PB80-142706
 Atmospheric  Distributions, Sources and Sinks  of Selected Halocarbons
 Hydrocarbons, SF6, and N20                                       *
SRI International,  Menlo Park,  CA
Prepared  for


Environmental Sciences Research Lab, Research Triangle Park,  NC





November  1979
                  U.S. DEPARTMENT OF COMMERCE
                National Technical Information Service
                                 NTIS

-------
United States
Environmental Protection
Agency
 Environmental Sciences Research
 Laboratory
; " . Research Triangle Park NC 27711
, EPA-600 3-79-107
 November 1979
Research and Development
   stributions/
Sources,  and
SI n k s of $ elected
Halbca'rbons,
Hydrocarbons, SF<

-------
 V                RESEARCH REPORTING SERIES             ,

 .  Research reports of the Office'of Research and Development, U.S. Environmental
   Protection Agency, have been grouped into nine series. These nine broad cate-
«,,gories were established to/facilitate further development and application of-en-
   vironmental technology.  Elimination of traditional grouping  was  consciously.
   planned to foster technology transfer and a maximum interface in related fields.
   The nine series are:

:,,,        1.  Environmental  Health Effects Research.
         2.  Environmental  Protection Technology
         3.  Ecological Research
         4.  Environmental  Monitoring
         5,  Socioeconomic Environmental Studies    -.                   «-
         6.  Scientific and Technical Assessment Reports (STAR)
         7.  Interagency Energy-Environment Research and Development   .    .
         8.  "Special" Reports
 ~ '      9r; Miscellaneous Reports       •          '•          - .       .-.L

   This report has been assigned to the ECOLOGICAL RESEARCH series. This series
   describes research on the effects of pollution on humans, plant and animal spe-
   cies, and materials. Problems are assessed for their long- and short-term influ-
   ences. Investigations include formation, transport, and pathway studies to deter-
   mine the fate of pollutants and their effects. This work provides the technical basis
   for setting standards to minimize undesirable changes in living organisms in the
   aauatic. terrestrial, and atmosoheric environments.
   This document is available to the public through the National Technical Informa-
   tion Service, Springfield, Virginia 22161.          *          v        t

-------
                NOTICE





THIS DOCUMENT HAS BEEN REPRODUCED



FROM THE BEST  COPY  FURNISHED US BY



THE  SPONSORING AGENCY.  ALTHOUGH IT



IS RECOGNIZED THAT CERTAIN PORTIONS



ARE  ILLEGIBLE,  IT IS  BEING  RELEASED



IN THE  INTEREST  OF MAKING  AVAILABLE



AS MUCH  INFORMATION AS POSSIBLE.

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA-600/3-79-107
                             2.
                                                          3. RECI
4. TITLE AND SUBTITLE
 ATMOSPHERIC DISTRIBUTIONS, SOURCES, AND SINKS OF
 SELECTED HALOCARBONS, HYDROCARBONS, SF& AND N.,0
             5". REPORT DATE
               November 1979
             6. PERFORMING ORGANIZATION CODE
 . AUTHOR(S)
  H. Singh,  L.J.  Salas, H. Shigeishi,  A.H.  Smith,
  E. Scribner,  and L. A. Cavanagh
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 SRI  International
 333  Ravenswood Avenue
 MenIo:Park,  California  94025
             10. PROGRAM ELEMENT NO.   ;„
              1AA603A AF-006  (FY-78)
             11. CONTRACT/GO ANT NO.
                                                              8038020
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental Sciences Research Laboratory - RTF, NC
 Office -of Research and Development             v
 U.S.  Environmental Protection  Agency
 Research Triangle Park, North  Carolina  27711	
             13. TYPE OF REPORT AND PERIOD COVERED
                Final	•	• '
             14. SPONSORING AGENCY CODE
                  EPA/600/09
16. SUPPLEMENTARY NOTES
16. ABSTRACT
 burdens  of CC1-F
 small  and that o
                                        is
     Global distributions of a large number of  halocarbons, hydrocarbons, SF, and
N20 are presented.   These data are complemented with,measurements in the polluted
environments to  establish urban-nonurban relationships and daily doses of these
species.  Measurements conducted over a 3-year  period  show that the atmospheric
                    CCl-F, and CH3CC13 are  increasing rapidly.  The growth of CC1,
                    ^O is essentially undetectable.   Long residence times ofVCCUF, ...,£
and CC1J? rule out  the existence of any major removal  processes in the troposphere.
The gloBal distribution of CC12F2 and CC1JF is  used to calculate a fast inter-
hemispheric exchange rate of 1.2 years.  CC1, was shown to have an important oceanic
sink,Awhile both CH3C1 and N20 have a major source in  the ocean.  CHLCC1. atmospheric
data was used to demonstrate that its long residence time makes it a potential
depletor of stratospheric ozone.  The hydroxyl  radical concentration in the tropo-
sphere was shown to be significantly lower (= 4 x 10  molec./cm3)'than previously
believed and probably asymmetric in the two hemispheres..  Measurements show'that
the troposphere  contains slightly under 3  ppb of organic Cl and about 77 percent of ;
this is of man-made origin.  The organic Br and I abundances are much lower and
much less certain.   The troposphere contains about  1 ppb of organic F and all of it
appears to be man-made.                              „       ,*•        ;        .       ^
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                          c.  COSATI Field/Croup
 *  Air  pollution
 *  Halohydrocarbons
 *  Hydrocarbons
 *  Chemical analysis™
 *  Troposphere
 *  Stratosphere
                             13B
                             07C
                             07D
                             04A
18. DISTRIBUTION STATEMENT

 RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
  UNCLASSIFIED
                                                                       21. NO. OF PAGES
                                              20. SECURITY CLASS (Thispage)
                                               UNCLASSIFIED
                          22. PRICE
EPA Form 2220-1 (»-7J)

-------
                                               EPA-600/3-79-107
                                               November 1979
   ATMOSPHERIC DISTRIBUTIONS, SOURCES AND SINKS
OF SELECTED HALOCARBONS, HYDROCARBONS, SFg , AND N.O
                        by
                   H. B.  Singh
                   L. J.  Salas
                   H. Shigeishi
                   A. J.  Smith
                   E. Scribner
                  L.  A.  Cavanagh
            Atmospheric  Science Center
                 SRI International
           Menlo Park, California  94025
                Grant Number 8038020
                   Project  Officer

                     John Spence
     Atmospheric  Chemistry  and  Physics  Division
     Environmental  Sciences Research Laboratory
        U.S.  Environmental  Protection Agency
   Research Triangle Park, North Carolina  27711
     ENVIRONMENTAL  SCIENCES  RESEARCH  LABORATORY
         OFFICE  OF  RESEARCH  AND DEVELOPMENT
        U.S.  ENVIRONMENTAL PROTECTION AGENCY
   RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711

-------
                                  DISCLAIMER
#..     This report has been reviewed by the Environmental Sciences Research
 Laboratory, U.S. Environmental Protection Agency and approved for,.publication.
 Approval does not signify that the contents necessarily reflect the views and
 policies of the U.S. Environmental Protection: Agency:, nor does mention of
ttradc names or commercial products constitute endorsement or recommendation
 for' use.
                                       ii

-------
                                  ABSTRACT

     Global distributions of a large number of halocarbons, hydrocarbons, SFg,
and N20 are presented.  These data are complemented with measurements in the
polluted environments to establish urban-nonurban relationships of these
species.  Atmospheric growth rates of selected species are characterized based
on measurements conducted between 1975 and 1978.
     Results indicate that inert man-made species are relatively uniformly
distributed in the global atmosphere, while the more reactive species show sig-
nificant interhemispheric gradients.  Atmospheric residence times of 65 to
70 years for fluorocarbon-12 and 40 to 45 years for fluorocarbon-11 best fit
the observational data.  These residence times rule out the possibility of any
major removal mechanisms in the troposphere.  Between 1975 and 1978
fluorocarbon-12 and fluorocarbon-11 concentrations in the troposphere increased
at an average rate of 22 ppt/year and 14 ppt/year, respectively.  The global
distribution of fluorocarbon-12 and fluorocarbon-11 was used to establish a
relatively fast average interhemispheric exchange rate of 1.2 years.  Unlike
fluorocarbons, carbon tetrachloride appears to have an oceanic sink that was
found to be about half as effective as the stratospheric sink.  A major source
of methyl chloride, sufficient to account for nearly all the atmospheric methyl
chloride, was identified as the ocean.  Atmospheric measurements of methyl
chloroform support an 8- to 11-year residence time.  These long residence times
would allow a significant amount of methyl chloroform to enter the stratosphere,
thus making it a potentially important depletor of stratospheric ozone.  A long
methyl chloroform residence time suggests serious disagreement with global
models and points to reduced average hydroxyl radical (HO) concentrations of
only 3 x 10^ to 4 x 10^ molecules/cm^.  In addition, methyl chloroform global
distribution supports southern hemispheric HO values that are a factor of 1.5
or more larger than northern hemispheric values.  These HO values were used
to estimate the residence times of a large number of molecules.

                                     iii

-------
     Measurements show that the troposphere contains slightly under  3  ppb
 >-f  .     '-   .  .   ^;..- •••••-   • "      .--.        '•••••, -.  -•••     ' ">••      %""
of  organic Cl  and about 75 percent, of this is of man-made origin.  The organic
Br  and I contents are much lower and much less certain.  The troposphere con-
tains about  1  ppb of  organic F and all of it appears to be man-made.
  -*^;.       •.*         • •                 " %                                  *
     The Pacific Ocean (45°N - 45°S) was found to be a significant source of
^0 despite  the  high  oxygen content in the South Pacific.  This oceanic flux,
                   •f         ~            '      • '             -            '•'•
when extrapolated to  global waters, suggests an oceanic N20 source of  20' to
30 million tons  (N20)/year (13 to 19 million tons (N)/year).  The atmospheric
residence time of ^0 was found to be very long (20 years < TN'   < 120 years).
  .,„ h! "'••                         •                           ,..*•"  '.;.-       t*;. '•"' f>.
During this1 study, no detectable.increase in atmospheric N20 was observed.
Measurements in  urban-suburban locations were used to characterize daily
doses of the measured chemicals, many of which are suspected to be highly.,
 *JS..         '        '* -       I?'""       • • '.'   ' '••'•-   .        -i'di-,       »••.«.'- .,'      .n.^itS.if™  „
toxic.
     This report was  submitted in fulfillment of Grant No. 8038020 by SRI
International  under the sponsorship of the U.S.. Environmental Protection-
Agency.  This  report  covers a period from 7/75 to 6/79, and work was com-
pleted as of 6/79.
                                       iv

-------
                                  CONTENTS

ABSTRACT	   iii
FIGURES	 .  viii
TABLES	     x
 1   INTRODUCTION	 	 . .     1
 2   OVERALL OBJECTIVES .  	     2
 3   SUMMARY OF MAJOR FINDINGS AND CONCLUSIONS  	     3
 4   ANALYTICAL METHODOLOGY .... 	 .  	     8
     A.  Trace Constituents of Interest 	     8
     B.  Experimental Procedures  	     8
         1.  Air Analysis	     8
         2.  Water Analysis 	 ........    10
         3.  Calibrations		    18
 5   EXPERIMENTAL PLAN	 .  ; . .    24
     A.  Air Sampling with the Instrumental Mobile Laboratory 	    24
     B.  Global Sampling	    28
         1.  Air Sampling .......................    28
         2.  Water Sampling		    33
 6   DISTRIBUTION, SOURCES, AND SINKS OF NITROUS OXIDE	    35
     A.  Atmospheric N20 Measurements	    35
     B.  Pacific Seawater Measurements  . . 	    38
     C.  Results and Discussion	    43
 7   FLUOROCARBONS-12 AND -11 IN THE GLOBAL ATMOSPHERE  	    47
     A.  Global Emissions	    47
     B.  Global Burden and Distribution 	    47
     C.  Atmospheric Growth	    50
     D.  Sources and Sinks	<	    52
         1.  Atmospheric Residence Times  	    52

-------
         2.  Calculation of a Mean  Interhemispheric  Exchange
  . ,£. ,       ,.Rate . . ..,		 .  .  .  ..'.•-.  .  .  .-..,:-	54
         3.  Oceanic Sink	     55
     E.   Discussion of Results   .....  	  ...     55
 8   FLUOROCARBONS-22, -113, -114,  -21 AND SF6  IN  THE GLOBAL
   "ATMOSPHERE	.................'    59
     A.   Global Emissions  	     59
     B.   Global Burden and Distribution  	     59
  •-.»C.'   Discussion of Results	  .  .  .  .'    61
 9   ATMOSPHERIC CARBON TETRACHLORIDE	  .     65
     A.   Global Emissions	  ... ,  .......	     65
     ?B.   Global Burden"and Distribution.  .  .  .  .  .  .* 7 .  .  ...  ...  T*    68
     C.   Atmospheric Growth	"..-.'	,•.,;•	.::.^    69
     D.   Sources and Sinks	  .     69
    *;'   1.  Atmospheric Residence  Time;.-.	1   >  69
     E.   Discussion of Results   .  .	 ,.-, .     73
10   ATMOSPHERIC METHYL CHLOROFORM	,'•[.	     74
   :  ^A.   Global Emissions	v  .  .  .  .  .  .  .'    74
     B.   Global Burden and Distribution  .  .  .	     76
     C.   Atmospheric Growth 	     77
   ft-D.   Sources and Sinks of Methyl Chloroform .  .  . .  .......  .     78
         1.  Atmospheric Residence  Time  of Methyl  Chloroform  .....     78
         2.  Methyl Chloroform as an Indicator  of  the Hydroxyl
               Radical Abundance in the  Northern and Southern
   V*-.'        ^Hemispheres	Z    82
     E.   Discussion of Results   .	     84
11   ATMOSPHERIC METHYL CHLORIDE  ...  	     85
    ,IA.   Global' Emissions	...;.........'    85
     B.   Global Burden and Distribution  ......  	     86
     C.   The Oceanic Source	     87
   4D:   Residence Time 
-------
     B.  Atmospheric Measurements 	     93
         1.  Chloroform	     93
         2.  Methylene Chloride	     93
         3.  Trichloroethylene and Tetrachloroethylene  	  ...     93
         4.  Hexachloroethane 	     94
         5.  Phosgene	     94
     C.  Discussion of Results:  Sources and Sinks  	     94
13   ATMOSPHERIC BROMINE AND IODINE SPECIES 	     97
14   GLOBAL BUDGETS OF ORGANIC CHLORINE, BROMINE, IODINE,  AND FLUORINE.    100
15   ATMOSPHERIC HYDROCARBONS, CO, NOX, PAN, AND 03	    104
     A.  Hydrocarbons	    104
     B.  Carbon Monoxide	    107
     C.  NOX, PAN, and 03	    109
16   BEST ESTIMATES OF THE MEAN RESIDENCE TIME OF SELECTED
     MOLECULES BASED ON THE PRESENT STUDY 	    112
17   URBAN-NONURBAN RELATIONSHIPS OF MEASURED TRACE CONSTITUENTS  ...    114
     A.  Chlorofluorocarbons and SF5	    115
     B.  Halogenated (Nonfluorinated) Species	    115
     C.  N20, Hydrocarbons, CO, NOX, and Ozone  . . .	    116
18   RECOMMENDATIONS FOR FUTURE RESEARCH	  .    123
19   GENERAL OBSERVATIONS	    125
LIST OF PUBLICATIONS	    126
REFERENCES	    128
                                     vii

-------
                                      FIGURES                 .   •—-




•Number           '                                                           Page
1

2
3
« •$$? £1
* 5
6

7
- . 8
•1?''
- 9
10
11

12
•S, J,
'•"13
14
15
16
17
,^18
ft *
19

20
21
22
j^3'
^24
25

Fluorocarbon-11, Cti^CCly, CCl^, and CC14 along with
Other Halocarbons in the Ambient Air 	 	 	
Fluorocarbon-12 and ^0 in the Ambient Air 	 	
Flurocarbons-12 , -114, -21, -11; C^Cl, and O^Br Separation^ . .
Ethylene Dibromide and C2C16 Separation from' Ambient Air ... .
Phosgene Separation from the Ambient Air 	 	 	
PAN and PPN Separation from the Ambient Air Using
a Coulometric EC-GC 	 .... 	
Permeation-Tube Holder 	
Demonstration of Linear Response of Frequency-Modulated
EC Detector! ; 	 / :. ....... /' '". 	 	
Location of Monitoring Sites on the Continental United States . .
Map Showing the Sampling Locations for the Four Experiments . . .
Specially Treated 1-liter Stainless-Steel Sampling Vessel
for Collecting Grab Samples 	 	 	
One-liter Glass Sampling Vessel 	 	 ."•= . .
-**-'. " • ;,VW, '.,..."• ' »«" '••••"• •'•
Deepwater" Sampling in the Pacific Ocean 	 . . .
Global Distribution of N20 	 	 	 	
Atmospheric Growth of ^0 with Time 	
N20 Supersaturation in the Pacific Ocean (46°N to 40°S) .....
Distribution of N20 in the Pacific 	 	 	 . . . .
Global Distribution of Fluorocarbons-12 and -11 	 	 "." . .

Atmospheric Growth and Residence Times
of Fluorocarbons-12 and -11 	
Global Distribution of F-113, F-114, F-21, and SFg .......
Production and Emission Trends for CC14 	 	
Global Distribution of Carbon Tetrachloride 	 . .T . -.
Atmospheric Growth of CC14 in the Northern Hemisphere . . .". ; .
Cumulative CC14 and F-ll Emissions 	 	 	
Identical CC13F and CC14 Behavior as an Indicator
of a Common Urban Source 	 	 	

12
13
•• 14
15.
•'•'• 16

17
19

19 '•
26
29

30
31

34
36
37 !
41
44
50


51
62
66
68
69
71

72
                                       viii

-------
26   Global Distribution of Methyl Chloroform  ....	 .       77
27   Atmospheric//Growth of Methyl Chloroform	       78
28   Global Distribution of Methyl Chloride	       86
29   Global Distribution of Methane	      104
30   Global Distribution of Ethane 	      105
31   Ethane Diurnal Variation at Point Arena (Site 11) 	      106
32   Ethane Diurnal Variation at Jetmore (Site 12)	      107
33   Carbon Monoxide Diurnal Variation at Point Arena (Site 11).  . .      108
34   Carbon Monoxide Diurnal Variation at Jetmore  (Site 12)   ....      108
35   PAN Diurnal Variation at Jetmore (Site 12)   	      110
36   PAN Diurnal Variation at Point Arena (Site 14)   	      110
                                       ix

-------
                                      TABLES
Number                                                                      Page
   1    Average Background Concentrations of Important Atmospheric
   j-    Trace Constituents .........  .  .  ...........  .  .       4
   2    Average Atmospheric Growth Rates of  Selected Species
       from 1975 to 1978 in the Northern Hemisphere  ..........       5
   3    Tropospheric Chlorine,  Bromine , Iodine ,  and Fluorine
       Organic Budgets ............  .............       6
   4    Best Estimate of the Seasonally Averaged Atmospheric
       Residence Time of Selected Molecules Based on Results From
       This Study  ........ ...................       7
   5    Chemicals Measured in the Clean Troposphere and the Pacific
 • "*"    Ocean ...... . .......... " '"'.  "....-. •; .,-....".       9
   6    Identification and Separation of Selected Trace Constituents  .  .      11
   7    Permeation Rate Data for Halocarbon  Primary Standards ......      21
   8    Analysis and Intercomparison of an Identical Air Sample
       by SRI and WSU  .....................  ....      23
   9    Environmental Mobile Laboratory Instrumentation .........      25
  10    Measurement Sites Over the Continental United States  ......      27
  11    Ambient N20 Concentrations (ppb) in  1975 and 1977 .......  .   .   38
  12  , N£0 Concentration in Pacific Surface Water  ..........  .      40
  13    N0 Concentration in the Pacific Ocean  .............      42
  14   Global Annual Production and Release of
       Fluorocarbons-12 and -11. .  .  .......... ........     48
  15   Least-Squares Error Coefficients of a Third-Order
       Polynomial Used to Define the Global Distribution
       of Trace Constituents ......................     49
  16   F-12 Concentrations in the Pacific Ocean  .  .  ..........     56
  17, -" Concentrations of Halocarbons in Pacific Seawater .......  .     57
  18   Annual Global Production and Release of Fluorocarbon-22 .....     60
  19   Annual Global CC14 Release Since 1910 ..............     67
  20   Annual Global Production and Release Rate of Methyl Chloroform  .     75
  21   Methyl-Chloroform Global Average Residence Time

-------
22   Residence Times of Methyl Chloroform in the Northern and
     Southern Hemispheres  	  ...      81

23   Global Atmospheric Release of Methyl Chloride	      85

24   Average Concentrations of Selected Trace Constituents
     in Clean Environments		      88

25   Global Atmospheric Release of Chloroform, Methylene Chloride,
     Trichloroethylene, and Tetrachloroethylene  	      9]

26   Phosgene Production, Sales, and Release in the United States  .  .      92

27   Estimated World-Wide Production and Release
     of Ethylene Dibromide and Methyl Bromide  	      98

28   Average Background Concentrations of Important
     Trace Constituents  	 .............     101

29   Tropospheric Chlorine, Bromine, Iodine, and Fluorine
     Organic Budgets	     103
30   Estimated Tropospheric Residence Times of Selected Molecules  .  .     113

31   Urban-Nonurban Pollutant Relationships:  Concentrations
     and Daily Outdoor Doses of Chemical Species 	     117
                                      xi

-------
                               ACKNOWLEDGMENTS

     This project was supported in part by federal funds from the United States
Environmental Protection Agehcy under Grant No. R-8038020.  We thank many of
the personnel from the U.S. Coast Guard and especially the crew of the Burton
Island for their cooperation and help during this study.  Suggestions and
comments from Dr. P. L. Hanst and Mr. J. Spence of the EPA were appreciated.  We
thank Prof. E. Robinson of Washington State University for kindly providing us
with air samples from the South Pole.  Constructive suggestions from F. L.
Ludwig, R.T.H. Collis, and W. B. Johnson of SRI International are also
acknowledged.
                                      xii

-------
                                  SECTION 1
                                 INTRODUCTION

     Over the past two decades, an increasing number of hazardous chemicals
has been released into the environment.  Many of these chemicals persist in
the atmosphere for such long periods of time that they bring about nearly
permanent changes in the biosphere.  Others are found to be hazardous to human
health and are responsible, at least in part, for the rising cancer rates in
the industrialized countries.
     The primary group of pollutants of interest in this study was halogenated
organics, although other halogenated as well as nonhalogenated species (such
as SF((, N20, light hydrocarbons, and CO) were also studied.  Halogenated
organics are important atmospheric constituents for a number of reasons,
Halocarbons that are stable in the troposphere (such as fluorocarbons-12, -11,
-113, -114 and carbon tetrachloride) are suspect as precursors of stratospheric
ozone-destroying chlorine atoms.  Vinyl chloride has been linked to angiosar-
coma and is possibly mutagenic.  Many chloroethanes and methanes are suspect
as potential human carcinogens.  In addition, atmospheric halocarbons and
nitrous oxide absorb infrared heat radiation, and these may exacerbate the
warming effects caused by C02-
     It should be added that the same chemicals that threaten to cause global
pollution are tracers that can provide valuable information about the dynamics
and the chemistry of the natural atmosphere.
     The present study was therefore launched to develop information on the
composition of the clean and the polluted atmosphere to determine the distri-
bution, sources, and sinks of a large set of pollutants of interest.

-------
                             SECTION 2

                        OVERALL OBJECTIVES


The overall objectives of this study were:

•  To characterize the burden of selected trace constituents such as
   halocarbons, hydrocarbons, SF^, ^0, 0-j, NOX, and CO in the clean
   troposphere and the polluted urban environments.

•  To determine the atmospheric distributions of selected halocarbons,
   hydrocarbons, SF^, and ^0 in the northern and the southern.hemi-
   spheres.

•  To determine the growth rates, residence times, sources, and sinks of
   the trace constituents of interest.

•  To analyze measured atmospheric data, along with other available in-
   formation, to understand better the dynamics and the chemistry of the
   global atmosphere.

-------
                                  SECTION  3

                  SUMMARY OF MAJOR FINDINGS AND CONCLUSIONS


     A number of conclusions with wide-ranging implications to our understand-
ing of atmospheric chemistry were derived from the conduct of this study.   The

following are some of the salient results:

     •  Global analysis of carbon tetrachloride (CC14) emissions data since
        1914 showed that atmospheric CC14 was predominantly of man-made
        origin, contrary to the prevalent view that CCl^ was largely of
        natural origin (Singh et al . ,  1976a).
     •  Phosgene was measured for the first time in the urban and the clean
        atmosphere and its sources and sinks were characterized (Singh, 1976;
        Singh et al., 1977c).

     •  The global distribution of a comprehensive set of reactive and inert
        trace constituents was measured.  The average concentrations in the
        northern hemisphere, the southern hemisphere, and the globe are sum-
        marized in Table 1.  We demonstrated that relatively inert trace
        constituents (such as fluorocarbons-12, -11, -113, -114; CC14, SF6,
        N20) are well-mixed within each hemisphere and the average concentra-
        tion in the southern hemisphere differs only marginally (10 to 15 per-
        cent lower) from the northern hemisphere values.  More reactive man-
        made species (such as CH3CC13, C2C14, CHC13) were found to have more
        pronounced hemispheric gradients (Singh et al., 1979a).  Global
        distributions of many species (e.g., F-113, F-114, CI^Cl,
        and CH) were obtained for  the first time.
     •  Average atmospheric growth rates of important chlorocarbons (and N20)
        were characterized based on field studies conducted between 1975 and
        1978.  These are summarized in Table 2.
     •  Unlike other reactive chlorinated organics, methyl chloride was shown
        to possess no north-south concentration gradient.  This implied a
        large natural source.  We collected data from the Pacific Ocean and
        identified in the ocean a major source of methyl chloride (3 million
        tons/year) sufficient to account for nearly all of the atmospheric
        methyl chloride (Singh et al., 1979a).
     •  The distribution of fluorocarbons-12 and -11 was used with the global
        emissions data to establish quantitatively a mean interhemispheric
        exchange rate of 1.2 years (Singh et al., 1979a).
     •  The residence times of fluorocarbons-12 and -11 were shown to be so
        long (40 to 45 years for fluorocarbon-11 and 65 to 70 years for
        f luorocarbon-12) as to rule out the possibility of any significant

-------
                                   Table  1
               AVERAGE  BACKGROUND CONCENTRATIONS OF  IMPORTANT
                        ATMOSPHERIC TRACE CONSTITUENTS
Compound
N20
CC12F2 (F-12)
CC13F (F-ll)
CC12FCC1F2 (F-113)
CC1F2CC1F2 (F-114)
CHC1F2 (F-22)
CHC12F (F-21)
SF6
CC14
CH3CC13
- CH3C1
CH3I
CH3Br
CHC13
CH2C12
C2HC13
C2C1A
C2C16
CH2BrCH2Br
CH4
C2H6
C2H2
Major"*"
Source
N
A
A
A
A
A
A
A
A
A
N
N
A, N
A
A
A
A
A
A
N, A
N, A
A
Concentration*
Northern
Hemisphere
Average
311 ppb
230 ppt
133 ppt
19 ppt
12 ppt
20-30 ppt
5 ppt
0.31 ppt
122 ppt
113 ppt
611 ppt
2 ppt
5-20 ppt
14 ppt
44 ppt
16 ppt
40 ppt
<5 ppt
'<5 ppt
1430 ppb
1060 ppt
<200 ppt
Southern
Hemisphere
Average
311 ppb
210 ppt
119 ppt
18 ppt
10 ppt
—
4 ppt
0.27 ppt
119 ppt
77 ppt
615 ppt
2 ppt
—
<3 ppt
20 ppt
<3 ppt
12 ppt
—
— •
1390 ppb
524 ppt
<200 ppt
Global
Average
311 ppb
220 ppt
126 ppt
18 ppt
11 ppt
--
4 ppt
0.29 ppt
120 ppt
95 ppt
613 ppt
2 ppt
—
8 ppt
32 ppt
8 ppt
26 ppt
—
—
1410 ppb
792 ppt
<200 ppt
 *For those species where  significant latitudinal variations within the
 hemisphere were observed,  the  average  concentration within each hemisphere
 is the concentration  that, when uniformly mixed in the hemisphere, repre-
 sents the total burden of  the  species  in that hemisphere.  The concentra-
• tion data are  for late 1977:   ppt  = 10~12 v/v; ppb =  10~9 v/v.
 ^N «= Natural; A » Anthropogenic.

-------
                              Table 2

       AVERAGE ATMOSPHERIC GROWTH RATES OF SELECTED SPECIES
           FROM 1975 TO 1978 IN THE NORTHERN HEMISPHERE
Compound
N20
CC12F2 (F-12)
CC13F (F-ll)
cc.i4
CH3CC13
Average
Growth Rate*
<1 ppb/yr
22 ppt/yr
14 ppt/yr
3 ppt/yr
15 ppt/yr
               Based on data collected between
               30°N - 40°N.
   tropospheric sinks that may prevent these species from entering the
   stratosphere almost in their entirety (Singh, 1977a; 1977b; Singh
   et al., 1979a).

•  We demonstrated for the first time that the methyl ^chloroform resi-
   dence time is 8 to 11 years, contrary to the belief based on model
   predictions that methyl chloroform residence time must be 1 to 2
   years.  This new information suggested that 15 to 20 percent of all
   methyl chloroform released at ground level could enter the strato-
   sphere (Singh, 1977a, 1977b; Singh et al., 1979a) thus making methyl
   chloroform a potential depletor of stratospheric ozone.

•  The utility of using methyl chloroform as an indicator of the hydroxyl
   radical (HO) abundance was recognized.  We showed that the global
   average HO abundance was about 3 to 4 x 10^ molecules/cm^, or about
   a factor of 5 lower than that predicted by models.  This finding has
   wide implications for all classes of species for which reaction with
   HO is the dominant removal mechanism.  Thus, it appears that a large
   number of chemicals persist in the atmosphere for significantly longer
   periods that previously believed.

•  Methyl chloroform distribution in the northern hemisphere and the
   southern hemisphere also supports the view that the HO abundance in
   the southern hemisphere is at least 1.5 times more than in the
   northern hemisphere.  We inferred that higher levels of carbon monox-
   ide (CO), an effective sink for HO in the northern hemisphere, are
   the reason for this asymmetry.  Increasing CO emissions in the future
   have the potential to deplete further the scavenging ability of the
   atmosphere (Singh, 1977b).

-------
•  We demonstrated that the oceanic sink for CCl^ can provide an atmo-
   spheric turnover time of about 100 years, thus making this sink about
   half as effective as the stratospheric sink (Singh et al., 1979a).

•  Nitrous oxide (N20) was shown to have a uniform concentration in the
   northern hemisphere and southern hemisphere and an atmospheric resi-
   dence time of greater than 20 years (Singh et al., 1979b).

•  We found a dominant source of ^0 in the Pacific Ocean.  The global
   oceanic source was quantified to be 20 to 30 million tons (N20)/year
   (Singh et al., 1979b).

•  No change in the atmospheric abundance of N20 was detectable during
   the three-year period of this study.

•  The total organic halogen budgets were determined and are summarized
 * in Table 3.  These contradict free chlorine levels of greater than
   3 ppb (max 8 ppb) that have occasionally been reported in the upper
   stratosphere.  It is clear from Table 3 that man-made activities have
   already significantly perturbed the global .atmosphere.
                              Table 3       ,  ;. -\

            TROPOSPHERIC CHLORINE, BROMINE, IODINE AND
                     FLUORINE ORGANIC BUDGETS

Species

ci
Br
I
F
Budgets
Northern
Hemisphere
2.9 ppb
10-30 ppt
<2 ppt
1.0 ppb
Southern
Hemisphere
2.4 ppb
—
<2 ppt
0.9 ppb
Global
Average
2.7 ppb

<2 ppt
1.0 ppb
Percentage
Source
Contributions ;
Natural
23
50-90
100
0
Man-Mad e
77
10-50
0
100
•  We determined levels of a large number of halogenated pollutants by
   conducting field experiments at several characteristic urban sites.
   Many of the chemicals are either highly toxic or suspected carcino-
   gens.  The human intake (dose) of these chemicals was quantified.

•  Results from this study allowed us to provide our best estimates for
   the seasonally averaged atmospheric residence times of a large number
   of molecules.  These are given in Table 4.

-------
                                   Table  4

     151-ST  ESTIMATE  OF  THE  SEASONALLY AVERAGED ATMOSPHERIC  RESIDENCE TIME
           OF  SELECTED MOLECULES  BASED ON RESULTS  FROM THIS  STUDY
Molecule
CC12F2
CC13F
CH3C1
CH2C12
CHC13
CC14
CHFC12
CHF2C1
CH2C1F
CH3Br
CH4
CO
CC12CC12
CHC1CC12
CH3CC13
CH2C1CH2C1
CH3CHC12
CH3CH2C1
CH3CF2C1
CH2BrCH2Br
C2H6
N20
COS
cs2
Mean
Atmospheric
Residence Time,
T (years)
b5
40
2.7
1.0
1.2
>25 (<40)
4.3
34.8
3.3
2.9
21.5
0.6
0.8
0.04
8.3
0.6
0.5
0.3
50.6
0.6
0.5
>20
1.6
0.4
Major
Sinks*
S
S
T,S
T
T
S,0
T,S
S,T
T,S
T,S
T,S
T
T
T
T,S
T
T
T
S
T
T
S
T
T
Major
Source
A
A
N(0)
A
A
A
A S
A <
A
N.(0),A
N,A
N,A
A
A
A
A
A
A
A
A
A,N
N(0)
U
U
S-tUrntosphcre, T-troposphcre, 0-oceans, A-anthropogenic, N-natural, U-unknown
 Jn the instance more than one symbol is defined; the first, symbol represents
 the dominant sink, the second symbol represents a less important sink.   The
 latter si\\k is identified only if it contributes at least 5% to the total
 loss.   Unless otherwise identified in parentheses, the tropospheric sink is
 due to reaction with hydroxyl radical.  The percentage of (material released
 at ground level that will enter the stratosphere is estimated to be between
 JT to  2.5r.   For r>40 years nearly all of the material will enter the strato-
 sphere.
                                      7

-------
        •   '                       SECTION 4
                           ANALYTICAL METHODOLOGY

A.   Trace Constituents of Interest
     We attempted to measure as many halogenated compounds of interest as
possible.   These were complemented by measurements of other nonhalogenated
species that may affect the chlorine chemistry of the stratosphere.  Table 5
gives the list of chemicals that were measured during the conduct of this
study in urban, suburban, and background environments.  For all halogenated
species, N_0, and organic nitrogen compounds, electron-capture gas chroma-
tography was the primary means of analysis.  The hydrocarbons and CO were
measured using flame-ionization gas chromatography.  Continuous chemilumi-
nescent instrumentation was used for the analysis of 0.,, NO, and N0~.

B.   Experimental Procedures                   ,_
     1.   Air Analysis
     Because of the extremely high sensitivity of the electron-capture
detector (BCD) a number of species listed in Table 5 can be measured with a
direct 5-ml injection of air.  In general, however, preconcentration of
samples was necessary.  We chose to use as small a sample size as necessary
to measure a maximum number of chemicals of interest on a given GC column.
In no case did a sample volume of more than 500 ml have to be used.  The
cryogenic trapping of air was conducted in 1/16 in. diameter stainless-steel
traps packed with a 4 in. bed of glass beads or glass wool.  During sampling
these traps were maintained at liquid oxygen temperature.  The actual volume
of the sample was measured at the cryogenic trap exit.  The collected aliquot
was thermally desorbed and injected directly into the GC.  Both electrical
heating and hot water desorption techniques were used for injecting the
trapped sample and both were found to be satisfactory.  A typical sample size
was 250 ml.  Because of the dominant water response of the BCD, a post-column

-------
                               Table 5




CHEMICALS MEASURED IN THE CLEAN TROPOSPHERE AND IN THE PACIFIC OCEAN
Compound Name
Fluorocarbon-12
Fluorocarbon-11
Fluorocarbon-22
Fluorocarbon-21
Fluo rocarbon- 11 3
Fluorocarbon-114
Sulfur hexafluoride
Carbon tetrachloride
Chloroform
Methylene chloride
Methyl chloride
Methyl iodide
Methyl bromide
Hexachloroethane
Methyl chloroform
Ethylene dibromide
Tetrachloroethylene
Trichloroethylene
Phosgene
Nitrous oxide
Nitric oxide
Nitrogen dioxide
Peroxyacetyl nitrate (PAN)
Peroxypropionyl nitrate (PPN)
Ozone
Carbon monoxide
Methane
Ethylene
Acetylene
Ethane
Propane
Butane
Pentane
Chemical
Formula
CCl.F,
CC13F
CHC1F2
CHCljF
CC1F2CC12F
CC1F2CC1F2
SF6
cci4
CHC13
CH2C12
CH3C1
CH3I
CH3Br
cci3cci3
CH3CC13
CHjBrCHjBr
cci2cci2
CHC1CC12
coci2
N20
NO
N02
8
CH^COONO,
CH CH COONO
°33
CO
CH,
C2H4
C2H2
C2H6
C3H8
C4H10
C5H12
Measured in
Air
x
x
X
x"
X
X
X
x
X
X
x \
X
X
X
X
X
X
X
X
X
X
X
X
X

X
X
X
X
X
X
X
X
Sea Water
x
X





X
x

x
x







x













-------
 Ascarite trap was inserted to remove water for halocarbon analysis.   No water
 trap was used for the analysis of NjO, peroxyacetyl nitrate (PAN),
 peroxypropionyl nitrate (PPN), and phosgene.
      The ECDs were maintained at 325°C because we discovered that  in most
 cases the ECD response increased with an increase in temperature.   In the
 case of PAN, PPN, and phosgene, the EC-GC coulometer was maintained at 30°C.
 For these species, the reduced temperature seemed to give the best response.
 A number of GC columns were used during the conduct of this study.   Table 6
 shows the columns that can adequately be utilized for the separation and
 analysis of these chemicals.   Table 6 also shows;-the recommended columns for
.                                                  '"               ''
 routine analysis.  The other columns listed in Table 6 were mainly used for
 identification purposes.  In addition, the identiy of chemicals was confirmed
 by determining their EC ionization efficiency as well as EC thermal response.
 Methyl chloride identity was further established by GC/MS analysis.   The
 light hydrocarbons were measured by employing standard methods from the
"          »
 literature (Westberg et al.,  1974).  Details of these measurement  methods
 have already been provided (Singh et al., 1977d; 1978d) and need not be
 repeated here.
      Figures 1-6 provide representative chromatograms for the atmospheric
 separation of selected species.  The separation and analysis procedures for
 light hydrocarbons .are well documented (Westberg et al., 1974).

      2.   Water Analysis
      As shown in Table 5, we also performed water analysis for a limited
 number of species, namely fluorocarbons-12 and -11, CH^Cl, CHCl3»  CCl^, CHjI,
 and N20.  In the initial stages of the research effort a large volume of water
 was purged with ultrapure helium at a very slow flow rate.  The outcoming
 helium, which was in equilibrium with the water content, was analyzed and this
 concentration reported.  In the latter stages (third year) of this project,
 most of the water analysis was performed aboard ship.  For this cruise, we
 devised additional quantitative techniques.
      The analysis on the seven selected chemicals in seawater (Table 5) was
 conducted by quickly enclosing a 25-ml volume of seawater and an equal volume
                                       10

-------
                                                           Table  6
                                IDENTIFICATION AND SEPARATION OF SELECTED TRACE CONSTITUENTS
Chemlcal»

CCl F
CC12 2
CC1 F
CHClFj
CHC1,F
CClFjCCljF
CC1F,CC1F,
2 2
Sf6
cci4
CHClj
CHjClj
CHjCl
CH3I
CHjBr
CCl, CCl.
3 3
CH CC1
CH,lrCH,Br
2 2
2 -
COClj
"2°
PAN
PPN
C,H,
2 6
C'>K>
C'HJ
c)"«
... t) C.H10
(,,. 1) C,MU
X Coluan





X



X
X


X

























X












X












X
X
X
X


X

X
X
X
X



















X

X



X


X

X






X


















X
X


X



X





















X












X


















X
X







































































X










Nunb*r
10





















X
X







11




X
X


X
X






X













12




X



X

X



















13























X

X
X
X
X
X
14




















X









IS


'





X





X















16






























17






























EC
lonlsatlon
Efficiency




X


X
X
X


X






X

X
1







EC
Thermal
RMpona*





X



X
X
X

x !i
ft
•




X









CC/MS












X


















R«comm*nd*d
Coluttn

3, 4

3, 4
3
3, 12
3

2
1, 3
1, 11
12
3
1
3
15
1, 3
11
1 11

9
2, U
10
10
13

n
13
13
13
13
i:,.|>inni 1,    Silicon* oil IOX DC 200, SS, 6 ft « 1/4 in.,  30/60 Mih  ChroMinrb W(A/W)
>:ultMi .'.    Alwliu, SS. 20 fc » 1/S in., 80/100 Mih
Column I.    Silicon* oil 20! DC 200, SS, 30  ft » 1/4 in.,  30/60  m«h ChroBoiorb W(A/W)
Column 4.    Chropoill 310, SS, 1J  tt *  1/4 in., 80/100  H*h  Chroooiorb  W(A/W)
Column ).    SI 1C  )0, SS, 7 ft » 1/4 in., 80/100 vth Oiromoiorb W(A/U)
Column l).    Por*p*k Q, SS, 6 ft x  1/8  in., (0/100  M*h
Column 7.    0.4X Carbovix 1500, SS, IS ft '  1/4 in., 80/100  m.ih Chromo.orb W(A/W)
Column 8,    10! C*rbov*x 400, SS,  10 ft > 1/4  in,, 80/100 Mih Chromoiurb W(A/W)
Colu-wi 9.    302 Dldtcyl phtluUt*. SS, 5 ft  »  1/4  in.,  100/120 m*fh  Chromogorb P(A/U)
lolumn 10.  31 C*rbow*x ' 400, Ttflon, 10 in.  * 1/4  in.,  60/80 m««h  Chromoiorb W(A/W)
i:.'lumn 11.  20Z SP-2100, SS, 15 ft « 1/8 in.,  80/100 null Sup«lcoport
Column 12.  Cucbupack C/0.21 CW 1500,  SS, b  ft «  1/8  in.,  80/100 math
Column 1J.  Ourapak n-oct*n
-------
                                                             CCI3F-
                                                             1 38 i>pt
    t
               10.0-ml SAMPLE
               SUNDAY OCT. 5, 197t)
               ATT. 2
               MENLOPARK.CA
                                               C2HCI3
                                               6"20 ppt
                                     I
40
       36
32
       28
24
                                    20

                                  MINUTES
16
                      12
         FIGURE 1  FLUOROCARBON-11, CH3CCI3, AND CCI4 ALONG WITH
                  OTHER FLUOROCARBONS IN THE AMBIENT
                                     12

-------
                        N2O
                              2-mi AIR SAMPLE
                              (40 N   CLEAN
                              MAHINE AIH)
                                  CCI2F2(F12)
                             5         10
                            TIME   min
FIGURE 2  FLUOROCARBON-12, AND N2O IN THE AMBIENT AIR
                            13

-------
I
         Ftl
                200 ml AIR SAMPLE
                (40 ft X Vt" DC 200)
                                             SA-44B7-3
 FIGURE 3  FLUOROCARBONS-12. -114, -21. -11; CH3CI AND CH3Br
          SEPARATION FROM THE AMBIENT AIR
                            14

-------
                                                          CH2BrCH2Br
         100 ml AIR INJECTION
         U.C. RIVERSIDE
         (Sampla coll*cMd from
         • contested parking lot)
     98
90             26
   TIME (mm)
                                                               22
                                                                            18
FIGURE 4  ETHYLENE DIBROMIDE AND C2CI6 SEPARATION FROM AMBIENT AIR

                                      15

-------
                  Flurocarbon 11
                  Retention time: 6.9 minutes
                           p -  0.40
                    Sample size:  13.0 me
                    (Menlo Park, California Air)
                       Phosgenn
                       Retention time: 5.1 min
                                p =  0.65
987654
                TIME — minutes

 FIGURE 5  PHOSGENE SEPARATION FROM
            THE AMBIENT AIR
                     16


                      I

-------
               WATER
                           6 ml SAMPLE
                           FROM U.C. NIVIMSIOC
                           MAV 2, 1»77
                           COULMETMIC IC-OC
FIGURE 6  PAN AND PPN SEPARATION FROM AMBIENT AIR USING A
          COULOMETRIC EC-GC
                            17

-------
of ultrapure air in an all-glass syringe of 100-ml volume.  The speed with
which samples were transferred ensured minimal temperature change and species
loss during sample transfer.  Once in the syringe the sample was allowed to
reach equilibrium in about 30 minutes.  This also allowed the water to reach
the room temperature, which was carefully recorded.  Fluorocarbons-12 and -11,
CC1/, CH-C1, CHC1-, and N20 were analyzed in the air, the corresponding
equilibrium concentration in seawater was determined from solubility data
(Junge et al., 1971; Singh et al., 1978b; Billing, 1978) at the measured room
temperature, and the two were added to obtain the concentration of the
chemical in seawater.  Replicate analyses of water samples suggested a pre-
cision of better than 5 percent.  Details of these techniques have already
been presented (Singh et al., 1978b).  All water samples were analyzed within
an hour or two of collection.

     3.   Calibrations
     Calibrations for halocarbons, hydrocarbons, PAN, PPN, N00, and SF, were
                                                            *.         o
performed using permeation tubes, gas-phase coulometry and multiple dilutions
as necessary.
     Permeation tubes (3.2 in. long) for 18 halocarbons of interest, con-
structed from standard FEP Teflon tubing of varying thicknesses, were obtained
from AID Inc. (Avondale, PA).  The methods of manufacture and sealing of
permeation tubes are proprietory (AID, private communication) but are very
similar to those used by O'Keeffe and Ortman (1966) in principle.  Each
permeation tube was contained in a specialized glass holder (Figure 7) which
was held in a 37-1 water bath (18 * 12 * 10.5 in.) maintained at 31.0 ± 0.05°C,
Permeation holders were flushed constantly with a purge gas flowing through
at a rate of approximately 15 ml/min.  The purge gas was prepurified helium
and was further passed through a sequence of traps containing charcoal,
anhydrous calcium sulfate, and molecular sieve.  The permeation holder inlet
coil length was adequate to allow a helium flow rate of as much as 5 1/min
over the permeation tubes without causing any heat transfer problems.  All
tubing materials were either aluminum or glass.  Fittings were either brass
Swagelok or all-glass ball joints.
                                      18

-------
             MIXTURE   PURIFIED
               OUT    HELIUM IN
                             MALE BALL JOINTS
                       GLASS SIEVE PLATE
                       (Permeation tub« stands
                       vertically on this piatel
        FIGURE 7  PERMEATION TUBE HOLDER
                   CO/ S«M»Lt SIZE. Pi I10*"|l

FIGURE 8   DEMONSTRATION OF LINEAR RESPONSE
           OF FREQUENCY-MODULATED EC DETECTOR
                         19

-------
     Permeation tubes were weighed at least once a week and in many cases
twice a week on a semimicro (10  g) balance.  The water bath was filled to a
constant level every week with the additional water at 30-32°C to make up for
evaporation losses.  Table 7 shows the permeation rates of a number of
chemicals measured in our laboratory.  Thus, direct ppb concentration mixtures
could be obtained from these tubes.  Extensive analysis also suggested that
the frequency-modulated ECDs in use at SRI were highly linear in the low ppb
range.  Figure 8 demonstrates this linearity for fluorbcarbon-11.
     In addition to permeation tubes, commercially available standards were
obtained for nearly all chemicals of interest.  These were obtained at higher
concentrations («5 to 10 ppm) for long-term stability reasons.  All of the
standards were stored in aluminum cylinders except CH«C1, which was contained
in stainless-steel cylinders.  Extreme care had to be used in selecting
cylinder materials since some of the chemicals (e.g., CH^Cl) form unknown
complexes with aluminum leading to potentially explosive conditions.  All of
the commercially available standards were rechecked with permeation tubes or
our own multiple-dilution techniques.  The hydrocarbon standards were checked
against those available from NBS and found to agree within ±5 percent.
     In addition to these, an exponential dilution system was installed in our
mobile laboratory.  This system was devised at SRI and is based on ultrasonic
sound-wave principles.  The exponential dilution system worked satisfactorily
down to low ppb levels.  At sub-ppb levels surface sorption problems made the
use of this system unacceptable.  However, it was very useful for species such
as NpO that are present at concentrations of several hundred ppb's.
     In addition, the measured concentrations of some of the species
(fluorocarbon-11 and CC1,) were further checked by conducting absolute
analysis with the aid of the gas-coulometric technique (Lillian and Singh
1974, Singh et al., 1975b).  In the case of PAN and PPN, only gas-phase
coulometry was used and the data must be considered preliminary until the
confirmation of the reliability of PAN and PPN determination using gas-phase
coulometry can be established.  In the meantime, these data do appear reason-
able.  Calibration for total hydrocarbons, CO, and CH, were conducted in the
field at least twice a day.  N02 calibrations were conducted using a perme-
ation tube as primary standard.  Standard secondary mixtures of NO and N0?

                                      20

-------
                                         Table 7

                   PERMEATION  RATE  DATA FOR HALOCARBON  PRIMARY STANDARDS
Compound
CHjCl
CH2C12
CHC13
CC13F (F-ll)
CH3Br
CH31
coci2 .
C2H5C1
CH2CHC1
CH2C1CH2C1
CHC1CHC12
CC1F,CC1F, (F-114)
i i
CC1,FCC1F, (F-113)
i i
cci2cci2$
cci45
CH3CC1 §
CH2BrCH2Br5
CC12F2 (F-12)"
Tube wall
thickness*
(inch)
0.125
0.030
0.030
0.030
0.030
0.030
0.125
0.030
0.062
0.030
0.030
0.030
0.030
0.030
0.030
0.030
0.030
0.125
Permeation rate,
ng/min-952
confidence limit
1350.0 ± 80.4
580.0 ± 34.2
132.0 ± 12.7
1050.0 ± 78.4
2020.0 ± 102.2
256.6 ± 9.6
2460.0 ± 183.6
460.0 ± 37.8
1170.0 ± 64.1
71.3 ± 6.0
246.0 ± 20.1
14160.0 ± 240.0
480.0 ± 49.6
64.8 ± 26.1

...


Mean permeation
rate, ppb/1/min
(25°C, 1 atm)
653.6
166.8
27.0
186.9
520.2
44.2
607.6
174.4
457.7
.17.6
45.7
2025 . 6
62.6
9.5

...


Status'*'
S
S
S
S
S
S
S
S
S
S
S
S
S
u
u
u
u
u
Conditioning
time,
weeks
<1
<1
1-2
5-6
1-2
5-6
1-2
1-2
1-2
1-2
1-2
1-2
5-6


* • *


 Tubes were made of FEP Teflon tubing (3.2 in. long) and were maintained at 31.0 - 0.05°C.
 For tubes with wall thickness of 0.125 in., the tube o.d. was 0.375 in.; for all others
 the o.d.  was 0.250 in.
0.
'S » satisfactory; U • unsatisfactory.
t
 There are short time periods when CC12CC12 permeation tubes perform satisfactorily; however,
 errors of less than 10% are difficult to obtain.  TFE Teflon is expected to perform better
 and will  be tested.

3Leak rate is too slow to quantify accurately and may be used as a secondary standard.

 Leak rate Is too fast at 31°C; new permeation tube is under test.
                                             21

-------
 were used in the field; 0~ primary calibrations were performed using the
                          J                        t>
 tneutral-buffer Kl method and then monitored with the Bendix 0_ calibrator.
      In all field experiments, we calibrated 3 to 4 times a day by using
 secondary standards.  These secondary standards were samples of atmospheric
"air that had been carefully characterized by comparing them against primary
 standards.  Only these secondary standards were used during field operations.
 The frequency of use, however, was variable.  For N20 analysis, where we were
 aiming for a less than 1 percent precision, every analysis of an air sample
 was followed by the analysis of a standard.
      Because of the long-term study period and the need to characterize the
 troposphere accurately, several air standards were retained in specially
 polished 5-1 stainless-steel vessels.  These were used to ensure long-term
 internal consistency.  For species such as fluorocarbons-12, -11, -113, and
 -114, CC1,, CH-CC1-, N20, and SF   they performed excellently.   In most
 other cases, however, the trace constituents showed signs of deterioration
 .with.time.  When this occurred, we had to depend largely on our ability to
 generate standards repeatedly.
      Because halocarbons are relatively new pollutants and present at very
 low concentrations, the generation of primary standards is a difficult and
 tedious task.  At the same time, the needs are such that an extremely high
 degree of accuracy is required.  There are no agreed-upon sources of primary
 standards.  To test the uncertainties in measurements, identical samples of
 air of unknown composition were analyzed by SRI and Washington State Uni-
 versity (WSU) independently.  The results of this intercomparison are shown
 in Table 8 and indicate reasonable agreement.
                                       22

-------
              Table 8

ANALYSIS AND INTERCOM?ARISON OF AN
IDENTICAL AIR SAMPLE BY SRI AND WSU
Compound
N20 (ppb)

F-12 (ppt)
F-ll (ppt)
CH3CC13 (ppt)
CC14 (ppt)
Concentration
SRI
312

224
124
106
122
WSU
329

250
150
104
150
                 23

-------
                  I
    SECTION 5
EXPERIMENTAL PLAN
      Our  first attempt was  to develop  analytical methodology for a large
 number of species  identified in Table  5.  Having developed the methodology, we
 conducted a number of short-term  field programs in urban, suburban, and clean
 background environments.  In this chapter, we  shall  identify the nature of the
;.,field experiments  and the mode of operation.

 A.   Air Sampling'with the  Instrumental Mobile Laboratory
      All measurements over  the continental United States were performed with
 the help of an instrumented mobile laboratory.  Table  9 summarizes the equip-
 ment installed in  this mobile laboratory  for this program.  This laboratory
 was air conditioned  for  temperature control and operated on a 220-V, 80-A
 circuit.  Provisions were also devi/sed to operate on 110-V input.  A 200-m
 electrical cord wa's  always  used to station the laboratory away from the build-
 ing of electrical  source.   The sampling manifold was all stainless steel with
 a variable inlet height.  In almost all cases, sampling sites were chosen to
 have no nearby buildings that exceeded a  height of 5 meters.  The inlet of the
 sampling manifold  was adjusted to be higher than nearby structures.  A typical
 manifold  inlet height was 5 m above ground.
      We conducted  a  total of 14 field  studies  in the western half of the
 United States with the help of the mobile laboratory.  The site locations are
 shown in  Figure 9.  Table 10 provides  additional information on the nature of
 the sites selected.  A wide variety of sites and locations was considered.
 The Pacific Coast  provided an ideal location for clean background conditions
 because of westerly  winds and virtually no source within several thousand
 miles upwind.  Occasionally, however,  the winds were easterly and urban
 effects were encountered.   Since  some  of  the species being measured in this
 study are excellent  indicators of urban contamination, such data are not con-
 sidered in determining the  geochemical background of pollutant levels.

                                        24

-------
                                   Table 9
               ENVIRONMENTAL MOBILE LABORATORY INSTRUMENTATION
     Instrumentation
        Feature
         Analysis
Perkin Elmer 3920 GC1
Perkin Elmer 3920 GC2
Perkin Elmer 3920 GC3
Coulometric dual EC-GC

Beckman 6800
Horiba AIA-24
Bendix 8101-B  «
Dasibi Model 1003 AH
AID Model 560
Monitor Labs Model 8440E
Bendix 8002
Eppley Pyranometer
Eppley UV Radiometer
Meteorological equipment
Auto Lab TV data
System (No. 1)
Digitom data system
(No. 2)
Stainless steel manifold
Teflon manifold
2 ECD* and 1 dual FID1'
2 ECD and 1 dual ECD
2 ECD and 1 dual ECD
Coulometric ECD

FID
NDIR*
Chemiluminescent
Photometric principle
Chemiluminescent
Chemiluminescent
Chemiluminescent
Trace constituents
Trace constituents
Trace constituents
Halocarbons, PAN, PPN,
COCX.2 analysis and
calibration
CO-CH4-THC
CO, C02
NO, N02
Ozone
Ozone
NO and NOX
03
Solar flux
Ultraviolet radiative flux
Wind speed, wind direction,
temp, pressure, dew point,
relative humidity, etc.
All gas chromatography
data Input
All continuous air quality
and meteorological data
Sampling of HCs, halocar-
bons, etc.
Sampling of 03, NO, NOX
 Electron capture detector.
"^Flame ionization detector.
TNondispersive infrared.
                                      25

-------
                                            KANSAS  •
       SCALE
    0 20   60   100 miles

    i  '.''.'  I
    0  60 100  161 kllomettn
FIGURE 9  LOCATION OF MONITORING SITES ON THE CONTINENTAL UNITED STATES
                                   26

-------
                                                                 Table 10




                                           MEASUREMENT SITES OVER THE CONTINENTAL UNITED STATES
Site Ho.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Location
Name
Stanford Hills, CA
Point Reyes, CA
Los Angeles, CA
Palm Springs, CA
Yosemite, CA
Menlo Park, CA
Mill Valley, CA
Riverside, CA
Badger Pass, CA
Reese River, NA
Point Arena, CA
Jetraore , KA
San Jose, CA
Point Arena, CA
Lat.
CN)
37° 23'
38° 07'
34° 04'
33° 48'
37° 39'
37" 24'
37° 59'
33° 54'
37° 39'
38° 59'
38° 57'
37° 59'
37° 20'
38° 57'
Long.
<°W)
122" 14'
122" 56'
118° 11'
116° 33'
119° 40'
122° 12'
122° 39'
117C 23'
119° 42'
117° 28'
112° 44'
99° 53*
121° 53'
123° 44'
Approximate
Elevation
MSL, meters
120
10
20
800
2380
20
760
250
2380
1982
10
850
20
10
Monitoring
Period
11/23 - 11/30, 1975
12/01 - 12/12, 1975
4/28 - 5/04, 1976
5/05 - 5/11, 1976
5/12 - 5/18, 1976
5/24 - 5/28, 1976
1/11 - 1/27, 1977
4/25 - 5/04, 1977
5/05 - 5/13, 1977
5/14 - 5/20, 1977
5/23 - 5/30, 1977
6/01 - 6/07, 1978
8/21 - 8/27, 1978
8/30 - 9705, 1978
Nature of Site
Clean site subject to urban trans-
port
Clean marine
Urban
Urban-suburban
Remote-high altitude
Suburban
Clean marine-subject to urban
transport
Urban-suburban
Remote-high altitude
Remote-high altitude
Clean-marine
Remote continental
Urban
Clean marine
K>

-------
                                               i
      During each of the field studies, the mobile laboratory was  installed  in
 a carefully selected place and measurements were conducted for a  period  of  one
 to two weeks.  Despite the difficulty in logistics,  we decided that 24 h
 operation is necessary for proper characterization of the atmosphere.  There-
 fore, in all cases a round-the-clock measurement schedule was followed.

 B.   Global Sampling
      The global atmospheric measurements extended from 64°N to 90°S latitudes.
 The Pacific Ocean water measurements were conducted  between 46°N  to 40°S
 latitudes.   Figure 10 describes the sampling locations of the four experiments.
 In all the global experiments it was our objective to obtain the  cleanest
 possible samples that would be representative of background conditions.   The
 global sampling-and analysis program was composed of four experiments.  Three
 of these involved air analysis and the fourth on.  involved water  analysis.

      1.   Air Sampling
           a.   Whole-Air Batch Sampling
      All batch sampling was conducted in specially designed stainless  steel
 and glass vessels of 1 liter volume each.  Both  type| of sampling vessels were
 so designed that they could be easily pressurized to 40 psi.  Figures  11 and
 12 show these sampling vessels.  The stainless-steel sampling vessels  were
 electrochemically polished to have an inert surface  and were commercially
 obtained from D and F Instruments.  The glass vessels were constructed at SRI
 International.   Prior to use, they were heated to 250°C and flushed with ultra-
 pure helium that was cryogenically cleaned.  A hundred volumes of the  sampling
 vessel were exchanged over 100 minutes (helium flow rate =1 liter/minute).
 The sampling vessels were then checked for background contamination, and the
 procedure was repeated until the background contamination of the  sampling
 vessels was reduced to less than 2 to 5 percent  of the expected background
 concentration of a given trace constituent.  Tin- sampling vessels were also
 placed in a contaminated room and rechecked for  pressure stability and back-
.ground contamination.  Typically, the vessels that maintained the pressure
 were not affected by room contamination because  of the high positive pressure
 maintained in these vessels.
                                       28

-------
                                                          (IB Sept -   Oet 1977)
                                                             E»p*i«nkm
                                     LAS PALM
                                    	4_._
                                         DAK/
        -j/   TRIP !
        /I* N» - '3 I .- 19771
       /   E> w
180  I60W   140    120    100    80    60    40   20W
20E   40    60    80  100
 FIGURE 10   MAP SHO VING THE SAMf LING LOCATIONS FOR THE FOUR EXPERIMENTS
             ''he open -ircles indicate th  deep seawatcr  ) to 300 m) sampling sites.
                                      29

-------

FIGURE 11  SPECIALLY TREATED 1-LITER STAINLESS-STEEL SAMPLING VESSEL
          FOR COLLECTING GRAB SAMPLES
                              30

-------




FIGURE 12  ONE-LITER GLASS SAMPLING VESSEL
                   31

-------
     Prior to air sampling, the sampling vessels were carried at a positive
pressure of helium (=5 psi) to prevent any contamination en route.  During
sampling each vessel was flushed with 100 to 150 liters of ambient air and-  , '
then pressurized to 40 psi with a special stainless-steel Metal Bellows com-
pressor pump (Model MB 158), which could maintain a.continuous flow rate of
25 liters/minute.  A 200-m electrical cord was carried so we could sample away
from the source of electricity, usually a farmhouse or an abandoned building
near the ocean.  Two*blank samples were also carried to ensure that no con-
tamination occurred during sample collection and analysis.

     Experiment 1—Pressurized air samples were collected between 64°N and
20°S in specially treated 1-liter stainless-steel and glass vessels.  The
                                           t
sampling vessel and the sampling procedure are discussed in detail below.   ;     •
                                                                       » r. •   «•*'"»
Typically, four samples—two with glass and two with stainless-steel vessels---
were collected at each site.  Figure 10 shows the locations from which air was
collected (Trip 1).  The air samples were always collected at the cleanest *
possible locations under the most favorable meteorological conditions.  A few
urban samples were also collected.  Trip 1 sampling continued from 15 September
to 26 October 1977.  These samples were analyzed after 26 October 1977.

     Experiment 2—Air samples were collected on board the U.S. Coast Guard
icebreaker, Burton Island, which traveled from Oakland, California to Wellington,
New Zealand, covering 36°N to 42°S latitude between 120°W and 175°E longitude.
The actual1 itinerary is shown in Figure 10 (Trip 2).  Air samples were collected
In stainless-steel vessels, as in Experiment 1.  This experiment was conducted
from 20 November to 13 December 1977, and these samples were latet analyzed aih
our laboratories in Menlo Park, California.

          b.   In Situ Sampling
     Experiment 3—This experiment was conducted during the same oceanogrnphic
cruise and at the same time as Experiment 2.  The major difference in Experi-
ment 3, however, is that the air samples were analyzed in situ with two 3920
Perkin Elmer gas chromatographs, each equipped with two electron-capture
detectors and one flame-ionization detector, and a coulometric gas chromato-
graph that were installed on board.
         + --  -                        32

-------
     Aboard the Burton Island, in situ air sampling was conducted using a
stainless-steel manifold with two inlet positions.  We hoped that contamination
from the ship exhaust could be avoided by switching from one manifold inlet to
the other if meteorological conditions turned unfavorable.  Both manifold
inlets were at the front of the vessel, about 10 m above sea level and 20 m
apart.  At any given moment, only the manifold inlet that would obtain the
cleanest possible air was operational.  The decision to select the right
manifold inlet was made on the basis of wind data.  On extremely calm days
(wind speed <5mph) it was impossible to avoid exhaust contamination from either
of the sampling manifolds, and air samples were collected in glass syringes
from the best possible location.  If clean air samples could not be obtained,
analysis of air samples was terminated.  Because the Burton Island is round-
bottomed and rolls excessively, we could not monitor on highly stormy days.
           j                                                ?
     2 •   Viator Samp line
     Experiment 4—The task of water sampling from depths of 0 to 300 m was
handled by the Burton Island crew.  Paired samples were obtained from 0 to
300 m at 50-m intervals.  Water temperature at these depths was also measured.
Each water sample provided 0.5 liters of seawater and was usually analyzed
within an hour or two of collection.  Figure 13 shows the water sampling being
conducted.
     Water samples from the Pacific Ocean were analyzed on-site aboard the
Burton Island (Trip 2).  In addition, water samples from 40°N to 46°N in the
1'cirlfic Ocean were also analyzed.  Seawater samples were collected from the
surface (0 to 2 m) and from depths of 0 to 300 m at intervals of 50 m.  Analy-
sis of water samples was limited to six key species.  These were ^0, CHjCl,
CC12F2, CHC13, CC^F and CCl^.  One CC^F and CC14 paired surface sample a day
was collected during Trip 2.  The deep-water sampling sites are identified in
Figure 10.
                                      33

-------

                                   .
                                             •
FIGURE 13  DEEPWATER SAMPLING IN THE PACIFIC OCEAN

-------
                                  SECTION 6
               DISTRIBUTION,  SOURCES,  AND SINKS OF NITROUS OXIDE

     Nitrous oxide (^0) has received attention from atmospheric scientists
in iveent years because of the hypothesis that continued use of nitrogen
fertilizers ran increase the atmospheric abundance of N?0, thereby causing a
depletion in the stratospheric ozone (McElroy et al., 1976; Crutzen, 1976;
Liu ot al., 1977).  Estimates of the depletion of stratospheric ozone by ^0
are 
-------
     Figure 14 shows the  global  distribution of N~0 from data obtained in
November-December of 1977.  The  Northern  Hemisphere average background
concentration of 311 ±2.3  ppb  (ppb  =  10~9  v/v), the Southern Hemisphere
average background concentration of  311 ± 2.8 ppb,  and the global average
of 311 :t 2.6 ppb are scarcely different from one another.  Thus, it appears
that N,,0 is very uniformly  distributed, and the standard deviation of the
      t-                                                           V
global variability is 0.8%  (o).   Because part of this variability is due to'
the precision of our measurements (o  = 0.4%) and to the small variability
of about 0.2% (ODU) associated with  the RH  corrections, a net atmospheric
                                              2222
variability of 0.7% (o  )  can be  calculated  (o  =0-0  - 0DU).  Using the
                      cl                      el    t    p    Cul
approximate statistical derivation of  Junge (1974), an atmospheric residence
                          14
time of 20 years (TN «  =  —) can be  calculated.
 O.
 a.
340

320

<
300
280
260
| I | 1 | 1 1
SH AVG. N20 CONC. GLOBAL AV(
"~ =. 311 ± 2.3 ppb = 311 ± 2.6 p
7
r-f
*

—
I I I 1 1 1 I 1
1 1 | 1 | 1 ]
3. N2O CONC. NH AVG. N20 CONC.
pb = 31 1 ± 2.8 ppb
0 __
D
Wn!8^!* $""T" fi
*
—
I I I I I I I
      S  -80     -60     -40      -20       0      20
                                  LATITUDE — degrees
40
60
80  N
       v TRIP 1. EXPERIMENT 1. STAINLESS-STEEL VESSELS
       A TRIP 1, EXPERIMENT 1, GLASS VESSELS
       O TRIP 2, EXPERIMENT 2, STAINLESS-STEEL VESSELS
       Q TRIP 2, EXPERIMENT 3, IN SITU AIR SAMPLING AND ANALYSIS

                    FIGURE 14  GLOBAL DISTRIBUTION OF N2O
                                       36

-------
     We have been conducting N~0 measurements  since  1975,  and have also
retained air samples since September  1975.   In the interim, we  have  conducted
a number of short-term studies to measure N~0  background in various  seasons
on the West Coast of the United States.  Our measurements  between November 1975
and August 1978 (Figure 15) show no systematic average  increase in ^0.   How-
ever, the precision of our early measurements  was not very high because  of  in-
frequent calibrations (Figure 15).  To alleviate this deficiency we  carried  the
air samples we collected in September 1975  and compared them with in situ air
samples between 24°N and 35°S on Trip 2  (November, 1977).  Table 11  shows a
comparison of replicate analysis of air  samples from 1975  and 1977.   The
average N»0 concentration of air samples from  1975 was  found to be 312.3 ±
0.81 ppb; ir. the in situ samples of November-December 1977 it was 311.8  ±
1.14 ppb.  These averages are statistically indistinguishable.   It thus
appears that any change in that atmospheric abundance of N«0 since 1975  has
been well within the atmospheric variabilities of N~0.
              -O
              a
              a
              O
               M
                 340
                320
                 300
                280
                260
                              10
                                    15    20
                                  TIME — months
25
      30
35
                 FIGURE 15  ATMOSPHERIC GROWTH OF N20 WITH TIME
                           Time period beginning 1 November 1975
                                       37

-------
                                   Table 11
               AMBIENT N20 CONCENTRATIONS (ppb)  IN 1975 AND 1977
t










Latitude
24°N
18°N
14°N
4°N
1°N
,4°S
15°S
•25°S
30°S
35°S
Average
concentration (±o)
September
1975
312
311
313
312
313
313
313
312
313
311
312.3 ±0.81
November-
December
|* 1977
310
312
313
311
312
312
313
310
313
312
311.8 ±1.14
B.   Pacific Seawater Measurements
     Pacific seawater N_0 measurements are of considerable interest because
preliminary data from this region collected by Craig and Gordon (1963) led
McElroy et al. (1976) to propose the possibility of oceans as a sink for N-0.
The Pacific data of Craig and Gordon were preliminary because the measurement
sensitivity was inadequate, and samples from various depths had to be lumped
together.  Rasmussen et al. (1976) have reported measurements from the eastern
tropical Pacific (31°N to 11°S), but their results may be low (Hahn and Junge,
1977) because of the use of incorrectly published solubility data (Junge
et al., 1971).  Here .we provide a much wider coverage (46°N to 40°S) of N_0
data from the North and South Pacific.  To the extent that the Pacific Ocean
is representative of world oceans in terms of microbial activity (McElroy
et al., 1976), our data tend to remedy some of the past deficiencies in
available oceanic N»0 data.
                                      38

-------
     Rop] icat.e seawat-er  samples  from the  Pacific  Ocean from 45°N to 40°S were
analy/>'il during Trip  2.   Sampling  locations and  the sampling and analysis
methodo ' ugy have already been  discussed  in Section IV.  A total of 16 paired
suriac'.: si-awater samples (0  to 2 m)  were  analyzed.   In eight instances, N?0
analyses were conducted  down  to  300-m depth at 50 m intervals (see Figure 10).
In a major ily of cases,  rep] icate  samples indicated concentration differences
that \M-re within '3 percent.   Occasionally a greater difference of 15 percent
was uru-ounf.f'L'd.  In  all cases the concentrations were averaged.  There were
also some  uTstauces when only  a  single sample could be analyzed.  Such situ-
ations occurred iargvly  because  of unfavorable weather conditions resulting
in unsafe working condition.';.  The variation of  surface seawater N_0 concen-
trations, percentage  supersaturation,  and .'seawater temperature with latitude
is shown in Ta^lo \2  and plotted in  Figure 16.  It is clear from Figure 16
that in ail instances an N,,0  SMpersaturation was  observed.  . Tl^e. average surface
N00 concentration was found  to be  0.38 ug/1 and  the average surface saturation
was 133 percent, with a  maximum  of 190 percent near the equator (Table 12).
     From the surface NO data it  is possible to  calculate  the flux, F, of
N ,0 into the atmosphere  by using a simple film diffusion model:

                              F „
where 1)(T)  Is; the-  film diffusion  c;oc?f 1 icient for NO in water and is a function
of temperature; /  is  the  stagnant film depth (Broecker and Peng, 1974) which
h.a.-- bi?i-n ,nt?,isurc-:d  to  br 60  ±  30 }im (pi'\ = 10  in)  for the world ocean; C  is the
con.-entiai Ion of N,,0  in water; and Cv   is the N?0 concentration in equilibrium
with the: atuiosplieric  abundance of N«0.
     For uui calculations we  have selected Z = 60 m and used the solubility
data presented ir  Singh et  al.  (1978b) .   The variation of D(T) with tempera-
ture was obtained  from Broecker and Peng (1974).  Our choice of Z is the one
recommended by 1-kihn and Junge (1977) ,  but the actual value of Z may be higher
than 60 imi  for a biologically nettye  g«s such as N«0.  The mean NyO concentra-
Li on in the atmosphere was  taken  to be 311 ppb.   The variation of calculated
flux with latitude is also  shown  in Table 12 and Figure 16.  An average N_0
iJnx of 3(> x KJ-14 g/ctn^/s  was calculated.  Table 13 shows the; NO data
                                       39

-------
                                        *•    Table 12
                               CONCENTRATION  IN PACIFIC SURFACE WATER*
Date
10-29-76
10-28-76
11-25-77
11-26-77
11-27-77
11-28-77
11-29-77
11-30-77
12-01-77
12-02-77
12-04-77
12-05-77
12-06-77
12-07-77
12-08-77
12-11-77
Average
Lat
45°35'N
41°35'N
19°02'N
15°43'N
11°52'N
06°35'N
02°40'N
01°26'S
06°07'S
09°07'S
18°24'S
22°38'S
25°40'S
30°14'S
33'40'S
39°05'S

Long
125°55'W
125°35'W
137°48'W
140° 05 'W
142°05'W
145°14'W
147°43'W
150°10'W
153°13'W
155°35'W
161°42'W
164°26'W
167°33'W
171°18'W
170°50'W
179°i-3'W

Surface
Water Temp

-------
  1000
   900
   800
   700
   600
   500

   400

   300
   200
   100
                                                 I
       50   40    30    20
               SOUTH
   10    0    10
LATITUDE — degrees
20   30   40   50
   NORTH
     	" N20 FLUX (g/cm2- s X 1015)
      —— — SURFACE WATER CONCENTRATION (/ug/B X 103)
     	 SURFACE WATER TEMPERATURE ( 'C x 10)
     ———— PERCENT N2O SATURATION (200-300 m AVG.)
   - —— —— PERCENT N2O SATURATION IN SURFACE WATER (0 TO 3 m)

FIGURE 16  N20 SUPERSATURATION IN THE PACIFIC OCEAN (46°N-40°S)
                             41

-------
                Table 13




N20 CONCENTRATION IN THE PACIFIC OCEAN


Depth
(•)
0
50
100
130
200
2JO
300
Date 10-29-76
Lat 45*35'H
Long 125*55 'W
Water
Temp
CC)
1J.5
12.2
10. 5
9.6
9.0
8.6
8.2
N,0
(ug/D
0.46
0.56
0.73
0.95
1.27
1.37
1.62
X
Sat
115
136
166
213
270
291
296
Date 10-23-76
Lat 41'35'N
Long 125'55'U
Water
Teap
CC)
14.2
13.2
12.0
10.7
10.0
9.5
9.2
N,0
(1*8/1)
0.1.6
0.58
0.71
0.96
1.08
1.16
1.18
I
Sat
120
145
169
218
237
252
257
Date 11-25-77
Lat 19*02 'N
Long U7*48'W
Water
Tt»p
CC)
24.3
24.0
24.0
20.6
19.2
17.3
12.4
N20
(ug/1)
0.36
0.78
0.58
0.42
0.40
0.51
0.93
t
Sat
133
289
215
133
1M
146
227
Date 11-27-77
Lat ll'52'S
Long 142*05 'W
Water
Teoip
CC)
25.4
17.4
12.2
11. B
11.1
10.6
10.2
M,0
(l»g/D
0.41
0.83
2.48
2.83
2.09
2.10
2.31
I
Sat
152
237
605
690
486
477
519
Date 11-30-77
Lat 01'26'S
Long 150'10'W
Water
Temp
CO
27.6
27.5
27.5
25.8
19.0
13.6
13.4
HjO
(»S/»
0.48
0.34
0.39
0.40
0.93
2.08
1.35
I
Sit
190
135
1S4
149
282
527
342
Date 12-6-77
lat 25-40 'W
Long 167*33 'W
Water
Te«p
CC)
23.2
21. 8>
19.*
1ft. S
17.7
16.3
15.2
SjO
(»g/l>
0.36
0.40
0.40
0.48
0.42
0.46
0.49
I
Sat
124
133
123
143
120
127
130
Date 12-8-77
Lat 33*40'S
Long 170'50'W
Water
Temp
CO
18.5
14.6
13.5
12.7
12.2
11.7
10.7
HjO
(u«/O
0.40
0.44
0.50
0.58
0.59
0.58
0.65
%
Sat
119
144
125
141
»144
141
148
Date 12-11-77
Lat 39'05 'S
Long 179*13'W
Water
Te^>
CC)
16.9
16.8
14.7
13.7
13.4
13.1
13.0
KjO
(»g/D
0.42
0.47
0.41
0.47
0.47
0.4S
0.41
*
Sat
118
132
113
119
118
113
103

-------
between 0 and 300 m in the Pacific Ocean at 50-m intervals.  These data are
plotted in Figure 5.  It is clear from Figure 17 that the maximum N?0 concen-
tration below the top mixed layer was 1 to 7 times the surface N~0 concentration.
The most intensive synthesis of N..O was near the equator.  Near the southern
midlatitudes, the supersaturation was found to be marginal.  The average
variability of N20 saturation between 200- and 300-m depth with latitude is
also shown in Figure 16.

C.   Results and Discussion
     The atmospheric distribution of N_0 was found to be uniform around the
globe with an average global concentration of 311 ppb.  A number of investi-
gators have recently reported N.O concentrations that vary from 290 to 330 ppb,
and have be&n summarized by Liu et al. (1977).  This is largely because of
the unavailability of primary standards at this time.  A ±6 percent uncertainty
in the absolute concentration of N»0 currently exists.  Some of this uncer-
tainty (between 0 to 2 percent) can be attributed to the manner of reporting
data.  We have reported our data at 50 percent relative humidity and 25°C
while some have reported data at 0 percent relative humidity and yet others
have made no relative humidity correction (Rasmussen et al., 1976).  However,
the most important aspects of atmospheric measurements rest on the variability
of relative concentrations of N_0 rather than its absolute value.
     The small atmospheric variability of KLO (o  = 0.7%)  suggests a t^ Q of
greater than 20 years.  The T« Q is higher than 20 for two major reasons:
     •  All our data were collected within 15 m of sea level and can be
        expected to show a larger oa because of the close  proximity of the
        oceanic source.
     •  The atmospheric variabilities are approaching the  precision of the
        analysis and the true oa is probably less than 0.7 percent.
In our opinion, the past estimates of Tjq?Q of 2 to 10 years (Hahn and Junge,
1977) were derived  from measurements of less precision, which result in the
calculation of a large a  and a short TJJ o-
     Our data, collected over a period of about 3 years, while limited, suggest
an unchanging background of ^0.  There appears little doubt that during this
period (1975 to 1978) the change in N20 concentrations has been less than the

                                      43

-------
           % N,O SATURATION
   100  200  300   400   500   600  700
 60


100


160


200


260

300
  0

 GO


100


ISO


200


250

300
  0

 SO


100


ISO


200


250


300
 50 [i-

100

ISO

200
250 ~\
300
        T__T    (     r

//      LAT, OI°26'S
T I       LONG. 150°10'W
 \\
    I     I  J>
  TT
   I
   I
               LAT. 33°40'S
               LONG. 170°50'W


              I    I     I
             10         20
           TEMPERATURE -- "C
                                        % NjO SATURATION
                                tOO  200  300  400 ^00  600 TOO
                                                LAT. 41"3S'N    _
                                                LONG 12S"55'W
•  \r  r
\  .    LAT. 25°40'S
I  I    LONG. 167°33'W
ii
                                 ;   i
                LAT. 39°OS'S
                LONG. 174°13'W


                    I	I
       05   1.0   1.5   2.0   2.6  3.0   0
                                 30  0
                                    0.5    1.0   1.5   2.0   2.S  3,0
                                            N2O (j-B/t)
                                          |	1	)
                                          10        20        30
                                        TEMPERATURE — °C
     FIGURE  17   DISTRIBUTION  OF N20 IN THE PACIFIC
                                 44

-------
atmospheric variability of N~0.  This is in disagreement with the upper limit
of a 4 percent projected increase in N_0 abundance during this time because of
nitrogen fertilizer applications (Crutzen, 1976).  It thus appears that the
lower limits of projected N_0 increases, which suggest essentially no change
during this period, are more likely to be reliable (Crutzen, 1976; Hahn and
Junge, 1977).
     Based on our measurements in the Pacific, we obtained an average N_0
flux of 36 x 10~14 g/cm^/s and an average saturation of surface water of
133 percent.  Hahn (197A) reported a saturation of 123 percent in the Atlantic
between 38°N and 50°N.  This can be compared with a 118 percent saturation
measured by us between 41°N and 46°N.  Comparison of data between the Pacific
and Atlantic (Hahn, 1974) would suggest that our measured N~0 saturation of
133 percent is reasonably representative of global waters between 45°N and
45°S.  Additional data are required to make this generalization, but based on
currently available information this assumption does not seem to be unreason-
able.  In the rest of the world ocean body we assume an average saturation
of 115 percent based on Hahn's measurements.  Using the average 133 percent
saturation for the 74 percent of the seawater between 45°N and 45°S, and
115 percent for the remaining ocean surface, a net N~0 exchange of 30 million
tons  (N90)/year can be calculated.  This estimate of oceanic N?0 flux is less
than half of the best estimates of 70 million tons (N,,0) /year derived by Hahn
and Junge (1977) .
     We wish to further add that this flux estimate is critically dependent
on the film thickness used in the film model.  While we have used a film
thickness of 60 pm for comparative purposes (Hahn and Junge, 1977), the
possibility that this film thickness could be .as high as 90 urn cannot be
ruled out.  Therefore, our estimated oceanic flux would be between 20 and
30 million tons (N-O^year (or 13 and 19 million tons (N)/year).  This source
alone allows a turnover time of 75 to 110 years  for atmospheric N~0.  Despite
the uncertainties, our findings support the conclusion that oceans are most
likely a net source of N20.  This is in agreement with the findings of Junge
et al. (1971) and Hahn (1974) but in disagreement with the assertions of
McElroy et al.  (1976) who extrapolated  the Pacific N»0 measurements of Craig
                                                    ^         V
and Gordon (1963) to world oceans to conclude that oceans are probably net
sinks of N~0.
                                      45

-------
     The high N_0 synthesis in equatorial waters is not surprising because
microbial activity is known to be most pronounced in the warm equatorial
regions.  What is surprising, however, is the amount of N?0 synthesis in the
region of the Pacific Ocean where oxygen content is typically 4 to 4 to 5 ml/1,
or an order of magnitude higher than required for efficient denitrification
(liahn and Junge,. 1977) .  This high CL content could, however, lead to efficient
nitrification processes.  It would thus appear that either nitrification
processes play a more important role in N-0 synthesis in the ocean than
hitherto believed, or denitrification processes are poorly understood and can
proceed under conditions that are significantly different from anoxic.

-------
                                  SECTION  7
              FLUOROCARBONS-12  AND -11  IN  THE  GLOBAL ATMOSPHERE

A.   Global Emissions
     Fluorocarbons-12 and -11 are the two most dominant fluorocarbons in use
and have received the most attention.  Table 14 shows the global emissions of
these chemicals since 1931, when they first came into commercial use.  The
emissions data are considered accurate to ±5 percent (E. I. du Pont de Nemours,
1978).  From Table 14 it is obvious that by the end of 1977 about 5.13 million
tons of F-12 and 3.45 million tons of F-ll had been released into the atmo-
sphere.   Because of delays associated between  production and release, the
cumulative emissions at the end of 1977 were about 86 percent of the produc-
tion for both F-12 and F-ll.  We should emphasize that it would only be a
matter of time before the remaining 14 percent would be released.  The best
current estimate also suggests that 96 percent of the total release of fluoro-
carbons-12 and -11 occurred in the northern hemisphere and the remaining 4
percent in the southern hemisphere (Baur, 1978).

B.   Global Burden and Distribution
     FiRure 18 shows the global distribution of F-12 and F-ll for late 1977
and a third order polynomial fit to the data.   The equations describing the
polynomial fit are given in Table 15.  The average concentrations within each
hemisphere are also shown in Figure 18.  The northern hemisphere average con-
centration of 230 ppt for F-12 is about 9.5 percent higher than the southern
hemisphere average value of 210 ppt.  Interestingly enough, the northern hemi-
sphere average concentration of 133 ppt for F-ll is also 10 percent higher
than the average northern hemisphere concentration of 119 ppt.  The global
average burden of F-12 and F-ll by the end of 1977 was represented by tropo-
spheric concentrations of 220 ppt and 126 ppt, respectively.  These should be
compared with the global average concentration of 245 ppt for F-12 and 145 ppt
                                      47

-------
                              Table  14

              ANNUAL GLOBAL PRODUCTION AMD  RELEASE OF
            FLUOROCARBONS-12  AND -11  (In Units of 106kg)
Year
1931
1932
1933
1934
1935
1936
1937
1936
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
F-12* f
Production
0.5
0.1
0.3
0.7
1.0
1.7
3.1
2.8
3.9
4.5
6.3
5.9
8.2
16.7
20.1
16.6
20.1
24.8
26.1
34.6
36.2
37,2
46.5
49.1
57.6
68.7
74.2
73.4
87.6
99.4
108.5
128.1
146.4
170.1
190.1
216.2
242.8
278.8
311,4
336.9
360.5
401.7
447.5
473.6
419.7
449.8
424.4
5934.7
Released
0.0
0.0
0.1
0.2
0.2
0.4
0.6
0.9
1.3
1.7
2.3
2.9
3.5
4.7
6.1
11.9
19.0
22.2
24.2
27.1
30.2
31.5
35.5
4
-------
                              Table 15

 LEAST-SQUARES ERROR COEFFICIENTS OF A THIRD-ORDER POLYNOMIAL USED
      TO DEFINE THE GLOBAL DISTRIBUTION OF TRACE CONSTITUENTS
Compound
N20
CC12F2 (F12)
CC13F (Fll)
CCUFCC1F. (F113)
2 1 2
CCIF-CCIF- (F114)
i i
CHC12F (F21)
SF6
CCIA
CH3CC13
CH3C1
CH3I
CH4
C2H6
Coefficients of Polynomial*
a
3.112 (2) I"
2.206 (2)
1.290 (2)
1.787 (1)
1.153 (1)
4.212
0.292
1.198 (2)
8.971 (1)
6.190 (2)
1.711
1.409 (3)
0.769
b
-1.451 (-2)
0.363
0.498
5.249 (-2)
5.591 (-2)
2.952 (-2)
9.221 (-4)
4.412 (-2)
0.818
0.697
-9.422 (-3)
0.828
9.926 (-3)
c
-3.325 (-4)
1.096 (-3)
-3.178 (-3)
2.346 (-4)
-4.059 (-4)
5.583 (-5)
-5.817 (-6)
6.337 (-4)
7.584 (-4)
-3.128 (-3)
6.931 (-4)
4.435 (-5)
6.526 (-5)
d
-8.675 (-8)
-1.574 (-5)
-8.510 (-5)
-6.662 (-6)
-7.693 (-6)
-3.038 (-6)
1.370 (-7)
4.967 (-6)
-7.894 (-5)
-1.204 (-4)
-5.878 (-6)
-8.113 (-5)
5.561 (-8)
'The polynomial is Y - a+bL-fCL^-WL^, where Y is the concentration in
 ppt for all species except N£0, 0*4, and C2H$, for which it is ppb.
 The independent variable L is the latitude in degrees and varies
 from -90° to +64° (N.H. - 0 to +90°, S.H. - 0 to -90°).

f3.112 (2) - 3.112 X 102.

-------
                     600
                     400
                     200 "
                           210* (25.1)t ppt
230* (25,51* ppt
                     300
                   a
                   a
                     200
                     100
                           119* (11.71* ppt
133* (13.4)* ppt
                      -90°-80C -60° -40°  -20"  0°   20"  40° 60'  80° 90°
                        S          LATITUDE — degrees          N      .
                        Average hemispheric concentration       -
                        *Standsrd deviation               .  :.
           FIGURE 18   GLOBAL DISTRIBUTION OF FLUOROCARBONS-12 AND-11

for F-ll that  would result if the cumulative  emissions were to be  uniformly
mixed  in the global atmosphere.  Clearly,  our data would suggest  that some
loss has indeed  taken place.

C.   Atmospheric Growth
     It is also  clear from Figure 18 that  within each hemisphere  the  concen-
tration of F-12  and F-ll is essentially  uniform.  Thus, any latitude,  especially
the temperate  latitudes from 30-60°,  can be adequately used to characterize
the hemisphere.   Figure 19 shows the average growth rate of F-12  and  F-ll
based on several field studies conducted between 35 and 45°N latitude.   The
dotted lines in  Figure 19 represent  the  burden and growth in the  northern
hemisphere that  would result from man-made sources in the absence of  any
removal mechanisms.  It is clear from Figure 19 that the atmospheric  burden
of F-12 and  F-ll between 1975 and 1978 increased at an average rate of 22
ppt/year and14  ppt/year, respectively.  Using late 1975 as the base,  this
represents an  average linear growth  rate of 11 percent/yr and 13  percent/yr
                                       50

-------
  300


  250


  200


  150


  100


   50


    0

  200



  150
AVERAGE GROWTH RATE - 22 ppt/yr
TF-12 " 65 T0 70 YEARS; Tt - 1.1 TO 1.2 YEARS
g
I
!  100
   50
                  AVERAGE GROWTH RATE • 14 ppt/yr
                  Tp.,, • 40 TO 45 YEARS; T, -  1.1 TO 1.2 YEARS
                  10
                         15
                         TIME
               20     25
               months
30
                                                    35
  FIGURE 19  ATMOSPHERIC GROWTH AND RESIDENCE TIMES
             OF FLUOROCARBONS12 AND 11
             Time period beginnning 1 November 1975
                           51

-------
for F-12 and F-ll, respectively.  The  corresponding emissions burdens increased
                                                                     4
similarly.  The difference between  the dashed and the solid lines in Figure 19
is indicative of  the amount  lost  to atmospheric sinks.

D-   Sources and  Sinks
     While the sources of F-12 and  F-ll have been adequately characterized,
the sinks are still a matter of considerable uncertainty.  One of the major
reasons for this  uncertainty has  been  because of insufficient and often con-
flicting data on  the global  distributions of these species.  Calculation of
the residence time of fluorocarbons is critically linked with this distribu-
tion and is an essential input into models  that attempt to simulate the
stratospheric 0-j  depletion problem.  In addition, inert species such as
fluorocarbons and SFg can be used as tracers to characterize quantitatively
some of the fundamental features  of global  circulation.
     In this study we were able to  provide  needed data that substantially
alleviated the problems.of the past and allowed us to develop quantitative
source-sink relationships.

     1.   Atmospheric Residence Times
     Before we proceed to calculate the atmospheric residence times of F-12
and F-ll, it is essential to state  that both these species are completely
man-made in origin and no natural sources are known or suspected to exist.
The growth of these chemicals in  the atmosphere, as shown in Figure 19, is also
consistent with an exclusively man-made source.  Samples from the past decades,
.although very scarce, further confirm  this  well-accepted notion that fluoro-
carbons-12 and -11 are not a part of the unperturbed natural atmosphere.
     The only known destruction mechanism for F-12 and F-ll is photolysis in
the stratosphere.  F-12 and  F-ll  are not photolyzed in the troposphere and do
not react with free radicals.  No tropospheric sinks of any significance are
either known or postulated.  However,  it has always been speculated that as
yet unknown removal mechanisms for  F-12 and F-ll may exist and prevent these
chemicals  from  entering  the  stratosphere  in significant quantities.  Data  col-
lected in this study provided the best means to assess the question of unknown
sinks by the determination of residence times from budget and growth considera-
tions.
                                      52

-------
     It is clear from Figure 18 that both F-12 and F-ll are relatively uni-
formly distributed in the northern and southern hemispheres.  This allows the
use of a simplified two-box model to describe the burden of F-12 and F-ll in
the northern hemisphere and the southern hemisphere.  The two-box model can be
mathematically described as follows:
     If N and S are the cumulative masses and tin and Tis are the
     residence times (years) of any given species i at time T in the
     northern and southern hemispheres, respectively, then
dN,
                                          (Nrsi)
                                        _
                         dt     in   T,       T
                                     in      e
                        dS,
                              F,
                         dt  -  is                    >
                                     is       e

     where Te is the interhemispheric exchange rate and F^n and T?±a
     describe the emissions of species i in the northern and the south-
     ern hemispheres, respectively.
     In the case of fluorocarbons-12 and -11, it is reasonable to assume
     that T^n = ^iS'  Adding equations (2) and (3) we get

                                dM         M
                                — - = F  -- -                             (4)
                                 dt     i   T                              v '
     where M.^ * Nj + S±'t F.^ = Ffn + FIS, and t± « tin « t^g.  In the
     case of F-12 and F-ll, the initial condition is Mi « 0 at t = 0
     (year 1930).
     The global emissions data for F-12 and F-ll  (F^) is accurately known
 (±5 percent) and is reported in Table 14.  Equation  (4) can be easily solved
 with the help of a computer program for known values of i±.  The selected TJ|_
 is the value when modeled and measured estimates of  the F-12 and F-ll atmo-
 spheric burden agree.
     Since most of the measurements we made were conducted in the northern
 hemisphere, we show these results for the northern hemisphere in Figure 19.
                                      53

-------
The quantity N^ is estimated by using Equation 4 and the measured distribution
of F-12 and F-ll in the northern and southern hemispheres (Np_i2 * 1.095 Sp_i2»
NF-11 = 1-118 SF_n).
     The solid line in Figure 19 shows the average growth based on data mea-
sured in the northern hemisphere, a region of essentially uniform concentra-
tion.  The dashed line indicates the concentration in the northern hemisphere
that would result if the cumulative emissions were distributed globally accord-
ing to the distribution shown in Figure 18.  The difference between the dashed
and the solid lines is indicative of the amount lost to atmospheric sinks.
The circles indicate the predicted concentration in the northern hemisphere
from emissions data and the range of indicated residence times.  It is clear
from Figure 19 that residence times of 65 to 70 years for F-12 and 40 to 45
years for F-ll best fit the observational data.

     2.   Calculation of fa Mean Interhemispheric Exchange Rate
     Once the residence times for F-12 and F-ll are calculated, it is easy to
solve Equations (2) and (3) simultaneously to determine An average value for the
interhemispheric exchange rate (Te).  Here we compared the measured distribu-
tion of F-12 and F-ll in the two hemispheres with that predicted by Equations
(2) and (3).  Since our interhemispheric data was obtained in late 1977, conver-
gence was sought only for that time.  As should be expected, the model did
reflect some changes in the north-south distribution of F-12 and F-ll with
time.  This was mainly because of the changing emission patterns of F-12 and  ,
F-ll.
     The north-south distribution of F-12 and F-ll shown in Figure 18
(NF-12/SF-12 = l-°95; NF_II/SF_-Q = 1.118) were best described by a Te of 1.1
to 1.2 years, as shown in Figure 19.  This estimate of Te is in good agreement
with recommended estimates of Czeplak and Junge (1974) but in disagreement
with other studies that calculate a Te of 2 to 4 years (Czeplak and Junge,
                                                                          y.
1974).  Thus, the fluorocarbon data suggest that the northern hemisphere and
southern hemisphere air masses are exchanged in a relatively fast time of 1.2
years.
                                      54

-------
     3.   Oceanic Sink
     It has also been postulated that fluorocarbons-12 and -11 may find a
potential sink in the oceans.  During our Pacific trip we analyzed a limited
number of surface-water samples to assess the role of oceans, in a preliminary
fashion, as a source or a sink for halocarbons.
     Table 16 shows data for the concentration of F-12 between 0- and 300-m
depths at intervals of 50 m.  The average surface-water halocarbon concentra-
tions as they vary with latitude (0- to 2-m depth) and the 300-m concentration
data are summarized in Table 17.  The average measured surface-water concentra-
tions of F-12 and F-ll were 0.28 and 0.13 ng/liter, respectively.  With the*
surface-water concentrations of halocarbons known, a simple film-diffusion Tiodel
described in Equation 1 can be applied to estimate the flux of halocarbons
into or out -of the ocean.  Solubility data for F-12 and F-ll'from Junge (1976)
suggest that if the surface water is in equilibrium with the atmospheric burden,
the concentrations of F-12 and F-ll in water should be about 0.05 and 0.06
ng/liter, respectively.  These concentrations are lower than the measured
average concentrations of 0.28 and 0.13 ng/liter, so it appears that ocean
water is supersaturated with F-12 and F-ll.  This means that either the solu-
bility data are inaccurate or the water samples were inadvertently contaminated.
It is also possible that the ocean surface waters have been contaminated by
man-made activities on a global scale.  The lowest concentration of F-12 and
F-ll measured (Table 17), 0.07 ng/liter, is about what one would expect if  the
surface water were saturated with F-12 and F-ll.  If the surface water were
saturated, the ocean would be a relatively ineffective sink for F-12 and F-ll
but could act as a reservoir containing less than 0.5 percent of the atmospheric
burden of F-12 and F-ll in a steady-state situation.

E.   Discussion of Results
     Atmospheric residence times of 65 to 70 years for F-12 and 40 to 45 years
for F-ll best fit the observational data.  These long residence times can be
accounted for from the stratospheric photolytic loss alone, and rule out the
possibility of any significant missing sinks that may prevent these fluoro-
                                                              f>
carbons from entering the stratosphere almost  in their entirety.
                                      55

-------
                                                        Table 16




                                        F-12 CONCENTRATIONS IN THE PACIFIC OCEAN
Depth
(m)
0
50
100
150
200
250
300
Date 11-25-77
Lat 19e02'N
Long 137°46'W
ng/1
0.40
0.20
0.32
0.15
0.15
0.22
0.20
Temp
24.3
24.0
24.0
20.6
19.2
17.3
12.4
Date 11-27-77
Lat 11°52'N
Long 142°25'W
ng/1
0.32
0.23
0.23
0.56
Q.42
0.24
0.07
Temp
25.4
17.4
12.2
11.8
11.1
10.6
10.2
Date 11-30-77
Lat 01°26'S
Long 150° 10 'W
ng/1
0.35
0.25
0.28
0.33
0.31
0.24
0.24
Temp
27.6
27.5
27.5
25.8
19.0
13.6
13.4
Date 12-6-77
Lat 25°40'S
Long 167 °33'W
ng/1
0.25
0.20
0.20
0.25
0.25
0.45
0.45
Temp
23.2
21.8
19.4
18.5
17.7
16.3
15.2
Date 12-8-77
Lat 33°40'S
Long 173° 50' W
ng/1
0.19
0.20
0.22
0.25
0.24
0.27
—
Temp
CO
18.5
14.6
13.5
1.297
12.2
11.7
10.7
Date 12-11-77
Lat 39°05'S
Long 179° 13 'W
ng/1
0.53
0.18
0.10
0.09
0.09
0.09
0.09
Temp
CO
16.9
16.8
14.7
13.7
13.4
13.1
13.0
Ul

-------
                     Table 17




CONCENTRATIONS OF HALOCARBONS IN PACIFIC SEAWATER
Date
11-25-77
11-26-77
11-27-77
11-28-77
11-29-77
11-30-77
12-01-77
12-02-77
12-04-77
12-06-77
12-07-77
12-08-77
12-11-77
Lat
19°02'N
15°43'N
11°52'N
06°35'N
02°35'N
01°264S
06°07'S
09°07'S
18°24'S
25°40'S
30°14'S
33°40'S
39°05'S
Long
137°48'W
140°05'W
142°25'W
145°14'W
145°14'W
150°10'W
153°13'W
155°75'W
161°42'W
167°33'W
171°18'W
173°50'W
179°13'W
Water
Temp
(°C)
24.3
25.5
25.4
28.4
26.5
27.6
28.1
28.8
26.7
23.2
21.5
18.5
16.9
Average Surface (0-2 m) concentration
Average 300-m Concentration
Concentration
(ng/1)
F12
0.40
0.32
0.32
0.47
0.07
0.35
0.14
0.07
0.14
0.25
0.40
0.19
0.53
0.28 ± 0.15
0.21
Fll
--
--
--
--
--
--
--
0.15
0.24
0.07
0.10
0.10
0.12
0.13 ± 0.06
0.06
cci4
__
. __
--
--
--
--
--
0.41
0.45
0.35
0.38
0.40
0.40
0.40 ± 0.03
0.15
CRjCl
_ _
--
--
--
--
--
--
34.5
85.8
21.2
1.4
12.5
5.3
26.8 ± 31.2
3.3
CHC13
_ _
--
--
--
--
--
--
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05

-------
     Jesson et al. (1977) have calculated a F-ll residence time of 10 to 20
years and implied missing sinks for F-ll.  We attribute this difference to
the use of southern hemisphere data as measured by Lovelock et al. (1973) in
1972, in which the southern hemisphere concentrations of F-ll were only half
the northern hemisphere values.  Our measurements of the global distribution
of five fluorinated species confirm that the southern hemisphere concentrations
of all inert species, including F-12 and F-ll, are only about 10 percent lower
than the northern hemisphere values and, contrary to the findings of Lovelock
et al. (1973), no gradients within each hemisphere exist.  In addition,
analysis of Lovelock's F-ll data would suggest a te of more than 2 years.
This is not supported by available information (Czeplak and Junge, 1974).
Thus, overall budget considerations do not allow for the postulation of major
sinks for F-12 and F-ll, and the oceans certainly do not appear to be such a
sink.  The reasons for the supersaturation of oceanic waters with F-12 and
F-ll currently cannot be understood.
                                      58

-------
                                  SECTION 8
                FLUOROCARBONS-22, -113, -114, -21 AND SF6 IN
                           THE GLOBAL ATMOSPHERE
A.   Global Emissions
     The emissions data for F-22, the third most commonly used fluorocarbon,
is given in Table 18.  By the end of 1977, 0.93 million tons of F-22 had been
produced and about 0.43 million tons or 46 percent had been released.  Unlike
F-12 and F-]l, nearly all of F-22 is used as a refrigerant.  In such applica-
tions, long delay times between production and release can be expected.  The
remaining 54 percent of F-22 produced would eventually be released to the
atmosphere.
     Accurate emissions data for F-113, F-114, F-21 and SFg are currently not
available.  Production estimates suggest that cumulative F-113 production by
the end of 1977 was about 0.7 million tons, and about 0.6 million tons of
F-113 were released.  Since almost all of the F-113 is used in solvent appli-
cations, there are no long-term delay periods in production and end usage.
F-114, on the other hand, finds a major application as a refrigerant.  We
estimate that by the end of 1977 about 0.3 million tons of F-114 had been
produced and about 0.25 million tons had been released.  Fluorocarbon-21 is
u very minor industrial product.  Release estimates through 1977 from known
product and by-product sources are probably less than 0.6 million kg, though
large unknown sources of F-21 may exist.  The total SFg released to date can
be roughly estimated to be somewhere between 20 and 40 million kg.

B.,   Global Burden and Distribution
     The global distribution of F-22 has not yet been characterized because
of measurement difficulties.  No atmospheric data on F-22 are available from
any source.  Our limited attempts involving preconcentration of large amounts
of air samples during this study suggested F-22 levels in the neighborhood of
20 to 30 ppt in clean background air.  Poor sensitivity of F-22 to electron

                                      59

-------
                       Table 18
ANNUAL GLOBAL PRODUCTION AND RELEASE OF FLUOROCARBON-22
                 (in units of 106 kg)
Year
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
Production
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0,0
0.1
0.0
0.1
0.2
0.3
0.8
1.0
1.6
2.2
2.9
3.7
6.1
6.3
7.6
11.2
12.2
12.2
15.4
17.6
22.4
25.1
31.5
37.3
45.6
55.9
58.6
64.5
70.0
76.5
87.7
73.6
94.1
101.9
946.2
Release
0.9
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0 .
0.0
0.1
0.1
0.2
0.3
0.4
0.6
0.8
1.2
1.1
1.4
2.7
3.3
3.6
4.6
5.6
6.8
8.1
9.8
12.3
16.3
20.3
23.2
26.2
30.9
35.7
41.8
47.1
60.2
65.2
429.9
        Source:  Bauer (1978)

                          60

-------
capture has made its measurement difficult.  It also appeared that F-22 was
significantly destroyed on Ascarite traps.  At this time no reliable atmo-
spheric data base for F-22 exists.
     Figure 20 shows the global distribution of F-113, F-114, F-21, and SFg,
and illustrates a third order polynomial fit to the data.  (The nature of the
polynomial is described in Figure 15.)  The global average concentrations of
F-113, F-114, and SFg are 18, 11, and 0.3 ppt, respectively.   These values
are comparable to the estimated cumulative emissions of these species.  An
accurate emissions inventory for these species is not available, but their
residence times can be expected to be several decades.
     The stratosphere is expected to be the major sink for F-113 and F-114.
Sulfur hexafluoride (SF$) is transparent to ultraviolet in both the tropo-
sphere and the stratosphere.  Continued emissions of, and increases in, SFg
could result in the deposition of sulfur above 50 km (destruction via electron
capture) with unknown effects.  Unlike the fully halogenated fluorocarbons,
F-21 is quite reactive and we estimate its lifetime to be about 4 years (see
Section XVI).  The northern hemisphere and southern hemisphere concentrations
of 5 ± 3 ppt and 4 ± 1 ppt would not be consistent with a northern hemisphere
source.  It remains unclear at this time whether F-21 is a true atmospheric
constituent or an unknown artifact.

C<   Discussion of Results
     As estimated in Section V1I1-A about 0.43 million tons of F-22 had been
released to the atmosphere by the end of 1977.  In the absence of any sinks
this would correspond to a background concentration of about 29 ppt.  However,
no measurements of F-22 in the global atmosphere are available.  Our measured
value of 20 to 30 ppt in the northern hemisphere is comparable to the esti-
mated emissions.  F-22 is expected to have a sink in the troposphere via reac-
tion with HO.  However this removal process is extremely slow and can provide
a residence time of about 35 years in the troposphere (see Section XVI).
     No tropospheric sinks for F-113 and F-114 of any significance are known.
Their structural similarity to F-ll and F-12 would also preclude the existence
of dominant sinks.  Photochemical destruction in the stratosphere is the only
known removal mechanism.  The atmospheric residence times of F-113 and F-114
                                      61

-------
          ..  60

          8


          I   40
             20




              0
18* (3.11* ppt
19* (3.5)* ppt
          i
          i
          u.
             30.
             20
             10
                      10* (1.31* ppt
                        12* (1.9)* ppt
             30
             10
                      4* (1.0)* ppt
                      5* (2.6)f ppt
            0.6
             0.4
          (0
         u.
            0.2
                     0.27* (0.01 )f ppt
                      0.31* (0.04)* ppt
           S -90° -80°  -60°  -40°  -20°   0°   20°   40°  60°  80° 90° N

                               LATITUDE — degrees



          Average hemispheric concentrations


         'standard deviation




FIGURE 20    GLOBAL DISTRIBUTIONS OF F-113, F-114, F-21, AND SF6
                                    62

-------
can bo expected to be greater than 40 years.  The cumulative emissions of
K-113 and F-114 are estimated to be 0.6 million tons and 0.3 million tons,
respectively.  If these were to be globally distributed instantly,  an average
background concentration of 19 ppt for F-113 and 9 ppt for F-114 would result.
Tins is in excellent agreement with the measured global average concentration
of 18 ppb and 11 ppt, respectively, for F-113 and F-114.  The agreement is
somewhat fortuitous since the emissions data for F-113 and F-114 are rather
approximate.
     SFg is present in extremely small quantities and is an excellent tracer
of atmospheric motions.  We do not expect SFg to have any significant sinks
in the troposphere or the stratosphere.  Unlike chlorinated species, fully
fluorinatod species (e.g., SFg, CF^) appear to be transparent to UV even in
the stratosphere.  The major sink for SFg is likely to be because of destruc-
tion by electron absorption in the ionosphere.  Injections of sulfur and/or
fluorine in the ionosphere may have an unknown environmental impact.  In the
meantime, the atmospheric burden of SFg continues to increase at a rate of
5 to 10 percent/year.
     Since the first reported measurement of F-21 (Singh et al., 1977c), some
researchers have speculated whether or not F-21 is a product of F-ll decompo-
sition in the troposphere.  This possibility has been advanced because of two
major considerations:
     •  The direct emissions of F-21 appear to be so low that they could not
        account for more than 0.1 ppt of atmospheric F-21.
     •  Laboratory experiments do show some thermal and photochemical decom-
        position of F-ll on sand with F-21 as one of the products of decom-
        position.
Both of these observations are based on qualitative evidence that is not easy
to extend to the atmosphere.  Indeed, secondary sources of F-21 may exist but
have not yet been adequately characterized.  Recent experiments conducted in
Italy have used mass fragementometry to confirm the identity of F-21.
     It is therefore acknowledged that more information on F-21 is needed to
characterize adequately its role in the atmosphere.  In the meantime, it is
possible to estimate the significance of F-21 as a sink for F-ll by using a
simplified model.
                                      63

-------
                                  F-21 -->   unknown
                                               products
Assuming F-21 to be in a steady state one can calculate that
                                                 •   I
                             TF-11 = ^2   [F-ll]
                                   -    -
            f
In 1977 the average background concentration of F-ll was 126 ppt and that of
F-21 was 4.5 ppt.  Tp-21 has been estimated to be 4.3 years (see Section XVI).
Inserting these in Equation (5) we can calculate a tropospheric F-ll residence
time (TJ; -,) of 125 years.  Thus, even if the postulate that F-21 is a product
       r — 1JL
of F-ll decomposition is correct, this tropospheric sink would be relatively
small.   The possibility that F-21 is truly an atmospheric constituent that is
represented by our measurements is yet to be convincingly demonstrated.

-------
                                  SECTION  9
                      ATMOSPHERIC CARBON TETRACHLORIDE

'^ •    (' 1 o l)a 1 Em i ss 1 ons
     Unlike the fluorocarbons, the sources of atmospheric CC2-4 have been a
matter of considerable uncertainty.  Lovelock and coworkers (1973) were the
first to suggest that CCl^ must be of natural origin.  These conclusions were
based on two observations:
     •  Their inability to uncover significant man-made emissions.
     •  The*global distribution of CCl^ was much more uniform that F-ll
Thus, up until late 1975 the emissions of CC14 were believed to be negligibly
small and it was considered to be a species of natural origin.  When attempts
were made to develop an emissions data base it became clear that large amounts
of CCl^ had been released to the atmosphere between 1930 and 1960.  Indeed,
it appeared that CCl^ was predominantly of man-made origin.  Three almost
simultaneous and independent attempts at quantifying the sources of CCl^
(Altshuller, 1976; Singh et al., 1976a; Galbally, 1976) arrived at nearly
identical conclusions.  Here we shall describe our results (Singh et al.,
1976a) with the proviso that these are not significantly different from the
results of the other two investigators.
     Figure 21 shows the yearly production and release data for CCl^ for the
United States, Western Europe, and Japan.  The global emissions were estimated
to be 1.1 times the emissions from these  three regions.  Details of these
estimates and the sources of information have already been published (Singh
et al., 1976a).  Table 19 has been constructed from the same data used to plot
Figure 21.  When year to year details were not available, data between years
was interpolated.  It is clear from Figure 21 and Table 19 that an estimated
3,04 million tons of CCl/^ has been released  to date and over  75 percent of
these emissions have originated in the United States.  It should also be added
that secondary sources of CCl^ are known  to exist but are much harder to
                                      65

-------
500
450
400
350
300
250
200
150
                  U.S. CCU production  (curve 1)
    ••—«• Japanese  CCI4  production (curve  2)
    • •••• Western European CCU production (curve 3)
    1     Total  U.S. dispersive  usage  (curve 4)
    >ooooo Total U.S.  emissions (curve 5)
o—o—o Total  Japanese emissions (curve  6)
•—•—• Total  European emissions (curve  7)
100
                LJJ  U.S. agricultural fumigant usage
       •J U.S. fire  extinguisher usage
       3 U.S. dry-cleaning and spotting usage
                J  U.S. industrial solvent usagt
  1910      1920       1930     1940       1950      1960      1970
          FIGURE 21   PRODUCTION AND EMISSION TRENDS FOR CCI,
                                   66

-------
                               Table 19

                 ANNUAL GLOBAL CCL4 RELEASE SINCE 1910
                           (in 106 kg/year)
Year
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
J934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
CC1, Release
0.0
0.5
1.1
.1.6
2.2
3.3
4.4
5.0
5.5
6.6
7.7
8.8
9.9
10.0
10.1
11.6
13.2
14.8
16.5
18.2
19.8
20.9
22.0
25.3
28.6
31.9
35.2
40.7
46.2
52.2
58.3
63.8
69.3
72.4
75.5
Year
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
CC1, Release
75.0
74.5
72.6
70,6
67.9
65.2
63.0
| 60.8
I 58'6
56.5
54.8
53.2
52.7
52.2
53.0
53.9
54.4
55.0
56.6
58.3
62.2
66.0
68.8
71.5
75.4
79.2
82.0
84.7
89.1
94.0
85.0
81.0
82.0
3037.3
Source:  Singh et al.  (1976)
                                   67

-------
quantify.  Chlorination  of  water and accidental spillage of  large  amounts  of
CCJ^, which are often  not  reported, are only two sources that are  not  included
in Table 19.  The  possibility that some 8 percent of the fy^h  released  to
the atmosphere may be  converted to CC14 has also been proposed  (Singh  et al.,
1975) based on laboratory  smog-chamber data.  This source alone could  account
for up to 0.7 million  tons  of CCl^.  However, the applicability of these
laboratory results to  the  real atmosphere are still in question and therefore
not included in Table  19.   We suspect that the CCl^ emissions inventory  could
be an underestimate by as  much as 30 percent.  No natural sources  of CCl^  are
known to exist.

B.   Global Burden and Distribution
     Figure 22 clearly shows that unlike fluorocarbons, CC14 is rather uni-
formly distributed with  a  northern hemisphere and southern hemisphere  concen-
tration difference of  2.5  percent.  The global average concentration of  120
ppt is indicated from  Figure 22.  This corresponds to an atmospheric burden
of 3.2 million tons.   (A third order polynomial used to fit  the global dis-
tribution of CCl^  is described in Table 15.)
                     150
                   g TOO
                   8  50
                            119* (4.0)f ppt
                      0 I	L.
122* (4,9)* ppt
                     S-90°-80°  -60° -40° -20°  0"  20°  40°  60°  80° 90° N
                                  LATITUDE — degrees
                    Average hemispheric concentrations
                   ^Standard deviation

             FIGURE 22  GLOBAL DISTRIBUTION OF CARBON TETRACHLORIOE
                                       68

-------
C.   Atmospheric Growth
     Figure 23 shows a small average growth rate of about  3 ppt/year  during
the period 1975 and 1978.  It appears from our measurements that  the  growth
rate of CCl^ is only about a quarter of the fluorocarbon growth rates.

D.   Sources and Sinks
     1.   Atmospheric Residence Time
     Since the emissions data for CCl^ are not highly accurate and all  the
sources are not characterized, it is not possible  to determine the residence
time of CCl^ directly.  Therefore, here we shall use direct information on
sinks to estimate the residence times of
     The global background CCl^ concentration of  120  ppt  that  we measured
corresponds*to an atmospheric burden of  3.2 million tons.   Tne known emissions
reported in Table 19 are estimated to be 3.0 million  tons.   Thus, an incon-
sistency exists and must be attributed to  inaccuracies  in the  emissions in-
ventory, largely from sources that currently cannot be  quantified.
              8
                200
                 150
                100
                 50
                                 AVERAGE GROWTH RATE > 3 ppt/yr
                               10     15     20     25
                                     TIME — months
30
35
          FIGURE 23  ATMOSPHERIC GROWTH OF CCI4 IN THE NORTHERN HEMISPHERE
                    Time period beginning 1 November 1975
                                      69

-------
     The stratosphere is by  far the- most impnrta.nt sink for CClv,.  The inert-
ness of CC14 in the troposphere is well known and its lifetime hero has been
estimated to be in excess of 300 years.  The stratospheric sink from photolysis
is estimated to be between 50 and 75 years.  Figure 24 shows the atmospheric
burden of CCl^ based on a 50-year residence time.  Curve 5 shows the net
calculated worldwide profile of CCl^ cumulative atmospheric loading, resulting
solely from emissions from anthropogenic sources with concurrent removal by
transport to the stratosphere and photolysis.       t
   *It is clear from Figure 24 that the cumulative build-up of CCl^ from
known anthropogenic sources up to 1973, was about 1.9 million tons or about
72 ppt.  When extrapolated to 1977 this predicted value would be about 2.1
million tons or about 80 ppt.  Thus, known emissions data, when coupled with
a 50-year atmospheric residence time, can account for 80 ppt of CCl^.  This
can be compared with an actual atmospheric abundance of 120 ppt.  Therefore,
about a third of the atmospheric CC14 is currently unaccounted for.  Further-
more, examination of the atmospheric accumulation rate of CCl^ (curve 5) and
that of F-ll (curve 6) reveals that 65 percent of the total CCl^ released to
the atmosphere was emitted before 1960, whereas 96 percent of the F-ll was
released after 1960.  This additional time available to CC14 for global dis-
persion and its slow release to the atmosphere explains the relative global
uniformity of CCl^ as compared with F-ll.

          a.   Other Evidence for Man-Made Sources of CC1&
     Another way to ascertain existing man-made sources of CCl^ is through
actual field measurements.  Based on measurements from California (site 2)
we clearly demonstrated instances of CCl^ increase that coincided with a
corresponding increase in F-12, the latter being an excellent indicator of
urban air masses.  These results were subsequently published (Singh et al.,
1977a).  The phenomenon, however, is not limited to California.  Figure 25
shows CC14 increases that coincide with F-ll increases for two days of data
from the eastern United States.  These data clearly support the view that
man-made CC14 sources still exist in the urban environments within the United
States.
                                      70

-------
4.0
             T
                          T
           T
3.6
3.2
2.8
•—•—•  U.S. cumulative CCU emissions (curve 1)
• •—*•—••  Japanese  cumulative CCU emissions (curve 2)
• ••••••••  Western European  cumulative CCU emissions (curve 3)
————-  Worldwide cumulative CCU emissions (curve 4)
«—•—o  Worldwide cumulative CCU buildup (curve 5)
—•<—• •——  Worldwide cumulative F-ll emissions (curve 6)
fc;'v7£y-V-.1  Net stratospheric CCU loss
          Worldwide CCU emissions
          = 1.1 (U.S.  +  Western Europe + Japan)
            1920
               1930
1940      1950
    YEAR
1960
                                                       1970
                                                                        1960
              FIGURE 24  CUMULATIVE CCI4 AND F-11 EMISSIONS
                                    71

-------
         800
                                                         JULY 18F-11
                                                     	JULY 18 CCI4
                                                     	 JULY 31 F-11
                                                     	JULY 31 CCI4
         100
                                                               21
                                                                    23
            FIGURE 25  IDENTICAL CCI3F AND CCI4 BEHAVIOR AS AN INDICATOR
                      OF A COMMON URBAN SOURCE
          b/   Oceanic Compared With Stratospheric Sink of
     At the time of our earlier analysis of historic CCl^ data  (Singh et al.,
1976a) the only known sink was the photolysis in  the stratosphere.  The esti-
mated CCl^'.residence time of 50 to 75 years (best value of  60 years) based on
the photolytic sink was also considered to be the overall residence time.
     In Table 17 we have shown the CCl^ concentrations in the Pacific seawater.
The average surface water concentration for CCl^ was 0.40 ng/liter.  The flux
of CC14 into the ocean can be calculated from Equation (1)  with D « 10"-' crtr-/
sec, Z «= 90 urn, and Sccl  = 0.85.  The solubility, S, in seawater is defined
as the ratio of the species concentration at the air-sea interface  (0^ ) to
the atmospheric concentration at standard temperature and pressure.  A high Z
is used because CCl^ is rapidly absorbed in fatty tissues and may be biologi-
cally active.  For such species, the upper limit of the stagnant film thick-
ness calculated from radon data (63 ± 30 \im) is more appropriate.  Using
Equation (1), we can calculate a CCl^ flux into the ocean of 2.8 * 10~16g
ctiT^/sec.  if we assume this flux to be typical of all oceans,  we obtain an
                                      72

-------
exchange rate of 3.2 x 1010g/year.  The atmospheric burden of CCl^ from our
measurements is calculated as 3.2 * lO^^g.  Thus, the ocean is a sink for
CCl^ that can provide a turnover rate of 100 years, as in

                                         12
                      T (CC1.) = 3'2 * 10.n = 100 years
                                 3.2 x 1Q1U

Our measurements, although preliminary because of a limited number of samples,
do suggest the possibility of an oceanic sink for CCl^ that is about half as
effective as the stratospheric sink.  Consequently, it would appear that an
overall CCl^ atmospheric residence time of 35 to 45 years is the best value.
Our use of the 50-year residence time may be a slight overestimate.

Et   Discussion of Results
     Based on our results, our understanding of the atmospheric fate of CCl^
has undergone a significant change.  Prior to our findings, it was believed
that CCl^ was of natural origin, even though no natural sources were known.
     More careful quantitation of CCl^ emissions has shown the following:
             appears to be essentially of man-made origin, and known emissions
        data, when coupled with known residence time information, can account
        for about two-thirds of the atmospheric CCl^ burden.  Large unquan-
        tified man-made sources of CC14 probably exist.
     •  The atmospheric release of CCl^ has been at a much slower rate than
        that of f luorocarbons over a longer period of time.  This has allowed
        CCl^ to be relatively uniformly distributed in the global atmosphere.
        The present atmospheric growth of CCl^ (2 to 3 percent/year) is almost
        entirely attributable to man-made emissions.
     •  It is probable that an oceanic CCl^ sink exists that is about half as
        effective as the stratospheric sink.
     It is pertinent to add that the emissions data for CCl^ is not of very
high quality and most likely underestimates the emissions by as much as 30
percent.  The arguments used to suggest a natural origin of CCl^ are no longer
valid.  The bulk of the evidence points to dominant man-made sources.
                                      73

-------
                                 SECTION 10
                        ATMOSPHERIC METHYL CHLOROFORM

     Methyl chloroform (CH.,CC1,) is a commonly used solvent and degreasing
agent.  It has undergone rapid growth in recent years because it was the
recommended substitute for trichloroethylene and tetrachloroethylene, whose
usage had been significantly curtailed for toxicity reasons and because of
their involvement in producing smog.
     Until recently, it was commonly believed that the atmospheric lifetime
of CH^CCl, was 1 to 2 years (NAS, 1976) and this was too long to cause photo-
chemical pollution but too short to allow significant amounts to enter the
stratosphere.  Based on the present study, the fate of CH,CCl.,..has become a
matter of considerable debate.  We analyzed our measurements in light of
available emissions data to demonstrate the role of CH.CC1- as an indicator
of hydroxyl* radical and to prove that the lifetime of CH-CCl, is sufficiently .
long to allow large quantities to enter the stratosphere.  In this section we
discuss the sources and sinks of CH.CC1,, and show how this man-made pollutant
provides a unique opportunity to study the chemistry of the natural atmosphere.
A.   Global Emissions                     •          f     •
     The global production and release figures for CH_CC1» are shown in
Table 20.  The data was provided by Dow Chemical Company and is expected to
he accurate to within ±5 percent (Neely and Plonka, 1978; Farber, 1979).
Table 20, however, does not include production and release figures from the
Soviet Union and East European countries.  These emissions are estimated to
be less than 5 percent of those reported in Table 20.  Plonka (Dow Chemical)
suggests that the production figures in Table 20 could be used as an upper
limit on thettotal global emissions of CH_CC1_.  We have used Plonka's
suggestion in the calculations that follow in the subsequent sections, since
this also leads to a conservative analysis (shortest residence time).
                                      74

-------
                           Table 20

ANNUAL GLOBAL PRODUCTION AND RELEASE RATES OF METHYL CHLOROFORM
                        (in 10& kg/yr)
Year
1951
1952
1953
19 5 A
1955
1956
1957
1958
VJ59
1960
1961
1962
1963
1964
1965
J966
1967
1968
1969
1970
1971
1972
1973
1974
1975
19/6
1977
Total
Production
i
0.1
0.2
1
1)
8
14
21
22
32
38
40
59
58
64
83
116
146
158
!I68
181
101
267
350
390
374
438
470
3692
Release
0.1
0.2
1
3
8
12
20
21
30
36
38
56
51
57
73
109
131
145
148
155
167
230
340
363
365
416
446
3421
             Sources:  Neely and Plonka (1978);
                       Farber (1979)

-------
     As is clear from Table 20, the growth of CH_CC1_ has been very rapid.
Between 1955 and 1977 an average yearly growth rate of about 16 percent
can be calculated.  In more recent years this growth rate has been closer
to 12 percent.  By the end of 1977 about 3.4 million tons of CH^Cl^, or
about 90 percent of production, had been released.  The remaining 10 percent
would be released with some delay time.  Our best estimate also suggests that
96 percent of CH.CC1., release took place in the northern hemisphere while the
remaining 4 percent occurred in the southern hemisphere.

B.   Global Burden and Distribution
     Figure 26 shows the detailed global distribution that was measured
during the conduct of this study.  Prior to this all measurements were point
measurements and most of these were in the northern hemisphere.  Figure 26
shows the global distribution of CH»CC1_ and a third-order polynomial fitted
to it.  (The nature of the polynomial is described in Table 15.)  This compound
shows a latitudinal distribution quite different from that of the fluorocarbons.
Up to about 30°N, the CH_CC1~ is well mixed in the northern hemisphere with
an average concentration of about 123 ppt.  A fairly sharp decline seems to
occur between 20°N and 20°S, and the concentration of CH-CC1- then levels
off to about 75 ppt.                                           .
     The weighted averages of CH«CC1» concentrations of about 113 ppt in the
northern hemisphere and 77 ppt in the southern hemisphere best describe the
burden of CH«CC1^ in each hemisphere.  It is obvious from Figure 26 that
while measurements at one point may be used to characterize the distribution
of inert species, such is not the case for relatively reactive species.  The
reactivity of CH~CC1~ is clearly demonstrated by its large north-south ratio
as compared to the ratio of F-ll and F-12.
     While no detailed CH-CC1- latitudinal profile has been available from
other sources, comparison of results from Figure 26 with point measurements
of Lovelock (1977) and Rasmussen et al. (1976) shows significant disagreement.
These investigators have reported a much larger north-south difference than
              i
indicated by our measurements.  Our results, however, are in much better
agreement with the more recent measurements of Rowland (1979), Krasneck (1979)
and Rasmussen (1979).
                                                            ^
                                      76

-------
200
i 150
;> 100
) i
o
5 50
0
77 ppt*
*"•--.. ..*4*&v*

i • i * i v i
113 ppt"
^" «.--••» fr •-
?«""*""
i i i , 1 . 1
                      -90-80' -60° -40"  -20"  0"  20°  40  60'  80° 90°
                        S          LATITUDE — degrees           N

                FIGURE 26  GLOBAL  DISTRIBUTION OF METHYL CHLOROFORM
                          'Corresponds to a uniformly distributed concentration
                          that is consistent with  the measured hemispheric burden.
C.    Atmospheric Growth
     The dark line  in  Figure  27  shows the measured growth of CH-CC1.,  during
1975 and 1978 in the northern hemisphere.  These measurements were  all  con-
ducted between  the  latitudes  of  35°N to 64°N, a region where the distribution
of CH-CCl., is essentially  uniform, as shown in Figure 26.  The  dashed line in
Figure 27 shows the growth of CH-CCln that would result if the  man-made emis-
sions were to be distributed  according to the distribution shown in Figure 26
and in the absence  of  any  removal mechanisms.  The dotted line  is deduced  from
the dark line and indicates the  global average concentration of CH-CC1.,.   The
average concentration  corresponds with the actual measured burden of  CH,,CC1-.
This is obtained by scaling the  northern temperate measured concentrations to
an average global concentration  according to Figure 26.  A northern temperate
CH..CC1., concentration  of  123  ppt corresponds to a northern hemisphere concen-
tration of 113  ppt, a  southern hemisphere concentration of 77 ppt,  and  a global
average concentration  of  95 ppt.  Thus, the global average CH-CC1-,  concen-
tration is about 0.77  times the  northern temperate region concentration.
     It is clear from  Figure  27  that the atmospheric burden of  CH-CC1-  in-
creased at an average  rate of 15 ppt/year («17 percent/year) during the course
of this study.  The available emissions data for CH-CC1- suggest an increase
of about 25 ppt/year.
                                       77

-------
                 200
                 150
               a.
               a.
              _P 100
              O
              O
               r)
              U
                 50
                                AVERAGE GROWTH RATE • 15 ppt/yr
                                TCH3C03 - 8 TO 11 YEARS
                               10
                                     15     20     25
                                     TIME — months
                                               30
35
              FIGURE 27  ATMOSPHERIC GROWTH OF METHYL CHLOROFORM
                        Time period beginning 1 November 1975
D-   Sources and Sinks of Methyl  Chloroform
     1.   Atmospheric Residence Time  of  Methyl  Chloroform
     The first interest in the residence time of methyl chloroform was gen-
erated by the qualitative possibility that  CH.CC1.  may be used as an indicator
of the. HO radical abundance.  During  this study we  soon recognized that the
global budget and distribution of methyl chloroform (CH-CC1-)  offer a unique
means of determining the average  HO radical concentrations in  the troposphere
(Singh, 1977).  The uniqueness stems  from three major considerations:
     •  Sources are entirely man-made and reliable  emissions data are
        available.
     •  Reaction with HO is the major removal mechanism and the rate constant,
        which is accurately known (NASA, 1977)  is relatively fast and not
        strongly dependent on temperature.
CH3CC1,
availat
                is easy  to measure  and  a relatively reliable data base is
                                       78

-------
The exclusively man-made sources of CH CC1_ were confirmed by two observations:
no natural sources are known, and measurements by Lovelock (1977) in the
southern hemisphere in 1970 indicated a background CH,CC1- concentration of
12 ppt consistent with man-made emissions for that period.  The preliminary
analysis by Singh (1977a; 1977b) clearly showed that atmospheric residence
times of CH-CC1- based on measurements were about a factor of 5 longer than
those predicted by models.  Thus, we reported that the residence time of
CH.CC1, was about 8 to 11 years, contrary to all earlier estimates which
placed the residence time between 1 and 2 years.  A longer residence time
meant that CH~CC1~ was a potential depletor of stratospheric ozone, but even
more importantly it meant that perhaps a large group of organic and inorganic
chemicals resided in the atmosphere1 significantly longer than previously
believed.  However, more reliable calculations of the CH~CC1, residence time
had to await the data shown in Figure 26, illustrating the detailed latitudinal
variation of CH,CC1_ that is by no means uniform.
     It is clear from Figure 26 that the application of the two-box model as
described by Equations 2 and 3  (Section VII-D-1) is somewhat of a simplifi-
cation since the concentrations within the northern hemisphere and southern
hemisphere are not uniform.  The exact simulation of the global CH«CC1- pro-
file would require a two-dimensional (2--D) global model, which is currently in
a state of active development.  However, as will be shown in this section,
valuable information is also gained by considering a two-box model.  Let us
define an average global residence time, T.
                 TCH3CC13 ~ Tia " T
                                      [-(I)]
Solving Equations 2 and 3 for several cases makes it clear that while T.  and
                                                                       is
T.  are quite sensitive to the average north-south ratio of CH»CC1-, T.  is
relatively insensitive to this ratio.  Thus T.  is determined by the budget
                                             13
rnther than by the distribution of CH-CC1-.
     The nort-hern temperate concentration  that would result from the emitted
CHALl- if there were no removal mechanisms is also shown in Figure 27 as the
dashed line.  Clearly a great deal of CH_CC1~ has been lost to atmospheric sinks.

                                      79

-------
     The application of the two-box model shows that the best estimates of
TCHjjCCll ^*e ketween 8 and 11 years.  The good agreement between measurements
and project values using an 8 to 11 year residence time is shown in Figure 27.
The large circles represent the global average concentration projected by
models.  Thus, despite the uncertainties, it is clear that the global average
        residence time is between 8 and 11 years.
     Subsequent to our findings a number of other investigators have supported
our results.  Table 21 provides a summary of the various estimates T_
                                                                      _s~
     Considering all of the results reported in^Table 21, it appears that the
best current estimates of CH»CC1_ residence time lie between 6 and 12 years,
excluding the outlier point from Neely and Plonka (1978) .  This should be
compared with the CH_CC1_ residence time of 1.4 years reported by NAS (1976).
A 6- to 12-year residence time allows 12 to 25 percent of the CH.CCl. released
at ground level to enter the stratosphere.  This long tropospheric residence
time, when coupled with the rapidly increasing emissions, suggests that
CH,CC1., may be a potential depleter of stratospheric 0~  in the decades ahead.
Worldwide release of CH.,CC1_ to the atmosphere currently approaches
8 x 10^-g/year and is increasing at a rate of 10 to 15 percent per year.
Continued, uncontrolled release of CH.CC1,, to the atmosphere should be a
matter of future concern.
     Another somewhat less certain finding also emerges  from the application
of the two-box model.  It appears that the residence times of CH.CC1, in the
northern hemisphere (T, ) are significantly longer than  in the southern
hemisphere (T . ,) .  The ratio of t. /T, , however, is quite sensitive to the
interhemispheric exchange rate (T ) and the northern hemisphere and southern
hemisphere concentration gradient of CH..CC1-.  Table 22  summarizes the results
based on measurements by several investigators.  Here we have selected the
best value of T  to be 1.2 years.  Lower values of T  increase the T.  and T,
               e           }                        e               in      is
asymmetry, while higher values tend to reduce it.
     Our results from Table 21 and 22 can be summarized  as follows:
  (TCH3CC13) NH +  (-rCH3CCl3)  SH
= (TCH3CC13)                - 8 to 11 years,
                                            global average
                                                                 -'         (7)
                                      80

-------
                                Table  21
                    METHYL-CHLOROFORM  GLOBAL AVERAGE
                        RESIDENCE TIME (TCH3CC13)
Estimated from
Field Data
Singh (1977) §
Lovelock ^1977)
Singh (1977)§
McConnell & Schiff (1978)
Chang & Penner (1978)
Singh et al. (1979a)§
Krasnec (1979)
Rowland (1979)
Neely & Plonka (1978)
^TCH3CCl3^
(Years)
7 ± 1
5-10
8-11
8
12
8-11
9-12 (est)f
*6 (est)t
3*
Model
Estimates
All estimates
Prior to 1977
NAS (1976)
Crutzen &
Fishman (1977)


(TCH3CCl3)
(Years)
1-3

1.4
10


 Same data as used by Lovelock (1977).
 Estimated by Singh.
?Based on data from present study.
                                 Table 22
              RESIDENCE TIMES OF METHYL CHLOROFORM IN THE
                    NORTHERN AND SOUTHERN HEMISPHERES
Data Source
Loveloc^k (1977)
Rasmussen et al. (1976)
Singh et al. (1979a) )
Rowland (1979) >
Rasmussen (1979) /
Krasnec (1979)
NH/SH Concentration
Ratio of CH3CC13
2.2
2.2
1.3-1.5
1.2
(TCH3CC13)*NH
(TCH3CC13)SH
>4
>4
1.5 to 3
= 1
    Estimated during the present study.
                                    81

-------
                           ^TCH3CC13

                                        - 1.5 to 3    .                    (8)

                                    /SH

In short we find that:

     •  The average global residence time of 0130013 is significantly longer
        than predicted by models.  Our data supports a TCH3CCl3 value of
        8 to 11 years.  When estimates of other investigators are included a
        range of 6 to 12 years is suggested.  This is a factor of 5 to 10
        longer than the 1.4 year TCHoCCl-i values reported by NAS (1976).

     •  A 6 to 12 year TCH3CCl3 allows an estimated 12 to 25 percent of all
        OH30013 to enter the stratosphere.  The long tropospheric residence
        time, when coupled with the rapidly increasing emissions, suggests
        that CH3CC13 may be a potential depletor of stratospheric 03 in the
        decades ahead..  The impact may be further escalated in the future as
        more and more nations substitute CH3CC13 to replace the declining
        usage of C2HC13 and to some extent 02014 (for toxicity and health
        reasons), thereby resulting in extraordinary increases in CH3CC13
        emissions.  Worldwide release of CH3CC13 to the atmosphere currently
        approaches 7 x lO^-g/year and is increasing at a rate of 10 to 15
        percent per year.  Continued uncontrolled release of CH3CC13 to the
        atmosphere should be a matter of future concern.

     •  The residence time of CH3CC13 appears to be longer in the northern
        hemisphere when compared to the southern hemisphere.
     2.   Methyl Chloroform as an Indicator of the Hydroxyl Radical Abundance
          in the Northern and Southern Hemispheres

     As stated earlier the major removal mechanism for CH.CC1- is expected to .
be reaction with the hydroxyl radical (HO).  The rate constant of CH-CC1- + HO

is well known (K = 3.5 x 10~12 exp(-1562/T) cm^/molecule/sec), the residence
time of CH_CC13 is significantly long, and the tropospheric reservoir suffici-
ently large that short-term perturbations in global emissions are unlikely to
alter significantly the distribution of CH-CCl,.  Therefore, the residence

time of CH-CCl- could be used to determine average hydroxyl radical concentra-

tions.  Once the average HO abundance is determined, the residence times of a
vast number of chemicals can be determined.

     In the past, the HO abundance has been derived from 1-D models which

have never been satisfactorily validated.  The uncertainty in these models
                                      82

-------
allows for an uncertainty in the HO abundance of a factor of 10 or more
(Isakscn and Crutien, 1977).  Therefore, the. HO estimates from CH,CC13 resi-
dence Lime data should be used to validate existing models (Crutzen and
Fishman, 19/7).
     Using the rate constant for the CH.CC1., + HO reaction at a weighted-
                                                        -15    •*
average tropospheric temperature of 265°K  (k = 9.64 x 10    cnr«molec/sec) we
can calculate the HO burden from CH.CC1- residence times given in Equations
7 and 8.
                             (HO)   + (HO)
        (HO)global average =	~~	~ = 3 to 4 x 105 molec/cm'   ,
             .                 (HO)SH - 1.5 to 3 .            fi             (9)
                              (»0)NH

                                                   5         3
     The global average HO abundance of 3 to 4 x 10  molec/cm  is a factor of
5 to 10 lower than the values estimated by many modelers (Weinstock et al.,
1979).   More recently, Crutzen and Fishman (1977) have revised their photo-
chemical model to conform with these inferred HO values.
     The significance of reduced HO levelo cannot be overemphasized, since it
is a major tropospheric loss mechanism for a majority of hydrocarbons and halo-
carbons.  A five-fold increase in the lifetime of these trace constituents is
likely to increase substantially the estimates of stratospheric chlorine con-
tribution due to F-22, CH.CCl.,, CH-Cl., etc., as well as result in an increased
intrusion of nonmethane hydrocarbons (such as C0H,, C_H0, and C«H0) which have
                                               £ D   Jo       I. i
been all but ignored in existing models.  The high activation energy of the
C-H^ + HO reaction rate (3600 cal/mole) also permits a long lifetime at the
reduced temperatures of the lower stratosphere.  In addition, the rate con-
stant for the C-H, + Cl reaction is about 1000 times faster than for CH. + Cl
               26                                                     4
and a small  fraction of C_H, could produce an important competing reaction.
These factors could change the estimates of stratospheric ozone losses.
Methane oxidation reactions, proposed to control the tropospheric 0, and CO
              *                                                    J
              'i
distribution, would be impacted by these considerations.  Reduced HO levels
                                      83

-------
further help to resolve the current discrepancies in the tropospheric budget
and distribution of CO.
     Although some differences in the northern hemisphere and southern hemi-
sphere levels of HO are to be expected because of natural differences in water
vapor, ozone, and particulate concentrations, a much larger difference can be
attributed, at least in part, to CO (a principal sink for HO) whose levels are
three times higher in the northern hemisphere when compared with the southern
hemisphere (Seiler, 1974).  If the additional CO in the northern hemisphere
is predominantly from man-made sources, as the current data seem to suggest,
continued release of CO has and will further reduce the HO levels, thereby
depleting the scavenging ability of the atmosphere.  Additional dependence on
fossil fuels in coming years is likely to cause more severe depletions in the
HO abundance.  These HO losses will not only allow a larger stratospheric input
of anthropogenic halocarbons and hydrocarbons, but may already have altered
the stratospheric chemistry by permitting increased intrusion of natural
ubiquitous tropospheric species, such as CH, and CH..C1.  Vie wish to add, how-
ever, that the possibility of a natural hemispheric CO (and HO) gradient from
unknown CO sources in the northern hemisphere, although unlikely, cannot be
completely ruled out at this time.

E.   Discussion of Results
     The major results of this section can be summarized as follows:
     *  Contrary to previous belief the observational data best supports an
        atmospheric residence time for CH3CC13 of 8 to 11 years.  This allows
        about 15 to 22 percent of all CH3CCl3 released at ground level to
        enter the stratosphere.  The long residence time and the rapid growth
        of CH3CC13 cause it to be a potential depletor of stratospheric ozone.
     •  Methyl chloroform appears to have excellent properties to act as an
        indicator of seasonally average hydroxyl radical concentration.  Our
        data suggests a global average HO concentration of 3 to 4 x 10^ molec/
        cm3.  When data by other investigators is included an HO concentration
        range of 3 to 6 x 10^ molec/cm^ can be calculated.
     •  The distribution of CH3^13 in the northern hemisphere and southern
        hemisphere points to an .asymmetric HO distribution with higher average
        HO levels in the SHCCH^SH/CHO)^ >1.5).   The  uncertainty  in  this
        result is larger because of limited data and greater sensitivity of
        calculations.  The conclusion is qualitatively justifiable and can be
        attributed to carbon monoxide (a principle sink for HO) and its
        asymmetric hemispheric distribution.  More data is required to confirm
        these findings.
                                      84

-------
                                 SECTION 11

                         ATMOSPHERIC METHYL CHLORIDE


A.   Global Emissions

     Table 23 shows the estimated emissions of CH3C1 from man-made sources.

The accuracy of this data is not expected to be better than a factor of two,
                                  Table 23

                GLOBAL ATMOSPHERIC RELEASE OF METHYL CHLORIDE
                               (in 106 kg/yv)               ?'
Year
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
Methyl Chloride
Release
0.5
0.6
0.7
0.6
1.0
1.2
1.5
1.6
1.7
1.9
2.7
3.4
4.0
4.4
5.9
6.1
6.3
6.6
7.9
6.7
5.3
5.5
6.0
82.1
                       Source-   B;iuer  (1978)

                                      85

-------
However,  for  the  purposes of the present discussion the emissions data serve
a qualitatively useful purpose.  It is clear from Table 23 that the total
release of Cl^Cl  has been about 0.082 million tons.  An upper limit of 0.16
can be estimated.   If all of the 0.16 million tons were to be uniformly dis-
tributed  in the troposphere, an average global concentration of less  than  20
ppt would result.   In actuality, since Ct^Cl is rather rapidly removed, the
emissions from Table 23 could not lead to an atmospheric concentration of  more
than 4 ppt.   It suffices to say that man-made emissions of CH-jCl have been
quite small.

B.   Global Burden  and Distribution
     A number of  investigators have measured CI^Cl in the northern hemisphere
and background concentrations have varied between 400 and 1500 ppt  (Lovelock,
1975; Grimsrud and  Rasmussen, 1975, Singh et al., 1977a, 1977c).  All measure-
ments indicated a large burden of CI^Cl that could not be accounted for by man-
made sources, which appear to be quite small.
     Figure 28 shows the global distribution of CH3C1 as measured by us.
                                                                    r
It is clear from  Figure 28 that CH3C1 is essentially uniformly distributed
over the  globe.   Since no other profile of the global distribution of CH3C1
has been  published, comparisons with other investigators are not possible.
                     1500
                      -90°-80° -60°  -40°  -20°  0°   20° 40°  60°  80° 90°
                        S          LATITUDE — degrees         N
                      'Average hemispheric concentration
                      1 Standard deviation

                FIGURE 28 GLOBAL DISTRIBUTION OF METHYL CHLORIDE
                                       86

-------
It is obvious from Figure 28 that CH^Cl has an average global background con-
centration of about 615 ppt.  Unlike Ctt^CCl^, there is no north-south gradient
for CH-jCl, despite the fact that Cl^Cl is at least 2 to 3 times more reactive
than CH-jCCl3 (see Section XVI).  This only points to a very large natural source
of CHjCl that is relatively uniformly distributed.
     It is also interesting to note that CH^Cl shows a large atmospheric
variabilit/ with a standard deviation of about 15 percent.  Within the northern
hemisphere we have observed higher values in the Pacific marine environment
that fall off as we enter deep into continental atmospheric areas.  Table 24
shows the variability of CH-jCl as measured at clean continental and marine
environments.
     Table 24 shows the lowest CH3C1 concentration of 586 ppt at Site 12, which
is a truly continental surface site at about 800 m above sea level.  The cor-
responding concentration at the marine Site 11 in roughly the same season is
730 ppt.  It further appears that a seasonal gradient in CH3C1 levels exists,
with'higher concentrations in spring or early summer.  A comparison of data
from Sites 11 and 14, which are identical for all practical purposes, indicates
CH3C1 levels that are about 15 percent lower in September when compared to May.
The data base, however, is too sparse to generalize these observations.  It is
adequate to say that Cl^Cl must have large natural sources.  These natural
sources give it a complex atmospheric variability that cannot be adequately
characterized with the limited existing data.  Our atmospheric measurements
over the continental United States appeared to indicate an oceanic source of
CH3C1.  Indeed, the highest concentrations of Cl^Cl in the clean environments
approaching 2 ppb were measured by us in the marine boundary layer.  This in
no way implies that other natural or secondary man-made sources of Q^Cl do
not exist.  Here it is pertinent to add that in urban areas we have found
significantly higher levels of Ct^Cl indicating as yet unknown sources associ-
ated with man-made activities.

C.   The Oceanic Source
     As suggested by our field measurements, a preliminary attempt was made to
measure Cl^Cl in Pacific seawater.  These measurements are shown in Table 17.
                                      87

-------
                                                                                            Table 24
                                                           AVERAGE CONCENTRATIONS Of SELECTED T3ACE CONSTITUENTS IN CLEAN ENVIRONMENTS
00
00
Chemicals
CC12P2 (ppt)
CC13P (ppt)
CHC1F2 (ppt)
CHC12P (ppt)
CC12FCC1F2 (ppt)
CC1F2CC1F2 (ppt)
SFft (ppt)
CC14 (ppt)
CHC13 (ppt)
CH2C12 (ppt)
CH3C1 (ppt)
CH3I (ppt)
CBjBr (ppt)
CC13CC13 (ppt)
CHjCClj (ppt)
CH23rCU2Br (ppt)
CC12CC12 (ppt)
CHClCClj (ppt)
COC12 (ppt)
»20 (ppb)
HO (ppb)
"02 (ppb)
CH3(CO)OON02, PAH (ppt)
CH3CH2(CO)OON02, Pro (ppt)
Oj (ppb)
CO (ppb)
CH4 (ppb)
C2H2 (ppt)
C2Hfi (ppt)
Sita 5
High Altitude
Continental
(May 1976)
204 (19)+
116 (5)
•-
14 (4)
20 (3)
—
0.23 (0.03)
114 (6)
17 (2)
—
713 (51)
9 (5)
5 (3)
—
103 (9)
—
31 (10)
15 (3)
22 (5)
312 (18)
1.3 (0.9)
4.0 (1.3)
— '
—
53 (12)
—
1412 (103)
—
—
Site 11
Sur face-Marine
(Hay 1977)
220 (20)
126 (13)
20-30
5 (2)
23 (6)
12 (5)
0.26 (0.04)
123 (8)
20 (5)
45 (22)
730 (134)
5 (3)
20 (6)
7 (1)
111 (18)
5 (1)
34 (5)
11 (1)
15 (4)
313 (8)
<5.0
<5.0
100 (55)
40 (17)
35 (5)
118 (15)
1479 (19)
<400
1951 (281)
Site 12
Surface Continental
(June 1978)
253 (17)
138 (6)
. . 20-40
6 (1)
23 (5)
12 (3)
0.39 (0.06)
128 (12)
16 (5)
54 (15)
586 (64)
2 (1)
5 (1)
6 (2)
130 (16)
<5
41 (10)
13 (4)
—
311 (5)
1 (2)
3 (2)
252 (103)
--
31 (11)
175 (50)
1600 (80)
— • ' .
9980 (4950)
Site 14
Surface-Marine
(Sept. 1979)
252 (16)
152 (19)
—
7 (1)
22 (2)
13 (3)
0.30 (0.14)
125 (8)
—
39 (17)
628 (50)
—
13 (5)
—
132 (11)
<5
34 (11)
17 (7)
15 (5)
312 (6)
3 (3)
8 (5)
80 (57)
—
—
" —
1563 (284)
<200
1595 (986)
Hemispheric Averages
(Nov. -Dec, 1977)
Northern
Hemisphere
230 (26)
133 (13)
—
5 (3)
19 (4)
12 (2)
0.31 (0.04)
122 (5)
14 (7)
44 (14)
611 (84)
2 (1)
—
—
113t
~
40 (12)
16 (8)
'
311 (2)
—
—
—
—
—
—
1430 (5)
<200
1060*
Southern
Hemisphere
210 (25)
119 (12)
_
4 (1)
18 (3)
10 (1)
0.27 (0.01)
119 (4)
O
20 (4)
615 (103)
2 (1)
—
—
77
—
12 (3)
O
—
311 (3)
—
—
—
—
—
—
1390 (51)
<200
524 (15)
                                ppt - 10"12 v/v; ppb « 10-9 v/v
                                ^Standard deviation in parentheses.
                                tNonunifor* concentration within each hemisphere,  the average represents that concentration in the hemisphere which would result if the
                                 hemispheric mass of-the species were to be uniformly distributed.

-------
It is clear from Table 17 that the surface concentration of CHjCl in the
1'ncific is quite variable, with values somewhat higher near the equator.  The
average surface concentration found was 26.8 ng/liter.  Using an SCH-,CJ of
2.65 (Dilling, 1971), D = 10~5 cm2/sec, and Z = 90um, the simple film-diffusion
model described in Equation A can be used to estimate an oceanic flux.  Using
Equation 4 and other parameters described above, we calculated CH-jCl flux from
ocean to the atmosphere of 2.6 x lO"-^ g cm~2/sec.  Extending this to the world
                                             •I O
ocean body gives an exchange rate of 3.0 x 10   g/year.  From our measurements,
the atmospheric burden of CH3C1 can be estimated to be 5.5 x H)12 g.  Thus, on
the basis of our limited data, the ocean appears to be a significant source of
CH3C1 , which can provide an atmo.-spher Lc turn* ver rate of about 2 years.  This
is in reasonable agreement with the estimated CH^CI residence time of about
2 to 3 years because of attack (HO - 3 x 1Q5 to 5 x ifl5 moleulces/cm3) .
Although the data base is too scarce and variable to make a strong case, it
may turn out that CH-jCl atmospheric burden is dominated by the oceanic source.
It should be pointed out that microbial fermentation and combustion of vege-
tation as sources of CH3C1 have also been proposed  (Palmer, 1976) but remain
largely unquantif led.  The mechanism by which CHgCl may be produced in the
ocean is unclear but reactions involving CA^I and the chloride ion of seawater
have been suggested as a possible source of oceanic CH^Cl (Zafiriou, 1975) .

D .   Residence Time of
     The residence time of ' CH_3C1 can be estimated in two principal ways:
by atmospheric variability of CH^Cl and removal via reaction with HO.
     It is clear from Figure 28 that CHjCl is relatively uniformly distributed
and a variability of about 15 percent  (ot) is associated with its global
distribution.  We estimate that on the average about 7 percent of this vari-
ability (op) is associated with analytical methodology.  Thus, the true atmo-
                                                                       n
spheric variability of CH3C1 can be estimated to be about  13 percent  (o|  =
°t ~ °P^^ '  ^or sPecies with a uniform source distribution,  the approximate
statistical relationship of Junge  (1974) can be used:
                                      89

-------
Since the above relationship is accurate to within a factor of 3, we estimate
a CH-jCl atmospheric residence time of 1 to 3 years.  The upper limit of this
estimate is probably more correct since ot of 15 percent is partially high
because of the proximity of our measurements to the oceanic source.  The best
estimate of average HO derived by us from the CH3CC13 data is 3 to A x 105
molecules/cm^.  From the known reaction rate of CX^Cl with HO (see Section XVI)
we estimate that CHjCl would have a mean residence time of 2.5 to 3.5 years.
     Thus, two completely independent estimates of ^CHiCl are *n S°°d agree-
ment and point to an estimated ^cHoci °f 2 to 3 years.

E.   Discussion of Results
     Unlike other halocarbons we measured, CH-jCl is the only one with a
large natural source.  The atmospheric abundance as well as the relatively
uniform global distribution of C^Cl (despite its high reactivity) point to
the existence of large natural sources.  Our limited data support the con-
clusion that one of the most significant natural sources of CH^Cl is the ocean.
This oceanic source is estimated to be 3 million tons/year and can account for
nearly all of the atmospheric burden of CH3C1.  The possibility that unknown
man-made sources yet remain to be quantified also exists.  It is unlikely,
however, that these would be comparable to the natural sources in quantity.
Atmospheric variability data as well as the removal by hydroxyl radical point
to a 2- to 3-year tropospheric residence time.  Thus, less than 6 percent of
all the CH3C1 released at ground level will enter the stratosphere.  The
mechanisms by which CH^Cl is produced in the ocean are not understood, and
the distribution of CH3C1 in the ocean is highly variable and poorly charac-
terized.
                                      90

-------
                                 SECTION 12
             CHLOROFORM, METHYLENE CHLORIDE, TRICHLOROETHYLENE,
             TETRACHLOROETHYLENE, HEXACHLOROETHANE AND PHOSGENE
A.   Global Emissions
     The estimated global emissions of the first four chemicals (CHC13, CH2C12,
C2HC13, and C2C1^) are listed in Table 25.  These data are estimated from
production figures and usage patterns.  The accuracy of this data is probably
better than ±50%.
                                  Table 25
                                                            f,
             GLOBAL ATMOSPHERIC RELEASE OF CHLOROFORM, METHYLENE
       CHLORIDE, TRICHLOROETHYLENE AND TETRACHLOROETHYLENE (in 106 kg)
Year
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
Chloroform
2
3
3
2
4
4
4
5
5
6
8
9
9
9
11
12
11
12
12
15
13
14
15
188
Methylene
Chloride
49
63
63
59
75
75
77
96
99
119
141
178
175
202
244
268
267
314
346
394
331
356
375
4,366
Trichlor-
ethylene
453
496
485
423
516
506
443
511
528
531
624
689
703
744
856
877
738
613
648
622
420
453
435
13,304
Tetrachlor-
ethylene
154
160
170
161
175
180
194
276
280
316
370
400
460
549
548
610
608
633
609
630
586
>577
'570
9,216
       Source:  Bauer  (1978)
                                      91

-------
     Hexachloroethane is not manufactured in large quantities in the United
States and only 500 tons/year are importedtfrom Europe.  Most of the C2C16
produced is released to the atmosphere.  Since the molecule is very stable, it
may tend to accumulate in the atmosphere.  Thus, the emissions of C2C16 as a
small impurity in C2C14 may be one of the sources.  No global emissions
inventory is currently available.
     Phosgene is predominantly manufactured for captive use and very little
                                          +
direct release takes place.  Table 26 shows that even though the production of
phosgene has been significant, the sales have never exceeded 6 million kg/year.
While the release of phosgene may create & local health hazard, there Is no
evidence to support the assumption that the direct emissions of phosgene are
large.  The major source of phosgene in the atmosphere is likely to be from
the photooxidation of chloroethylenes (Singh, 1976).  This secondary source is
difficult to quantify accurately but may be as large as 0.3 million tons/year
(Singh, 1976).
                                  Table 26
                   PHOSGENE PRODUCTION, SALES, AND RELEASE
                       IN THE UNITED  STATES  (in  106  kg)
Year
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
Production
2.4
4.1
16.7
16.6
26.3
50.9
96.4
111.4
129.4
149.9
169.1
203.0
229.5
280.7
241.2
289.5
331.0
Sales
2 to 6
















Emissions
negligible
















               Source:  CIS (1979)
                                      92

-------
K•   Atmospheric Measurements
     1.  ChjLor£form
     There is little doubt that large man-made sources of chloroform exist.
Table 24 shows a typical background concentration at northern  temperate
latitudes of 16 to 20 ppt.  The northern hemisphere average, as determined
from an analysis of a limited number of samples, was found to  be 14 ± 7 ppt.
We suspect that the background concentration of CHC13 is latitude-dependent
and is unlikely to be uniform in the northern hemisphere.  In  the southern
hemisphere CHC13 was barely detectable and a background concentration of <3
ppt was reported.  A small gradient away from the marine environment was
observed.  This gradient cannot be easily attributed to natural oceanic
sources of CHC^.  As will be discussed in Section XVII, the Pacific coastal
waters appeared to be significantly contaminated with CHC13 and may have
caused this apparent marine air/continental air concentration  gradient.

     2.   Methylene Chloride
     Methylene chloride is a well-known solvent and is released in significant
quantities.  Over the northern hemisphere average background concentrations in
the range of 40 to 55 ppt have been measured  (Table 24).  The  northern hemi-
sphere background concentration in 1977 was measured to be 44  ± 14 ppt.  The
southern hemisphere average concentration was 20 ± 4 ppt.  Thus, a significant
north-south gradient is apparent.  The emissions of CH2C12 from man-made
sources alone (Table 25), when coupled with 1-year atmospheric residence time
(#»'<•• ;,ci I. l(»n XVIJ, HIT rwifllst on!, with l.hf. mefjefi/'ed afflionpheflc global bufd&it
A short average residence tJme of the order of 1 year can cause the atmo-
s|>li«'flf abundance of CH^CL, to vary st-fiHonally.  Superimposed  on  this  variation
is the variability in emissions.  Additional data is required  to characterize
more adequately the abundance and variability of these relatively short-lived
species.

     3.   Trichloroethylene and Tetrachloroethylene
     Despite the high emission strength of these species, as suggested by  data
in Table 25, the background concentrations of these species were quite low.
Table 24 clearly shows an average northern hemisphere background concentration
                                      93

-------
for CjCl. of about 40 ppt.  Between 1975 and 1978 they have typically varied
between 30 and 40 ppt.  Table 24 also shows that the southern hemisphere C2C14
abundance is about a third of its northern hemisphere value.  No global dis-
tribution profile is available but significant latitudinal gradients can be
expected.  These gradients are expected to be even larger for €2^13.  The
northern hemisphere average of 11 to 17 ppt for C2H Cl-j can be compared to a
southern hemisphere value of less than 3 ppt.  The low background concentra-
tions of C2C1^ and C2HC1-J as well as the large northern hemisphere and south-
ern hemisphere gradients are indicative of short residence times.

     4.   Hexachloroethane
     As stated earlier, C-Cl, is not directly emitted in large quantities.
Typical background concentrations have varied between 5 and 7 ppt.  C2Clg
measurements in the northern hemisphere are highly limited and are not avail-
able from any source other than our own.  No southern hemisphere data is
available.

     5.   Phosgene
     Phosgene was measured for the first time at several clean sites at con-
centrations of about 15 ppt.  As is clear from Table 26, the direct emissions
of phosgene are very small.  No measurements in the free troposphere and the
southern hemisphere are currently available.

C.   Discussionof Results;  Sources and Sinks
     The total emissions of CHCl-j from man-made activities are estimated to
be 0.2 million tons.  The accuracy of this estimate is low.  CHC13 is rela-
tively rapidly removed from the atmosphere via reaction with HO and a residence
time of about 1 year is estimated (Section XVI)  No natural sources of CHC13
are known, and the northern hemisphere and southern hemisphere gradient further
suggests that these sources, if they exist, are not uniformly distributed.
There is little doubt that dominant emissions of CHC^ occur in the northern
hemisphere.  If the measured concentration of CHC13 is representative of the
northern hemisphere troposphere, then emissions data in Table 25 must be a
gross underestimate.  It is also possible that secondary anthropogenic sources
                                      94

-------
of CHC1-3 (such as chlorination of water) exist.  These sources have not yet
been quantified.
     The atmospheric fate of CH2C12 is similar to CHC13 and it also persists
in the atmosphere for about 1 year (see Section XVI).  Unlike CHCl^, an esti-
mated 4.2 million tons of CH2C12 have been released to the atmosphere (Table 25).
Because of its increasing use in solvent applications, its growth has been
rapid.  In 1965 the emissions of CH2C12 were 25 percent of those of C2HC13.
In 1977 this fraction had increased to 85 percent.  The measured concentration
of 40 to 50 ppt in the northern hemisphere and about 20 ppt in the southern
hemisphere are not inconsistent with the available emissions data.  There is
little doubt that essentially all CH2C12 is of man-made origin.  It is useful
to add that CH2C12, in a manner similar to CH-jCClj, could be used as an
indicator of HO radical abundance.  Because of its relatively short residence
time of one year, a better resolution of the emissions data would be needed.
The advantage of CH2C12 would be its greater reactivity which could potentially
be used.to establish seasonal trends in the HO abundance.
     Both C2HC13 and CoCl, have been large volume chemicals.  These have been
in disfavor in recent years, largely for toxicity reasons, but also because of
their photochemical involvement with other hydrocarbons in smog formation
processes.  To date an estimated 13 million tons of C2HC13 and 9 million tons
of C2C1^ have been released to the atmosphere.  There is no doubt that both
these chemicals are exclusively of man-made origin and at least 95 percent of
their release occurs in the northern hemisphere.  The southern hemisphere data
is virtually nonexistent but our limited measurements are consistent with man-
made sources in the northern hemisphere.  The northern hemisphere background
concentration of 30 to 40 ppt of 0201^ is in agreement with a man-made source
and a residence time of 0.8 years  (see Section XVI).  On the other hand,
C2HC13 is highly reactive, and the measured background concentration of about
15 ppt is higher than can be explained from emissions data.  The measurements
of Cronn et al. (1976) suggest few vertical gradients in the troposphere for
species with residence times of greater than a few months.  For C2HCl3» however,
the residence time is estimated to be only about 2 weeks (Section XVI) and
significant vertical gradients can be expected.  Thus, the ground-level mea-
sured background concentration of a very short-lived species such as C2HC13

                                      95

-------
may not be representative of the entire troposphere.  The data of Cronn et al.
(1976) showing virtually no vertical gradients for C2HC13 probably suffers
from contamination problems.
     The sources of C2Clg are not known well enough to assess its source-
sink relationship.  C2C1^ can be expected to 'be essentially inert in the
troposphere.  It is possible, however, that it may eventually accumulate in
marine biota.  The possibility exists that C2C1/- abundance represents an
accumulation of emissions that have occurred over a period of many years.
This opens up the possibility that it may be present or may have been present
as an impurity in a high-volume chemical such as C2HC13 or 0201^.
     The sources and sinks of phosgene have already been discussed (Singh,
1976; Singh et al., 1977c).  We concluded that the direct primary emissions of
COC12 are negligible.  Its major source is atmospheric photooxidation of
C;;>HCl3 and C2C1^.  The major sinks appear to be liquid-phase hydrolysis or
heterogeneous decomposition on surfaces.  Gas-phase hydrolysis was found to be
a negligible sink.  The reaction rate of COC12 with HO has not yet been mea-
                                      t
sured.  In any case, it appears that COC12 residence time is of the order of
several months.
                                      96

-------
                                  SECTION 13
                   ATMOSPHERIC BROMINE AND IODINE SPECIES

     Only two brominated organics have been identified in the lower troposphere.
These are ethylene dibromide and methyl bromide.  The concentrations of both
these species are low and variable.  Their tropospheric distribution remains
poorly characterized.
     Table 27 provides estimates of the worldwide production and release of
CH2BrCH2Br and CHjBr.  A great deal of Cl^BrCH-Br is used in gasoline but is
converted to particulate matter (largely PbBr2) that is lost to the ground via
dry and wet deposition processes.  One can only roughly estimate that between
5 and 25 percent of CH,,BrCH7Br escapes to the atmosphere.  Table 27 clearly
shows that this source of bromine can be between 15 and 75 million kg(Br)/year.
     Table 27 also shows the production data for CH^Br.  It is clear from this
table that the production of CH.,Br is much smaller than CH?BrCH_Br.  However,
most of CH Br is used as a soil fumigant and a significant fraction is released
to the atmosphere.  We estimate that 50 to 75 percent of all CH Br manufactured
is released to the atmosphere.  Thus, a man-made bromine source of 13 to 20
million kg (Br)/year is attributable to CH.Br.  Together, CH-BrCH2Br and CH-Br
add up to a man-made source of 28 to 95 million kg (Br)/year.  Other forms of
organic bromine probably exist.  CHBr_ has been identified to be present in
drinking water.  Fluorocarbon-13Bl (CBrF_) is a commonly used and an increas-
ingly popular fire extinguisher.  It is also used as a common tracer in
scientific studies.  Among the various brominated species, CBrF- is the only
one that is likely to enter the stratosphere essentially intact.  The current
production of CBrF_ is too small to be of any consequence.
     The only organic iodine species to have been identified in the atmosphere
is CH^I.  No significant man-made sources of CH_I are known to exist.
     The global distribution of CH2BrCH2Br and CH_Br has not been characterized,
Indeed no atmospheric data from the southern hemisphere is available.  No
                                      97

-------
                            Table 27
         ESTIMATED WORLD-WIDE PRODUCTION AND RELEASE OF
              ETHYLENE DIBROMIDE AND METHYL BROMIDE
                      (in units  of  106 Kg)
Year
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
Total
1955-1976
^Ethylene Dibromide*
Production
120
127
124
121
126
108
112
116
121
144
170
164
§ 173
178
191
189
181
205
217
192
174
150f
3380
Release
5 to 25%
of
Production


















Methyl Bromide
Production
:
—
—
~

—
—
—
9
10
12
13
12
14
14
16
19
20
23
26
188
Release








50 to 75%
of
Production










Based on U.S. data (CIS, 1979).  U.S. Production is taken to be 75
percent of world production between 1955 and 1969 and 70 percent
between 1970 and 1976.
    the year 1976 it is assumed that the production outside U.S. was
the same as in 1975.
                               98

-------
measurements other than our own have been published even for the northern
hemisphere.  Table 24 clearly shows that the surface background concentration
of CH2BrCH2Br is less than 5 ppt.  The corresponding concentrations for
CH-Br are highly variable.  In May 1977 a mean concentration of 20 ppt was
measured at a marine site in Point Arena (Site 11).  At locations away from
the marine environment (Sites 5 and 12) and in about the same season, average
concentrations of 5 ppt were measured (Table 24).  In September 1979 at a lo-
cation identical to Site 12 (Site 14) the average of measured CH-jBr concentra-
tions was about 13 ppt.  Briefly then, the data from Table 24 suggest higher
values of CH~Br in the marine air compared to continental air and a possible
seasonal variation of CH~Br in the marine air.
     Preliminary measurements of Lovelock (1975) in coastal water provide
further support for the oceanic source of CHoBr.  No measurements for the
open oceans are available.
     It is clear from Table 24 that the concentration of CH-I in the marine
environment typically varies between 1 and 5 ppt and has an average value of
about 2 ppt.  CHnI seems to be of oceanic origin.  Limited coastal water
samples from the California coastal waters were found to contain about 50 ppt
of CH»I in helium in equilibrium with water (Singh, et al., 1977c).  Lovelock
(1973; 1975) has measured CH.,1 concentrations in excess of 100 ppt in coastal
waters elsewhere.  Ambient values as high as 61 ppt were measured at Site 2
during a thunderstorm.  It is likely that the ocean spray provides a localized
source of CH.,1 during turbulent conditions.  Because of the high reactivity of
CH-I found in past analyses, its atmospheric residence is limited to about 3
days.  No measurements of CH I from the free troposphere are available.  The
northern hemisphere and southern hemisphere marine levels are essentially
indistinguishable.
     On the whole, the atmospheric distribution of both brominated and iodated
organics has not yet been satisfactorily characterized.  Additional research
in this area is clearly needed.
                                      99

-------
                                  SECTION 14
       GLOBAL^BUDGETS OF ORGANIC CHLORINE, BROMINE, IODINE, AND FLUORINE

     Because of continual release of pollutants into the atmosphere, the
halogen budget is a constantly changing quantity.  In its present condition
the atmospheric halogen burden is sufficiently large that yearly perturbations
are not expected to alter it by more than 5 to 10 percent.  Table 28 lists the
concentrations of a large number of chemical species that we measured at the
end of 1977.  We choose this period because it is^the only period for which we
have both northern hemisphere and southern hemisphere data.  The major sources
of species are also defined in Table 28.
     Table 29 uses the data from Table 5 to develop a Cl, Br, I, and F budget
for the northern hemisphere, southern hemisphere and globe.  In addition, the
                                       ,y
contributions from man-made and natural sources are defined.
     It is clear from Table 29 that the global average Cl budgetVat the end of
1977 was 2.7 ppb while that in the northern hemisphere was 2.9 ppb.  This im-
plies that the odd chlorine available in the stratosphere could not exceed 3
ppb.  The possibility that other, as yet unquantified, chlorinated species
exist in the troposphere cannot be completely dismissed.  However, measurements
of total organic chlorine, as measured by Berg and Winchester (1976), show
relatively good agreement with our results and do not support the existence of
large unknown organic chlorine sources.  We further calculate from the source
categories shown in Table 28 that 77 percent of the tropospheric organic
chlorine is man-made while the remaining 23 percent is of natural origin.  Thus
a significant perturbation of the natural atmosphere from man-made activities
has already taken place.
     Anderson (1979) has measured free chlorine concentrations in the strato-
sphere and reports maximum levels at about 40 kms that vary between 1 to 2 ppb
in late 1977.  In one flight a maximum of 7 ppb is measured.  This latter
value would be in serious disagreement with our findings that indicate an
upper limit of 3 ppb of total odd free chlorine.
                                       *
                                     100

-------
                                Table  28

             AVERAGE  BACKGROUND CONCENTRATIONS OF IMPORTANT
                           TRACE CONSTITUENTS
Compound
N20
CC12F2 (F-12)
CClaF (F-ll)
CCl2FCClF2 (F-113)
CC1F2CC1F2 (F-114)
CHC1F2 (F-22)
CHC12F (F-21)
SF6
CC14
CH3CC13
CH3C1
CH3I
CH3Br
CHC13
CH2C12
C2HC13
C2C1A
C2C16
CH2BrCH2Br
CH4
C2H6
C2H2
Major*
Source
N
A
A
A
A
A
A
A
A
A
N
N
A, N
A
A
A
A
A
A
N, A
N, A
A
v Concentration*
Northern
Hemisphere
Average
311 ppb
230 ppt
133 ppt
19 ppt
12 ppt
20-30 ppt
5 ppt
0.31 ppt
122 ppt
113 ppt
611 ppt
2 ppt
5-20 ppt
14 ppt
44 ppt
16 ppt
40 ppt
i5 ppt
£5 ppt.
1430 ppb
1060 ppt
<200 ppt
Southern
Hemisphere
Average
311 ppb
210 ppt
119 ppt
18 ppt
10 ppt
t
4 ppt
0.27 ppt
119 ppt
77 ppt
615 ppt
2 ppt
—
<3 ppt
20 ppt
<3 ppt
12 ppt
—
—
1390 ppb
524 ppt
<200 ppt
Global
Average
311 pph
220 ppt
126 ppt
18 ppt
11 ppt
—
4 ppt
0.29 ppt
120 ppt
95 ppt
613 ppt
2 ppt
—
8 ppt
32 PPt
8 ppt
26 ppt
—
—
1410 ppb
792 ppt
<200 ppt
For those species where significant latitudinal variations within the
hemisphere were observed, the average concentration within each hemisphere
is the concentration that, when uniformly mixed in the hemisphere, repre-
sents the total burden of the species in that hemisphere.  The concentra-
tion data are for late 1977:  ppt = IQ~^ v/v; ppb = 10~9 v/v.
N « Natural; A = Anthropogenic.
                                    101

-------
     The organic bromine and iodine budgets are much less certain.  In the
case of bromine, the only measurements available were.taken during the conduct
of this study.  The only two brominated species that we measured are CH3Br and
CH2BrCH2Br.  However, even for these species the data are very scarce.  A pre-
liminary interpretation suggests an organic Br content of 10 to 30 ppt in the
northern hemisphere, of which 50 to 90 percent may be of natural origin
(Table 29).  No data from the southern hemisphere are available.
     There do not appear to be any significant sources of organic iodine.
The only species identified to date is methyl iodine, which is present at an
average concentration of 2 to 5 ppt in the marine boundary layer.  The mean
tropospheric concentration is expected to be much less than 2 ppt because of
the high reactivity of C^I.  Nearly all of CH3I is of natural origin
(Table 29).
     Table 29 also shows that there is about 1 ppb of F and almost all of it
is attributable to man-made sources.  In developing the F budget we have
added the contribution from OF*, which is present at ta concentration of about
                                                    I
65 ppt (Rasmussen et al., 1979) but was not measured during this study.
     Thus it appears that nature contributes virtually no organic fluorine and
nearly all of the iodine to the troposphere.  It is pertinent to add that
sources and abundance of Br and I may be poorly characterized and as yet
unknown species may be present.  Special emphasis is needed to undertake the
identification of new bromine and iodine species and better characterization
of those already measured.
                                     102

-------
                 Table 29

TROPOSPHERIC CHLORINE, BROMINE,  IODINE AND
         FLUORINE ORGANIC BUDGETS
Species
Cl
Br
I
F
Budgets
Northern
Hemisphere
2.9 ppb
10-30 ppt
<2 ppt
1.0 ppb
Southern
Hemisphere
2.4 ppb
—
<2 ppt
0.9 ppb
Global
Average
2.7 ppb
—
<2 ppt
1.0 ppb
Percentage
Source
Contributions
(%)
Natural
23
50-90
100
0
Man-made
77
50-10
0
100
                    103

-------
                                  SECTION 15
               ATMOSPHERIC HYDROCARBONS, CO, NOX, PAN, AND 03
A.   Hydrocarbons
     While these species  were not of principal interest in this study,  they
were also measured  at  several urban and nonurban sites.  The major  reason  for
including hydrocarbons (HCs)  was largely based on our finding of  reduced HO
levels in the  troposphere.   These low HO levels could allow some  nontnethane
                                           *'
hydrocarbons to enter  the stratosphere.  Since the Cl atom reaction with
molecules such as C2H5 and  C^2 was 10  to I'O  times faster than  reaction
with CH^, we felt that nonmethane hydrocarbons could compete with CH^ as a
sink for Cl atoms.
     Figure 29 shows the  global distribution of CH^ in the northern and
southern hemispheres.   The  global distribution of CH^ is essentially uniform .
with southern  hemisphere  levels, and about 3 percent lower than the northern
hemisphere values.   (Table  15 describes the polynomial used to fit  the  CH^
global data.)  The  global average concentration of 1.41 ppm for CH^ was also  *
measured.  Even though large man-made sources of Cfy exist in the northern
                  £>
                  g
                     1600
                     1000
                  O
                      500
1 * 1 ' 1 * I *
9
1390* (51.4)t ppb
1 1 ' 1 ' 1 1
.«.«*.— #"«»-*— 8— - "
1430* (64.7)t ppb
i.i.i. i
                      -90° -80° -60" -40°  -20°  0°  20° 40°  60°  80° 90°
                        S         LATITUDE — degrees          N
                         Average hemispheric concentration
                        * Standard deviation

                    FIGURE 29 GLOBAL DISTRIBUTION OF METHANE
                                      104

-------
hemisphere, CH^ is  largely  of  natural origin with a high background and a long
atmospheric lifetime.   This long lifetime allows for a nearly uniform distri-
bution of CH^ .  While  residence  times of  2 to 20 years have been estimated for
CH/,, our data are in good agreement  with the highest values
                                                                  « 21 years).
     The next most ubiquitous  hydrocarbon present in the global atmosphere
was found to be C2H^.  The  global  distribution of £2^6 was dissimilar to that
of CH^ , probably because  C2H5  has  a much shorter tropospheric lifetime than
CH/, does.  A third order  polynomial is used to fit the C2^(, global distribu-
tion and is described  in  Table 15.   Figure 30 shows the nonlinear C^ti^ dis-
tribution in the northern hemisphere.   At least in the southern hemisphere
the concentration of C2Hg is relatively uniform, averaging about 0.5 ppb.  In
the northern hemisphere,  however,  the  concentration approaches 2 ppb at mid-
northern latitudes but drops off  to nearly 1 ppb near the equator.  The
weighted average concentration of  C^tif, that represents its burden in the
northern hemisphere is calculated  to be 1.1 ppb.  Thus, a very large gradient
of €2^6 between the northern hemisphere and southern hemisphere exists.  The
latitudinal profile of C£Hg suggests that there are significant sources in
the northern hemisphere.
     Figure 31 shows the  diurnal variation of Z^^t* at Point Arena (Site 11).
The C2Hg background is 1.9  ppb and no  diurnal variation is evident.  The
sources of £2^6 are complex but there  is little doubt that both man-made and
                    10
                    d"  2
                            0.5* (0.1 )r ppb
                                               1.1* ppb
                      -90° -80° -60° -40° -20°  0°  20°  40°  60°  80° 90e
                        S          LATITUDE — degrees          N
                         Average hemispheric concentration
                        * Standard devietion

                     FIGURE 30  GLOBAL DISTRIBUTION OF ETHANE
                                      105

-------
               8000
             i
6000
              
-------
                 50
                 40
                 30
              u>
              51  20
                 10
                          AVERAGE CONCENTRATION (ppb) — 9.98 ± 4.95
                                    10        15
                                     TIME — hours
                                                      20
25
                     FIGURE 32  ETHANE DIURNAL VARIATIONS
                               ATJETMORE (SITE 12)
and southern hemisphere concentrations of €2^, for example, were measured to
be 1.4 ± 1.1 ppb and 1.0 ± 0.6 ppb, respectively.  In situ measurements at
Sites 12 and 14 indicated average concentrations of only 0.3 ppb.  We consider
the latter to be more reliable.  Additional in situ measurements must be con-
ducted to determine if C2H4 is a ubiquitous component of the global atmosphere.
B.   Carbon Monoxide
     Carbon monoxide is a well-established component of the global atmosphere,
and its global distribution and latitudinal gradients have been quite well
characterized (Sieler, 1974).  Our measurements point to additional features
that may suggest a far more complex distribution of CO.  Figures  33 and  34
show the diurnal variations of CO at two sites (Sites 11 and 12)  that are
located at roughly the same latitude but have different characteristics  be-
cause of their marine (Site 11) and continental natures (Site 12) .  At Site 12
no diurnal variation of CO is observed and an average concentration of
175 ± 50 ppb was measured.  Site 11, however, did show slightly lower values
in the afternoon, but more importantly indicated CO levels of 118 ± 15 ppb.
                                      107

-------
   500
   400
 a 300
 S
   200
   100
AVERAGE CONCENTRATION (ppb) — 118 * 15
                        10        15
                         TIME — hours
                             20
25
      FIGURE 33  CARBON MONOXIDE DIURNAL VARIATIONS
                 AT POINT ARENA (SITE 11)
8
   600
   400
   300
   200
   100
               AVERAGE CONCENTRATION (ppb) — 175 ± 50
                        10        15
                         TIME — hours
                             20
25
   FIGURE 34  CARBON MONOXIDE DIURNAL VARIATIONS
              AT JETMORE (SITE 12)
                          108

-------
A ne.irly 50 percent difference between Site 11 and Site 12 at roughly the
same season (see Table 24) cannot be attributed to the slight difference in
latitude.  Out- could speculate that this gradient is because of a CO source
on the North American continent.  Similar differences between F-ll and F-12,
which are indicators of authropogenic activity, were not observed.  It is
possible that continental land masses provide a large source of natural CO.

C.   NOV, PAN, and 0^
     The NO and N02 levels we measured were typically less than 5 ppb each at
clean sites (Table 24).  While in principle the chemiluminescent instruments
we used  (see Table 9) had adequate sensitivity to measure 1 ppb of NO, we are
unsure of the specificity of this detector.  Since the current instrumenta-
tion cannot be considered very reliable at concentrations of less than 5 ppb,
perhaps  it would be more prudent to say that levels of NO and N0£ in clean
atmospheres were less than 5 ppb.  Recent NOX measurements of Noxon (1978)
and McFarland (1979) suggest NOX levels that are less than 0.1 ppb.  NO levels
of about 4 ppt have been measured by McFarland in the clean marine environ-
ment.  The atmospheric abundance and the fate of NOX in the free troposphere
is poorly understood at present.
     If McFarJand's measurements are correct, it is possible that inorganic
nitrogen is tied up and transported in an organic form.  PAN is one such
species.  We have attempted to measure PAN at some of the clean remote sites.
Figures  35 and 36 show the diurnal variations of PAN at Site 12 (clean con-
tinental) and Site 14 (cluan marine).  The higher levels of PAN (maximum of
0.5 ppb) at Site 12 were probably from the oxidation of alkanes.  At Site 14
the levels were much lower.  The data, however, were quite limited and general
conclusions cannot be drawn.  Meanwhile, if NOX levels in the troposphere are
indeed less than 20 ppt, as McFarland's data might suggest, organic nitrogen
carriers such as PAN could play an important role in the nitrogen chemistry
of the troposphere.  From our data it would appear, however, that PAN levels
in the troposphere are likely to be much less than 100 ppt (Table 24).  Other
organic nitrogen species (such as CH30N02) may also play a similar role.
     Ozone was also measured at several sites at typical background concen-
trations of 30 to 60 ppb.  The very short lifetime of 03 does not allow any

                                     109

-------
    1000
     800
  S  600
  z
    400
     200
                          10        15
                          TIME — hours
20
25
   FIGURE 35  PAN DIURNAL VARIATIONS AT JETMORE (SITE 12)
    1000
     800
     600
     400
     200
                          10        15
                           TIME — hours
FIGURE 36  PAN DIURNAL VARIATIONS AT POINT ARENA (SITE 14)
                            110

-------
      l  cone 1us inns from our limited data.  A much more extensive analysis of
0-j data  from several remote sites in the northern hemisphere has already been
presented by Singh et al. (1978c).
                                      Ill

-------
                                  SECTION 16
            BEST ESTIMATES OF THE MEAN RESIDENCE TIME OF SELECTED
                    MOLECULES BASED ON THE PRESENT STUDY
     The analysis of CH-jCCl^ data supports an average tropospheric HO concen-
tration of 3 to 4 * 1Q5 molecules/cm^.  The role of HO as a primary removal
mechanism for a number of atmospheric chemicals has already been discussed by
a number of investigators (Singh, 1977a, 1977b; Weinstock et al., 1979).  At
this time a seasonally averaged HO abundance of A x 10^ molecules/cm-^ in the
troposphere is our best estimate and has been- supported by other investigators
(Crutzen and Fishman, 1977; Campbell et al., 1979).  A weighted global average
temperature of 265°K was used to estimate the average residence time of a
large number of C^ and €2 molecules.  The rate data for reaction with HO was
taken from two sources (NASA, 1977; Hampson and Garvin, 1978).  In some in-
stances when no rate constants were available, it was possible to estimate
these from theoretical considerations and from comparisons with similar
molecules.  Table 30 reports our best estimate of the mean tropospheric resi-
dence time of 20 selected molecules of atmospheric interest.  It is worthwhile
                                                                            »•
to add that, as a rule of thumb, the percentage of a given material released
at ground level that will enter the stratosphere is roughly twice its resi-
dence time.  As an example, 16.6 percent of C^CCl-j released at ground level,
according to Table 30, would enter the stratosphere, while only 1.2 percent
of CH2BrCH2Br would.  The residence times listed in Table 30 are optimized
towards the tropospheric distribution of CH3CC13.  Residence times of species
with significantly different global distributions could be subject to errors
of ±50 percent because of the expected latitudinal gradients of HO in the
troposphere.
                                      112

-------
                         Table 30
ESTIMATED TROPOSPHERIC RESIDENCE TIMES OF SELECTED MOLECULES
Molecule
CH Cl
CH2C12
CHC13
CHFC12
CHF2C1
CH2C1F
CH^Br
CH4
CO
cci2cci2
CHC1CC12
CH3CC13
CH2C1CH2C1*
CH3CHC12*
CH3CH2C1
CH3CF2C1
CH2BrCH2Br*
C2H6
COS
cs2
HO Reaction Rate
Constant at 265°K
2.96 •
8.38 •
6.51 •
1.85 •
2.28 •
2.42 •
2.74 •
3.70 •
1.40 •
1.02 •
2.30 •
9.64 •
1.41 •
1.71 •
2.57 •
1.57 •
1.41 •
1.79 •
5.60 •
1.90 •
io-14
io-14
io-14
io~14
io-15
io-14
io-14
io-15
io-13
io-13
'io-12
io-15
io-13
io-13
io-13
io-15
io-13
io-13
io-14*
io-13t
Mean Residence
Time''" T (years)
2.7
1.0
1.2
4.3
34.8
3-3.'!
2.9
21.5
0.6
0.8
0.04
8.3
0.6
0.5
0.3
50.6
0.6
0.5
1.6
0.4
     *Rate constant with HO estimated from theoretical considerations.
     'Based on an average HO abundance of 4 * 10-* molecules/cm3.
     ^Calculated at 298°K.
                             113

-------
                                 SECTION 17 |
         URBAN-NONURBAN RELATIONSHIPS OF MEASURED TRACE CONSTITUENTS

     Because many of the trace constituents measured in this study are of man-
made origin, a knowledge of their abundance in urban areas where the actual
releases take place is a useful way to confirm the nature of the source.  In
addition, many of the trace constituents measured here are toxic and otherwise
detrimental to human health.  Since urban centers have high population densi-
ties, the need to characterize the urban atmospheres to establish the intake
of a complex mixture of toxic chemicals clearly exists.  Aware of the poten-
tial toxicity of several chemicals, we undertook a number of field studies in
urban and suburban locations.
     In Table 31 we have summarized data from all sites monitored during this
study.  As stated earlier, Site 6 was merged with Site 1 since only two pollu-
tants were measured at Site 6 (CH-jBr and COC^) and were not measured at
Site 1.
     In Table 31 we present the maximum, minimum, and the average concentra-
tions for each of the species.  The quantity in parentheses is the standard
deviation associated with the average concentration.  The concentrations are
based on hourly averages which were obtained from instantaneous measurements.
In addition, we have presented an average daily outdoor dose for each of the
species.  The standard deviation associated with the daily dose is also
reported based on the collected data.  The daily outdoor dose is reported in
Ug/day and is calculated based on an air intake of 23 m^/day at 25°C and 1 atm
(NAS, 1978).  Table 31 contains a great deal of information that is self-
explanatory; therefore only salient observations will be made here.  In the
following sections we will discuss the urban-nonurban relationships of
chemicals in three subgroups.
                                     114

-------
A .    Chlorof luorocarbons and SF,-

     None of the chlorof luorocarbons nor the SFg that we studied are known  or

suspected to be toxic at the concentrations that were measured.   The following

observations from Table 31 are pertinent:

     •  F-12 and F-ll maximum concentrations did not exceed 4.2  ppb.  The
        highest average concentrations measured in urban environs were less
        than five times the background levels.

     •  F-1.13 and F-114 maximum concentrations  did not exceed 1.2 and 0.2 ppb,
        respectively.  The highest average concentrations were once again less
        than six times the background values.
            maximum levels never exceeded 10 ppt.
        The combined concentration of all f luorocarbons,  based on our data,
        should not exceed 10 ppb even in a highly  polluted urban environment.
        The maximum dose of all f luorocarbons at typical  urban-suburban sites
        should not exceed 250 ug/day.  The typical daily  doses for each
        fluorocarbon are listed in Table 31.
B.    Halogenated (Nonfluorinated) Species

     Unlike the chlorofluorocarbons,  a number of the chlorinated hydrocarbons

are suspected to be highly toxic.  Once again we shall briefly discuss the

levels of exposure to these chemicals.

     •  Typical CCl^ average concentrations in urban areas were about 25 per-
        cent higher than background levels.  This provides evidence for the
        existing urban sources of CCl^ but also suggests that these sources
        are not very large.  The maximum measured concentration of CCl^ did
        not exceed 0.25 ppb.  Because of the relatively uniform distribution
        of CC]/,, an average urban dose of 19 ug/day was calculated.

     •  Unlike CCl^, wide fluctuations in the abundance of CHC13 were observed.
        The maximum concentration, however, never exceeded 0.8 ppb.  Among all
        the sites monitored, the highest average dose of 13 pg/day was calcu-
        lated for Site 8.  Coastal water samples at Site 2 were found to
        contain very high levels of CHC13 (3 ppb of CHC13 in helium that was
        in equilibrium with water).

     •  Since both CCl^ and CHC13 are suspected toxic chemicals, a combined
        30 ug/day dose in urbanized areas is possible.  This can be compared
        to a possible intake of 15 to 30 pg/day from drinking water (NAS,
        1978).  Thus, the intake of CCl^ and CHC13 from the polluted atmo-
        sphere can approach or even exceed that from drinking water.

     •  The CH2C12 concentration did not exceed 2 ppb.  The highest calculated
        average daily dose was 35 yg/day at Site 13, but typically it was much
        lower.  The maximum average concentration at Site 13 was an order of
        magnitude larger than the background concentration of about 40 ppt.


                                     115

-------
        The maximum CH^Cl concentration was less than 3 ppb.  A typical daily
        dose was between 40 to 60 yg/day.  As stated earlier, Cl^Cl is a con-
        stituent of the natural atmosphere but man-made sources do appear to
        be significant.  In Lisbon, as in Los Angeles, the CHjCl concentra-
        tions were found to be almost 3 times the background levels.

        Methyl bromide levels in urban areas are definitely elevated, but
        natural sources may also exist.  The highest CH3Br concentration was
        less than 1 ppb (Site 3) although typicalgCH3Br levels did not exceed
        0.2 ppb.  Among all the sites monitored, the highest dose was calcu-
        lated for Los Angeles and was about 9 yg/day.

        The toxicity of CH3CCl3 is still in question (Farber, 1979).  Concen-
        trations as high as 8 ppb were measured.  Among the various sites
        monitored, the highest average daily dose of 180 yg/day was measured
        at Los Angeles (Site 3) .
                   is a suspected carcinogen.  Its levels,- however, never
        exceeded 0.1 ppb.  Since Cl^BrCH^Br was only measured at a few sites,
        much higher urban concentrations can be expected.

     •  Maximum concentrations of CC12CC12 and CHC1CC12 were 3.7 ppb and
        3.1 ppb, respectively.  Both of these species are suspected to be
        toxic and the latter has been identified as a potential carcinogen
        (Greenberg and Parker, 1979).  The highest average daily dose for
        CC12CC12 and CHC1CC12 was 170 yg/day (Site 8) and 45 yg/day (Site 13).
        respectively.

     •  COC12 was measured at several sites but the concentrations never
        exceeded 0.13 ppb.  The average concentrations were considerably
        lower.  Our data points to a dose that is always less than 5 yg/day.
        The distribution of COC12 is consistent with a photochemical secondary
        source.  This source has been identified to be photooxidation of
        chloroethylenes, particularly CC12CC12 ami CHC1CC1-  (Singh, 1976;
        Singh et al., 1977c).        '                     Z
C.   N00, Hydrocarbons, CO, NO.., and Ozone
     "™"£            L1          X ~ ~""   A/

     The urban-nonurban relationships of these chemicals are self-evident

from Table 31.  The ^0 and the light hydrocarbons measured here are not

suspected to be toxic.  A great deal of literature on the distribution and

effects of CO, NOX, and 03 is available in the literature (EPA, 1977).
                                     116

-------
                        Ifccproefaced from

                        best available copy.
                                                                                                                  m


                                                                                                                  M9ftZl
                                                                                                                                                                  -S"1
121(7)

13*

IAS
                                                                                         2.1(0)    S?W,*
19 •U.6-   m(8»

       ii£
                                                                                                                                                                    J.»(0. 51







                                                                                                                                                                    1.11,0.1) '






                                                                                                                                                                   O.OKO.O3J






                                                                                                                                                                   ia.iw-.jj
                                                                                                                        L,
                                                                                                                                                   f

                                                                                                                                                  , j-
                                                                                                                             »u
                                                                                                                             s

                                                                                                                           "jt


                                                                                                                           !?•• '
                                                                                                                                                                                        117-118

-------
_ 	 ™ 	 _, 	 	 „ 	 _ 	 — 	 	 — — _-- - - 	 	 — 	 —
*"" i --'I -• ! - I - I « i -
f i~i iri j i*. efi-T1 j *i'-^*-. i .•:.*;. i. { i**» is*, j :/•-.*. :-w *•
>:;. - | ±:-i ;
s-^. ! ?••• | -^ j a—
ii» •>- j ;*.i ..c' ! .i-,l- I ...»f',..-.
1* . «_ ; -•• ' | .- • *•* •->+ | j •/* f
f~'T •' * - f ~ *
llr* '
>r-. j *«
;;j.i)C' j ?*.>«:.»•.


i ; - ; , ;.-!..
»i. -
'^*.
1J*(»0.
**.r
-BJ
....••
. " ; ' '''•', i i ' 1 '
*,.
^... li../



-, j -»K

i>: 1
i*> )
j
<«. , ;
^J^l ! X» I
.<;..i t - : •: • •
.3H |
**' j
i 1
l ! ' ', * :
!  :.„.:
                    •j I  n.i :3.- }  •-•.• .'  i .jj.i..-.-!, ; m,:Ji   --.  ;-.;
                      :        j""    '        i1;;
                              i   .            'i
                                                                                              I £,-.. | ^ •

                                                                                              i  11  ;    * L
                                                                                              i  J    !
                                                                                              ! __
                                                                                                                             *  -•
                                                                                                                             (  '
                                                                                                                                                    i        !'
US)(iW
in;
ir»7
                                                                                                                                                            i"
                                                                                                                             J 1*»»*W]
                                                                                                                              IUJ
                                                                                                                             i IMB
                                                                                                                                                       119-120
                                                                                                                                                Precedng page Uank

-------

: - r

: . i
; i
j 	 j ..
;

I'-"'"' ! "
i ^ 4 •-». i
j -.': ' i "
iC_k,_tr7T- j
!
l



"a..-- w**** **. «-!«•

. *

i














;
_


[

" !•
:
i
}
«







1
1












...







"







(
""
1












-•»*
-







_.




'.'-

"












5
*-
-










.tlu*.

S!
,-« .
..
anstsii1
ii»c

;i"








t* i
^•»«
-.
53.9^.11











*1
•?=:
*Ct.'
"

Ka'



"





»Sl« 3i '^icc'-jdXl
!
iit • t-~ , •*--«
.:. ; -- . '?• . -.? .
! *ct ,
; :•*>
i
,,. jy.,!^.
!i-« |
1 ij^T t
! ;* !
! !
]
i j
j !
— i j
i
i i
" r "
: - i
i i




«
.,.,.
-
-:.';.-"
i»*



-






-.. ::
*M

•,-.G:».?s






"




!l


;«lf»i>
im

.«



"




^ '.:
;-«
-
M »(J.n




~






IK
,-

.»..«
S;-i

.,..„.
'"•33 ->0fc)
•"•
.i>-J-!J«JI
i-ir?s
i".





T :j t ««. - ,... , t
"*>*' '• "^^ ! ' -* -- !fc«« *
•-:;- i ::::""' '.--
':' ;;i ! ; ' '
i :!-: i ' • '
„. . . i .. ,-,).,. . ; . , !
"' • •-" " .;.-; i ' ' ' !
1 - • ; ;
*.iK*.9f 1 ''t-k*tU:>tl *5-*-.j. • i
1 *«• [ J I
. --- ! j
;..-*;
u.tur -,• | J:JT )^w»h»- . -
'?»->» - ' -
i - i !
«.i;:r}.,' t ->*-ii'.«jt:^ -
-• ~ i
_u4- -..
(-•--j r { - j
i "*'" i 1



1Z1-122I

-------
                                 SECTION 1C
                     RECOMMENDATIONS FOR FUTURE RESEARCH

     While a great deal about the budgets and fates of chlorinated species
has been learned over the past three years, the data base has been less than
sufficient to quantify the biospheric sources and sinks of several species.
Spot measurements have not been adequate to determine if seasonal variations
of species even exist.  Considerable disagreement over the global distribution
of both reactive and unreactive species exists.  The past research can ade-
quately be characterized as work of an exploratory nature.,
     The best method for understanding and controlling global pollution would
require long-term measurements of chemical species in the free troposphere in
both hemispheres.  Continuous measurements at fixed stations can provide the
high precision that is essential to quantifying tropospheric perturbations of
man-made as well as natural species.  A minimum set would require two stations
in the northern hemisphere and another two in the southern hemisphere.  The
scope of measurements at these stations could be expanded as needs arose.
The characterization of global pollution from a single site, while qualita-
tively useful, does not allow for quantitative analysis of sources and sinks.
     Intel-hemispheric profiles of important trace constituents (such as
  o., and C^C^) need to be better characterized to ascertain seasonal
variations in the global distribution of these species.  Such latitudinal
profiles can be used to confirm possible asymmetries in the hemispheric HO
distribution and the reasons for such asymmetries.  These also provide an
essential means for validating global 2-dimensional models that are cur-
rently in a state of active development.  The possibility exists that
species such as CH2C12 could be used to understand seasonal variations of
HO in the natural atmosphere.
     The ocean appears to be a major source of methyl halides and N20.  Very
little oceanic data on methyl halides is available.  It is still unknown
                                     123
Preceding page blank

-------
whether ocean fluxes of methyl halides and ^0 have any seasonal fluctuations.
The role of oceans as sources or sinks of N£0 remains unresolved for lack of
an extensive data base.  The bromine and iodine atmospheric distributions and
budgets are at be!st poorly understood.  A number of trace constituents mea-
sured in the clean atmosphere are also toxic.  The urban abundances of these
and other toxic chemicals need to be characterized to determine population
exposure.
                                     124

-------
                                 SECTION 19
                            GENERAL OBSERVATIONS

     We concluded that the atmosphere is contaminated by a large number of
man-made chemicals that have never existed in nature.  These man-made chemi-
cals interfere with the chemistry and radiation balance of the atmosphere
with potentially deleterious consequences.  Fully three-quarters of the
chlorine in the atmosphere is attributable to man-made activities.  In many
coses, nature appears to have no means to cleanse itself of pollutants in a
harmless fashion.
     It appears that fluorocarbons are increasing rapidly in the global atmo-
sphere.  The only removal mechanism appears to be photolytic destruction in
the stratosphere.  Unfortunately, the products of these fluorocarbons
(chlorine atoms) can catalytically destroy stratospheric ozone.  This could
cause an increase in the rate' of skin cancer as well as other unknown climatic
effects.
     Halocarbons do, however, serve a beneficial purpose.  We have used them
to study the dynamics of the atmosphere as well as the fundamental removal
processes that are operative.  We found that removal mechanisms in the gas
phase are almost an order of magnitude slower than previously believed.  The
chances that man's activities are further reducing the scavenging ability of
the natural atmosphere are evident.
     Nitrous oxide, unlike halocarbons, is found to have major sources in
ocean waters and does not appear to be changing.  Nitrous oxide, as well as
most halocarbons, absorbs infrared radiation.  These chemicals have the
potential to exacerbate the warming that may be caused by excessive C02.
     Because of significant man-made sources of halocarbons, levels in urban
areas are one to two orders of magnitude higher than in clean environments.
It is further becoming apparent that many of these chemicals are toxic and
carcinogenic, thus compounding their potential harmful effects.
                                     125

-------
                            LIST OF PUBLICATIONS


     The following reports and articles have been published to date from the

results of this study.


Journal Articles:

Singh, H. B., 1976:  "Phosgene in the Ambient Air," Nature, 264, pp. 428-429.

Singh, H. B., D. P. Fowler, and T. 0. Peyton, 1976:  "Atmospheric Carbon
     Tetrachloride:  Another Man-Made Pollutant," Science, 192, pp. 1231-1234.

Singh, H. B., 1977:  "Atmospheric Halocarbons:  Evidence in Favor of Reduced
     Average Hydroxyl Radical Concentration in the Troposphere," Geophys.
     Res. Let. , 4_, pp. 101-104.

Singh, H. B., L. Salas, and L. A. Cavanagh, 1977:  "Distribution, Sources,
    .and Sinks of Atmospheric Halogenated Compounds," J. Air. Poll. Contr.
     Assoc., 27, pp. 332-376.

Singh, H. B., 1977:  "Preliminary Estimation of Average Tropospheric HO
     •Concentrations in the Northern and Southern Hemispheres," Geophys. Res.
     Lett.. 4, pp. 453-456.

Singh, H. B., L. Salas, D. Lillian, and R. R. Arnts, 1977:  "Generation of
     Accurate Halocarbons Primary Standards Using Permeation Tubes," Environ.
     Sci. Technol., 11, pp. 511-513.

Singh, H. B. , L. Salas, H. Shigeishi, and A. Crawford, 1977:  "Urban-Nonurban
     Relationships of Halocarbons, SF6, N20, and Other Atmospheric Trace Con-
     stituents," A tin. Env. . 11, pp. 819-828.

Singh, H. B., L. Salas, H. Shigeishi, and E. Scribner, 1979:  "Atmospheric
     Halocarbons, Hydrocarbons and SFg:  Global Distributions, Sources and
     Sinks," Science. 203, pp. 899-903.

Singh, H. B., L. J. Salas, and H. Shigeishi, 1979:  "The Distribution of
     Nitrous Oxide (N20) in the Global Atmosphere and the Pacific Ocean,"
     Tellus, 31, 313-320.
                                     126

-------
Reports:

Singh, H. B., L. J. Salas, H. Shigeishi, and L. Cavanagh, 1976:  "Atmospheric
     Fates of Halogenated Compounds:  First Year Summary Report," EPA Grant
     No.  R-80380201, SRI Project No. 4487, Interim Report, Stanford Research
     Institute, Menlo Park, California.

Singh, H. B., L. Salas, H. Shigeishi, and A. J. Smith, 1978:   "Fate of
     Halogenated Compounds in the Atmosphere," EPA Publication No.  EPA-600/
     3-78-017, Environmental Protection Agency, Research Triangle Park, North
     Carolina.

Singh, H. B., L. J. Salas, H. Shigeishi, and E. Scribner, 1978:  "Global
     Distribution of Selected Halocarbons, Hydrocarbons, SF^, and N20," EPA
     Publication No. EPA-600/3-78-100, Environmental Protection Agency,
     Research Triangle Park, North Carolina.
                                      127

-------
                                  REFERENCES
Altshuller, A. P., 1976?  "Average Tropospheric Concentration of Carbon Tetra-
     chloride Based on Industrial Production, Usage, and Emissions," Env.Sci.
     Tech., 10, -pp. 596-598.

Anderson,  J. G., et al., 1979:  "Stratospheric Free Chlorine Measured by
     Balloon Borne In Situ Resonance Fluorescence," Jour. Geophys. Res.
     (submitted for publication).

Bauer, E., 1978:  "A Catalog of Perturbing Influences on Stratospheric Ozone,
     1955-1975," Contract DOT FA 7WA3965, Prepared by Institute of Defence
     Analysis (Paper P-1340), Arlington, Virginia.

Berg, W. W. and J. W. Winchester, 1976:  "Organic and Inorganic Gaseous
     Chlorine in Relation to the Particle Size Distribution of Chloride In the
     Marine Aerosol," Proceedings of the Nonurban Tropospheric Composition
     Symposium, Hollywood, Florida, November 10-12, 1976.

Blackmer, A. M. and J. M. Bremner, 1976:  "Potential of Soil as a Sink for
     Atmospheric Nitrous Oxide," Geophys. Res. Lett., _3, pp. 739-742.

Broecker, W. S. and T. H. Peng, 1974:  "Gas Exchange Rates Between Air and
     Sea>" Tellus, 26., pp. 39-46.

Campbell, M. J., J. C. Sheppard and B. F. Au, 1979:  "Measurement of Hydroxyl
     Radical Concentrations in Boundary Layer Air by Monitoring CO Oxidations,"
     Geophys. Res. Lett. , _6, pp. 185-188.

Chang, J.  S. and J. E. Penner, 1978:  "Analysis of Global Budgets of Halo-
     carbons," Atm. Env., 12, pp. 1867-1873.

CIS, 1976:  Chemical Information Service, Stanford Research Institute, Menlo
     Park, California.

Craig, H.  and L. I. Gordon, 1963:  "Nitrous Oxide in the Ocean and in the
     Marine Atmosphere,"  Geochimica et Cosmochimica Acta, 27, pp. 949-955.

Cronn, D. A., et al., 1976:  "Measurement of Tropospheric Halocarbons by Gas
     Chromatography-Mass Spectrometry," Washington State University, Interim
     Report, Submitted to EPA, Grant R-0804033.

Crutzen,  P. J., 1976:  "Upper Limits of Atmospheric Ozone Reductions Following
     Increased Application of Fixed Nitrogen to the Soil," Geophys. Res. Let t.,
     _3, pp. 169-172.

                                     128

-------
Crutzen, P-. .1 .  and J. Fishman, 1977:  "Average Concentration of OH in the
     Northern Hemisphere Troposphere, and the Budgets of CH4 , CO, and H2,"
     ilVJlPJ.'-ys.--- Kcs . Le _tt_L , 4^, pp. 321-324.
C/.c'plak, G. and C. Junge, 1974:  "Studies of Interhemispheric Exchange in the
     Troposphere by Diffusions Model,"  Adv.. Geophys . , 18B , pp. 57-67.

Davis, 1). D., W. Heaps and T. McGee, 1976:  "Direct Measurement of Natural
     Tropospheric Levels of OH via an Aircraft-Borne Tunable-Dye Laser,"
     Geophys. Res. Lett., 3>, pp. 331-333.

Dilling, W. L., 1977:  "Interphase Transfer Processes II:  Evaporation Ratio
     of Chloro Methanes, Ethanes, Ethylenes, Propanes, and Propylenes from
     Dilute Aqueous Solutions.  A Comparison with Theoretical Predictions,"
     Env. Sci. Tech. . pp. 405-409.

K. 1. du Pont de Nemours, 1978:  "World Production and Release of Chloro-
     f luorocarbons-11 and -12 through 1977," E. I. du Pont de Nemours and
     Company, Wilmington, Delaware, 26 June 1978.

Ellsaesser, H. W., 1977:  "Has Man Increased Stratospheric Ozone?" Nature , 270,
     pp. 592-593.

EPA, 1977:  "Proceedings, International Conference on Photochemical Oxidant
     Pollution and Its Control," Raleigh, North Carolina, Sept. 12-17, 1977,
     U.S. Environmental Protection Agency, Research Triangle Park, North
     Carolina.

Farber, H. P., 1979:  "1, 1, 1-Trichloroethane As An Industrial Solvent:  A
     Review of Current Health and Environmental Knowledge," Dow Chemicals
     U.S.A., Midland, Michigan.

Gnlbally, I. E., 1976:  "Man-Made Carbon Tetrachloride in the Atmosphere,"
     Science , 193 , pp. 573-576.

Greenherg, M. M. and J. C. Parker, 1979:  "Health Assessment Document for
     Tetrachlorethylene," U.S. Environmental Protection Agency, External
     Review Draft No. 1, Research Triangle Park, North Carolina.

Grimsrud, E. P. and R. A. Rasmussen, 1975:  "The Analysis of Chlorofluoro-
     carbons in the Troposphere by Gas Chromatography-Mass Spectrometry , "
     At mos. Environ. , 9, pp. 1010-1016.

Hahn, J., 1974:  "The North Atlantic Ocean as a Source of N20," Tell us , 26,
     pp. 160-168.

Hahn, J. and C. Junge, 1977:   "Atmospheric Nitrous Oxide:  A Critical Review,"
     Z. Naturforsch, 32a, pp. 190-214.

Hampson, R. F. and D. Garvin, 1978:  "Reaction Rate  and Photochemical Data
     For Atmospheric Chemistry-1977 ," National Bureau of  Standards, Washington,
     D.C. , May 1978.

                                      129

-------
Hanst, P. L., et al., 1975:  "infrared Measurements of Halogenated Hydrocarbons
     and Other Atmospheric Trace Gases," J. of Air Poll. Control Assoc. ,  25,
     pp. 1220-1226.

Isaksen, I.S.A. and P. J. Crutzen, 1977:  "Uncertainties in Calculated
     Hydroxyl Radical Densities in the Troposphere and Stratosphere,"
     Geophysica Norvegica, 31, pp. 1-10.

Jesson, J. P., P. Meakin and L. C. Glasgow, 1977:  "The Fluorocarbon-Ozone
     Theory II:  Tropospheric Lifetimes—An Estimate of the Tropospheric
     Lifetime of CC13F," Atm. Env., .U, pp. 499-508.

Johnston, H. S., 1977:  "Analysis of Independent Variables In The Perturbation
     of Stratospheric Ozone by Nitrogen Fertilizer," Jour. Geophys. Res., 82,
     pp. 1767-1772.

JPL, 1979:  "Chemical Kinetics and Photochemical Data for Use in Stratospheric
     Modeling," JPL Publication 70-27, Pasadena, California.

Junge, C., et al., 1971:  "N20 Measurements In Air and Seawater Over the
     Atlantic," Meteor. Forsch. Ergebrisse. Reihe B 6, pp. 1-11.

Junge, C. E., 1974:  "Residence Time and Variability of Tropospheric Trace
     Gases," Tellus, 26, pp. 477-488.

Junge, C. E. , 1976:  "The Role of the Oceans As A Sink for Chlorofluoro-
     methanes and Similar Compounds," Z. Naturforsch, 31a, pp. 482-490.

Krasnec, J., 1979:  private communication, National Center for Atmospheric
     Research (NCAR), Boulder, Colorado.

Lillian, D. and H. Singh, 1974:  "Absolute Determination of Atmospheric Halo-
     carbons by Gas Phase Coulometry, Anal Chem., 46, pp. 1060-1063.

Linnebom, V. J., J. W. Swinnerton and R. A. LaMontagne, 1973:  "The Ocean as
     a Source of Atmospheric Carbon Monoxide," Jour. Geophys. Res., 78,
     pp. 5333-5340.

Liu, S. C., et al., 1977:  "Sources and Sinks of Atmospheric N20 and the
     Possible Ozone Reduction Due to Industrial Fixed Nitrogen Fertilizers,"
     Tellus, 29,, pp. 251-263.

Lovelock, J. E., R. J. Maggs and R. J. Wade, 1973:  "Halogenated Hydrocarbons
     In and Over the Atlantic," Nature, 241. pp. 194-196.

Lovelock, J. E., 1974:  "Atmospheric Halocarbons and Stratospheric Ozone,"
     Nature, 252, pp. 292-294.

Lovelock, J. E., 1975:  "Natural Halocarbons in Air and in the Sea," Nature,
     256, pp. 193-194.
                                      130

-------
Lovelock, J. E., 1977:  "Methyl Chloroform in the Troposphere as an Indicator
     of OH Radical Abundance," Nature, 267, pp. 32-33.

Markham, A. E. and K. A. Kobe, 1941:  "Solubility of Carbon Dioxide and
     Nitrous Oxide in Salt Solution," J. Amer. Chem. Soc., 63, pp. 499-454.

McConnell, J. C. and H. I. Schiff, 1978:  "Methyl Chloroform:  Impact on
     Stratospheric Ozone," Science, 199, pp. 174-177.

McElroy, M. B., et al., 1976:  "Sources and Sinks for Atmospheric N20," Rev.
     Geophys. Space Phys., 14, pp. 143-150.

McFarland, G., 1979:  private communication, National Oceanographic and
     Atmospheric Administration (NOAA), Boulder, Colorado.

Molina, M. J. and F. S. Rowland, 1974:  "Stratosphere Sink for Chlorofluoro
     Methanes:  Chlorine Atom Catalyzed Destruction of Ozone," Nature, 249,
     pp. 810-812.

NAS, 1976:  "Halocarbons:   Effects on Stratospheric Ozone," Panel on Atmo-
     spheric Chemistry, National Academy of Sciences, Washington, D.C.

NAS, 1978:  "Chloroform, Carbon Tetrachloride, and Other Halomethanes:  An
     Environmental Assessment," National Academy of Sciences, Washington, D.C.

NASA, 1977:  "Chlorofluoromethane," Assessment Workshop Report, NASA Goddard
     Space Flight Center,  March 1977.

Neely, W. B. and J. H. Plonka, 1978:  "An Estimation of the Time-Averaged
     Hydroxyl Radical Concentration in the Troposphere," Env. Sci. and Techno.,
     12, pp. 317-321.

Newell, R., et al., 1969:   "Interhemispheric Mass Exchange from Meteorological
     and Trace Substance Observations," Tellus, 21, pp. 641-647.

Noxon, J. F., 1978:  "Tropospheric N02," J. Geophys. Res., 83, pp. 3051-3057.

O'Keefe, A. E. and G. C. Ortman, 1966:  "Primary Standards From Trace Gas
     Analysis," Anal. Chem., 38, pp.  760-762.

Pack, D. H., et al., 1977:  "Halocarbon Behavior From a Long Time Series,"
     Atm. Env. , 11., pp. 329-344.

Palmer, T. Y., 1976:  "Combustion Sources of Atmospheric Chlorine," Nature,
     263. pp. 44-46.

Perner, D., et al., 1976:  "OH-Radicals in the Lower Troposphere," Geophys.
     Res. Lett., 3^, pp. 466-468.

Ramanathan, V., 1975:  "Greenhouse Effect Due  to Chlorofluorocarbons:
     Climatic  Implications," Science, 190, pp. 50-52.
                                      131

-------
Rasmussen, R. A., et al., 1976:  "Trip Report on the Cruise of the Alpha Helix
     Research Vessel," Grant No. OCE 75 04688 A03, Washington State University,
     Pullman, Washington.

Rasmussen, R. A., R. B. Chatfield and M. W. Holdren, 1977:  "Hydrocarbon
     Species in Rural Missouri Air," College of Engineering, Washington State
     University (unpublished manuscript).

Rasmussen, R. A., 1979:  private communication, Oregon Graduate Center.

Rowland, F. S., 1979:  private communications, University of California,
     Irvine, California.

Seiler, W., 1974:  "The Cycle of Atmospheric CO," Tellus, 2£, pp. 116-135.

Singh, H. B., et al. , 1975a:   "Atmospheric Formation of Carbon Tetrachloride
     from Tetrachloroethylene," Env. Letters. 10, pp. 253-256.

Singh, H. B., D. Lillian and A. Appleby, 1975:  "Absolute Determination of
     Phosgene:  Pulse Flow Coulometry," Anal. Chem., 47, pp. 860-864.

Singh, H. B., 1976:  "Phosgene in the Ambient Air," Nature, 264, pp. 428-429.

Singh, H. B., D. P. Fowler and T. 0. Peyton, 1976a:  "Atmospheric Carbontetra-
     chloride:  Another Man-Made Pollutant," Science, 192, pp. 1231-1234.

Singh, H. B., et al., 1976b:  "Atmospheric Fates of Halogenated Compounds:
     First Year Summary Report," EPA Grant R-80380201.  SRI Project No. 4487,
     Interim Report, Stanford Research Institute, Menlo Park, California.

Singh, H. B., 1977a:  "Atmospheric Halocarbons:  Evidence in Favor of Reduced
     Average Hydroxyl Radical Concentration in the Troposphere," Geophys.
     Res. Lett., <4, pp. 101-104.

Singh, H. B. 1977b:  "Preliminary Estimation of Average Tropospheric HO Concen-
     trations in the Northern and Southern Hemispheres," Geophys. Res. Lett.,
     _4, pp. 453-456.

Singh, H. B., L. Salas and L. A. Cavanagh, 1977a:  "Distribution, Sources,  and
     Sinks of Atmospheric Halogenated Compounds," J. Air. Poll. Contr. Assoc.,
     21, pp. 332-376.

Singh, H. B., et al., 1977b:  "Generation of Accurate Halocarbons Primary
     Standards Using Permeation Tubes," Environ. Sci. Technol., 11,
     pp. 511-513.

Singh, H. B., L. Salas, H. Shigeishi and A. Crawford, 1977c:  "Urban-Nonurban
     Relationships of Halocarbons, SFfc, N20, and Other Atmospheric Trace Con-
     stituents," Atm._Jnv. , 11, pp. 819-828.
                                      132

-------
Singh, H. B., et al., 1977d:  "Atmospheric Fates of Halogenated Compounds:
     Second Year (Phase I) Summary Report," EPA Grant R-80380202, SRI Project
     4487, Stanford Research Institute, Menlo Park, California.

Singh, H. B., et al., 1978a:  "Fate of Halogenated Compounds in the Atmo-
     sphere," EPA-600/3-78-017, Environmental Protection Agency, Research
     Triangle Park, North Carolina.

Singh, H. B., L. J. Salas, H. Shigeishi and E. Scribner, 1978b:  "Global
     Distribution of Selected Halocarbons, Hydrocarbons, SFfc, and N20,"
     EPA-600/3-78-100, Environmental Protection Agency, Research Triangle
     Park, North Carolina.

Singh, H. B., F. L. Ludwig and W. B. Johnson, 1978c:  "Tropospheric Ozone:
     Concentrations and Variabilities in Clean Remote Atmospheres," Atm.
     Env. , 12_, pp. 2185-2196.

Singh, H. B., et al., 1979a:  "Atmospheric Halocarbons, Hydrocarbons and SF6:
     Global Distributions, Sources and Sinks," Science, 203, pp. 899-903.

Singh, H. B., L. J. Salas and H. Shigeishi, 1979b:  "The distribution of
     Nitrous Oxide (N20)  in the Global Atmosphere and the Pacific Ocean,"
     Tellus (in press).

Sze, N. D., 1977:  "Anthropogenic CO Emissions:  Implications for the Atmo-
     spheric CO-OH-CH4 Cycle," Science, 195. pp. 673-675.

Wang, W. C., et al., 1976:  "Greenhouse Effects Due to Man-made Perturbations
     of Trace Gases," Science, 19A, pp. 685-690.

WeinsLock, B., H. Niki and T. Y. Chang, 1979:  "Chemical Factors Affecting the
     Hydroxyl Radical Concentration in the Troposphere," Advances in Env. Sci.
     and Tech.  (in press).

WesthiM-g, H. H., R. A. Rasmussen and M. Holdren, 1974:   "Gas Chromatographic
     Analysis of Ambient  Air for Light Hydrocarbons Using a Chemically Bonded
     Stationary Phase," Anal. Ghent. , 46, pp. 1852-1854.

Wofsy,  S. C., 1976:  "Interactions of CH4 and CO in the  Earth's Atmosphere,"
     Annual Review of Earth and Planetary Sciences, pp.  441-468.

Wofsy,  S. C., M. B. McElroy and Y. L. Yung, 1975:  "The  Chemistry of Atmo-
     spheric Bromine," Geophys. Res. Lett., 2^ pp. 215-218.

Zafiriou, 0., 1975:  "Reaction of Methyl Halide with Seawater and Marine
     Aerosols," J. Marine Res., 33, pp. 73-49.
                                      133

-------