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 * * *
------- |