric species
includes (1) knowledge of the atmospheric concentration of the
species [A], and (2) knowledge of the rate constants k for the
reaction of the atmospheric species with the chemical substance
                                158

-------
under conditions of temperature, pressure, relative humidity,
etc., relevant to the troposphere.

     Accurate calculation of the tropospheric lifetime requires
knowledge of [A] and k throughout the troposphere for all seasons;
however, for the present purpose, it is adequate to use globally
averaged quantities.  The tropospheric concentration profiles of
most reactive species are sufficiently uncertain to allow a
global average to be used.  Further, the only significant tropo-
spheric variable associated with nearly all rate constants is
temperature.  It is recommended that for estimating the lifetimes
of chemical substances in the troposphere, t = l/k[A], the follow-
ing quantities be used:  k = k(265K), [OH] = 5xl05 molecule cm"3,
and [03] = 7x101X molecule cm"3.  Rate constant data should be
taken from standard references such as Atkinson et al. (1979),
Hudson and Reed (1979), and Baulch et al. (1980).  When rate
constant data are not available, estimation methods such as
those given by Hendry and Kenley (1979) may be used.  These
estimates consider the principal mechanisms by which hydroxyl
radicals and ozone attach chemical substances and assume that
similar structural characteristics in different molecules will
have the same reactivity toward such attack.  On the basis of
known rates of reaction for processes such as hydrogen atom
abstraction, addition to olefinic bonds and addition to aromatic
rings, reasonable estimates of reaction rate constants for sub-
stances with similar structural features can be made.


3.3  Scavenging by Atmospheric Aerosols

     Vapor-aerosol partitioning in the atmosphere depends on the
vapor pressure of the organic compound, the size and surface area
of the suspended particulates, and the organic content of the
aerosol.  For vapor-aerosol distributions calculated from vapor
pressure and the quantity of atmospheric particulates see the
workshop report on Toxic Substances in Atmospheric Deposition.


3.4  Removal by Precipitation

     Wet removal of vapors is governed by the Henry's law con-
stant H, which can be calculated as the ratio of the substance's
vapor pressure to solubility.  Washout ratios calculated as the
reciprocal of H are given for a number of substances in the
workshop report on Toxic Substances in Atmospheric Deposition.


3.5.  Dissolution Into the Oceans

     Methods for calculating the flux of a chemical to a water
body have been described in the workshop report on Toxic Sub-
stances in Atmospheric Deposition.
                                159

-------
3.6.  Surface Removal Processes

     The vapor pressure of a substance bears upon this poorly
characterized removal process, but it is difficult to express  a
quantitative relationship.
4.  DATA SOURCES

     Numerous sources of required data are available in the
published literature, and privately maintained data bases may be
purchased.  A partial summary of sources of visible and ultra-
violet absorption spectra, vapor pressure, and solubility data
is presented in the Appendix.  For those substances for which
data are not available, the required information often can be
estimated on the basis of structure-correlation relationships
and similarities to substances for which data are available.
Provided that the specific chemical identity (molecular com-
position and structure) is known, the photochemical and chemical
reactivity of some substances can be estimated according to
methods developed by Hendry and Kenley (1979).  Other physical
and chemical properties of primary importance to determining
stratospheric significance by the proposed methodology can also
be estimated.

     Numerous methods for estimating environmentally important
physical and chemical properties of chemicals are available in
the published literature (Reid et al.  1977; Arthur D. Little,
Inc. 1978; Lyman 1978).
5.   REFERENCES

Arthur D Little, Inc. 1978.  Study performed for the US Army
Medical Bioengineering Research and Development Laboratory.

Atkinson R, Darnall KR,  Lloyd AC,  Winer AM, Pitts JN, Jr. 1979.
Kinetics and mechanisms  of the reaction of the hydroxyl radical
with organic compounds in the gas  phase.  In:  Advances in Photo-
chemistry,  Vol. 11, pp.  375-488,  John Wiley and Sons, New York,
NY.

Ausloos P,  Tebbert RE, Glasgow L.  1977.  Photodecomposition of
chloromethanes adsorbed  on silica  surfaces.  J. Res. Natl.  Bur.
Stds. 82:1.

Baulch DL,  Cox RA,  Hampson RF,  Jr.,  Kerr JA,  Troe J, Watson RT.
1980.  Evaluated kinetic and photochemical data for atmospheric
chemistry.   J. Phys.  Chem. Ref.  Data 9:295-471.
                                160

-------
Broecker WS, Peng TH. 1974'.  Gas exchange rates between air and
sea.  Tellus 26:21-35.

Calvert JG, Pitts JN. 1966.  Photochemistry-  John Wiley and
Sons, New York, NY.

Chou CC, et al. 1978.  Stratospheric photodissociation of several
saturated perhalochlorofluorocarbon compounds in current tech-
nological use.  J. Physical Chem., 82:1.

DeMore WB, ed. 1979.  Chemical, kinetic and photochemical data
for use in stratospheric modeling.  Jet Propulsion Laboratory
Pub. 79-29, Pasadena CA.

Gordus AA, Berstein RB. 1954.  Isotope effect in continuous
ultraviolet absorption spectra.  J. Chemical Phys.  22(5):290-295 .

Hales JM, Sutter SL. 1973.  Solubility of sulfur dioxide in water
at low concentrations.  Atmospheric Environment 7:997-1001.

Hendry DG, Kenley RA. 1979.  Atmospheric reaction products of
organic compounds.  EPA-560/12-79-001 report, U.S.  Environmental
Protection Agency, Washington DC.

Hudson RD, Reed El, eds. 1979.  The stratosphere:  present and
future.  NASA Ref. Pub. 1049, Washington, DC.

Junge CE. 1963.  Air chemistry and radioactivity.  Academic,  New
York, NY.

Junge CE. 1977.  Fate of global pollutants.  In:  Fate of
Pollutants in the Air and Water Environments, Part 1, Suffet IH,
ed., John Wiley and Sons, New York, NY.

Lacher JR, et al. 1950.  The near ultraviolet absorption spectra
of some fluorinated derivatives of methane and ethylene.  J.
American Chemical Soc. 72:5486-5489.

Lyman WJ. 1978.  Paper presented at the 1978 Spring Meeting of
the American Chemical Society, Anaheim, CA.

Moore HE, Poet SE, Martell EA. 1973.  222Rn, 210Pb, 210 Bi, and
210P profiles and aerosol residence times versus altitude.
J. Geophys. Res.  78:7065-7075.

Panel on Atmospheric Chemistry of the National Research Council.
1976.  Halocarbons:  Effects on stratospheric ozone.  National
Academy of Sciences, Washington,  DC.

Reid RC, Prausnitz JM, Sherwood TS. 1977.  The properties of
gases and liquids, their estimation and correlation.  Third
Edition, McGraw-Hill, New York, NY.
                                161

-------
Slinn WGN,  et al.  1978.   Some aspects of the transfer of
atmospheric trace  constituents past the air-sea interface.
Atmospheric Environment 12:2055-2087.

Tsubomura H,  et al.  1964.   Vacuum ultraviolet absorption spectra
of saturated organic compounds with non-bonding electrons.  Bull.
Chem. Soc.  Japan 37:417-423.
                               162

-------
                             APPENDIX:

               Sources of Physical and Chemical Data


Ultraviolet Absorption Spectra

(1)  Sadtler Research Laboratories, Inc.
     3316 Spring Garden Street
     Philadelphia, PA 19104
     215/382-7800
     Contact — Alen Bloom

Sadtler publishes data bases of chemical UV spectra.  The fol-
lowing data bases are available:

     (a)  Standard UV spectra;
     (b)  Pharmaceutical UV spectra;
     (c)  Biochemical UV spectra;
     (d)  Commonly abused drugs UV spectra; and
     (e)  Prepared and prescription drugs UV spectra.

Each data base provides UV absorption spectra of the chemicals
included and indices for accessing the spectral information.
Data are indexed alphabetically for chemical name,  chemical
class,  molecular formula, and UV absorption locator.

     (a)  Standard UV spectra.  Now comprises 102 volumes,  each
          containing approximately 500 spectra of pure chemical
          compounds.  Current price:  $135 per volume plus  $500
          for the complete index (Total = $14,300).  Four addi-
          tional volumes of new spectra are added each April.

     (b)  Pharmaceutical.  Now comprises 4 volumes with a total
          of 2,000 spectra.  The materials included are pharma-
          ceuticals, and therefore, may be composed of one  or
          several pure chemicals.  Total price is $964, including
          index.

     (c)  Biochemical.  Now comprises 2 volumes with a total  of
          650 spectra.  Total cost is $324, including index.

     (d)  Abused drugs.   Now comprises 1 volume with 300 spectra.
          Total cost is $318, including index.

     (e)  Prepared and prescription drugs.  Now comprises 2
          volumes with a total of 600 spectra.  Total cost is
          $293, including index.

Unlike  data base 1,  data bases 2 through 5 are not updated an-
nually.   Rather,  they may be updated by occasional issuing of a
new volume.
                                163

-------
All prices are subject to a price increase of approximately 10%
in September 1979.

(2)  The Thermodynamics Research Center
     Data Distribution Office
     Texas A & M Research Foundation
     F.E. Box 130
     College Station,  Texas 77843
     713/846-8765
     Contact — Dr.  Wilhoit

Publishes tables of thermodynamic and physical property data
(including vapor pressure) and spectral data (including UV) in
the form of loose-leaf data sheets.

Approximate costs of data files are:

     Thermodynamic and physical property data — $1,200.
     Spectral data — $400.

Supplements are issued twice each year for the thermodynamic and
physical property tables and at irregular intervals for the
spectral data.

TRC is building a computer data file that can be accessed by
remote terminal.  The only costs for these data would be the
nominal computer charges by the University Data Processing
Center.

(3)  Phillips JP, Feuer H, Thyagarajan BS (Eds.). 1945-1970.
     Organic Electromagnetic Spectral Data.  Volumes 1-12,
     John Wiley and Sons, New York,  NY.

          Gives ultraviolet-visible  spectra of over 250,000
          organic compounds.

(4)  DMV. UV Atlas of Organic Compounds. (Butterworths),
     1966-1971, Volumes 1-5,  Plenum  Press, New York, NY.

          Gives spectra of approximately 1,500 compounds.
          Cost:  $65 per volume.

(5)  Hirayama K. 1967.  Handbook of  Ultraviolet and Visible
     Absorption Spectra of Organic Compounds.  Plenum Press,
     New York, NY.

(6)  Friedel RA, Orchin M. 1951.  Ultraviolet Spectra of
     Aromatic Compounds.  John Wiley & Sons, New York, NY.

          Gives spectra of 579 compounds.

(7)  Ail-Union Synthetic Rubber Research Institute. 1966.
     Ultraviolet Spectra of Elastomers and Rubber Chemicals,
     Plenum Press,  New York,  NY.


                                164

-------
          Gives spectra of 141 compounds.

(8)  Calvert JC,  Pitts JN. 1967.  Photochemistry-   Second
     Edition,  Wiley and Sons, New York,  NY.

Photochemical and Reaction Rate Data

     Baulch DL, Cox RA, Hampson RF Jr.,  Kerr JA,  Troe J,
     Watson RT. 1980.  Evaluated kinetic and photo chemical data
     for atmospheric chemistry.  J. Phys.  Chem.  Ref.  Data,  9:295-
     471.

Solubility

(1)  Linke WF. 1965.  Solubility of Inorganic and Metal-
     Organic Compounds, Volumes I and II,  Fourth Edition, 1965,
     American Chemical Society, Washington,  D.C.

          Gives solubilities, in g/100 g solution, of approxi-
          mately 5,000 compounds.
          Index in Volume II.
          Cost:  Approximately $80.

(2)  Stephen H, Stephen T. 1963.  Solubilities of Inorganic
     and Organic Compounds, Volume I, Parts 1 and 2.   Pergamon
     Press, New York, NY.

          Gives solubilities, in wt. % or g/liter, of over 2,000
          compounds.  Index in Part 2.

(3)  National Academy of Engineers. 1972.   CAChe Physical
     Properties Book.  Washington, D.C.

(4)  JANAF Thermochemical Tables.  Published in looseleaf form,
     updated yearly.

Vapor Pressure

(1)  Thermodynamics Research Center

          See entry under Ultraviolet Spectra.

(2)  Thermodynamics Research Center. 1978.  Handbook of Vapor
     Pressures and Heats of Vaporization of Hydrocarbons and
     Related Compounds.

          Gives vapor pressure of over 700 compounds. Doubly
          indexed — alphabetically according to  subject index of
          Chemical Abstracts and by boiling point.

(3)  Weast RC  (Ed.). 1979.  Handbook of Chemistry and Physics.
     Chemical Rubber Co., Cleveland, OH.
                                165

-------
          Gives  vapor  pressure  of inorganic  ( «350)  and organic
          ( w1,000)  compounds.

(4)   Dreisbach RR. 1952.   Pressure - Volume  - Temperature
     Relationships of  Organic Compounds.   Handbook Publishers,
     Inc.,  Sandusky, OH.

          Gives  vapor  pressure  data on  compounds  in  23  families
          of  organic chemicals.

(5)   Jordan TE.  1954.  Vapor Pressure of Organic  Compounds.
     Interscience Publishers, New York, NY.

          Gives  vapor  pressure  of 1,492 chemicals. Indexed alpha-
          betically  and by family.

(6)   JANAF Thermochemical  Tables.

          See  Solubility References.
                               166

-------
    Report of a Workshop on




THE IMPACT OF CHEMICALS ON THE




 RADIATIVE TRANSFER IMBALANCE
       Boulder, Colorado



        September 1979
          Edited by



       John J. DeLuisi
              167

-------
                   PARTICIPANTS AND CONTRIBUTORS
Thomas Ackerman
Space Science Division
NASA - Ames Research Center
Moffett Field, CA  94035
Kirby Hanson
U.S. Dept. of Commerce
NOAA/ERL/ARL/GMCC, RF3 292
325 Broadway
Boulder, CO  80303
Marshall Atwater
Center for the Environment & Man
275 Windsor Street
Hartford, CT  06120
John Miller
NOAA/ERL/ARL, R32
Silver Spring, MD  10910
Robert Charlson
Dept. of Civil Engineering
Water and Air Resources
University of Washington
Seattle, WA  98195
Rudolph Pueschel
U.S. Dept of Commerce
NOAA/ERL, R31
325 Broadway
Boulder, CO  80303
John J. DeLuisi - Convener
U.S. Dept. of Commerce
NOAA/ERL/ARL/GMCC,  RF3 292
325 Broadway
Boulder, CO  80303
Hal Rosen
Lawrence Berkeley Laboratory
University of California
Berkeley, CA  94720
James Friend
Drexel University
32nd and Chestnut
Philadelphia, PA  19104
William P. Wood
Office of Toxic Substances
U.S. EPA       (TS-792)
401 M Street, S.W.
Washington, D.C.  20460
James Hansen
Institute for Space Studies
GSFC-NASA
New York, NY  10025
                               168

-------
                             CONTENTS

                                                            Page

PARTICIPANTS AND CONTRIBUTORS                                i.68

PREFACE                                                      170

1.   INTRODUCTION                                            171

2.   CRITERIA FOR DETERMINING SIGNIFICANT RADIATIVE EFFECTS  172

3.   GASES AND SECONDARY PRODUCTS                            173

     3.1.  Reaction With Atmospheric Constituents            175
     3.2.  Gas-Particle Transformations                      176

4.   PARTICLES                                               176

     4.1.  Direct Effects                                    176
     4.2.  Indirect Effects                                  179

5.   THE EFFECT OF CHEMICALS ON THE OPTICAL PROPERTIES
      OF CLOUDS                                              179

     5.1.  The CCN Problem                                   180
     5.2.  The Absorption Problem                            182

6.   REFERENCES                                              184

APPENDIX:  Atmospheric Chemical Processes                    187
                                169

-------
                             PREFACE


     This workshop was given the task of delineating the signif-
icant factors, possessed by any chemical, that may directly or
indirectly affect the radiation balance of the atmosphere.
Atmospheric physics research has already given information on
direct and indirect effects.  An example of direct effect is the
cumulative infrared greenhouse effect of several trace gases,
some of which are anthropogenic (Wang et al. 1978), and the short
wave radiative effects of unactivated aerosol particles in clouds
(Ackerman and Baker 1977).  Examples of indirect effects are
those of aerosols as they change cloud droplet size distributions
and hence albedo and radiative properties (Twomey 1976), and as
found in observations of albedo and absorption anomalies in
clouds (Robinson 1958),  and cloud modification by urban pollution
(Barrett et al. 1979).
                              170

-------
   THE IMPACT OF CHEMICALS ON THE RADIATIVE TRANSFER IMBALANCE
 1.  INTRODUCTION

     There are several general factors that have a bearing on
whether a chemical will have a significant effect on the radia-
tion balance.  These factors may be posed as questions:

     (1)  Does the chemical strongly absorb infrared or visible
          radiation?
     (2)  Will great quantities of the chemical be released to
          the atmosphere?
     (3)  Does the chemical possess efficient water/ice particle
          nucleation properties, and
     (4)  Does the chemical possess catalytic properties that
          will influence/alter gas-to-particle conversion proces-
          ses which ultimately affect the radiation balance?
     (5)  Is the expected lifetime of the chemical long (years)
          or short (days)?

     It is obvious that these factors are not independent of each
other.  For example, a very minute amount of a trace gas having
powerful infrared radiative properties (1) may influence the
radiation balance; on the other hand, large quantities of gas (2)
with extended lifetime (5) and with small infrared radiative
properties might produce a similar effect.  It is conceivable
that (1) may be influenced by (4) which results in efficient ice
nucleating material (3).  In this report we do not consider
photochemical influence on natural gases (primarily ozone) since
it has been the subject of another workshop.

     The Appendix describes a comprehensive organized scheme of
atmospheric chemical processes which covers all aspects of the
source-sink fate of any chemical introduced into the atmosphere.
There is a need to consider the short lifetimes of chemicals in
certain intermediate states which may also play an important role
in determining the ultimate effect on the radiation balance.

     Ah absolute reference does not exist from which to define
significant radiative effects.  Any concentration of any sub-
stance will have an effect on the radiation field, although it
may be physically imperceptible and climatologically insignifi-
cant as well.  Consequently, the need for a reference scale stim-
ulated workshop participants to adopt a set of criteria based
upon a prespecified perturbation to atmospheric temperature in-
cluding its dependence on a spatial scale.  The current set of
criteria is subject to change with new results from modeling and
experimental efforts since the state of the art in our under-
standing of radiation and climate is not well advanced.

     With regard to the potential effect of industrial chemicals,
it is important to maintain a perspective of what is possible and

                               171

-------
what is probable.  It is highly possible that a chemical may
possess some adverse radiative property, but the question that
must follow is whether a sufficient amount will be released to
produce a significant effect.  The probability of this happening
is undoubtedly low.  There is a greater, but still low, probabi-
lity that the radiative effects of several or more substances may
combine in a cumulative way to produce a non-negligible tempera-
ture perturbation.  Cumulative effects can be estimated to a
first approximation by simple addition of the individual tempera-
ture perturbation of each substance released.  No specific recom-
mendation was given by the panel for action to be taken if in a
particular circumstance the potential exists for a significant
cumulative effect.  However, it would seem reasonable to maintain
an inventory on those released substances having potentially
significant radiative effects in order to reduce the chance of
being caught unaware of the situation.

     The sections that follow represent a consensus of experts in
the fields of aerosol chemistry and physics, atmospheric gas
chemistry, atmospheric radiative transfer and thermodynamical
processes, and aerosol optical properties.
2.  CRITERIA FOR DETERMINING SIGNIFICANT RADIATIVE EFFECTS

     Anthropogenic chemical products released in the atmosphere,
both gases and aerosols, are potentially capable of absorbing
solar and/or terrestrial thermal radiation and thereby affecting
the atmospheric temperature and climate.  Furthermore, it is
necessary to consider not only primary pollutants (chemicals
released into the atmosphere),  but also secondary products which
may be formed by transformation of the primary pollutant.  Esti-
mates of effects of secondary products must be based on our
current knowledge of atmospheric chemistry.  Since there is still
much to be desired in our understanding of gas-to-particle-forma-
tion chemistry, we can expect changes in such estimates as our
knowledge grows.

     The magnitude of radiative impact that constitutes a signi-
ficant threat depends upon the horizontal scale over which the
change in atmospheric composition is effective.  Changes that are
global in scale, as in the case of chemicals with a tropospheric
lifetime of years or longer, require a relatively smaller radia-
tive impact to yield a significant climate change.  On the other
hand,  regional and local changes must be relatively large in
order to be significant compared with natural variability and
with other known or anticipated effects of human activity.1


xThe panel acknowledges the fact that the global climatic effects
of anthropogenic pollutants (despite the tremendous amounts
released to the atmosphere) are still under debate.
                               172

-------
     For those released constituents or atmospheric modifications
that will be global in extent, the threshold level for a signif-
icant temperature effect is -^0.1-0.5 K.  This is based primarily
on the fact that the variation of global temperature over the
last century, which has witnessed significant climate variation,
is -^0.5 K (cf. Figure 1).  In addition, climate models suggest
that the largest human-induced climate impact during the next few
decades will be due to increasing atmospheric C02, and will be of
this order of magnitude.  For dispersals on a synoptic scale the
magnitude for a significant temperature impact is -^0.5-2 K (Ball
and Robinson 1979), as a result of the larger natural variability
on such a scale.  For a local dispersal, e.g., over a given city,
the change required for a substantial impact is -^1-5 K (Clarke
1969; Atwater 1975, 1977).

     These temperature criteria can be converted to specifica-
tions on gaseous, aerosol, or cloud properties by employing
simple radiative models that give order-of-magnitude estimates of
the temperature change accompanying a given change in atmospheric
composition.  As a uniform basis of comparison the radiative/con-
vective temperature profile models, as described by Manabe and
Wetherald (1967), Wang et al. (1978) and Ramanathan and Coakley
(1978), can be used.  These models are reasonably consistent; for
example, they yield a warming of -^2 K for a doubling of atmos-
pheric C02.   In fact, more complex models of the climate system
generally agree with the radiative/convective models within a
factor of 2 or so.  The real climate system contains many feed-
back effects, as in snow/ice albedo, cloud properties, and mixing
of heat with the deep ocean layers.  With our present ignorance
of these effects, we are forced to assume that the simple models
yield appropriate order-of-magnitude estimates of temperature
effects.

     Under certain circumstances, for example in highly polluted
regions, a trend in the surface irradiance or visibility might be
measurable.   For the same situation, a study of temperature
fluctuations may be too noisy for establishing a cause and effect
case against a specific pollutant (e.g., Husar et al. 1979).  In
this case, it may be necessary to resort to model analysis for a
solution; however, the optical parameters for the pollutant must
be known.  Clearly, this approach is severe in terms of complex-
ity and expense, and should be avoided unless there is ample
justification for an undertaking.  Generally, these situations
are unique because of local topological conditions, so special
consultation might be required for an investigation.
3.  GASES AND SECONDARY PRODUCTS

     A primary pollutant is defined as a substance that is di-
rectly released to the atmosphere.  Secondary products are de-
fined as transformed primary pollutants which may be in a gaseous
                               173

-------
                            Northern
                           Hemisphere
                               Southern
                               Hemisphere
     I88O    1900      I92O      1940

                          Date
I960   1976
Figure 1.  Global  and hemispheric variations  of surface
air temperature during the past century, based on data
for surface stations  archived by the National Climate Data
Center.  (Graph supplied by panel member J. Hansen.)
                          174

-------
or particulate form.  The potential radiative effects will depend
upon the five factors stated in the introduction.

     For gases the most common case requiring examination will be
absorption bands in the thermal infrared window region (8-14 pm).
For this case an upper limit of the possible temperature change
can be obtained simply from the total absorptance of the infrared
bands.  The thermal impact depends upon the location of the
absorption in the window region, with bands in the most trans-
parent regions having the largest effect.  An estimate of the
impact of an infrared absorbing gas on the surface temperature
can be obtained from the relation


                         AT = 0.18 AB,   .


where A is the band absorptance in inverse centimeters, i.e., the
product of the band intensity and the vertical column amount of
the absorbing gas, and B(v ) is the Planck flux (W m~2s~1cm~1)
at the frequency of the band center computed for a global mean
surface temperature of 288 K.  The gas amount can be found as the
product of the projected gas release rate and its estimated life-
time.  If AT found in this way is not negligible, e.g., if it is
>^ 0.1 K for a globally dispersed gas, then the gas should be ex-
amined in more detail.  As a first step a more accurate estimate
of the temperature impact could be obtained from a radiative/
convective model.

Lifetime

     One of the most powerful and reliable guides to estimating
the potential importance of a substance is its lifetime in the
atmosphere because it defines the time and space scales over
which the substance will be distributed.  The four factors listed
below are quantities that can be used to estimate lifetimes.

     (1)  Reactivity with OH and other reactive substances (e.g.,
          S02) in the atmosphere.
     (2)  Solubility in fresh and sea water.
     (3)  Absorption in biological surfaces and reaction char-
          acteristics on surfaces.
     (4)  Association with aerosols.

3.1.  Reaction With Atmospheric Constituents

     The lifetimes of pollutants can be controlled by reactions
with one or more atmospheric constituents as indicated above.
Also the transformation of one substance into another  (generation
of secondary pollutants) usually occurs by means of reaction with
an atmospheric constituent.  Information for establishing crite-
ria could be obtained by testing for reactions with atmospheric
constituents and identifying the reaction products.  Ultimately,
                               175

-------
it may be necessary to understand the entire chain of atmospheric
reactions, not merely the results of the first transformation.

3.2.  Gas-Particle Transformations

     If the primary pollutant undergoes transformations that lead
to (1) new particle formation and (2) growth of existing parti-
cles, then clearly it will affect radiative transfer, and the
composition of aerosols and precipitation (for which the parti-
cles act as CCN) may be altered.   An assessment of such effects
could be aided by testing the primary pollutant for (1) its
ability to form particles under atmospheric conditions and for
(2) its ability to cause a test aerosol to grow when subjected to
atmospheric conditions.
4.  PARTICLES

4.1.  Direct Effects

     Suspended particulate matter affects the radiation balance
directly by scattering and absorbing both visible and infrared
radiation.  These optical effects have already led to visibility
degradation (Husar et al . 1979) and a reduction in surface irra-
diance in urban areas (Ball and Robinson 1979; Peterson and
Stoffel 1979) and may ultimately lead to regional and global
climate modification.  In order to understand and quantitate
these effects, detailed knowledge of the optical properties of
aerosol particles (i.e., absorption coefficient and angular
distribution of scattering coefficient) as a function of particle
size, chemical composition, and wavelength is necessary-  Such
detailed information is difficult to obtain, and it would not be
appropriate from a cost benefit point of view to require such
analyses.  We, therefore, have considered an approximate three-
parameter characterization of the aerosol optical properties to
screen particulate emissions,  which could have a significant
impact on the radiation balance.  Such a screening procedure
would permit rapid processing of the vast majority of emissions
which do not come close to being a radiative hazard.  The three
parameters that are relatively straightforward to measure are


                                               dm
     (1)  Submicron aerosol mass emission rate -rr- ,
          where m_ =        ±-  N(r) dr;
                 s     Q    j

          N(r) is the number distribution function, and
          r is particle radius.
     (2)  Integrated absorption coefficient per unit mass in the
          visible spectral region between 0.3 and 0.7 |j .
                               176

-------
 0.01
Figure 2.  Warming and cooling
of the atmosphere by aerosols
as a function of the ratio of
absorption (a) to backscatter-
ing (P) and the surface albe-
do (a) based on the results
of Chylek and Coakley (1974).
            Surface Albedo (a)
      (3)  Integrated  absorption  coefficient per unit mass in the
          infrared window between  8  and  14 |j .

      The optical properties  of an  aerosol are critical to the
warming or cooling of the atmosphere.  This is demonstrated in
Figure 2 which shows  warming or  cooling  of the atmosphere depend-
ing on the surface albedo and the  absorption-to-backscatter ratio
of aerosols.

      Since the real part of  the  refractive index does not vary
significantly from one substance to  another  (1.33  - 1.6), one can
make  an estimate of the scattering coefficient from the submicron
mass  alone (Bhardwaja et al.  1974).  This estimate will be based
on using the maximum  scattering  coefficient per unit mass, which
is found for particles ^ O.I  - 0.4 p in  radius, and will therefore
give  an upper limit on the scattering contribution.  The choice
of particle density and size  distribution is left  open at this
time; however, on the basis  of experience a density of 2 g cm 2
and size distribution exponent of  v=3 (based upon  the Junge law
d N/d log r a r"v) would seem reasonable for calculation of a
mass  scattering coefficient  until  representative size distribu-
tions are made available.  The other two parameters will give a
direct measure of the absorption coefficient in the visible and
infrared regions.  These three parameters, when coupled to ap-
propriate dispersion  models,  should  allow an approximate estimate
of the radiative temperature  impact  of these emissions on a local
and regional basis.   As a rule of  thumb, if these  emissions do
not make a significant contribution  to the ambient submicron
                               177

-------
mass, they will not make a significant perturbation  on  the  scat-
tering effects of particulates.  The situation could be somewhat
complicated by the absorption effects since a very small amount
of highly absorbing material may produce large optical  effects.
This  is demonstrated by the fact that "graphitic" carbon domi-
nates the optical absorption coefficient in urban areas and in-
dustrial regions, yet it represents a relatively small  fraction
of the aerosol mass (Rosen et al . 1978 a and b, and  Weiss et al .
1979).  On the other hand, the likelihood of large releases of
highly absorbing particles is seen as rather remote.

      It is not possible to make a reliable estimate  of  the  direct
effect of aerosols on the radiation balance, or the  indirect
effects of gases or aerosols through modification of cloud  prop-
erties, at our present state of knowledge.  However, we can
identify extreme cases that are likely to perturb the radiation
balance.  In the case of a purely absorbing aerosol, the optical
thickness required to cause a warming of -vO.lK is -^0.002.   Thus
with  the magnitude of a significant temperature perturbation
defined as above, the limit on optical thickness for absorbing
aerosols is given by

                         At  -\> 0.02 AT,
                           a

with  AT in °C.  For a purely scattering aerosol the  relation is

                         At  -v -0.09 AT
                           o

with  increased aerosol loading causing a cooling effect in  this
case.  These approximate relations are based on the  assumption
that  the aerosol radiative effects in the visible are larger than
in the thermal infrared, which is normally the case  for aerosols
small enough to remain airborne for a climatologically  signifi-
cant  time.  The optical thickness of the aerosols is given  by
                       0
                          0

                        / Qext(r) rtr2 n(r)dr
                            exT:
where n(r)dr is the number of particles above unit area with radius
between r and r + dr, and Qext(r) is the extinction efficiency
factor (van de Hulst 1957; Hansen and Travis 1974); as an approx-
imate upper limit Qext can be set equal to 2.  Reliable size
distribution measurements are important for the estimation
of t .   If the temperature impact estimated in this way is found
to be significant (as defined above) for a projected aerosol
release,  it should be subjected to more detailed study.

     Gas and aerosol chemical releases may also indirectly impact
the earth's radiation balance by modifying cloud properties.
However,  the physics involved is not adequately understood, and
it is  thus not possible to specify what gas or aerosol releases
                               178

-------
would be sufficient to cause a significant problem through radia-
tive effects of cloud changes.  Research in cloud and aerosol
physics and the interactions with radiation must be pursued
extensively before it will be possible to recommend constraints
on gas and aerosol releases because of their potential effects on
cloud properties.  More will be said on this subject in Section


     With the exception of stratospheric particles, the lifetime
of suspended particles in the troposphere is usually on the order
of a few days to a few weeks at most.  This does not completely
void the possibility of subtle global radiative effects from
gases that eventually end in the chemical makeup of clouds after
long-range transport.

     Chemicals released to the atmosphere are unlikely to react
(if they do at all) with natural aerosols (ocean salts, crustal
materials and biotic effluent) to produce any serious imbalance.
If reactions should occur, then in all likelihood, larger parti-
cles may result and removal will occur sooner.

4.2.  Indirect Effects

     Aside from the effects of particulate emissions on clouds
(discussed in section 5.1), probably the most important indirect
impact of these emissions could be due to their catalytic activ-
ity-  It is well known that+very+small concentrations of appro-
priate catalysts (e.g., Fe3 , Mn 4, soot) can dramatically change
reaction rates (see, e.g., Middleton and Kiang 1979; Chang et al.
1978).  These catalytic agents could significantly enhance atmos-
pheric gas-to-particle loading and have a large impact on radia-
tion imbalance.  A screening procedure must therefore be devel-
oped to assess the possible catalytic role of various particulate
emissions in effecting the transformation processes of S02 to
particulate sulfur, NO  to particulate nitrogen, and hydrocarbon
to particulate carbon.x

     A well-defined generally applicable procedure for such
screening is not readily available, and at present each effluent
will have to be assessed individually with the advice of experts
in the field if significant amounts of emissions are expected.
5.  THE EFFECT OF CHEMICALS ON THE OPTICAL PROPERTIES OF CLOUDS

     Before dealing with the specifics of this problem, several
general observations are worth noting.  First of all, the inter-
actions between atmospheric chemicals and the hydrologic cycle
are not well understood, even in the case of ubiquitous quanti-
ties such as sulfur and nitrogen compounds.  Consequently, one
can hardly expect to make definitive statements with regard to
                               179

-------
the somewhat more esoteric compounds being produced by the chem-
ical industry.  Second, deducing the radiative effects of an
absorbing particulate in an atmosphere containing clouds is not a
trivial problem.  Even answering the simplest possible problem,
which is whether the particles are likely to warm or cool the
environment, is complicated by many factors such as the index of
refraction of the particles, the location of the particle layer,
the optical properties of the clouds, and the albedo of the
underlying surface.  Finally, our understanding of the nature of
gas-to-particle conversion in the atmosphere is still incomplete.
This makes it difficult to predict with certainty whether or not
released gases will become particles and, if they do so, what the
chemical constituency of the particle would be.

     These preliminary comments should not be construed as imply-
ing that no assessment of the effects of chemicals on clouds can
be made.  Rather, they are meant to convey some of the difficul-
ties that must be faced in making these assessments.  Since our
knowledge of cloud formation systems is incomplete, any assess-
ment is likely to contain some aspects that are based more on
scientific speculation and educated estimates than on scientific
facts.  Additionally, there will be many cases in which a quali-
tative assessment can be made with some degree of certainty but
where a quantitative assessment is out of the question.

     Turning now to the specifics of the problem, there are two
principal questions to be considered.  The first, and perhaps the
most important, is whether particulates, either primary or sec-
ondary, are likely to act as cloud condensation nuclei (CCN) and
what effect these nuclei would have on the cloud droplet size
distributions.  The second is whether gases or particles will
affect the radiative properties of the clouds directly.  The
effect of gases can be treated fairly simply but the particulates
present a substantially more difficult problem.

5.1.  The CCN Problem

     An assessment of the effect of particulates as CCN can be
made as follows:

     (1)  What is the size distribution of the particles?

     For the particles to act as effective CCN, the particle
radius must be in the general range 0.11 < r < 5.0 microns.  If
the particles are smaller than this they are unable to attract
water because of curvature effects.  (Note the growth character-
istics illustrated in the Kohler curves of Figure 3.)  Larger
particles tend to fall out of the atmosphere before they achieve
significant concentrations.

     (2)  What is the solubility of the particles?

     Chemical substances with low solubility are poor CCN since
high water vapor supersaturations are required to make them grow.


                               180

-------
                          Drop Radius (cm)

   Figure 3.   Variation of equilibrium vapor pressure over an
   aqueous solution drop with drop size, for various amounts
   of NaCI (solid lines) and (NH4)2S04  (dashed lines) in solu-
   tion, and for 20°C (after Pruppacher and Klett 1978).
(Note the growth curves for pure water and weak solution droplets
as opposed to those for strong solutions in the figure.)  Con-
versely, hygroscopic substances such as salts and some acids
(e.g., NaCI and H2S04) are very effective CCN.  Thus, if the
chemical under study is a polymer or forms a particulate composed
of organic compounds, it is unlikely to be important as a CCN.
But if it is a sulfate or nitrate or reacts with such species, it
may have a significant effect.

     (3)  What is the quantity of the added CCN relative to the
          ambient CCN?

     If the number of additional CCN is less than 1% of the
ambient CCN on the local or regional scale CCN are probably
unimportant.  This is not a rigid condition because a small
number of hygroscopic CCN can cause a substantial change in the
size distribution of a local cloud.  (For an example see the
study by Barrett et al.  1979.)  Obviously a local effect must be
evaluated in relation to surrounding regional effects.

     (4)  What are the radiative effects?
                               181

-------
     Given that the particles act as CCN and are sufficiently
numerous to affect the cloud droplet distribution, their radiative
effects must be assessed.  This assessment cannot be made a
priori since changes in the cloud droplet spectra will depend on
both the composition and number of the particles.  In addition,
the particles will usually change the index of refraction of the
droplets.  Twomey (1977) suggests that for all but the thickest
clouds (optical depth greater than 100),  adding particles will
increase the albedo of the system.  Since his calculations
neglect some of the complexities of the problem, his conclusions
should be used with care.  (Note: the basic logical chain of this
and subsequent sections is outlined in Figure 4).

5.2.  The Absorption Problem

     As we mentioned above, the effect of gaseous absorption can
be treated in a straightforward manner:

     (1)  What is the solubility of the gas?

     Obviously, if the gas is not soluble,  it will not have a
large effect on the radiation balance of a cloud.

     (2)  What are the spectral absorption characteristics of the
          gas in solution?

     Pure water droplets are non-absorbing at visible wavelengths
but many solutions absorb radiation because of the properties of
the solute.  If the solution formed by the released gas is moder-
ately or strongly absorbing,  it can have a noticable effect on
the heating in the cloud.

     (3)  What are the radiative effects?

     Typical estimates of radiative cooling from stratus clouds
are ~0.5 K per hour.  Consequently, a heating rate of ^-0.1 K per
hour due to dissolved gases would be a significant addition to
the radiation budget of stratus clouds.  It is very unlikely that
heating rates would reach the rates associated with cumulus
clouds since these rates are an order of magnitude larger than
those for stratus clouds.

     Dealing with the effect of particulates on the absorption
characteristics of a cloud is difficult because we know very
little of what happens to the particulates in the cloud.  While
we can sketch a general approach, much of the quantitative know-
ledge needed is not yet available.  Questions asked in the ap-
proach are as follows:

     (1)  Is the particulate soluble?

     If the particulate is soluble, it acts as a CCN (see section
5.1).  In general, it will also change the optical properties of
                               182

-------
00
u>
                          Released gas
                  No
           Exit-
          Yes
                  No
           Exit-*-
Absorption spectra
   of solution?
                       Radiative effects
                                                Gas-to-particle
                                                  conversion
CCN quantity
>0.01 ambient?
                                                           Yes
                                                     Released particulate
                                            No
-Exit
                          Radiative  effects
                                                                                   Size:
                                                                               0.01
-------
the cloud drops,  and this effect should be handled as outlined in
the procedure for gases.

     (2)  Is the particle light-absorbing?

     If the particle is neither soluble nor light-absorbing, it
is unlikely to have climatic implications.  If it does absorb
visible radiation, then it is potentially important as a modifier
of the radiation field of the cloud. Because of the high number
of scattering events in a cloud which usually is optically thick
(t  > > 1.0), a small quantity of material can absorb a signif-
icint amount of energy.  The absorbing particle can exist then on
the surface, within the droplet, or in the air between droplets
and absorption effects will still occur.

     (3)  What are the radiative effects?

     Some general answers to this question for nonsoluble parti-
cles have been proposed by Ackerman and Baker (1977).  They
conclude that pure scattering particles increase the system
albedo for all types of clouds.  Absorbing particles increase the
albedo for thin clouds but decrease it for moderate to thick
clouds.  This latter effect is due to the dominance of absorption
over backscatter effects as the optical depth increases.  These
results are only qualitative and need to be greatly extended
before they could be used to assess the impact of absorbing
particles of clouds.  It would also be necessary to relate the
magnitude of these effects to others that affect clouds on the
local and regional scale.
6.   REFERENCES

Ackerman TP, Baker MB.  1977.   Shortwave radiative effects of
unactivated aerosol particles in clouds.  J.  Appl.  Meteorol.
16(1):63.

Atwater MC. 1975.  Thermal changes induced by urbanization and
pollutants.  J. Appl. Meteorol.  14(6):1061.

Atwater MA. 1977.  Urbanization and pollutant effects on the
thermal structure in four climatic regimes.  J. Appl. Meteorol.
16(9):888.

Ball RJ, Robinson GD. 1979.  The origin of haze in central
U.S.A. and its effects on solar irradiation.   Center for Environ-
ment and Man,  Hartford,  CT, Rep. 4222-670, 74 pp.

Barrett EW, Parungo FP,  Pueschel RF. 1979.  Cloud modification by
urban pollution:  A physical demonstration.  Meteorol. Rundsch.
32(5):136.
                               184

-------
Bhardwaja PS, Herbert J, Charlson RJ. 1974.  Refractive index of
atmospheric particulate matter:  An in situ method for determin-
ation.  Appl. Opt. 13(4):731.

Butcher SS, Charlson RJ. 1972.  An introduction to air chemistry,
Academic, NY.

Chang SG, Brodzinsky R, Toosi R, Markowitz SS, Novakov T. 1979.
Catalytic oxidation of S02 on aqueous suspensions.  Conf. on
Carbonaceous Particles in the Atmosphere, Mar. 20-22, 1978,
Lawrence Berkeley Lab., Berkeley, CA, LBL-9037, pp. 122-130.

Chylek P, Coakley J. 1974.  Aerosols and climate.  Science
183:72-75.

Clarke JF. 1969.  Nocturnal urban boundary layer over Cincinnati.
Won. Weather Rev. 97(8):582.

Hansen JA, Travis LD. 1974.  Light scattering in planetary at-
mospheres.  Space Sci. Rev. 16(4):527.

Husar RB, Paterson DE, Holloway JM. 1978.  Trends of eastern U.S.
haziness since 1948.  Preprint Proc. 4th Symp. on Turbulence,
Diffusion, and Air Pollution, Reno, Nev., Jan. 15-18, 1979, AMS,
Boston, MA, pp. 249-256.

Manabe S, Wetherald RT. 1967.  Thermal equilibrium of the atmos-
phere with a given distribution of relative humidity.  J. Atmos.
Sci. 24(3):241.

Middleton P, Kiang CS. 1979.  Relative importance of nitrate and
sulfate aerosol production mechanisms in an urban atmosphere.
In:  Grosjean D, ed. Nitrogenous air pollutants, chemical and
biological implications.  Ann Arbor Science Publishers Inc., Ann
Arbor, MI, p. 269.

Peterson JT, Stoffel TL. 1979.  Urban-rural solar radiation
measurements in St. Louis, Missouri.  NOAA Tech. Memo. ERL-ARL-
76, 51 pp.

Pruppacher HR, Klett JD. 1978.  Microphysics of clouds and
precipitation.  D. Reidel Publishing Co., Dordrecht, Holland;
Boston, MA, 714 pp.

Ramanathan V, Coakley JA. 1978.  Climate modeling through  radia-
tive-convective models.  Rev. Geophys. Space Phys. 16(4):465.

Robinson GD. 1958.  Some observations from aircraft  of surface
albedo and the absorption of clouds.  Arch. Meteorol. Geophys.
Bioklimatol., Ser. B,9:28-41.
                               185

-------
Rosen H,  Hansen ADA,  Gundel L,  Novakov T. 1978.  Identification
of the optically absorbing component in urban aerosols.  Appl.
Opt. 17(24):3859.

Rosen H,  Hansen ADA,  Gundel L,  Novakov T. 1979.  Identification
of the graphitic carbon component of source and ambient particu-
lates by Raman spectroscopy and an optical attenuation technique.
Proc. Conf. on Carbonaceous Particles in the Atmosphere, Mar. 20-
22, 1978, Lawrence Berkeley Lab,  Berkeley, CA, LBL-9037, pp. 49-
55.

SCEP- 1970.  Man's impact on the  global environment:  assessment
and recommendations for action.  Report on the Study of Critical
Environmental Problems.  Massachusetts Institute of Technology
Press, Cambridge,  MA, 319 pp.

Twomey S. 1977.  The role of aerosols in influencing radiative
properties of clouds.  In:  Bolla HJ, ed. Symposium on radiation
in the atmosphere, IAMAP, Garmisch-Partenkirchen,  Germany, August
19-28, 1976, Science Press, Princeton, NJ, pp. 171-175.

Twomey S. 1977.  The influence  of pollution on the shortwave
albedo of clouds.   J. Atmos. Sci. 34(7) .-1149-1152.

van de Hulst HC. 1957.  Light scattering by small  particles.
John Wiley and Sons,  New York,  NY, 470 pp.

Wang WC,  Yung YL,  Lacis AA, Mo  T, Hansen JE. 1978.  Greenhouse
effects due to man-made perturbations of trace gases.  Science
194(4266):685-690.

Weiss RE, Waggoner AP, Charlson RN,  Thorsell DL,  Hall JS,  Riley
LA. 1979.  Studies of the optical, physical and chemical prop-
erties of light absorbing aerosols.   Proc. Conf.  on Carbonaceous
Particles in the Atmosphere, Mar. 20-22,  1978, Lawrence Berkeley
Lab., Berkeley, CA, LBL-9037, pp. 257-262.
                               186

-------
            APPENDIX:  Atmospheric Chemical Processes


     The material below is quoted from Atmospheric Chemistry,
Problems and Scope, NAS, Washington D.C., 1975, pp. 108-109:


The Role of Aerosol

     Figure [A-l] shows the atmosphere as a chemical system with
a flow of trace substances from source to sink via a large number
of possible routes.  Various classes of trace substances or their
physical state are shown as separate boxes in the diagram.

     These figures show two basic types of meteorological effect
resulting from the presence of specific trace substances in the
atmosphere:  (1) direct radiative interactions and (2) water in
nucleation processes and the like.  Nucleation processes are
manifested in two general ways:

     (1)  Modification of albedo by changes in clouds, or other
          changes in radiative processes, including modification
          of surface albedo by snow accumulation, conversion to
          desert, and dry soil to wet soil;
     (2)  Hydrometeor modification, including amounts, composi-
          tion, and optical properties.

     It is clear that most current efforts center on direct
climatological involvement via albedo changes and modification of
other radiative processes (SCEP 1970).  While a great deal of
attention has been given to the study of cloud and nucleation
microphysics, little of it has developed to the point at which
global or even mesoscale or synoptic-scale effects can be un-
equivocally explained.

     There still are uncertainties about whether increases in
atmospheric particulate matter will act to heat or to cool the
earth.  The uncertainty exists for several reasons.  First, there
are various approaches to ascertaining the effects of particulate
matter even when the physical and chemical properties are known.
Unfortunately, in the case of the atmosphere they are not known
well enough.  It is particularly important to know the sizes,
Shapes, and indices of refraction of the particles.  The complex
part of the refractive index is governed by the molecular charac-
ter of the particles, which is a problem of certain  interest to
air chemistry.

     The following topics represent a partial  list of relevant
problems for which increased effort seems justified.

     (1)  Rate of conversion of gases to particles and the  clima-
          tological consequence of the aerosols so created;
     (2)  The possible  stabilization of  fog or clouds by  organic
          matter;


                                187

-------
(3)   The role  of  deliquescence  in  radiative  balance of
      clouds, fog,  and haze;
(4)   Relative  importance  of  reactions  in cloud  drops  to  gas
      phase  reactions  in  1;
(5)   Role of nucleation  (both cloud condensation nuclei
      (CCN),  and ice nuclei  (IN)  in changing  albedo of
      clouds;
(6)   Role of CCN  and  IN  in changing area covered by clouds
      or cloud  height.
                                            Primary
                                            meteorological
                                            effect
            Goseous, Nonoerosol
            Precursors
            CO,CO
            Gaseous Aerosol
            Precursors
            SO.,, H2S,NO, N

            HC, NH3,(H20)
            Low RH Aerosol
            RH<~0.7
High RH Aerosol
07
-------
Figure A-l.  The troposphere as a chemical system.  Rectan-
gles are recognizable entities in the atmosphere.  Triangles
represent processes that have a single direction of material
flow, and diamonds (two triangles) represent reversible pro-
cesses,  a, Sources;  b, sinks; c, gas-to-particle conversion;
d, sorption; e, deliquescence; f, efflorescence; g, Raoult's
equilibrium; h, reaction in concentrated solution droplet;
i, nucleation and condensation of water; j, evaporation; k,
capture of aerosol by cloud drops; 1, reaction in dilute
solution; m, rain; n, freezing of supercooled drop by ice
nucleus; q, precipitation.  (Butcher and Charlson 1972, p.8,
with permission of the authors and the copyright holder,
Academic Press; primary meteorological effects were added for
publication in Atmospheric Chemistry:  Problems and Scope,
National Academy of Sciences Panel on Atmospheric Chemistry,
Washington, DC, 1975.)
                              189

-------
         Report of a Workshop on
ANTHROPOGENIC CHEMICALS AS MODIFIERS OF
        CLOUDS AND PRECIPITATION
             Boulder, Colorado
              September 1979
                 Edited by
              Earl W. Barrett
                   191

-------
                   PARTICIPANTS AND CONTRIBUTORS
Dr. Earl W. Barrett
NOAA-ERL-APCL, R31
Boulder, CO  80303
(303) 497-6260
(FTS 320-6260)
Dr. Barry A. Bodhaine
NOAA-ERL-GMCC, R329
Boulder, CO  80303
(303) 497-6659
(FTS 320-6659)

Dr. Richard D. Cadle
National Center for
Atmospheric Research
Boulder, CO  80303
(303) 494-5151

Dr. William G. Finnegan
Dept. of Atmospheric Science
Colorado State University
Fort Collins, CO  80523
(303) 491-8667 or -8257

Dr. Norihiko Fukuta
Dept. of Meteorology
University of Utah
Salt Lake City, Utah  84108
(801) 581-6136

Dr. Edward E. Hindman
Dept. of Atmospheric Science
Colorado State University
Fort Collins, CO  80523
(303) 491-8511 or -8675
Dr. Wallace E. Howell
U.S. Bureau of Reclamation
P.O. Box 25007
Denver, CO  80225
(303) 234-3384
(FTS 234-3384)
Dr. John A. Kadlecek
Atmospheric Sciences Research
Center
State University of New York
at Albany
1400 Washington Ave.
Albany, NY  12222
(518) 457-4930

Dr. John M. Miller
NOAA Air Resources Labs.
8060 13th St.
Silver Spring, MD  20902
(301) 427-7645

Dr. Farn P. Parungo
NOAA-ERL-APCL, R31
Boulder, CO  80303
(303) 497-6460
(FTS 320-6460)

Dr. Russell C. Schnell
NOAA-ERL-APCL, R31
Boulder, CO  80303
(303) 497-6822
(FTS 320-6822)

Charles Van Valin
NOAA-ERL-APCL, R31
Boulder, CO  80303
(303) 497-6279
(FTS 320-6279)

Dr. Helmut K. Weickmann
CIRES,  PSRB #3
3100 Marine St.
Boulder, CO  80302
(303) 497-6289
(FTS 320-6289)
                               192

-------
                             CONTENTS

                                                            Page

PARTICIPANTS AND CONTRIBUTORS                                 192

PREFACE                                                       194

1.   HOW DO CLOUDS FORM AND PRODUCE PRECIPITATION?            195

2.   HOW CAN CHEMICALS AFFECT THESE PROCESSES?                198

3.   EFFECTIVENESS OF CHEMICALS AS NUCLEI                     201

4.   BACKGROUND NUCLEUS CONCENTRATIONS:  SOME QUANTITATIVE
      CONSIDERATIONS                                          204

5.   SENSITIVITY OF PRECIPITATION TO VARIATION  IN NUCLEUS
      CONCENTRATIONS                                          207

6.   CONCLUSIONS AND RECOMMENDATIONS                          216

7.   TECHNIQUES FOR TESTING NUCLEATION EFFICIENCY             218

8.   ACKNOWLEDGMENTS                                          224

9.   REFERENCES                                               224
                                193

-------
                              PREFACE
     This workshop was sponsored by the Office of Toxic Substances
of the Environmental Protection Agency and held at the Boulder
Environmental Research Laboratories of the Department of Commerce,
Boulder, Colorado on 24-25 September 1979.  The workshop panel
consisted of eleven full-time and three part-time participants
with expertise in atmospheric chemistry,  nucleation physics and
chemistry, aerosol physics and chemistry,  and weather modification
theory and practice.

     The charge given to the panel was (1) to evaluate the possi-
bility of significant environmental impact of chemical -substances
on precipitation patterns by virtue of their influence on the
cloud-microphysical processes that produce precipitation, and (2)
to recommend laboratory test procedures for screening chemicals
in case such impacts are identified.  Other environmental impacts
that could conceivably arise from the interactions of chemicals
with clouds, such as changes in pH of precipitation or changes in
the Earth's radiation balance by alteration of the reflectances
of clouds, were not treated by the panel.
                              194

-------
                     ANTHROPOGENIC CHEMICALS
            AS MODIFIERS OF CLOUDS AND PRECIPITATION


1.  HOW DO CLOUDS FORM AND PRODUCE PRECIPITATION?

     A brief discussion of the processes leading to precipitation
is helpful in understanding how anthropogenic chemicals might
influence these processes.  The reader who is interested in a
more complete treatment is referred to any of the standard texts
on cloud physics, such as Mason (1971), Byers (1965), or Fletcher
(1962).

     Clouds are formed by cooling of moist air.  Either, or both,
of two cooling mechanisms may be involved, viz., radiative cool-
ing or cooling by expansion when air ascends from high pressure
to low.  The first is effective in generating fogs near the
ground and stratiform clouds aloft; the second is dominant in
convective cumuliform clouds.  Generally speaking, clouds that
produce significant amounts of precipitation are formed by expan-
sion in updrafts.  Because the equilibrium, or saturation, vapor
pressure of water decreases strongly with temperature, the rela-
tive humidity of the cooling volumes of air increases as the
temperature falls, and eventually reaches saturation  (100%).

     Everyday experience might lead one to believe that water
should condense into drops when saturation is reached.  This,
however, is not necessarily the case in the free air  away from
solid surfaces.  In perfectly clean air containing no solid or
liquid particles, droplets will not begin to form until the
relative humidity rises to about 400%  (300% supersaturation).
At this point, condensation will take place on ions created by
cosmic rays or terrestrial radioactivity-  This is the basis of
a variety of "cloud chambers" that are used in the study of
nuclear particles.  In order for drops to form at relative hu-
midities near 100%, the air must contain small solid  or liquid
particles that can serve as cloud condensation nuclei (CCN).
Other factors being equal, the larger the airborne particle the
more effective it will be as a CCN.  (The effectiveness of a
particle as a CCN is expressed in terms of the relative humidity
—or supersaturation—that the air must reach in order to grow a
drop on that particle.)  On the other hand, the larger the parti-
cle the faster it will settle out of the atmosphere.  As  a re-
sult, CCN in the atmosphere consist of particles with diameters
in the range 0.01 micrometer (pm) to several (jm.  Particles
smaller than that require supersaturation of several  or even tens
of percent to work as nuclei and therefore are not considered to
be effective CCN.

     The number-vs.-size spectrum, or  size distribution,  of
droplets in a young cloud is therefore determined by  the  size
distribution of the available particles, the total concentration
(number per unit volume of air) of particles, and the physical
                                195

-------
and chemical properties of the particles that determine  their
nucleating effectiveness.  The theory of nucleation of cloud
droplets by pure soluble substances is well-developed; it  is
discussed in section 3.

     Once the droplets have condensed on a nucleus they  continue
to grow by condensation of vapor that diffuses to the drop.  It
is characteristic of diffusional growth that the initial rate of
increase of drop radius is rapid, but decreases as the drop gets
larger.  Under steady-state conditions the drop diameter is
proportional to the square root of elapsed time.  This means that
drops that are nucleated later in time tend to catch up  in size
with those nucleated earlier, thereby giving rise to a rather
narrow droplet-size distribution in the cloud.  The actual drop
growth rate is proportional to the supersaturation, which  in turn
tends to be proportional to the cooling rate and therefore (usu-
ally) to the updraft speed.  In many cases the growth rate is so
slow that the droplets cannot grow big enough to fall through the
updraft and are carried out of the cloud top into drier air where
they evaporate.  This is why many small cumulus clouds, or shal-
low stratus clouds, never produce precipitation.

     A typical droplet in a young cloud has a diameter of  order
10 |jm, whereas an average raindrop has a diameter of order 1 mm.
Assuming that the conditions in a cloud are such that a newly-
nucleated droplet grows to 5 urn diameter in the first second
after nucleation, the time required for the same drop to reach a
diameter of 1 mm by diffusional growth would be 11.1 h. This is
longer than the lifetime of most clouds and greater than the
depth of the clouds divided by the mean updraft speed.  Other
mechanisms must therefore exist to account for the much speedier
onset of precipitation.

     Looked at in another way,  the volume of an average raindrop
is roughly equal to that of 1 million average cloud droplets.  If
a cloud is to generate rain in a reasonable time, it is necessary
that this many cloud droplets aggregate to form the raindrop.
That is possible only if the distribution of droplet diameters is
broad enough so that the fall speed of larger droplets relative
to the updraft is significantly greater than that of the smaller
ones.  The larger droplets collide with smaller ones; in a sig-
nificant fraction of these collisions the smaller merge with the
larger.  The added mass increases the fall speed of the large
drop, thereby increasing the collision frequency, etc.  This
gives rise to an avalanche process that can grow raindrops
provided that the cloud is deep enough and contains sufficient
total liquid water and an adequate vapor supply.  Many clouds do
not meet the requirements and so do not precipitate at all.

     The mechanism of raindrop formation outlined above is known
as the collision-coalescence process,  or, more commonly,  the
coalescence process.  There is another mechanism that operates to
produce precipitation-size particles in a short time in clouds
                               196

-------
that extend to altitudes above the 0°C isotherm.  This process
depends on the existence of an energy barrier to the freezing of
pure liquid water that is analogous to the barrier that impedes
the condensation of pure water vapor unless a nucleus is avail-
able.  Water droplets cannot freeze unless they contain or come
into contact with a suitable solid body or unless the temperature
is below about -40°C.

     In middle and high latitudes, most clouds are deep enough
that their tops are colder than 0°C; cumulonimbus clouds that
produce severe thunderstorms and tornadoes have top temperatures
of -60°C or colder.  The middle and upper parts of young,  grow-
ing, middle- and high-latitude clouds therefore consist mainly of
small supercooled droplets.  These can freeze only if they con-
tain or come into contact with an effective ice nucleus (IN).
The effectiveness of an IN is expressed in terms of the tempera-
ture at which it causes a drop to freeze or nucleates an ice
crystal directly from the vapor phase; the warmer the nucleation
threshold the higher the effectiveness.  Natural IN have activa-
tion temperatures ranging from -1°C to -30°C or lower, but very
few natural airborne IN are active at temperatures warmer than
-10°C.

     In a supercooled cloud the prevailing humidity is close to
water saturation at the existing temperature.  Because the sat-
uration vapor pressure over water is always greater than that
over ice at temperatures below 0°C, a newly-frozen droplet or
embryonic ice crystal finds itself in a much more supersaturated
environment than that of the unfrozen droplets.  The diffusional
growth of the ice crystals will therefore be very rapid; even
with the (time)"5 dependence of the particle diameters the ice
crystals will grow to precipitation size in a minute or less.
The crystals grow at the expense of the unfrozen drops; the lat-
ter evaporate some of their water to feed the growing crystals if
the humidity drops below water saturation.  The crystals will
fall through the cloud and reach the ground as snow if the tem-
perature at the ground is colder than 0°C; otherwise they will
melt and fall as rain.  This mechanism of precipitation produc-
tion is known as the ice-crystal process, or the Bergeron-
Findeisen process (after the two meteorologists who first pro-
posed it) .

     The coalescence process operates alone only in clouds that
are (1) warmer than 0°C throughout, and (2) deep enough for the
process to be efficient.  These conditions are met only in trop-
ical clouds and in subtropical clouds in summer.  The process
does, however, play a role in initiating precipitation in summer
mid-latitude clouds in the United States (Battan 1953, Braham
1957, Changnon et al. 1977).  Likewise, the ice-crystal process
operates alone only in relatively thin clouds at temperatures
below 0°C.   In clouds with high concentrations of snow crystals,
aggregation of these into larger snowflakes by collision and
sticking acts to form larger snowflakes.  Also, crystals can
                                197

-------
sweep up supercooled droplets in their fall; these freeze on
contact with the crystal and increase its mass.  This accretion
or riming process leads to quick formation of large ice particles
and raindrops and,  in strong updrafts, to hailstones.

     A special case of the ice-crystal process is found fre-
quently in the warm-frontal portions of low-pressure storm sys-
tems, especially in temperate-maritime climates such as the
Pacific Northwest and western Europe.  In these systems, two (or
more) stratiform cloud decks separated by a layer of clear air
are frequently found.  The upper layer,  or "releaser" cloud is
cold, possibly near -40°C at the top, and hence is well-glaciated
but contains only a small mass concentration of water-substance
because of the low temperature.  The lower deck is warmer and so
contains much liquid water in the form of small supercooled
drops, but these would not normally freeze because of a lack of
IN active at the in-cloud temperature (-10°C to -15°C, for exam-
ple).  Ice crystals that fall from the upper deck enter the lower
"spender" cloud and serve as 100% effective IN for that cloud,
thereby releasing the water in it faster and ultimately in
greater amount than would have occurred in the absence of the
releaser cloud.  The same thing happens when convective (cumulus)
clouds grow inside a pre-existing stratiform cloud deck contain-
ing much ice; the cumuli are seeded much more effectively than if
they had developed in clear air (Hall 1957).

     From the foregoing it should be clear that: (1) only those
clouds that are sufficiently deep and have long enough lifetimes
(or strong enough updrafts) will produce precipitation; and (2)
that atmospheric particulate matter with diameters in the range
0.01 pm to a few pm is intimately involved in the mechanisms of
precipitation production.
2.  HOW CAN CHEMICALS AFFECT THESE PROCESSES?

     Chemicals introduced into the atmosphere can alter the rate
of precipitation in three basic ways.

     (1)  They may act as effective CCN or IN or both.
     (2)  They may "poison" nuclei that are already present in
          either of two ways:
          (a)  They may coat dry nuclei with a hydrophobic
               (water-repellent) or passivating layer, or
          (b)  They may dissolve in droplets and react with
               potential IN so as to reduce or destroy their
               effectiveness.
     (3)  They may dissolve in supercooled droplets endother-
          mically,  thereby chilling them so that they freeze
          spontaneously.  Such substances can be called "pseudo-
          IN."
                              198

-------
     It is now necessary to consider the possible effects of
addition (or removal) of nuclei on the clouds and on the precipi-
tation mechanisms.  In all cases the effects depend on the popu-
lation of nuclei present prior to introduction of an anthropo-
genic substance.  CCN, for example, vary widely in size distribu-
tion and chemical composition, depending on geography and prox-
imity to more or less steady natural or artificial sources.  CCN
counts in tropospheric air above the boundary layer range from
less than I/cm3 to 100/cm3 or so, with most of the particles
being found in the small size range with diameters less than 0.2
|jm.  In the boundary layer over the oceans, the number of smaller
particles is about the same but the number of large (diameters
from 0.2 to 2 pm) and giant (diameters greater than 2 pm) parti-
cles increases, especially when the sea is rough and whitecaps
are present (Byers 1965, ch. 5; Allee et al.  1976).  In urban
areas the concentrations and size spectra vary widely depending
on the automotive and industrial emissions.  Typical concentra-
tions are of the order of 1,000/cm3 in the St. Louis area (Auer
1975, Fitzgerald and Spyers-Duran 1973) and 2,000-3,000/cm3 in
Los Angeles as estimated from particle size-distribution measure-
ments (Barrett et al. 1979).  With the exception of the Los
Angeles data, the numbers are for CCN active at about 1% super-
saturation; the number active at typical cloud supersaturations
of the order of 0.1% is undoubtedly lower.

     Maritime clouds form on a CCN population that is lower in
number concentration but broader in size distribution than con-
tinental or urban clouds.  The maritime clouds therefore tend to
have fewer droplets per unit volume but a broader initial size
distribution.  They are characterized by higher coalescence
efficiency and would therefore start to precipitate earlier than
continental clouds, other factors being equal.  Addition of
anthropogenic CCN composed mainly of small particles to the
maritime boundary layer will render the clouds more continental
in character (more droplets with narrower size distribution) and
so delay the onset of rain insofar as this onset is determined by
the coalescence process.  Addition of anthropogenic large and
giant CCN, on the other hand, will cause clouds to become more
maritime in character and speed up the start of rain.  This has,
in fact, been observed in Los Angeles (Barrett et al. 1979).
Unpolluted stratocumulus clouds over the sea had droplet spectra
that were heavily skewed toward large sizes; the nuclei were
mostly sea-salt particles.  Clouds nucleated by the urban smog
(containing most small sulfate particles) had narrower spectra
and were dominated by small drops.  Clouds nucleated by a mixture
of smog and oil-refinery effluent containing both small sulfate
particles and large and giant nitrate and sulfate particles had
droplet size spectra that were intermediate between the other two
extremes.  All three sets of clouds had about the same liquid-
water content  (0.1 g/m3) and depth (-300 m).  Similar effects
have been reported by Hindman et al. (1976) and Braham and Dungey
(1976) for clouds polluted by urban and industrial aerosols.
                               199

-------
     Although these measurements have shown clearly that modula-
tion of the CCN concentration and particle size distribution does
affect the droplet size spectra and therefore the colloidal sta-
bilities of clouds, it is not at all clear if the amount of pre-
cipitation that falls from polluted clouds differs significantly
from the amount that falls from "natural" clouds.  Some observa-
tions (e.g., Warner 1968, Ramana Rao and Ramana Murty 1973) of
rainfall up- and downwind of artificial CCN sources (burning
sugar-cane fields and steel mills, respectively) seem to show
decreases of rainfall downwind.  On the other hand, the results
of the METROMEX project at St. Louis do not show any clear-cut
influence of CCN from urban pollution (Ackerman et al.  1978)
except that radar echoes (showing the presence of precipitation-
size drops) first appeared at levels between 600 and 1,100 m
closer to the cloud base in the polluted clouds.  Rainfall mea-
surements downwind (east) of St. Louis were found to be generally
higher than those upwind, but factors other than CCN,  such as
heat and moisture input from the city, are undoubtedly involved
in causing the downwind excess; it cannot be ascribed solely to
the added CCN.

     On the basis of the evidence to date, it may be concluded
that addition of artificial CCN to the atmosphere does not have
significant impact on precipitation amounts in the middle and
high latitudes, although strong, concentrated local CCN sources
might produce local anomalies.

     The clouds of major storm systems in middle and high lati-
tudes (which provide most of the annual precipitation totals) are
cold enough so that the ice-crystal process is dominant in pro-
ducing precipitation.  These clouds should be more susceptible to
modification by increasing or decreasing the IN concentration,
particularly IN that are effective at warmer temperatures.  If a
cloud or cloud system is deficient in IN, then introducing arti-
ficial IN will convert more of the small droplets into large ice
crystals and thereby speed up the formation of precipitation.
This is the basis of intentional precipitation-enhancement pro-
jects or experiments.  Unfortunately, the total amount of precip-
itation that ultimately falls out of the clouds is affected by a
large number of other physical factors and is not a simple
function of the IN concentration.  If, for example, a great
excess of IN is introduced into a cloud, a large number of ice
crystals will form and compete for the available water.  The
supersaturation with respect to ice will drop to a low value and
the crystals will remain small.  In this case, one would expect
the onset of precipitation to be delayed and the total amount to
be decreased.  It might be expected, then, that there should
exist an optimum IN concentration that would maximize precipita-
tion from a particular cloud system.  This optimum would, how-
ever,  vary from one cloud system to another, and cannot be pre-
dicted with any degree of confidence at the present state of
knowledge.
                              200

-------
     Real clouds are much more complex than the simple models
that are used to describe them.  As an example, an attempt was
made by Weickmann (1974) to delay the onset of snowfall from
stratocumulus clouds over the Great Lakes in order to reduce the
heavy snowfalls in shoreline cities such as Buffalo, New York,
and to deposit more of the snow farther inland.  The procedure
used was to "overseed" with artificial IN.  One expected result
was obtained:  the snow crystals in unseeded portions of the
cloud systems were large and heavily rimed with frozen cloud
droplets and fell rapidly.  On the other hand, the crystals in
the seeded volumes were initially small and unrimed, as pre-
dicted.  Unfortunately for the experiment, however, these small
crystals were present in such large numbers that they aggregated
very rapidly into large snowflakes that fell out just as soon as
those from the unseeded cloud volume.

     Chemicals that are hydrophobic may coat or be absorbed on
existing nuclei and lower their effectiveness or make them total-
ly inactive.  In the case of CCN, reducing their numbers should
give rise to clouds with greater percentage of large droplets and
a higher coalescence efficiency.  On the other hand, natural IN
are often in short supply, so poisoning of IN would tend to
inhibit precipitation by cold clouds.

     IN that are already present in cloud drops can also be
poisoned by chemicals that dissolve in the drops and react with
the IN.
3.  EFFECTIVENESS OF CHEMICALS AS NUCLEI

     In order to act as a nucleus of either type, a substance
must be dispersed into the atmosphere  as a solid or liquid
aerosol with particle diameters  in the approximate range  0.01 to
2 |jm, or as a gas that is subsequently converted to an  aerosol by
reaction with atmospheric constituents.  Any  substance  for which
this is not the case can be  eliminated as a possible precipita-
tion modifier.

     A substance must be hydrophilic to act as  an efficient
nucleus of either type.  Any hydrophobic substance can  be elim-
inated as_ a nucleus but may  act  cis a nucleus  poison.

     The theory of condensation  nucleation by soluble aerosols  of
a single chemical species is well-developed and can be  found in
the cloud-physics textbooks  referenced at the beginning of this
report.  The equation for the humidity required for equilibrium
between a solution droplet and the air may be written as


                ,  c  _ „      2a
                ln S  - mw
                               201

-------
where S is the saturation ratio (relative humidity expressed as
a fraction) at equilibrium,  m  the molecular weight of water, a
the surface tension of the droplet relative to air (Gibbs free
energy per unit area of drop surface), p the drop density, R the
universal gas constant, T the Kelvin temperature, r the drop
radius, n the number of ions formed when the solute is fully
dissociated,  the osmotic coefficient (function of chemical
species and concentration of solution),  and f the molality of the
solution.  It can be seen that S involves a balance between
opposing effects.  Lowering the radius while keeping everything
else constant increases the saturation (because of the effect of
radius of curvature on the free energy).  On the other hand,
increasing the concentration of ions lowers the equilibrium vapor
pressure in accordance with Raoult's Law.  The second term on the
right side of Eq. (3-1) can often exceed the first, so that
droplets of highly concentrated solution can form and be in
equilibrium at relative humidities less than 100%.  This is the
phenomenon known as deliquescence; it accounts for such familiar
facts as the stickiness of table salt in muggy weather (the
equilibrium relative humidity for a saturated NaCl solution at
25°C is 72%).

     The theory for insoluble particles, or mixed particles
consisting of soluble and insoluble material is more complex.
The case of spherical particles has been treated by Fletcher
(1958); he found that the critical saturation ratio was a func-
tion of particle radius and contact angle between a water drop
and the bulk material.  This latter is a quantitative measure of
the "wettability" of the substance; a contact angle of zero means
the water spreads out on the surface as thinly as possible, while
one of 180° means the drop remains spherical and does not wet the
surface at all.  The equilibrium saturation ratio decreases with
increasing particle radius and with decreasing contact angle, but
can never be less than unity as is the case for soluble CCN.

     On the basis of this brief and incomplete exposition of the
theory, it is possible to determine that effective CCN have these
characteristics:

     (1)  Easily wettable (especially if insoluble).
     (2)  Highly soluble in water.
     (3)  High degree of dissociation into ions in solution.
     (4)  Large number of ions per molecule.
     (5)  Highly hygroscopic (low deliquescence humidity).
     (6)  Large size.

     From the above, it is easy to see that highly polar sub-
stances such as acids, bases, and salts of low molecular weight
and high solubility are the most effective CCN, although insol-
uble particles that are highly wettable can be moderately effec-
tive .
                               202

-------
     Examples of good CCN include phosphorus pentoxide, sulfuric
acid, nitric acid, aluminum chloride, ammonium nitrate, ammonium
sulfate, and alkali chlorides, sulfates and nitrates.  Particles
with diameters greater than 1 or 2 (jm (so-called giant particles)
are most effective in generating large cloud droplets.  However,
the rate of loss of particles to the earth's surface by sedimen-
tation increases with particle diameter, so that the larger
particles have a shorter time to act as nuclei.

     The criteria for effective IN are less clear-cut.  In
principle one can apply the general statement of thermodynamic
equilibrium, namely that the Gibbs free energy must attain a
stationary value.  Unfortunately, some important parameters such
as the interfacial energy between a crystalline solid and liquid
water are hard to determine and are not at all constant over the
particle surface.  The situation is further complicated by the
fact that particles can nucleate ice crystals in three ways,
viz.,

     (1)  Deposition.  Water vapor molecules attach themselves to
          the particle to form a monolayer, after which succes-
          sive layers build up to form an ice crystal directly
          from the vapor phase.
     (2)  Contact freezing.  The particle impinges on a super-
          cooled droplet and initiates freezing, either at the
          point of contact or after entering the droplet.
     (3)  Condensation-freezing.  The particle acts first as  a
          CCN to grow a drop and then as an IN when the tempera-
          ture is sufficiently low.

     The weight of the evidence is that (2) and (3) are the most
important mechanisms, but that (1) may contribute significantly
at lower temperatures in cloud tops.

     It is apparent that the thermodynamic parameters for (I)
will differ from those of (2) and (3).  They will also be differ-
ent for crystals of different habits (axis ratios).  Some at-
tempts to predict this activation temperature as a function of
nucleus properties have been carried out (e.g., Fletcher 1958,
1970).  They result in highly complicated formulae that will not
be given here.  Some qualitative conclusions from the theory are
these:

     (1)  Wettability is a favorable factor for ice nucleation.
     (2)  Epitaxy (approximately equal spacing of atoms in the
          crystal lattice of the nucleus and in one plane of the
          ice crystal) is a favorable factor and is probably
          essential for deposition nucleation.
     (3)  Insolubility in water is generally favorable because of
          the freezing-point lowering effect of solutes.  How-
          ever, soluble substances with large negative heats of
          solution can chill droplets to the spontaneous freezing
          temperature.
                               203

-------
     (4)  Except for some biological IN, all effective substances
          are crystalline solids.  Glassy, amorphous substances
          can be eliminated as possible IN.  Some pure substances
          that are effective ice nuclei are given below.

          Inorganics:  Silver iodide, lead iodide, cupric sul-
          fide, cupric and cuprous oxides, aluminum oxide.

          Organics:  Metaldehyde, phloroglucinol, 1-5 dihydro-
          xynaphthalene, copper acetylacetonate.  Urea acts as a
          pseudo-IN by virtue of its high negative heat of solu-
          tion.

          Biological:  Some bacteria (Pseudomonas syringae,
          Erwina herbicola); unidentified constituents of decom-
          posing leaf tissue (Schnell 1976, Schnell and Vali
          1976, Vali et al. 1976).

     Natural IN consist mainly of crystalline soil minerals of
various kinds (e.g., kaolinite).  The biological IN mentioned
above are the only non-crystalline IN discovered to date.  Be-
cause the precise nature of the nucleation sites on these biolog-
ical particles has not yet been determined, the possibility that
they too are crystalline remains open.

     The apparent common factor for high ice-nucleation effi-
ciency of organic compounds is the presence of one or more polar
groups such as -COOH, -CHO, =CO, or -OH in the molecule.
4.   BACKGROUND NUCLEUS CONCENTRATIONS:  SOME QUANTITATIVE
     CONSIDERATIONS
                                                 >
     The environmental impact of introducing nuclei to the
atmosphere, or removing existing nuclei, will obviously depend in
some way (not necessarily simple) on the amount of increase or
decrease relative to preexisting background levels.  This section
is devoted to obtaining reasonable estimates of present-day
global background concentrations of tropospheric aerosols and
nuclei.

     Natural atmospheric aerosols arise from several sources:
wind-raised soil particles,  evaporated sea-spray, volcanoes,
forest fires, meteorites, and oxidation of gases such as S02 and
organic vapors emitted by living and decaying vegetation.  All of
these sources are subject to considerable variation with time and
meteorological conditions.  The sink strength also varies with
time, location,  and altitude.  The main mechanism for removal of
tropospheric aerosols is precipitation scavenging;  the particles
may serve as nuclei or may simply be swept out by the falling
rain or snow.  The mean residence time of a particle is a func-
tion of the annual precipitation total; aerosols dispersed in
                               204

-------
arid climates will persist longer than those in more moist re-
gions .

     As of 1971 the global annual direct production of atmos-
pheric aerosol was estimated to be in the range 425 to 1,100
million (M) metric tons  (t) per year; the additional source due
to gas-to-particle conversion was estimated as 345 to 1,100
Mt/yr.   The human contributions were estimated at 10 to 90 Mt/yr
for direct generation and at 175 to 325 Mt/yr by gas-to-particle
conversions (SMIC 1971, p. 189).  The range and mean for the
natural sources are 770-2200 and 1485 Mt/yr and those for the
anthropogenic are 185-425 and 300 Mt/yr.  So as of 1971 the total
mean aerosol production was 1785 Mt/yr of which about 20% was
anthropogenic.

     Estimates of tropospheric residence time have been made by,
among others, Weickmann and Pueschel (1973).  They arrived at a
range of 1 yr to 5.25 h and a mean of about 8.75 days (d).  In
view of the wide spread of these figures, the round figure of
10 d is assumed in this report.

     The theory of first-order processes leads to the simple
relation between mean loading Q (Mt), mean (1/e) residence time
t(yr),  and source strength S (Mt/yr):

                         Q = tS                        (4-1)

     With t = 10 d or 0.0274 yr and S = 1,785 Mt/yr one obtains
Q = 48.9 Mt as the mean global atmospheric aerosol burden.  Some
of this is located in the stratosphere where (because precipita-
tion scavenging does not occur) the residence time is several
years.   Except in years with strong volcanic eruptions,  nearly
all of the aerosol is in the troposphere.  Assuming a mean tropo-
pause height of 13 km and a mean Earth radius of 6,371 km, and
noting that the volume of a thin spherical shell of thickness Ar
and inner radius r is given by

                    AV - 4rtr2Ar                        (4-2)

one obtains a tropospheric volume V = 6.63 x 1018 m3.  Dividing
this into the burden Q = 4.89 x 1013g or 4.89 x 1019 pg, gives a
mean aerosol concentration of 7.4 |jg/m3 as a background level for
the troposphere.  Most of this is found in the boundary layer
(lowest 1 km or so of the troposphere) and over continents,
although Saharan and Indian dusts are frequently found throughout
the entire depth of the troposphere  (Bryson and Wendland  1970,
Allee et al. 1976).  Generally, the background loading above the
boundary layer and over the oceans is of order 0.1 (jg/m3; in
urban-industrial areas it is of order 100-300 ng/m3 and can ex-
ceed 1,000 |jg/m3 during some stagnation episodes.

     If all of the aerosol particles were active as IN or CCN,
there would be no reason for any concern about global impacts  of
                                205

-------
manufactured chemicals or precipitation because the combined re-
leases of all chemicals to the atmosphere would have to be at
least of the same order of magnitude as the existing source
strength of 1,785 Mt/yr to have any appreciable effect.  Measure-
ments show, however,  that only a small fraction of the total
aerosol population serves as nuclei.  Data taken within the
boundary layer over the North Atlantic and Mediterranean by
Hoppel (1979) show concentrations of CCN active at 0.16% super-
saturation in the range 100 to 3,000/cm3 with a (geometric) mean
near 700/cm3.  No IN measurements were made on this cruise.
Allee et al. (1976) made CCN and IN measurements aboard aircraft
over the North Atlantic above the boundary layer during the BOMEX
and GATE field programs and obtained CCN counts (at 1% supersatu-
ration) of 10 to 200/cm3 with a mean near 100/cm3.   Their IN
counts covered the range 1 to 9/£ (1,000 to 9,000/m3).  They also
made measurements in the boundary layer at an altitude of 600 m
over Florida and found CCN counts in the range 240 to 2,500/cm3
and IN counts of 103  to 104/m3.  Recent measurements in Colorado
cumuli (Heymsfield et al. 1979) gave low IN counts in the range
50 to 500/m3.

     For CCN, a representative mean concentration of 100/cm3
appears to be reasonable.  If one assumes a characteristic mass
of 10~15 g (corresponding to a spherical particle of diameter
0.1 pm and bulk density of 2 g/cm3) one obtains a mean mass
loading of 0.1 (jg/m3.  Thus apparently only about 1.35% of the
tropospheric particles act as CCN below 1% supersaturation.
Taking this fraction of the annual aerosol production rate gives
24.2 Mt/yr for the mean rate of generation of CCN;  multiplication
by the residence time of 10 d (0.0274 yr) yields a mean tropo-
spheric CCN burden of 663 kt (kilotonnes).

     It is less easy to arrive at a reliable mean concentration
of IN because measurements of IN in air and measurements of ice-
crystal concentrations in clouds show wide discrepancies.  The IN
measurements cited above suggest a mean value of order 5,000/m3.
On the other hand, ice-crystal counts in clouds are frequently
higher than this figure.  Koenig (1963) reports one case in which
the measured number of ice particles with diameters greater than
100 |jm was 2 x 105/m3 while at the same time the number of IN
active at temperatures warmer than -20°C was only 5,000/m3; the
number of ice particles was 40 times greater than the available
nucleus supply.  Mossop et al. (1968) measured ice particles in a
cloud whose coldest temperature was -4°C and found an approximate
mean count during a cloud traverse of 4 x 104/m3 when the cloud
reached maturity.  IN counts on air samples taken below cloud
base showed zero IN active at -4°C, 100/m3 active at -15°C, and
1,000/m3 active at -20°C.  In an effort to circumvent the pro-
blems involved in catching and sizing particles on board an air-
craft,  Koenig (1968)  sampled orographic clouds over the Califor-
nia Coast Range at a fixed mountain-top location and found con-
centrations of ice crystals in the range 3 x 104 to 4 x 105/m3 in
clouds with top temperatures around -8°C.  On the other hand, the
                               206

-------
observations of Heymsfield et al.  (1979) that were carried out
only in the updraft portions of  relatively young clouds yielded
ice-particle counts of the same  order  (50 to 500/m3) as the IN
counts.  However, the sample is  rather small to be statistically
significant.

     The large discrepancies cited above have been explained by
various investigators (Koenig 1963, 1968, Hallett and Mossop
1974, Hobbs 1969, Mossop 1976, Mossop  and Hallett 1974, Mossop
1978) as being due to ice-crystal multiplication processes of
various kinds.  These processes  involve the breaking off of ends
of crystals or of spicules formed when droplets freeze on contact
with other crystals or nuclei; the small fragments then grow into
new large crystals.  On the other hand, the techniques for in-
situ measurements of both IN and ice particles on board aircraft
are rather imperfect, so that the data may have substantial
errors.  It is also possible that ground-based IN counts may not
be representative of in-cloud conditions.

     Since this whole question will probably not be resolved for
some time, a figure of 105 IN/m3 effective at temperatures
warmer than -20°C will be assumed as an upper bound to the mean
tropospheric IN loading.  Assuming again a mean particle mass of
10~9 |jg, the IN mass loading works out to be 10~4 |jg/m3, or
(1.35 x 10~4)% of the total particulate loading.  This corre-
sponds to an annual IN production rate of 24.1 kt/yr and a mean
tropospheric burden of 663 t.
5.   SENSITIVITY OF PRECIPITATION TO VARIATION IN NUCLEUS
     CONCENTRATIONS

     The figures arrived at in the previous section can serve as
a crude measuring-stick for assessing the impact of adding new
nuclei to the atmosphere.  If anticipated annual releases of
chemicals to the atmosphere are smaller than the present source
strength, then there is no need to be concerned about significant
modification of global precipitation patterns.  On the other
hand, if the expected releases are larger than the current source
strength, then the possibility of significant weather modifica-
tion effects must be investigated.  The question immediately
arises: how great an increase in nucleus concentration is re-
quired to produce a significant change in global or regional
precipitation?  Unfortunately, the existing data are much too
sparse to provide a truly definitive answer to this key question.
The best that can be done is to make an educated guess based on a
limited number of field studies and well-designed intentional
weather-modification experiments.

     First of all, it must be stated categorically that there is
no sound evidence whatever that man's emissions of waste chemical
substances to the atmosphere in aerosol form (i.e., the 300 Mt/yr
                                207

-------
referred to in the previous section) have had any effect on
global precipitation totals.  If there have been any such ef-
fects, they are totally buried in the natural variability of
precipitation.

     Inasmuch as most of the anthropogenic aerosol is introduced
into the atmosphere from the approximately 2.5% of the Earth's
surface that is occupied by urban-industrial complexes in the
developed countries (Weickmann and Pueschel 1973), one should
look first for regional rather than global effects of man-made
aerosols on precipitation.  Several claims for such effects have
been made; there is, however,  only one fairly complete field
study of a metropolitan area that has been carried out up to this
time.  This is the METROMEX project, whose objective was to
determine as quantitatively as possible the total climatic
impact of the St. Louis metropolitan area.  Past climatic records
within the region and at greater distances were supplemented by
ground and airborne measurements of many variables,  including IN
and Aitken-particle counts.  CCN were apparently not measured
routinely,- but were said to comprise roughly 10% of the total
particles measured by the Aitken counters.  The Aitken counts in
the boundary layer at 600 m AGL upwind of the urban-industrial
area averaged around 16,000/cm3, increasing to about 70,000/cm3
downwind, a nearly fivefold increase.  It may be assumed that the
increase in CCN was at least of the same order.

     In the case of IN, counts of particles active at -20°C were
greater than 105/m3 in relatively small plumes associated with
industrial sources and the downtown area.  Upwind background
levels were not given in the project report (Ackerman et al.
1978, Part C); one can only assume that they were below 2.5 x
104/m3.  Thus the plumes contained at least 4 times the back-
ground level; the ratio is probably more like 10.

     The urban effect on summer rainfall for the years 1972-1975
was studied using data from a dense rain-gauge network.  A per-
sistent excess in the northeast and southeast quadrants (downwind
in most precipitation situations) relative to the western (up-
wind) region was noted.  Rainfall to the east control area
ranged from 90 to 100 cm during June through August; the range in
the west was 80 to 90 cm.  The urban effect was thus an approxi-
mate 10% increase downwind.  Studies of the distribution of the
increase show  that it is quite local; most of it is contained
within 40 km of the city center, and the effect is lost in the
noise at about 80 km to the east.  Much of the downwind excess
was found near concentrations of heavy industry; increases of
more than 100% above upwind levels were found in small local
areas (Changnon et al. 1977, Part B).

     This study has shown that a metropolitan area does tend to
increase precipitation locally, at least in summer.   The in-
crease,  however,  cannot be tied directly to the downwind in-
creases in nuclei,  except to say that in some cases (but not all)
                              208

-------
the local precipitation maxima lie near the spots with highest
Aitken counts.  It is likely, however, that the heat and water-
vapor outputs from the industrial sources have at least as much
to do with the augmented precipitation as do the excess nuclei.

     Analysis of radar measurements showed that 45% of these
summer cumuliform clouds had first echoes (corresponding to
regions of precipitation-size drops) that lay entirely below the
0°C isotherm  (Braham and Dungey 1976).  This indicates that the
coalescence process is active in initiation of precipitation in
summer cumuli.  One might therefore expect that clouds nucleated
by the plumes from industrial sources would be less apt to gener-
ate rain by coalescence and that the first echoes would appear at
higher elevations.  While this tended to be true of organized
cloud systems, isolated clouds tended to have lower first echoes
when they formed in the pollution plumes.  Thus the role of the
excess industrial nuclei in augmenting summer rain in the St.
Louis area is not at all clear even though an urban-industrial
effect on precipitation in a small downwind region is well-
established.

     Other observations suggest possible effects of steady aero-
sol emissions on regional precipitation.  Parungo et al. (1978)
made measurements of the size distributions, morphology, chemical
composition,  and nucleation activity of aerosols emitted from the
stacks of a large copper smelter west of Salt Lake City, Utah.
They found that most (^70%) of the particles in the plume con-
tained sulfur; much of this was in the form of H2S04 haze.  The
percentage of small sulfate particles tended to increase down-
wind, indicating that new H2S04 haze droplets were being formed
from the S02  in the plume.  Total Aitken counts (at 200% super-
saturation) ranged from 1.66 x 105/cm3 at 3.2 km downwind to 1.7
x 104 at 30 km.  CCN measurements were not made at the time
because of instrument malfunctions, but later laboratory tests on
collected aerosol samples indicated that about 50% of the parti-
cles were active as CCN.  Background  (upwind) Aitken counts were
of order 300/cm3.

     This smelter began operations in 1904, but emission inven-
tories of sulfur are available only since 1941.  Peak outputs  of
S were about  1,900 t/d in 1941 and have decreased to about 200
t/d in 1978 as a result of emission-control measures.   If one
assumes that  all of this S is converted to H2S04 and other sul-
fate salts, and that half of these are active as CCN at less  than
1% supersaturation, then the annual source strength for CCN  has
ranged from 1.06 Mt/yr in 1941 to 0.112 Mt/yr in 1978.

     Measurement of IN in the plume showed increases in concen-
tration with  distance downwind from the stacks.  This is  due to
poisoning of  the  IN by adsorbed oxides of sulfur and H2S04.
These poisons evaporate as the aerosol ages so that more  IN
become active as the plume drifts downwind.  This was confirmed
by exposing samples of the aerosol to vacuum evaporation  and to
                                209

-------
baking at 200°C.  The treated samples increased in -20°C nucleat-
ing activity by as much as 600% for particles collected near the
stack, but treating produced no increase in activity for parti-
cles collected 30 km downwind, showing that the poisoning had
completely disappeared by then.  IN concentrations at 30 km
downwind were found to be about 7 x 104/m3.   The ratio of IN to
total Aitken particles was approximately 1 particle in 2,100
active at -20°C after aging.

     From this data one can estimate the annual production rate
of IN.  Given the sulfur output and the approximation that this
constitutes 70% of the total aerosol by mass, the aproximate IN
source strength of this single source has decreased from 471 t/yr
in 1941 to 49 t/yr in 1978.

     Analysis of precipitation records for Utah show that there
was a rather abrupt increase in precipitation in 1904.  The mean
annual statewide precipitation for the period 1892-1904 was 27.43
± 3.81 cm.  (The spread is one standard deviation each side of
the mean.)  The corresponding figures for 1905-1946 were 36.0 ±
6.81 cm, an increase of 8.57 cm or 31.24%.  After the emissions
control program was instituted, the mean for 1945 to 1978 was
28.1 ± 5.1 cm.  When only the Colorado river basin region of the
state (the area most frequently downwind of the smelter) is
considered, the 1892-1904 figures are 19.48 ± 3.78 cm and those
for 1905-1949 are 30.51 ± 5.74 cm, an increase of 11 cm or 56%.
It thus appears possible that the very high emissions that
prevailed during the period 1904-1941 may have brought about a
non-trivial (and beneficial) increase in downwind precipitation
in this arid region.

     A follow-up study (unpublished) suggests that the regional
effect may not have been quite so large.  When precipitation
records for the neighboring state of Idaho were examined, it was
found that the mean annual precipitation for the state as a whole
for 1892-1904 was 44.5 ± 5.9 cm and that for 1905-1949 was 46.15
± 6.8 cm, an increase of 1.65 cm or 3.7%.  This is not statisti-
cally significant.  When, however, only the southeastern part of
the state, which is climatically similar to the Colorado River
Basin of Utah, is considered, the figures for the two periods are
29.7 ± 4.9 cm and 33.2 ± 6.8 cm, an increase of 3.5 cm or 11.8%
as an approximation of the "natural" difference of the two peri-
ods.  Furthermore, the jump in precipitation in 1904 appears to
have been a continental or even a global event.  So, while it is
possible that the smelter effluent may have caused a 45% increase
in precipitation during the years 1904-1949, the statistical
significance is not sufficiently high to induce much confidence
in this conclusion.  The drop back to approximately the pre-1904
levels of precipitation after the introduction of emissions
control gives a little more support to the possibility of an
effect,  and also indicates that present levels of emission are
not affecting precipitation significantly.
                               210

-------
.     Another source of data relevant to the question of human
influences on precipitation can be found in the data from various
intentional weather-modification experiments that have been
carried out all over the world in the last thirty-odd years.
These experiments have involved seeding of clouds with IN or CCN
for the purposes of precipitation enhancement or hail reduction.
The operational procedures have varied; in some cases individual
clouds were seeded internally from aircraft, and in others whole
cloud systems were seeded for extended periods from arrays of
ground-based IN generators.  Silver iodide (Agl) was the most
frequently used substance.  It has a nucleation threshold of -4°C
and, with the usual methods of dispersal, yields about 1014 to
1015 nucleating particles per gram of Agl, active at -20°C or
warmer.  For some of the experiments, the amounts of Agl emitted
are well-documented.

     In one of the longest and best conducted rain-enhancement
projects (Gagin and Neumann 1974), winter storm systems in Israel
were seeded with Agl.  Increases in rainfall of as much as 20.5%
in seeded areas relative to control areas were noted.   The seed-
ing rate was 800-900 g/h of Agl dispersed along a line by air-
craft.

     These emissions of nucleating agent were,  of course,  inter-
mittent; the generators were turned on only when suitable clouds
were present.  The equivalent steady source strength required to
maintain the same concentration of IN at all times would be of
order 7.5 t/yr emitted from a line source some 60 km in length.

     Studies of drop-size distributions and nucleus concentra-
tions in these Israeli cloud systems showed that these particular
clouds are particularly amenable to increasing rain by IN seed-
ing.  They have an abundance of small CCN and therefore cannot
readily generate rain by the coalescence process.  On the other
hand, their tops are not usually very cold, so that they are
lacking in naturally efficient IN.  This is not true of the great
majority of precipitation-producing clouds.  These facts should
be borne in mind when considering the apparent large increase in
rainfall in response to a relatively low source strength of IN.

     In project "Whitetop", carried out by the University of
Chicago in the southern part of Missouri (Braham 1966), seeding
rates of the order of 2,700 g/h, or 23.6 t/yr,  of Agl were used.
In this experiment, precipitation on seeded days was actually
less than on unseeded days.

     The Florida Area Cumulus Experiment (FACE), conducted by the
National Hurricane and Experimental Meteorology Laboratory of the
National Oceanic and Atmospheric Administration  (NOAA), was also
a randomized rain-enhancement project in which groups of develop-
ing cumuli were seeded from aircraft (Woodley et al. 1977).
Typical seeding ranges were 3,000-4,000 g/d of Agl, for an
                               211

-------
equivalent annual emission rate of 1.1-1.4 t/yr.  However, be-
cause the releases were localized inside the clouds, this  is
equivalent to a far higher diffuse source strength.  This  is  also
true for "Whitetop".  Precipitation data showed a positive seed-
ing effect of 70%, i.e., seeded days had 70% more precipitation
on the average than unseeded days.  However, the statistical
significance of the increase has been questioned (Nickerson
1979).  He showed that natural variability in Florida precipita-
tion could account for the observed seed/no-seed ratios.

     Several experiments on so-called orographic cloud systems
(clouds formed in air that is flowing over mountains) have been
carried out.  The advantage of working with these clouds is that
they are always in the same place and that one can sample them
with ground-based equipment.  A 17-yr program of seeding such
clouds over the California Coast Range in the Kings River water-
shed (Henderson 1968) used 25 Agl smoke generators distributed
within a 3,000 km2 area with outputs of 12 g/h each; the equiva-
lent source strength is 2.6 t/yr.  Stream runoff was used as a
measure of seeding effect; two neighboring unseeded watersheds
were used as controls.  A mean annual increase of 6% relative to
the controls was reported.

     The Climax project (Grant and Mielke 1967) was carried out
on clouds that formed over the Continental Divide in Colorado.
The seeding was done with six Agl generators positioned on the
ground at various distances between 13 and 59 km from the target
area; generator output was 20 g/h.  The corresponding steady
source strength was 1.05 t/yr.  Increases of up to 100% in the
target area were claimed when the temperature at the 500-mb
pressure level (-\-5.6 km ASL) was in the range -11° to -20°C.

     Winter storm systems over the Lake Almanor drainage area in
the northern Sierra Nevada mountains of California were seeded in
another experiment (Eberly and Robinson 1967).   Six generators
with outputs of 2.7 g/h were positioned within the 500 km2
watershed;  the equivalent annual emission rate was 1.4 t/yr.
Seeding effects of up to 80% increases were reported for storms
with westerly winds;  decreases were noted for southerly winds.

     In addition to experiments and operational programs for
increasing precipitation,  there have been several that had as
their purpose the reduction in hailfall,  or decreasing the size
of hailstones.   The rationale behind these experiments is that
heavy seeding of growing cumulonimbus clouds with IN will result
in a larger number of smaller ice particles and also a reduction
of available supercooled liquid water so that ice particles
cannot grow to  large sizes by riming.   In these operations, the
desired IN concentration is considerably higher than in the case
of "rainmaking".   In some countries,  notably the USSR,  the Agl is
delivered to the desired part of the cloud by rocket or artillery
shell.   In  other countries where the hazards of falling shrapnel
                               212

-------
from these carriers is considered too great, the shotgun ap-
proach, using large numbers of ground-based Agl aerosol genera-
tors, is used.  The largest operational program is in France.  It
has been in operation since 1952, with the object of protecting
some crops with high cash value, namely, wine-grapes and tobacco.
At present, the seeding network comprises 482 generators distrib-
uted over 7 x 104 km2 in the southwestern and central parts of
the country.  The emission rate is 13.6 g/h of Agl.  If all
generators were operating at once, the emission rate would be
6,554 g/h, or 28.7 t/yr.  However, the Meteorological Service
alerts only the part of the network where the probability of
large hail is high, so this figure is not achieved in practice.
Fortunately, the project keeps an Agl inventory, so that the
actual usage is known for each year.  For the last decade, the
maximum emission was 4.379 t in 1971; the minimum was 0.736 t in
1974 (Anonymous 1978).

     Because the only concern of the organization that conducts
the operation is reduction of hail damage to crops, the amount of
insurance claims filed is the only statistic used for verifica-
tion of the program's effectiveness.  However, if any major
effect on precipitation had occurred, it would have shown up in
the Meteorological Service records and would have received
comment in the literature.  No such comment has been published.
Even the effectiveness of the program in reducing hail has been
questioned (Boutin 1970) and reaffirmed by the chief scientist of
the project (Dessens 1974).

     Some attempts have been made in low latitudes to increase
precipitation from warm clouds by seeding with large and giant
CCN.  Common salt (NaCl) ground in a ball mill at high tempera-
ture and kept dry until dispersed, was the source of CCN.   One
such experiment, in the Rio Nazas catchment in Mexico (Fournier
d'Albe and Aleman 1976) involved seeding of clouds in a 2 x 104
km2 region with about 1 t/d of salt for an annual rate of 365
t/yr.  The effect on rainfall, a slight decrease on seeded days
relative to unseeded days, was statistically insignificant.  A
similar experiment on summer-monsoon clouds in India (Kapoor et
al. 1976) involved salt releases of up to 1,975 kg/d; duration of
seeding was up to 3 h 40 min.  Assuming that these two maxima
occurred together, the seeding rate was 4,694 t/yr.  Statistical
analysis gave a 16.5% increase relative to one control area in
1974 and 17.3% and 38.6% relative to two controls in 1973, but
neither result is statistically significant.

     The experiments reported above have shown that concentrated
IN sources with strengths of the order of a few tons per year can
affect local or regional precipitation totals at barely satis-
factory level of statistical significance, under favorable
circumstances.  On the other hand massive CCN sources do not
appear to have any detectable influence on precipitation.
                               213

-------
     In order to obtain an estimate of nucleus source strengths
that would be needed to produce detectable changes in precipita-
tion on a global scale, it is useful to calculate the emissions
used in the various cloud-seeding experiments on an areal basis.
This has been done by Weickmann (unpublished internal report
1968); the figures are given in Table 1.
  Table 1.  Source Strengths of Agl Emissions Per Unit Area
             for Various Cloud Seeding Projects
                                        Source strength
          Project                          (g/h/km2)
Snowy Mountain
Canadian rain-making
Whitetop
Swiss hail suppression
Bavarian hail suppression
French hail suppression
Argentine hail suppression
Various commercial seeding operations

Gunn et al .
ACN Chicago*
Australian rain-making
Kraus-Squires cumulus seeding*
Florida cumulus (J. Simpson)
Australian convective cloud seeding
0.4
0.5
1.5
0.07
0.25
0.35
0.1
0.01 to
0.11
25
150
10
300
800
0.15
*Dry-ice seeding actually used; conversion to Agl is based
 on laboratory comparisons of number of ice crystals formed
 by the two techniques.
This table shows that seeding rates,  expressed in this way,  have
been highly variable.  In projects where the seeding has been
done with ground-based generators distributed over an area,  the
figures are low, while in projects involving treatment of indi-
vidual clouds they are higher.

     The total global emission that would be required to give the
same elevation of IN concentration as was achieved in the various
projects is arrived at by multiplying the areal emission rates by
the surface area of the earth, 5.1 x 108 km2.  Considering only
the projects that used networks of ground generators distributed
over an area, one gets total source strengths in the range 5.1 x
107 to 1.78 x 108 g/h, or 0.45 to 1.56 Mt/yr.  If it is assumed
that releases occur only over land and that all influences on
                               214

-------
precipitation also occur only over land, then the figures are
reduced to about one-third of the above, or 150 to 520 kt/yr.
Comparing this with the upper bound of 21.4 kt/yr for current IN
production arrived at in Section 5 shows that in order to pro-
duce even minimal effects on continental precipitation,"The
annual IN production rate would have to be increased by_ at least
a factor of from 1_ to 24.  Furthermore, because 21.4 kt/yr is in
fact an upper bound to the current production rate, and because
the actual current rate could be as much as two orders of magni-
tude smaller than that figure, it is^ very likely that the actual
increase in IN source strength required to affect even minimally
the world-wide precipitation would be a factor of 1000 or more.
Put another way, in order to affect precipitation clouds signifi-
cantly, the IN concentration must be increased relative to the
existing levels by a factor of the order of 1,000.

     There is, of course, a great deal of uncertainty in the
foregoing analysis.  It is unfortunately true that, even after
34 years of cloud-seeding experience, we still have no firm
quantitative relationship between IN concentration and precipi-
tation.  The figures presented in this report are the best esti-
mates that can be reached on the basis of the very uncertain data
base available to us.

     An estimate of total continental CCN production can be
arrived at on the basis of Braham's (1966) figure of 2 to 4 x
101 CCN/cm2/s for the St. Louis area emission rate.  Taking a
mean particle mass of 10~15 g, multiplying by one-third of the
planetary surface area, and converting to an annual basis gives
an equivalent source strength of 2,144 Mt/yr.  However, as Braham
points out, there is a systematic increase in precipitation down-
wind of the city in spite of the fact that clouds over the city
do have narrower drop-size distributions and smaller mean drop
sizes than upwind clouds.  Clouds 10 to 40 km downwind have
already lost those characteristics and contain the usual number
of larger drops.  Braham suggests that this may be due to a
higher percentage of giant CCN in the urban aerosol relative to
the upwind aerosol.  On the other hand, Ochs and Semonin (1977),
using a mathematical model of cloud development containing
nucleation microphysics, found that in-cloud processes were
insensitive to the presence of giant CCN in concentrations found
at St. Louis.  It is therefore still uncertain whether or not the
CCN concentrations existing at St. Louis are influencing even the
local precipitation.

     The unsuccessful Rio Nazas experiment had a CCN  release rate
of 0.02 t/km2/yr.  Multiplying by the approximate planetary  land
area of 1.7 x 108 km2 gives 3.4 Mt/yr, much less than the St.
Louis emission rate.  On the other hand, most of these CCN were
in the giant size category and should have been more  effective  in
modifying the clouds.
                               215

-------
     Comparing the St. Louis emission figure with the estimated
existing global production rate of 24.2 Mt/yr shows that an
increase of CCN production by a factor of 89 above current back-
ground is_ insufficient to produce an unequivocal effect on warm-
cloud precipitation.Again, the uncertainty is great, but it
appears that CCN emission rates would have to be increased by a
factor of >100 to have any significant impact whatever.

     Unfortunately, no data exist for assessment of the nucleus-
poisoning hazard.  Parungo et al. (1978) did observe that the
concentration of IN in the plume of the Utah copper smelter
increased by a factor of slightly more than two as the aerosol
moved from near the stack to 35 km downwind, and that if the
aerosol samples collected near the stack were outgassed the IN
activity increased sixfold.  So for this particular type of
poisoning by oxides of sulfur and H2S04, the effect was only
temporary.- and natural outgassing soon restored the particles to
full activity.

     Since the only nucleus poisons that are likely to attach
themselves to aerosol particles in the atmosphere are gases or
condensible vapors, it follows that when the affected particles
move to lower concentrations of the gas or vapor the volatility
of the absorbed poison will eventually cause it to leave the
nuclei.  It therefore appears safe to conclude that no global
nucleus-poisoning problem will arise, and that local effects will
be restricted to the vicinity of concentrated sources of nucleus-
poisoning vapors or gases.
6.   CONCLUSIONS AND RECOMMENDATIONS

     The following conclusions can be drawn on the basis of the
information given in the previous sections.

     (1)  The potential impact of man-made chemicals on precipi-
          tation is by virtue of their droplet or ice nucleating
          ability or their action as poisons of nuclei.
     (2)  In order to function as nuclei,  chemicals must be
          dispersed into the atmosphere as aerosols with particle
          diameters in the range of 0.01 (jm to a few pm.
     (3)  Only substances that are dispersed as gases or conden-
          sible vapors are likely to act as nucleus poisons.
          This is because the probability of liquid droplets or
          particles colliding and coalescing with nuclei is much
          lower than the probability of adsorption of gases or
          vapors on particles.
     (4)  The chemical and physical properties that characterize
          effective CCN and IN are different.  They have been
          summarized in Section 3.
     (5)  Estimates of existing burdens and source strengths of
          nuclei (Section 4) are these:  for CCN, 663 kt and 24.2
                               216

-------
          Mt/yr; for IN, upper bounds of 663 t and 24.1 kt per
          year.
     (6)   Examination of the available data from field studies of
          inadvertent weather modification and cloud-seeding
          experiments and operations shows that massive injec-
          tions  of CCN of the order of 100 times the background
          levels have no statistically significant effects on
          precipitation.  Therefore, unless total releases of
          chemicals in solid or liquid aerosol form to the
          atmosphere in amounts greater than 2,400 Mt/yr are
          expected, there will be no need to screen chemicals
          for CCN activity.
     (7)   The data for IN are more uncertain;  an increase by a
          factor of from 7 to 24 over the upper bound to the
          background quoted in (5) above appears to be necessary
          to cause detectable increases in precipitation,  but the
          factor is probably much higher if statistically signif-
          icant  increases are to be produced.   It is therefore
          proposed (rather arbitrarily) that a criterion of about
          10 times the background production rate, or 250 kt/yr,
          be established as the limit above which screening for
          IN activity should be required.  This figure might have
          to be  lowered if many different substances active as IN
          eventually become dispersed at the same time, but it
          will suffice initially.

     In consideration of these points,  and the characteristics of
effective IN, listed in Section 3, the logic scheme, or flow
chart,  shown in  Figure 1 is recommended for screening chemicals
as potential precipitation modifiers.  Although most of this
diagram is self-explanatory, a few additional comments that are
too long to fit  in the boxes are appended here.

     (1)   In the top box, "released to the atmosphere" means
          under  normal conditions of use of the chemical.   Ac-
          cidental releases due to fire, explosion, etc.,  even if
          large, are unimportant from the weather-modification
          point  of view because any effects will be transient and
          will not affect long-term precipitation statistics.
          The short average residence time of tropospheric aero-
          sols and the absence of any recycling reactions such as
          those  involved in the ozone-Freon problem insure that
          no long-term effects will ensue from transient emis-
          sions  of active nuclei.
     (2)   The side-branch at the second box from the top provides
          for situations in which the primary substance is a gas
          or vapor that might undergo gas-to-particle conversion
          by reaction with atmospheric gases.  The  query "will it
          react  with atmospheric gases?" should have appended to
          it the qualification "at a high enough rate so that its
          mean lifetime as a gas is less than the 10-d residence
          time of tropospheric aerosols."
                               217

-------
     Most of the questions are very easy to answer and do not
require any special tests.  There is,  of course, the possibility
that solid aerosols might undergo reactions that converted them
to active IN.  Reactions between dry solids and gases, however,
are usually very slow except for such things as alkali metals,
etc., which are not likely to be dispersed as aerosols.  This
category of reactions may therefore be dismissed as an acceptably
tiny hazard.  It follows that very few substances will be dis-
persed as fine aerosols in quantities greater than 250 kt/yr.
Therefore very few, if any, products would require mandatory
testing for IN activity.

     An obvious exception to the last statement is, of course,
stack emissions from chemical factories or processing plants.  As
has been shown in Section 5, local or regional effects can be
(and are) produced by IN emitted as by-products of plants such as
ore smelters and steel mills.  It is not clear to the panel
members whether control of stack emissions is within the scope of
the Toxic Substances Control Act.  If it is, and if local and
regional weather modification is a matter of concern, then air-
borne in-situ measurements of IN in stack plumes, such as those
done by Parungo et al. (1978) in Utah, would be needed.

     The problem of nucleus poisoning does not appear to be a
serious one.  The gaseous poisons such as SO2 will evaporate off
the nuclei once the ambient partial pressure of the poison be-
comes low.  The only persistent poisons would be hazes or mists
of nonvolatile, oily, hydrophobic liquids that might collide with
and coat IN or CCN.  It is hard to conceive of any circumstance
in which such substances would be deliberately released to the
atmosphere, except possibly as stack effluents.  In the latter
case, the adverse effects on health and other nuisance effects
would demand prompt abatement measures.  Therefore the box
labeled "Test for nucleus poisoning" in Figure 1 should have
appended to it "if releases to the atmosphere will exceed the
250 kt/yr figure proposed as a criterion for solids."
7.   TECHNIQUES FOR TESTING NUCLEATION EFFICIENCY

     As explained in Section 3, ice nuclei in the air may nucle-
ate ice in clouds by acting as deposition, freezing, or contact
nuclei.  A deposition nucleus is one on which ice is deposited
directly from the vapor phase; a freezing nucleus acts by nucle-
ating a supercooled droplet in which it is embedded; and a con-
tact nucleus initiates the freezing of a supercooled droplet with
which it collides.  The threshold temperature of an ice nucleus
depends on the mechanism by which it nucleates ice as well as its
previous history.  Consequently, the concentration of ice nuclei
in the air that will be active in forming ice in a cloud is an
extremely difficult quantity to measure.  All the many techniques
developed to measure concentrations of ice nuclei active at a
                               218

-------

Will substance
be released to
the
atmosphere?
1


As an c
with p
diamel
ran
0.01 to

i
Y
r
lerosol
article
ge
2 //.m?
> Y
> I
r
In quantity
greater than
250 metric
kilotons
per year?
i
Y
r
Is the
material
easily
wettable?
^
Y
r
Is it crystalline
at
atmospheric
temperatures?
^
Y
r
Is it
insoluble
or nearly so
in water?

N

ls '* a 8as Does it react To yield solid
N ^ at Y with Y aerosol N ^
r atmospheric " ^ atmospheric reaction
temperature? gases? products?
N N Y
i r
Test for ice
l\l nucleus r~\
~| v poisoning v ^-^
(see text). Go to point (A)
for each such product.
• ^11 ^ i
N ^
N ^

Is heat of
N solution in ™ ^ ^
water large and
negative?
Y . . . Y

w " •"
Test for Testing
ice nucleus
unnecessary.
Figure 1.  Ice nucleus screening flow chart.
                     219

-------
given temperature process the  ice nuclei  in particular manners.
In general, therefore,  it is not to be  expected  that  the  differ-
ent techniques will give the same number  for the concentrations
of ice nuclei in a given sample of air, or that  any one of  them
will give the number of ice nuclei effective in  a natural cloud.
Probably the best that  can be  expected  at the present time  is
that a particular device might measure  the relative concentra-
tions of ice nuclei in  the air that are effective at  different
temperatures.  These reservations should  be borne in  mind in the
following discussion.

     Hobbs  (1974, p. 631) has  described approaches to the study
of ice nuclei:

          Ice nuclei which exist as free  particles in the air
     have been studied  by three general methods.  The first
     involves isolating a quantity of air, cooling it below the
     dew-point, and noting the freezing points of the condensa-
     tion products.  In the second method, a known volume of air
     is cooled until a  cloud is produced  and the number of  ice
     crystals which form at a  particular  temperature  is deter-
     mined.  In expansion chambers the cooling is produced by
     compressing the air and then allowing it to expand rapidly.
     In mixing chambers the cooling is by refrigeration.  Several
     techniques are available  for counting the number of ice
     particles which appear in the cloud.  For example, the cloud
     may be illuminated by a light beam and the  concentration of
     ice crystals estimated visually.  Alternatively,  the small
     ice crystals which form in the chamber may be allowed to
     fall into a dish or film  of supercooled water (Cwilong
     1947), supercooled soap solution (Schaefer  1948), or super-
     cooled sugar solution (Bigg 1957), where they can be detec-
     ted and counted by the larger ice crystals they produce.
     Another technique  for counting ice crystals, known as the
     acoustic particle  counter, has been  described by Langer
     (1965) and has been used  in an automatic ice nucleus counter
     (Langer et al. 1967).   In this instrument the ice particles
     which form in a cold box  pass through a capillary tube where
     they produce audible clicks.   These  clicks are counted
     electronically and the number of ice nuclei in a given
     volume of air active at the temperature of the cold box is
     recorded.  The third type of method which has been used for
     counting ice nuclei in the air is to draw a known volume of
     air through a membrane filter (Bigg  et al. 1963;  Stevenson
     1968).  The membrane filter retains  those particles in the
     air with dimensions in excess of about one-tenth of the
     diameter of the pores  of  the filter.   The number of ice
     nuclei on a filter is  determined by holding the filter at a
     known temperature,  exposing it to a given supersaturation,
     and counting the number of ice crystals which grow on it.

     A more recent technique for measuring ice nuclei has
been developed by Schnell (1979).   Aerosols are captured on
                               220

-------
hydrophobia membrane filters, and the ice nuclei in/on the filter
are detected by setting an array of nucleant-deficient water
drops on the filter and then cooling the filter.  The presence of
nuclei on the filters is detected by the freezing of the drops.

     Today there are probably 75 unique instruments available for
measuring ice nuclei, each with their proponents and detractors.
No instrument measures all types of ice nucleation, nor do any of
them measure in all of the three general methods.  Only the Mee
ice nucleus counter is sold commercially, and in 10 years of
sales, only 10 have been purchased.  The rest of the ice nucleus
counters are in experimental laboratories or are used in some
field programs to monitor relative changes in ice nucleus concen-
tration.  In the latter case, cloud chambers of the NCAR/E.
Bollay type are used most frequently (Langer 1965).  These count-
ers are no longer being manufactured.

     Several international workshops have been held with the
express purpose of comparing ice nucleation measurement methods
and to find a common ground for measuring ice nucleation modes.
These goals have not as yet been attained (Grant 1971, Vali
1976).

     The following list presents some key information on sources
of ice nucleation measurement techniques and results, as well as
comparisons between instruments and techniques that are not
listed, discussed, or compared in the two volumes from the
International Workshops (Grant 1971, Vali 1976) or are not refer-
enced above.

Cloud chamber comparisons

       Isothermal Cloud Chamber              (Garvey et al. 1976)
       Dynamic Cloud Chamber
       Diffusion Chamber                (Schaller and Fukuta 1979)

       Diffusion Chamber                 (Langer and Rodgers 1975)

       Diffusion Chamber                     (Jiusto et al. 1976)

Expansion chamber comparisons

       Mee Counter                        (Langer and Garvey 1980)
       CSU Isothermal
       NCAR Counter
       Mee Counter                           (Hindman et al. 1980)

Drop freezing techniques

       Chemical effects on ice nuclei              (Reischel  1973)

       Chemical effects on ice nuclei      (Schnell and Vali  1974)
                               221

-------
     In view of the rather chaotic state of the art, it is not
possible to establish a "standard" instrument or procedure for
nucleation measurement at the present time; the recommendations
that follow are tentative.

     The least expensive and easiest-to-implement procedure for
IN testing is the method of Schnell (1979).  The aerosol to be
tested is collected on a membrane filter treated with a hydro-
phobic coating.  The filter is then partially covered with small
(millimeter-sized) distilled and deionized water drops and
chilled slowly by a thermoelectric cooler equipped with a therm-
istor or thermojunction for monitoring the filter temperature.
The number of drops that freeze in a given temperature interval
is recorded by hand.  A filter that has not been exposed to the
aerosol is processed in the same way to serve as a blank and thus
eliminate the effect of the substrate and any impurities in the
water.

     The cloud-chamber techniques eliminate any possible sub-
strate effect and create a closer approximation to a real cloud
environment, but are more expensive and require trained opera-
tors.  One instrument of this type is available commercially from
Mee Industries, Inc., 1629 S. Del Mar Ave, San Gabriel, CA
91776.  It is similar to the NCAR/Langer chamber, but differs
from it in the technique used to count the ice crystals automati-
cally.  The crystals fall between two crossed polarizing filters
that are placed between a light source and a photodetector and
produce light pulses by virtue of the depolarization associated
with crystals.

     A very recent design of thermal-diffusion cloud chamber is
the "wedge" chamber of Schaller and Fukuta (1979).  It has the
advantages of being relatively easy to build and being able to
detect the mode of nucleation (deposition, condensation-freezing,
or contact), as well as the temperature of activation.  It is the
instrument recommended at present for exhaustive testing for ice-
nucleating potential.  Unfortunately, it is not in commercial
production and the information given in the referenced paper is
not complete and detailed enough to serve as a construction
manual.  Further information about this device may be obtained
from Professor Norihiko Fukuta,  Department of Meteorology,
University of Utah, Salt Lake City, Utah 84108.

     For CCN measurement, the thermal-diffusion cloud chamber is
the only practical technique at present.  Many varieties exist
and are described in the two workshop reports  (Grant 1971, Vali
1976).  They differ in dimensions, methods of achieving thermal
and vapor stratification, and in the methods for counting the
droplets.  Mee Industries produces one such instrument (Model
140) with optical drop counting, continuous analog voltage out-
put, and continuous sample flow.  Another commercial version of a
chamber originally designed by Radke (Grant 1971, pp. 41-42) is
manufactured by Meteorology Research, Inc., Box 637, 464 W.
Woodbury Road, Altadena, CA 91001.


                               222

-------
     In order to reproduce  conditions  in  natural  clouds, the
supersaturation with respect  to water  must be  low,  less than
0.25% or so.  Many cloud chambers  do not  meet  this  requirement;
they operate at around  1%.  Fukuta and Saxena  (1979) have de-
signed a chamber for CCN counting  in which a well-defined spatial
distribution of supersaturation is maintained  while sample air
flows continuously through  the system.  An optical  scanner se-
quentially examines regions of different  supersaturation to give
a spectrum of the number of nuclei as  a function  of supersatura-
tion.  Again, this is the preferred instrument for comprehensive
testing, but it is not  commercially available; inquiries should
be directed to Prof. Fukuta.

     Techniques for generating test aerosols are  described in the
nucleation workshop reports (Grant 1971,  Vali  1976).  Except for
materials that decompose irreversibly  below their boiling or
sublimation temperatures, thermal  evaporation  is  the preferred
method.  A measured amount of bulk material is made into a paste
with pure water and coated on a platinum  or Nichrome wire in a
clean glass vessel.  A  clean, dry  carrier gas  is passed through
the vessel.  The wire is quickly heated electrically above the
vaporization point of the substance under test.  The material
condenses to a smoke that is  transported  by the carrier gas into
a storage volume, e.g., a Mylar bag contained within a rigid
vessel.  The aerosol may be allowed to age in  the bag so as to
coagulate the aerosol into larger  particles if desired.

     When a test is run, the  Mylar bag is squeezed by pressuriz-
ing the outer vessel to transfer the aerosol to the cloud chamber
or filter.  The carrier gas may be air or nitrogen; it must be
rendered oil- and water-free  by passage through silica gel and
then cleaned of all particles by passage  through an absolute
filter.  Provision for monitoring  particle size of the generated
aerosol is highly desirable when thermal-diffusion cloud chambers
are used.  This is because particles that act  as contact IN must
exist in the smaller size range (diameters from about 0.01 |jm to
0.1 urn) so that Brownian motion can be effective  in bringing them
into contact with supercooled droplets in a reasonably short
time.  On the other hand, CCN are  more effective when they are
larger than 0.1 pm because no collision process is involved in
drop nucleation from the vapor.

     For IN, the desired information is the number of nuclei per
gram of substance effective in forming ice crystals in various
class intervals of temperature from 0°C to -25°C.   It would also
be desirable to establish the relative effectiveness in each of
the nucleation modes, but this is  not  essential in routine test-
ing.  For CCN, the desired data are the numbers of droplets
produced per gram of substance at  water supersaturations between
0.1% and 1%.

     To test for IN poisoning, a sample of a known good IN such
as Agl is processed first and the  number-vs.-temperature curve
                               223

-------
recorded.  Then another sample is run with the same aerosol mixed
with the vapor of the suspected poison, and the decrease,  if  any,
in the numbers of ice crystals formed at the various temperatures
is noted.
8.  ACKNOWLEDGMENTS

     The panel chairman thanks Drs. Helmut K. Weickmann and
Russell C. Schnell for assistance in collecting information on
emission rates in weather-modification experiments and on the
techniques for nucleus measurement.
9.  REFERENCES

Ackerman B, Changnon SA, Jr., Dzurisin G. et al.  1978.  Summary
of METROMEX, Volume 2:  Causes of Precipitation Anomalies.
Illinois State Water Survey, Urbana, IL.

Allee PA, Sax RI,  Delgado E. 1976.  Concentrations of Aitken,
cloud droplet condensation, and ice nuclei over selected oceanic
and continental sites.  In:  Preprints, International Conference
on Cloud Physics,  July 26-30, 1976, Boulder, Colorado.  American
Meteorological Society, Boston, MA, pp. 1-8.

Anonymous. 1978.  Programme de prevention centre la grele.
Association Nationale d'Etude et de Lutte Centre les Fleaux
Atmospheriques 27:2-20.

Auer AH, Jr. 1975.  The production of cloud and Aitken nuclei by
the St. Louis metropolitan area (Project METROMEX).  J. Rech.
Atmos. 9:11-22.

Barrett EW, Parungo FP, Pueschel RF. 1979.  Cloud modification by
urban pollution: a physical demonstration.  Meteorol. Rundsch.
32:136-149.

Battan LJ. 1953.  Observations on the formation and spread of
precipitation in convective clouds.  J. Meteorol. 10:311-324.

Bigg EK. 1957.   A new technique for counting ice-forming nuclei
in aerosols.  Tellus 9:394-400.

Bigg EK, Mossop SC,  Meade RT, Thorndyke NSC. 1963.  The measure-
ment of ice nuclei concentration by means of millipore filters.
J. Appl. Meteorol. 2:266-269.
                               224

-------
Boutin C. 1970.  Controls statistigue et controle physique des
operations de prevention de la grele.  These 3  Cycle, Faculte
des Sciences de Paris.

Braham RR, Jr. 1957.  In:  Cloud and Weather Modification: A
Group of Field Experiments.  Meteorol. Monogr. 2(11):67.

	•  1966.  Project Whitetop - A Convective Cloud Randomized
Seeding Project.  Part I - Design of the Experiment.  Part II -
Summary of Operations.  University of Illinois, Urbana, IL.

Braham RR, Jr., Dungey M. 1976.  Effects of a large city upon
convective clouds and coalescence rain.  In:  Preprints, Int.
Conf. on Cloud Physics, July 26-30, 1976, Boulder, CO. 'American
Meteorological Society, Boston, MA, pp. 275-278.

Bryson RA, Wendland WM. 1970.  Climatic effects of atmospheric
pollution.  In:  Global Effects of Environmental Pollution.
Singer SF, ed.  Springer-Verlag, New York, NY, p.  130.

Byers HR. 1965.  Elements of Cloud Physics.  University of
Chicago Press, Chicago, IL.

Changnon SA, Jr., Huff FA, Schickedanz PT, Vogel JL. 1977.
Summary of METROMEX, Volume 1:  Weather Anomalies  and Impacts.
Illinois State Water Survey, Urbana, IL.

Cwilong BM. 1947.  Sublimation in a Wilson chamber.  Proc. R.
Soc. London, Ser. A 190:137-143.

Dessens J. 1974.  La lutte contre la grele.  Commentaires sur une
phrase: "Rien ne permet d'affirmer que les interventions aient eu
un effet. ..."  Association Nationale de Lutte Contre les
Fleaux Atmospheriques 22:43-44.

Eberly DZ, Robinson LH. 1967.  Design and evaluation of random-
ized winter time cloud seeding at high evaluation.  Proc. Fifth
Berkeley Symp. on Mathematical Statistics and Probability, Vol.
V, Weather Modification Experiments.  University of California,
Berkeley, CA,  pp. 65-90.

Fitzgerald JW, Spyers-Duran PA. 1973.  Changes in cloud nucleus
concentration and cloud droplet size distribution associated with
pollution from St. Louis.  J. Appl. Meteorol. 12:511-516.

Fletcher NH. 1958.  Size effect in heterogeneous nucleation.  J.
Chem. Phys.  29:572;  1959. Erratum. 31:1136.

	.  1962.  The Physics of Rainclouds.  Cambridge University
Press,  Cambridge, England.
                                225

-------
Fletcher, NH. 1970.  The Chemical Physics of Ice.  Cambridge
University Press, Cambridge, England.

Fournier d'Albe EM, Mosino Aleman P- 1976.  A large-scale cloud
seeding experiment in the Rio Nazas catchment area, Mexico.  In:
Papers Presented at the Second WMO Scientific Conference on
Weather Modification.  (WMO Publication No. 443).  Secretariat of
the World Meteorological Organization, Geneva, Switzerland,
pp. 143-150.

Fukuta N, Saxena VK. 1979.  A horizontal thermal gradient cloud
condensation nucleus spectrometer.  J. Appl.  Meteorol.  18(10):
1352-1362.

Gagin A, Neumann J. 1974.  Rain stimulation and cloud physics in
Israel.  In:  Weather and Climate Modification.  Hess WN, ed.,
Wiley, New York, NY, pp. 454-494.

Garvey DM, Lilie LE, Grove TC, Grant LO. 1976.  Determination of
the rates of ice crystal formation in two large cloud chambers.
Preprints, Int. Conf. on Cloud Physics, Boulder, Colorado,  July
26-30, 1976.  American Meteorological Society, Boston MA,
pp. 121-125.

Grant LO, ed. 1971.  The Second International Workshop on Con-
densation and Ice Nuclei.  Colorado State University, Department
of Atmospheric Science, Fort Collins, CO.

Grant LO, Mielke PW, Jr. 1967.  A randomized cloud seeding exper-
iment at Climax, Colorado 1960-1965.  Proc. Fifth Berkeley Symp.
on Mathematical Statistics and Probability, Vol. V, Weather
Modification Experiments.  University of California,  Berkeley,
CA, pp. 115-131.

Hall F. 1957.  In:  Cloud and Weather Modification: A Group of
Field Experiments.  Meteorol. Monogr. 2(11):33.

Hallet J, Mossop SC. 1974.  Production of secondary ice crystals
during the riming process.  Nature 249:26-28.

Henderson TJ. 1968.  Cloud seeding on the Kings River watershed,
final report on the 1966-1967 operations with review of results
for water years 1955-1967.  Report submitted to Kings River
Conservation District,  Fresno, CA, by Atmospherics, Inc.

Heymsfield AJ,  Knight CA, Dye JE. 1979.  Ice initiation in un-
mixed updraft cores in northeast Colorado cumulus congestus
clouds.  J.  Atmos. Sci. 36:2216-2229.

Hindman EE,  II, Tag PM, Silverman BA, Hobbs PV. 1976.  Calcula-
tions of the effects on rainfall caused by cloud condensation
nuclei from a paper mill.  In:  Papers Presented at the Second
WMO Scientific Conference on Weather Modification (WMO Publica-
tion No. 443).   Secretariat of the World Meteorological Organiza-
tion,  Geneva, Switzerland, pp. 21-24.

                               226

-------
Hindman EH,  Odenkrantz  FK,  Finnegan WG.  1980.   Cloud nuclei from
launches  of  solid  rocket  boosters.   J.  Weather Modif.  12 (in
press).

Hobbs PV. 1969.  Ice multiplication in  clouds.   J.  Atmos.  Sci.
26:315-318.

Hobbs PV. 1974.  Ice Physics.   Clarendon Press,  Oxford.

Hoppel WA. 1979.   Measurement  of  the size distribution  and  CCN
supersaturation spectrum  of submicron aerosols  over the  ocean.
J. Atmos. Sci. 36:2006-2015.

Jiusto JE, Lala GG, Zamurs  J.  1976.   Filter measurement  of  ice
nuclei concentrations and selected  comparisons.  Preprints,  Int.
Conf. On  Cloud Physics, Boulder,  Colorado, July  26-30, 1976.
American  Meteorological Society,  Boston,  MA, pp. 53-56.

Kapoor RK, Krishna K, Chatterjee  RN,  Ramana Murty AS, Sharma SK,
Ramana Murty BV. 1976.  An  operational rain stimulation  experi-
ment using warm technique over  Rihand catchment  in  northeast
India during summer monsoons of 1973  and 1974.   In:  Papers pre-
sented at the Second WMO  Scientific  Conference of Weather Modifi-
cation, (WMO Publication  No. 443).   Secretariat  of  the World
Meteorological Organization, Geneva,  Switzerland, pp. 15-20.

Koenig, LR. 1963.  The  glaciating behavior of small cumulonimbus
clouds.   J. Atmos. Sci. 20:29-47.

	. 1968.  Some observations suggesting ice multiplication
in the atmosphere.  J.  Atmos.  Sci.  25:460-463.

Langer G. 1965.  An acoustic particle counter; preliminary
results.  J. Colloid Sci. 20:602-609.

Langer G, Garvey DM. 1980.   Intercomparison of Mee  and NCAR ice
nucleus counters and the  Colorado State  University  Isothermal
Chamber.  J. Weather Modif.  12  (in press).

Langer G, Rodgers  J. 1975.  An  experimental study of the detec-
tion of ice nuclei on membrane  filters and other substrata.  J.
Appl. Meteorol. 14:560-570.

Langer G, Rosinski J, Edwards  CP. 1967.   A continuous ice nucleus
counter and its application to  tracking  in the atmosphere.  J.
Appl. Meteorol. 6:114-125.

Mason BJ. 1971.  The Physics of Clouds.   Clarendon  Press, Oxford.

Mossop SC. 1976.   Production of secondary ice particles  during
the growth of graupel by  riming.  Q.  J.  R- Meteorol. Soc. 102:45-
47.
                               227

-------
        1978.  The  influence  of  drop  size  distribution  on the
production  of  secondary  ice particles  during  the  growth of
graupel by  riming.  Q. J. R. Meteorol.  Soc. 104:323-330.

Mossop SC., Ruskin RE, Heffernan KJ. 1968.  Glaciation  of  a
cumulus at  approximately -4C.  J. Atmos.  Sci.  25:889-899.

Mossop SC,  Hallet J.  1974.  Ice crystal concentration in cumulus
clouds: influence of  the drop spectrum.   Science  186:632-634.

Nickerson EC.  1979.   FACE rainfall results: seeding effect or
natural variability?  J. Appl. Meteorol.  18:1097-1105.

Ochs HT, III,  Semonin RG. 1977.  The sensitivity  of cloud  micro-
physics to  an  urban environment.  Preprints,  6th  Conf.  on  Inad-
vertent and Planned Weather Modification.  American Meteoro-
logical Society, Boston, MA, pp. 41-44.

Parungo FP, Pueschel  RF, Ackerman E, Proulx H, Wellman  D.  1978.
Characteristics and meteorological impacts of pollutants from
the Kennecott  Copper  Smelter.  NOAA Tech  Rept. ERL 395-APCL 42.

Ramana Rao  T,  Ramana  Murty BV. 1973.  Effect of steel mills on
rainfall at distantly located stations.   Indian J. Meteorol.
Geophys. 24:15-26.

Reischel MT. 1973.  Influences of dissolved salts on the effec-
tiveness of freezing  nuclei.  Report AR 105, Dept. of Atmos.
Sci., University of Wyoming, Laramie, WY.

Schaefer VJ. 1948.  The  detection of ice  crystals in the free
atmosphere.  Proj. Cirrus Occasional Report No. 9, General
Electric Company, Schenectady,  NY.

Schaller RC, Fukuta N. 1979.  Ice nucleation by aerosol parti-
cles: experimental studies using a wedge-shaped ice thermal dif-
fusion chamber.  J. Atmos. Sci. 36:1788-1802.

Schnell RC. 1976.  Naturally occurring biological ice nucleants:
a review.   In:  Preprints, Int. Conf. on  Cloud Physics,  July 26-
30, 1976, Boulder CO.  American Meteorological Society,  Boston, MA,
pp. 57-60.

	. 1979.  A new technique for measuring atmospheric ice
nuclei active at temperatures from -20°C  to approaching 0°C,
with results.  Proc.   Seventh Conf. on Inadvertent Weather Modi-
fication, Banff, Alberta, Oct.  8-12.  American Meteorological
Society,  Boston, MA,  pp.  110-111.

Schnell RC,  Vali G.  1974.  Heterogeneous  ice nucleation in the
presence of aliphatic amines and ammonia.  Report AR 113, Dept.
of Atmos.  Sci., Univ.  of Wyoming,  Laramie, WY.
                               228

-------
         .  1976.   Biogenic ice nuclei.  Part I:  terrestrial  and
   marine  sources.   J.  Atmos. Sci.  33:1554-1564.

   SMIC  1971.   Study of Man's Impact on Climate.  Wilson CL, ed. ,
   MIT Press,  Cambridge,  MA.

   Stevenson,  CM.  1968.  An improve millipore filter technique for
   measuring  the  concentrations of freezing nuclei in the atmosphere
   Q. J. R. Meteorol.  Soc.  94:35-43.

   Vali  G,  ed.  1976.   The Third Int. Workshop on Ice Nucleus Mea-
   surements.   Univ.  of Wyoming, Laramie, WY.

   Vali  G,  Christensen M, Fresh RW, Galyan EL, Maki LR,  Schnell RC.
   1976.   Biogenic  ice nuclei.   Part II:  bacterial sources.  J.
   Atmos.  Sci.  33:1565-1570.

   Warner  J.  1968.   Reduction in rainfall associated with smoke from
   sugar-cane  fires.   J.  Appl.  Meteorol. 7:247-251.

   Weickman HK. 1974.   The  mitigation of Great Lakes storms.  In:
   Weather and Climate Modification, Hess WN, ed.  Wiley,  New York,
   NY, pp.  318-354.

   Weickman HK, Pueschel  RF. 1973.   Atmospheric aerosols:   residence
   time, retainment factor, and climatic effects.  Contrib.  Atmos.
   Phys. 46:112.

   Woodley WL,  Simpson J, Biondini  R,  Berkeley J. 1977.   Rainfall
   results, 1970-1975:   Florida Area Cumulus Experiment.   Science
   195:735-742.
U.S. GOVERNMENT PRINTING OFFICE: 1981— 778-709/228 Region No. 8
                                   229

-------
50277-101
 REPORT DOCUMENTATION  ^..REPORT NO.
       PAGE          j  EPA  560/5-80-001
                                                                3. Recipient's Accession No.
                                                                5. Report Date

                                                                  12/80  (Date of  Issue
4. THIe »nd Subtitle
 THE POTENTIAL  ATMOSPHERIC IMPACT OF CHEMICALS
 RELEASED TO THE  ENVIRONMENT.
 Proceedings of Four Workshops	
7. Authorts)                                                          8- Performing Organization Rept. No.

 JOHN M.  MILLER (editor)  NAT' 1.  OCEANIC  &  ATMOSPHERIC ADMiiN.
 9. Performing Organization Name and Address
 AIR  RESOURCES  LABORATORIES
 NAT'L OCEANIC  & ATMOSPHERIC  ADMINISTRATION
 SILVER SPRING,  MD   20910
                                                                10. Proiect/Task/Work Unit No.
                                                                11. Contract(C) or Grant(G) No.

                                                                (C)
                                                                (G;
                                                                 'V.AG 98120AOTE
 12. Sponsoring Organization Name and Address

 ENVIRONMENTAL PROTECTION AGENCY
 401  M STREET, SW
 WASHINGTON,  DC  20460
                                                                13. Type of Report & Period Covered

                                                                WORKSHOP REPORT
                                                                14.
 IS. Supplementary Notes
 16. Abstract (Limit: ZOO words)
 Four  workshops are  reported:   toxic  substances in atmospheric deposition,
 screening chemicals  for inadvertent  modification of  the stratosphere,
 the  impact  of chemicals on  the radiative transfer imbalance,  and  the
 impact of anthropogenic chemicals on precipitation processes.  These
 workshops were convened as  part of an effort  to assess the  impact  of toxic
 chemicals on  the abiotic environment (specifically,  the atmosphere), to
 assess the  feasibility of screening  chemicals for these impacts,  and
 whenever possible to develop  a screening logic.  Report contains  an  extensive
            survey and bibliography.
 17. Document Analysis «. Descriptors
   O. Identifiers/Open-Ended Terms
 Toxic Substances,  screening  for effects, atmospheric deposition,  stratosph
 ozone depletion, radiative transfer  imbalance, precipitation modification
 atmospheric  modification.                                                       '
                                                                                    eri
   t. COSAT1 Field/Group
 18. Availability Statement
 Release Unlimited
                                                   19. Security Class (This Report)
                                                    Unclassified
                                                   20. Security Class (This Page)
                                                                          21. No. of Pages

                                                                          	240
                                                                          22. Price
(See ANSI—£39.181
                                     See Instructions on Reverse
                                                                         OPTIONAL FORM 272 (4-77)
                                                                         (Formerly NTIS-35)

-------