AN  ASSESSMENT  OF THE  RISKS  OF
          STRATOSPHERIC  MODIFICATION
              Volume II:   Chapters  1-6
                   Submission to the

                Science Advisory Board
        U.S.  Environmental  Protection  Agency
                           By

             Office of Air  and Radiation
        U.S.  Environmental  Protection Agency
                     October 1986
          Comments  should be  addressed to:

                    John  S. Hoffman
   U.S.  Environmental Protection Agency,  PM 221
                  401  M Street,  S.W.
                Washington, D.C.   20460
                          USA
The following report is being submitted to the Science Advisory Board and
to the Public for review and comment.  Until the Science Advisory Board
review has been completed and the document is revised, this assessment
does not represent EPA's official position on the risks associated with
Stratospheric Modification.  This report has been written as part of:'the
activities of the EPA's congressionally-established Science Advisory Board,
a public group providing extramural advice on scientific issues.  The Board
is structured to provide a balanced independent expert assessment of scientific
issues it reviews, and hence, the contents of this report do not necessarily '-
represent the views and policies of the EPA nor of other agencies in the^
Executive Branch of the Federal Government.  Until the final report is
available, EPA requests that none of the information contained in this.
draft be cited or quoted.  Written comments should be sent to: John S. Hoffman
at the EPA by November 14, 1986.                                       ..' '

-------
                             ACKNOWLEDGEMENTS
    Many persons helped make possible this risk assessment  and  the  original
work upon which it is based.  Contributors include:   Steve  Anderson, Michael
Gibbs, Peter Gleick, Kevin Hearle,  John Hoffman,  Project  Director  for  this
risk assessment, Larry Kalkstein,  Patricia Lill,  Janice Longstreth, Johnathan
Overpeck, Neil Patel, Hugh Pitcher,  Cynthia Rosenweig,  Holly  Stallworth, Alan
Teramura, Dennis Tirpak, Jim Titus,  Kathleen Valimont,  John Wells,  and Robert
Worrest.

    Numerous reviewers provided assistance and guidance during  the  preparation
of these materials.  Although too  numerous to name,  we  would  like  to thank
them for their keen insights and the time they devoted  to this  endeavor.

    The production of this document  was made possible through the  considerable
efforts of Susan Farris, Cynthia Whitfield, Lee Neff and  Susan  MacMillan.
                                 DRAFT FINAL  * * *

-------
Table of Contents

-------
                             TABLE OF CONTENTS

                                                                        PAGE

VOLUME I

EXECUTIVE SUMMARY 	   ES-1

VOLUME II

ACKNOWLEDGEMENTS 	   i

INTRODUCTION 	   1

    The Rise of Concern About Stratospheric Change 	   1
    Concern About Public Health and Welfare Effects of Global
      Atmospheric Change 	   1
    Need for Assessments 	   2

1.   GOALS AND APPROACH OF THIS RISK ASSESSMENT 	    1-1

    Analytic Framework 	    1-1
    Supporting Documents and Analysis  for this Review 	    1-2
    Chapter Outlines	    1-2
    Errors and Corrections 	 	    1-3

2.   STRATOSPHERIC PERTURBANTS:  PAST CHANGES IN CONCENTRATIONS
        AND FACTORS THAT DETERMINE CONCENTRATIONS 	    2-1

    Summary 	    2-1
    Findings 	    2-3
    Measured Increases in Tropospheric Concentrations of
        Potential Ozone Depletors 	    2-4
    Measured Increases in Tropospheric Concentrations of
        Potential Ozone Increasers 	   2-13
    Factors that Influence Trace Gas Lifetimes 	    2-18
    Long-Lived Trace Gases 	    2-21
    Trace Gases with Shorter Lifetimes 	    2-25
    Carbon Dioxide and the Carbon Cycle 	    2-25
    Source Gases for Stratospheric Sulfate Aerosol (OCS, CS2) 	    2-26
    Appendix A: CFC Emissions-Concentrations Model 	    2-27
    References 	    2-30

3.   EMISSIONS OF INDUSTRIALLY PRODUCED POTENTIAL OZONE MODIFIERS 	    3-1

    Summary 	    3-1
    Findings 	    3-3
    Introduction 	    3-5
    Chlorofluorocarbons 	    3-5
    Chlorocarbons 	    3-55
    Halons 	    3-57
    References 	    3-62


                          * * -  DRAFT FINAL  * * *

-------
                              TABLE OF CONTENTS
                                 (Continued)

                                                                        PAGE

4.  FUTURE EMISSIONS AND CONCENTRATIONS OF TRACE GASES WITH
        PARTLY BIOGENIC SOURCES 	    4-1

    Summary 	    4-1
    Findings 	    4-2
    Factors Influencing Future Trace Gas Concentrations 	    4-4
    How Trace Gases Influence the Stratosphere and Troposphere 	    4-4
    The Lifetime of Emissions and the Predictability of
        Future Concentrations 	    4-7
    Scenarios of Trace Gases	    4-7
    Effects of Possible Future Limits on Global Warming 	    4-24
    Conclusion	    4-27
    References 	    4-29

5.  ASSESSMENT OF THE RISK OF OZONE MODIFICATION 	    5-1

    Summary 	    5-1
    Findings 	    5-3
    Introduction	    5-7
    Equilibrium Predictions for Two Dimensional Models 	    5-22
    Time Dependent Predictions for One Dimensional
        Models for Different Scenarios of Trace Gases 	    5-36
    Time Dependent Predictions for Two Dimensional Models
        with Different Scenarios of Trace Gases 	    5-59
    Models Fail to Represent All Processes That Govern
        Stratospheric Change in a Complete and Accurate Manner 	    5-72
    The Implications of Ozone Monitoring for  Assessing Risks
        of Ozone Modification 	    5-100
    References 	    5-124

6.  CLIMATE 	    6-1

    Summary 	    6-1
    Findings 	    6-2
    The Greenhouse Theory 	    6-6
    Radiative Forcing by Increases in Greenhouse Gases 	    6-7
    Ultimate Temperature Sensitivity 	    6-14
    The Timing of Global Warming 	    6-18
    Regional Changes in Climate Due to Global Warming 	    6-19
    Effects of Possible Control of Greenhouse Gases
        on the Stratosphere 	    6-26
    Appendix A:  Description of Model to be Used in Integrating
                 Chapter 	    6-27
    Appendix B:  Trace Gas Scenarios 	    6-32
    References 	    6-33
                                 DRAFT FINAL  *

-------
                              TABLE OF CONTENTS
                                 (Continued)

                                                                        PAGE

VOLUME III

7.   NONMELANOMA SKIN TUMORS 	    7-1

    Summary 	    7-1
    Findings 	    7-2
    Introduction 	    7-4
    Biology of Nonmelanoma Skin Tumors:  Links to UVB 	    7-4
    Epidemiological Evidence 	    7-18
    Dose-Response Relationship 	    7-30
    Appendix A 	    7-37
    References 	    7-46

8.   CUTANEOUS MALIGNANT MELANOMA 	    8-1

    Summary 	    8-1
    Findings 	    8-3
    Introduction 	    8-7
    Background Information 	     8-7
    Epidemiologic Evidence	 . .    8-16
    Experimental Evidence	    8-29
    Dose-Response Relationships 	    8-30
    References 	    8-40

9.   EFFECTS OF SOLAR RADIATION ON THE IMMUNE  SYSTEM
        AND RESISTANCE TO INFECTIONS 	    9-1

    Summary 	    9-1
    Findings 	    9-3
    Introduction 	    9-5
    Effects of Ultraviolet Radiation on the Immune System	 ..    9-5
    Antigen Presentation In Vitro 	    9-11
    Langerhans Cells and Antigen Presentation 	    9-13
    Human Studies 	    9-16
    Effects of Ultraviolet Radiation on Infectious Diseases 	    9-17
    Conclusions 	    9-22
    Appendix 9A -- Radiation Sources 	    9-24
    References 	    9-25

10.  CATARACTS AND OTHER EYE DISORDERS 	   10-1

    Summary 	   10-1
    Findings 	   10-2
    Cataracts 	   10-3
    Potential Changes in Senile Cataract Prevalence for
        Changes in UV-B 	   10-30
    Other Eye Disorders 	   10-35
    References 	   10-39

                          * * *  DRAFT FINAL   * * *

-------
                              TABLE OF CONTENTS
                                 (Continued)

                                                                        PAGE

11.  RISKS TO CROPS AND TERRESTRIAL ECOSYSTEMS FROM ENHANCED
        UV-B RADIATION 	   11-1

    Summary 	   11-1
    Findings 	   11-2
    Introduction 	   11-4
    Issues and Uncertainties in Assessing the Effects
        of UV-B Radiation on Plants 	   11-4
    Issues Concerning UV Dose and Current Action Spectra
        for UV-B Impact Assessment 	   11-4
    Issues Concerning Natural Plant Adaptations to UV 	   11-6
    Issues Associated with the Extrapolation  of Data from
        Controlled Environments to the Field  	   11-9
    Uncertainties in Our Current Knowledge of UV-B Effects on
        Terrestrial Ecosystems and Plant Growth Forms 	   11-11
    Uncertainties with the Ability to Extrapolate Knowledge to Higher
        Ambient C02 Environment and Other Atmospheric Pollutants 	   11-11
    Risks to Crop Yield Resulting from an Increase in
        Solar UV-B Radiation 	   11-14
    Risks to Yield Due to a Decrease in Quality 	   11-19
    Risks to Yield Due to Possible Increases  in
        Disease or Pest Attack 	   11-21
    Risks to Yield Due to Competition with Other Plants 	   11-22
    Risks to Yield Due to Changes in Pollination and Flowering 	   11-23
    References 	   11-25

12.  AN ASSESSMENT OF THE EFFECTS OF ULTRAVIOLET-B
        RADIATION ON AQUATIC ORGANISMS 	   12-1

    Summary 	   12-1
    Findings 	   12-2
    Introduction 	   12-4
    Background on Marine Organisms and Solar  Ultraviolet
        Radiation 	   12-4
    Effects of UV-B Radiation in Phytoplankton 	   12-9
    Effects on Invertebrate Zooplankton 	   12-11
    Effects on Ichthyoplankton (Fisheries)	   12-24
    Conclusions 	   12-31
    References 	   12-32
                            * *  DRAFT FINAL  * * *

-------
                              TABLE OF CONTENTS
                                 (Continued)

                                                                        PAGE

13.  EFFECTS OF UV-B ON POLYMERS 	   13-1

    Summary 	   13-1
    Findings 	   13-2
    Photodegradation of Polymers 	   13-4
    Polymers in Outdoor Uses and the Potential for Degradation 	   13-7
    Damage Functions and Response to Damage 	   13-14
    Effect of Temperature and Humidity on Photodegradation 	   13-29
    Future Research 	   13-30
    References 	   13-31

14.  POTENTIAL EFFECTS OF STRATOSPHERIC OZONE  DEPLETION ON
        TROPOSPHERIC OZONE (SMOG) 	   14-1

    Summary 	   14-1
    Findings 	   14-2
    Introduction 	   14-3
    Potential Effects of Ultraviolet Radiation and Increased
        Temperatures on Urban Smog 	   14-8
    Conclusions and Future Research Directions 	   14-11
    References 	   14-14

15.  CAUSES AND EFFECTS OF SEA-LEVEL RISE 	   15-1

    Summary 	   15-1
    Findings 	   15-2
    Causes of Sea-Level Rise 	   15-5
    Effects of Sea-Level Rise 	   15-17
    Conclusion 	   15-33
    Notes 	   15-34
    Appendix A 	   15-35
    References 	   15-36

16.  POTENTIAL EFFECTS OF FUTURE CLIMATE CHANGES ON FORESTS
        AND VEGETATION, AGRICULTURE, WATER RESOURCES
        AND HUMAN HEALTH 	   16-1

    Summary 	   16-1
    Findings 	   16-5
    References 	   16-10
                          * * *  DRAFT FINAL  * »

-------
                              TABLE OF CONTENTS
                                 (Continued)

                                                                        PAGE

17.  MODELS FOR INTEGRATING THE ANALYSES OF HEALTH AND
        ENVIRONMENTAL RISKS ASSOCIATED WITH OZONE MODIFICATION  	   17-1

    Summary 	   17-1
    Introduction 	   17-2
    The Model as a Framework 	   17-2
    Analysis Procedure 	   17-4
    Model Limitations 	   17-9
    References 	   17-12
    Appendix A:  Model Design .and Model Flow 	    A-l
    Appendix B:  Scenarios of Chemical Production, Population,
        and GNP  	    B-l
    Appendix C:  Evaluation of Policy Alternatives 	    C-l
    Appendix D:  Emissions of Potential Ozone-Depleting Compounds  	    D-l
    Appendix E.  Atmospheric Science Module 	    E-l
    Appendix F:  Health and Environmental Impacts of Ozone
        Depletion 	    F-l

18.  HUMAN HEALTH AND ENVIRONMENTAL EFFECTS 	   18-1 .

    Summary 	   18-1
    Findings 	   18-2
    Introduction 	   18-4
    Methods for  Estimating Health and Environmental Risks  	   18-9
    Health and Environmental Risks:  Central Case 	   18-11
    Comparison of Central Case with Results Using Alternative
        Assumptions 	   18-17
    Sensitivity  of Effects to Parameter Uncertainty 	   18-36
    Relative Importance of Key Uncertainties 	   18-46
    References 	   18-52

VOLUME IV

    Appendix A 	

VOLUME V

    Appendix B 	
                              *  DRAFT FINAL  * * »

-------
                            LIST OF EXHIBITS

                                                                    Page

1-1    Relationships Among the Chapters 	    1-4

2-1    Measured Increases in Tropospheric Concentrations
       of CFC-11 (CFC-13) 	    2-5

2-2    Measured Increases in Tropospheric Concentrations
       of CFC-12 (CF2CL2) 	    2-6

2-3    Measured Increases in Tropospheric Concentrations
       of CFC-22 (CHC1F2) 	    2-7

2-4    Measured Increases in Tropospheric Concentrations
       of CFC-113 (C2CL3F3) 	    2-7

2-5    Measured Increased in Tropospheric Concentrations
       of Carbon Tetrachloride (CC14) 	    2-9

2-6    Measured Increases in Tropospheric Concentrations
       of Methyl Chloroform (CH3CC13) 	    2-10

2-7    Measured Increases in Tropospheric Concentrations
       of Halon-1211 (CF2ClBr) 	    2-11

2-8    Measured Increases in Tropospheric Concentrations
       of Nitrous Oxide (N20) 	    2-12

2-9    Measured Increases in Tropospheric Concentrations
       of Nitrous Oxide (N20) 	    2-14

2-10   Ice Core Measurements of Historical Nitrous Oxide
       (N20) Concentrations 	    2-15

2-11   Measured Increases in Tropospheric Concentrations
       of Carbon Dioxide (C02) 	    2-16

2-12   Ice Core Measurements of Historical Carbon Dioxide
       (C02) Concentrations 	    2-17

2-13   Measured Increases in Tropospheric Concentrations
       of Methane (CH4) 	    2-19

2-14   Ice Core Measurements of Historical Methane (CH4)
       Concentrations 	    2-20
                         * * *  DRAFT FINAL  * * *

-------
                                                                    Page

2-15a  CFC-12:  Constant Emissions 	    2-22

2-15b  CFC-12:  Atmospheric Concentrations 	    2-22

2-16a  CFC-12:  Emissions 	    2-23

2-16b  CFC-12:  Atmospheric Concentrations 	    2-23

2-17   CFC-12:  Atmospheric Concentrations from Different
       Emission Trajectories 	    2-24

3-1    Selected Properties of CFCS 	    3-7

3-2    CFC Characteristics and Substitutes 	    3-9

3-3    Companies Reporting Data to CMA 	    3-10

3-4    Production of CFC-11 and CFC-12 Reported to CMA 	    3-12

3-5    Historical Production of CFC-11 and CFC-12 	    3-14

3-6    CFC-11 and CFC-12 Used in Aerosol and Nonaerosol
       Applications in the EEC . . .•	    3-15

3-7    Comparison of Estimated CFC-11 Use: 1985 	    3-17

3-8    Comparison of Estimated CFC-12 Use: 1985 	    3-18

3-9    Estimates of Production and Emissions of CFC-11
       and CFC-12 	    3-20

3-10   Published Estimates of U.S.S.R. Production of CFC-11
       and CFC-12 	    3-21

3-11   Historical Production of CFC-11 and CFC-12 in the U.S	    3-23

3-12   EEC Production and Sales Data 	    3-24

3-13   The Bottom Up Approach 	    3-27

3-14   Nonaerosol Application of CFC-11 and CFC-12 (OECD)	    3-28

3-15   Range of Population and GNP Per Capita Projections 	    3-31

3-16   Summary of Demand Projection Estimates 	    3-33

3-17   Summary of Demand Projection Estimates
       (Average annual rate of growth in percent) 	    3-34
                                DRAFT FINAL  *

-------
                                                                    Page

3-18   Long Term Projections CFC-11 and CFC-12
       World Production (2000-2050) 	    3-36

3-19   Nonaerosol Production Per Capita of CFC-11 and CFC-12
       Has Been Correlated with Gross Domestic Production (GDP)
       Per Capita in Developed Countries (1962 to 1980) 	    3-39

3-20   Global Population and GNP Scenarios Used in
       Gibbs'  Analysis 	    3-41

3-21   Range of Future Global Demand for CFC-11 and CFC-12 	    3-45

3-22   Consensus Projections: World CFC-11 and CFC-12
       (1985-2000) 	    3-47

3-23   Growth Rates for Global Aerosol and Nonaerosol
       Applications of CFC-11 and CFC-12 	    3-49

3-24   Scenarios of Global Production and Emissions
       CFC-11 and CFC-12 	    3-50

3-25   Current and Projected Future CFC-11 and CFC-12 Use per
       Capita and GNP per Capita 	    3-52

 3-26  Scenarios of Global Production and Emissions
       CFC-22 and CFC-113 	    3-53

3-27   Scenarios of Global Production and Emissions of
       Carbon Tetrachloride and Methyl Chloroform 	    3-56

3-28   Scenarios of Global Production and Emissions of
       Halon-1301 and Halon-1211 	    3-60

4-1    Effects of Changes in Composition of Atmosphere 	    4-5

4-2    Historical Carbon Dioxide Emissions from Fossil Fuels
       and Cement 	    4-9

4-3    A Schematic of the Carbon Cycle 	    4-10

4-4    Projected Carbon Dioxide Emissions and Doubling
       Time of Concentrations 	    4-11

4-5    Two Ways That CH4 Concentrations Could Have Changed 	    4-14

4-6    Estimated CH4 Emission Sources (1012 grams per year) 	    4-15

4-7    Possible Changes in CH4 Sources and in
       Emission Factors 	    4-17
                         * * *  DRAFT FINAL  » * *

-------
                                                                    Page

4-8    Current Sources and Sinks of Carbon Monoxide
       (1984 concentration of CO:   30-200 ppb) 	    4-19

4-9    Scenarios of Carbon Monoxide (CO) Emissions from
       Combustion 	    4-20

4-10   Scenarios of Methane (CH4)  Concentrations 	    4-22

4-11   Preliminary Scenario of Future Growth in N20
       Emissions by Source 	    4-25

4-12   Projected Nitrous Oxide (N20) Concentrations 	    4-26

4-13   Summary of Standard Scenarios Proposed for Assessment 	    4-28

5-1    Temperature Profile and Ozone Distribution in the
       Atmosphere 	    5-8

5-2    Steady-State Scenarios Used in International Assessment ...    5-12

5-3    Change in Total Ozone from Representative One-Dimensional
       Models for Steady-State Scenarios Containing Clx
       Perturbations 	    5-13

5-4    Change in Total Ozone at 40 Kilometers for Steady-State
       Scenarios Containing Clx Perturbations 	    5-14

5-5    Change in Total Ozone for Steady-State Scenarios 	    5-15

5-6    Effect of Stratospheric Nitrogen (NOy) on Chlorine-Induced
       Ozone Depletion 	    5-17

5-7    Effect of Doubled C02 Concentrations on Ozone
       Temperature 	    5-19

5-8    Calculated Changes in Ozone Versus Altitude 	    5-21

5-9    Two-Dimensional Model Scenarios Used in International
       Assessment 	    5-23

5-10   2-Dimensional Model Results: Globally and Seasonally
       Averaged Ozone Depletion 	    5-25

5-11   Ozone Depletion by Latitude, Altitude, and Season for Clx
       Increase of 6.8 ppbv (MPIC 2-D Model)  	    5-26

5-12   Ozone Depletion by Latitude, Altitude, and Month for Clx
       Increase of 6.8 ppbv (AER 2-D Model) 	    5-27
                           * »  DRAFT FINAL  * * *

-------
5-13   Ozone Depletion by Latitude, Altitude, and Month for Clx
       Increase of 14.2 ppbv (AER 2-D Model)	    5-28

5-14   Ozone Depletion of Latitude, Altitude, and Month for Clx
       Increase of 6.8 ppbv (GS 2-D Model) 	    5-29

5-15   Ozone Depletion by Latitude and Season for Clx Increase
       of -6.0 ppbv (IS 2-D Model) (AER 2-D Model) 	    5-30

5-16   Change in Ozone by Altitude for CFC-11 and CFC-12
       Emissions at 1980 Levels (LLNL 1-D Model) 	    5-31

5-17   Change in Ozone by Altitude for Clx Increase of 6.7 ppbv
       (LLNL 1-D Model) .	    5-32

5-18   Change in Ozone by Altitude for Clx Increase of 13.7 ppbv
       (LLNL 1-D Model) 	    5-33

5-19   Change in Ozone by Latitude and Season for Clx
       Perturbations (MPIC 2-D Model) 	    5-34

5-20   Change in Ozone by Latitude and Season for Clx
       Perturbations (AER 2-D Model) 	    5-35

5-21   Latitudinal Dependence of AER and MPIC 2-D Models 	  5-37

5-22   Change in Ozone by Latitude, Altitude, and Month for
       Coupled Perturbations (GS 2-D Model) 	    5-38

5-23   Changes in Ozone by Latitude, Altitude,  and Season for
       Coupled Perturbations (MPIC 2-D Model) 	    5-39

5-24   Changes in Ozone by Latitude and Altitude in Winter for
       Coupled Perturbations (MPIC 2-D Model) 	    5-40

5-25   Models With Reported Time Dependent Runs 	   5-42

5-26   LLNL 1-D Model Versus Parameterization Fit 	    5-43

5-27   Time Dependent Change in Ozone for Low CFC Growth and
       Coupled Perturbations 	    5-44

5-28   Time Dependent Change in Ozone by Altitude for Low CFC
       Growth and Coupled Perturbations (LLNL 1-D Model) 	    5-45

5-29   Trace Gas Assumptions for Results in Exhibit 5-30
       (Brasseur and DeRudder 1-D Model, 1986)   	    5-46
                                DRAFT FINAL

-------
                                                                    Page

5-30   Time Dependent Change in Ozone for CFC Growth and Coupled
       Perturbations (Brasseur and DeRudder 1-D Model) 	    5-47

5-31   Time Dependent Change in Ozone for Constant CFC Emissions
       and Growth in Other Trace Gases (Brasseur and DeRudder
       1-D Model) 	    5-48

5-32   Sensitivity of 1-D Models to Representation of Radiative
       Processes (Brasseur and DeRudder 1-D Model) 	    5-49

5-33   Model Comparison: Time Dependent Change in Ozone for CFC
       Growth and Coupled Perturbations 	    5-50

5-34   Trace Gas Assumptions for Results in Exhibit 5-35
       (AER 1-D Model,  1986) 	    5-52

5-35   Time Dependent Change in Ozone for Various Scenarios of
       Coupled Perturbations (AER 1-D Model) 	    5-53

5-36   Trace Gas Scenarios Tested in LLNL 1-D Model 	    5-54

5-37a  Time Dependent,  Globally Averaged Change in Ozone for
       Coupled Perturbations (LLNL 1-D Model) "Reference Case" ...    5-56

5-37b  Time Dependent,  Globally Averaged Change in Ozone for
       Coupled Perturbations (LLNL 1-D Model) 	    5-57

5-38   Effect of Potential Greenhouse Gas Controls on Ozone
       Depletion (Results from 1-D Parameterization)  	    5-58

5-39   Comparison of the Calculated NOy Profile at the
       Equator in the 2-D Models of Stordahl and Isaksen,
       and Ko 	    5-60

5-40   Calculated Ozone Depletion for 1970 to 1980 Versus
       Umkehr Measurements 	    5-61

5-41a  Time Dependent Globally and Seasonally-Averaged Changes
       in Ozone for Coupled Perturbations (IS 2-D Model) 	    5-63

5-41b  Time Dependent Globally and Seasonally-Averaged Changes
       in Ozone for Coupled Perturbations (IS 2-D Model) 	    5-64

5-42a  Time Dependent Seasonally-Averaged Change in Ozone for
       1980 CFC Emissions and Coupled Perturbations (IS 2-D
       Model) 	    5-65
                         * * *  DRAFT FINAL  * *

-------
                                                                    Page

5-42b  Time Dependent Seasonally-Averaged Change in Ozone for
       1.2% Growth in CFC Emissions and Coupled Perturbations
       (IS 2-D Model) 	    5-66

5-43   Time Dependent Seasonally-Averaged Change in Ozone for
       3% Growth in CFC Emissions and Coupled Perturbations
       (IS 2-D Model) 	    5-67

5-44   Time Dependent Seasonally-Averaged Change in Ozone for
       3.8% Growth in CFC Emissions and Coupled Perturbations
       (IS 2-D Model) 	    5-68

5-45   Temperature Feedback Experiment: Time Dependent,  Globally
       and Seasonally Averaged Change in Ozone for 3% Growth in
       CFC Emissions and Coupled Perturbations (IS 2-D Model) ....    5-70

5-46   Model Comparison for Coupled Perturbations Scenario 	    5-71

5-47   Calcualted Ozone -- Column Change to Steady-State for Two
       Standard Assumed Perturbations 	    5-73

5-48   Latitudinal Gradients in Odd Nitrogen: Models vs
       Measurements 	    5-75

5-49   Hydroxyl Radical (OH) Measurements 	    5-77

5-50   H02:  Models versus Measurements 	    5-78

5-51   Variability of Observed CIO Concentrations 	    5-81

5-52   CIO Vertical Profiles: Models Versus Measurements 	    5-83

5-53   HC1:  Models Versus Measurements 	    5-84

5-54   Logical Flow Diagram for Monte Carlo Calculations 	    5-87

5-55   Histogram of Measurements for a Rate Constant 	    5-89

5-56   Recommended Rate Constants and Uncertainties Used in
       Monte Carlo Analyses 	    5-90

5-57   Monte Carlo Results: Change in Ozone Versus Fluorocarbon
       Flux 	    5-91

5-58   Monte Carlo Results: Change in Ozone Versus Fluorocarbon
       Flux 	    5-92

5-59   Monte Carlo Results: Ozone Depletion for Coupled
       Pertubations 	    5-94
                         * * *  DRAFT FINAL  * * *

-------
                                                                    Page




5-60   A Monte Carlo Distribution of Column Ozone Changes
5-61
5-62
5-63
5-64
5-65
5-66
5-67
5-68
5-69
5-70
5-71
5-72
5-73
5-74
5-75
5-76
6-1
6-2
6-3
Monte Carlo Results: Changes in Ozone by Altitude 	
Monte Carlo Results: Changes in Ozone by Column and
Altitude, Unscreen Data 	
Monte Carlo Analysis With the LLNL 1-D Model 	
Monte Carlo Results : Changes in Ozone by Altitude 	
Ozone Trend Estimates by Latitude 	
Changes in Ozone from 1970 to 1980: Umkehr Measurements
and Model Calculations 	
SBUV Zonal Trends Estimates Versus "Umkehr Station
Blocks" 	
Ozone Trend Estimates and 95% Confidence Intervals 	
Ozone Balloonsonde Stations 	
Correction Factors for Balloonsonde Measurements 	
Ozone Trend Emissions (% per year) As Determined from
Balloon Ozonesondes Versus Those Determined from Dodson
Measurements (Tiao, et al. , personal communication) 	
Monthly Means of Total Ozone at Hal ley Bay 	
Nimbus 7 Antarctic Ozone Measurements: 12-Day Sequence ...
Nimbus 7 Antarctic Ozone Measurements: Mean Total
Ozone in October 	
Global (60°N-60°S) Monthly Ozone Determined from
NOAA TOVS System 	
Preliminary Ozone Trend Data (Health versus 2-D Model
Results) (Isaksen) 	
Stratospheric Perturbants and Their Effects 	
Absorption Characteristics of Trace Gases 	
Radiative Forcing for a Uniform Increase in Trace Gases . . .
5-96
5-97
5-98
5-99
5-103
5-105
5-107
5-109
5-111
5-112
5-113
5-116
5-117
5-119
5-121
5-122
6-8
6-9
6-10
                           * *  DRAFT FINAL  * *

-------
6-4    Effects of Vertical Ozone Distribution on
       Surface Temperature 	    6-12

6-5    Water Vapor, Altitude, and Radiative Forcing 	    6-13

6-6    Temperature Sensitivity to Climatic Feedback Mechanisms ...    6-15

6-7    Empirical Estimates of Climate Sensitivity are Sensitive
       to Estimates of Historical Temperature Increases and
       Trace Gas Concentrations 	    6-17

6-8    Relationship of Radiative Forcing, Ocean Heat Uptake,
       and Realized and Unrealized Warming 	    6-20

6-9    Transient Estimates of Global Warming 	    6-21

6-10   Expected Temperature Increases 	    6-22

6-11   Results of Transient Analysis Using a General
       Circulation Model 	    6-23

6-12   Regions of U.S. :   Change in Runoff 	    6-25

7-1    Organization of the Adult Skin 	    7-5

7-2    Ultraviolet Absorption Spectra of Major Epidermal
       Chromophores 	     7-8

7-3    Skin Types and Skin Tanning Responses 	    7-9

7-4    Ultraviolet Action Spectra for DNA Dimer Induction 	    7-13

7-5    Comparison of Age Adjusted Incidence Rates Per 100,000
       for Squamous Cell Carcinoma (SCC) and Basal Cell Carcinoma
       (BCC) Among White Males and Females in the United States ..    7-20

7-6    Percentage of Tumors by Anatomic Site for Nonmelanoma
       Skin Cancer Among White Males and Females in the United
       States (1977-1978 NCI Survey Data) 	    7-22

7-7    Distribution by Sex and Anatomic Site of Nonmelanoma
       Skin Tumors: Canton of Vaud, Switzerland (1974-1978) 	    7-23

7-8    Annual Age Adjusted Incidence Rates for Basal and Squamous
       Cell Carcinoma (1977-1978 NCI Survey Data) and Melanoma
       (1973-1976 SEER Data) Among White Males 	    7-26
                         * * *  DRAFT FINAL  *

-------
7-9    Annual Age Adjusted Incidence Rates for Basal and
       Squamous Cell Carcinomas (1977-1978 NCI Survey Data)  and
       Melanoma (1973-1976 SEER Data) Among White Females 	    7-27

7-10   Estimates of  Relative Risks of Basal and Squamous Cell
       Carcinomas for 32 Combinations of Risk Factors 	    7-29

 A-l   Correlation of Alternative Measurements of UVB Radiation
       for Ten Locations in the United States 	    7-39

 A-2   Population Weights for Ten Locations in the United
       States 	    7-40

 A-3   Estimated Dose-Response Coefficeints for Basal and Squamous
       Cell Skin Cancers (UVB Dose Skin Cancer Incidence) 	     7-41

 A-4   Estimated Percentage Changes in UVB Radiation in San
       Francisco for a Two and Ten Percent Depletion in Ozone ....    7-43

 A-5   Percentage Change in Incidence of Basal and Squamous  Cell
       Skin Cancers for a Two Percent Depletion in Ozone for
       San Francisco 	    7-44

 A-6   Percentage Change in Incidence of Basal and Squamous  Cell
       Skin Cancers for a Ten Percent Depletion in Ozone for
       San Francisco 	    7-45

8-1    Variation in UV Radiation by Latitude as Percent of
       Levels at the Equator on March 21 at Noon 	    8-9

8-2    UV Radiation by Month in Washington, D.C	    8-10

8-3    Relative Change in UV Flux by Hour in Washington, D.C.,
       on June 21 	    8-11

8-4    Average DNA Action Spectrum 	    8-12

8-5    Location of Melanocyte in the Epidermis 	    8-14

8-6    Comparative Transmittance of UV Radiation 	    8-15

8-7    Increases in Incidence and Mortality Rates from
       Malignant Melanoma in Different Countries 	    8-17

8-8    Anatomic Site Distribution of Cutaneous Malignant
       Melanoma 	    8-19
                             *  DRAFT FINAL  * * *

-------
                                                                     Page

 8-9    Anatomic Site Distribution of Cutaneous  Malignant
        Melanoma by Gender 	    8-20

 8-10   Malignant Melanoma Risk Factors by Measures  of Skin
        Pigmentation Within the Caucasian Population 	    8-27

 8-11   Summary Statistics for Regressions of SIC in Cancer
        Incidence and Mortality on Latitude 	    8-31

 8-12   Estimated Relative Increases in Melanoma Skin Cancer
        Incidence and Mortality Associated with  Changes in
        Erythema Dose 	    8-33

 8-13   Summary of Fears,  Scotto,  and Schneiderman (1977)
        Regression Analyses of Melanoma Incidence Dose-Response ...    8-34

 8-14   Biological Amplification Factors for Skin Melanoma by
        Sex and Anatomical Site Groups, Adjusting for Age and
        Selected Constitutional and Exposure Variables 	    8-36

 8-15   Biological Amplification Factors for Melanoma Incidence
        by Sex and Anatomical Site Groups, Adjusting for Age
        and Combination of Selected Constitutional and
        Exposure Variables 	    8-37

 8-16   Estimated Percentage Change in Melanoma  Mortality for
        Different Percentage Changes in UVB 	    8-39

10-1    Cataract Prevalence by UV Zone 	   10-7

10-2    Comparison of Cataract Prevalence for Aborigines and
        Non-Aborigines'	   10-7

10-3    Composite Transmittance Curves for the Rabbit 	   10-9

10-4    Calculated Total Transmittance of the Human Eye 	   10-10

10-5    Percent Transmissivity Through the Entire Rhesus Eye 	   10-11

10-6    Transmittance of the Total Rabbit Cornea, the Total
        Human Cornea, and the Rabbit Corneal Epithalium 	   10-12

10-7    Transmittance of the Anterior Ocular Structures of
        the Human and Rabbit Eyes 	   10-13
10-8    UV Radiant Exposure Threshold Data for the Cornea
        H.,, Lens HL, Cataracts, and Retina H
         0                                  K
        for the Rabbit and Primate 	   10-14
                          * * *  DRAFT FINAL  * * *

-------
10-9    The Action Spectra for Photokeratitis and Cataracts
        for the Primate and Rabbit 	   10-15

10-10   Free Radicals and Oxidation:  Reduction Systems 	    10-17

10-11   Enzyme Systems Involved in Oxidation: Reduction 	    10-19

11-1    A Summary of Studies Examining Cultivars Differences
        in UV-B Radiation Sensitivity 	   11-7
11-2    Survey of UV Studies by Major Terrestrial Plant
        Ecosystems (after Whittaker 1975) 	   11-12

11-3    Summary of UV-B and C02 Effects on Plants 	   11-13

11-4    Summary of Field Studies Examining the Effects of
        UV-B Radiation on Crop Yields 	   11-15

11-5    Details of Field Study by Teramura (1981-1985) 	   11-16

11-6    Summary of Changes in Yield Quality in Soybean
        Between the 1982 and 1985 Growing Seasons
        (Teramura 1982-1985) 	   11-20
12-1    Solar Irradiance Outside the Earth's Atmosphere and at
        the Surface of the Earth for a Solar Zenith Angle of 60°...   12-5

12-2    Relationship Between Ozone Depletion and Biological
        Effectiveness of Increased UV-B Radiation 	   12-6

12-3    Solar Spectral Irradiance at the Surface of the Ocean
        and at Four Depths 	   12-7

12-4    Lethal Effects on Shrimp Larvae for Various Combinations
        of UV-R, Dose-Rate, and Total Dose 	    12-13

12-5    Estimated Effective UV-B Solar Daily Dose at Various
        Atmospheric Ozone Concentrations Based on a 4-Year
        Mean of Medians, Manchester, Washington, 1977-1980 	   12-14

12-6    Estimated Biologically Effective UV-B Doses Leading to
        Significant Effects in Major Marine Zooplankton Groups ....   12-17

12-7    Annual Fish Production in Coastal Waters Baseline
        Data for Coastal Waters 	   12-18

12-8    Percentage of Total Dose Limit to be Reached on Any
        Particular Day:  Lethal Doses Accumulated Only After
        Dose-Rate Threshold is Exceeded 	   12-21
                          * * *  DRAFT FINAL  * » *

-------
                                                                     Page

12-9    Effect of Increased Levels of Solar UV-B Radiation on
        the Predicted Loss of Larval Northern Anchovy from Annual
        Populations,  Considering the Dose/Dose-Rate Threshold
        and Three Vertical Mixing Models 	   12-30

13-1    Wavelengths of UV Radiation and Polymers with Maximum
        Sensitivity and Corresponding Photon Energies 	   13-4

13-2    Plastics Used in Applications Where Exposure of the
        Material to Sunlight Might Be Expected 	   13-8

13-3    Modes of Damage Experienced by Polymers Used in
        Outdoor Application	  13-11

13-4    PVC Siding Compound Composition 	   13-12

13-5    UV Screening Effectiveness of Selected Pigments 	   13-14

13-6    Domestic Consumption of Light Stabilizers,  1984-85 	   13-15

13-7    Increased Stabilization Market (1970-2020)  	   13-17

13-8    Ozone Depletion Estimates	   13-18

13-9    Cumulative Added Cost 	   13-19

13-10   Diagrammatic Representation of the Effect of
        Pigment/Fillers as Light Shielders
        (Monodisperse Spherical Filler) 	   13-21

13-11   Relative Damage Indices for Yellowing of PVC Under Miami
        (March 22nd)  Conditions, at Different Extents of
        Ozone Layer Deterioration 	   13-25

13-12   Estimated Ranges of Factor Increase in Damage and the
        Factor Increase in Stabilizer Needed to Counter the
        Change of Yellowing of Rigid PVC Compositions 	   13-27

13-13   PVC Damage with Ozone Depletion 	   13-28

13-14   Projections of Future Demand for Selected Years
        (Thousands of Metric Tons) 	   13-29
14-1    Ozone Concentrations for Short-Term Exposure That
        Produce 5% to 20% Injury to Vegetation Growth Under
        Sensitive Conditions 	   14-6
                                 DRAFT FINAL  * *

-------
14-2    Ozone Concentrations at Which Significant Yield
        Losses Have Been Noted for a Variety of Plant
        Species Exposed Under Various Experimental Conditions ....    14-7

14-3    Ozone Concentrations (ppm) Predicted for Changes in
        Dobson Number and Temperature for Three Cities 	    14-10

14-4    Global Warming Would Exacerbate Effects of Depletion
        on Ground-Based Ozone in Nashville 	   14-12

15-1    Snow and Ice Components 	   15-6

15-2    Worldwide Sea Level in the Last Century	   15-9

15-3    No Title 	   15-11

15-4    Estimates of Future Sea Level Rise 	   15-15

15-5    Local Sea Level Rise 	   15-16

15-6a   Evolution of Marsh as Sea Level Rises 	    15-18

15-6b   Composite Transect -- Charleston, S.C	    15-19

15-7    Louisiana Shoreline in the Year 2030 	    15-21

18-8    Distribution of Population in Bangladesh 	    15-23

15-9    The Bruun Rule 	    15-25

15-10   Percent of Tidal Cycles in Which Specified
        Concentration is Exceeded at Torresdale During a
        Recurrence of the 1960's Drought for Three Sea Level
        Scenarios 	   15-30

16-1    Summary of Findings from the WMO/UNEP/ICSU Conference
        on Global Climate Held in Villach, Austria, October 1985 ..   16-1

17-1    Modular Structure 	   17-3

17-2    Major Model Input Choices 	   17-6

17-3    Effects Not Quantified 	   17-11

  A-l   Flow of Analysis Program 	    A-4

  A-2   Files Required to Specify a Run 	    A-7
                            * *  DRAFT FINAL  * * *

-------
                                                                   Page

B-l   Future Global Production Scenarios:  Middle Scenario 	    B-3

B-2   Regional Use Shares -for CFC-11 	    B-4

B-3   U.S. End Use Shares for CFC-11 	    B-6

B-4   Middle U.S. Population Scenario 	    B-7

B-5   Middle U.S. GNP Scenario 	    B-7

B-6   User-Modified Scenario Specifying a Growth Rate of 2
      Percent Annually for CFC-11 in the U.S	    B-10

B-7   A User-Modified Scenario: Production as a Function of
      Population	    B-10

D-l   Release Tables for CFC-11 	    D-2

D-2   Emissions from a Hypothetical 100 Million Kilograms of
      Production in 1985 	    D-4

D-3   Emissions from Production Over a Series of Years
      (Millions of Kilograms)	     D-5

D-4   Sample Table of Exogenously Specified Emissions 	     D-7

E-l   Trace Gas Abundances Used to Develop the Ozone Depletion
      Relationship 	     E-2

E-2   Comparison of Total Column Ozone Depletion Results from
      the 1-D Model and the Parameterized Numerical Fit 	    E-3

E-3   Hypothetical Table of User-Specified Ozone Depletion 	    E-6

E-4   Example Ozone Depletion Scaling Factors 	    E-7

F-l   Coefficient Used to Estimate Ozone Depletion at Latitudes
      when Global Average Depletion Exceeds 3.0 Percent 	    F-4

F-2   Comparison of Estimate of Latitudinal Ozone Depletion
      when Global Ozone Depletion Exceeds 3.0 Percent: Isaksen
      Versus the Linear Relationship 	     F-5

F-3   Coefficients Used to Estimate Ozone Depletion of Latitudes
      when Global Depletion Exceeds 1.5 Percent and is Less than
      3. 0 Percent  	    F-7
                          * *  DRAFT FINAL  * *

-------
                                                                   Page

F-4   Comparison of Estimates of Latitudinal Ozone Depletion when
      Global Ozone Depletion Exceeds 1.5 Percent and is Less than
      3.0 Percent: Isaksen versus Linear Relationship 	   F-8

F-5   Coefficients Used to Estimate Ozone Depletion at Latitudes
      when Global Depletion is Less than 1.5 Percent 	   F-10

F-6   Comparison of Latitudinal Ozone Depletion when Global
      Ozone Depletion is Less than 1.5 Percent: Isaksen Versus
      the Linear Relationship 	   F-ll

F-7   Cities Used to Evaluate Changes in UV Flux for the Three
      Regions of the U.S	   F-13

F-8   States Included in the Three Regions of the U.S	   F-14

F-9   Percent Change in UV as a Function of Change in Ozone
      Abundance for Three U.S. Regions 	   F-15

F-10  Age Distributin of the U.S. Population Over Time in the
      North Region 	   F-17

F-ll  Baseline Incidence for Nonmelanoma Skin Cancers	 ..   F-19

F-12  Basel Incidence for Melanoma Skin Cancers 	   F-22

F-13  Mortality Rates for Melanoma Skin Cancers 	   F-23

F-14 .Baseline Prevalence of Senile Cataracts  	   F-24

F-15  Sample Table for Specifying Relative Weights for Exposure
      During a Person's Lifetime 	   F-26

F-16  Coefficients Relating Percent Change in UV to Percent
      Change in Incidence 	   F-29

F-17  Coefficients Relating Percent Change in UV to Percent
      Change in Melanoma Mortality 	   F-31

F-18  Coefficients Relating Percent Change in Senile Cataract
      Prevalence for a One Percent Change in UV 	   F-32

F-19  Damage Index and Increase  in Stabilizer  for Ranges of
      Ozone Depletion 	   F-36
                          * *  DRAFT FINAL  * * *

-------
                                                                     Page

18-1    Types of Human Health and Environmental Effects
        Estimated	 .   18-5

18-2    Global Average Ozone Depletion:   Central Case 	   18-13

18-3    Human Health Effects:  Central Case 	   18-14

18-4    Materials,  Climate and Other Effects:   Central Case 	   18-16

18-5    Global Average Ozone Depletion:   Emission Scenarios 	   18-19

18-6    Human Health Effects:  Emissions Scenarios Additional
        Cumulative Cases and Deaths Over Lifetimes of People
        Alive Today 	   18-20

18-7    Human Health Effects:  Emissions Scenarios Additional
        Cumulative Cases and Deaths Over Lifetimes of People
        Born 1985-2029 	   18-21

18-8    Human Health Effects:  Emissions Scenarios Additional
        Cumulative Cases and Deaths Over Lifetimes of People
        Born 2030-2074 	   18-22

18-9a   Materials,  Climate, and Other Effects:
        Emissions Scenarios  	   18-25

19-9b   Equilibrium Temperature Change:   Emissions Scenarios 	   18-26

18-10   Global Average Ozone Depletion:   Comparison to Results
        with a 2-D Dimensional Atmospheric Model 	   18-27

18-11   Materials,  Climate, and Other Effects:   Sensitivity to
        Relationship Between Climate Change and C02 Emissions  	   18-29

18-12   Summary of Effects of Greenhouse Gases  on Ozone  Depletion
        and Global Equilibrium Temperature 	   18-30

18-13   Global Average Ozone Depletion:   Scenario of Limits
        to Future Global Warming 	   18-31

18-14   Human Health Effects:  Scenarios of Limits to Future
        Global Warming Additional Cumulative Cases and Deaths
        Over Lifetimes of People Alive Today 	   18-32

18-15   Human Health Effects:  Scenarios of Limits to Future
        Global Warming Additional Cumulative Cases and Deaths
        Over Lifetimes of People Born 1985-2029 	   18-33
                          * * *  DRAFT FINAL  * * *

-------
                                                                     Page

18-16   Human Health Effects:   Scenarios of Limits  to Future
        Global Warming Additional Cumulative Cases  and Deaths
        Over Lifetimes of People Born 2030-2074 	   18-34

18-17   Materials,  Climate,  and Other Effects:   Scenarios
        of Limits to Future  Global Warming 	   18-35

18-18   GLobal Average Ozone Depletion:   Methane Emission  Cases  ...   18-37

18-19   Human Health Effects:   Methane Emissions Cases
        Additional  Cumulative Cases and Deaths  Over Lifetime
        of People Alive Today 	   18-38

18-20   Human Health Effects:   Methane Emissions Cases
        Additional  Cumulative Cases and Deaths  Over Lifetime
        of People Born 1985-2029 	   18-39

18-21   Human Health Effects:   Methane Emissions Cases
        Additional  Cumulative Cases and Deaths  Over Lifetime
        of People Born 2030-2074 	   18-40

18-22   Materials,  Climate and Other Effects:   Methane
        Emissions Cases 	 	   18-41

18-23   Human Health Effects:   Sensitivity to Dose-Response
        Relationship Additional Cumulative Cases and
        Deaths Over Lifetimes of People Alive Today 	   18-43

18-24   Human Health Effects:   Sensitivity to Dose-Response
        Relationship Additional Cumulative Cases and
        Deaths Over Lifetimes of People Born 1985-2029 	   18-44

18-25   Human Health Effects:   Sensitivity to Dose-Response
        Relationship Additional Cumulative Cases and
        Deaths Over Lifetimes of People Born 2030-2074 	   18-45

18-26   Global Average Ozone Depletion:   Sensitivity to
        Relationship Between Ozone Depletion and Emissions 	   18-47

18-27   Human Health Effects:   Sensitivity to Relationship
        Between Ozone Depletion and Emissions Additional
        Cumulative Cases and Deaths Over Lifetimes  of People
        Alive Today 	   18-48
                          * * -  DRAFT FINAL  * * *

-------
18-28   Human Health Effects:   Sensitivity to Relationship
        Between Ozone Depletion and Emissions Additional
        Cumulative Cases and Deaths Over Lifetimes of People
        Born 1985-2029 	   18-49

18-29   Human Health Effects:   Sensitivity to Relationship
        Between Ozone Depletion and Emissions Additional
        Cumulative Cases and Deaths Over Lifetimes of People
        Born 2030-2074 	   18-50
                            * *  DRAFT FINAL  * * *

-------
Introduction

-------
                            INTRODUCTION
    Ozone in the stratosphere helps protect humans,  biological organisms, and
useful materials by partially blocking ultraviolet  radiation  in wavelengths
from 295 nanometers to 320 nanometers  from reaching the  earth's surface.  The
vertical distribution of ozone in the  stratosphere  and the  relative dryness of
that atmospheric region also play a role in sustaining the  current radiative
balance of the planet; changes in vertical distribution  of  ozone  and
stratospheric water vapor can influence the average surface temperature of the
earth and its weather and climate.   As such,  the composition  and  structure of
the stratosphere are of human concern.

THE RISE OF CONCERN  ABOUT  STRATOSPHERIC CHANGE

    In the early 1970s, Johnston  and others expressed concern that the
exhausts of the then proposed fleets of supersonic  transports would deposit
nitrogen oxides and water vapor into the stratosphere where these substances
could decrease the abundance of ozone.  In 1974, Rowland and  Molina raised the
concern that chlorofluorocarbons  (CFCs),used as  aerosol  propellants,  in foam
blowing, as solvents, and as refrigerants, would pose a  similar threat.  Since
then, others have raised concerns about nitrous  oxides produced from
combustion,  soil processes, and fertilizers,  about  brominated compounds
produced as fire extinguishing agents, and about a  variety  of other
substances,  including carbon dioxide (C02) and methane  (CH4), which have the
potential to perturb stratospheric composition and  structure.

    An understanding of the role  of infrared gases  in maintaining earth's
climate started much earlier.  In 1861 Tyndall discovered that water  vapor
absorbs infrared radiation.  Not  much  attention  was paid to infrared  absorbing
gases because they were assumed more or less constant in the  atmosphere.  In
1957, accurate measurements of C02 started.  Since  then, concern  about the
rise in the atmospheric level of  greenhouse gases and their potential to warm
the earth and change its climate  has steadily increased. In  the  last ten
years, measurements of many radiatively important trace  species other than C02
have been made and have further heightened concern  about global warming.  In
addition, recent studies of stratospheric modification on climate have begun
to consider more than uniform alterations in the whole column of  ozone that
had been considered in early studies of stratospheric modification.   These
studies have analyzed how possible changes in vertical structure  of ozone and
in the addition of water vapor could add to or in some cases  subtract from a
greenhouse warming.  As a result  of these efforts,  it has become  clearer that
considerations of stratospheric modification must include the potential
contribution to the total greenhouse warming.

CONCERN ABOUT  PUBLIC HEALTH AND WELFARE  EFFECTS
OF GLOBAL ATMOSPHERIC CHANGE

    Concern about the effects of ozone depletion first focused on skin
cancers, crops, and aquatic organisms  during the Climate Impact Assessment
Project in the 1970s.  These concerns  still exist and much  has been learned
                              ***DRAFT FINAL'-**

-------
                                   -2-
about them since then.  In addition a number of new concerns have been added,
including the potential, effect of UV radiation on materials useful to man,  on
ecosystems, on air pollution problems like ground-based oxidants and acid rain
precursors, on the immune system of humans, and on cataracts and other eye
problems.  In fact, there has been a general recognition that ultraviolet
radiation can alter many different kinds of molecules,  changing structures and
even ultimately perturbing the dynamics of the ecosystems.   Despite advances,
however, our understanding of the possible effects of increased ultraviolet
radiation on the earth remains in its infancy.

    Concern about the impacts of climate change, including sea level rise,
alterations in hydrology and ecosystems, and a variety of other human
endeavors has grown stronger since Revelle's famous statement that mankind is
performing a great uncontrolled experiment on the earth.  Knowledge of
specific effects of climate change however, also remains rather limited.

NEED FOR ASSESSMENTS

    Protection of the stratosphere is an environmental issue that differs from
other concerns.  Most environmental problems are local or regional in scale.
Increases in ultraviolet radiation or changes in climate would produce effects
worldwide.  Most environmental problems can be eliminated as quickly as
emissions can be reduced.  For instance, the lifetime of particulates in the
atmosphere can be measured in days.  Once a problem is perceived and action
decided upon, concentration will drop as quickly as reductions in source
emissions can be made.  For stratospheric modification this is not true.   The
lifetimes of many possible stratospheric perturbants are very long.  If a
problem begins to be observed, it will not disappear immediately, even if
emissions are totally curtailed.  In fact, even with a total termination of
emissions of the long lived gases, decades would pass before concentrations
declined.  In the case of CFCs, if depletion starts to occur, actions that did
not drastically curtail or reduce CFC levels would, in fact, only tend to slow
the rate of depletion.

    These two features, the global impact of change, and the difficulty in
preventing, slowing, or reversing change if it occurs, create a unique problem
for scientists and decisionmakers attempting to assess risks of continued
unrestrained emission of potential ozone modifers.  Errors may take decades or
more to correct.  Complicating the assessment process is that fact unlike many
areas of risk assessment, experiments cannot be done to verify model
predictions.  There is no experimental earth to test theories and models --
the only experiment that can be done is the one now underway on the planet.

The Basis for  This Assessment

    Progress has been made, however, in the science of the stratosphere,  the
scientific disciplines related to effects, and in the process of doing risk
assessments.  Much has been learned since the the Climate Impact Assessment
Project  (CIAP) done in the early 1970's and later assessments.  Not only has
there been a steady improvement in atmospheric science, but atmospheric
science has reached out to other geophysical systems to include analysis of
                              — -DRAFT FINAL***

-------
                                   -3-
biogenic inputs into the atmosphere.  During the past two years, under the
leadership of Dr. Robert Watson and with the joint sponsorship of many
international bodies, including the World Meterological Organization (WHO),
the United Nations Environment Programme (UNEP), the National Aeronautics and
Space Administration (NASA), the Federal Aviation Administration (FAA),  the
National Oceanic and Atmospheric Administration (NOAA), the Commission of
European Communities, and the Bundesminsterium Fur Forchang and Technologic,
the most comprehensive assessment of atmospheric science ever conducted has
been completed.  This review considered a wider range of issues, a greater
number of scenarios of trace gases, and more uncertainties than in any
previous assessment.  The atmospheric chapters in this Risk Assessment draw
heavily from that document and consequent documents developed by NASA from
it.  In addition, work by Ivar Isaksen of Norway, using two dimensional time
dependent models, has helped to produce a much more powerful understanding of
the atmosphere.

    Advances have also been made in health and welfare effects.  In the field
of melanoma, major advances have been made by Holman, Armstrong, Scotto, and
Elwood.  In materials, Androtti has significantly pushed the frontier
forward.  In tropospheric air pollution, Whitten has opened new paths.  In
climate, Hansen, Manabe, Washington, Ramanathan and others have greatly
increased our understanding.  In crops and aquatics, Teramura, Caldwell,
Tevini, Hoder, Worrest and others have improved our understanding.   In
addition, the chapter on cataracts represents a new assessment extracted from
a group coordinated by Morris Waxier of the Food and Drug Administration to
understand that issue.

    This risk assessment attempts to build on the work of these and other
scientists, in order to understand the joint implications of stratospheric
modification through time.
                              ***DRAFT FINAL***

-------
Chapter 1

-------
                               CHAPTER 1

             GOALS AND APPROACH OF THIS  RISK ASSESSMENT
    Under Part B of the Clean Air  Act,  the Administration of EPA "shall
propose regulations for the control  of  any substance, practice, process or
activity (or any combination thereof) which  in his judgment may reasonably be
anticipated to affect the stratosphere,  especially ozone in the stratosphere,
if such effect in the stratosphere may  be reasonably anticipated to endanger
public health or welfare."

    This risk assessment seeks to  summarize  the state of scientific knowledge
so that the Administrator can make a considered judgment about the need for
additional controls.   Scientific evidence is therefore reviewed and evaluated
for two purposes:

        (1)  to assess the likelihood that different human
             activities could alter  the stratosphere in ways that
             altered ultraviolet radiation reaching earth's surface
             or that changed climate.

        (2)  to assess the likelihood that changes in ultraviolet
             radiation or climate  due to modifications in column
             ozone or stratospheric  water vapor would have
             detrimental effects on  human health or welfare.
ANALYTIC FRAMEWORK

    There are four key parts to this  risk  assessment:

        •   Part 1 (Chapters 2, 3,  4)  assesses  possible changes in
            the composition of the  atmosphere

        •   Part 2 (Chapters 5, 6)  assesses  possible  responses of
            the stratosphere to such  changes in atmospheric
            composition

        •   Part 3 (Chapters 7-16)  assesses  possible  public health
            and welfare to changes  in the  stratosphere

        •   Part 4 (Chapters 17, 18)  assesses the  joint
            implications of knowledge and  uncertainties about
            atmospheric composition,  stratospheric modification, and
            effects,  to produce a range of outcomes consistent with
            current knowledge.
                              ***DRAFT FINAL*—

-------
                                   1-2
    Appendix A presents a separate volume on assessing the risks  that  the
incidence and mortality of melanoma will increase if stratospheric ozone
levels are modified.   Appendix B presents a separate volume on the effects  of
climate change.  Every chapter tries to summarize the most likely science as
well as the implications of uncertainty.  The most likely case is called the
"central case."  High and low cases representing uncertainties are also, in
general, developed.   For example, in the chapter that reviews  studies  of the
growth of chlorofluorocarbons emissions, a central,  a high, and a low  scenario
are identified, encompassing the range of likely trajectories  of  emissions,
not just the single best case growth estimate.

    Each chapter begins with a summary description and a set of general
findings.  Following this the relevant science is critically reviewed.   The
second to last chapter integrates all the findings from the previous chapters
into a comprehensive modeling framework.  The last chapter presents the  joint
implications of the prior work in terms of risk.  Not only is  the most likely
or "central" case described, but the implications of high and  low assumptions
are examined.  In this way the risk assessment attempts to provide
decisionmakers with a range of estimates of overall  risk, based on various
combinations of the joint uncertainties, not just a  single estimate.

SUPPORTING DOCUMENTS AND ANALYSIS FOR THIS REVIEW

    As mentioned earlier, this assessment draws heavily for atmospheric
science from the WMO/UNEP assessment.1  Other documents that were used
include those produced by UNEP's Coordinating Committee on the Ozone Layer.
In addition analyses from the Fluorocarbon Program Panel of the Chemical
Manufacturers Association (CMA), which continues to  fund research primarily
related to atmospheric modelling were used.  The climate change conference
convened by WHO in Villach, Austria in October 1985, to examine potential
changes in climate also provided valuable information.  Finally a risk
assessment supervised by Dr. Donald G. Pitts and Morris Waxier of FDA  are used
for cataracts. All four documents should be consulted by those interested  in
more detailed treatments of these subjects.

CHAPTER OUTLINES

    Chapter 2 focuses on past changes in stratospheric perturbants (trace
gases that can change the structure and composition of the stratosphere).
Chapter 3 reviews a variety of studies completed in the last two years on the
future emissions of trace gases that are solely produced by man,  and
constitute potential ozone depleters.  Chapter 4 centers on studies done to
understand future concentrations of trace gases at least partly on biogenic
sources -- carbon dioxide, methane, and nitrous oxide.
    *To the maximum degree possible this report directly used the results of
other assessments.  In many places large sections of text are taken from those
reports.  Such text has been put into BOLDFACE,  to  facilitate  recognition
of their quotation.
                              — -DRAFT FINAL*-*

-------
                                    1-3
    Following chapters on how atmospheric composition might change,  Part II
consists of Chapter 5 on stratospheric modification and Chapter 6 on climate
change.

    The next part of the analysis examines effects of stratospheric
modification on public health and welfare.  In each chapter, the likelihood of
effects are evaluated and attempts made to identify the quantitative
relationships that relate atmospheric change to impact.  Chapter 7 reviews
basal and squamous skin cancers, and Chapter 8 melanoma.  Chapter 9 addresses
immune suppression and the diseases it might produce, Chapter 10 cataracts and
other eye impairments.  Chapter 11 analyzes possible crop and terrestial
ecosystem disruption, while Chapter 12 reviews potential damage to aquatic
life.  Chapter 13 covers materials damage while Chapter 14 completes the
review of possible UV-B induced damage by examining possible air pollution
effects of ozone modification.  Chapter 15 then reviews possible effects of
global warming on sea level rise and Chapter 16 summarizes Appendix B's review
of the effects of global warming and associated climate change.

    Part 4 consists of two chapters.  Chapter 17 explains the models used to
integrate the analyses, so that a comprehensive time dependent assessment of
risks of stratospheric modification for conditions of uncertainty can be
developed.  Chapter 18 summarizes the risk assessment by presenting estimates
of possible damage in the absence of affects to limit the emission of
greenhouse gases.  It constitutes the integration of all that has proceeded.

    Exhibit 1-1 summarizes the relationships among the various chapters.

ERRORS AND CORRECTIONS

    Inevitably errors will have crept into this draft risk assessment.
Reviewers should send comments and corrections to:

                    John S. Hoffman
                    EPA, PM 221
                    401 M Street, S.W.
                    Washington, D.C.  20460
                              ***DRAFT FINAL***

-------
                                   1-4
                               EXHIBIT 1-1

                     Relationships Among  the  Chapters*
Emissions

• Chapter 1
• Chapter 2
• Chapter 3
• Chapter 4
Atmospheric Response

• Chapter 5
• Chapter 6
-•* Effects
     Chapter  7
     Chapter  8
     Chapter  9
     Chapter  10
     Chapter  11
     Chapter  12
     Chapter  13
     Chapter  14
     Chapter  15
     Chapter  16
Year-by-Year
  Impacts
 Cumulative
  Impacts

• Chapter 18
    * Chapter 17 develops model representations that integrate all components
of the analysis.

    Various emissions scenarios are used with different atmospheric models  to
    estimate changes in ozone and global temperatures.   These outputs  of the
    atmospheric models are used in effects models that  produce quantitative
    and qualitative estimates of impacts.
                              ***DRAFT FINAL***

-------
Chapter 2

-------
                               CHAPTER 2

         STRATOSPHERIC  PERTURBANTS: PAST CHANGES  IN CONCENTRATIONS

SUMMARY

    Recent measurements of  trace gas  abundance  demonstrate that the
concentrations of many trace gases  are  increasing throughout the  lower
atmosphere (the troposphere).   These  gases  can  directly  influence the upper
atmosphere (the stratosphere)  when  they are transported  there  and become
involved in chemical reactions.   Or they can  play an  indirect  role in
stratospheric issues by influencing the atmosphere's  radiative balances
(temperatures) or dynamics.

    Tropospheric concentrations  of  the  two  most important chlorofluorocarbons,
CFC-11 and CFC-12, are growing at approximately 5 percent per  year.  In the
last ten years, their concentrations  have more  than doubled.   Tropospheric
concentrations of other chlorine bearing perturbants  are also  growing.
Measured increases in concentrations  of chlorofluorocarbons  include CFC-22 at
11 percent per year and CFC-113  at  10 percent per year.  Measured increases in
chlorocarbons include carbon tetrachloride  (CC14) at  1 percent per year and
methyl chloroform (CH3CC13)  at 7 percent per  year.

    Tropospheric concentrations  of  Halon-1211,  a bromine bearing  compound, are
increasing at.23 percent per year.  Trends  in the concentrations  of
Halon-1301, another brominated compound,  have not been reported.

    Tropospheric concentrations  of  gases with some sources that are partially
biogenic have also been increasing:   nitrous  oxide (N20) at  0.2 percent per
year, carbon dioxide (C02)  at  0.4 percent per year, and  methane (CH4) at  1
percent per year.

    Trace gases vary in their  atmospheric residence times.   Lifetimes of
long-lived trace gases include 75 (+32/-17) years for CFC-11 and  111 years for
CFC-12;1 90 years for CFC-113; 50 years for CC14; 110 years  for Halon-1301
and 25 years for Halon-1211; and 150  years  for  N20.   Lifetimes of
shorter-lived trace gases include,  for  anthropogenic  gases,  20 years for
CFC-22 and 6.5 years for CH3CC13, and for the biogenic gas CH4, approximately
11 years.

    After a gas is emitted,  its  fate  depends  on deposition processes, chemical
interactions and transport  to  the stratosphere.  Gases with  long  lifetimes are
inert, being less influenced by chemical loss mechanisms and are  thus more
    1 The uncertainty associated with CFC-11  is  +32  and  -17 years;  for
CFC-12 it is +289 and -46 years.  Thus for 100 kilograms  of CFC-11  released in
1987, the best estimate is that 37 kilograms  will  still  be in  the atmosphere
in 2087, with 66% probability that that amount will  be reached between 2060
and 2119.
                            * *  DRAFT FINAL  *  *

-------
                                   2-2
likely to survive until they can affect the stratosphere.  Since the loss rate
to the stratosphere is relatively slow, once concentrations of inert gases
rise, their influence will be long lasting.  If a given level of ozone
depletion occurs, it will diminish slowly even if there are no new emissions.
In fact, unless new emissions are less than the quantity of molecules lost to
the stratosphere, concentrations and depletion would increase.  The
concentrations of gases that are chemically active are more dependent on a
continued renewal by emissions.  Thus their concentrations are subject to
greater short term perturbation.  The difference between long and short
lifetimes has implications for the kinds of risks associated with rising
concentrations of different trace species.  Long lived gases will result in a
situation which may only be reversible over a period of decades or more.
                            * *  DRAFT FINAL  * *

-------
                                   2-3
FINDINGS

    1.   MEASUREMENTS SHOW THAT THE TROPOSPHERIC CONCENTRATIONS OF MANY
        STRATOSPHERIC PERTURBANTS (TRACE GASES THAT ALTER THE STRATOSPHERE)
        ARE INCREASING.

        la.  Concentrations of chlorine bearing perturbants are growing as
             follows:  CFC-11 and CFC-12 at 5% per year, CFC-22 at 11%,
             CFC-113 at 10%, carbon tetrachloride (CC14) at 1%, methyl
             chloroform (CH3CC13) at 7%.

        Ib.  The global concentration of Halon 1211, a bromine bearing
             compound, is growing at 23%.

        Ic.  Changes in concentrations of Halon 1301, a potential depleter,
             have not been reported.  It may be possible to estimate past
             changes using available archived air, but this has not been done.

        Id.  Concentrations of other relevant gases are growing as follows:
             nitrous oxide (N20) at 0.2%, carbon dioxide (C02) at 0.4%, and
             methane (CH4) at 1%.

    2.   THE TROPOSPHERIC LIFETIME OF TRACE GASES VARY

        2a.  Gases which are chemically inert accumulate in the atmosphere.
             Most emissions reach the stratosphere.  Concentrations of these
             trace gases are difficult to reduce quickly.  The best estimate
             of the lifetime (the time when 39% of the compound still remains
             in the atmosphere) of CFC-11 is 75 years, of CFC-12 is 111 years,
             of CFC-113 is 90 years, of CC14 is 50 years, of Halon 1211 is 25
             years, of N20 is 150 years and of Halon 1301 is 110 years.
             Uncertainty exists about these lifetimes.

        2b.  Gases which are chemically active have shorter lifetimes.  Most
             emissions tend to be lost from the atmosphere before influencing
             the stratosphere.  Their future concentrations will depend on the
             emissions that occur in time periods more immediately prior to
             the time of concern.  Their concentrations can fall sharply if
             emissions change in the future.  Methane has an estimated
             lifetime of 11 years, methyl chloroform of 6.5 years and CFC-22
             of 22 years.
                          * * *  DRAFT FINAL  * * *

-------
                                   2-4
MEASURED  INCREASES  IN TROPOSPHERIC CONCENTRATIONS OF
POTENTIAL OZONE DEPLETERS

    The tropospheric concentrations  of many  potential ozone-depleting trace
gases have been measured as increasing throughout the world.  Due to the
strength of global circulation and the chemical  stability of many of these
gases, they are rapidly well mixed in the troposphere.  Gases with longer
lifetimes are more equally mixed between the Northern and Southern
Hemispheres.  The longest-lived trace gas, N20,  has the smallest
interhemispheric mixing ratio:   1.003;  and CH3CC13 a gas with a short
lifetime, has the greatest interhemispheric  ratio: 1.37 (Prinn et al. 1983).
After ozone depleters are transported to the stratosphere, they are
photodissociated.  The NOx, Clx, or  Brx released can then enter into catalytic
cycles that would reduce stratospheric ozone (see Chapter 5).

CFC-II

    The global average concentration of CFC-11  (CFC13) is approximately 200
parts per trillion volume (pptv) (Prinn et al.  1983).  Tropospheric
concentrations show little latitudinal variation, with a North/South
interhemispheric mixing ratio of 1.09 (Prinn et  al. 1983).  CFC-11 has been
measured at the five globally distributed sites  of the Atmospheric Lifetime
Experiment program.  Average tropospheric concentrations are increasing at
approximately 5 percent per year (WHO 1986).   (Exhibit 2-1)  In the last 10
years tropospheric concentrations have more  than doubled (Rasmussen and Khalil
1986).

CFC-12

    The global average concentration of CFC-12  (CF2CL2) is approximately 320
pptv and is increasing at approximately 5 percent per year (WHO 1986).  It has
been measured at the five Atmospheric Lifetime Experiment stations (Exhibit
2-2).  Concentrations are globally well mixed, with a North/South
interhemispheric mixing ratio of 1.08 (Prinn et  al. 1983).  In the last ten
years tropospheric concentrations have more  than doubled (Rasmussen and Khalil
1986).

CFC-22

    The global average concentration of CFC-22  (CHC1F2) is approximately 52
pptv (NAS 1984).  Concentrations have been measured in the Pacific Northwest
and at the South Pole.  Global increases in  tropospheric CFC-22 concentrations
have averaged approximately 11.7 percent per year  (Khalil and Rasmussen 1982)
(Exhibit 2-3).

CFC-113

    The global average concentration of CFC-113  (C2C13F3) is approximately 32
pptv (NAS 1984).  Recent increases in tropospheric CFC-113 concentrations,
measured at the South Pole and Pacific Northwest, have averaged approximately
ten percent per year (Rasmussen and  Khalil 1982) (Exhibit 2-4).
                          * * *  DRAFT FINAL  *  *

-------
                                   2-5
                               EXHIBIT 2-1


Measured  Increases in Tropospheric Concentrations of CFC-11  (CFCI3)


               CFCL, IP) ADRIGOLE. IRELAND
|210
£200
; 190
U 180
1 "°

|l'''ll
,,„.,„ 	 ' 	 '" 	 '
' ' ' ' '
i , H1"'"'"'1'"'

-
                 197B    1979      1980      1981
               CFCL, IP) CAPE MEARES. OREGON
                                                       1983    1984
           < '90 j
           EC   |
           0 180

           X 170
           i
                 1978    1979      1980     1981
               CFCL, |P) RAGGED POINT, BARBADOS
                                                       1983   1984
           ? 210 •
           a   i
           S 200;
           11 180 i


           il7o!
                 1978    1979      1980      1981      1982
               CFCL, |P) POINT MATATULA, AMERICAN SAMOA
                                                       1983    1984
                 1978    1979      1980      1981      1982      1983    1984
               CFCL, IP) CAPE GRIM. TASMANIA
"* 200
& 190
|180
I170
5 '60
S1SO
. i ' '
	



                1978   1979
                                               1982     1983    1984
Average concentrations  of CFC-11 are  increasing at  approximately
five  percent per  year.   Data are from the Atmospheric Lifetime
Experiment.


Source:  World Meteorological Organization,  1986; Figure 3-2.
                        * •'• *  DRAFT FINAL  * *

-------
                                    2-6
                                EXHIBIT 2-2

Measured  Increases in Tropospheric Concentrations  of CFC-12 (CF2CL2)
                 CF,CL, ADRIGOLE, IRELAND
                                                  '"
              < 320 ;
              cc   I
              O 300 •
                   J978    1979      1980      1981
                                                          1983    1984
                  CF,CL, CAPE MEARES. OREGON
              £ 380 .

              O 360 !

              5 340-

              § 320'

              | 300.
              s  (
                   1978    1979      1980      1981      1982      1983    1984
                  CF,CL, RAGGED POINT, BARBADOS
              a. 350-
              Q.
              5330.

              5 310 •
              ac
              O .290•

              X 270 ! i I 1 i ; ' ! ' ' '
                    1978    1979      1980      1981
                                                          1983    1984
                  CF.CL, POINT MATATULA, AMERICAN SAMOA
              | 330.

              O 310-

              < 290 •
              a   i
              a 270 •

              * 250 •  , I ;
              S   i'"
                    1978    1979      1980      1981
                                                          1983    1984
                  CF,CL, CAPE GRIM. TASMANIA
              | 330 '

              9 310i
              < 290 •
              cc
              O 270 •

              X 250 .
              E
                    1978    1979      1980      1981
                                                           1983    1984
Average concentrations of  CFC-12 are  increasing at approximately
five percent  per year.  Data are from the Atmospheric Lifetime
Experiment.

Source:   World Meteorological Organization,  1986;  Figure 3-3.
                                 DRAFT FINAL  * *  *

-------
                               2-7
                            EXHIBIT 2-3


 Measured  Increases  in Tropospheric  Concentrations of CFC-22 (CHCIF2)
                    1975     1977     1979     1981     1983

  Average concentrations  of  CFC-22 are increasing at approximately 12

  percent per year.



  Source:  Rasmussen  and  Khalil 1982.
                           EXHIBIT 2-4


Measured Increases  in Tropospheric  Concentrations of CFC-113 (C2CL3F3)
              0


            §§
            0 -H
§
B
8
                 60
                 40
                 20
                                   South
                                         pole
                    1975
                1977
1979
1981
                                                    1983
  Average concentrations  of CFC-113 are increasing at approximately

  10 percent per year.



  Source:  Rasmussen  and  Khalil 1982.
                       » * *  DRAFT FINAL  * * *

-------
                                   2-8
Carbon Tetrachloride

    The global average concentration of carbon tetrachloride (CC14) is
approximately 140 pptv and is increasing at approximately one percent per year
(WHO 1986).   It has been measured at the five Atmospheric Lifetime Experiment
stations.  The North/South interhemispheric mixing ratio is 1.05 (Prinn et al.
1983)  (Exhibit 2-5).  In the last ten years, tropospheric concentrations in
both hemispheres have increased by more than a factor of 1.2 (Rasmussen and
Khalil 1986) (Exhibit 2-5).

Methyl Chloroform

    The global average concentration of methyl chloroform (CH3CC13) is
approximately 120 pptv and is increasing at approximately seven percent per
year (WHO 1986).  It has been measured at the five Atmospheric Lifetime
Experiment stations  (Exhibit 2-6).  Because methyl chloroform has a relatively
shorter lifetime than the other halocarbons, concentrations show a greater
geographic variation.  The North/South interhemispheric ratio is 1.37 (Prinn
et al. 1983).  In the last ten years, tropospheric concentrations have more
than doubled (Rasmussen and Khalil 1986).

Halon-1301

    The current concentration of Halon-1301 (CF3Br), measured at the South
Pole,  is approximately one pptv (Khalil and Rasmussen 1985).  Because
researchers have only recently attempted to measure its concentration, no
trend  estimates are available.  Archived air exists, however, and it may be
possible to use it to establish trends for this species.

Halon-1211

    The current concentration of Halon-1211 (CF2ClBr), measured at the South
Pole,  is approximately one pptv.  Increases in the tropospheric concentrations
measured at the South Pole have averaged 23 percent (Khalil and Rasmussen
1985)  (Exhibit 2-7).

Nitrous Oxide

    The mean global concentration of nitrous oxide (N20) is approximately 300
parts per billion volume (ppbv).  Weiss (1981) collected separate measurements
between 1976 and 1980 at several monitoring stations and aboard oceanographic
vessels in the major world oceans.  His analysis demonstrated that the
tropospheric N20 concentration was increasing at approximately 0.2 percent per
year (Exhibit 2-8).

    Recent data for N20 from the five Atmospheric Lifetime Experiment stations
are shown below.  The results are consistent with Weiss (1981).  The measured
trends at the five network stations were:   0.77, 0.27, 0.24, 0.09, and 0.18
percent (WHO 1986).  While the rates of increase appear to be larger in the
northern hemisphere  (0.25-0.7 percent per year) than in the southern
hemisphere (0.1-0.2 percent per year) (WHO 1986), there is little variation in
actual concentrations.  The North/South interhemispheric mixing ratio is 1.003


                          * * *  DRAFT FINAL  * * *

-------
                                    2-9
                               EXHIBIT  2-5

                    Measured  Increases  in Tropospheric
             Concentrations of Carbon  Tetrachloride  (CCI4)
                CCL. ADRIGOLE, IRELAND
            |150 i

            2140i
            ; 130

            a 120
            5 no •
                  1978    1979
                                        1981      1982      1983    1984
                CCL. CAPE MEARES, OREGON
             O 140J

             < 130
             giio
                  1978    1979      1980      1981
                                                        1983    1984
               CCL. RAGGED POINT, BARBADOS
            | ,40

            O 130

            < 120
            c
            o no

            I100!
                 1978    1979      1980     1981      1982      1983    1984
               CCL. POINT MATATUtA, AMERICAN SAMOA
            O 100'

            I 90!
            s
                 1978    1979
                                       1981      1982
                                                        1983   1984
               CCL. CAPE GRIM. TASMANIA
             100 I
            O

            x  90
                 1978   1979     1980      1981
                                                        1983    1984
Average concentrations of  carbon  tetrachloride are increasing at
approximately one percent  per year.   Data  are from the Atmospheric
Lifetime Experiment.


Source:   World Meteorological Organization,  1986;  Figure  3-5.
                            *   DRAFT FINAL  *  * *

-------
                                 2-10
                             EXHIBIT 2-6

                  Measured Increases in Tropospheric
           Concentrations of Methyl  Chloroform  (CH3CCI3)
               CHiCCL, ADRIGOLE, IRELAND
           !'50i
           ; 130 i
           z   i
           g.20

           5 '10 • .
                 1976    1979      1980      1961      1982      1983   1984
               CH.CCL, CAPE MEARES. OREGON
            £ 150 •

            | ,40.;

            < 130:
            E   I
            a 120!

            S no j
            I   l
                 1978    1979      1980      1981      1982      1983    1984




               CH,CCL, RAGGED POINT. BARBADOS	

               i
            7 140(                                               , . •
                 1978    1979      1980      1981      1982      1983    1984
               CHiCCL, POINT MATATULA. AMERICAN SAMOA
            > 120
            a   I
            3 110 ;

            8100!
            <   i
            1 90 I
            z   I
                 1978    1979      1980      1981      1982      1983    1984
               CH,CCL, CAPE GRIM. TASMANIA
             tioo
            _
            o  90
            a  70

            I  6o
                 1978    1979      1980      1981      1982      1983    1984
Average concentrations of methyl  chloroform are increasing  at
approximately seven percent per year.   Data are from  the Atmospheric
Lifetime Experiment.


Source:  World Meteorological Organization, 1986; Figure 3-4.
                        *  * *  DRAFT FINAL  * *

-------
                                2-11
                            EXHIBIT 2-7

                  Measured  Increases in Tropospheric
                Concentrations of Haion-1211  (CF2CIBr)
   1,4
C
0
n  1.2H
c
t
r   ,8-
a
o
n
P
P
   .0-
    1979
1980
1981
1982
1983
1984
Measurements from South Pole.    Average concentrations of Halon-1211
are increasing at approximately 23 percent  per year.

Source:  Khalil and Rasmussen,  1985.
                         *  DRAFT FINAL  * * *

-------
                                  2-12
                              EXHIBIT 2-8

                     Measured Increases in Tropospheric
                   Concentrations of Nitrous Oxide (N2O)
300
304
302
300
208
206
     •  Q
                             °°o°«?°8 "I
                                     o
          1076
1077
1078
                                                      1070
                                            1080
       Measurements of tropospheric nitrous oxide concentrations.  Data
       were collected at several monitoring sites and aboard vessels in the
       major world oceans.   Concentrations are increasing at approximately
       0.2 percent per year.

       Source:  Weiss, 1981.
                          * * *  DRAFT FINAL  * *

-------
                                   2-13
(Prinn et al. 1983).  If statistically significant,  this discrepancy may be
due to a growing imbalance between Northern sources  and global sinks (WMO
1986) (Exhibit 2-9).

    Pearman et al.   (1986) measured N20 concentrations trapped in polar ice and
provided the first evidence of changing N20 concentrations over hundreds of
years.  They found that "the ice data for recent times agree with modern
measurements" and that concentrations have increased by 6 ppbv per 100 years
over the last three centuries.  They conclude that their recent trend estimate
of 0.5-1.8 ppbv per year is not characteristic of the 400-year ice record, but
results from an increase since the time of the Industrial Revolution (Exhibit
2-10).

MEASURED  INCREASES IN TROPOSPHERIC CONCENTRATIONS OF
POTENTIAL  OZONE INCREASERS

    Carbon dioxide  (C02) and methane (CH4) may, in some cases, contribute to
increasing ozone abundance in the stratosphere (C02  and CH4) or in the
troposphere  (CH4).   They may also contribute to global warming.  Atmospheric
measurements indicate that their concentrations of potential ozone increasers
are increasing.

Carbon Dioxide

    Continuous and accurate measurements of C02 concentrations began in 1958
at Mauna Loa Observatory in Hawaii (Keeling and GMCC/NOAA).   The data show an
increase of approximately ten percent since then (Exhibit 2-11).

    Other measurements of the rise of C02 have been  made and are consistent
with the Mauna Loa time series, each varying in seasonal amplitude depending
on the distance from annual biospheric sources (Keeling 1978; Keeling et al.
1976; Peterson et al. 1982; and Sundquist and Broecker 1985).

    Direct measurements of C02 concentrations were attempted as early as 1872
(Callendar 1958).  Because they are subject to substantial error, their
utility is limited.  Indirect methods, however, such as measuring the
concentrations of C02 trapped in polar ice (Oeschger et al.  1982; Neftel et
al. 1982; and Delmas et al. 1980), and analyzing the isotopic carbon ratios  in
tree rings (Freyer  1978; and Peng et al. 1983), show that C02 concentrations
before 1850  (pre-industrial times), remained approximately constant at 270 ppm
+10  --25 percent lower than today.  Recent ice core measurements (Pearman
et al. 1986) are consistent with the upper end of this range and indicate that
as early as the seventeenth and eighteenth centuries, the average C02
concentration was approximately 281 ppm + 7 ppm (Exhibit 2-12).

Methane

    Ambient atmospheric measurements in the last several decades show that the
concentration of methane is increasing (Rasmussen and Khalil 1984; Blake et.
al. 1982; Fraser, Khalil and Rasmussen, 1984; Fraser et. al. 1981; and Ehhalt,
Zander, and Lamontague 1983).  Estimates of the rate of increase vary.
                            * *  DRAFT FINAL  * * *

-------
                                   2-14
                              EXHIBIT 2-9

                  Measured  Increases  in Tropospheric
                Concentrations of Nitrous  Oxide (N20)
              N,0 ADRIGOLE, IRELAND
          a 300               •| i
          a 290 i
          5 280 !

          S                                    .».-'...  ,-,,..  '
                1978   1979     1980      1981      1982      1983    1984


             ' N,0 CAPE MEARES, OREGON          	          	
          7   I
          a 330 ;
          S   I
          O 320 ;
          < 310'
          OC   ;
          0 300;
          S 290.
          5

                1978   1979     1980      1981      1982      1983    1984


              N,0 RAGGED POINT. BARBADOS	
          _   .-|
          S 330
          a
          O 320


          g33oo:'::	

          I "°;	

                1978   1979     1980      1981      1982      1983    1984


              N,0 POINT MATATULA. AMERICAN SAMOA	^^_

          a 320 i

          g3'°i  .;!!i	;,,,,,,	,,.   ,,;.,	 'I,	
          < 300 • •i i : '
          g 290;
          2 280 •


                1978   1979     1980      1981      1982      1983    1984


              N,0 CAPE GRIM, TASMANIA	

          | 320;
          O 310 !         .	, • .
          530oiii!!1'	
          S290!

          i"0i___s_r

                1978   1979     1980      1981      1982      1983    1984
Rates  of increase in  tropospheric nitrous oxide at the  five
monitoring  stations above are:   0.77%,  0.27%.  0.24%, 0.09%, and
0.18%.   While  the rates of  increase  appear to  be larger in the
Northern hemisphere than in  the southern hemisphere, there is  little
variation in actual concentrations.  Data are  from the  Atmospheric
Lifetime Experiment.

Source:   World Meteorological  Organization, 1986;  Figure 3-7.
                        * *
                               DRAFT FINAL  * * *

-------
                                      2-15
                                   EXHIBIT 2-10

                         Ice Core Measurements  of Historical
                         Nitrous Oxide (N2O) Concentrations
   Q)
   O  350
   C
0.3
-r-j

•
rd -p-r
     300
g;S
U &
849
     250
                                  :  •
             1600
1700
                                             1800
1900
        Ice core data  show that nitrous oxide concentrations were relatively
        constant until the time of the Industrial  Revolution.

        Source:   Pearman et al., 1986.
                            * * *  DRAFT FINAL  * * *

-------
                                            2-16
                                       EXHIBIT 2-11

                             Measured Increases in Tropospheric
                           Concentrations of Carbon Dioxide (CO2)
    350
    345
1   310
o
o
3   335
l_>
SE
O
    330
    325
    320
    315
    310
       1955
1960
1965
1970
1975
                                                                        1980
            Monthly measurements from Mauna Loa, Hawaii.  Data collection began
            in 1958 with the International Geophysical Year.  Since that time,
            concentrations have increased 10 percent.

            Source:   Keeling,  C.D., and GMCC/NOAA, unpublished.
                                     *  DRAFT FINAL  * * *

-------
                                       2-17
                                   EXHIBIT 2-12

                         Ice Core Measurements of Historical
                        Carbon Dioxide  (CO2) Concentrations
   cy
     320
  t-H
   O
   c

OH 300
fO
     280
    260
                               •.  i
                                /

                                o
            1600
1700
1800
1900
        Ice core data show that  carbon dioxide concentrations were
        relatively constant until  the time of the Industrial Revolution.

        Source:   Pearman et al.,  1986.
                            * * *  DRAFT FINAL  * * *

-------
                                   2-18
Rasmussen and Khalil (1984) had computed an average annual increase of 1.3
percent, which at current concentrations is an increase of 0.02 ppm per year.
Their recent work (Khalil and Rasmussen 1986), however, indicates that this
trend estimate was influenced by interannual variability which may be linked
to the effects of the El Nino-Southern Oscillation.  Their corrected estimate
of the long-term rate of increase is about 0.016 ppm per year, or one percent
annually.  Measurements of Blake et. al. (1982) are consistent with a linear
rise of 0.018 ppm per year.  Recently, spectrographic analyses by Rinsland,
Levine and Miles (1985) indicate an annual growth rate since 1951 of 1.10
percent.  Rowland (in NASA 1986) shows an increase of one percent per year
since 1977 (Exhibit 2-13).

    The examination of methane trapped in ice cores.reveals that its
concentration was relatively constant until approximately 150-200 years ago.
Since that time it appears to have roughly doubled (Pearman et al. 1986;
Rasmussen and Khalil 1984) (Exhibit 2-14).

FACTORS THAT INFLUENCE TRACE GAS LIFETIMES

    Trace gas molecules, both natural and man-made, are released into the
atmosphere at the earth's surface.  Unless some rapid process exists for their
removal, these molecules are soon distributed throughout the troposphere.  The
time scale for local vertical mixing is a few weeks;  that for east-west mixing
about the globe, a few months; and that for exchange between northern and
southern hemispheres, a year or two (NAS 1976).

    Trace gases can be removed from the troposphere by a number of processes
that occur either at the earth's surface or in the atmosphere.  Physical
processes, such as absorption in the oceans, can remove trace gases from the
atmosphere.   This is an important process for C02.  Chemical processes, such
as reactions with the OH radical, can also remove trace gases from the
troposphere.  These processes are important for a wide array of gases,
particularly methane, CFC-22, and methyl chloroform.   Other trace gases,
including CFC-11, CFC-12, CFC-113, Halon-1211 and Halon-1301, are chemically
inert.  For these gases, transport to the stratosphere is the only significant
removal process (NAS 1983).

    Transport of trace gases within and between the troposphere and
stratosphere involves the vertical motion of large air masses.  It is thus
independent of the molecular weight or structure of the species involved (NAS
1976).  It is a slow process because air exchange is inhibited by the
temperature structure of the atmosphere.  The stratosphere sits as a permanent
air inversion over the troposphere with temperature rising with altitude
rather than falling.  The troposphere is heated from the surface and is
vertically unstable.  This inversion inhibits transport.

    The average concentration of a trace gas in the troposphere is determined
by the balance between the source strength of the trace gas, its rate of
chemical transformation in tropospheric destruction,  its deposition and the
net flux through the tropopause.
                          * * *  DRAFT FINAL  * * *

-------
                              2-19
                          EXHIBIT 2-13

               Measured Increases in  Trospospheric
                 Concentrations of Methane  (CH4)
      1.65
   o

   §
  •H
   I
   fl
   I
   g
1.60
       1.55
   8
       1.50
            1977
                        1979
                                     1981
                                                  1983
                                                               1985
Average concentrations of methane are increasing at approximately
one percent per year.

Source:  Rowland, in NASA, 1986; Figure 5.
                     * * *  DRAFT FINAL  * *

-------
                               2-20
                           EXHIBIT 2-14

                      Ice  Core Measurements  of
              Historical  Methane (CH4) Concentrations

0)
iH
o
5
^1,400
c c
0 5
•H -H
•P rH
ro H
£3
§ MVOO
B£
8 u]
8
^600
ll| 1
// .
!/
r "'
•«/
• • C*
r ^7
. • o
t* .-af
• * •!• • #
• • • . »» o ./»
• * o • $
• • a •* * ci^0
o °; ..'""
oo , o °a o 0° °.o9a|" °
o o 8 c_o— -S"
0 0 ..02 	 S- 	 " 	 rt 0
o e o - o °
           1600
                        1700
                                     1800
                                                  1900
Ice core data show that methane concentrations  were  relatively
constant until the time of the Industrial  Revolution.   Since  that
time, they have roughly doubled.

Source:   Pearman et al., 1986.
                     * -  *  DRAFT FINAL  * *

-------
                                   2-21
    The lifetime of a trace gas in the atmosphere can be defined as the
average time its molecules reside there.  If at any time, the release of a
particular trace gas is suddenly terminated, the residence time, or
atmospheric lifetime, is the time required for the concentration to drop to
1/e (37%) of its value at the time of termination (NAS 1976).

LONG-LIVED TRACE GASES

    Many stratospheric perturbants are relatively chemically inert in the
troposphere.  Despite an intense ten year effort to search for alternative
sinks for these long-lived gases, their only known significant sink is
transport to the stratosphere (NAS 1976; and NAS 1984).  Lifetimes for these
gases are:  75 (+32/-17) years for CFC-11 and  111 (+289/-46)  years for CFC-12
(WMO 1986); 90 years for CFC-113 (NAS 1984); 50 years for CC14 (WHO 1986); 110
years for Halon-1301 and 25 years for Halon-1211 (NAS 1984); and 150 years for
N20 (WMO 1986).  Because loss to the stratosphere is slow relative to their
current source strength, long lived gases would be expected to accumulate in
the troposphere and increase in concentration.

    Clearly the rise in concentrations of gases demonstrates this fact.
Current concentrations of these trace gases appear far from equilibrium, and
given their current emission levels, concentrations will continue to rise even
if emissions do not increase (that is, if emissions do not move further from
atmospheric equilibrum).  Thus, if emissions are held constant, tropospheric
concentrations will continue to increase— although ultimately the increase
will slow and concentrations will gradually reach equilibirum (steady state).
At steady state, the tropospheric concentration will exceed today's
concentration.

    Neglecting atmospheric feedbacks, a simplified model (See Appendix A, this
chapter) of source and loss terms implies that if current CFC-12 emissions are
held constant (Exhibit 2-15A), the concentration in the year 2100 would be 1.9
ppbv (Exhibit 2-15B), almost four times higher than the current concentration.
To halt a rise in concentrations would require a large shift in emission
rates.   A reduction of 85 percent in CFC-12 emissions would be required to
prevent an increase from current concentrations (Exhibit 2-16A).  To hold
tropospheric concentrations of CFC-11 constant would require an 80 percent cut
in current emissions (Exhibit 2-16B).

    Reductions in emissions less than 85 percent would only slow increases in
concentrations.  Exhibit 2-17 shows the concentrations of CFC-12 associated
with different emission trajectories.  For example, a 15 percent decrease in
current emissions of CFC-12 would imply a tropospheric concentration in 2100
of 1.6 ppb, approximately 3.9 times higher than the current concentration.  A
50 percent decrease in current emissions of CFC-12 would still result in
increases in tropospheric concentrations, and by 2100 a concentration of 1.01
ppbv, 2.5 times the current concentration.

    The disequilibrium for these long-lived gases between current emissions
and tropospheric concentrations has implications for the certainty of
estimating future concentrations.  Future tropospheric concentrations will
                            * *  DRAFT FINAL  * * *

-------
                             2-22
                        EXHIBIT 2-15A

              CFC-12:  Constant Emissions
                           (mill kg)
          500


          400


          300


          200


          100


           0
            1930
1985
2100
                       EXHIBIT 2-15B
        CFC12:  Atmospheric Concentrations
                           (ppbv)
        1.6
        1.2
        0.8
        0,4 -
         0
          1930
                      2100
If future emission of CFC-12 were held constant at  today's levels
(A),  atmospheric concentratons would continue to rise for over  100
years (B).  Computed with simplified model of source and loss
terms.  See Appendix to Chapter 2.
                   * *
                         DRAFT FINAL  * * *

-------
                         2-23
                     EXHIBIT 2-16A
             CFC-12:   Emissions
                      (mill kg)
      500


      400


      300


      200


      100  -


        0
         1930
 1985
  EXHIBIT 2-16B
                                           2100
    CFC-12 Atmospheric  Concentrations
                      (ppbv)
    0.40
    0.30
    0.20  -
    0.10 -
       0
       1930
1985
2100
A reduction of 85% in CFC-12 emissions (A) would be required to hold
concentrations constant (B).  Computed with simplified model of
source and loss terms.  See Appendix to Chapter 2.
                 * - *  DRAFT FINAL * * *

-------
                             2-24
                         EXHIBIT 2-17

               CFC -12:   Atmospheric Concentrations
               from Different Emission Trajectories
  -5  2
  I
  ~  1

  §
  g
  I
     0
                                     Constani,
                                     emissions
                                     15% Cul.
                                                           50% Cut
                                     85% Cut
        930
1905
2100
For CFC-12, holding emissions constant at 85% of today's level (a  15
percent  cut), would allow concentrations to increase by a factor of
four.  Only an emission cap at 15 percent of today's level (an 85
percent  cut) could hold concentrations constant.  Computed with
simplified model of source and loss  terms.  See Appendix to Chapter
                       *  DRAFT FINAL  * *

-------
                                   2-25
depend heavily on past emissions (e.g., closer to the present),  the value of
which we know or can predict with some certainty.  It is true that the further
one moves into the future, the greater the uncertainties associated with
emission estimates.  With long lived gases, however, it is also  true that the
further one moves into the future,  the smaller is the variation  in
concentrations that could occur because of variation in future emissions.  In
addition, with long lived gases, preventing rises in concentration will be
difficult because only large reductions will bring emissions and gases back
into equilibrium.  Freezes or small reductions will allow concentrations to
rise for decades.

TRACE GASES WITH SHORTER  LIFETIMES

    Several other stratospheric perturbants have relatively shorter
lifetimes.  Their concentrations will depend on emissions from relatively
shorter time periods.  Among the anthropogenic trace gases, CFC-22 and methyl
chloroform (CH3CC13) have lifetimes shorter than 50 years.

    CFC-22 and CH3CC13 are partially halogenated (i.e., they contain C-H
bonds) and react with OH, the hydroxyl radical, in the troposphere.  This
chemical removal process is responsible for their shorter lifetimes: 20 years
for CFC-22 (HAS 1984), and 6.5 years for CH3CC13 (WHO 1986).  (These lifetimes
are long, however, compared with those of molecules like carbon monoxide (CO),
which .have a 16-week average lifetime (Ramanthan et al. 1985).)

    Methane (CH4), a trace gas with both natural and anthropogenic sources,
has a lifetime of approximately 11 years (WHO 1986).  The major documented
removal processes for methane is also reaction with OH.  This removal process
appears to account for 98 percent of the total removal.  The remaining 2
percent is accounted for by consumption by aerobic soils.

    The current concentrations of these trace gases are not as far from
equilibrium with current emissions as are gases like CFC11 or 12.
Concentrations twenty or more years from now will rise at current rates only
if emissions increase or if chemical sinks increase so as to lengthen chemical
lifetimes.  Changes in emissions can, within a relatively shorter time, alter
future concentrations significantly.

CARBON  DIOXIDE AND THE CARBON  CYCLE

    The carbon cycle controls the distribution of carbon dioxide (C02)
throughout the biosphere.  Emissions from fossil fuels are only one part of
this complex combination of biogeochemical processes.  Other components of the
carbon cycle are the uptake of carbon by the terrestial biosphere and the
uptake, absorbtion, and outgassing of C02 in the oceans.  The carbon cycle is
discussed in greater detail in Chapter 4 of this assessment.  Interested
readers may also wish to consult the "state-of-the-art" report Atmospheric
Carbon Dioxide and the Global Carbon Cycle, recently issued by the U.S.
Department of Energy (Trabalka 1985) and The Carbon Cycle and Atmospheric
C02: Natural Variatons Archean to the Present (Sundquist and Broecker 1985).
                            * *  DRAFT FINAL  * *

-------
                                   2-26
SOURCE GASES FOR STRATOSPHERIC SULFATE AEROSOL

    Stratospheric sulfate aerosols (OCS,  CS2)  and volcanically injected
chlorine can in some cases reach and perturb the stratosphere.   While  most
sulfur gases emitted into the troposphere from natural  and anthropogenic
sources are too reactive and/or too soluble to reach the stratosphere, OCS  is
an important exception (Crutzen 1976; Sze and Ko 1979).   Apart from volcanic
injection, the major source of sulfur to  the stratosphere is  OCS from  the
troposphere.

    Carbonyl sulfide (OCS), the most abundant gaseous sulfur  carrier in the
atmosphere, has a nearly uniform concentration in the free troposphere of 500
pptv and an estimated lifetime of more than one year (WHO 1986).  Current
estimates of the global sources and sinks of OCS have been derived by
extrapolating a very limited data base and are subject  to large
uncertainties.  Natural sources include oceans, soils,  and coastal salt
marshes and an important secondary source -- the oxidation of CS2 by OH
(Harris and Niki 1984).

    Turco et al. (1980) conclude that "the total global source of OCS  is  1  to
10 tg per year, and that a large fraction of this, as much as 50%, may result
from anthropogenic activities related both to fuel processing and
consumption."  They note that the size of the OCS combustion  source is
sensititve to. the use of sulfur recovery  systems:  untreated  stack gases have
very low OCS concentrations, while the use of sulfur recovery systems
generates OCS.  Increases in the future use of scrubbers could thus increase
OCS emissions.

    A small contribution to OCS emissions may result from the use of catalytic
converters in automobiles, which occasionally generate large  quantities of
(H2S) and OCS.  Negligible anthropogenic  sources include direct commercial
production, cigarette smoke, vapours from cooking grain mashes,  and Kraft
mills.  Future concentrations of OCS cannot now be reliably predicted  due to a
lack of research in this area.  Given the possible importance of aerosols in
the stratosphere, this deficiency needs to be remedied (Turco et al. 1980).

    Volcanoes have long been recognized as dominant sources of stratospheric
sulfate and aerosol.  It has also been clear for some time that they could
also be sources of stratospheric chlorine.  Little research has been done to
quantify this source.  The amount of volatile material in the pre-eruption
magma varies from volcano to volcano, the amount of chlorine  in the volatile
material varies similarly, and soluble, polar compounds like  hydrochloric acid
(HC1) can be removed during the rapid rise (and condensation) of a volcanic
plume.  Hence volcanoes are sporadic sources not easily described by annual
averages.  It is clear, however, that only a fraction of volcanic eruptions
penetrate the stratosphere (WHO 1986).
                            * *  DRAFT FINAL  * *

-------
                                   2-27



                             APPENDIX A

                 CFC EMISSIONS-CONCENTRATIONS MODEL
    A simplified atmospheric parameterization,  taken from Rind and Lebedeff
(1984), can be used to relate future tropospheric CFC-11  and CFC-12
concentrations to their future surface emissions.1 Their  documentation
follows:

         ... we estimate these residence times  t,  to be 75 years
    for Fll and 150 years for F12.2  Thus,  for the year m the
    concentratioi
    is given by:
concentration C, (m)  as a function of the annual  release  R, (m)
                                                         K.

Ck(m) = 1

m -(m-e)/tk
Ek I e Rk(e).
e=1940
                                                         (A)
         According to our estimate the annual release for both
    f luorocarbonc during the year 1940 was zero.   The constant f,

    relates the mixing ratios C,  in ppbv of f luorocarbons to their

    annual release R,  (in millions kg/year).   These constants of

    proportionality were determined by comparing the computed
    concentrations with observed globally averaged values of Fll and
    F12 for the years 1977-1979 as reported by NOAA (1979) and (1980)
    in Geophysical Monitoring for Climatic Change No. 7 and No.  8.
    Global average concentrations were computed from the results of
    measurements of concentrations at five stations.  Locations  of
    these stations and the measured values are summarized in Tables 4
    and 5 (Exhibits A-l and A-2, this paper).  We assumed the
    concentrations to be zonally uniform and fitted expressions
1 The Rind and Lebedeff method does not consider changes in atmospheric
processes that may occur as a result of potential stratospheric ozone
depletion.  Results from one-dimensional (1-D) model calculations indicate
that if stratospheric ozone levels decrease, the increased ultraviolet
radiation that penetrates to the troposphere may reduce CFC-11 and CFC-12
lifetimes and hence the growth in their tropospheric concentrations.   This
process is unlikely to be significant in scenarios of low CFC growth
(Stolarski, personal communication).

2 While the Rind and Lebedeff estimate for the CFC-12 lifetime differs from
that of WHO (1986), the estimate is within the uncertainty band given by WMO.
                          * * *  DRAFT FINAL  * *

-------
                               2-28
                  Ck(X,0) =  I ae sin6"1
                            e=l
to the data in Exhibit 2-4.  Here  denotes latitude and X
longitude.  The global averages are shown in the last column of
Table 4 (Exhibit A-l, this paper).  The constants of
proportionality f  were obtained by fitting expressions (A) to

the data, yielding f   = 4.6395*10   and f   =

5.3279*10~5 ppb/(millions kg/year).
                      * * *  DRAFT FINAL  * * *

-------
                       2-29



                   EXHIBIT A-1

       Concentrations of Fluorocarbons (ppbv)
YEAR
 BRW
                           STATION
 NWR
MLO
SMO
SPO
GLOBAL
1977
1978
1979
                    Fluorocarbon 11  (CC13F)
0.159
0.172
0.182
0.155   0.148
0.168   0.162
0.175   0.174
       0.140   0.139
       0.153   0.154
       0.164   0.175
                    Fluorocarbon 12  (CC12F2)
                   EXHIBIT A-2

             Locations of the Stations
               0.145
               0.159
               0.171
1977
1978
1979
0.292
0.302
0.301
0.275
0.296
0.301
0.270
0.291
0.296
0.256
0.273
0.276
0.248
0.271
0.306
0.262
0.282
0.290
       NAME
            ABREV.
               LONGITUDE
                    LATITUDE
Point Barrow
Niwot Ridge
Mauna Loa
American Samoa
South Pole
BRW
NWR
MLO
SMO
SPO
130.60°W
105.63°W
155.58°W
170.56°W
24.80°W
70.32°N
40.05°N
19.53°N
14.25°S
89.98°S
  Source:  National Oceanic and Atmospheric
           Administration (1979) Geophysical Monitoring
           for Climate Change No.  7,  Summary Report
           1978, B.G. Bendonca, (ed.).
                * *  DRAFT FINAL  * * *

-------
                                   2-30
REFERENCES

Blake, D.R., E.W. Mayer, S.C. Tyler, Y.  Makide,  D.C.  Montague,  and F.S.
Rowland, (1982), "Global Increase in Atmospheric Methane Concentrations
between 1978 and 1980", Geophysical Research Letters, 9, 243-248.

Callendar,  G.S., (1958), "On the Amount  of Carbon Dioxide in the Atmosphere",
Tellus. 10, 243-248.

Crutzen, P.J. 1976, "The Possible Importance of  CSO for the Sulfate Layer of
the Stratosphere", Geophysical Research  Letters, 3, 73-76.

Delmas, R.J., J.M. Ascencio, and M. Legrand, (1980),  "Polar Ice Evidence that
Atmospheric C02 20,000 Yr.  BP was 50% of Present", Nature.  284, 155-157.

Ehhalt, D.H., R.J. Zander,  and R.A. Lamontagne,  (1983), "On the Temporal
Increase of Tropospheric CH4", Journal of Geophysical Research, 88, 8442-8446.

Eraser, P.J., M.A.K. Khalil, and R.A. Rasmussen, (1984), "Tropospheric Methane
in the Mid-Latitudes of the Southern Hemisphere", Journal of Atmospheric
Chemistry,  I, 125-135.

Freyer, H.D., (1978), "Preliminary Evidence of Past C02 Increase as Derived
     13
from   C Measurements in Tree Rings", in J. Williams  (ed.), Carbon Dioxide,
Climate and Society, Pergammon Press, New York.

Harris, R., and H. Niki, (1984), "Sufur  Cycle",  in Global Tropospheric
Chemistry:   A Plan for Action, National  Academy  of Sciences, Washington, D.C.

Hoffman, J.S., (1986), "The Importance of Knowing Sooner" in J.G.  Titus,
(ed.), Effects of Changes in Stratospheric Ozone and  Global Climate.   Volume
I: Overview, U.S. Environmental Protection Agency, Washington,  D.C.

Kavanuagh,  M., (1986), "Estimates of Future CO,  N20,  and NOx Emissions from
Energy Combustion", Atmospheric Environment, in  press.

Keeling, C.D., (1978), "Atmospheric Carbon Dioxide in the 19th Century",
Science. 202, 1109.

Keeling, C.D., J.A. Adams,  C.A. Ekdahl,  Jr., and P.O. Guenther, (1976),
"Atmospheric Carbon Dioxide Variations at the South Pole",  Tellus, 28,
552-564.

Keeling, C.D., and GMCC/NOAA, (n.d.), Monthly Concentrations of Carbon
Dioxide at  Mauna Loa, Hawaii, unpublished.

Khalil, M.A.K., and R.A. Rasmussen, (1981), "increase of CHC1F2 in the Earth's
Atmosphere", Nature, 292, 823-824.

Khalil, M.A.K., and R.A. Rasmussen, (1984), "Trichlorofluoroethane (F-113):
Trends at Pt. Barrow, Alaska", Geophysical Monitoring for Climate Change, No.
13, Summary Report, U.S. Department of Commerce, Washington, D.C.

                          * * *  DRAFT FINAL  *  * *

-------
                                   2-31
Khalil, M.A.K., and R.A. Rasmussen, (1985), "The Trend of Bromochlorodifluoro-
methane (CBrClF2) and the Concentrations of other Bromine Containing Gases at
the South Pole", Antarctic Journal of the United States:  1985 Review.

Khalil, M.A.K., and R.A. Rasmussen, (1986), "interannual  Variability of
Atmospheric Methane: Possible Effects of the El Nino-Southern Oscillation",
Science, 232, 56-58.

National Academy of Sciences, (NAS), (1976), Halocarbons: Effects on
Stratospheric Ozone, National Academy of Sciences, Washington, DC.

National Academy of Sciences (NAS), (1984), Causes and Effects of Changes in
Stratospheric Ozone, National Academy Press, Washington,  DC.

National Aeronautics and Space Administration (NASA), (1986),  Present State
of Knowledge of the Upper Atmosphere:  An Assessment Report.   Processes that
Control Ozone and Other Climatically Important Trace Gases, NASA, Washington,
DC.

Neftel, A., H. Oeschger, J. Schwander, B. Stauffer, and R. Zumbruun, (1982),
"Ice Core Sample Measurements Give C02 Content during the Past 40,000 Yr.",
Nature, 295, 220-223.

Oeschger, H., B  Stauffer, A. Neftel, J. Schwander, and R. Zumbruun, (1982),
"Atmospheric C02 Content in the Past Deduced from Ice-Core Analysis",
Annals of Glaciology, 3, 227-232.

Pearman, G.I., D. Etheridge, F. de Silva, and P.J. Eraser, (1986),  "Evidence
of Changing Concentrations of Atmospheric C02, N20 and CH4 from Air Bubbles  in
Antarctic Ice", Nature, 320, 248-250.

Peng, T.H., W.S. Broecker, H.D. Freyer, and S. Trumbore,  (1983), "A
                                     J 1 Q
Deconvolution of the Tree Ring Based    C Record", Journal of Geophysical
Research. 88, 3609-3620.

Peterson, J.T., W.D. Kohmyr, T.B. Harris, and L.S. Waterman,  (1982),
"Atmospheric Carbon Dioxide Measurements at Barrow, Alaska, 1973-1979",
Tellus. 34,  166-175.

Prinn, R.G., P.G. Simmonds, R.A. Rasmussen, R.D. Rosen, F.N.  Alyea, C.A.
Cardelino, A.J. Crawford, D.M. Cunnold, P.J. Fraser, and J.E. Lovelock,
(1983), "The Atmospheric Lifetime Experiment. 1. Introduction,
Instrumentation, and Overview", Journal of Geophysical Research, 88(C13),
8353-8367.

Ramanathan, V., R.J. Cicerone, H.B. Singh, and J.T.Kiehl, (1985), "Trace Gas
Trends and Their Potential Role in Climate Change," Journal of Geophysical
Research. 90(D3), 5547-5566.
                          * * *  DRAFT FINAL  * * *

-------
                                   2-32
Rasmussen, R.A., and M.A.K.  Khalil, (1982), "Atmospheric Fluorocarbons and
Methyl Chloroform at the South Pole",  Antarctic Journal of the United States:
1982 Review, 203-205.

Rasmussen, R.A., and M.A.K.  Khalil, (1984), "Atmospheric Methane in the Recent
and Ancient Atmospheres: Concentrations, Trends, and Interhemispheric
Gradient", Journal of Geophysical Research, 89(D7),  5547-5566.

Rasmussen, R.A., and M.A.K.  Khalil, (1986), "Atmospheric Trace Gases: Trends
and Distributions over the Last Decade", Science, 232, 1623-1624.

Rind, D., and Lebedeff, S.,  (1984), Potential Climatic Impacts of Increasing
Atmospheric C02 with Emphasis on Water Availability and Hydrology in the
United States, U.S. Environmental Protection Agency, Washington, DC.

Rinsland, C.P., J.S. Levine, and T. Miles, (1985), "Concentration of Methane
in the Troposphere Deduced from 1951 Infrared Solar Spectra", Nature, 318,
245-249.

Sze N.O. and Ko. M.K.W. (1979), "CS2 and COS in the Stratospheric Sulfur
Budget," Nature, 280, 308-310.

Thompson A.M., Cicerone R.J., "Possible Perturbations to Atmospheric CO, CH4
and OH," Journal of Geophysical Research. Vol. 91, #D10 Pg 10,853-10,864,
Sept. 20, 1986.                           .

Trabalka, J.R., (1985), Atmospheric Carbon Dioxide and the Global Carbon
Cycle, DOE/ER-0239, U.S. Department of Energy.  Washington, D.C.

Turco, R.F., R.C. Whitten, O.B. Toon,  J.B Pollack, and P. Hammill, (1980),
"DCS, Stratospheric Aerosols and Climate", Nature, 283, 283-286.

Weiss, R.F., (1981), "The Temporal and Spatial Distribution of Tropospheric
Nitrous Oxide", Journal of Geophysical Research, 86(C8), 7185-7195.

World Meteorological Organization, (1986), Atmospheric Ozone 1985.
Assessment of Our Understanding of the Processes Controlling its Present
Distribution and Change, World Meteorological Organization Global Ozone
Research and Monitoring Project -- Report No. 16, World Meteorological
Organiztion, Geneva, Switzerland.
                          * * *  DRAFT FINAL  * * *

-------
Chapter 3

-------
                                CHAPTER 3

                  EMISSIONS OF INDUSTRIALLY PRODUCED
                      POTENTIAL OZONE MODIFIERS
SUMMARY

    Human activities are the only source of emissions  for  three  important
classes of chemicals that may modify the amount  of  ozone in the  atmosphere:
(1) chlorofluorocarbons (CFCs);  (2)  chlorocarbons;  and (3) Halons.  CFCs and
chlorocarbons release free chlorine  into the stratosphere, and Halons release
free chlorine and/or bromine into the stratosphere  (there  are other natural
sources of chlorine in the atmosphere).

    Historically, CFCs have contributed  more to  potential  ozone  modification
than have chlorocarbons and Halons.   The most widely used  CFCs are CFC-11 and
CFC-12, which account for 80 percent of  today's  CFC production worldwide.
CFC-113 and CFC-22 are also commercially produced in large quantities.  A
variety of specialty CFCs are produced in small  quantities.

    Due to their unique physical properties, CFCs are  employed in a diverse
set of applications, including:   aerosol propellants;  refrigeration; air
conditioning; solvents; and foam production.  Substitutes  for CFCs exist for
many of these applications, but  generally are inferior due to hazards that
they pose,  increased costs, or reduced product quality.  In some applications
(such as refrigeration), the CFCs remain captured in the product for many
years before being released.

    Data describing the historical production and use  of CFCs for various
parts of the world have been collected by a variety of organizations.  The
most comprehensive data on the production of CFC-11 and CFC-12 have been
compiled by the Chemical Manufacturers Association  (CMA) from reports by CFC
producers.   These data indicate  that production  of  CFC-11  and CFC-12 peaked in
1974, and subsequently declined through  1982. Production  since  1982 has
increased each year.

    The decline in the production of CFC-11 and  CFC-12 from 1974 to 1982 is
attributable primarily to the reduction  in their aerosol propellant use.  The
United States implemented a ban on the use of CFCs  in  nonessential aerosol
applications in 1978.  Several other countries and  groups  of countries have
instituted controls on aerosal applications with varying degrees of stringency.

    During the period when aerosol applications  were declining,  nonaerosol
applications (refrigeration, air conditioning, solvents, foam production)
continued to increase.  In 1978, nonaerosol usage exceeded aerosol usage, and
by 1985 nonaerosol usage is expected to  be in excess of 70 percent of total
CFC applications.  Therefore, CFC usage  is increasingly dominated by
nonaerosol applications.
                          * * *  DRAFT FINAL  * *

-------
                                   3-2
    Long term projections of CFCs are needed for risk assessment because of
their long lifetimes.  Projections of the potential future demand for CFCs
were presented at a UNEP economics workshop on CFCs in May of 1986, and are
found in the trade literature.  Indications are that in the absence of
additional controls on production, use, or emission, total CFC production is
expected to increase in the foreseeable future.  Based on the UNEP workshop,
a range of zero to five percent growth is accepted as possible for CFC-11 and
CFC-12 for the long term.  A rate of approximately 2.5 percent per year is
recognized as a figure in the middle of the range.  However, long term
projections of potential growth are viewed as very uncertain.

    Underlying this range of growth rates for CFCs is the expectation that
aerosol applications will continue to decline and level off.  The future
growth in production is expected to be driven by increases in nonaerosol
applications.  Chlorofluorocarbon 113 and 22 are expected to grow faster than
CFC-11 and CFC-12.

    The chlorocarbons (carbon tetrachloride and methyl chloroform) are used
primarily as solvents and chemical intermediates.  In the United States carbon
tetrachloride is primarily used to make CFCs and only a small amount of this
use is emitted to the atmosphere.  In developing countries carbon
tetrachloride is believed to be used as a solvent, resulting in considerable
emissions.  Methyl chloroform is primarily used as a general purpose solvent
worldwide.  Future demand for methyl chloroform is expected to grow at a rate
similar to rates of growth of economic activity.  Future demand for carbon
tetrachloride is expected to grow at the same rate as the demand for CFCs.

    Halons have been used in hand-held and total-flooding fire extinguishers
since the 1970s.  Their unique properties make Halons valuable for protecting
delicate electronic equipment such as computers.  Annual production has been
estimated as small (approximately 20,000 kilograms) and emissions have been
assumed, by some authors, to be very small because the Halons remain in the
fire extinguishers for many years.  Recent discussions with fire protection
experts, however, have raised concern about the possibility that Halon
emissions are underestimated.  Halon production and emissions are expected to
grow significantly over the next 15 years, although the potential for
long-term growth is uncertain.
                          * * *  DRAFT FINAL  * * *

-------
                                   3-3
FINDINGS

1.  HUMAN ACTIVITIES ARE THE ONLY SOURCE OF EMISSIONS FOR THREE IMPORTANT
    CLASSES OF POTENTIAL OZONE MODIFIERS:  CHLOROFLUOROCARBONS (CFCS);
    CHLOROCARBONS (CARBON TETRACHLORIDE AND METHYL CHLOROFORM); AND HALONS.

    la.  Since their development in the 1930s, CFCs have become important
         inputs for the production of consumer and industrial goods,
         including:  aerosol propellants; air conditioning; refrigeration;
         foam products (e.g., cushions, insulating foams); solvents used in
         the manufacture of electronic components; and a variety of
         miscellaneous uses.

    Ib.  Historically, CFC-11 and CFC-12 have dominated the use and emissions
         of CFCs, accounting for over 80 percent of today's CFC production
         worldwide.  CFC-113, however, is becoming increasingly important.

2.  NONAEROSOL USES OF CFCS HAVE GROWN CONTINUALLY SINCE THE DEVELOPMENT OF
    CFCS AND APPEAR CLOSELY COUPLED TO ECONOMIC GROWTH.

    2a.  From 1960 to 1974, the combined emissions of CFC-11 and CFC-12 from
         both aerosol and nonaerosol applications grew at an average annual
         compounded rate of approximately 8.7 percent, resulting in the peak
         annual emissions in 1974 of these CFCs, over 700 million kilograms.

    2b.  From 1976 to 1984, the annual production of CFC-11 and CFC-12 for
         aerosol applications has declined at an average annual compounded
         rate of over 8 percent.  During the same period, production for
         nonaerosol applications grew at an average annual compounded rate of
         5 percent.  As a result total production in 1984 was nearly that in
         1976.

3.  STUDIES ON FUTURE GLOBAL PRODUCTION AND EMISSIONS OF CFCS PROJECT AN
    INCREASE AT AN AVERAGE ANNUAL RATE OF APPROXIMATELY 1.0-4.0 PERCENT OVER
    THE NEXT 15 TO 65 YEARS.
    3a.  Aerosol propellant applications are expected to remain constant or
         decrease in many portions of the world.

    3b.  Growth in developed countries for nonaerosol applications is expected
         to be driven by uses for foam blowing and solvent use.

    3c.  The most likely annual rate of growth for CFC-11 and CFC-12 over the
         next 65 years is 2.5 percent.  Such a rate assumes that CFC growth
         slows from its historical relationship with economic growth even in
         the absence of regulation. CFC-113 and CFC-22 are expected to grow
         much faster.

    3d.  Key uncertainties include technological changes over the long term
         and the patterns of use in developing countries.
                          * * *  DRAFT FINAL  * *

-------
                                   3-4
4.  THE CHLOROCARBONS ARE USED PRIMARILY AS SOLVENTS AND CHEMICAL
    INTERMEDIATES.

    4a.  Methyl chloroform is primarily used as a general purpose solvent.
         Use in 1980 has been estimated at nearly 460 million kilograms.
         Projections of growth have been quite limited.  Those made indicate
         that it is expected to grow at the rate of growth of economic
         activity (as measured by GNP).

    4b.  Carbon tetrachloride is primarily used to make CFCs in the U.S.  In
         developing countries it may also be used as a general purpose
         solvent.  Future production of carbon tetrachloride is expected to
         follow the pattern of production of CFCs.  Emissions are expected to
         remain small.

5.  HALONS HAVE BEEN USED IN HAND-HELD AND TOTAL-FLOODING FIRE EXTINGUISHERS
    SINCE THE 1970s.  ANNUAL PRODUCTION HAS BEEN ESTIMATED AS SMALL
    (APPROXIMATELY 20,000 KILOGRAMS) AND EMISSIONS HAVE BEEN ASSUMED SMALL
    BASED ON THE CONSERVATIVE ASSUMPTION THAT THE HALONS REMAIN INSIDE THE
    FIRE EXTINGUISHERS.  RECENT DISCUSSIONS WITH FIRE EXPERTS CAST SOME DOUBT
    THAT THIS ASSUMPTION IS TRUE.

    5a.  A single projection of the future demand for these Halon-based fire
         extinguishers indicates that their demand is growing rapidly and that
         production may double by the year 2000.  In that study, long-term
         demand is judged uncertain and may range from an average annual
         decline of 1 percent from 2000 to 2050 to an annual increase
         exceeding 5 percent.

    5b.  The expected rate of emissions from these Halon applications is not
         known precisely.  In the one study done, an assumption of low
         emissions is made.  That study has been the basis for scenarios used
         in this analysis and in the risk estimates presented in Chapter 8.

    5c.  Atmospheric concentrations of Halon-1211 have been shown to increase
         in excess of 20 percent per year in recent years, indicating
         significantly higher emissions than assumed in the one existing
         study.  No measurements are available of Halon-1301.

    5d.  Discussions with Halon users indicate that emissions may be much
         larger than assumed.  Possible future standards for full scale
         testing of Halon 1301 systems in the U.S. would result in emissions
         much larger than the levels assumed here.
                              *  DRAFT FINAL

-------
                                   3-5
INTRODUCTION

    Man's activities are the primary source of emissions for several important
potential modifiers of stratospheric ozone, including:

        •   Chlorofluorocarbons (CFCs), including:   CFC-11;
            CFC-12; CFC-22; and CFC-113;

        •   Chlorocarbons, including:  methyl chloroform and
            carbon tetrachloride;  and

        •   Halons, including: Halon-1211 and Halon-1301.

The atmospheric concentrations of these gases are driven by their emissions,
and by the natural physical processes that destroy and transport them (thereby
removing them from the atmosphere).   This chapter describes  the historical and
expected future use and emissions of each of these three classes of potential
ozone-modifying substances.  A long term perspective on emissions is essential
for assessing the risks associated with various emission trajectories because
of the long lifetime of most of these chemicals.
                            i
    CFCs are discussed first.   Historically, these compounds have contributed
the most to potential modification of stratospheric ozone.1   Chlorocarbons
are next in importance (despite the fact that their production volume is
larger than CFCs), and are discussed second.  Finally,  the information on the
relatively small amount of Halon production and use is presented.  Despite
their relatively small production volume, Halons are considered to be growing
in importance due to their strong potential for modifying stratospheric ozone.

CHLOROFLUOROCARBONS

    CFCs are a class of man-made chemicals that contain chlorine, fluorine,
and carbon.  There are four major CFCs that are produced in commercial
quantities:2

            •   CFC-11;
                CFC-12;
                CFC-22; and
                CFC-113.
    1 The relative importance of the compounds in terms of stratospheric
ozone modification depends on a variety of factors relating to the chemical
composition of each substance.  Camm et al. (1986) report the following
relative importance as of 1985:  CFCs -- 77.8 percent; Chlorocarbons -- 16.7
percent; and Halons -- 5.6 percent.

    2 Other CFCs produced in limited quantities include:  CFC-13, CFC-14,
CFC-21, CFC-23, CFC-114, CFC-115, CFC-142b, CFC-152a.  (See Hoffmann and
Klander 1978, p. 8).  Other CFCs with potential applications that are not
currently available commercially include CFC-123 and CFC-134a.
                          * * *  DRAFT FINAL  * *

-------
                                   3-6
CFC-11 and CFC-12 account for the largest share of total CFC production,
approximately 80 percent in recent years.

    CFCs are used in a variety of applications that require chemicals with a
fairly unique set of physical properties.  The properties of CFCs that make
them valuable include:

    •   CFCs can be used with a variety of materials,
        including plastics;

    •   CFCs are safe to use, they are not flammable and they
        are relatively nontoxic;

    •   CFCs have a low boiling point; this factor is
        important in manufacturing foams and for refrigeration
        applications; and

    •   CFCs have good thermodynamic properties.

Exhibit 3-1 displays selected properties of three of the major CFCs.

    These unique properties enable CFCs to be used in a variety of important
applications.  Initially, CFCs displaced ammonia in early refrigeration
applications.  Ammonia was used in refrigerators that were replacing ice boxes
in the early twentieth century.  However, ammonia is more toxic and
combustible than CFCs, and consequently the switch to CFCs reduced the hazards
of owning a refrigertor.

    In the 1950s, aerosol propellant uses of CFCs were first introduced.
These applications grew into large markets of personal care products such as
hair sprays and deodorants.  Also during the 1950s, refigeration and air
conditioning applications grew rapidly.  By the 1960s, these applications were
well established.

    Two new important uses for CFCs were introduced in the 1960s.  During this
decade, CFC-11 was first used to make plastic foams.  These foams are used as
seat cushions (such as in furniture and cars) and as carpet backings.  The
foams are also used as insulation in appliances (such as refrigerators) and
buildings.  The use of CFC-11 in these foam applications has grown so rapidly
that this is now the dominant nonaerosol use of CFC-11.  The other major use
introduced in the 1960s is the application of CFC-12 to automobile air
conditioners.  In the U.S., the market for automobile air conditioners has
grown rapidly and now is an important use of CFC-12.

    The most recent market for CFCs is for applications that require a
solvent.  In particular, the use of CFC-113 as a solvent in the manufacture of
electronic components and computer chips has been expanding rapidly.  CFCs are
preferred as solvents in these applications because they are mild and are not
toxic.  Its potential as a solvent for metal cleaning may also be large, but
was not explicitly analyzed in this chapter.
                          * * *  DRAFT FINAL  * * *

-------
                                                    EXHIBIT 3-1
                                            Selected Properties  of CFCs

Chein i ca 1 Forniu 1 a
Molecular V/eight
Bo i 1 i rig Po int, " F
Freezing Point, "F
Vapor Pressure, psig
At 70"F
At 130"F
3
Liquid Density, gm/cm
At 70 "F
At 130~F
Atmospheric Lifetime,
yea rs
CIC-11
CO I F
3
137.4
74.8
-168
13.4*
24.3
1.485
1.403
84
CFC-12
CCI
2
120
-21
-252
70
181
1
1
148
F
2
.9
.6

.2
.0
.325
. 191

CFC-22
CHC I F
86.
-41 .
-256
122.
300.
1 .
1 .
33
2
5
4

5
0
209
064

CFC-1 13
CC1 FCC I F
2 2
187.4
117.6
-31
21.2 (at 77 F)
NA
1.565 (77 F)
NA
88
CFC-142b
CH CCIF
3 2
100.5
14.4
-204
NA
NA
1 .113 (77~F)
NA
NA
CFC-1 23 CFC-1 34a
CHC I CF
2 3
152.9
80.7
NA
NA
NA
1.475 (at 60 F)
NA
NA
CF CM F
3 2
102
-15.7 F
NA
NA
NA
OJ
i
NA
NA
6.4
* psia.
The unique properties of CFCs make them a  highly valued  class  of chemicals.
Sources:   Hoffmann,  B.L.,  and D.S.  Klander (1978),  Final  EIS FIuorocarbons:   Environmental  and Health Implications.
          FDA,  p.  9.
          Cormell,  P.S.  (1986),  "A Parameterized Numerical  Fit to Total  Column Ozone  Changes Calculated by the LLNL
          1-D Model  of the Troposphere  and Stratosphere,"  Lawrence Livermore  National  Laboratory,  Livermore,
          CaI i forn i a, p.  4.
           WHO Criteria  Document on ChIorofIuorocarbons  (1986),  Draft  Report,  Environmental  Criteria and Assessment
           Office,  U.S.  EPA,  Cincinnati, Ohio,  p.  2-10.
           Memo from Lynn Erikson,  Radian  Corporation to  Neil  Patel,  EPA,  July 1986.
           Anderson,  L.G.  (1980),  "The  Atmosphere  Chemistry of 1,1,1-2-tetrafIuoroethane,"  General  Motors Research
           Laboratory, Warren,  Michigan, GMR-3450.
                                              *  *  *  DRAFT  FINAL  *  *  *

-------
                                   3-8
    Exhibit 3-2 provides the following information about the four major CFCs
used by industry today; the major applications; the properties that make them
useful; the potential for non-CFC substitutes (there may be CFC chemical
substitutes such as CFC-123 and CFC-134 that are not listed); and the
consequences of switching to the substitutes.  For example, refrigeration
applications use CFCs 11, 12, and 22.  These compounds are used because of
their thermodynamics properties, safety, and low cost.  Potential alternatives
to these compounds include ammonia, sulfur dioxide, and methyl chloride.
However, these chemicals are more toxic, flammable, corrosive, explosive,
and/or less energy efficient than CFCs.  Analogous descriptions are shown for
the other applications.

    Historical Production and Emissions

    Data on the production of CFC-11 and CFC-12 have been collected by:

        •   The Chemical Manufacturers Association (CMA) for the
            years 1931 to 1985;

        •   The United States International Trade Commission
            (ITC); and

        •   The European Economic Community  (EEC).

The CMA data were supplied by 21 CFC-producing companies that operate plants
around the world.  These companies and the locations of their plants are
listed in Exhibit 3-3.  These reporting companies represent CFC production in
almost all of the world except for India, Argentina, the East Bloc countries,
and China.  Production is known to take place in Argentina and India, but data
on their production is sparse, and the level of production is believed to be
low.

    The most recent report issued by the CMA is "Production, Sales, and
Calculated Release of CFC-11 and CFC-12 Through 1985," September, 1986.
Exhibit 3-4 lists the historical production data for CFC-11 and CFC-12
reported to the CMA.  Exhibit 3-5 displays these data for CFC-11 and CFC-12
combined in a graphical form for 1960 to 1985.  As the exhibits show, total
production increased rapidly throughout the  1960s and early 1970s, reaching a
maximum of 813 thousand metric tons in 1974.  During this period, the combined
production of these two CFCs grew at an average rate of 14 percent per year.

    However, beginning in the mid-1970s, public attention focused on the
potential for CFCs to deplete stratospheric ozone.  In response to these
concerns, the United States, a major producer of CFCs, instituted a ban on the
use of CFCs in nonessential aerosol propellant applications in 1978.  Several
other countries also imposed controls on the production and use of CFCs with
varying degrees of stringency.  For example, the members of the European
Economic Community (EEC) agreed to reduce the use of CFCs in aerosols to 70
percent of what they had been in 1976.  The  subsequent use of CFC-11 and
CFC-12 in aerosols as reported by the EEC (1985) is displayed in Exhibit 3-6.
                          * * *  DRAFT FINAL  * * *

-------
                                                           EXHIBIT 3-2

                                               CFC Characteristics and Substitutes
        AppIica t ion
'lype of
CJ c Used
  Key Characteristics
      of CFC Used
      Potential Alternatives
  Consequences of Using
      Al terna t ives
Refrigerat ion
Ai r Cond i t ion i ng
Plastic Foam
So I vents
Aerosol  Propellants
CFC-I I
crc-12
CI'C-22
crc-ii
CFC-12
CFC-22
CFC-I 13
Automobile Air Conditioning    CFC-I 2
CFC-I I
CFC-12
CFC-I I

CFC-I 13
CFC-11
CFC-12
Thermodynamic properties
Safety
Cost
Thermodynamic properties
Safety
Cost
           Thermodynamic properties
           Safety
           Cost
Thermodynamic properties
Safety
Cost
Ability to displace all
   contam i nants
Chemica My inert
Safety

Thermodynamic properties
Safety
Ammon i a
SuIphur D iox ide
Methyl  Chloride
Ammon i a
Sulphur Dioxide
Methyl  Chloride
                           Ammonia
                           Sulphur Dioxide
                           Methyl  Chloride
None for high-efficiency insulation
Pentane (some foams)
Methylene Chloride (some foams)
Perch Ioroethylene
Trichlorethylene
Trichloroethane
Hydroca rbons
Carbon Dioxide
More toxic
Combust i bIe
Co rros i ve
Explos ive
Less energy efficient

More toxic
Combust i ble
Corros ive
Fxplos ive
Less energy efficient

More toxic
Combust i ble
Corros ive
Explos ive
Less energy efficient

Less effective  insulation
Combust ible
Processing difficulties
Toxi c i ty

More toxic
Uses more energy
Combust ible
Reduced product quality
CFCs are used in a  diverse  set of applications.  Substitutes for CFCs have undesirable consequences in many applications.

Source:   Alliance for Responsible CFC  Policy  (I960), "An Economic Portrait of the CFC-UtiIizing Industries  in the United  States,"
         Washington,  D.C.
                                                     * * * DRAFT FINAL * * *

-------
                                   3-10



                               EXHIBIT 3-3

                      Companies  Reporting Data to CMA


    The following is a listing of the reporting companies inclusive of any
related subsidiaries and/or joint ventures that reported CFC production and
release data:

        1.  Akzo Chemie B.V. (Holland)

        2.  Allied Corporation (U.S.)

            (a) Allied Canada Inc. (Canada)
            (b) Quimobasicos, S.A. (Mexico)

        3.  Asahi Glass Co., Ltd. (Japan)

        4.  ATOCHEM, S.A. (France)

            (a) Pacific Chemical  Industries Pty.  Ltd.  (Australia)
            (b) Ugimica S.A. (Spain)
            (c) Produven (Venezuela)

        5.  Australian Fluorine Chemical Pty/Ltd. (Australia)

        6.  Daikin Kogyo Co., Ltd. (Japan)

        7.  Du Pont Canada, Inc.  (Canada)

        8.  Du Pont Mitsui Fluorochemicals Co., Ltd.  (Japan)

        9.  E.I. du Pont de Nemours & Company, Inc. (U.S.)

            (a) Du Pont de Nemours (Netherlands)  N.V.
            (b) Ducilo S.A. (Argentina)
            (c) Du Pont do Brasil S.A. (Brazil)
            (d) Halocarburos S.A. (Mexico)

        10. Essex Chemical Corporation (Racon) (U.S.)

        11. Hoechst AG (West Germany)

            (a) Hoechst Iberica  (Spain)
            (b) Hoechst do Brasil Quimica e Farmaceutics S.A.

        12. Imperial Chemical Industries PLC  (England)

            African Explosives & Chemical Industries,  Ltd.
                                 DRAFT FINAL  * *

-------
                                   3-11



                         EXHIBIT 3-3 (Continued)

                      Companies  Reporting Data to CMA


        13.  I.S.C.  Chemicals Ltd.  (England)

        14.  Kaiser Aluminum & Chemical Corporation (U.S.)

        15.  Kali-Chemie Aktiengesellschaft (West  Germany)

        16.  Kali-Chemie Iberia SA (Spain)

        17.  Montefluos S.p.A.  (formerly  Montedison S.p.A.)  (Italy)

        18.  Navin Fluorine Industries  (India)

        19.  Pennwalt Corporation  (U.S.)

        20.  Showa Denko, K.K.  (Japan)

        21.  Societe des Industries Chimiques du Nord  de la  Grece, S.A.  (Greece)

        22.  Union Carbide Corporation  (U.S.)

            (Union Carbide ceased production in 1977.   CFCs resold by Union
            Carbide are included).
Source:   CMA (1985),  "Production,  Sales,  and Calculated  Release  of  CFC-11
         and CFC-12 Through 1984," Schedule 1,  Listing of  Reporting Companies,
         Washington,  D.C.
                            * *  DRAFT FINAL  * *

-------
                    3-12
                EXHIBIT 3-4

Production of CFC-11 and CFC-12 Reported to CMA
            (millions of kilograms)
Years
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
CFC-11
_
-
-
-
-
0.1
0.1
0.1
0.1
0.2
0.3
0.3
0.4
0.4
0.4
0.7
1.3
3.0
4.5
6.6
9.1
13.6
17.3
20.9
26.3
32.5
33.9
29.5
35.6
49.7
60.5
78.1
93.3
111.1
122.8
141.0
159.8
183.1
217.3
238.1
CFC-12
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
267.5
297.3
321.1
           * * *  DRAFT FINAL  * * *

-------
                    3-13
           EXHIBIT 3-4  (Continued)

Production of CFC-11 and CFC-12 Reported to CMA
            (millions of kilograms)
       Years          CFC-11          CFC-12
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
263.2
306.9
349.1
369.7
314.1
339.8
320.5
308.9
289.5
289.6
286.9
271.4
291.7
312.4
326.8
341.6
379.9
423.3
442.8
381.0
410.7
382.8
372.1
357.2
350.2
351.3
328.0
355.3
382.1
376.3
       Source:  CMA, 1986, "Production,  Sales,
               and Calculated Release of
               CFC-11 and CFC-12 Through
               1985," Schedules 2 and 3,
               Washington, B.C.
           * * *  DRAFT FINAL  * * *

-------
                                            3-14



                                        EXHIBIT 3-5

                          Historical Production of CFC-11 and  CFC-12
tt
o
0
o
       900
      800 -
      700 -
      600 -
      500 -
                                                                                         TOTAL
                                                                                         NONAERO!
                                                                                         AEROSOL
      100
        0
          1960
1965
1970
1975
1980
1985
                                            YEAR
         Total  reported production of CFC-11 and CFC-12 increased rapidly throughout
         the  1960s and 1970s, reaching a maximum of 813 thousand metric tons in 1974.
         Aerosol applications declined since the mid-1970s, while nonaerosol
         applications continued to increase.   (Note:  aerosol/nonaerosol divisions
         prior  to 1976 are estimates.)


         Source:  CMA (1986), "Production, Sales, and Calculated Release of CFC-11 and
                  CFC-12 Through  1985,." Washington, D.C.
                                     * *  DRAFT FINAL  * * *

-------
                                3-15
                            EXHIBIT 3-6

                  CFC-11 and  CFC-12 Used in Aerosol  and
                   Nonaerosol  Applications  in  the  EEC
Year
1976
1977
1978
1979
1980
1981*
1982*
1983*
1984*
Aerosol Use
(millions of
kilograms)
176.9
162.5
150.4
136.5
126.4
116.1
111.7
113.9
114.7
Aerosol Use
Relative to 1976
(percent)
100
92
85
77
71
66
63
64
65
Nonaerosol Use
(millions of
kilograms)
67.1
70.4
81.0
83.0
90.4
93.6
95.1
102.5
103.1
Nonaerosol Use
Relative to 1976
(percent)
100
105
121
124
135
139
142
153
154
* Includes data for Greece, whereas the 1976 to 1980 data do not.

Source:   EEC (1985),  "Chlorofluorocarbons in the Environment:   Updating the
         Situation," Communication from the Commission to the Council.
                        *  * *  DRAFT FINAL  * * *

-------
                                   3-16
    The effect of these controls on aerosol applications, as well as the
public perception of CFC dangers, is clearly evident; total production
reported to CMA began to decline in 1975.  This decline was driven by
reductions in the use of CFC-11 and CFC-12 in aerosol propellant applications,
which occurred in both the U.S. and the EEC.

    In contrast to the decline in production for aerosol propellant applica-
tions, production of CFCs for nonaerosol applications reported to CMA continued
to increase throughout the 1970s and the 1980s (1976-1985) at an average rate
of over 5 percent per year.  Reported use of CFCs for nonaerosol applications
in the EEC has also increased during this period (see Exhibit 3-6).

    The CMA reports that the largest nonaerosol use of CFC-11 is currently as
a blowing agent for making closed cell foam (CMA 1986, Schedule 5).  In this
application, the CFCs not only help form the cells in the foam, but they also
increase the insulating properties of the foam.  The CMA estimates that the use
of CFC-11 for this application currently accounts for approximately 36 percent
of total CFC-11 production, and that it more than doubled from 1976 to 1985,
growing at an annual average rate of over 9 percent.  The second largest
nonaerosol application is reported by CMA to be as an auxilliary blowing agent
in the production of open cell foams, accounting for approximately 19 percent
of current CFC-11 production.  Refrigeration is reported as accounting for 8.2
percent, and all other uses account for 5.7 percent.

    Hammitt (1986) developed independent estimates of the shares of CFC-11
production going to each of the applications identified by the CMA.  These
estimates were developed by estimating the intensity of use of CFCs in each of
its applications and then estimating the amount of production of each applica-
tion  (e.g., foams, refrigerators, aerosol products, and miscellaneous).
Hammitt's results are very similar to the CMA estimates, and are compared in
Exhibit 3-7.

    As shown in the exhibit, Hammitt's estimates conform closely with the CMA
numbers.  However, Hammitt's method reportedly resulted in estimates of CFC-11
use that are 8 percent less that the reported total production.  This
discrepancy is listed by Hammitt under "unallocated" in Exhibit 3-7, and it
indicates that while there is general agreement on the approximate magnitudes
of the use of CFC-11 in its various applications, there is some difficulty in
accounting for the total production.

    Similar work has been done to estimate the current use of CFC-12.  The CMA
reports the largest current use to be in refrigeration applications,
accounting for 49 percent of the total production in 1985.  Between 1976 and
1985  the use of CFC-12 in refrigeration applications (reported to CMA) grew at
an average rate of 4.2 percent per year.  Aerosol propellant applications are
reported as accounting for 32 percent, and the remaining nonaerosol uses are
reported as accounting for approximately 19 percent.

    Hammitt's estimates for the share of CFC-12 going to various applications
are compared to the CMA estimates in Exhibit 3-8.  Compared to CFC-11, there
is less agreement on the amount of CFC-12 currently being used in its various
                              *  DRAFT FINAL

-------
                        3-17
                    EXHIBIT 3-7

       Comparison of Estimated CFC-11  Use:   1985
Application
Aerosol Propellant
Foam Production:
Closed Cell
Open Cell
Refrigeration and
Air Conditioning-
All Other Uses
Unallocated Use
Total
N/A = not applicable
Sources: Hammitt, J
Potential
Hammitt
Use
(103 kilograms)
93,700

115,800
57,000
9,900
N/A
23,600
300,000
•
.K. , et al. (1986)

(percent
of total)
31

39
19
3
0
8
100

, Product
Ozone-Depleting Substances,
CMA
Use
(103 kilograms)
100,500

117,300
63,500
26,900
18,600
N/A
326,800


(percent
of total)
31

36
19
8
6
0
100

Uses and Market Trends for
1985-2000, The RAND

Corporation,  p.  5  and p.  94.  CMA (1986), "Production,  Sales,  and
Calculated Release of CFC-11 and CFC-12 Through 1985,"  Schedule  5,
                * *  *  DRAFT FINAL  * *

-------
                                3-18
                           EXHIBIT 3-8

              Comparison of Estimated CFC-12 Use:  1985
                            Haramitt
                                         CMA
Application
      Use          (percent
(103  kilograms)    of total)
                                                     Dse
                  (percent
(103  kilograms)   of total)
Aerosol Propellant
Foam Production:
Closed Cell
Open Cell
Refrigeration and
Air Conditioning
All Other Uses
Unallocated Use
Total
N/A = not applicable
Sources: Hammitt, J
Potential
115,600
42,800
98,300
25,400
82,900
365,000
•
.K., et al. (1986),
32
12
27
7
22
100

Product
Ozone-Depleting Substances,
119,700
30,200
20,400
185,000
21,000
N/A
376,300

Uses and Market Trends
1985-2000, The RAND
32
8
5
49
6
0
100

for

       Corporation,  p.  6 and p.  95.  CMA  (1986), "Production, Sales, and
       Calculated Release of CFC-11  and CFC-12 Through 1985," Schedule 6.
                              DRAFT  FINAL  * * *

-------
                                   3-19
applications.   The primary difference between the CMA and Hammitt estimates is
that Hammitt has only 27 percent of the production going to refrigeration
applications,  while CMA has 49 percent.  This difference appears to be
accounted for by a 22 percent share listed as "unallocated" by Hammitt.

    The differences in the Hammitt and CMA allocation estimates indicate that
there is some uncertainty regarding the current use of CFC-12.

    Describing the allocation of CFC production across its possible
applications is required for assessing the current (and future) emissions of
the compounds  to the atmosphere.  Some applications emit their CFCs
immediately upon use, and are termed "prompt emitters."  These applications
include aerosol propellant, production of open cell (or flexible) foam,
solvent applications, and some miscellaneous applications.  Other applications
store their CFCs for periods of time, often years.  These applications include
refrigeration, air conditioning, and the production of closed cell rigid
insulating foams.

    Estimates  have been made of the duration for which CFCs are stored in
these products (see, for example, Quinn et al. 1986,  Gamlen et al. 1986,
McCarthy et al. 1977, and Khalil and Rasmussen 1986).  Exhibit 3-9 displays
the estimates  of emissions for 1971 to 1985 developed by the CMA for CFC-11
and CFC-12 based on their assessments of the allocation of production across
the end uses and the duration for which CFCs are stored in the various
products.  As  shown in the exhibit, because CFCs are stored in many products,
emissions generally do not equal production in any given year.

    Data describing CFC production in the East Bloc countries and China are
sparse.  Very little has been published about the historical and current rates
of production in these countries.  Published estimates of production in the
U.S.S.R. are displayed in Exhibit 3-10 (these values are not included in the
CMA total production values shown in Exhibit 3-4).  Data have not been
reported that divide these production figures into aerosol and nonaerosol
applications.

    Zhijia (1986) presented estimates of current CFC production in China at
the May 1986 UNEP Workshop on the Control of Chlorofluorocarbons.  Zhijia
reports annual production of "18 thousand tons" for CFCs 11, 12, 13, 14, 21,
22, 112, 113,  and 114.  Half of this production "is used for high-molecular
polymer as raw materials;' another half for the production of refrigerant,
aerosol propellant etc." (Zhijia 1986, p. 1).

    Based on the data reported in Exhibit 3-10 and assumed rates of growth
since 1975, Hammitt estimated the 1985 production of CFC-11 and CFC-12 in the
"communist" countries to be 41.5 thousand metric tons and 78.7 thousand metric
tons, respectively.  Using Hammitt's assumption that 15 percent of total
CFC-11 and CFC-12 production in these countries is outside the U.S.S.R., the
implied annual production for CFC-11 and CFC-12 in the U.S.S.R. is 105
thousand metric tons, for an annual average growth rate from 1975 to 1985 of
10.3 percent.   This rate of growth is much lower than the average of 19.6
percent per year from 1968 and 1975.  However, the growth rate for production
                          * * *  DRAFT FINAL  * * *

-------
                          3-20
                     EXHIBIT 3-9

Estimates of Production and Emissions of CFC-11  and CFC-12
                   (millions of kilograms)
CFC-11
Year
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
Annual
Production
263.2
306.9
349.1
369.7
314.1
339.8
320.5
308.9
289.5
289.6
286.9
271.4
291.8
312.4
326.8
Annual
Emissions
226.9
255.8
292.4
321.4
310.9
316.7
303.9
283.6
263.7
250.8
248.2
239.5
252.8
271.1
280.8
CFC-12
Annual
Production
341.6
379.9
423.3
442.8
381.0
410.7
382.8
372.1
357.2
350.2
351.3
328.0
355.3
382.1
376.3
Annual
Emissions
321.8
349.9
387.3
418.6
404.1
390.4
371.2
341.3
337.5
332.5
340.7
337.4
343.3
359.4
368.4
    Source:   CMA  (1986), "Production, Sales,  and Calculated
             Release of CFC-11 and CFC-12 Through 1985,"
             Table 3.
                 * * *  DRAFT FINAL  * * *

-------
                            3-21
                        EXHIBIT 3-10

Published Estimates  of U.S.S.R.  Production of CFC-11  and CFC-12
                     (millions of kilograms)
                    Year          Production
1968
1969
1970
1971
1972
1973
1974
1975
11.2
14.4
16.0
19.1
22.2
24.2
32.1
39.2
              Source:   Borisenkou, U.P., and Kazakou,
                       Y.E.,  1980, "Effect of Freons
                       and Halocarbons on the Ozone
                       Layer  of the Atmosphere and
                       Climate," Tr. (51. Geofiz
                       Obs. 1980 438, pp. 62-74, as
                       reported in CMA (1986).
                    *  * *  DRAFT FINAL  * *

-------
                                   3-22
relative to the growth rate of GNP is similar for the two periods.From 1968 to
1975 real GNP in the U.S.S.R. grew at annual rate of approximately 4.3 percent
(CIA 1983).  The ratio of the CFC rate of growth to the GNP rate of growth was
approximately 4.6 during this period.  From 1975 to 1983 real GNP grew at
about 2.5 percent per year (CIA 1983), for a ratio of about 4.1.  Nevertheless,
a recent oral report by Soviet attendees at a UNEP meeting indicates that the
current production capacity in the U.S.S.R. may be 60,000 metric tons for
CFC-11 and CFC-12.

    Production data for the U.S. have been collected by the International
Trade Commission (ITC) in its annual series Synthetic Organic Chemicals.   In
the early 1960s, production in the United States accounted for over 70 percent
of total CFC-11 and CFC-12 production reported to the CMA.  The U.S. share of
production dropped in the 1960s, and since 1975 it has dropped both as a share
of total production and in absolute terms.  By 1984, the U.S. share of
reported production had dropped to 34 percent.  Most of the absolute decline
in production in the U.S. can be attributed to the 1978 U.S. ban on
nonessential aerosol propellant uses of CFCs.  Data on U.S. production of
CFC-11 and CFC-12 are displayed in Exhibit 3-11.  Because of the ban on
nonessential aerosol propellant applications of CFC-11 and CFC-12 in the U.S.,
nearly all of the current production of these CFCs in the U.S. is used for
nonaerosol applications.

    The EEC publishes data on the production and sales of selected CFCs by EEC
producers.  Exhibit 3-12 displays these EEC data.  As- shown in the exhibit,
total EEC production of CFC-11 and CFC-12 has remained below the reported 1976
value.  Also of note is that a large fraction (approximately one-third) of all
CFC-11 and CFC-12 production in the EEC is exported.  Unlike the decline in
CFC-11 and CFC-12 production since 1976, the data in Exhibit 3-12 indicate a
strong increase in the production of CFC-113 and CFC-114 in the EEC.

    Less information is available on the historical production and use of other
CFCs.  Relative to CFC-11 and CFC-12, the quantities produced are small.
CFC-113 is an important solvent used in the manufacture of electronic
components.  Hammitt (1986) estimated the current world production of this CFC
at approximately 158 thousand metric tons (due to the lack of data, this
estimate is recognized as somewhat uncertain, however).  Of note is that the
production of this CFC is believed to be growing rapidly, in part in support
of the recent growth of the computer industry.

    CFC-223 is used in air conditioning applications (particularly in home
air conditioners) and in the production of fluoropolymers.  When CFC-22 is
used to make fluoropolymers, it is destroyed, and is consequently not emitted
to the atmosphere.  Gibbs (1986a) reports the production of CFC-22 in the U.S.
to be approximately 110 thousand metric tons in 1983, and that approximately
25 to 35 percent of the production is used in flouropolymer production.  Gibbs
    3 Of note is that CFC-22 has a much lower ozone-depleting potential than
CFC-11, CFC-12, and CFC-113.
                          * * *  DRAFT FINAL  * *

-------
                        3-23
                   EXHIBIT 3-11

Historical  Production  of CFC-11 and CFC-12 in the U.S.
                 (millions of kilograms)
            Year          CFC-11          CFC-12
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
32.8
41.3
56.6
63.6
67.4
77.3
77.3
82.7
92.7
108.2
110.9
117.0
135.9
151.4
154.7
122.3
116.2
96.4
87.9
75.8
71.7
73.8
63.7
73.1
84.0
73.8 p
75.5
78.7
94.3
98.6
103.4
123.1
129.9
140.5
147.7
166.8
170.3
176.7
199.2
221.7
221.1
178.3
178.3
162.5
148.4
133.3
133.8
147.6
117.0
135.4
154.1
. 127.9 p
            p = preliminary estimate.

            Source:   ITC  (1986), Synthetic
                     Organic Chemicals.
                * * *  DRAFT FINAL  * *

-------
                                   3-24
                              EXHIBIT 3-12

                       EEC  Production  and Sales Data
                           (millions of kilograms)
CFC-11 and CFC-12
Year Production a/ EEC
1976
1977
1978
1979
1980
1981 b/
1982
1983
1984
326
319
307
304
295
300
289
310
322
.4
.1
.0
.2
.7
.1
.0
.2
.2
244
233
231
219
216
209
206
216
217
Sales
.0
.0
.4
.6
.8
.7
.8
.4
.7
Exports
83
81
82
81
79
88
82
91
103
.6
.2
.2
.6
.4
.2
.0
.2
.4
CFC-113 and CFC-114
Production EEC Sales Export
23.5
23.9
N/A
N/A
N/A
N/A
N/A
N/A
53.6
17.6
19.3
N/A
N/A
N/A
N/A
N/A
N/A
38.6
5.
5.
N/A
N/A
N/A
N/A
N/A
N/A
12.
2
7






9
a/ Includes imports.  Production may not equal  EEC  sales  plus  exports.

b/ Data for 1981 to 1984 include Greece; previous years do  not.

N/A = Not available.

Source:  EEC (1985), "Chlorofluorocarbons in the Environment:  Updating the
         Situation," Communication from the Commission to the  Council,
         Annex I.
                          * * *  DRAFT FINAL  *

-------
                                   3-25
also reports production in the OECD countries in 1975 at approximately 75
thousand metric tons.  This is the most recent estimate for the OECD countries
reported.  No estimate is avaliable for the entire world at this time.

    Of note is that the trade literature recently reported that a major CFC
producer has plans to double the CFC-113 capacity and add 75 percent to its
CFC-22 capacity in Japan.*  These increases in capacity may indicate
expectations of continuing demand for these CFCs.

    Other specialty CFCs are produced in small quantities (e.g., CFC-114,
CFC-115).  Comprehensive information on these CFCs is not available at this
time.

    Projections of Future Demand for  CFCs

    A series of studies have examined the potential future demand for CFCs in
various parts of the world, 11 studies were submitted to the UNEP Workshop on
the control of chlorofluorocarbons (Rome, Italy, May 1986).  Most of these 11
studies analyzed CFC-11 and CFC-12 demand over the next 10 to 15 years.  Only
three studies examined the potential demand beyond 2000.  The general
conclusions from the studies can be summarized as follows:

        •   Future global production and emission of CFCs are
            expected to increase at an annual average rate of
            approximately 1.0 to 4.0 percent over the next 15 to 65
            years.

        •   Aerosol propellant applications are expected to remain
            constant or decrease in many portions of the world.

        •   Growth in nonaerosol applications is expected to be
            driven by uses for making foams and electronic equipment.

        •   Key uncertainties include technological changes over
            the long term and the patterns of use in developing
            countries.

This section first describes the methods that were used to project demand, and
then summarized the key findings of the studies.

    Methods Used for Projecting Future Demand

    Two complementary methods have been used to project the future demand-for
CFCs:  (1) a "bottom up" approach; and (2) a "top down" approach.  Although
the specifics of each author's method vary, the two approaches can be
characterized as follows:
      Chemical Marketing Reporter, March 10, 1986, p. 3.
                            * *  DRAFT FINAL

-------
                                   3-26
        •   Bottom Up.  The bottom up approach relies on a
            detailed specification of all the products that use CFCs
            or are made with CFCs (such as aerosols, foams,
            electronic equipment, and refrigerators).  Each of the
            products is examined to identify its expected future
            demand and the potential for changes in the intensity
            with which CFCs are used in the products (i.e., changes
            in the amount of CFCs required to produce the given
            product, such as foam).  The potential future demand for
            each of the products is driven by a variety of factors
            (such as population growth, economic growth, government
            regulation, and technological change).  The intensity of
            use of CFCs in the identified products may change (i.e.,
            increase or decrease) due to technological change or
            changes in the relative prices of substitutes.  The.
            projection of future CFC demand is compiled by adding
            the estimates of demand for CFCs in each product.
            Exhibit 3-13 shows graphically the bottom up approach.

        •   Top Down.  The top down approach does not rely on
            detailed specifications of the products that use CFCs.
            Instead, an aggregate relations-hip between CFC use and
            general descriptors of overall demand for goods and
            services is used.  These relationships vary, but are
            generally of the form where GFC use is related to
            overall economic activity (such as gross national
            product (GNP) or GNP per capita).  The relationships may
            be defined on a chemical-specific basis (e.g., CFC-11
            versus CFC-12), and may differentiate between different
            classes of uses (e.g., aerosol versus nonaerosol
            applications).  Historical data and statistical methods
            are generally used to specify these relationships.  A
            range of adjustments to the relationships may be made
            based on a range of assumptions about future changes in
            the underlying factors affecting the demand for CFCs.
            The projection of future CFC demand is compiled by
            applying the (adjusted) relationships to a projection
            (or set of projections) of the descriptors of overall
            economic activity (e.g., GNP).  Exhibit 3-14 compares
            predicted nonaerosol CFC production against actual
            production for one such equation.

    The strength of the bottom up approach is that it examines the details of
the demand for products that use CFCs.  This detailed examination may reveal
recent or impending changes in technology or demand patterns for these
products.  These changes or ongoing trends can be reflected in the projections
of CFC demand made using the bottom up approach (the top down approach will
not reflect impending changes and may not discern recent trends if they are
small).
                                 DRAFT FINAL  * * *

-------
                                               EXHIBIT  3-13

                                         The  Bottom  Up Appronch



Aerosol
Propellants



i
Furnitun
Be
^






Rigid
Foams


End Use
Products

i
i Carpet
Underlay
dding
Auton
Se


i i
Dash- Others
boards
Packaging
nobile
ats
o





CFC-1 1



Flexible
Foams


Substitute
Products
Spr



Pliable
Plastics
ngs Pa
Prod
ier
lucts





•

Refrigeration






Miscellaneous


Substitute
Technologies


i
Methylene CO2
Chloride (High Density)
<•
Technologii
Innovatior



cal
i

i i
Alternative Other
Blowing
Process
(Belgium)
Journals, Company Estimates, Market Studies, Conversations with Experts.Trade Associations
                                                                                                                    I
                                                                                                                   NJ
                                                                                                                   vj
Source:  "Overview Paper for Topic *2:  Projections of Future Demand,"

         UNEP Workshop, May 1986


                                         * * * DRAFT FINAL * * *

-------
         400-
         300-
Millions of
Kilograms
         200-
         100-
                                           EXHIIJIT 3-1U


                            Nonaerosol Application of CFC-11  and CFC-12
                                              (OECD)
                                                                                                           i
                                                                                                          to
                                                                                                          oo
            1962
1965
1968
1974
1977
1980
                                   -»- Actual            	Regression

  Source:   "Overview Paper for Topic  #2:   Projections of Future Demand,

           UNEP Workshop, May 1986
                                    * * * DRAFT FINAL * *  *

-------
                                   3-29
    The weaknesses of the bottom up approach can be characterized as follows:

        •   Data Intensity.  The bottom up approach requires
            considerable data on every product that uses CFCs.   The
            data may be difficult to obtain or verify for certain
            products or parts of the world, thereby limiting the
            applicability of this approach.

        •   Bias.   The bottom up approach tends to be limited to
            examining existing products that use CFCs.  The
            potential demand for new (as yet unknown) products
            generally is difficult to estimate.  By omitting
            potential new products (while including the potential
            displacement of existing products), the approach is
            biased toward underestimating future demand.
            Additionally, data limitations may limit the ability to
            project future demand for known products.  For example,
            the inability to allocate all of current production to
            existing uses shows the potential difficulty in
            implementing the bottom up approach.

        •   Time Frame.  The detailed data needed to implement
            the bottom up approach necessarily limit it to the  short
            run (10 to 15 years).  Detailed estimates of demand for
            the individual products that use CFCs are generally not
            feasible beyond this time frame.

    The major strengths of the top down approach are that it is much less data
intensive than the bottom up approach, and it can be used to produce
projections of demand beyond the short term.  The ability to produce long-term
projections is particularly important for the analysis of stratospheric
protection because potential ozone-modifying substances (such as CFCs) remain
in the stratosphere for very long times (up to 100 years or longer).
Consequently, choices regarding what level of stratospheric protection to
adopt necessarily require trade offs between short-term and long-term uses of
substances like CFCs.  Therefore, long-term projections are required.

    The top down approach reflects the potential for product innovation and
displacement, albeit in an implicit manner.  The historical data upon which
the method is based include new product introduction, product maturation, and
product displacement.  For example, during the 1960s and 1970s  the use of CFCs
for producing foams grew rapidly (product introduction).  During the 1970s,
the use of CFCs in certain refrigeration applications matured and leveled off
in the developed countries.  Aerosol propellant applications of CFCs have
matured or declined since the mid-1970s.  The top down approach implicitly
assumes that these types of product introduction, maturation, and displacement
(or some adjusted representation of these) will continue into the future.

    The top down approach also allows the uncertainty in the long-term role of
CFCs to be reflected.  Using short-term detailed analyses as a guide, the
long-term projections based on the top down approach may be adjusted to
                          * * *  DRAFT FINAL  * * *

-------
                                   3-30
reflect alternative assumptions regarding how the key factors affecting future
demand may unfold (such as technological change, changes in relative prices,
changes in consumer tastes, and changes in regulations).  Because these
factors are very uncertain over the long term, and may be very different from
current conditions, long-term projections based on the top down approach must
use a range of assumptions to identify the possible paths of future CFCs use
that may be expected.

    The weaknesses of the top down approach can be characterized as follows:

        •   Reliance on Other Projections.  Because the top down
            approach relates CFC demand to overall indicators of
            economic activity (such as GNP), the method must rely on
            projections of these overall indicators.  Although these
            aggregate indicators may be easier to project into the
            long term than more detailed factors, they will
            necessarily be somewhat uncertain, particularly in the
            out years.  Therefore, the CFC demand projections using
            this method can only be as good (and as precise) as the
            economic and population projections upon which they are
            based.  Exhibit 3-15 shows the range of population and
            economic projections used in these studies.

        •   Bias.   The top down approach will be biased if the
            period of historical data upon which the method is based
            is not representative of the future.  In the case of
            CFCs,  the recent historical data cover a period that
            includes both new product introduction (e.g., foam
            blowing), product displacement (reduction in the use of
            aerosol propellants, in part due to government
            regulation) and product improvements.  If the relative
            rates of introduction and displacement during this
            period are not expected to continue into the future,
            then the top down approach may be biased either upward
            or downward.

    The strengths and weaknesses of the two methods make them natural
complements.  The bottom up approach provides valuable information about
current and ongoing trends, which allows short-term projections to be made.
Although these projections may be biased downward, they are useful for
calibrating the long-term projections based on the more aggregate top down
approach.  Together the two approaches provide a rich set of information that
indicates the likely range of future CFC demand.

    Of note is that both methods are subject to uncertainty.  With three
exceptions, the papers reviewed did not explicity quantify the range of
uncertainty surrounding their projections.  Nevertheless, the uncertainty
exists, in part due to uncertainty regarding future economic and population
growth (even in the short term), and uncertainty in the future role that CFCs
will play in existing and potential new products.
                          * * *  DRAFT FINAL  * *

-------
                                                EXHIBIT 3-15

                                Range of  Population and GNP Per Capita Projections
         12
         10
          8
Population
(Billions)
                Population Projections
          0
          1985  1995  2005  2015  2025  2035   2045
                                                     Real
                                                     GNP
        9000


        8000


        7000


        6000
Capita
(Dollars/
person) 4000

        3000


        2000


        1000


           0
                     Economic Projections
                                                                                                              I
                                                                                                              to
           1985  1995  2005   2015   2025   2035   2045
        Source:   "Overview Paper for Topic 12:  Projections of  Future Demand,"

                 UNEP Workshop, May 1986
                                           * * * DRAFT  FINAL * *  *

-------
                                   3-32
    Summary of CFG Projections

    Exhibit 3-16 summarizes the methods used in the 11 papers reviewed that
contain projections of future demand of CFCs.  As shown, six authors used the
bottom up approach, four used the top down approach, and one author used a
mixed approach.  Also shown in the exhibit are the CFCs covered by each
author, the portion(s) of the world covered, and the time period analyzed
(short term or long term).

    Exhibit 3-17 summarizes the estimates of potential future demand for CFCs
from all the authors.  The exhibit is divided into short-run and long-run
estimates.  The short-run estimates are divided into aerosol applications of
CFC-11 and CFC-12, nonaerosol applications of CFC-11 and .CFC-12, nonaerosol
applications of CFC-22, and nonaerosol applications of CFC-113.  Within each
of these categories, projections of future demand for various portions of the
world are listed.

    The long-term estimates portion of the exhibit reports on all the
applications of CFC-11 and CFC-12 for the entire world.  The uncertainties in
these long-term estimates are listed.  Exhibit 3-18 shows the long term
projections made.

    The short-term projections listed in Exhibit 3-17 show distinct themes.
Aerosol applications of CFCs are expected to decline or remain constant in
several portions of the world (EEC, U.S., Sweden).  The growth in aerosol
applications in other portions of the world is generally expected to be slight
(Japan, 1.5 percent per year, and "Western" Countries, an average of less than
1.0 percent per year).  Gibbs (1986a) reports a large rate of increase for the
non-U.S. use of CFC-11 and CFC-12 in aerosols, but his aerosol estimates are
higher than the other authors' estimates are not considered representative of
the overall aerosol projections.

    These projections of aerosol applications also indicate an important
point.  Since the middle 1970s, the total use of CFCs in the Western Countries
category  (including the U.S.) has declined and remained fairly flat.  This
decline and subsequent flattening out of total use was driven by substantial
reductions in aerosol applications, which balanced off increases in nonaerosol
applications during this time.  The projections of flat or slight growth for
aerosol applications in the Western Countries indicate that:   (1) as aerosol
applications become a smaller share of total use, further aerosol application
reductions will be less effective in balancing off growth in nonaerosol
applications; and  (2) future large reductions in aerosol applications are not
anticipated under current regulations.  Together, these two points indicate
that the total use of CFC-11 and CFC-12 will increasingly be driven by
expected rates of growth in nonaerosol applications.

    The short-term nonaerosol applications of CFC-11 and CFC-12 show a trend
that is different than the trend for aerosol applications.  All estimates
indicate continued strong growth in nonaerosol applications during the short
term.  The estimates range from 2.1 percent per year in the EEC (Bevington
1986; EFCTC 1985) to 7.9 percent per year in Canada (Sheffield 1986, upper
                          * * *  DRAFT FINAL  * *

-------
                                                         EXHIBIT 3-16
                                            Summary of Demand Projection Estimates

COM tjojj N DS ._coyi R_EJD :
AEROSOL
API' 1. 1 CAT IONS:
Cf-C-11
crc-iz
NONALROSOL
APPI 1 CAT IONS:
ore- i i
Cf C-12
Cl C-22
crc-i 13
REGIONS COVERED :



PERIOD COVERED:

SHORT TERM
LONG TERM
METHOD USED:



DEVI NGTON



YES
YES


YES
YES
NO
YES


EEC



YES
NO


BOTTOM
UP
CAMM



YES
YES


YES
YES
NO
YES

a/
WORLD


e/
YES
YES

e/
MIX

EFCTC



YES
YES


YES
YES
NO
NO

b/
VARIOUS



YES
NO


BOTTOM
UP
GIBBS



YES
YES


YES
YES
YES
YES

C/
WORLD



YES
YES


TOP
DOWN
HAMMITT



YES
YES


YES
YES
NO
YES

a/
WORLD



YES
NO


BOTTOM
UP
OSTMAN



Yf'.S
YES


YES
YES
NO
YES


SWEDEN



YES
NO


BOTTOM
UP
KNOLLYS



YES
YES


NO
NO
NO
NO


EEC



YES
NO


BOTTOM
UP
KUROSAWA



YES
YES


YES
YES
NO
NO


JAPAN



YES
NO


BOTTOM
UP
NORDHAUS



YES
YES


YES
YES
NO
NO

d/
WORLD



YES
YES


TOP
DOWN
PERSKY



NO
NO


YES
YES
NO
NO


U.S.



YES
NO


TOP
DOWN
SHEEFIELD



NO
NO


YES
YES
YES
YES


CANADA



YES
NO


TOP
DOWN














W
i
tO







a/ Camm and Htiinmitt break their world estimates into tho following regions:  (1) U.S.; (2) Other Reporting Countries;  and  (3)
"Communist" Countries.
by Regions covered by EFCTC include:   (1)  Western Europo,  South Africa,  Australia,  and New Zealand; and (2) the  "Rest  of the
World," excluding North America (Canada,  U.S.,  Mexico),  Japan,  East Bloc countries, and Peoples Republic of China.
c/ Gibbs reports results separately for:   (1)  U.S.;  (2)  non-U.S.  OECD countries; and (3) the "Rest of the World."
d/ Nordhaus reports results separately for the  U.S.  and  the Rest  of the  World.
e/ Caturn's projections prior to 2000 are based  o'n the bottom up  approach  results reported in Hammitt.  The post-2000  estimates
are based on the top down approach results reported  in Quinn (1986).
Sources:  Bevington 1986; Camm ct al.  1986;  EFCTC 1985;  Gibbs 1986a;  Hammitt et a I . 1986;  Osttnan, Hedenstrom, and  Samuelsson,
          1986;  Knollys 1986;  Kurosawa and Imazeki  1986;  Nordhaus and Yohe 1986; Persky, Weigel, and V/hitfield  1985;
          Sheffield 1986 (see  references  for complete source).
                                                    * *  * DRAFT FINAL *  * *

-------
                                                        EXH!
                                               Summary of Demand Projections
                                         (Average annual rate of growth  in percent)
                       BEV ING I ON  CAMM
                                      a/
                EFCTC  GIBBS  HAMMITT
OSTMAN
                                  KNOLLYS  KUROSAWA  NORDHAUS   PERSKY  SHEFFIELD
                                       SHORT-TERM PROJECTIONS:  APPROXIMATELY 1985 TO 1993/2000
    SO I  App l_ i ca.t i on s
of CFC:'i 1 "
  ma;
  Japan
  Sweden
  U.S.
  "Western" Countries d/
  Wo r I d

M°_0 a 9. £°_ § o l__Ap p I i cat \o DJS
o f " C'l'c- ll" a rid' CFC- 12 :"
  Canada
  EEC c/
  Japan
  Sweden
  U.S.
  "Western" Countries d/
  Wo r I d
                     £/
                 -0.6   0.0     -3.9
                                                                               -0.6
2. 1
                       0.0
                 0.2   2.5
                       H.O
2. 1
                 2.5   't.3
               0.0
               0. 1
               1.7 e/
                                          0.0
                                                              1.5
                               2.2
                                                                                         U.U-7.9
                                          U.2
                                                              U.9
               2.5
               3.3
               3.3 e/
                               3.2       3. '*

                               3.7        '. .
Noriaerosol Applications
2JL_C_!iC_.-_22. :
Canada . .
Wo r 1 d
Nonaerosol Applications
of CFC-113:
Canada
EEC 7.5
Sweden . .
U.S.
"Western" Countries d/
Wo r 1 d


• • • * •
5.


. . . . .
.
. .
'. . .. 5.
5.
5.


. .. . • •• ••
1


« -
• •• •• •• ••
2.5
3 5.9
3 7.1
4 6.5
*

H. 2-5.1



14.5-8.9
• • • • • •
. . • • . .
• . • . • •
. . . * . .
• •
a/  Camm's short-term projections are based on Hammitt, and are therefore not  listed here separately.

b/  Values listed under long-term projections are for the 25th, mean, and 75th percent! le.

c/  EFCTC projection includes Western Europe, South Africa, Australia, and New Zealand.

d/  Coverage varies by author.  Approximate coverage includes non-U.S., non-East Bloc countries.   EFCTC  also excludes Japan,
Canada, Mexico.

e/  Assumes that one-half of East Bloc country use reported in Hammitt is in aerosol applications.

.. = Not reported.  Only "base case" or "middle" values are listed here for Camm, Gibbs, and Nordhaus.

                                                      * * * DRAFT FINAL * * *

-------
                                                     EXHIBIT  3-17  (Continued)

                                                  Summary of  Domand  Projnet ions
                                            (Average annu.'i I  r;ite of  growth  in  percent)
                       I3LVINHION  CAMM
                                         EFCTC  GIBBS  HAMMITT
                               OSTMAN
                             KNOLLYS   KUROSAWA  NORDHAIJS
                                                                         PERSKY   SHEFFIELD
A M _ A|jp I_LC;I I. i qns or
                "
      --...
  Lower Bound
  Low
  Med i UIH
  High
  Upper Bound
                                    LONG-TERM PROJECTIONS:  APPROXIMATELY  2000  TO 2050
0.5
1 .6
2.4
3.2
4.3
1.8
2.4
3. 1
3.'I
U.O
0.8
3.6
3.6
1 1 Apj> 1 i c.'i t ions of
KG- 12 Tor~Tfie~ Wo~r7d
Lower Bound
Low
Med i urn
High
Upper Bound


-0.4
1.6
2.4
3.2
4.4


1.6
2.0
..2.6
2.8
3.4


. . .* .. •*
1.0
3.6
3.9
• •


. . . .
. . . .
. • • •
. .
• •
                                                                                                                                 i
                                                                                                                                (-0
                                                                                                                                In
Sources:   Bevington 1986; Cnimn ot al. 1986; EFCTC 1985; Gibbs  1986a;
          Samuelsson,  1986; Knollys 1986; Kurosawa and  Imnzeki  1986;
          1985;  Sheffield 1986 (see references for complete  source).
                                   Hammitt  et  a 1.  1986;  Ostman,  Hedenstrom, and
                                   Nordhaus and  Yohe 1986;  Persky, Weigel, and Whitfield
                                                      * * * DRAFT  FINAL  *  *

-------
                                        EXHIIilT 3-18

                                   Long Term Projections
                           CFC-11 and CFC-12 -- World Production
                                        (2000-2050)
CAMM
GIBBS'
NORDHAUS
                             cFc-12 •mmmmmmm.
-1.0      0.0
                                              -i2;
                                                2.0
   Range = 5th lo 95ih petcentile
   Probabilities not reported; range reliocts five scenarios
  'Range = 2Slh to 75th percenule; nonaerosol apphcalory only
  1.0      2.0       3.0
Annual Rate of Change (%)
4.0
                                                                                                             UJ
                                                                                                             I
                                                                                                             to
5.0
6.0
 Source:  "Overview Paper  for Topic  »2:  Projections of Future Demand,1
          UNEP Workshop,  May 1906
                                  * * * DRAFT FINAL * * *

-------
                                   3-37
bound).  No estimates of declines or flat demand were reported.  These
estimates are consistent with a recent trade journal projection of growth of 4
percent to 5 percent per year through 1990.5

    It is interesting to note that the underlying patterns of demand reported
by the various authors show considerable similarity.  The bottom up estimates
all reported that the largest nonaerosol application growth is expected in the
production of foams, both flexible and rigid.  The important insulating
properties of CFCs used in the production of insulating rigid foams were often
cited as the driving factor behind this growth.  Refrigeration and air
conditioning applications were generally reported as being fairly mature, with
only small growth anticipated in the short term.  Due to changing workplace
practices and modifications to equipment, future use of CFCs in some air
conditioning applications was projected to decline (Hammitt et al. 1986, p.
56).  Overall, foam production is expected to account for the majority of
nonaerosol applications of CFC-11 and CFC-12 in the future.

    The estimates do not include detailed evaluations of developing countries
(e.g., Africa, portions of Asia).  The pattern of future use of CFCs in
developing countries may differ from the pattern in developed countries
because the refrigeration and air conditioning markets in these countries may
not be as mature.  Because the developing countries have fairly large
populations, the potential growth in the demand for CFCs for refrigeration and
air conditioning applications in these countries may be significant.

    The several estimates of future demand for CFC-22 and CFC-113 are also
shown in Exhibit 3-17.  All estimates show strong growth.  The demand for
CFC-113 is reportedly driven by expected strong growth in solvent
applications, particularly in solvents used for cleaning electronic
components.   Although these CFCs are expected to grow rapidly in the short
term, their overall level of use will remain well below the level of use of
CFC-11 and CFC-12 (less than half).

    Exhibit 3-17 also displays long-term demand projections by Camm et al.,
Gibbs, and Nordhaus and Yohe.  These projections are necessarily more
uncertain than the short-term estimates, and consequently the full range of
estimates provided by each author is listed.  As one check on the validity of
these long-term projections, one should note that in the short term (through
2000) they are consistent with the range of short-term estimates provided by
the other authors.

    As shown in Exhibit 3-17, the overall picture of long-term demand for CFCs
is one that may range from modest growth (on the order of 1.0 percent per year
or less), to sustained strong growth (nearly 4.0 percent per year).  The only
possibility that appears to have been considered unlikely by the authors is
substantial declines in the use of CFCs over the long term.  As the authors
caution, however, currently unforeseen technological advances could produce a
      Chemical Marketing Reporter, March 10, 1986, p.54.
                          * * *  DRAFT FINAL  * * *

-------
                                   3-38
situation where use declines (or increases) considerably, as unforeseen
applications develop.  Although the likelihood of either a considerable
decline or increase is difficult to quantify at this time, it is believed to
be very small.

    The specific methods used by the individual authors varied.  Bevington
(1986), EFCTC (1985), Kurosawa and Imazeki (1986), and Knollys (1986) all used
trade association and company estimates of expected market conditions over the
next 10 years as the basis for their short-term projections.  Detailed
estimates of demand for individual CFC applications were reportedly collected
and summarized.   Due to the confidentiality of the original estimates, the raw
data upon which the projections were based could not be reported.  In some
cases, judgment was reportedly required to reconcile estimates from different
sources, but the details of the process used, and the extent of the
reconciliation required, were not reported (see EFCTC 1985).

    Estimates for CFC use in Sweden (Ostman, Hedenstrom, and Sammuelsson,
1986) represent a synthesis of detailed data collected from CFC users in
Sweden.  This study is distinguished as the only regional study that
explicitly estimated the amount of CFCs contained in products that are
imported and exported.  CFCs are imported to Sweden in bulk as well as in
finished products (such as refrigerators).  Sweden does not produce CFCs.

    Not included in the Ostman study (or in any other study) are the regional
CFC demand implications of trade in products that do not retain their CFCs
after manufacture (e.g., flexible urethane foam).  The methods generally used
for estimating demand for CFCs assigns the demand to the country of
manufacture of the foam, not to the country that uses the foam.  The extent to
which the incorporation of this factor would modify the current perception of
regional CFC demand is not known at this time.  Flexible foam products (cars,
furniture) and electronic components (cleaned with CFC-113 solvent) are traded
extensively on the world market.

    Gibbs (1986a), Sheffield (1986), and Persky, Weigel, and Whitfield (1985)
used similar statistical methods to project future demand.  These statistical
methods are based, to various extents, on the strong historical correlation
between CFC production and economic activity in developed countries.  Exhibit
3-19 displays how the production per capita of CFC-11 and CFC-12 for
nonaerosol applications has been well correlated with GDP per capita from 1962
to 1980.  Persky, in a review of an early draft of the work by Camm et al.
(1986), used historical data for the U.S. to develop a relationship between
the U.S. per capita demand for nonaerosol applications of CFC-11 and CFC-12
and GNP per capita.  This relationship was applied to a scenario of GNP and
population for the U.S. through 2010.  The GNP and population scenario was not
described, but the result reported was similar to results reported by other
authors (e.g., Gibbs 1986a; Hammitt et al. 1986; and Nordhaus and Yohe 1986).

    Gibbs developed relationships between per capita CFC usage and per capita
GNP based on historical data, controlled for the price of CFCs, and included
estimates for OECD countries.  Ordinary least squares regression was used, and
serial correlation reportedly biased downward the estimates of the standard
deviations of the coefficients.
                          * * *  DRAFT FINAL  * * *

-------
                                                EXHIBIT  3-19

                            Nonaerosol  Production  Por Capita  of CKO11  and CFC-12 Has Been
                     Correlated With Gross Domestic Product (GDP) Per Capita in Developed Countirnii
                                               (196;> to  1980)
                                               • Un i led  St;i tus
                                               d Ul CU-Uni u.'d Status
                 0.8-
Production
      PGP      0.6H
  Capita

      (Kg)      °-4-i


                 0.2-)
                            0
                                                                                                       VO
                              0246
                                         GDP  PGP  Capita
                                   (Thousands  1975  US  $)
                                                                   8
Production per capita of Cf'C-11 and CFC-12 for nonaerosol applications has been correlated with GDP per capita  in the United
States and other Ot'CD countries.


Source:  CFC production, population, and GDP data obtained from:  Gibbs, Michael J., (1986), Scenarios of CFC Use:   1985 to 2075.
       IGF Incorporated,  prepared for the U.S. Environmental Protection Agency.
                                            * * * DRAFT FINAL * * *

-------
                                   3-40
    Gibbs applied these relationships to the OECD countries using a range of
population and GNP scenarios from published sources.  The ranges used were
quite large, resulting in wide ranges of potential CFC demand in the future.
Population varied by a factor of nearly two by 2075, and GNP per capita varied
by nearly a factor of five.  Gibbs1 population and GNP scenarios are
summarized in Exhibit 3-20.  In his highest and lowest scenarios, Gibbs also
adjusted the coefficients by one standard deviation to reflect potential
technological innovation.  The potential implications of innovation were not
explicitly evaluated.

    Gibbs estimated demand in the non-OECD countries by assuming that per
capita demand in these countries by the end of the projection period (2075)
would be the ratio of (1) the GNP per capita in the non-OECD countries to (2)
the current GNP per capita in the OECD countries times the current use per
capita in the. OECD countries.  Between 1985 and 2075 Gibbs fitted an
"S-shaped" curve that reflects the standard product lifecycle.

    This method of projecting non-OECD demand has a large influence on the
overall results produced by Gibbs because the non-OECD countries account for a
large fraction of the total growth in demand (nearly 50 percent).  The
potential bias of the method is ambiguous.  Differences between the OECD and
non-OECD countries (in terms of culture, climate, and other factors) could
produce different patterns of CFC use relative to OECD countries, even at
similar levels of per capita GNP.  Whether the levels would be higher or
lower, is unknown.

    There is some indication that Gibbs1 method may produce estimates of
demand in the non-OECD countries that are biased downward, because evidence
indicates that at the same level of income per capita, later developing
countries tend to have higher levels of consumption of industrial products
when compared to the countries that developed first.  The explanation given is
that product penetration into the market place generally proceeds faster in
later developing countries.  Whether this would be the case for CFCs in
non-OECD countries is unknown, although evidence indicates it may have been
true in the non-U.S. OECD countries as compared to the U.S.6

    Sheffield (1986) projected the potential future demand in Canada by
estimating a relationship between per capita CFC use and per capita GNP using
historical data.  Two functional forms were examined, linear and log-log.  The
log formulation produced rapid rates of growth during the entire projection
period.  Both forms fit the historical data.

    Although their methods differed considerably, Hammitt et al. (1986), Camm
et al. (1986), and Nordaus and Yohe (1986) each explicitly considered
uncertainty, and its effects on estimates of future demand for CFCs.  The work
by Hammitt et al. and Camm et al. was performed jointly, and the long-term
      Gibbs (1986b), p. 17.
                            * *  DRAFT FINAL  * * *

-------
                           3-41
                        EXHIBIT 3-20

Global Population and GNP  Scenarios  Used in Gibbs' Analysis


            GLOBAL  POPULATION SCENARIOS  (millions)
    Year
Lowest
Low
Medium
High
               GLOBAL GNP PER CAPITA  SCENARIOS
                        (1975 U.S.  $)
Highest
1985
2000
2025
2050
2075
4536
5377
6505
7324
7131
4745
5901
7384
7664
7944
4745
5901
7384
8223
8491
4835
6147
8160
9496
9960
5000
6500
9500
12100
13600
    Year
Lowest
Low
Medium
High
Highest
1985
2000
2025
2050
2075
1900
2206
2499
2831
3051
1900
2375
3201
4105
5264
1900
2447
3729
5683
8662
1900
2557
4195
6883
11292
1900
2752
5102
9458
17535
     Source:   Gibbs,  Michael  J.  (1986),  Scenarios of CFG
              Use:  1985  to 2075,  ICF  Incorporated, prepared
              for U.S. Environmental  Protection Agency,
              Washington,  D.C.
                  * * *  DRAFT FINAL *  » *

-------
                                   3-42
estimates by Camm et al. build on the Hammitt et al. results for the short-
term.  For short-term estimates, Hammitt et al. examined each of the
applications of CFCs in detail.  For each application, they identified
possible changes in the intensity of use expected as a consequence of market
penetration (of both CFC-using products and substitutes for CFC-using
products), market saturation, innovation, population growth, and growth in GNP
per capita.  Based on this detailed assessment for each application, the range
of potential growth for each CFC was summed across the applications.
Uncertainty was quantified by assuming that the factor that relates CFC growth
in each application to growth in GNP was normally distributed.  The
convolution of the uncertainties across all the applications was estimated
analytically using this assumption of normality, and assuming that the
uncertainties across applications were independent.

    The Hammitt et al. method is the only one reviewed that explicitly
considered the potential implications of technological innovation in a
detailed manner.  Additionally, the explicit quantification of uncertainty
provides guidance as to the likelihood of the different scenarios.  The
shortcomings of the method are that:

        •   it was unable to associate all current production with
            an application -- a significant portion of production
            remained unallocated across applications, implying that
            better information on specific uses could be obtained,
            and that the method may underestimate future demand; and

        •   the assumption of normally distributed uncertainty may
            be unrealistic, given evidence from the results by
            Nordhaus and Yohe that indicate that a skewed result may
            be possible.

    The work by Camm et al. builds on the Hammitt et al. results, and uses a
top down approach to specify rates of growth in demand relative to rates of
growth in GNP.  The approach used by Hammett et al. was also used by Camm et
al. to quantify uncertainty, and the intervals reportedly reflect a subjective
interval of probability of given limits.  The subjective nature of the method
of quantifying the uncertianty is an important characteristic of the method.

    The work by Camm et al. makes an important contribution to the development
of scenarios of future demand because it not only provides detailed scenarios
of the future use of each of the CFCs, but it is the only method that attempts
to reflect a convolution of uncertainties across chemical compounds.  Camm et
al. correctly assert that taking the fifth fractile estimate of demand for
each potential ozone-depleting substance at the same time does not necessarily
produce the fifth fractile estimate of ozone-depleting potential.  Because
ozone-depleting potential is the key policy variable of interest, Camm et al.
developed a method for identifying joint scenarios across a group of
compounds that convolute each compound's uncertainties in conjunction with
each compound's ozone-depletion potential.
                          * * *  DRAFT FINAL  * *

-------
                                   3-43
    The results of this approach for creating joint scenarios are scenarios of
chemical production that have narrower uncertainties than the independent
scenarios for each compound .(reported in Exhibit 3-17).  This result is
expected when the uncertainty across compounds is not perfectly correlated (a
reasonable assumption).  The degree of correlation among the uncertainties is
unknown, and the Camm et al. assumption of complete independence across the
chemicals (except for the common driving factor of economic growth) may be
extreme.  Nevertheless, the general result of their work is indicative that
the scenarios of production used for policy analysis should possibly be
narrower in uncertainty than the scenarios estimated for each compound
separately.

    The approach used by Nordhaus and Yohe (1986), called probabilistic
scenario analysis, explicitly considers uncertainty and is based on
statistical analyses of historical data.  This method was originally designed
by the authors to investigate the problem of forecasting carbon dioxide
emissions and concentrations from an economically consistent model of future
energy markets.  The method provides not only a best-guess estimate, but a
quantified range of uncertainty, built up from uncertainty regarding the
individual factors affecting demand.

    Using statistical methods and historical data, Nordhaus and Yohe specified
the relationship driving the intensity of use of CFC-11 and CFC-12 over time
in the U.S.   A family of logistic curves is fit to the data to describe the
manner in which the .use of CFCs approaches what is called a. "frontier" over
time.  The frontier represents the total possible use of CFCs at a given level
of wealth, and it changes over time.  A family of curves is used to represent
the uncertainty in the location of the frontier, which is not known with
precision.

    This method is able to capture the implicit effects of product
introduction, innovation, and maturity.  It reflects the uncertainty in our
ability to estimate (from historical data) the relative importance of these
factors.  Because data are not available for performing this analysis on the
rest of the world outside the U.S., the authors assume that the rest of the
world will develop in a pattern similar to the lead provided by the U.S., but
more slowly.

    The relationships developed using this method are used to project future
demand.  First, the basic building blocks of the relationships, the rate of
change of the frontier, the increase in labor productivity, and population
growth are projected.  The future values for labor productivity and population
are derived from probability distributions.  The distributions are based on
the range of published estimates, and studies that indicate that ranges of
published estimates often understate true uncertainty.  As recognized by the
authors, however, the distributions are considered to be subjective.

    Estimates of the future growth or decline of the frontier for CFC use have
not been published.  Nordhaus and Yohe write that a reasonable range for
industrially demanded goods is between a 2 percent per year average decline to
a 2 percent per year average increase.  A decline in the frontier implies that
                                 DRAFT FINAL  * * *

-------
                                   3-44
the amount of CFCs that can be used at a given level of economic activity
(e.g., measured as gross national product or GNP) is declining overtime.  In
other words, due to technological innovation, product displacement, or other
factors, the intensity with which CFCs would be used per dollar of GNP is
simulated to decrease.  An increase in the frontier implies the opposite, an
increase in intensity.  This range of a 2 percent decrease to a 2 percent
increase is used in the study.

    Given the relationships based on historical data and the probability
distributions for the future values for population, labor productivity, and
the CFC frontier, a Monte Carlo analysis is performed by drawing from the
distributions using pseudo-random numbers, and inserting the drawn values into
the relationships.  A number of independent trials are performed, the results
of which form a distribution.  This distribution of results reflects the
uncertainty in each of the steps of the method.

    The strengths of this approach include its statistical analysis of
historical data and its explicit consideration and modeling of uncertainty.
Of interest is that the results indicate that the potential for future demand
for CFC-11 and CFC-12 does not have a symmetrical range of uncertainty.  The
possibilities for values that are much larger than the median appear to exceed
the possibilities that the values will be far below the median.  Also, there
does not appear to be a significant likelihood that CFC-11 and CFC-12 demand
will decrease from current levels, even though a decline in the CFC frontier
is modeled as being equally likely as an increase.

    The shortcoming of the analysis is the lack of detail with which the use
of CFCs is modeled in countries outside the U.S.  Applying the U.S.-based
relationships may over- or understate demand in these countries; the bias is
ambiguous.

    Despite the diversity of the methods used by the various authors to
project possible future demand, the results presented above in Exhibit 3-14
indicate a considerable degree of similarity and consensus.  The overall
picture of long-term demand for CFCs is one that may range from modest growth
(on the order of 1.0 percent per year or less) to sustained strong growth
(nearly 4.0 percent per year).  The only possibility that appears to have been
considered unlikely by the authors is substantial declines in the use of CFCs
over the long term.

    Of note is that potential constraints on future production were
investigated in two studies, Gibbs and Weiner  (1986) and Mooz, Wolf, and Camm
(1986).  Both studies investigated the potential for the supply of fluorspar
to constrain future CFC production.  Fluorspar is the primary source of
fluorine used in the production of hydrofluoric acid (HF), which is used to
make CFCs.  It is considered a strategic mineral because it is also used in
the production of steel, and HF is also used to make aluminum.  Although not
mined extensively in the U.S., fluorspar is abundant in Mexico, South Africa,
China, Mongolia, and portions of Southeast Asia (Gibbs and Weiner 1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   3-45
    To assess whether fluorspar availability would likely constrain future CFC
production, Gibbs and Weiner and Mooz,  Wolf and Camm compared U.S.  Bureau of
Mines estimates of fluorospar resources to the potential requirements of
future production implied by the demand scenarios developed in Gibbs (1986a)
and Camm et al. (1986).   Both studies found that currently identified economic
reserves would be sufficient to supply expected fluorspar demand (from all
applications) for the next 20 years.   The high quality of the ore currently
being mined (e.g., in Mexico and Southeast Asia) and the fact that fluorspar
has historically been abundant indicate that it is likely that sufficient new
reserves will be available to meet demand far beyond the next 20 years.   Gibbs
and Weiner found that the rate at which new reserves would have to be added in
order to meet demand over the next 65 to 90 years was less than the rate of
addition to reserves over the last 20 years.  Additionally, Mooz, Wolf,  and
Camm point out that the abundance of fluorine in other forms (such as
phosphate rock) could provide a potentially large .source of fluorine.
However, it is expensive to recover fluorine from these sources.

    Mooz, Camm, and Wolf also examined possible constraints posed by the
availability of production capacity.   Although current production capacty may
be reached by approximately the year 2000, they found that in the absense of
government regulation, increases in capacity would likely take place.

    These two studies indicate, therefore, that currently there do not appear
to be supply constraints that would limit the ability to meet the future
demand for CFCs.

    Consensus Scenarios of Production and  Emissions

    The projections of the future demand for CFC-11 and CFC-12 discussed above
indicate that (1) demand for aerosol applications is likely to decline
slightly, remain flat, or grow slightly in the foreseeable future,  and (2)
demand for nonaerosol application is likely to increase.  Exhibit 3-21
displays the range of potential future demand implied by these studies.


                              EXHIBIT 3-21

            Range of Future Global Demand for CFC-11  and CFC-12
                        (Average Annual Growth Rates)
                                          Low      Middle     High
Short Term (1985-2000)
Long-Term (2000-2050)
Total Period (1985-2050)
1.5%
1.1%
1.2%
2.5%
2.5%
2.5%
4.1%
3.7%
3.8%
             Source:  UNEP (1986), Workshop on the Control of
                      Chlorofluorocarbons Rome, Italy, Annex II.
                          * * *  DRAFT FINAL  * *

-------
                                   3-46
    As shown in the exhibit, the overall growth over the next 65 years is
expected to range from 1.2 percent to 3.8 percent per year.  These values are
influenced heavily by the three long-term estimates by Camm et al. (1986),
Gibbs (1986a), and Nordhaus and Yohe (1986).  The short-term growth rates,
which are larger than the long-term rates, represent the range reported by all
the studies.

    These rates of growth, and the papers upon which they are based, were
presented and discussed at the UNEP Workshop on the Control of Chlorofluoro-
carbons held in Rome, Italy, May 1986.   Due to the difficulty in making
projections,  and the uncertainty in the data and methods used to develop the
projections,  the workshop concluded that a wider range of estimates of the
future demand for CFC-11 and CFC-12 was appropriate, ranging from no change
from 1985 levels (0.0 percent) to increases at a rate of 5.0 percent per year
from 1985 to 2000 (UNEP 1986, Annex II).  These values, 0.0 percent and 5.0
percent, can consequently be regarded as upper and lower bounds on the
expected future demand of CFC-11 and CFC-12 from 1985 to 2050.

    A variety of approaches were examined to developing consensus projections
(UNEP 1986).   As shown in Exhibit 3-22 they produce similar results.  To
develop the actual scenarios of production and emissions used in later
analyses, the following four steps were performed to apply the above growth
rates (UNEP 1986):

        1.  Estimate Global 1985 Production Level..  Estimates of
            global production of CFC-11 and CFC-12 by Hammitt et al.
            (1986), Gibbs (1986a), Nordhaus (1986) and Yohe, and CMA
            (1986)7 differ by over 15 percent.  The average values
            reported are adopted here for the middle scenario:  375
            million kilograms and 475 million kilograms for CFC-11
            and CFC-12, respectively.  The range of uncertainty in
            the 1985 value is reflected by using a range of plus and
            minus 8 percent of their middle values.

        2.  Estimate End Use Allocation. .Because emissions rates
            from different end uses vary, the end use allocation of
            production is required.  As described above, the CMA
            (1986) and Hammitt et al. (1986) end use allocations are
            very similar.

        3.  Estimate Emissions Rates.  The emissions rates for
            hermetic refrigeration, non-hermetic refrigeration,
            flexible foam, and rigid foam reported in Quinn et al.
            (1986) are used.  These rates may overstate emissions at
            disposal for refrigeration applications, but the bias is
            expected to be small (see Gamlen et al. 1986).  These
    7 CMA estimates reportedly represent 80 to 85 percent of global
production.  See UNEP (1986), Annex I.
                          * * *  DRAFT FINAL  *

-------
                                        EXHIBIT 3-22

                          Consensus Projections:  World CFC-11 and CFC-12*
                                        ( 19»!i-2050)
    Eliminate Extremes
               Average
             Composite
   Economic Growth3
                         0
    Nonaerosol applications only
                                  Low=1.2%  Medium=2.5% High=3.8%
   ./.,
                                                   '
1234
  Annual Rate of Change (%)
"Estimates based on eleven studies.
 Source:  "Overview Paper for Topic 12:  Projections of Future Demand,"

         UNEP Workshop, May 1986        * * * DRAFT FINAL * * *

-------
                                   3-48
            rates do not include estimates of fugitive emissions
            during production, which are expected to be on the order
            of 3.3 percent of CFC-12 production and 2.0 percent of
            CFC-11 production (Gamlen et al. 1986, p. 1080).
            Emissions from aerosol and miscellaneous applications
            are assumed to be prompt.

        4.  Divide Growth Rates into Aerosol and Nonaerosol
            Applications.   The projections of demand indicate that
            the outlooks for the future demand for aerosol and
            nonaerosol applications are very different.  The demand
            for aerosol applications is expected to be flat, while
            the demand for nonaerosol applications is expected to
            grow.  Exhibit 3-23 displays the assumptions used.  Five
            scenarios are shown, lowest to highest.  The total
            growth rates in the short and long term are the same
            values displayed in Exhibit 3-21.  The range of
            assumptions used for aerosol growth in the five
            scenarios are as follows:

            (1)  Lowest:  no growth;

            (2)  Low:  decline at 0.6 percent per year in the short
                 term, decline at 0.3 percent per year in the long
                 term;

            (3)  Middle:  no growth;

            (4)  High:  grow at 0.6 percent per year in the short
                 term, grow at 0.3 percent per year in the long
                 term; and

            (5)  Highest:   grow at 2.5 percent per year in the short
                 term, grow at 1.0 percent per year in the long term.

            The growth rates for the nonaerosol applications shown
            in Exhibit 3-23 are the rates needed to reach the total
            growth rates associated with the scenarios.  The aerosol
            and (implied) nonaerosol growth rates are representative
            of the range of estimates reported in the papers
            discussed in the previous section.

    Based on these four steps, five scenarios of global production and
emissions were estimated for CFC-11 and CFC-12.  These values are displayed in
Exhibit 3-24.  The range displayed across the five scenarios is large by
2050.  The production values are driven by the overall growth rates discussed
above.  The emissions estimates are also influenced by the allocation of
production to end use applications and by the release rates from those
applications.  Two factors could significantly alter the annual emissions
estimates:  (1) allocation of production to end uses with different emissions
rates; and (2) changes in emissions rates from current end uses (e.g., due to
recycling or to collection followed by destruction).


                          * * *  DRAFT FINAL  * * *

-------
                         3-49
                    EXHIBIT 3-23

Growth Rates for Global Aerosol and Nonaerosol Applications
                  of CFC-11 and CFC-12
                    (percent per year)
Scenario
Lowest:
Low:
Middle:
High:
Highest:
Aerosol
Nonaerosol
Total
Aerosol
Nonaerosol
Total
Aerosol
Nonaerosol
Total
Aerosol
Nonaerosol
Total
Aerosol
Nonaerosol
Total
Short Term
(1985-2000)
0.0
0.0
0.0
-0.6
2.3
1.5
0.0
3.5
2.5
0.6
5.4
4.1
2.5
6.0
5.0
Long Term
(2000-2050)
0.0
0.0
0.0
-0.3
1.4
1.1
0.0
2.9
2.5
0.3
4.1
3.7
1.0
5.5
5.0
Total Period
(1985-2050)
0.0
0.0
0.0
-0.4
1.6
1.2
0.0
3.0
2.5
0.4
4.4
3.8
1.3
5.6
5.0
                * * *  DRAFT  FINAL  * * *

-------
                                 3-50
                             EXHIBIT 3-24

               Scenarios  of Global Production and Emissions
                           CFC-11 and  CFC-12
                          (millions of kilograms)
                                     Scenario
         Lowest        Low          Middle          High          Highest
Year   PROD  EMIT   PROD  EMIT   PROD    EMIT   PROD    EMIT   PROD     EMIT
                               CFC-11

1985   340   272    358    285      375    298     393    311      410     323
2000   340   322    447    393      543    462     717    578      853     680
2025   340   340    589    555     1,008    897   1,782  1,513    2,885   2,352
2050   340   340    775    729     1,870  1,650   4,428  3,714    9,764   7,806
                               CFC-12
1985
2000
2025
2050
440
440
440
440
412
440
440
440
458
574
754
993
426
564
744
978
474
687
1,274
2,365
438
668
1,236
2,291
494
901
2,237
5,556
452
861
2,143
5,308

1
3
12
511
,061
,593
,170
465
1,008
3,398
11,462
                           * *  DRAFT FINAL  * * *

-------
                                   3-51
    The reasonableness of these potential growth rates in global production
can be assessed by comparing the range of future use to current use patterns.
Exhibit 3-25 displays the current (i.e., 1985) world, EEC, and U.S. use per
capita of CFC-11 and CFC-12 (see the shaded boxes).  As shown, the world use
per capita is approximately 0.2 kg, while the EEC and U.S. use per capita
exceeds 0.8 kg.

    The exhibit also shows estimates of the current GNP per capita for the
world, EEC, and the U.S.  The current world GNP per capita is under $2,000,
while the EEC and U.S. GNP per capita exceed $5,000 and $8,000 respectively
(1975 U.S. dollars).  The locations of the shaded boxes reflect the historical
correlation between CFC use and economic activity:  higher GNP per capita has
been correlated with higher use per capita.

    The implications of the scenarios of future CFC growth are displayed as
the shaded ovals in Exhibit 3-25.  Between 1985 and 2000, the CFC-11 and
CFC-12 use per capita will likely change very little; both CFC use and
population are expected to grow at approximately the same rate.  By 2050, CFC
use per capita may range from over three times the current global level, or
stay roughly the same as today.  Throughout this entire range, however,
projected use per capita remains below levels currently observed in the EEC
and the U.S., at comparable levels of economic wealth.  In other words, the
scenarios imply less use of CFCs than would be indicated by extrapolating
historical experience alone.

    Similar scenarios of production and emissions were also developed for
CFC-22 and CFC-113 which are shown in Exhibit 3-26.  Because only two
projections for CFC-22 were available, the range of growth rates for CFC-11
and CFC-12 was assumed to apply to the non-fluoropolymer applications of
CFC-22.8  These growth rates may be underestimates because the two
projections for CFC-22 both indicated more rapid growth than for CFC-11 and
CFC-12.  The use of CFC-22 is assumed to be associated with non-hermetically
sealed refrigeration and air conditioning applications, with release rates as
described in Quinn et al. (1986).  The estimate of 1985 production is taken
from Gibbs (1986a).

    The estimates of future demand for CFC-113 indicate that in the short term
(1985-2000) the demand for CFC-113 is expected to grow much more rapidly than
demand for CFC-11 and CFC-12 (see Bevington 1986; Hammitt et al. 1986; Gibbs
1986a; and Sheffield 1986).  Only Ostman, Hedenstrom, and Samuelsson,(1986)
reported a lower rate of growth for CFC-113.  Because this estimate is for
Sweden, it may not be representative of the overall growth expected globally.
Over the long term (2000 to 2050) the two estimates of CFC-113 demand that
have been prepared (Gibbs 1986a and Camm et al. 1986) indicate that CFC-113
demand will grow at a rate similar to the growth of CFC-11 and CFC-12.
    8 Because fluoropolymer applications of CFC-22 result in their
destruction, this use of CFC-22 is not included in the estimates presented
here.
                          * * *  DRAFT FINAL  * * *

-------
                                                   EXHIBIT  3-25
               1.0-
               0.8-
     Use per   0.6 •
     Capita
     (kg)

               0.4 H
               0.2-
                                   Current and Projected Future CFC-11 and CFC-12
                                        Use Per Capita  and  GIIP Per Capita
                                World
                                 1985
                                      World
                                       2000
World
 2050
                                                          i
                                                          Ln
                                                          N>
                          1,000     2,000    3,000    4,000    5,000     6,000   7,000    8,000
                                  9,000
                                                     GNP per Capita
                                                    (1975 U.S. Dollars)

Current  patterns of use pur capita  in  the world, EEC,  and  U.S.  reflect the historical  correlation between use per
capita and  GNP par capita:   higher  use per capita is associated with higher GNP per capita.  The scenarios of future
CFG use  (shaded ovals) imply ie_ss use  per capita at comparable  levels of GNP per capita  than would be  indicated by
current  use patterns  in the EEC and U.S.

Source:  "Summary: Overview Paper, Topic #2," presented at  UNEP Workshop on the Control of  ChIorofIuorocarbons, Rome,
        Italy, May 1986.
                                                    * * *  DRAFT  FINAL * * *

-------
                                3-53
                           EXHIBIT 3-26

             Scenarios of  Global Production and  Emissions
                          CFC-22  and CFC-113
                        (millions  of  kilograms)
                                    Scenario
         Lowest        Low         Middle         High         Highest
Year   PROD  EMIT   PROD  EMIT   PROD   EMIT   PROD   EMIT    PROD   EMIT
                              CFC-22

1985   102    79    106    80    111     81       115     82     120     84
2000   102   102    133   128    161    151   .    211     184     250    219
2025   102   102    175   170    298    279       525     455     846    741
2050   102   102    231   224    552    515     1,304   1,111   2,862  2,506


                             CFC-113

1985   150   127    156   133    163    138       170     144     177    150
2000   150   127    218   185    283    240       416     354     523    444
2025   150   127    287   244    525    446     1,032     877   1,770  1,505
2050   150   127    378   322    974    828     2,558   2,174   5,994  5,094
                       * * *  DRAFT FINAL  *  *  *

-------
                                   3-54
    To reflect this information regarding the expected growth rates of
CFC-113, the short-term growth rate is assumed to be 1.5 times the rates used
for CFC-11 and CFC-12, and the long-term rates are equal to the CFC-11 and
CFC-12 rates.  Based on Quinn et al. (1986), the emissions of CFC-113 are 85
percent of annual production; the remaining 15 percent is destroyed or
otherwise never emitted.  The 1985 production of CFC-113 was estimated by
Hammett et al. (1986).

    The resulting scenarios of production and emissions for CFC-22 and CFC-113
are displayed in Exhibit 3-26.  Again, the ranges displayed across the five
scenarios are large.  Because there are fewer projections for these CFCs,
there is more uncertainty regarding their potential future demand.  The method
of using the projected rates for CFC-11 and CFC-12 may under- or overstate the
future demand for these CFCs.  Two factors, which may not be adequately
reflected in these projects, may influence the future demand of these CFCs:
(1) new solvent applications for CFC-113 in certain personal computer
applications and (2) controls on the use and disposal of substitute
chlorinated solvents in the U.S.

    Scenarios of future production and emissions of other CFCs including
CFC-13, CFC-14, CFC-21, CFC-23, CFC-114, CFC-115, CFC-142b, and CFC-152a are
not presented here because these chemicals are only produced in limited
quantities and their use is not expected to increase.

    Representatives from DuPont, Allied, and Imperial Chemical Industries have
reported that they' had investigated chemical substitutes beginning a decade
ago.  CFC-123 and CFC-134a were identified as attractive candidates for
replacing CFC-11 and CFC-12.  CFC-123 has a hydrogen atom and would probably
have a short atmospheric lifetime.  CFC-134a has no chlorine and thus no
depletion potential.  CFC-123 and CFC-134b are not now commercially produced.
At the current time it is believed that producers will not produce these
chemicals unless environmental controls lead to increases in the prices of
existing CFCs, unless there is an improvement in the technology of producing
these chemicals.

    Between 1975 and 1980 DuPont spent 15 million dollars searching for
alternatives for CFCs.  DuPont required that the alternative meet technical
and economic criteria.  The technical criteria included low heat of
vaporization, nonflammability, low chemical reactivity, and low toxicity.
These properties result in low energy consumption, very few material
compatibility problems, and safety-in-use.  The economic criteria included
that a commercial process for its manufacture be available, or be developed,
that an economic incentive for manufacture exist, and that the compound be
competitive with existing CFCs and value-in-use.  According to a recent policy
statement by DuPont, these substitutes probably could be available in 5 years,
assuming no unforeseen problems and that existing obstacles can be removed
(DuPont, 1986).

    In May, 1986 industry representatives stated that the costs of
alternatives to CFCs would be 3 to 10 times the current price of CFCs and
therefore production was not feasible (UNEP 1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   3-55
    In June, 1986 Dr. S. Robert Orfeo (Orfeo, 1986) speaking for Allied
Chemical Corporation, reported that they too had investigated alternative
chemicals beginning in 1975.  He said that the new CFC substitutes would cost
a minimum of five times more and therefore were not profitable to produce.

CHLOROCARBONS

    Two chlorocarbons have been identified as potentially important ozone
depleters:

        •   carbon tetrachloride (CC14); and
        •   methyl chloroform (CH3CC13).


    Carbon tetrachloride is an excellent solvent in many applications, and was
once used extensively as a solvent and grain fumigant in the U.S.  Because of
its toxicity, it is only used in small amounts in such applications today
(Hammitt et al.  1986).  The major use of carbon tetrachloride in the U.S. is
in the manufacture of CFC-11 and CFC-12.  During the manufacturing process,
almost all of the carbon tetrachloride is consumed or destroyed; very little
is emitted.

    Hammitt et al. (1986) estimate current U.S. production of carbon
tetrachloride at 280 thousand metric tons, and world production at 870,000
metric tons (this estimate does not include production in East Bloc
countries).  Hammitt et al. and Camm et al. project the future demand for
carbon tetrachloride to remain closely tied to the demand for CFC-11 and
CFC-12.  Consequently, the future rates of growth of this compound are
expected to be approximately the same as the rates of growth in the demand for
these CFCs.  Of note is that the uses of carbon tetrachloride as a solvent and
grain fumigant may persist in other countries.  The projections of demand for
those compounds are therefore uncertain.

    Because the major use of carbon tetrachloride is in the production of
CFCs, scenarios of future demand for this compound must be related to the
scenarios for CFCs described in the previous section.  Using stochiometric
factors and estimates of conversion losses reported in Quinn et al. (1986,
p. 66), the carbon tetrachloride needed to produce the amounts of CFC-11 and
CFC-12 described in the five scenarios above were estimated.  The amount of
carbon tetrachloride used in other applications (grain fumigation in the U.S.,
which is being phased out, and miscellaneous pharmaceutical applications) was
added to these estimates to get total production.  The emissions associated
with this level of production'were estimated based on rates in Quinn et al.
(1986), which indicate that annual emissions are 6.44 percent of annual pro-
duction, plus grain fumigation use (Quinn et al. 1986, p. 70).  These estimates
of carbon tetrachloride production and emissions are reported in Exhibit 3-27.

    Methyl chloroform is produced in large quantities around the world, and is
used as a cleaning solvent in vapor degreasing and cold cleaning applications.
Small amounts are also used in adhesives, aerosols, and coatings (Hammitt et
                                 DRAFT FINAL  * *

-------
                                 3-56
                             EXHIBIT 3-27

               Scenarios  of Global Production and Emissions
              of  Carbon Tetrachloride and Methyl Chloroform
                          (millions of kilograms)
Scenario
Lowest Low Middle
Year PROD EMIT PROD EMIT PROD EMIT
High Highest
PROD EMIT PROD EMIT
                          Carbon Tetrachloride
1985
2000
2025
2050
1,021
1,021
1,021
1,021
65
65
65
65
1,066
1,333
1,757
2,315
68
85
112
148
1,112
1,610
2,985
5,534
71
103
191
354
1
2
5
13
,157
,114
,256
,069
74
135
336
836
1,203
2,500
8,466
28,670
77
160
541
1,831
                            Methyl  Chloroform
1985
2000
2025
2050
500  425
500  425
500 . 425
500  425
522  443
546  464
603  512
666  566
  545    463
  848    720
1,533  1,303
2,771  2,356
  567    482
1,021    867
2,242  1,906
4,928  4,189
   589     501
 1,335   1,134
 3,913   3,326
11,480   9,759
                         * *  *   DRAFT FINAL  * * *

-------
                                   3-57
al. 1986).  Hammitt et al. estimate 1985 U.S. production at 270 thousand metric
tons and world production at 545 thousand metric tons (Hammitt et al. 1986,
p. 80).

    Hammitt et al. (1986) and Camm et al. (1986) estimate a range of future
growth in the demand for methyl chloroform.   These growth rates reflect alter-
native assumptions regarding the influence on use and recycling of the expected
land disposal ban on chlorinated solvents (including methyl chloroform).  These
annual rates range from 0.4 percent to 4.7 percent for the period 1985 to 2040.
Using this range, and adding 0.0 percent (i.e., no growth) as a lower bound,
five scenarios of future global production were developed.  These estimates of
future production are shown in Exhibit 3-27 along with emissions estimates
based on estimates by Quinn et al. (1986) that annual emissions equal 85 per-
cent of annual production (the remaining 15 percent is destroyed or otherwise
not released).  Due to the lack of data on the production and use of methyl
chloroform outside the United States, these estimates of production and
emissions are recognized as uncertain.

HALONS

    Halogenated extinguishing agents are made from hydrocarbons in which one
or more hydrogen atoms have been replaced by halogen atoms.  The common
halogen elements are fluorine, chlorine, bromine, and iodine.  These
extinguishing agents include:
            Chemical Name
      Methyl bromide
      Methyl Iodide
      Bromochloromethane
      Dibromodifluoromethane
      Bromochlorodifluoromethane
      Bromotrifluoromethane
      Carbon tetrachloride
      Dibromotetrafluoroethane
 Formula

CH3Br
CH31
BrCH2Cl
Br2CF2
BrCClF2
BrCF3
CC14
BrF2CCBrF2
Halon No.

   1001
  10001
   1011
   1202
   1211
   1301
    104
   2402
     Use
Fixed

Hand
Fixed a/
Fixed/Hand b/
Fixed b/, d/
Hand
Hand c/
      &l  Declining use in the United States Air Force applications.
      b/  Recognized by NFPA Standards.
      c/  Limited use in Italy and Russia, no U.S. use.
      d/  Halon-1301 is also mixed with Halon-1211 in small hand-held
          extinguishers.
      Halon-1301 and Halon-1211 have excellent flame extinguishment properties
and acceptable toxicity under favorable circumstances.  However, there are
significant health risks if the fire is not quickly extinguished.  Halons
undergo pyrolysis when subject to flame or surface temperatures greater than
480°C.  The extent of pyrolysis depends on length of exposure to high
temperatures, the Halon concentration, and the degree of mixing.  Halon-1301
                          * * *  DRAFT FINAL  * * *

-------
                                   3-58
decomposition products are halogen acids, hydrogen fluoride (HF),  hydrogen
bromide (HBr), the free halogens (Br2), and small amounts of carbonyl halides
(COF2), COBr2).  Halon-1211 has the same decomposition products as Halon-1301
plus hydrogen chloride (HC1), chlorine gas (C12), and phosgene (COC12) (NFPA
1986).

      Halon-1301 and Halon-1211 are the only halogenated extinguishing agents
recognized by the National Fire Protection Association (NFPA) for use in the
United States.  They have a combination of desirable characteristics:

        •   low toxicity in occupied spaces (in most
            circumstances);

        •   low electrical conductivity;

        •   high visibility during use;

        •   little corrosive or abrasive residue; and

        •   high effectiveness per pound of chemical.

    In stationary or fixed applications the gaseous Halon-1301 is widely used
while the liquid Halon-1211 is preferred in portable applications, where
liquid range is desirable.

    These chemicals were not used extensively in fire protection until the
1970s.  Although these compounds are currently produced in small quantities,
they are of major concern to ozone protection because use is growing rapidly
and because they are believed to be as much as ten times more potent per pound
in depleting stratospheric ozone.

    Halon-1301 use has increased rapidly for applications like computer
centers and art collections because:   (1) it can be discharged in an occupied
space without significant risk to occupants; (2) it does not damage chemically
sensitive contents of the space; and (3) it is very effective against many
types of fires.  These advantages reduce the financial consequences of
accidental extinguisher discharge and  allow faster extinguisher deployment
since evacuation need not proceed discharge.  In contrast, extinguishing
agents like carbon dioxide (C02) can reduce oxygen so quickly that the
occupants of the area may be endangered.  The absence of residue not only
protects property against damage, but  allows quicker return to work spaces.

    The future demand for Halon-1211 and Halon-1301 will depend on the need
for fire extinguishers with their unique properties.  The choice depends on
property value, fire and fire fighting consequence, probability of fire, the
consumer awareness, and product marketing.

    The expected continued growth in the use of computing and other expensive
electronic equipment indicates that the use of the Halon-1301 may be expected
to grow rapidly.  The backlog of property with substantial value such as art,
                          * * *  DRAFT FINAL  * * *

-------
                                   3-59
antiques, and business records offers large new markets for Halon use.
Military uses, including routine preemptive releases in situations of high
fire danger, could drastically increase use.

Studies that Project  Halon  Emissions

    Analysis by the RAND Corporation experts (Hamraitt et al., 1986; Camm et
al., 1986; and Quinn et al., 1986) present preliminary estimates of the U.S.
and world production of these two compounds (estimates do not include
production in East Bloc countries).  Current U.S. production of Halon-1301 is
estimated at 5.4 million kilograms with historic growth rates of 15 to 30
percent per year and world production is estimated at 10.8 million kilograms
(Hammitt et al. 1986).

    RAND built its projections on limited information and simplifying
assumptions:

        •   U.S. Halon-1301 production was from industry sources.

        •   World Halon-1301 production was thought to be about
            twice U.S. production (unknown growth rate).

        •   Halon-1211 production was thought to be about the same
            as Halon-1301 production.

        •   Future growth rates were based primarily on
            electronics expansion, not all property that could be
            effectively protected by Halon systems.

        •   Emissions from Halon-1301 systems are based on:

                19% for initial system testing;
                1%/year for filling and servicing;
            --  0.1%/year for leaks;
                1%/year for false discharge;
                1%/year fire discharge; and
            --  Recycle of Halon-1301 at 40 year building disposal.

        •   Emissions from Halon-1211 portable extinguishers are
            assumed to be the same as the emissions from fixed
            Halon-1301 systems.

    Exhibit 3-28 provides production and emissions scenarios based on the
projections RAND developed.  These scenarios are used in the analysis of risks
presented in Chapter 18.  Although the RAND effort is a contribution to our
understanding of the potential significance of Halons to stratospheric ozone
depletion, there are good reasons to believe that the estimates presented
below understate the emissions, perhaps by substantial amounts.
                          * * *  DRAFT FINAL  *

-------
                   3-60
               EXHIBIT 3-28

Scenarios  of Global Production and Emissions of
          Halon-1301  and Halon-1211
            (millions of kilograms)
Scenario
Lowest
Year
PROD
EMIT
Low
PROD
EMIT
Middle
PROD
EMIT
High
PROD
EMIT
Highest
PROD
EMIT
HALON 1301
1985
2000
2025
2050
9.9
9.9
7.5
5.7
2.4
5.6
7.8
6.0
10.4
15.0
17.9
21.3
2.5
7.6
14.5
17.6
10.8
19.7
32.3
53.0
HALON 121
1985
2000
2025
2050
9.9
9.9
8.3
7.0
2.4
5.6
9.0
7.7
10.4
15.0
19.7
25.9
2.5
7.6
16.2
21.8
10.8
19.7
36.5
67.8
2.6
9.3
22.7
37.9
1
2.6
9.3
25.4
48.1
11
26
58
128

11
26
65
163
.2
.6
.4
.2

.2
.5
.8
.2
2.7
11.8
36.6
81.7

2.7
11.8
40.7
102.9
11.7
39.2
160.3
656.2

11.7
39.2
149.4
570.0
2.8
16.1
83.3
344.0

2.8
16.1
80.1
310.2
                 DRAFT FINAL  * * *

-------
                                   3-61
    For Halon-1301 systems RAND assumes a 1 percent per year accidental
discharge and a 1 percent per year fire discharge.  No basis is given for
these estimates in the RAND Study.  If the actual accidental release was as
much as ten times higher and the probability of a fire in a Halon protected
space ten times lower (one in a thousand or less), the actual total emissions
would increase by more than three hundred percent over those estimated by
RAND.  In addition, RAND assumed that the volume of Halon protected space
would decrease as a result of computer downsizing.  RAND made no adjustment
for the possibility that other nations will rapidly computerize, thus
accelerating the world rate of growth in Halon systems.

    Other factors not included in these estimates could lead to even higher
releases.  For example, the National Fire Protection Association (NFPA)
Halon-1301 technical committee has proposed that all new systems be tested
with a full-scale release of Halon-1301.  This would increase emissions of
Halon-1301 by over four hundred percent.

    RAND assumed that Halon-1211 use for portable fire extinguishers is about
the same as use of Halon-1301 in fixed systems and that it will grow at a
similar rate.  The product uses are quite different.  Halon-1211 is used
primarily in hand-held fire extinguishers and U.S. Air Force rapid
intervention crash trucks (Hammitt et al. 1986, p. 68).  The sizes of these
devices reportedly range from 14 ounces to 150 pounds  (Quinn et al. 1986, p.
86).  Because the devices are small, it is unlikely that significant
quantities will be recovered from disposed or deactivated systems.  Portable
Halon-1211 units are not subject to the same testing by discharge.  However,
Halon-1211 units are now being marketed as relaviely low-cost consumer
products.  Their superior performance in particular circumstances means that
they may face rapid growth, particularly if they are specified for commercial
buildings and other business applications or if they are aggressively
marketed.  Furthermore, it is unlikely that the average useful life of
hand-held extinguishers will be as long as 40 years as assumed by RAND.
                          JL. .1. .J.
                                 DRAFT FINAL  *

-------
                                   3-62
REFERENCES

    Alliance for Responsible CFC Policy (1980),  "An Economic Portrait of the
CFC-Utilizing Industries in the United States,"  Washington,  D.C.

    Bevington, C.F.P. (1986), Projections of Production Capacity,  Production
and Use of CFCs in the Context of EEC Regulations,  Metra Consulting Group
Ltd., London, U.K., prepared for the European Economic Community.

    Camm, Frank et al. (1986), Joint Emission Scenarios for  Potential Ozone
Depleting Substances, working draft by The RAND  Corporation, prepared for the
U.S. Environmental Protection Agency, Washington,  D.C.

    Camm, F., J. Hammitt, and K. Wolf (1986), Reaction to DRI Critique of
Unpublished RAND Working Draft: "Projected Use,  Emissions, and Banks of
Potential Ozone-Depleting Substances," The RAND  Corporation, Santa Monica,
California.

    CMA (1986), Chemical Manufacturers Association, "Production,  Sales, and
Calculated Release of CFC-11 and CFC-12 Through  1984," Washington, D.C.

    Chemical Marketing Reporter, "Chemical Profile," March 10, 1986, Vol.
229, pp. 3 and 54.

    CIA (1983), Handbook of Economic Statistics, 1983, Directorate of
Intelligence, Central Intelligence Agency, Washington, D.C.

    Connell, P.S. (1986), "A Parameterized Numerical Fit to  Total Column Ozone
Changes Calculated by the LLNL 1-D Model of the  Troposphere  and Stratosphere,"
Lawrence Livermore National Laboratory, Livermore,  California.

    Department of Environment, Ministry of Science, Technology and
Environment, Malaysia (Malaysia) (1986), Country Report, Chlorofluorocarbon
Chemicals in Malaysia.

    DuPont (1986), Corporate Policy Statement.

    EEC (1985), European Economic Community, "Chlorofluorocarbons in the
Environment:  Updating the Situation," Communication from the Commission to
the Council.

    Erickson, Lynn (1986), Memo to Neal Patel, EPA, Radian Corporation,
Houston, Texas.

    European Fluorocarbon Technical Committee (EFCTC) (1985), Halocarbon
Trend Study 1983 - 1995, EFCTC (a CEFIC Sector Group).

    Gamlen, P.H., et al. (1986), "The Production and Release to the Atmosphere
of CC13F and CC12F2  (Chlorofluorocarbons CFC-11  and CFC-12), Atmospheric
Environment, pp. 1077-1085.
                                 DRAFT FINAL  * * *

-------
                                   3-63
    Gibbs, Michael J. (1986a), Scenarios of CFG Use:  1985 to 2075, ICF
Incorporated, prepared for the U.S. Environmental Protection Agency,
Washington, B.C.

    Gibbs, Michael J. (1986b), "Summary of Historical Chlorofluorocarbon
Production," Presented at UNEP Economic Workshop on the Control of Chloro-
fluorocarbons, Rome, Italy, May 1986.

    Gibbs, Michael J., and Robin Wiener (1986), Assessment of the
Availability of Fluorine for the Production of Chlorofluorocarbons, ICF
Incorporated, prepared for the U.S. Environmental Protection Agency,
Washington, B.C.

    Hammitt, James K., et al. (1986), Product Uses and Market Trends for
Potential Ozone Depleting Substances:  1985 - 2000, The RAND Corporation,
prepared for the U.S. Environmental Protection Agency, Washington, D.C.

    Henderson, L.G.  (1980), "The Atmosphere Chemistry of 1,1,1-2-tetra-
fluoroethane," General Motors Research Laboratory, Warren, Michigan, GMR-3450.

    Hoffman, B.L. and D.S. Klander (1978), Final EIS Fluorocarbons:  Environ-
mental and Health Implications, U.S.  Federal Drug Administration, Washington,
D.C.

    ITC (1986), U.S. International Trade Commission, Synthetic Organic
Chemicals, ITC, Washington, D.C.

    Khalil, M.A. and R.A. Rasmussen (1986), "The Release of Trichlorofluoro-
methane from Rigid Polyurethane Foams," JAPCA, pp. 159-163.

    Knollys, R.C. (1986), Fluorocarbon Use in Aerosols -- A Trend Study
1984-1995 in Member Countries of the European Economic Community, prepared on
behalf of The Federation of European Aerosol Associations.

    Kurosawa, Kimio, and Katsuo Imazeki (1986), Projections of the Production
Use and Trade of CFCs in Japan in the Next 5-10 Years, Japan Fluoride Gas
Association and the Japan Aerosol Association.

    McCarthy, R.L., et al. (1977), "The Fluorocarbon -- Ozone Theory -- I.
Production and Release:   World Production and Release of CC13F and CC12F2
(Fluorocarbons 11 and 12) through 1975, Atmospheric Environment, pp. 491-497.

    Mooz, W.E., K.A. Wolf, and F. Camm (1986), Potential Constraints on
Cumulative Global Production of Chlorofluorocarbons, The RAND Corporation,
prepared for the U.S. Environmental Protection Agency, Washington, D.C.

    National Fire Protection Association (1986), National Fire Protection
Handbook, 14th Edition,  Boston, MA.
                                 DRAFT FINAL  * *

-------
                                   3-64
    Nordhaus, William D., and Gary W. Yohe (1986), Probabilistic Projections
of Chlorofluorocarbon Consumption:  Stage One, Yale University and Wesleyan
University, prepared for the U.S. Environmental Protection Agency, Washington,
B.C.

    Orfeo, S. Robert (1986), "Response to questions to the panel from Senator
John H. Chafee," Ozone Depletion, the Greenhouse Effect,  and Climate Change,
Hearings before the Senate Subcommittee on Environmental  Pollution of the
Committee on Environment and Public Works, Washington, DC:  US G.P.O., pp.
189-192.

    Ostman, Anders, Olle Hedenstrom, and Sture Samuelsson (1986), Projections
of CFC Use in Sweden, prepared for Statens naturvardsverk, Sweden.

    Persky, Susan, C. Miles Weigel, and Ronald M. Whitfield (1985), A Review
of the Study:  "Projected Use, Emissions, and Banks of Potential Ozone
Depleting Substances" (Rand Report No. 2483-3-EPA), prepared by Data
Resources, Inc., Washington, D.C., at the request of E.I. duPont de Nemours.

    Quinn, Timothy H., et al. (1985), Projected Use, Emissions, and Banks of
Potential Ozone Depleting Substances, draft working paper by The RAND
Corporation, prepared for the U.S. Environmental Protection Agency,
Washington, D.C.

    Quinn, Timothy H., et al. (1986), Projected Use, Emissions, and Banks of
Potential Ozone-Depleting Substances, The RAND Corporation, prepared for the
U.S. Environmental Protection Agency, Washington, D.C.

    Sheffield, A. (1986), Canadian Overview of CFC Demand Projections to the
Year 2005, Commercial Chemicals Branch, Environmental Protection Service,
Environment Canada, Ottowas, Canada.

    Strobach, Donald (1986), "A Search for Alternatives to the Current
Commercial Chlorofluorocarbons," in Protecting the Ozone  Layer:  Workshop on
Demand and Control Technologies.  Washington, DC:  U.S. Environmental
Protection Agency.

    UNEP (1986), Report of the First Part of the Workshop on the Control of
Chlorofluorocarbons, UNEP/WG.148/2, Nairobi, Kenya.

    Weigel, C.M., and R.M. Whitfield (1986), Reply to the RAND Corporation's
Response to DRl's Review of RAND's Working Draft, "Projected Use, Emissions,
and Banks of Potential Ozone-Depleting Substances," Data  Resources, Inc.,
Washington, D.C.

    WHO Criteria Document on Chlorofluorocarbons (1986),  Draft Report,
Environmental Criteria and Assessment Office, U.S. EPA, Cincinnati, Ohio, p.
2-10.
                          * * *  DRAFT FINAL  *

-------
                                   3-65
    Yarrow, G.K. (1986), The Reliability of Very Long Run Forecasts of
Chlorofluorocarbon Production and Emissions, Hertford College, Oxford.

    Zhijia, W. (1986), "Country Paper for Topic I:   UNEP Workshop on the
Protection of the Ozone Layer," National Environmental Protection Agency of
the People's Republic of China.
                            * *  DRAFT FINAL  *

-------
Chapter 4

-------
                                CHAPTER 4

          FUTURE EMISSIONS AND CONCENTRATIONS OF TRACE GASES
                    WITH  PARTLY  BIOGENIC SOURCES
SUMMARY

    Assessment of the risks of stratospheric modification must be based on
realistic projections of atmospheric  change.   Because the stratosphere can be
perturbed by many gases simultaneously,  models that attempt to simulate such
changes require projections of future concentrations of all relevant trace
gases including those that are wholly industrially produced and those that are
partly biogenic.

    This chapter summarizes evidence, gathered by a variety of researchers, on
past emissions and possible future  changes  in  the concentrations of three
gases:  methane (CH4), carbon dioxide (C02), and nitrous oxide (N20) -- all
potentially important stratospheric perturbants.  Future concentrations of
these gases are difficult to project. Estimates must be made of the growth in
emission sources, which depend on future economic, political, and physical
forces.  Both current and future emission factors must be estimated for each
source.  The complex biogeochemical cycles  that control the fate of emissions
once they are released into the atmosphere  must be considered.  Possible
changes in these cycles must be projected.

    Despite these uncertainties, scenarios  have generally been used by
atmospheric modelers.  While recommending the  use of these standard scenarios
for a risk assessment, this chapter also suggests additional sensitivity
scenarios.

    For carbon dioxide, the 50th percentile scenario prepared for the National
Academy of Sciences is recommended.

    For methane, concentrations are more uncertain.  The standard scenario of
the atmospheric modeling community, one  percent growth per year based on
recent historic changes, is suggested, and  three sensitivity scenarios which
cover a wide range of possibilities are  also proposed.

    For nitrous oxide, the standard scenario of continued growth in
concentrations of 0.25% per year is proposed.

    While these scenarios are consistent with  current knowledge and reasonable
estimates of future trends in the absence of regulation, they do not reflect
the considerable uncertainty introduced  by the possibility that governments
may attempt to limit "greenhouse warming" by controlling the concentrations of
these gases.  In assessing the risks  of  stratospheric modification, some
assumption must be made about the growth of greenhouse gases.  The assumption
that has been implicitly made by extrapolating past rates of growth of these
trace gases indefinitely into the future is that decision makers will never
decide to limit global warming.  In Chapter 18, the implications of
alternatives for this assumption on the  assessment of risks is analyzed.


                          * * *  DRAFT FINAL   * * *

-------
                                   4-2
FINDINGS

    1.   FUTURE CONCENTRATIONS OF STRATOSPHERIC PERTURBANTS THAT HAVE AT LEAST
        SOME BIOGENIC SOURCES ARE DIFFICULT TO PROJECT.

        la.   The size of existing source terms (wetland areas,  for example) is
             not known with certainty today for all these trace species.   The
             growth of source terras (e.g.,  acreage of rice paddies, wetlands
             area), which will be determined by many political  and social
             factors, must be estimated.

        Ib.   Current emission factors for each source term must be estimated;
             many are not known today.

        Ic.   Possible changes in emission factors due to changes in the
             environment must be projected.  This is difficult  because the
             underlying physical processes  are not well understood and because
             changes in the environment that could alter emissions are not
             easy to project.

        Id.   Biogeochemical cycles that control the fate of emissions once
             released into the atmosphere must be understood;  there are severe
             limitations to our current understanding of these.

        le.   Possible changes in these biogeochemical cycles due to changes in
             the environment must be projected; this again will be very
             difficult.

    2.   DESPITE THE UNCERTAINTIES ASSOCIATED WITH EACH OF THESE FACTORS,
        RESEARCHERS HAVE DEVELOPED SCENARIOS FOR THREE GASES WHICH IS USED IN
        THE  STANDARD ANALYSIS IN THIS RISK ASSESSMENT.

        2a.   For carbon dioxide, a single scenario developed by the National
             Academy of Sciences (its 50th percentile), was considered
             sufficient for use in this risk assessment.

        2b.   For methane, future changes in concentrations are  more uncertain,
             and several sensitivity scenarios are suggested for use in this
             analysis in addition to the standard case of 1% per year growth.

        2c.   For nitrous oxide, future concentrations are assumed to continue
             to increase at present rates of growth, approximately 0.25% per
             year.

        2d.   Additional research is needed, however, to validate these
             conclusions.
                            * *  DRAFT FINAL  * *

-------
                               4-3
3.   TO THE EXTENT FUTURE DECISIONMAKERS BECOME CONCERNED ABOUT GLOBAL
    WARMING. GOVERNMENTS MAY TAKE ACTION TO LIMIT THE RISE IN
    CONCENTRATIONS OF CARBON DIOXIDE, METHANE, AND NITROUS OXIDE.

    3a.  The standard assumption that has been made by default is that
         greenhouse gases will be allowed to increase without limit.

    3b.  In order to provide decision makers with adequate information to
         assess the risks of ozone modification due to rising CFCs and
         Halons, alternative assumptions about the future of greenhouse
         gases need to be examined.
                             DRAFT FINAL  * * *

-------
                                   4-4
FACTORS  INFLUENCING FUTURE TRACE  GAS CONCENTRATIONS

    To project future tropospheric concentrations  of  trace  gases  the  following
estimates must be made:  1)  the growth of source  terms (e.g.,  rice paddies  for
CH4 or cars for carbon monoxide [CO]);  2)  the current emission  factors  for
each type of source;  3)  possible changes over time in emission  factors; 4) the
effects of physical and chemical transformation  and deposition  that influence
the fate of emissions; and  5)  the ways  these latter processes may change over
time.   There are some uncertainties associated with each of these five  factors.

    For example, one  source of CH4 is rice paddies.   Consequently, future  CH4
concentrations will depend, in part,  on the addition  of  new lands to  rice
acreage (a change in  the source term);  oh the choice  of  cultivars for rice
acreage, (a change in the emission factor, as cultivars  vary in the amount of
CH4 they emit), and on the  cropping practices chosen  for each area and
cultivar, which may alter the  biomass available  for anaerobic processes to
generate CH4 (these factors could influence emission  rates).  Consequently,  to
project the quantities of rice acreage in the future, forecasts must  be made
of changes in population, biotechnology, climate,  income, development
decisions, and human  tastes.   Additional factors make projecting  CH4  emissions
from rice paddies more complex:  emission rates  could change.  For example,
most of the CH4 that  reaches the atmosphere from rice paddies appears to be
transported through plants  (Cicerone and Shetter 1981).   Atmospheric
concentrations of C02 are rising worldwide, and  it is known that  higher C02
levels alter stomatal resistance in plants and reduce gas exchange  (Acock  and
Allen, Jr. 1985).  This means  that C02 could possibly alter the gas exchange
through stomates, thereby reducing CH4 emissions rates.   Ij: stomatal  exchange
is a major transport  mechanism for the escape of CH4  emissions  from plants
(and this may not be  the case), the rise in concentrations  of C02 may lower
emission rates from rice paddies over time, thereby decreasing  the flux of CH4
from this source and  other  biosystems.   Similarly, a  global warming could
possibly increase emission  rates from sources now  already important or  from
sources, such as methane hydrates (Revelle 1983),  that do not now contribute
CH4 to the atmosphere.  Thus,  changes in the environment could  alter  CH4
emission rates.

    Even this does not describe the full complexity of estimating future CH4
concentrations.  The  possibility exists that nations  may undertake policies
intended to limit emissions of CH4 or other gases  in  an  attempt to  limit
global warming or other types  of environmental change (Mintzer  and Miller
1986;  Hoffman, Wells, and Titus 1985).   For example,  nations may  begin  to
encourage rice cropping practices that diminish  CH4 emissions,  thereby
reducing atmospheric  concentrations.   With all these  factors to consider for a
variety of gases, the process  of estimating future concentrations will  not be
easy.   Similar complexities surround the problems  of  estimating future  growth
of C02 and N20 concentrations.

HOW TRACE GASES  INFLUENCE THE STRATOSPHERE AND  TROPOSHERE

    Rising concentrations of CH4, C02,  and N20 may influence a  number of
natural atmospheric and environmental processes  (see  Exhibit 4-1).  CH4 has
                              *  DRAFT FINAL  *

-------
                                                          EXHIBIT 4-1


                                        EFFECTS Ol: CHANGES IN COMPOSITION OF ATMOSPHERE
RISING
LEVELS
EFFECT ON PLANTS
  EFFECT ON
SURFACE CLIMATE
      EFFECT ON
TROPOSPHERIC CHEMISTRY
             EFFECT ON
STRATOSPHERIC COMPOSITION AND STRUCTURE
C02
Cll'l
Changes physiology;
increases photo-
synthesis; changes
water relations
(Stra i n and Cure
1985)

     None
N20
                 None
CO
                 None
Greenhouse gas
(Ramanathan et a I
1985)
Greenhouse gas
(Ramanathan et
1985)
                                                  a I
                        Greenhouse  gas
                        (Ramanathan et
                        1985)
                        Ind i rectly
                        increases C02  and
                        03,  two  green-
                        house  gases  (World
                        Meteorolog ica I
                        Organization,  1986),
No d i rect effect
Creates ozone (NAS 198'l);
alters Oil abundance
(KhnI iI and Rasmussen
1985
-------
                                   4-6
two important effects on the stratosphere:  (1) it decreases column ozone
destruction in a way that also enhances ground warming due to column ozone at
lower altitudes; (2) it also adds water vapor to the stratosphere, which
amplifies global warming and climate change.  In the troposphere, CH4 will
increase tropospheric ozone, resulting in a further enhancement of the
greenhouse effect and possibly an increase in oxidants that affect surface
processes (World Meteorological Organization 1986).

    Carbon monoxide (CO) is a pollutant that can cause health problems.  Its
abundance in part controls and is, in part, controlled by the abundance of CH4
(Thompson and Cicerone 1986; Levine, Rinsland, and Tennille 1985).  By
altering the abundance of OH, increases of CO will indirectly increase CH4;
decreases in CO will have the opposite effect (Levine, Rinsland and Tennille
1985).

    N20 is emitted by natural soil processes, fertilizer applications, and
combustion (McElroy and Wofsy 1984).  Its tropospheric concentrations are
increasing (Weiss 1981).  Increasing concentrations of tropospheric N20 are a
major source of global warming (Ramanathan et al. 1985).  By itself, an
increase in stratospheric N20 would also decrease stratospheric ozone
abundance.  However, in tandem with increasing chlorine, N20 interferes with
chlorine's catalytic cycle and causes catalytic reductions of ozone (03).  The
net effect depends on the exact scenario (Stolarski, personal communication).

    .Increases in C02 emissions have been produced primarily from combustion,
with some biogenic contributions  (Rotty and Marland 1984).  C02 concentrations
have three major effects:

        •   C02 is an important greenhouse gas (Ramanathan et al.
            1985);

        •   C02 affects the physiological development of plants,
            enhancing photosynthesis, decreasing gas exchange and
            transpiration (water  loss), and altering the
            morphological and biochemical development of plants
            (Strain and Cure 1985).

        •   C02, by cooling the stratosphere, increases ozone
            abundance there (World Meteorological Organization 1986).

    The fact that C02, CH4, and N20 are all gases that contribute to global
warming is important.  To the extent that global warming becomes a concern,
future emissions of these gases may depend on whether governments undertake
policies to directly or indirectly alter these atmospheric concentrations.
Nations may, for example, undertake policies to limit carbon monoxide
emissions in order to lower CH4 and reduce global warming.  Similarly, nations
may control emissions in order to prevent the buildup of oxidants in the
troposphere.  Future policies may play an important role in determining the
emissions and concentrations of these gases.  As Mintzer and Miller (1986)
note, if "restraints were imposed on the buildup of C02, CH4 and N20 in order
                          * * *  DRAFT FINAL  * * *

-------
                                   4-7
to control their contribution to global  warming,  the  role of these gases as
moderators of potential ozone depletion  in  high CFC emission scenarios could
be severely limited."

THE LIFETIME OF EMISSIONS AND THE PREDICTABILITY OF
FUTURE CONCENTRATIONS

    Of the three gases discussed in this chapter, N20 and C02 have long
lifetimes (World Meteorological  Organization 1986), which guarantes that once
their concentrations increase,  they will not change dramatically in a short
time.  CH4 has a much shorter lifetime,  approximately 10 years  (World
Meteorological Organization 1986).   Therefore, the concentrations of CH4 could
respond fairly quickly to a sudden  loss  in  production.  This means that
estimates of future CH4 concentrations require examination of the likely
stability of forces emitting CH4.   For concentrations of CH4 to continue to
rise, emission fluxes have to keep  growing  (or CH4's  lifetime must continue to
increase).  To the extent that future fluxes may  stop growing or fall,
concentrations could stabilize or even fall relatively rapidly.    To the
extent that the flux of CH4 begins  to increase sharply, concentrations could
rise quickly.  Because neither of these  possibilities can be rejected, the
short lifetime of CH4 adds a major  uncertainty to projecting CH4
concentrations, making the projections inherently less certain  than for the
other gases.

SCENARIOS  OF TRACE GASES

    The remainder of this chapter reviews scenarios that researchers have
developed to describe possible trends in concentrations of CH4, C02, and N20.
Efforts to measure past and current concentrations were discussed in Chapter
2; here we examine the current understanding of source terms and fluxes for
each gas or its precursors and how  these might change over time.  For this
review, the scenarios generally in  use in the atmospheric community have been
adopted, and sensitivity scenarios  for CH4  have been  added.  These scenarios
can be used in analyzing future policy decisions.

Carbon Dioxide (CO2)

    Measured Increases in C02 Concentrations

    Chapter 2 summarizes historical changes in the concentrations of C02:
relatively constant concentrations  until the onset of the Industrial
Revolution, followed by continuous  increases (Pearman et al. 1986; and Keeling
and GMCC/NOAA 1985).

    Historical C02 Emissions

    Before 1850, the natural losses of C02  roughly balanced natural emissions,
leading to a seasonal cycle still visible each year.   After 1850 and the onset
of the Industrial Revolution, the heavy  use of coal and other fossil fuels
(and possibly massive deforestation) led to a large increase in C02
emissions.  Rotty and Marland (1984) have analyzed historical records of
                            * *  DRAFT FINAL  * *

-------
                                   4-8
fossil fuel use and reconstructed the emissions profile shown in Exhibit 4-2.
While fossil fuels are the major source of emissions, such activities as
forest clearing and deforestation may have redistributed C02 from biomass
sinks to the atmosphere.  However, estimates of the amount of C02 emitted to
the atmoshpere as a result of deforestation have declined substantially in
recent years (Houghton et al. 1983; Brown and Lugo 1981).

    The Carbon Cycle:  Biogeochemical Factors

    Emissions from fossil fuels are only one part of the biogeochemical
processes known as the carbon cycle.  Other components of the carbon cycle are
the uptake of carbon by the terrestial biosphere and the uptake, absorption,
and outgassing of C02 in the oceans (Trabalka 1985).  Exhibit 4-3 provides a
schematic of these components.  Over time the fluxes between compartments may
change.  For example, the ocean's absorption of C02 may change as it becomes
more saturated with C02 and as ocean circulation patterns are modified.
Photosynthesis may increase, leading plants to store additional carbon
standing biomass, thereby reducing the percentage of C02 emissions that
remains in the atmosphere (Gates et al. 1983).  Respiration may increase with
warming, altering C02 emissions from previously inactive storage compartments
(Woodwell et al. 1983).  Understanding the carbon cycle presents an enormous
scientific challenge to which much effort has been devoted.  The Department of
Energy has recently issued a state-of-the-art report Atmospheric Carbon
Dioxide and the Global Carbon Cycle, which interested readers should consult
for more information (Trabalka 1985).

    Projections of Future C02 Emissions

    Future concentrations of C02 will depend on C02 emissions, the future
operation of the carbon cycle, and its impact on the fraction airborne.
Consequently, predictions of future C02 levels are subject to considerable
uncertainty.  Various researchers have used long-term energy models to
represent the socioeconomic factors that contribute to energy use and
therefore C02 emissions.  While models differ, in some fashion each model
connects world economic production and population growth with energy use and
economic efficiency. . By making different assumptions about future population
and economic growth, the modelers generate projections of future energy use
and fuel mix.  They then apply average carbon content ratios to compute the
associated C02 emissions (Edmonds and Reilly 1985a).

    Perhaps the most widely accepted set of projections was developed for the
National Academy of Sciences by Nordhaus and Yohe (1983).  In their most
likely scenario, C02 emissions rise from five gigatons per year in 1975 to
nearly 20 by the year 2100.  Other researchers with published projections of
C02 emissions include Seidel and Keyes (1983) and Edmonds, et al. (1984).
Exhibit 4-4 shows the low, medium, and high projections of each group.

    Projections of Future C02 Concentrations

    As C02 emissions increase, some researchers believe that the airborne
fraction will also increase as the upper ocean (now a major sink) b.ecomes
                          * * *  DRAFT FINAL  * * *

-------
                                  4-9
                             EXHIBIT 4-2

     Historical  Carbon Dioxide Emissions from Fossil Fuels and  Cement
     r
                        1 900
                               1920
                                      I 940
                                             1 960
                                                    19B0
Carbon dioxide emissions  have  risen  rapidly since the outset of the Industrial
Revolution.

Source: Rotty and Marland,  1984.
                            * *   DRAFT FINAL  * * *

-------
                                      4-10



                                 EXHIBIT 4-3

                        A Schematic of the Carbon  Cycle
                                       Atmosphere
711 (335 ppm of C02)
t
t

•
                                     56
56
                    2-3
90
            12,000
            (7,500
           ultimately
          recoverable)
     1,760
90
                                                                      580
                                 38.400
                                            Surface
                                 Intermediate
                                 and deep
        Fossil fuels
        and shales
Terrestrial
biosphere
                  Oceans
        Reservoirs in 10 metric tons

        Fluxes in 109 metric tons/year
The stocks of carbon  in each natural  reservoir and  the annual  flows  that take
place  among reservoirs  are shown.  Taken together,  these stocks  and  flows
constitute the carbon cycle.  Carbon  distributed through this  cycle  plays an
important role in the ecological balance of the atmosphere, biosphere, and
oceans.

Source:  Council on Environmental Quality, 1981.
                            * * *  DRAFT FINAL  * *

-------
                                   4-11
                               EXHIBIT 4-4

   Projected  Carbon Dioxide Emissions and Doubling Time  of Concentrations
                Car-ban Dioxide Pi-o j «e t i or»» :
                Em i ss i oi-ts and Doub I < n
-------
                                   •4-12
saturated (Emmanuel, Kilbugh, and Olson 1981; Baes, Jr., Bjorkstrom, and
Mulholland 1985).  Thus, both rising C02 emissions and an increasing airborne
fraction should contribute to future increases in C02 concentrations.  A
widely adopted measure of concentration projections is the "doubling time,"
the year in which C02 concentrations reach twice the pre-industrial level.
While the range in emissions projections looks wide, the range in doubling
time is smaller.  Exhibit 4-4 shows that mid-range projections estimate that a
doubling of C02 could occur as early as 2050 or as late as 2065.

    Standard Scenario Proposed for Assessing Risks and Later Policy Testing

    For the purposes of assessing risks of ozone depletion due to increases in
CFCs, Halons, and other chemicals, a standard scenario for C02 is proposed:
the NAS 50th Percentile (Nordhaus and Yohe 1983) which is generally consistent
with that used in most atmospheric modeling  (Ramanathan et al. 1985; Wuebbles,
MacCracken, and Luther 1983; and World Meteorological Organization 1985).  The
NAS 50th Percentile Scenario projects the following C02 concentrations from
1975 to 2100:

                          1975               (340 ppm)
                          2000               (366 ppm)
                          2025               (422 ppm)
                          2050               (508 ppm)
                          20.75               (625 ppm)
                          2100               (770 ppm).


This scenario assumes that no efforts are made in this period to limit
greenhouse warming by limiting C02.

Methane  (CH4)

    Measured Increases in CH4 Concentrations

    Chapter 2 summarizes historical changes  in CH4 concentrations: relatively
constant levels until approximately 150 to 200 years ago, followed by
increases that now average approximately 1% per year (Pearman et al. 1986; and
Khalil and Rasmussen 1986).

    CH4 Photochemistry and Biogeochemical Cycling

    Changes in CH4 concentrations depend on  changes in CH4 fluxes and the many
physical and chemical processes that control the fate of CH4 emissions.
Anaerobic processes that produce CH4 require biomass and are regulated by
temperature and water (Cicerone and Shetter  1981).  Once generated, CH4 must
be transported to the atmosphere.  The success of that transport depends on
many environmental and atmospheric conditions.

    CH4 emitted to the atmosphere (1) combines with the OH radical to form C02
and water vapor, (2) is reabsorbed by soils  and water, and (3) is transported
to the stratosphere where it is oxidized to  add water vapor and other
compounds in that relatively dry region (Thompson and Cicerone 1986).


                          * * *  DRAFT FINAL  * * *

-------
                                   4-13
    Once in the atmosphere, the key factor controlling the fate of CH4 is its
reaction with OH (Khalil and Rasmussen 1985b).  OH concentrations in turn are
determined by CH4, CO, and possibly non-methane hydrocarbons on a localized
scale.  As such, OH provides a close coupling between CO and CH4:    CO
concentrations must be considered when interpreting the behavior of CH4 in the
atmosphere.  One possible cause for a temporal increase in CH4 concentrations
is a declining supply of OH brought about by increasing CO (Khalil and
Rasmussen 1985b; and Levine, Rinsland, and Tennille 1985).

    An example of the interaction between CO and CH4 appears in Exhibit 4-5.
Note that increases in CO emissions, by decreasing the OH abundance, increase
the CH4 lifetime and concentrations; decreasing CO will tend to have the
opposite effect.  Unfortunately, there are two factors that complicate
analysis of actual CO and CH4 trends and predictions of future changes in
these trace gases.  The first .is that atmospheric CO trends are harder to
discern than are CH4 trends, although there is accumulating evidence that at
least in some locations CO is increasing (see below) (Rinsland and Levine
1985; and Khalil and Rasmussen 1983).  The second is that measurements of OH
are nearly nonexistent and both present day and historical OH can only be
deduced from model calculations.  Calculations show that there are numerous
factors besides concentrations of CH4 and CO that determine OH concentrations,
and hence methane lifetimes (Thompson and Cicerone, 1986).  These include 03,
water vapor distributions, the amount of sunlight and concentrations of
soluble OH end-product gases that .can be rained out (e.g., HN03, .H202).  All
of these factors vary widely in time and with locality and season.  Reactive
non-methane hydrocarbons (presumably both those emitted by human activities
and those occurring naturally from vegetation) may control OH more effectively
than CO and CH4 in areas where their concentrations are high.  Ozone and NOx
(NOx = NO + N02) levels are critical to OH concentrations since 03 is the
precursor species for OH.  For example, at low NOx concentrations, increasing
CO and CH4 supresses OH almost uniformly throughout the troposphere, but at
high NOx (as would be found in a continental or urban environment), increasing
CH4 or CO contributes to photochemical ozone formation and increases OH near
the ground and in the upper troposhere (Hameed, Pinto, and Steward 1979;
Thompson and Cicerone 1985).

    Consequently, to estimate future CH4 concentrations requires estimating
not only future CH4 fluxes, but future fluxes of many other gases, as well as
a variety of environmental conditions.

    Current and Historical CH4 Emissions

    The extent of scientific agreement on the probable sources and quantities
of CH4 emissions (Exhibit 4-6) is quite limited (World Meteorological
Organization 1986).  In fact, the discrepancies shown in the exhibit probably
underestimate uncertainty -- not reflecting, for example, questions emerging
about the potential magnitude of tundra as a source of emissions.  Our
understanding of past changes is limited.  Expanded human activities, such as
rice cultivation, agriculture, raising cattle, and the use of oil and natural
gas, may have increased fluxes from these sources (World Meteorological
Organization 1986).  Decreases in wetlands may have reduced fluxes from this
                          * * *  DRAFT FINAL  * * *

-------
                                   4-14
                              EXHIBIT 4-5
          Two Ways That CH4 Concentrations Could Have Changed
              1.5        2.0
           CO FLUX (Normalized)
2.5
   1.5       2.0
CH4 FLUX (Normalized)
As a consequence of coupling between CO and CH4, due to their reactions with
OH, concentrations of CH4 can increase both through increases in CO fluxes (a)
and CH4 fluxes (b).
Source: Thompson and Cicerone, 1985.
                          * * *  DRAFT FINAL  * * *

-------
                       4-15
                   EXHIBIT 4-6



           Estimated  CH4 Emission Sources

                  12
               (10   grams per year)

Enteric fermentation
(livestock)
Rice paddies
Wetlands
Biomass burning
Freshwater lakes
Oceans
Tundra
Anthropogenic/
fossil fuel
Other
TOTAL
a/
COLUMN I
100-200
~280
190-300
1-25
1-17
0.3-3
16-50
--
586-825
b/
COLUMN II
100-150
100 + 50
150 + 50
10-60
--
--
— -
10-150
390-765
  Column I is taken from  Ehhalt  (1975) and Ehhalt and

    Schmidt (1978).




  Column II is taken from Khalil and Rasmussen (1983).




Source:   World Meteorological Organization (1986).
                 * *   DRAFT FINAL  * * *

-------
                                   4-16
source.  However, because the size of change in each source is uncertain
(e.g., rice acreage added), and the flux associated with different conditions
for each source is unclear (for example, emissions from different types of
rice cultivation vary), we cannot directly estimate past changes in CH4 fluxes
through time.

    Future Emissions of Methane

    If CH4 emissions are to continue to grow, increases in source terms and/or
emission rates would be required.  Empirical work on this subject is limited,
although certain educated guesses may be possible (Hoffman and Wells 1986)
(see Exhibit 4-7).

    Consequently, as Exhibit 4-7 outlines, it is possible that CH4 emissions
will cease growing (if they are growing now) if the growth of various sources
slows, if source terms are eliminated or if emissions factors stay constant or
decline (which may not happen).  However, it is also possible that increases
in temperature or changes in hydrology will increase emission factors,
compensating for, or even overwhelming, decreases in source term quantities
(Hoffman and Wells 1986).

    Estimates of Changes of Historic CO Concentrations

    Information about past CO trends would allow models to be used to estimate
how CO and CH4 fluxes have changed in the past.  The following is a summary
from World Meteorological Organization's international assessment (World
Meteorological Organization 1986) (references cited refer to the WMO document):

            Khalil and Rasmussen (1984b,c) indicated CO increases of
        about 5% yr-1 at Cape Meares, Oregon, between 1979 and
        1982.  Subsequent measurements at this site lowered the
        estimated mean trend.  Results presented by Dvoryashina et
        al. (1984), based on spectroscopic measurements of CO over
        the U.S.S.R. between 1971-1983, suggest a 1-2% increase
        during that period.  Rinsland and Levine (1985) deduced
        values for CO over Switzerland in 1951 (from Migeotte's
        plates), and they estimated a mean annual increase of ~ 2%
        yr-1 between 1951 and 1981 at that site.

            Mean concentrations and variability are smaller for CO
        in the Southern Hemisphere than in the north.  The dominant
        sources of CO in the south may be the oxidation of CH4 and
        transport from the north.  Comparison of recent measurements
        of CO at Cape Point, South Africa, (1978-81) with shipboard
        data obtained in 1971-1972 indicates that CO in the southern
        hemisphere may have increased by 0.5-1% yr-1 (Seiler et
        al., 1984b).  Measurements of CO taken at Tasmania and the
        South Pole do not yet show statistically significant
        increases (Fraser et al., 1984).
                            * *  DRAFT FINAL  * * *

-------
                                   4-17
                              EXHIBIT 4-7

      Possible  Changes  in CH4 Sources and in  Emission  Factors
SOURCE TERMS
 POSSIBLE CHANGE IN LEVEL
     OF SOURCE TERMS
POSSIBLE CHANGE IN EMISSION
  FACTOR FOR SOURCE TERMS
Rice
Livestock
Deforestation

Wetlands



Mining

Tundra
To slow and then stabilize
with population
To slow and then stabilize
To slow, then disappear

Loss of wetlands should
increase, then slow and
stabilize

Increasing

None
Decline possible if:

•   increased C02 reduces
    emissions by reducing
    transport through plants

•   cultivars shift

•   cropping practices
    remove wastes

Increases possible if warmer
temperatures raise emissions
per unit biomass

Possible decrease with
biotechnology to reduce
methanogenesis and "wasted"
energy
Possible increase with
temperature/decrease with
C02

No information available

Changes are possible:

•   Increase with rising
    temperature if
    methanogenesis increases

•   Decrease with rising
    temperature if several
    microorganisms are
    cold-tolerant and subject
    to adverse selection if
    warming occurs
Source:  Hoffman and Wells 1986

                          * * *  DRAFT FINAL  * * *

-------
                                   4-18
    Unfortunately the extent to which these measurements can be extrapolated
with confidence to estimate global CO change is unclear given the spatial and
temporal variability of CO.  The importance of establishing a definitive trend
for CO concentrations is clear, however, since once the CH4 and CO trends are
both understood, the fluxes generating them can more easily be derived.

    For example, modeling analyses by several investigators (Khalil and
Rasmussen 1985a; Levine, Rinsland, and Tennille 1985; and Thompson and
Cicerone 1986) imply increases in CH4 fluxes from 1850 to 1985 ranging from
15% to 70%, depending on the magnitude of assumed CO changes during that
period.  Levine, Rinsland, and Tennille (1985) used their estimates of changes
in CO and CH4 concentrations (1% and 2%, respectively), to estimate a change
of 14% in methane fluxes and 86% in CO fluxes to reconstruct observed
atmospheric changes in these gases.  The uncertainties in measurements and
atmospheric models, however, are still too large to rely upon.  Their paper,
interesting as it is, does not convincingly explain why methane concentrations
have been rising.

    Current and Historical CO Emissions

    Information about past and current CO emissions is sparse.  Logan (1985)
has made estimates of the current sources and sinks of CO emissions (Exhibit
4-8).  Estimates of past global CO emissions have not been made, except by use
of models that make assumptions about concentrations of CH4 and CO.  Kavanaugh
(1986) has recently completed a study of historical CO emissions from •
combustion.  He estimates that from 1960 to 1975, emissions of CO from fossil
fuel combustion increased at a rate of 1.7% per year.

    Future Emissions of CO

    Estimates of future emissions of CO require projecting changes in all
source terms and their associated emission factors.  A search of the
literature revealed no published literature of future CO emissions from
natural sources.  Kavanaugh (1986), however, argues that there is likely to be
a substantial reduction in future CO emissions from combustion.

    Exhibit 4-9 shows three such scenarios.  The cause of the expected decline
in CO emissions is the penetration of efficient engines into world vehicle
markets, the substitution of ethanol, diesel, and unleaded gasoline for leaded
gasoline, and the adoption of pollution-reduction equipment in some parts of
the world's transportation sector (Kavanaugh 1986).

    Furthermore, to the extent that CO emissions stem from forest clearing,
this source of emissions can be expected to slow and eventually terminate, as
clearable forest areas that can be cleared are exhausted.

    Past, Current, and Future Emissions of NOx, OH, and
    Non-Methane Hydrocarbons

    NQx and non-methane hydrocarbons, through their effect on OH and CO, play
a role in determining CH4 concentrations.  Unfortunately, there is little
historical information on these compounds and even present-day ambient data
                            * *  DRAFT FINAL  * * *

-------
                              4-19
                         EXHIBIT 4-8

           Current  Sources and Sinks of Carbon Monoxide
             (1984 Concentrations of CO:  30-200  ppb)
           SOURCE GASES
Atmospheric burden (10s tons as carbon)                            200

Sinks + Accumulation (10s tons per year  as carbon)
    Reaction with OH                                          820+300
    Soil uptake                                                   100
    Accumulation (5.5%/year)                                       10

TOTAL                                                         940+330

Sources (10s tons per year as carbon)
    Fossil fuel combustion                                        190
    Oxidation of anthropogenic hydrocarbons                        40
    Wood used as fuel                                              20
    Oceans                                                         20
    Oxidation of CH4                                              260
    Forest wild fires (temperate zone)                             10
    Agricultural burning (temperate zone)                          10
    Oxidation of natural hydrocarbons  (temperate zone)             100
    Burning of savanna and agricultural  land (tropics)             100
    Forest clearing (tropics)                                     160
    Oxidation of natural hydrocarbons  (tropics)       .             150

TOTAL                                                            1060

Tropical Contribution
    Burning                                                       100
    Forest clearing                                               160
    Oxidation of hydrocarbons                                     150

TOTAL      '                                                       410
Source:  Logan et al., (1981), updated by Logan et al.  (private
         communication, 1984), as reported in World Meteorological
         Organization (1986).
                       * *  DRAFT FINAL  * * *

-------
                                   4-20
                                EXHIBIT 4-9

        Scenarios of Carbon  Monoxide (CO) Emissions from Combustion
                900
eoo -



700 -



600 -



500 -


   I
IOC -
                                                              High
                                                              Medium
                              —r—
                               I960
                           	1	
                            2000
Future carbon monoxide emissions from combustion, as projected by Kavanaugh
(1986).  The case of the expected decline is the penetration of efficient
engines into world markets, the substitution of cleaner fuels, and the
increased use of pollution reduction equipment in transportation.

Source:  Kavanaugh (1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   4-21
are uneven in quality and spatial resolution.  An evaluation of past NOx and
non-methane hycrocarbon combustion sources is underway (Dignon and Hameed
1986; and Hameed, personal communication to Anne Thompson, 1986).  These gases
are also relevant to the formation of tropospheric ozone, which may be
increasing (Logan 1985).

    Future Concentrations of Methane

    The standard modeling scenario used by the atmospheric community is a
growth rate of 1% per year for CH4 (World Meteorological Organization 1986).
Accurately predicting future concentrations of CH4 would require a much better
understanding of all of the areas discussed earlier in this chapter,
including:  how the CH4 sources would change over time (how much more rice
paddies, land cattle; how much less wetland area); how emission rates (e.g.,
emissions per acre of rice) would change because of alterations in
temperature, water, cultivation practices, and even C02; how CO, NOx and
non-methane hydrocarbon emissions will change; and how tropospheric chemical
balances will change as a result.  Current science is not yet capable of
making estimates for all of these and is unlikely to be able to do so for some
time.

    Scenarios Proposed for Assessing Risks and Later Policy Testing

    In the absence of such information, Hoffman and Wells (1986) suggested the
following scenarios of CH4 concentrations (shown in Exhibit 4-10):

        Standard Case

        •   assume 1% growth for the whole forecast period.  This
            is the extrapolation used most commonly in the
            scientific community.

        Sensitivity Cases

        •   assume CH4 concentrations stop growing in 2010, a year
            in which deforestation, a source of CO and CH4 fluxes,
            is likely to have slowed or stopped if current rates of
            deforestation continue (Woodwell et al. 1983).

        •   assume that the rate of growth slows from 1% per year
            to no growth at the same rate at which world population
            growth slows (this scenario assumes that CH4 growth has
            been co-linear with population growth).

        •   assume that CH4 concentrations increase to 1.25% and
            then 1.5% in 2020 and 2050, respectively.  (This
            scenario asserts that temperature increases overwhelm
            other forces that would tend to reduce the growth of CO
            and CH4 fluxes.)

These scenarios are suggested in order to illuminate the effects of
uncertainty about CH4.
                          * * *  DRAFT FINAL  * * *

-------
                                   4-22



                               EXHIBIT 4-10

                 Scenarios of Methane  (CH4) Concentrations
                                          2060
                                                       i I3O
Hoffman and Wells (1986) suggested the above scenarios for methane (CH4)
concentrations.
Reference
   Case 1
Sensitivity
   Cases

   Case 2
   Case 3
   Case 4
                       Assumptions
1% growth.  This is the scenario commonly used by the
atmospheric modelling community.
1% growth, to 2010, constant concentrations thereafter.
This scenario models decreased rates of deforestation.

1% growth, slowing to no growth as world population growth
slows to zero.

1% growth to 2020, 1.25% growth from 2021 to 2025,  and
1.5% growth from 2026 to 2050.   This scenario models
positive temperature feedbacks.
   Source:  Hoffman and Wells, 1986.
                          * * *  DRAFT FINAL  * * *

-------
                                   4-23



Nitrous Oxide  (N2O)

    Measured Increases in N20 Concentrations

    Chapter 2 summarizes that historical changes in N20 concentrations have
increased approximately 6 ppb over the last 300 years (Pearman et al. 1986).
Currently, concentrations are increasing at about 0.2% per year (Weiss 1981).

    Historical N20 Emissions
    The major sources of N20 are emissions from soils, fossil fuel combustion,
and fertilizer use (McElroy and Wofsy 1984).  Researchers estimate that fossil
fuel combustion currently accounts for between 20% and 30% of the emissions,
with fertilizers and natural sources contributing the remainder (McElroy and
Wofsy 1984).

    N20 and the Biogeochemical Cycle

    The only known sink for N20 is stratospheric photolysis and reaction with
singlet atomic oxygen.  Both processes occur only in the upper atmosphere, and
the atmospheric residence time of N20 is consequently long -- approximately
100 to 175 years (World Meteorological Organization 1986).  As with the CFCs,
the long lifetime means that curtailing the growth of N20 emissions so that
N20 emissions are held constant would not prevent an increase in N20
concentrations for many decades since N20 is far from the equilibrium value
with its current emissions.

    Projections of Future N20 Emissions and Concentrations

    Future emissions of N20 will primarily depend on fossil fuel combustion,
natural emissions, and agricultural activity.  Estimates of future N20
emissions from energy combustion were prepared by Kavanaugh (1986).  Kavanaugh
used the Institute for Energy Analysis/Oak Ridge Associated Universities
Long-Term Global Energy Model to project fossil fuel use (Edmonds and Reilly
1985b).  He compiled current emission factors, and considered possible changes
in emission factors (from increased adoption of catalytic converters, for
example).  Kavanaugh found large increases in N20 emissions from combustion:
"a doubling in emissions from 1975 to 2000 and a 44% increase from 2000 to
2025."  He stated that the "driving force in these changes is rapid economic
and population growth in [the rest of the world] and China."

    Weiss (1981) developed a simplified atmospheric model that computes
atmospheric N20 concentrations from increases in N20 emissions.  Input to the
Weiss model is the aggregate emissions growth rate in each time period.  To
compute this aggregate growth rate, emissions from all sources must be
projected.  Estimates for future changes in the flux of agricultural and
natural emissions have not been developed, however.

    To conduct a preliminary analysis of future N20 concentrations, we adopted
simplified scenarios for each N20 source term.  Combustion was assumed to
contribute 31% of the N20 source term (McElroy and Wofsy 1984).  Projections
of the combustion source were taken from Kavanaugh (1986).  The agricultural
                          * * *  DRAFT FINAL  * *

-------
                                   4-24
source was assumed to contribute 58% of the source term (McElroy and Wofsy
1984).  In the absence of independent projections, it was assumed that the
agricultural source would increase at the same rate as world population.   The
population projections, consistent with those in the energy model, were based
on the work of Keyfitz and are discussed in Gibbs (1986).  Natural fluxes,
assumed to contribute 11% of total emissions (McElroy and Wofsy 1984), were
held constant.

    Using the base allocation of source terms and the future increases in each
source, aggregate growth rates were computed.  Exhibit 4-11 presents the
growth in emissions from each source and the aggregate growth rate.

    These increases in N20 emissions were used as inputs to the Weiss (1981)
model.  Exhibit 4-12 shows that the computed concentrations for this
preliminary scenario are consistent with the N20 concentrations scenarios
generally used by the atmospheric modeling community.  Estimates of growth in
concentrations for scenario testing include Ramanathan et al. (1985), who
project increases of 0.3% in concentrations and Wuebbles et al. (1984), who
suggest an increase of 0.25%, with an uncertainty range of 0%-0.3%.

    Standard Scenario Proposed for Assessing Risks and Later Policy Testing

    Because our preliminary emission-concentrations analysis gives results
consistent with the scenario commonly used by the atmospheric modeling
community, we propose using this scenario -- 0.25% growth per year -- for the
standard case for risk assessment.  The standard scenario assumes no effort is
made to limit emissions caused by human activity.

EFFECTS OF  POSSIBLE FUTURE LIMITS ON  GLOBAL WARMING

    C02, CH4, and N20 are radiatively active trace gases that may contribute
to future global warming (Ramanathan et al. 1985; Cicerone and Dickinson 1986;
and Wang, Wuebbles, and Washington 1985).  To the extent that decisionmakers
become concerned about global warming, goverments may take action to limit the
rise in their concentrations.

    Mintzer and Miller (1986) suggest that several measures are available to
reduce greenhouse gases either directly, by limiting their emissions, or
indirectly, by manipulating their atmospheric loss terms.  To reduce C02
emissions from biogenic sources, they suggest reforestation efforts;  To
control CH4 concentrations, they suggest regulations to limit emissions and
natural gas pipelines, coal mines, and oil fields.  Buildup in the atmosphere,
they contend, could also be limited by efforts to reduce CO emissions, which,
as this chapter earlier discussed, may deplete OH concentrations, which are a
major loss term for CH4.  A recent workshop report indicated that N20 might be
controlled through better combustion practices (EPA, 1986).

    It is beyond the scope of this chapter to evaluate the likelihood of
successful government action to limit greenhouse gas increases.  As Hoffman
and Wells (1986) point out, however, there are physical, economic, and
political obstacles to reducing global warming by limiting trace gas
                              *  DRAFT FINAL  * * *

-------
                             4-25
                         EXHIBIT 4-11

            Preliminary  Scenario of Future Growth in
                    N2O Emissions by Source
                                  PERCENT ANNUAL GROWTH RATE
SOURCE (% of total)      1975-2000    2000-2025   2025-2050   2050-2075
Combustion   (58)           2.9          1.6          1.6         1.6
Agricultural (31)           1.6          0.9          0.4         0.1
Natural      (11)           0.0          0.0          0.0         0.0

AGGREGATE   (100)           1.9          1.1          1.0         1.0
                      * *  DRAFT FINAL   * * *

-------
                                        4-26



                                   EXHIBIT 4-12

                    Projected Nitrous  Oxide  (N2O)  Concentrations
     400
                                                      Preliminary emissions-
                                                      concentrations model
     380 -
o>
_D
5
c
o
m
o
0.
                                                                          \
                                                              0.25% annual  increase
                                                              in concentrations
360 -
     340 -
     320 -
     300
        1980
                 2000
2020
2040
2060
2080
                                     DRAFT FINAL  * * *

-------
                                   4-27
concentrations.  The long atmospheric lifetimes of C02 and N20 make it
difficult to reduce concentrations even if one can reduce emissions.  CH4 has
a shorter lifetime, which presents little difficulty.  Emissions sources of
CH4, however, may be more difficult to control.  It is difficult to imagine
control policies that could affect emissions from such natural sources as
tundra.  Similar difficulties may exist for natural sources of C02 and N20.

    The ties between trace gas emissions and growth in population and the
economy (Hoffman and Wells 1986) also present obstacles to limiting trace gas
concentrations.  The major source of C02 emissions, fossil fuel use, is
tightly intertwined with modern economies.  Edmonds and Reilly (1985c) used a
model of energy consumption and carbon dioxide to test the .effectiveness of
fossil fuel control policies.  They found that unilateral action to discourage
fossil fuel use would lower world energy prices, which would then spur
increased fossil fuel use in other regions.  The net result would be that
fossil fuel use would remain almost unchanged.  They found that global
policies, which are themselves subject to political obstacles, would be more
effective, but even severe taxes would delay the doubling time of C02
concentrations by only 10 years, at most.

    There is no certainty about how future decision makers will respond to
greenhouse warming.  The default assumption of the standard scenarios has been
that no response will develop.  In order to provide decision makers with an
assessment of the risks of ozone.depletion for various emission scenarios of
ozone depleting substances, it may be useful to develop alternative
scenarios.  We would welcome reviewers comments on what scenario or scenarios
of C02, CH4, and N20 should be used as standard cases in the risk assessment.
In Chapter 18, one possible example is provided.

CONCLUSION

    Because N20 and C02 are long-lived gases, their concentrations are likely
to rise despite significant uncertainties about future trends.  Future
concentrations of CH4 are another matter, however.  CH4 has a relatively short
lifetime and the factors that determine its production and loss terms are
quite, difficult to predict.  In general, CH4 concentrations are not likely to
continue increasing at existing rates unless rising global temperatures
increase emissions enough to compensate or overwhelm the reductions in source
term growth or lifetime due to changing sinks.

    Exhibit 4-13 presents the combined scenarios proposed for use in risk
assessment analysis.  In Chapter 18 an alternative scenario is examined, which
examines how risks would change if future decisionmakers decide to limit
global warming to 3°C + 1.5°C.
                            * *  DRAFT FINAL  * * *

-------
                       4-28
                   EXHIBIT 4-13

Summary of Standard Scenarios Proposed for Assessment



    Carbon  dioxide (CO2) concentrations (ppm)
             YEAR

             1975
             2000
             2025
             2050
             2075
             2100
          NAS 50th  Percentile
           C02 Concentration

                 340
                 366
                 422
                 508
                 625
                 770
             Source:   Nordhaus and Yohe (1983)
    Methane (CH4) concentrations  (ppmv)
               REFERENCE
            SENSITIVITY  CASES
        1980
        1990
        2000
        2010
        2020
        2030
        2040
        2050
        2060
        2070
        2080
        2090
        2100
                Case  1   Case 2   Case 3   Case 4
1.65
1.82
2.02
2.23
2.46
 ,72
 .01
3.32
3.67
4.06
4.49
4.96
5.48
2.
3.
1.65
1.82
2.02
2.23
2.23
2.23
2.23
2.23
2.23
2.23
2.23
2.23
2.23
  65
  82
  02
  15
  30
  46
  62
  80
  30
2.46
2.62
2.80
2.62
1.65
1.82
2.02
2.23
2.46
2.79
3.16
3.58
4.16
4.83
5.62
6.53
7.58
    Nitrous oxide (N2O) concentrations:
        0.25% annual  increase
    Source:   Proposed  scenarios taken from Hoffman
             and Wells,  1986.
                 * *   DRAFT FINAL
                                    * *

-------
                                   4-29
REFERENCES
        \
Acock, B., and L.H.  Allen, Jr.,  (1985),  "Crop Responses  to Elevated Carbon
    Dioxide Concentrations",  in  Strain,  B.R., and J.D.  Cure (eds.)  Direct
    Effects of Increasing Carbon Dioxide on Vegetation,  DOE/ER 0238,  U.S.
    Department of Energy, Washington,  D.C.

Baes, C.F. Jr., A. Bjorstrom,  and P.J.  Mulholland,  (1985), "Uptake  of Carbon
    Dioxide in the Oceans," in Trabalka, J.R.,  (Ed.),  Atmosphere, Carbon
    Dioxide and the Global Carbon Cycle, DOE/ER-0239,  U.S. Department of
    Energy, Washington, D.C.

Brown, S., and A.E.  Lugo, (1981), "The Role of the Terrestial Biota in the
    Global Carbon Cycle," Div. Petrol.  Chem., 26, 1019-1025.

Cicerone,  R.J., and J.D. Shetter, (1981), "Sources of  Atmospheric Methane:
    Measurements in Rice Paddies and a Discussion," Journal of Geophysical
    Research. 86(C8), 7203-7209.

Cicerone,  R.J., and R.  Dickinson, (1986), "Future Global Warming for
    Atmospheric Trace Gases,"  Nature,  319,  109-115.

Connell, P.S., and D.J. Wuebbles, (1986), Ozone Perturbations in the LLNL
    One-Dimensional Model—Calculated Effects of Projected Trends  in CFC's,
    CH4, C02, N20 and Halons  over 90 Years, draft report submitted  to U.S.
    Environmental Protection  Agency, Lawrence Livermore National Laboratory,
    Livermore, CA.

Council on Environmental Quality, (1981), Global Energy Futures and the Carbon
    Dioxide Problem, U.S. Government Printing Office,  Washington, D.C.

Dignon, J. and S. Hameed, (1986).  "A Model Investigation of the Impact of
    Increases in Anthropogenic NOx Emissions Between 1967 and 1980  on
    Tropospheric Ozone," J. Atmos.  Chem, in press.

Edmonds, J.A., J.M.  Reilly, (1985a), "Future Global Energy and Carbon Dioxide
    Emissions," in Trabalka,  J.R.,  (Ed.), Atmospheric  Carbon Dioxide and the
    Global Carbon Cycle, DOE/ER  0239,  U.S.  Department  of Energy, Washington,
    D.C.

Edmonds, J.A., J.M.  Reilly, (1985b), The IEA/ORAU Long-Term Global  Energy
    Model, Oak Ridge National  Laboratory, Oak Ridge, TN.

Edmonds, J.A., J.M.  Reilly, (1985c), Global Energy: Assessing the  Future
    Oxford University Press,  New York.

Edmonds, J.A., J.M.  Reilly, J.R. Trabalka,  and D.E. Reichle,  (1984),  An
    Analysis of Possible Future  Atmospheric Retention  of Fossil Fuel C02.
    U.S. Department of Energy, Washington,  D.C.
                          * * *  DRAFT FINAL  * * *

-------
                                   4-30
Emanuel, W., G. Kilbugh, and J. Olson, (1981), "Modeling the Circulation of
    Carbon in the World's Terrestial Ecosystems," in B.  Bolin et al., (Eds.),
    Modeling the Global Carbon Cycle -- Scope 16, John Wiley and Son, New
    York.

Ehhalt, D.H., (1974), "The Atmospheric Cycle of Methane,"  Tellus,  26,
    58-70.

Ehhalt, D.H., and U. Schmidt (1978), "Sources and Sinks of Atmospheric
    Methane," Pageopho, 116, 452-464.

EPA (1986), Proceedings of a Workshop on N20, Office of Research and
    Development, U.S. EPA, Washington, B.C.

Gates , D.M., B.R. Strain, and J.A. Weber, (1983), "Ecophysiological Effects
    of Changing C02 Concentrations," in O.L. Lange, P.S. Nobel, C.B. Osmond,
    and H. Ziegler (Eds.), Physiological Plant Ecology,  IV, Encyclopedia of
    Plant Physiology (new series, volume 12D), Springer Verlag, New York.

Gibbs, M.J., (1986), "Summary of Alternative Population and GNP Projections,"
    presented at the UNEP Workshop on Protection of the Ozone Layer, Rome,
    Italy, May 26-30, 1986.

Hameed, S., J.  Pinto, and R.W. Stewart (1979).  Sensitivity of the predicted
    CO-OH-CH4 Perturbation to Tropospheric NO  Concentrations, J.  Geophys.

    Res., 84, 763-768.

Hoffman, J.S.,  and J.B. Wells, (1986), "Trace Gas Scenarios," presented at
    United Nations Environment Programme Workshop on the Ozone Layer, Rome,
    Italy, May 26-30, 1986.

Hoffman, J.S.,  and J.B. Wells, "The Momentum for a Greenhouse Effect:
    Physical, Economic and Political Limits to Curtailing the Magnitude of
    Global Warming" in Shands, W., and J.S. Hoffman (Eds.), Proceedings of the
    C02 Conference of the National Forest Products Association, June 25-26,
    1984, Boulder, Colorado.  In press.

Hoffman, J.S.,  J.B. Wells, and J.G. Titus, (1985), "Future Global Warming
    and Sea Level Rise," in G. Sigbjarnarson (Ed.), Iceland Coastal and River
    Symposium Proceedings, National Energy Authority, Reykjavik, Iceland.

Houghton, R.A., J.E. Hobbie, J.M. Melillo, B. Moore, B.J. Peterson, G.R.
    Shaver, and G.M. Woodwell, (1983), "Changes in the Carbon Content of
    Terrestial Biota and Soils Between 1860 and 1980:  A Net Release of C02 to
    the Atmosphere," Ecological Monographs, 316, 617-620.

Isaksen, I.S.A., and F. Stordal, (1986), "Ozone Perturbations by Enhanced
    Levels of CFCs, N20 and CH4: A Two-Dimensional Diabatic Circulation Study
    Including Uncertainty Estimates," Journal of Geophysical Research, 91
    (D4), 5249-5263.
                            * *  DRAFT FINAL  * *

-------
                                   4-31
Kavanaugh, M. (1986) "Estimates of Future CO, N20 and NOx Emissions from
    Energy Combustion," Atmospheric Environment, in press.

Keeling, C.D., and GMCC/NOAA, (1985), Monthly Concentrations of Carbon Dioxide
    at Mauna Loa, Hawaii, unpublished.

Khalil, M.A.K., and R.A. Rasmussen, (1983), "Sources, Sinks, and Seasonal
    Cycles of Atmospheric Methane," Journal of Geophysical Research, 88(C9),
    5131-5144.

Khalil, M.A.K., and R.A. Rasmussen, (1985a), "Causes of Increasing Atmospheric
    Methane:  Depletion of Hydroxyl Radicals and the Rise of Emissions,"
    Atmospheric Environment, 19(3), 397-407.

Khalil, M.A.K., and R.A. Rasmussen, (1985b), "Carbon Monoxide in the Earth's
    Atmosphere:  Increasing Trend," Science, 224, 54-56.

Khalil, M.A.K., and R.A. Rasmussen, (1986), "interannual Variability of
    Atmospheric Methane:  Possible Effects of the El Nino-Southern
    Oscillation," Science, 232, 56-58.

Levine, J.S., C.P. Rinsland, and G.M. Tennille, (1985), "The Photochemistry of
    Methane and Carbon Monoxide in the Troposphere in 1950 and 1985," Nature,
    318, 254-257.

Logan, J.A., (1985), "Tropospheric Ozone:  Seasonal Behavior, Trends and
    Anthropogenic Influence," J. Geophys. Res., 90, 10463-10482.

McElroy, M.B., and S.C. Wofsy, (1984), "Tropical Forests:  Interactions with
the Atmosphere," unpublished.

Mintzer, I.M., and A.S. Miller, (1986), "Spicing Up the Soup:  Multiple
    Perturbation Scenarios and Their Implications for Global Warming,"
    discussion draft prepared for the U.S. Environmental Protection Agency
    Workshop "Protecting the Ozone Layer:  Workshop on Demand and Control
    Technologies," March 6-7, 1985.

National Academy of Sciences (NAS), (1984), Causes and Effects of Changes in
    Stratospheric Ozone, NAS, Washington, D.C.

National Aeronautics and Space Administration, (NASA), (1986), Present State
    of Knowledge of the Upper Atmosphere:  Processes that Control Ozone and
    Other Climatically Important Trace Gases. NASA, Washington, D.C.

Nordhaus, W.D., and G.W. Yohe, (1983), "Future Paths of Energy and Carbon
    Dioxide Emissions," Changing Climate, National Acadmey Press, Washington,
    D.C.

Pearman, G.I., D. Etheridge, F. deSilva, and P.J. Fraser, (1986), "Evidence of
    Changing Concentrations of Atmospheric C02, N20 and CH4 from Air Bubbles
    in Antarctic Ice," Nature, 320, 248-250.
                            * *  DRAFT FINAL  * *

-------
                                   4-32
Ramanathan, V., H.B. Singh, R.J. Cicerone, and J.T. Kiehl, (1985),  "Trace Gas
    Trends and Their Potential Role in Climate Change," Journal of
    Geophysical Research, 90(D3), 5547-5566.

Reichle, D.E., J.R. Trabalka, and A.M. Solomon, (1985), "Approaches to Studying
    Global Carbon Cycle", in Trabalka, J.R. (Ed.), Atmospheric Carbon Dioxide
    and the Global Carbon Cycle, DOE/ER-0239,  U.S. Department of Energy,
    Washington, D.C.

Revelle, R.R., (1983), "Methane Hydrates in Continental Slope Sediments and
    Increasing Carbon Dioxide," in Changing Climate:  Report of the Carbon
    Assessment Committee, National Academy of Sciences, Washington, D.C.

Rinsland, C.P., and J.S. Levine, (1985), "Free Tropospheric Carbon Monoxide
    Concentrations in 1950 and 1951 Deduced from Infrared Total Column Amount
    Measurements," Nature, 318, 250-253.

Rotty, R.M., and G. Marland, (1984), The Changing Pattern of Fossil Fuel
    C02 Emissions, DOE/OR/21400-2, U.S. Department of Energy, Washington, D.C.

Seidel, S., and D. Keyes, (1983), Can We Delay A Greenhouse Warming?
    Government Printing Office, Washington, D.C.

Strain, B.R., and J.D. Cure, (1985), Direct Effect of Increasing Carbon
    Dioxide on Vegetation, DOE/ER-0238, U.S.  Department of Energy,
    Washington, D.C.

Thompson, A.M., and R.J. Cicerone, (1985), "Possible Perturbations to
    Atmospheric CO, CH4, and OH," Journal of Geophysical Research,  in press.

Thompson, A.M., and R.J. Cicerone, (1986), "Atmospheric CH4, CO and OH
    from 1860 to 1985," Nature, 321, 148-150.

Trabalka, J.R., (1985), Atmospheric Carbon Dioxide and the Global Carbon
    Cycle. DOE/ER-0239, U.S. Department of Energy, Washington, D.C.

Wang, W., D.J. Wuebbles, and W.M. Washington,  (1985), "Potential Climatic
    Effects of Perturbations Other than Carbon Dioxide," in MacCracken, M.C.,
    and F.M. Luther, (Eds.) Projecting the Climatic Effects of Increasing
    Carbon Dioxide, DOE/ER-0237, U.S. Department of Energy, Washington, D.C.

Weiss, R.F., (1981).  "The Temporal and Spatial Distribution of Tropospheric
    Nitrous Oxide," Journal of Geophysical Research, 86(C8), 7185-7195.

Woodwell, G.M., J.E. Hobbie, R.A. Houghtons,  J.M. Melillo, B. Moore,
    B.J. Peterson, G.R. Shaver, (1983), "Global Deforestation:  Contribution
    to Atmospheric Carbon Dioxide," Science,  222:1081-1086.
                          * * *  DRAFT FINAL  * * *

-------
                                   4-33
World Meteorological Organization (WHO), (1985), Report of the International
    Conference on the Assessment of the Role of Carbon Dioxide and of Other
    Greenhouse Gases in Climate Variations and Associated Impacts, WMO - No.
    661, Villich, Austria, 9-15 October, 1985.

World Meteorological Organization (WMO), (1986), Atmospheric Ozone 1985:
    Assessment of Our Understanding of the Processess Controlling its Present
    Distribution and Change, WMO Global Ozone Research and Monitoring Project
    Report No. 16, WMO, Geneva, Switzerland.

Wuebbles, D.J., M.C. MacCracken, and F.M. Luther, (1984), A Proposed Reference
    Set of Scenarios for Radiatively Active Atmospheric Constituents,
    DOE/NBB-0066, U.S. Department of Energy, Washington, D.C.
                          * * *  DRAFT FINAL  * * *

-------
Chapter 5

-------
                               CHAPTER 5

            ASSESSMENT OF THE RISK OF OZONE MODIFICATION
SUMMARY

    Models project that the average global column of  ozone will  deplete  if
chlorofluorocarbon emissions grow from current  levels and if  the greenhouse
gases that counter ozone depletion continue to  grow at current rates.  For
latitudes greater than 40°N, column depletion is  projected even  for  scenarios
in which CFC emissions are reduced to their 1980  levels and halon emissions
are eliminated.   At 60°N, models project that the depletion for  this latter
scenario would exceed 3 percent by 2030.

    Some uncertainty exists about the validity  of model predictions.  Although
models do  a good job of explaining the distribution  and abundance of most
chemical species in the atmosphere, discrepancies between model  predictions
and apparent observations of the current atmosphere lower confidence in  model
predictions.

    Another source of uncertainty is that laboratory  measurements provide
inexact values for various inputs to the models,  such as rate constants  or
cross-sections.   Over the past 15 years, better laboratory estimates of  rate
constants and cross sections have produced several alterations in depletion
estimates.   In order to understand the likelihood that estimates will change
again with new laboratory measurements, formal  analyses have  been done to
determine the sensitivity of model projections  of ozone depletion to the
currently accepted set of best estimates for such inputs.  Based on  these
analyses, depletion appears likely if CFCs grow,  although a small possibility
exists that no depletion would occur. Additionally, the analyses indicate that
a depletion significantly greater than the current predictions is more likely
than a depletion that is significantly smaller.

    Formal uncertainty analyses take into account only some factors  that could
alter model predictions.  Existing models may omit processes  important to
determining the future evolution of the stratosphere.  No method exists  to
quantitatively evaluate the impacts of such potentially missing  factors.
Clearly, the possibility exists that models have  omitted a key process and
that their predictions under or overestimate the  depletion levels that would
be associated with any scenario of trace gas emissions.

    From 1970 to 1980 ground-based and balloon  monitoring of  ozone in the
stratosphere shows depletion roughly predicted  by models at high and low
stratospheric altitudes, as well as the small increases that  would generally
be expected in the troposphere.  Models have failed,  however, to predict the
observed rapid springtime depletion of ozone over and adjacent to Antarctica.
The implications of the Antartic depletion for  depletion over the rest of the
world cannot be assessed, however, until the processes causing it are
understood.  At this time, it is not known whether the Antarctic 'hole'  is
caused by natural factors or by human activity.  Nor  is it possible  to
                          * * *  DRAFT FINAL  * * *

-------
                                   5-2
determine whether the depletion occurring there is a precursor to future
global conditions or whether it is merely an anomaly that has been created by
Antarctica's unique characteristics.  Until the causes and implications for
other regions are better understood, the rapid depletion over Antarctica
should not be used as a basis for decision making.  The Antarctic depletion
does, however, demonstrate that the atmosphere can change rapidly and in
unexpected ways.

    Recent satellite measurements from Nimbus 7 appear to show a decrease in
global ozone greater than called for in models, especially in Artie regions.
At this time, analysis of this data is still in its preliminary stages,
however, and it cannot be inferred that models are underpredicting depletion.
If further analysis shows that the ozone decreases observed in the satellite
data are real, and that the decreases are caused by manmade chemicals, it may
be necessary to increase depletion estimates from those made with currently
available models.  At this time, however, current models provide the most
reliable estimates of stratospheric response and should be used in assessments
to project risks of ozone depletion.

    Most model projections, including those presented above, have assumed that
the atmospheric growth of carbon dioxide, methane and nitrous oxide will go
unchecked for the period being examined.  Since these gases add ozone to the
atmosphere or prevent ozones depletion from occurring, their growth counters
ozone depletion that CFCs and halons are predicted to cause.  Future efforts
to limit these gases, which might be taken in order to reduce the greenhouse
warming, would make the stratosphere significantly more vulnerable to
depletion from CFCs and Halons.  Thus, the assessment of column depletion
risks associated with CFCs and other depleters depends strongly on the
assumptions made about whether future decisionmakers will act to limit
increases in those greenhouses in the future.

    In conclusion, current knowledge indicates that average global ozone will
deplete if chlorine continues to grow at current rates and that depletion over
much of the United States will exceed the global average.  The possibility
that models underestimate depletion appears larger than the possibility that
models overestimate depletion.  The risk of depletion is much higher if the
assumption that C02, methane, and N20 continue to grow at past levels is
altered because of limits ultimately necessitated by concern over global
warming.
                            * *  DRAFT FINAL  * * *

-------
                                   5-3
FINDINGS

1.   STRATOSPHERIC MODELING PROJECTS THAT THE COMBINED EFFECTS OF A VARIETY OF
    TRACE GASES (CHLOROFLUOROCARBONS.  NITROUS OXIDE.  CARBON DIOXIDE,  HALONS.
    AND METHANE) ARE LIKELY TO REDUCE  THE COLUMN DENSITY OF OZONE UNLESS
    EMISSIONS OF OZONE DEPLETERS ARE PREVENTED FROM GROWING.

    la)  Photochemical theory continues to support the conclusion that
       '  chlorine, nitrogen, and hydrogen can catalytically destroy ozone in
         the stratosphere, depleting column levels.

    Ib)  One-dimensional (1-D) models  currently predict a 5-9 percent
         depletion for the equilibrium concentrations of chlorine that would
         result from constant emission of CFCs at 1977 levels.  While useful
         for intercomparing models, these values cannot be used to assess the
         risks of depletion in an atmosphere in which other gases are also
         changing.

    Ic)  One-dimensional (1-D) models  predict average column ozone will
         decrease if CFG emissions continue to rise from current levels, even
         if concentrations of methane, carbon-dioxide, and nitrous oxide
         continue to grow at past rates.  For a 3 percent growth of CFCs,
         models predict over a 25 percent depletion by 2075 if the other gases
         continue to grow.

    Id)  Two-dimensional models (2-D)  used in steady state multi-perturbant
         studies that include chlorine, methane, and N20 project depletion
         higher than global averages at latitudes greater than 40°N,
         especially in the spring.

    le)  Time dependent simulations of stratospheric change using a 2-D model
         predict that depletion over 4% will occur at some latitudes  for all
         cases of positive growth in CFC emissions.  Such models even predict
         ozone depletion of up to 3% at inhibited latitudes for a scenario in
         which emissions are reduced from current to 1980 levels and in which
         halon emissions are eliminated, but in which the greenhouse gases
         that counter depletion are allowed to grow at historical rates.

    If)  Time dependent simulation using a 2-D model with CFCs growing at 3
         percent, methane rising at 1  percent, N20 at 0.25 percent and C02
         growing at 0.6 percent, projects annual average depletion at 40°N of
         approximately 1.1 percent by 2000 and 5.2 percent by 2030.  At 50°N
         depeletion is projected to be 1.5 percent by 2000 and 6.5 percent by
         2030.  At 60°N depletion is projected to be 2.1 percent by 2000 and
         8.1 precent by 2030.  Springtime depletion would be higher.

    Ig)  Time dependent simulation with a 2-D model with CFC 11 and 12
         emissions rolled back to 1980 levels, CFC 113 capped, other
         chlorinated emissions and bromine emissions eliminated, CH4 rising at
         1 percent, N20 at 0.25 percent, and C02 growing at 0.6, project
                          * * *  DRAFT FINAL  *

-------
                                   5-4
         depletion by 2030 of about 0.5% at 40°N, 0.7 percent at 50°N, and
         1.1% at 60°N (these depletions values are from 1985 levels).   If C02
         concentrations are prevented from growing from current levels,
         depletion would be anticipated to be higher.

    Ih)  A two dimensional model that relies on eddy diffusion for all
         transport, rather than including advective processes, would predict
         depletion somewhat lower north of 40°N than the model used for time
         dependent analysis and somewhat higher south of that latitude.
         However, this model also projects a latitudinal gradient from
         equatorial to polar and northern temperate regions.

    li)  Because of possible increases in the emissions of bromine molecules
         (see Chapter 4), Halons present a more important risk for
         stratospheric depletion than have generally been appreciated.

2.   CURRENT THEORY AND MODELS FAIL TO REPRESENT ALL OBSERVATIONAL
    MEASUREMENTS OF THE ATMOSPHERE AND PROCESSES THAT WILL INFLUENCE
    STRATOSPHERIC CHANGE IN A COMPLETE AND ACCURATE MANNER.

    2a)  While accurately reproducing many measurements in the current
         atmosphere, current models fail to reproduce some measurements; ozone
         at 40 kilometers is underestimated, for example.

    2b)  While including representations .of most atmospheric processes>
         current models fail to include all the processes that influence
         stratospheric composition and structure in a realistic manner.
         Transport processes, for example, are represented in a simplified
         manner that does not emcompass all the complications of movement in
         the real atmosphere.

    2c)  The inability of models to wholly reproduce measurements of the
         current atmosphere lowers our confidence in them to predict the
         future; it is possible that models are over or underpredicting future
         depletion.

3.   UNCERTAINTY ANALYSES THAT CONSIDER A RANGE OF POSSIBLE VALUES FOR
    CHEMICAL AND PHYSICAL INPUTS CRITICAL FOR MODEL ESTIMATION OF DEPLETION
    INDICATE THAT DEPLETION IS LIKELY IF CFCS CONTINUE TO GROW.

    3a)  Uncertainty analyses conducted with one dimensional models predict
         depletion for a variety of CFC levels.  These analyses indicate that
         the probability of a depletion that is significantly less than the
         most likely case (e.g., the depletion estimated with recommended
         kinetics and cross-sections) is much smaller than the probability of
         a depletion that is significantly larger.

    3b)  Uncertainty analyses using different sets of kinetics and cross
         sections have not been tested in two dimensional models.  However,
         different 2-D models have used different approaches for transporting
         species.  This provides a useful test of the sensitivity of model
                            * *  DRAFT FINAL  * * *

-------
                                   5-5
         predictions to the uncertainty of how transport actually works.
         While differing somewhat in the latitudinal gradients of depletion,
         the models with different transport both predict depletion that
         increases with distance from the equator.

    3c)  Not all uncertainties can be tested in the modeling process.  The
         possibility that missing factors may lead to a greater or lesser
         depletion than indicated in formal uncertainty analyses can not be
         excluded.

4.  OZONE MONITORING SHOWS CHANGES IN OZONE ROUGHLY CONSISTENT WITH MODEL
    PREDICTIONS, WITH TWO EXCEPTIONS.

    4a)  Measurements by balloons and Umkehr show 3 percent depletion at
         mid-latitudes in the upper atmosphere, 1.3 percent depletion in lower
         stratosphere, and 12 percent increases in the lower troposphere.
         Uncertainty exists about the accuracy of all these observations.
         These results, however, are roughly consistent with the expectations
         generated by one-dimensional and two-dimensional models.  The ground
         based measurement system covers only a small part of the earth and is
         severely limited at high latitudes.

    4b)  Nimbus 7 measurements appear to show a decrease in global ozone,
         especially at both poles.  However, the decrease in the Arctic from
         1978 to 1984 may have occurred only in the last several years.
         Concern exists about calibration problems which make an exact
         determination of the absolute magnitude of depletion difficult.
         However, the latitudinal variations in depletion seem to indicate
         that a real phenomenon is being observed, not just instrumental drift.

    4c)  The cause of these apparent ozone decreases measured by Nimbus 7 has
         not been sufficiently analyzed to determine whether the changes (if
         they are real) can be attributed to manmade chemicals.  Other
         possible explanations include natural variations caused by solar
         cycles or other processes.  The latitudinal gradients of the changes,
         are, however, roughly consistent with those projected by 2-D model
         results, although the magnitude is substantially larger than models
         predict.  Until further analysis is performed to determine whether
         depletion is actually occurring and whether it can be attributed to
         man-made chemicals, models to assess risks to the stratosphere should
         not be revised.

    4d)  Measurements in the Antarctic spring show that the gradual depletion
         that occurred in the mid-1970's over and near Antartica has given way
         to a steep non-linear depletion from 1979 to 1985.  The ozone maximum
         outside Antarctica (between SOS and 70S) appears to be showing a
         decline.  The depletion of all areas south of 80°S appears to be 16
         percent.
                            * *  DRAFT FINAL  * * *

-------
                               5-6
4e)  Models with conventional chemistry do not predict "the Antarctic
     ozone hole".  Care should be exercised in interpreting the meaning of
     the phenomenon.  Several hypotheses have been put forward, including
     chemical explanation with manmade sources (bromine and chlorine),
     chemical explanation with natural source (NOx, solar cycle), or an
     explanation that claims the phenomenon is entirely due to the change
     in climate dynamics.  Until more is understood about the true causes
   .  of the hole, it is impossible to determine whether the hole is a
     precursor of atmospheric behavior that will occur in other regions of
     the world.  Until a better understanding of the mechanisms creating
     the depletion is  obtained, the Antartic ozone hole should not be
     utilized for making regulatory decisions.
                          *  DRAFT FINAL  *

-------
                                   5-7
INTRODUCTION

    "The  ozone layer  has  a  continuous  distribution  with  a  peak concentration
in the lower stratosphere  between about 20 and 25  kilometers altitude.
Exhibit 5-1 illustrates the standard  definitions of the troposphere,
stratosphere and mesosphere in  terms of  the  profile of temperature  with
altitude and shows an average ozone distribution,"  (NASA,  1986).  Because
the stratosphere constitutes a permanent  inversion  of air, with relatively low
mixing of air with the troposphere below,  the two must be  viewed as two
separate but related zones.

    "Ultraviolet solar  radiation produces ozone in the atmosphere.   Wave-
lengths of less than 242 (nanometers) nm  possess sufficient energy  to
dissociate molecular oxygen,  O2, into its  component O atoms.  These  O atoms
in turn react rapidly  with O2 to form ozone,  O3.  The O3 formed can
subsequently absorb ultraviolet  radiation  in the 200-320  nm wavelength
region, dissociating into an  O atom  and an O2 molecule.   Ozone is  also
dissociated,  by visible and near  infrared radiation,  although to a much
lesser extent.  These processes form a long chain in which the oxygen atom
alternately attaches itself  to  an  O2 to form O3 and  then  is detached,  until
finally, the O  atom and an O3 molecule react  to  re-form  two O2 molecules,"
(NASA, 1986).

    "According to  this very  simple model, the ozone concentration  is
controlled by its production  and loss rates.   The ratio of  the frequency of
the ozone loss to production  rate  is  directly dependent on the effective
length of the chain (catalytic efficiency)  and  thus, the O3 concentration
depends directly on the chain length," (NASA, 1986).

    "There  are chemical processes which  can  shorten this  chain.  Among these
are the catalytic processes of the nitrogen, chlorine and hydrogen  oxides
(NOx, ClOx,  HOx).   These processes have the same net effect as  the direct
reaction  of  O and O3.  For  example, the  simplest catalytic cycle for
involving ClOx is the two-reaction set:
                                      Cl + 03  -» CIO +  02
                                      CIO + 0  •* Cl + 02
                                 Net   0 + 03  -> 02  +  02

                       for NOx:

                                      NO + 03  •* N02 + 02
                                      N02 + 0  -* NO  +  02
                                 Net   0 + 03  -»• 02  +  02
                                 DRAFT FINAL  * * *

-------
                               5-8
                           EXHIBIT 5-1

   Temperature  Profile and Ozone Distribution  in the Atmosphere

                          OZONE CONCENTRATION (cm-1)
        IU
        Q

        H

        <
                     10'°
 10'
   10'
             140
             120  -
             100  -
                     100
200
300
                                                 400
10"
  500
                               TEMPERATURE (K)

The ozone layer has a peak  concentration in the lower stratosphere
between about 20 and 25 kilometers  altitude.   In the troposphere,
temperature decreases with  altitude.   In the stratosphere,
temperature increases with  altitude.

Source:  NASA (1986).
                     * * *   DRAFT FINAL  * * *

-------
                                   5-9
These sets  (or cycles) are catalytic because, at the end, the Cl atom or  NO
molecule is  again available to continue converting O and O3 back to O2"
(NASA, 1986).

    The significance of the catalytic cycles is that  small amounts of the
catalytic species  can affect ozone  in a substantial manner. ... We  now
find that an even smaller amount of the catalytic oxides (in the  part per
billion range) are significant in controlling the amount  of ozone  (NASA,
1986).

Modeling Stratospheric Change

    To estimate the stratosphere's response to increasing concentrations  of
stratospheric perturbants requires modeling  the physical and chemical  forces
that will bear on the catalytic cycles that control ozone abundance.  Chemical
interactions between gases are controlled by their  rate of combination  and
dissociation.  These reactions are controlled by kinetic rate coefficients.
NASA has established an extensive program and review procedure to  establish
the database of rate coefficients.  While subject to continual update and
improvement, this database has been progressively strengthened by  many  years
of laboratory experiments.

    Radiation is  also a driving force on the structure  and evolution of the
stratospheric system, playing an important role in  chemistry  as  well as the
dynamics that transport various species from region to  region.   Uncertainties.
exist in some of  the cross sections of various molecules as well as the energy
received from the sun at different wavelengths.  We refer interested readers
to Chapter 7 of the Word Meteorological Organization assessment  (WHO, 1986)
for details about these uncertainties.

    Transport determines how various species move from  one area  to another.
It attempts to understand how the distribution and  abudance of species  is
changed by motions of the atmosphere.

    [Ideally models would simulate the actual]  structure of the
stratosphere, [which includes the] complex interplay among radiative,
dynamical and  chemical processes  [described.   Such a model would have  a]
relatively complete description  of  all of the relevant processes in a
three-dimensional time-dependent  manner.  This is  not  yet possible, both
because of  limitations in  computer resources and limitations in the complete
understanding of all of the  relevant processes.   Thus,  the problem of
understanding the complete  stratosphere is attacked with a hierarchy of
models which vary in complexity and vary in completeness of description  of
each of the major aspects of the atmospheric system. (NASA,  1986)

    One convenient method of classifying  stratospheric  models is according
to their dimensionality.   A zero-dimensional model considers  chemistry in  a
box,  i.e., at single  points in the atmosphere decoupled from all other
points.  Such  models (or sub-models) allow a detailed description  of the
chemical evolution of the system.   They generally consider situations in
which  a chemically dominated system is driven  from equilibrium.   One example
                            * *  DRAFT FINAL  * *

-------
                                   5-10
is  the  diurnal variation  of the solar input which can be isolated from the
complexity of the overall system so that the problem of diurnal correlation
of reacting  species can  be evaluated.   This leads to accurate computation of
the modification of the effective diurnal average rate of reaction as
compared to that computed from average constituent concentrations.   With
zero-dimensional models, the impact of changes in  rate coefficients or
reaction  mechanisms on  the  chemical system are readily evaluated.   Of course,
care must be taken  to apply these  models  only to chemically dominated regions
of the  atmosphere   (NASA, 1986).

    The  next level of complexity in  atmospheric models is the
one-dimensional model which  considers  variations in the vertical dimension.
In this type of  model, transport in the vertical dimension  is parameterized
as Fickian diffusion in which the same diffusion coefficient  is used for each
chemical  species.   In  general, optimal values for the  diffusion coefficient
vs. altitude  are obtained by fitting  the distribution of one  or  more of the
source gases.   [Since the atmosphere exists in more than one dimension,
there are various ways to interpret the output of one-dimensional models.]
... In  one sense, ... one-dimensional  models  represent the globally  and
annually averaged stratosphere, in that the diffusion  representation  of
transport attempts to account for globally averaged motions in  which  all
horizontal motions average out  ...  in  another  sense,  the models purport to
represent a  specific latitude (30°N) and a  specific time of  year (equinox) in
that they use a diurnal  march of zenith angles for the Sun appropriate to
those conditions.  Thus, the photochemical driver is  not globally averaged
properly, and the models represent a hybrid situation. (NASA,  1986).

    The  next level of complexity in  the dimensionality of atmospheric models
are two-dimensional models.   These models attempt to  simulate the latitudinal
and seasonal variations  in atmospheric  structure.  Some of these models  use
specified two-dimensional transport dynamics while others seek to model  the
evolution  of  the transport by zonally averaged meridional and  vertical dimen-
sions.   A characteristic difficulty  with such models is how  to  rationally
specify the transport effects of asymmetric motions.  (NASA, 1986).

    Three different broad classes  of two-dimensional photochemical models
are presently in use.   These are models  in which transport is  accomplished by
specified  diffusion  coefficients which are quite large.  There are also those
two-dimensional models with either  specified or internally computed advective
circulations with specified small diffusion coefficients;  and  finally, there
are some  models in which an effort is being made to use a  consistent
formulation  of the advective  and diffusive transports.  (NASA,  1986).

    Two-dimensional  models  do allow  latitudinal and seasonal variations of
constituents to  be calculated.  Thus,  these features may be tested against
observations for models of present-day conditions.   These  models also allow
predictions  of the latitudinal and seasonal effects of constituent scenarios
to be made.   Many present  two-dimensional models include  quite complete
chemical  schemes.   In fact,  in  several  cases,  the chemical  scheme  being  used
                          * * *  DRAFT FINAL  * *

-------
                                   5-11
in a two-dimensional model is identical to the one being used  in a
one-dimensional model by  the same group. (NASA, 1986).

    No three-dimensional models with complete  chemistry yet exist  (NASA,
1986).  Efforts to use general  circulation models to test transport of species
have been performed, but no assessments  have been made with them.  Thus for
the purposes of this assessment, no further  mention will be made of them.

Equilibrium Predictions of One  Dimensional Models

    In the World Meteorological Organization (1986) assessment, a  series of
steady state calculations were  done with a variety of one dimensional models.
Exhibit 5-2 shows the atmospheric concentrations used in various "steady state
scenarios" of different trace gas levels.  Exhibits 5-3, 5-4, and  5-5 show the
effect of these trace gas levels on globally averaged ozone depletion.  The
following section highlights the results from  these modeling  experiments.

Interpretation of Prediction

    A  standard scenario for comparing  model results for chlorine
perturbations has historically been the  calculation of CFC-11  and  CFC-12
constant emissions  to steady state.  Results in [Exhibit 5-3]  indicate a
range  in model-calculated  change in total ozone of -5  to -7% for models
without temperature  feedback and -6 to -9%  with temperature  feedback; these
results ...  [assume]  a constant flux of  CFC-11 and CFC-12 at 1980 levels
to steady state.  These values  may be  compared to the -5 to  -9%  ozone change
in WMO (1982), and  -2 to -4% determined in NRC (1984).   Small changes in a
number of chemical  rate constants have  tended to increase the calculated
impact on ozone for this  scenario since the 1984 NRC  assessment.   (World
Meteorological Organization, 1986).

   Until  recently, it was generally thought that the change in total ozone
calculated  in 1-D stratospheric  models for chlorocarbon  perturbations  was
approximately linear,  that  is, the percentage change in total  ozone relative
to the amount of stratospheric  chlorine  (Clx)  was nearly constant.  (World
Meteorological Organization, 1986).

   Chlorine Perturbations

    For large Clx perturbations (> 12   [parts  per billion volume]
ppbv [relative to 1980 Clx values],  Prather et al. (1984) found a
significant nonlinearity in the ozone-Clx relationship, with a  rapid decrease
in the total ozone column  occurring  for  incremental  additional Clx  when the
Clx level  approximately exceeds that of  stratospheric  odd-nitrogen.  The
nonlinearity for large Clx perturbations may have significant  implications
for the interpretation of effects if chlorocarbon emissions increase
substantially.  Other models (e.g.  Wuebbles and  Connell, 1984;  Stolarski,
1984)  have found a  [qualitatively] similar .behavior, (World
Meteorological Organization, 1986)  although  not as sharply non-linear or
occuring at as low Clx value as in Prather et  al. (1984).
                          * * *  DRAFT FINAL  * *

-------
                                   5-12
                               EXHIBIT 5-2

          Steady-State Scenarios Used in  International  Assessment


'•'SO:  Definition of 1980 reference, ambient atmosphere:

                                       Assumptions
                	Assumptions	
C02          =  340 ppm,     N20     =  300 ppb,     CH4      =1.6 ppm,
CO           =  100 ppb,     CH3C1   =  0.7 ppb,     CC14     =  100 ppt,
CFC-11       =  170 ppt,     CFC-12  =  285 ppt,     CH3CC13  =  100 ppt.
CHSBr        =   20 ppt (assumed only stratospheric source of bromine),
CFC-11 flux  =  309 Gg/yr = 8.4E6/cm2/sec,
CFC-12 flux  =  433 Gg/yr = 1.34E7/cm2/sec.

SOA:    Definition of background chlorine, circa 1960 atmosphere:   Same  as
        above without CFC-11, CFC-12., CH3CC13.

Scenario Identification letter and number

 S1A:   CFC-11 and -12 in steady state in 1980 fluxes.
*S2A:   Clx =  8 ppb (approx:  from CFC-11 = 0.8 ppb, CFC-12 = 2.2 ppb).
*S2B:   Clx =  8 ppb plus 2 x CH4 (concentration),  1.2 x N20.
 S2C:   Clx =  8 ppb plus 2 x CH4, 1.2 N20 and 2 x C02.
*S3A:   Clx = 15 ppb (approx:  from CFC-11 = 1.6 ppb, CFC-12 = 4.4 ppb).
-S3B:   Clx = 15 ppb plus 2 x CH4 (concentration),  1.2 x N20.
 S3C:   Clx = 15 ppb plus 2 x CH4, 1.2 x N20, and 2 x C02.
 S4:    1980 with 2 x CH4 concentration.
 S5:    1.2 x N20.
 S6:    2 x CO.
 S7:    2 x C02.
 S8:    NOx injection from stratospheric aircraft 1,000 molec cm-3S-x or
        2,000 molec cm^S-1 at 17 km and 20 km.
 S9:    Bromine increase from 20 ppt to 100 ppt.
    * Also used as 2-D model scenario.

    A set of scenarios for the future evolution of the atmosphere was
    selected by the World Meteorological Organization for 1-D and 2-D
    model simulations of stratospheric ozone.  Each scenario has a
    label.  For instance, S1A refers to the scenario in which
    equilibrium concentrations are reached for CFC-11 and -12 without
    increases in other gases.

    Source:  Adopted from World Meteorological Organization, (1986).
                            * *  DRAFT FINAL  * * *

-------
                                   5-13
                                EXHIBIT 5-3

    Change in Total Ozone from Representative One-Dimensional Models for
            Steady State Scenarios Containing Clx Perturbations
                   	Change in Total Ozone (%)	
                      LLNL      Harvard     AER     DuPont     IAS      MPIC
   Scenario        (Wuebbles)  (Prather)    (Sze)   (Owens)  (Brasseur)  (Bruehl)
S1A CFC 1980 Flux
only
S2A 8 ppbv Clx
only
S2B 8 ppbv Clx
+ 2 x CH4
+ 1.2 x N20
S2C 8 ppbv Clx
+ 2 x CH4
+ 1.2 x N20
+ 2 x C02
-7.0 -5.3
(-7.2)
-5.1 -2.9
(-5.7)
-3.4 -3.0
(-2.8)

(+0.2)
(-2.8)


-5.3 -4.9
(-6.1) (-7.9) (-9.4)
-4.6
(-4.1) (-9.1)
-3.3 -3.1
(-2.3) (-6.0)

(-1.4) (0.0) (-5.2)



S3A  15 ppbv Clx      -12.2       -17.8    -15.
              only   (-12.4)                                  (-8.8)   (-22.0)

S3B  15 ppbv Clx       -7.8        -8.2     -8.8      -7.2
     + 2 x CH4        (-7.2)                                  (-5.6)   (-13.7)
     + 1.2 x N20

S2C  15 ppbv Clx      (-4.6)                                  (-3.5)   (-13.6)
     + 2 x CH4
     + 1.2 x N20
     + 2 x C02
Calculated ozone changes from six modeling groups participating in
international assessment.  Results are from 1-D models and are relative to an
atmosphere with about 1.3 ppb background Clx and with no CFC (Scenario #
SOA).   Numbers in parentheses refer to calculated changes when including
temperature feedback.

Source:  World Meteorological Organization, (1986).
                            * *  DRAFT FINAL  * * *

-------
                                 ,  5-14
                               EXHIBIT 5-4

           Change in  Total Ozone at 40 kilometers for Steady-State
                   Scenarios  Containing Clx Perturbations



LLNL
Scenario (Wuebbles)
S1A

S2A

S2B


S2C



S3A

S3B


S2C



CFC 1980 Flux
only
8 ppbv Clx
only
8 ppbv Clx
+'2 x CH4
+ 1.2 x N20
8 ppbv Clx
+ 2 x'CH4
+ 1.2 x N20
+ 2 x C02
15 ppbv Clx
only
15 ppbv Clx
+ 2 x CH4
+ 1.2 x N20
15 ppbv Clx
+ 2 x CH4
+ 1.2 x N20
+ 2 x C02
-63
(-56)
-55
(-50)
-50
(-45)

(-35)



-74
(-68)
-69
(-64)

(-58)



Change in 40 km Ozone (%)
Harvard AER DuPont IAS MPIC
(Prather) (Sze) (Owens) (Brasseur) (Bruehl)
-64 -62 -62
(-57) (-81) (-59)
-57 -56
(-67) (-57)
-50 -49 -58
(-62) (-50)

(-49) (-55) (-45)



-78 -77
(-83) (-76)
-73 -64 -74
(-81) (-71)

(-78) (-67)



Calculated changes in upper stratospheric ozone (40 kilometers)  from  six
modeling groups participating in international assessment.   Results are from
1-D models and are relative to an atmosphere with about 1.3 ppb  Clx and with
no CFC (Scenario # SOA).   Numbers in parentheses  refer to calculated  changes
when including temperature feedback.

Source:  World Meteorological Organization,  (1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   5-15



                               EXHIBIT 5-5

              Change in Total  Ozone for Steady-State Scenarios
Scenario
S4 2 x CH4
S5 1.2 x N20
S6 2 x CO
S7 2 x C02

LLNL
(Wuebbles)
+2.0
(+2.9)
-2.1
(-1.7)
+1.1
(+1.1)
(+3.5)
Change in
Harvard AER
(Prather) (Sze)
+0.3 +0.9
-2.6 -1.8
+0.3 +0.6
(+2.6)
Total Ozone
DuPont
(Owens)
+1.7
-2.3
+0.8
(+2.8)
(%)
IAS
(Brasseur)
(+1.6)
(-1.1)

(+3.1)

MPIC
(Bruehl)
(+1-4)
(1.2)
(+0.8)
(+1.2)
S8a  NOx, injection    -1.8
     17 km, 1,000     (-1-3)                                  (-1.4)
     molec. cm-3s-1

S8b  NOx, injection    -5.7
     17 km, 2,000                                             (-3.4)
     molec. cm-'s-1

S8c  NOx, injection    -5.7
     20 km, 1,000     (-4.6)                                  (-3.9)
     molec. cm-3s-1

S8d  NOx, injection   -12.2
     20 km, 2,000                                             (-8.8)
     molec. cm-3s-1

S9   Brx               -3.0
     20 to 100 pptv
Calculated ozone changes from six modeling groups participating in
international assessment.  Results are from 1-D models and are relative to the
present atmosphere (Scenario #SO), except for the AER and DuPont calculations,
which are relative to a background atmosphere with no CFCs (Scneario # SOA).
Numbers in parentheses refer to calculated changes when including temperature
feedback.

Source:  World Meteorological Organization, (1986).
                            * *  DRAFT FINAL  * * *

-------
                                  5-16
    [Exhibit 5-6]  shows recent model results from Owens and  Fisher of
DuPont,  which illustrates a number of interesting points.   The independent
variable  is total Clx in the upper stratosphere, regardless  of  the scenario
by which it was achieved.  The calculated ozone change is  presented as a
function  of  stratospheric Clx  for several different values of stratospheric
NOy [NO, N02, N03,  N205, C10N02, NH04 and HN03].  For large NOy  (about 30
ppbv)  the decrease  of ozone  is small  and very nearly linear with increasing
Clx, but for small background  NOy (13 ppbv)  ozone  is strongly and
non-linearly reduced by Clx.   This effect  of NOy is  large: with 31  ppbv NOy,
18 ppbv  Clx is calculated to reduce ozone by 4.5%; but with 13 ppbv NOy, 18
ppbv  Clx is calculated to reduce the  ozone column  by 45% (World
Meteorological Organization,  1986).  This analysis  demonstrates that the
value  of NOy (NO,  N02, and N03)  is important to understanding  the  atmosphere's
future behavior.   Current estimates for NOy are 20  ppm (Connell and Wuebbles,
1986).
       •

Methane  (CH4) Perturbations

    Model results  for a doubling of methane, from approximately 1.6 to 3.2
[parts per million volume]  ppmv, give an increase of ozone ranging from
0.3% to 2.9%, as shown in  [Scenario  S4, Exhibit 5-5].  Results from most
of the  models are  relative to  the I960 reference atmosphere (SO, 2.5 ppbv
Clx) but results from AER  and DuPont are relative to non-CFC reference case
(SOA,  1.3 ppbv Clx). (World Meteorological  Organization,  1986)  A doubling
of CH4 is. not expected however,  in the next 50 years, unless CH4
concentrations rise  faster than 1% per year (see Chapter 4).   Methane has two
impacts:   it creates ozone  in the troposphere and it has a  role in  suppressing
Clx  (Cl + CH4 --HC1).  The  former is most important in balancing ozone
depletion that might occur  if chlorine levels rise (World Meteorological
Organization, 1986).

Nitrous Oxide  (N2O) Perturbations

    The  reaction  O(ID) * N2O = 2  NO  provides the major source of odd  nitrogen
(NOx)  in the stratosphere.   Stratospheric formation  of NOx from N2O  occurs
primarily in  the middle stratosphere,  from about 20 to 40 km.   ...   From
Table  [Scenario S5 of Exhibit 5-5] model results indicate that  an increase in
the  background concentrations  of N2O by 20% from about 300 to 360 ppbv, gives
a decrease  in total ozone ranging from. I.I  to 2.6%.    (World  Meteorological
Organization, 1986)   If  the rise in N20 of 0.25% per year continues, a 20%
increase is likely in the next  50 years  (see Chapter 4).

Carbon Monoxide (CO)  Perturbations

    The  pertubation scenario considered for  CO  is  the doubling of  present
surface concentrations (from  approximately 100 to 200 ppbv).   For five I-D
models,  the  resultant changes in calculated total ozone vary from an increase
of 0.3% to 1.1% (see  [Scenario S6 of Exhibit 5-5]).   ...most of the change
in ozone occurs in the troposphere.   Carbon monoxide participates  in the
chemistry of the free troposphere  as  a  sink for OH by its  oxidation to CO2,
                          * *  *   DRAFT FINAL  * *

-------
                              5-17
                          EXHIBIT 5-6

              Effect of Stratospheric  Nitrogen (NOy)
               on Chlorine-Induced Ozone Depletion
                     30.6 ppbv N0y

                     27.6 ppbv N0y

                     24.1 ppbv N0y

                        2 x CH4

                      20.8 ppb N0y
                                  10   12  14  16  18  20
                          CHANGE IN Clx (ppbv)
Results from Owens and Fisher of Dupont.   Calculated ozone change is
presented as a function of stratospheric  Clx  for  different values of
stratospheric NOy (NO, N02, N03, N205,  C10N02,  NH04,  and HN03).   For
large background NOy  (about 30 ppb)  the decrease  of ozone is small
and very nearly linear with increasing  Clx, but for small background
NOy (13 ppb), ozone is strongly and  non-linearly  reduced by Clx.

Source:  World Meteorological Organization,  (1986).
                     * * *  DRAFT FINAL   *

-------
                                   5-18
and as a source  (or sink) for ozone by the  "smog" reactions.  For the  LLNL
model, doubling CO increases  tropospheric ozone  source terms by about 14% and
the total atmospheric column by 1.1%.  (World Meteorological  Organization,
1986)

   There is a close relationship among OH, CH4,  and CO concentrations;
therefore,  an increase  of any  one of these species has  significant effects on
the others  and on  other important trace gases (Levy,  1971,  1972; Wofsy  et
al., 1972;  Wofsy, 1976;  Sze, 1977;  Chameides et al.,  1977).  This calculation
of doubled carbon  monoxide with other surface concentrations held constant
seems especially  artificial, and it is  to be emphasized that these  single
scenarios are artificial  sensitivity  studies (World Meteorological
Organization, 1986) intended to show us how  models behave, rather than
attempts to simulate future atmospheric evolution.

Carbon  Dioxide

    The maximum percentage effect  in ozone for changes in atmospheric
concentrations of C02  is near  40 km.  Unlike the  other trace  gases that can
perturb stratospheric  ozone, carbon dioxide  (CO2) does not affect ozone
through direct chemical interactions.  Absorption of  solar radiation by
stratospheric  ozone and infrared emission to space by  carbon dioxide  are
primarily responsible for balancing radiative energy  process  in the
stratosphere.  Thus, an increase  in CO2 concentration  alters the heat
balance, reducing  stratospheric temperatures,  and leading to a slowing  down
of temperature-dependent (O  + O3,  NO + O3)  ozone-destruction reactions.   This
results  in  a  net  increase in stratospheric ozone concentrations.  (For very
high chlorine perturbations,  (over 25 ppb Clx) the opposite effect of CO2
may occur; lower temperature reduces the rate of Cl +  CH4,  increasing the
concentration of  ozone  destroying Cl and CIO  relative to inert HCI).  (World
Meteorological Organization, 1986)

    For a  doubling of CO2 the various  models  calculate ...  temperature at
40 km between -7 and  -9  K, calculate  changes in local  ozone  at 40 km between
+9 and  +19%,  and calculate changes in  the ozone column between  1.2 and 3.5%
[Exhibit 5-7].  All of the I-D  radiative convection models, except that of
LLNL, calculate increases in surface temperature  also;  whereas the LLNL model
has a fixed surface temperature.  A sensitivity study  by Wuebbles (I983a)
indicates that this  feature causes  the  LLNL  model to overestimate the total
ozone increases by about 0.4%.  (World Meteorological Organization,  1986)

Nitrogen Oxides  (NOx)  Perturbations

    Historically,  concern about the possible  impact of anthropogenic trace
gas emissions on ozone began  in the early I970's with studies of the  effects
from potential emissions of nitrogen  oxides (NOx) from high-flying supersonic
aircraft (e.g., see Johnston,  1971; CIAP, 1974; NRC, 1975).   Although no such
fleets are  currently proposed, the scenarios assumed at that  time for
hypothetical  fleets  of stratospheric aircraft flying at altitudes  of 17 and
20 km remain useful as  an  indication of the  effects of  nitrogen oxide
                          * * *  DRAFT FINAL  *  *

-------
                         5-19
                      EXHIBIT 5-7

Effect of Doubled  CO2 Concentrations on Ozone Temperature
Ozone Ozone
Column at 40 km
LLNL (Wuebbles) +3.5 +19.3
AER (Sze) +2.6 +9.4
DuPont (Owens) +2.8 +11.5
IAS (Brasseur) +3.1 +18.8
MPIC (Bruehl) +1.2 +13.
Temp.
at 40 km
K
-8.0
-8.4
-7.4
-9.0
-7.1
    Percentage changes  in  total ozone column, ozone at 40
    km,  and temperature at 40 km are shown for a doubling
    of C02 (Scenario  #57), relative to the present
    atmosphere,  as  calculated by 1-D models.

    Source:   World-Meteorological Organization, (1986).
                 * *  *   DRAFT FINAL  * *

-------
                                  5-20
emissions on atmospheric ozone.  Results from the LLNL model for  NOx
emissions of 1,000 molecules  cm-3s-l and 2,000 molecules cm-3s-l injected at
altitudes of 17 and 20 km are given in [Scenarios S8a,  S8b,  S8c, and S8d of
Exhibit 5-5].  Also shown in [Exhibit 5-5]  are results from the model by
Brasseur.   Calculated changes  in total ozone  are  comparable to model results
in the mid-1970's. (World Meteorological  Organization,  1986)

   Of more  immediate concern are impacts on  ozone from surface emissions of
odd nitrogen and from  the emissions of NOx from subsonic aircraft in  the
troposphere  and  lower  stratosphere.  Several studies suggest that  these
emissions may be influencing tropospheric ozone concentrations, with a net
increase in  ozone generally expected from the methane-NOx-smog reactions
(e.g. Logan et al., 1981; Liu et al., 1983; Callis et al., 1983;  Wuebbles et
al.,  1983; Wuebbles,  I983a).  (World  Meteorological Organization, 1986)

Bromine

...Although  bromine chemistry  is in many respects similar to  that for
chlorine,  there are  also significant  differences.  Dissociation and reactions
of CHSBr and other important bromine sources occur at lower altitudes than
for the major chlorine sources.   While the reaction of Cl with  CH4  to produce
HCI limits the abundance of  active chlorine  radical species in  the strato-
sphere, the  reaction of Br with CH4  is endothermic and  therefore negligibly
slow.   Also, the  photolysis of HBr  is more  rapid  than  that of HCI, and the
reaction of OH  with HBr is more rapid than  its rate  with  HCI.  Consequently,
the majority of BrX is  present  as the active species  BrO.  On a  molecule  for
molecule basis, bromine is a much  more efficient sink for stratospheric odd
oxygen than chlorine.   (World Meteorological Organization,  1986)

    The bromine.. .sensitivity test considered  is  an increase in surface
mole fraction of CHSBr from 20 to 100 pptv.   As seen in [Scenario  S9  of
Exhibit 5-5] the  LLNL  model was the only model used to calculate the
perturbation.  It gave  a total ozone change of -3% (without temperature
feedback),  in good  agreement with  the  -4%  calculated change  in total ozone
for the same scenario by Prather et al.  (1984).  The major contribution to
the change  in the ozone column occurs  around 20 km.   The largest relative
change in ozone  (7% decrease)  is at 15 km with a secondary peak at about 40
km.  (World  Meteorological Organization, 1986)

    Because current increases in concentrations are at 23% per year, and
emissions may increase  greatly,  bromine  from  halon molecules constitute a
larger risk for stratospheric depletion  than  generally recognized.

Combined Steady State Perturbation Scenarios

    The calculated steady-state  changes in  total ozone and ozone at 40 km are
shown in  [Exhibits 5-3 and  5-4] respectively, for several combined scenarios
(S2, b, c,  and S3 b and c) involving chlorocarbon  emissions  to give about 8
ppbv or 15  ppbv of upper stratospheric Clx,  doubled methane, nitrous oxide
increased by 20%, and,  in some cases,  doubled carbon dioxide.  Calculated
changes in ozone versus altitude are shown in [Exhibit 5-8].   Each of the
                          »  »  *  DRAFT FINAL  * * *

-------
                              .5-21



                           EXHIBIT  5-8

           Calculated  Changes  in  Ozone  Versus Altitude
           55
            50 —
            45
            40
            35
         E
         £  30
         UJ
         O
         H
         E  25
            20
            15
            10
                     S3B
                    LLNL 1-D MODEL
                    COMBINED STEADY-STATE
                    SCENARIOS (S2B, C AND S3B. C)
                      -  FIXED TEMPERATURE

                      •-  TEMPERATURE FEEDBACK
                       I
              I
I
I
-80  -70  -60  -50  -40  -30  -20  -10   0

           CHANGE IN LOCAL OZONE (%)
                                                   T
                                                   10   20
Changes in vertical  distribution of ozone  for  steady-state changes
in ozone as  calculated by the LLNL 1-D model for  the following
scenarios:

         S2B:   Clx = 8 ppb plus 2 x CH4  (concentration), 1.2 x N20
         S2C:   Clx = 8 ppb plus 2 x CH4, 1.2 x N20,  and 2 x CO 2
         S3B:   Clx = 15 ppb plus 2 x CH4,  1.2  x N20
         S3C:   Clx = 15 ppb plus 2 x CH4,  1.2  x N20, and 2 x C02.

Source:  World  Meteorological Organization,  (1986).
                        - *  DRAFT FINAL  * *

-------
                                  5-22
models used tend  to show similar behavior,  with  large ozone  decreases
calculated in the upper stratosphere, and ozone  increases in the troposphere
and lower stratosphere.   The  addition of the CO2 perturbation reduces the
ozone decrease in  the upper stratosphere.  (World Meteorological
Organization,  1986)

    Several  aspects  of these multiple perturbant runs are notable:   all  result
in some depletion  except  the lowest  cases of Clx.  The variation between
models is small, except for the  MPIC model.  Care should be taken in inferring
too much from these  scenarios, however.  The runs are steady-state
calculations that  utilize various  combinations of chlorine and other
substances without reference to  actual emissions of the various substances
over time; thus, they do  not represent the probable state of the atmosphere at
any future point in  time.  Rather, they  represent various perturbation tests
that allow us to examine  the sensitivity of  the models to specified changes.
As such, their primary benefit is  to demonstrate that the models reach very
similar conclusions.   For the  purposes of assessing risks in the future,  time
dependent runs are much more useful.

EQUILIBRIUM PREDICTIONS FOR  TWO DIMENSIONAL MODELS

    In recent years, many detailed photochemical  and dynamical
two-dimensional models of the  stratosphere have  been developed.  These  have
achieved a measure  of success in simulating  the zonally and  seasonally
averaged distribution  of constituents influenced both by photochemistry  and
transport in the stratosphere, such  as methane and  nitrous oxide (see Miller
et al.,  1981;  Gidel et al.,  1983; Garcia and Solomon,  1983; Jones and Pyle,
1984;  Guthrie,  et  al., 1984; Ko et  al., 1984,  1985).  Ozone densities below
about 20-25 km are  predominantly  controlled by transport of ozone from  the
middle and  upper  stratosphere.  Since most  of the ozone column abundance  at
extra-tropical latitudes is located  in this dynamically dominated  region, it
is important to examine ozone  perturbations  using multi-dimensional  models
that include at least a first order  representation of transport in the
meridional  (height-latitude) plane.   Such studies  reveal latitudinal
variations in ozone depletions, which are of  importance  for ozone monitoring
programs and they provide insight beyond that obtained with comparable
one-dimensional model studies.   (World Meteorological Organization, 1986)

    Here the results of  four two dimensional models are examined.  The model
referred to here as  MPIC  is that of  Gidel et al.  (1983);  the calculations
were  done by Schmailzl and  Crutzen.  The model (GS)  is that of Garcia  and
Solomon  (1983) and Solomon and  Garcia  (1984).  The  model  (AER)  is that
described by  Ko et  al. (1985).  The photochemical reaction  rates used were
those of JPL/NASA  (1985)  [Appendix 1],  and the solar flux, oxygen and
ozone cross sections were taken  from Chapter 7  of the WMO,  1986, but the
authors  used  different methods  in the treatment  of the  Schumann-Runge  bands
and used different boundary conditions.  [Exhibit 5-9]  shows the  scenarios
used  in  each model.  (World Meteorological Organization, 1986)

    A fourth model that  is examined  (IS) is  that of Isaksen  and Stordal.   It
has been added to  the analyses in  Exhibit 5-9.
                              »   DRAFT FINAL  * * »

-------
                                  5-23
                              EXHIBIT 5-9

     Two-Dimensional Model  Scenarios  Used in  International  Assessment
Clx/ppb
Scenario # Total
S2A 8.
S2A 8.2
S3A 15.5
S2C 8.
SMA 9 . 5
2.7
9.5
8MB 9 . 5
SMC 18.
IS 7.2
Reference
1
1
1
1
2
1
1
2
2
1
.3
.3
.3
.3
.7
.3
.3
.7
.7
.0
Increase
6
6
14
6
6
1
8
6
15
6
.7
.9
.2
.7
.8
.4
.2
.8
.3
.2
2xCH4
1.2xN20 Model Symbol
no Garcia and Solomon GS
(1983)
no Ko et al. (1985) AER
no AER
yes Garcia and Solomon GS
(1983)
no Gidel et al. (1983) MPIC
no MPIC
no MPIC
yes MPIC
yes MPIC
no Isaksen and Stordal IS
Scenarios used for 2-D model  simulations of stratospheric ozone.  The GS,  AER,
and MPIC models were used in  the WMO  international assessment.  The IS model has
been added for this analysis.

Source:   Adopted from World Meteorological Organization (1986).
                            * *  DRAFT FINAL  » * *

-------
                                  5-24
    For each of these scenarios, the global,  seasonal average  reduction of
ozone is  given  in  [Exhibit 5-10].  These  results are analyzed and
discussed at a  later section, but first  the two-dimensional structure of the
ozone reductions is presented  by various  graphical means.

Pertubation by Clx Only

    [Exhibit 5-11]  presents  latitude-altitude  cross  sections of the
percentage ozone depletion obtained for winter and spring from the MPIC model
for scenario  SMA (Clx  increase, 6.8 ppbv; reference Clx, 2.7 ppbv).  Similar
plots are shown for all four  seasons according to the AER model  for scenario
S2A  (Clx increase, 6.8 ppbv;  reference Clx, 1.3 ppbv) in  [Exhibit 5-12]
and for scenario S3A (Clx increase, 14.2  ppbv; reference Clx, 1.3 ppbv) in
[Exhibit  5-13]  .   Results of the GS model for  winter and spring  are  given
as a latitude-altitude plot of percentage ozone  reduction  in [Exhibit  5-14]
 for scenario S2A.   (World Meteorological Organization,  1986)  Isaksen's
model results for a similar  scenario is presented in Exhibit  5-15, along with
a comparison to an  equivalent  run  from AER's 2-D model.  Clearly, the results
are similar.

    For these four models,  essentially  the same Clx perturbation  is
represented  by [Exhibit 5-lla]  (winter,  MPIC),  [Exhibit 5-12a]
(January,  AER), and [Exhibit  5-14a]  (December, GS).   Results  for Winter
in IF can also  be seen  in Exhibit  5-15.  Certain similarities and  differences
can be noted.   In all four models, the maximum percentage ozone reduction
at 40 km is 50  to 60%,  which is  in agreement with I-D results  ([Exhibit 5-4]
and  [Exhibits 5-16, 5-17,  and 5-18].   At  this  altitude, the AER  and  GS
models show similar latitude  profiles with  ozone-reduction maxima near the
poles.  Both show  a  saddlepoint minimum  of  ozone  reduction of about 35% near
the equator.  The  MPIC maximum  percentage ozone reduction in an almost
uniform ridge from 85°N at 45 km  to 85°S at 35 km.   In  all three models the
-20% contour is flat almost from pole to pole  at an  altitude of about 30 km.
At 20 km altitude the three  models show qualitatively  similar  features, an
ozone increase  in the tropics and  ozone reduction at extra-tropical
latitudes; but the quantitative  values differ:   GS varies  as -5% at 90°S,
+15% at the equator,  -5% at  90°N;  AER values  are  -10% at the  south pole,  +20%
at the  equator, -30% at the  north  pole; and  MPIC varies  as  -10%  at 80°S,  +5%
at the  equator, -5% at  80°N. (World Meteorological Organization,  1986)  IS
varies from >-5% 80°S to +10%  at the equator to -10% at 80°N.

    The  percentage changes in the ozone  vertical column are shown  as a
function  of latitude  and season in  Dobson contour maps.   The result of the
MPIC model for Clx perturbation SMA  is  given in [Exhibit 5-19a] and the
results of  the AER model are shown for scenarios S2A  and S3A in [Exhibit
5-20].   The  scenario for  [Exhibit  5-19a]  (MPIC) is essentially the same
as that for [Exhibit  5-20a]  (AER), which provides a direct comparison
between  the  models.  For the  MPIC model, larger ozone reductions are obtained
at high latitudes than at  low latitudes  by almost a  factor of two  in the
winter ([Exhibit 5-19a] .  However, for the AER model there are much
greater differences with latitude,  more than  a  factor of four  in February,
for example  ([Exhibit 5-20a]).   The contour intervals  of ozone change are
the same (every 2%)  for [Exhibits 5-l9a and 5-20a]  and  it is  obvious by


                         *  *  * DRAFT FINAL  * * *

-------
                               5-25
                          EXHIBIT 5-10

                     2-Dimensional  Model Results:
            Global and Seasonally-Averaged Ozone  Depletion

Initial
1.3
2.7
1.3
2.7
2.7
1.3
1.3
1.0
Clx/ppbv
Final
2.7
9.5
9.5
9.5
18.
8.2
15.5
7.2

Increase
1.4
6.8
8.2
6.8
15.3
6.9
14.2
6.2
2 x CH4
1.2 x N20
no
no
no
yes
yes
no
no
no
% Ozone
Decrease
1.9
7.2
9.1
4.5
11.1
8.5
18.
7.1
S
-%/ppbv
1.36
1.06
1.11
0.66
0.73
1.23
1.27
1.16
Model
MPIC
MPIC
AER
IS
Results for 2-D models  used  in  international assessment.

Source:  World Meteorological Organization, (1986).
                       * * *  DRAFT FINAL  * * *

-------
                              5-26
                         EXHIBIT 5-11

         Ozone  Depletion by  Latitude, Altitude,  and Season
                   for Clx  Increase  of 6.8 ppbv
                         (MPIC 2-D  Model)
       0
       D
                                                      500
Results from MPIC 2-D model,  for  scenario # SMA  (see Exhibit 5-9).
Panel a shows percent change  in ozone  in winter; panel b in spring.

Source:  World Meteorological Organization, (1986).
                            DRAFT  FINAL  * * *

-------
                                      5-27
                                 EXHIBIT 5-12

               Ozone  Depletion by Latitude,  Altitude, and Month
                          for Clx Increase of 6.8 ppbv
                                 (AER 2-D Model)
a)  0.1

LU
V)
CO
cc
Q.
    10
   100
.-	--40     -40 JL5JLJ...''.
  1000 &
                      JANUARY
                                   -- 60
                           -20
                               I
                                      b)  0.1

                                                                      APRIL     ..60
                                                  :••-"•";•/	-10	   "--.
                                                  £'••—>.V	-20	_"*"* „
                                                  L-.—»'<-	- 30	-V;-= - 50
                               ^o  S  1

                             - 30  5  =

                                  X  cr
                               20  g  (L
                                  jfc
                                                10
                                               100
                                    -10
                                        1000
                                                         -40
                                                                  -40-
                                                                       -50—-..-
                                                  i^l'...„._	- 30...;;-,./---.	.:
                                                      10         o   -..
                                                                               h30
      <
      X
      a.
      o.
                                                                                -10
    -90  -60  -30   EQ   30    60   90
            S               N
                 LATITUDE
                                          -90  -60 -30    EQ  30    60
                                                 S                N
                                                       LATITUDE
90
c)  0.1
                                      d)  0.1
                                                                    OCTOBER
                                                  : _. ............... ------ - 10 ...... •'/
                                                  L- ................... - ...... -30-"'
                                               100
                                              1000
                                                      ""-50       ^,. "-"-"-•*•
                                                    .......... '  _.—-'"    ..-30'''
                                                  r--i-^-~-~~--~-~:---~ — 2° ...... '""
                                                     .- ............... -10
                                                                         -60
                                                                         ^-50  f


                                                                         -40  S
                                                                         ^30
                                                                              X
                                                                         \-20  g
                                                                              a.
                                                                              a.
                                                                                - 10
   -60  -30   EQ   30   60  90
       S             N
           LATITUDE
                                                -90  -60  -30   EQ   30   60
                                                         S             N
                                                             LATITUDE
                                                                               90
      Results  from AER  2-D model,  for scenario # S2A  (see Exhibit  5-9).
      Panel  a  shows percent change in ozone  in January;  panel b in April;
      pancel c in July,  and panel  d in October.

      Source:   World Meteorological Organization, (1986).
                                 *   DRAFT FINAL

-------
                                     5-28
                                EXHIBIT 5-13

               Ozone Depletion by Latitude, Altitude,  and Month
                          for  CLx  Increase of 14.2  ppbv
                                 (AER 2-D Model)
   0.1
v>
ec
o.
    10
   100
                        JANUARY
               	-15	-.^.,
               -30	ir""----
               	45	 ,-, -
               --60
                    -60	'
                    -45
                             .-"**  •-*•
  1000
    -90
                            -15
                                     0.1
                           60
                           50  .
                                   -10
-60  -30    EQ  30   60   90
                 N
        LATITUDE
                                       ~     "'--60               *"*-=
                                                             APRIL
                                                                 60


                                                                 50


                                                                -40


                                                                -30
                                                                                  a
                                                                        x
                                                                        o
                                                                        cc
                                                                        a.
                                                                        a.
                                                           LATITUDE
   0.1
UJ
cc
a.
    10
   100
  1000
                            JULY



               60	-"„"_	-60.~__
._	--45..
                         - 15
                              I
                           60


                           50


                          -40


                          -30


                          -20


                          -10
                                        O
                                        l-
<
x
o
cc
OL
Q.
     -90 -60
     - 30   EQ    30

        LATITUDE
                60  90
                                           LU
                                           CC
                             UJ
                             CC
                             a.



                                                           ,---15-1..
       -90  -60  -30
                                              100 -
                                             1000&
                                                                            90
     Results  from AER 2-D models,  for scenario #S3A (see Exhibit  5-9).
     Panel a  shows percent change  in ozone in January, panel b  in April;
     panel c  in  July; and panel  d  in October.

     Source:  World Meteorological Organization,  (1986).
                           * * *   DRAFT FINAL  * *  *

-------
                               5-29
                          EXHIBIT 5-14

         Ozone Depletion by Latitude,  Altitude,  and Month
                    for  Clx Increase of 6.8 ppbv
                           (GS  2-D Model)
                                          DECEMBER
                   -90  -70  -50 -30  -10  10   30  50   70  90
                  -90  -70  -50  -30  -10   10  30   50   70  90
                                 LATITUDE
Results from GS  2-D  models,  for scenario #S2A (see Exhibit  5-9).
Top panel shows  percent  changes in ozone in December; and bottom
panel in March.

Source:  World Meteorological Organization, (1986).
                      *  * *  DRAFT FINAL

-------
                                  5-30
                              EXHIBIT 5-15

          Ozone Depletion by  Latitude and  Season for CLx Increase
                            of ~6.0 ppv
                               (IS 2-D Model)
                              (AER 2-D Model)
Isaksen:
AER:




CFC Releoses,
1980 Production,
Steody Stote
























2 BU
60

40
u 20
o
t o
t-
-1 -20


-40

-60

tfl _ort
,\V -I'/'-'iz^y'//" ''"vYO^-PO
~^s'/,'~'?-* '/'/ /AV-"JI9\S
^--Iy^T-8'/ ,' \ •- — v >
:;^<5^'' \ \
r~~l*'''\ ~~'
\ / '
1 f
/' • •
/ 1
"\ ' A ^ —4^^'
V.^ 1 *•"" \ V N
~ -1) ' 	 ' '' *~7"*\^^ S"
:^N | ,' ^ 	 */C-'la~vV\ *% ~X\ "
\ 1 / ' ~<" / S * \ S. "* ^> V-
iJ/'V' -~' J -ID'' ^v^ """vX»\
i i.'/.v v^i i L' A \ vr-~i\v
«»3
45
41
37
1 "
g "
§ 25

5 21
< 17

13
9

5
1
. ' A-L ' """^ — '- J---so1--'""1 ' ' ' -
>' l' 	 .-50 ^0 -40'
~ vv ^ ^" ^^^ -
- ~~^--~~- 	 	 30 	 1 'J"J
h ,'^ 	 	 25 	 rZ 	 ;
x' ^^--------^5"------~~'V^^
-/x r'r i l /' Z-"-5 	 =
*<'"''•' '' ' ' '' '"^s^ 	 5° 	 ' 	 ^J
-''"' '' '' (f "N "s^-
~ *-**** * 1 \ £1~~~~ *®~" ' ••-». \ f
*~ '* \ y_ 	 ^r ( ^
• ^ \. ^> 	 ^
^^^^"^^\^'^\
~_ ./ v ^- *~""' \^-i2.rvc:
\ \ **™*fc *^^
'*'•! I I I f I >% I I I I '* I
B.














                          JFMAMJJASOND
                                         MONTH

     Results  from IS  2-D model, for an increase in Clx from 1.0  ppbv  to
     7.2 ppbv.   The bottom panel depicts changes of 1.3 to 8.2 ppbv Clx
     for the  AER 2-D  model.  Both panels show change in the total  ozone
     column as  a function of latitude and time of the year.

     Source:   Isaksen and Stordal (1986a); WMO (1986).
                         * *
                                DRAFT FINAL
                                               * *

-------
                                5-31
                           EXHIBIT 5-16

             Change  in Ozone by  Altitude for CFC-11 and
                   CFC-12 Emissions  at 1980 Levels
                           (LLNL 1-D Model)
             55
             50
             45
             40
          |  30
          LU
          Q

             25
             20
             15
             10
LLNL 1-D MODEL

1980 CFC FLUX

SCENARIO S1A


WITH FIXED TEMPERATURES
                   	WITH TEMPERATURE FEEDBACK
               -70  -60  -50   -40   -30   -20   -10

                         CHANGE IN LOCAL OZONE (%)
                                                       10
Calculated percent  change  in vertical ozone at steady-state  relative
to atmosphere with  no  CFC.

Source:  World Meteorological Organization, (1986).
                     * * *  DRAFT FINAL  * *

-------
                                5-32
                          EXHIBIT 5-17

      Change in Ozone by Altitude for Clx Increase of 6.7 ppbv
                          (LLNL  1-D Model)
            55
            50
            45
            40
            35
            30
         Q

         I  25
            20
            15
             10
LLNL 1-D MODEL
8 ppbv Clx
SCENARIO S2A
                          WITH FIXED TEMPERATURE
                     	WITH TEMPERATURE FEEDBACK
                     I
              -60    -50    -40    -30   -20    -10

                         CHANGE IN LOCAL OZONE (%)
                                                        10
Calculated percent  change in vertical ozone at steady-state  for 8
ppbv Clx relative to  reference atmosphere with 1.3 ppbv  Clx.

Source:  World Meteorological Organization, (1986).
                        *  *  DRAFT FINAL  * *

-------
                               5-33
                          EXHIBIT 5-18

     Change  in Ozone by Altitude for  Clx  Increase of 13.7 ppbv
                          (LLNL 1-D  Model)
          01
          O
          3
             55
             50
             45
             40
             35
             30
             20
             15
             10
LLNL 1-D MODEL


15 ppbv Clx X

SCENARIO S3A
                          WITH FIXED TEMPERATURES
                   	WITH TEMPERATURE FEEDBACK
              -80 - -70   -60  -50  -40  -30  -20   -10

                          CHANGE IN LOCAL OZONE (%)
                                     10
Calculated percent  change in vertical ozone at  steady-state for 15
ppbv Clx relative to reference atmosphere with  1.3  ppbv Clx.


Source:  World Meteorological Organization, (1986).
                      * *
                             DRAFT FINAL  *

-------
                              5-34
                         EXHIBIT 5-19

   Change in  Ozone by  Latitude and  Season for  Clx  Perturbations
                         (MPIC 2-D  Model)
                      80 -
                                             -J  *
Steady state changes in ozone by latitude  and season as calculated
by the MPIC 2-D model for the following  scenarios:

   Panel A, Scenario SMA:  Clx = 9.5  ppbv
   Panel B, Scenario 8MB:  Clx =9.5  ppbv  plus 2 x CH4, 1.2 x N20
   Panel C, Scenario SMC:  Clx =18.0 ppbv plus 2 x CH4, 1.2 x N20

Changes are relative to reference atmosphere with 2.7 ppbv.

Source:  World Meteorological Organization,  (1986).
                     * - *  DRAFT FINAL

-------
                               5-35
                          EXHIBIT 5-20

   Change in Ozone by Latitude and Season for Clx Perturbations
                          (AER  2-D Model)
                90 °N
                60 °N
                30 °N
               0  EQ
                30 °S
                60 °S
                 90 °S
                     -18'
                        -16

                                   -6
\ '•*-.  """'   \  "-12  ""'"
                    JFMAMJJASONO
                                MONTH
                90 °N
                60°N
                 30°N
               0  EQ
                 30 °S
                60 °S
                90 °S
                     -32
                               -10
                                     -10
                           *  /       - lu
                      ''-•—•-'' .-•''..-••' -12-.  '"••-.
                             Ox *3°  \ • 2i "vOx'••
                             i  '• i '' i  'l  i  '''i "T xi  .
                    J  FMAMJ   JA  SOND
                                MONTH
Steady state  changes  in ozone by latitude and  season as calculated
by the AER 2-D model  for the following scenarios:

   Panel A, Scenario  S2A:   Clx =  8.2 ppbv
   Panel B, Scenario  S3B:   Clx =15.5 ppbv

Changes are relative  to reference atmosphere with  1.3 ppbv Clx.  The
global average ozone  changes are -8.5% and  -18%, respectively.

Source:  World Meteorological Organization,  (1986).
                             DRAFT FINAL  *  *  *

-------
                                  5-36
inspection that the AER model shows more  variation  with  latitude  and season
than the MPIC model.   A direct comparison is given by  [Exhibit 5-21] which
gives the latitude dependence of ozone column  reduction for these two models
(spring) (World Meteorological Organization, 1986).   The differences  can be
attributed to differences in  transport especially the rate  of horizontal
mixing.   The results shown  in Exhibit 5-15 show latitudinal depletion
approximately equal in  the  Isaksen/Stordhal 2-D model as  in the AER 2-D
model.   The Isaksen model (see Exhibit 5-15) has a latitudinal gradient in
between the MPIC and AER models.

Mixed Scenarios

    Two-dimensional steady-state  model  studies were carried out in which it
was assumed that methane increased by a factor  of two and nitrous oxide
increased by a factor of 1.2 while Clx increased  by  6.7 ppbv (GS), 6.8 ppbv
(MPIC),  and 15.3 ppbv (MPIC).   Latitude-altitude contour  maps of percentage
ozone change are presented for the GS  calculation for (NH) winter and spring
([Exhibit 5-22,a,b] for the MPIC  model  with 6.8 ppbv increase of Clx for
(NH)  winter ([Exhibit  5-23a] and for spring ([Exhibit 5^23b]  and for
15.3 ppbv Clx for winter ([Exhibit  5-24]   (World Meteorological
Organization, 1986).

    The  general  effect  of increasing the methane abundance is to reduce the
magnitude of the calculated ozone changes.  For the GS model, the qualitative
features  of  the  altitude-latitude contours are unchanged, ([Exhibit 5-14]
vs [Exhibit 5-22]) but there are  interesting quantitative differences.
The high altitude polar maxima of ozone  reduction are 55% without increasing
CH4  and N2O and 45% with  the combined scenario.   The saddlepoint near the
equator  is  reduced from 35-40% (S2A) to 25-30% (S2C, combined scenario).  The
calculated latitudinal gradient in ozone depletion  at 40 km is slightly
greater than in the chlorine only  case.   In both cases the  ozone  depletion
near 30  km is almost  independent of latitude.  The  region of increased ozone
in the lower stratosphere (at 18 km) is  very nearly  the  same in both  cases,
but it covers a  slightly greater range of latitude at low altitude for the
case of the combined  perturbation.   In  the region of ozone increase,  the
maximum value of the  increase is  15% for S2A and 10% for S2C.  These maximum
values near the equator are consistent with the interpretation  as  ozone self
healing  (greater penetration of oxygen-dissociating  radiation to lower
altitudes as ozone is  reduced),  since larger ozone reduction in the upper
stratosphere shows larger ozone increase at 20 km at the equator.  The  GS
model results in  [Exhibits 5-14 and 5-22]  extend down only to 18 km,  and
thus do  not show effects in the lowest stratosphere. (World Meteorological
Organization, 1986)

TIME  DEPENDENT PREDICTIONS  FOR ONE DIMENSIONAL MODELS FOR
DIFFERENT SCENARIOS OF TRACE GASES

    Time-dependent calculations including  multiple-specie perturbations are
regarded as the most nearly  realistic of the one-dimensional model
assessments.  Several studies have considered  such time-dependent
multiple-species scenarios (e.g. Wuebbles et al., 1983; Callis et al., I983a;
                             *  DRAFT FINAL

-------
                               5-37
                          EXHIBIT 5-21

         Latitudinal Dependence of AER  and MPIC 2-D Models
   20
-  16
LU
tfl
<
LU
tr
o
LU
a
111
o
N
o
12
                      I
       -80
           -60
-40
S
                           -20
20
                                            40
60
80
                                LATITUDE
Ozone column decrease for Clx increase of about 7 ppbv.  The AER 2-D
model shows more variation with latitude than the MPIC model.  This
difference can be attributed to differences in transport, especially
the rate of horizontal mixing.

Source:  World Meteorological Organization, (1986).
                       * *  DRAFT FINAL  * * *

-------
                               5-38
                          EXHIBIT 5-22

         Change  in Ozone by Latitude,  Altitude, and Month
                      for  Coupled Perturbations
                           (GS 2-D Model)
                             COUPLED CH..N.O
                                          DECEMBER
                  20 -
                   -90 -70  -50 -30  -10.  10
                           S     LATITUDE
                                        30
   50   70
   N
                                            MARCH
                  20 -
                   -90 -70 -50 -30  -10  10
                           S     LATITUDE
30  50
   N
                                                70  90
Steady-state changes  in ozone by latitude and altitude as calculated
by the GS 2-D model.   For Scenario S2C:  Clx = 8 ppbv plus  2 x  CH4
and 1.2 x N20.  Top panel shows results for December; bottom panel
for March.  Changes are relative to reference atmosphere with 1.3
ppbv Clx.

Source:  World Meteorological Organization, (1986).
                          *  DRAFT FINAL  * * *

-------
                              5-39
                         EXHIBIT 5-23

        Changes in  Ozone by  Latitude, Altitude,  and Season
                     for  Coupled Perturbations
                         (MPIC 2-D Model)
                                                     500
Steady-state changes in ozone by  latitude and altitude as calculated
by the MPIC 2-D model.   For  Scenario  8MB:  Clx =9.5 ppbv plus 2 x
CH4 and 1.2 x N20.   Panel  a  shows  results for winter; panel b for
spring.  Changes are relative to  reference atmosphere with 2.7 ppbv
Clx.

Source:  World Meteorological Organization,  (1986).
                       * *   DRAFT  FINAL

-------
                              5-40
                         EXHIBIT 5-24

        Changes in Ozone by Latitude and Altitude in Winter
                     for Coupled Perturbations
                         (MPIC  2-D Model)
  50
  45
     -80
  0      20

LATITUDE
                                                             -  500
                                                            80
Steady-state changes in ozone by latitude and altitude  in  winter  as
claculated by the MPIC 2-D model for Scenario SMC:   CIO -  15.3  ppbv
plus 2 x CH4 and 1.2 x N20.   Changes are relative to reference
atmosphere with 2.7 ppbv CIO.

Source:  World Meteorological Organization,  (1986).
                     * * *  DRAFT FINAL

-------
                                   5-41
Sze et al.,  1983; DeRudder and Brasseur, 1984;  Owens et  al., 1984;  Owens  et
al.,  I985a, b;  Brasseur et al., preprint 1985)  (World Meteorological
Organization, 1986).  Exhibit 5-25 shows  a list  of 1-D and 2-D models used  for
time dependent runs that have been reported  in this analysis.  Recent studies
by Connell using a parameterized version  of  the  Lawrence Livermore National
Laboratory model had added to the available  analysis of 1-D runs.  See Chapter
17 and Connell (1986) for details of the  parameterization.   Exhibit  5-26 shows
the fit between the Connell parameterization and the LLNL  1-D model  results.

    Shown in [Exhibit 5-27a]  are the calculated  changes in total ozone
from several models for scenario T2B, where CFC-II and CFC-12 emissions are
assumed to  increase  1.5% per year, CH4 concentrations to  increase 1% per
year, N2O concentrations to  increase 0.25% per year,  and  CO2 to increase
about 0.5% per year,  corresponding to the analyses of Edmonds et al.  (1984)
as discussed in Wuebbles et al. (1984).  Calculations with  temperature
feedback  tend  to give a  smaller decrease in  total ozone for this scenario
than calculations with fixed temperatures, primarily due to the  impact of
temperature-ozone  interaction  from increasing  CO2  concentrations.

    [Exhibit 5-27b] shows the  change in  ozone at 40 km for this same
scenario.  With the exception  of  the  Brasseur model, similar changes in ozone
at this  altitude are found in  those models with similar temperature
treatments.   [Exhibit 5-28] shows the change in ozone with altitude for
this  scenario at selected times  and calculated with the  LLNL model.  (World
Meteorological Organization,  1986)       .

    Several  time dependent runs have recently been done using the models of
Brasseur, AER, and Connell with greater levels of trace gases.  Exhibit 5-29
shows the Brasseur scenario and Exhibit 5-30 the results  (Brasseur and
DeRudder, 1986).  The run shows global ozone depletion increasing to around 5%
by 2040 and to 30% by 2080.  Note that these runs assumed  methyl chloroform
and carbon tetrachloride are capped in 1986, and that Halon-1211 and -1301  and
CFC-22 are essentially eliminated.  Brasseur also ran scenarios of constant
CFC emissions, other chlorine emissions at 1985  levels (and halons
eliminated).  The results (Exhibit 5-31)  show depletion increasing to almost
two percent  before the increase in C02 (0.5%), Methane (1%) and N20  (0.25%)
overwhelm increases in chlorine concentrations (Brasseur and DeRudder,  1986).

    In the same paper, Brasseur and DeRudder demonstrate a very  important
sensitivity of the models to representation of radiative processes.   Exhibit
5-32 summarizes the results.   Clearly, the different radiative  codes produce
different depletion estimates.  Brasseur points  out that the differences  in
radiative code may explain differences in 1-D model calculations in  the WMO,
1986 report.  In Exhibit 5-33, the results of Connell's parameterization  and
Brasseur's model are presented.  Connell's results, which are lower, are based
on a model that uses an assumption of fixed equilibrium temperature  for
regions above 14km and fixed temperature below.   Clearly,  Brasseur may be
correct in his assertion that differences in model results may  stem  from  this
assumption.   In general, one might expect the approach used by  Brasseur  in  the
model C.I (that is, not assuming fixed equilibrium temperatures) to  be more
                          * * *  DRAFT FINAL  *

-------
                               5-42








                            EXHIBIT 5-25



              Models With Reported Time Dependent  Runs
Wuebbles (Lawrence Livermore National. Laboratory Model) ID




Connell (Parameterization of Lawrence Livermore National Laboratory) ID




AER Model ID, partly funded by Chemical Manufacturs Association




Brasseur ID, Belgium (partly funded by European Community)




Isaksen 2-D, Norway
                      * * *  DRAFT FINAL  * *  *

-------
                                      5-43





                                 EXHIBIT 5-26


                      LLNL 1-D  Model  versus  Parameterized Fit
c
0)
u

c
Q>
LU
CD
I
o
LU   -20  -
o
M
O
O
U
t—
O
                           LOW CFC SCENARIO
                                                     1-D MOObL
                                              	PARAMETERIZED FIT
                                          HIGH CFC SCENARIO
                                                    (  years )
                10
   30      40      50

TIME  FROM PRESENT
         Time  dependent change in ozone as  calculated by LLNL 1-D model  and

         parameterized fit for two divergent  scenarios of growth in CFCs and

         other trace gases.  "Differences between the parameterization and

         the full  1-D model are generally less  than one percent except at

         very  large depeletions."


         Source:   Connell, (1986).
                               * *  DRAFT FINAL  * * *

-------
                               5-44
                          EXHIBIT 5-27

                Time Dependent Change in Ozone for
             Low CFC Growth and  Coupled Perturbations
               o
               z
               <
               I
               u
               0
               z
               3
               o
               u
                      10  20  30  40  50  60  70  80  90  100

                             YEARS FROM PRESENT
                                        WITH TEMPERATURE
                                        FEEDBACK
                 -80
                   0  10  20  30  40  50   60  70  80  90  100

                             YEARS FROM PRESENT
Time dependent  changes  in  ozone as calculated by several modeling
groups in international assessment.   Panel A shows change in total
column ozone; panel  B  in ozone at 40 km.  Assumptions for Scenario
T2B are:
Source:
         Compound
            CFCs
            CH4
            N20
            C02
                                Growth Rate
                                 (%/year)

                            1.5   (emissions)
                            1.0   (concentrations)
                            0.25  (concentrations)
                          ~  0.5   (concentrations)
World Meteorological Organization,  (1986).
                          *  DRAFT FINAL
                                            * *

-------
                               5-45
                          EXHIBIT 5-28

      Time Dependent Change in Ozone  by Altitude for  Low CFC
                   Growth and Couple  Perturbations
                          (LLNL  1-D Model)
               55
               50
               45
               35
I 30

UJ
Q

t 25
              . 15
               10
                  100
                0
                -50
                   LLNL 1-D MODEL
                 — WITH TEMPERATURE FEEDBACK
                   SCENARIO T2B
                          I
                    -40  -30   -20  -10   0    10

                          CHANGE IN LOCAL OZONE. |%)
                                               20
                                                   30
Time dependent changes  in  ozone  by altitude as calculated by LLNL
1-D model.  Change  in ozone  distribution is shown for five years,
ten years, 20 years, 40 years, 60  years, 80 years, and 100 years
from the present.   Assumptions for Scenario #T2B are:
         Compound
            CFCs
            CH4
            N20
            C02
                               Growth Rate
                                (%/year)

                           1.5  (emissions)
                           1.0  (concentrations)
                           0.25 (concentrations)
                         ~ 0.5
Source:
                                 (concent rat ions)

World Meteorological Organization,  (1986).
                     * * *   DRAFT FINAL  » * *

-------
                                  5-46
                             EXHIBIT 5-29

              Trace Gas Assumptions for Results in  Exhibit 5-30
                  (Brasseur  and DeRudder 1-D Model, 1986)
                    CHLOROFLUOROCARBON EMISSIONS
                         (mill kg/year)
                     CFC-11    CFC-12    CFC-113
         OTHERS
Compound
Growth Rate
  (%/year)
1985
1990
1995
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
279.5
253.5
298.0
347.7
455.8
615.3
830.5
1121.1
1513.3
2042.8
2757.4
3722.1
5024.4
6782.2
397.0
396.2
450.6
514.8
665.7
898.6
1213.0
1637.4
2210.2
2893.5
4027.3
5436.2
7338.1
9905.4
148.4
198.6
265.7
347.7
465.8
615.3
830.5
1121.1
1513.3
2042.8
'2757.4
3722.0
5024.4
6782.2
C02
CH4
N20
CC14
CH3CC13









-0.5
1.0
0.25
constant
constant









Average % growth,
1985-2000
2000-2100
1.5
3.0
1.7
3.0
5.8
3.0
Source:   Brasseur and DeRudder, (1986)
                         * * *  DRAFT FINAL  * *

-------
                               5-47
 c
 
-------
                                5-48
                           EXHIBIT 5-31

      Time Dependent Change in Ozone for Constant CFC Emissions and
                      Growth  in Other Trace Gases
                  (Brasseur and DeRudder 1-D Model)
 c
 V
 u
LU
O
X
CJ
LJ
LU
cr.
-2

-3

•4

-5
                                     i    i   l
                              I
                                      i    i    i
                                                     !    I    1    I
          1950.
                       2000
2050
2100
   Time dependent change in ozone as calculated by Brasseur and
   DeRudder,  (1986), for constant CFC emissions at the 1985 level.
   Growth  in  other trace gases is:
           Compound
             CH4
             N20
             C02
                     Growth Rate  in Concentrations
                                (%/year)

                                 1.0
                                 0.25
                               -0.5
   Depletion  increases to almost two percent before the effect of
   increasing CH4, N20, and C02 concentrations overwhelms the effects
   of increase chlorine.

   Source:  Brasseur and DeRudder,  (1986).
                              DRAFT FINAL * *

-------
                                        5-49
                                   EXHIBIT 5-32

                       Sesitivity of 1-D Models to Representation
                                of Radiative Processes
                           (Brasseur  and DeRudder 1-D Model)
 c
 Q>
 O
 0)
LjJ
O
 X
 o
i
 !5
LU
           OZONE
           Total  column
                       Model  C.I
                       Model  C.2
          1950
                                  2000
2050
2100
                                            YEAR
           Time dependent changes in ozone as calculated  for two trace  gas
           scenarios  with two different model representations of radiative
           processes.  Solid lines represent a treatment  with radiative code,
           that is, run iteratively until radiative equilibrium conditions are
           reached in the stratosphere  (above approximately 14 km).   Below 14 km
           a convective adjustment is performed such that the lapse  rate never
           exceeds -5.9 K/km.  The dashed lines show the  case where  the
           radiative  code has fixed equilibrium temperatures at the  surface.

           Source:  Brasseur and DeRudder, (1986).
                               * * *  DRAFT FINAL * *

-------
                                  5-50
                              EXHIBIT 5-33

              Model Comparison:   Time Dependent Change in  Ozone
                  for CFC Growth and  Coupled Perturbations
    -10 -
o>
o
.c
u
o
V
Q.
    -20 -
    -30
        1980
                   2000
                             2020
                                       2040
                                                 2060
                                                                     2100
     Time dependent change in ozone,  as calculated by the 1-D model  of
     Brasseur and DeRudder,  (1986) and Connell's parameterization of the
     LLNL 1-D model.  Trace  gas assumptions are shown in Exhibit 5-30 and
     may be summarized as:
       Compound

     CFC-11
        and CFC-12
     CFC-113
     CC14
     CH3CC13
     CH4
     N20
     C02
                Growth Rate (%/year)
  1.5% to 2000, 3% from 2000 to 2100 (emissions)
  1.0% to 2000, 3% from 2000 to 2100 (emissions)
  constant
  constant
  1.0
 0.25
-0.50
(emissions)
(emissions)
(concentrations)
(concentrations)
(concentrations)
     Note that Connell's parameterization produces lower results than
     Brasseur's 1-D model.

     Source:  Brasseur and DeRudder, (1986); and Cormell, (1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   5-51
likely to describe the real world,  since radiative  equilibrium  is not  likely
to prevail in that region with ozone changes.   Brasseur  does point out,
however,  that both radiative models do as  well in  explaining the current
atmosphere.

    The time dependent AER scenarios are shown in Exhibit  5-34  and the results
in Exhibit 5-35.  They are roughly  consistent  with  Brasseur's runs,  indicating
that global average depletion over  1% will  occur even  if greenhouse  gases are
not limited unless CFCs and other depleters do not  grow  beyond  current levels.

    The LLNL 1-D model has been used to estimate global  average changes  in
ozone over time (Connell and Wuebbles, 1986).   Exhibit 5-36 shows trace  gas
emissions over time for three scenarios in  which concentrations of CH4 grow at
one percent per year, N20 at 0.25%  per year, C02 at approximately 0.6% per
year and halocarbon emissions vary  in each  scenario:   a  "reference case"
(-2.5% growth), "low growth", (-1.4%), and  "high growth" (-4.1%).
Exhibit 5-37a shows the results for the reference case and Exhibit 5-37b for
the low and high cases.

    [For the reference case,] it  is  striking that although this global and
annual average estimate of ozone depletion  reaches  about 20%  after 90 years,
half of the depletion occurs  in the  last  15 years of the simulation.
Depletion over the first 50 years is less than  5%.   This  very slow initial
decrease results from both the mitigating effects of the  concomitant
increases in  CH4 and CO2 and nonlinearity in  the ozone response to Clz
increase [See Exhibit 5-37a]  (Connell and Wuebbles,  1986).

    For a given multiple species  coupled scenario,  the nature of the ozone
response will depend  on the  details of specified emission increases for the
various source species, which chiefly determine whether odd oxygen loss will
remain dominated  by NOx reactions or shift to control by  ClOx....  Assumed
N2O,  CO2  and CH4 trends for [the "high case]  are identical to the
reference scenario.  In the "high"  case, CFC  emissions  increase so rapidly
that trends in other species  play a minor role in affecting the  ozone
change.   The ozone  column response with time is strongly nonlinear as a
result of the  exponential  nature  of the CFC increases.  A depletion  of 20% is
reached after 35 years  (2020) and  60% after 40 years  (2025),  by which time
the validity of the model  has probably broken down [See Exhibit 5-37b]
(Connell and Wuebbles, 1986).

    The "low" case ozone response is also  somewhat nonlinear with time, with
an ozone depletion of 1.5% over the first 50 years  and an  additional  2.3%
over the subsequent 40 years.   Interaction among families and  increases in
N2O,  CO2  and CH4 with time are relatively more important in the "low"  case,
given the gradual  CFC  increase  [see Exhibit 5-37b] (Connell  and Wuebbles,
1986).

    Exhibit 5-38 shows the results  of a run using the  Connell parameterization
in which, instead of assuming greenhouse gases grow without  limit,  it  is
assumed they are eventually  limited (Gibbs, 1986).   In that run, it  is assumed
that C02, CH4 and N20 are halved from the growth rate  assumed in other
                                 DRAFT FINAL  * *

-------
                            5-52
                       EXHIBIT 5-34

       Trace Gas  Assumptions for Results in  Exhibit 5-35
                    (AER 1-D Model, 1986)
                     CH4*                        CFC**
SCENARIO     1960-2020   After 2020      1985-2008    After 2008
1A
2A
3A
3B
4A
4B
1
1
1
1
1
1
.0
.0
.0
.0
.0
.0
1
1
1
0
1
0
.0
.0
.0
.5
.0
.5
No CFC
Constant
3.0
3.0
3.0
3.0
Emission
at 1984 Rates
Constant at
2008 rate
Constant at
2008 rate
Constant at
1984 rate
Constant at
1984 rate
*   Methane growth rate is in percent  per  year  for  the periods
    1960-2020 and after year 2020.

**  Assumed CFC release rates after  year  1985.   Prior to  1985,
    historic release data are used.  The  emission rate at the
    year 2008 corresponds to double  the present-day CFC
    production if 3% annual growth were maintained  through
    1985-2008.

Source:  Chemical Manufacturers  Association,  (1986).
                       *  DRAFT FINAL

-------
                             5-53
                        EXHIBIT 5-35


                Time Dependent  Change in Ozone for
             Various Scenarios of Couple Perturbations
                        (AER 1-D Model)
              2-r
          rO
          O

          c
          E
          J3
          ~O
          O
          O)
          O1
          c
          O
          £
          O
          c
          0>
          O
          0-   -I —
                1960   1980   2000   2020  2040  2060
             -4
                1960   1980   2000   2020  2040  2C6O
Time dependent change  in ozone as calculated by the AER 1-D  model
for several scenarios,  shown in Exhibit 5-34, of trace gases.  The
results indicate that  global average depletion over one percent will
occur if growth in CFCs and other potential ozone depleters
continues.


Source:  Chemical Manufacturers Association, (1986).
                      * *  DRAFT FINAL  * * *

-------
                EXHIBIT 5-36


Trace Gas Scenarios Tested in LLNL 1-D Model
     (Emissions in millions of kg/year)







CFC1 1
3
(CF CICFCI )


Year

1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
2055
2060
2065
2070
2075
CF C11

Reference

325
428
553
717
870
1024
1173
1322
1480
1637
1826
2006
2237
2468
2702
2937
3188
3440
3711
(CFC1 )
3
Low

324
1422
476
522
558
594
626
657
691
725
762
798
837
876
918
960
1006
1053
11 04


High

594
1184
1486
2428
5456
8485
17422
26358
35841
45323
47666
50009
51790
53570
54795
56020
57044
58068
69555
CFC12

Reference

449
528
625
746
870
995
1130
1266
1415
1564
1742
1919
2131
2342
2556
2770
2998
3226
3472
(CF Cl )
2 2
Low

446
519
573
611 .
647
683
723
763
808
853
902
951
1004
1058
1 118
1177
1245
1312
1386


High

527
722
1042
1406
1997
2587
3483
4378
5219
6058
6499
6940
7299
7658
7890
8122
8278
8434
8547
2
Reference,
High

102
142
210
277
311
344
378
411
445
493
540
588
635
683
753
823
893
963
1033
2

Low

102
142
157
180
202
224
245
267
289
320
351
381
412
443
489
534
580
625
671


CFC22 (CHF C1 )

Reference,
High
52
84
122
167
221
273
332
394
463
541
626
715
807
900
998
1097
1195
1292
1387
1495
2

Low
54
71
89
107
127
137
148
155
163
170
178
187
196
205
216
226
238
250
263
276
                                                                                     l/l
                                                                                      I
         * * *  DRAFT FINAL  * * *

-------
                                EXHIBIT  5-36  (continued)


                      Trace Gas  Scenarios  Tested  in  LLNL  1-D Model
                           (Emissions  in millions  of kg/year)
Year
1085
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
20140
20*45
2050
2055
2060
2065
2070
2075
6
CC14 10
Reference
and High
153
188
206
226
250
275
300
325
350
38<4
419
454
489
524
576
628
680
732
784

Low
131
41
'15
49
5 '4
59
64
70
76
83
91
99
107
115
126
137
1148
160
172
CH3CCI3
(methyl chloroform)
Reference, Low,
and High
510
569
634
708
793
878
963
1049
1135
1256
1377
1499
1621
1743
1922
2101
2280
2459
2638
HALON 1301
(CF3Br)
Reference,
High
2
3
4
6
9
12
16
19
22
25
29
32
36
39
43
47
51
55
59

Low
1
2
2
3
4
5
7
9
1 1
12
14
16
18
20
21
23
25
27
29
HALON 121
(CF2BrCI 1
Reference,
High
0
1
1
1
1
2
3
4
5
5
6
7
8
9
9
10
11
12
13
1
Low
0
0
0
1
1
1
2
2
2
2
3
3
4
4
4
5
5
6
7
                                                                                                                    I
                                                                                                                    l/l
                                                                                                                    tn
Biogenic Trace Gases:   CH4  concentrations  at
at approximately 0.6% per year.


Source:   Connell  and Wuebbles (1986).
per year,  N20 at 0.25% per year,  and C02
                                        *  * *   DRAFT  FINAL  * * *

-------
                                  5-56
                             EXHIBIT 5-37a


              Time Dependent,  Globally Averaged Change in Ozone
                         for Coupled Perturbations
                             (LLNL 1-D Model)
                             "Reference Case"
c
OJ
u

c.
0)
Q.
Ul
e>

<
DC
U

Ul

o
M
O

z

u
_J
o
u
         1990   2000  2010   2020   2030   2040   2050   2060   2070  2080.
     Total column ozone  change  for "reference case" scenario of trace

     gases:   ~ 2.5% growth  in CFCs, concentrations of CH4 at 1%,  N20 at

     0.25%,  and C02 at -0.6%.


     Source:   Connell  and Wuebbles (1986).
                          * * *  DRAFT FINAL  * * *

-------
c

-------
                                     5-58
                                EXHIBIT 5-38

                    Effect of Potential  Greenhouse Gas  Controls
                               on Ozone Depletion
                       (Results from 1-D Parameterization)
             -2.0
  Global
   Ozone   -4.  0
Depletion
    «>
             -8.0-
           -10. 0-
           -12. 0-
                                            S.l'C
                      CONTROL OPTION 1
     CONTROL OPTION 1 WITH
     THE GROWTH RATES OF
     CONCENTRATIONS OF OTHER
     TRACE GASES REDUCED BY
     50 PERCENT STARTING
     IN 2000
                 1980      2000
        2020     2040
2060
 Source:   Gibbs (1986).  "Analysis of the Importance of
         Various Design Factors in Determining
         the Effectiveness of Control  Strategy Options."
         UNEP Workshop
        Change in global average ozone as calculated by  a  parameterized
        version of the LLNL  1-D model (see Connell,  1986).   Control option 1
        is:
                 CFCs
                 CH4
                 N20
                 C02
 2.5%  (emissions)  until CFCs double
 1.0%  (concentrations)
 0.25% (concentrations)
~0.6%  (concentrations)
        In "Half Trace  Gas" curve, action is simulated in the  year 2000 to
        cut the growth  rates in half for concentrations of CH4, N20, and C02,
        The temperature equilibrium uses 3°C as the temperature sensitivity
        for doubled  C02.
                                *  DRAFT FINAL  * * *

-------
                                  5-59
scenarios starting in the  year 2000.  As the results show,  the result  still
allows a large greenhouse  warming,  (the change in temperature equilibrium  for
the earth for these scenarios is estimated using the equations of  Lacis,
modified by coefficients from Ramanathan (see Chapter 6)  (Hoffman,  1985)).
Clearly, assumptions about future greenhouse warming are  critical.   Unless one
assumes no efforts are ever made to reduce the greenhouse warming  by limiting
C02, N20, or methane,  models runs that merely extrapolate the growth of these
gases from past increases  in concentrations will seriously underestimate ozone
depletion.

TIME DEPENDENT  PREDICTIONS FOR TWO  DIMENSIONAL MODELS WITH
DIFFERENT SCENARIOS OF TRACE GASES

    Isaksen and Stordal have conducted two sets of time dependent  runs using
their two dimensional model.  The two sets of Isaksen runs  differ  in one
important respect.   In one set of model runs, temperature feedback is  not
considered and C02 was implicitly assumed not to grow.  In the second set  of
runs, however, the radiative cooling for rising C02 is considered  using values
C02 cooling obtained from  the Goddard Institute for Space Studies  general
circulation model.   Their  model is  the only 2-D model that  has been  run in
such a manner.  As shown earlier, the Isaksen model (IS)  compares  very well
with other 2-D models in equilibrium runs.  In addition,  this model  appears to
simulate apparent atmospheric observations of NOy better  in equatorial regions
than other two dimensional models, which have had a difficult time reproducing
LIMS data (see Exhibit 5-39).  While Ko et al (1986) have suggested  that
failures of other 2-D models to reproduce apparent NOy observations  can be
eliminated by using lightening as a source of NOy, it appears that the IS
model's diabatic circulation does as good a job of explaining NOx  levels.

    For the period 1960-1980, Isaksen used CFC release rates from  "Cunnold
et al.  (1983a, CFCI3), Cunnold et al. (1983b, CF2CI2), Prinn et al. (1983b,
CH3CCI3) and Simmonds et al. (1983, CCI4).  The atmosphere  was assumed to
initially  (1960) contain 0.6 ppb of  CH3CI and 0.1  ppb of CCI4, resulting in a
1 ppb  in  content of stratospheric chlorine.   The CH3CI surface flux needed to
obtain  the 1960 mixing ratio was kept constant in all the  computations.  For
the CFCI3,  CF2CI2 and CH3CCI3, the integrations were started in  1960 with
zero abundances and  releases corresponding to  amounts accumulated in the
years prior to 1960,  which is a reasonable assumption since the releases were
small before 1960"  (Stordal and Isaksen,  1986).

    In looking at the time period in which actual measurements exist,  the  IS
model performs well in comparison to Umkehr measurements  (Exhibit  5-40).
"Estimates with  temperature feedback (curve 1) give upper  stratospheric ozone
depletion larger than  what is deduced from Umkehr observations.  We have,
however, also done one model  study where we included temperature  changes
between  1979  and 1980.  The stratospheric temperature decrease adapted for
this period  is taken from  one-dimensional model studies by Brasseur,
DeRudder,  and Tricot (1985).   Their calculations are based on a combined
scenario, where temperature decreases in the 1970-80 period are a result of
increases in CO2 and other trace species.   We have assumed that  approximately
one-third of the estimated  change  up to  1983 given  by Brasseur,  DeRudder, and
                                DRAFT FINAL

-------
                             5-60
                        EXHIBIT 5-39

          Comparison of the Calculated NOy Profile at the
    Equator in  the 2-D Models of Stordal and  Isaksen, and  Ko.
50
45

40
                -   EQUATOR
          t30
             20
                    i  4 11*411
                                              n  -
                    \  ttftMf     t  flflftM
               .1              1             10    30
                      MIXING RATIO (ppbv)
Odd nitrogen (N04 = NO, N02, N03, N205,  C10N02,  HN04,  and HN03)
concentrations  by altitude at the equator.   Curve 1  computed by
Stordal and Isaksen's 2-D model; curve 2 by Ko's 2-D model, and
Curve 3 represents observed nightime concentrations  of two
components  of N04:  N02 and HN03.  The comparison indicates that the
IS 2-D model does as good a job of explaining odd nitrogen  levels.

Source:  Stordal and Isaksen, (1986).
                      * *  DRAFT FINAL  » *

-------
                              5-61
                         EXHIBIT 5-40

            Calculated Ozone Depletion for 1970 to 1980
                     vs. Umkehr Measurements
            UmkeHR
             Layer
                                                        MB

                                                        1-2

                                                        2-4

                                                        4-8

                                                        8-16


                                                       16-32
                                 I    I	>
             -8     -6       -4       -2       0

                          DECADAL OZONE CHANGES
"Calculated ozone depletion between 1970 and 1980 without (curve 1)
and with (curve 2) temperature changes from radiatively active gases
in the stratosphere.  The adapted temperature change is taken from a
1-D model study by Brasseur, DeRudder, and Tricot (1985).  Estimated
global trends (Reinsel et al. 1983; Reinsel et al. 1984) are based
on Umkehr data (triangles with error bars).  A one-dimensional model
calculation by Wuebbles, Luther, and Penner (1983) is included (3)."

Source:  Stordal and Isaksen (1986).
                     - - *  DRAFT FINAL  * * *

-------
                                   5-62
Tricot (1985) takes place between 1970 and  1980.  This  gives  temperature
decreases of approximately 1°C  in the 45-50 km region and less than 1°C below
35 km.   When the  temperature feedback is considered (curve 2) ozone depletion
becomes  approximately two-thirds of the estimated depletion when  no
temperature  effect is  considered."  (Stordal  and Isaksen,  1986).

     Four scenarios were run with the IS 2-D model with temperature feedback.
In scenario  IT,  CFC 11 and 12 are rolled back to 1980 levels,  CFC  113 is
assumed to grow at the same rate as 11 and 12 and to have the  characteristics
of CFC-12, halon emissions do not occur,  methyl chloroform concentrations and
CC14 rise, as in Quinn (1985), at about a 2% rate, methane concentrations rise
at 1.0%,  C02 at 0.6%, and N20 at 0.25% yearly.   Scenario 2T allows CFC11 and
12 to grow at 1.2%, Scenario 3T has 3.0%,  and Scenario 4T has  3.8%, with all
other assumptions the same as IT.  Exhibit 5-41a shows the globally and
seasonally averaged change in ozone calculated for these four  scenarios.

     Scenarios 1WT, 2WT, 3WT, 4WT are scenarios without temperature feedback,
assuming essentially that C02 does not grow.  These scenarios  may  be thought
of as the case in which greenhouse warming is limited, although its unlikely
that C02 would be reduced to no growth and N20 and CH4 allowed to  grow.   The
main value of these scenarios is demonstrating the additional  susceptibility
of the stratosphere to depletion if greenhouse warming is limited  by reducing
C02 or other greenhouse gases.  Of course, the effects of limiting N20 or CH4
would be quantitatively somewhat different than limiting C02,  but  would still
excaberate depletion at some latitudes.  Exhibit 5-41b shows the results for
global averages.

     The results of the various scenarios that include the stratospheric
cooling  (temperature feedback) associated with rising C02 are  included in
Exhibits 5-41a,  5-42, 5-43, and 5-44.  Several important results are clear
from these runs.  If CFCs grow at 3.8%, while halons are eliminated; and CH4
grows at 1%  and N20 at 0.25%, depletion will exceed 4% (from a 1985 base) at
50°N before  the year 2010.  Since near term growth estimates do not preclude
growth at 3.8% (see Chapter 3), these results are particularly important.  For
growth of 3.0% for CFC11 and 12 (with all other assumptions about  limiting
depleters and allowing greenhouse gases to grow), depletion will exceed 4% at
60°N by 2015.  For CFC11 and 12 growth of 1.20% (and other assumptions
limiting depleters and allowing greenhouse gas growth), 2% depletion will be
exceeded at  50°N in 2015.  Even if CFC11 and 12 emissions are  reduced 10% to
1980 levels  (with other depleters limited and greenhouse gases growing), 1%
depletion will be reached at 50°N by 2015, although the ozone  layer would
cover if greenhouse gases grow thereafter.

     Examining the shape of the depletion curves is also instructive.  In
Scenario IT, the rollback case, the rise of C02 and methane eventually limits
depletion to 50°N.  In scenario 3T, with moderate CFC11 and 12 growth, a
turnaround never comes -- depletion continues to increase at an accelerating
rate.
                            * *  DRAFT FINAL  * * *

-------
                                 5-63
                            EXHIBIT 5-41a

              Time Dependent Globablly and Seasonally Averaged
                Changes in Ozone for Coupled Perturbations
                             (IS 2-D Model)
   Ozong
DoplQtion
                 1960  1970  1980  1990  2000  2010  2020  2030
     Results show for four scenarios of trace gas growth:

              Scenario       CFC-11 and CFC-12
              IT
              2T
              3T
              4T
1980  levels
1.2%  growth
3.0%  growth
3.8%  growth
     Assumptions  for other trace gases are  the same in each scenario:
     constant emissions of CFC-113, CC14, and CH3CC13, zero emissions  of
     halons,  one  percent growth per year in CH4, and 0.25 percent growth
     per year in  N20.  C02 concentrations grow at 0.5 percent.

     Source:   Stordal and Isaksen, (1986).
                               DRAFT FINAL

-------
                              5-64
                         EXHIBIT 5-41b

         Time Dependent Globablly and Seasonally Averaged
            Changes in Ozone for Coupled Perturbations
                          (IS 2-D Model)
Results show for four scenarios  of  trace gas growth:

         Scenario       CFC-11 and  CFC-12
         1WT
         2WT
         3WT
         4WT
1980 levels
1.2% growth
3% growth
3.8% growth
Assumptions for other trace gases  are  the  same  in each scenario:
constant emissions of CFC-113,  CC14, and CH3CC13, zero emissions of
halons, one percent growth per  year  in CH4,  and 0.25 percent growth
per year in N20.  C02 concentrations are held constant.

Source:  Stordal and Isaksen,  (1986).
                       * *  DRAFT FINAL * * *

-------
                              5-65
                           EXHIBIT 5-42a

         Time Dependent Seasonally-Averaged Change in Ozone
          for  1980 CFC Emissions and  Coupled Perturbations
                           (IS 2-D  Model)
I960
                   1980
                                                         2020
 Results shown from constant  CFC  emissions at the 1980 level
 (approximately 10% less  than current  emissions); CH4 concentrations
 at 1% per year,  N20 concentrations  at 0.25% per year, and C02
 concentrations at approximately  0.5%  per year.  Changes shown for
 40°N, 50°N,  and  60°N.  Temperature  feedback considered in model.

 Source:  Isaksen (personal communication)
                        * *   DRAFT  FINAL  * * *

-------
                                5-66
                           EXHIBIT 5-42b

           Time Dependent Seasonally-Averaged Change in Ozone
       for  1.2% Growth in CFC Emissions  and Coupled Perturbations
                             (IS 2-D Model)
-6
  I960
                     1980
                                                            2020
   Results  shown  for  1.2% growth per year in CFC emissions;  1% growth
   in  CH4 concentrations; 0.25% growth in N20 concentrations,  and
   approximately  0.5% growth in C02 concentrations.  Changes shown for
   40°N, 50°N,  and  60°N.  Temperature feedback considered in model.

   Source:   Isaksen (personal communication)
                       * * *  DRAFT FINAL  * *

-------
                                 5-67
                            EXHIBIT 5-43

           Time Dependent Seasonally-Averaged Change in  Ozone
        for 3% Growth in  CFC  Emissions and  Coupled Perturbations
                             (IS 2-D Model)
-4 -
—8 —
  1960
                                           I
                                         2000
2020
   Results shown for 3% growth  per year  in CFC emissions; 1% growth in
   CH4 concentrations;  0.25% growth  in N20 concentrations, and
   approximately 0.5% growth in C02  concentrations.  Changes shown for
   40°N,  50°N,  and 60°N.   Temperature feedback considered in model.

   Source:  Isaksen (personal communication)
                              DRAFT FINAL  * * *

-------
                               5-68
                          EXHIBIT 5-44

          Time Dependent Seasonally-Averaged Change in  Ozone
      for  3.8% Growth in CFC  Emissions and Coupled Perturbations
                           (IS 2-D Model)
-10 -
-12 -
    i960
                       1980
                                         2000
  Results  shown for  3.8% growth per year in CFC emissions;  1% growth
  in CH4 concentrations; 0.25% growth in N20 concentrations,  and
  approximately 0.5% growth  in C02 concentrations.   Changes shown for
  40°N,  50°N,  and 60°N.  Temperature feedback considered in model.

  Source:   Isaksen (personal communication)
                          *  DRAFT FINAL  * * *

-------
                                   5-69
    Globally and seasonally average ozone changes for scenario 3WT (moderate
CFC growth) are shown in Exhibit 5-45.   Also shown in this exhibit is the
results for model calculations in which C02 cooling and temperature feedback
are excluded.

    Clearly C02 plays an important role in countering depletion.   The risks of
depletion will be increased if decisionmakers in the future decide not to
accept global warming of the magnitude that would be implied by allowing C02
to rise at 0.6%, N20 to rise at 0.25%,  and CH4 to rise at 1.0% for the rest of
the next century.  More discussion on this point is presented in Chapter 18.

    The Isaksen runs underestimate depletion because they do not consider all
depleting chemicals.  CFC 113 which is not explicitly represented in the model
was predicted to grow at 3% and was assumed to have equal effect as adding CFC
12 emissions to the atmosphere due to their similar photochemical
characteristics (NASA/IPC, 1985).  Clearly CFC 113 is likely to grow faster
than CFC 12 (see Chapter 3).  As mentioned earlier, brominated compounds
(Halon 1211 and 1301) are omitted in the Isaksen runs, thus making all runs
equivalent to an assumption that emissions will be prohibited.  (Currently,
concentrations of 1211 are growing at 23% (Khalil and Rasmussen,  1985).   If
bromine compounds grow at current or faster rates, depletion would be greater
than predicted by these model runs (see earlier discussion on bromine).

A  Comparison of Results of One-Dimensional and Two-Dimensional  Models

    One dimensional and two dimensional models differ in their treatment of
transport.  One dimensional models, at best, project average global
depletion.  Two dimensional models project depletion by latitudes, which of
course, can be averaged together to estimate average global depletion.
Exhibit 5-46 shows the averaged global depletion results of Isaksen's 2-D
model, Brasseur's 1-D model, and the Connell parameterization of the Lawrence
Livermore 1-D model for very similar scenarios.  The outcomes are relatively
similar, indicating that the 1-D models and Isaksen's 2-D models do not differ
in their fundamental projections for global results, and that Isaksen's model
distributes depletion in a manner consistent with total global depletion from
1-D projections.  The higher estimate from Isaksen's 2-D model is to be
expected.  In model comparisons done in the WMO, 1986 assessment, the 2-D
model of AER exhibited similar behavior to Isaksen's.  The global average for
AER's 1-D model, with exactly the same chemistry and an almost identical
perturbation study, was twenty percent higher.  While much remains to be
learned about 2-D models, it is clear that they offer more information for
monitoring, for validation  (and invalidation) and for impact analyses than 1-D
models.  The WMO (1986) report stated "In summary, while 1-D models remain
useful assessment tools for  assessment,  it is becoming clear that 2-D models
provide a much more  detailed picture of  atmospheric response to
perturbations."
                            * *  DRAFT FINAL  * * *

-------
                                   5-70
                               EXHIBIT 5-45


                       Temperature Feedback Experiment:
             Time Dependent, Globally and Seasonally Averaged Change
        in Ozone for 3% Growth  in CFC Emissions and Coupled  Perturbations
                                (IS 2-D Model)
a)
01
c
U

4J
C

       Results  shown are for globally and seasonally  average depletion  for
       model experiment with and without temperature  feedback from C02
       cooling.


       Source:   Isaksen and Stordal  (1986)
                               *  DRAFT FINAL

-------
                                         5-71
                                    EXHIBIT 5-46

                   Model Comparison for Coupled Perturbation  Scenario
Ozone
Depletion
               1985
1995
2005
2015
2025
           Global average change in total  column ozone as calculated by several
           modeling groups for a common scenario of:

                    Compound       Growth  Rate  (% per year)

                    CFCs            3.0 (emissions)
                    CH4             1.0 (concentrations)
                    N20             0.25 (concentrations)
                    C02            ~0.60 (concentrations)

           Results shown for 2-D models of Isaksen and AER, 1-D models of
           Brasseur and Wuebbles, and Connell's parameterization of the LLNL
           1-D model.

           Source:  Chemical Manufacturers Association,  (1986); World
                    Meterorological Organization,  (1986); Connel, (1986);
                    Brasseur and DeRudder, (1986); and Isaksen and Stordal,
                    (1986).
                                * * *  DRAFT FINAL

-------
                                   5-72
MODELS FAIL TO REPRESENT  ALL PROCESSES THAT GOVERN
STRATOSPHERIC CHANGE IN A COMPLETE  AND  ACCURATE  MANNER

    Current models fail to represent  all  the processes that will influence
stratospheric change in either  a complete or accurate manner.  While the
models can reproduce many of the measurements  that describe today's
atmosphere, they fail to reproduce some observations very accurately.  This,
by itself, lowers our confidence in current  model predictions, suggesting the
need for careful sensitivity testing  for  various uncertainties and
assumptions.   In addition, theory suggests that models do not describe all
processes in a completely accurate manner, suggesting the possibility that
missing factors or simplified processes could  undermine the accuracy of model
projections.

The  Nature of Model Validation for Geophysical Systems

    In many science disciplines, the  predictive power of theories and models
can be confirmed by doing and redoing experiments.  In this way experimenters
can gain confidence by actually manipulating the  system under study.  For
geophysical systems, this approach generally cannot be taken  -- there is no
experimental earth on which to  conduct tests.  Thus modelers  attempt to
validate their models by comparing them to the actual atmosphere.  Such tests
impose strong restrictions on models  in the  sense that there  are many
variables to be predicted and fixing  one  aspect of a model's  deficiencies
often results in a new deficiency emerging.  Even if a model  reproduces the  •
current world with complete fidelity, however, it may have substantial
problems.  There may be several models that  reproduce existing condtions, yet
projections of change from each could diverge.  Models that  'fit well' may
lack adequate power to predict  the effects of  perturbations if factors are
missing from the model that are not critical in describing current conditions,
but will be critical in describing future conditions.  There  always exists the
possibility that some factor is missing,  that  some relationship that now holds
will be invalid under future conditions.   Without an experimental earth to use
for model validation, it will always  be impossible to eliminate that
possibility.

    Nevertheless, there are mechanisms for testing models and their
robustness.  Past assessments have reported  changes in the prediction of ozone
depletion (Exhibit 5-47) for the same scenarios.  These changes, in
themselves, reduces our confidence in the models.  An examination of the cause
of those changes, however, reveals that most occurred primarily as a product
of new estimates in rate constants or cross  sections from the laboratory.
While it was impossible for laboratory kineticists to predict the changes that
occured in their laboratory estimates, kineticists certainly  were not
surprised that changes occurred, nor  that they had implications on model
outputs.  Even today, after enormous  improvements in laboratory procedures,
uncertainty about kinetics and  cross-sections  still exist.  Clearly then, to
avoid surprise in changes in the estimates of  depletion in the future, we must
examine a range of rate constants and cross  sections to determine the extent
to which model results depend on a unique set  of  inputs.
                          - * *  DRAFT FINAL  * *

-------
                                5-73
                           EXHIBIT 5-47

          Calculated Ozone -- Column Change to Steady-State
               for Two Standard Assumed Perturbations
UJ
O
z
<

o
z
5
O
o
UJ
Z
o
IM
o
Q
UJ
3
O
O
 10



  5



  0



 -5



-10



-15



-20
CFC STEADY
  1974
 EMISSION
  RATE
                                             STRATOSPHERIC

                                             AIRCRAFT 20 km
      1974     1976      1978      1980     1982     1984

                   YEAR IN WHICH CALCULATION WAS MADE
                                                          1986
                                                              1988
 Calculations made  at  LLNL over 11 years showing changes in 1-D model
 results for two  standard scenarios.   Calculations used the
 photochemical parameters,  eddy diffusion functions, and boundary
 conditions current when  made.


 Source:  World Meteorological  Organization, (1986).
                         *  *   DRAFT FINAL  * * *

-------
                                   5-74
    Another source of uncertainty that  needs  to be  considered is transport.
Chemical interactions may be described  quite  well by models, but if molecules
do not move to areas they are expected  to,  the outcome of models may be
inaccurate.

    To meet these concerns,  this  section addresses  these three questions:

        (1)  How well do models  reproduce the current atmosphere?

        (2)  How dependent are models predictions on a narrow range
             of kinetics and cross-sections?

        (3)  How dependent are models on a  single formulation of
             transport?

By answering these questions, it  should be  possible to establish range of
concern that should govern judgments  about  model reliability.

Agreements and  Deficiencies  In Model Representations of the Current
Atmosphere

    The extent to which models reproduce observed concentration will have a
bearing on the confidence we place in them.   There  are a few discrepancies
between model predictions and the observed  atmosphere which lowers our
confidence in the predictions of  one  and two  dimensional models.  One is
that there is  an  inconsistency between  modelled  (using both  one- and
two-dimensional models) and  measured (by  both ground-based and satellite
instrumentation)  ozone abundances above 35 km by as  much as 30-50%.  For
some, it is particularly troubling  that this problem  occurs in an altitude
region in  which the ozone concentrations are  photochemically controlled.
Others feel that because this is  a region with  low  ozone, the discrepancy is
not that important.  Furthermore,  if the observed  ozone concentrations are
used  in radiative models of the upper stratosphere  the predicted temperatures
are higher than  measured.  Another problem  is that there are considerable
differences in the  distribution of odd nitrogen species that  are  computed by
the models of different  groups ([Exhibit 5-48]).  This is true for both
one-  and  two-dimensional  models,  with  the  exception of Isaksen's model.
This  disagreement is understandable for  the two-dimensional models given the
differences in their formulations of transport, but  is very disturbing for
the one-dimensional models in which these differences cannot  be attributed to
differences in transport.  One source for these differences in odd nitrogen
distributions are the differing treatments for the penetration  of solar
radiation  in the  Schumann-Runge  bands.  However, models  with  similar
radiation  schemes still  show significant differences  in odd nitrogen.  We do
not presently understand the source  of these different odd nitrogen
distributions.  They are important, however, since nitrogen species interfere
with chlorine catalysis of ozone destruction and also because of the
important catalytic role of the odd nitrogen species themselves  (NASA,
1986).  Except for ozone, the data base of  oxygen species is very limited and,
therefore, not adequate to critically test  these models.
                          * * *  DRAFT FINAL  * * *

-------
                              5-75
                         EXHIBIT  5-48

              Latitudinal Gradients in  Odd Nitrogen:
                      Models vs  Measurements
                   3mb SUMMER
            - 16


            20
                     -32
  -48       - 64

16 mb SUMMER
          O
                              UMS
                               -32
                                       -48       - 64

                                     30 mb SUMMER
                                                          -64
"There are considerable difference in the distribution of odd
nitrogen species that are computed by models of different groups."

Source:  NASA, (1986).
                     * * *  DRAFT FINAL  * * *

-------
                                  5-76
    Solar  proton  events  have been the classical test of ozone perturbation by
catalytic chemistry.   Simulations of the perturbation of ozone by NOx
chemistry in  the  upper stratosphere following the August 1972 event  provided
a significant  success  for model calculations.   Recently, particle  events
which penetrated only as far as the meososphere, and should  product effects
on O3 through HOx chemistry,  have been examined and  indeed  found to produce
the expected short-term effect.  There remain, however, some inconsistencies
in the magnitude of the  effects when compared  to models.   (NASA, 1986)

   Although the odd-hydrogen species  play a vital  role  in  stratospheric
chemistry, our current knowledge of the atmospheric concentrations of  the key
species  OH, H02, and H2O2 is woefully inadequate  and not  useful for
validating or invalidating models.  Only  very uncertain  measurements  of OH
have been made...  While the total column content of OH  is  in general
agreement with model predictions, most features of  the data await a
theoretical interpretation.

    Data from the LI MS  instrument has been used to  derive global  OH fields
from the measured HNO3/NO2 ratio, and  by using the production and loss
approach  with fixed  temperature,  H2O, O3, and HNO3 [Exhibit  5-49].
However,  while this  is extremely valuable given the dearth  of OH data, and
encouraging given that the results are consistent with two-dimensional  model
predictions, it must  be remembered that  we do  not  fully  understand  the
quality  of the LIMS HNO3  data above 35  km,  and the observed  HNO3 profile
above 30  km  is still  not  adequately simulated  by the theoretical  models.
(NASA,  1986).

   There  have been  recent measurements of HO2 between 16 and 34 km, and
between 35 and 60 km using in-situ balloon-borne cryogenic grab  sampling and
ground-based mm wave emission techniques, respectively.   This data
complements the earlier data which covered the altitude  range 30 to 35  km.
While there is only a  small overlap in  the altitude range  of  the  two reported
data sets, the mm emission data  is not consistent with the high
concentrations of HO2 below 35 km that have been  observed by  all of the
in-situ measurements.  The  in-situ HO2 data  is significantly higher (a  factor
of two  at 34  km and  an  order  of magnitude at and  below 25 km) than predicted
from theory,   indicating a serious  problem with  either the measurements or our
current understanding of both the total odd hydrogen budget and  the
partitioning between  OH and HO2 ([Exhibit] 5-50).   In  addition, if the HO2
concentrations are as high as  reported from the in-situ  studies  then  we
should have observed high concentrations of  species  such  as H2O2 and HO2NO2
(NASA,  1986).

Nitrogen Containing  Species

    The odd-nitrogen species considered  to be  important in the  chemistry of
the stratosphere  are  N,  NO,  NO2, NO3,  N2O5, HNO3, HO2NO2, and  CIONO2.
All of these specifies, except N,  have  either been  measured or  at  least
detected.   HCN has  also been detected but is not thought to be important  in
the photochemistry of the  stratosphere.  The recent  observations of  N2O5,
HO2NO2,
                           - *  DRAFT FINAL  * * *

-------
                              5-77



                         EXHIBIT 5-49

                Hydroxyl Radical (OH) Measurements
          12

          14

          16
          18
                     MNOyNOj RATIO
                     SOURCES AND SINKS iPyl* and Zavodvl
                     SOURCES AND SINKS Ulckmin. It ill
42
41
40
39
18
37
36 1

»5
34 5
33 <
32
31
30
29
28
                          10             100

                           HVDBOXVl MIXING RATIO Ipml
                                                      1000
"OH concentration versus  altitude.   The circles represent  all
stratospheric OH measurements prior to 1983:  filled circles were
obtained using  lamps  to  excite fluorescence and have typical
uncertainties of +30%.; open circles and crosses are results  from
two different flights of  balloon-borne LIDAR system.  The  height of
the crosses represents the altitude range of the measurements;  the
width of the crosses  present the precision  (lo).  The bar  at the
bottom of the figure  represents the accuracy estimates  for all  the
point. ... The  -.-,  - -  -, 	,  lines represent the OH profiles
inferred by Pyle and  Zavody from the HN03/N02 ratio; by Pyle and
Zavody from the sources  and sinks;  and by Jackman et al. from  the
sources and sinks."

Source:  NASA,  (1986).
                      * * *  DRAFT FINAL  * *

-------
                              5-78



                         EXHIBIT 5-50

                 HO2:  Models  versus Measurements
    35
    30
 §  25
    20
    15  —
              HO,
              44 °N
KFA-JULICH
HELTEN ET AL. 11984. 1985)
    •  9-26-80
            D • 10-21-80

            A A 9-9-83

            O 4 9-20-83
           i   i  i i  i i I i
                                                           i   i  i i  i i
     10
       - 13
                     10
                      - 12
        2-D MODEL/RANGE
        WINTER - SUMMER
        DAYTIME AVERAGE
                          10-"
                      MIXING RATIO
10
                                                      -10
                                                                   10
                                                                    -9
Observations of  H02 are indicated by cross-hatched triangles,
diamonds, and  squares.   The dashed line  represents a typical
altitude profile for H02 as calculated by  Prather.  The shaded area
represents the range of H02 profiles calculated in 2-D models.  Note
that the measurements are significantly  different than those
predicted by the theory.

Source:  NASA,  (1986).
                        * *  DRAFT FINAL

-------
                                  5-79
and CIONO2 are extremely  important as they will allow vital tests  of certain
facets of stratospheric photochemistry,  i.e. the coupling between  the
odd-nitrogen, odd-hydrogen, and odd-chlorine families in the lower
stratosphere.   As stated earlier, the availability of satellite data for N02,
HNO3, and N2O has been a major achievement  allowing us to quantify our
knowledge  of the temporal  and  spatial variability of these species.   (NASA,
1986).

   There is a large  balloon-borne data base for NO, NO2,  and HNO3.  This
data exhibits considerable  scatter,  and  it is  difficulty to determine changes
in the  vertical  distribution  with either season or latitude.   It is still not
totally clear whether the scatter  is due to atmospheric variability  or to
experimental  inaccuracies.   Recent intercomparisons of balloon-borne
techniques for  each  of these species has augmented and  greatly added to the
value of the existing data,  and  has led to an improved knowledge of the
detailed profiles of these species at mid-latitudes.   However,  it should be
noted that there are still some  unresolved  issues regarding data accuracy,
particularly between the  remote sensing  techniques for NO2, and  between the
in-situ chemiluminescence instruments for NO in  the mid-stratosphere.
(NASA,  1986). .

   Direct measurements of  the NO to NO2 ratio are also available from
balloon-borne  instruments,  and while consistent with theoretical predictions,
they are not really of adequate  accuracy to critically constrain them.
(NASA, .1986).

   Data on diurnal variations of NO2 from  balloons  and satellites,  while  of
considerable  use,  are  not yet sufficiently accurate  nor temporally detailed
to definitively test photochemical theory.  Ground-based, balloon, aircraft,
and satellite data  have all  ben  used to address the variability of NO2 at
high latitude in winter in much  more detail  than previously possible.  These
data have greatly quantified our understanding of the distribution of  N2O5
and its  role in the Noxon "cliff"  phenomena.   Balloon data have shown that
the cliff region  of "missing"  NO2 occurs at altitudes from about 15-30  km.
Theoretical studies have  revealed that the Noxon "cliff"  is primarily due to
meridional  transport processes  and  combined photochemical effects  in the
presence of large-scale waves in the vicinity of the polar night region (NASA,
1986).

    Satellite observations of NO2 in the  polar night region have quantified
our knowledge of  thermospheric sources of odd-nitrogen  to the  stratosphere.
They demonstrate that while thermospheric odd-nitrogen  cart be transported
down  to  the upper stratosphere, and may be important on a local scale,  it  is
probably unimportant  to the  global budget.  (NASA, 1986).

   NO3  has been observed  both by ground-based and balloon-borne
instrumentation,  revealing  an inconsistency with photochemical theory over
much  of  the annual  cycle.   (NASA,  1986).
                          » * *  DRAFT FINAL  * * *

-------
                                  5-80
   While the  data for HNO3 in the low and mid-stratosphere is in good
agreement with the theoretical models, our understanding  at both high
altitudes (above the main layer) and at high latitudes, has been shown to be
incomplete.   In the latter case, the existence of  an  additional  polar source
of HNO3 has been suggested.   Studies of the latitudinal variation of HNO3,
both from aircraft and  satellite observations,  have led to some important
insights into transport  processes,  showing in some detail how  advection and
dispersion compete in tropical  latitudes.   (NASA,  1986).

   The  combination of available balloon and satellite data fro NO, NO2 and
HNO3 has greatly improved our knowledge of total odd-nitrogen.   This
information is particularly important both  because it enables us to better
understand the budget of the  family as a whole,  and because odd-nitrogen
plays a major role in model predictions of ozone perturbations.   It is
important to  emphasize  that the odd-nitrogen  derived from the data is in
marked  disagreement with two dimensional  model  predictions below about 30
km,  with the  exception of Isaksen's model.  On the other hand these same
models  seem to predict  the odd-nitrogen  concentrations above 30 km quite
well.  This suggests that boundary conditions or low-altitude  sources of
odd-nitrogen are not properly treated in the models.  (NASA, 1986).

Chlorine Containing Species

    The odd-chlorine species of importance are Cl and CIO (radical species),
HOCI and CIONO2, (temporary reservoir species),  and HCI (sink species).
(NASA,  1986).

   While there are only three  measurements of the vertical distribution  of
Cl over a limited altitude range, there is a relatively  large data base for
CIO, with the majority of the measurements being made with the balloon-borne
in-situ  resonance fluorescence and ground-based millimeter wave emission
techniques.   Additional data have  been obtained  using two remote sensing
balloon-borne instruments; one which sensed the millimeter wave emission; the
other using infrared  laser heterodyne absorption techniques.  The  mean  of the
resonance fluorescence  data  and the ground-based mm data are in good
agreement.   However, there is one substantive difference between the two  data
sets and that is the degree of variability observed in CIO  concentrations.
The in-situ  resonance fluorescence data  indicates significant variability in
the CIO concentration profiles.  With  the exception of the two  apparently
anomalous profiles in July 1976 and July  1977, the vertical profiles span a
range of approximately a factor of 4 between  25  and 40  km (Exhibit 5-51).

    In contrast, the ground-based mm wave data, which has been obtained
between January 1980 and December  1984, with observations taken  at 20,  32, and
42°N  in winter and summer,  shows less than a +20% variation about the
mean  above  30 km.  While some of this difference may be  ascribed to a lower
reported precision  in the in-situ data this is  clearly not the complete
explanation.   Theoretical models predict  a  variability in  CIO that is
consistent with the ground-based data, but not with the in-situ data.
However, we  should remember that the in-situ measurements are made within a
                                 DRAFT FINAL

-------
                               5-81


                          EXHIBIT 5-51

            Variability of Observed CIO Concentrations
          O 28 JULY
          Q  8 DECEMBER
            14 JULY
            20 SEPTEMBER
            25 OCTOBER
             2 DECEMBER
            16 NOVEMBER
            15 JUNE
             ^ AUGUST
        _ O26 SEPTEMBER
          • 15 SEPTEMBER 1984
            (REEL DOWN)
                                                              10-'
                            CIO VOLUME MIXING RATIO
"With the  exception of the  two apparenty  anomolous profiles  in July
1976 and July 1977, the vertical profiles  [of CIO] span  a  range of
approximately a factor of 4 between 25 and 40 km."

Source:  NASA, (1986).
                      * *  *  DRAFT FINAL   *

-------
                                  5-82
few minutes with high vertical resolution (less than I  km), whereas the
ground-based measurements  take  several hours and are averaged vertically over
approximately 7 km.  The mean of the in-situ and  ground-based data agrees
with one- and two-dimensional model predictions to within a factor of two at
25 km and better at higher altitudes ([Exhibit] 5-52).  The CIO data from
the balloon-borne millimeter emission instrument are in good agreement with
the other measurements,  as are the reevaluated balloon-borne  infrared laser
heterodyne data.  The  diurnal variability of CIO has been  measured by using
both the ground-based  and balloon-borne millimeter wave techniques and  is in
reasonable agreement with model predictions,  although  the  observations
indicate a somewhat slower rise than theoretically expected.  The existing
CIO data base is not yet  adequate to establish  seasonal and latitudinal
variations or long-term increases  which are all predicted by theoretical
models.   (NASA, 1986).

   The  few measurements of Cl in the  upper stratosphere (40  km) are limited
yet consistent with the observations of CIO and the photochemical
partitioning between Cl and  CIO.  The measurements of ethane  (C2H6) in the
lower stratosphere  are  an indicator  of  atomic chlorine concentrations (the
major stratospheric loss process for C2H6 is its reaction  with Cl) and  are
reasonably consistent with the CIO measurements and  the theoretical
predictions of Cl.   (NASA, 1986).

   Direct evidence  for the presence of CIONO2 in the stratosphere  has now
been obtained with  balloon-based  infrared absorption measurements  being made
in a second spectral region.  The vertical  distribution determined  from these
observations  is in reasonable agreement with model calculations.  In
addition,  CIONO2 has been observed during the  recent space shuttle flight of
the ATMOS instrument  (a high resolution infrared  interferometer  used in  the
absorption mode).  These direct observations are supported by the observed
diurnal  behavior of CIO which is  thought to be due to the formation and
destruction of CIONO2.   (NASA, 1986).   There have been no observations to
date of  HOCI.  (NASA, 1986).

   Several different remote sensing techniques for HCI were carefully
intercompared during a series of  balloon intercomparisons (BIC) carried out
in 1982  and 1983.   The  vertical concentration profile of HCI can now be
measured to  15% accuracy with confidence.   These observations complement the
earlier measurements of the  vertical distribution  of HCI which  were largely
made  using remote  sensing spectroscopic techniques.   The earlier measurements
had a significant amount  of scatter  which could have been  due to either
atmospheric variability  or measurement inaccuracies.   As the cause  of  the
scatter  in the observations was, and still is,  unknown it is difficult to
meaningfully  compare them to model calculations.  It should be noted,
however,  that the mean HCI profile obtained from the  BIC campaigns is quite
similar to the mean of the earlier data and  has been  compared to the results
of a two-dimensional model.   The observations and theory  agree very well at
both 20 and  40 km, but the predicted  HCI  is lower  by a factor of two at 30 km
([Exhibit] 5-53).   However,  the  discrepancy cannot be considered  to be
very significant at  present,  considering the lack of data on HCI variability
and on  simultaneous observations for  HCI,   CIONO2, and  CIO.   (NASA,  1986).
                              *  DRAFT FINAL  *

-------
                                 5-83
                           EXHIBIT 5-52

        CIO Vertical  Profiles:   Models  versus Measurements
     40
  g   35
 o
 <   30
     25
MODEL PREDICTIONS. 30 °N

   	  SUMMER

   •	  WINTER


                          I
                                   MEASUREMENTS

                             IN-SITU RESONANCE FLUORESCENT

                              • WEINSTOCK ET AL. (19811
                              • BRUNE ET AL. 11985)

                             BALLOON-BASED REMOTE

                              -fr1 WATERS ET AL. 11981)
                             <
-------
                              5-84



                         EXHIBIT 5-53

                 HCI:  Models  versus Measurements
     50
     40
  |
  O
  D
  <  30
     20
     15
                 WMO
                            HCI
                            (MODEL)
                                                     BIG-2
                                          HCI + CIONO2
                                          (MODEL)
              10-1
                10-9
HCI VOLUME MIXING RATIO
"The observations and theory [for HCI concentrations] agree very
well at both 20 and 40 km, but the predicted HCI is  lower by a
factor of two at 30 km.  However, the discrepancy cannot be
considered to be very significant

Source:  NASA, (1986).
                       * *  DRAFT FINAL  *

-------
                                  5-85
   Total column measurements  of HCI from the surface  (including the
tropospheric component) can be determined with  a  precision approaching 5%.
Significant spatial and temporal variability in  the column  amount of HCI in
the stratosphere (observed from an  aircraft platform) has been established.
Column  values  for the total HCI seen from the surface  show even greater
temporal variability.  The expected trend in stratospheric HCI  has  not yet
been observed  and has presumably been masked  by this  variability, some of
which may have been due to recent  volcanic eruptions  such  as  El Chichon.   It
should be noted that the observed long-term  trend of HF is compatible with
theoretical predictions.  (NASA, 1986).

   There  have  been very few  measurements of the  total amount of chlorine in
the stratosphere.   However, the few measurements that do exist appear to be
consistent with calculations.   (NASA,  1986).

   The  data  base for the vertical profiles of the  halocarbon  source gases
has expanded over the last few years.  There is now general  agreement between
calculated and  observed profiles for  these gases.   Some differences  remain
but these are most likely due  to imperfections in the modeling of transport
processes.  The  least satisfactory  agreement  occurs with CH3CI,  a  species
removed predominantly by reaction with OH.  However, CH3CI  is among the
halocarbons presenting the greatest  measurement difficulty.   (NASA,  1986).

    In summary, models  can  reproduce many, but not all atmospheric
measurements.  Uncertainty  in  the reliability of atmospheric measurements
hampers  this effort.   Nevertheless,  the cases in which models  do  not reproduce
apparent observations  lowers our confidence in the predictive ability of
models.

Uncertanties in Chemical  and Photochemical  Parameters

    Clearly there are  uncertainties in the chemical and photochemical rate
parameters and in the mechanisms  involved  in the atmospheric chemistry.
They are one of the major factors in  limiting  the  accuracy of model
calculations of  species concentrations and ozone perturbations in the
atmosphere.  Most  of the changes  in the  predicted ozone depletion  due to
chlorofluoromethanes that have occurred in  recent  years  have resulted  from
changes in the values of kinetic parameters used in model calculations
(World Meteorological  Organization,  1986).

    The uncertainty in the kinetic parameters for the key atmospheric
reactions  has been reduced greatly over the  last 10-15  years due  mainly to
the rapid development of the techniques used for the direct measurement of
radical  species in the  gas phase and for  investigation of their reaction
kinetics.  Whereas 20 years ago the rates of  most  radical-molecule reactions
were only known to within  a factor of 10, today the room temperature rate
constants of atmospherically important reactions of this type can  be measured
within an accuracy of + 10%.   Moreover  the number of  reactions for  which
good  kinetic data  are  available have increased tremendously.  The consistency
in the experimental  measurements gives confidence in  the data  base.  There
remain  problems  in reaction  rate theory which is not able to explain some of
                          * *  *   DRAFT FINAL  » * *

-------
                                   5-86
the observed  temperature and pressure  dependencies.  Although there  is
improved  reliability of the data  it should be recognized that the errors in
the rate coefficients  increase as the temperature diverges  from room
temperature and that certain reactions e.g.,  radical + radical  reactions are
intrinsically more difficult to study and  consequently are always likely to
carry more  uncertainty  than straightforward  radical + molecule reactions
(World Meteorological Organization,  1986).

    Difficulties  also arise in the study of very slow reactions between
radicals and molecules,  due  to  complications such as those arising  from
heterogeneous effects (World Meteorological Organization,  1986).

    The uncertainty  in  the  rate coefficients for atmospheric  reactions
results  primarily from systematic errors  arising from the chemical  systems
and the techniques used for their determination rather than measurement error
of a  statistical nature.   Consequently it is  not straightforward to  assign
uncertainties  to preferred values given  in an evaluation.   Errors quotes in
the NASA or  CODATA  evaluations are assessments  based on such  factors as the
number of independent  determinations made and the number and reliability of
the different  techniques employed.  Furthermore  in most cases,  the
probability of an error  of a given magnitude falls  off more slowly  than  a
normal Gaussian function  (World Meteorological Organization, 1986).

    For the key elementary  reactions identified as  being important for the
stratosphere many of which  are radical  + radical reactions, the prospect of
reducing  uncertainties  in the rate coefficients to  less than + 10%
cannot be considered realistic.   Some reduction in  uncertainty can be
expected from further temperature and  pressure dependence studies, and  a
further understanding  of product channels and reaction mechanisms  can be
anticipated  in the  future  (World Meteorological Organization, 1986).

For a Range of  Testable Uncertainties,  Models Project Depletion

    The depletion predicted by earlier models uses a single set of input
values for kinetics, cross sections,  and transport.  As discussed  above,
significant uncertainty exists  about  the value of  these inputs.  Past changes
in these inputs  that occurred with improved laboratory experiments have
significantly changed estimates of ozone depletion.  Given past history,  it
makes sense to test a range of kinetic and cross  section inputs, rather than
relying solely on best case estimates.

    Fortunately, such uncertainty analyses  have been conducted by  Stolarski
and Douglass  (1986) and  Grant  et al. (1986).  These analyses  recognizing  the
uncertainties in reaction rate coefficients, absorption cross-sections,  solar
fluxes and boundary conditions, have  sampled from  a range  of different inputs,
using the probability of the value for each input  as a selection criteria.   By
doing this many times and for a variety  of emission scenarios, these modelers
have essentially explored the robustness of the model predictions  to these
uncertainties.  The approach to doing the uncertainty analyses is  illustrated
in Exhibit 5-54.  None of the uncertainty analyses included higher levels  of
C02.   Different sampling strategies have been used, and different  CFC levels
analyzed.


                          * * *  DRAFT FINAL  * *  *

-------
                                        5-87




                                  EXHIBIT  5-54

              Logical Flow Diagram for Monte  Carlo  Calculations
    NASA Codate Panel
                              For Given Fluxes of
                               CFCS, N2O, CH4
    Range & Probability
of Kinetics & Cross Sections
 Sampling Strategy
 (n cases where n
 was large enough
   to develop a
reasonable measure
of the distribution)
Run  1 Dimensional
  Model n Times

                                                                   Output Range of
                                                                  Depletion Nufnbers
                                                                   Screening System
                                                                  (Must Meet Current
                                                              Atmospheric Measurements
                                                                    fo X Variables)
                        Probabilities of
                       Various Levels  of
                  Depletion without Screening
                                Probabilities of
                               Various Levels of
                          Depletion with    Screening
                           * * *  DRAFT FINAL  *  * *

-------
                                  5-88
    Exhibit 5-55 is included in  order to give readers a better feel for the
nature of this analysis;  that is, how are uncertainty ranges derived for rate
constants.  It shows a "histogram of all of the individual  measurements of
the room  temperature rate coefficient of the  reaction O *  NO2 - O2  + NO by
three different groups (Bemand  et al.,  1974; Slanger et a!., 1973 and  Davis
et al., 1973).   These 62 data points  form a distribution that is fit well by
either a  Gaussian or a log-Gaussian  distribution  with parameters given by
                                                           -12
DeMore et al.  (1985).   These are a central value of 9.4 x 10    and a  one-
sigma uncertainty of +10%.   Also shown in the  [exhibit]  is  a vertical
line  representing the mean  value which  is  normally  used  in  a photochemical
model.  This  is not a complete representation of the known  information on
this  reaction" (Stolarski and Douglass,  1986).

    In the uncertainty analysis,  rather than using  a single value for  the rate
coefficient described above,  different values are used.   Some modelers used
Monte-Carlo method, to choose coefficients, others used  other sampling
methods.   ..."this  illustration.   Other reactions have only been measured  a
few times (or  only the mean  of the  actual  runs  were reported) and  the
measurements  may  disagree with one another by substantial  amounts.   For these
cases, it  is not obvious how  much improvement  is obtained by  attempting  to
represent an  entire distribution  (Stolarski and Douglass,  1986).

    Most of the reactions used in the Stolarski  and Douglass model  used the
chemical  kinetic and cross section data  and the 1-sigma uncertainties from
evaluation 7 of the  NASA Panel  In Data Evaluation  (DeMore  et  al.,  I985).   For
O2 dissociation, the  1-sigma  uncertainty is assumed to be 20% with  the
Schumann  Runge bands and the Herzberg  continuum treated independently.   The
1-sigma uncertainties in the boundary conditions and assumed  dissociation
rates are given in  [Exhibit  5-56] .

    Stolarski  and Douglas then applied  the Monte-Carlo technique to this
model  ...to generate a total of 329 cases  of  varied  inputs for calculation of
the atmosphere  without fluorocarbons, with enough  fluorocarbons to give  a
reasonable steady-state representation of the present atmosphere and  with the
sequence of increasing fluorocarbon  perturbation to investigate the  model
response  to large chlorine amounts.   [Exhibit 5-57] shows  the calculated
change in  the column content of ozone compared to the no fluorocarbons case
as a function  of the injected fluorocarbon  flux shown in  units  of the present
flux  of fluorocarbons 11 and 12  (PFF).   The solid curve, labeled the  base
case, gives the results using the mean  values  for  each input parameter.   The
dashed curve is  the mean ozone depletion  obtained  from the 329 cases.  The
mean curve is significantly more linear  than the base case curve (Stolarski
and Douglass,  1986).

    Stolarski and Douglass  (1986) used  current observations to screen their
model runs.  None of the 329 runs was able to satisfy  all the  constrainst
posed by all the observations.   Exhibit 5-58 shows  the results for the
equilibrium values of various levels  of CFCs alone.  For  current emissions,
the most common depletion is 3%. For 1 sigma deviation (64% probability),
depletion always occurs.   Exhibit 5-58  shows the distribution  of results for
various levels of CFC alone,  with the shaded areas  being  cases that passed all
                                 DRAFT FINAL   * * *

-------
                            5-89



                       EXHIBIT 5-55

         Histogram of Measurements for a Rate Constant
   7.5     8.0    8.5   9.0    9.5   10.0   10.5  11.0   11.5 12.0x10~12

       RATE COEFFICIENTfCM3 MOLECULE"1 SEC"1)
Histogram of individual  data points for the rate coefficient of the
reaction 0 + N02 	-*• 02 + NO.  The measurements have been fit to
a smooth curve that forms a probability distribution to be used in
Monte Carlo calculations.

Source:  Stolarski and Douglass, (1986).
                    * * *  DRAFT FINAL

-------
                                  5-90
                              EXHIBIT 5-56

                Recommended  Rate  Constants and Uncertainties
                        Used  in Monte Carlo Analyses
Experimental
Sensitivity Factors Uncertainty
No.
1.
2.
3.
4.
5.
6.
Reaction

CIO
Cl
OH
OH
OH
OC1

+ 0 =
+ CH4
+ HC1
+ HN03
+ HN04
D) + M

Cl + 02
= HC1 + CHS
= Cl + H20
= H20 + N03
= H20 + 02 + N02
= 0(3P) + M
+16
+0.60
-0.48
+0.56
-0.51
-0.16
+0.60
-16
+0.68
-.046
+0.79
-0.56
-0.33
+0.63
Ave
+0.
-0.
+0.
-0.
-0.
+0
64
47
68
53
25
.62
Factor
1
1
1
1
2

.43
.16
.32
.30
.20
1.32
Error
Contribution
Factor
+0.
-0.
+0.
-0.
-0.
+0
23
07
19
14 •
20
.17
7.   0(aD) + N20 = 2 NO
                = N2 + 02        -0.51   -2.00   -1.26      1.30         -0.33

8.   CIO + N02 + M = C10N02  + M  -0.31   -0.37   -0.34      1.56         -0.15

9.   02 + hv (S-R) = 2.0         -0.58   -0.85   -0.72      1.40         -0.24
     Rate constants recommended  in  evaluation 7 of the NASA Panel In Data
     Evaluation.   Also shown are the  1  sigma  (6) uncertainty bands.

     Source:   World Meteorological  Organization, (1986).
                          * * *  DRAFT  FINAL  * * *

-------
                             5-91
                        EXHIBIT 5-57

                      Monte Carlo Results:
            Change in Ozone Versus Fluorocarbon Flux
        -10
      3  -20
      O
      o
      o
      z
      D.  -30
      O
      U
        -40
        -50
                        1           2
                      FACTOR TIMES PRESENT
                       FLUOROCARBON FLUX
The calculated change in total  column ozone as the fluorocarbon flux
increases.  The sold line shows the base case, which  uses the mean
value for  each input parameter.  The dashed line shows  the mean
depletion  from the 329 model runs.  The vertical bars show the
one-sigma  uncertainty limits.

Source:  Stolarski and Douglass,  (1986).
                      *
                           DRAFT FINAL  * * *

-------
                                 5-92
                            EXHIBIT 5-58

                          Monte  Carlo Results:
               Change  in  Ozone Versus Fluorocarbon Flux
-60   -40  -20   0
 1O] COLUMN/O] COLUMN 1%)
-80  -60   -40   -20    0
     SO) COLUMN/O] COLUMN (*l
-80  -60   -40   -20    0
     SO) COLUMN/O] COLUMN <%l
-*>   -40   -20   0
SO, COLUMN/Oj COLUMN l%l
    -60   -40   -20    0
     SO) COLUMN/0) COLUMN l%l
-80   -60   -40  -20    0
     SO, COLUMN/0) COLUMN !%l
Frequency distributions of Monte  Carlo Model  results for changes in
total column ozone  as a function  of fluorocarbon flux.  The  shaded
areas show the cases  for which  the NO, N02, and  CIO concentrations
all  fell within  the rage of measurements at 25 km.

Source:   Stolarski  and Douglass,  (1986).
                              DRAFT FINAL  » * *

-------
                                   5-93
the screen except 03 in the upper atmosphere.  Note that the screens reduce
the variation of results substantially for 2x -CFC from +1 to 60% depletion is
reduced to 24 to +1 %.  Exhibit 5-59 shows the distribution of runs for
equilibrium concentrations of 3.5 x current CFC, 1.2 x N20, and 2x CH4.  Note
that distribution is weighted towards higher values more than lower, and that
the worst cases are eliminated if only the cases meeting the screens are
used.  Exhibit 5-60 represents the change in depletion that would occur for
moving from 3.5x CFC to 3x, from 3 to 2.5, etc.  This exhibit shows that any
reduction in emissions of one half of current emissions (that is, from 3x to
2.5x, from 2.5x to 2x, etc.) is likely to lead to less ozone being depleted.
The distributions of changes in the ozone column look quite similar; for
example, going from 3.5x current to 3.Ox current (panel e) produces very
similar distribution as going from 3.Ox to 2.5 (panel d).   The average amount
of depletion prevented by reducing emissions by 0.5x current emissions is
about 5%.  The benefits of a shift in 1/2 current CFC flux are relatively
stable across the absolute level of CFCs.   Even in the case of going from 1.5
current emissions to current emissions (Panel A), the mean gain in column
ozone is close to 4%.  And the distribution shows that there is -little chance
that a 0.5x current cut in CFCs will not produce a thickening of the ozone
column.  In addition, the common shape of the distribution shows that expected
change in column ozone is relatively independent of the absolute depletion
across the range of input uncertainties Stolarski and Douglass investigated.

    Grant et al. (1986) performed two uncertainty analyses.  One used 15 ppb
Cl, 2 x CH4/ 1.2 x N20; and NASA/CODATA kinetics distribution similar to
Stolarski and Douglas, and one that assumed that all uncertainties in kinetics
were reduced to 10%, the level the panel has suggested is the practical limit
for improving kinetic estimates.  Exhibit 5-61 shows the changes in ozone with
altitude due to the uncertainty tests (as a percent).  Exhibit 5-62 shows the
results for unscreened data.  When only cases meeting all observational
screens are used, uncertainty is reduced to 7.7 + 6%.  It is relevant to note
that at least eight sets of kinetics and cross sections adequately describe
all the apparent observations of the atmosphere.

    Exhibit 5-62, the distribution of unscreened cases,  reinforces the
conclusions of Stolarski and Douglas, that the probability of a depletion of
much greater than the "central or best" case (e.g., recommended kinetic
values) is substantially higher than much lower.  Exhibit 5-63 shows the
cumulative probability of various depletion levels.  Approximately 10 out of
100 cases were at around 0.2 depletion.  For an approximately equal number of
cases depletion was equal to or greater than 20.2%.  Exhibit 5-64 demonstrates
that even if a low level of uncertainty is achieved about kinetics (that is
+10%), uncertainty will still exist about outcomes.

    The results of the two studies indicate that across a wide range of CFC
scenarios and current uncertainties about  kinetics, cross section and solar
flux, average global depletion can be reasonably anticipated for cases in
which CFCs grow.  The studies also demonstrates that the possibility for
depletion levels significantly larger than the central case is much higher
than for values substantially lower.  Stolarski and Douglass also demonstrated
                                 DRAFT FINAL  * * *

-------
                           5-94
                       EXHIBIT 5-59

                     Monte Carlo Results:
           Ozone Depletion for Coupled  Perturbations
en
LU
en
<
o
LL
O
CC
LU
ffl
   20
   15
   I0
   10
               I
              3.5 x PRESENT
          FLUOROCARBON FLUX
           PLUS 1.2xN2OFLUX
           PLUS 2.0 x CH4 FLUX
    -80     -
                60     -40      -20        0
                A03 COLUMN/OS COLUMN (%)
+20
Frequency distribution of Monte Carlo Model results for changes in
total column ozone for 3.5 times the present fluorocarbon flux, 1.2
x N20 flux,  and 2 x CH4 flux.  The shaded area shows the cases for
which the NO, N02, and CIO concentrations all fell within the range
of measurements at 25 km.

Source:  Stolarski and Douglass, (1986).
                         DRAFT FINAL  * *

-------
                                   5-95
                              EXHIBIT 5-60

             A  Monte Carlo Distribution of  Column Ozone  Changes
                       for Changes  in CFC Production
DU
50
ft 40
CO
jj 30
L^.
O
i 20
10
0
_j

-
-
-
i
T_
Pi


r ,


J


5 0 '5
1 1 1
-
-
-
i ^— 'Hn L
ou
50
ft 40
CO
o 30
O
i 20
10
0

-
-
-
r
i i i i
J]
s
V
%
U
\ -

K
1 1 1 HI n 1
10 15 20 25 -505 10 15 20 2!
A03 COLUMN/03 COLUMN (%) A03 COLUMN/03 COLUMN (%
                                                          60

                                                          50
                                                       ft 40
                                                       CO

                                                       £ 30
                                                       O
                                                       520
                                                  10

                                                   0
                                                                            ffl i-i nl
                                                          -5   0   5   10   15   20  25
                                                            A03 COLUMN/OS COLUMN (%>
   60

   50
ft  40
to
o  30
O
O
z
           20
           10
                        U
                                  60

                                  50
                                        {2 40
                                        CO
                                          30
                                  20
                                  10
                                                i    i    i

           -5   0   5   10   15   20  25
             A03 COLUMN/03 COLUMN (%)

                         d
                                   -5   0   5   10  15  20  25
                                    A03 COLUMN/03 COLUMN (%)
                                               e
Panel e shows  the  distribution of change in column ozone  for Monte Carlo cases
for going from 3.5x  current CFC to 3.Ox current.   Panel d shows 3x to 2.5x; c
from 2.5x to 2x, b from 2x to 1.5x, a from 1.5x to Ix.  For example in Panel
e,  the majority of cases produce a change in column ozone of 5-10% for going
from 3.5x CFC  to 3x  current CFC.
                                DRAFT FINAL

-------
                                5-96
                           EXHIBIT 5-61

                        Monte Carlo Results:
                    Changes  in Ozone by Altitude
                But lint CM*
                RccoiMndtd Unctrtilnty Ftcton
                101 Uncertainty P»ctor«
-80   -70   -
60   -50   -40   -30   -20   -10
             Ozone     cm** (-3)
10
 Results from Monte Carlo  analysis  of uncertainties regarding rate
 constants.  Solid line  shows  calculated changes in ozone by altitude
 for mean values of input  parameters.   Dashed lines show calculated
 changes for recommended uncertainty factors.   Dotted lines show
 calculated changes for  uncertainties which have been reduced to ten
 percent.  Perturbations are:   15 ppbv Clx,  2 x CH4, and 1.2 x N20.

 Source:  Grant, Connell,  andWuebbles,  (1986).
                      * * *   DRAFT FINAL  * *

-------
                                           5-97
                                     EXHIBIT 5-62


                                   Monte Carlo Results:

                Changes  in Ozone by Column and Altitude, Unscreen Data
           Distribution of Percent Ozone Change
                                                              Distribution of Percent Ozone Change
-
n
§ «
tr
01
o w
o
o
•* ?4.
Number from
0 • •








^^








P

















•K\>CvN\\NXS^N
I
1
1
1
l
I


1
1
1
1
l
1

\

y
/,
y

S
\








1
                                                    c
                                                    D
                                                    cr









||
1
1
1
1
1
I
i
I
1
1
I
1
                Percent Change at 25 km
                                                          -3.0 -4.3 -4.0 -3.3 -3.0 -2.3 -2.0 -1.3 -1.0 -0.3 0.0 0.3

                                                         TlM> I.OE'OO!

                                                                    Percent Change  at 32 km
           Distribution  of Percent Ozone Change
                                                   Distribution of Percent Ozone Change
§  «
tr
o  3?
i

o
o
—  24

E
1
1
1



1

1
1
.1
i
1
1
1
1
1
1
1
1


I
,


m
c
3
IT
„
01
0
o
0
o
i.
l_
u
1
2
-8.2  -7.7  -7.2 -6.7 -6.2  -S.7  -3.2 -4.7 -4.2


     Percent Change at 40  km
1
1

1

I
1
1
1
1
1
1
1
1
1


                                                           -33.2 -30.2 -23.2 -20.2 -13.2 -10.2 -3.Z -0.2  4.»  9 1


                                                                 Percent Change for Total Column
        Frequency distributions  of Monte Carlo model  results for  changes  in

        ozone.   Top left  panel  shows  results  for 25 km; top right  for 32  km;

        bottom  left for 40 km;  and bottom right for total  column  ozone.

        Perturbations are:  15  ppbv Clx, 2 x  CH4, and 1.2  x N20.


        Source:   Grant, Connell,  andWuebbles,  (1986).
                                *  * *  DRAFT FINAL
                                                         *  *

-------
                                       5-98





                                  EXHIBIT 5-63


                    Monte Carlo Analysis  With  the LLNL 1-D Model
     100
1
4->
0
3

E
D
O
      90 -


      ao -
Scenario analyzed:


  15  ppbv C1X
  1.25 x present cone, of N2O
  2 x present cone, of CH.
      1O -
               /\
                                 >2O
                                                         >O
                   TOTAL COLUMN OZONE DEPLETION  (PERCENT)
       SOURCE; Adopted  from K.E.  Grant et al;  (1986)  "Monte Carlo Uncertainty Analysis of
              Stratospheric Ozone in Ambient and Perturbed Atmospheres",  LLNL.'
   The  cumulative probability of ozone depletion cases unscreened for meeting
   current  atmospheric measurements  is shown  above.   For example, the chances of
   depletion greater than 10% is about 50%.   The dashed line indicates the value
   of depletion predicted for current kinetics  and other inputs.
                                     DRAFT FINAL  * * *

-------
                                               5-99
                                          EXHIBIT 5-64

                                       Monte  Carlo  Results:
                                   Changes in Ozone  by  Altitude
        Dlstr of Ozone  X Change for  10X Uncertainty
                                                            Distr of Ozone X Change for 10X Uncertainty
c
I
"
40
n
11







. rn.rn V






7
/^
1

X









I
^
/






0
48

n
§ 40.
cc
„
D
0 M'
1
O
o
-« 24-
E
o
c.
*" IS-
C.
0>
i
i ••
0 •
1
1
1
1
1

1
1

I

I
1
1
1
1
1
1
1
1
m


1
       -37.0 -J2.« -Z7.0 -rc.O -17.0 -18.0 -7.0 -J.O  3.0  «.0 13.0 l«.0

                Percent Change at 25 km
 -S.O -4.3 -4.0 -3.3 -3.0 -2.5 -2.0 -1.3 -1.0 -O.S 0.0 0.3
Tlmi 1.0FKJ01
           Percent Change  at 32 km
       Distr of Ozone X Change for  10X Uncertainty
a
E
3
1
1

1
1
1
1
1
1
1
1
1
1
1
1

vA &7\
  Dlstr of Ozone X Change  for 10X Uncertainty
                                                     c
                                                     3
                                                     n.
       -8.7 -B.2  -7.7  -7.2  -6.7 -6.2 -3.7 -3.2 -4.7  -4.2

                Percent Change at  40 km
1
1

„ I
'/,
//
\
\
\
\
\
\
1
1
1
1
1
1
\

  -35.2 -30.2 -25.2 -20.2 -13.2 -10.2 -3.2 -0.2  4.8  9.8

       Percent Change for Total Column
            Frequency  distributions  of Monte Carlo model results for changes in
            ozone,  when uncertainties in  input  parameters are reduced to  ten
            percent.   Top left panel shows  results for  25 km;  top  right for 32
            km;  bottom left  for 40 km; and  bottom right for  total  column  ozone.
            Perturbations are:  15 ppbv Clx, 2  x  CH4, and 1.2 x N20.

            Source:   Grant,  Connell, andWuebbles, (1986).
                                      *  *  DRAFT FINAL  *  *

-------
                                   5-100
that the change in column, for a fixed reduction in CFCs,  will vary very
little.  Thus, even though there is uncertainty about the  absolute depletion
for any given level of CFCs, there is less uncertainty about how much the
column will thicken for reductions in CFCs.  This means that the change in
column ozone obtained by cutting back emissions is less uncertain than the
absolute depletion caused by any level of CFCs.

Uncertainty Analysis of Two Dimensional  Models

    To date, the computation costs of performing uncertainty analysis for two
dimensional models have prevented such studies from being  undertaken.
However, because two-dimensional modelers have used very different transport
mechanisms, in effect, creating an opportunity to compare  the outcomes of
their runs to test the sensitivity of predictions to the transport
uncertainties.  Exhibits 5-21 shows the results of the AER and MPIC models;
clearly transport makes some difference.   The quantity of  depletion varies at
different latitudes, with the AER model predicting greater depletion further
from the poles and the MPIC model predicting greater depletion towards the
equator.  Both, however, have a gradient of increasing depletion as one moves
away from the equator.  Thus, at least for the differences in transport
between these models, the results robustly predict depletion.

Missing Factors

    It is difficult to deal with the possibility of missing factors..  The most
likely "missing factor" is heterogenous chemistry.  Several reactions have
been suggested in recent years for such reactions and several of the
hypothesis put forth to explain the Antarctic ozone hole include them.
Clearly the possibility of missing factors exists.  It is, and always will be,
impossible to rule out "missing factors".  It is imperative, however, to try
to encompass all factors that suggest themselves as possibly important.
Consequently, it is important to undertake research into the possibility that
aerosols in the stratosphere are providing a place for heterogenous reactions
to occur.  The search for other potentially missing factors should also be
continued and intensified.

THE IMPLICATIONS  OF OZONE  MONITORING  FOR ASSESSING  RISKS
OF OZONE MODIFICATION

    Monitoring can do several things:  it can validate or  invalidate models;
it can suggest a path for future stratospheric change; and it can serve as a
warning system that helps initiate action.  Monitoring of ozone is currently
done with a variety of ground-based and satellite systems.  This section
analyzes research based on ground-based measurements of total and upper
atmosphere ozone trends, balloon measurements, and satellite measurements.
The first part of the analysis focuses on total column ozone as measured by
ground-based instruments.  Observations made with groundbased instruments
appear roughly consistent with one-dimensional model predictions.  The second
part of this section reviews ozone measurements for different altitudes as
measured with ground-based instruments, which also appear  to show depletion
roughly consistent with models.  The third section reviews altitude data
                          * * *  DRAFT FINAL  * *

-------
                                   5-101
collected by balloons, which apparently shows depletion inconsistent  with one
dimensional models, but consistent with two-dimensional models.   The  fourth
part reviews ground-based and satellite data for Antarctica,  which appear
inconsistent with models.  Finally, satellite data for total  column ozone for
the globe and different latitudes is reviewed.   This data shows  decreases in
ozone greater than models predict.

Analyses of Global Total Ozone with Ground  Based Instruments

    Analyses of total column ozone using Dobson instruments must convert  the
irradiances received by the spectrometer into a total column  ozone reading.
Reading from each station must be combined to represent seasonal and  yearly
averages for that location.  To develop global averages, station values must
be areally weighted and averaged.  At each one of these stages,  assumptions
and choices must be made which can influence the outcome of the  analysis.

    Ground Based Global Ozone Depletion Estimates Appear Consistent with  One
    Dimensional Model Predictions

    Over the  past several  years  there  have been several studies conducted in
an attempt to detect any evidence  of a total  ozone trend (e.g.  Hill et al.,
1977; Reinsel, et al., 1981; St. John et al., 1982,  Angell and Korshover,
1983; Bloomfield et  al., 1983).  Most recent  statistical analyses have
adopted a time  domain approach to estimating a  global trend.   The ozone  value
X. .  at time t and observing station j is  represented by X. .  = w.
 ^/J                                                       ^'J     J
 h. + £.  . where h. represents a predicted global trend ... at
   L    U, J         I
station j; and e.  . represents an error process  to  account for
                *'J
other influences on ozone.   The  E.  . series  is typically assumed to
                                  *rl
be an  autoregressive  process  (e.g., Hill et al.,  1977; Reinsel et al.,  1981;
St. John et al., 1982).   This model is  fit separately to  the ozone record
from each station, and a global trend estimate  is obtained by combining the
station values e..  Bloomfield  et  al., (1983), on the other hand,

introduce a frequency domain statistical model.   This  model extends
variance-components analysis  to the time series  case and incorporates both
temporal and  spatial association found in the ozone data.  A  key feature  of
this  model  is  the inclusion  of  a common  global term  representing  natural
global  variations in ozone.   (World Meteorological Organization,  1986).

    Reinsel  et al.,  (1981) found  an increase of  0.28%  in global total ozone
over the period  1970-1978 with a standard  error of 0.67%, while St. John et
al.,  (1982)  found an  increase of 1.5% with  a standard error of 0.5% from
'70-'79 and  Bloomfield et al.,  (1983) found an increase of 0.1% for the same
period with a standard  error  of 0.55%.  Thus there is little overall support
for the suggestion of a statistically significant trend in total  ozone.
(World Meteorological Organization, 1986).

    Reinsel  et al.,  (1985 - personal communication)  have recently extended
the analysis using data through  1983.  Time series  models  were used  to obtain
a trend estimate for each station,  where level shifts to account  for
instrument  recalibration were  also  included  in the model for five stations


                          * * *  DRAFT FINAL  » * *

-------
                                   5-102
(Mt. Louis, Mt. Abu, Lisbon, Buenos Aires,  and Hradec Kralove).   The overall
trend estimate for total  ozone change over the entire period 1970-1983,  with
associated  95% confidence  limits, is (-0.003  + -1.12)  % per decade.
(World Meteorological Organization,  1986).

    Trend  analyses  were also performed using the f10.7 solar flux series  and
the sunsport  number series separately  as explanatory variables for the  total
ozone series at each station.  Results were quite similar in both cases,  and
the overal  total ozone trend estimates for the period  1970-1983 are
summarized as follows, (World Meteorological  Organization, 1986):

        (-0.00 + 1.12) % per decade with no solar effect in model
        (-0.17 + 1.10) % per decade with f10.7 solar  flux  in model
        (-0.14 + 1.08) % per decade with sunspot series in model

    The  f10.7  solar flux  series (as well as the  sunspot series) was found to
be mildly related to total ozone overall, with the estimated effect of  f10.7
solar flux  on  total  ozone (averaged over 36 stations)  equal to  (0.63  +
0.53)% ozone change per 100 units of F10.7 solar flux.  This estimate
corresponds to about  a one percent  change in total ozone  from solar cycle
minimum  to maximum.   (World Meteorological Organization,  1986).

    These  analyses  all indicate no significant overall  trend in total ozone
during the fourteen year  period 1970-1983, and  suggest a mild  relation
between  total  ozone and f10.7 solar  flux.  The trend estimates are about  0.2%
per decade more negative with the inclusion of data for 1983 than comparable
estimates based on data through 1982.   (World Meteorological Organization,
1986).

   A major question of the statistical analyses has been the general  lack of
global coverage of  the ground-based observations suggesting possible spatial
biases.   For the most recent trend estimates  given above by  Reinsel et al.,
(1985,  personal communication).   [Exhibit  5-65]  shows  a plot of  the  trend
estimates as a function of latitude.  Altogether, there does not seem to be
any latitudinal effect, but if we examine this diagram by region  we see an
interesting pattern.   For  example, all of the  North American  stations are
below the  Indian network  as are 6 of 7 European stations.  While the analysis
of Reinsel  et  al.,  (1981) takes this type of regional networking  into
consideration,  it suggests that more consideration should be given to the
representativeness  of the  data set.   (World Meteorological Organization,
1986).   In addition, it is clear  that there  is  very poor coverage in northern
latitudes,  with no stations in the sample  significantly above  60°N.  Coverage
in the southern hemisphere is worse.

    Analysis of Ozone Trends at  Different Altitudes with Ground Based
    Instruments

    Although  there  have been several  recent analyses of Umkehr data  (eg Angel
and Korshover, 1983b; Reinsel et  al.,  1983; Bloomfield et  al., 1982),
these have been  limited in that they did not consider the  impact  of
stratospheric  aerosols on the  observations.  This has been discussed by
                          * * *  DRAFT FINAL  * *

-------
                              5-103



                         EXHIBIT 5-65

                Ozone Trend Estimates by Latitude
0
0.4
0
ooo
° 0.2
oooo
0
oooooo
ooooo n -.
oo
oooo
00 -0.2
ooo
o
-0.4
oo
—
1
-
A
A ' J
J N
S , NE J EN E E
N
AS E E N
E
N
N NEE

_
H
A
1 1 1 1 1 1 1
60 -40 -20 0 20 40 60 80
LATITUDE
A AUSTRALIA 1 INDIA N NORTH AMERICA H HAWAII
E EUROPE J JAPAN S SOUTH AMERICA
Recent ozone trend estiamtes  from ground-based  observation
networks.  Horizontal axis shows  latitude  of  measurement.  Legend
identifies each site.  Vertical  axis  shows histogram of trend
estimates.

Source:  World Meteorological Organization,  (1986).
                            DRAFT FINAL  *  *

-------
                                   5-104
Deluisi (1979) and  Dave et al.,  (1979) and  it has been concluded  that
aerosols  tend to  induce  significant negative errors in the Umkehr
measurements in  the  uppermost layers 7-9,  with  the  largest percentage error
occurring in layer  9.   (World Meteorological Organization, 1986).

   Based on this,  Reinsel et al., (1984) have completed  a statistical
analysis  of  the Umkehr  data where  atmospheric  aerosols  are taken into
account.   In the statistical trend analysis,  time  series models have been
estimated using monthly averages of Umkehr data over the past 15 to 20  years
through  1980 at each of 13 Umkehr stations and  at each  of the five  highest
Umkehr  layers,  5-9,  which cover an altitude range of approximately 24-48  km.
The  time series regression models  incorporate seasonal,  trend, and noise
factors and an additional factor  to explicitly account for  the  effects  of
atmospheric aerosols  on  the Umkehr  measurements.   At each  Umkehr  station, the
explanatory series  used in the statistical  model  to account for the aerosol
effect is  a 5 month running average of the  monthly atmospheric transmission
data at Mauna  Loa, Hawaii,  the only long running  aerosol data available.   A
random effects model is  used to  combine the 13 individual station  trends  for
each Umkehr layer.  The  analysis indicates statistically significant trends
in the  upper Umkehr layers 7 and 8 of the order of  -0.2 to  -0.3% per year
over the period 1970-1980, with  little trend in  the lower  layers 5  and 6.
(World Meteorological Organization,  1986).

   The results are [repeated here and] shown graphically in  [Exhibit
5-66] (Reinsel  et al., 1983) where we  have added, for comparison numerical
model calculations from Wuebbles et al.,  (1983).  We see that there  is
substantive agreement between the  observations  and  the  model calculations.
(World Meteorological Organization,  1986).  Readers  may  also wish to refer
back to Exhibit 5-40  for a 2-D comparison.

   There are several points to be  raised  on these  results of  the Umkehr
analysis.   The first is  that the  results,  including the sign, are very
sensitive to the inclusion  of a stratospheric aerosol impact (i.e. Reinsel  et
al.,  1983,  1984).   For this data  record, the major impact is due to the
volcanic  eruption at Mt  Agung in 1963 and  to lesser  extent that of Tiera  Del
Fuego in 1974 and  possibly  Mt.  St. Helens  in 1980.  The use of a single
station such a Mauna Loa  to account for the aerosol  on a global basis is
fraught  with difficulty  and great care must be taken on  the  interpretation  of
the results.   This  was  evidenced most recently by the events of the volcanic
eruption  of El Chichon.   Attempts (DeLuisi, personal communication) to
consider  this even  within  the framwork of the  previous  analysis have not been
free of difficulty.  One  suggestion  by DeLuisi  is that there existed  an
actual decrease  of  ozone associated with the volcanic debris,  possibly due to
heterogeneous chemistry or injected  chlorine, but this is still quite
uncertain.   Considerably more work will be required before we will be able to
utilize  the Umkehr  data  through the El Chichon  event with some confidence.
(World Meteorological Organization,  1986).

   Another  point for consideration is the global  representativeness of the
13 station  network, an  element touched on  in the total ozone trend section.
For the  Umkehr data, Reinsel  et al., (personal communication) have  examined
                          * * *  DRAFT FINAL

-------
                               5-105
                          EXHIBIT 5-66

                 Changes  in  Ozone from  1970  to 1980:
             Umkehr Measurements and Model  Calculations
 LYR
                                                                MB
              REINSEL ET AL. (1984)
              WUEBBLES ET AL. (1983)
            A REINSEL ET AL. (1984)
                                                               1-2
                                                               2-4
                                                               4-8
                                    8-16
                                                              16-32
        -6
              -5
                     -4
-3
-2

 %
                                       - 1
Changes in ozone by altitude  for  1970  to  1980.   The data "analysis
indicate statistically significant  trends  in the upper Umkehr layers
7 and 8 of the order of  -0.2  to  -0.3 percent per year over the
period 1970-1980, with little trend in the lower layers 5 and 6.

Source:  World Meteorological Organization,  (1986).
                         *  DRAFT FINAL   *  *  *

-------
                                  5-106
the spatial sampling by examining  the approximate  4  year  period of the global
SBUV data.  Their results are  presented in  [Exhibit 5-67] where  the  trends
over the period November '78 - April '82 have been determined from SBUV  zonal
averages.  Superposed on these curves  are the values determined in  10°  x  10°
longitude/latitude boxes from the SBUV  data that includes the Umkehr site.   In
layer 8,  for example,  we see that  the North  American and European stations are
biased on the high side of the  curve and that Japaneses stations tend to bring
the overall average into line.  Thus, the overall results can  be  very
sensitive to the availability of the  stations.  This is  taken to extreme  in
the Southern Hemisphere where we see  that the Australian station happens  to
coincide  with the  relative  peak  in  the zonal average.   Because of this
fortuitous sampling the overall  station average is rather close to the total
SBUV area weighted trend.  That  the ground-based  results are  so sensitive to
the spatial sample is,  of course, precisely why the NOAA  operational satellite
ozone measurement program was initiated.  This will  be  discussed  further
below.   (World Meteorological Organization,  1986).

   As the final element in this  section we discuss the inclusion of a solar
flux variation within the trend  analysis.  For the Umkehr data,  Reinsel et
al.,  (personal communication) have included  the f10.7 cm  solar radio flux as
an additional independant  variable.  (World Meteorological Organization,
1986).

   For each Umkehr layer, 5 to 9,  and  for a  given Umkehr  station, the model
was  used:
Where
         .
        xt
                        = y
Y.  = monthly average for month  t of a station's observations  at a
      given layer
      seasonal component (annual and semi-annual)
      0 to t < T (T denotes  12/1696)
      (t -  T)/12 for t > T
      annual rate  of change  parameter (trend)
      a smoothed version of  the transmission data at Mauna Loa
      a parameter  providing  an empirical measure of the aerosol
      effect on Umkehr data
      montnlv mean °f 10-7  cm solar radio flux  (2800 MHz)
      a parameter  for an empirical measure  of the solar effect
      on Umkehr data
      an autocoreleated  noise terms, modelled as an autoregressive
      process to account for non-independence of data
         ..
         1


        V2t =
         -   =


        N.   =


        M
                                DRAFT FINAL  * » •'•

-------
                               5-107
                          EXHIBIT 5-67

    SBUV  Zonal Trends Estimates  Versus  "Umkehr Station Blocks"

                                  LAYER 9
       - 1 5
       - 2 0
               -60
     Q
     Z
        0.5
       -0.5
       - 1.5
         -80
               -60
        1.0
        0.5
        0.0
       -0.5
         -80    -60
     o
A AUSTRALIA
E EUROPE
                      -40
                            -20
                                    0
                                  LAYER 8
                                          20
                                                 40
                                                       60
                                                              80
                      -40
                             -20
                                    0
                                  LAYER 7
                                          20
                                                 40
                                                       60
                                                              80
-
1.0
0.5
0.0
0.5
1 0
BO -60 -40 -20 0 . 20 40 60 ' 80
LAYER 6
/^ 	 '""\ _;^J-K i
E \ !
x 1
                      -40
                              20     0
                                  LAYER 5
                                          20     40      60     80
1.0
0.5
0.0
0.5
1.0
E
r 	 1 E^N^
/ \^/
80 -60 -40 -20 0 20 40
^\
E
\
60 8
                                LATITUDE
                                1 INDIA
                                J JAPAN
N NORTH AMERICA
S SOUTH AMERICA
Ozone trend estimates for November 1978 to April 1982, as determined
by SBUV data,  are indicated  by solid line.   Superimposed letters
indicate  results from Umkehr stations.
Source:   World Meteorological Organization,  (1986).
                      * *  *  DRAFT FINAL  *  * *

-------
                                  5-108
   A term to account for a shift in mean  level due to instrument
recalibration was also  included  in the model for some stations,  as  described
above.   (World Meteorological Organization,  1986).

   The trend estimates including the  solar flux series are shown in
[Exhibit  5-66]  as the  solid triangles  (Reinsel et al., personal
communication).  We see that there .is very little impact on the decadal trend
with this term  added.   The  overall estimates for the regression coefficients
of the  solar flux series with  accompanying 95 percent confidence intervals
are 0.81  + 2.56,  0.46  +  1.80, 2.04  +  1.32, 1.53 + 0.78 and
1.18 + 0.86  percent,  respectively,   for layers  9 through 5.  These
coefficient estimates represent percentage change in  ozone per 100 units
change in solar flux.  The solar effect is larges in  layer  7,  and
statistically  significant in  both  layers 7 and 6.  The estimates  correspond
to a solar effect  on ozone  from solar  cycle minimum  to maximum of about  1.3,
0.8, 3.4, 2.5,  and 2.0 percent respectively,  for layers 9 through 5.   (World
Meteorological  Organization,  1986).

   It is not at  all clear as to why  the solar flux term should have a
significant regression  coefficient in layers 7  and 6 and not in layers  9 and
8, and rises serious questions on the effect  of the solar flux variations.
(World Meteorological  Organization, 1986).

Ballon  Date  for Ozone Trends at Different Altitudes

    In  a  recent update of  ozone variations determined from balloonsondes,
Angell  and  Korshover (183)  determined that  in the troposheric layer  of north
temperature latitudes, 2-8 km,  the  data  suggest a 12%  increase in ozone
between  1970 and 1981.  This is accompanied  by a 1-3% decrease in the region
16-32 km.   Since then, several on-going  studies have focused  on  the quality
of the  ozonesonde data for trend detection (Tias et al.,  personal
communication;  Logan,  1985)  and the  discussion  is presented here with the
author's  kind permission.   Formal  publication is planned  for  the near
future.   From  Tias et al.,  Ozonesonde data from  13  stations  have been
processed to obtain monthly averages of ozone in 14 fractional  Umkehr layers,
(1A, IB,  1C, ID,  2A,  2B, 3A, 3B, 4A,  4B, 5A,  5B, 6A, 6B), (e.g. [Exhibit
5-68]  and an additional layer above 6B.   For each station, the daily  opv0e
data (in  partial pressure)  were first  integrated into  ozone readings  \-\\>
Dobson  units)  within each layer,  and monthly averages  for each  layer  were
then computed  from the integrated  readings.   The data were screened  to meet
the following criteria:  (World Meteorological  Organization, 1986).

            The correction factor was between  .8 to 1.3  for  ECC
            and  .8 to 1.4  for Brewer.

        --   The  balloon reached a  burse  level of  15.8 mbar (top of
            layer 5B).

            There  were  no zero partial pressure readings recorded
            in  each daily  sondes.
                                 DRAFT FINAL   -• * *

-------
                             5-109



                        EXHIBIT 5-68

       Ozone Trend Estimates and 95% Confidence  Intervals
                        95% Interval  (%/yr.)       95%  Interval  (?i/yr.)
 Layer        KM       (Without  Intervention)    •  (With  Intervention)
Above 6B
6B
6A
5B
5A
4B
4A
3B
3A
2B
2A
ID
1C
IB
1A
35+
30-35

25-30

20-25

15-20

10-15

5-10

0-5

-0
0
0
0
0
-0
-0
-0
-0
-0
-0
0
1
1
1
.05 +
.04 +
.22 +
.15 +
.05 +
.07 +
.21 +
.33 +
.56 +
.30 +
.17 +
.93 +
.44 +
.43 +
.72 +
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
.38
.33
.31
.17
.18
.15
.22
.25
.31
.53
.67
.04
.23
.87
.17
0
0
0
0
-0
-0
-0
-0
-0
-0
-0
0
0
0
0
.22 +
.27 +
.46 +
.18 +
.02 +
.16 +
.30 +
.48 +
.71 +
.48 +
.64 +
.08 +
.57 +
.66 +
.75 +
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.56
.54
.51
.21
.27
.22.
.21
.26
.27
.49
.75
.88
.74
.53
.83
Source:  WHO (1986),  p.  799.
                    * •'- »  DRAFT FINAL  * * *

-------
                                  5-110
   --  The total ozone  reading  for the daily  sonde data had a
        nonzero value.

    The  station  locations,  data  span  and methods  of measurement are given in
[Exhibit  5-69]  (Tiao et al.,  personal  communication),  where the Canadian
stations changed from the Brewer  system to  the ECC at the designated
times.   (World Meteorological Organization,  1986).

   One of the first elements examined were the correction factors for the
various instruments and,  as examples of this parameter,  we present in
[Exhibit  5-70]  (Reinsel et al.,  personal  communcation), the results at Goose
Bay, and Hohenpeissenberg.  At Goose Bay  we  see that  the individual months
show large variations with a small  tendency for  decrease till 1980 where  the
change to ECC was  effected.  The impact of  this  change will be discussed
further below.   For Hohenpeissenberg the diagram also shows some interesting
month-to-month  variations and we  note,  in particular,  the tendency  for the
corrections to increase  during the first few years followed  by the strong
minimum  in the late '70's.   (World  Meteorological  Organization. 1986).

   The cause of these tendencies in  the  correction factors  is unclear and
may  be related to instrument manufacture, personnel changes  or  changes in  the
Dobson system.   The major point is that we  can not expect the correction
factors to be random about some average value and that  we will have to
consider, in detail,  the possible impacts  of these  variations.   As Hilsenrath
et al., (1984) have  indicated, this brings into  question whether or not the
factors should be  applied as they  are, as a percentage change to the  profile,
or in some height dependent manner.  (World Meteorological Organization,
1986).

   The above  notwithstanding,  trend  estimates (1970)  have been  obtained from
the monthly averages using  standard  models  reported  previously  with  and
without an intervention at the 4 Canadian sites for  the changes of
measurement method.  As  a cross-validation  an overall  trend estimate for each
station was  obtained  by calculating a  weighted  average of the individual
estimates of the 15 layers.   These estimates  may then be compared with  the
corresponding trend  estimates obtained from  the Dobson total  ozone  reading  on
the ozonesonde file.  The results are shown  in  [Exhibit 5-71] (Tias et al.,
personal  communication).   For the nine non-Canadian  stations, the trends from
weighted averages are in close  agreement with  those from Dobson total ozone
readings on the sonde file.  For Churchill,  Edmonton  and Resolute,  the
agreement seems  much better without the intervention  level  adjustments.   This
is in direct contrast to what we would expect and reflects,  again, the
question  as to whether the correction factors between the Brewer Mast and ECC
sondes are applied in a consistent manner.

   Finally, in [Exhibit  5-68  above] (Tias et al., personal communication)
we present the combined  ozone  trend estimates ([Exhibit 5-71]) in the
various layers along with their  95% confidence limits.   We see  that with and
without the data  intervention at the  4 Canadian sites the negative trends in
the lower stratosphere  appear statistically significant and of the  same
magnitude, about  -5% per  decade.   In the lower troposophere, however,  the
                          .'-  J,  JL.  r>TD A T?T T? T XT A T  * --V •*•
                                DRAFT FINAL

-------
                      5-111
                   EXHIBIT 5-69



           Ozone Balloonsonde Stations
    Station
Data Span
Source:   WHO (1986),  p.  797.
Methods
Hohenpeissenberg
Biscarrosse
Lindenberg
Payerne
Aspendale
Churchill
Edmonton
Goose Bay
Resolute
Wallops Island
Kagoshima
Sapporo
Tateno
1/70-2/83
3/76-1/83
1/75-2/83
9/68-12/81
6/65-5/81
10/73-1/83
10/70-12/82
6/69-12/82
1/66-1/8-3
5/70-4/82
1/70-12/82
12/68-12/82
3/68-12/82
Brewer
Brewer
Brewer
Brewer
•Brewer
Brewer/ECC 9/79
Brewer/ECC 9/79
Brewer/ECC 12/80
Brewer/ECC 12/79
ECC
ECC
ECC
ECC
                    DRAFT FINAL

-------
                                5-112









                           EXHIBIT  5-70




          Correction  Factors for Balloonsonde Measurements
                                GOOSE BAY
 l.-J
 1.2
 1.0
 0.8
   65
70
                                    75
                              HOHENPEISSENBERG
                                                     80
                                                                     85
 1.4
 1.2
 1.0
 0.8
   65
                    70
                                    75
                                                     80
                                                                     85
Shown for Goose  Bay  and Hohenpeissenberg.




Source:  World Meteorological Organization,  (1986)
                       ft  ft   DRAFT FINAL  * * *

-------
                                5-113
                           EXHIBIT 5-71

Ozone Trend Emissions  (% Per Year)  As Determined from Balloon Ozonesondes
            Versus Those Determined from Dobson  Measurements
                (Tiao,  et al., personal communication)
            Station
Ozonesonde
 Total  Ozone
  Readings
on Sonde  File
Aspendale
(6/65-5/82, Brewer)
Biscarrosse
(3/76-1/83, Brewer)
Hohenpeissenberg
(1/70-2/83, Brewer)
Lindenberg
(1/75-2/83, Brewer)
Payerne
(9/68-12/81, Brewer)
Kagoshima
(1/70-12/82, ECC)
Tateno
(3/68-12/82, ECC
Sapporo
(12/68-12/82, ECC)
Wallops, Isl.
(5/70-4/82, ECC)
Churchill
(10/73-1/83, Brewer/ECC 9/79)
Edmonton
(10/70-12/82, Brewer/ECC 9/79)
-.115
(.055)
-.416
(.114)
-.174
(.052)
-.287
. (-128)
-.149
(-045)
.136
(.090)
.085
(.082)
.148
(.103)
.034
(-075)
.473*
(.118)
. 405*
(.095)
-.162
(.075)
-.587
(.185)
-.220
(.088)
-.269
• (.266)
-.181
(.074)
.215
(.155)
-.055
(.094)
.203
(-138)
.037
(.122) •
-.310** .282
(.197) (.234)
.027** .360
(.152) (.149)
                         * *  DRAFT FINAL  * * *

-------
                                5-114
                      EXHIBIT 5-71  (Continued)

Ozone Trend Emissions (% Per Year) As  Determined from Balloon Ozonesondes
            Versus Those Determined from Dobson  Measurements
                 (Tiao, et al., personal  communication)
                                                              Total Ozone
                                                               Readings
            Station                Ozonesonde                on Sonde File
 Goose Bay                            .029*       -.095**       -.123
 (6/69-12/82,  Brewer/ECC  12/80)       (.060)       (.077)         (.099)

 Resolute                            -.194*       -.240**       -1.64
 (1/66-1/83,  Brewer/ECC  12/79)        (.045)       (.070)         (.074)
  *  Without intervention  adjustment.

 **  With intervention  adjustment.

 Source:   WHO (1986), p. 798.
                        * *  *  DRAFT FINAL  * * *

-------
                                  5-115
values are very different although they  remain  positive.   Thus, there is
evidence to suggest the existence  of overall  negative trends at layers 3A and
3B, and perhaps also at 2B  and 4A.   (World Meteorological Organization,
1986).

   The  results for the troposhere  have recently been  examined by  Logan
(1985) and her analysis indicates that  the  surface ozone  at mid-latitudes
displays two modes of seasonal  behavior:   a  broad  summer maximum within a few
hundred kilemeters of populated and industrialized regions in  Europe and the
United States; and a summer maximum  in sparsely populated regions  remote from
industrial activity, in Canada and  Tasmania, for example.  She argues,  in
addition,  that the current data base for different regions, in  combination
with limited historical data indicates  that summertime concentrations of
ozone near the surface in  rural areas  of Europe and the  central and eastern
U.S.  may have increased by approximately 10-20 ppb  (30-100%) since the
1940's.   The seasonal cycle  of ozone in the middle troposhere  over  Europe,
the United States,  and northern Japan is very  similar to that at the surface,
with a summer maximum,  but it is  quite  different from that a  300 mb, which is
characterized  by a maximum  in  spring.  There  is good evidence for an increase
in ozone in  the middle troposphere over Europe during the past 15 years, and
weaker  evidence for  a  similar increase  over Northern America  and  Japan.  From
this she argues that the summer maximum  in ozone and the observed trends are
due to photochemical production associated with anthropogenic emissions of
NOx hydrocarbons and CO from combustion of fossil fuels.  A strong seasonal
variation  in ozone observed  at Natal, Brazil  (6°S) may also result from
emissions of NOx and hydrocarbons, in this  case from  agricultural  burning.
Maximum concentrations at Natal are similar to values found at mid-latitudes
in summer.   (World Meteorological Organization, 1986).

   With  the limited network  of ozonesonde  stations,  however,  the question
remains as to whether  the tropospheric ozone increase  is due  to local
pollution effects or  is symptomatic of a more general atmospheric behavior.
(World Meteorological Organization,  1986).

    Ground Based  and Satellite Data  for Antarctica

    Recently,  Farman et al.  (1985) have published data, displayed  in Exhibit
5-72,  showing a large secular decrease in total ozone for the month of October
over their station at Halley Bay,  Antarctica  (76 degrees  south).   Since 1957
the mean total ozone for the month of  October over Halley Bay has  decreased by
40% with most of  the decrease occuring since the mid-1970's.   Since 1978,
satellite data from the Nimbus & TOMS  and SBUV  instruments confirm  these
findings.  Along  with data in the  early  1970's  from the Nimbus 4  BUY instru-
ment,  the satellites show that  in  the  southern  hemisphere total  ozone is
generally at minimum during  the springtime over the Antarctic continent (or
generally poleward of about  70 degrees south).  This minimum is  surrounded by
a range of maximum total ozone values  centered  at about 55 degrees  south.  The
surrounding maximum region displays  significant wave structure.  The minimum
often becomes distorted into an oblong shape and rotates  along with the maxi-
mum.  Exhibit 5-73 shows  a twelve  day  sequence  for October of 1985  which
                                 DRAFT FINAL  * * *

-------
                               5-116






                          EXHIBIT 5-72


             Monthly Means  of  Total Ozone at Halley  Bay
        300
       UJ


       O
       rsi
       O
        200
— OCTOBER
                1960
                                1970
                                               1980
Total Ozone in October for the years 1957 to  1984.


Source:  World Meteorological Organization, (1986)
                     * * *  DRAFT FINAL  * * *

-------
                                       5-117
                                  EXHIBIT 5-73

              Nimbus 7 Antarctic Ozone Measurements:   12  Day Sequence
11 October 1984
15 October 1984
19 October 1984
12 October 1984
16 October 1984
20 October 1984
13 October 1984
17 October 1984
                                            21 October 1984
14 October 1984
18 October 1984
                      22 October 1984
         "Twelve-day sequence (11-22 October 1984) of TOMS measurements of
         total ozone content.  The data are shown in south polar projections,
         with the pole indicated by a cross (SP) and Halley Bay shown by an
         asterisk.  Contours are every 30 Dobson units (1 DU = 10-3 atm cm).
         The region shown extends to ~45° latitude and the Greenwich
         meridian is towards the top of each diagram.  Shaded regions
         indicate total ozone values <180 and 210 DU and >390 and 420 DU."

         Source:  Stolarski et al. (1986)
                              * * *  DRAFT FINAL  * * *

-------
                                   5-118
displays a 7 day rotation period.  There is a marked tendency for the minimum
to be displaced off the pole towards the direction of Halley Bay (Stolarski et
al. 1986).

    Since 1979, the Nimbus 7 data show decreases in both the maximum and
minimum region.  The largest decreases (of order 40%) are in the minimum
region leading to the lowest total ozone values ever recorded (less than 150
milli-atmosphere-centimeters -- Dobson units).  Exhibit 5-74 (taken from
Stolarski, et al., 1986) shows a map of the mean total ozone for each of the 7
Octobers measured by the TOMS instrument.   The map is a polar projection with
the outer edge at 30 degrees south latitue (Stolarski et al. 1986).

    The other major feature of the data is the seasonal variation of the ozone
minimum.  The minimum values of total ozone appear in the spring, after the
long polar night, at approximately the values entering the polar night.  The
rapid decline then takes place during September with the minimum being reached
in October and a rapid recovery in November when the polar vortex breaks down
(Stolarski, et al., 1986).  "These  data indicate that some mechanisms or
mechanisms are at work in the cold  southern polar night or twilight that are
not generally included in  models.  This clearly warrants further
investigation"  (WHO, 1986).

Alternate Hypotheses  of Antarctic Depletion

    Explanation of the Antarctic ozone phenomenon still remain in the category
of hypotheses.  Many interesting ideas have been put forward which are, in
principle, capable of explaining parts of the observed phenomena.  Further
data is clearly necessary to differentiate among competing ideas before a
scientific consensus can be reached on the cause of the changes.  The first
concept put forward was that the decrease was caused chemically by the action
of chlorine which is known to have increased due to chlorofluorocarbon release
(Farman et al. 1985; Solomon et al.  1986;  McElroy et al. 1986).  For this
explanation to work, the cold polar night and the polar stratospheric clouds
must be invoked to modify the previously known chemistry in a major way.  The
rate limiting radical, CIO, must be present in the part per billion range
(almost 100 times what would have been assumed based on previously known
chemistry) in the lower stratosphere during September and October.  Bromine
has also been introduced as a possible enhancement for the chlorine theory.
The chlorine theory is susceptible to test by the microwave instrument which
is in place at McMurdo in the spring of 1986.  The primary aspect of this
hypothesis is that the driving force, chlorine chemistry, has a known secular
variation.

    Another primarily chemical hypothesis is that the depletion is due to
nitrogen oxide (NOx) chemistry which is varying due to solar cycle variations
in the production of NOx in the thermosphere.  This idea requires the
transport of this excess NOx downward such that polar night and twilight
values are increasing to show a maximum a number of years after solar
maximum.  For this idea to work, NOx would have to be high, CIO normal, and
the effect should show an 11 year cycle.  Thus the major problem is the lack
of an apparent cyclic behavior in the Halley Bay data.
                          * * *  DRAFT FINAL  * *

-------
                              5-119
                         EXHIBIT 5-74

         Nimbus  7  Antartic Ozone Measurements:   Mean Total
                         Ozone in October
                   1979
                                               1981
"Six-year sequence of October monthly means  of  total ozone.  South
polar projections, with the pole indicated by a cross and 30°S
latitude by a dashed circle.   The Greenwich  meridian is towards the
top of each panel.  Contours  are every 30 DU.   The  shaded regions
indicate monthly mean total ozone amounts of <240 and >390 DU."

Source:  Stolarski et al.  (1986)
                     * * *  DRAFT FINAL  *  *  *

-------
                                   5-120
    A third class of hypotheses is that the depletion is caused primarily by a
modification of the climatological pattern of the wind fields in the southern
hemisphere.  One suggestion is that increased aerosols from the El Chichon
eruption and their associated stratospheric heating lead to lifting of the
atmosphere and low total ozone values (Tung et al., 1986).   Analysis of the
data indicates that there are variations that exist that are superposed on the
secular decrease.  These year-to-year differences may be due to variations in
atmospheric dynamics.  Furthermore, secular decrease in Antarctic
stratospheric temperatures has accompanied the ozone decrease.  Present
understanding would seem to indicate that these are not a result of the ozone
decrease.  Thus, it is possible that a secular change in the climatology of
the south polar region is taking place.  A test of these ideas would be that
the chemistry would be relatively normal but analyses of temperature and
constituent data should clearly show effects of modifications in transport
properties.  The primary problem with this hypothesis is to trace such
modifications back to a cause that shows an apparent secular variation of the
time period under consideration.

    Several satellite systems, Firos, Nimbus 7, and NOAA9,  exist for measuring
column ozone at different latitudes and global ozone.  The TOVS system on
Tiros, and TOMS and SBUV-2 on Nimbus 7 and SBUV on the NOAA series.  SBUV-2
can also measure ozone at various altitudes.

    All satellite systems rely on interpreting irradiances as ozone and
require corrections for aerosols and instrument drift:  Thus, the major
problem in determining ozone trends from these instruments is maintaining the
calibration of the systems.

    Exhibit 5-75 shows global ozone measurements for the TOVS.

    The main problem in this system is that the TOVS system requires
regression against the ground-based Dobson network.  This data seem to
indicate a drop in ozone, perhaps attributable to El Chichon.  The data record
is too short to determine if the trend is a short term fluctuation or part of
a longer trend.

    Nimbus 7 Results

    Results from Heath's analysis of Nimbus 7 have not yet been published, but
results have been mentioned by NASA at Senate hearings and shown at several
meetings.

    Heath shows a .56 + 19% per year global ozone decline from 1978 to 1984
(Heath, personal communication).  Heath's interpretation of this trend is
qualitatively consistent with seasonal and latitudinal predictions in Isaksens
two dimensional time dependent runs, although of greater magnitude.  Exhibit
5-76 shows the 1978 to 1984 estimates of depletion from Isaksen's model and
according to Heath's interpretation of the Nimbus 7 data.  Clearly the pattern
is similar, although Heath's interpretation of the satellite data suggests
depletion eight times greater at 80°N than Isaken's model.   Isaksen's
estimates of Antarctic ozone (without special chemistry), however, is smaller
than Arctic estimates, contrary to Heath's data.


                          * * *  DRAFT FINAL  * * »

-------
                                  5-121
                              EXHIBIT 5-75

                   Global (60°N-60°S) Monthly Total  Ozone
                      Determined from NOAA  TOVS System
       310
      270
         MAY  OCT  MAR  AUG   JAN  JUN  NOV  APR  SEP  FEB  JUL  DEC MAY

          1979        1980        1981        1982        1983        1984
Source:   WHO (1986), p.  792.
                            * *  DRAFT FINAL  * * *

-------
                                        5-122
                                   EXHIBIT 5-76

                         Preliminary Ozone Trend Data (Health
                          versus  2-D Model  Results  (Isaksen)
zu-
18
16
14
Ozone 12
Depletion JQ.
(%) 0
O "
6-
4
2-
0
Latitude: South
Pole










-

17.
•\
''•-.-._


'-•„

\
•\'
'\
80

4

















6.C

\J
;'••>
•60

//ear/i




7
1 F

2:4 2:4 i-s pq ?
0.4 n n |\N f\N F^ • ' v
. — , «-u [•••.•-.] k\1 KM l\i >
-40 -20 Equator +20 +40 +60 +i






.2

^
s
>
JO Nort
Pole
Ozone
Depletion
i.u-



0.5-



0
Isaksen

0.7






O.f

'HA
•-J '•
''<>'''
0.4 0.4 0.4




§
;v
V"





0.2
^
'''. '"'•:
'"•:'"'




0.3
\\

v%




0.2
c\

../'.,,.
1



1
\>
X'-
.V




1
\\

'','




•A
1

••S".'
\>







0.9
\ '*''
^
'•Sv
V"1
1
^
*, ':
\\









Latitude:  South  -80      -60
                              -40
-20    Equator    +20     +40      +60
+80
                                                                                 Pole
         Predicted ozone changes  from  Isaksen's 2-D model are qualitatively
         consistent with preliminary data from Heath.

         Source:  Heath (personal  communication);  and  Isaksen and Stordal
                  (1986).
                              * * -  DRAFT FINAL  *  * *

-------
                                  5-123
    One problem with interpreting data is the shortness of the satellite
data.  It is not clear whether these "trends" are long-term in nature, or
short-term fluctuations.  Additional analysis is needed and 1985 and 1986 is
needed quickly.

    Use of Antarctic and Satellite Measurements in Risk Assessments

    The published Antarctic and yet unpublished satellite data present a
difficult challenge to us in this risk assessment.  Clearly, if the Antarctic
trends are chemical in origin and extrapolatable to the world, or if the
Nimbus 7 data has been correctly interpreted and portends a chemically induced
trend, the risks of depletion are significantly greater than predicted in
models being used in this assessment.  Given the disequilibrium between
current emission and concentrations of CFCs, if these results portend the
future, depletion can be expected to increase with a large cutback in
depleting chemicals.

    In our judgment, however, it is premature to reach the conclusion that
these observations can be used for assessing global trends.  The scientific
analysis to support such a judgement is incomplete.  Until the analysis
reaches the stage where the Antarctic situation can be extrapolated to the
rest of the world or until the Nimbus 7 data is shown to be real and
representative of the rates of depletion in the future, these data should be
treated with caution.  Thus, for decisionmaking purposes we propose that
current models constitute the appropriate tools for assessing future risks.

    The satellite and ground-based Antarctic data do, however, raise a
"blinking yellow alert", which may require reassessment of risks.  Because
both sets of data present sufficiently stark pictures of a possible future,
efforts need to be made on an expedited basis to better resolve our
understanding of these data sets.  The Nimbus 7 data, already global in
character, needs the earliest analysis so that resolution of its implications
can be reached.  In particular, a very strong effort must be made to expedite
the flow of data from satellites to analysis to peer review.  The flow of data
must be as close to "real time" as possible, not a year or two behind.
                                 DRAFT FINAL  * *

-------
                                  5-124
REFERENCES
Angell, J.K.j and J. Korshover, Global temperature variations in the
    troposphere and stratosphere, Mon. Weather Rev.,  111,  901-921,  1983a.

Angell, J.K., and J. Korshover, Global variations in  total ozone and
    layer-mean ozone: An update through 1981,  J.  Glim.  Appl.  Meteor.,  22,
    1611-1627, 1983b.

Bemand, P.P., M.A.A. Clyne, and R.T. Watson, Atomic resonance fluorescence
    and mass spectrometry for measurements of the rate constant for elementary
    reactions, J. Chem. Soc. Faraday Trans., 2, 70,  564-576,  1974.

Bhartia, P.K., D.F. Heath, and A.J.  Fleig, Observation of  anomalously  small
    ozone densities in South Polar Stratosphere during October 1983 and 1984,
    paper presented at 5th General Assembly, IAGA Symposium,  Prague,
    Czechoslovakia, July, 1985.

Bloomfield, P., G. Oehlert, M.L. Thompson, and S. Zeger, A frequency domain
    analysis of trends in Dobson total ozone records, J. Geophys.  Res., 88,
    8512-8522, 1983.

Brasseur, G. and A. DeRudder, The potential impact of atmospheric ozone
    and temperature of increasing trace gas concentrations, final report to
    the Commission of the European Communities, Contract #85-66602-11-
    010-11-N, EC, Brussels, Belgium, 1986.

Brasseur, G., A. DeRudder, and C. Tricot, Stratospheric response to chemical
    perturbations, J. Atmos. Chem.,  3, 261-288, 1985.

Brune, W.H., E.M. Weinstock, J.J. Schwab, R.M. Stimpfle,  and J.G.  Anderson,
    Stratospheric CIO: In situ detection with a new approach, Geophys.  Res.
    Lett., 12, 441-444, 1985.

Callis, L.B., J.M. Russell III, K.V. Haggard,  and M.  Natarajan, Examination of
    winter-time latitudinal gradiants in stratospheric N02 using theory and
    LIMS observations, Geophys. Res. Lett., 10, 945-948,  1983b.

Callis, L.B., M. Natarajan, and R.E. Boughner, On the relationship between the
    greenhouse effect, atmospheric photochemistry and species distribution,
    J. Geophys, Res., 88, 1401-1426, 1983a.

Chameides, W.L., D.H. Stedman, R.R.  Dickerson, D.W.  Rusch, and R.J. Cicerone,
    NOx production in lightning, J.  Atmos. Sci.,  34,  143-149, 1977a.

Chameides, W.L., S.C. Liu, and R.J.  Cicerone,  Possible variations in
    atmospheric methane, J. Geophys. Res., 82, 1795-1798,  1977b.
                                 DRAFT FINAL

-------
                                  5-125
Chemical Manufacturers Association, Atmospheric ozone:  response to combined
    emissions of CFCs, N20, CH4, and C02,  prepared for the United Nations
    Environment Programme Workshop on the  Control of Chlorofluorocarbons,
    Leesberg, VA, September 8-12, 1986.

Climatic Impact Assessment Program, Report of findings:  The effects of
    stratospheric pollution by aircraft, DOT-TST-75-50,  edited by A.J.
    Grobecker, S.C.  Coroniti, and R.H.  Canno, Jr., 551 pp., Department of
    Transportation,  Washington, DC, 1974.

Connell, P.S., A parameterized numerical fit to total column ozone
    changes calculated by the LLNL 1-D model of the troposphere and
    stratosphere, prepared for the U.S.  EPA, Lawrence Livermore National
    Laboratory, Livermore, California,  1986.

Connell, P.S., and D.J. Wuebbles, Ozone perturbations in the LLNL one-
    dimensional model -- calculated effects of projected trends in CFCs, CH4,
    C02, N20 and halons over 90 Years,  prepared for the U.S. EPA, Laurence
    Livermore National Laboratory, Livermore, California, 1986.

Cunnold, D.M., Fluorocarbon lifetime and releases from 5 years of ALE data,
    paper presented at CSIRO symposium,  The Scientific Application of Baseline
    Observations of Atmospheric Composition, Aspendale,  Australia, 7-9
    November, 1984.

Cunnold, D.M., R.G.  Prinn, R.A. Rasmussen, P.G. Simmonds, F.N. Alyea, C.A.
    Cardelino, and A.J. Crawford, The atmospheric lifetime experiment, 4,
    Results for CF2C12 based on 3 years of data, J. Geophys. Res., 88,
    8401-8414, 1983b.

Cunnold, D.M., R.G.  Prinn, R.A. Rasmussen, P.G. Simmonds, F.N. Alyea, C.A.
    Cardelino, A.J.  Crawford, P.J. Fraser, and R.D. Rosen, The atmospheric
    lifetime experiment, 3, Lifetime methodology and application to 3 years of
    CFC13 data, J. Geophys. Res., 88, 8379-8400, 1983a.

Dave,  J.V., J.J. DeLuisi, and C.L. Mateer, Results of a comprehensive
    theoretical examination of the optical effects of aerosols on the Umkehr
    measurements, Spec. Environ. Rep. 14,  pp. 15-22, WHO, Geneva, 1979.

Davis, D.D., J.T. Herron, and R.E. Huie, Absolute rate constants for the
    reaction 0(3P) + N02 	-* NO + 02 over the temperature range 230-339K,
    J. Chem Phys, 58, 530-535, 1973.

DeLuisi, J.J., Umkehr vertical ozone profile errors caused by the presence of
    stratospheric aerosols, J. Geophys.  Res., 84, 1766-1770, 1979.

DeMore, W.B., D.M. Golden, R.F. Hampson, M.J. Kurylo, C.J. Howard, J.J.
    Margitan, M.J. Molina, and A.R. Ravishankara, Chemical Kinetics and
    Photochemical Data for Use in Stratospheric Modeling, Evaluation Number
    I, JPL Publication 85-37, 1985.
                            * *  DRAFT FINAL  * *

-------
                                  5-126
DeMore, W.B., Rate constants for the reactions of hydroxyl and hydroperoxyl
    radicals with ozone, Science, 180, 735-737, 1973.

DeMore, W.B., and 0. Raper, Hartley band extinction coefficients of ozone in
    the gas phase and in liquid nitrogen, carbon monoxide and argon, J. .Phys.
    Chem., 68, 412-414, 1964.

DeMore, W.B., J.J. Margitan, M.J. Molina, R.T. Watson, D.M. Golden, R.F.
    Hampson, M.J. Kurylo, C.J. Howard, and A.R. Ravishankara, Chemical
    kinetics and photochemical data for use in stratospheric modeling,
    Evaluation Number 7, JPL Publication 85-37, 226 pp., Jet Propulsion Lab.,
    Pasadena, CA 1985.

DeMore, W.B., M.J. Molina, R.T. Watson, D.M. Golden, R.F. Hampson. M.J.
    Kurylo, C.J. Howard, and A.R. Ravishankara, Chemical kinetics and
    photochemical data for use in stratospheric modeling, Evaluation number
    6,  JPL Publication 83-62, 219 pp., Jet Propulstion Lab., Pasadena, CA
    1983.

DeMore, W.B., Rate constant and possible pressure dependence of the reaction
    OH + H02, J. Phys. Chem., 86, 121-126, 1982.

DeRudder, A., and G. Brasseur, Ozone in the 21st century: Increase or
    decrease?, in Atmospheric Ozone, edited by C.S. Zerefos and A. Ghazi, pp.
   . 92-96, D. Reidel, Dordrecht, 1984.

Edmonds, J.A., J. Reilly, J.R. Trabalka, and D.E. Reichle, An analysis of
    possible future atmospheric retention of fossil fuel C02, DOE/OR-21400/1,
    169 pp., Institute for Energy Analysis, Washington, DC, 1984.

Garcia, R.R., and S. Solomon, A numerical model of the zonally averaged
    dynamical and chemical structure of the middle atmosphere, J. Geophys.
    Res.. 88, 1379-1400, 1983.

Gidel, L.T., P.J. Crutzen, and J. Fishman, A two-dimensional photochemical
    model of the atmosphere. 1: Chlorocarbon emissions and their effect on
    stratospheric ozone, J. Geophys. Res., 88, 6622-6640.

Grant, K.E., P.S. Connell, and D.J. Wuebbles, Monte Carlo uncertainty
    analysis of stratospheric ozone in ambient and perturbed atmospheres,
    submitted to Journal of Geophysical Research, 1986.

Guthrie, P.O., C.H. Jackman, J.R. Herman, and C.J. McQuillan, A diabatic
    circulation experiment in a two-dimensional photochemical model, J_^
    Geophys. Res., 89, 9589-9602, 1984.

Hill, W.J., P.N. Sheldon, and J.J. Tiede, Analyzing worldwide total ozone for
    trends, Geophys. Res. Lett., 4, 21-24, 1977.
                          * - *  DRAFT FINAL  * * *

-------
                                  5-127
Hilsenrath, E., J. Ainsworth, A. Holland, J.  Mentall, A.  Torres,  W.
    Attmannspacher,  A. Bass, W. Evans, W. Komhyr, K.  Mauersberger,  A.J.
    Miller, M. Proffitt, D. Robbins, S. Taylor and E. Weinstock,  Results from
    the balloon ozone intercomparison campaign (BOIC), in Atmospheric Ozone,
    edited by C.S. Zerefos and A. Ghazi, pp.  454-459, D.  Reidel,  Dordrecht,
    1985.

Isaksen, I.S.A., and F.  Stordal, "Ozone perturbations by enhanced levels
    of CFCs,  N20, and CH4: A two-dimensional  diabetic circulation study
    including uncertainty estimates," Journal of Geophysical Research,
    91(D8), 5249-5263, 1986.

JPL: See DeMore, et al.  and NASA-JPL listings.

Johnston, H.S., Reducton of stratospheric ozone by nitrogen oxide catalysts
    from supersonic transport exhaust, Science, 173,  517-522, 1971.

Jones, R.L.,  and J.A. Pyle, Observations of CH4 and N20 by the Nimbus 7 SAMS:
        A comparison with in-situ data and two-dimensional numerical model
        calculations, J. Geophys. Res., 89, 5263-5279, 1984.

Khalil, M.A.K., and R.A. Rasmussen, The trend of bromochlorodifluormethane
    (CBrClF2) and the concentrations of other bromine containing gases at the
    South Pole:, Antarctic Journal of the United States,  Annual Issue, 1985.

Ko, M.K.W., K.K. Tung, D.K. Weinstein, and N.D. Sze,  A zonal-mean model of
    stratospheric tracer transport in isentropic coordinates: Numerical
    simulations for nitrous oxide and nitric acid, J. Geophys.  Res., 90,
    2313-2329, 1985.

Ko, M.K.W., N.D. Sze, J. Livshits, M.B. McElroy, and J.A. Pyle, The seasonal
    latitudinal behavior of trace gases and 03 as simulated by a two-
    dimensional model of the atmosphere, J. Atmos. Sci.,  41, 2381-2408, 1984.

Ko, M.K.W., and N.D. Sze,, Diurnal variation of CIO:  Implications for the
    stratospheric chemistries of C10N02, HOC1, and HC1, J. Geophys.  Res., 89,
    11619-11632, 1984.

Levy II, H.,  Normal atmosphere: Large radical and formaldehyde concentrations
    predicted, Science,  173, 141-143, 1971.

Levy II, H.,  Photochemistry of the lower troposphere, Plant: Space Sci.,
    20, 919-935, 1972.

Liu, S.C., M. McMcFarland, D. Kley, 0. Zafiriou, and B.J. Huebert,
    Tropospheric NOx and 03 budgets in the equatorial Pacific,  J. Geophys.
    Res.. 88, 1360-1368, 1983.

Logan, J.A.,  M.J. Prather, S.C. Wofsy, and M.B. McElroy,  Tropospheric
    chemistry: A global perspective, J. Geophys. Res., 86, 7210-7254, 1981.
                          * * *  DRAFT FINAL  * * *

-------
                                  5-128
Logan, J.A., Tropospheric ozone: Seasonal behavior, trends and anthropogenic
    influence, J. Geophys.  Res., 90, 10463-10482, 1985.

NcElroy, M.B., R.J. Salawitch, S.C. Wofsy, and J.A. Logan, Reductions of
    Antarctic ozone due to synergistic interactions of chlorine and bromine,
    Nature, 321, 759-762.

Henzies, T., A re-evaluation of laser heterodyne radiometer CIO measurements,
    Geophys. Res. Lett., 10, 792-732, 1983.

Miller, C., D.L. Filken, A.J. Owens, J.M. Steed, and J.P. Jesson, A
    two-dimensional model of stratospheric chemistry and transport, J_._
    Geophys. Res.. 86, 12039-12065, 1981.

NASA-JPL, Chemical kinetics and photochemical data for use in stratospheric
    modeling, Evaluation No. 7, NASA Panel for Data Evaluation, JPL
    Publication 85-37, Jet Propulstion Laboratory, Pasadena, CA, 1982.

National Aeronautics and Space Administration, Present State of
    Knowledge of the Upper Atmosphere:  An Assessment Report.  Processes that
    Control Ozone and Other Climatically Important Trace Gases, NASA
    Reference Publication 1162, NASA, Washington, DC, 1986.

National Research Council,  Causes and Effects of Changes in Stratospheric
    Ozone:  Update 1983, National Academy Press, Washington, DC, 1984.

National Research Council,  Environmental Impact of Stratospheric Flight,
    National Academy of Sciences, Washington, DC, 1975.

Owens, A.J., C.H. Hales, D.L. Filkin, C. Miller, and M. McFarland, Multiple
    scenario ozone change calculation: The subtractive perturbation approach,
    in Atmospheric Ozone, edited by C.S. Zerefos and A. Ghazi, pp. 82-86, D.
    Reidel, Dordrecht, 1985a.

Owens, A.J., C.H. Hales, D.L. Filkin, C. Miller, J.M. Steed, and J.P. Jesson,
    A coupled one-dimensional radiative-convective, chemistry-transport model
    of the atmosphere 1. Model structure and steady-state perturbation
    calculations, J. Geophys. Res., 90, 2283-2311, 1985b.

Prather, M.J., M.B. McElroy, and S.C. Wofsy, Reductions in ozone at
    high concentrations of stratospheric halogens:, Nature, 312, 227-231,
    1984.

Prinn, R.G., P.G. Simmonds, R.A. Rasmussen, R.D. Rosen, F.N. Alyea, C.A.
    Cardelino, A.J. Crawford, D.M. Cunnold, P.J. Fraser, and J.E. Lovelock,
    The atmospheric lifetime experiment, 1, Introduction, instrumentation and
    overview, J. Geophys. Res., 88, 8353-8367, 1983a.

Prinn, R.G., R.A. Rasmussen, P.G. Simmonds, F.N. Alyea, D.M. Cunnold, B.C.
    Lane, C.A., Cardelino,  and A.J. Crawford, The atmospheric lifetime
    experiments, 5, Results for CH3CC13 based on three years of data, J_^
    Geophys. Res., 88, 8415-8426,  1983b.
                          -v * *  DRAFT FINAL  * * *

-------
                                  5-129
Reinsel, G.C., G.C. Tiao, J.L. De Luisi, C.L. Mateer, A.J.  Miller, and J.E.
    Frederick, Analysis of upper stratospheric Umkehr ozone profile data for
    trends and the effects of stratospheric aerosols, J.  Geophys.  Res.,  89,
    4833-4840, 1984.

Simmonds, P.G., F.N. Alyea, C.A. Cardelino, A.J.  Crawford,  D.M.  Cunnold, B.C.
    Lane, J.E. Lovelock, R.G. Prinn, and R.A. Rasmussen,  The atmospheric
    lifetime experiment, 6, Results for carbon tetrachloride based on 3  years
    data, J. Geophys. Res., 88, 8427-8441, 1983.

Slanger, T.G., B.J. Wood, and G. Black, Investigation of the rate constant
    for 0(3P) + N02 	-> 02 + NO, Int.  J. Chem. Kinet., 5,  615-620, 1973.

Solomon, S., R.R. Garcia, F.S. Rowland, and D.J.  Wuebbles,  On the depletion
    of Antarctic ozone, Nature, 321, 755-758, 1986.

Solomon, S., and R.R. Garcia, Transport of thermospheric NO to the upper
    stratosphere, Planet. Space Sci.,  32, 399-409, 1984a.

Solomon, S.,and R.R. Garcia, On the distribution of long-lived tracers and
    chlorine species in the middle atmosphere, J. Geophys.  Res., 89,
    11633-11644, 1984b.

St. John, D., W.H. Bailey, W.H. Fellner, J.M. Minor, and R.D. Sull, Time
    series analysis of stratospheric ozone, Commun..  Statist. Theory Methods,
    11, 1293-1333, 1982.

Stokarski, R.S., A.J. Krueger, M.R. Schoeberl, R.D.  McPeters, P.A. Newman and
    J.C. Alpert, Nimbus 7 satellite measurements of the springtime Antarctic
    ozone decrease, Nature, 322, 808-810,'1986.

Stolarski, R.S., Impact of large amounts of chlorine on stratospheric ozone,
    paper presented at the International Workshop on Current Issues in our
    Understanding of the Stratosphere and the Future of the Ozone Layer, BMFT,
    NASA, FAA, WMO, Feldafing, FRG, June  11-16, 1984.

Stolarski, R.S., and A.R. Douglass, Sensitivity of an atmospheric
    photochemistry model to chlorine perturbations including consideration of
    uncertainty propagation, Journal of Geophysical Research, 91(D7),
    7853-7864, 1986.

Stordal, F., and I.S.A. Isaksen, (1986), Ozone perturbations due to increases
    in N20, CH4, and chlorocarbons: two-dimensional time dependent
    calculations, J. Titus (ed.), Effects of Change in Stratosphere Ozone and
    Global Climate. Volume I. Overview, U.S. EPA, Washington, DC.

Sze, N.D., Anthropogenic CO emissions:  Implications for the atmospheric
    CO-OH-CH4 cycle, Science, 195, 673-675, 1977.
                          * * *  DRAFT FINAL  * * *

-------
                                  5-130
Sze, N.D., M.K.W. Ko, M. Livshits, W.C. Wang, P.B.Ryan, R.E. Specht, M.B.
    McElroy, and S.C. Wofsy, Annual Report on the Atmospheric Chemistry,
    Radiation and Dynamics Program, Atmospheric and Environmental Research,
    Inc., Cambridge, HA, 1983.

Tung, K.K., M.K.W. Ko, J.M. Rodriquez, and N.D. Sze, Are Antarctic ozone
    variations a manifestation of dynamics or chemistry? Nature, 322, 811-814.

WHO, The Stratosphere 1981. Theory and Measurements, WHO Global Ozone Research
        and Monitoring Project Report No. 11, 516 pp., WHO, Geneva, 1982.

Waters, J.W., J.C. Hardy, R.F. Jarnot, and H.M. Pickett, Chlorine monoxide
    radical, ozone, and hydrogen peroxide: Stratospheric measurements by
    microwave limb sounding, Science, 214, 61-64, 1981.

Weinstock, J., Nonlinear theory of gravity waves: Momentum depostion,
    generalized Rayleigh friction and diffusion, J. Atmos.  Sci., 39,
    1698-1710, 1982.

Wofsy, S.C., Interactions of CH4 and CO in the earth's atmosphere, Ann.  Rev.
    Earth Planet. Sci., 4, 441-469, 1976.

Wofsy, S.C., J.C. McConnell, and M.B. McElroy, Atmosphric CH4, CO and C02, J_._
        Geophys.  Res., 77, 4477-4493, 1972.

World Meteorological Organization, Atmospheric Ozone 1985.   Assessment
    of Our Understanding of the Processes Controlling its Present
    Distribution and Change,  WMO Global Ozone Research and Monitoring
    Project - Report No. 16, WMO, Geneva, Switzerland, 1986.

World Meteorological Organization, The Stratosphere 1981.  Theory and
    Measurements, WMO Global Ozone Research and Monitoring Project Report No.
    11, WMO, Geneva, Switzerland, 1982.

Wuebbles, D.J. A theoretical analysis of the past variations in global
    atmospheric composition and temperature structure, UCRL-53423, Lawrence
    Livermore National Laboratory, Livermore, CA, 1983b.

Wuebbles, D.J., Chlorocarbon emission scenarios: Potential impact on
    stratospheric ozone. J. Geophys. Res., 88, 1433-1443, 1983a.

Wuebbles, D.J., F.M. Luther, and J.E. Penner, Effect of coupled anthropogenic
    perturbations on stratospheric ozone, J. Geophys, Res., 88, 1444-1456,
    1983.

Wuebbles, D.J., and P.S. Connell, Interpreting the  1-D Model cclculated
    nonlinearities from chlorocarbon perturbations, presented at the
    International Workshop on Current Issues in Our Understanding of the '
    Stratosphere and the Future of the Ozone Layer, BMFT, NASA, FAA, WMO,
    Feldafing, FRG, June 11-16, 1984.
                          * * *  DRAFT FINAL  * * *

-------
                                  5-131
Wuebbles, D.J., and P.S. Connell, Interpreting the 1-D model-calculated
    nonlinearities from chlorocarbon perturbations, paper presented at
    International Workshop on Current Issues in Our Understanding of the
    Stratosphere and the Future of the Ozone Layer, BMFT, NASA,  FAA, WHO,
    Feldafing, FRG, June 11-16, 1984.
                                 DRAFT FINAL

-------
Chapter 6

-------
                               CHAPTER 6

                                CLIMATE
SUMMARY

    Stratospheric perturbants are trace gases that are emitted by industrial
or biogenic processes that can change the amount and location of ozone or
water vapor in the stratosphere.  Increases in these trace gases can be
expected to warm the earth directly by reducing the escape of infrared
radiation to space.  As stratospheric perturbants, they will indirectly warm
the earth by altering the vertical distribution of ozone1 and the abundance
of water vapor in the stratosphere.

    The magnitude of warming anticipated from the expected increases in
greenhouse gases is quite large compared to past natural variations in global
temperature.  The anticipated speed of the warming is more rapid than has
occurred before.  The temperature increase from the Wisconsin Ice Age, 18,000
years before present (B.P.), is believed to be 4°C.  Global average
temperature increase in the next 100 years, without curtailment of greenhouse
gas growth, is likely to be somewhat comparable to that amount.

    The major uncertainty in predicting the magnitude of global warming is the
extent to which clouds will amplify or possibly reduce the warming.  The major
uncertainty in predicting the timing is the rate at which the oceans will
absorb heat.  Despite the numerous uncertainties, the realized global
temperature rise would be expected to exceed 1.6°C by 2075 and be as high as
4.5°C if no steps are taken to limit the rise of greenhouse gases.

    Major weather and climatic changes would be expected to accompany global
warming, although it is impossible to specify details of how the changes will
occur in different regions of the world.  In addition to temperature
increases, changes can be expected in precipitation, storms, and weather
patterns.  Current evidence suggests that significant changes will occur in
the dryness or wetness of many regions and in the frequency and location of
extreme weather.

    Policies aimed at reducing chlorofluorocarbons (CFCs) in order to limit
ozone depletion would also reduce global warming.  Increases in other
stratospheric perturbants (e.g., methane, and carbon dioxide) while also
contributing to the greenhouse effect are expected to buffer the earth from
ozone depletion.  Reductions in these gases to limit global warming would make
the stratosphere more vulnerable to depletion from CFCs.
1 If ozone depletion grows to a sufficient magnitude that there is a
decrease in ozone at altitudes below 28 kilometers, something which occurs
only in cases where depletion is large, the net effect of the ozone change
would, from that point on, produce a cooling effect.
                          -'• * *  DRAFT FINAL  * *

-------
                                   6-2
FINDINGS

1.   INCREASES IN THE ABUNDANCE OF TRACE GASES THAT ARE STRATOSPHERIC
    PERTURBANTS CAN INCREASE GLOBAL TEMPERATURES.   CHANGES IN THE
    STRATOSPHERE CAUSED BY THESE GASES CAN ALTER THE VERTICAL DISTRIBUTION OF
    OZONE AND INCREASE STRATOSPHERIC WATER VAPOR,  THEREBY INFLUENCING GLOBAL
    WARMING.

    la.   Trace gases that act as stratospheric perturbants also are greenhouse
         gases -- as the concentrations increase in the troposphere they will
         retard the escape of infrared radiation from earth,  causing global
         warming.

    Ib.   Increases in methane (CH4) will also add water vapor to the
         stratosphere, thereby further enhancing global warming.   Methane
         increases will also add ozone to the troposphere, where it acts as a
         strong greenhouse gas that will further increase global warming.

    Ic.   In all scenarios of ozone depletion, ozone decreases in the
         stratosphere above 30 km.   This allows more ultraviolet radiation to
         penetrate to lower altitudes, where the "self healing effect"
         increases ozone to partially compensate for the ozone loss above.  In
         some scenarios, sufficient depletion occurs that ozone eventually
         decreases at all altitudes.

    Id.   Decreases in ozone at 28 kilometers or above will have a warming
         effect on the earth.  There is a small net gain in energy because the
         increase in ultraviolet radiation (UV-B)  allowed to reach the earth's
         surface more than compensates for the infrared radiation that is
         allowed to escape due to depletion of ozone above that altitude.

    le.   Below 28 km, increases in ozone are more effective as absorbers of
         infrared radiation.  Consequently, increases in ozone below 28 km
         will produce a net warming.  In this case, the additional UV blocked
         by more ozone is less than the additional infrared that is blocked
         from escaping the earth.  Conversely, a decrease in ozone below 28
         kilometers will tend to cool the surface.

    If.   The effect of column depletion of ozone will depend on the magnitude
         of the depletion.  Until the depletion is of sufficient magnitude
         that the lower part of the column is depleted, ozone depletion will
         contribute to global warming.  If the stratosphere continues to
         deplete so that ozone is depleted below 28 km, this depletion will
         cause a cooling.

    Ig.   Because the radiative forcing will vary strongly with altitude and
         latitude, estimates of the effects of changes in the vertical column
         of ozone on global warming made with one dimensional models must be
         viewed cautiously.
                                 DRAFT FINAL  * * *

-------
                                   6-3
2.  EXPECTED INCREASES IN STRATOSPHERIC PERTURBANTS ARE LIKELY TO WARM THE
    EARTH SIGNIFICANTLY.

    2a.   Two National Academy of Science panels have concluded that the
         equilibrium warming for doubling atmospheric concentrations of C02,
         or for an equivalent increase in the radiative forcing of other trace
         gases, will most likely be between 1.5° and 4.5°C.

    2b.   Agreement exists about the magnitude of warming that would be
         directly associated without feedback enhancement with radiative
         forcing for given increases in stratospheric perturbants.   Thus, if
         all else remained constant (i.e., there were no changes in the
         geosystems of the earth), temperatures would rise by 1.26°C for a
         doubling of C02, and 0.45°C for a simultaneous doubling of N20 and
         CH4.  Direct radiative forcing from a uniform 1 ppb increase in
         CFC-11 and CFC-12 would increase temperature by about 0.15°C.

    2c.   Agreement exists that any initial warming from direct radiative
         forcing would change some of the geophysical factors that determine
         the earth's radiative balance (e.g., feedbacks will occur) and that
         these changes will amplify the initial warming.  Increased water
         vapor and altered albedo effects (snow and ice melting, reducing the
         reflection of radiation back to space) can be expected to increase
         the warming to as much as 2.5°C.  Large uncertainties exist about the
         influence of warming on global clouds, which could further amplify,
         or possibly reduce, the magnitude of warming.

    2d.   The three major general circulation modeling groups in the U.S.
         estimate an average global warming of around 4°C for doubled C02 or
         its radiative equivalent.  However, the modelers who produce these
         estimates do not feel confident enough in their representation of the
         cloud contribution to rule out greater or lesser amplifications,
         including a negative feedback that would reduce the warming to 2°C or
         an even lower value.

    2e.   Global average temperature has risen about 0.6°C over the last
         century, consistent with general predictions of temperature
         sensitivity.  It would be useful if this data could be used to derive
         the temperature sensitivity of the earth to a greenhouse forcing
         empirically.  This is not possible.  Uncertainty about the past
         concentrations of trace gases in the atmosphere, other exogenous
         factors that affect the climate (such as aerosols or solar input),
         and oscillation and instabilities in the internal dynamics of the
         climate system (such as ocean circulation), prevent such a derivation
         of the earth's temperature sensitivity from examination of the
         historic rise of temperature.  As a result of uncertainty about these
         factors, it will be difficult to ascertain the earth's temperature
         sensitivity using realized warming for more than another decade.
                                 DRAFT FINAL  * * *

-------
                                   6-4
    2f.  Efforts have begun to gather worldwide time series data for clouds.
         If adequate, these data, within the next decade,  may narrow estimates
         of the cloud contribution to temperature sensitivity.   However,
         because of the complexity of this issue, this effort may fail to
         provide a complete understanding of this aspect of climate.

3.   THE TIMING OF GLOBAL WARMING DEPENDS ON THE RATES AT WHICH GREENHOUSE
    GASES INCREASE, THE RATES AT WHICH OTHER FORCINGS, SUCH AS VOLCANOES  AND
    SOLAR RADIATION CHANGE, AND THE RATE AT WHICH OCEANS TAKE UP HEAT AND
    DELAY THE FULL TEMPERATURE EFFECTS.   WARMING IN EXCESS OF VARIATIONS
    WITHIN THE LAST CENTURY IS EXPECTED IN THE NEXT TEN YEARS IF VOLCANIC AND
    SOLAR FORCINGS DO NOT CHANGE GREATLY.

    3a.  The delay introduced by absorbtion of heat by the oceans can only be
         roughly described.  The simple one-dimensional models of oceans  that
         have been used for this purpose do not realistically portray the
         mechanisms for heat transport into the oceans.  Instead, these models
         use eddy diffusion to treat heat in a parameterized manner so that
         heat absorbtion is consistent with data for the paths of transient
         tracers.  These models indicate that the earth will experience
         substantial delays in experiencing the full warming from greenhouse
         gases.

    3b.  The earth's current average temperature is not in equilibrium with
         the radiative forcing from current concentrations of greenhouse
         gases.  Consequently global average temperature would increase in the
         future even if concentrations of gases did not rise any further.  For
         example, if 2°C is the actual sensitivity of the earth's climate
         system to a C02 doubling, simple models estimate the current
         "unrealized warming" to be approximately 0.34°C;  for a 4°C
         temperature sensitivity the unrealized warming would be estimated at
         approximately 1.0°C.

    3c.  Only one three-dimensional general circulation model has been used to
         simulate the increasing concentrations of greenhouse gases as they
         occur over time.  This simulation shows a faster warming than
         predicted by simpler one-dimensional models that use ocean box models
         to simulate time dependent warming.

    3d.  Future uptake of heat by the oceans may change as global warming
         alters ocean circulation, possibly altering the delaying effect  of
         the oceans as well as reducing their uptake of C02.

    3e.  No information exists to predict how volcanic or solar forcings  may
         change.  Analyses done of transient warming assume that past levels
         of volcanic aerosols will continue into the future and that solar
         forcing changes will average out over relatively short periods of
         time.
                          * * *  DRAFT FINAL  * * *

-------
                                   6-5-
4.  WITH A FEW GENERALIZED EXCEPTIONS, THE CLIMATIC CHANGE ASSOCIATED WITH
    GLOBAL WARMING CANNOT BE RELIABLY PREDICTED ON A REGIONAL BASIS.

    4a.  In general, as the earth warms, greater temperature increases will be
         experienced with increasing distance from the equator.

    4b.  Global warming can be expected to increase precipitation and
         evaporation, intensifying the hydrological cycle.  While models lack
         sufficient reliability to make projections for any single region, all
         perturbation studies with three-dimensional models (general
         circulation models) show significant regional shifts in dryness and
         wetness, which suggests that significant shifts in hydrologic
         conditions will take place throughout the world.

    4c.  Current general circulation models represent oceanic, biospheric, and
         cloud processes with insufficient realism to determine how extreme
         weather events and climatic norms are likely to change on a  regional
         basis.  For example, one analysis of general circulation model
         outputs suggests that the frequency of extreme conditions will change
         in many regions of the world.  Another shows increased summer drying
         in mid-latitudes for pertubation studies that utilized two different
         representations of clouds.

5.  LIMITING GLOBAL WARMING BY REDUCING EMISSIONS OF STRATOSPHERIC
    PERTURBANTS THAT TEND TO ADD OZONE OR COUNTER DEPLETION WOULD INCREASE
 ..  THE STRATOSPHERE'S VULNERABILITY TO OZONE DEPLETION.

    5a.  Decreases in substances with potential to deplete stratospheric ozone
         -- that is, chlorofluorocarbons and nitrous oxides -- would  decrease
         the rate and quantity of global warming.

    5b.  Decreases in methane emissions, which have potential to increase
         stratospheric and tropospheric ozone and thereby buffer ozone
         depletion, would decrease warming in three ways:   by reducing direct
         radiative effects from its presence in the troposphere; by lowering
         water vapor in the stratosphere; and by reducing ozone build-up below
         28 km.

    5c.  Decreases in C02 emissions that would decrease global warming, but
         would also have the effect of increasing the stratosphere's
         vulnerability to ozone depletion.

    5d.  Decreases in carbon monoxide concentrations, which may occur as
         energy practices change, could result in decreases in methane
         concentrations by increasing off radical abundance which in  turn
         would shorten the lifetime of methane.
                                 DRAFT FINAL

-------
                                   6-6
THE GREENHOUSE THEORY

    The earth is heated by solar radiation received in a variety of
wavelengths, with the greatest quantity of energy in the visible part of the
spectrum.  In turn, the earth is cooled by the infrared radiation given off to
space.  The average surface temperature that results is a balance of this
heating and cooling.

    Since Tyndall discovered in 1861 that water vapor can absorb infrared
radiation (Tyndall, 1861), it has been understood that increases in the
abundance of infrared absorbing gases (the "greenhouse" gases) in the
atmosphere can raise the earth's temperature by reducing the escape of this
energy to space.  It has also long been known that other aspects of
atmospheric composition, including ozone and aerosol abundance, can influence
the quantity of radiation (ultraviolet and visible) that reaches the earth's
surface.

    A variety of studies and international assessments (NAS, 1979; NAS, 1982;
WMO, 1986a) have been conducted over the last two decades concerning the
potential for increasing concentrations of trace gases to raise global
temperature and alter weather and climate.  All of the reviews conducted have
come to general agreement that increases in greenhouse gases will lead to a
global warming.

    This chapter reviews studies on how the gases that can alter the
stratosphere- create a greenhouse effect and how the alterations in the
stratosphere itself can add to or reduce the magnitude of the greenhouse
warming.  The chapter is divided into five sections:

    Section (1), on radiative forcing,  examines evidence that gases that
could perturb the stratosphere are also gases that could retard the escape of
infrared radiation, thereby forcing an increase in the earth's surface
temperature.  In particular, this section analyzes potential indirect
greenhouse effects of these gases by reviewing how changes they cause in the
stratosphere can add to or subtract from the direct radiative forcing of any
of the gases.  This section also reviews studies that estimate the magnitude
of warming that would directly result from radiative forcing in the absence of
feedbacks (that is, changes in the .earth's surface or atmosphere that could
alter the warming).

    Section (2), on the earth's ultimate temperature sensitivity to
radiative forcing, reviews evidence that bears on the magnitude of feedbacks
that would amplify or dampen the initial radiative warming.  Uncertainties in
current knowledge about the magnitude of the feedbacks is highlighted.

    Section (3), on timing,  reviews those factors in the earth's
geophysical system that will delay the full experience of the temperature
equilibrium associated with specific increases in greenhouse gases.  In
particular, this section reviews evidence on the uncertainties associated with
oceanic heat absorbtion.
                                 DRAFT FINAL  *

-------
                                   6-7
    Section (4),  on regional climate changes  associated with  global  warming,
focuses on changes in regional weather and climatic patterns  which could alter
the basic environmental conditions of the planet.

    Section (5),  on the effects of possible control of  greenhouse  gases,
analyzes the possible impact that curtailing various greenhouse gases (C02 and
CH4) to limit global warming would have on stratospheric ozone.

    Together the five sections cover the major issues that will determine how
the greenhouse effect will unfold with time and how that evolution is inter-
related to risks to the stratosphere.

RADIATIVE FORCING BY INCREASES IN GREENHOUSE  GASES

    By their influence on chemical or physical processes, a large  number of
gases can influence the structure and composition of the stratosphere .   During
their residence in the troposphere, these gases, alternatively called strato-
spheric perturbants, trace species, radiatively important trace species, or
greenhouse gases retard the flux of infrared radiation from the earth's  surface
to space.  Exhibit 6-1 shows the multiple effects that these  gases have  on
atmospheric processes.  Carbon dioxide,  the most important greenhouse gas,
cools the stratosphere and in most models,  increases ozone (NAS, 1984 and WMO,
1986b).   Methane has three effects on warming:  its direct radiative effect in
the troposphere; its effect in adding ozone to the atmosphere below 28 kilo-
meters;  and its effect adding water vapor to the stratosphere (NAS,  1984).
Nitrous oxide warms the surface by its presence in the stratosphere and
depletes ozone in the stratosphere, unless interfered with by high chlorine
levels (Ramanathan et al., 1985 and Stolarski, personal communication).

    While small uncertainties still exist in our understanding of  the radiation
that is absorbed by each gas, with a few exceptions (CFC 113, for  instance),
the absorption feature of these gases are well characterized, as shown in
Exhibit 6-2 (see NASA, 1986; WMO, 1986b; Luther and Ellingson (1985); and
Ramanathan et al. (1985) for discussion).  Consequently, increases in the
concentrations of these gases will reduce the escape of infrared radiation
from earth, providing a radiative forcing of the earth's climate system.
When computing the magnitude of that forcing for multiple perturbants, care
must be taken in considering possible overlaps that would reduce net forcing
(Wang et al., 1985; Luther and Ellingson, 1985).

    Studies of the direct radiative effects of increases in infrared absorbing
gases all produce approximately the same result -- a doubling of C02 (or
equivalent) eventually will raise global temperatures 1.26°C  if no other
factors in the geosystem are changed (Ramanathan et al., 1985; Schlesinger and
Mitchell, 1985; Hansen et al., 1984).

    Exhibit 6-3 shows the relative effectiveness of a 1 ppm increase in  such
greenhouse gases.  Notice that chlorofluorocarbons and methane have much
greater impact per molecule than C02.  The reason is that they  'cover' a part
of the infrared spectrum that is  'open'  or 'transparent1 to much of the
infrared radiation given off by earth, whereas the region of C02 absorption is
already somewhat 'opaque' or 'closed' part of the infrared spectrum.
                          * * *  DRAFT FINAL

-------
                                   6-8
                               EXHIBIT 6-1

                Stratospheric  Perturbants and Their  Effects
                         Direct Effect
                           on Global
                        Temperature from
                          Tropospheric
                            Presence
   Physical
  Effect on
 Stratosphere
  Effect on
Column Ozone
Carbon Dioxide
Methane
Nitrous oxide
Chlorof luorocarbons
Other Trace Gases
Increases l
Increases i
Increases l
Increases *
Increases l
Cools2
Adds water
vapor;
hydrogen *
Adds nitrogen*
Adds chlorine''
Adds catalytic
Increases differently
at different latitudes3
Increases at some
latitudes5
Decreases6
Decreases'*
Decreases'*
(methyl chloroform,
carbon tetrachloride,
halons)
species to
stratosphere'*
1 Ramanathan et al., 1985.

2 Connell and Wuebbles, 1986.

3 Isaksen, personal communication.

* National Academy of Sciences, 1984.

5 Isaksen and Stordal, 1986.

6 National Academy of Sciences (1984), notes the direct effect of N20 on
column ozone.  In the presence of high levels of chlorine,  N20 may interfere
with the catalytic cycle of chlorine,  reducing net depletion (Stolarski,
personal communication).
                          * * *  DRAFT FINAL  * * *

-------
                                    6-9
                                EXHIBIT 6-2

                 Absorption Characteristics of Trace  Gases
                                         (|am)
          20  18 16   14
                          12
                                  10
II! I
FI3
FII6
CHCIj
CH2CI2 CH2CI2
F22
FI4 <— >
CHjCCIjCCU
<-» *=— >
Fl!
SO, «=— »
**" FI2
N20
i
I I I I
I fu c . 	 ^ I I
CH2F2< — >^ 	
£ , y ^^~-~+^^
FI3BI FI38I CHF3
f -^ -f -y
~ FI3 F!3
FII6 FII6
.CHCIj
CH2Cl2
< — >
F22 F22 p|4
CHjCCIj ,CHjCCI3
Fil
<-> S02 S02
p «_> FI2 <-:>
N20 • ' N,0
C H a CHfl
03 03
HO *»
i i i i i i i i
          500  600  700  800   900  1000 1100 1200  1300  1400  1500  1600 1700
Atmospheric trace gases absorb outgoing  infrared  radiation of various
wavelengths.  Carbon dioxide  absorbs  much  of  the  radiation in the spectral
region between 13 ym and 20 ym.   Several trace  gases,  including
chlorofluorocarbons 11 and 12  (Fll  and F12) have  strong absorption features in
the spectral region of 7 ym to 13 ym.  This spectral  region is called the
atmospheric "window" since in this  region  the atmosphere is relatively
transparent.

Source:  World Meteorological Organization, 1986.
                                 DRAFT  FINAL   *  »

-------
                        6-10
                    EXHIBIT  6-3

Radiative  Forcing for a Uniform  Increase in Trace Gases
Temperature Sensitivity
Compound (°C/ppb)
C02
CH4
N20
CFC-11
CFC-12
CFC-13.
Halon 1301
F-116
CC1.
4
CHC13
F-14
CFC-22
CH2C12 .
CH3CC13
C2H2
so2
.000004
.0001
.001
.07
.08
.10
.10
. .08
.05
.04
.04
.03
.02
.01
.01
.01
           Source:  Adapted from Ramanathan
                   et al., 1985.
                      DRAFT FINAL

-------
                                   6-11
Changes In Stratospheric Structure and Composition
Can  Change  Radiative  Forcing

    In photochemical models, the structure and composition of the stratosphere
changes as gases such as CFC and methane increase in the atmosphere.   These
changes can influence the radiative forcing of the earth by altering fluxes of
infrared or ultraviolet radiation from or to it's surface.  The critical
factors of the stratosphere's organization that influence radiative balance
are the amount and vertical distribution of water vapor located in the
stratosphere and the vertical distribution of ozone.   One-dimensional and
two-dimensional models of the stratosphere project changes in the vertical
organization of ozone in response to increasing concentrations of CFCs and
other gases (Connell and Wuebbles, 1986; Stordal and Isaksen, 1986).   These
changes are significant.  Two dimensional models also show the changes are
latitude dependent (Stordal and Isaksen, 1986).  Increases in water vapor (due
to increasing methane) are also projected for most scenarios of trace gas
growth (Connell and Wuebbles, 1986).

    Changes in ozone abundance can have two effects on radiation.  They can
alter the penetration of UV-B and visible radiation from space and the escape
of infrared from the surface.  Ozone decreases anywhere in the column will
always allow more UV-B and visible radiation (in the Hartley-Muggins and
Chanpuis bands) to reach earth's surface.   Ozone increases will decrease UV-B
penetration.   Ozone decreases will allow more infrared to escape, but the
effect of ozone changes on infrared escape varies substantially with
altitude.  Ozone increments added near the tropopause produce the largest
increases in surface temperature.  This is because greenhouse efficiency is
proportional to the difference in the temperature of the radiation that is
absorbed by the ozone (essentially the ground temperature) and the radiation
that is emitted by the ozone (the local atmospheric temperature) (Lacis et
al., 1986).  As a result, a net column depletion that consists of depletion of
ozone in the upper atmosphere and of increases of ozone in the lower
atmosphere, will greatly increase the infrared radiation that is blocked from
escaping from earth.

    Exhibit 6-4 shows the relationship of ozone and altitude to increases in
temperature (Lacis, Wuebbles and Logan, 1986).  At 12 kilometers, the net
sensitivity of the surface warming to a given increase in ozone is
approximately fifteen times greater than at 20 km.  Above 28 km an ozone
decrease actually would produce a small net temperature decrease.  Exhibit 6-5
shows the relationship between water vapor, altitude, and radiative forcing.
However, increases in water vapor always produce a warming, with maximum
effect for water vapor increases at 14 kilometers (Lacis and Wuebbles, 1986).

    Scenarios in which total column ozone depletion is small, combine two
effects:  a large decrease above 28 km and a smaller increase in the lower
stratosphere (self healing effect and methane) and troposphere (from methane
increases).  In these cases, the total net radiative forcing of the planet
will be increased (Lacis and Wuebbles, 1986).
                          * * *  DRAFT FINAL  * * *

-------
                                   6-12
                               EXHIBIT 6-4

       Effects  of Vertical Ozone  Distribution on  Surface Temperature
   Height
(Kilometers)
                50.0
                40.0
                30.0
               20.0 -
                    —  (ozone increases
                       cause warming)
                10.0
                          REGION I
                D.Dr '  '  '  '
                 -0.01
                                                    REGION II
                                                 (ozone increases  .
                                                  cause cooling)
o.oo
0.01
0.02
               Relative  Temperature Change (°C per Dobson  Unit)
Sensitivity of global  surface temperature to changes  in vertical ozone
distribution.  The  heavy solid line is a least square fit  to model results for
an increase in ozone of 10 Dobson units (approximately 3%)  added to each
vertical  layer.   Above 28 kilometers, increases in ozone cause a net cooling
of the surface.   Below 28 kilometers, increases in ozone cause a net warming
of the surface.

Source:   Lacis,  Wuebbles, and Logan, 1986.
                             * *  DRAFT FINAL  » *

-------
                                   6-13



                               EXHIBIT 6-5

                Water Vapor,  Altitude, and  Radiative Forcing
DU • U
^n n
-jLJ. U
40.0
qn n
JU • U
Height
(Kilometers)



?n n





10.0
i
n.n
- 	 v • • i 	
\
: • L ' ~:
' • I ' ^
\
; \ :
— • ' • — \ — ,
v 1 :
' |\!
\ 1 i
r- ' X
\
1 V
: ^H
r _______ ^ ~
	 	 j- — — "•
- ^ ' \
— i — i — : — i 	 1 — I — i — i — ! — i — i — i — i — 1 — i — i — i — i 	 1 — i — i — i 	 1 	 i 	 i 	 i 	 i 	 i 	 L i 	 i_ i _i 	 i 	
                  -0.1    -0.0    0-1    0.2     0-3     0.4     0.5    0.6     0.7

                     Temperature Sensitivity (°C/CM(STP))
Increases in water vapor at all altitudes will warm the surface.
warming occurs for increases at 14 kilometers.

Source:  Lacis, personal communication.
The maximum
                                 DRAFT FINAL  * *

-------
                                   6-14
    In scenarios in which total depletion becomes large enough that ozone
decreases at altitudes below 28 kilometers, the total radiative forcing from
changes in vertical ozone will begin to favor cooling.

    Care must be taken in estimating vertical changes in ozone from
one-dimensional models; the forcing may be different than predicted.   The
reason is that, as two-dimensional chemical models show, depletion levels will
vary with latitude.  At higher latitudes, there may be  less ozone increase at
low altitudes.  At mid-latitudes, where industrial pollution is important,
ozone increases are already evident.  Nearer the equator, tropospheric
increases in ozone from CH4 will be more important and  help balance depletion
more.  Consequently, the radiative forcing is not likely to be equal at all
latitudes.  As a consequence, the positive feedbacks from albedo changes (snow
and ice melting) may be lower (Lacis and Wuebbles, 1986).

ULTIMATE TEMPERATURE SENSITIVITY

    To understand the possible role of feedbacks on the warming caused solely
by the initial radiative forcing, one must model the effects of that warming
on the earth.  A number of modeling analyses have been  used to examine the
earth's responses to radiative forcing and how those responses can increase or
decrease the global temperature rise (Hansen et al., 1984; Washington and
Meehl, 1983; Manabe and Wetherald, 1975).  Exhibit 6-6  shows an analysis by
Hansen et al. (1984), of the changes that would occur with the initial
warming.  While the exact responses are particular to their model,
quantitatively they are representative of the feedbacks found by other
modeling groups.  Thus, as Exhibit 6-6 shows, water vapor increases due to
initial warming and snow and ice melting will enhance warming significantly.
Water vapor has this effect because it is a greenhouse  gas.  Snow and ice
melting enhance warming because they reflect more visible radiation back to
space than land and water.  As such, snow and ice melting will decrease
visible radiation reflected back to space (Hansen et al., 1984).

    Consistency between models is quite strong.  In Hansen's model (Hansen et
al., 1984), the magnitude of the global warming is 2.0°C without cloud or snow
and ice feedback.  Manabe and Wetherald (1975) obtain a 2°C warming for
perturbation experiments in which their model assumes fixed clouds.  When
Manabe allows clouds to vary, his model shows that the  warming is amplified to
4°C (Manabe and Wetherald; 1986).  Washington and Meehl (1984), who also allow
clouds to vary, obtain a global temperature sensitivity of 3.5°C.

    Despite this apparent consensus, however, care should be taken in
interpreting this agreement.  The exact nature of future cloud changes remain
uncertain.  Somerville and Remer (1984) propose a mechanism by which clouds
might dampen the warming.  Manabe, Washington, and Hansen all counsel caution
(personal communications), and are uncertain of their model's portrayal of
cloud feedback processes.  The uncertainty associated with clouds is, of
course, completely consistent with the National Academy of Sciences' estimate
of temperature sensitivity.  Instead of using the range of temperature
responses produced by models, 2° to 4°C, the NAS has recommended a range of
                          * * *  DRAFT FINAL  * *

-------
                                   6-15
                               EXHIBIT 6-6

          Temperature Sensitivity  to Climatic Feedback Mechanisms
                                     Lapse
                                     Rate
                                   (-0.2K/km)
                C02
                
-------
                                   6-16
1.5°C to 4.5°C, which characterizes their judgement about the possible range
of temperature sensitivity.  Unfortunately, the possibility that the earth's
temperature sensitivity lies outside this range cannot be excluded (NAS, 1982).

Past and Near Term  Rises  in Temperature  Cannot be Expected
to Resolve Uncertainties  about  Actual  Climate  Sensitivity

    A different approach other  than modeling has been suggested as a means for
estimating the temperature sensitivity of the earth to a given radiative
forcing -- to compare historical temperature rises to past increases in
greenhouse gases (Webb and Wigley, 1985; Broecker, 1986).  Unfortunately,  this
approach cannot provide precise estimates of temperature sensitivity because
many of the factors that may have contributed to past temperature changes  are
themselves uncertain.  Exhibit  6-7 from Hansen et al. (1984) shows the extent
to which the derived temperature sensitivity would depend on knowing initial
trace gas concentrations and the rise in temperature.  Uncertainty exists
about both past temperature change and past concentrations of C02 and trace
gases.  Estimates of pre-industrial values of C02 based on ice core and other
data range from 260 to 290 ppm  (Oeschger et al., 1982; Neftel et al., 1982:
Pearman et al., 1986; and Peng  et al., 1983).   To complicate matters further,
uncertainty also exists about changes in other kinds of forcing that can
influence radiative balance, such as changes in solar radiation and aerosols
from volcanoes (Bradley and Jones, 1985).  Failure to include these possible
changes in forcing could bias estimates of temperature sensitivity.  In
addition, uncertainties about instabilities and oscillations in the internal
dynamics of the climate/ocean system are also unknown, and could produce
variation in observed temperatures (Hoffert and Flannery, 1985).  For example,
in a recent run of the Goddard  Institute for Space Studies (GISS) general
circulation model, without any  change for radiative forcing, global average
temperature varied 0.2°C from the 100 year mean in a period of 100 years
(Hansen et al., 1986).  Based on this estimate of "natural variation which is
entirely expected according to  model runs  (Hansen et al., 1986b)," an accurate
assessment of temperature sensitivity will not be possible even after the
temperature has risen another 0.5°C.

Clouds are the Major  Uncertainty

    The International Satellite Cloud Climatology Project is seeking to
collect a global database of radiation information that can be used to improve
our understanding of clouds and their response to global warming (WHO, 1984).
Several research obstacles must be overcome for this endeavor to be
successful, including insufficient satellite coverage, difficulties in
converting radiation data to accurate representation of clouds, and the
complexity of modeling cloud systems and their effects.  The goal of this
effort is to provide data that  allows us to narrow our uncertainty about
clouds considerably.   However,  the possibility exists that the data from this
international project will provide a useful beginning, but not a complete
resolution of the important climate feedback (Luther and MacCracken, 1985).
                                 DRAFT FINAL  * * *

-------
                          6-17
                      EXHIBIT 6-7

Empirical  Estimates  of Climate  Sensitivity are Sensitive to
       Estimates of Historical Temperature Increases
               and  Trace Gas  Concentrations
                 \z
                 •c
                 IJO

               1M
               en
               006
               CD
          Ciptcltd Wormng in AM - IMO Ou* t» CO, • Troe*
          . '   n Function of Equiferiunt ClimoO Smrtnrily
                                          CO,k>«90
                                            9
                                            (b)
                                                               "b"
The vertical axis measures  the  increase in global average temperature from
1850 to 1980.  The hdrizontal axis  indicates the earth's temperature
sensitivity to doubled C02  concentrations.  The curves represent temperature
changes consistent with  different historical changes in trace gas
concentrations.  Panel "a"  shows  the uncertainty for C02 alone.  Panel
incorporates other trace gases.   If one assumes an historical temperature rise
of 0.5°C, Panel "b" shows that  temperature sensitivity could vary from
approximately 2.3°C to 4.5°C, within a reasonably accepted range of 260 ppm to
280 ppm for pre-industrial  C02  concentrations.

Source:  Hansen et al.,  1984.
                        DRAFT FINAL   *

-------
                                   6-18
THE TIMING OF GLOBAL WARMING

    The timing of global warming is of particular importance to assessing
risks associated with it.  Knowing the radiative forcing and the temperature
sensitivity of earth to that forcing are insufficient for assessing the time
dependent evolution of the climate system.   Oceanic heat absorption must be
considered to determine timing.

    In this section, the results of two groups that have considered all three
components of climate change -- changes in greenhouse gases, temperature
sensitivity, and oceanic heat absorption -- are reviewed.

    These estimates can be broken down into efforts that utilized
one-dimensional and three-dimensional efforts.  Hansen et al.  (1981) used a
one-dimensional model to estimate the rise of temperatures from historic
levels.  Details of the model, a parameterized version of which was also used
by Hoffman et al. (1985), are presented in the appendix to this chapter.
Before proceeding, however, a brief discussion of the radiative forcings and
how the model deals with oceanic heat uptake is warranted.

    The rate of increase in expected increases in infrared absorbing gases has
been presented elsewhere in this document (Chapters 2, 3, and 4).  While
changes in volcanic aerosols and solar radiation could accelerate or retard
the increase in net radiative forcing, changes in these factors will not be
examined here, except to say that the radiation output of the sun has been
measured since, early 1980 by the NASA Solar Maximum Mission Satellite and
unpublished data show a decrease in the last four years.  Previous studies
have speculated on various cycles for changes in solar radiation (Newkirk,
Jr., 1983; Hoyt,. 1979; and Gilliland and Schneider, 1984).  At this time,
there is no basis for predicting future trends (Hoffert and Flannery, 1985).

    Aerosols from volcanoes depend on the rate and nature of eruptions.  In
general the eruptions must propel aerosols into the stratosphere for there to
be a climate effect.  For example, despite the magnitude of the eruption, Mt.
St. Helens had neglible effects on climate, due to a lack of sulfur in the
explosion and its sideway ballistics.  In 1982 El Chichon had a very large
climatic effect, possibly placing more aerosols into the stratosphere than any
since Krakatoa (MacCracken and Luther, 1984; NASA, 1982; and Pollack et al.,
1983).  Future eruptions cannot yet be predicted, but volcanic aerosols are
unlikely to have more than a two to three year influence on climate (Hoffert
and Flannery, 1985).

    The oceans are the major factor that could influence the timing of the
warming because of their enormous capacity as a heat sink (NAS, 1979).  At
this time, it is impossible to describe accurately all the processes that will
influence the ocean system's ability to take up or give off heat.  Data and
knowledge are both lacking.  The ocean is a complex three dimensional system,
in which heat exchange processes operate by many means, including both
convection and diffusion.  The specific topography of ocean floors is
important, as is the salt and water density.  Initial attempts are now being
                          * * *  DRAFT FINAL  * * *

-------
                                   6-19
made to model the ocean in more of its complexity (Hoffert and Flannery,  1985;
Sarmiento and Bryan, 1982; and Woods, 1985).   However,  an adequate ocean
general circulation model does not yet exist.

    In the absence of an adequate three-dimensional representation of oceanic
processes and heat uptake, representation of  the oceans has been relegated to
simple parametric models of oceanic heat uptake that use diffusion as a
surrogate for other mechanisms to absorb heat.  Such models, while clearly not
portraying mechanisms,  have the advantage of  allowing different assumptions to
be used about the quantity of heat uptake in  perturbation studies, thereby
providing some understanding of the variation in estimates of warming that is
attributable to uncertainties in oceanic heat absorption.  In one-dimensional
(1-D) models, models of the oceans that use eddy diffusion to transport heat
have been developed for this purpose.  The models have been calibrated to
reproduce the behavior of transient tracers (Broecker and Peng, 1982).  Such
models have allowed experiments to be conducted that allow for changes in the
rate of heat absorption.  Exhibit 6-8 shows the logical flow of how these
models can be used for estimating transient atmospheric temperatures.  Note
the parameter "k", which can be exogenously manipulated to test differing
rates of heat absorption.

    Hansen et al. (1984) used a one-dimensional radiative convective model
coupled to an ocean box model to estimate future realized and unrealized
warming through time (Exhibit 6-9).  The scenarios used for trace gas growth
are approximately the same as in Chapters 2 and 4.  Readers should consult
Appendix B for details.  The distance between the dotted and solid lines
represents the unrealized warming for two different assumptions about heat
uptake.  Note that in the time period examined, that is, until the year 2007,
the projected temperatures are quite similar  for different heat absorption.
Hoffman et al. (1985) also use a modified version of this model to make
estimates of temperature change further into  the future (see Exhibit 6-10).
They simulated a similar scenario, using 2°C  and 4°C as the earth's
temperature sensitivity to doubled C02.

    Only two time-dependent simulations have  been conducted using a general
circulation model (Hansen et al. 1986).  The  representation of the oceans in
that model, while still unrealistic, does use three-dimensional heat
absorption based on .transient tracers.  Exhibit 6-11 shows their results,
which depicts a faster warming than their 1-D models.  Since the results  are
from the same modeling group, the question arises of whether 1-D time7
dependent representations underestimate the actual rate of global warming.

REGIONAL CHANGES IN CLIMATE DUE TO GLOBAL  WARMING

    Only one change in regional climate can be projected with a high degree of
confidence.  Temperatures will rise more for  latitude bands as distance from
the equator increases.   For model runs in which the atmospheric C02
concentration is doubled, Hansen et al. (1986) and Washington and Meehl (1983)
find a ratio of approximately two to one between warming near the poles and
near the equator.  Manabe and Wetherald (1975) show a larger gradient of up to
seven to one.  Almost as certain as the latitudinal gradient is that global
                                 DRAFT FINAL  *

-------
                                             6-20
                                         EXHIBIT  6-8

                   Relationship of  Radiative Forcing,  Ocean Heat Uptake,
                            and Realized and Unrealized Warming
                                            Input:

                                                 Set
                                             Temperature
                                              Sensitivity
Input:

  Scenarios of
  Greenhouse
   Gas Rise
 Atmospheric
Concentrations
    Rise
                2°C to 4° C
 Equilibrium
Temperature
 Estimated
                                    Heat Flux
                                  Air
                                Warms
                                Slightly
                             Heat
                                               Heat
                                 Air
                             Temperature
                             Slowly Rises
                                                                               Difference at Any Time
                                                                                 Unrealized Warming
           Input:
           Set Diffusivity Surrogate for Oceanic Heat Absorption
           (1.7 cm2/sec used in this analysis)
                             Top Layers of
                             Ocean Slowly
                            Absorb Heat and
                                Warm
                                               Heat
                              Passed to
                            Deeper Layers
                              of Ocean
   The model allows input parameters  to be  set and then  consistently simulates
   time dependent  evolution of  the  system.
                                           DRAFT  FINAL   * * *

-------
                                    6-21



                                EXHIBIT 6-9

                    Transient Estimates of Global Warming
                      3.C
                      2.5
                      2.0
	— Mned Lift' (nOm) Heat Capacity
                    u
                      1.5
                      l.O
                      0.5
                               Ho Oceon
                                           X
                                             /
                              x^,,:::::---'----- "*
                                 i	t	i	i	i	i_
                       1850  1830    1920    I960    2000
                                   Date
Change in global  average surface temperature due to C02  and trace gas
increases from  1850  to 2007.   The solid line shows the warming that would
occur if oceanic  heat  absorption did not delay global warming.   The dashed
line represents oceanic heat  absorption by only the top  layers of the ocean.
The dotted  lines  represent heat absorption by the lower  layers of the ocean,
and test two assumptions about the rate of heat uptake,  "k".   The distance
between the solid line,  which shows "equilibrium" temperature,  and the dotted
lines, which show "transient" temperature, equals the "unrealized" warming.
Note that changes in this exhibit are relative to 1850.

Source:  Hansen et al.,  1984.
                           ••'•  •'• *  DRAFT FINAL  * * *

-------
                                    6-22



                               EXHIBIT 6-10

                        Expected Temperature Increases
              5 r-
           O
           •o
           V
           L.
           01
           c
           o
           u
           o
           t_
           m
           u
           o
Unrealized
Warming
(in pipeline)
                                                                      Equilibrium
                                                                      Terrperature
                                                                      Rise
                                    Realized
                                    Temperature
                                    Rise
1980
                        199O
2OOO
20 1O
2O2O
203O
 Change in global average  surface  temperature due to C02 and trace gas
 increases from  1980 to  2030.   A modified version of the Goddard Institute for
 Space Studies model (Hansen  et al.,  1981) was used.  The model's thermal
 sensitivity is  4°C for  doubled C02  concentrations.   Note that changes in this
 figure are relative to  1980.
 Source:  Hoffman, Wells and  Titus  (1985)
                            •'•- * *  DRAFT FINAL  * *

-------
                                    6-23
                               EXHIBIT 6-11

       Results  of Transient Analysis Using a General Circulation Model
   u
    i
   CD
   LJ
   Q
   UJ
   Q
    <
    CC
             OBSERVATIONS

             SCENARIO A

             SCENARIO 8
                                        DATE
Only two time-dependent simulations have been conducted  using  a  general
circulation model.  The results, shown above, indicate an  increase  in global
average temperature of approximately 0.9°C by the year 2000  for  Scenario  A
(which is a continuation of current rates of growth  in trace gases).   Scenario
B (which reflects reduced rates of trace gas growth)  indicates a warming  of
about 0.5°C by 2000.  Scenario A achieves a radiative forcing  equivalent  to
that of doubled C02 about 40 years from now; Scenario B  requires 75  years.

Source:  Hansen et al.. 1986.
                                 DRAFT FINAL  * * *

-------
                                   6-24
warming can be expected to intensify the hydrological cycle, increasing global
mean precipitation and evaporation (Schlesinger and Mitchell, 1985).  In
response to a quadrupling of C02 (with a model with a 2°C sensitivity; no
cloud feedback allowed), Manabe and Wetherald (1985) find a 7% increase in
both precipitation and evaporation.  Washington and Meehl (1984), in a similar
experiment, show an increase of 670 in precipitation.  For a doubled C02
experiment which yields a temperature increase of approximately 4°C, the
Goddard Institute for Space Studies model predicts an 11% increase in
precipitation and evaporation (Rind and Lebedeff, 1984).

    All three-dimensional models predict significant shifts in dryness and
wetness throughout the world.  For the U.S., the Goddard Institute for Space
Studies model (Rind and Lebedeff, 1984) predicts an average increase in runoff
of 18%.  Exhibit 6-12 shows the distribution of changes predicted by that
model.  Estimates of the distribution of runoff change from the general
circulation model of Princeton's Geophysical Fluid Dynamics Laboratory and
National Center for Atmospheric Research are similar to the Goddard Institute
for Space Studies run, even though they vary greatly in regional detail
(Manabe and Wetherald, 1985; Washington and Meehl, 1984).

    The reliability of regional projections from general circulation models is
weak, however.  Current general circulation models lack realism in
representing important ocean, biospheric, and cloud processes that are
critical to establishing precise estimates of hydrological change for
particular regions.  At this time, no interactive climate/ocean model has been
successfully developed and published.  Yet we know from events such as the
1983 El Nino that changes in oceanic circulation can influence regional
climates (Wigley, Angell, and Jones, 1985).  None of the models adequately
represent the hydrological process, especially the role of the biosystem.  In
fact, in all models the biosystem remains constant as climate changes and C02
increases.  In reality we can expect large alterations in ecosystems in
response to the effects of climate change and the response to the direct
physiological effects of C02 on photosynthesis (Strain, 1985).  These changes
would alter hydrological processes in important ways.

    The possibility still exists, however, that for large enough areas, some
robust projections can be made.  The general circulation model at Princeton's •
Geophysical Fluid Dynamics Laboratory has been run using two very different
cloud regimes -- one in which clouds are prescribed and one in which they
change.  Manabe and Wetherald (1986) find that in both simulations summer
drying occurs at mid-latitudes:  "... it seems signifcant that all the
experiments discussed in this report indicate C02-induced summer reduction and
winter enhancement of soil wetness over extensive, mid-continental regions in
middle and high latitudes ... it is likely that the basic conclusion is valid
despite the imperfections of the model."  If Manabe is correct, it means that
while regional changes are difficult to state with confidence, some can be
projected.
                          * * *  DRAFT FINAL  * * *

-------
                                    6-25



                               EXHIBIT 6-12

                      Regions of U.S.:  Change in Runoff
         60

         50

         40

         30
     UJ
     2  20
     a
     i
     u  |0
     X
          0

        -10

        -20

        -30
The general circulation model of the Goddard  Institute  for  Space Studies was
used to simulate changes in hydrology due  to  increases  in C02  and other trace
gases.   Estimates were made for 23 gridded regions  of the United States.  The
graph above shows the distribution of changes  in precipitation runoff:
increases in 14 regions and decreases in 9 regions.

Source:  Rind and Lebedeff, 1984.
                              -   DRAFT FINAL

-------
                                   6-26
EFFECTS OF POSSIBLE CONTROL OF GREENHOUSE  ~ASES ON  THE  STRATOSPHERE

    Stratospheric perturbants that are greenhouse gases may also alter
stratospheric structure and composition.  Limiting chlorofluorocarbon
emissions in the future would decrease both stratospheric ozone depletion and
global warming.  This relationship is straightforward.  For other perturbants,
however, the situation is more ambiguous.

    Miller and Mintzer (1986) point out that emissions of some stratospheric
perturbants may eventually be reduced by governments in an effort to limit
global warming.  Targets could include nitrous oxide (N20), carbon dioxide
(C02) or methane (CH4).  Decreases in nitrous oxide would lower global
temperatures (Ramanathan et al.,  1985).  The net effect on ozone depends on
chlorine concentrations (Stolarski, personal communication).  In many
scenarios considered plausible, lowering N20 to reduce global warming might
actually excaberate depletion.  Decreases in methane would reduce global
warming in three ways:  by reducing its direct radiative effects while
resident in the troposphere (Ramanathan et al., 1985); by altering its
indirect effects on increasing tropospheric and lower stratospheric ozone
(NAS, 1984); and by reducing water vapor added to the stratosphere (NAS,
1984).  Methane emissions might be limited by altering rice cropping
.practices, by altering livestock  rearing practices (and through biotechnology
to reduce livestock methaneogenis), by reducing forest burning, or controlling
pipeline leaks.  Or methane might be limited through indirect means, that is
by controlling carbon monoxide, emissions that would reduce the lifetime of
methane molecules (Miller and Mintzer, 1986).  As such, a decrease in methane'
would allow  increasing CFC concentrations to deplete column ozone more
effectively.

    Carbon dioxide cools the stratosphere, slowing the process of ozone loss
(Connell and Wuebbles, 1986).  Consequently, model scenarios that include C02
growth reduce depletion (Connell  and Wuebbles, 1986).  Efforts to limit the
growth of C02 could include altering energy mixes (more nuclear or solar
energy used), conservation (less  energy to do the same work), and altering
land clearing practices (Seidel and Keyes, 1983).  Such actions to limit C02
could inadvertantly increase the  vulnerability of the stratospheric ozone
layer.

    Decreases in carbon monoxide  (CO) emissions are projected from combustion
(Kavanaugh,  1986).  Decreases in  CO emissions may also be possible if forest
burning decreases (Hoffman and Wells, 1986).  Since carbon monoxide plays a
large role in OH concentrations,  which in turn, influence the lifetime of
methane, changes in carbon monoxide emissions could alter methane
concentrations (Khalil and Rasmussen, 1985).  If CO concentrations decrease,
the lifetime of methane would fall, and there would be a tendency to lower
concentrations (Thompson and Cicerone, 1986).  The impact of decreasing
methane (CH4) concentrations would not only be to decrease global warming, but
also to make the ozone layer more vulnerable to depletion.
                          * * *  DRAFT FINAL  * * *

-------
                                   6-27
                             APPENDIX A

       DESCRIPTION OF MODEL TO BE USED  IN  INTEGRATING  CHAPTER
    The model used in the integrated assessment chapter at the end of this
document was adapted from a one-dimensional radiative-convective model
developed by the Goddard Institue for Space Studies for estimating temperature
increases associated with atmospheric C02 rises (Hansen et al., 1981).

    The model computes vertical temperature profiles over time from net
radiative and convective energy fluxes.  Radiative fluxes, in turn, depend on
changes in atmospheric gases, and on the associated feedback effects.  The
parameterized equation used here is based on an empirical fit to the
radiative-convective model.

    The fitted equation, described in Hansen et al. (1981), was developed for
C02 only.  It was later modified by Lacis (personal communication) to
incorporate CFC-11, CFC-12, CH4, and N20.  Other minor trace gases were
incorporated based on the work of Ramanthan et al. (1985).

Radiative Forcing Equations

    The equations in the model are:

C02:      FC02 = In [1 + .942*C02/(1+.00062*C02) + .0088-C022 +
          3.26*10-6*C023 + .156-C021'3 exp (-C02/760)]


CH4:      FCH4 = [ . 394*CH4'66 + . 16--CH4 exp (-1.6--CH4] /  (1 +  .169-CH4'62)


N20:      FN20 = 1.556 In [1 + 1.098 (1+.032*N20)*N20'?? / (1+.0014*N202)]

.Overlap:  FOVL = .14 In [1 + .636 (CH4*N20)'75 + (.007*CH4 (CH4*N20)l'52j

Fll:      FF11 = .066*F11

F12:      FF12 = .084*F12

ATrad = FC02(now)-FC02(ref) = FCH4(now)-FCH4(ref) + FN20(now)-FN20(ref) -
FOVL(now)+FOVL(ref) = FFll(now)-FFll(ref) + FF12(now)-FF12(ref)

For C02 from 300 to 600:  ATrad = FC02(600) - FC02(300) = 1.26°

Where:  ATrad is the radiative forcing in year t, expressed in degrees
centigrade (C).  The model is initialized to the year 1880 ("ref" = 1880).
                              »  DRAFT FINAL  * * *

-------
                                   6-28
C02 is the C02 concentration in year t, in parts per million (ppm).
C02(ref) = 270.

CH4 is the CH4 concentration in year t, in ppm.  CH4(ref) = 1.02.

N20 is the N20 concentration in year t, in ppm.  N20(ref) = 0.2853.

Fll is the CFC-11 concentration in year t, in parts per billion  (ppb).
Fll(ref) = 0.0

F12(t) is the CFC-12 concentration in year t, in ppb. F12(ref) = 0.0

    For the other trace gases incorporated in the model, the radiative forcing
in any year is computed in the same fashion as CFC-11 and CFC-12:
                         ATrad,.. = a. » X.
                              (j)    J    Jt

where:  ATrad,.. = radiative forcing in degrees C in year t for compound "j1

        a.       = sensitivity to a uniform 1 ppb increase in the trace gas

                   concentration (0 to 1 ppb) for compound "j".

        x        = the concentration in ppb in year t for compound "j".


The sensitivities were extracted from the work of Ramanathan et al.,  (1985)
and are:

                                                 "a"
                                      Direct radiative forcing
                                      for 0 to 1 ppb increase
                  Trace Gas                 (degrees C)

                  S02                           0.01
                  F14                           0.06
                  F116                          0.13
                  CFC-22                        0.05
                  CFC-13                        0.22
                  CH2C12                        0.03
                  CHC13                         0.06
                  CC14                          0.08
                  CH3CC13                       0.02
                  Halon 1301                    0.17
                  C2H2                          0.02
                          * * *  DRAFT FINAL  *

-------
                                   6-29
    The total radiative forcing (the sum of the radirtive forcing from each
trace gas) is the basis for estimating heat flux ito the earth's surface. The
following equations, from Hansen et al. (1984), are used:

ATeq = f.ATrad

           Fo(2*C02)
F (W/m2) = 	 (ATeq - AT) = global flux into the ocean
           ATeq(2*C02)

AT = is the current global temperature change from initial reference state
We use the flux conversion efficiency of Model II for global flux into the
ocean as determined by the 2*C02 experiment.

F (W/m2) = 3.58.(ATrad - AT/f)*(0.7)
F (W/m2)/697.4 = F (cal*min-^cm-2)
    The heat flux is estimated for time periods ranging from each month to
each year ( a semimonthly time step was used in this study).   The appropriate
AT value for calculating F(t) in each time period (t=n) is the value
estimated for the previous period (t=n-l).  For a simple one-layer ocean
model, AT is obtained by solving the following differential equation:

                                d A T   F(t)
                                 dt      C
                                          o

Where:  C  is the heat capacity of the mixed layer of the ocean per unit

area (cal cm-2).

Diffusion  of  Heat in  the  Ocean

    The ocean model consists of a mixed layer of depth H  = 100m and a
thermocline with 63 layers and depth H = 900m.  The mixed layer temperature is
assumed to be independent of depth, while the thermocline temperature is
defined by a diffusion equation with constant thermal diffusivity.

    The temperature change in the mixed layer (AT) is a solution of the
                                                   m
equation:


                 CHm   d  A  Tm  = F(t) + F
                          dt
                          * * *  DRAFT FINAL  * * *

-------
                                   6-30
where C is the heat capacity of water, H  is the depth of the mixed layer,
F(t) is the heat flux from the atmosphere into the ocean, and

                 FD(t) = - X 8 A T  |
                               9 Z I  Z=Hm
is the heat flux from the thermocline into the mixed layer.  Note that our
z-axis is directed toward the bottom of the ocean.  Also, we use g, cm, sec
and cal, so the heat conductivity lamda is numerically equal to the heat
diffusivity K.  Different values for diffusivity may be chosen.

    The temperature change in the thermocline (AT) is determined by the
diffusion equation:


             c  3A  T(z.t')  =  X  32A T(z,t)
                    6A               5   z2
    The boundary conditions for  A T are:
                        AT = AT   atz = H
                                 m          m
and zero heat flux at the bottom of the thermocline:
             X3A  T  =  0 at z = H -1- H  .
                                      m
Thus it is assumed that no energy escapes through the lower boundary of the
thermocline.  Note that A T  and A T are temperature changes of the
                           m                         	 —
mixed layer and the thermocline between the initial time (1880) and time t.
It is assumed that in the year 1880 AT  = A T = 0 and thus that the
                                       m
ocean temperature was in a state of equilibrium with the atmosphere at that
time.
                            * *  DRAFT FINAL  * * *

-------
                                   6-31
    The use of diffusivity coefficients as a surrogate for all circulation
processes that transport heat may fail to describe the time paths well, with
the downward heat transport in the latter period probably being overestimated
because of increasing oceanic stability.  Nevertheless this method simplified
the problem substantially and a range of different possible coefficients can
be used to investigate the sensitivity of the overall estimate to the rate of
downward heat transport.

    Using transient tracers, Broecker and Peng (1982) developed an average
eddy diffusion coefficient for the ocean of 1.7 cm2/sec.   This value,
accepted by many as a reasonable surrogate measure for estimating heat
absorption, is the reference value in our model.   To test the sensitivity of
the results to this choice, other values may be chosen.

    The range of coefficients tested, from 0.85 to 3.4 cm2/sec, covers the
range of mean oceanwide mixing rates as determined by the National Science
Foundation-sponsored transient tracer experiment and others.  Variation in
estimates of the exact value of the data depend on the tracer used and the
statistical method of computing the global mean.   The values of 0.2 and 4.0
cm2/sec were tested in earlier work to account for the possibility of
dramatic changes in ocean transports due to deglaciation and climate change
(Hoffman et al., 1983).
                          * * *  DRAFT FINAL  * * *

-------
                            6-32
                      APPENDIX B
 Trace Gas Scenarios Used in Hansen  et al.,  1984
YEAR
1850
1900
1950
1960
1970
1980
1990
2000
2010
C02
(ppm)
270
291
312
317
326
338
353
372
396
CFG 11
(ppt)
0
0
7
33
126
308
479
638
787
CFC12
(ppt)
0
0
1
11
62
178
280
369
447
CH4
(ppb)
1400
1400
1400
1416
1500
1650
1815
1996
2196
N20
(ppb)
295
295
295
295
295
301
307
313
320
Trace Gas Scenarios Used  in Hoffman et al.,  1985
YEAR
1880
1900
1950
1960
1970
1980
1990
2000
2010
2020
2030
C02
(ppm)
280
288
311
316
326
339
355
373
398
431
470
CFC11
(ppt)
0
0
1
11
64
185
341
599
992
1500
2116
CFC12
(ppt)
0
0
8
36
137
337
577
903
1371
2023
2863
CH4
(ppb)
1402
1448
1568
1593
1619
-1645
1822
2093
2406
2764
3177
N20
(ppb)
295
296
298
299
299
300
307
313
320
328
338
                   * * *  DRAFT FINAL  * * *

-------
                                  • 6-33
REFERENCES

Bradley, R.S., and P.D.  Jones,  (1985),  "Data Bases for Isolating the Effects
of the Increasing Carbon Dioxide Concentration," in MacCracken,  M.C.,  and  F.M.
Luther (eds.)> Protecting the Climatic  Effects of Increasing Carbon Dioxide,
DOE/ER-0235, U.S. Department of Energy, Washington, D.C.

Broecker, W.S., (1986),  Statement for Ozone Depletion, the Greenhouse Effect,
and Climate Change, Hearings before the Subcommittee on Environmental
Pollution of the Committee on Environment and Public Works,  United States,
Senate, Ninety-Ninth Congress,  Second Session, June 10 and 11,  1986.

Broecker, W.S., and T. Peng, (1982), Tracers in the Sea.   Palisades,  NY:
Lament-Doherty Geophysical Observatory.  Columbia University.

Connell, P.S., and O.J.  Wuebbles, (1986), Ozone Perturbations  in the LLNL  One-
Dimensional Model -- Calculated Effects of Projected Trends  in  CFC's,  CH4,
C02, N20, and Halons Over 90 Years, Lawrence Livermore National  Laboratory,
Livermore, CA.

Dickinson, R.E., and R.J. Cicerone, (1986), "Future Global Warming from
Atmospheric Trace Gases", Nature: 319,  109-115.

Gilliland, R.L., and S.H. Schneider, (1984), "Volcanic, C02 and Solar Forcing
of Recent Climatic Changes" Nature: 310, 38-41.

Hansen, J.E.,  (1986), Statement for Ozone Depletion, the Greenhouse Effect,
and Climate Change, Hearings before the Subcommittee on Environmental
Pollution of the Committee on Environment and Public Works,  United States,
Senate, Ninety-Ninth Congress,  Second Session, June 10 and 11,  1986.

Hansen, J., A. Lacis, D. Rind,  G. Russell, I. Fung, and S. Lebedeff, (1986),
"Evidence for Future Warming:  How Large and When," presented at the C02
Conference of the National Forest Products Association, June 25-27, 1984,
Boulder, Colorado, in press.

Hansen, J., A. Lacis, D. Rind,  G. Russell, P. Stone, I. Fung,  R. Ruedy, and  J.
Lerner, (1984), "Climate Sensitivity:  Analysis of Feedback Mechanisms,"  in
Hansen, J.E., and T. Takahashi, (eds.), Climate Processes and Climate
Sensitivity, Geophysical Monograph 29,  Maurice Ewing Volume 5,  American
Geophysical Union, Washington,  DC.

Hansen, J., D. Johnson,  A. Lacis, S. Lebedeff, P. Lee, D. Rind,  and G. Russell,
(1981), "Climate Impacts of Increasing Atmospheric C02, Science, 213:957-966.

Hoffert, M.I., and B.P.  Flannery, (1985), "Model Projections of the
Equilibrium Climatic Response to Increased Carbon Dioxide," in MacCracken,
M.C., and F.M. Luther (eds.), Projecting the Climatic Effects of Increasing
Carbon Dioxide, DOE/ER-0237, U.S. Department of Energy, Washington, D.C.
                                 DRAFT FINAL  * * *

-------
                                   6-34
Hoffman, J.S. and J.B. Wells, (1986) Trace Gas Scenarios,  presented at the
UNEP Economic Workshop on Protection of the Ozone Layer, May 26-30, 1986,
Rome, Italy.

Hoffman, J.S., J.B. Wells, and J.G. Titus, (1985) "Future Global Warming and
Sea Level Rise," in G. Sigbjarnarson (ed.), Iceland Coastal and River
Symposium, Proceedings, Library of the National Energy Authority, Reykjavik,
Iceland.

Hoyt, D.V.,  (1979), "Variations in Sunspot Structure and Climate," Climatic
Change: 21,  79-92.

Isaksen, I.S.A., and F. Stordal, (1986), "Ozone Perturbations by Enhanced
Levels of CFCs, N20, and CH4:  A Two-Dimensional Diabatic Circulation Study
Including Uncertainty Estimates," Journal of Geophysical Research, 91(D4):
5249-5263.

Kavanaugh, M. , (1986), "Estimates of F.uture CO, N20, and NOx Emissions from
Energy Combustion," to appear in Atmospheric Environment (in press).

Khalil, M.A.K., and R.A. Rasmussen, (1985), "Causes of Increasing Atmospheric
Methane:  Depletion of Hydroxyl Radicals and the Rise of Emissions,"
Atmospheric Environment, 19(3): 397-407.

Lacis, A.A.  and D.J. Wuebbles, (1986) "Radiative Forcing of Global Climate
Changes in Stratospheric Water Vapor" to be submitted to Geophysical  Research
Letters.

Lacis, A.A., D.J. Wuebbles, and J.A. Logan, (1986), "Radiative Forcing of
Global Climate by Changes in the Vertical Distribution of Atmospheric Ozone,"
to be submitted to Geophysical Research Letters.

Luther, F.M., and R.G. Ellingson, (1985), "Carbon Dioxide and the Radiation
Budget," in MacCracken, M.C., and F.M. Luther (eds.), Projecting the  Climatic
Effects of Increasing Carbon Dioxide, DOE/ER-0237, U.S. Department of Energy,
Washington, D.C.

Luther, F.M., and M.C. MacCracken, (1985), "Recommendations for Research and
Modeling Activities for Projecting the Climatic Effects of Increasing Carbon
Dioxide," in MacCracken, M.C., and F.M. Luther (eds.-)  Projecting the Climatic
Effects of Increasing Carbon Dioxide, DOE/ER-0237, U.S. Department of Energy,
Washington, D.C.

MacCracken, M.C. and F.M. Luther, (1984), "Radiative and Climatic Effects of
the El Chichon Eruption," Geofisica International 23(3):385-401.

Manabe, S.,  and R.T. Wetherald, (1975), "The Effect of Doubling the C02
Concentration in the Climate of a General Circulation Model," Journal of
Atmospheric  Sciences 32:3-15.
                          * * *  DRAFT FINAL  * * *

-------
                                   6-35
Manabe, S., and R.T. Wetherald, (1986), "Reduction in Summer Soil Wetness
Induced by an Increase in Atmospheric Carbon Dioxide," Science, 232: 626-628.

Miller, A.S., and I.M. Mintzer, (1986), '"Draining the Sink1 Policy
Implications of CFC Growth Trajectories," discussion draft prepared for the
U.S. Environmental Protection Agency Workshop, '"Protecting the Ozone Layer'
Workshop of Demand and Control Technologies", March 6-7, 1986, Washington, D.C.

National Academy of Sciences, (1984), Causes and Effects of Changes in
Stratospheric Ozone and Update 1983, National Academy Press, Washington, D.C.

National Academy of Sciences, (1982), Carbon Dioxide and Climate:  A Second
Assessment, National Academy Press, Washington, D.C.

National Academy of Sciences, (1979), Carbon Dioxide and Climate:  A
Scientific Assessment, National Academy Press, Washington, D.C.

National Aeronautics and Space Administration (NASA), (1982), "Radiative
Effect of the El Chichon Volcanic Eruption:  Preliminary Results Concerning
Remote Sensing" (NASA TM 84959) NASA, Washington, DC.

Neftel, A., H. Oeschger, J. Schwander, B. Stauffer, and R. Zumbrunn, (1982),
"ice Core Sample Measurements Give Atmospheric C02 Content During the Past
40,000 Yr.," Nature, 295: 220-223.

Newkirk, G., Jr., (1983), "Variations in Solar Luminosity", Annual Reviews of
Astronomy and Astrophysics.  21: 429-467.

Oeschger, H., B. Stauffer, A. Neftel, J. Schwander, and R. Zumbrunn, (1982),
"Atmospheric C02 Content in the Past Deduced from Ice-Core Analyses,"  Annals
of Glaciology, 3: 227-232.

Pearman, G.I., D. Etheridge, F. deSilva, and P.J. Fraser, (1986), "Evidence of
Changing Concentrations of Atmospheric C02, N20, and CH4 from Air Bubbles in
Antarctic Ice", Nature, 320: 248-250.

Peng, T.H., W.S. Broecker, H.D. Freyer, and S. Trumbore, (1983), "A
Deconvolution of the Tree Ring Based d!3C Record," Journal of Geophysical
Research, 88: 3609-3620.

Pollack, J.B., O.B. Toon, E.F. Hoffman, O.J. Rosen, and J.M. Rosen, (1983)
"The El Chichon Volcanic Cloud:  An Introduction," Geophysical Research
Letters 10:989-992.

Ramanathan, V., R.J. Cicerone, H.B. Singh, and J.T. Kiehl,  (1985), "Trace Gas
Trends and their Potential Role in Climate Change," Journal of Geophysical
Research, 90(D3):55457-5566.

Rind, D., and S. Lebedeff, (1984), Potential Climatic Impacts of Increasing
Atmospheric C02 with Emphasis on Water Availability and Hydrology in the
United States, U.S. Environmental Protection Agency, Washington, D.C.
                          * * *  DRAFT FINAL  * *

-------
                                   6-36
Sarmiento, J., and K. Bryan, (1982), "An Ocean Transport Model for the North
Atlantic," Journal of Geophysical Research 87:9694-9698.

Schlesinger, M.E., and J.F.B. Mitchell, (1985), "Model Projections of the
Equilibrium Climatic Response to Increased Carbon Dioxide," in MacCracken,
M.C., and F.M. Luther (eds.)> Projecting the Climatic Effects of Increasing
Carbon Dioxide, DOE/ER-0237, U.S. Department of Energy, Washington, D.C.

Seidel, S. and D. Keyes,  (1983), Can We Delay a Greenhouse Warming? U.S.
Environmental Protection Agency, Washington, D.C.

Somerville, R.C.J., and L.A. Remer, (1984), "Cloud Optical Thickness Feedbacks
in the C02 Climate Problem," Journal of Geophysical Research, 89(d6):
9668-9672.

Stordal, F., and I.S.A. Isaksen, (1986), Ozone Perturbations Due to Increases
in N20, CH4, and Chlorocarbons:   Two-Dimensional Time Dependent Calculations,
prepared for UNEP Economic Workshop on the Ozone Layer, Rome, Italy, May
26-30, 1986.

Strain, B.R., (1985), "Background on the Response of Vegetation to Atmospheric
Carbon Dioxide Enrichment," in Strain, B.R., and J.D. Cure, (eds.) Direct
Effects of Increasing Carbon Dioxide on Vegetation, DOE/ER-0238, U.S.
Department of Energy, Washington, D.C.

Thompson, A.M., and R.J.  Cicerone, (1986), "Atmospheric CH4, CO and OH'from
1860 to 1985," Nature, in press.

Tyndall, J., (1861), "On the Absorption and Radiation of Heat by Gases and
Vapours, and on the Physical Connexion of Radiation, Absorption, and
Conduction," Philosophical Magazine and Journal of Science, 22(146): 169-194.

Wang, W.C., D.J. Wuebbles, and W.M. Washington, (1985), "Potential Effects of
Perturbations Other than Carbon Dioxide," in MacCracken, M.C., and F.M. Luther
(eds.), Projecting the Climatic Effects of Increasing Carbon Dioxide,
DOE/ER-0237, U.S. Department of Energy, Washington, D.C.

Washington, W.M., and G.A. Meehl, (1984), "Seasonal Cycle Experiment on the
Climate Sensitivity Due to a Doubling of C02 With an Atmospheric General
Circulation Model Coupled to a Simple Mixed-Layer Ocean Model," Journal of
Geophysical Research, 89(D6): 9475-9503.

Webb, T., Ill and T.M.L.  Wigley, (1985), "What Past Climates Can Indicate
about a Warmer World," in MacCracken, M.C., and F.M. Luther (eds.), Projecting
the Climatic Effects of Increasing Carbon Dioxide, DOE/ER-0237, U.S.
Department of Energy, Washington, D.C.

Wigley, T.M.L., J. Angell, and P.O. Jones, (1985), "Analysis of the
Temperature Record," in MacCracken, M.C., and F.M. Luther (eds.), Projecting
the Climatic Effects of Increasing Carbon Dioxide, DOE/ER-0237, U.S.
Department of Energy, Washington, D.C.
                          * * -  DRAFT FINAL  * * *

-------
                                   6-37
Woods, J.D., (1985), "The World Ocean Circulation Experiment," Nature,  314:
501-511.

World Meteorological Organization, (1986a), Report of the International
Conference on the Assessment of the Role of Carbon Dioxide and of other
Greenhouse Gases in Climate Variations and Associated Impacts; Villach,
Austria, 9-15 October 1984, WMO-No. 661, International Council of Scientific
Unions, United Nations Environment Programme, and World Meteorological
Organization.

World Meteorological Organization, (1986b), Atmospheric Ozone 1985.
Assessment of Our Understanding of the Processes Controlling its Present
Distribution and Change, Global Ozone Research and Monitoring Project --
Report No. 16,  WMO, Geneva, Switzerland.

World Meteorological Organization (WMO), (1984), "The International  Satellite
cloud Climatology Project (ISCCP)", Report of the Third Session of the
International Working Group on Data Management, March 6-8, 1984, Tokyo.  World
Climate Programme Report WCP-82, WMO, Geneva, Switzerland.
                          * * *  DRAFT FINAL  * * *

-------