United States
         Environmental Protection
             United Nations
             Environment Programme
August 1986
Volume I: Overview  170OPPE861

Library of Congress Cataloging - in - Publication Data
Effects of changes in stratospheric ozone and global climate.
Proceedings of a conference convened by the United Nations Environment
Programme and the U.S. Environmental Protection Agency.
Contents:  v. 1. Overview — v. 2.  Stratospheric ozone — v. 3.  Climate
change. — v. 4.  Sea level rise
1. Atmospheric ozone—Reduction—Congresses.  2. Stratosphere—Con-
gresses.   3. Global  temperature  changes—Congresses.    4.  Climatic
changes—Congresses.  5. Sea level—Congresses.   6.  Greenhouse effect,
Atmospheric—Congresses.  7.  Ultraviolet   radiation—Congresses.   I.
Titus, James G    II. United States Environmental Protection Agency.  III.
United Nations  Environment  Programme.

                  Volume 1:  Overview
                          Edited by
                        James G. Titus
              U.S. Environmental Protection Agency
        This report represents the proceedings of the INTERNATIONAL CON-
        United Nations Environment Programme and the U.S. Environmental
        Protection Agency. The purpose of the conference was to make available
        the widest possible set of views. Accordingly, the views expressed herein
        are solely those of the authors and do not represent official positions of
        either agency.  Mention of trade names or commercial products does not
        constitute endorsement or recommendation for use.
                                    U.S. EnvirosE&rtsl Frr/tec-tira
                                    Great Lakes National  rro£ra,u
                                              GLKPO Li

     This report  examines  the possible consequences of  projected changes in
stratospheric ozone  and global  climate  resulting from  emissions of chloro-
fluorocarbons,  carbon  dioxide,  methane,  and other  gases released  by  human
activities.    During  the week  of June 16-20, the United Nations Environment
Programme  and   the   U.S.   Environmental   Protection   Agency  sponsored  an
International Conference  on  the Health  and  Environmental Effects  of  Ozone
Modification  and  Climate   Change,   which  was  attended   by  scientists  and
officials from  approximately  twenty countries  from  all  areas  of the world.
The four volumes of this report comprise the proceedings of that conference.

     Volume  1 provides  an  overview  of  the issues  as well as the  introductory
remarks and  reactions  from top  officials of the  United  Nations Environment
Programme and the United  States Environmental  Protection Agency,  two  U.S.
Senators,  and  representatives  from  industry,  academia,  and  environmental
groups.   Volumes 2,  3, and  4  provide more  detailed  investigations on the
effects of  ozone depletion, climate change,  and  the rise in  sea level  that
might result from a global  warming.

     We wish to thank the many people who helped  us  to publish  this report on
time.     Mary  Dalrymple  and   Wendy  Martin  of  Technology  Applications,
Incorporated  (TAI)  coordinated  a team  of  copy  editors  that  included  Karen
Spear,  Linda Stiles,  June  Leatherman,  and Susan MacMillan.    Sandra Miller
directed a  team of  word processors that included  Lynette  Dunaway,  Felicia
Jenkins, and Patricia Maddox.   All of these  people worked many long nights.

     We also  wish  to  thank Maria Tikoff  and  Stephen Seidel,  who coordinated
the presentations at the conference  and whose perpetually good nature helped
keep us enthusiastic,  and Judy Salmon of TAI who coordinated the logistics for
the conference.   Peter Usher of  UNEP and  John Hoffman of EPA provided overall

     With the exception of  the papers  by  Lee Thomas and Genady Golubev,  this
report does  not  present the official views of  either  the U.S. Environmental
Protection Agency or  the United Nations Environment Programme.


Overview  of the Effects of Changing the Atmosphere
     James G. Titus and Stephen Seidel 	     3

Global Environmental Change:  The UNEP Perspective
     Genady N. Golubev 	    21

Global Environmental Change:  The EPA Perspective
     Lee  Thomas 	    27

Global Environmental Change:  The International Perspective
     Richard E. Benedick 	    31

Man's Impact on Earth's Atmosphere
     James P. Bruce 	    35

The Importance of Knowing Sooner
     John S. Hoffman 	    53

Our Global Environment:  The Next Challenge
     John H. Chafee 	    59

The Greenhouse Effect
     Albert Gore,  Jr	    63

Atmospheric Ozone
     Robert T.  Watson 	   69

Ozone Perturbations Due to Increases in NpO, CHj, and Chlorocarbons:
  Two-Dimensional Time Dependent Calculations
     Frode Stordal and Ivar S.A.  Isaksen 	   83

The Ultraviolet Radiation Environment of the Biosphere
     John E. Frederick 	   121

Health Effects of Ultraviolet Radiation
     Edward A.  Emmett 	   129

Ozone Depletion and Ocular Risks from Ultraviolet Radiation
     Morris Waxier 	:	   14?

Overview of Our Current State of Knowledge  of UV Effects on Plants
     Alan H. Teramura 	   165

The Effect of Solar UV-B Radiation on Aquatic Systems:  An Overview
     R.C. Worrest 	   175
                                   CLIMATE CHANGE

The Greenhouse Effect:  Projections of Global Climate Change
     J. Hansen, A. Lacis, D. Rind, G. Russell, I. Fung,
     P. Ashcraft, S. Lebedeff, R. Ruedy, and P. Stone	   199

The Causes and Effects of Sea Level Rise
     James G.  Titus 	   219

Reduction in Summer Soil Wetness Induced by an Increase in Atmospheric
  Carbon Dioxide
     S. Manabe and R.T. Wetherald 	   249

Effects of Climatic Changes on Agriculture and Forestry:
  An Overview
     Martin L. Parry and Timothy R. Carter 	   257

The Water Resource Impact of Future Climate Change and Variability
     Max Beran 	   299

                                REVIEW AND  ASSESSMENT
Determining the Options - The Role of UNEP in Addressing Global Issues
     Peter Usher [[[  331

The Policy Context
     Gus Speth [[[  335

The Need for Responsible CFC Policy
     Richard Barnett [[[  341

Cooling the Chemical Summer:  Some Policy Responses to Ozone-
  Destroying and Greenhouse Gases
     David D. Doniger and David A. Wirth .....................................
Climate Change and Stratospheric Ozone Depletion:
  Need for More Than the Current Minimalist Response


Overview of the Effects of Changing the Atmosphere

James G. Titus and Stephen R. Seidel
Environmental Protection Agency
Washington, DC USA

     Society is  conducting  a global  experiment  on  the earth's  atmosphere.
Human activities are increasing the  worldwide  atmospheric  concentrations of
chlorofluorocarbons,  carbon  dioxide,  methane,  and  several other  gases.   A
growing body of  scientific evidence suggests that if  these trends  continue,
stratospheric ozone may decline and global temperature  may  rise.   Because  the
ozone layer shields  the  earth's surface  from  damaging ultraviolet  radiation
(UV) future depletion could  increase the incidence of  skin cancer  and  other
diseases,  reduce  crop yields, damage materials, and place additional stress on
aquatic plants  and animals.    A global warming from the "greenhouse  effect"
could also threaten human health,  crop yields,  property, fish, and  wildlife.
Precipitation and storm  patterns could change, and  the level of  the oceans
could eventually  rise.

     To  improve  the  world's  understanding  of  these and  other  potential
implications of  global atmospheric  changes,  the  United Nations  Environment
Programme (UNEP)  and  the  U.S. Environmental Protection  Agency  (EPA)  sponsored
an International  Conference  on  the  Health and  Environmental Effects of  Ozone
Modification and Climate  Change during  the week  of June  16-20,  1986.    The
conference brought  together  over  three hundred  researchers  and policy makers
from approximately  twenty nations.   This  four-volume  report  presents  the
seventy-three papers that were  delivered at the  conference  by  over eighty
speakers,  including  two U.S.  Senators,  top  officials from  UNEP and  EPA, some
of  the  leading  scientists   investigating  the implications  of  atmospheric
change,   and  representatives  from  industry and  environmental  groups.    In
addition to this summary  of  the entire report,  Volume  1  presents  a  series of
overview papers  describing each of  the major areas of research on  the  effects
of atmospheric  change, as well as policy assessments  of these  issues by  well-
known  leaders  in  government,  industry,   and  the  environmental   community.
Volumes 2,  3,  and  4 present the more  specialized papers  on  the  impacts of
ozone modification,  climate  change,  and sea  level  rise,   respectively,  and
provide some of the latest research  in these areas.


Atmospheric Processes

     The ozone  in the  upper part of  the   atmosphere—known as  the  strato-
sphere—is  created  by  ultraviolet radiation.   Oxygen  (02)  is  continuously
converted  to  ozone  (Oo)  and back to  02 by numerous  photochemical  reactions
that  take  place  in  the  stratosphere,  as  Stordal  and  Isaksen  (Volume  1)
describe.   Chlorofluorocarbons  and other  gases released by  human activities
could  alter  the  current  balance of   creative  and  destructive  processes.
Because  CFCs  are  very  stable  compounds,  they  do  not break up in  the lower
atmosphere  (known as  the troposphere).   Instead,  they  slowly  migrate  to the
stratosphere,   where  ultraviolet   radiation  breaks  them   down,  releasing

     Chlorine acts as a  catalyst to destroy  ozone;  it promotes reactions that
destroy ozone without being  consumed.  A chlorine  (Cl) atom reacts with ozone
(Oo) to  form  CIO  and 02.   The CIO later  reacts  with another  0,  to form two
molecules  of  Op,  which  releases  the  chlorine  atom.   Thus,  two molecules  of
ozone are converted  to three molecules  of ordinary oxygen,  and the chlorine is
once again free  to start the process.   A.  single  chlorine  atom  can  destroy
thousands  of  ozone  molecules.   Eventually,  it returns to  the  troposphere,
where it is rained out as hydrochloric  acid.

     Stordal and Isaksen  point out that  CFCs are  not the only gas released by
human   activities   that   might    alter   the   ozone   balance.     Increasing
concentrations of methane in the  troposphere increase the  water  vapor  in the
stratosphere,  which helps create  ozone.   Carbon dioxide and other greenhouse
gases  (discussed  below)  warm  the  earth's   surface  but  cool   the   upper
atmosphere; cooler stratospheric  temperatures slow the process of ozone  deple-
tion.  Nitrous oxide (N20) reacts  with  both chlorine and ozone.

     Stordal  and  Isakson  present  results  of  possible  ozone  depletion  over
time, using their two-dimensional atmospheric-chemistry model.   Unlike  one-
dimensional models which  provide changes in  ozone  in the global average,  this
model calculates  changes for specific  latitutdes and seasons.   The  results
show that  if concentrations of the relevant trace gases grow at recent levels,
global  average  ozone   depletion  by 2030 would  be  6.5  percent.   However,
countries  in the higher latitudes  (60°N)  would experience 16 percent depletion
during  spring.   Even  in  the  case of  constant CFC  emissions, where  global
average  depletion would  be  2  percent  by  2030,  average  depletion would  be 8
percent in  the high northern latitudes.

     Watson (Volume 1) presents evidence that ozone has been changing recently
more than  atmospheric models had predicted.   As Plate 1  shows, the ozone over
Antarctica  during the  month  of   October  appears  to have  declined over  40
percent  in the  last six  to eight  years.  Watson  also discusses observations
from ozone monitors that suggest  a 2 to 3 percent worldwide reduction in ozone
in the upper portion of the stratosphere (thirty to forty kilometers above the
surface),  which  is  consistent with model  predictions.  Finally,  he presents
preliminary data  showing a  small  decrease  since  1978  in  the total (column)
ozone worldwide.  However,  he strongly  emphasizes  that  the data have not yet
been  fully reviewed  and that it  is  not  possible to conclusively  attribute
observed ozone  depletion to  the  gases  released  by human  activities.   While

there  are  several  hypotheses  to  explain  why  ozone  concentrations  have
declined,  none  have  been  adequately   established;   nor   did  any  of  the
atmospheric models predict the measured loss of ozone over Antarctica.

Ultraviolet Radiation

     Many  of  the  chemical  reactions  investigated  by  atmospheric  scientists
take  place only  in the  stratosphere  because  they  are caused  by  types  of
radiation  only  found  in  the  upper  atmosphere.    As  Frederick  (Volume  1)
explains, the sun  emits radiation  over a  broad  range of wavelengths, to which
the human eye responds in the  region from approximately 400 to 700 nanometers
(nm).  Wavelengths from 320  to 400 nm are known as UV-A; wavelengths from 280
to 320 nm are called UV-B,  and wavelengths from  200 to 280  nm  are known  as

     Frederick explains why  attention has primarily focused on  the UV-B part
of the spectrum.   The  atmosphere absorbs  virtually  all UV-C,  and is expected
to continue to do  so under all foreseeable circumstances.   On the other hand,
UV-A  is  not absorbed by  ozone. .   By  contrast, UV-B is partially absorbed  by
ozone, and future depletion would reduce the effectiveness of this shield.

     We  now   examine  the  potential  implications  of  such  changes  on  human
health, plants, aquatic organisms, materials, and air pollution.

Effects on Human Health

     The  evidence  suggests  that   solar  ultraviolet  radiation  induces  skin
cancer,  cataracts,  suppression of the  human  immune  response  system,  and
(indirectly  through  immunosuppression)  the development  of  some  cutaneous
infections, such as herpes.  Emmett (Volume 1)  discusses the absorption of UV
radiation by  human tissue and  the mechanisms  by which  damage  and  repair may

     Emmett also  examines UV radiation as  the  cause of aging  of the skin and
both  basal and  squamous  skin cancers.  In reviewing the role  of UV radiation
in melanoma   (the  most frequently fatal  skin  cancer), he  states  that  some
evidence suggests  this link, but  that  currently there is no acceptable animal
model that can be used  to explore or validate this relationship.  He concludes
that  future   studies must focus  on  three major  factors—exposure  to  solar
radiation, individual  susceptibility,  and  personal  behavior.   Waxier (Volume
1) presents evidence of a link between UV-B exposure and cataracts.

     Volume 2 presents specific research results and  provides  more detail  on
many  of  the aspects  covered  in this volume.   Scotto presents epidemiological
evidence linking solar radiation with  skin cancers,  other  than melanoma.  His
analysis suggests that  Caucasians in the United States have a 12 to 30 percent
chance of developing these cancers within their lifetimes,  even without ozone
depletion.   Armstrong  examines the role  of UV-B exposure  to melanoma  in  a
study  of  511 matched  melanoma  patients  and  control subjects  in  Western
Australia.   He  shows  that  "intermittent exposure"  to sunlight was closely
associated with this type of cancer.
  However, $2 an<* ^ reflect some UV-A back to space.

     In a paper examining nonmelanoma skin cancer in Kuwait, Kollias and Baqer
(Volume  2)  show  that despite  the  presence  of  protective pigmentation,  75
percent of cancers occur on the 10 percent of the skin exposed to sunlight.  A
second paper on skin cancer presents experimental evidence suggesting that the
mechanism  by  which  skin  cancer   could  occur  involves  disruption  of  the
cytoskeleton  from  exposure  to UV-A  and  UV-B  light  (Zamansky  and  Chow,
Volume 2).

     The  pathways  by which  suppression  of  the  immune   response  might  be
triggered are  explored  in  papers  by  DeFabo  and Noonan,  Daynes et  al.,  and
Elmets et  al (all  Volume 2).   Davies and Forbes  (Volume 2) show  that mice
exposed to UV-B radiation had a decrease  in lifetime that was proportional to
the quantity of  radiation and  not  directly related to the  incidence of skin

     Possible  implications   of  immune   suppression  of   diseases   and  the
mechanisms by which it occurs are still uncertain.  However, several papers in
Volume 2 suggest that in addition to skin cancer and contact hypersensitivity,
diseases influenced  by  UV-B induced immune suppression  include  leishmaniasis
and herpes  infections.   Fisher  et.  al  (Volume 2)  show  that  at  least  one
sunscreen effectively protects  mice exposed  to UV-B radiation  from sunburn;
but it does  not stop the immune suppression from  interfering with  a contact
hypersensitivity (allergic) reaction.

Effects on Plants

     The effects of  increased exposure to UV-B  radiation  on plants has been a
primary  area  of research  for  nearly  a decade.   Teramura  (Volume  1) reports
that of the two hundred plants tested for their sensitivity to UV-B radiation,
over  two-thirds  reacted  adversely;  peas,  beans,  squash,   melons,  and cabbage
appear to  be  the  most  sensitive.   Given  the complexities  in  this  area of
research, he warns that  these  results may  be misleading.   For  example, most
experiments have  used growth chambers.   Studies of plants  in the field have
shown them to be less sensitive to UV-B.   Moreover,  different cultivars of the
same  plant  have   shown  very   different   degrees   of   sensitivity  to  UV-B
radiation.   Finally, Teramura  suggests  that  potential effects  from multiple
stresses  (e.g., UV-B exposure plus water stress or  mineral deficiency) could
substantially alter a plant's response to changes in UV-B alone.

     In  Volume  2,  Teramura draws  extensively  from  the  results of  his five
years of  field  tests on soybeans.    His data  show that  a  25 percent depletion
in ozone  could result in a 20 to  25  percent reduction in  soybean  yield and
adverse  impacts on  the quality  of  that yield.  Because soybeans are the fifth
largest  crop  in  the  world,  a  reduction  in  yields  could  have  serious
consequences for world food  supplies.  However,  some soybean cultivars appear
to be  less  susceptible  to UV-B radiation,  which  suggests  that selective crop
breeding might reduce future  losses,  if  it  does not increase vulnerability to
other environmental stresses.

     BJorn  (Volume 2) examines the  mechanisms  by which plant  damage occurs.
His research relates specific wavelengths with those aspects  of plant growth
that might  be  susceptible,  including  the destruction of  chloroplast, DMA, or
enzymes necessary  for photosynthesis.

Aquatic Organisms

     Aquatic plants would also  be  adversely affected by increased ultraviolet
radiation.  Worrest (Volume 1) points out that most of these plants, which are
drifters  (phytoplankton),  spend much  of their  time  near the  surface  of the
water   (the   euphotic  zone)   and  are  therefore   exposed   to  ultraviolet
radiation.   A reduction  in their productivities  would be  important because
these plants  directly and  indirectly provide  the  food for almost  all fish.
Furthermore,  the  larvae  of many  fish found  in the  euphotic  zone  would  be
directly affected, including crabs, shrimp, and anchovies.  Worrest points out
that fish account for  18  percent of  the animal protein that people around the
world consume, and 40 percent of the protein consumed in Asia.

     An important  question  is  the extent  to which current UV-B  levels are a
constraint on aquatic  organisms.   Calkins  and  Blakefield  (Volume 2) conclude
that some  species are already  exposed to  as much UV-B as  they can tolerate.
Thomson (Volume  2)  shows that  a  10  percent decrease  in  ozone  could increase
the number of abnormal larvae as much as 18 percent.  In a study of anchovies,
a 20 percent  increase in UV-B  radiation over  a 15-day period  caused the loss
of all the larvae within a  10-meter mixed layer in April and August.

     Many other factors could affect  the magnitude  of the impacts on specific
species,  ecosystems,  and the  food  chain.   An important mechanism  by which
species could adapt to higher UV-B incidence would be to reduce their exposure
by moving  further away from  the  water's surface during  certain  times  of the
day or  year  when  exposure  is greatest.  Haeder (Volume 2)  suggests, however,
that for certain species such avoidance may be  impaired by UV-B radiation.

     Even  for  those  organisms  that  could  move to  avoid  exposure,  unwanted
consequences may result.  Calkins and Blakefield present model results showing
that movement by  phytoplankton  away  from sunlight  to reduce exposure to a 10
percent  increase  in  UV-B  would  result in a  2.5  to  5  percent  decrease  in
exposure  to the  photosynthetically   active radiation  on  which  their  growth
depends.   Increased  movement  requires  additional  energy  consumption,  while
changes in location may affect the availability of food for zooplankton, which
could cause other changes in shifts in the aquatic food chain.

     To a  certain  extent,  losses within a  particular species  of plankton may
be compensated by gains in other species.  Although it is possible that no net
change  in  productivity will occur, questions  arise  concerning  the ecological
impacts  on species  diversity  and community  composition  (Kelly,  Volume 1).
Reductions in diversity  may make  populations  more  susceptible  to  changes  in
water temperatures, nutrient availability,  diseases, or pollution.  Changes in
community composition could alter  the protein  content,  dry  weight,  or overall
food value of the initial stages of the aquatic food chain.

Polymer Degradation and Urban Smog

     Current sunlight can cause paints to  fade, transparent window glazing to
yellow,  and  polymer  automobile roofs  to  become chalky.   These  changes are
likely  to  occur more  in places closer  to  the  equator where UV-B radiation is
greater.   They  are  all examples  of  degradation  that  could accelerate  if
depletion of  the  ozone layer occurs.   Andrady and  Horst  (Volume 2) present a
case study  of the  potential  magnitude of  loss due to increased  exposure  to

UV-B radiation on polyvinyl chloride  (PVC).   This  chemical is used in outdoor
applications  where  exposure   to   solar  radiation  occurs  over  a  prolonged
period.   It is also  used  in  the  construction  industry in siding  and window
frames and as a roofing membrane.

     To analyze  the potential  economic  impact of  future ozone  depletion on
PVC, the  authors assumed  that  the future service  life of polymers  would be
maintained by  increasing the quantity of light  stabilizers (titanuim dioxide)
used in the product.   As  a result,  the costs associated  with  increased UV-B
radiation  would   be  roughly  equal  to  the   costs  of  increased  stablizers.
Preliminary  results  show  that for  a  26  percent  depletion  by  2075,  the
undiscounted costs would be $4.7 billion (1984 dollars).

     Increased penetration  of  UV-B radiation  to the earth's surface could play
an  important  role  in  the  formation  of  ground level  oxidants  (smog).   UV-B
affects smog  formation  through the  photolysis of formaldehyde,  from  which
radicals  are   the  main  source for   deriving  chain  reactions  that  generate
photochemical  smog.   Whitten and  Gery (Volume  2)   analyze  the  relationship
between UV-B, smog, and warmer  temperatures.  The  results of this preliminary
study of  Nashville, Philadelphia,  and Los  Angeles  show that  large depletions
in stratospheric ozone and increases  in temperature would increase smog by as
much as 50  percent.   In addition, because oxidants would form  earlier in the
day  and   closer  to population  centers (where  emissions  occur),  risks  from
exposure  could increase  by an  even  higher percentage  increase.   Whitten and
Gery also report a  sensitive relationship  between  UV-B  and hydrogen peroxide,
an oxidant and precursor to acid rain.


The Greenhouse Effect

     Concern about a possible  global warming  focuses largely on the same gases
that may  modify  the stratospheric ozone:  carbon dioxide, methane,  CFCs, and
nitrous oxide.  The report of  a recent conference  convened by UNEP, the World
Meteorological  Organization,   and  the International  Council  of  Scientific
Unions  concluded  that  if  current trends  in  the  emissions  of   these  gases
continue,   the  earth  could  warm a few degrees  (C)  in  the next  fifty  years
(Villach  1985).   In the next century,  the planet  could warm as  much as five
degrees (MAS  1983),  which  would leave  the planet  warmer  than  at any time in
the last  two million years.

     A planet's temperature is  determined  primarily by  the amount of sunlight
it  receives,  the  amount  of sunlight  it reflects, and the extent  to which the
atmosphere  retains heat.   When   sunlight  strikes the earth,  it warms the
surface, which then reradiates the heat as infrared radiation.  However, water
vapor, COp, and other gases in the atmosphere absorb some of the energy rather
than allowing  it  to pass undeterred  through  the atmosphere to space.  Because
the atmosphere traps  heat  and warms  the earth  in  a manner somewhat analogous
to  the glass panels of a greenhouse,  this  phenomenon  is commonly known as the
"greenhouse effect."  Without  the  greenhouse effect  of the gases  that  occur
naturally in the atmosphere, the earth would be approximately 33°C  colder  than
it  is currently (Hansen et al.  1984).

     In  recent decades,  the  concentrations  of  greenhouse  gases have  been
increasing.  Since the beginning  of  the industrial  revolution,  the combustion
of  fossil  fuels,  deforestation,  and  a  few other  activities  have  released
enough COp  to  raise atmospheric  concentrations by  20 percent;  concentrations
have risen  8 percent  since 1958 (Keeling,  Bacastow,  and Whorf  1982).   More
recently, Ramanathan  et  al. (1985)  examined the  greenhouse gases  other  than
C02 (such as methane, CFCs, and nitrous oxide), and concluded that these other
gases  are  likely  to  double the  • warming  caused  by  C02  alone.    Using  these
results, the Villach Conference estimated  that an "effective doubling" of C02
is likely by 2030.

     Hansen et  al.  (Volume 1) and  Manabe & Wetherald  (Volume  1)  present the
results  that their climate models predict for an  effective  doubling of COp.
Both models  consider a  number of "climatic feedbacks" that could  alter the
warming that would directly result from COp  and other gases released by human
activities.  Warmer temperatures would allow  the  atmosphere to  retain  more
water vapor, which  is  also a greenhouse gas,  thereby resulting in additional
warming.   Ice   and  snow  cover  would  retreat,  causing sunlight  that is  now
reflected by these bright  surfaces to  be absorbed instead,  causing additional
warming.  Finally, a change in cloud cover might  result,  which could increase
or  decrease  the projected  warming.    Although  the  two models differ in  many
ways, both conclude that  an effective  doubling of greenhouse gases would warm
the earth's surface between two and four degrees (C).

     Hansen et  al. project  the doubling to occur  between  2020 and 2060.   They
also  provide  estimates  of  the  implications  of  temperature  changes  for
Washington,  D.C.,  and seven  other  U.S.  cities  for  the middle  of  the  next
century.   For  example, Washington would have  12 and 85  days per  year  above
38°C (100°F) and 32°C  (90°F),  respectively,  compared with 1 and 35 days above
those levels today.  While evenings in which the thermometer fails to go below
27°C (80°F)  occur  less than once per year  today  in that  city,  they project
that such  evenings would occur  19  times per  year.   (See  Plates  2  and  3 for
worldwide maps  of historical and  projected temperature changes.)

Water Resources

     Manabe  and Wetherald (Volume  1)  focus  on  the  potential  changes  in
precipitation patterns  that might result  from the greenhouse warming.   They
project substantial  increases  in  summer dryness at  the middle  latitudes that
currently support most of  the  world's  agriculture.   Their model also projects
increased rainfall for late winter.

     Beran  (Volume  1)  reviews the  literature  on  the hydrological  and  water
resource impacts  of  climate  change.   He expresses  some  surprise  that  only
twenty-one papers  could  be found  that address  future water resource impacts.
One  of  the problems,  he notes,   is  that  there   is  a  better  scientific
understanding of  how global average  temperatures and rainfall might change,
than for the changes  that specific regions  may experience.  Nevertheless, he
   Studies on the greenhouse  effect  generally  discuss  the impacts of a carbon
   dioxide doubling.   By  "effective doubling" we  refer  to any combination of
   increases  in  concentrations  of  the  various  gases  that  causes  a  warming
   equal to the warming of a doubling of carbon dioxide alone.

demonstrates  that  useful  information  can  be  extracted  by  studying  the
implications of particular scenarios.

     Nicholson (Volume  3)  shows how historical changes  in  water availability
have caused  problems  for society  in the past.   The best  lesson  of climatic
history,  she writes,  "is  that agricultural  and  economic  systems must  be
flexible enough to adapt to changing conditions and,  in the face of potential
water  scarcity,   systems  must  be  designed  that  require  minimum  use  of
resources."  Wilhite (Volume 3) examines drought policies in Australia and the
United  States,  concluding  that  the  lack  of  national  drought plans  could
substantially impair the ability of these two nations to successfully adapt to
hydrologic changes resulting from the greenhouse warming.

     Cohen  (Volume 3)  examines  the  potential  implications  of  the  global
warming  for  water levels  in  the  Great Lakes that  separate Canada  from the
United States.  Using  results  from the models of both Hansen et al. and Manabe
& Wetherald,  he  concludes that lake levels could drop  10  to 30 centimeters.
This drop would significantly reduce the  capacity of ocean-going vessels that
enter the Great Lakes.   On the other  hand,  such  a drop  might  be  viewed as a
benefit by the owners  of critically eroding property whose homes are currently
threatened by historically high lake levels.  Street-Perrott et al. (Volume 3)
discuss  the  historic  impacts  of changes in climate  on  the  levels  of lakes in
North America, South America,  Australia, and Africa.

     Gleick  (Volume  3) uses  scenarios from  the  Hansen  et al. and  Manabe &
Wetherald models  (as  well  as a  third developed  by  the National  Center for
Atmospheric  Research)  to  drive  a water-balance model of the Sacramento Basin
in California.  He finds  that  reductions  in runoff could occur even in months
where precipitation increases  substantially, because of the increased rates of
evaporation  that  take  place at higher  temperatures.  He also points out that
the models  predict that changes  in  monthly runoff patterns  will  be far more
dramatic  than  changes  in annual averages.   For  seven  of ten scenarios, soil
moisture would be  reduced every month of  the  year; for  the other three cases,
slight  increases  in moisture are  projected  for winter months.  Mather (Volume
3)  conducts  a  detailed  analysis  for southern  Texas   and  northern  Mexico;
examines  in  less  detail twelve regions around the world; and projects shifts
in global vegetation zones.

Agriculture and Forestry

     The  greenhouse  warming  could  affect  agriculture   by  altering  water
availability,  length  of  growing  season,  and the  number  of  extremely  hot
days.    Increased COp  concentrations  could  also  have  two  direct  impacts
unrelated to climate change:  At least  the laboratory, plants grow faster (the
COp fertilization effect) and retain moisture more efficiently.  The extent to
which  these  beneficial  effects  offset the  impacts of climate  change will
depend  on the extent  to which  global warming is  caused  by COp as opposed to
other greenhouse gases, which do not have these positive  impacts.

     Parry (Volume 1)  provides an overview of the  potential  impacts of climate
change  on agriculture  and  forestry.   He points  out that commercial farmers
plan according to the  average  year,  while family and subsistence farmers must
ensure  that  even  in  the  worst years  they can  make  ends meet.    Thus,  the
commercial farmer would be concerned about the impact of future  climate  change

on average conditions and average yields, while farmers at the margin would be
most  concerned  with  changes  in  the  probability  of  (for  example)  a  severe
drought that causes  complete crop failure.   Parry  notes  that the probability
of  two or  more  anomalous  years  in  a row  could  create  disproportionately
greater problems  for agriculture. -For example,  a  persistent  drought  in the
U.S. Great Plains from  1932  to 1937  contributed to about two hundred thousand
farm bankruptcies.

     Parry discusses  a  number Of historical  changes in  climate.   The  Little
Ice  Age  in western  Europe  (1500-1800 A.D.)  resulted  in the  abandonment of
about half the  farms in Norway, an  end to cultivation  of cereals in Iceland,
and some farmland in Scotland being permanently covered with snow.  Concerning
the late medieval cooling (1250-1500) he writes:  "The failure to adapt to the
changing circumstances  is believed to  explain much of the Norse decline.  The
Norse continued to emphasize stock-raising in the  face of reduced capacity of
the already limited  pastures.  The  option  of exploiting  the rich seas  around
them, as the Inuit (Eskimos)  successfully  did,  was not taken up ... This is
an  extreme  example  of how  governments can  fail  to identify  and  implement
appropriate policies of response."   It also suggests that effective responses
can reduce damages from climate change.

     The paper  reviews a  number  of studies  that project impacts  of climate
change on agriculture.   "Warming  appears  to  be detrimental  to  cereals  in the
core  wheat-growing  areas of North  America and Europe."  If  no precipitation
changes take place,  a one-degree  warming would decrease yields  1 to 9 percent
while a two-degree  (C)  warming would  decrease  yields 3 to  17 percent.   Parry
also discusses how particular crop zones might  shift.  A doubling of COp would
substantially expand the wheat-growing area  in Canada due   to  higher  winter
temperatures and increased rainfall.    In Mexico, however, temperature stresses
would increase,  thereby reducing yields.

     A number  of studies  have been conducted  using the models of  Hansen et
al.,  Manabe  & Wetherald,  and others.   Although these projections  cannot be
viewed as reliable forecasts, they do provide consistent scenarios that can be
useful for  examining vulnerability  to climate change.    Parry  indicates that
investigations of  Canada,  Finland,   and the northern USSR using  the model by
Hansen et al. show reduced yields of spring-sown  crops such as wheat, barley,
and  oats,  due  to  increased  moisture stress  early in  the growing period.
However, switching  to  winter wheat  or  winter rye might reduce  this stress.
Parry  goes  on to  outline  numerous  measures  by which farmers  might adapt to
projected climate change.

     Waggoner (Volume 3) points out that  the global warming would not affect
plants  uniformly.    Some  are more  drought-resistant  than  others,   and some
respond to higher COg concentrations more  vigorously than others.   C^ plants,
such  as  wheat,  respond to  increased COp  more  than Ch plants  such  as  maize.
Thus, the COp fertilization  effect woula not  help the farmer growing C^ crops
accompanied by Co weeds.   Waggoner also examines  the impact of future climate
change on  average crop  yields and  pests, and the  probability of successive
drought years.   He concludes that although projections of  future changes are
useful, historical evidence  suggests that  surprises may be  in store, and that
"agricultural scientists will be  expected   to  aid  rather  than watch mankind's
adaptation to an inexorable  increase in C02 and its greenhouse effect."

     The  impact  of  future  climate  change on  yields  for  spring  wheat  in
Saskatchewan,  Canada,  is  the subject  of  the  paper by  Stewart  (Volume  3).
Using the output  from the  Hansen  et  al.  model  (Volume 1), which projects that
the  effective  doubling   of   carbon   dioxide  would  increase  average  annual
temperatures  in  that region  by  4.7°C,  he  estimates that  the  growing season
would start  two or  three  weeks  earlier  and  end  three  or four  weeks later.
Although average  precipitation during the  growing season  would  increase,  he
also finds  that  the  area would become more prone to drought.  The  impact of
climatic change would be  to reduce yields 16 to 26 percent.  Stewart estimates
that the fertilization effect of a COp doubling  would reduce the losses to 6
to  15 percent.   Cooter  (Volume  3) examines the  economic  impact  of projected
climate change  on the economy of Oklahoma, concluding  that the  Gross State
Product would decline 75 to  300 million dollars.   (The  state's gross product
in 1985 was approximately 50  billion dollars.)

     Fritts  (Volume  3)  examines  tree  rings  to  assess  how  past changes  in
climate have  affected  forests, and  concludes  that tree  rings  are useful  for
estimating past changes  in climate.   Solomon and  West (Volume  3) discuss  the
results of their  efforts to model the future impacts.  Considering the impact
of climate change caused  by doubled C02  without the fertilization effect, they
find that  "biomass  (for  boreal  forests)  declines for 50-75  years as warming
kills off large boreal forest  species, before new northern hardwoods can grow
into the plot."

     "Warming at  the transition  site causes an  almost immediate  response in
declining biomass from dieback of mature trees,  and in decline of tree mass as
large trees  die  and are  temporarily replaced  by  small  young  trees,"  they
write.    "The deciduous  forest site  .  .  .  results in permanent loss of dense
forest.    One  might  expect  the  eventual  appearance  of  subtropical  forests
similar to  those  in Florida  today,  but the real difficulty is  the moisture
balance  (which  is)   more  similar  to  those  of  treeless  Texas today,  than to
those  of  southern   Florida."    Solomon  and  West  go  on to  show  how  the
fertilization effect from increased  concentrations  of C02 could  offset part
but not all of the drop in forest productivity.

Sea Level Rise

     One of  the most widely recognized consequences  of a global warming would
be  a rise  in sea level.    As  Titus (Volume 1)  notes,  global  temperatures  and
sea level  have fluctuated over  periods of one  hundred   thousand  years,  with
temperatures  during  ice  ages  being  three  to five degrees (C) lower  and  sea
level  over  one  hundred  meters  lower  than  today.    By  contrast,   the  last
interglacial  period  (one hundred  thousand  years  ago)  was  one  or two degrees
warmer than today, and sea level was five to seven meters higher.

     The projected global warming could raise  sea level by heating and thereby
expanding  ocean  water,   melting  mountain  glaciers,  and by  causing  polar
glaciers  in  Greenland and Antarctica  to  melt  and possibly slide  into  the
oceans.    Thomas  (Volume  4)  presents  new  calculations  of  the  possible
contribution  of Antarctica and combines them with previous estimates for  the
other sources,  projecting that  a worldwide rise in sea  level of  90 to  170
centimeters  by  the  year  2100  with 110 centimeters  most  likely.   However, he
also estimates  that  if  the global warming  is  substantially delayed, the rise
in  sea  level could  be  cut in half.  Such a  delay might  result either from

actions to  curtail  emissions or  from  the thermal lag  induced  by the oceans'
ability to absorb heat.

     On  the other  hand,  Thomas  also  estimates that  if a  warming  of  four
degrees  results  from a  C02  doubling, (which  the  model  of  Hansen et  al.
projects) and concentrations continue to grow after 2050, the rise could be as
great as 2.3 meters.   He also notes that an irreversible deglaciation of the
West Antarctic  Ice  Sheet might, begin  in the next century, which would raise
sea level another six meters in the following centuries.

     Titus  (Volume  1) notes  that  these  projections  imply that sea level could
rise 30 centimeters by 2025,  in addition to local subsidence trends that have
been important  in Taipei, Taiwan;  Venice,  Italy; the  Nile Delta,  Egypt;  and
most of the Atlantic and Gulf Coasts of the United States.  The projected rise
in  sea level  would  inundate  low-lying  areas,   destroy  coastal marshes  and
swamps,  erode   shorelines,   exacerbate   coastal   flooding,  and   increase  the
salinity of rivers,  bays, and aquifers.

     Bruun  (Volume  4)  argues that  with a combination  of coastal engineering
and  sound   planning,  society can  meet  the  challenge  of a  rising  sea.    He
discusses  a number of  engineering  options,  including  dikes   (levees)  and
seawalls, and  adding sand  to  recreational  beaches  that are eroding,  with a
section on  the  battle that  the Dutch  have  fought with the  sea  for  over  one
thousand years.   Goemans  (Volume  4)  describes  the  current  approach  of  the
Dutch  for  defending  the  shoreline, and estimates that  the  cost  of raising
their dikes for a one meter rise  in  sea level  would be  10  billion guilders,
which is less  than  0.05 percent of their Gross  National  Product for  a single

     Goemans concludes that there  is no need to anticipate such a rise because
they could  keep up  with it.  However,  he is more concerned  by  the two-meter
scenario: "Almost immediately after detection, actions  would  be  required.   It
is not  at  all  certain that decision-makers  act  that fast.  .  .  .   The present
flood protection  strategy came  about  only after  the tragic  disaster  of 1953.
When nobody can  remember a specific disaster,  it  is  extremely  difficult to
obtain  consensus  on  countermeasures."   For  his  own country,  Goemans  sees  one
positive impact:    Referring to  the  unique  experience of Dutch engineering
firms in the battle with the sea,  he suggests that "a rising sea may provide a
new global market for this expertise."  But  he predicts that "the question of
compensation payments may come up," for the poorer countries who did not cause
climate change  but must face its consequences.

     Broadus et al.   (Volume 4) examine  two such countries  in  detail:  Egypt
and Bangladesh.  The inhabited areas of both countries are river deltas, where
low-lying land  has been created by the sediment washing down major rivers.   In
the case of Egypt,  the damming of  the  Nile  has  interrupted  the sediment,  and
as the  delta sinks, land is lost to the Mediterranean Sea.   Broadus et  al.
estimate that  a  50-centimeter  rise in  global  sea  level, when  combined  with
subsidence  and  the  loss of  sediment, would  result -in  the loss of  0.3 to  0.4
percent  of the  nation's land  area;  a 200-centimeter  rise  would flood  0.7
percent.  However, because Egypt's population is concentrated in the low-lying
areas,   16  and  21  percent of the  nation's population currently  reside in  the
areas that would be lost in the two scenarios.

     The situation would be even more severe in Bangladesh.  As Plate 4 shows,
this nation, which is already overcrowded,  would  lose  12 to 28 percent of its
total  area,  which  currently  houses  9  to  27  percent  of  its  population.
Moreover, floods could penetrate farther  inland,  which  could  leave the nation
more vulnerable  to the type of  tropical  storm that killed 300,000  people in
the early 1970s, especially if the frequency of tropical storms doubled due to
warmer water temperatures, which deSylva  (Volume  4)  projects.   Broadus et al.
conclude  that  the vulnerability  of Bangladesh to  a rise in sea level  will
depend in large measure on whether future water projects disrupt land-creating
sediment washing down the Ganges.

     Bird (Volume  4) examines  the  implications  of  sea level rise  for  other
African and Asian  nations,  as  well as Australia.  While holding  back the sea
may be viable in Australia, he  shows  areas  in New Guinea where people live in
small cottages  on  the  water's  edge on a  barrier  island that  almost  certainly
would be  unable to  justify  construction of  a dike.   He also points  to the
Philippines, where many people have literally  "taken to the water,"  living in
small boats and maintaining fishing nets  in their own  plots of bay instead of
land.  Current wetlands, he suggests, may convert to these shallow bays,  with
people converting to a more water-based economy.

     Leatherman  (Volume  4)  examines  the  implications   of  sea level  rise for
South  America.    He  notes that  such popular resorts  as Copacabana  Beach,
Brazil;  Punta  del Este,  Uruguay;  and Mar  del Plata,   Argentina,  are already
suffering  serious  erosion.     He   concludes   that   because  of  the  economic
importance  of  resorts,  governments  will  allocate  the  necessary   funds  to
maintain  their  viability.   However,  he  predicts  that  "coastal wetlands  will
receive benign neglect" and be lost.

     Park et al. (Volume 4) focus on the expected drowning of coastal wetlands
in the United  States.   Using a computer  model of over  50 sites,  they project
that 40-75  percent of existing  U.S.  coastal wetlands   could  be lost by 2100.
Although  these  losses  could  be reduced to  20-55  percent if new wetlands form
inland  as  sea  level  rises,   the  necessary  wetland  creation would  require
existing  developed  areas  to  be  vacated  as   sea  level  rises,   even  though
property  owners would  frequently  prefer to  construct bulkheads to protect
their property.  Because coastal wetlands are important for many  commerically
important seafood  species,  as well as  birds  and  furbearing  animals, Park et
al. conclude that  even  a  one-meter  rise  in  sea level would have major impacts
on the coastal environment.

     DeSylva  (Volume 4) also  examines  the  environmental  implications  of sea
level  rise, noting  that  in  addition  to  wetlands  being  flooded,  estuarine
salinity would  increase.  Because 66 to 90 percent of U.S. fisheries depend on
estuaries,  he writes that  these  impacts  could be  important.  He also suggests
that coral  reefs  could  become  vulnerable  because  of sea level rise,  increased
temperatures, and  the decrease in the pH  (increased  acidity) of the ocean.

     Kuo  (Volume  4)  examines  the implications  of sea  level rise  for flooding
in  Taipei,   Taiwan,  and  coastal  drainage  in  general.   Although  Taipei is
upstream  from the  sea, Kuo concludes that projected  sea  level rise would  cause
serious  problems,  especially  because Taiwan  is also sinking.   He recommends
that engineers  around the world take "future sea  level  rise into consideration
... to  avoid  designing a system that may  become prematurely obsolete."


     Gibbs (Volume 4)  estimates that sea level  rise  could result in economic
damages in Charleston,  South Carolina,  equal to as much  as 25 percent of the
annual product of the community.  Anticipatory measures, however, could reduce
these impacts by half.  Gibbs finds that in some areas actions should be taken
today, in spite of the current uncertainty regarding future rates of sea level
rise, while for other  areas  it  would be more prudent to wait until uncertain-
ties are resolved.

     Ken Smith, a realtor  from  coastal  New  Jersey,  reacts to the other papers
presented in  Volume  4.   He argues that the  issue of  sea  level rise should be
taken seriously today,  but laments the fact  that many  of his fellow realtors
make comments such as  "What  do  you care?   You won't be around to see it!" and
the scientific community  is  "a bunch of eggheads who don't want us (to build
on  the  coast) anyway." Smith  suggests  that part of  the  resistance to taking
the  issue  seriously   is  that there are a  number of  "naturalists"  who oppose
building near the shore,  and  "most  of  the discussion seems  to  come from the
'naturalist'   camp."   Nevertheless,  Smith argues that  "the solutions—if there
are  any—should  be  contemplated  now as  part  of  a concerted  global effort.
This is a beautiful world, and we are its stewards."

Human Health and Ecological Impacts

     Climate  and  weather  have  important  impacts on  human health.   A global
warming would  increase  the stresses due to  heat, decrease those due to cold,
and possibly  enable  some  disease  that require warm  year-round temperatures to
survive  at  higher  latitudes.    Kalkstein  et  al.   (Volume  3)  present  a
preliminary statistical  assessment of the relationship of mortality rates to
fluctuations  in  temperature  in New York City.   They  find  that  a  two to four
degree  (C) warming  would  substantially increase mortality rates  in New York
City, if nothing  else  changed.  However,  they caution that if New Yorkers are
able to acclimatize  to temperatures as  well as people who currently live in
U.S. cities  to the  south, fewer  deaths would occur.   Kalkstein  et al. write
that  knowledgeable  observers  disagree  about whether  and  how  rapidly people
adapt to higher temperatures; some people undoubtedly adjust more readily than

     Although people may be able to adapt to changes in climate, other species
on  the planet would  also be affected and may not be  as able to control their
habitats.   Peters and  Darling  (Volume 3) examine the  possibility that changes
in  climate would  place multiple  stresses  on some species which would become
extinct, resulting in a significant decline in biodiversity.   (Mass extinc-
tions appear to have accompanied rapid changes in temperatures in the past.)

     Throughout the  world  reserves have been set aside  where targeted species
can remain relatively  free of  human  intrusion.   Peters  and Darling ask:  Will
these reserves continue to serve the same function if the climate changes?  In
some cases,  it will depend on whether the reserve's boundaries encompass areas
to  which  plants  and  animals  could migrate.    Some  species  may  be  able  to
migrate "up the mountain"  to find cooler temperatures;  coastal wetlands could
migrate  inland.    A northerly  migration  of  terrestrial  species  would  be
possible in the undeveloped  arctic regions of  Alaska,  Canada,  and the Soviet
Union; but human  development would block migration of  larger animals in many


     Papers by UNEP  Deputy Director Genady Golubev and  EPA  Administrator Lee
Thomas (both in Volume  1)  provide  official  views  on  the nature of the effects
from projected changes in the atmosphere and the role of their institutions in
addressing  those  changes.    Golubev notes  that  while "the global  issues are
complex, uncertainty  exceeds understanding, and patience  is  prudence," there
is an other side  to the story:   "Our  legacy  to the  future  is  an environment
less benign than that inherited from our forbearers.   The risks are sufficient
to generate  a collective  concern  that  forebodes too  much  to  wait out the
quantifications of scientific  research.   Advocating  patience  is an invitation
to be a spectator to our own destruction."

     Golubev also points out that UNEP has worked for  the achievement  of the
Vienna Convention for the Protection of the Ozone Layer, in which many nations
have agreed to act in concert  to  address  an environmental  issue whose impacts
have not yet  been  detected.   Yet he notes  that  the  agreement is for coopera-
tion in research and does not yet bind nations to observe limits in production
and emissions of gases that could deplete stratospheric ozone.

     Thomas points  out  that both  the  potential depletion  of ozone and the
global  warming  from the  greenhouse  effect  are examples  of  environmental
problems that involve the "global commons."  Because all nations contribute to
the problem and  experience the consequences, only an international agreement
is likely  to  be effective.   He  urges  scientists around the  world  to discuss
this issue with their colleagues and key officials.

     Richard Benedick,  Deputy Assistant  Secretary  in the U.S.  Department of
State (Volume  1), describes  the  emerging  international  process addressing the
ozone issue.   Although  the process for addressing climate change has not yet
proceeded  as  far,  he writes,  "from my perspective  as  a  career  diplomat, it
appears  that  the greenhouse  effect has all  the  markings  of becoming  a high
visibility foreign policy  issue.  .  .  .  How we address this issue internation-
ally depends  to  a  great extent on  our success or failure  in  dealing with the
ozone depletion issue."

     J.P.  Bruce  (Volume  1)  of  Environment  Canada presents   the   issue  of
atmospheric change in the context of air pollution in general.  He writes that
ozone modification and  climate change  are "urgent issues," especially because
important  long-term  decisions are  being  made today  whose outcomes  could be
strongly affected by changes in climate and the ozone layer.  Bruce recommends
that emissions of  CFCs  be  reduced, and concludes that  "a  new approach, a new
ethic towards  discharging  wastes and  chemical  materials  into  the  air  we all
breathe must soon be adopted on a international scale."

     Two U.S.  Senators  also  provide their  reactions.   John  Chafee from  Rhode
Island  (Volume 1)  describes  hearings  that his Subcommittee  on Environmental
Pollution  held June  10-11,  1986.   "Why  are policy makers  demanding  action
before  the scientists  have resolved all  of the  questions  and uncertainties?"
he  asks.   "We are  doing  so  because  there  is a  very  real  possibility  that
society—through ignorance or  indifference, or both—is irreversibly altering
the ability of our  atmosphere  to perform basic life support functions for the
planet."   Albert Gore, Jr.  from Tennessee, who has chaired three congressional
hearings on  the greenhouse  effect,  explains  why he  has introduced a bill in


the U.S. Senate  to  establish an International Year  of  the Greenhouse Effect.
"The  legislations  would  coordinate  and  promote  domestic and  international
research efforts on  both the scientific  and policy aspects of  this problem,
identify strategies to reduce the  increase  of  carbon dioxide and trace gases,
investigate  ways  to  minimize   the   impact of  the  greenhouse  effect,  and
establish long-term  research plans."  Senator Gore closes  by quoting Sherwood
Rowland (discussed below): "What's the use  of  having developed a science well
enough to make  predictions,  if  in  the  end  all  we're  willing  to do is stand
around and  wait for  them to come  true?"   Both Senators call  for  immediate
action to reduce global use of CFCs.

     John S. Hoffman  (Volume 1)  emphasizes  the  inertia  of the atmosphere and
oceans.   Because  there  are time  lags  between changes  in   emission  rates,
atmospheric  concentrations,  and   changes  in   ozone  and   global  warming
temperatures, the types of management strategies must be different  from those
that  are  appropriate  for controlling,  for example,  particulate  pollution,
where  the problem  goes away as  soon  as  emissions are halted.    CFC  emissions
would  have  to  be  cut  80 percent simply   to   keep  concentrations  from
increasing.    Although constant  concentrations would prevent  ozone  depletion
from worsening,  Hoffman points out that even if  we hold the concentrations of
greenhouse  gases  constant once  the  earth  has  warmed  one degree,  the  planet
would warm another degree as the oceans come into  equilibrium.  Thus it might
be  impossible  to  prevent a substantial  warming  if  we  wait  until a  small
warming has taken place.^

     The final section of this  volume presents  the  papers from  the  final day
of the  conference.   Peter Usher of UNEP  recounts the evolution  of  the ozone
issue.   Following Rowland  and  Molina's  hypothesis that  chlorofluorocarbons
could cause a depletion of stratospheric ozone  in 1974,  UNEP held a conference
in 1977 that led  to a world plan  of  action to assess the issue  and quantify
risks.  Since that time,  UNEP has  held  numerous  coordinating meetings leading
up the the  Vienna Convention.  However, Usher  suggests  that motivating inter-
national effort  on the greenhouse effect will be more difficult:   "Prohibition
of  nonessential emissions   of  relatively  small  amounts  (to control  ozone
depletion)  is one  thing,  limiting  emissions of  carbon  dioxide from coal- and
oil-burning is quite another."

     Dudek  and  Oppenheimer   of the  Environmental Defense  Fund  (U.S.)  analyze
some  of the  costs  and   benefits  of  controlling   emissions  of  CFCs.    They
estimate that by holding emissions constant, 1.65 million cases of nonmelanoma
skin cancers could be prevented  worldwide, and  that the cost of these controls
would be 196 to  455 million dollars,  depending  on the availability of alterna-
tive chemicals.

     Two  former  high-ranking  environmental officials   in  the United  States
argue  that  we should be  doing more  to  address these problems.   John Topping
recommends that  CFCs in aerosol  spray cans,  egg cartons, fast-food containers,
and other  nonessential uses be phased  out,  and  that   people recognize that
   Titus (Volume  1)  and Thomas  (Volume  4)  also explore  inertia,  noting that
   even  if  temperatures remained  constant  after warming somewhat,  sea level
   would rise at an accelerated rate as the oceans, mountain glaciers, and ice
   sheets came into equilibrium with the new temperature.

along with energy  conservation,  nuclear power is the most  likely alternative
to  fossil  fuels over  the  next"  generation  or two.   He also  recommends that
society take  steps to  minimize  the impacts  of  climate change and  sea level
rise, for  example, by  requiring environmental  impact  statements  to consider
the likely impacts.

     Gus  Speth,  president  of  the  World  Resources  Institute,  recommends
international  efforts  to  stop  tropical deforestation;  a  production  cap  for
chlorofluorocarbons;  increased energy  conservation;  advanced  technologies  for
producing  electricity   from  natural  gas;   and  tighter  regulations  to  limit
carbon monoxide  from automobiles,  which would  indirectly limit  increases  in
atmospheric  methane.    He  agrees  with  Topping  that  environmental  impact
statements for  projects that  could contribute  to  or be affected by  climate
change or ozone modification should consider these impacts.

     Doniger  and  Wirth, from  the  Natural  Resources Defense  Council  (U.S.),
argue  that the  current uncertainties  are  no  longer  a reason  to wait  for
additional information: "With the stakes so high,  uncertainty  is  an  even more
powerful argument for taking early action."  These authors conclude that sharp
reductions in CFCs are  necessary,  pointing  out  that even with  a production
cap,  atmospheric  concentrations  of  these  gases  will  continue  to  grow.
Therefore, Doniger and  Wirth propose an 80  percent cut  in production over the
next  five years  for  CFCs  11  and 12,  the  halons,  and  perhaps some  other
compounds, with a complete phaseout in the  next  decade.

     Richard  Barnett of  the  Alliance  for  a Responsible  CFC Policy  (which
represents CFC  using industries) agrees that we should not delay all action
until  the effects of  ozone depletion  and  climate  change are  felt;   but  he
"would hardly characterize the activities over  the last twelve years as 'wait
and see'  . .  .   The  science, as we currently understand it, however, tells us
that  there  is additional  time  in  which to  solidify  international consensus.
This must  be  done  through discussion and negotiation,  not  through unilateral

     Barnett  adds  that industry  should  "take  precautionary measures  while
research  and  negotiations continue  at the international  level.   We  will
continue   to   examine   and  adopt   such prudent  precautionary  measures  as
recapturing,  recycling, and recovery  techniques to  control CFC emissions;
transition  to  existing alternative  CFCs  that  are considered   to be  more
environmentally  acceptable;  practices to   replace  existing  systems  at  the
expiration of their  useful  lives  to  equipment  using other  CFC formulations;
practices  in  the  field to prevent  emissions  where possible;  encouragement of
CFC users to  look  for processes  or  substances that are as efficient,  safe,  and
productive—or better—than what is presently available."

     Barnett  concludes that  "these environmental  concerns are  serious,  but
their  successful  resolution will  require  greater global cooperation  in con-
ducting the necessary  research and monitoring,  and in developing coordinated,
effective, and equitable policy  decisions for all nations."

     We hope  that  this  paper has provided the reader with a "road map" through
the papers of this four-volume report on the  potential  effects of changing the
atmosphere.   But  we have  barely scratched  the  surface of each,  just as the
existing   research has  barely  scratched   the  surface  in  discovering  and


demonstrating the possible risks of ozone modification  and  climate change.   A
continual evolution of  our  understanding will be necessary  for  our knowledge
to stay ahead of the global experiment that  society is conducting.


Hansen, J.E., A. Lacis,  D. Rind, and G. Russell.  1984.  Climate sensitivity to
   increasing greenhouse  gases.   In Greenhouse effect and  sea level  rise:   A
   challenge for this generation,  eds. M.C. Earth and  J.G.  Titus.   New York:
   Van Nostrand Reinhold.

Keeling, C.D.,  R.B. Bacastow,  and T.P. Whorf.   1982.   Measurements  of  the
   concentration of carbon  dioxide at  Mauna Loa,  Hawaii.   Carbon  Dioxide
   Review 1982. 377-382, ed. by W.  Clark. New York:   Oxford University Press,
   Unpublished data from NOAA after 1981.

NAS.  1983.   Changing Climate.  Washington,  D.C.:  National  Academy Press.

Nordhaus, W.D.,  and G.W. Yohe.   1983.  Future carbon  dioxide emissions from
   fossil fuels.   In Changing  Climate.   Washington,  D.C.:   National Academy

Villach.  1985.  International assessment of the  role  of carbon dioxide and of
   other  greenhouse  gases   in  climate  variations  and  associated  impacts.
   Conference Statement.  Geneva:  United Nations Environment Program.

Global Environmental Change: The UNEP Perspective
Genady N. Golubev
Office of Environment Programs
United Nations Environment Programme
Nairobi, Kenya
     It is, perhaps,  paradoxical  that  the global  environmental  threat, with
the potential  to influence  for good  or ill the lives  of every  individual on
the planet, should generate such ambivalent attitudes among people with so
much at stake.

     It has been constantly  said,  and  will be  said again:  the  global  issues
are complex; uncertainty exceeds understanding;  and  patience is prudence.

     That  coin  has another  side.   The issue,  stripped to its essentials, is
simple and  unequivocal in its message, our legacy to the future is an environ-
ment less  benign  that  that  inherited  from our forebearers.   The  risks are
sufficient  to  generate  a collective  concern that  forebodes  too  much to wait
out the quantifications of scientific  research.   Advocating patience  is an
invitation  to be a spectator  to our own destruction.

     One approach,  adopted by  the United  Nations Environment Programme, aids
the selection  of an appropriate course of  action.   This concerns the concept
of  "outer  limits," which  postulates  that the  biosphere has only  a limited
tolerance to the demands placed upon it.

     Two global outer limits have preoccupied  UNEP during the  last decade:
the risk to  the ozone  layer  and  the  issue of climate  change  occurring in
response to increasing concentrations of  trace  gases in the atmosphere  due to
the actions of humanity.   Of course, all  environmental sectors  have an outer
limit  of  exploitation,  but  in this  paper,  I   restrict  my  comments  to the
vulnerability  of  the  atmosphere  and  its climatic  processes  to  possible
changes.   The role of this report and the associated  conference, as seen by
UNEP,   can  be  expressed  in  terms far  beyond a mere  exchange of  scientific
information and viewpoints on  the subject  of ultraviolet radiation  damage or
of  climatic variability.   It is  looked  upon  more as  a necessary step in
formulating international policy to protect the  common  atmospheric heritage.

     In  1974,  the  possibility  of stratospheric  ozone  depletion was  first
noticed.   Simple  chemical  formulae  for  the  catalytic destruction  of  the
earth's ozone shield were unfolded to an astonished  world.   The repercussions
of a weakened ozone layer are  as  significant  as any  threat  we are ever likely
to face.   It does not  need  hyperbole  to emphasize  its  significance  for man-
kind.  Human health,  fish and animals,  plants  and food crops are at risk.  The
principal suspect is the family of manmade  chemicals,  chlorofluorocarbons, of
benign  property  (at  least in  the lower atmosphere),  and of  vast commercial
usefulness.   The enormity of the issue  and  the seemingly contradictory results
of the  scientific examination  of the  ozone depletion  hypothesis  have to some
extent  been  counterproductive  to  attempts  to deal  with the problem.   Ozone
depletion estimates have  ranged from 5% in the mid-1970s to more  than  20% a
few years later.   Indeed, some  scenarios  have produced  estimates indicating
increases in the  total ozone  column.   Such divergences  have  bred skepticism
and  fear  that  subjective  and selective interpretation are being introduced
into  scientific  analysis  to  accommodate   rapidly  polarizing  viewpoints  of
different nations (and  factions) on the problem of the ozone layer.  You have
read the headlines;  perhaps you have  even written them:  "The ozone problem is
no problem"  or  "The  end of the world  is nigh."

     I am happy  to  report that UNEP,  at least, kept  its  head  in  the  midst of
controversy  and,  supported by  its Governing  Council  and most of the inter-
national  community,  established  a program of  activities  that  has,  in  the
shortest  possible  time, led  to a unique  milestone   in  the annals of inter-
national legislation:   the  Vienna Convention  for  the  Protection  of the  Ozone
Layer.  This covenant  represents an agreement  among  nations to act in concert
to address  an  environmental issue still  awaiting scientific  confirmation of
its actuality.   Impressive as this agreement is, it remains  a tenuous response
to the potential threat posed  by  a changing ozone layer.  In present form, it
represents an agreement to  cooperate  in research  and  in monitoring the  ozone
layer  and  other gaseous  constituents  of  the  atmosphere and  in  information
exchange regarding production,  emissions, usage,  trade,  and national  legisla-
tion on substances suspected  of modifying  the  ozone  layer.   It does  not bind
states to observe limits in production  and  emission of these substances,  which
will require a specific protocol to the Convention and for nations to be party
to both the protocol  and the Convention—a   thorny path  still  to  be  nego-

     Let  us look  briefly  at  the program UNEP  is  operating regarding  the
atmosphere,   which  will  explain  the reason  why  it  cosponsored  the  "Effects
Conference"   and  this  report.   In 1977,  in response to calls  for a  rational
approach to  the  ozone  layer  issue, UNEP convened  an International Conference
on the Ozone Layer  in Washington, D.C.   A main  purpose was  to  elaborate a
world  plan of action on the  ozone layer.   Under the plan, coordination of the
research effort was to  be the province  of UNEP and, to assist it in this task,
UNEP established a  Coordinating Committee  on  the  Ozone  Layer  (CCOL)  composed
of international, governmental,  and nongovernmental  organizations with active
ozone  layer  research  programs.   This  committee  makes and  publishes through
UNEP,  assessments of ozone layer modification and  its  impact based on national
research  results  reported to  it.   To  date,  the  CCOL has met  eight  times to
refine  its  estimate of projected  ozone layer change.   As I noted above, the

CCOL's predictions  have shown wide  variations from year  to year.   Over the
last five years, predictions have moved  in  a  much narrower range and even the
most skeptical  of  governments  accept the  validity of UNEP's  approach  to the

     I have referred to the UNEP role as coordination  and catalysis.  I might
add that the UNEP approach to an  environmental problem is to assess the nature
and scale of the issue  and, based on  the assessment,  to provide advice on the
range of managerial options available to address  the  problem.   In the case of
ozone, the assessment is a continuing one and one that has yet to resolve the
many uncertainties that still exist.   Let us not minimize these uncertainties:
A  statistically significant depletion  of  ozone  has  been  observed at  40 km
consistent with the theory  of ozone  modification.   Yet we  may still  be  a
generation away from finally confirming or denying ozone modification as fact.

     The  assessment  of  the  physical  state  of  the   atmosphere  and,  in
particular, of  the  ozone  layer,  has  taken  a  giant step  forward over the past
year and  a half.   Nine national  and international  organizations cosponsored
and  cooperated   in  a giant—there  is no  other  appropriate  word—assessment
orchestrated  by  the  U.S.   National  Aeronautics  and  Space  Administration
(NASA).  The CCOL drew  on this assessment  to  make its own estimation of ozone
layer change.  The committee did  not,  however, address the issue of effects of
ozone layer modification.   UNEP  will next  arrange for the summarizing of the
conference findings  by  experts for consideration by government  experts  at  a
special  session of the  CCOL  later  in  1986  with a  view  to  preparing  an
assessment of  impacts  of  ozone  layer  depletion  equivalent  to  that  already
produced on the physical ozone layer.

     But  assessment,  in  the  UNEP  context,  is  to be  followed  by  rational
environmental management.   An  ad  hoc  working group  of legal  and technical
experts was  convened  by UNEP in 1982  and  met seven times  before agreeing on
the framework for an international convention for the  protection of the ozone
layer.  This group  is  now to be reconvened to  determine  the  need for and the
form of a protocol of chlorofluorocarbons to the Vienna Convention.  Technical
information  on   chlorofluorocarbons  will  become  available  from a  two-part
workshop on aspects of production,  use and release of those chemicals, and the
consequences and policies of their control.   The  first workshop took place in
May  1986  in  Rome.   The second part  will be  in Washington,  B.C.  in September

     If there is need of a protocol on CFCs, the Vienna Group, for that is how
it is known,  will determine it; hopefully,  the agreed protocol will be adopted
by diplomatic conference in 1987.

     It  is a  long  road,  full  of pitfalls,  demanding  goodwill  and  indeed
sacrifice  in  the form  of  agreements to limit national  industrial activity,
trade, and a loss  of  convenience that  CFCs currently  provide.   Forget the
frivolous  cosmetic  aerosol  use  of  the  chemicals—refrigeration,  rigid and
flexible  foams   for - insulation,  packaging  and upholstery,  solvents  for the
electronic  industry,  and  pharmaceutical  products  also  depend  on the CFC
industry.    More than  thirty-five  million  tons  of  these  non-toxic,  non-
reactive, convenient gases have  already been  emitted  to  the  atmosphere to

remain there for tens or even hundreds of years.   Each molecule has the poten-
tial to dissociate as many as four hundred of the essential ozone molecules in
the fragile shield protecting us  from  the  ravages  of unrestrained ultraviolet

     The  Convention  was  a  necessary and  timely  precaution  and UNEP  will
encourage any  measure—such as  the elaboration  of suitable  protocols—that
will ensure its effectiveness in protecting the environment.

     I believe  that  the  goodwill  of most nations  supports  UNEP in its objec-
tive; but it needs more  than goodwill.   Adherence  to the spirit as well as to
the letter of the Convention and its protocols will be necessary prerequisites
for  its  effectiveness.    My  concern here  arises from a  request  by  UNEP for
factual data on current  production capacity, production,  emissions,  use, and
regulatory measures concerning chlorofluorocarbons  to all member states of the
United Nations,  all  regional economic  groups,  and  relevant  non-governmental
organizations.    The  purpose of  the request  was to make  available  necessary
background information to the  Workshop on Chlorofluorocarbons.   The response
was  extremely  poor.   Only  twenty relevant  replies  were  received out  of the
more than  170  requests  sent.   It hardly provided a  representative  sample of
the  nations  involved  in  the  chlorofluorocarbon  industry—hardly enough  to
promote useful  analyses  of the use,  trade,  and emission  patterns.   Probably
all countries use  CFCs  even if they do not  manufacture them.   After all, the
substances are  widely used  in  refrigeration and  air conditioning.    Can the
lack of replies be attributed  to ignorance of the extent  of  trade and use in
CFC containing  products,  to the  designation  of  low priority  to  the  issue by
states with more pressing environmental problems begging attention?  UNEP will
do  all  it can  to  raise  the consciousness of  individuals  and  nations  to the
potential gravity of the situation.  But meanwhile, the ozone  problem is still
"up  in   the   air."      The   improvement  of  scientific   knowledge   is  the
responsibility  of  you,  your fellow  scientists,  and  others who  care  for our

     I have  focused  on  the  ozone layer problem mainly  because it  is  at the
center of an  intense effort by  UNEP at the  present time; but  it is not the
only global issue receiving the attention of the international community.  Nor
even is the ozone issue a problem addressed independently of other atmospheric
environment  problems.    UNEP is  active in  the fields  of long-range  trans-
boundary  air  pollution,   acid  deposition,  and  the greenhouse  effect climate
change issue.   Carbon dioxide  is generally  regarded as the  agent of climate
change; but ozone, projected to increase in the troposphere, and ozone modify-
ing  gases,  such as methane  and  the ubiquitous  chlorofluorocarbons,  are also
greenhouse gases.   Despite their relatively low concentrations compared with
carbon dioxide, some are vastly more efficient absorbers of infrared radiation
than COp.  UNEP operates a major program on the greenhouse gas  issue, parallel
and  closely  interwoven with the ozone layer program.   Indeed, action to pro-
tect  the  ozone  layer,  for  example,  by limiting  emissions  of chlorofluoro-
carbons, will surely have a profound impact upon the rate of climatic change.

     If anything,  the climate change  issue  is  more resistant  to solution than
ozone  modification.   Although  there is already  scientific  consensus  on the
probable  future consequences of doubling carbon dioxide or its equivalent in
other  greenhouse gases—at  least in  the  global  sense—the  understanding of
regional  climatic change is  closer  to speculation  rather   than scientific


prediction.   Nevertheless,  preparations for a  future climate change  need to
begin now and UNEP is developing a program that will address the socioeconomic
consequences  of  climate  change and of  the range of  policy alternatives that
governments might  consider  in response.   Whether  there will be a  need for a
future legislative process  similar to  that  developed under the  Risks to the
Ozone Layer Programme is not  yet  known.   One major difference between the two
issues is  that  the consequences  of ozone depletion  are  universally harmful;
climatic change, on  the  other hand,  will  provide benefits  to  some  as well as
disadvantages to  others.    Some  countries,  for  instance,  those  with densely
populated  low-lying  coastal  areas  or with  agricultural  areas postulated to
receive less  rainfall, may  be  big losers.   Increased  rainfall  or longer grow-
ing  seasons   in  the  present marginal areas  for crop production  can  only be
greeted with  delight  by  the affected  countries.   Again,  while regulations of
the emissions and  use of CFCs are likely  to be  relatively  easy,  the price of
regulating COp  emissions  through energy controls  may be  too high  a price to
pay, and the future lies  in adaptation to rather than  in prevention of climate
change.  UNEP's  present  role  in this issue  is  the  provision of international
assessments.   It does so in cooperation  with its partners,  the World Meteoro-
logical Organization  (WHO)  and the International Council  of Scientific Unions
(ICSU), the three organizations that  implement  the  World  Climate  Program.  In
October  1985, UNEP,  WHO,  and  ICSU held a second  conference on the  issue at
Villach,  Austria, where a new  assessment of  the  issue was made (the first was
also made at Villach, five years earlier).

     A recommendation of  the  Conference is  the establishment of an Advisory
Group on greenhouse gases,  composed of a small group of eminent scientists, to
advise the international organizations on  a  program to  address the  greenhouse
gas/climate change issue.   This group of "seven  wise  men" will meet in Geneva
during the  first  few days of July.    However, no matter  what  program they
suggest  to  the  international  community,  it  will  again  require an  enormous
complementary scientific effort.   Most  importantly,  there  is  need  of a part-
nership of effort involving all nations to consider the consequences of green-
house gas-induced climate change and  to respond in concert to provide whatever
solution may be required.

     In  the   face  of these  global   issues,  the  world's  commitment  to  the
environment is on  trial.   I trust in the  final  analysis  that  commitment will
not be found wanting.

Global Environmental Change:  The  EPA Perspective
Lee Thomas
Administrator, Environmental Protection Agency
Washington, DC USA
     This report  focuses  on  two  global  environmental  problems  that might
result from  the  worldwide  emissions  of  certain  gases:    the  depletion  of
stratospheric  ozone  and  a  global warming  from  the  "greenhouse effect."  Both
issues are clear examples  of a "global  commons" environmental problem.   All
nations are  responsible for contributing  to recent changes in our atmosphere—
although   the   industrially  developed  nations  must  shoulder  most  of  the
responsibility.   All nations will be affected by depletion  of the ozone layer
and by global  climate changes.

     Based on  our  current understanding, we believe that a small change in the
amount of UV-B  radiation  striking  the earth and/or  a change in the earth's
mean   temperature   could    have    significant   environmental   and   health
consequences.   The fundamental scientific uncertainty  is  no  longer "are these
phenomena real?" but rather, "at what rate  are  they  likely  to  occur?"    We
need to better understand  the  full  nature of the impact  of  these changes and
the possible  options, both  in the  near  and long term,  for managing these
risks.   We  do  not  yet  have complete answers  to  these questions,  although
recent efforts have resulted in substantial progress.

     It was  only after extended discussion and considerable  thought that  the
United Nations Environment Programme (UNEP)  and the  Environmental Protection
Agency chose to proceed  with a conference on both ozone depletion and climate
change.   As  the  other  papers  in   this  volume show,  our  understanding  of
atmospheric  science  tells  us  that  ozone depletion  and  global  warming  are
inexorably interconnected.    However,  the domestic  and  international politics
surrounding  each  issue  are  separate and unique.  Combining the  two  in  one
conference had the potential to confuse and compound the political controversy
surrounding  each issue.    Separating  the  issues would fail to address their
physical  interdependence.  In the end the  choice was  clear:   we resolved this
issue by  recognizing that  this  conference is first and foremost a scientific
meeting,  not a political one, and therefore it should be organized around our
understanding  of the science.

     As a  scientific  meeting our objective  was  not to find solutions  to the
problems but rather to do the best we  can  to organize and present the current
state of  science.   In trying to achieve this goal the conference  was  not an
isolated  event  but part  of a  sequence of  environment  meetings  intended to
systematically build our understanding  of these problems.

     In  October   1985   in   Villach,  Austria,  atmospheric   scientists  and
government  officials  prepared  a major  document  summarizing  the  scientific
consensus that  is  gradually emerging regarding  the greenhouse effect.   That
conference,  sponsored  by  the  World  Meteorological  Organization  (WHO),  the
International Council of Scientific Unions (ICSU),  and UNEP, provided a sound
scientific foundation for  this  report's discussion  of the effects  of  global
climate change.

     In  February  1986  in  Nairobi,  Kenya,  UNEP  sponsored a  meeting  of the
Coordinating  Committee  on  the  Ozone  Layer   (CCOL)  to explore  the  state of
knowledge  on  the   atmospheric   science  of   ozone  depletion.    A  text  was
unanimously adopted based extensively on the  NASA Assessment Report.  The CCOL
Report and the  NASA  documents  both serve as  the foundation  for  this report's
discussions on the  effects  of ozone depletion.

     Finally, in May  1986  in Rome,  UNEP held  another  meeting  that focused on
projections   of   future   chlorofluorocarbon   production   and   emissions.
Information  from that meeting  will  help us  to  address  the problem  of ozone

     This sequence  of meetings  is  no  accident.   Each meeting  is part  of a
conscious effort  to build  a common  understanding among  the  nations  of the
world about  the nature of  these problems before  we  sit  down to  talk about
possible solutions.  At the U.S.  Environmental Protection Agency,  we recognize
that  it  is  important  to separate  the process  of  risk  assessment  from the
decision-making task that we call  "risk management."  Risk  assessment  refers
to  the  gathering  of scientific  data  in  order  to understand  what type of
problems may  confront us  if we fail to  intervene.   Risk  management refers to
the evaluation of options to avert those potential problems.   Although various
interests may have  different points of view regarding the level of action that
may  be  appropriate,  it  is  my  strong belief that  the  process  will be more
constructive  if everyone works  from a common  understanding  of the scientific
facts.   We  rely on the separation of these tasks  as  a basic approach to most
of the major environmental decisions.  Before we can sit down together to find
a common  solution  to  a problem,  we must first have a common understanding of
the problem we are  trying to solve.

     In the last decade we made major advances in assessing the risks of ozone
depletion and the greenhouse effect.   This report  presents and summarizes the
major portion of our understanding of these two problems.

     However, before  we move too  quickly  into the task of  risk  management I
must raise  a word  of caution.    If the risk management part  of the process is
to  work  effectively,  environmental  policy-makers  from each of  our countries
must be made aware of and understand the science involved in these issues.  If
not,  then  much  of  the  progress  we  have  made  in  developing  a  common
understanding of those problems  will be  lost.  It serves no useful purpose to
have decision-makers  like  myself approach  problems such as  these with either


an  inflated  or  deflated view  of the  problem, and  it  is  only  through  the
patient but persistent efforts of the experts that decision-makers like myself
can get the facts we need to make intelligent decisions.  I urge the reader to
actively share this information with your  colleagues  and  key officials.  Tell
them what  we  know,  what we don't know,  and what we are  likely to understand
soon.  Let them  know  where we have  consensus  and  where we  disagree.   Let me
also ask you to share  with us your data and scientific information.

     In the U.S.  we are still gathering the facts;  we are still in the role of
risk assessor.  Domestically,  we have committed ourselves to making a decision
on the need for  further regulations  of CFCs in the fall  of  next  year,  and we
have  committed ourselves  to  actively  supporting  and  participating  in  the
Vienna  Convention  process  for  the  protection  of  the   ozone  layer.    Our
commitment to  this  process  should not be  seen  as a precedent to  a particular
outcome.   At  the same  time  we have  not made  any decisions  either  as  to  the
need for  additional domestic  regulation or the  content  of  an international
protocol.    In  both  cases  we  will  wait  until we  have  completed  the  risk
assessment, before we  make the risk management decision.

     We recognize that several important facts must not be overlooked.  To the
extent that action is  needed,  it is essential that the international community
move forward  to  deal  with these  issues together.   Our analysis  must include
all trace gases  that  may modify the  ozone layer and not  just CFCs.   CFCs  are
extremely important chemicals  used across  a broad spectrum  of industrial  and
consumer goods.   For  some uses,  no  effective  alternative chemical  currently
exists.   Finally,  the  potential risks  we  face are  generally long-term;  if
action proves  warranted, any regulatory approach selected should be structured
in a way that  minimizes costs and disruption to producers and users.

     I am deeply  concerned about  the potential  environmental risks associated
with these issues and  the complexity  of developing a response that effectively
addresses their  unique characteristics.   Despite  these  complexities,  we  are
moving forward in a timely  manner to responsibly address  these issues.   I am
committed to making a  decision even in the  face of the many uncertainties that
are likely to  persist  for the next several  years.

Global Environmental  Change:
The International Perspective

Richard E. Benedick
Deputy Secretary of State for Environment,
Health and Natural Resources
Washington, DC USA
     As a career diplomat  who  deals with scientific issues,  and who strongly
believes in  their  growing  relevance  to  foreign  policy,  I  appreciate  the
opportunity  to  present my views.  I will briefly  sketch how the international
community can confront such complex, long-term issues as global climate change
and ozone layer modification.

     International  law  is  relatively  well equipped for  dealing with  trans-
boundary environmental problems, where  one nation  can be  identified as  "the
polluter" and the other as  "the polluted."  But in the case of global environ-
mental problems, such as ozone  depletion and climate change, the international
community is in fact entering uncharted territory.

     This emerging  set  of problems  contains a  number  of unique  elements.
First, almost  every nation  is  both perpetrator and victim:  many nations  are
producers and/or consumers of  chlorofluorocarbons  (CFCs), most  nations  burn
fossil  fuels  or otherwise  emit  carbon dioxide,  and  all nations  will be
affected by  ozone  depletion  and  climate  change.   Secondly,  the  adverse
impacts, and  the   costs of  mitigating those  impacts,   are  not distributed
equally  among   nations—nor  necessarily   in  proportion  to  each  nation's
contribution to the problem.   Finally,  for both  ozone  depletion and climate
change the  time-lag between  cause  and  effect  spans decades—which  is  longer
than the normal policy-making horizon of most governments.

     Given  these factors, designing an  efficient  and equitable control  system
would be exceedingly difficult,  even as a theoretical exercise; in practice, a
new paradigm for international  cooperation likely  will be needed.

     But what form might this paradigm take?  To answer this question, I think
it would be useful  to look  at international activities of the last decade that
relate to the  ozone  depletion  issue and—with the benefit of  hindsight—to
reflect on what has worked  and  what has not.

     International attention to  the  question  of  stratospheric ozone depletion
began shortly after Rowland and  Molina  published their now-famous theoretical
paper.   A  number of  international  meetings were held on  this  subject during
the late  1970s,  at the same time  as several nations  began  to  issue domestic
CFC  regulations.   In  1980,   the  Governing  Council  of  the United  Nations
Environment Programme decided  to convene a Working Group of Experts to discuss
appropriate international action to address this  potential problem.

     The UNEP Working Group decided  early  in  their  deliberations to develop a
convention—that is, an international treaty—on  the  ozone layer.   Four years
and several long negotiating sessions later, they completed their work on such
a treaty.  In March  1985, at a Plenipotentiary Conference in Vienna, Austria,
where I  represented  the  United States,  twenty-one  nations  signed  the  Con-
vention  for the Protection of  the Ozone Layer. This was a landmark event:  it
was the  first time  that  the  international community acted in  concert  on an
environmental issue  before  substantial damage to the environment  and health
was done—in effect, acting  together in anticipation of potential problems.

     The "Vienna Convention" creates a framework  for international cooperation
on  research,  monitoring,  and  information  exchange  concerning  the  ozone
layer.   It also  creates  general obligations to  protect the ozone  layer and
provides procedures for eventually adopting protocols to the Convention, which
could contain specific  measures to  control,  limit,  prevent, or  reduce emis-
sions of  ozone-modifying substances—should  such measures  be  deemed neces-
sary.   In  the United States,  the Senate has  held hearings on the Convention,
and we expect it  to be  ratified  soon.   After twenty nations have ratified it,
the Convention becomes international law.

     The question of a control  protocol was the  subject of considerable debate
during the UNEP  Working Group  negotiations.  In  April 1983,  Norway, Finland,
and Sweden tabled  a draft protocol for controlling all  CFC uses.   In October
1983,  the  United  States voiced  its  support for  that part  of  the "Nordic
Proposal"  dealing with  CFCs   used  as  aerosol  propellants.   Eventually the
Nordics,   along   with  Canada  and  Switzerland,   joined  us  in  supporting  an
international aerosol ban protocol.

     On   the  other side  of  the debate, the nations of  the  European Economic
Community—who  represented  the  other  major  source  of CFC  production—were
initially opposed  to  any  further controls  on CFCs.   However, they eventually
supported a.  protocol—but,  like  us,  and not  surprisingly,  they  supported one
that  mirrored what  they  already  had  in  place:   a  30 percent  reduction in
aerosol  use and a cap on future CFC production capacity.

     During the debate on these two alternative  control approaches, we pointed
out many of  the problems with their approach, and  they  in turn  noted many of
the flaws  in our approach.   The  result was total gridlock,  and there was no
possibility for agreement on a protocol text at  the Vienna Conference.

     However, because  most  of the  disagreement  near  the  end of the negotia-
tions had  centered on the economic  and policy aspects of alternative control
strategies,  participants generally agreed  that  a  series  of  international
workshops—outside  of  a formal negotiating context—would  be useful to  these
questions in detail.  All parties also agreed to approach these workshops with
an  open  mind,   unwedded to   their  previous negotiating  positions.    This

certainly represents  the U.S. position:   we  are prepared to  re-examine the
costs  and  benefits  of  various  protocol  options,  ranging  from no  further
controls to a full cap on production or use.

     The first international workshop was held in Rome during the last week of
May  1986.  A  second  international  workshop was  planned  for September 1986 in
the  United  States  to  evaluate  alternative  strategies  for  international
control.   After  the  workshops,  the next  round  of protocol  negotiations was
planned for November  1986,  with  a  second Diplomatic  Conference scheduled for
April 1987.

     Scientific assessments  are  proceeding on a  parallel track:   earlier in
1986,  the  National  Aeronautics  and  Space Administration  (NASA),  the  World
Meteorological Organization (WMO),  UNEP, and  other organizations completed two
years of work on  an  international  assessment of atmospheric  processes.   The
UNEP Coordinating Committee  on the Ozone Layer met  in January 1986 to review
the  state of  the science  and will meet again  in August.   The International
Conference on  the Effects  of Ozone Modification  and Climate  Change further
contributed to this process.

     Where have we come on the ozone layer internationally?  Looking back with
the benefit of hindsight on the past round of negotiations, I believe that the
international community  may  have tried to put the cart  before the horse:  we
were trying,  in effect, to make a risk management decision before conducting a
risk assessment.

     The first  part  of  this  approach, risk assessment,  involves identifying
the range of potential adverse effects and the affected population, estimating
the  timing  and  magnitude  of  those  effects,   and   assessing  the  attendant
uncertainties.  The second  step, risk  management,  entails developing a set of
possible  regulatory   options,  analyzing  the  efficacy  of  these options  in
reducing the  risks  identified,  evaluating the  economic  and  social  costs and
benefits of the various options,  and then choosing the most efficient option.

     Some of the virtues of this approach  are that it recognizes cost/benefit
analysis as an integral  part  of decision-making,  but it  separates such analy-
sis from assessment of  the  risk.   It recognizes  the  need to assess scientific
and other uncertainties, but  it does  not  allow  the existence of uncertainties
per se to automatically stop the  decision-making process.

     The two-part  UNEP workshop  uses  a format similar  to  that prescribed by
the "risk assessment-risk management" approach.  As far as I am aware, this is
the  first time  that  such  a process has been deliberately  established at the
international level and, as such,  it is truly an innovative approach.

     Whether  this  approach will  work  is  at  this  point  an open  question.   A
primary aim of the first part of the workshop was to try to find a reasonable
range of estimates of future growth in CFCs  and  the other trace gases,  which
could then be plugged  into  the  various global atmospheric models  to give us
estimates of the timing and magnitude of ozone depletion.

     The Rome workshop was only partially successful  in this regard.  Although
I wasn't there, I gather that the  Workshop's  limited success was due in large
part to  the  view of  some  delegations that because  making future projections


inherently entails uncertainties, nothing  can  be said about  the  future.   The
pitfalls inherent in  this  kind of all-or-nothing thinking  should be obvious.
For example, the notion that "future  risk  cannot be  estimated with certainty"
is sometimes translated, by  a  curious leap of  faith,  into  the assertion that
"there is no future risk"—when that may not be the case!

     We should suspend judgement on whether this approach  can in fact work on
the international level until  after Part II of  the  Workshop.   However,  if it
does  work,   I  believe  that  the  international  community  could adopt  this
paradigm in order to come  to grips with the issue of global climate also.

     As noted  at the UNEP/WMO/International  Conference on Scientific Unions
(ICSU) Conference on Climate Change last fall  in Villach,  Austria, it is well
established scientifically that  the earth's average surface  temperature will
rise as  the  result of increasing  concentrations of "greenhouse"  gases.   But
how much  warming will  occur,  and  when,   is  less certain.   And  what global
warming will mean for regional  temperature  and  rainfall patterns is still very
much uncertain.

     Our knowledge of the  impacts of climate change is  thus a few years behind
that of ozone  depletion.   It is therefore probably  too early for the kind of
structured workshops  that  we are undertaking with respect  to the ozone layer
issue  to  be undertaken  to deliberate  climate  change.   In  the  meantime,  I
believe that we  should  follow the  Villach Conference recommendation  to give
higher priority to researching  and assessing the impacts of climate change.

     From my perspective as a  career  diplomat,  it appears  that the greenhouse
effect  has  all  the  markings  of becoming  a high  visibility  foreign policy
issue.   The time  scale  between cause and  effect,  between policy action and
desired  result  is  distressingly  long.     Somehow,  political  leaders  and
government processes  and budget  makers must accustom  themselves  to a new way
of thinking.   The scientific issues are  complex, interrelating many different
scientific disciplines,  often  at the  frontiers  of discovery.   The uncertain-
ties are still a subject of debate.  How we address this issue internationally
depends to a great  extent  on our success or failure in dealing with the ozone
depletion  issue.    And  how we  deal  with  global climate  change  may  in turn
influence how we deal with other similar international issues.

     Thus, the activities  in which we are engaged—including this report and
the International Conference on  the Effects  of Ozone Modification and Climate
Change—are  of no  small   consequence to  the  world.   I  wish  these efforts
success and wise counsel.

 Man's Impact on  Earth's Atmosphere
James P. Bruce
Atmosphere Environment Service
Downsview, Ontario, Canada
     The astonishing  thing about earth is that it is alive in a corner  of the
cosmos where all  other  bodies  are  lifeless.  This was never more dramatically
demonstrated than 17 years ago, when  men  first  walked  on the moon and  sent
home  those  marvellous  photographs of  earth-rise  over  the  dead  moon.    Our
beautiful planet  home is  blue  with water, green and brown with vegetation and
land, and wreathed in  bands  of white  clouds of our atmosphere.  This  atmo-
sphere,   "as  thin as the dew  on  an  apple" (Watson 1984),  makes  the  rest

     We  now  know  that our species  has  the capability of modifying, in a major
way, the constituents of  the  atmosphere and  the climate it produces.  We have
inadvertently assumed a great  responsibility.  If we are to continue  to alter
the atmospheric balances, we  should be sure  we can predict and understand the
consequences and  be sure  that these changes  are tolerable.


     More than three decades  ago,  the great climatologist R.  Geiger  (1950)
wrote:   "the blue sky laughs  over  the  landscape,  while  in the  city all  is
covered  with grey and  the sun shines  along with  a weak yellowish-red  light.
Outside, it  is possible to see church  towers several kilometers away; inside,
the houses on  long streets soon disappear  in  impenetrable grey.  The  larger
the city, the denser, heavier and more  resistant is its haze hood."

     Thirty-five   years  after  Geiger's  observations,   the  Organization  for
Economic Cooperation  and  Development (OECD  1985)  was able  to list twenty-one
cities in its industrialized member countries where the  situation  had changed
drastically.    In  these major  cities between 1975 and  1983,  sulphur dioxide
concentrations have  declined  by 19$ to 65%  (the  latter  figure for Newcastle
and Oslo), with only  one, Rome,  showing an  increase in concentrations.   Simi-
lar  trends  are evident in concentration  of carbon monoxide  and  particulate
matter,  but  trends are not so positive for nitrogen oxides and photochemical
oxidants, such as  tropospheric ozone.

     These  reductions   have  been  achieved  by  enforcing  pollution  control
measures; by reducing economic  activity over  the measured period in the heavy,
most polluting,  industries;  and by building higher  stacks to  carry  the air-
borne wastes  beyond  the cities.  So while the. air quality in  our  cities has
generally improved,  there  has   been  a  growing  realization that much  of this
improvement may be causing contamination of a  large  part of the earth's atmo-
sphere as  well as environmental  damage in  the countryside far removed from
urban centers.   How  much the  situation  has  deteriorated  in the regions away
from cities is difficult to  tell, since  very few measurements  of rural air or
precipitation quality go back before  the 1970s.

     However,  we do  know that   some  of the  effects of airborne pollutants on
lakes and  soils  are cumulative year after  year;  we know that many  chemical
species are increasing  in the whole  global atmosphere; and we  know that stag-
nant  weather   conditions,  with  high  pressure  systems   over   the  industrial
regions of Europe and eastern North  America, result  in  "megalopolitan plumes"
of smog that can  blanket half a continent.  With these newer perceptions, much
of the scientific and  regulatory attention  on  air pollution has  shifted from
the cities,  becoming  global  and regional in scale.   We are now  concerned about
what Watson (1984) has  called the "international trade  in pollutants" brought
about by "the democracy of world winds."

     We are altering the chemical  balances  of the earth's atmosphere and the
relationship  between  our veil  of air  and  the earth's  biological  systems.
These biological  communities, which sustain our species,  have grown in harmony
with the air  we  breathe.  Or if you are a  follower of  Lovelock and  the Gaia
hypothesis, the biota have largely created the atmosphere, over the millenia,
with a chemical composition  designed to  sustain  the  life forms of earth.  But
we are upsetting  the balances   in a  relatively short time.  What  are some of
the ways  our  activities have been changing  the air  we  live in and  what are
some of the consequences?


     At  a  conference on the Great  Lakes in  1970,   I  felt justified,  after
reviewing some data  on  chemical contamination  of precipitation of the basin,
in saying that it  was  no longer a compliment  in Toronto  to say that a woman
was  "as  pure  as the  driven snow."   Chemical  composition of precipitation
reflects contamination  that  is  carried by the winds  for both  short distances
and distances  of thousands   of  kilometers.   Furthermore,  evidence  of serious
contamination of the rain and snow has  accumulated  alarmingly over the past 15

     The major  concern  of  the  media,  and with justification, has  been acid
rain, a problem in large regions such as eastern North America and Europe, and
an  increasing global  concern.    The  pH of  "clean"  rain  is  about  5.6 with
present COp concentrations in the atmosphere.  But major areas of the Northern
Hemisphere have been experiencing for some years—probably decades—precipita-
tion  with  acidity  averaging lower than  4.6  and as  low as 4.2—ten  to forty
times as  acidic  as  "clean"   rain.  The main  contributor  to the suite of acids
is  sulphuric  acid derived  from sulfur  dioxide  (SO^) emissions  from thermal
power  plants  and smelters.    The relationship  between  low pH  areas  and high
sulfate deposition is very strong in  Europe and eastern North America.

     The data from the Background Air Pollution Monitoring Network (BAPMoN) of
the World Meteorological Organization, part  of the  Global Environmental Moni-
toring System  (GEMS)  sponsored  by  the U.N.  Environment Programme,  has  been
used to assess the sulfur  in  precipitation on a global scale.  The areas with
average values greater  than  1.5 to 2.0 mg/1  in 1980-82  include eastern North
America; much of  Europe;  central USSR on both sides  of  the Ural mountains; a
region in the Far East  over eastern  China,  Japan,  Taiwan and the Philippines;
the east  coast  of South America between  Sao Paulo  and Rio  de  Janeiro;  and a
smaller less well documented area in southern Africa.   It is hardly surprising
that high  rates  of  sulfur deposition in rain and  snow are becoming nearly
global,  if  one  considers  the  tenfold  increase  in  anthropogenic  sulfur
emissions since 1900 (see Figure 1).

     The acidification  of unbuffered  lakes  is of  course closely  related to
sulfur  deposition.     Many   lakes   in   Canada,   the   northeastern  USA,   and
Scandinavia have become too acidic to  sustain fish.   Recent findings that the
droplets at  the  base of clouds  and  in some  fogs,  for example  in Japan,  are
often ten times  as  acidic as the rain have  generated concern about damage to
forests and  vegetation at high elevations  that are frequently  shrouded in
clouds and fog for long periods.

     But  it  is not  just  acidic substances  that are  contaminating  the rain;
toxic metals  and organics do so as  well.  The next  newspaper  headlines will
probably feature  "toxic rain."   For example,  atmospheric  loadings  of toxic
contaminants are  known to  be the  dominant  source  of some  pollutants  in  the
Great Lakes, particularly PCBs,  lead, zinc,  and copper (see Figure 2).

     Recent  evidence  suggests  that  in addition  to  being  a source  of toxic
contaminants for  the  lakes,  the atmosphere is  also a "sink," at least tempo-
rarily, for  toxic substances volatilized to  the  atmosphere from  the waters
(see Figure 3).  In fact,  recent studies of the budget of chlorinated benzenes
introduced by the  Niagara  River into Lake Ontario  suggest that more than 80$
of these compounds are volatilized into the atmosphere.  As much as 50% of the
PCBs  that  get   into  Lake  Ontario  water may return  to  the  atmosphere  by
volatilization.   What  is  most  disturbing  is  that  measurements of  a highly
toxic contaminant such  as  Mirex, known to have  been  only discharged directly
into Lake Ontario  waters,  have  been measured  in  crops,  cattle,  and milk from
land areas surrounding  the lake.  These represent a far more important vehicle
by which  toxic  chemicals  are likely  to  be  consumed  by most people living in
the basin  than  the  fish  and treated  drinking water.   Thus,  the scientific
questions of the volatilization processes and their implications for the toxic
chemical budget of the  region are under intensive research.


     These problems of  contamination of  rain  by acidic  and toxic substances,
while gradually becoming global, tend to be concentrated regionally.  However,
there  are  a  number  of atmospheric  contaminants that have a  nearly uniform
worldwide distribution  and are  beginning  to  have global effects, particularly
on  the earth's  climate,  on  the  vertical  temperature  distribution of  the
atmosphere,   on  stratospheric ozone,  and on  the  total amount  of ozone  in a
column above  the  earth at any location.  These increases in contamination of
the earth's  atmosphere are  probably decreasing  the  health-protecting  ozone
layer and can cause an unprecedented global warming.   Let us review, briefly,


               Tg  S/y»or









® Cullis ond Hirschler (1980)
© Katz (1956)
x Robinson ond Rabbins  (1971)
                   1860 70 80 90190010 20 30 40 50 60 70 80 yeor
Figure 1.   Estimates of Global Anthropogenic Emissions of Sulfur
           into  the Atmosphere (millions of metric  tons per  year)

Metric tons
 per year
 500 -
point sources

Rural runoff

Urban runoff

                                                         point sources
               Lead            Zinc              Copper

      Industrial loadings are not indicated for lead, zinc, and copper.
        Figure 2.  Metal  Loadings to  the Great Lakes,  by Source

                   Source:  Great  Lakes Basin  Commission and
                            Great  Lakes Science  Advisory Board

      s <   •-:.•>.'•••»..•  •.
      o £:;.?..».•. .•..•:.•/• -S .
      £; co 	^-'.'••*.• '•••?•'•
      °o-.'•••' • •«  • 't"" '•*•'.*.'./I
      v     •.•..•••;••;--••.•'
      o o
<5 iti
Z l"
C o
5 T

a «










some of the wonderful things humanity  has  achieved so far in changing the air
around us.

     The  substances  of  primary  concern  in bringing  about ozone  and climate
changes are  carbon  dioxide  (COp),  the  chlorofluorocarbons  (CFCs),  methane
(CH|j), and the  nitrogen oxides  (NO  ).   These are all  increasing  on a global
scale and current  annual  rates of increase in  the atmosphere are very large.
The most  important of the  CFCs  are  increasing at 5% to 1%  per  year;  C02 is
increasing at about  0.5%  per year and methane  at about  1%.   To  compound the
problem,  the  residence  times  of some of  these  gases  are very  long;  CFClo,
CFpCl2, and nitrous  oxide  remain in  the atmosphere 75  to 150 years.  C02 re-
mains  even  longer.   These long lifetimes  imply  that  annual increases are
cumulative over a long period of time and that any changes in climate or ozone
distribution  brought about  by such substances will last for centuries even if
remedial actions are taken.

     To complicate matters,  other   gases  introduced  into  the atmosphere by
human activities or  formed in the atmosphere by  reactions  involving anthro-
pogenic substances  also  have direct or indirect effects on temperatures and on
ozone  formation  or  dissipation.   These  include  the hydroxyl radical  (OH),
which reacts with  carbon  monoxide (CO) and methane  (CH|j).   With  increases in
both CO and CH^ the  cycle  between these  gases and the hydroxyl radical may be
reducing OH concentrations, and thus  the capacity of the troposphere to remove
many gases may be reduced.

     The  increase   in  methane  concentrations  is  particularly  interesting
because of difficulties in determining why it  is  occurring.   The data on the
upward trend summarized in  Figure 4  indicate  a  1.\% increase per year between
1976 and  1985.   Methane  concentrations are affected by natural microbiological
processes in water-logged  soils  and  in  the  intestines of  herbivores,  and by
industrial activities involving  the  use of natural gas, burning  of biomass,
and burning  of coal.   Figure  5 (Bolle,  Seiler, and  Bolin  1985)  shows the
probability that increases in atmospheric CHjj  are caused by both  industrial
activities and  society's  impact  on  natural processes.    The  figure shows the
close parallel between  world  population  trends and CHj,  mixing ratios for the
past  600  years,  as  determined  from  ice  cores  in  Greenland.  Table  1  gives
estimates by  various workers  of the  range  of values  of emissions from the
various sources.  Increasing numbers  of ruminants and rice paddies,  as well as
biomass and  natural  gas   exploitation,  are  the  main  sources of  increased
methane in  the  atmosphere.   Increased methane concentrations  will certainly
add to  the greenhouse effect,  but may also decrease  chlorine (Cl)  concentra-
tions  in  the  stratosphere, which  would  slightly reduce  the  rate  of  ozone
depletion resulting from CFC emissions.

     Nitrous  oxide  (N20)  is also increasing  in the  atmosphere but  at a con-
siderably slower rate.   The source  is primarily  microbial  processes in soil
and water, with both denitrification  and nitrification contributing  to release
of  N20.    Production is  enhanced by  application of fertilizers  of mineral
nitrogen,  especially ammonia-based fertilizers, and  from increases in the use
of lands  for agriculture.    Both  sources  are significantly affected  by farming
activities.   Natural processes in the  oceans  and  fossil fuel burning are also
N20 emission sources.   Global  emissions over the  past  century have increased
about 55% (see Figure 6).   In addition to its   relatively  small  effect  as a




                1  1.65
                .£• 1.60
                          NORTHERN  HEMISPHERE

                      1976  1977 1978  1979  1980  1981  1982  1983  1984  1985

     Figure  4.     Trend  of  tropospheric  CHh  mixing  ratios  (concentration)
observed  at midlatitudes of  the Northern  Hemisphere.   Data  are obtained  from
measurements carried  out on  aircrafts  (circles),  on ships  (squares),  and  at
different land  based stations during  clean air  conditions  (triangles).   The
figure  includes data  (dots)  measured  by Rowland  and  colleagues  (see  Blake
1984) at  similar latitudes in air  from the  Pacific Ocean  (from Bolle,  Seller,
and Bolin 1985).
                              CH4 MIXING  RATIO [ ppmv ]
                       0.6   0.8   1.0   1.2   U  1.6   1.8
                                           o Rasmussen and Kham. 1984

                                             Craig and Chou. 1982

                                           * Robbing et al.,1973

                                           • Present day level
                            1.0     2.0    "3.0     4.0
                              WORLD  POPULATION
      Figure 5.  Growth of Human Population (in billions)  and Increase of Atmo-
 spheric Cfy  Mixing  Ratios  (concentrations)  During  the  Last 600  years  (from
 Bolle,  Seiler, and  Bolin 1985).

Table 1.  Estimated Trend  of CHjj  Emission Rates from Individual Ecosystems
          (Tg per Year)  (from Bolle,  Seller,  and Bolin 1985).
Rice paddies
Other biogenic
Biomass burning
Natural gas
Coal mining
           1880  1890  1900  1910  1920  1930  19^0 1950 1960  1970 1980

                                TIME [years]
Figure 6.  Trend of Estimated Total  N20 Emission Rates Between 1880 and  1980
           (from Bolle, Seller, and  Bolin  1985).

"greenhouse" gas,  it also reduces stratospheric ozone  through chemical reac-
tions  that  form nitric  oxide.    It  is estimated  that  a doubling of  the N20
concentration alone would reduce stratospheric ozone by 3% to 5%.

     Increases in carbon dioxide  caused primarily  by burning fossil  fuels are
well documented  from the Pacific  to  the  Atlantic, and from  the  North to the
South Pole (see Figure 7).  Projection of future CCU concentrations has become
one of the  favorite  indoor  sports of  scientists and economists.   Eighty per-
cent of  the global  energy  demand  is currently met by burning  fossil fuel.
Projections of  energy demand and  of  the mix of  energy sources  that  will be
used are uncertain.   As a result, estimates of when C02 concentrations  will
double the preindustrial levels of about  275  ppm   are uncertain, ranging from
the middle to the end of the next century.

     The chlorofluorocarbons  (CFCs)-used as  refrigerants,  solvents,  propel-
lants  for  spray  cans, and  in  other ways-are exclusively manmade.   Since the
mid-1970s they  have  been identified  as a  source  of chlorine  in the  strato-
sphere, and thus as  a threat to  the  ozone  layer.    The most  important of the
CFCs are  CFClo  (F11) and  CF2C12  (F12).   Global emissions  of  F11  and F12
increased rapidly until  the early  70s,  followed by a slight decline caused by
restrictions on  spray can uses in North America  and Scandinavia (see Figure
8).  However, other  uses have continued  to increase and as noted, concentra-
tions  in the  atmosphere of these  long-lived  substances grow by  5%  to 1% per
year (see Figure 9).

     The effects of these changes on ozone  in the troposphere and stratosphere
have been observed.   Ozone  is  a  radiatively active gas that shields the earth
from ultraviolet B  (UV-B)  radiation from the sun.   A  1$ decrease in ozone in
the atmosphere results in a 2% increase in UV-B radiation, which  in turn would
cause a 4%  increase  in skin cancer.   There are  considerable uncertainties in
trends in ozone  concentrations since  individual measurements at times require
significant corrections  and since  the limited  number  of  individual measure-
ments  for   large segments  of  the troposphere  or  stratosphere  may   not  be
entirely representative.  The data available, for example, from Resolute, NWT,
and  Hohenpeissenberg,  Germany  (see   Figure  10),   strongly  suggest  that  an
increase in  ozone  near  the ground in the  middle  and  high  latitudes of the
Northern Hemisphere  has  occurred  over the  past few  decades because  of indus-
trial activities at  a rate  of 1%  to  2% per year.    At  the same time,  strato-
spheric ozone concentrations  between  28 and 35 km above the ground  appear to
be declining; satellite  data  suggest   that  levels  have  declined as much as 5%
at the 45-km (1  mb)  level.   Because about 90% of the atmosphere's ozone  is in
the  stratosphere,  some  estimates  of   total column ozone  indicate  that  small
decreases have occurred since 1960.

     However,  the  most  startling change  in  stratospheric  ozone  has  been
observed over Antarctica.   The Dobson spectrophotometer  at Halley Bay (76°S)
operated by  British  scientists has shown decreases in  ozone in the Southern
Hemisphere spring,  especially since the  early  70s, amounting to more  than 40%
for  the month of October.  Recent satellite  measurements from-Nimbus 7  indi-
cate that  this  is  a  much more widespread  effect  than  suspected  and that the
decline appears  to  extend  over  the  whole  area of  the winter polar  vortex.
While several workers have suggested that this is  caused, at  least in  part, by
chlorine   increases   in  the   stratosphere  from   the  chemicals  discussed
previously,  the  modeling of  the  mechanisms  to  explain such  a rapid decline
remains a major scientific problem.



           —  335
           •  330
                    Figure 7.   Carbon Dioxide Concentrations
CFC 1 1 Rl
- 900
_! B00
£ 700
z 600
K 500
§ «••
"• 300
§ 200
a 100
1 9
sJD CFC12: X^«-



/ v







60 1965 1970 1975 1980 1 985
SOUHCC: cut* ttaayjt KRND 119001;
us t re ri96f-i9au
     Figure  8.   World  production has  remained nearly  constant since  the mid-
1970s.   While aerosal production has steadily declined,  nonaerosol  production
has consistently  increased.


        1 200

        2 190

        2 180

        Z 170

        5 160
               1978     1979
                                                                1983     1984
s. 350

2 23°
5 310
O 290

5 270
               1978     1979        1980       1981        1982        1983     1984
        a 130


        s. 110
        5 100
        I  90
               1978     1979
                                                                1983     1984
        g 130

        < 120

        O 110

        B 100
           II,  m,,,""""""1"""1
               1978     1979
                                                                 1983     1984
     Figure 9.  Monthly Mean  Mixing Ratios (concentrations)  and Monthly Vari-
ances  of CFC-11,  CFC-12, CHoCClo, and  CCljj Measured Four Times Daily  on a Gas
Chromatograph at  the Ragged Point, Barbados ALE/GAGE Station During the First
Six Years of the  ALE/GAGE Network (parts per  trillion volume).
Source:   NASA Assessment Report

                               A OZONE (%yr"1}
     Figure  10.   Percent increase or decrease in ozone at various altitudes
(expressed as pressure in millibars)  at Resolute (75°N,  Canada,  1966-1979) and
Hohenpei-ssenberg (47°N, Germany,  1970-1983).  The horizontal bars  give the 90$
confidence  interval.    The  shaded  area  shows  the  range of  the tropopause
heights (from Logan 1985).

Source:  NASA Assessment Report

     Models of ozone changes caused by these atmospheric contaminants indicate
that  the  maintenance of  stratospheric  ozone is  critically dependent  on  the
rate of growth  of  CFC use.   If  CFC  emissions could be returned  to  and main-
tained at the 1980 levels, the models predict only very small changes in total
column ozone.   Increases of  1.5%  per  year and 3% per  year make  increasingly
large depletions likely.   However, two-dimensional models  show strong latitu-
dinal difference with depletions throughout the stratosphere north of 40°N and
some  ozone  increases  in  the  lower  stratosphere nearer  the tropics.   Polar
depletions in the ozone column are predicted  to  be  two to four times those at
the equator.

     The future concentrations of gases affecting the  ozone layer and climate
are difficult to predict  since they  depend on regulatory  decisions by govern-
ments, energy policies and consumption, and other human interventions as well
as a more thorough understanding of physical and chemical  atmospheric process-
es.   However,  it  is easier to predict what  concentrations of some substances
will be if people  and governments take no action.   Ramanathan and colleagues
in  1985  published  a comprehensive  set  of projections of  concentrations  of
forty-seven  trace  gases  of  the  atmosphere  that  affect  radiation  balances.
Many  of  these have  a very minor  role to  play.   The  relative importance  of
these gases  to the  "greenhouse" effect  is  summarized in  Figure 11,   which
gives the  cumulative equilibrium surface  temperature warming  caused by these
gases from  1980 to  2030  as  computed  from a one-dimensional  model  following
Ramanathan's work.   Because of positive  feedbacks predicted by general circu-
lation models,  the  most  probable  temperature changes  at  the  earth's surface
are one and a half  to two times  those given in the diagram.  The  figure shows
that  CFCs  are  next in  importance  to carbon  dioxide;  that  methane,  nitrous
oxide, and ozone are significant  contributors; and  that the forty or so other
radiatively active  trace gases  add  little to the  total.    Carbon  dioxide  is
about half the story.

     In October 1985 in Villach,  Austria, a group of nearly  100 of the world's
leading scientists  and policy analysts  in the  field  gathered to  assess  the
present state of  knowledge of greenhouse  gases and  climate change.   The con-
ference concluded that "The role  of  greenhouse  gases other than C02 in chang-
ing  the  climate is  already as  important  as  that of C^.   If present trends
continue, the combined concentrations of  atmospheric COp  and other greenhouse
gases would be radiatively equivalent to a doubling of COp from pre-industrial
levels possibly as  early  as the  2030's."   Hansen et al. tthis volume) suggest
a date as early as the 2020s.

     The consensus  of the results from  general circulation  models  about  the
effects of such  a   doubling  of  COp  was   also reviewed at  Villach.    It  was
concluded  that  the  models indicate  that  the global mean  equilibrium surface
temperature would  rise by 3°C ± 1.5°C.    In  the shorter  term, a further 1°C
rise  beyond  the 0.5°C  of the past  century  could  occur  within  the next  two
decades,  making the earth's  climate  warmer  than   at  any  time  in  the past
100,000 years.  However,  the  scientists working with these models are quick to
point out imperfections of and problems with the models, particularly that the
hydrologic cycle,  clouds, and  ocean-atmosphere  interactions  are not modeled
with  much  confidence.   Thus,  while  model  results are  converging  on the range
of values indicated, uncertainties in the  modeling make many of the scientists
most  involved cautious  about the  results.   They believe  that we can use the

                                     Other  trace
                                     gases  -^
                                        Nitrous oxide
                   Chi orof1uorocarbons
                    Carbon dioxide
 Range of
 uncerta inty

1 .5
     Figure  11.   Cumulative  Equilibrium  Surface  Temperature Warming Due  to
Increase  in  Carbon Dioxide  and  Other  Trace  Gases from  1980 to  2030 AD  as
Computed  by  a  One-Dimensional Model  (after  Ramanathan et al.  1985).   Due  to
positive feedback mechanisms as revealed by general circulation models, actual
warming may  be 1.5 to  2  times the values given in this  figure (from Bolle,
Seller, and Bolin 1985).

outcome as  some  indication of what  may happen, but  not as a  fully  reliable

     In addition, the models do not agree on regional distribution patterns of
temperature changes or on changes in regional precipitation.   The indications
are, however, that the larger  temperature  increases  will be in high latitudes
in winter and autumn  and that summer dryness may become more  frequent in the
midlatitude Northern Hemisphere  (see Manabe and Wetherald,  this volume).   The
Villach projection of  sea-level  rise,  based  simply  on  extrapolation  of  past
experience,   is that  for  global  temperature  increases from  1.5°  to  4.5°C sea
level would  rise 20  to  140  centimeters (see Titus,   this  volume,  and Thomas,
Volume 3).

     This  has  been a  very brief  and  simplified overview  of  a  very complex
matter, i.e., the changes in  the  global atmosphere.   It might  be useful to
reflect on  the  implications of  the  present  state  of knowledge for all  of us
and  especially   for  those concerned  with public  policy.   Four main  points
should be  considered.

     First,  these are not, as  some have assumed, problems for  the next gener-
ation, but  urgent issues  for  us  today.   The protective  stratospheric  ozone
layer is decreasing,  and  preventive measures taken  so far have  been inadequate
to arrest the global increases in  the  substances causing the change.   Climate
warming caused by greenhouse gases  appears  to  be  under way and  is  likely to
reach very significant proportions in the first half of the next century,  even
if the inertia of the system caused by  the oceans delays warming by  a decade
or two.   As the  Villach Conference statement says,  "Many important  economic
and social decisions are  being made today on long-term projects—in irrigation
and  hydro-power;  combatting  of drought;  land  use; agriculture;  structural
designs; coastal engineering projects;  and  energy  planning."    The  statement
points out  that  even if  the  consequences of global  warming are  ignored, the
projects are based on some assumption  about  the  climate of the next 50 to 100
years.  It  is a  matter  of real urgency to  be able  to  provide more  reliable
predictions and guidance.

     Second, the need is  great for governments and  funding agencies to support
research on  the  changing  atmospheric composition and its effects.  Uncertain-
ties  in our predictions of  climate  change  and   ozone  layer  modification
continue  to be  disturbing.    It  is especially important that  major  inter-
national efforts such as the World Climate Research  Programme, the  Global
Tropospheric  Chemistry  Program,  and  the  proposed   International  Geosphere-
Biosphere Programme be adequately supported.

     Third,  remedial,  or at least  preventive,  measures seem  to  be  possible.
As the Villach meeting noted,  "The rate and degree of future warming could be
profoundly  affected  by  governmental  policies on energy  conservation,  use of
fossil fuels, and the emission  of some greenhouse gases."  I  am reminded of
the  debate  of the early  70s on  how to  control the  rate of eutrophication of
lakes.  A  number  of  nutrients were over-stimulating algal  growth  in many of
our  lakes,  but  one key  nutrient,  phosphorus, was most  readily controllable.
In our present case,  CFCs are  one  group of substances that clearly have major
impacts on  both  climate  and  on the ozone layer.  Some of their uses  for spray
cans  and  foam blowing can hardly  be considered essential  and are  therefore
controllable.   It is my view that  governments  should  move quickly  to  a CFC

control protocol  under  the  Vienna Convention  on the  Ozone Layer,  and  that
individual governments  and companies  should  be urged  to limit their  use  of
these chemicals.  To advance  the  development  of such an international control
program, the directors  of environmental programs from many  countries need  to
be briefed on  these  issues and should  discuss  international remedial action.
An S02  control protocol  signed  in Helsinki,  July  1985, by  twenty countries
under the  Economic  Commission for Europe  and ministerial meetings  in Ottawa
and Munich were essential  steps in moving  the process forward.   If the scien-
tists and bureaucrats make little progress  in achieving  a CFC control program
by early  1987,  we  should ensure that  the most  senior elected and  politically
appointed officials become involved.

     Finally, we must  accelerate  acceptance  of a more  comprehensive  view  of
the  earth's  atmosphere  and   the need to   protect  its  beneficial  charac-
teristics.   The problems of the ozone  layer,  of greenhouse  gases  and climate
change, of  long-range  transport  of  acidic and toxic substances  are closely
interrelated.   They are all  problems caused  by human contamination  of  the
atmosphere—the only one we  have.  A  new approach,  a new ethic,  towards  dis-
charging wastes and  chemical  materials into  the air  we  all  breathe must  soon
be adopted  on an  international  scale.  Without this we will  do  irreparable
damage to our earthly home.

Bolle, H.J., W. Seller, B. Bolin.  1985.  Other greenhouse gases and aerosols.
    In  Greenhouse  Gases  and  Climate  Change:    An  Assessment  For  Villach
    Conference 1985.  Stockholm: IMI,  (to be SCOPE publication).

Geiger,  R.   1950.    The  Climate Near  The  Ground.    Cambridge,  MA:  Harvard
    University Press.

National Aeronautics and Space Administration (with WMO). 1986.  An Assessment
    Report:   Processes That  Control  Ozone  And Other  Climatically Important
    Trace Gases.   NASA Ref.   Publ.  No. 1162.

Organization for Economic Cooperation and  Development.   July  1985.   The State
    Of The Environment 1985. OECD Observer No. 135.

Wallen, C.C. 1986.   Sulphur and  Nitrogen In Precipitation:   An Attempt To Use
    BAPMoN   And   Other  Data  To  Show  Regional   And   Global  Distribution.
    Environmental Pollution Monitoring and Research Programme, Publication No.
    26, Geneva: World Meteorological Organization.

Watson, Lyall.   1984.   Heaven's  breath:  A  Natural History Of The Wind.  U.K.:
    Coronet Books.

World Meteorological  Organization.    1986.   Report of  the  International Con-
    ference  on Assessment of the Role  of Carbon Dioxide and  of  Other Green-
    house Gases  in  Climate  Variations and  Associated  Impacts.  WMO publication
    No. 661. Villach, Austria:  WMO.

The Importance  of  Knowing Sooner

John S. Hoffman
U.S. Environmental Protection Agency
Washington, DC USA

     Unlike most  environmental  problems,  the  greenhouse  warming and  ozone
depletion are  global  in  extent, possess enormous momentum for change,  and may
be  almost  irreversible  once   they  occur.    Recent scientific  assessments
uniformly agree that the changes that are  likely to occur in this century will
exceed the range  of natural variation experienced  in the whole last  century.
Given our  relative  ignorance  of the  possible  effects  of these  changes,  it
would appear that  the value of knowing about  effects  sooner is substantial.


     The relative  paucity  of scientific  research that has  been undertaken to
understand the health and  environmental effects of ozone depletion and global
warming  can, in part, be explained by the  belief that has been held that these
phenomena were relatively distant and uncertain.  Clearly, that view is now in
doubt.   Recent  scientific  reviews  sponsored  by  the   World  Meteorological
Organization,  the United Nations Environment Programme, and  others  indicate
that the speed of  atmospheric change has  been greater than thought (NASA 1986;
WMO  1986a; WMO 1986b).   These reviews contend that the magnitudes of ozone
depletion and  global warming are likely to  be large enough to emerge above the
level of natural  variations  that have  occurred within this century,  which for
climate, at least, has been quite large.

     In   this  volume,  papers  by   Hansen  et  al.  and   Stordal  and  Isaksen
strengthen that  message by  presenting  some  important  new  results.    Both
research teams used,  for the first time,  the most  complex  models to  simulate
atmospheric change  with  realistic  time-dependent  scenarios  of  emissions.
Their models predict  not just  global  average changes in ozone and climate but
the  changes  that will  occur  through  time  at  different  latitudes.      The
results   indicate  that at  some latitudes,  changes  will be  much larger than
many had expected.  As such, these results  are certain to heighten our  concern


about  environmental  and  health  effects  of  ozone  depletion  and  climate
change.  As  such,  a new sense of urgency  is  likely  to  develop about the need
to understand the effects of global atmospheric change.

     Two  other  important  justifications  exist  for  understanding  effects
sooner.   A  large  amount of  momentum  has developed  for ozone  depletion and
global warming, and these phenomena may  be essentially  irreversible once they
occur.  Momentum is relevant because the more that exists, the harder it is to
stop.   Irreversibility  is  important  because  if  one   cannot  go  backwards,
momentum threatens  to take  one forever  beyond  the  point  one  wishes  to go.
This paper  explores the forces  that  create  momentum and  irreversibility for
global atmospheric systems.


     Most pollution problems  are local  in character  and can be quickly solved
by reducing emissions.  For example, if the concentration of particulates in a
locality is  judged too  high, the emissions  from sources  that  contribute to
pollution  in that  area can  be reduced  proportionally  to the  reduction  in
concentrations desired.  Within days  of  turning  on  the controls, the problem
will be solved.   The  situation for global atmospheric problems is radically

     Chlorofluorocarbons  (CFCs) 11,  12,   and 113;   nitrous  oxide; and the
halons have long lifetimes (NASA 1986).  Emissions of these gases do not leave
the atmosphere quickly, but linger for decades or more.   In fact, for CFCs 11,
12, and 113; halons 1211 and  1301;  and nitrous oxide, the only real sink  (loss
mechanism)  is the stratosphere itself,  where high energy ultraviolet radiation
is capable  of  photodissociating these  gases.  Consequently,  for any level of
emissions  into  the atmosphere,  concentrations   will  increase  in  the  lower
atmosphere  until the  point   in  time  that the  transport to  the stratosphere
equals the global emissions.   We are  currently far from that point.  Thus, if
current emissions did  not grow,  concentrations would continue to increase for
over  one  hundred years, as  shown  in Figure  1  for CFC 12.    If we wanted to
stabilize the concentration of CFC 12,  total emissions  in the world would have
to be reduced approximately 85 percent immediately, as  shown in  Figure 2.

     This  disequilibrium between  emissions  and   concentrations  constitutes a
kind  of momentum for  change  in the world.   Change  is  built  into the status
quo.   Putting  on   the  "brake," except  in the most  extreme  case,  would not
prevent change,  but would only  slow  it down. (Figure  3  shows the results of
various reductions  in emissions on the increase in 2  concentrations.

     In point of fact,  the  world is still gathering momentum.   Emissions have
been increasing and are likely  to continue to do so according  to most studies.
Figure  4  shows  fifteen  year projections  of  CFC  11  and 12 for a variety of
papers  presented at the UNEP conference in May.   In developed nations, many
uses  of CFCs  are  growing  faster  than  the  economy—production  of personal
computers,  in  which CFC 113  is  used as  a solvent for  the  chips, is just one
example.    Insulation  is  another.     In many   developing nations,  current
consumption can  be  expected  to  grow  faster than the  economy,  as  segments of

    CFC-12:  Constant Emissions
                (mill kg)





CFC12:  Atmospheric Concentrations
       Figure  1.    If  future  emissions  of CFC-12 were held constant at  today's
  levels (Figure  1a),  atmospheric concentrations would continue to rise  for  over
  one hundred  years (Figure 1b).
       CFC-12:   Emissions
               (mill kg)





 CFC-12 Atmospheric Concentrations
0.20 -
0.10 -
       Figure  2.   Atmospheric concentrations of  CFC-12 are  far  from  equili-
  brium.   Only if emissions were cut 85% and  not  permitted  to  increase  (Figure
  2a)  could  atmospheric concentrations be held constant (Figure  2b).

            CFC-12:  Atmospheric Concentrations
            from Different  Emission Trajectories
                                                     15% Cut
                                                     50% Cut
                                                     85% Cut
     Figure  3.   Atmospheric concentrations  of CFC-12 will  continue  to  rise
unless emissions are cut.   Holding emissions constant  at today's level or even
15$ or 50% lower would still allow atmospheric concentrations to grow.  Only a
cut of 85% or more could stablize atmospheric concentrations.
                             SHORT TERM
                   Nonaerosol Projections: CFC-11 and CFC-12
Bevlngton f^

Hedenstrom [_ 	

Kurosawa 1'""* 1 Canada


           0.0     1.0    2.0     3.0    4.0     5.0
                            Annual Rate of Change (%)
     Figure 4.  A variety of researchers have estimated  future demand for CFCs
in various countries and  regions.   Despite their diversity of  methods  and
data, they are in  substantial  agreement  that  short-term demand  for  CFCs in
non-aerosol uses will continue to  increase.

the populations reach the state of being able to afford goods that use CFCs or
other chemicals.   People are likely to want  refrigerators,  air conditioning,
and electronics as their  wealth  increases.   As  populations and economies grow
we can expect  the demand for goods and.services  to grow in developed nations
too,  including those  that  use  potential  ozone  depleting  chemicals.    This
process creates a major source of momentum for accelerated atmospheric change.

     Another form of momentum exists for  global warming.  At the present time
the earth has not experienced all the warming that would be expected from past
increases in greenhouse  gases.   Instead,  some  of  the infrared energy trapped
by  past  increases in  greenhouse gases  has  been  used  to  heat up  the  upper
layers of the  ocean  (MAS 1979).  While this  effect has retarded the speed of
the  warming   to   date,   eventually  we  can   expect  ocean  and  atmosphere
temperatures to  rise  to their  equilibrium values.   The  current  difference
between the  realized and equilibrium temperature constitutes  a  major form of
momentum  for  global  warming—a  form that is inescapable.    For   the  model
simulation of  global  warming performed by  Hansen et al.  in  this volume, the
temperature change realized by 2000 is 1°C.  The unrealized warming also turns
out to be 1°C.   This means that even if the concentrations of greenhouse gases
were totally stabilized  by that year the world would eventually experience a
2°C average global warming.


     Clearly momentum makes  it difficult  to stop  any system quickly.  It is a
different question whether it is possible to go back once one has stopped.  As
it  turns  out,  however,  restoring lost ozone or cooling  the world  would  be
difficult.   As discussed earlier,  the  gases likely  to cause these phenomena
have long lifetimes.   Even  if  emissions are totally eliminated, it could take
decades  for  the  concentrations  to  fall.    A total  elimination  of emissions
seems unlikely, but if it did  occur in  the case of ozone depletion, one would
expect  to eventually  see  the  concentrations  of  ozone  restored,   all  other
things being equal.  For global  warming,  this may not be the case.   It may be
that  warming  makes  it  impossible  for all things  to  go  back to  what  they
were.   In particular,  the  water  vapor  that has  entered the  atmosphere and
amplified  the  radiative  warming from  other  greenhouse gases  may  not  just
disappear from the atmosphere.   If it resides  in the atmosphere permanently,
it will  continue  to  enhance global temperature,  perhaps forever.   Regardless
of whether this happens,   the warming will last a long time.


     A system  that is  hard  to  stop  and  that may be impossible to reverse pre-
sents a very different issue from those facing decisionmakers currently in the
environmental  arena.   A  problem  that  is global in nature, in which any change
can expose  billions  of  people,  all  ecosystems, and  all outside materials to
change, raises issues that are also very different  from environmental problems
in localized systems.   It is not difficult to  imagine  that as recognition of
the  unique   aspects  of  global  atmospheric  change becomes  more  widespread,
decisionmakers will  begin to  focus on the effect of  depletion and warming,
recognizing the importance of knowing sooner.


National  Academy  of  Sciences (MAS).  1979.    Carbon dioxide  and climate:  A
     scientific assessment.  Washington,  D.C.: MAS.

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.   Wash ing ton,
     D.C.: NASA.

World  Meteorological  Organization  (WHO).   1986.    Atmospheric  ozone  1985:
     Assessment of our understanding of  the  processes controlling its present
     distribution  and change.    WHO  Global  Ozone  Research  and  Monitoring
     Project Report No. 16.  To be published.

World Meteorological  Organization (WMO).  1986.   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. Villach,  Austria: WMO.

Our Global Environment:  The Next Challenge
John H. Chafee
United States Senate
Washington, DC USA
     The problems  associated with ozone depletion,  the greenhouse  effect, and
climate change  have  extraordinary  ramifications for  the  world we  inhabit.
Although these problems have not yet been  elevated to the same level of public
awareness  as  toxic waste or acid rain,  more  and more people are becoming aware
and concerned about  them.   The  press that these  issues received in the second
week of June 1986 is evidence of mounting awareness and  concern.   There have
been stories and editorials in newspapers  across the  country  and  in  Canada,
national television, radio, and magazines  like Newsweek and The New Yorker.

     On June 10 and  11,   1986 we conducted two  days of hearings  before the
Senate Subcommittee on  Environmental Pollution to explore the nature of these
problems and to examine what is  being done by  the  U.S.  government, domesti-
cally and  internationally,  to both  improve  our understanding  of these  matters
and to respond to them.   Why did we decide to spend  time  on these problems?
Why is all of  this  so  important?   Why are  the  policymakers  demanding action
before the scientists have resolved all of the  questions  and uncertainties?
We are doing  so  because there is a very real possibility that society—through
ignorance  or indifference, or both—is irreversibly altering  the  ability of
our atmosphere to perform  basic life support functions for the planet.

     This  is  not a matter  of Chicken Little  telling us the sky is falling.  At
our hearings, we heard graphic,  powerful and disturbing testimony from  distin-
guished scientists such as Bob Watson  and Jim Hansen of NASA, Sherry Rowland,
George Woodwell, Stephen Leatherman, and Carl Wunsch.  The scientific evidence
is telling us we  have  a  problem,  a serious problem.  There  is much  that we
know.    There is  a  great  deal  that  we  can predict with a fair amount of

     As evidenced by  the  October  1985 Villach  Conference,  there is  now an
international  consensus among  the scientific community.   Although  there will
always be  dissenters—those who claim,  for example, that the earth is actually
cooling,   not  warming—the scientific  community has  told  us with  unusual


clarity that we have  a  problem.   We must  not  allow their message  to  fall on
deaf ears.

     It is  true that  we lack the tools  to close all of the  scientific gaps.
We  don't  completely  understand  our  climate systems  and we cannot  predict
precise outcomes.    But  we will always be  faced with a level  of uncertainty.
The current  gaps  in  scientific knowledge  may not  be  closed for many years.
Scientists have characterized our  treatment of  the greenhouse  effect as  a
global experiment.   It strikes me  as a form of planetary Russian roulette.

     We should  not be  experimenting with  the earth's  life support  systems
until we  know that  when the  experiment  is  concluded, the  results  will  be
benign.    As  Russell Peterson (former  Chairman  of the  President's  Council on
Environmental Quality) who worked as a chemist  for  twenty-six years has said,
"We cannot  afford  to give  chemicals the  same  constitutional rights  that we
enjoy under the law ...  chemicals  are not innocent until proven guilty."

     By not  making policy  choices   today,  by sticking  to a  "wait and  see"
approach,  we  may  in  fact  be making  a passive choice.   By  allowing  the  so-
called "greenhouse"  gases to  continue  to  build up  in the   atmosphere,  this
generation may be  committing all  of  us  to severe economic  and  environmental
disruption without ever  having decided that the value  of  "business as usual"
is worth the risks.

     Those who believe  that  these are problems  to be left  to  future  genera-
tions are misleading  themselves.   Human activities  to date  may  have  already
committed  us to some  level  of temperature change.   If  historical evidence is
any  guide,  a  slight warming may  be enough  to  turn  productive,  temperate
climates  into  deserts.   To  quote from a recent Department  of  Energy  report,
"large changes in both precipitation and the extent  of  deserts and grasslands
can be associated  with relatively  small variations in the global mean tempera-

     The  path  that society  is following  today  is  much  like  driving a  car
towards the  edge  of a cliff.   We have  a  choice.   We  can go ahead,  take no
action and drive off  the edge—figuring  that, since the car  will not  hit the
bottom of the canyon until our generation is already long gone, the problem of
coping  with   what  we   have  made   inevitable  is  a  problem  for   future
generations.   We  can  hope that they will learn  how to adapt.   On the other
hand, we can put the brakes on now,  before the car gets any closer to the edge
of  the cliff and before  we reach  a  point where momentum will take us over the

     What do we do about all  of this?

     First,  it  is  important  to  focus attention  on the potential  effects of
ozone modification and climate change.   Simply  telling  people that there will
be  a change  is not  enough.   They  need to understand how they will be affected
by  it.  The  International Conference on  the Effects of Ozone Modification and
Climate Change was an excellent  step  in that direction.  Many  people came a
long way to participate in that gathering.

     In addition, we must  begin to consider the choices  that  will have to be
made  if we  are  going  to  avoid further  build-up  of harmful  gases  in  the
atmosphere.   All  of us must  recognize  that these are no  longer just science
issues.  They are not policy issues.  They demand solutions.

     Many  of the  government witnesses  who appeared  before  my Subcommittee
argued that we need more studies.  They contend that there are too many scien-
tific  uncertainties  to warrant  action.   Only EPA  Administrator  Lee  Thomas
recognized what  is at  stake  here'.    He was the  only decisionmaker  from the
government who  appreciated  the fact  that  there  will  always  be  scientific
uncertainty  and  that  policymakers  and  those  who  make  regulatory decisions
cannot allow themselves to be paralyzed by these gaps in knowledge.

     Sure, we can  continue to  study  the problem—and we  should continue our
studies—but  we  cannot  wait  until these  studies  are  completed.   We  need

     Those who argue against action like to recite the caveats, uncertainties,
and margins of error that accompany scientific  reports and projections in this
area.  For those people, it  is  worth  noting that the margins of error usually
associated with the greenhouse  effect are  in terms  of decades.  The issues of
magnitude and timing need  to be resolved but  they do  not need to be resolved
before  we  take  action.   It  is  important to  talk  about  projections  of,  for
example, global  warming over the next 50 to 100 years.  But disagreements over
the accuracy of projections for that time frame obscure the issue:  Do we have
the right to pollute the atmosphere today in a  manner that will wreak havoc in
as few as 100 to 300 years?

     Obviously,  no  single  step  is  going to solve  the multitude  of problems
associated with these matters.   But we must not let  the  enormity of the task
keep us from taking a series of small steps.

     The growing  use  of chlorofluorocarbons (CFCs)  is a  prime  candidate for
controls.   Given  the  risks  associated  with  ozone  modification  and  climate
change,  a  treaty  that  limits  the  availability of  CFCs  would  seem to  be a
sensible course  to pursue.   The good news is that,  under the leadership of the
United  Nations  Environment  Programme,   negotiations  aimed  at  developing  an
international agreement  to limit future growth in CFCs—to impose controls—
are  currently underway.    The  not-so-good  news  is  that we  don't have  an
agreement yet.

     At the  Environmental  Pollution Subcommittee hearings, Dr. Rowland raised
an interesting suggestion  that  I intend  to  examine  more closely:   In addition
to  imposing  controls on  the use  of CFCs  generally, why don't  we  redesign
equipment such as car air-conditioners  to use  alternative, less harmful forms
of CFCs?   Although CFCs are  not the only  source  of the  problem,  they are a
significant  factor  and  controlling them would represent  a major accomplish-

     The need  for  international  cooperation  in  addressing these  matters  is
obvious.  We are  dealing with a global  issue.  But  the  difficulty of getting
international agreement should  not  keep the United States from taking action
unilaterally.  By taking such action, we can move  once again into  the role we
once proudly  held:   we can  be  world  leader in environmental  protection.   We


can set the example.  A  country  as  large and as powerful as the United States
can afford  to  make temporary sacrifices  that ultimately will  benefit all of

     The  importance  of  these  issues  is matched  only by  the complexity  of
them.   But  we  have created  complex  problems before and we have  responded to
them with creative solutions.    There  is no  reason  to believe that  the only
answer is to throw up our hands and  wait for the inevitable.

     We lose so very little by  trying.   We lose everything by  doing nothing.

     Tough choices need  to be  made  and we should start what  promises to be a
long, sometimes  frustrating process of  making policy choices  and regulatory
decisions.  All of us will have key  roles to play in that process.  By working
together, we  can  attempt  to  minimize  the  disruptions  that  go along  with
changes in  the  status quo.  But there will  be changes.    Present and future
generations  of all life forms depend on our making the right choices.

The Greenhouse Effect

Albert Gore, Jr.
United States Senate
Washington, DC USA
     I am pleased  to  have  the opportunity to join the other authors of this
report to accelerate  our progress  on the global problems of climate change.
In order  to resolve  these  issues,  we  will  need the  kind  of  international
cooperation   and  leadership  that   organizations  like   the  United  Nations
Environment   Programme  and  the  U.S.  Environmental  Protection  Agency   can
provide.   Let  me briefly discuss  one  of our gravest climatic  concerns—and
perhaps the  most  significant  environmental  problem  of the next  century—the
greenhouse effect.

     The   buildup  of  atmospheric  carbon  dioxide  and other  trace  gases   is
causing a global  warming that  could have disastrous results  on sea  level,
agriculture, and the  climate  in general.  Other  papers  present some of  the
forecasts—a 3  to  8  degree  (F)  temperature  increase by  2030,  a 10 percent
decrease  in  crop yields, more frequent  tropical  storms.   And  if today's worst
case  scenarios  come true we  may  have only  a  few decades left  in  which  to
decrease  these  impacts.

     My interest in the greenhouse  effect goes back  several years, and  I have
chaired three  congressional  hearings on  the phenomenon.  During the first
hearing in  1981, a  number of  prominent  scientists,  including Professor Roger
Revelle of the  University of California,  testified that the greenhouse  effect
was not a hypothesis  but a  reality.  This marked perhaps the first time that
the scientific  community had reached a consensus on this point.

     The  evidence  to substantiate that  conclusion was not long  in coming.   In
19§2,  Dr.   James  Hansen of  NASA  and  Professor George Kukla  of   Columbia
University  testified   that  the  rise  in  carbon  dioxide levels could   be
correlated  very closely with a  rise  in the  earth's   mean  temperature, a
shrinking of glacial  ice,  and  the resulting rise  in  the earth's  mean   sea
level.   At  that time the Washington Post, in  an editorial,  stated  that  the
greenhouse effect  was no longer just something  for  the "sandals  and  granola"
crowd; it was in the mainstream of scientific thought.


     The  most  comprehensive  evaluations  of   the   greenhouse   effect  were
presented at  the  third  hearing in 1984.   One  was an EPA report  predicting a
global temperature increase of 3.6 degrees Fahrenheit by  the  year 2040, and a
rise of as much as 9 degrees Fahrenheit by the year 2100.  This  would result
in sea  level  increases ranging  from 4.5  to  7 feet  by the  end  of  the  next

     A 1983 report from the National  Academy  of Sciences was more  conservative
in its  projections.   The  WAS predicted  a 2 to  8  degree Fahrenheit  rise in
temperature along  with a "modest" 2.5 foot rise  in  the  sea  levels by the year
2100.  The major difference between  the  two  studies  was that  EPA  included the
impact of  all the greenhouse  gases, including CFCs  and  methane, as  well as
carbon dioxide.  The Academy projections were based  solely on  increased carbon
dioxide concentrations.

     In October  1985, a conference on the greenhouse effect  brought scientists
from twenty-nine  countries to Villach,  Austria.   These  scientists  concluded
that "...greenhouse gases  are likely  to  be the  most  important  cause of climate
change over  the  next century."   They went  on  to recommend  that governments
should "...take  into account the  results  of  this  assessment in their policies
on social  and economic development, environmental  programs,  and control of
emissions of radiatively active gases."

     We must  explore  measures  to   prevent  and  mitigate  the  impact  of  the
greenhouse effect.   If there is  to  be  an effective global effort  to address
this major  environmental  problem,   then all  nations  must coordinate their

     To address  this issue,  I have  introduced Senate Concurrent  Resolution
96.  The  bill calls  for  an International Year  on  the Greenhouse Effect  and
would be  the  beginning of  a  long-term  cooperative  effort  by  scientists from
all over the world.

     The legislation  would coordinate and promote domestic and  international
research efforts  on  both   the scientific  and  policy aspects  of this problem,
identify strategies to reduce the increase of  carbon  dioxide  and  trace gases,
investigate  ways   to minimize   the   impact  of   the  greenhouse  effect,  and
establish long-term research plans.

     The concept of an  international  year  of study  has  been used  successfully
in the past.  In  fact,  the best  and  most complete collection  of data based on
atmospheric concentrations of carbon  dioxide began  in  1957  as a result of the
International  Geophysical  Year.    At   that  time   a  sampling  station  was
established on Mona  Loa,  Hawaii.   Today  scientists are still collecting data
from that station.

     I  believe   this  legislation   will  markedly   improve  national   and
international research efforts.   In  time,  it will produce the vital data  that
the  United  States and  other  nations need  to  establish  a  long-term workable
corrective policy.

     Paul Brodeur  pointed  out in the June 9,  1986 issue of the New Yorker  that
we  have known about the   greenhouse effect and  ozone depletion for  over a
decade, yet our efforts to develop corrective solutions have been minimal.  He


quotes Professor  Sherwood  Rowland of the  University of California  at Irvine
who  discovered   the  impact  of   chlorofluorocarbons   on   the  ozone  layer,
"...what's  the  use  of having   developed a  science  well  enough  to  make
predictions, if  in  the  end all we're willing  to do is stand  around and wait
for them to come true?"

     The  greenhouse  effect  is   complicated  and  worrisome,  and  seems  too
overwhelming to  resolve.   But its  potential  impact is far  too  disastrous to
ignore.  The longer we  wait,  the worse the impact will be  because the effect
of the greenhouse gases is  cumulative.   If we  wait much longer to act, it may
be too late.


Atmospheric Ozone

Robert T. Watson
Office of Space Science and Applications
National Aeronautics and Space Administration
Washington, DC USA

     For several  decades scientists  have  sought  to  understand the  complex
interplay among the chemical, radiative, and  dynamical  processes  that govern
the structure of the earth's atmosphere.   During  the  last  decade  or  so there
has been  particular  interest in  studying  the  processes  that  control  atmo-
spheric ozone,  since it  has  been  predicted that human activities  might cause
harmful effects to the  environment  by modifying the  total  column  amount  and
vertical distribution  of atmospheric ozone.   Most of the ozone in the earth's
atmosphere resides in a  region  of the atmosphere  known as  the stratosphere.
The stratosphere  extends from  about 8  km at  the poles  and  17  km at  the
equator, to about 50 km  above the earth's  surface.  Ozone  is  the  only gas in
the atmosphere  that prevents  harmful  solar ultraviolet radiation from reaching
the surface of the earth.  The total amount of ozone  in the atmosphere would,
if compressed to  the  pressure  at the earth's surface, be a layer  about one-
eighth  of  an inch thick.    Figure  1 schematically illustrates the  vertical
distribution of  ozone and  temperature.    Unlike  some  other   more  localized
environmental issues  (e.g.,  acid deposition) ozone  layer  modification is  a
global phenomenon that affects the well-being of  every  country in the world.
Changes  in  the  total  column amount  of atmospheric  ozone  would modify  the
amount  of  biologically  harmful  ultraviolet radiation  penetrating  to  the
earth's surface,  resulting  in potential  adverse  effects  on  human health (skin
cancer  and  suppression   of  the   immune  response system) and  on aquatic  and
terrestrial ecosystems.   Changes  in  the  vertical  distribution of  atmospheric
ozone, along with changes in  the atmospheric concentrations  of other infrared-
active  (greenhouse)  gases,   could  contribute to  a change  in  climate on  a
regional and global  scale by modifying the atmospheric temperature structure,
which  could  lead to  changes  in atmospheric  circulation,  in precipitation
patterns,  and  in sea  level.    (See Hansen  et al;  Titus;  and  Manabe  and
Witherald,  all in this   volume,   for  discussion  of the greenhouse  effect  and
consequences of global warming.)

                      OZONE CONCENTRATION

               10'°         10"          101
     •   80
     <  60
200        300
                                              400      500
Figure  1.   Temperature Profile and Distribution of Ozone
            (molecules per  cubic centimeter)  in the Atmosphere

     The ozone  issue and  the greenhouse warming  issue are  strongly coupled
because ozone itself is a  greenhouse  gas,  and because the same gases that are
predicted to  modify  ozone are also  predicted to produce  a climatic warming.
These gases include carbon monoxide (CO), carbon dioxide (COo), methane (CH^,),
nitrous  oxide   (NpO),  and   several   chlorofluorocarbons  (CFCs),   including
chlorofluorocarbons  11  (CFClo)  and  12  (CFpCl2),  and  are precursors  to  the
hydrogen, nitrogen, and chlorine oxides, which can catalyze the destruction of
ozone in the  stratosphere  by a  series  of  chemical  reactions.  Concentrations
of these gases  in  the  parts  per billion range  control  the abundance of ozone
whose concentration  is in the  parts  per  million  range.   (For  example,  one
molecule of  a chlorofluorocarbon  destroys  thousands of molecules of ozone.)
CO and C02 can affect ozone  indirectly.  CO controls the concentration of the
hydroxyl  radical   (OH)   in   the  troposphere,  which   itself  controls  the
atmospheric concentrations of some of  the  gases that can affect stratospheric
chemistry.   C02  plays  a key  role  in  controlling the temperature structure of
the stratosphere, which is itself  important in  controlling the rates at which
the hydrogen,  nitrogen,  and chlorine oxides destroy ozone.


     There  is   now  compelling  observational   evidence  that  the  chemical
composition of  the atmosphere  is  changing  at  a rapid  rate  and  on  a global
scale.   The  atmospheric concentrations of  C02,  CHh, NpO, and  CFCs  11  and 12
are currently increasing at rates ranging from 0.2% to 5.0% per year  (Figure 2
shows the  trends for four chlorinated  gases,  and Figure  3 shows  the  global
trend for methane).  The concentrations of  other gases  important  in  the ozone
depletion  and global  warming issues  are  also increasing,  some  at  an  even
faster rate.   These changes in atmospheric composition reflect  in  part  the
metabolism of the  biosphere  and in part a broad range of human activities,
including agricultural and combustion  practices.   It should be noted that the
only  known  source of  CFCs  is   industrial  production.   CFCs are used  for  a
variety  of applications,   including aerosol propellants,  refrigerants,  foam
blowing agents, and solvents.  At  present  one of our greatest difficulties in
accurately predicting future changes in ozone concentrations or global climate
is  our  inability  to  predict  the   future  evolution  of  the  atmospheric
concentrations  of   these   gases.    We  need  to understand  the  role of  the
biosphere,  and  the  impact  of  human  activities upon  it,  in  regulating  the
emissions  to  the  atmosphere  of  gases such  as CH^,  C02,  N20,  and  methyl
chloride  (CH^Cl)  and  we   need  to  know the  most  probable future industrial
release rates of gases  such  as  the CFCs, NpO,  CO,  and  COp, which depend upon
economic, social, and political factors.

     One important aspect  of the ozone and  global  warming issues is that the
atmospheric lifetimes of gases such as  NpO,  CFC1,,  and  CF2C12 are known to be
very long.   Consequently,  if there is a change in atmospheric ozone or climate
caused  by  increasing  atmospheric  concentrations  of  these  gases,   the  full
recovery of the  system will  take  decades  or centuries  after the  emission of
these gases into the atmosphere is terminated.


     Numerical models  are  used  as tools  to  predict  to  what extent  human
activities  will  modify  atmospheric  ozone and climate.  The types of models

                CFCI, (S) RAGGED POINT, BARBADOS

            a 200

            2 190

            K 180 I
            a   '
            2 1701

            S 160
                  1978    1979
                                                                1983    1984
           U 350
           S    I
           0 330
           L_    1
           < 310

           0 290

           X 270
                  1978    1979       1980

                                                                1983     1984
7 140
3 130
2 120
1C 110
* 100
S 90



t t I I 1 I 1
i ,, III,!'! 1,1 'I
n I 'I l!l||.IHl!il!Il l! l!il
T I 1 I I ' ' I ' 1 '
M ..lli!!1!! c - !• i '•

1 1 i ' '
i i

S. 140

g 130

< 120

O 110

X 100
                                             111 i I	I' III'''' 11,1'
                  1978     1979
                                                                1983    1984
     Figure  2.    Monthly-mean mixing  ratios  and  monthly  variances  of CFC-11,
CFC-12,  CFUCClo and  CCln measured 4-times-daily on  a gas  chromatograph  at the
Ragged  Point,  Barbados  ALE/GAGE  station during  the  first  six  years  of  the
ALE/GAGE network.

              I        I      I        I
                   I       I       I       I
         1977   1978    1979
1980    1981    1982    1983    1984    1985
Figure  3.   Globally  averaged concentrations of CH^ from  1977 to 1985.

most commonly used  to predict ozone  modification  are known as one-  and two-
dimensional photochemical models.   One-dimensional models predict  changes in
the column  content and  vertical distribution  of  ozone,  but cannot  predict
variations in ozone modification  with latitude,  longitude, or season.   Major
progress has  been made  over the  past few  years  to develop  two-dimensional
models which can predict the variation of ozone change as a function of season
and latitude.

     Three-dimensional models which include  longitudinal  variations are being
developed to study the coupling  between the chemical,  radiative,  and dynamical
processes that control the distribution of ozone, but these models are not yet
ready to perform perturbation calculations.

     Because  it  is now  well recognized  that the chemical  effects  of these
gases on atmospheric  ozone  are  strongly coupled and  should not  be  considered
in  isolation,  the  most  realistic  calculations  of  ozone change  take  into
account the impact  of simultaneous  changes  in the atmospheric  concentrations
of C02,  CHh,  N20,  the CFCs, and  possibly  other  gases,  such as CO,  oxides of
nitrogen (NO ),  and bromine-containing substances.   The effects  of these trace
gases on ozone are not simply additive.   Increased atmospheric  concentrations
of CFCs and NpO are predicted to decrease the column  content of  ozone, whereas
increased atmospheric concentrations of C02  and CHn are  predicted to increase
the column  content  of ozone.  Therefore,  it can be  seen  that the  effects of
increasing  concentrations  of  CFCs  and  NpO are  to  some degree  offset  by
increasing concentrations of COp and CHjj  as shown in Figure  4.    This is in
contrast   to   the    global   warming   issue  where   increased   atmospheric
concentrations of  the  same trace  gases  are  all  predicted  to  increase  the
temperature of  the  atmosphere   in  an  approximately  cumulative  manner  (see
Figure 7).

     One-dimensional  model  calculations have  been performed  to predict  how
ozone would change with time (assuming  that  the atmospheric concentrations of
COp, CHh, and N20  continue  to increase at their current  rates  of 0.5%, 1.0%,
and 0.2% per  year,  respectively,  for  the  next  100 years)  in  conjunction with
three different assumptions  for annual  increase rates in  the  emission of CFCs
11 and  12  to  the atmosphere (i.e., 0.0%,  1.5%,  and  3.0%).  For  CFC emission
increases of up to  1.5% per  year, the ozone  column changes were calculated to
be less  that  3%  over  the next 70 years.   With a CFC growth  rate of 3.0% per
year, the predicted ozone depletion is 10%  for the next  70 years and rapidly
increasing  thereafter (see  Figure  5).    The  results  of these  calculations
demonstrate the  strong chemical  coupling between  these  gases  and  the  time
scale on  which  ozone changes are predicted  to occur.   In essence,  when the
growth rates  of  the  CFCs  are less  than  the  growth rates of CHh  and  COp only
small column ozone changes are predicted because the  CFC efffects on ozone are
temporarily masked.   However, when  the growth rates  of  the CFCs  exceed those
of CHh and  C02,  these gases can no longer offset  the impact  of the CFCs—and
large ozone depletions are predicted.

     Even when the predicted column ozone  changes  are small,  and hence little
change is expected in the amount of ultraviolet radiation reaching the earth's
surface,  major   changes   in  the  vertical  distribution   of  ozone  are  still
predicted, which may  have potential consequences for  climate.   Figure 6 shows
that ozone is predicted to decrease in the middle to upper  stratosphere—

     Figure  4.    Ozone depletion  for particular  changes  in atmospheric  con-
centrations, as estimated  by the Lawrence Livermore model.




i -2
          I  "3
          5  -4





                                                           I      I
                   LLIML 1-D MODEL
                    10    20    30    40    50    60

                                    YEARS FROM PRESENT
                                                     70    80    90   100
     Figure  5,    Calculated  changes  in  ozone  column with  time  for  time-
dependent scenarios:  A  (CFC flux continues at 1980 level,  CH^  increased \% per
yr,  NoO  increases  0.25$  per  yr,  and  CO^  increases  according  to   the  DOE
scenario); B  (CFC emissions  begin at  1980  rates and increase at  1.5$ per yr,
other  trace  gases  change as  with  A);  C  (same  as  B except  CFC  emissions
increase at 3% per yr).

                  E  30

                  t  25
 -40   -30   -20   -10    0     10

                                   20   30
     Figure 6.  Calculated percentage change in  local ozone  at selected  times
5 to 100 years in the future (scenario B of Figure 5).

primarily   because of increasing  concentrations  of CFCs—and  to  increase in
the troposphere primarily because of increasing concentrations of CH^.

     Two-dimensional  models  predict  a  significant  variation  in  the  ozone
column, with  the  greatest depletions  occuring  at high  latitudes.   Depending
upon the  exact  trace-gas scenario used to predict  ozone change, the pole-to-
equator  ratio of  ozone  depletion  can range  from  a  factor  of two  to ten.
Seasonal  effects  are  predicted  but  are  somewhat  less  pronounced  than  the
latitudinal  effects.    For  a  given   trace-gas  scenario  the  two-dimensional
models predict larger column  ozone  depletions  at  mid-  and high-latitudes,  but
less  in  the  tropics,  than  the global average  predicted by  one-dimensional

     One  important  question  that has  been  debated  during the  last couple of
years  is  whether  the magnitude  of  the predicted  ozone change  is  a linear or
non-linear function of atmospheric chlorine concentration.  The issue is still
somewhat  unresolved,  although  it  now  appears  from  most  one-  and  two-
dimensional  model  calculations  that  ozone  change  is a  linear  function  of
atmospheric  chlorine  for atmospheric  concentrations  of chlorine  of 25  parts
per billion  or  less.  Thus,  sudden catastrophic global  ozone depletions  are
unlikely  if  only  small  changes  occur in the  atmospheric  concentrations  of
chlorine.    However,  this  statement  assumes  that  the  models  accurately
represent the real  world,  yet prudence tells us  that  we should remember that
these models  are not perfect and that they  cannot  explain the observed large
changes  that  are   currently   occurring   in  ozone  over  Antarctica  during
springtime, which we discuss below.


     A crucial task for scientists is to assess the extent to which changes in
global ozone have already  taken  place, and  to  compare  the changes to what has
been predicted by theory.  The search for global ozone trends involves looking
for small  secular  changes amidst  the  large  natural variations  that occur on
many time  scales.   Observations of the total  column content and the vertical
distribution  of  ozone have  been made for  several  decades  using  networks of
different measurement techniques.   Unfortunately, each of these observational
techniques has certain limitations, which  tends to  restrict  our confidence in
the  results.    These  limitations  arise  from  factors  such as  the  lack  of
continuity of reliable  calibration and the  uneven  geographic  distribution of
stations.   Statistical  analyses  of the data  are required  to  identify  small
trends  amongst   high natural  variability   using   data  from  relatively  few

     In general, analyses for the trends in the total global column content of
ozone, using data from the ground-based Dobson spectrophotometer network, show
no  statistically  significant  trend since  1970,  which  is in  agreement with
model  predictions  for the same  period when the  changes  in  all of the  trace
gases  are  taken  into account.   It should be  noted that  the  values of total
global column ozone  in  the last three years have exhibited  significant vari-
ability.   Abnormally low values  of total  column ozone  were observed in 1983
following  the eruption  of  El-Chichon and  the  largest  El-Nino   event  this
century.  However,  the values of total column ozone recovered in 1984, only to
decrease significantly in 1985.

     Trend estimates  have also  been  made for  the altitude profile  of ozone
from the network of stations using the Umkehr technique.   Deriving an accurate
trend for changes in the vertical distribution of ozone is more difficult than
for the total column because there are  fewer stations  and the Umkehr measure-
ments  are very  sensitive  to  the  presence  of  aerosols  in the  atmosphere.
Recent volcanic  eruptions,  such  as that  of El-Chichon,  have  deposited large
quantities of aerosols into the stratosphere, and thus  the Umkehr data must be
corrected.  After correcting the derived  ozone amounts for the aerosol inter-
ference, an estimate of  the ozone trend  in  the  middle and upper stratosphere
(30-40 km) gives a 2% to 3% decrease for the period 1970-80.  The magnitude of
the change is broadly consistent  with the predictions of  photochemical models,
which predict that chlorine will  have its maximum effect  at this altitude.

     A  recent  preliminary  analysis  of Nimbus  7 satellite  Solar  Backscatter
Ultra-Violet (SBUV)  data has  indicated that there  has  been  a  statistically
significant change  in  both the  total column  content  (a  decrease)  and the
vertical  distribution  (a decrease  in  the middle  and  upper  stratosphere)  of
atmospheric  ozone  between   1978  and  1984.    Further analysis of   the  data
indicates that most of  the change has  occurred  since  1981.   It  is crucial to
evaluate whether the decrease  is due  to natural  causes,  such as a decrease in
solar radiation,  the  1982 eruption  of  El-Chichon, or  the  1982 El-Nino event,
or whether  it  is due  to human  activities such  as  the  use  of  chlorofluoro-
carbons.  At this time,  any of these explanations are plausible.

     Important new  observational  evidence on ozone changes  has  recently been
obtained.  Data  from  a  single Dobson instrument  at Halley  Bay  76°S,  27°W has
indicated  a  considerable  decrease  (greater  than 40$)  in  the  total  column
content of  ozone above  Antarctica  during the  spring  period  (late  August to
early November) since 1957, with  most of the decrease occurring since the mid-
1970s.  Satellite measurements using both the Nimbus  7 Total Ozone Monitoring
System  (TOMS)   and  the  SBUV  instruments  have  verified  this  trend  over
Antarctica  since   1979   and   have  demonstrated  the  spatial  and  temporal
variations in  this  feature.   Plate  1  shows the decrease  in  the monthly mean
concentrations of the column amount of ozone over Antarctica  in the month of
October from  1979 to 1985.  Similar  ozone changes are  not  observed  over the
Arctic.    Satellite measurements   of   the  vertical   distribution  of  ozone,
nitrogen  dioxide, water  vapor, and aerosols over Antarctica during  the 1985
spring  period  have  been obtained  using  the Stratospheric  Aerosol  and Gas
Experiment (SAGE),  which was  launched  in 1984  on the Earth Radiation Budget
Satellite (ERBS).  This data is now being analyzed and interpreted.   It is not
yet evident whether the behavior  of ozone above Antarctica is an early warning
of future changes in global ozone or whether  it will always be confined to the
Antarctic  because of  the  special  geophysical  conditions  that exist there.
While it has been suggested that  these Antarctic ozone decreases are caused by
increasing  concentrations  of  chlorine   in  the  stratosphere,  no  credible
mechanism  has   been demonstrated—since   the  models  using present  chemical
schemes are unable  to  simulate this effect.  Until  the  processes responsible
for  the  decrease in spring-time Antarctic ozone are  understood,  it will not
be  possible to state with any certainty whether it is- a precursor of a global
trend.    A  number  of  theories,  some  chemical,  others  dynamical,  have been
advanced  to  explain these  observations.   A major field measurement has been
planned to take place in  1986  to study  the ozone layer above Antarctica.  This
campaign  is being cosponsored and coordinated by NASA,   the National  Science

Foundation,  the  National Oceanic  and Atmospheric Administration  (NOAA),  and
the Chemical Manufacturers Associations (CMA).


     There are now observations of increases in the atmospheric concentrations
of the gases predicted  to affect  ozone,  and there are observations indicating
that the total column content of ozone has changed significantly on a regional
and, possibly,  global  scale.   In addition,  there  are indications  that  the
vertical distribution  of ozone  may  also  have changed.   The  question still
remains  concerning  the  reliability  of  the  model  used  to  predict  ozone
change.  Given  that we  cannot  directly  test the accuracy of  a prediction of
the future state of  the  atmosphere,  including  the distribution of atmospheric
ozone,  we  must  test the models  by trying  to  simulate  the  present atmosphere
(including the distribution of atmospheric ozone) or by trying to simulate the
evolution  of the atmosphere,  in  particular ozone,  over the  past  few years.
This is done by comparing model predictions with atmospheric observations.

     We  should  note that nearly  all  the  key chemical  constituents  that  are
predicted  to be present  in  the atmosphere  (and  that are important  in ozone
photochemistry) have  now been observed.   In general,  the models  predict  the
distribution  of   the   chemical  constituents  quite   well.     However,   the
measurements are not adequate for critically  testing the reliability  of  the
photochemical  models.     Close   examination  of   the  intercomparison   of
measurements and model simulations  of the present atmosphere  reveal several
disturbing disagreements.  One  of the major disagreements appears  to be  that
modelled ozone  concentrations are typically 30% to  50% lower  than  measured
ozone concentrations in  the upper  stratosphere—where  it should be easiest to
predict the  concentration of  ozone (Figure 7)  and where chlorine is predicted
to have its maximum effect.   These types of disagreements limit our confidence
in the predictive capability of these models.

     In  the  end,  our  predictive  capability will  be tested by  measuring  the
changes  taking  place in the  atmosphere.   This will  require  careful  measure-
ments  of critical  species  to be  carried  out over  long time  periods (i.e.,
decades).   NASA, NOAA,  and  the  Chemical  Manufacturers  Association  recently
cosponsored a workshop to design an  "Early Detection  of Stratospheric Change"
system.   This  system would  be designed  primarily  to provide the  earliest
possible detection  of  changes in  the chemical and physical structure of  the
stratosphere, and  the means  to understand them.   Implementation of  such  a
system is a high priority.


     As stated earlier, the observed  increase in atmospheric concentrations of
the CFCs,  CH|j,  COp,  and N20  also have direct  implications  for  the earth's
radiative balance through the so-called greenhouse effect.  These gases absorb
infrared radiation  in a  part of  the spectrum  for  which  the  atmosphere is
otherwise  transparent.   Presently,  and in the  near  future,  changes  in  the
concentrations of trace gases other than COp are thought to be contributing to
the greenhouse  warming  of  the  earth's surface  and lower  atmosphere by an
amount that  is about equal to that due to  changes in the concentration of COp
(Figure 8).  The cumulative effect of the  increase in all trace gases for the
period 1850-1980 is a predicted equilibrium warming (ignoring the heat


         0 04
                (PER DECADE)

• • • J + * •

                                                            - lN20
• • • m • • •
      Figure  8.   Decadal  additions to  global mean greenhouse forcing  of the
climate  system.      AT0 is   the  computed  temperature  change  at  equilibrium
(t -MO ) for  the  estimated decadal increases  in trace gas  abundances,  with no
climate feedbacks included.

capacity of  the oceans)  in the   range  of 0.7°-2.0°C,  about  half of  which
should have occurred to date.   Model  calculations indicate that  the greenhouse
warming predicted to occur during the next 50 years should be about twice that
which has  occurred during  the  previous 130 years.   Thus, problems of  ozone
change and climate change should be  considered  together.   It is also apparent
that what  has been  previously  thought of  as  the COp/climate  problem  should
more properly be thought of as  the trace gas/chemistry/climate problem.


     Given  what we  know  about  the  ozone and  trace  gas/chemistry/climate
problem, we should recognize that we are  conducting  a global-scale experiment
on  the  earth's atmosphere without   fully  understanding  the   consequences.
Significant progress has  been  made  in  our  understanding of  the  physical and
chemical  processes  that  control  the  distribution  of  atmospheric  ozone.
However, we  must  recognize  that significant uncertainties  in  our  knowledge
remain,  and  that  these  can only be  resolved  through  a vigorous program  of
research.   It  is  essential that  the U.S.  government and  industry  continue
their  strong  commitment   to  study  the  earth's  atmosphere,  and  that  the
scientific agencies  continue  their close  collaboration at both  the  national
and international level.

Ozone Perturbations Due to Increases

in N2O, CH4, and Chlorocarbons:

Two-Dimensional Time-

Dependent Calculations

Frode Stordal and Ivar SA Isaksen
Institute of Geophysics
Binbern, Oslo, Norway


    Time-dependent  ozone  perturbation  estimates  are  performed  in  a  2-D
diabatic circulation model with ozone photochemistry.   Future increases  in the
emissions  of NpO, CH^,  and the Chlorocarbons  CFCl^,  CFpCl2, CHC1F2,  C2Cl^^,
CHqCClo, and Culj, are considered.   The  altitude variation of  the  calculated
ozone change in  the upper stratosphere is within the uncertainty limits  of the
observed  trend  reported  from  Umkehr  data  for  the  period 1970-80 when  the
effect of  temperature changes from  COp is included.   Future ozone predictions
consider only the photochemical  response to  source  gas  increases.  For  the
integrations through the  year 2030,  four  different  emission scenarios  for
chlorine releases have  been used.   The  average global ozone column  would be
reduced  by 6.5% in  the  year 2030  compared to the  1960  level,  if  chlorine
emissions  are  increased  by 3.0%/yr  and the  concentrations of  N20  and  CHn
increase by 0.25 and 1.0/S/yr, respectively.

    In a  case with  low chlorocarbon emissions and a modest reduction  in the
total ozone  column,  the altitude profile is highly distorted.   In addition,
the latitude gradients are pronounced.   If all the  chlorocarbon emissions are
stopped in the year 2000, a minimum in the total ozone column density  would be
reduced  at different times at  low  and  high latitudes.   Although the average
total global density would decrease for  almost 10 years,  the minimum  would be
reached  after only  5  years at  high latitudes where the  depletion  in total
ozone would be largest.  Accelerated ozone depletion would  occur approximately
when the  total  chlorine exceeds  the  amount of nitrogen in the stratosphere.
This situation would not occur before the year 2030  in any of  the  model runs
with the  scenarios  selected here.   Ozone  predictions  were performed  for an
additional period of 50  years for  the chlorocarbon scenario with the highest
emissions.  In this case, nonlinear growth in ozone  depletion was  projected to
occur after the  middle of the next century.


     Possible modifications of stratospheric ozone have attracted considerable
interest  in  the  last  decade.    Increased concentrations  of  stratospheric
chlorine  from industrial  emissions  of  chlorocarbons, as  well as  increased
levels of nitrogen accompanying  an observed increase  in the  concentration of
nitrous oxide  (^0),  have the potential  to significantly  deplete ozone.   An
increase in  C02  and  other radiatively  active  trace gases  that  lead  to lower
temperatures  in  the  stratosphere  will  have  the  opposite effect  through  a
decrease in ozone destruction.   Enhanced  ozone production  is  believed to take
place in the troposphere due to the observed increase in methane (CHh).

     A number of ozone perturbation experiments have been performed  using two-
dimensional (2-D) computer models  (Brasseur  and Bertion 1978;  Vupputuri 1979;
Borucki et  al.  1980;  Haigh and  Pyle  1982;  Gidel, Crutzen, and  Fishman 1983;
Whitten, Bourucki,  and Woodward 1983;  Haigh  1984;  Solomon,  Garcia,  and Stordal
1985).  In  a recent  study (Isaksen and Stordal 1986)  we used  the  present 2-D
diabatic  circulation  model  for  steady-state  ozone  perturbation  studies.
Perturbations due  to  increases  in CHh  and  NpO,  as well  as  a wide  range of
chlorine levels were  investigated.   The main features  in  the  distribution of
ozone  changes were similar  to  those  obtained by  the models  of Garcia  and
Solomon (1983), Solomon and Garcia (1984),  and Ko  et al.  (1985), as presented
in  the National  Aeronautics  and  Space  Administration/World  Meteorological
Organization  (NASA/WHO) (1986) compilation.   This  is  not surprising  since the
transport  representation   is  similar  in  the  model  (as  discussed   in  NASA
1986).    It  deviates   from  most  2-D  models  using   the  Eulerian  transport
formulation  with  higher  diffusion (e.g.,  Gidel  et  al.  1983)  in  important
ways.   For  instance,  a more pronounced latitudinal  gradient  in the chlorine-
induced reductions of  the total ozone  column  is  predicted when the diabatic
circulation and slow  diffusion  are used for tracer  transport.   All  the above
calculations  are based  on  steady-state  assumptions.   Time-dependent long-time
simulations have so far been confined to one-dimensional (1-D)  models.

     In this  paper, a  set of 2-D  time-dependent  experiments  is described, to
our knowledge the first reported since the experiment of Vupputuri (1979).  He
calculated  ozone  changes  through the  year  2015  under  the  assumption  of
constant releases  of  chlorofluoromethanes at 1975 levels.  Our understanding
of the  photochemistry of the  stratosphere has  increased considerably over the
last   few   years,   leading   to   ozone-perturbation   estimates   that  differ
considerably  from what  Vupputuri obtained.   Besides  offering  a more realistic
tracer  transport  than  1-D models, 2-D  models employ predictions  of future
ozone  variations at  specific  latitudes.   When the  time history of  the future
ozone  distribution is considered,  the difference  in  latitudinal response is a
main   concern  since   the  ozone  depletion  has  been   shown   to    increase
significantly  from lower  to  higher  latitudes.    As  discussed in  Stordal,
Isaksen, and  Horntveth  (1985),  the model gives a  credible description of the
distributions and lifetimes of long-lived gases,  such as Oo, NpO, CHh, CFpClp,
CFC13, and CCljj.

     A  set  of scenarios of the  future  evolution  of some source gases has been
selected for  the present  study.   It  has been necessary to restrict the number
of cases.   Four different scenarios for  the chlorocarbons have been  selected
in  combination with  increases  in NpO  and CHh.    The source  gas  scenarios
adopted are presented  in  the Ozone Loss  Mechanisms  section.   As the  observed


CHh increase  is  not  well understood, it  is  far  from  certain that the present
rate of increase  (1% per year) will  continue  over  the time period considered
in these  calculations.   One of  the experiments has  therefore  been performed
with future CH^  constant at the 1980 level, making a total of five scenarios
that have been  selected  for model simulations for  the time period 1980-2030.
The  reference year  for  the  model  study  is  1960,  when  the  time-dependent
calculations were started.

     The  present model   extends  from  the  surface  to  50  km  altitude,  and
includes  the  photochemistry  of  oxygen,  hydrogen,   nitrogen,  and  chlorine
compounds.  The distribution of ozone-loss rates due to each of these chemical
families  is  important for  the response in  ozone  to  changes in  the  chemical
composition  of  the  stratosphere.    Photochemical  terms  for  the  long-lived
tracers are calculated four times a year  when fully  diurnal computations are
undertaken.    The  photochemical data have  been updated  in accordance  with the
latest NASA/Jet  Propulsion Laboratory  (JPL)  (1985)  recommendations.    In the
discussion on ozone  balance  in the  Ozone Loss  Mechanisms section, latitudinal
as well as  altitudinal  variations in the  influence of each of  the main ozone
loss cycles are demonstrated.

     In the  middle   and  lower  statosphere,  nitrogen  species are  believed to
contribute  largely   to  the  loss  of  ozone.    In  our   previous  study we  also
demonstrated that nitrogen compounds have a strong influence on the efficiency
of ozone  depletion due  to chlorine  increases  (Isaksen  and Stordal 1986).  In
light of  reports that 2-D  diabatic  circulation  models  tend to underestimate
the amount of odd nitrogen present  in  the lower  stratosphere (NASA/WHO  1986;
Ko  et al.  1986), we have  included a discussion of how  well  nitrogen is
distributed in the lower stratosphere in the Ozone Loss Mechanisms section.

     In the section  on  the trend in  ozone for the  period 1970-80, we compare
estimated changes in upper tropospheric  ozone  as compared to  the  observed
decadal  trend  in ozone  in   the  period  1970-80,  as  reported   in  NASA/WHO
(1986).    Estimated   changes  in  ozone  during  the  decade  based  on  Umkehr
measurements at thirteen observational locations were  reported.

     Results  describing   the   future   evolution  of   the  tropospheric  and
stratospheric ozone  are  described  in the section titled Time-Dependent Future
Ozone Experiments.  Figures showing the globally and seasonally averaged total
column, latitudinal  and  yearly  variation of  the ozone  columns, as  well as
altitudinal  distributions are  presented.    For some of  the  scenarios,  the
effects of a  cease of reductions  in  chlorine reductions are demonstrated  (see
the  section  on   the response  in ozone  to  future  regulations   in  chlorine
releases),  which could   offer  useful   information  with  respect  to  future
emission control.

     All ozone predictions presented in this study represent the photochemical
response  to source gas increases  except upper  stratospheric depletion between
1970 and  1980, which includes cooling effects from (X>2.  Feedback from changes
in temperatures and  tracer transport have not been considered.



     Observations show  that  the concentrations of several halocarbons  in the
atmosphere are  increasing.   In an extensive measurement  program (Atmospheric
Lifetime  Experiment  [ALE]),  the  concentrations of  CFClo,  CF2C12,  CCljj,  and
CHoCClo have  been monitored  at five  coastal  or  island  stations  exposed to
little  polluted  air   (Prinn  et  al.   1983a).    A  steady increase in  these
components has been reported at all locations  during the  observational period
     The  Chlorocarbons  included  in  the  present  study  are methyl  chlorine
(CHoCl),  carbon  tetrachloride  (CClh),  methyl chloroform  (CHoCClo),  and  the
chlorofluorocarbons  F11-CFC13,  F12-CF2C12,  F22-CHClFp,   and   Fl13-C2Clo_F3.
These compounds  are  believed to be  the main sources of  future  stratospheric
chlorine.  The source  for  these gases  (except CH^Cl)  is  industrial releases.
As the residence times are of the order of 50-100 years (see Stordal,  Isaksen,
and Horntveth  1985),  except  CHoCClo  which  has a  residence time of  about 10
years, the present increases are due to past as well as present  emissions.

     Release rates  of Chlorocarbons  for  the period  1960-80 are taken from:
Cunnold  et al.  (1983b,  CFClo); Cunnold  et  al.  (1983b, CFpCl2);  Prinn  et al.
(1983b,  CHoCClo);  and  Simmonds et  al.   (1983,  CC1).    The atmosphere  was
assumed  to  initially  (1960)  contain  0.6 ppb  of CHoCl and 0.1   ppb  of CCl^,
resulting in a 1 ppb  in content of  stratospheric chlorine.   The CHoCl surface
flux  needed  to  obtain  the  1960  mixing  ratio  was kept  constant  in all  the
computations.  For CFClo,  CFpClp,  and CHoCClo, 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.

     Scenarios for  future  industrial releases are necessarily uncertain.   We
have  performed experiments with four  different  chlorocarbon scenarios,  listed
in Table  1 .   Our computations are partly based on  the  evaluation of  Quinn et
al.  (1986)  from whom  we have  adopted emissions  for CFClo, CF2C12,  CHClFp,
C2CloFo, CCljj, and CHoCClo in the period  1985-2030.   Based on economic models
that  consider  industrial  usage,   competitive  products,  and   economic  and
population growth, they developed  seven scenarios for the  Chlorocarbons.   In
our  scenario  C we have used emission rates  between  the  values  in their two
highest  scenarios (VI and VII).  As  can be  deduced from Table 2, the releases
of the  largest contributions  to stratospheric  chlorine, CF2C12  and  CFClo, are
projected  to  increase  3.5$  and  5%  per  year,  respectively,  until  the  year
2000.   Thereafter the growth rate declines  and  reaches 2% per year by the end
of the  period  for both gases.   Our  scenario  D  is chosen  to be  identical to
their scenario II.   In this scenario  the reduction  in emission growth starts
earlier  (in the  year  1990).   By the end of the period in question,  the growth
is 1% per year for  CFpClp  and CFClo.   We have also included one scenario with
constant emissions at 198T) levels (scenario A) as well as one with  a steady 3%
yearly   increase  in  emissions   after  1980  (scenario  B).    C2CloF3   is  not
explicitly represented in the model.  The release of this compound  is added to
the  releases  of CFpCl2  (on a Cl  base)  since  the  two  gases  have  similar
photochemical  characteristics (NASA/JPL  1985).   Table 2 demonstrates that the
releases of CpCloFo are modest compared to the CF2C12 releases.


              Table 1.  Source Gas Scenarios 1980-2030

0.25$/yr increase in the period.

Either 1.0%/yr increase or constant surface mixing
ratio at 1980 level.
              Chlorinated source gas emission (10^/yr)
SCENARIO A:  Constant releases 1980 level
SCENARIO B:  A 3$/yr increase in the emissions, starting with the
             values of scenario A in 1980

SCENARIO C:  Emission rates between the values of scenarios VI and
             VII in Quinn et al. (1986)
2F3C13 c
SCENARIO D:  Emission rates as in scenario II in Quinn et al. (1986)






                    Table 2.   Source Gas Scenarios 1960-80
      N20    :      1980 level 305 ppb,  0.2%/yr increase in the period.

      CHjj    :      1980 level 1.70 ppm,  1.1$/yr increase in the period.

      CHoCl  :      1960 level 0.6 ppb,  constant surface boundary flux.

      Emissions for Chlorine Source Gases Taken From:
             :      Cunnold et al.  (1983a)
             :      Cunnold et al.  (1983b)
          . U:      Prinn et al.  (1983b)
      CCI4   :      Simmonds et al.  (1983)
     Chlorocarbons  decompose  in  the stratosphere,  where  chlorine  (Cl)  and
chlorine monoxide  (CIO)  radicals  catalytically destroy ozone.   Reductions of
the stratospheric ozone  layer  is  therefore presumed to accompany  a continued
growth in the chlorocarbon release rates.

Nitrous Oxide

     The atmospheric  concentrations of  nitrous oxide  are increasing  (e.g.,
Weiss  1981;  Khalil and  Rasmussen  1983).   Natural  as  well  as  anthropogenic
sources  have  been  identified.     The  two  main  anthropogenic  sources  are
fertilizers (Crutzen 1974; Logan et  al.  1978)  and  combustion  processes (Weiss
and Craig 1976; Pierotti and Rasmussen 1976).

     Weiss (1981) observed an increase of about 0.2% per year  in the four-year
period  1976-80  and suggested  that this  increase  was  due  to  increasing
anthropogenic releases.  In the present study this growth has  been adopted for
the entire period  1960-80  (Table 2).   Isaksen  and  Stordal (1986)  argued that
the anthropogenic  releases  constitute  only a small fraction  of  the total N20
source at present.   In the future  the anthropogenic  fraction can be expected
to increase  markedly  and also play  a  more significant role  in  the total N20
budget, causing  an accelerated growth in  NpO.   Despite  this speculation, we
take  a more conservative  approach; a yearly growth  of 0.25% is  assumed for
future  NpO  releases in  accordance  with  the recommendation  of Quinn  et al.

     NpO has a long atmospheric  lifetime (about 150  years, see discussion in
Stordal, Isaksen, and Horntveth 1985) as the loss processes in the troposphere
are very slow.   In the  stratosphere, NpO  decomposes,  and is  partly converted
to  oxides  of   nitrogen   (NO )  which  catalytically  destroy   ozone.    Future
increases in nitrous oxide  will therefore  be  associated with  ozone reductions
in the stratosphere.

     The  efficiency of  stratospheric  chlorine in  reducing   the  atmospheric
ozone  layer depends strongly on the level of stratospheric NOX through various
interactions between the NOX and  Clx chemistry.  N20 increases will therefore
also   influence  the  ozone  reduction due  to  the  presence   of  Clx  in the

stratosphere.  The rates at which  chlorocarbons  and  nitrous oxide are assumed
to increase  in  combined scenario  experiments  thus play an  important role in
the estimation of future ozone depletion.


     Atmospheric measurements have  clearly  established  that the concentration
of methane is increasing (Rasmussen and Khalil 1981;  Blake et al. 1982; Khalil
and Rasmussen 1982; Ehhalt, Zander, and Lamontagne 1983).  Since 1979, methane
data  have  been  obtained  from  at  least  six measuring  sites,  including  the
continuous record  from the Cape  Meares Region  (Khalil  and  Rasmussen 1983).
After an  increase  of about 1.8$  per  year  in the  period  1979-82,  a temporary
decline in methane was obtained at  this  station.   This  feature was associated
with an El  Nino Southern  Oscillation  phenomenon  (Khalil and  Rasmussen 1985)
and reduced  the average trend in  the  period 1979-84 to about  1.0% per year.
The  increase in methane  prior to  about 1979  is more  uncertain.    From  the
period after  1965  values  from 0.5%  per  year (Ehhalt, Zander,  and Lamontagne
1983) to  1.7$ per year  (Rasmussen  and Khalil 1981)  have been  reported.   In
this  study  we used  an  annual increase  of   1.1$  for  the period  1960-80  (see
Table 2).

     Our understanding of the  global methane budget is still insufficient.   An
extensive review  of the present  knowledge   is presented in  NASA/WHO (1986).
The total sink  is  estimated  to be approximately  450  TG/year (Isaksen and  Hov
1986), due  to  the  loss through  reaction   with  the   hydroxyl  radical  in  the
troposphere.   Total production of methane should  be approximately the same; it
is commonly  thought  that there is  a balance between  total production and loss
of methane   in  the  atmosphere.    The  relative  strength of each  source  is,
however, highly uncertain  (Ehhalt and Schmidt 1978; Khalil and Rasmussen 1983;
Seller 1984).   The  latitudinal distribution of  these sources  is  also poorly
known.  It is therefore difficult  to  explain the current rise in methane,  and
predicting future  evolution   of  methane concentrations  in   the  atmosphere is
highly uncertain.   We  have chosen here  to  assume  a  future  annual  increase of
1.0$ (Quinn et al.  1986) in most of the scenarios, but we have also included a
scenario with constant level  of CHj, in the  future  (see Table 1).

     Through methane's  interaction  with  tropospheric chemistry,  increases in
emissions will  lead to  increased  ozone concentration  in this  region  of  the
atmosphere.   Methane also  plays an  important role in the partitioning between
the chlorine constituent Cl,  CIO,  and HC1 in the  stratosphere.  Future methane
changes will  therefore  also  affect chlorine species in the stratosphere  and
thereby the depletion of ozone through these species.


     Chemical  loss  of  ozone   in   the  stratosphere  is   characterized by  the
efficiency of  chemical  cycles involving  nitrogen,   hydrogen,   chlorine,  and
oxygen  species.    To  avoid   the  strong   coupling   between  the  individual
components,   particularly among  the oxygen  and nitrogen  species, but  also in
the chlorine  cycles  at  the enhanced levels  expected  to  appear  in  the future,
we introduce an odd oxygen  family  quantity:   Ov = 0Q  + 0 + 0('D) - NO - Cl.
                                              x    j

     The rationale  for  introducing  such a  numerical quantity  is  thoroughly
discussed  by  Hesstvedt and  Isaksen  (1978)  where   it  was applied  to  air


pollution  modeling   to   avoid  stiffness   in   the  oxygen-hydrogen-nitrogen
system.   In our  previous  studies on  stratospheric chemistry  (Stordal 1983;
Stordal, Isaksen, and Horntveth 1985;  Isaksen and  Stordal  1986),  chlorine was
added to the quantity,  strongly  improving the stiffness problem  (and thereby
allowing for longer  time  steps to be used in the  chemical  calculations).   We
have  considered  the  following   loss  reactions  in  the   oxygen,  hydrogen,
nitrogen, and chlorine cycles  (the parenthesized number  to  the left-hand side
of the reaction denotes the number of CL molecules  destroyed in the reaction):

     R1  Oxygen:    (2)   0 + 0^   -»•    02 + 02

     R2  Hydrogen:  (1)  OH + Oo   •*   H02 + 02
     R3             (1) H02 + 0^   *    OH + 02
     R4             (1)  OH + 0   -.     H + 02
     R5             (1) H02 + 0   +    OH + 02

     R6  Nitrogen:  (2) N02 + 0   -»•    NO + 02

     R7  Chlorine:  (2) CIO + 0   *    Cl + 02

     In our previous  paper (Figure 3  in Isaksen and Stordal  1986)  we demon-
strated the strong  variations with height  in  the  relative  importance of the
different  chemical  families.    There   are also  some  marked   latitudinal
variations in  their relative  importances, which are demonstrated  in Figures
1a-d, where the relative loss of  ozone from the  oxygen,  hydrogen,  nitrogen and
chlorine  families  is  given  as  a  function  of latitude  and  height.   Most
important are the latitudinal differences in  ozone  destruction by  the nitrogen
and  the  chlorine  reactions,  particularly because  the concentrations of these
two  families probably will  increase  noticeably in  the  future.   The figures
represent the situation in 1980 when  the chlorine  content  of the  stratosphere
still  was  low  (about  2.3  ppb).    Nevertheless  it  is seen that  chlorine
reactions become  increasingly important  going   toward  high  latitudes,  where
they dominate upper stratospheric ozone loss  in  the present atmosphere (Figure
1d).    Ozone   loss  by  nitrogen  species  is systematically  moved   to  lower
latitudes as we proceed toward high  latitudes.   This will  be of importance,
not only for the  response  in ozone concentrations  to increased levels of NO
or  Clx,  but also  to the  latitudinal  response   in  ozone due  to  simultaneous
changes  in  NOX and  Clx  (combined scenario  studies).   The  efficiency  of the
chlorine cycle  at high  latitudes  with these moderate Clx  levels  explains the
fast, high-latitude response  in  upper stratospheric ozone and in total ozone
to  chlorocarbon  perturbation  that  is  discussed  in this  paper.    It  seems
reasonable in this  early  state of ozone  depletion,  before  chlorine reactions
are  the  dominant  loss  process   at   low latitude,  that the  differences  in
depletions are  largest at high and low latitudes.

     The importance of the NOX chemistry for the ozone  loss in the middle and
lower  troposphere  is  clearly  demonstrated  in  Figure 1c.    In addition,  NO
plays  a  significant  role   for   chlorine-induced  ozone   depletion  through
interaction with- the chlorine chemistry (Isaksen  and  Stordal 1986).   Ozone
perturbation studies  from  chlorocarbon  releases   rely  heavily  on  a proper
representation of NO  levels  in the models used.   Unfortunately,  2-D diabatic
circulation models  tend  to underestimate NO  (total odd nitrogen)  levels in
the  lower  stratosphere   in   equatorial  regions   compared  to  observations
(NASA/WMO 1986).   At  higher  latitudes and altitudes  the agreement is better.

Ko et al.  (1986)  suggest that the discrepancy between  observed  and 2-D model
calculated NO  values can be  removed  by  taking  into account NO  production by
lightning.    This  production  is  efficient  in  the  upper  troposphere  in
equatorial regions, where a significant  amount  (compared  to the  stratospheric
N0y production from NpO) can  be  transported  up  through  the troposphere by the
process of diabatic circulation applied in these 2-D models.

     The NO  distribution in  the actual  area is  very sensitive,  for instance,
to the horizontal eddy diffusion coefficients applied in the model at this low
latitude region in the lower stratosphere.  Our  2-D diabatic circulation model
tends to give NO  levels in better agreement with observations even without a
lightning source.

     In Figure  2 we  have  taken one  of  Ko  et  al.'s figures showing  the NO
profiles at the Equator from  their model along  with observations presented by
Callis, Natarajan, and  Russell   (1985) and  included  the  NOV profile from our
model.  In contrast to the model of Ko et al. (1986), our mcrael underestimates
NOV only at very low  latitudes.   At  the  important 25-km level there is a very
good agreement between our model and  the measurements.   The figure shows that
there is in general a higher NO  level in our model than in the model of Ko et
al.  (1986).    The  discrepancy  between   the models increases  toward  lower
altitudes in the lower stratosphere.   This is at least partly due to a balance
between vertical advection and horizontal diffusion  which is more in favor of
the latter process in our model than  in theirs.


     Trend estimates based on thirteen Umkehr stations indicate that there has
been a statistically  significant  reduction  in ozone  between  1970  and 1980 in
the upper stratosphere between approximately  30  and  40  km.  This reduction is
estimated to  be  on the order of 0.2^-0.3/5  per  year  (NASA  1986),  and  is most
pronounced in the height regions where the  chlorine  chemistry has its strong-
est effect on the ozone (Figure 1d).

     Although the Umkehr data are hampered  by large uncertainties—the result
is particularly  sensitive  to the corrections made for the  aerosol burden in
the  stratosphere—they  nevertheless  represent  important  indications  that
chlorine-induced perturbations can be detected in  the stratosphere.   For this
reason it  is very  important  to compare  the  estimated  ozone trends  based on
observations with model-calculated ozone  depletion for the same period.

     Figure 3 shows height profiles at 40°N for  the decadal depletion from our
2-D model  studies.    Global  trend estimates  reported in  (NASA/WHO 1986) are
also given in the figure.  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 include
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  COp  and other trace species.   We  have  assumed that
approximately one-third of the estimated  change  up to 1983 given by Brasseur,
DeRudder,  and  Trico  (1985)  takes  place  between  1970  and 1980.    This  gives
temperature decreases of approximately  1°C in the 45-50 km  region and less

,  198O
        •is .
        -IB .
        26 .
                                     to .
          i~eo .  -so
                                -20 .   O .    20 .
                                                            60 ,
         2S .
    -2O .   O .    20
                                                      -»O .    8O .    80
     Figure  1.   Percentage  ozone  loss  rates  due  to  0 ,  HOX,  NOX and  C10X
chemical cycles,  relative to  the  total  loss  rates.   Results  are given  as
altitude-latitude contours for near  equinox  conditions.

         as .
         «~i-ao.  —ao.  -to.
                                -20 .   O .    20 ,
                                                    to.    so.   eo
SEflSON=  2
         •49 .
          «-!_eo.   -so,
     Figure  1   (Cont).
following reactions:
-tO.   -2O.   O.    20.    to.   BO.    80.

 The  individual  loss  rates  are  computed from  the

x "



+ Oo
+ Oo
+ 0
+ 0
+ 0
+ 0







      » I • i •ii
                            / .  . .....I    .   .	I
                       '.]              \             10    30
                              MIXING RATIO (ppbv)
     Figure 2.   Comparison  of the calculated NO   profile  at Equator in this
model (curve 1) and the model  of  Ko et  al.  (1986)   (curve 2)  with the observed
nighttime concentration  of  NOP+HNO?  (Callis et al.,  1985;  LIMS-data)  (curve





                 -8     -6      -4      -2       0

                              DECADAL OZONE CHANGES
     Figure  3.    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.

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.

     The calculated  profile  is in,  better  agreement with  the result  of the
trend analysis when temperature effects are considered.   However, almost twice
as much  depletion  in the maximum  area at high latitudes takes  place than at
low latitudes.  Comparisons with observations are therefore strongly dependent
on how  representative the Umkehr  observations are  for mid-latitudes  in the
Northern Hemisphere (40°N).

     The temperature effect from CCU is estimated to be more pronounced at low
than at  high  latitudes,  thus increasing the latitudinal differences  in ozone
depletion when temperature feedback  is  included.   The  reason is that nitrogen
reactions that dominate ozone loss at low  latitudes become  less efficient at
lower temperatures (Isaksen,  Hesstevdt,  and Stordal 1980),  while  this  is not
the case for  chlorine reactions that  dominate at high  latitudes  (see Figure
     Temperature changes will  have  a pronounced effect on  the  trend in total
ozone as  shown  in Table 3.   Global ozone  decrease between  1970  and 1980 is
reduced from  0.62% to  0.22$  when temperature  feedback  is  included.   If the
observed methane increase is included there is even a slight increase (+0.03$)
in  the  total column  of ozone.   At  high  latitudes  (60°N) the  depletion is
pronounced in all three model studies discussed above.
     Table  3.   Time-dependent  2-D model  calculations of  ozone  column
     depletion between 1970 and 1980.   The numbers are for March and are
     given in %.
60° N
         case 1.  Without temperature feedback and constant CHn release.
         case 2.  With temperature feedback and constant CHh release.
         case 3.  With temperature feedback and 2% annual increase in CHh

     The distribution of  ozone at  three  different points of  time,  the years
1980, 2000,  and  2020, will  be used to illustrate  some main  features  in the
development  of  ozone changes.   Figures  4a-4f show the  change in  the total
ozone column  as  a function  of latitude and time  of the year.   The depicted
changes  result   from  a model  run  with  a  chlorine increase  as assumed  in
scenario C.   Simultaneous increases in N^O  and  CHn (see Tables  1 and  2) are
assumed.   The  figures  also  show  latituae-altitude cross  sections for  the
solstices  (Northern  Hemisphere winter) in the same years.  The latitudinal,
altitudinal,  and yearly patterns are similar to  those  obtained in the steady-


state ozone calculations reported earlier with this (Isaksen and Stordal 1986)
and other  2-D  models  (especially  the diabatic  type  models, WASA/WMO  1986):
Ozone reductions take  place  in  the middle and upper  stratosphere  as a result
of  chlorine  chemistry  (Figure   4d).    The  balance   between  active  chlorine
compounds (Cl and CIO), and the  inactive compound HC1  is shifted toward higher
values of  Cl  and CIO.   This is due  to  low CH^ mixing  ratios  resulting from
downward transport (Cl+CHh   -»•    HCl+CHo is a main reaction converting active
chlorine to  HC1,  see  Solomon and Garcia  1984;  Solomon, Garcia,  and Stordal
1985).  The maximum reductions are  15$,  40$,  and 65%  in the years 1980, 2000,
and  2020,   respectively.     In   the  lower  stratosphere  at  low  and  middle
latitudes,  increased  penetration  of UV radiation, resulting  in  the so-called
self-heating effect,   leads  to   increased  levels  of  ozone.    The  growth  in
methane leads  to  increases in ozone  in  the entire troposphere,  with maximum
increases  in  the upper  tropical troposphere.   The  model  predicts  a  steady
increase from 5% to 20$ between  the years 1980 and 2020.

     In the lower stratosphere,  where ozone is mainly  transport-dependent, the
isolines for constant percentage change slope poleward and downward due to the
prevailing  circulation pattern (see discussion in Solomon, Garcia,  and Stordal
1985).  This  transport feature   is  the main cause for  the  marked  latitudinal
gradient in  changes of  the  total  ozone  columns due  to increased  levels  of
chlorine and nitrogen  in  the stratosphere.  Maximum  depletions  occur at high
latitudes during winter and spring when the poleward and downward transport is
most pronounced.

     As Isaksen  and  Stordal  (1986) point  out,   the  latitudinal  gradient  in
total  ozone depletion caused  by  increases  in  stratospheric  chlorine  and
nitrogen is enhanced by the effect of CHj, increases.   However, even if the low
latitude ozone  columns  are only modestly changed, the  height distribution of
ozone is highly  distorted.   These changes are believed to  have the potential
for  large  climatological  effects  (NASA  1986).    In  the  present  (1986)
atmosphere  the  estimated   ozone changes   are  modest,  as  discussed  in  the
section titled  Trend  in  Ozone  for  the  Period  1970-80.   The  situation may,
however, be  quite different  already  at the  turn  of  the century,  when large
depletions  in  the upper  stratosphere  are estimated  to take place.   By that
time, depletion  in ozone  column  densities may  exceed  5% at middle and high
latitudes.    Toward   the  year  2020 ozone  depletion  will  continue to  grow
strongly,  as  a  result of the  pronounced  increase  in  release  rates  of the

     Figure 5  shows the  time-dependent  depletion in  globally  and seasonally
averaged total ozone column densities.  The curves represent the four chlorine
scenarios (see Table 1).  NpO and CHh increases are assumed in all the results
depicted in Figure 5.  An ozone  reduction of 1$ compared to the 1960 level was
reached between 1980  and 1985.   The effect of  temperature changes  is not
included in  the results  presented  in  this and  the forthcoming  sections.   In
the scenario where the chlorine  releases are kept constant at 1980 levels, the
ozone continues  to  deplete to about 2% by 2010.   For the rest of the period,
the global  ozone  is practically unchanged.   In  the  three remaining scenarios
with  increasing  chlorine  releases,  ozone  columns    in  the  entire  period
decrease.   At the year  2020 the ozone  reductions are 8.5$  in  the reference
case, 5.0$  in  the  low scenario,  and 6.5$ in the case with 3$ annual increases
in chlorine emissions.   All  values  refer to average  global reductions.  These
values can be compared to the time-dependent, 1-D study of Wuebbles  (1983)


         O3 PERT
                                 I .  4O . 4 4 . 42 . 4J
Figure  4.   Percentage  ozone
changes in  the  year 1980 (see
scenario  Table   2)  and  the
years  2000  and  2020  in  the
chlorine   emission   reference
case     with     simultaneous
increases in NpO  and CH^ (see
Table 1).  Panels a, c ,  and e
depict  changes  in   the  total
ozone columns as a function of
latitude  and   time   of  the
year.  Panels b, d,  and f show
altitude-latitude        cross
sections    of    local    ozone
changes  for  solstice  condi-
 '-eo.  -eo.  -HO.  -20.
         O3  PERT



    5 30.


                                         \ \

  O3  PERT
  SEOSON=  1

          O3  PERT
       BO .



     ji 20.




             a.  *.
                        . a.
                                             Figure  4 (Cont.)
03  PERT
                   --40 .cr
 : « .
  '-80.  -80.  --»0.  -ZO .

                                 GLOBAL OZONE COLUMN
                       Chlorine Scenarios A, B, C, D
                       N2O and CH4 Increase
                          i  	i	i










     Figure  5.   Calculated  changes in  globally and  seasonally  averaged
     ozone  column as a  function of  time for  the  four chlorine  scenarios
     (Table  1).   Increases  in  N20  (0.25%/yr) and  CH^ (1$/yr) have been
which has been  run  for our  scenario  C  (Wuebbles  1986, private  communication)
with, as  well as without,  temperature changes.   In the latter  case the reduc-
tions are slightly less in his model compared  to ours.  The  time evolution  is,
however,  very similar in the 1-D and 2-D model runs.

     Due  to considerable  uncertainties  in the CHh  scenario  one  experiment  was
performed with  constant CHj,  releases (see  the Source  Gas Scenarios section).
In this  case, ozone  is depleted substantially after  1980.   In  the year 2030,
the reduction in the global  ozone column density reaches 4.5%  (see also Figure
                                 GLOBAL OZONE COLUMN
                         Chlorine Scenario A
                         No Chlorine Releases After 2OOO: A'
                         70%/Yr Red In Chlorine Rel. After 20OO: A"
                         N2O Growth
                         CH4 Constant After 1980
     Figure  6.   As Figure 5.   Also  included are ozone changes  after the
     year  2000 when  either  a  cease or  annual reduction of 10%  per year
     in emissions is introduced.   CHj, is kept  constant after  1980.

     Figure 4  shows a pronounced  latitudinal gradient  in the  ozone changes.
At high latitudes we  also find marked seasonal variations.   To illustrate the
latitudinal  effect  we show  time  variations  of ozone   column densities  at
different  latitudes in the  constant chlorine  release case  (Figure  7).   The
spring values  have  been  chosen  since the  latitudinal gradient  is  largest at
that time  of the year (see  Figure  4).   Even  if the globally  and  seasonally
averaged ozone  column densities show small  changes  toward the  year 2030, the
column  densities are  still   decreasing  markedly  at high  latitudes.    For
instance,  the reduction in the year 2030 is 8.3% at  60°N latitude and 3.8% at
40°N, while the global average reduction  is  only  2.0$.

     The ozone  reductions at 60°M  latitude during   spring  are presented  in
Figure 8 for the  four chlorine scenarios.   In all cases  NpO  and CHh increases
are assumed.   Ozone  reductions  are more than  twice  as large as  the globally
and seasonally  averaged  reductions  shown  in  Figure  5.   Particularly  in the
scenario C case, the  springtime ozone  reduction is  pronounced at  60°N.   At
this latitude  the ozone  column  density decreases at a rate of approximately
0.4$ per  year  in  the  year  2030.   Note  that high  latitude depletion  shows
little dependence on  the  scenarios  chosen  up to year 2000.  The depletion is
pronounced in all four cases,  contrary to what is seen  for the global average
reductions (Figure  7).
                                OZONE COLUMNS SPRING







                 J -7





Chlorine Scenario A
N2O and CH4 Increase











                                     1990    2OOO

     Figure  7.   Changes  in  total  ozone column  with time  for various
     latitudes.   The  results are  for the  spring in  the case  of  1980
     level constant  chlorine  emission and growth in   0 and CH.

                              OZONE COLUMN 60°N SPRING
             o"  ~1
                -16 -
                -18 -
- -10
                     Chlorine Scenarios A, B, C, D
                     N2O and CH4 Increase
                                                              -  -12
                                                              - -16
                 1960   1970    1980   1990   2000    2010    2020   2030
     Figure 8.  As  Figure 5,  but for  the ozone column at 60° during  spring.

     The most  dramatic ozone  reduction takes  place at high  latitudes in  the
upper stratosphere  (see Figure 4).  Figure  9 compares local ozone changes with
time at the 43 km level at  60°N during the  winter for different scenarios.   In
all cases the ozone reduction  is more than  30% by the turn of the century,  and
the depletion continues  towards the year 2030.   The reductions range  from  45%
in scenario A to 12%  in scenario C.

     Based  on  what  has been  shown  in the  previous two  figures,  it can  be
concluded that the  ozone depletion  at high  latitudes in the upper stratosphere
will be pronounced  toward  the  beginning of the  next century regardless which
scenario unfolds.   This  statement applies to  total column ozone  as well  as
local depletion.

     Temperatures are expected to change   in  the  future  as  a result  of  an
altered distribution  of radiatively active  gases.   Predicted ozone  reductions
for the middle and  upper stratosphere will  lead to  temperature reductions,  as
will  COp   and  other   radiatively active  gases  that  are  increasing  in   the
atmosphere  (see,  e.g.,  Ramanathan   et  al.   1985;  NASA/WHO   1986).    These
temperature changes will  influence the ozone  distribution,  as we have seen in
Figure 3,  leading to  smaller ozone reductions  from  increases in stratospheric
chlorine and nitrogen.

                             LOCAL OZONE 6O°N 43 KM WINTER
               I -3°

                       Chlorine Scenarios A, B, C, D
                       N20 and CH4 Increases
                                     1990   2000

     Figure 9.  As Figure 5,  but for local  ozone densities  at 60°  N at
     the 43 km level during winter.

     It is also likely that when the stratospheric temperature  distribution is
altered, changes  in  atmospheric  circulation will  take  place, which  in turn
will influence the calculated ozone distribution.  Ozone  is jointly  controlled
by  dynamical  and photochemical  processes.   Even in  regions where  the ozone
chemistry is fast,  the dynamics  can be of importance  through the  distribution
of  other  long-lived species.  The  works of Fels et al.  (1980) and Schoeberl
and Strobel (1978), however, indicate a  thermal rather than dynamical response
to GO and COp  temperature  perturbations.  Their findings support  the approach
taken  in  this study, where  we have used  fixed dynamics  in  all  the calcula-
tions.  In their study, Fels et al. (1980) assumed uniform perturbations in 0^
and COp, making the application of  their results somewhat uncertain  for the 0^
perturbations presented in this paper.


     The results  presented  in  this and  other studies show  that large ozone
depletions, especially at high latitudes (see  below),  will take place early in
the  next  century  if  the  chlorine  emissions continue  to  grow.    Extended
emission  control  could  then be   required.    In  some  experiments  we  have
therefore studied the effect of reductions in  chlorine releases.  In Figure 10
the effect of  stopping all chlorine releases  is  demonstrated.   In  two of the
cases,  the  constant 1980  level  release case  (scenario A) and the  3% yearly
emission increase  case (scenario B),  the emissions  are  assumed  to  come to a
halt  in the  year  2000.     (However,  NpO  and CHh   are  assumed  to  continue
increasing.)   In  both  cases  the  globally and  seasonally averaged  ozone column
density  continues  to  decrease  for  5-10 years.    The  relatively  efficient
recovery of the ozone layer after chlorine releases  are stopped is a result of
the continued  increase in CHh releases after  the year 2000.

     Figure 6  shows that  in  the case with  a constant CH^  concentration the
recovery of  the ozone  layer  is slower.   Obviously the future  CHh  evolution
will have a pronounced  influence on the recovery of  the ozone  column after a
potential stop  in  chlorine releases.   In  the experiment  described  in Figure
6, constant releases of chlorine source  gases have  been assumed in the period
1980-2000 (scenario A).   If chlorine emissions are stopped  in  the year 2000,
the  ozone  would  not  relax  back to  the  1960 situation,  mainly  due  to NpO
increases that  have been  assumed (also  after the year 2000).   In the period
2020-30  in  this   experiment,  ozone  continues  to   recover  in   the  upper
stratosphere due  to decreases in stratospheric  chlorine.    In  the middle and
lower stratosphere  there  is  a weak ozone  reduction,  mainly  attributed to the
NpO  growth,  compensating  for  the  production  in  the  upper  stratosphere  such
that the  globally  averaged  total   column  is  almost  constant  throughout the

     So  far results  of  a  total   removal  of  chlorine emissions have  been
demonstrated.    A prompt cease in the emissions at  a  certain time, as we have
assumed in the  year 2000,  must be considered unlikely.  A gradual reduction of
the emissions could be more conceivable.   A scenario with a 1% yearly decrease
in CFC emissions starting in  the year 2000 (after constant release from 1980,
scenario A) is  therefore  included in Figure  6 in  the  constant CHj,  case.  This
case is identical to scenario  IX in the  Organization  for  Economic  Cooperation
and Development (OECD)  (1981)  report.   In the particular  case  it  takes about
15  years  to reach a  minimum  in  ozone  column  densities,  and the  increase
thereafter  is   extremely  slow,  in  agreement with  the  1-D calculations  of
Wuebbles  (1983).    Thirty  years  after  the  implementation of  an  emission
reduction,  the  global ozone  depletion  would  be 3.1$,  compared  to  4.5$ in the
case without emission reductions.

     As discussed  in the  previous section   (Figure  7),  the ozone  column  is
projected  to   decay  more   rapidly  at  higher  than  at  lower   latitudes.
Correspondingly, when  chlorine  releases are stopped, the  increase  will  be
faster.   This   is  demonstrated  in  Figure  11,  which  shows  results  from 60°
latitude in the experiments where  chlorine  releases  were halted  in  the year
2000 (scenario  B).   In the scenario A and C cases, the rate of increase is not
only larger  than  the globally averaged  column;  it also  starts faster.   The
ozone minimum is reached  after 5 years, while the  corresponding time for the
globally averaged column  was  shown  to be  10-15 years  (Figure 10).  The local
behavior of ozone  at  60°  northern  latitude  at the  43 km  level  is similar  to
the behavior of total  ozone.   Minimum  ozone is obtained  5-6 years  after the
emissions are ceased,  and  ozone recovery is fast thereafter.   If emissions are
reduced by  1%   per  year instead of being completely  stopped,  the  period  of
decreasing ozone is prolonged to 13 years,  according to our results.


     Since Cicerone, Walters,  and  Liu (1983)  reported  a  pronounced  nonlinear
relationship between ozone reductions and increases in stratospheric chlorine,
this relation has  been  subject to  extensive  studies.   Cicerone, Walters, and
Liu (1983)  found a more rapid growth in ozone reductions than in stratospheric

                                GLOBAL OZONE COLUMN








                       Chlorine Scenarios A, B
                       No Chlorine Releases After 2000: A', B'
                       N2O and CH4 Increase
1990    2000











     Figure  10.   As Figure  5  for the chlorine  scenarios A and  B.  Also
     included  is  the   change  in  ozone  column  after  a  cease  in  the
     chlorine emissions  after  the year 2000.   Growth in NpO and CHh  is
chlorine for Cl  mixing ratios up to about  10 ppb.   Rather,  McElroy,  and Wofsy
(1985); Herman and McQuillan  (1985);  and Isaksen and Stordal  (1986)  found the
relationship to  be  almost linear  in  this Cl  range.   They found, however,  a
faster growth  in ozone  reductions than  in Cl   mixing  ratios from  the point
where  Cl   becomes  larger than  the  amount of  stratospheric  NO  (total  odd
nitrogen).  It was interesting  to  see if this situation  occurred in  the time-
dependent experiments described  in  this  paper.   The values  for Cl  and NO  in
case C, the scenario with the largest chlorine emissions, are shown  in Figure
12.  The Clx concentration increases rapidly after  the  turn  of the century and
reaches 12 ppb in the year 2030, well below the predicted 25  ppb of NO .   The
point  of  accelerating  ozone  reductions  is therefore reached  much later  than
the year 2030 in all the  calculations presented here.   As discussed in Isaksen
and Stordal (1986),  there is still a  considerable uncertainty  in the modeled
NO   (as  well as in NO   measurements).    The  lower the values  of  NO ,  the
earlier the accelerated ozone depletion  will start.  The  NO  results  presented
include an  increase  in  NpO of 0.25$  per year,  a growth  rate  that is somewhat
uncertain.  The  accompanying increase  in NO  values  is much  slower  than the
growth in Cl .   In the 50-year time period  from  1980 to 2030 the stratospheric
level of NO  is projected to  increase 16.5$.
     To  study  the  alterations in  the ozone  chemistry when  Cl   reaches the
level of NO  ,  we have run the  scenario C case for an additional  period of 50
years.  The chlorine release rates for this period  are  also contained in Table
1.    The  yearly  growth  in  emissions  for CF2Clp  and  CFClo  are assured  to
decrease from about 2% in the year 2030 to less  than  1$ 50 years later.  Since

                            OZONE COLUMN 60°N SPRING
                  Chlorine Scenario B
                  No Chlorine Releases After 2000:B
                  Na and CH4 Increase
                                                                 - -10
                                                                 - -12
                                                                 - -18
            1960    1970    1980    1990    2000    2010    2020    203O

     Figure  11.   As Figure 10, but  for the ozone column at  60°  northern
     latitude during  spring.
the predicted increases  in  CH^  and N20 are based on currently  observed trends,
we find  their scenarios  through the  middle  of the next  century to  be highly
uncertain.   We  have therefore  assumed constant surface mixing  ratios for CHh
and NpO  in  the  period 2030-80.   This assumption  is also  convenient  for the
                                  then  will  exceed  the  level  of
current  study since  Cl  levels
With the  adapted scenario, Cl   catches up with  NO
reaches 34 ppb in the year  2080.
                                                                   NO   earlier.
                                                      between 2060 and  2070 and
     Isaksen's  and Stordal's  (1986)  steady-state  experiments showed  that for
high levels of  stratospheric  chlorine,  the most significant ozone  reduction

                            Clx and NOy 37 km Equator





    ~  6



                  Chlorine Scenario C
                  N2O and CH4 Increase






     Figure  12.    Maximum  stratospheric  levels  of  Clx  and  NO     as
     function of  time,  in  the chlorine  scenario C case  (see Table  1)
with simultaneous increases in
                                       and CHh
will be  in  the lower  stratosphere.   Figures  13a-c  illustrate that  the  time-
dependent calculations  confirm  this  behavior.  The  absolute changes  in  ozone
concentration  in  three different   10-year  periods  are  given  as  altitude-
latitude  contour  plots.    In  the  first  period,   1980-90,  the  poleward  and
downward transport will  result  in maximum ozone depletions at high  latitudes
in  the  lower  stratosphere.    At   middle  and low  latitudes,   the  largest
depletions would be  confined to the middle  stratosphere.   Forty  years later,
in the period 2020-30, the depletions in the middle  and low  latitudes would be
intensified.   The  zero line separating the  two regions would descend  2-3 km.
As in  the former period, the maximum decrease in  the upper region  (at middle
and low  latitudes) would be  similar in magnitude  to the  maximum  increase in
the lower region.   This is no longer the case  in the period  2060-70 when the
nonlinear  growth  in  ozone  depletions  at  low  and  middle  latitudes  would
start.   Maximum depletions  would take place  (as  in  Isaksen and Stordal  1986)
in the  25 km  region  where  the zero  line  was  located  80  years earlier.   The
large  projected  ozone  depletions  are  due   to   increase   in   stratospheric
chlorine, which is  shown  to influence  the  loss rate  of  ozone  increasingly,
even in the lower stratosphere.

 O3  RED
              199O   VS    1980
•48 .

•« 4 .
as .
24 .
•-•o.  -.o.  -«o.  -ao. Lfl?fTUDE*°.
                                              eo.   so,
  03  RED
              2O3O   VS   2O2O

 t 1
                                 » ™ V* ^
      ~~'\-i\o '*   .~^*"~   *'      '•nr~:—"'"*•.   .'-""'I	•'•
      :--:::<'"":^'/_	              ^  \ \ \,4.* i.-^o <-
      -:..-• ^/                    KT    \ ^.>N:s:.;>
   >-ao.  —so.  -
                 •to.  -20.   o.    20.
                                        •4O.    SO.
Figure   13-     Changes   in
ozone densities (10   molec
em"-*)  as   a  function   of
altitude  and  latitude   in
the experiment discussed  in
the section on the response
in    ozone    to     future
regulations   in   chlorine
releases.   The changes are
given for the following  10-
year   intervals:   1980-90,
2020-30, and 2060-70.
  O3 RED
               2O7O   VS   2O6O
 •4S .
 SI .

  4B .


  B .

         -ao.   -«o. -20. Lf,?fTUDE*°.    "o-   eo.   ao.

     We now focus on the rather dramatic changes in the chemistry of the lower
stratosphere  that  are  assumed  to  take  place by  the  year  2080 when  the
stratospheric  chlorine  level  becomes  very large.   Other  chlorine reactions
than reaction  R7 below must then be  considered.   Wuebbles  and  Chang (1981)
listed four different chlorine cycles  resulting  in  0   destruction (the number
in brackets to the left-hand side of each equation represents the number of 0
molecules lost  in the reaction, as  described in the  section titled Trend in
Ozone for the Period 1970-80).

             Cycle 1


Cl +
0 +
Cl +
NO +
C10NO +
N03 +

NlL + M •*•
Cl + 02
CIO + 02
2 02

CIO + 02
N02 + 02
C10NO + M
Cl + N03
NO + 02
3 02
     The  dissociation  of NOo  has an alternative  pathway to  reaction  (R12).
Only  the  fraction  of  the  dissociation  forming NO  (1/9, see  NASA/JPL  1985)
leads to 0  destruction.  The net loss term for cycle 2 is then 2/9 J-i-i  ClONOp
since the ClONOp  dissociation  is the rate-limiting  step  (see,  e.g., Wuebbles
and Chan  1981; Cicerone, Walters, and Liu 1983).
Cycle 3
Cl +
N02 +
0 +
Cl +
NO +
2 02

Cycle 4
Cl H
H 0,
K H02
i- hu
h °3

OHC1 +
OH +
H02 +
* 302


     At very high chlorine levels, OHC1 dissociation (R15) becomes faster than
the rate of  the  reaction OH+CU (R2).   Another  reaction is then converting OH
to H02, namely OH+C10, constituting a 5 0  loss cycle:
             Cycle 5


Cl -
OHC1 -
OH -
1- Oo
H H02
K hu
H02 H
i- 0
h 02
h Cl
H Cl
H 02
     Net: 2
     Cycle 1 is the only  pure  chlorine cycle,  while the four remaining cycles
involve nitrogen  (2 and  3)  and  hydrogen  (4 and  5)  compounds.   As  noted in
Wuebbles  and  Chang  (1981),  the  distribution  of  the  ozone loss  between the
individual families is then not obvious.  We have taken the following approach
in this study:

     Cycle 3  has  the 0   loss  reaction (R6) in common  with the pure nitrogen
catalytic cycle:

     N02  +  0   ->•   NO  + Op                    R6
     MO   +  03  -*   N02 + 02                    R17

     We distribute the QX  loss  between the  two cycles so that T17/(T1+T17) is
attributed  to  the  pure  nitrogen chain and the  remaining T13/(T13+T10)  to
chlorine  cycle  (3).   Here Tn represents the rate K XY of  reaction  R  with X
and Y as reactants.

     The rate-limiting process in cycle r is reaction (R16).  The loss rate of
cycle (5) is therefore set to the rate of R16 times two, since two 0  are lost
in each  cycling.   The OHC1  dissociation  (R15)  is common for  cycles  4 and 5.
The  part of  the  rate of  this  reaction   that  exceeds  the rate  of  R16  is
attributed to  cycle (4).   This  cycle has   reaction  R2 common  with  the pure
hydrogen loss chain

     OH  + Oo   -    H02  -f 02               R2
     H02 + 0^   ->-    OH   + 2 02             R3

     Only  the  part  of the  rate  of  R2  that has  not  been involved  in the
chlorine cycles (4) and (5) is attributed to the pure hydrogen caused 0  loss.

     Figures I4a-f show the  fraction of the 0  loss attributed to the various
chlorine cycles in the year  2080.   Altitude-latitude cross sections are given
for the season March-May.   In  the region of chemical control, chlorine cycles
account  for  at least  80$ of the total loss.   The contribution  of  the pure
hydrogen  cycle  is  reduced to  less  than 10%  except around  the stratosphere
where a  10$  influence remains.   The influence of the  pure nitrogen  cycle as
well  as   the  oxygen cycle is  also  less  than  10$.    Chlorine cycle  (1)  is
dominating, even in the lower stratosphere.   At low latitudes, this cycle


           CYCLE  1
  n .
S «»•
  «~»-ao.  -eo.  --»o.  -20.  o.   20.
                                         i.   ao,
CLOX      CYCLE  2    2O8O
                                         SEflSON=  2
  ia .
  •is .
  2S .
  2B ,
  '"'-•O.  -ao .  --4O.  -2O.  O.   2O.   -»O .
                                            >O.   BO.
                                                    Figure 14.   Percentage ozone
                                                    loss  rates   due  to   C10X
                                                    cycles  (see  text,  the  sec-
                                                    tion  on   nonlinear   ozone
                                                    reductions    at   very   high
                                                    chlorine levels).  Altitude-
                                                    latitude contours  are given
                                                    for near equinox conditions
                                                    in  the   year  2080  of  the
                                                    experiment  discussed  in the
                                                    section  on  the  response  in
                                                    ozone to future  regulations
                                                    in chlorine releases.
    CLOX      CYCLE  3    2O8O
                                    SEflSON=  2
    —ao.   —eo.  —^o.  -20.  o.   20.   -40.  BO.   ao.

  CLOX      CYCLE  4     2O8O
  •46 .
  1 4 .
  a4 .
                                            SERSON=  2
   CLOX      CYCLE  5     2O80
JS ««
'"'— *o .  -BO.  — -«o .  — ao .   o.   ao .
                                            SERSON=  2
                                               »o .   BO
                                                                    Figure  14  (Cont.)
CLOX      TOTflL
                                         SEnSON=  2
           eo.  —-»o.  -ao.   o.    ao.   -»o.    eo
                                                   eo .

stands for more than 60% of  the  total  loss at the 25 km level.  In the middle
stratosphere cycle  (3) is  the  second  largest depletion  cycle,  attaining  a
maximum of  more than  30%.   Below  the 25  km level the two  cycles involving
OHC1, cycles (4) and  (5),  gradually increase  their  influence.   Each of these
cycles  contributes  to the  0   loss with  about  1Q%-20%   in the  transition
region.  The chlorine  cycle  (2)  involving  chlorine  nitrate is effective only
at very  low altitudes  at  high  altitudes  where contributions of  1Q%-20% are

     Since the  chemistry of  0   in the year 2080  is  dominated by the chlorine
cycles, as demonstrated in Figures  15a-d, the  explanation  to the accelerated
growth in ozone depletion  must  be  sought in  the chlorine  chemistry.   At all
altitudes above  the 30 km level more  than  half of the ozone has  vanished in
the  year  2080,  leading   to  a  slower   growth  in  ozone  depletions  there.
Accelerated depletion  of the  total column  can  only  take place  due  to the
highly nonlinear increase  in  ozone reductions  in the lower  stratosphere, which
was demonstrated in Figures 13a-c.

     As discussed  in  Isaksen an   Stordal  (1986),  the explanation for  the
accelerated chlorine  influence  can  be found in the  changes in  the  balance
between ozone-depleting components  and  passive  reservoirs  in the  chlorine
chemistry  in  the  lower stratosphere.    The  calculations   show  that HC1 has
effectively been converted mainly to CIO  but  also ClONOp  in  the region in the
period 2030-80.

     The explanation for the  increase in the fraction of CIO must be sought in
changes in components  determining the balance  between the chlorine components.
Figures 15a-d show the  change  in the period 2030-80 in some  key components as
well as in  the  ratio  between CIO and HC1.  The increase in the ratio C10/HC1
is large in the lower  stratosphere.  At the 25 km level at low latitudes this
ratio has  increased by a factor 3-4 in the 50-year period in  question.   The
ratio  can  be  expressed as  a  product  of the  ratios  C10/C1  and C1/HC1.   The
first of these  ratios,  C10/C1, increases by  a  factor  of about 2 since the NO
concentrations  drop  (Figure   15b),  lowering  the  rate  of  reaction  R13.   The
C1/HC1  ratio  increases by a  factor  1.5-2 in  the  lower  stratosphere  at low
latitudes,  since  CHh  decreases  (Figure  15c)   and OH  increases  (Figure  15d).
Both  favor  an  increase in  the  decreases  since  OH  increases and  since the
reaction Cl + CHh   •* HC1 +  CHo  (R18)  becomes a significant loss reaction for
CHn in the upper stratosphere.   OH increases due to lowered levels of HNOo and
HOpNOp  (Prater  et  al.   1985) as  well as  an efficient  transformation from H02
via HOC1 (see detailed discussion in Isaksen and Stordal 1986).

     The time-dependent experiments described here confirm the  point made by
Isaksen and Stordal  (1986)  that  accelerated  ozone  depletion  at high chlorine
levels occurs as a result of  the  fractional  CIO increase.   OHC1  and to some
extent ClONOp contribute to the nonlinearity.

     According   to  the  present  understanding  of   the   chemistry  of  the
stratosphere  (NAS/JPL  1986),  the  transformation  of the active  components Cl
and  CIO  to  the  main reservoir HC1  takes  place through Cl  reactions.  At high
chlorine  levels,  when the  C1-C10 balance  is  shifted   towards   the  latter
component,  the  loss mechanisms  for C1+C10 are less  efficient.   Isaksen and
Stordal (1986) demonstrated,  however, that only small yields  of  reactions like
C10+OH  and  OHCl+hu for pathways leading  to  HC1  formation could decrease the
nonlinear ozone depletion significantly.


                                                           SERSON=  2
         4"I-eo.  —ao.  ~
                                   so .    eo ,
2O8O    VS    2O3O
         •4S .
           « .
           T—eo .  -so.   — -»p
     Figure 15.  Changes in ratios and components in the period 2030-80 in  the
experiment considered  in the  section  on nonlinear  ozone  reductions at very
high  chlorine  levels.   Altitude-latitudes  cross sections  for near  equinox
conditions are given for the quantities log, (HOQQQ/HPQOQ)  where H represents,
firstly, the  ratio  a) C10/HC1, and  secondly,  the mixing ratios of b)  NO,  c)
CH4, and d) OH.

                2O8O   VS   2O3O
SEflSON=  2
24 .
                      .. ..... -o.-s
*~T—eo.  — eo.  -HO.  -20.
                                          •4O.    so.   eo.
                2O8O   VS   2O3O
                                             SERSON=  2
                                          1	1	1	r-
        —eo.  --4O.  -20.
                                        •40 .    60 .   00 ,
                   Figure 15  (Cont.)


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.  Geophys. Res.  Lett. 9:477-80.

Blake,  D.R.,  and F.S.  Rowland.    1985. Worldwide  increases in  tropospheric
     methane,  1978-1983.  Submitted to J. Atm. Chem.

Borucki, W.R., R.C.  Whitten,  H.T.  Woodward, L.A. Capone,  C.A.  Riegel,  and S.
     Gaines.   1980.  Stratospheric ozone decrease  due  to chlorofluoromethane
     photolysis:   predictions of latitude dependence.  J. Atmos.  Sci.  37:686.

Brasseur, G.,  and M. Bertion.  1978.  The action of chlorine on the ozone layer
     as  given by  a  zonally  averaged  two-dimensional  model.    Pure  Appl.
     Geophys.  117:436.

Brasseur, G.,  A.  De Rudder,  and C.  Tricot.  1985  Stratospheric  response to
     chemical  perturbations. J. Atmos. Chem. 3:261-88.

Callis,  J.B.,  M.  Natarajan,  and  J.M.  Russell III.  1985.   Estimates  of  the
     stratospheric distributions of odd nitrogen from the LIMS-data.   Geophys.
     Res. Lett.  12:259-62.

Cicerone, R.J.,  S.  Walters,  and  S.C.  Liu.    1983.    Nonlinear  response of
     stratospheric ozone  column  to   chlorine  injections. J.  Geophys.    Res.

Craig,  H.,  and C.C.  Chou.  1982.  Methane:   the record  in  polar ice  cores.
     Geophys.  Res.  Lett. 9:1221-24.

Crutzen, P.J.  1974.   Estimates of  possible variations in total  ozone  due to
     nature  cause and human activities.  Ambio. 3:201, 210,

Cunnold, D., R. Prinn,  R. Rasmussen, P. Simmonds,  F.  Aleya,  C.  Cardelino, A.
     Crawford, P.  Fraser,  and  R.  Rosen.    1983a.   The  atmospheric  lifetime
     experiment 3:   lifetime  methodology and application  to  3  years  of CFClo
     data.   J. Geophys.   Res. 88:8379-8400.

Cunnold, D., R. Prinn,  R.  Rasmussen,  P.  Simmonds,  F.  Aleya,  C.  Cardelino,  and
     A. Crawford.  1983b.  The  atmospheric lifetime experiment 4:   results  for
     CF2C12  based on 3 years of data.  J. Geophys.  Res. 88:8401-14.

Dickinson,   R.E.,  and   R.J.  Cicerone.   1986.    Future  global  warming  from
     atmospheric trace gases.  Nature. 319:109-15.

Dunkerton,  T., 1978.   On  the mean meridianal  mass  motion of the stratosphere
     and mesosphere.   J. Atmos. Sci.  35:2325-33.

Ehhalt,  D.J.,  and  U.   Schmidt.    1978.    Sources  and  sinks of  atmospheric
     methane.  Pure and Appl Geophys., 116:452.

Ehhalt, D.H.,  R.J.  Zander, and R.A. Lamontagne. 1983.  On the temporal increase
     of tropospheric CH^.   J. Geophys. Res. 88:8442-46.


Pels,  S.B.,  J.D.   Mahlman,  M.D.  Schwarzkopf,  and  R.W.  Sinclair.    1980.
     Stratospheric sensitivity  to  perturbations in ozone and  carbon dioxide:
     radiative and dynamical response.   J.  Atmos.  Sci.  37:2265.

Frederick, J.E., and R.J. Cicerone.  1985.  Dissociation  of  metastable Op as a
     potential  source  of atmospheric odd  oxygen.    Submitted  to  J.  Geophys.

Garcia, R.R., and S. Solomon.   1983. A  numerical model  of the  zonally average
     dynamical and chemical  structure of the middle atmosphere.   J.  Geophys.
     Res. 88:1379-1400.

Gidel,  L.T.,  P.J.  Crutzen,  and   J.   Fishman.     1983.   A  two-dimensional
     photochemical model  for the  atmosphere,   1,  chlorocarbon emissions  and
     their effect on stratospheric ozone. J.  Geophys.  Res.  88:6622.

Groves,  K.S.,   and  A.F.  Tuck.     1980.  Stratospheric  O^-COp   coupling  in  a
     photochemical-radiative column  model.   I:   without chlorine chemistry.
     Quart. J.R. Met.  Soc. 106:125-40.

Haigh,  J.D.  1984.    Radiative  heating in  the  lower   stratosphere  and  the
     distribution of ozone in a two-dimensional model.   Q.J.R.  Meteorol.   Soc.

Haigh,  J.D.,  and  J.   Pyle.  1982.  Ozone  perturbation   in  a  two-dimensional
     circulation model.  Q.J.R.  Meteorol. Soc.  108:551.

Herman,   J.R.,   and   C.J.   McQuillan.   1984.     Atmospheric  chlorine  and
     stratospheric  nonlinearities and   trend  detection.    J.  Geophys.  Res.

Hesstvedt, E., and I.S.A. Isaksen.  1978. Quasi steady-state approximations in
     air pollution modeling:  comparison of  two numerical  schemes for oxidant
     prediction.  Int. Journal Chem.  Kin. X:971-84.

Isaksen,  I.S.A., and  0.  Hov. 1986. Calculation of  trends  in the  tropospheric
     concentration of Cs, OH, CO, CH^,  and NO^, submitted to Tellus.

Isaksen,  I.S.A.,  E.   Hesstvedt,  and  F.   Stordal.    1986.    Influence  of
     stratospheric cooling from COp on the ozone layer.   Nature. 283:189-191.

Isaksen, I.S.A., and F. Stordal.  1986.  Ozone perturbations by  enhanced levels
     of  CFCs,  NpO  and CHj,:   a  two-dimensional  study  including uncertainty
     estimates.  J. Geophys. Res. 91:5249.

Khalil,   M.A.K.,  and R.A.  Rasmussen.   1982.   Secular   trends  of atmospheric
     methane (CH^). Chemosphere. 11:877-83.

Khalil, M.A.K.,  and R.A.  Rasmussen.   1983.  Sources,  sinks and  seasonal cycles
     of atmospheric methane. J. Geophys. Res. 88:5131.

Kida,  H.     1983.   General   circulation  of   air  parcels   and   transport
     characteristics derived from a hemispheric  GCM,  Part  1.   A determination
     of advective  mass  flow in  the  lower stratosphere.  J. Met.  Soc.  Japan.

Ko, M.K.W.,  K.K.  Tung,  O.K. Weisenstein, and N.D.  Sze.  1985.   A  zonal mean
     model  of  stratospheric  tracer  transport  in  isentropic  coordinates:
     numerical simulations  for nitrous  oxide  and  nitric  acid.  J.  Geophys.
     Res. 90:2312.

Ko, M.K.W.,  M.B.  McElroy,  K.K. Weisenstein, and N.D.  Sze.   1986.   Lightning:
     A  possible   source of stratospheric  odd   nitrogen.    To  appear   in  J_._
     Geophys. Res.

Logan, J.A.,  M.J. Rather,  S.C.  Wofsy,  and  M.B. McElroy.    1978.  Atmospheric
     chemistry:    response  to  human influences.   Phil.  Trans.  R. Soc.  London,
     Ser. A. 290:187-234.

McFarland, M., D.  Kley, J.W.  Drummond,  A.L.  Schmeltekopt, and  R.H.  Winkler.
     Nitric oxide measurement  in  the  equatorial  Pacific region. Geophys. Res.
     Lett. 6:605-8.

NASA/JPL.  1985.     Chemical  kinetics  and  photochemical   data  for  use  in
     stratospheric modeling.  JPL publication 85-37.

NASA/WMO.  1986.  Atmospheric ozone:   Assessment  for our understanding  of the
     processes   controlling   its   present   distribution   and  change   (in

OECD.  1981.  Scenarios  for  chlorofluorocarbons,   ENV/CHEM/PJC/81.93.   Paris,
     France: Organization for Economic Cooperation and Development.

Pierotti, D., and R.A.   Rasmussen.   1976.   Combustion as a source  of nitrous
     oxide in the atmosphere.  Geophys. Res. Lett.  3:265.

Prinn,  R.G.,  P.G.  Simmonds,   R.A.  Rasmussen,   R.D. Rosen,  F.N. Aleya,  C.A.
     Cardelino,  J.A.  Crawford, D.M.  Cunnold, P.J.  Fraser, and J.E. Lovelock.
     1983a.     The   atmospheric   lifetime  experiment  1:      introduction,
     instrumentation and overview.  J. Geophys.  Res. 88:8353-68.

Prinn, R.G.,  R.A. Rasmussen,  P.G.  Simmonds, F.N.  Aleya,  D.M.  Cunnold,  B.C.
     Lane,  C.A.   Cardelino, and  A.J.  Crawford.    1983b.    The  atmospheric
     lifetime experiment 5:  results for CHoCClo based on three years of data.
     J. Geophys.  Res. 88:8415-26.

Pyle,  J.A.  1980.     A  calculation  of  the  possible  depletion  of ozone  by
     chlorofluorocarbons using a two-dimensional model.  Pure  Appl.   Geophy.

Quinn, T.H., K.A. Wolf,  W.E. Mooz, J.K.  Hammitt, T.W. Chesnutt, and S.  Sarma.
     1986.  Projected  use,   emissions,  and banks of potential  ozone-depleting
     substances.  WD-2483-1-EPA. Santa Monica:  Rand Corp.

Ramanathan, V.,  R.J.  Cicerone,  H.B. Singh,  and J.T.  Kiehl.   1985.  Trace gas
     trends and  their potential  role  in climate   change.   J.  Geophys.  Res.

Rasmussen, R.A.,  and  M.A.K.  Khalil.  1981.  Atmospheric  methane:   trends and
     seasonal cycles.   JGR. 86:9826-32.

Rather, M.J., M.B.  McElroy,  and S.  Wofsy.  1985.  Reductions  in  ozone at high
     concentrations of stratospheric halogens.  Nature.  312:227-31.

Rowland,  F.S.,   E.W.  Mayer,  D.R.  Blake,  and  Y.  Makide.  1982.  Trends  in
     atmospheric  methane  concentrations since  1978.    Proceedings  of  the
     Symposium on the Composition of the Nonurban Troposphere.   Williamsburg,
     Virginia, May 25-28.

Schoeberl, M.R., and D.F. Strobel. 1978.   The response  of the  zonally averaged
     circulation to stratospheric ozone reductions.   J. fttmos.   Sci. 35:1751.

Seiler, W.  1982.   The cycle  of methane  in  the troposhere.   Presented  at the
     Symposium on the Composition of the Nonurban Troposphere.   Williamsburg,
     Virginia, May 25-28.

Seiler, W. 1984.   Current perspectives  in microbial ecology,  eds.   M.J.  Klug
     and  C.A.  Reddy,  pp.   468-77.  Washington  D.C.:  American  Society  for

Simmonds,  P.G.,  F.N.  Aleya, C.A.  Cardelino,  A.J. Crawford,  D.M.  Cunnold,  B.C.
     Lane,  J.E.  Lovelock,   R.G.  Prinn,  and  R.A.   Rasmussen.    1983.    The
     atmospheric  lifetime experiment  6:   results  for carbon  tetrachloride
     based on three years data.   J.  Geophys. Res. 88:8427-41.

Solomon S., and R.R.  Garcia.  1984.  On  the distributions of  long lived tracers
     and  chlorine  species  in  the middle  atmosphere.    J.  Geophys.  Res.

Solomon, S.,  R.R. Garcia, and F. Stordal.  1985. Transport processes and ozone
     perturbations.  J.  Geophys. Res.,  90:12981-990.

Stordal,  F.  1983.     The  stratospheric  ozone   layer:     numerical   model
     investigations of  human influences.   Ph.D.  thesis, University  of Oslo,

Stordal,  F.,  I.S.A.  Isaksen, and K. Horntveth. 1985.   A  diabatic  circulation
     two-dimensional  model with  photochemistry:   simulations  of  ozone and
     ground released tracers.  J.  Geophys. Res. 90:5757-76.

Tung,  K.K.  1982.   On  the  two-dimensional   transport  of stratospheric   trace
     gases in isentropic coordinates.   J. Atmos. Sci.  39:2330.

Tung,  K.K. 1984.   Modeling of tracer transport  in  the middle atmosphere.  In
     Dynamics of the middle atmosphere, pp.  412-444. Japan:  Terra Scientific.

Vupputuri, R.K.R.  1979.    The  structure of  the natural stratosphere  and  the
     impact of chlorofluoromethanes  on  the ozone layer investigated  in  a  2-D
     time dependent model.  Pure Appl. Geophys. 117:448-85.

Weiss, R.F., and  H.  Craig. 1976.  Production  of  atmospheric  nitrous  oxide by
     combustion.   Geophys. Res. Lett. 3:751-53.

Weiss,  R.F.   1981.  The  temporal  and  spatial distribution  of  tropospheric
     nitrous oxide.  JGR. 86:7185-95.

Whitten, R.C., W.J. Bourucki,  and H.T. Woodward. 1983.  Revised predictions of
     the  effect   on  stratospheric  ozone  of  increasing atmospheric  N^O  and
     chlorofluoromethanes:     a   two-dimensional  model  study.     Atm.   Env.

Wuebbles, D.J., and J.S. Chang.  1981. A study  of the effectiveness of the  Clx
     catalytic ozone loss mechanisms. J. Geophys.  Res. 86:9869

Wuebbles, D.J., 1983.   Chlorocarbon  emission scenarios:   potential impact on
     stratospheric ozone.  J.  Geophys. Res. 88:1433-43.

Wuebbles,  D.J.,   P.M.   Luther,  and   J.E.  Penner.   1983.  Effect  of   coupled
     anthropogenic perturbations  on  stratospheric ozone.   J. Geophys.Res.

The Ultraviolet Radiation  Environment
of the Biosphere

John E. Frederick
Department of the Geophysical Sciences
University of Chicago
Chicago, Illinois USA

     The flux of biologically  significant ultraviolet radiation that reaches
the surface of the earth varies markedly with season and  latitude.  A compari-
son of computed  UV-B (280 to 320  nm)  and UV-A  (320  to 400 nm) surface fluxes
shows  the greatest  latitudinal   and  temporal   gradients  at  the  shorter
wavelengths where  atmospheric  ozone  is an efficient absorber.   Calculations
performed  for four cities in the United States  define  the  respective roles of
the  solar  zenith angle  and  the column  ozone   amount  in  determining  the
biologically  damaging  flux  of solar  radiation  incident  on  the  biosphere.
These  variables,  together  with  surface  elevation,  combine  to  produce
significantly different radiation  environments in  the cities analyzed.


     The amount  of solar  ultraviolet  radiation that reaches the biosphere is
controlled by several factors of which the  best known  is the  amount of atmos-
pheric ozone.  This  paper defines  the major  variables  required to predict the
flux of biologically  relevant solar radiation at  the surface of the earth and
shows how  they combine to produce  large latitudinal and seasonal variations in
the energy received  at the ground.   The sun emits  electromagnetic radiation
over a  broad range  of  wavelengths,  of which  the human eye  responds  to the
region from approximately 400  to  700  nm.  This work focuses on shorter wave-
lengths located  in  the near ultraviolet.   The region from 320  to  400 nm is
termed the UV-A, while the  UV-B lies shortward of this, from 280  to 320 nm.
Absorption of solar  radiation by ozone is significant  primarily at UV-B wave-
lengths .

     One can  think of the atmosphere  as  a filter  located between the sun and
the earth's surface.   The transmission of this filter for solar radiation is a
very  sensitive  function  of  wavelength,  geographic  location,  and time  of
year.   Figure 1 presents a  reference solar spectrum  both at  the  top  of the

atmosphere  and  at  the  earth's surface  for conditions  typical of  afternoon
illumination at a middle latitude site.   The extraterrestrial  flux represents
a combination of  recent  rocket and  satellite measurements.  The  total energy
flux  incident  on the  atmosphere  over the  entire UV-B  and  UV-A  combined  is
approximately 109 watts  per  square  meter.  This is similar to  the power of a
standard 100 watt light  bulb being  directed onto  each  square meter of area at
the top of the atmosphere.

     The lower  curve  in  Figure 1 illustrates  the solar UV-B and  UV-A energy
flux reaching the ground for clear  sky conditions.  The solar  zenith angle in
this  calculation  is 60°.  (The solar zenith angle is the angle  between the
point directly  overhead  and  the  sun.   A  solar zenith  angle of  90° means that
the sun is  on  the horizon.)  The most obvious feature  in Figure  1  is the very
sharp decline in energy  flux as one moves toward  shorter wavelengths from the
UV-A  into the UV-B.   This  is the well-known "ozone cutoff" in  which the flux
at  the  ground  drops  to  negligible  levels  over  a wavelength  range of  a few
nanometers.   For most purposes one  does not  consider radiation  at wavelengths
less  than 295 to  300  nm as reaching  the  surface  of the earth.   However, the
surface flux  in  the  region  of  the  rapid  dropoff is  very  sensitive  to the
atmospheric ozone amount.

     At wavelengths longer  than 320  nm,  absorption by ozone becomes weak.  Yet
Figure 1 shows  that the  radiation flux reaching  the ground is  still much less
than  is incident  at the  top  of the atmosphere.   This  attenuation arises from
scattering by atmospheric molecules,  mainly 0^ an<^ N?*  Some of  the incoming
solar  radiation  is  backscattered  and   exits  the  earth's  atmosphere  into
space.  This constitutes the albedo of the  planet in the ultraviolet.   Hence,
even  in the absence of an  ozone layer the  earth's surface would  be partially
shielded from  the sun's ultraviolet  emissions.    However, in  such  a  circum-
stance  the  lethal  radiation  in the  UV-B would  be vastly  greater than the
values to which life is now adapted.

     Photobiological  responses  are  sensitive  functions  of the wavelength  of
incident light.   This wavelength  dependence is described  mathematically by a
biological  "action  spectrum,"  which  is  simply a  quantitative  way  of saying
that  some  wavelengths are more  efficient than  others at causing  a specific
type of photobiological  response.    Figure 2  is a  sample action  spectrum for a
type of radiation damage to the DMA  molecule (Setlow 1974).  This figure shows
that  biological sensitivity  is  greater   at  the  wavelengths where  the  solar
radiation  reaching  the  ground is  smaller and  vice-versa.    For  biological
applications the  relevant  quantity  is the  product of  an  action  spectrum and
the solar flux  at the ground added over  all wavelengths (mathematically, the
"convolution" of  the  action  spectrum with  the solar  flux).   This provides a
single  number   (rather  than  a  series of numbers at  many wavelengths)  that
depends on  the  biological  process  under  study and varies with  the radiative
transfer properties of the earth's atmosphere.   A combination of the values in
Figures 1  and  2  implies that  radiation  at wavelengths  in the 300  to 315 nm
region  is  overwhelmingly the  most  important  in  damaging the DMA  molecule.
This  is a region where atmospheric ozone exerts a major influence on the solar
flux reaching the ground.

            E  10°
                                  1    '    I    '   I    '    T
                                -EXTRATERRESTRIAL FLUX
                               FLUX AT  GROUND
                               (SZA=60 ,R = O.I5
                                 CLEAR  SKIES)
                  280     300
                    320     340     360
                   WAVELENGTH (nm)
380     400
     Figure  1.   Reference  solar  ultraviolet spectrum  incident on  the top of
the atmosphere (extraterrestrial flux)  and  at the  ground for typical afternoon
conditions and  a middle  latitude  location.  The  calculations assume  a solar
zenith angle (SZA) of 60°,  a ground albedo  (R)  of  15/6,  and clear  skies.

              cr  !0~2
E  '"-4
              E  I

                                    DNA  ACTION SPECTRUM
      280     300    320    340    360
                    WAVELENGTH (nm)
     Figure 2.   Action  spectrum for damage to  the  DMA  molecule.   The vertical
scale  is  a relative  measure of  biological sensitivity  to radiation  at each

The Latitudinal Distribution of Radiation

     Figure  3 presents  the  global  distribution  of  UV-B  radiation  at  the
surface  of the  earth  computed for  clear sky  conditions.   Contours  in  the
figure are  in  watts  per square meter and appear as  functions  of latitude and
month.   All  values  refer to a local time of  10:00 A.M. and  represent  the sum
of all  radiation at wavelengths between  280  and  320 nm.   The  largest energy
fluxes,  four watts  per  square meter,  exist in the tropics because the sun is
almost  directly  overhead,  and  the atmospheric ozone  amounts  are relatively
small.  The large variation in radiation flux with latitude,  especially during
the winter season  is  illustrated  in Figure  3.    In the  Northern Hemisphere
during December and January,  the flux decreases by a  factor of ten between the
equator and 50° latitude.  If one considered a calculation in which the radia-
tion  at each  UV-B  wavelength is  weighted  by an  action  spectrum  value,  the
gradients  in  latitude  would be still more  pronounced  than  those  of  Figure 3
because  shorter  wavelengths,  where ozone absorbs  most efficiently, would be
relatively  more  important.    During  summer  the latitudinal  changes are much
less  pronounced  than  in winter,  and one must move from the  tropics to 60° to
experience a factor of  two decrease  in  radiation  at  the  ground.  Accompanying
this  latitudinal  behavior  are seasonal  variations.   In  the  tropics  there is
very  little  change  in  the  10:00  A.M.  fluxes over the course of  a year.   At
middle  latitudes, however,  the seasonal  cycle  can range between  a factor of
two and  ten depending on location.

      A  calculation  analogous  to  Figure  3 could be done for  the UV-A spectral
region.  Although the contours would be similar  in shape,  the gradients would
be much  less pronounced  because of the  greatly  reduced absorption by ozone at
wavelengths longward of 320  nm.   Figure  4  illustrates this  behavior  with the
contours of the  ratio of the  UV-B  and UV-A  fluxes.  Clearly, the UV-B flux is
much  smaller  than  the  UV-A,  with  the ratios  ranging  between  2   and  7.5
percent.    The most  significant  information  in   Figure  4  is  the differing
latitudinal and  seasonal gradients shown by the UV-B  and  UV-A.   As one moves
from  the tropics to 60° latitude  in winter,  the UV-B  shows  a decrease, which
is  a  factor  of  three  to  four  greater than  the  UV-A,  while  in  summer  the
relative variation is much less than a factor of two.

Radiation  Received by Selected Cities

      For  many medical  applications the relevant quantity  is  the radiation
received  at  a  specific  locality weighted  by the  action  spectrum  for  the
particular  photobiological process under study.   The calculations that follow
use the DNA action  spectrum of Figure  2, and  therefore one  must not view the
results  as  applicable to all  cases of biological  interest.  Figure 5 presents
the annual  cycle of "biologically damaging" flux (flux at the ground convolved
with  the  DNA  action  spectrum)  for  four cities  in the  United States.   The
calculations refer  to local  noon  for  clear  sky conditions, and results appear
in  relative units.   Miami  represents  a low  latitude   site   (25.8°  N),  and
consistent  with  the global  contours of  the  previous  section,  the radiation
flux  here  is  significantly  larger than  that  received by the  more northerly
locations.   Boston  is  the highest  latitude  site in  Figure 5  (42.3°  N).  A
comparison  of  Miami and Boston reveals  a factor of  1.5 difference  in damaging
radiation  during July, and a factor of  8.5 in winter.





      Total  UV-B Flux  at  Ground (W-(vf2)
                                  Ratio UV-B  UV-A (Units: 10
     N  D   J  F  M
 A   M
                         J   J  A  S   0
N  0
M  A  M  J  J  A
S  0
  Figure  3.    The   latitudinal   and
  monthly   distribution    of    UV-B
  radiation at the ground  computed for
  clear  sky  conditions  and  a  local
  time of  10:00 A.M.   Contour values,
  in watts  per  square meter,  include
  all  wave lengths  between  280  and
                         Figure  4.   The ratio  of solar energy
                         flux  in the UV-B  (280-320 nm) to  that
                         in  the  UV-A (320-400 nm) as a function
                         of  latitude and  month.   Values refer
                         to  radiation  reaching the  ground  for
                         10.00  A.M.  local  time and  clear  sky
                         conditions.    Contours are  in percent
                         (7.5 means that the UV-B energy is  7.5
                         of  that in the UV-A).
                                   MIAMI (25.8°N)
                                   WASHINGTON (38.9°N)
                                   DENVER (39.7°N)
                                   BOSTON (42.3° N)
       Figure 5.    Biologically  damaging solar  flux  received  at  the  ground
  computed for four cities  in  the United States during  each month of the year.
  Results refer  to  clear  skies at  local  noon.    The  damaging  flux  is  the
  convolution of  the  solar  flux at  the  ground with  the DNA  action spectrum.
  Values  are in  relative units,  normalized to a maximum of 100.

     Denver (39.7° N) and Washington,  D.C.  (38.9° N) are at similar latitudes,
although  based  solely  on their  0.8° separation  one  would expect  slightly
larger  fluxes   in  Washington  (for  clear  skies and  the  same  local  times).
Figure 5  shows  that this  is  not  the case,  and  in July  the Denver radiation
level  is  approximately  12 percent  greater than  that  in  Washington.   This
offset shrinks  to  9  to  10  percent  in January.   The  elevations of  the two
metropolitan areas  are  the source of  the  computed differences.   The surface
atmospheric  pressure at   Denver  is  approximately  85  percent   of  that  for
Washington, D.C.   This  means  that over Denver  there  is less atmospheric mass
available  to  scatter  incoming  solar  radiation  back  to  space  than  for
elevations closer to sea level.  The  result is  a greater radiation flux at the
ground.   In summer  for  clear  sky  noon  conditions,  a one kilometer increase in
elevation is accompanied by roughly a  7 to 8 percent increase in biologically
damaging  flux.    In winter the  computed   change  is closer  to  6  percent per
kilometer, although the absolute radiation  levels are vastly smaller here than
in summer.

     Based  on   the  annual  cycle  in  solar  zenith  angle,  one  would  predict
maximum damaging  radiation levels  at the summer solstice  (June 21).   Yet,
inspection of Figure  5  reveals that  the peaks lie  closer to mid-July.  (Note
that the  tick marks on  the horizontal scale of Figure  5 refer  to the centers
of each month.) Figures 6 and 7  allow one to  interpret  the  annual  cycles in
radiation.  First,  Figure 6  illustrates  the noon  solar zenith  angles for the
four cities  in  each  month,  where the minimum  values obviously  occur at the
summer solstice.    The  annual cycle  in radiation  flux at the  ground clearly
varies inversely with the solar zenith angle.  This dependence arises from the
varying slant path  taken  by the sunlight through  the absorbing  and scattering
atmosphere during  different  times of the  year.  Both  absorption  by ozone and
scattering  by  atmospheric molecules  act   to  decrease  the  flux  reaching the
ground as the  radiation path  through  the  atmosphere  increases  (i.e., as the
solar zenith angle  increases).

     The  shift  in  time  between minimum solar zenith  angle  and  maximum radia-
tion flux arises from the  annual  cycle in  atmospheric ozone depicted for each
city in  Figure  7.   Note  that  the  total  ozone  in a  vertical  column  of the
atmosphere is not simply related to  the solar zenith angle.  Although ozone is
produced  by  the action of ultraviolet sunlight on Op  molecules  in the upper
atmosphere, much  of the  ozone that  exists  in  a vertical column has drifted
downward  from higher  altitudes.   At altitudes below  thirty  kilometers, where
most of the ozone  resides, the lifetime before  chemical conversion back to $2
is long.  The abundance here  is  controlled by  atmospheric transport processes
that carry ozone from low to  high latitudes during the winter,  leading to the
mid-latitude maxima of  Figure  7 in February and March.  The rapid decrease in
ozone  taking place between 35° and  45°N  during late  spring  and  early summer
effectively  suppresses  a  radiation  maximum  in the  time  immediately before
summer solstice, and  promotes a maximum shortly after  solstice.   This is the
mechanism  that  shifts  the maximum radiation  fluxes  of Figure  5  into July.
Note also that  the behavior  of  total ozone over  Miami in Figure  7  is more
representative  of  a tropical  than a mid-latitude  site.   The relatively small
ozone  amount combined with an annual  behavior different  from  that at higher
latitudes  leads to a seasonal  cycle  in  surface ultraviolet radiation with a
shape distinct  from that at more northerly locations.

         	MIAMI (25.8°N)
         	DENVER (39.7°N)
         	 WASHINGTON (38.9°N)
         	BOSTON (42.3°N)

S 60








     We can  conclude from  this  analysis that:   (a) the  gross shape  of the
annual cycle in biologically damaging radiation at a  given location is deter-
mined by the variation in solar  zenith angle;  (b)  the ozone amount influences
details of  the shape shown by  the  annual  radiation cycle  and is  the major
determinant of the absolute magnitude of the radiation  flux for a given solar
zenith angle;  and  (c)  the  biologically  damaging  radiation flux at latitudes
between 35° and 45°  varies  by  approximately an order  of magnitude from winter
to summer.

     The availability of global scale ozone  measurements and accurate informa-
tion  concerning  the  extraterrestrial solar  irradiance permits  more thorough
studies of  the ultraviolet  radiation environment  of  the biosphere  than have
been possible  in the past.  However, a major deficiency of our results is the
neglect of cloudcover and other forms of  atmospheric  turbidity.  Data sets now
exist that  could provide valuable information  on the reflectivity  and trans-
mission  of  clouds   in  the  near ultraviolet.    This  information   should  be
exploited both for applications  in  atmospheric science and photobiology.   In
addition, it is  essential  that a long-term measurement program be capable of
providing unambiguous evidence of any changes  in  the  global ozone amount.


     Portions  of  this work were supported  by the  National  Aeronautics  and
Space  Administration under  grant  NAGW-873-    The  author thanks   George  N.
Serafino (Applied  Research Corporation)  for  many of the  calculations shown
here  and  Daniel  Eiblum (University  of  Chicago)  for  assistance  in  data


Setlow, R.  B.  1974.   The wavelengths in sunlight  effective in producing skin
     cancer:   A theoretical analysis.  Proceedings of the  National Academy of
     Science. USA 9:3363-3366.

Health Effects of Ultraviolet Radiation
Edward A. Emmett
Johns Hopkins Medical School
School of Hygiene and Public Health
Baltimore, Maryland USA

     People are exposed to  ultraviolet  radiation  from  both natural sunlight
and artificial  sources;  the most  important source for  humans  is  the  sun.
Virtually  all  of us  are  exposed  to  solar  radiation  at  some  time  in  our
lives.   In the United States an estimated eight million outdoor workers have
occupational  exposure  to  solar  radiation.   The  sun  is  a  complex source.
Although we refer to  "sunlight,"  solar  radiation contains not  only visible
rays but also substantial portions of ultraviolet radiation—near, middle, and
far infrared and  other portions of  the electromagnetic  spectrum,   including
cosmic  rays, microwaves, and radiowaves.

     In this  paper I present a general  introduction to the health effects of
ultraviolet radiation  emphasizing photobiologic principles that are necessary
to  understand  the significance of  the  work  presented at  this conference.
Because many  scientists have contributed to  our knowledge  of this  field, my
references are  not exhaustive  and generally  direct readers  to major reviews.
Ultraviolet radiation,  particularly its UV-B component,  has   major effects on
the skin  and  eye.    Other effects,  including systemic  effects  have  been
claimed, but these are not discussed in this  paper.

     For convenience,  we customarily divide  the ultraviolet into three wave-
bands,  UV-A, UV-B, and  UV-C, whose properties are  indicated in Table 1.  This
table is an oversimplification.  The distinction between the wavebands is not
sharp,  and shorter UV-A rays can behave like UV-B.  Interactions between the
various wavebands of ultraviolet radiation  and with  other  wavebands outside
the ultraviolet may occur,  particularly involving ultraviolet radiation and
heat.   However,  this shorthand  notation  is convenient and serves to highlight
UV-B, which is  the focus of Volume 2.

                     Table 1.   Wavelengths of Ultraviolet
          Waveband	Properties
UV-A     320-400 nm                         Not carcinogenic at usual exposure
                                            levels.   Present in solar radi-
                                            ation at earth's surface. Respon-
                                            sible for most photosensitivity
UV-B     290-320 nm                         Usually responsible for sunburn
                                            and skin cancer.   Present in solar
                                            radiation at earth's surface.
UV-C     200-290 nm                         Virtually totally removed by
                                            stratospheric ozone.  Exposures
                                            occur only from artificial
                                            sources.  Can cause erythema and
                                            skin cancer.
     Artificial  sources  of  ultraviolet  radiation  are  varyingly  rich  in
different regions of  the ultraviolet.   The total  population  exposed to these
sources to  any significant extent  is  much smaller  than  for  solar radiation.
The  sources  include  welding  arcs;  germicidal  lamps;  sunlamps;  therapeutic
sources; environmental test chambers;  printing,  plastic,  and  paint curing and
drying processes; ultraviolet  lasers;  plasma  torches; and others.   Relatively
little is known about their long-term effects on exposed human groups.

     Much of  our knowledge about  the  effects of  ultraviolet  radiation comes
from experiments  on cellular  systems, whole  animals,  and a  relatively small
number  of  experiments   in  human  volunteers.    The experimental  ultraviolet
sources used  to develop this  information were  generally monochromators that
deliver exposures to very narrow wavebands of ultraviolet, fluorescent sources
including sunlamps  (UV-C + UV-B + UV-A),  germicidal lamps (predominantly 254
nm  radiation), blacklights  (predominantly UV-A),  and  xenon  arc  solar simu-
lators (generally designed to  simulate solar  radiation  in the UV-B but not in
the visible or infrared).  Most of our information about deoxyribonucleic acid
(DNA) photochemistry has been  generated using germicidal sources.  There is an
important discrepancy  between  solar radiation and the  sources generally used
in  experimental  studies.  Furthermore, the spectral distribution  of sunlight
is not constant;  it constantly varies  within  certain limits,  depending on the
latitude, altitude, time  of year, time  of  day,  and atmospheric  conditions.
Thus our  observations about  the effects  of  ultraviolet  radiation-  in humans
have been collected  mainly  through clinical case  reports and epidemiologic
studies  related  to  exposure  to ambient  solar  radiation  that  has a spectral
distribution  different from that of sources used for experimental  studies.


     The  ultraviolet radiation that  reaches  target cells  in  the body depends
not  only  on the characteristics  of the source and  the radiation reaching  the
surface  of the  body,  but also  on the  optical  properties  of  the  cells  and
tissues overlying  the  biological target.  In  these overlying  tissues ultra-
violet may  be reflected,  scattered,  absorbed,  or  transmitted.

     Figure  1  displays   the  seminal  work  of  Boettner  and  Wolter  (1962)
measuring the total  transmission  of direct and  scattered radiation through  the
tissues of the eye.    The figure shows  that  very little  ultraviolet shorter
than 300 nm penetrates  the cornea,  and  that very  little radiation shorter than
400  nm penetrates  deeper  into  the  eye  than  the lens,  whereas most  of  the
visible radiation  (400-780  nm)  reaches  the  retina.   To a large  extent this
explains  why  the effects of  UV-C  on  the  eye are  virtually confined  to  the
cornea, which explains  why we are concerned about  the  possible effects of  the
longer ultraviolet on the lens.   Interestingly,  the  rabbit eye often used  for
experimental studies  shows transmission similar to the  human eye (Geeraets  and
Berry 1968).
                              TOTAL TRANSMITTANCE AT THE
                              VARIOUS ANTERIOR SURFACES
                            4OO  500 600   800  10OO 12OO
                               WAVELENGTH MILLIMICRONS
                                                    1600 2000
     Figure 1.   Total  Transmittance at the Various Anterior  Surfaces of
     Structures Within the Human Eye  (from Boettner and  Wolter 1986).

     In the  case of  the skin, wavelengths  from approximately  250 nm  in  the
ultraviolet to  approximately 3000  nm in the  infrared  penetrate  the  surface,
but reach  vastly different  depths within  the tissues  (Anderson and  Parrish
1982).   The optical properties of skin are dynamic and  vary  with many factors
including  humidity,   thickness  of  the  skin   layers,  and  the   state   of
vasodilatation.   Epidermal  characteristics  are of  particular  significance,

since  the  cells  at  greatest  risk  for ultraviolet  carcinogenesis  in people
appear  to  be  keratinocytes  in  the  lowest layer  and melanocytes  within the
lower few layers  of  the  epidermis.   An important variable in modulating  indi-
vidual  susceptibility  to ultraviolet  is  melanin pigmentation  which provides
maximum  protection  against wavelengths  shorter than  1200 ran,  including the
entire  ultraviolet  range.   Melanin  is  located primarily  in the  basal  cell
layer  of the  epidermis  in pigment  granules which  can  be  seen  to directly
overlie the nuclei of many basal  cells, like  caps  protecting the nucleus.  It
is very  likely  that  this  supranuclear,  intracellular,  distribution of melanin
results  in  more  protection  against  nuclear  DNA  damage than  would  random
distribution of pigment in the skin.


     The sequence of events  by which  ultraviolet  produces health  effects is
indicated in Table 2.   Ultraviolet radiation must be  absorbed  to produce any
response in  tissues.  This  concept is central to photochemistry  and photo-
biology.  It was first proposed in 1818 by Grotthus and Draper and is known as
"the  first  law  of  photochemistry."   Further,  as   Frederick   (this  volume)
describes,   each  absorbing  molecule  (chromophore)  is  capable  of  absorbing
radiation only in specific wavelength  ranges, a relationship  described by its
absorption spectrum.   Each photon or quantum of radiation absorbed activates
one molecule in the excitation step of a photochemical sequence.
              Table 2.   Sequence by Which Ultraviolet Radiation
                        Produces a Health Effect
                                               Elapsed Time
              Absorbed by Chromophore
              Excited State
              Photochemical Reaction
              Biologic Response(s)
10'9 - 101 sec
10'3 - 1(T7 sec
                                                 sees - years

     The electromagnetic  energy absorbed  in  the chromophore  is  converted to
chemical energy,  resulting in molecular excitation and  an  excited electronic
state.  Some  electronically  excited states, notably  triplet  states,  are much
more  long  lived  than  others,  and  these  are  much more  likely to  result in
photochemical changes.   This electronic energy  may be  dissipated  by heat or
fluorescence, or  in the formation of  photoproducts  either by rearrangement of
the bonds  within the  excited molecule or  by  the interaction  of  the excited
molecule with another molecule.    The excited  molecule  can also  initiate a
photosensitized  reaction,  in which the  molecule that absorbed the radiation
transfers  the energy  to  another  molecule, which  then forms  products (Smith

     Photochemical changes occur in many key biologic molecules and structures
following  ultraviolet  irradiation.    These include  nucleic  acids,  proteins,
lipids, steroids, melanin, urocanic acid,  and others  (Smith  1977;  Harber and
Bickers 1981).   Damage occurs to  the  cell  nucleus,  cytoplasm, organelles, and
membranes.   Some (but  not all) of  the photochemical  changes that immediately
follow  irradiation  are fairly  well  characterized.   However,  the subsequent
biologic changes are less understood.

     Immediate  biologic   changes  after  exposure to  ultraviolet  include  the
release  of   chemical  substances   that  mediate  inflammation,  inflammatory
changes, numerous biochemical interactions involving  reactive oxygen species,
and immunologic  changes  including the suppression of certain immune response
and various  repair  processes.   But  the precise sequence of  these  and other
changes leading  to  the various health effects are  not satisfactorily charac-
terized in  any  instance.   Clearly  for some effects,  such  as carcinogenesis,
these changes take place over many years.


     Lesions  in  DNA,  the  essential  genetic material  in  all  living organisms,
are of  particular interest.    The  purine and pyrimidine  rings  of the nucleic
acids, adenine, guanine,  thymine, and cytosine, absorb maximally in the 245 to
280 nm  range; and the major  part of  skin  absorption  of  UV-C is by DNA rather
than  by  protein or  other  molecules.   Because of the extensive absorption of
ultraviolet by  DNA,  the  role of  DNA damage in  mutagenesis,  and DNA's postu-
lated role in carcinogenesis,  considerable attention  has been given to ultra-
violet-induced  DNA  damage.   Regardless of any  long-term effects, ultraviolet
damage  to  DNA produces substantial short-term effects  including blocking the
nucleic acid  dependent processes of  transcription  and  translation  which are
essential to  cellular function and cell division.  Above critical levels, this
damage is irreversible and therefore leads to cell death.

     Utraviolet-induced  photolesions   in  DNA  have  been  extensively reviewed
(Rahn  1979;  Saito et  al. 1983).   The best  studied  lesions  are  those where
ultraviolet radiation  causes a linkage of adjacent pyrimidines in DNA giving
rise  to  cyclobutane-coupled  pyrimidine  adducts,   often  called  pyrimidine
dimers.   Thymine-thymine, thymine-cytosine,  cytosine-thymine,  and  cytosine-
cytosine dimers are formed,  and  each exists  in four stereo  isomers.   These
lesions are normally repaired in humans and other species described below.  In
addition to  cyclobutane-coupled pyrimidine dimers, ultraviolet induces other
photoproducts  in  DNA.    These  include  other  dimers   (single-bond-coupled-
dipyrimidines,   thyminyldihydrothymidine)   and   nondimer  products  including


thymine glycol, cytosine hydrate,  DMA-DMA  cross-links,  and DMA-protein cross-
links.  Cells are usually able  to  repair at least some of these lesions, with
important exceptions.

     The repair of cyclobutane-coupled pyrimidine dimers in DMA has been given
substantial attention  in  recent years.   This  repair seems to  have  biologic
significance  since   the  reversal  of  pyrimidine  dimers   at  least  partially
restores  colony-forming  ability;  transcriptional capacity;  semiconservative
DNA synthesis; and normal, low, mutagenesis  rates in  irradiated cells.  Three
types of  repair have been  defined:  Excision repair,  postreplication repair,
and photoreactivation; other repair systems may  exist.

     In excision  repair,  a  battery of  enzymes  act in a  sequential manner to
remove the dimer and subsequently  to  resynthesize undamaged DNA.  This repair
system  is well developed  in  humans and  is clearly deficient  in  the  rare
genetic disease xeroderma pigmentosum (XP),  which is  characterized by extreme
sensitivity to  the  effects  of ultraviolet,  including  an  extraordinary sensi-
tivity to ultraviolet-induced cancers.

     The  sequence  of  events  in this  repair has been shown  to  include  the
following:    (a)    pyrimidine  dimer produced   by ultraviolet  distorts  DNA
molecule; (b)   the  backbone  of the damaged  DNA  chain  is  broken by an enzyme,
endonuclease; (c)    a small  region containing the thymine dimer is excised by
another enzyme, exonuclease;  (d)   synthesis of new DNA  strand,  with correct
bases corresponding to those on the  remaining intact  DNA  strand; and   (e)  the
two ends  of  the new strand  are joined by a  further set of enzymes designated
as polynucleotide  ligase.

     Less  is  known  about  the  other forms  of   DNA  repair.   Postreplication
repair allows restoration of DNA synthesis  without dimer removal.   Gaps left
in  daughter  DNA strands are  filled  in  by  de novo synthesis.    Although this
form  of  repair  is   not  well  understood,  it does occur in humans   (Lehmann
1975).  Photoreactivation is a  light-dependent  process that requires  a single
enzyme  and  either UV-A or  visible radiation.   The original  pyrimidines  are
reconstituted  in  situ without  DNA  strand  breakage  or  resynthesis.    Photo-
reactivation appears to occur in humans (Sutherland 1978), although its impor-
tance is not clear.

     Recently  scientists  have  found that  not   all  genes  are  repaired  with
uniform ease  (Bohr et  al. 1985)  and  that  certain chemical lesions (i.e., some
forms  of pyrimidine dimers)  may be  more  resistant to  repair  than others
(Lippke  et  al.  1981).   As  we understand  these processes  better,   we will
probably learn much about how cancer is actually  induced by ultraviolet.

     Humans with XP have a deficiency in DNA repair, clinical sun sensitivity,
and excessive ultraviolet-induced skin cancers (Kraemer 1983).   Cultured cells
from  patients with  XP are  unable to  repair ultraviolet-damaged DNA.   Cell
fusion  studies  have  demonstrated  the existence  of at  least nine types of DNA
repair defects, any  of which may  lead to  the disease.  Despite these  studies,
the  specific  responsible enzyme  defect has not yet  been  identified for  any
form  of XP.   Most  XP patients develop  freckles very early  in  life and on
continued  exposure   to  sunlight develop  abnormally  'aged'  skin,  eye abnor-
malities, and changes  on  the tip  of  the tongue.  These changes are generally
quite marked  before  the  age of 10.   Patients with XP have a greater   than one


thousand  increase   in   frequency  of  skin  cancers   including  basal  cell
carcinomas,  squamous  cell  carcinomas and melanomas.   The predilection of XP
patients to sun-induced  diseases is so  great that if a  disease condition is
seen  in  great excess  in XP,  it is  taken  as prima  facie evidence  that sun
exposure causes the condition.


     For any biological response  to  ultraviolet  radiation,  there is an action
spectrum which  is defined as the relationship  between the  intensity  of the
response  and   the spectrum  of  wavelengths  that  incite  the  reaction.    In
practice, an  action  spectrum  is most  easily  defined  for  acute and early
effects  of  ultraviolet and  is more difficult  to  determine for  chronic and
complex effects.

     Figure  2  shows  the action  spectra curve  developed  by  Cogan  and  Kinsey
(19^6) for  keratitis  (inflammation  of the  cornea  of the eye)  in  the rabbit.
The maximum sensitivity is at  about  290  nm.   The action spectrum for erythema
of  the  skin has peaks at  two wavelengths which are  longer  and  shorter  than
those that produce keratitis.  Neither of these  spectra correspond exactly to
the  absorption  spectra of  either nucleic  acids or  proteins which  are  also
shown in the figure.

                          I    I   II   T   T
                         220    240    260
                       I   I   I   I   T
                      280    300   320
                                   Wavelength (nm)

     Figure 2.   Action Spectra  for Experimental Photokeratitis  in the
     Rabbit Cornea and Erythema of the Skin.  The absorption spectra for
     nuclear and  serum  proteins  are also  shown (from Cogan  and  Kinsey

     One might  think the  action  spectrum to  be identical to  the absorption
spectrum for  the chromophore which  initiates the reaction;  determination of
the action  spectrum  has  enabled us in some  instances  to identify the respon-
sible chromophore  for  photobiologic  reactions.  However,  the action spectrum
for a response in vivo and the in vitro absorption spectrum of the chromophore
may differ  for many  reasons,  including that  the absorption spectrum  of the
chromophore in  vivo  differs from  that  in  vitro because of variations  in pH,
binding on  other phenomena, or that  the effective radiation received  by the
target chromophore in vivo  is modified by  the optical  properties of the over-
lying tissue.

     In addition to action spectra we can also describe biologic effectiveness
spectra.   These may  be used  to address  possible changes  in  the magnitude of
health  effects  of ultraviolet, such  as  those  that  might  occur with  solar
radiation  if  the stratospheric ozone layer  is altered.   A  biological  effec-
tiveness spectrum  is the relationship between the photobiologic response and
the wavelengths  present  in a specific source.   It is  obtained by multiplying
the relevant  points  on  the action  spectrum  curve  (expressed  as reciprocal
dose)  by the  energy  of the radiation of each wavelength present in  the light
source  used.     The  integral  of  the area  under  the  curve represents  the
biologically effective energy of that particular light source.

     Action spectra are generally determined using monochromators to study the
effects  of narrow  ultraviolet  wavebands   (approaching single  wavelengths).
However, interactions between different  wavebands also may occur in  producing
biological  effects.   Interactions  have  been  identified  between  UV-A and UV-B
in the production of erythema and  may occur  between  UV-A and UV-B in carcino-
genesis.    Such interactions  are not   normally  addressed  in   the  classical
available  action  spectra  but  should  be  considered   in  determining  likely
responses to any particular source of ultraviolet radiation.


     We now summarize the  effects  of  ultraviolet radiation on the skin, which
are listed  in Table 3-

Mon-Neoplastic Effects

     The role ultraviolet  radiation  plays  in  Vitamin  D metabolism is a bene-
ficial  effect of  ultraviolet  radiation (Haussler  and  McCain   1977).   Solar
ultraviolet  converts  epidermal 7-dehydrocholesterol  to  previtamin D  which
undergoes a nonphotochemical,  temperature-dependent,  isomerization to Vitamin
D3  which  is   transported  from  the   epidermis  by  plasma  Vitamin  D  binding
protein.   The bone disease rickets,  which is caused  by Vitamin D deficiency,
was common  in recent generations, particularly in those with  heavily pigmented
skin who lived  in  urban  surroundings  and were not exposed to sufficient solar
radiation.  However, now  that  Vitamin D  is added to milk, this  is no longer  a
significant problem in the United States.

     Sunburn  or solar  erythema (Hawk and Parrish  1982)  is  associated with
vasodilatation  and inflammation of the  skin.   A variety of chemical mediators
that  appear  to  be   responsible   for  erythema  are  released  from   the  skin
following  ultraviolet exposure;  the  actual  sequence   of  events  seems  to be
quite complex.   Among the  changes  is an  alteration and  depletion of  Langerhans


cells, the  major  epidermal cell  concerned  with antigen presentation  and the
development of allergic and immune responses in the skin.

     Ultraviolet radiation  produces  two types  of  tanning  reactions (Quevedo,
Fitzpatrick,  and  Jimbow  1985)  in those  genetically able  to  tan:  immediate
tanning due to darkening  and  redistribution  of existing pigment;  and the more
important process of  delayed  tanning,  one or  more  days after irradiation, in
which  there  is  production  of  new  melanin   pigment  and  proliferation  of
           Table 3.  Effects of Ultraviolet Radiation on Human Skin

        Thickening of the Epidermis
        Vitamin D Production

        Actinic Elastosis and Photoageing

        Monmelanoma skin cancer
               Actinic keratosis (precancerous)
               Squamous cell cancer
        Melanoma (?)

     Diseases Triggered by Ultraviolet, or Characterized by
               Xeroderma Pigmentosum
               Albinism, etc.
               Pellagra, etc.
               Herpes Simplex, Lymphogranuloma venereum, etc.
               Lupus erythematosus
               Pemphigus, etc.
               Porphyria cutanea tarda, etc.

     Immediately after sufficient exposure to UV-B reductions in DNA, RNA, and
protein synthesis occur.  Later, beginning within one day, increased synthesis
of these  substances  with thickening and hyperplasia of the  epidermis occurs.
Both this thickening and  the production  of new melanin pigment give a measure
of protection against further ultraviolet damage.

     Ultraviolet radiation  is the major  cause of aging  changes  in the skin,
particularly in those with  fairer-skin colorations.  These occur as a chronic
effect and presumably reflect cumulative sun exposure and individual suscepti-

bility.  These changes include patchy alterations in pigmentation,  loss of the
surface  architecture,  epidermal  atrophy,  telangiectasia  (widening of  small
blood vessels), and the premalignant change of actinic keratoses.   The elastic
fibers of the dermis are damaged so that they become thick and coarse and lose
their  elasticity  resulting in  coarse,  furrowed  skin,  a  condition known  as
actinic elastosis.  The action  spectrum for  actinic elastosis is  undetermined
but it is believed to include  at least ultraviolet.

     Apart from skin  cancers,  a number of skin  diseases  are  either triggered
by ultraviolet  exposure or  are manifested,  at least  in part, by excessive
sensitivity to ultraviolet (photosensitivity) so that sun exposure makes their
symptoms worse.  Examples are  listed  in Table  3,  including hereditary genetic
diseases  such  as  XP  or  albinism;  nutritional deficiency  states  such  as
Kwashiorkor  (protein-calorie   deficiency)  or  pellagra  (niacin  deficiency);
infectious diseases such as herpes  simplex;  immunologically mediated diseases
such as  lupus erythematosus;   altered  metabolic states  including  porphyrias;
and others (Fitzpatrick et  al.  1979).  Although these  diseases are common to
various  degrees,  when skin  cancers are  included,  sun-induced skin  diseases
account for more  than  half of all skin diseases seen by  dermatologists among
fair-skinned persons in areas  with high sun exposure, such as  rural Australia.

Nonmelanoma Skin Cancer

     The skin neoplasms unequivocably associated with sun exposure include the
premalignant  actinic  keratosis,  basal  cell  carcinomas, and  squamous  cell
carcinomas (Emmett 1973; Epstein  1983; Van der  Leun 1984).   The evidence that
such tumors might be associated with  solar radiation includes their increased
prevalence on  parts of the  body habitually  exposed to  sunlight;  in  lightly
pigmented and thereby less protected  individuals; in areas  of the world which
have the greatest insolation;  in individuals  who spend more time outdoors; and
especially in those who suffer  from  xeroderma  pigmentosum.   The highest inci-
dences of such tumors are seen amongst those  of Anglo-Saxon and Celtic descent
in such  areas  as  Queensland,  Australia,   and  Arizona and  Mew Mexico  in the
United States.  Complete statistics are not  available,  but there  is clearly a
relationship  between  latitude  and  incidence of nonmelanoma  skin  cancers  in
Caucasians in the United States.

     Although basal cell and squamous cell carcinomas are often grouped under
the designation nonmelanoma skin  cancer,  they  represent biologically distinct
diseases  that  should be  considered  separately and  indeed  have  different
relationships to ultraviolet radiation.  There  are  at least three reasons for
this distinction.   First,  basal  cell carcinoma and squamous  cell carcinoma
differ  in their  microscopic  and histologic appearance  as  well  as  in their
biologic  behavior,  pattern of growth, and  tendency to metastasize.   Second,
comparisons between  the  sites of occurrence of squamous cell  and basal cell
carcinomas on  the skin and the amounts of  ultraviolet  radiation  these sites
receive  indicate  that  while   squamous  cell  carcinomas almost  all occur  in
heavily  exposed skin, about  one  third  of  basal  cell  carcinomas  in  lightly
pigmented  individuals occur  on customarily shaded sites of the  face.   The
third  indication  of  differences   between   basal   cell   and  squamous  cell
carcinomas  arises out  of  recent  attempts  to  construct mathematical  models
relating  the occurrence of various types of skin cancer to the amount of solar
ultraviolet radiation  received  by the individual.    This  is  a difficult area,
partly because  we  do not have  sufficient  details about exposure  to make good


models.  However, the studies to date suggest that the basal and squamous cell
tumors  have  different  quantitative  relationships  to  sun exposure  (Scotto,
Fears,  and  Fraumeni 1981).   Data  from various centers around  the world also
indicate  that the  ratio  of  squamous cell  to  basal  cell  carcinomas becomes
higher as one approaches the equator.

     Some  individuals  are  much more  prone  than others  to  sun-induced skin
cancers of  all  types.   Phenotypic  characteristics  that are  risk factors for
skin  cancer  in  controlled  studies  include  fair complexion,  light eye color,
light  original  hair color,  poor ability to  tan, and easy and  repeated sun-
burns.  Persons of Celtic (Scottish,  Irish, Welsh) descent may be particularly
vulnerable,  possibly  even  in   the  absence  of  these  phenotypic  character-
istics.   The  risk is extreme in those with  XP.   Limited  studies also suggest
that subjects with  actinic  keratoses  have  less  capacity to repair damaged DMA
than controlled populations.  In contrast to light-skinned people, skin cancer
is relatively rare  in deeply  pigmented groups in  whom basal cell epitheliomas
are  uncommon  while  squamous cell  carcinomas  generally  occur  on the  lower
extremities and do not appear related to ultraviolet radiation.

     Much  information  concerning ultraviolet  carcinogenesis  and  the factors
that affect the  process  has been gained from studies in animals.   It is easy
to produce  fibrosarcomas   in  many  species  from repeated ultraviolet  irradia-
tion.   The hairless  mouse  has proven  a particularly useful experimental model
since,  like humans,  it develops squamous cell tumors.   Basal  cell tumors have
been  produced experimentally in rats  using  chemical  carcinogens, but  not to
date in any species using ultraviolet.

     Experimental  factors  shown to  be  important  in  causing   squamous cell
cancer  include  pigmentation,  presence or  absence  of  hair,  stratum corneum
thickness,  wavelength  of  the  irradiation,  ambient  temperature,  wind,  total
ultraviolet dose, and  the way  the ultraviolet  dose is  fractionated.   Experi-
ments  by  Forbes,  Davies,  and Urbach  (1978)  suggest that delivery of  the same
amount of ultraviolet in  divided doses,  i.e.,  five  exposures  a  week,  was more
effective at  producing  tumors  than when delivered  once a week.   UV-B is the
most  efficient  in producing  cancers;  UV-C  is also effective.   Although UV-A
markedly  accentuates acute skin damage  from UV-B,  conflicting  results exist
about whether UV-A augments UV-B carcinogenesis.

     A  number  of different  types  of  chemicals  can augment  the carcinogenic
properties of UV-B.  These include skin exposure to  chemical carcinogens (such
as  polycyclic aromatic  hydrocarbons, certain  nitrosureas,  and  N  mustard),
promoting  agents  (such  as  croton  oil,  dodecane,  and  possibly  all-trans
retinoic  acid),  some  but  not  all  phototoxic  agents  (for  example,  8
methoxypsoralen but  not  anthracene),  skin irritants, and  immunosuppressives.
Certain agents  have been  shown to inhibit  experimental ultraviolet  carcino-
genesis including  caffeine,  theophylline,  certain antioxidants  and  some sun-
screen agents.  Some of these observations appear to  relate directly to human
experience.   For example,  an  increase in  squamous cell carcinomas  has been
observed  in  patients  having   immunosuppressive  treatment  following  renal
transplantation   and  in   those   who   have   been   treated   for  psoriasis
with  8-methoxypsoralen and UV-A (Stern,  Zierler,  and  Parrish  1980).

     Immunosuppression may play an important role in the development of ultra-
violet-induced tumors.   Kripke and her colleagues  (Kripke  1981,  1982) have

shown that ultraviolet irradiation of mice  results  in  specific susceptibility
to transplantation  with  highly antigenic  ultraviolet-induced cancers.   This
effect does not result from general  loss of immune  competence but rather from
the  production  of  supppressor T  cells with  antigen  specificities  directed
toward  ultraviolet-induced  tumor  antigens.   The  action  spectrum for  this
effect is primarily in the UV-B.

Ultraviolet and Malignant Melanoma

     Melanomas (Lee 1982) are malignant neoplasms that  arise from melanocytes,
the melanin producing  cells  found both within the  skin  and some other sites.
Melanomas of  the  skin  are much less common  than  basal cell and  squamous cell
carcinomas but are  important  because of their tendency  to  metastasize easily
and kill  the  patient.    In contrast  to  all other cancers except lung  cancer,
both the  incidence and to a  lesser extent  death  rates  from melanoma appear to
have been rising  steadily over  the last several  decades  in  fair-skinned popu-
lations worldwide.   This effect  has  been  seen in  England,  Wales,  the United
States, Australia, New Zealand, Scandinavia, Israel, and  elsewhere.  A similar
trend has been seen  in Japan,  but no clear trend exists  in the mortality from
melanoma  of  the  skin among  U.S.  nonwhites.   In  the United States, the mean
annual increase in death  rates  from  malignant  melanoma over the  last 20 years
has been  about 3  percent for males and 2 percent for  females.   This increase
is not  due  to changes in  diagnostic proficiency;  it has occurred despite an
improvement  in  survival  rates  for  those  who  develop  melanomas.    If age-
specific  death rates over a  number  of years  are examined  in terms of birth
cohorts,  we   see  that  the  trends  are  dominated  by  year  of  birth.    This
indicates steadily  increasing risk.   Interestingly, no  such  trends have yet
been detected with regard to melanoma of the eye.

     Although there  is  evidence  that  sunlight  plays  some role  in  melanoma
genesis  in  white-skinned  individuals,  this relationship remains problematic
and  far  from definitely proven.   We lack  an  accepted animal model  in which
ultraviolet or sunlight reproducibly causes melanoma.  Available epidemiologic
studies are  almost all ecologic  and  associate melanomas with some potential
surrogate of  sun  exposure  such  as latitude  of  residence. Although it is often
presumed  that UV-B is the region of solar  radiation  responsible for  causing
melanoma  (because it damages DNA),  this  supposition lacks  proof,  even if we
accept a relationship to sun exposure.

     Perhaps  the  single most  convincing piece  of evidence for a role for UV-B
in the  induction  of melanoma  is  the high  frequency of  melanoma  in XP.  But
even  in this condition  the  repair  defect  is not   confined to  UV-B  induced
photoproducts, so that agents  other  than  UV-B could conceivably account for
this predisposition.

     Melanomas occur in  at least  three  distinguishable histologic forms which
appear  to  have  differing relationships to  solar radiation.   Lentigo  maligna
melanoma, which accounts for about  14  percent of  melanomas,  is slow growing
and occurs characteristically on  the heavily  sun-damaged skin of the face of
older,  light-skinned,  individuals.   Modular melanoma  (32  percent) and  super-
ficial spreading  melanoma (54 percent)  occur mostly on the  trunk and limbs and
are less clearly  related to sun exposure.

     There  are  several  observations  that  possibly  link melanomas  to  sun
exposure, but the evidence  is not  consistent.   Melanoma is much more frequent
in those who are light-skinned and  those  who  sunburn easily.   It is even more
frequent in those who have  certain  types  of pigmented moles called dysplastic
nevi (Kraemer and Greene 1985).  These moles  tend  to run in families and to a
large extent, but perhaps not entirely, the familial tendency  of melanomas is
determined by  the presence  of these lesions.   Melanomas  are more frequent in
those who  sunburn easily  and who  have   had  severe  or  repeated sunburns  in
childhood; however,  in  contrast to  most  nonmelanoma skin  cancers,  melanomas
are not usually  on  exposed  parts of the  body.  Latitude  studies  have  shown a
broad relationship between  melanoma  in Caucasians  and insolation,  but  studies
are  inconsistent.   Migrant  studies  strongly support  a childhood spent  in a
sunny location as  an important  factor in the  later  development  of melanoma.
Melanoma is a disease of higher  socioeconomic classes and is  not seen  as much
in those with the highest outdoor exposure.  It  has, therefore, been suggested
that a  causal factor  for   melanoma  might  be intermittent (i.e., weekend  or
vacation) high  sun  exposure, but  this hypothesis  awaits proof.    Indeed  the
risk of melanomas in individuals seems to  be less  in those with increasing sun
exposure, suggesting that individuals continually  exposed to the sun have less
risk either  because  of  tanning  or  because  susceptibles  purposely  avoid  sun
exposure, or both (Graham et al.  1985).

     In any  case,  if solar  radiation  is  a factor  in causing  human  melanomas
the relationship appears complex, and the distribution  of the  disease  popula-
tions  is  probably  not  directly  determined simply by  total  cumulative  sun


     The  effects of ultraviolet  radiation  on  the eye  (Lerman   1980)  are
affected both  by the amount of  ambient  radiation  incident on  the  surface of
the eye and by differential transmission  through the ocular tissues.  We have
been studying  the amount of radiation that may impact  the corneal  surface in
outdoor workers.   Only  a  small proportion  of  the  ambient solar  ultraviolet
impinges  on  the  eye;  the  amount  varies with season,   activity,  reflecting
surfaces  in  the  vicinity,   and  whether   or  not the subject   wears  a  hat  or
glasses.  Plastic  lenses are particularly  effective in  reducing the  flux of
UV-B at the surface of the  eye.   Although  direct effects of ultraviolet on the
retina are not generally expected  in normal individuals  because the radiation
is absorbed in the lens,  substantial ultraviolet reaches the retina in  aphakic

     The known effects  of   ultraviolet on eye are  indicated in Table  4.   The
best known effect on the conjunctiva and  cornea is acute photokeratitis, most
effectively caused  by  radiation  in  the   288- to  290-nm  wavelength  region in
which necrosis and death of corneal cells  occur.  The condition, known  by such
names as  "welders flash  burns" or  "snow  blindness,"  is characterized  by a
gritty sensation in  the eyes, tearing, photophobia (sensitivity to  light) and
blepharospasm  (spasm and closure of the   eyelids).  The  eyelids are  also often
red.  The symptoms occur after a latent period of  from 30 minutes to 24 hours,
which varies inversely with the  dose.   Symptoms are usually  self-limited and
heal within 48 hours.  Permanent injury is rare.

            Table 4.  Effects of Ultraviolet Radiation on the Eye

             Conjunctiva and Cornea

                 Chronic Actinic Keratopathy
                 Corneal Tumors
                   Benign Pterygia
                   Carcinoma of  the Cornea (very rare)
                 Diseases Triggered by Ultraviolet
                   Herpes Simplex Keratitis
                   Recurrent Corneal Erosions,  etc.


                 Nuclear Cataract


                 Retinal Damage  (aphakics, others?)
                   Ocular Melanoma (?)
     In chronic  actinic keratopathy there  are  white or yellow-brown nodular
opacities confined to the sites of the  cornea covered by the lids.   These have
a restricted geographic  location  and  are confined to areas  of  the world with
high ultraviolet exposure in the eye (e.g.,  "Labrador keratopathy").

     Ultraviolet radiation appears to  be a major factor  in developing pterygia
and pingueculae,  benign  tumors of the conjunctivae  and  cornea,  respectively.
It  can  also trigger  the development  of certain  corneal diseases  including
herpes simplex and recurrent corneal erosions.

     The  effect  of  ultraviolet  on the human  lens  is  of great  interest.
Cataracts can be  experimentally produced  in  rabbits  by  ultraviolet radiation,
which appears to  exert  an effect on the lens, at least  in part,  by acting on
protein-bound  tryptophan residues.    Nuclear  cataracts  appear   to be  more
frequent  in  human populations where   increased  ultraviolet exposure occurs,
although studies  that relate  this  exposure  to individual ultraviolet exposure
are very  few  (Pitts  et  al.  1986).  With age  there is an increase  in pigments
formed in the lens which  may  both  contribute to nuclear cataracts and serve a
photoprotective   function  in  screening  deeper  tissues   in   the  eye  from

     The retina  is  particularly sensitive to damage from  shorter  wavelengths
of ultraviolet.   There  is no  uniform  agreement  regarding  the  significance of
this  observation  for normal  people,   but  in aphakic individuals  substantial
opportunity for retinal  damage seems to exist.  Two recent studies have raised
the possibility that ocular melanoma,  a  relatively rare but serious neoplasm,
might be  related  to sun  exposure  (Tucker et al.  1985).  If  this finding is
correct,  it remains  speculative  whether  the  ultraviolet  portion  of solar
radiation is responsible.



     A  number of  foreign substances  act  as  photosensitizers  in  that  they
absorb  ultraviolet  (or visible)  radiation  and augment  or change  the tissue
response to this radiation (Emmett 1979).  Well-known photosensitizers include
several  components  of  coal  tars and  pitches; chemicals  produced  by plants
including psoralens;  a variety of dyes,  drugs,  and antibacterials  in common
use;  and   photoinitiators   used  for   industrial   photochemical  processes.
Potential photosensitizers are increasingly common components of products used
in modern life.

     The most common  reactions  are  designated as  "phototoxic" and  occur  as
direct  toxic  effect  from  the  chemical  and  ultraviolet radiation  without
mediation of  the  immunologic  system.   A -variety  of reactions  are produced on
the  skin,  including severe reactions  resembling  sunburn  and  severe blister-
ing.   In the  eyes,  photokeratoconjunctivitis  and  cataracts may  result.   As
mentioned  previously,  some   photosensitizers,  notably  8-methoxypsoralen  in
combination with UV-A, have been shown to be carcinogenic both in experimental
animals and humans.

     More rarely,  individuals can develop photoallergy,  an allergy of the skin
dependent on  exposure  to  both  the chemical  and ultraviolet radiation.  Photo-
allergic reactions are not known to result in cancer but  a so-called persis-
tent  light   reaction   can  ensue which  will persist for  years with  trivial
amounts  of  sun exposure,  even  in  the  absence  of  further  contact  with  the
inciting chemical.


     We have  reviewed  the various effects of ultraviolet on human health.   In
considering  the  health  effects of  ozone  modification,  we are  particularly
concerned with the  chronic effects of  ultraviolet radiation.   These are  not
determined  solely by  the amount  of ultraviolet  radiation but also  by  other
factors.   The  three  major  factors involved  are  exposure  to ambient  solar
radiation,  individual  susceptibility,  and personal behavior.   From the  point
of view of human health we cannot sensibly address the issues involved without
considering all three factors.

     Future studies  of chronic effects  will advance the  field to the extent
that they gather information  on individual susceptibility and behavior, and in
the  case  of  studies  of tumors, by  relating  to   or  distinguishing individual
tumor types.  Future risk assessments will only make medical or biologic sense
to the extent that they take  these factors into account.


Anderson, R.R., and J.A.  Parrish. 1982.   Optical  properties of human skin.  In
    The science of  photomedicine, ed.  J.D.  Regan and J.A.  Parrish. New York:
    Plenum Press.

Boettner, E.A.,  and  J.R.  Wolter.   1962.    Transmission  of the ocular media.
    Invest.  Ophth. 1:776-83.

Bohr, V.A., C.A. Smith,   D.S.  Okumoto,  and O.C.  Hanawalt.  1985.   DNA repair in
    an active gene.  Cell.  40:359-69.

Cogan, D.G., and V.E. Kinsey. 1946.  Action  spectrum  of keratitis produced by
    ultraviolet radiation.  Arch.  Ophthalmol.  35:670-77.

Emmett, E.A.  1973.   Ultraviolet  radiation as a  cause  of skin  tumors.  Grit.
    Rev.  Toxicol. 2:211-25.

Emmett,  E.A..   1979.    Phototoxicity  from endogenous   agents.   Photochem.
    Photobiol. 30:429-36.

Epstein,  J.H. 1983.  Photocarcinogenesis,  skin cancer and aging.  J.  Amer.  Acad.
    Dermatol. 9:487-502.

Fitzpatrick, T.B., et al.  1979.   Dermatology in  general medicine,  3d ed.  New
    York:  McGraw-Hill.

Forbes, P.D.,  R.E.  Davies,  and   F. Urbach,  1978.    Experimental  ultraviolet
    photocarcinogenesis:   wavelength  interactions  and   time-dose  relation-
    ships.  Nat. Cancer  Int.  Monograph.   50:31-38.

Geeraets,   W.J.,  and E.R.  Berry.  1968.   Ocular  spectral characteristics  as
    related  to  hazards  from  lasers  and  other light  sources.    Am.  J.  Ophth.

Graham, S., J. Marshall,  B. Haughey, B. Stoll,   M.  Zielezny,   J. Erasure,  and
    D. West.  1985.    An  inquiry  into  the  epidemiology of  melanoma. Am.  J.
    Epidem.  122:606-19.

Harber, L.C.,  and D.R. Bickers.  1981.   Photosensitivity  diseases: Principles
    of diagnosis  and treatment.   Part  two: Principles  of photobiology  and
    photochemistry,   11-41.   Philadelphia: Saunders.

Haussler,  M.R., and  T.A.  McCain.  1977.   Basic and clinical concepts related to
    vitamin D metabolism  and  action.  N.  Eng.  J.  Med. 297:974-83, 1041-50.

Hawk, J.L.M., and J.A.  Parrish.  1982.  Responses of  normal skin to ultraviolet
    radiation.   In  The  science  of  photomedicine,  eds.  J.D.  Regan  and J.A.
    Parrish,  216-60.  New York:  Plenum  Press.

Kraemer,   K.H.,   1983.    Hereditable  diseases  with  increased   sensitivity  to
    cellular injury.  In  Dermatology in general  medicine,  eds.  T.B. Fitzgerald
    et al., update,  pp.  113-142.  New York: McGraw-Hill.

Kraemer,  K.H. and M.H. Greene. 1985.   Dysplastic nevus  syndrome. Dermatologic
    Clinics. 3:225-37.

Kripke, M.L.  1981.  Immunologic mechanisms  in  UV  radiation  carcinogenesis.
    Adv.  Cancer Res. 34:69-106.

Kripke,    M.L.    1982.      The   role    of  ultraviolet   radiation-induced
    immunosuppression  in experimental  photocarcinogenesis.  Prog.  Dermatol.

Lehmann, A.R.  1975.   Postreplication repair  of DNA in mammalian  cells.  Life
    Sci. 15:2005-16.

Lerman, S.  1980.  Radiant energy and the eye.  New York: Macmillan.  861-68.

Lippke, J.A., L.K. Gordon, D.E. Brash,  and  W.A.  Haselture.  1981.  Distribution
    of UV  light-induced  damage  in a defined sequence  of  human DNA:  detection
    of   alkaline-sensitive   lesions   at    pyrimidine   nucleoside-cytidine
    sequences.  Proc. Natl.  Acad.  Sci.  USA.   78:3388-92.

Pitts,  D.G.,  L.L.  Cameron,   J.  Jose,   et al.  1986.    Cataracts.    In  Light
    toxicity, long-term  visual  health  and  optical  radiation,  eds.  M.  Wexler
    and V.  Hutchins.   Cleveland: CRC Press.

Quevedo, W.C., T.B. Fitzpatrick and K.  Jimbow.  1985. Human skin color:  origin,
    variation and significance.   J. Human  Evolution. 14:43-56.

Rahn,  R.O.   1979-   Non-dimer   damage   in  deoxyribonucleic  acid   caused  by
    ultraviolet radiation.  In Photochemical and Photobiological Reviews,  vol.
    4, ed.  K.C.  Smith, pp. 267-330. New York:  Plenum Press.

Saito,  I.,  H.  Sugiyama, and  T.   Matsura.  1983. Photochemical  reactions  of
    nucleic  acids   and   their  constituents   of photobiological   relevance.
    Photochem.  Photobiol. 38:735-43.

Scotto, J.,  T.R.  Fears,  and  J.F. Fraumeni.  1981.  Incidence  of  non-melanoma
    skin cancer in the United States.   Publication  number (NIH)  82-2433,  U.S.
    Department of HEW.

Smith, K.C. 1977. Photochemistry:  the science  of photobiology,  Chapters  3 and
    4. New York: Plenum Press.

Stern, R.S., S. Zierler,  and J.A.  Parrish.  1980.   Skin carcinoma  in patients
    with  psoriasis  treated   with  topical  tars  and  artificial  ultraviolet
    radiation. Lancet:732-725.

Sutherland, B.M.  1978.  Photoreactivation  in mammalian  cells.   In Aspects  of
    genetic action and evolution,  eds.  Bowine and  Danielli, Int.  Res.  Cytol.
    Suppl.  8:301-33.

Tucker, M.A., J.A.  Shields,  P. Hartge, J.  Augsburger, R.N. Hoover, and  J.F.
    Fraumeni. 1985. Sunlight  exposure as risk factor for intraocular  malignant
    melanoma. N. Eng. J.  Med.  313:789-92.

Van der Leun, J.C. 1984.   UV  carcinogenesis.  Photochem. Photobiol. 39:861-68.

Ozone Depletion and Ocular Risks  From
Ultraviolet Radiation

Morris Waxier
Center for Devices and Radiological Health
Food and Drug Administration
Rockville, Maryland USA

     Because depletion of  the  ozone  layer  (Watson et  al.  1986)  increases
ultraviolet  (UV)  radiation, and  since  ambient outdoor levels of UV  radiation
have  been   implicated  as  a  risk  factor  for  retinal  damage  and  cataract
formation (Waxier,  In press), expected atmospheric changes may increase the
risk of eye  damage.

     Ultraviolet  (UV)  radiation  can  damage  the  cornea  (Pitts  1973),  the
crystalline  lens  (Pitts et al. In press), the ocular photoreceptors (Massof et
al. In press), and the retinal pigment  epithelium (Ham et al.  In press).  This
damage may  contribute to  long-term  problems   such  as  retinal degeneration
(Newsome et al.  In  press),  visual  aging  (Owsley  et al. In press),  impaired
visual development  (Fulton  et al. In  press), and cataracts (Pitts et al. In
press).  These analyses were  made to determine  whether  optical  radiation has
the potential  to produce long-term visual  health problems and, if so, under
what conditions they might occur.   A risk assessment was conducted  (Waxier, In
press) to focus attention on estimating the risks that optical  radiation poses
for long-term  visual health problems,  the  conditions under which they might
occur, and the ways by which we can minimize the  occurrence of such problems
on the basis of what we know.   This  risk assessment will provide the  basis for
judging  the extent  to which  depletion of the ozone  layer  increases these


     To  provide  some  informed  guesses about  the  likelihood  that  long-term
visual health  problems could result from  exposure  to optical radiation, we
need to specify the visual  health problems  about  which  we  are  concerned, and
then  to  analyze  the evidence of  involvement  of  optical radiation  in these
problems.   First,  the potential health problems  already  have been  identi-
fied.    They are cataracts, stable retinal disorders,  retinal  degeneration,

visual aging,  and  developmental disorders.   Second,  because  these  long-term
problems stem from similar photobiological and visual science  principles,  the
following series of questions were  devised to help in  this risk assessment:

     •  Does the ocular tissue involved in the disorder absorb optical radia-

     •  Does the absorbed radiation induce bioeffects  in  this tissue?

     •  Are these radiation bioeffects a hazard  to this tissue?

     •  Can  the  disorder be  deduced from  the  damage induced in the ocular
        tissue by the radiation?

     •  Is there evidence confirming that  the disorder is  caused,  wholly or in
        part, by optical radiation?

     •  Can  the  probability  that optical  radiation  induces this  disorder be


     First,  we  know  that the  crystalline lens  absorbs ultraviolet  radiation
(Boettner and Wolter  1962; Kinsey 1948;  Cooper and Robson  1969).  Moreover, UV
radiation can  induce  biochemical and histological effects  in  the crystalline
lens  of  various animals  (Pitts  et  al.  1977;  Zigler and  Goosey 1981).   In
addition,  some  of these  bioeffects are  related  to  UV-induced  opacities in
human and animal lenses  in  vitro (Lerman   1983;  Varma et  al.  1983).   Although
there are a  number of limitations  to extrapolating these data to the In situ
lens of  the  human  eye (Pitts et al. In press),  the deleterious bioeffects do
provide a  basis  for  identifying cataracts as a hazard of  ultraviolet radia-
tion.  Various  studies  have  determined the wavelengths and  radiant  exposures
needed to  induce cataracts  (Pitts  1973;  Pitts  and  Cullen  1981;  Schmidt  and
Zuclich 1980; Zuclich and Connolly  1976).   Several photochemical pathways have
been described whereby cataracts  could be  induced  in  humans (Jose 1978; Zigman
1981).   Moreover,  there are  similarities between  UV-induced cataracts  and
human cataracts  of unknown  etiology  that provide an added basis for postu-
lating a risk  of UV  cataracts in humans  (Lerman  and  Borkman 1976).   Finally,
there are  clinical reports of photosensitized cataracts   in  humans  (Lerman et
al. 1983) and, more importantly,  epidemiological evidence  that demonstrates an
association  between  exposure  to  ultraviolet radiation and  cataracts (Hollows
and  Moran   1981;  Hiller,  Giacometti,   and  Yuan  1977;  Taylor 1980;  Hiller,
Sperduto, and Ederer, 1983; Ederer, In press).  Thus,  there is sufficient data
to  identify  cataracts  as a risk  of  ultraviolet  radiation.    However,  the
uncertainties  associated  with  this  evidence  limit the  utility   of   this
epidemiological  data for  estimating  the risk  of  human  cataracts  from UV

Stable Retinal Disorders

     Studies have been conducted that provide a chain of  evidence on which to
postulate  that  optical  radiation  can  contribute  to  human  retinal  disorders
that  are  not progressive, i.e.,  stable retinal problems.   This  chain starts
with the knowledge that UV-B, UV-A,  and light are absorbed by various retinal


tissues (Wolbarsht  1976;  McGuinness  and Proctor 1973)-  Moreover,  the amount
of radiant  energy  incident on  the  retina depends on  the  transparency of the
cornea and lens, and on pupil diameter.  Next, a large number of bioeffects of
these  wavelengths   have  been  identified as  being  hazardous  to  the  retina
(Massof et  al.  In   press;  Ham  1984).   Further, empirical  estimates  have been
made of the  domain  of time, intensity,  and wavelength that can induce damage
to the  pigment  epithelium, the  photoreceptors,  and the inner  retina.   These
estimates  provide  sufficient   quantitative   evidence  to  establish  safety
standards to protect people from the  short-term retinal damage  which could be
induced by  lasers  and other intense  sources  of optical radiation  (Ham 1984;
Sliney  and  Wolbarsht  1980).    There  is,  moreover,  sufficient  evidence  of
additive and persistent changes  in the monkey and  human retina  to hypothesize
that some  stable long-term visual impairments  in  humans  may  be related  to
optical radiation.   There  are reports  that  optical radiation produces retinal
damage  hours,  days,  months,  and  years  after  irradiation.   Although  it  is
sparse, this  evidence  is  instructive.   For example,  sungazing by  humans has
been reported to produce  persistent, evolving,  and delayed alterations in the
visual  system.    Craik   reported that   aniseikonia may  persist  for  years
(Cavonius, Elgin, and Robbins 1974).   Pigmentary patches and blind spots have
been reported to evolve in a complicated fashion over many months after solar
retinopathy  (Loewenstein  and Steel  1941).    In addition,   late  complications
resulting  in  reduced  visual   acuity  have  been   described  (McFaul  19&9).
Furthermore,  there  are   several  reports  in   humans  of   prolonged  and  even
cumulative impairments in dark  adaptation following extended viewing by humans
of skylight  (Hecht  et al. 1948; Clark,  Johnson,  and Dreher 1946)  or bright
blue, violet  (Brindley  1953),  and ultraviolet  (Wolf 1949)  radiation.   Also,
there is  at  least   one report of persistent  visual  impairment  resulting from
accidental exposure to ultraviolet  radiation  from a welding arc  (Naidoff and
Sliney  1974).    Even  more  important  is  the  description by Tso and Woodford
(1983) of  sub-RPE   neovascularization and late impairment   in fluid  transport
after  several  years   of   evolving  changes   in  retinal  pathology  following
excessive exposure  to  optical  radiation.  Although  long-term observations  of
the  monkey  retina  damaged by UV  have  not  been  conducted,  some  of  the
fluorescein leakage that has occurred within  days  or weeks following irradia-
tion suggest longer  term sequelae.

     Some of these  effects are  the result of injury to the retina,  which was
ophthalmoscopically  visible; others  resulted  from exposures to optical radi-
ation that did not  produce ophthalmoscopically visible damage.  Most important
are  two demonstrations  that the addition of  subthreshold  exposures  separated
by several days induces retinal  damage (Griess and  Blankenstein  1981; Kuwabara
and  Okisake  1976).  The  data strongly suggest  that  retinal damage induced  by
optical radiation  can accumulate  over  days.   This is especially  important
since the concern about long-term visual health problems necessarily involves
consideration  of  the  intermittent  nature   of  exposure  to  emissions  from
multiple  sources  over an  extended period  of  time.    Split-dose  experiments
using UV-A radiation of the monkey retina are  in progress and strongly suggest
that the  damage accumulates over  days  to  produce pathological changes that
persist  for  months.    Photobiological   principles  suggest that  a  similar
cumulative effect would occur in the UV-B region.

Retinal Degeneration.   There are some reasons to  suspect  that  retinal damage
induced by  ultraviolet radiation  can be unstable, that  is, it  can produce
progressive  damage.   The  retina  of monkeys can  be  damaged by  ultraviolet


radiation (Ham 1984;  Peyman,  Sloan,  and Lira 1983).   This damage to the retinal
pigment  epithelium,  to  the  photoreceptors,  and   to   the  neurons  is  dose-
dependent,   photochemically  mediated,  and  cumulative  over  hours  and  days.
Although very few long-term observations have been  made on  the primate retina
following  excessive  exposure  to  optical  radiation,   there  are  theoretical
reasons to suspect that  a  retina compromised by disease or age  would be more
vulnerable  to  any  degenerative  changes  that  could  be induced  by  optical
radiation  (Young  1981).     The  combination  of   short-wavelength   optical
radiation,  oxygen, chromophores, and photosensitizers  in the retina  is potent
with  possibilities  for  producing  retinal  degeneration,  although  direct
evidence of the exacerbation of retinal degeneration by optical  radiation has
been obtained only  in  rodents (LaVail  and  Battelle 1975).   In  addition, the
higher retinal  doses of UV  received by  aphakes  and pseudophakes make these
individuals  especially  vulnerable  to   cystoid  macular  edema  (Kraff  et al.
1985).   Fluorescein  leakage occurs  following exposure  of  the  retina  of the
aphakic monkey  eye  to  UV-A  radiation   (Peyman,  Sloan, and Lira  1983;  Tso  et
al.  Work  in  progress),  but  comparable studies using  UV-B  radiation  have not
been conducted.    Therefore,  retinal  degeneration should be considered  a risk
of excessive exposure to ultraviolet radiation.

     Very little  research  has  been conducted on  primates,  on diurnal  animal
models of human  retinal  degenerations,  or  on retinas compromised  by chemicals,
age, or other agents.   The  striking  evidence of long-delayed and persistent
changes in  the  monkey  retina following optical  radiation,  coupled with some
clinical evidence in humans,  suggests  that  research on primates  needs to be a

Aging Disorders.   Deterioration  of  the visual system  as individuals  age is a
fact  (Owsley et  al.   In  press).   The  likelihood   that  optical  radiation
contributes to  this  deterioration  is  high for  several reasons.  First, the
cellular pathology observed  in  older  individuals is similar in  appearance to
radiation-induced  damage  (Kuwabara   1979).     Second, the   ophthalmoscopic
appearance  of  the aged  retina  is sometimes  similar  to a  retina exposed  to
optical radiation  (Guyer  et  al.   1986).   Third,  some of the  kinds  of vision
loss  described  in  the  aged  are consistent  with  the losses  that  could  be
produced by  optical  radiation  damage  (Klein  1958;  Jaffe,  de  Monestario, and
Podgor 1982).  Fourth,  the accumulation of lipofuschin material in the RPE and
other changes in  the retina  suggest that the aged  retina has  less capability
of  repair  (Feeney-Burns, Berman,  and Rothman 1980).   The retina of the human
aphakic eye  has three times more macular drusen and  pigmentary changes than
the  retina  of the phakic eye,  suggesting that UV  plays a  role  in  aging the
human retina (Guyer et al.   1986).

     Optical radiation may have quite different visual health consequences for
eyes with  and without ocular disease.   Optical radiation may accelerate the
deterioration of central  vision in  older  individuals with  central retinal
disease,  whereas, in  those  free  of  ocular disease,  optical  radiation may
impair paramacular and peripheral vision more than foveal vision.  This inter-
pretation is offered to  reconcile the  facts of  visual loss  outside the macula
in  disease-free  aged eyes (Owsley et  al.  In press)  with  Young's prediction
that  the  central  retina will  exhibit  the  major effects of  lifetime  retinal
exposure to  optical  radiation (Young  1981).   This ad  hoc  explanation is not
very  satisfactory but  it does  indicate  the need  to temporize  theory with
data.   Other  interpretations also are possible, e.g.,  those  individuals with

central  retinal  disease may  have had  a different  exposure history  or some
genetic vulnerability.

     Given  these  facts and theoretical  considerations,  exposure  of  the aged
retina to more ultraviolet radiation than permitted by the phakic eye probably
is hazardous.  The retina of  the  aged individual is exposed to more radiation
than phakics receive when the crystalline lens  is  extracted without placing a
comparable  filter  in  front  of  the  retina,  even when  ordinary  sunglasses are

     Research is needed to determine  the extent to which optical radiation is
a factor in  the deterioration of  the  vision of  the elderly.  In the meantime,
we  should  take practical  steps  to  protect our  eyes from  excessive  optical
radiation,  and thus prolong  our visual lifetime.

Development  Disorders.  While  there  are several reasons  to suspect hazardous
effects  of  optical radiation on  visual development,  almost no  research has
been conducted  on the  developing primate  retina.   In  fact,  there  is very
little  experimental  data on  the  exposure  of  the  developing  retina  of any

     Glass et al.  (1976) have reported  that the probability of retinopathy of
prematurity  (ROP)  is  highest  in  infants exposed  to more  light  during their
hospital  stay.     This  evidence  is  consistent  with  the   evidence  for  the
interaction  of oxygen  and light in retinal damage  (Lerman  1980).   The action
spectrum for this  effect  is unknown  but UV-B  and UV-A probably  play  a role.
Furthermore,  there  is a  large variety  of  sources  of optical  radiation and
situations  in  which  infants are  exposed to optical  radiation.    Some  of the
same evidence that gave rise to suspicions  about the role of optical radiation
in retinal degeneration and visual aging is relevant to the potential hazards
to visual  development,  i.e.,  short-wavelength radiant energy  incident  on the
retina,  short-wavelength   chromophores  in   the  retina,  dose-dependence  of
retinal  damage,   cumulative  effects,   long-duration  effects,   and  delayed
effects.   In addition,  special  properties of  the  developing  visual  system
might  decrease  injury threshold,  such  as   more  transparent ocular media and
different densities of chromophores and  screening  pigments,  or  might increase
the  vulnerability  of  infants,  such  as long  periods  of  post-natal  retinal
development  (especially  foveal)  and  early  dependence  on nonfoveal  visual
fields.  Thus,  the long  periods required for many visual processes to mature
suggest  a  window-of-vulnerability during which time optical  radiation might
alter  later visual performance,  especially if  optical radiation  alters the
spatial and  temporal  summation properties of the neural retina.

     Even though there is little  experimental data on  the  effects of optical
radiation  on  the  developing  retina,   photobiological  and  visual  science
evidence is  sufficient to postulate  that   exposure  to optical  radiation may
cause disorders of visual  development.


Exposure Variables

     Several exposure  conditions affect  the likelihood  of long-term  visual
impairment.   The  fundamental  parameters  are  wavelength,  level of  radiant
exposure,  and  duration of  radiant  exposure.  Ocular  damage occurs at lower
radiant  exposures  and  shorter  exposure  durations  for  shorter  wavelengths.
More intense repetitive exposures at various  wavelengths  can cause eye  damage,
as  can exposures  of longer  duration at  intermediate  levels  of  radiation.
Depletion of the ozone layer increases the  ambient level  of UV-B radiation and
therefore also increases the risk of eye  damage.

Biological Variables

     A  number  of   biological   factors  influence  the  potential  for  visual
impairment from exposure to optical radiation.  Pupil  size and  the absorption
characteristics of ocular tissues are quite  important.   Lens transparency and
yellowing, aphakia,  and  pseudophakia  are important factors  in  estimating how
much  retinal  damage will  occur.    Lens  yellowing  may  promote  additional
opacification  of the  lens  by  increasing  the absorption of UV and blue light;
on  the other  hand,  the retinal  damage  threshold  will  be increased.    An
increase  in  oxygenation and   temperature   in  ocular  tissues  will  greatly
increase the potential for visual impairment.  Also,  the  developing infant eye
and  the  aging eye  may be  especially  vulnerable  to the  damaging effects of
radiation.  The retina may  be more  vulnerable  to  radiation effects when it is
degenerating because of genetic or other  toxic effects.  The cumulative effect
of  photochemical   events and   tissue  repair  rates  are  crucial  biological
variables influencing the likelihood of long-term  visual  health  problems.

Long-Term Risk—An Exercise in Its Estimation

     This section  attempts to  estimate the  domain  of  optical radiation  risk to
long-term visual health.   This domain refers  to a  subset  of the  wavelength and
radiant  exposure  parameters.    This  is  an  effort  to  establish a  boundary
between those  exposure conditions  that  are  likely  to produce visual  health
problems and those that are not.

Determine  Radiant  Exposure Associated  With  Disorders.   First,  to  establish
such a boundary,  the radiant exposures of each particular ocular disorder must
be determined.

The  Possible Effect  of UV  Exposure  on  Cataracts.  Some of the epidemiologic
data on sunlight and cataracts can be used  to estimate  the radiant exposure at
UV-B wavelengths (Hollows  and Moran  1981; Hiller,  Giacomett,   and Yuan 1977;
Taylor  1980).   Each of these  studies found  that  cataracts were  positively
associated with sunlight exposure when the latter exceeded an annual level of
2500 Minimal Erythemal Doses  (MEDs).  Luckiesh et al.  (1930) reported  the MED
in the UV-B to  be  about 0.004  joules per square  centimeter.  Therefore,  2500
MED  equals 10  joules  per  square  centimeter annually.    The  daily  radiant
exposure would be  the annual dose divided by  365 days,  0.027 joules per square
centimeter and a three-hour daily exposure would  be 0.0068 joules per square
centimeter.  On the  other hand,  if  we  use  3^8 joules per square centimeter as
the  value  (Parrish,  Jaenicke,  and  Anderson  1982)  for  the MED  at  300  nm, the

three-hour estimated dose becomes 0.06 joules per  square  centimeter.   Another
estimate can be  derived by using  the local ambient  level  at 310  nm  at 15°N
reported  by  Johnson  et al.  (1976)   and  the epidemiologic  data reported  by
Zigman et al.  (1979).   This estimate  is 0.9 joules per square centimeter.

     Several estimates of the UV-A radiant exposure needed to induce cataracts
can be derived  or have been reported.   If only the UV-A wavelengths emitted by
the sun  are  responsible for the  cataracts associated with  2500  MEDs  insola-
tion,   then  the annual  radiant exposure  associated  with  cataracts would  be
140,000  joules  per square centimeter,  approximately 395  joules  per  square
centimeter daily and 98.8 joules per  square  centimeter for  a three-hour daily
exposure; this assumes  (Johnson,  Mo,  and  Green  1976) that one MED at  365 nm
equals 57.7 joules  per  square centimeter.   Another  estimate  can be  obtained
from Lerman (1980).   He estimated that  accidental exposure of the the human
eye to a UV-A photopolymerization device produced cataracts at 21.6 joules per
square  centimeter.    Finally,  the  Zigman et  al.  (1979)  epidemiologic  data
suggest  an  estimate of 0.44  joules  per  square  centimeter,  using the local
ambient level  (Pitts,  Bergmanson,  and  Chu 1983)  at 340 nm at 15°N.

     Experimental levels  of radiant  exposure  needed  for UV cataractogenesis
are summarized  by Pitts et al. (In press).  Several  values  were  selected for
comparison  to   the above  epidemiologic  estimates.    These  estimates  are
summarized in Table 1.   The lowest radiant exposure for UV cataractogenesis is
at 300 nm as estimated  from the epidemiologic data.   Moreover, even  when the
joule equivalent for  the  MED used is 348  joules  per square  centimeter rather
than  4.0  joules  per   square  centimeter,   the  three-hour  radiant  exposure
estimate  of  the former is only  one  log  unit  higher than  the  latter.   The
Zigman et  al.  (1979)  estimate is only  0.3  log  units higher than this high
estimate.   Experimental  estimates   are   in this  intensity  range.     These
comparisons  strongly   suggest  that   decades  of  exposure  to  UV-B   induces
cataracts in humans at  doses comparable  to that  suggested  by the  experimental
evidence.  On the other hand, comparison of these UV-B estimates with the UV-A
epidemiologic  estimates clearly  demonstrates that  higher  doses  of UV-A than
UV-B are needed to  induce cataracts.   However, how  much higher  is  not clear.
The estimates based on  the Hollows and Moran (1981); Hiller,  Giacometti, and
Yuan  (1977); and  Taylor  (1980)  data  correspond  closely  to  the  experimental
data,  suggesting that 4.0  log  units  more  radiant  exposure are needed  at UV-A
than  at  UV-B  wavelengths.   On  the   other  hand,   the UV-A  estimate based  on
either the Zigman  or  the Lerman data suggest  that only 1.5  log  units higher
doses are needed.

The  Possible  Effect  of  UV  Exposure on  Retinal  Problems.   The  following
estimates of the level  of  radiant  exposure needed to induce  long-term retinal
problems  assume  a  three-hour  daily   exposure   additivity  over   days  and
additivity over wavelength.  When  the data used  for the estimation were based
on an  exposure  of less  than  three hours, the  level of radiant  exposure was
calculated based  on the  particular  duration used  in that  study  (i.e.,  1000
seconds).  Evidence that the damaging  effect would accumulate over a period of
time at least as long as  three hours  provides a critical bridge between these
shorter durations of exposure (higher  dose rate).   On the other hand,  when the
data were based on  exposures longer   than  three hours,  dose  was prorated to a
three-hour duration.  For  broad-band  sources of  optical radiation,  wavelength
additivity was  assumed and the estimate was based on the total integrated dose

    Table  1a.  Radiant Exposure for Cataractogenesis:  Experimental
1.2 x 10~1r3
1.26 x 101r3
191. 5r
6.27 x 10~4r
7.06 x 10~4r
2.65 x 1(T3r
5.06 x 101r3   3.6 x 104r

1.0 x 102r3
1.405 x 10~3r
    Table 1b.  Radiant Exposure for Cataractogenesis:   Epidemiologic
6.8 x 10~3d
6.0 x 10~2d
9.0 x 10"2d
3.24 x 10~1d
1.08 x 10^a
1.08 x 10^a
1.08 x 104a
1 .08 x 10^a
( Jem )
1.3 x 102r,a70'71
3.0 x 10~5Wcm-2r*
4.5 x 10-1d    1.08 x 104a     6.56 x 102r70'71

9.88 x 10~1d   1.08 x 104a     2500MED3r23'24'25
 r - reported
 d - derived
 a - assumed                      .
 1 - 1.0 MED = 4.0 x 10~3Jcm-2 r,a°jj
 2 - 1.0 MED = 3.48 x 102JcnT2 r,a°9
 3 - 1.0 MED = 5.77 x 101JcnT2 r,ab9
 * - Dose rate reported to be associated with cataracts induced by dental
     photopolymerizer (Cole et al. 1985)

     The  only  human  data from  which  one  could  derive an  estimate  of  the
radiant exposure  of UV  that could  damage the  retina  were  in Hecht  et  al.
(1948) and  Clark  et al.  (1946).   These data showed  that staring  at skylight
for several hours  induced an abnormal retardation of dark  adaptation.   Since
rhodopsin has an absorption  ban  in the  UV-A  (Kurzel,  Wolbarsht, and Yamanashi
1977), this deficit  could have  been  due to the ambient  outdoor UV-A.   If the
ambient  level   of  UV-A   at  the  cornea  was  about   0.09   joules  per  square
centimeter as  suggested  by Ham  and  Mueller  (1982),   then the  retinal radiant
exposure might have been about 0.009  joules per square centimeter in the Hecht
and Clark  studies.  Experimental  estimates  of  UV retinal  damage  in monkeys
have been few.   Schmidt and Zuclich (1980)  found the threshold at 325 nm to be
10 joules  per  square centimeter.  Recent  evidence in phakic  rabbits (Pitts,
Bergmanson, and Chu  1983)  and rats (Rapp,  Jose,  and  Pitts  1985) suggests that
the radiant exposure necessary  to  damage the retina  is  even  lower  at 300 nm.
The current threshold for  300 nm damage  to the monkey retina is approximately
0.6 joules  per square  centimeter.   Ham and  Mueller (1982)  found  that  the
aphakic monkey has  thresholds of 5.0 joules per  square  centimeter  at 325 nm,
5.4  joules per  square  centimeter  at  350  nm,  and  8.1   joules  per  square
centimeter at  380  nm.   These estimates are summarized in Table 2.   Estimates
are provided for wavelengths longer  than 400 nm  to  compare the effectiveness
of various spectral regions.

     Griess and Blankenstein (1981)  have shown  that  the damage at  441  nm is
cumulative over days, thus,  not  only  verifying  the radiant  exposure needed to
produce the  damage,  but  also  demonstrating  that  one-half  the energy  can be
delivered 24 hours  later  and still damage the  retina!   Preliminary data (Tso
et al. Work  in progress)  show  that  a similar effect occurs with  UV-A  radia-
tion.    UV-B   probably  produces a comparable  effect,  although this  is  an
extrapolation  based  on  photobiological  principles,  not  on direct  split-dose

Identify Ambient Levels  and  Compare  to  Radiant Exposure Estimates.   Ambient
levels of optical radiation  have been measured  at many geographical locations
around the world.   Johnson, Mo,  and Green (1976) have measured solar radiation
as a  function  of  latitude from  305  to 340 nm.   These  data are  adjusted for
three-hour daily exposures.  Since their data  included only wavelengths up to
340 nm, the  solar  spectral  radiance  reported  by Sliney and  Wolbarsht  (1982)
was used for wavelengths  longer than 340 nm.   Their  spectrum  for  the "summer
sun near zenith" was multiplied by a solar disc  size of 0.00006  sr to  obtain
the solar  spectral  irradiance  and  then multiplied  again  by  the  number  of
seconds  in a  three-hour  exposure  to  obtain  the   solar  radiant  exposure.
Several other  sources are available  for more detailed analytical  research on
ambient levels  of  solar  radiation  (Green,  Sawada,  and  Shettle  1974;  Green,
Cross, and Smith 1980).

     The relationship of  these outdoor ambient  levels to the  radiant exposure
estimates for cataracts and  retinal disorders  should  be  noted.   Cataracts can
result from  outdoor ambient levels  of UV-B and  UV-A; however, UV-B is much
more  effective.    Ozone  depletion  would   increase   ambient  levels  of  UV-B
radiation much more than  UV-A  radiation.    Within the  UV-A  region  there is
considerable variability   in  the estimates.    The possible  bases   for  these
differences will be discussed later.   Ambient  UV-A levels outdoors seem to be
sufficiently intense to damage the  retina.   Most damage thresholds between 400
and 760 nm are one or two log units  above  ambient light   levels outdoors with


            Table 2.  Radiant Exposure for Retinal Damage
Wavelength Radiant
(nm) Exposure
( Jem )

.0 x

. ir
.0 x
.0 x
.5 x
.0 x
.0 x
.2 x
.57 x
.52 x
.0 x
.2 x
.0 x
.0 x
.0 x
.0 x
.1 x
.08 x
.0 x
.0 x
• 3d
.0 x
.5 x
.08 x
.0 x
.5 x

10 r
10 r
10 r
10 r

10 V
10 r


1Q r
10 r
Exposure Retinal
Duration Irradiance
(s) (Wcm~2)


r 4

.0 x

.08 x
.08 x
.0 x
.0 x
.0 x
.0 x
.08 x
.0 x
.0 x

.8 x
.0 x
.2 x
.0 x
.0 x
.02 x
.2 x
.0 x
.08 x
.0 x
.0 x

.08 x
.0 x
.08 x
.0 x
.0 x


1Q a


10 r

10 r
10 r

.3 x
.0 x
.0 x
.0 x
.0 x
.5 x
.5 x
.5 x
.3 x
.0 x
.0 x
.0 x
.57 x
.3 x
.0 x
.0 x
.0 x_
.0 x
.3 x
.9 x
.5 x
.1 X
.0 x
.0 x
.0 x
.08 x
.08 x
.08 x
.3 x
.5 x
.0 x
.0 x
.5 x

1 O




7 »


c* f^
a. 2
a. 4

O r*i
r» A-.
5 0 x 10~2r77
. . j .\j iw i
5.1 x 10~^r''
5.4 x 10~2r77

8 I x 1 0 r
.«.973 x 10~3r77

7, **
4x 10" r'

8 x 10~4r49»121
•6 x 10-^9-121
1 2 x 10~4r84


r - reported
d - derived
a - assumed
* - assumes 10% corneal exposure
# - R.G. Allen, personal communication
     additivity over days at 1.0, 0.5 and 0.25 times threshold dose
** _
   - R.G. Allen, personal communication

the sun  at zenith.   However,  several estimates are  within 0.5 log  units of
outdoor ambient levels.

Determine  Proportion  of  Ambient  Likely to Be  Incident on  the  Cornea.   So far
we have  assumed  that  all of the ambient  radiation  in a  three-hour period was
incident on  the  cornea.   However,  Cole et al.  (1985) have demonstrated that
only 25  percent  of ambient  radiation reaches  the skin of  the  face of a human
standing upright  and  looking forward.   Furthermore,  in a  standing position,
the corneal  dose  may  be  only 10 percent  of ambient due  to shielding from the
brow  ridge  and   the  corneal  grazing  angle   in the ultraviolet  (Wolbarsht
1982).   However, an individual prone  on  his back with eyes open could receive
a  100  percent  corneal dose  of ambient.   On the  other hand,  Sliney (1986) and
Rosenthal  (1985)  report  higher  ocular  irradiances  of  UV-B  from reflective
surfaces on the earth's surface than from skylight.

     These considerations suggest  that  cataracts  associated  with  UV  ambient
levels  may  be  due to corneal  doses  1.0  log  unit  lower  than  the  ambient
level.   Since  the Hecht et  al.  (1945)  and Clark, Johnson,  and  Dreher (1946)
data are  based  on  staring at  the  sky, 100  percent of ambient  level  is an
appropriate value  for their  data.   However,  for  those  individuals who do not
stare at the sky,  the corneal dose may  be 1.0 log unit lower than the ambient

Determine  Proportion  of  Ambient  Level to Which  anIndividual  Is Likely to Be
Exposed  and _Number of Days  Exposed  Per Annum.   The  variety of physical,
cultural, and occupational environments in which people live is enormous.  The
previous  discussions  assumed  that  everyone   is  exposed  to  ambient  outdoor
levels 365 days per annum,  three  hours per day.  Situations exist where people
may be exposed to much more  or much less than  this amount.

Setting  Risk Boundaries.  Most  of the estimates of radiant exposure that can
damage the eyes are within 1.0 log unit of outdoor ambient levels.   Therefore,
one risk boundary can be established as  1.0  unit  below  the  average  outdoor
ambient  levels to establish a safety  factor.   This  boundary  should  be moved
lower  to take into account increases  in UV-B  radiation  that are  caused by
ozone  depletion.   The  implication  is  that  radiant  exposure  of the  eye for
three hours at levels above  this boundary (the pink area)  is probably a long-
term visual health risk,  i.e., the "probable  risk" area.

     Since there  are  some  estimates  that fall below this  region  of probable
risk,  a  lower  boundary was  constructed to take  these estimates  into account,
as well  as  certain  worst-case  conditions.    For  example,  the  epidemiologic
estimates for cataracts were derived as  if the effects  occurred  independently
at  only   two   wavelengths,   300  and   365   nm,  an unlikely   possibility;
cataractogenesis  due  to  broad-band  UV  exposure  is   probably  additive across
wavelengths.    (Note:   As a  matter of fact, there  isn't  even any experimental
evidence that  UV  at 365 nm  can  induce  cataracts  in  vivo at  doses below 100
joules per square centimeter.)   As a  consequence,  the lower boundary could be
set two  log units below  the  upper boundary.   In addition,  these epidemiologic
studies are not very  precise, so that  perhaps  cataracts  are associated with a
lower number  of annual MEDs.

     If certain worst-case assumptions are made, such  as subthreshold effects,
genetic  susceptibility,   exogenous   photosensitivity,   poor  nutrition,  and


frequent exposure  to ultraviolet radiation,  then  perhaps the  lowest radiant
energy at which cataracts might be induced is an order of magnitude lower than
the lower  boundary.   However, the  empirical estimates already may  represent
worst-case  conditions;   for   example,   cumulative  effects  over   time  and
wavelength  were  assumed  in  establishing  the  estimates.     If  worst-case
assumptions  are accepted,  then  perhaps  the  reported   estimates  should  be
lowered by an order of magnitude  at wavelengths  shorter  than  460 nm.   Some of
the worst-case  assumptions  are  aphakia and  pseudophakia without   blue  and
ultraviolet blocking  filters, transmittance  of infant  lenses  close  to 1.0,
genetic and disease  susceptibilities, exogenous  photosensitization,  and addi-
tivity  of  damage  over  days.    Each  of  these  is  a  special  case  requiring
clinical management.

     Although some  of these  estimates  are  based  on  sparse  data and assume
worst-case conditions, they provide some  prediction  of the wavelength-radiant
exposure domain  within  which  optical radiation  can  be  expected to produce
long-term visual health  problems in  the  primate eye.  Therefore,  to conduct
critical tests  of  long-term visual health  problems,  wavelengths in  the blue
and ultraviolet  regions  should  be  used  at  radiant  exposures  from as low as
approximately 0.1 joules  per  square centimeter to 0.000001 joules  per square
centimeter, respectively.

Reduce Invisible Radiation  of the Eye.    People  in outdoor occupations should
protect their eyes.   Reducing  the amount  of ocular  irradiation  by using hats,
umbrellas,  or  sunglasses is  a simple precaution that can be taken  to avoid
potential long-term  visual  impairment.    Shielding the eyes from  sunlight is
especially valuable  on  bright sunny days  on a  sandy beach  or snow field.
Sunglasses that  absorb  the proper wavelengths  certainly  should be  worn when
staring  into  skylight.   One  should never  gaze directly  at the  sun  or  an
eclipse, even when wearing sunglasses.

     While sunglasses have been  available  for many years,  questions  have been
raised  about  their  effectiveness  in protecting  the  eyes  from  ultraviolet
radiation (Sterne 1984).   Although voluntary standards (ANSI Z 80.3)  exist for
nonprescription sunglasses (American National Standards Institute 1977), there
are  no  standards  for   prescription  sunglasses.    Therefore,  there  is  no
assurance  that  there  is sufficient  filtration of  ultraviolet radiation  in
prescription  sunglasses.    While the Sunglass  Association of  America claims
that  those  nonprescription sunglasses made  in compliance with  the  voluntary
standard  provide  adequate  protection  from  UV radiation (Loomis,   personal
communication),  the  consumer  has no way of knowing  whether any  particular
brand  conforms  to this  standard.   Furthermore,  there is disagreement about
whether  this  standard   is  sufficiently  stringent,   especially  in   the UV-A
region.  In addition, there may  be  sunglasses on the market that are  so dense
in the  visible  region that the  pupil can be dilated sufficiently  so  that the
partial  UV  filtration is  negated by the increased  pupil diameter.   Various
proposals have  been  made to solve these  problems and they are  under  vigorous
discussion  within  the   ANSI   Z-80  Committee and  within  the  Food   and Drug
Administration.  Some of these proposals are as follows:

      •  Require  each pair  of sunglasses  to have  a tag  with the   spectral
        transmittance on it.

     •  Require  a  "sunscreen"  eye  protection  rating  with  each  pair  of

     •  Require certification that the sunglasses are made in conformance with
        the voluntary standard.

     •  Increase the voluntary density requirement in the UV.

     •  Require that prescription sunglasses  conform to a standard comparable
        to the voluntary nonprescription standard.

Some manufacturers  urge  consumers to purchase expensive  sunglasses  which may
not provide much more protection than much less expensive sunglasses.  Clearly
some kind of sunglass standard  is needed  to  ensure that both prescription and
nonprescription sunglasses reduce the corneal  level  of radiant exposure below
the levels associated with cataracts and retinal damage.


     Until there  is a more  precise understanding  about the  role of optical
radiation in  visual development, visual  aging,   retinal  degeneration,  photo-
sensitized reaction,  and cataracts, there will  be  uncertainty  about whether
particular protective  measures  provide  too  much or  too little  protection.
People are exposed  to optical  radiation  from a large  number  of sources over
varying periods  of  time  for  an entire lifetime.   In  addition,  depletion of
ozone over time  and in  certain  geographical regions  lends  more uncertainty.
No matter  where  the  maximum  permissible exposure  limit is  established for
immediate ocular damage, ambient outdoor  levels  of UV  may approach this limit
under some conditions.   Until these limits are  related to specific  long-term
visual health problems,  protective measures should be based on some worst-case
estimates.   This may mean  that "overprotection"  will  be  established,  but
people will  be  better   served  by  using  too  much  protection  than   by  using
borderline or insufficient protection.


American   National   Standards  Institute,   1977.      Requirements   for   non-
  prescription sunglasses and fashion eyewear.   Standard Z-80.3.   New  York:
  American National  Standards Institute.

Boetner,   E.A.,  and  J.R. Wolter.  1962.   Transmission  of the  ocular  media.
  Invest.  Ophthal.  1(6):766.

Brindley,  G.C.  1953.   The effects  on colour  vision  of adaptation to very
  bright  lights.  J.  Physiol.  (London).  122:332.

Cavonius,  C.R.,  S.  Elgin,  and  D.O.  Robbins.  1974.  Thresholds  for  damage to
  the human retina by white light. Exp.  Eye Res.  19:543-548.

Clark, B.A., M.L. Johnson, and R.E. Dreher.   1946.   The effect of sunlight on
  dark adaptation. Am. J. Ophth. 29:823-836.

Cole, C.,  P.D.  Forbes,   R.E.   Davies,  and  F.   Urbach.  1985.  Effect  of   indoor
  lighting on normal skin.  Unpublished manuscript.


Cooper, G.F., and J.G. Robson.  1969.  The  yellow  color  of the lens of man and
  other primates. J.  Physiol.  203:411.

Ederer, F. In press.   Epidemiologic considerations. In Optical  radiation and
  visual health, eds.  M.  Waxier and V.M.  Hitchins.   Boca Raton,  FL: CRC Press.

Feeney-Burns, L.,  E.R. Berman,  and H.  Rothman.  1980.   Lipofuscin  of  human
  retinal pigment epithelium.  Am.  J. Ophth. 90:783-791.

Fulton, A.,  I.   Abramov,  D.  Varner, P.  Raymond,  J. Gwiazda,  J.  O'Neill,  J.
  Allen, and  L. Mainline.   In press.   Optical  radiation  effects on  visual
  development.  In Optical radiation and  visual  health,  eds.  M.  Waxier  and V.M.
  Hitchins. Boca Raton, FL: CRC Press.

Glass,   P.A., G. Avery, G.,  K.N.  Sivasubramanian,  A.N.  Sostek,  M. Keys,  and
  D.S.  Friendly. 1985.  Effect of bright light  in the  hospital  nursery on the
  incidence of  retinopathy of prematurity.  New  Eng.  J.  Med.  313(7):401-404.

Green,   A.E.S.,   K.R.   Cross,  and  L.A.   Smith.  1980.     Improved  analytic
  characterization of  ultraviolet skylight. Photochem.  & Photobiol. 31:59-65.

Green,   A.E.S.,   T.  Sawada,  and E.P.  Shettle. The  middle  ultraviolet  reaching
  the ground. Photochem.  & Photobiol.  19:251-259.

Griess, G.A., and M.F. Blankenstein.  1981.  Additivity and  repair of actinic
  retinal lesions. Invest.  Ophth.  & Vis.  Sci. 20:803-807.

Guyer,   D.R., M.F.  Alexander,  C.L.  Auer,  M.B.   Hamill,  J.A. Chamberlain,  and
  S.L.   Fine.  1986.    A comparison  of  the  frequency and  severity  of macular
  drusen in  phakic and  non-phakic eyes.  Invest.  Ophth.  Vis.  Sci.  (Suppl).

Ham,  W.T.,  Jr.  1984.    Basic  mechanisms  underlying   the   production  of
  photochemical  lesions in the mammalian retina. Curr.  Eye Res.  3:165-174.

Ham, W.T., Jr., and  H.A.  Mueller.  1982.   Action  spectrum  for  retinal injury
  from near-ultraviolet radiation in the aphakic monkey. Am. J.  Ophth. 93:299.

Ham, W.T.,  Jr., M.L.   Wolbarsht,  P.H.  Proctor, L.  Parver, M.F. Marmor,  D.L.
  Sliney,  L. Feeney-Burns,  and  R.G.   Allen.    In   press.    Retinal  pigment
  epithelial damage induced  by optical radiation.   In Optical  radiation and
  visual health,  eds. M.  Waxier and  V.M. Hitchins.   Boca  Raton,  FL:   CRC

Hecht,   S.,  C.D. Hendley,  S.  Ross, and  P.V.  Richmond.  1948.   The effect  of
  exposure to sunlight on night vision.  Am  J. Ophth. 31:1573-1580.

Hiller,  R.,  L.   Giacometti,  and K.  Yuan.   1977.    Sunlight  and cataract:  An
  epidemiologic investigation. Am.  J. Epidemiol. 105:450-459

Hiller, R.,  R.D. Sperduto, and F.  Ederer. 1983.    Epidemiologic associations
  with  cataract in the  1971-1972  National Health  and Nutrition  Examination
  Survey. Am. J. Epidemiol. 118:239-249.


Hollows,  F.,  and  D.  Moran.  1981.   Cataract—the  ultraviolet risk  factor.
  Lancet. 5:1249-1250.

Jaffe,  J.J.,  P.M.  de Monestario,  and M.J.  Podgor.  1982.   Retinal  aging and
  blue-sensitive cone function.  Invest.  Ophth. Vis. Sci. (Suppl.) 22:218.

Johnson, F.S., T.  Mo, and A.E.S.  Green.  1976.   Average latitudinal variation
  in  ultraviolet  radiation at  the earth's  surface.  Photochem. &  Photobiol.

Jose, J.G. 1978.   The role  of DMA  damage, its repair and its misrepair in the
  etiology of the cataract: A review.  Ophthal. Res. 10:52.

Kinsey,  V.E.  1948.   Spectral transmission  of  the eye to  ultraviolet.  Arch.
  Ophth. 39:508.

Klien, B. 1958.  Diseases of the macula. Arch. Ophthal.  60:195-186.

Kraff, M.C.,  D.R.  Sanders,  L.M.  Jampol,  and H.L.  Lieberman.  1985.   Effect of
  an  ultraviolet-filtering   intraocular   lens  on   cystoid   macular   edema.
  Ophthalmol. In press.

Kurzel, R., M.L.  Wolbarsht, and  B.S. Yamanashi. 1977.  UV radiation effects on
  the human eye.  Photochem. Photobiol. Rev. 2:133.

Kuwabara, T.  1979.   Photic and  photo-thermal effects  on  the  retinal  pigment
  epithelium.  In  The retinal pigment  epithelium,  eds.,   K.M.  Zinn and  M.F.
  Marmor, 293-313.  Cambridge:  Harvard  University Press.

Kuwabara, T., and S.  Okisake.  1976.  Effect of electronic strobe flashlight on
  the monkey retina.  Jap. J. Ophthal.  20:9-18.

LaVail,  M.M.,  and  B.A.   Battelle.   1975.   Influence of eye pigmentation and
  light deprivation on  inherited  retinal  dystrophy in  the  rat.  Exp. Eye Res.

Lerman, S. 1980.   Human  ultraviolet radiation cataracts. Ophthal. Res.  12:303-

Lerman,  S.  1983.   NMR  and  fluorescence spectroscopy on  normal,   aging and
  cataractous lens. Lens Res.  1:175.

Lerman, S., and R. Borkman  1976.   Spectroscopic evaluation and classification
  of the normal,  aging,  and cataractous  lens.  Ophthalmic Res.   8(5):335-53.

Lerman,  S.,   J.W.  Megaw,  K.H.   Gardner,   and  L.  Drake.  1983.   Ocular  and
  cutaneous  manifestations of  PUVA  therapy—A  review.   J.  Toxicol.  Cutan.
  Ocular Toxicol.  1:257-266.

Loewenstein,   A.,   and  J.  Steel.   1941.    Macular  burning  and  ring scotoma.
  Glasgow Med. J.   135:73-85.

T. Loomis, personal communication.


Luckiesh, M.,  L.L.  Holladay,  and  A.H. Taylor.  1930.   Reactions  of untanned
  skin to ultraviolet radiation. J. Opt. Soc. Amer.  20(8):423-432.

Massof,  R.W.,  H.  Zwick,  W.G.  Robison Jr.,  B.  Hochheimer,  L. Rapp,  and S.A.
  Sykes.  In  press.    Photoreceptor damage induced  by  optical radiation.   In
  Optical radiation and visual health, eds. M. Waxier and  V.M. Hitchins.  Boca
  Raton, FL: CRC Press.

McFaul, P.A. 1969.  Visual prognosis after solar retinopathy.  Brit. J. Ophth.

McGinness, J., and P.  Proctor.  1973.   The importance of the fact that melanin
  is black.   J. Theor. Biol. 39:677-678.

Naidoff, M.A., and D.H. Sliney.  1974.   Retinal  injury  from a welding arc. Am.
  J. Ophth.  77(5):663-668.

Newsome, D.A., E.  Berson,  H.G. Sperling, M.O.M.  Tso, T. Lawwill,  M. LaVail, R.
  Bonner, L.  Hyman,   and  M.  Mainster.  In press.    Possible role  of optical
  radiation in retinal degenerations.  In Optical radiation and visual health,
  eds. M. Waxier and V.M.  Hitchins.  Boca Raton, FL:  CRC  Press.

Owsley,  C.,   D.  Kline,   J.  Marshall,  J.S. Werner,  and  V. Greenstein.  In
  press.   Optical  radiation  effects  on  aging  and visual  perception.   In
  Optical radiation and visual health, eds. M. Waxier and  V.M. Hitchins.  Boca
  Raton, FL:  CRC Press.

Parrish,  J.A.,  K.F.  Jaenicke,  and  R.R.  Anderson.   1982.    Erythema  and
  melanogenesis action  spectra of normal human skin.  Photochem.  & Photobiol.

Peyman,   G.A.,   H.D.  Sloan,   and  J.  Lim.   1983.     Ultraviolet-absorbing
  pseudophakos: An efficacy study. Am. Intra-Ocular Implant Soc.  J. 9:161-169.

Pitts,  D.G.   1973.    The  ocular  ultraviolet action  spectrum and  protection
  criteria.  Health Physics.  15:559.

Pitts,  D.G.,   and  A.P. Cullen.  1981.   Determination  of  infrared radiation
  levels for acute ocular cataractogenesis.  Albrecht. von Graefes  Arch. Klin.
  Ophthalmol. 217:285-297.

Pitts, D.G.,  J.P.G.  Bergmanson,  and  L.W-F. Chu.  1983.   Rabbit eye  exposure to
  broad-spectrum fluorescent light. Acta Ophthalmol. (Suppl). 159.

Pitts,  D.G.,  A.P.  Cullen,  P.O.  Hacker,  and  W.H.  Parr.  1977.    Ocular
  ultraviolet  effects  from  295  nm  to  400 nm in  the  rabbit  eye.   DHEW
  Publication  No.   77-175.     Washington,   D.C.:    National  Institute  for
  Occupational Safety and Health.

Pitts,  D.G.,  J.G.  Jose,  J.P.G.  Bergmanson,  L.W-F.  Chu, V.M. Hitchins, and M.
  Waxier.  Manuscript  in preparation.

Pitts, D.G.,  J.A. Zuclich,  S. Zigman,  S. Zigler,  S.D. Varma,  E. Moss,  S.
  Lerman,  L.L.   Cameron,   and  J.  Jose.  In  press.    Optical  radiation  and
  cataracts. In Optical radiation  and visual health, eds.  M.  Waxier and V.M.
  Hitchins.  Boca Raton,  FL: CRC Press.

Rapp, L.M., L.G. Jose, and  D.G. Pitts.  1985.   DNA repair synthesis in the rat
  retina following in vivo exposure to 300-nm radiation.  Invest. Ophthal. Vis.
  Sol. 26:384.

Rosenthal,  F.S.,  M.   Safran,   and  H.R. Taylor.    1985.   The  ocular dose  of
  ultraviolet  radiation   from  sunlight  exposure.    Photochem.  &  Photobiol.

Schmidt,   R.E.,  and  J.A.  Zuclich.   1980.    Retinal  lesion due  to  ultraviolet
  laser exposure. Invest.  Ophth.  & Vis. Sci. 19:1166-1175.

Sliney, D.H. 1986.  Physical factors in cataractogenesis - Ambient ultraviolet
  radiation and temperature.  Invest. Ophthal. Vis. Sci.   27:781-790.

Sliney,  D.,  and M.  Wolbarsht. 1980.   Safety with  lasers  and other  optical
  sources.  Mew York: Plenum Press.

Sterne, D. 1984.  How much UV protection should sunglasses provide?  Optometry
  Times.  2(9):3.

Taylor, H.R. 1980.  The environment and the lens. Br. J.  Ophth. 64:303-310.

Tso, M.O.M., M.  Waxier, W. Lee, R.  Bostrom, K. Alexander, and L. Jampol.  Work
  in  progress.    UVA  induced  fluorescein  leakage  in the  macula  of  aphakic
  monkey eyes.   (Work in  progress).

Tso, M.O.M., and B.J.  Woodford. 1983.   Effect of photic  injury on the retinal
  tissues. Ophthalmology.  90:952-963.

Varma, S.D., D.  Chand, Y.R. Sharma, J.F. Kuck,  Jr.,  and R.D.  Richards. 1983.
  Oxidative stress on  lens  and cataract formation:  Role  of light  and oxygen.
  Current Eye Res. 3(1):35-58.

Watson,  R.T., M.A. Geller,  R.S.  Stolarski,  and  R.H. Hampson.  1986.   Present
  state of  knowledge of  the upper  atmosphere:   Processes  that control ozone
  and  other  climatically  important  trace  gases.  Washington,  D.C.:  National
  Aeronautics and Space Administration.

Waxier,  M.  In  Press.   Long-term visual  health  problems:   Optical radiation
  risk.   In Optical radiation and  visual health,  eds. M.  Waxier and V.M.
  Hitchins. Boca Raton, FL: CRC Press.

Wolbarsht, M.L.  1976.  The function of intraocular  color filters.  Fed. Proc.

Wolbarsht, M.L., G.S.  George, W.A.,  Shearin Jr.,  J.A.  Klystra,  and M.B. Landes
  III. 1983.  Retinopathy of prematurity:  A new  look  at an old disease. Paper
  presented at the first International Symposium on Light and Oxygen Toxicity,
  April 8-10, 1983,  Baltimore,  Maryland.

Wolf, E.  19^9.    Effects  on visual threshold  of exposure to  radiation below

Young, R.W.  1981.   A  theory of central  retinal  disease.  In  Future directions
  in ophthalmic research,  ed. M.L.  Sears.  New Haven:  Yale University Press.

Zigler, J.S., and J.D. Goosey.  1981.   Photosensitized  oxidation in the ocular
  lens:    Evidence  for  photosensitizers  endogenous   to  the  human  lens.
  Photochem. Photobiol.  33:869.

Zuclich,   J.A.,  and  J.S.  Connolly.   1976.   Ocular  damage  induced by  near-
  ultraviolet laser  radiation.   Invest.Ophthal. 15:760.

Zigman,  S.  1981.   Photochemical  mechanisms   in  cataract  formation.    In
  Mechanisms of  cataract  formation in the human lens,  ed.   G.  Duncan,  117-
  149.  New York: Academic Press.

Zigman,  S.,  M.  Datiles,   and E.  Torczynski.  1979.    Sunlight  and  human
  cataracts. Invest. Ophth.  Vis. Sci.  18:462.

 Overview of Our Current State of  Knowledge
 of UV Effects on  Plants
Alan H. Teramura
University of Maryland
College Park, Maryland USA
     Fifteen  years ago, we  virtually  knew nothing about  the  effects of UV
radiation on  plants  (Nachtwey,  Caldwell, and Biggs  1975).   Therefore, one of
our first needs was to  estimate the degree of UV  sensitivity that occurs in
plants.   To  date,  over 200 different  plant species and varieties have been
screened  for  their  responsiveness  to  increased  UV  (Teramura  1983;  Van and
Garrard  1975;  Van, Garrard, and West 1976;  Tevini and Iwanzik  1982; Biggs and
Kossuth  1978).  Most  of this work was done on  crop  plants but also a fair
number of native plants and trees  have been examined.   Although this  initial
work was  performed by numerous independent  investigators under  a  wide array of
experimental  techniques, an alarming trend  emerged:  nearly two  out of  every
three plants  tested  showed some  degree  of UV  sensitivity (Teramura  1983;
National  Academy of  Sciences  1982).  Members of  the pea and bean, squash and
melon, and  cabbage families are especially  sensitive to UV damage (Teramura et
al. 1983; Biggs and  Kossuth 1978;  Sisson  1981;  Steinmuller  1986).  A  diverse
array of  UV   responses  can  be seen in  plants,  including  physiological and
biochemical  responses, as well as morphological  and anatomical ones (Teramura
1983; Klein,   Edsall, and  Gentile  1965;  Caldwell 1981; Tevini,   Duansik, and
Thomas 1981).   In general UV  radiation  deleteriously affects  plant growth
reducing  leaf size,  limiting the area available  for energy capture  (Table 1).
Plant stunting and a reduction in total dry weight are also typically  seen in
UV-irradiated  plants.   Of  course  this  large  degree of  sensitivity elicited
considerable  concern;  however,  during  the  last  five years  it has been  found
that  at  least  some  of this  sensitivity  was  directly  attributable  to the
experimental  growth  conditions  (Teramura  1982;  Teramura, Biggs, and  Kossuth
1980).  In general,  plants grown  in growth chambers and greenhouses are much
more  susceptible  to UV  than  plants  grown outdoors.    Although  not  fully
understood, at least part of this  is due to  the incomplete development  of some
of  the natural  plant  protective  mechanisms against UV  damage   (Mirecki and
Teramura  1984).  Therefore it is necessary to examine whole plant responses in
the field whenever  possible.   Unfortunately,  due  to  the greater costs and
complexities   of field  experiments,  less than a dozen or so studies have been

     Table 1.    A Summary of the Effects of UV-B Radiation
                on Crop Growth
A.  Physiological/biochemical effects

    1.   Photosynthesis
          Hill reaction
          Electron transport
          RuBP carboxylase
          PEP carboxylase
          Dark respiration
          Photosynthetic pigments
    2.   Soluble proteins
    3.   Lipids
    4.   Carbohydrates
    5.   Non-photosynthetic pigments
    6.   Plant hormones

B.  Morphological/anatomical effects

    1.   Leaf area
    2.   Specific leaf weight
    3.   Epidermal transmission
    4.   Bronzing/glazing/chlorosis
    5.   Seedling growth/stunting
    6.   Dry matter production/allocation
    7.   Yield

C.  Response differences

    1.   Interspecific (species differences)
    2.   Intraspecific (cultivar differences)

D.  Environmental interactions

    1.   Visible radiation (photoprotection)
    2.   Water stress

Source:  Teramura (1983)

attempted  (Teramura 1983).    Currently  there  are only  three ongoing  field
studies:  two sponsored by  the  United States Environmental Protection Agency
(EPA), and one  European  study is being  sponsored  by  the West German Ministry
for Research and Technology (BMFT).

     While  there  is  a  need  for  field  validations  to  examine   whole  plant
responses, it is also essential  to  perform some key growth chamber studies to
understand the  underlying causes and  mechanisms  of UV  damage.   Otherwise we
will  have to  blindly  screen  plants  in the  field.    Controlled  environment
studies have  provided  us with  insights  into natural mechanisms  that protect
plants  from  UV  (Beggs, Schneider-Ziebert,  and Wellman  1986;  Wellmann 1975).
These protective mechanisms include a wide  assortment  of diverse biochemical
and morphological adaptations  that  collectively suggest that UV radiation has
been  a  strong selective  force  during  the evolution of  plants.   For example,
many  plants  produce  UV-absorbing  pigments  in  their   upper  leaf  surfaces,
preventing UV  from penetrating  into the leaf  (Robberecht  and  Caldwell  1978;
Robberecht, Caldwell,  and  Billings  1980;  Wellmann  1982).   This  is somewhat
analogous to suntanning in human skin.   Yet despite these and other mechanisms
against UV  damage,  some  plants  still  appear to be sensitive  to  even current
outdoor levels  of  UV  radiation  (Sisson and  Caldwell   1976;  Bogenrieder  and
Klein 1978).

     Experimental evidence  indicates that the effectiveness of UV radiation on
plants  is  not  necessarily linear  and  there  is not  always a  one-to-one
relationship  between  UV  dose  and damage.   It  is especially  clear  in  field
studies  that  UV  effectiveness  is  modified  by other  environmental factors,
including the pattern of precipitation, temperature during the growing season,
soil  fertility,  and others  (Murali  and Teramura  1986).   For  example,  under
water stress  or  mineral  deficiency,  plants are  less  susceptible  to UV injury
(Murali and Teramura  1985). Therefore, the greatest UV  damage is anticipated
under well-fertilized,  irrigated situations.

     Although it is probably the most important issue  concerning  the general
public, the effects of UV radiation  on crop yield  has only been examined in a
handful of studies (Teramura 1983).  In fact, detailed information only exists
for one or  two crops.   Field  studies  at the  University of Maryland (U.S.A.)
have found that a  25% ozone depletion  can  result in up to a 2Q%-25% reduction
in  total  soybean crop yield (Table 2).   Additionally,  UV radiation reduces
seed quality, in terms of reduced  seed protein and oils  (Teramura unpublished
data).  This  UV-induced  reduction  in  yield is in  addition to current losses
from weeds, pests and diseases,  insect attack, etc.  Some preliminary evidence
suggests  that UV might  also heighten  losses  in these  other categories (Esser
1980), effectively producing even greater yield reductions.  If these data for
soybean are  also  generally applicable  for  other key  crops,  such  as  rice,
wheat,  or corn,  then there is no doubt that we should  be greatly concerned
with  the  potential   consequences  of  ozone  depletion  upon  global   food

     In addition  to direct  effects  on  plants,  UV radiation may  also  play a
more subtle role by indirectly  affecting the manner or  degree by which plants
interact  with one  another (Fox and Caldwell  1978;  Gold and Caldwell 1983). A
simple example of  this might be  seen in the competitive  interaction between a
crop plant and its  associated weedy species.   UV radiation may produce a

         Table 2.  Summary of UV Effects on Soybean Yield and Quality

Year                 % change in yield               % change in seed quality

                                                      protein          oils


+ 1
Source:  Teramura (unpublished)

subtle change,  perhaps an  increase  in rooting depth  of the weed,  which may
then allow the weed to outcompete  the  crop plant  for water or nutrients.  The
net  effect of  such  a change  would  be  an  increase  in  weediness and the
associated problems.   In  a natural ecosystem, this  weediness would lead to a
proliferation of  UV-tolerant organisms at the  expense of  UV-sensitive ones,
which  potentially  means   that  the  distribution  of  species  may  be  quite
different in a UV-enriched future.

     Despite  our  increased  knowledge  of   the   biological   effects  of  UV
radiation, there  is still  an  enormous amount  of uncertainty  concerning the
consequences of ozone  depletion (Table 3).   For  example,  we,, know that there
exists a tremendous diversity of responses among different plant species to UV
radiation. Yet we still do not have a good understanding why these differences
occur.   Although  over  200 species and varieties  of  plants  have been screened
for  UV sensitivity,  90%  of these were crop  plants.  Therefore,  very  little
information  exists  on   the   impacts  of   UV  on  native  terrestrial   plant
communities,  which account for over 90% of the total productivity in the world
(Table  4).   Therefore,  to  address  the  larger  question of  the  potential
consequences of enhanced  levels of  UV-B  radiation on global  communities and
ecosystems,  we  must  currently make  the  unlikely  assumption  that perennial
woody  trees,  such  as oaks,  respond  in  an  analogous  fashion as herbaceous
annual agricultural species, such as spinach.

     Although we  recently have  obtained some  information  on the effects of UV
in  combination  with some other  commonly  experienced stresses, such as water
stress and mineral deficiency, we have virtually no information on the effects
of  UV  in  combination  with other  atmospheric change,  such as  increased con-
centrations of ozone,  sulfur dioxide,  or  carbon  dioxide.   Although UV radia-
tion  generally produces   deleterious  effects  on  plants,  increased  ambient
levels of  C02  have beneficial   effects  (Table 5); the combination of the two
may  not  necessarily be compensatory.   One alternative possibility is  larger
plants   resulting  from   CC>2  enrichment,   which  in  turn  would  still  be
deleteriously  affected  by  UV  radiation.   Without  experimental  evidence,
critical uncertainties such as this cannot presently be resolved.

           Table 3.  Current Uncertainties of UV Effects on Plants
        •  Action spectra
        •  Reasons for tolerance and sensitivity
        •  Effects on natural communities and ecosystems
        •  Combined effects with pollutants
        •  Effects in high ambient CC>2 environment
          Table 4.  Survey of UV Studies by Major Terrestrial Plant
                    Ecosystems  (after Whittaker 1975)
Tropical forest
Temperate forest
Boreal forest
Woodland and scrubland
Temperate grassland
Swamp and marsh
Desert and semidesert
Tundra and alpine
Global NPP
(109 ton/yr)
Total Area
(10b knT)
Included in
UV Study ]_/
]_/  Only studies examining some aspect of growth.

             Table 5.  Summary of UV-B and C02 Effects on Plants
Plant Characteristic
Enhanced UV-B
Doubling of C02
Leaf conductance
Water use efficiency
Dry matter produc-
tion and yield
Leaf area

Specific leaf weight

Crop maturity


Drought stress
 Decreases in many
 Co and C|| plants
 No effect in many

 Decreases in most

 Decreases in many
 Decreases in many

 Increases in many

 Mo effect
In Co plants in-
crease up to 100% but
in C4 plant only a
small increase

Decreases both in Co
and Cjj plants

Increases in both Co
and Ch plants

In Co plants almost
doubles but in Cjj
plants only a small

Increases more in Co
than in Cjj plants

 May inhibit or stimu-   Flowers produced
 late flowering in some  earlier
 Species may vary in
 degree of response
 Response varies among
Major differences
occur between Co
and Cjj plants

Response may vary
among cultivars
 Plants become less      Plants become more
 sensitive to UV-B but   drought tolerant
 not tolerant to drought
Source:  Lemon (1983); Teramura (1983)

     In summary, we know  that  large  differences  exist  among species and these
are  manifested  in  a  diverse   array   of  physiological,   biochemical,   and
morphological  responses.    We  know  that  for  whole  plant responses,  field
validation experiments are  essential because plants are more  sensitive to UV
when grown  indoors.  Key  growth  chamber studies  are necessary to  provide an
understanding of the mechanisms of UV sensitivity in plants.  UV effectiveness
can be modified by the prevailing microenvironment.  For example,  water stress
can completely mask any UV  effects  seen in the field.   Moderately large ozone
depletions can reduce crop  yields in the field.   UV can produce subtle shifts
that alter the competitive balance among'plants.


Beggs,    C.J.,  U.  Schneider-Ziebert,  and E.  Wellmann.    1986.   UV-B radiation
  and adaptive mechanisms in plants.   In Stratospheric ozone reduction, solar
  ultraviolet radiation and plant life,  eds.  R.C.  Worrest  and M.M.   Caldwell.
  Berlin:  Springer-Verlag.   ISBN 13875-7.

Biggs,    R.H.,   and  S.V.   Kossuth.  1978.   Effects of  ultraviolet-B radiation
  enhancements  under  field conditions.    In  UV-B biological  and  climatic
  effects research (BACER),  Final Report.

Bogenrieder,  A., and R. Klein.  1978.   Die  abhangigkeit der UV-empfindlichkeit
  von der lichtqualitat bei der  aufzucht  (Lactuca  satiya  L.).  Angew.  Botanik

Caldwell, M.M.  1981.   Plant  response  to solar  ultraviolet  radiation.    In
  Encyclopedia of Plant Physiology,  New Series, vol. 12A.  Physiological Plant
  Ecology. I, eds. O.L.  Lange, P.S. Nobel, C.B.  Osmond, and H.  Ziegler,  169-
  197.  Berlin: Springer-Verlag.  ISBN 0-387-10763-0.

Esser,    G.   1980.      Einfluss   einer   nach  Schadstoffimission   vermehrten
  Einstrahlung von UV-B-licht  auf  Kulturpflanzen,  2.  Versuchsjahr.   Bericht
  Battelle Institute.  V.  Frankfurt,  F.R.G. BF-R-63.984-1.

Fox,  P.M.,   and M.M.  Caldwell.  1978.    Competitive  interaction  in  plant
  populations  exposed  to  supplementary ultraviolet-B  radiation.    Oecologia

Gold, W.G.,  and M.M.  Caldwell. 1983.   The effects of  ultraviolet-B radiation
  on plant competition in terrestrial ecosystems.  Plant Physiol.  58:435-444.

Klein,  R.M.,  P.C.  Edsall,  and A.C. Gentile. 1965.  Effects  of near ultraviolet
  and green radiations on plant growth.   Plant Physiol. 40:903-906.

Lemon,  E.R.   1983. C02 and plants.   The  response  of plants  to rising levels of
  atmospheric carbon  dioxide.   Boulder,  Colorado: Westview  Press Inc.

Mirecki, R.M., and A.H.  Teramura.  1984.   Effects  of ultraviqlet-B  irradiance
  on soybean.  V.   The dependence of plant  sensitivity  on the photosynthetic
  photon flux density during and after leaf expansion.   Plant Physiol.  74:475-

Murali,  N.S.,   and  A.H.  Teramura.  1985.    Effects  of  UV-B  irradiance  on
  soybean.   VI.    Influence  of phosphorus  nutrition  on growth and  flavonoid
  content.  Plant Physiol.  63:413-416.

Murali, N.S., and A.H. Teramura.  1986.   Effects  of supplemental  ultraviolet-B
  radiation on  the  growth  and physiology  of field-grown soybean.   Env.  Exp.
  Bot. in press.

Nachtwey,  D.S.,   M.M.  Caldwell,   and  R.H.  Biggs.  1975.    Climatic  Impact
  Assessment Program  (CIAP),  Monograph  5,  eds.  D.S.  Nachtwey, M.M.  Caldwell
  and  R.H.   Biggs.   U.S.   Dept.   of   Transp.,   Report  No.   DOT-TST-75-55,
  Springfield,  Virginia:   Natl. Techn.  Info. Serv.

National  Academy  of Sciences.    1982.   Causes  and effects of  stratospheric
  ozone reduction:  An update,  339.  Washington, D.C:  National Academy Press,
  ISBN 0-309-03248-2.

Robberecht,  R.,  and M.M.  Caldwell. 1978.   Leaf  epidermal transmittance  of
  ultraviolet  radiation  and  its   implications  for   plant   sensitivity  to
  ultraviolet-radiation induced injury.  Oecologia. 32:277-287.

Robberecht,  R.,  M.M. Caldwell,  and W.D.  Billings, 1980.   Leaf  ultraviolet
  optical  properties  along a  latitudinal  gradient in  the artic-alpine  life
  zone.  Ecology. 61:612-619.

Sisson,  W.B.  1981.    Photosynthesis,  growth,   and   ultraviolet  irradiance
  absorbance of  Cucurbita  pepo L.  leaves exposed  to ultraviolet-B  radiation
  (280-315 nm).  Plant Physiol.  67:120-124.

Sisson, W.B. and  M.M.  Caldwell.  1976.    Photosynthesis, dark  respiration, and
  growth of Rumex patientia L.  exposed  to ultraviolet  irradiance (290 to 315
  nanometers) simulating a  reduced atmospheric ozone column.   Plant Physiol.

Steinmuller, D.  1986.    Zur  Wirkung ultravioletter Strahlung (UV-B)  auf die
  Struktur   von   Blattoberflachen   und   zu  Wirkungsmechanismen   ben   der
  Akkumulation  und  Biosynthese  der  Kutikularlipide  einiger  Nutzpflanzer.
  Ph.D. Dissertation.  Botanisches Institut II, Universitat Karlsruhe, F.R.G.

Teramura,  A.H.  1982.   The amelioration of UV-B  effects on  productivity  by
  visible  radiation.   In  The role  of solar ultraviolet radiation  in marine
  ecosystems, ed.  J.  Calkins,  367-382.  New York:  Plenum Pub.  Corp.,  ISBN 0-

Teramura,  A.H.  1983.  Effects of  ultraviolet-B radiation on the  growth and
  yield of crop plants.  Physiol Plant.  58:415-427.

Teramura,  A.H.,  R.H.  Biggs,  and  S.  Kossuth.  1980.  Effects  of ultraviolet-B
  irradiances  on  soybean.    II.     Interaction  between  ultraviolet-B  and
  photosynthetically   active   radiation   on   net   photosynthesis,   dark,
  respiration, and transpiration.  Plant Physiol.  57:175-180.

Teramura, A.H.,  M.  Tevini,  and  W. Iwanzik.  1983.   Effects  of ultraviolet-B
  irradiation on  plants during  mild  water  stress.   I.   Effects  on  diurnal
  stomatal resistance.  Physiol Plant. 57:175-180.

Tevini, M.  and  W. Iwanzik.  1982.   The effects of UV-B  irradiation on higher
  plants.   In  The role of solar ultraviolet radiation  in  marine ecosystems,
  ed. J. Calkins,  581-615.  New York:  Plenum Pub.  Corp.

Tevini, M., and  W.  Iwanzik, 1982.  Untersuchungen uber  den Einfluss erhohter
  UV-B Strahlung  auf  Entwicklung, Zusammensetzung,  Struktur  und  Funktion von
  Pflanzen.  Munchen:   Bereich Projekttragerschaften,  GSF.

Tevini, M., W.  Iwanzik,  and U.  Thoma.  1981.   Some   effects  of enhanced UV-B
  irradiation on the growth and composition of plants.   Planta. 153:388-394.

Van,  T.K.  and  L.A.   Garrard,  1975.    Effect  of  UV-B  radiation  on  net
  photosynthesis  of some  Co  and Ch  crop  plants.  Soil Crop  Sci.  Soc.  Fla.
  Proc. 35:1-3.

Van, T.K.,  L.W.  Garrard,  and  S.H.  West,  1976.   Effects  of UV-B  radiation on
  net photosynthesis of some crop plants.   Crop Sci.  16:715-718.

Wellmann, E. 1982.  Phenylpropanoid pigment  synthesis and growth  reduction as
  adaptive  reactions  to  increased UV-B radiation.   In  Biological  effects of
  UV-B radiation, eds.  H.  Bauer,  M.M.  Caldwell,  M.  Tevini and R.C. Worrest,
  145-149. Munchen:  Gesellschaft fur Strahlen und Umweltforschung mbH.

Wellmann, E. 1975. UV  dose-dependent induction of enzymes related to flavonoid
  biosynthesis in cell suspension cultures of parsley.   FEES Lett. 51:105-107.

Whittaker,  R.H.   1975.     Communities and  ecosystems.   Mew  York:  MacMillan
  Publishing Co., Inc.

The Effect of Solar UV-B Radiation on
Aquatic Systems:   An Overview
R. C. Worrest
Oregon State University
Corvallis, Oregon USA

     Various  experiments have  demonstrated that UV-B radiation causes damage
to fish  larvae and juveniles,  shrimp larvae, crab larvae,  copepods, and plants
essential  to  the  aquatic food  web.   These  damaging  effects include decreases
in fecundity,  growth,  survival,  and  other  functions of these organisms.   In
natural  marine plant communities  a change in species  composition rather than a
decrease in net production would probably  result  of enhanced UV-B exposure.
The change in community composition  could  result in  a more unstable ecosystem
and would  be  likely to have an influence on higher  trophic levels.  A decrease
in  column ozone  would  diminish  the  near-surface  season   of  invertebrate
zooplankton populations.   Whether  the population  could endure a significant
shortening of the surface season  is unknown.

     The direct effect of UV-B radiation on food-fish larvae closely parallels
the effect on invertebrate zooplankton.  Information is required on seasonal
abundances and vertical distributions of  fish  larvae,  vertical  mixing,  and
penetration of UV-B radiation  into appropriate water  columns before effects of
incident or increased  levels  of  exposure to UV-B  radiation can be predicted.
However,  in  one  study  involving anchovy  larvae  a 20  percent  increase  in
incident biologically  damaging UV-B  radiation  (a result of about a 9 percent
decrease  in  the  atmospheric  ozone  column) would  result in  all  the larvae
within a ten-meter mixed layer in April and August being  killed after fifteen
days.  It  was calculated that about  8 percent  of the annual larval population
throughout the entire water column  would  be  directly  killed  by  a  9 percent
decrease in column ozone.


     Aquatic  environments include shallow freshwater  ponds, lakes and streams,
fresh and  saline inland seas, brackish-water  marshes,  swamps and estuaries,
and  a wide variety of  marine habitats.    The  marine  habitats  extend  from


inshore coastal regions, surf zones, and  fringing  reefs  to  the wide spaces of
the open sea  and  the abyssal deeps.  Inland waters  cover less than 1  percent
of the earth's  surface,  while the oceans extend over more  than 70 percent of
the earth's surface.

     Because of the relatively small volume of inland waters,  freshwater plant
production  contributes  only a  few  percent to  the total global  productivity
(Table 1).   Marine  systems,  covering a  vast area,  contribute  significantly to
global  productivity—an  amount  equivalent   to  half  the  terrestrial  plant
production.  The coastal kelp forests have productivities said to rival those
of tropical rain  forests,  generally considered  the world's  most productive
plant ecosystems (Mann  1973;  Lieth  1975).  Of the five  major  marine habitats
considered  in  Table 1,  the  microscopic plants  of the open sea are  the most
productive,  contributing  75  percent of the  marine primary   production,  24
percent of the global total.

     Photosynthesis,  the  conversion  of  radiant  energy from  the sun  into
chemical  energy  (food),  is  the  basis  of  almost  all  life   on   the  earth.
Photosynthesis requires light,  but adequate sunlight seldom penetrates to the
bottom of natural waters; thus, aquatic photosynthesis is largely confined to
a  relatively  thin   layer  at the water  surface  termed  the  euphotic zone,  the
surface layer where plants  create more  chemical  energy by photosynthesis than
they use for their own metabolism.

     Organisms  in  fresh water  or the  oceans are  either  swimmers (nekton),
bottom-dwellers  (benthos),   or  drifters  (plankton).  Plankton  can be  either
plants  (phytoplankton)  or  animals  (zooplankton).    An   important  zooplankton
group, the  icthyoplankton,   is  comprised  of the  drifting  eggs and larvae of
many  species  of fish.   Zooplankton in general  contain  nearly all groups of
aquatic animals, at least for  some phase in  their life history,  such as the
egg and/or  larval  stage.    Zooplankton  are found at all depths,  depending on
species and season, but are  most  abundant in  the upper  100  m sunlit where the
bulk of their food,  including phytoplankton,  is  found.   As  feeding progresses
from  the   phytoplankton  through  the  zooplankton  to fish  and other  higher
organisms   (the   food  web),   the   chemical   energy  generated   by   aquatic
photosynthesis  becomes  more  and  more  concentrated  into organisms of larger
size, reaching organisms suitable for harvesting for  human use.

     Marine ecosystems are highly dynamic and of bewildering complexity.  Many
of the fundamental processes affecting the abundance  of  marine resource stocks
are poorly  understood.  For  example,  the  complexity  of  food webs in the ocean
appears to  be much  greater than for  terrestrial  populations  (Larkin 1978).
Little is known in a quantitative sense about marine  ecosystems, although some
geographical areas are currently under  intensive investigations.

     Many  other sources  of complexity  in  marine  population  dynamics exist.
Simple density-dependence mechanisms,  for example,  can theoretically result in
so-called  "chaotic"  oscillations  in population  abundance (Levin and Goodyear
1980).   Most fish  stocks  undergo  significant  irregular  fluctuations,  the
underlying  causes   of  which  are  not well  understood.    In  most  cases both
density-dependent and density-independent effects are probably  involved.

     For any  given level of fish mortality,  the  productivity or  equilibrium
yield  of  a  stock  depends  upon growth,  recruitment,  and  natural mortality.


Table 1.  Representative Primary Production and Plant Biomass Summaries of the
          Earth (adapted from Whittaker and Likens 1975)
Ecosystem Type
(millions of
                                             Net Primary
                                           (grams per square
                                           meter per year)
               Mean Plant
             (kilograms per
              square meter)
  Tropical rainforest
  Tropical seasonal forest
  Temperate evergreen
  Temperate deciduous
  Boreal forest
  Woodland and shrubland
  Temperate grassland
  Tundra and alpine
  Desert and semi-desert scrub
  Extreme desert, rock,
    sand and ice
  Cultivated land
  Swamp and marsh
  Lake and stream

Percent of Worldwide

  Open ocean                   332.0
  Upswelling zones               0.4
  Continental shelf             26.6
  Algal beds and reefs           0.6
  Estuaries (excluding marsh)    1.4

Total/Average                   361
Percent of Worldwide             71$










Growth is relatively easy to measure and understand; recruitment is relatively
easy to  measure but  difficult  to understand;  and natural mortality  is both
difficult  to  measure  and  difficult  to  understand.   The phenomena  that are
difficult  to  measure  or   understand are  usually  masked  by  the  standard
productivity analysis.

     As early as 1925,  scientists were aware  of the damaging effects from the
ultraviolet component  of sunlight even  oq  aquatic organisms  (Huntsman  1925;
Klugh  1929,  1930;  Harvey 1930; ZoBell and  McEwen 1935;  Giese  1938;  Bell and
Hoar 1950;  Dunbar  1959; Marinaro and Bernard  1966).   It was shown  at these
early dates that different species have various sensitivities to UV radiation,
and that this differential  sensitivity might  relate  to the depth at which the
species were found.

     There  have  been  several  more  recent  studies on  the  effects  of  UV-B
radiation on a variety  of marine organisms (WAS  1979; MAS  1984).   Regardless
of species investigated, each study has potential  importance  through  the role
of the  particular  organism  in  its environmental or food  web context.   Some
studies,  in  addition,   have considered  economically  important  zooplankton
species,  such  as larval  stages of certain  shrimp, crab,  and  fish.    Recent
literature reviews  (Calkins  1982;  Worrest 1982,  1983a,b) have summarized the
results of most of these studies.

Physical Effects  on Organisms

     UV-B radiation is readily  absorbed by proteins and  nucleic  acids,  and is
effective in inducing photochemical reactions  in  plants  and animals (Caldwell
1971; Murphy  1975;  Giese 1976).   Even  though proteins and nucleic acids are
commonly  involved   in  biological  responses  to  UV-B  radiation,   the  action
spectra for tissue damage in many  organisms may differ because of wavelength-
dependent  refraction,  reflection,  or absorption,  and  hence protection,  by
outer tissues (Cheng,  Douek,  and Goring  1978).

     The  usual effects  of environmental  stress (e.g.,  pollution)  are changes
in overall  productivity and  reduction  in  species diversity.   Diversity  is
associated with stability in ecosystems,  allowing alternate routes and choices
within  food  webs.   In field  situations  under slight stress  one  often  cannot
measure changes in productivity  or size  of specific populations,  but sometimes
changes   in   species   diversity   can   indicate  that   adverse   effects  are
occurring.   With  loss   of  species an  ecosystem  loses  some  of  its  natural
resiliency and flexibility.   In aquatic  ecosystems we must  consider a great
number  of species  with  different  life  stages  and different trophic  levels.
With  increased  knowledge   of   ecosystems  and  the   physical   and  chemical
environments, the  effects of enhanced  solar UV  radiation  on single  species
could be placed in perspective.

     As already noted,  the biological  response to ultraviolet  radiation can be
very wavelength-dependent. This wavelength specificity makes it  necessary to
develop weighting  functions  to  express ultraviolet radiation in biologically
meaningful terms.   Biological action  spectra normally serve  as  the basis for
these weighting  functions.    The weighting function is  normally  taken  as an
action  spectrum  for a  particular response.   Several action spectra have been
developed  describing  the  response of  biological  systems  to UV-B radiation
(Caldwell et al.  1986).   Examples are a  representation  of  DMA  response  (Set-
low  1974), and a generalized plant response (Caldwell 1968).

     Based on an analytic characterization of ultraviolet sun- and skylight, a
10 percent  ozone  reduction  at 45°N latitude  would  result  in  only a 1 percent
increase  in total solar ultraviolet (290-360 nanometers)  daily exposure at the
surface  of the  earth.    This  would  be  of minimal  consequence  if radiation
throughout   the  solar   spectrum  were   of  equal  biological  effectiveness.
However,  when the  biological weighting  functions based on  the action spectra
are  employed,  a  very different picture  emerges.   Based  on DNA  damage,  a 10
percent ozone reduction  would result  iq  a 28 percent increase in biologically
effective radiation (Table 2).   The generalized plant  response  (damage)  would
increase  by 21 percent.

        Table 2.  Relationship Between Ozone Depletion and Biological
                  Effectiveness of Increased UV-B Radiation.
Increase in
290-320 nm
UV Radiation
290-320 nm
Increase in
aAction spectra are referenced to 300 nm - 1.00.

     The increased exposure to biologically effective (DNA,  plant) ultraviolet
radiation resulting  from a 10 percent  decrease in stratospheric  ozone would
correspond to  migrating  over  30° toward  the equator—a change  of ecological
significance.  Levels of UV-B irradiance  vary  latitudinally,  with the highest
exposures in  the  tropics.   The  current difference  between  the  extremes  of
exposures is  about three- to sixfold  but biota are presently adapted  to the
levels that are normally experienced at their current locations.

     There is increasing evidence that marine ecosystems change as a result of
natural as well as events  induced by  human  activities.   Many  of these changes
occur on time scales of tens of years.  The past changes may be associated,  in
some  general  context, with  gradual  climatic  trends,  but  ecological  changes
seem  to  occur quite rapidly,  with  relatively  persistent  communities existing
in the  intervening  periods.  In  any  region,  rather  than  having one ecosystem,
there are two, or possibly several,  alternatives—each resilient to some range
of  variability  but  capable   of  being  replaced  if  some  factors  in  the
environment pass a  threshold.   To this natural episodic  response  we must now
add the effects of anthropogenic perturbations.  The evidence  suggests that we
will  continue to  have  such  changes but  will have  additional questions  of
attributing  cause  to natural  stresses or  to  stresses  resulting  from human
activities.    Although we may not be able to  predict  when, or  possibly why,
such  changes   occur,  they  can   be  regarded  as  alternative   ecological
solutions.   Once  again, whether  we consider these  changes deleterious  or
beneficial is  a  matter of human values,  but we must keep  in  mind that these
problems are occurring on a global scale.


     The amount of  UV-B  radiation  reaching the ocean's surface  has long been
suspected  as  a  factor  influencing  primary  production.     Research  shows
convincingly that  ultraviolet radiation,  at levels currently  incident  at the
surface of natural  bodies of water,  has  such  an  influence (Steemann Nielsen
1964; Jitts,  Morel,  and Saijo  1976;  Hobson and Hartley  1983).    It  has been
calculated  that  near the  surface  of  the  ocean,  enhanced  UV-B  radiation
simulating  a  25  percent  reduction  in  ozone concentration  would  cause  a
decrease in primary productivity of about 35 percent  (Smith et al. 1980); the
estimated reduction  in production for the whole euphotic zone  would be about
      These  calculations are  based  on attenuation  lengths, i.e.,  the product

of depth  in  the water column  and  the diffuse attenuation coefficient  of the
water.   Therefore,  waters  of various  turbidities  and absorption  character-
istics can be compared.

     Effects induced by  solar  UV-B  radiation have been measured  to depths of
more than twenty meters  in  clear waters and more than five  meters  in unclear
waters.   The  euphotic  zone  (i.e.,  those  depths with light  levels  sufficient
for net photosynthesis  to be positive) is  frequently taken as the water column
down to the depth to which  1  percent  of the surface photosynthetically active
radiation  penetrates   (Figure  1).    In  marine  ecosystems,  UV-B  radiation
penetrates approximately  10  percent of  the upper marine  euphotic zone  before
it is reduced to 1 percent  of  its surface  irradiance (Figure 2).   Penetration
of UV-B  radiation into  natural waters  is a  key variable  in assessing  the
potential impact of this radiation on  any  aquatic ecosystem.

     If one assumes  that present phytoplankton populations  sense and control
their average vertical position  in  such a way as to limit UV-B  exposure to a
tolerable level,  then  a 10 percent  increase in solar  UV-B  radiation would
necessitate a downward movement  of  the  average position  sufficient to  reduce
the  average  UV-B exposure  by  10 percent.   There  would  be  a  corresponding
reduction of  light  for  photosynthesis;  the loss  of  photosynthetically  active
radiation is  the primary  limiting  factor  for marine productivity (Russell-
Hunter 1970).   The  loss  of photosynthetically active radiation  from optical
measurements has been estimated  to  be in  the range of 2.5 to  5  percent for a
10  percent  increase  in   UV-B  radiation   (Calkins,  Volume  2).     If  the
photosynthetic  base  of  aquatic ecosystems  were perturbed  one  would  expect
ramifications  to  extend  up  through  the  food  web  through  predator-prey

     Experiments  with  marine  diatoms  have shown  significant reductions  in
biomass,   protein,  and  chlorophyll  at  UV-B  irradiances  equivalent  to ozone
reductions ranging from 5 to 15 percent   In addition, studies on  chain-forming
diatoms and other phytoplankton in  the laboratory showed  that increased growth
occurs when  the UV-B radiation  is  filtered out of the incident  solar  radia-
tion,  indicating  that  existing levels of UV-B radiation  depress productivity
(Thomson, Worrest, and  Van  Dyke 1980;  Worrest 1982).   Furthermore,  indirect
effects  of  ambient  levels  of  UV-B  radiation  endanger  the   survival  of
microorganisms  (Euglena, slime  mold,  some blue-green algae) by  decreasing
their motility and by inhibiting phototactic and photophobic  response (Haeder,
see  Volume  2).    This   reduces  the  ability  of a  population  to  move  into
favorable environments,   which would result  in  damage  that may  impair develop-

     Direct  measurements of photosynthesis  that  span   a  single  day  could
underestimate the overall action of  solar  UV exposure by failing  to account
for the next day's population level.  Growth delay or direct  killing may cause
the  subsequent  population   to  fall  below  the  numbers  that  an  unexposed
population would attain.   Prolonged delays (about two days)  in  growth of the
survivors have been observed when two strains of the diatom Thalassiosira were
irradiated with  simulated solar  ,UV-B  radiation at doses  below lethal levels.
If unicellular organisms  are  in  a rapid growth phase,  a  growth delay equaling
the  time of  one growth cycle  produces  the  same  effect  on the  subsequent
population as would be produced by 50 percent killing (without growth delay).

                                               0.5 L
                               — Dc, Compensation depth (PC=RC)
                                  Ic, Compensation light intensity
                            a	Dcr, Critical depth

                               Dm, Depth of mixing
        Figure  1.  The Relationship Between the Compensation Depth or Euphotic
                   Depth,  the  Critical  Depth,  and the Depth of Mixing
            600 r
          o 500
          9 400
          - 300
          jZ 200
2   4  6810  20  4060100,
               0.2  0.4   0.6   0.8   1.0  1.2   1.4   1.6  1.8   2.0   2.2  2.4
     Figure 2.   Solar spectral  irradiance at the surface  of the ocean and at
four depths  (adapted  from Sverdrup  et  al.  1942).   Insert  illustrates per-
centage of energy transmitted at depth  in oceanic waters.

     In Hawaii, long-term photoinhibition of growth of six algal species under
natural sunlight  was  measured by Jokiel  and York (1984).  Two  strains could
not  grow  at  all at  the  levels  of  irradiance for  full  natural  surface
sunlight.  Of  the species tested, those  collected from  tropical surface water
showed  the  greatest  adaptive  power,  but  it is  reasonable  to  conclude  that
resistance to  solar UV radiation  is achieved  through  expenditure of resources
that can be better applied to other  needs in less-exposed  species.

     In   simulated   marine  ecosystems,   enhanced  UV-B  radiation   levels,
simulating  15  percent decreases  in ozone,  result in  shifts  of  the  species
diversity  and  community  composition  of  phytoplankton  communities  (Worrest,
Thomson, and Van Dyke  1981).   It  appears  that in  natural  communities a change
in  species  composition rather  than a decrease in net  production might  be a
more  likely result  of  an  enhanced  UV-B  exposure.    The decreased  species
diversity observed in  simulated field studies  is usually not  accompanied by
deleterious effects on biomass and chlorophyll  accumulation or  by deleterious
effects  on  total community  primary   productivity.     However,  a  change  in
community composition might result in  a more unstable  ecosystem and might have
influence on higher trophic levels (Kelly,  Volume 2).   One effect of enhanced
levels  of UV-B  radiation would  be  to  alter  the size  distribution  of the
component producers in a marine ecosystem.   Increasing  or decreasing the size
of  the  representative  primary   producers  upon  which  consumers  graze  can
significantly  increase the energy allotment required  for  consumption,  thereby
reducing  the  feeding  efficiency of  the  consumer.    In  addition,  the  food
quality of  the producers is  altered  by exposure  to  UV-B radiation.   It has
been  demonstrated   that  the  protein  content,   dry  weight,   and  pigment
concentration are all depressed by enhanced levels of  UV-B radiation.


     Zooplankton  are  critical   components  in  typical   aquatic  food  webs
(nutrient pathways)  that lead  to larger animals, including those comprising
commercial  fisheries  (finfish and shellfish)  and therefore humanity  itself.
With respect to  potential  impacts of  enhanced  solar  UV-B radiation,  only the
zooplankton living in  daylight in the upper  several  meters would be directly
affected.   Many  zooplankton  species normally live very close  to the surface,
even in daylight, while  others occupy  the  near-surface  layer during only part
of their life  cycle.

     The  near-surface  layer  is a very  important  zone  in  the  interactions of
the  physical,   chemical,  and  biological   components   of  aquatic  systems.
Zooplankton have apparently evolved  mechanisms and behavior by  which they have
adjusted to current levels of UV radiation  (Damkaer 1982), but  they may not be
able to  adjust to relatively  rapid increases.   If there  were  changes in the
abundance  of  zooplankton  species,  those  changes would  have  an  impact far
beyond  any direct  effects because of  the critical  role of   zooplankton in
energy transfer within the ecosystem.

     Karanas,  Van Dyke, and Worrest  (1979,  1981) presented evidence that acute
exposure of a  marine  invertebrate zooplankton resulted  in reduced survival of
the  organisms.  They  also showed that exposure  of this  species to sublethal
exposures of  UV-B radiation could also  reduce the fecundity  of the parents.
Thomson  (Volume  2) demonstrates that  current levels  of UV-B radiation are of
significance in  the  developmental life of  several species of  shellfish.  For


some species, a  10  percent decrease in column ozone could  lead  to as much as
an 18 percent increase in the number of abnormal larvae produced.

     Spawning of zooplankton  (and  fish)  is usually timed so  that survival at
hatching is  optimized.   Compared  to  birds and mammals, the  lack of internal
temperature control, such  as  prevails in most aquatic  animals,  makes ambient
temperature  exert  overriding  and continuous  accelerating  or  decelerating
influences on metabolic functions and, hence, on the growth of aquatic animals
at whatever stage.

     Before feeding by  zooplankton can begin, variations in  exposure to UV-B
radiation may cause stresses at various  stages of egg and very early larval
development, causing mass  mortalities  or  an alteration in  time  of occurrence
at the  surface.   The  matching in  time of  the onset of  larval feeding and the
spring phytoplankton bloom—which,  in turn, depends on climatic events for its
timely  occurrence—will,  in  fact, determine  the  sizes  of  year  classes  of
grazers (Gushing 1972).

     Fish stocks suffer most  from  the  vagaries  of climate  through the effects
that  changes in  ambient  conditions  have  on  the  larvae.    The  biological
dynamics at the  larval  stages in  the  life  cycles  of  fishes  and  aquatic
invertebrates have  a short  time  horizon  and are  highly  variable.   Instant
mortality coefficients  during the  larval  stages range  between  300  and 400,
while  they  are  0.2  to  1.5  in  the   juvenile  and adult  stages  (Rothschild
1981).  The  most  catastrophic and frequent  event that  can  befall larvae is a
mismatch in  time  between  their  food  needs  and  the prevalence  of their food
(Gushing 1972).    Larval  stages  last only a  few days. Small as  the biomass of
fishes is during very early life,  their future  numbers  are  largely determined
then and not later in life, when the biomass is large   and  when the commercial
fishery takes its toll.

     Regardless  of  the  cellular-level  responses  to  UV  irradiation,   it  is
usually noted that up to some level of UV exposure there is no apparent effect
on the organism  (Damkaer et al.  1980;  Damkaer and Dey  1982).  At greater doses
either the  repair systems  themselves  may become  inactivated  by  the radiation
or the  damage  to the general tissues  goes beyond the  capacity  of the repair
systems  (Damkaer  and  Dey  1983).   For such  impacts to occur  these threshold
levels  probably  must  be  exceeded  during  several consecutive  days.    The
apparent thresholds are near current levels.  The thresholds for all groups in
one test appeared to  be above the present median solar incident UV levels at
the test location, at least until  late in  the time span of natural occurrence
near  the  surface (Damkaer,  personal  communication).   This  season  could  be
significantly shortened by  a  20 percent ozone reduction.   Whether or not the
populations  could  endure  with  a  drastically  reduced  time  of near-surface
occurrence is not known.  Success of any year-class depends on the timing of a
great  number of  other  events   in addition  to  levels of exposure to UV-B
radiation (Hunter, Kaupp,  and Taylor 1982).

     It  has been  thought  that  the  superabundance  of eggs  spawned by most
marine  organisms  make. egg  production  an  irrelevant   factor   in population
control.  The superabundance of marine eggs, however,  is of clear evolutionary
significance and must interrelate with the variability of population size.

     The superabundance of eggs could have any number of  purposes.   First, it
could satiate predators,  thus  enabling  a certain fraction of  eggs  and larvae
to  slip through  the  predatory  network.   It could  also maintain  ecosystem
integrity and metabolism by directly serving  as  food for other  organisms or by
being  degraded   into   various  essential  nutrients,   sustain  cannibalistic
species,  or  contribute  to  a  patchy  distribution of  organisms,  which  is
apparently essential to marine ecosystem structure.

     As fish stocks decline in size, the biomass  of  eggs decreases at a faster
rate than  recruitment;  the ordinary functions of  the superabundance  of eggs
become  stressed   and  modified.   But  stocks compensate  to some extent  for
biomass decline  by growing faster,  tending  toward  an  earlier maturity,  and
increasing in fecundity.

     Hunter,  Kaupp, and Taylor   (1982) exposed anchovy  eggs  and  larvae  and
mackeral larvae  to high doses of  UV-B  radiation in small  closed containers.
These authors provided  data that  indicated  that anchovy  larvae  off southern
California  are   typically  centered  in  moderately  productive  ocean  water.
Baker, Smith and Green (1981)  calculated surface  DNA-effective  UV-B irradiance
levels  expected   for  this area,  and Smith  and  Baker  (1979)  calculated  the
penetration of UV-B radiation  into  the moderately productive ocean  water.   An
analysis of  Damkaer (personal communication) indicates  that  14.7  percent of
the anchovy  larvae would die  in  four to ten days  if there were no vertical
mixing of the larvae.   These  static situations no doubt approach "worst case"
scenarios.   However,  static  situations are  not likely  because of physical
vertical mixing  of water (through  the  action of wind, density  currents,  and
tides) and  the active  and passive movement of fish  larvae.  If  some vertical
mixing did not offer  relief from the effects of UV-B radiation,  it is likely
that even  today  there would  be  no anchovy larvae  above  a few meters depth,
which is clearly  not the  case.  The surface  mixed  layer probably extends to a
depth of  at  least twenty-five meters  from  December  to  February  (Sverdrup,
Johnson,  and  Fleming  1942).   A  mixed  layer of at least  ten meters  can be
assumed for the rest of the year.

     For the mixed situation in February, the larvae would spend  less than one
hour  each  day  in  the  upper  one  meter,  resulting  in  exposures   where  the
threshold doses would never be reached in the larval stage.   In  the February
model, with mixing to twenty meters, there would  have to be a sixfold increase
in  incident UV-B  radiation  before threshold  doses would  be  reached.  For the
months from March to  October,  with mixing to ten meters,  the  threshold doses
are also not reached within the twenty days available before the  larvae become
vertically migrating.   Only with  the  highest  surface irradiances  of April and
August are these doses approached within this available time (about twenty-two

     From March to October, with ten-meter mixing and a 20 percent increase in
incident  UV-B radiation  (a  result of  about  a 9 percent decrease  in  the
atmospheric ozone  column)  the  depth of the threshold dose-rate  is increased.
With the 20%  increase in  incident  irradiance  for the dose/dose-rate threshold
model, all of the anphovy larvae within the ten-meter mixed layer in April and
August would be killed, the threshold doses being achieved after  fifteen days.

     Apparently  at  all  months  about  36  percent  of  the  larval  anchovy
population is less than ten meters deep (Hunter,  Kaupp,  and Taylor 1981).  The


greatest numbers of anchovy  larvae  are  found below the surface in April.  (20
percent of  annual  population), so  that 7.2 percent of  the annual population
would be eliminated with a 9 percent ozone decrease.   Only 1.9 percent of the
annual population  is  present  in  August,  so the  loss  of those less  than ten
meters deep would  amount  to  0.7 percent of the  annual population.  The total
predicted loss, then,  due to a 9 percent decrease in total ozone column, would
be about 8 percent of the larval anchovy population (Figure 3).

     Because   of   complex   interactions   between  mixing   depths,   vertical
distribution of larval  anchovy,  seasonal changes in solar  irradiance and the
penetration of UV  radiation  into seawater, and  seasonal changes  in anchovy
abundance, there is no  linear  relationship between  mixing depth and predicted
annual loss of anchovy larvae.

     In addition to the direct effects upon the fisheries, it is possible that
with  less  primary production  of  organic  biomass  there  would   also  be  an
increased mortality in  larval  fish  due  to  food limitation.   And perhaps there
would be a synergistic effect on mortality in that some animals could die from
direct exposure, some from lack of  food,  and some from  the  combination  of a
reduction in food and  the weakening from exposure (or could be weakened and be
out-competed by other  fauna for limited food reserves).

     The  impact  on marine  fisheries  as  a food  supply to  humans  would  be
significant if the species of  phyto-  and zooplankton  that adapted to enhanced
levels of UV-B radiation were  of  different nutritional value (if  they altered
growth and fecundity of the consumers).   If the indirect impact of suppression
upon consumers were linear,  a  5 percent reduction of primary production would
result in  a 5 percent  reduction  in  fish  production.  A  question  still under
investigation is whether these relationships might be nonlinear.  For example,
there may  be  an amplification factor  involved  that  results  in  a relatively
greater impact at  higher  trophic levels  than at the  primary producer  level.
As illustrated  in  Table 3,  a  5  percent reduction  in  primary production (P1)
would give  an annual  yield  of  115 million metric  tons of  fish flesh  (a  5
percent reduction), whereas  a 5%  reduction in  the ecological efficiency  of
energy transfer (e) would  yield  103 million tons of fish flesh (a 14 percent

Potential Consequences to World Fisheries

     Well over 95 percent of the  sea's  living  resources  occur within belts of
ocean two  hundred  miles wide, surrounding  the coastal nations of the  world.
By international agreement each  nation exercises economic  jurisdiction  over
these Extended Economic  Zones (EEZs).    With  the establishment  of the  two
hundred-mile EEZs,  the  coastal nations  of the world have acquired an area of
sea surface amounting to about 25 percent  of the globe.   More than 99 percent
of the global  catch of marine  fisheries is currently taken  within  the EEZs.
These fish  are of  crucial importance  in world nutrition, making  up  about  18
percent of  the average world animal protein  intake.    That figure  rises  to
nearly 40  percent   in  Asia  where  more  than  half of  the  world's population
lives.  The  people of  Japan,  for  instance, derive about 60  percent  of their
intake of  animal protein  from the  ocean  (Clark 1981).   The  annual  income  to
the world's  fishermen from  marine  catches  is  roughly US$10 billion,  with  a
world fish catch of about  seventy  million metric tons.

                Larval  Northern Anchovy
           0      10     20     30     40      50      60     70
                   INCREASED  UV-B  RADIATION  (%)
    Figure 3,.   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  the vertical mixing models (based on data of
Hunter, Kaupp, and Taylor 1981, 1982).

     Table  3.   Annual  fish  production  in  coastal  waters and  impacts  of 5
percent reductions  in  primary production and ecological  efficiency of energy
transfer.   Baseline data for coastal waters adapted from Ryther (1960).
                                          Baseline      -5% P1       -5% e

Primary production, kJ/m2yr                 4200         4000        4200

Ecological efficiency per trophic
     level, (e)                             0.15          0.15       0.14

Number of trophic levels from plant
     production to fish production, (n)     3             33

Fish production, kJ m2yr                   14            13         12

Total fish production, millions of
     metric tons per year (using
     Winberg's transformation,
     4.2 kJ = 1 g fish flesh)               121           115        103

Percentage of baseline fish production     (100$)        (95%)      (86$)

     To address  the  economic impact  of  enhanced levels of  UV-B radiation on
global fisheries the principal concepts that need to be considered are supply,
demand, and the economic marketplace.  Supply  is the relationship between the
quantity of  a commodity that will be produced  and  the price of the  output.
For fish, the supply curve is heavily dependent on the biological productivity
of the stock  as well as economic  factors, because  supply  is derived from the
average cost  required  to  produce a pound of fish.  Supply  is also affected by
the fish or  uncertainty  associated with the various  costs  or availability of
the raw material or  other economic components  of the supply curve.  Demand is
the  relationship  between  the   price  consumers  are  willing  to pay  for  a
commodity  and  the  quantity  of  the  commodity  that  is available.    Global
production of fish has been predicted to be 92.5 million tons in the year 2000
(Robinson  1979).    Based on  per  capita  consumption  of  fish and  population
growth,  the  projected global demand  for  fish  in the  year  2000  will  be 97.1
million tons.  The economic marketplace is the  abstract setting where producer
interests in  the  form of the supply  curve are brought together with consumer
interests in the form of the demand curve.

     The Peruvian  anchoveta  fishery  is  an example of  the economic  impact
resulting from a stress on a fishery.  The fishery developed in the 1960s into
the world's  largest single  fishery.  Since the  export of  anchoveta  fishmeal
constituted Peru's second largest  source  of foreign exchange,  the loss of the
fishery had a major influence on the Peruvian economy.

     In  1970,  Peru's anchovy harvest amounted  to  12.3 million  metric tons,
more than  20 percent of  the  total world marine finfish catch  for  that year.
These  fish  rely on  phytoplankton  blooms  that  result  from the  surfacing of
colder nutrient-rich water.    In  1972,  the  Peruvian catch  plummeted  to  4


million metric tons and fell further  in  1973  to  1.8 million metric tons.  The
primary cause  of  this  drastic collapse  was  the  occurrence of  an El  Nino,
although heavy fishing  aggravated  the condition (Bardach and Santerre 1981).
The economic consequences to the fishery were catastrophic.   The  major use of
the anchoveta harvest was for  the production of  fishmeal, which was used as a
major source of  bulk  protein  in livestock feed.  A  shortfall of  the fishmeal
in the world's markets had severe ripple effects on other animal  feeds and on
livestock  prices.   The price  of  fishmeal  rose  127  percent,   the price  of
soybean meal  rose 99  percent, and  the  cost of pork,  for example,  rose  96
percent, adjusted for  inflation.

     The recovery  of  a depleted fishery population may require many years,
even if catches  are  prohibited.  Economic  survival of  the  existing resource
industry may,  however, depend upon continuing catches, even  though  these will
delay rebuilding  of   the  stock and  perhaps  increase  the  probability of the
population's collapse.  In the language  of the mathematical  theory of games,
common-property   resource  exploitation   has   the   characteristics   of  the
prisoner's dilemma.


Baker,  K.S., R.C. Smith, and A.E.S. Green. 1981. Middle  ultraviolet radiation
   reaching the ocean  surface,   Photochem Photobiol.  32:367-374.

Bardach, J.E., and R.M.  Santerre.    1981.   Climate  and  the fish in the sea.
   BioScience.   31:206-215.

Bell, G.M., and  W.S.  Hoar.   1950.   Some effects of ultraviolet  radiation on
   sockeye salmon eggs and alevins.   Can.  J.  Res.   28:35-43.

Caldwell,  M.M.  1968.   Solar ultraviolet radiation as an  ecological  factor for
   alpine plants.  Ecol. Monogr. 38:243-268.

Caldwell,  M.M.  1971.    Solar  UV irradiation and the  growth  and  development of
   higher plants.  In Photophysiology, vol.  6,  ed.  AC. Giese,  131-177.  New
   York: Academic Press.

Caldwell,  M.M.  et  al.   1986.   Action spectra and their  key  role  in assessing
   biological consequences of  solar  UV-B radiation  change.    In Stratospheric
   ozone  reduction,  solar ultraviolet  radiation  and  plant  life,   eds.  R.C.
   Worrest and M.M. Caldwell,  87-111.   Berlin,  Heidelberg, New  York, Tokyo:

Calkins,  J.  1982.   Some considerations  on  the  ecological and  evolutionary
   effects of solar UV.  In The role of solar ultraviolet radiation in marine
   ecosystems,  ed. J.  Calkins,  685-689.   New  York:  Plenum.

Cheng,   L., M.,  Douek,  and  D.,  Goring.   1978.    UV  absorption   by gerrid
   cuticles. Limnol.  Oceanogr.   23:554-556.

Clark,  C.W. 1981.  Bioeconomics of  the ocean.  BioScience.  31:231-237.

Cushing, D.H. 1972.  The production cycle and  the numbers of fish. Symp. Zool.
   Soc. London.  29:213-232.


Damkaer, D.M. 1982.  Possible influence of solar UV radiation in the evolution
   of marine zooplankton. In The role of solar ultraviolet radiation in marine
   ecosystems, ed. J. Calkins, 701-706.  New York: Plenum.

Damkaer, D.M., and  D.B.  Dey.   1982.   Short-term responses  of some planktonic
   Crustacea  exposed  to enhanced  UV-B  radiation.    In  The  role of  solar
   ultraviolet radiation in marine  ecosystems,  ed. J. Calkins,  417-427.   New
   York: Plenum.

Damkaer, D.M., and  D.B.  Dey.   UV damage and photoreactivation potentials of
   larval  shrimp,  Pandalus  platyceros,  and adult  euphausiids,  Thysanoessa
   raschii.  Oecologia.  48:178-182.

Damkaer, D.M., D.B. Dey, G.A. Heron,  and  E.F.  Prentice.  1980. Effects of UV-B
   radiation on near-surface  zooplankton  of Puget Sound.   Oecologia.  44:149-

Dunbar, C.E.  1959.   Sunburn  in  fingerling  rainbow-trout.   Progr.  Fish Cult.

Giese,  A.C.  1938.  The  effects of  ultra-violet  radiations  of  various  wave-
   lengths upon cleavage of sea urchin eggs.  Biol. Bull. 75:238-247

Giese, A.C. 1976.   Living with our sun's ultraviolet rays.  New York: Plenum.

Hairston,  N.   1980.    The  vertical  distribution of  diaptomid  copepods  in
   relation to body pigmentation.   In Evolution  and  ecologyofzooplankton
   communities, ed.  W.  Kerfoot,  98-110.   Hanover,  NH:  University  Press  New

Harvey, J.M.  1930.   The action  of light on Calnus finmarchicus  (Gunner)  as
   determined by  its effect on the heart rate.   Contr ib.  Can. Biol.  5:85-92.

Hobson,  L.,   and  F.  Hartley.    1983.    Ultraviolet  irradiance and  primary
   production  in  a  Vancouver Island  fjord, British  Columbia, Canada.    J_^
   Plankton Res.  5:325-331.

Hunter,  J.R.,  S.E.  Kaupp,   and   J.H.  Taylor.  1981.    Effects  of  solar  and
   artificial ultraviolet-B  radiation on larval northern anchovy,  Engraulis
   mordax.   Photochem.  Photobiol.   34:477-486.

Hunter, J.R., S.E. Kaupp, and J.H.  Taylor.  1982.   Assessment of effects of UV
   radiation  on   marine  fish  larvae.    In  The  role  of  solar  ultraviolet
   radiation in marine  ecosystems, ed. J.  Calkins, 459-497.   New York:  Plenum.

Huntsman,  A.G.  1925.   Limiting  factors  for marine  animals I.    The  lethal
   effect of sunlight.  Contrib.  Can. Biol.   2:83-88.

Jitts, H.R.,  A. Morel,  and Y. Saijo.  1976.   The  relation of oceanic  primary
   production  to   available   photosynthetic  irradiance.    Aust.  J.   Mar.
   Freshwater Res.  27:441-454.

Jokiel, P.L., and R.H. York,  Jr.    1984.  Importance of  ultraviolet radiation
   in photoinhibition of microbial growth.   Limnol. Oceanogr. 29:192-199.


Karanas,   J.J.,  H.  Van  Dyke,  R.C.  Worrest.    1979.    Midultraviolet  (UV-B)
   sensitivity of  Acartia  clausii  Giesbrecht  (Copepoda).  Limnol.  Oceanogr.

Karanas,  J.J., R.C. Worrest, and H. Van Dyke.  1981.   Impact of UV-B radiation
   on the fecundity of the copepod  Acartia clausii.  Mar.  Biol.   65:125-133.

Kaupp, S.E.,  and  J.R.  Hunter.   1981.   Photorepair  in larval  anchovy, Engraulis
   mordax.  Photochem.  Photobiol.   33:253-256.

Klugh, A.B.  1929.   The  effect  of the ultra-violet component of  sunlight on
   certain marine organisms.  Can.  J.  Res.   1:100-109.

Klugh, A.B. 1930.   The effect  of ultra-violet  component  of the sun's radiation
   upon some aquatic organisms.  Can.  J.  Res.   2:312-317.

Larkin, P.A.   1978.  Fisheries management—An essay for  ecologists.  Annu.  Rev.
   Ecol.  Syst.  9:57-73.

Levin, S.A.,  and C.P. Goodyear.   1980.  Analysis  of an  age-structured fishery
   model.  J.  Math. Biol.  9:245-274.

Lieth, H.  1975.    Primary  production  of the  major  vegetation  units of  the
   world.  In Primary  productivity of the biosphere.    Ecological  studies 14,
   eds. H. Lieth and R.  H.  Whittaker,  203-215.  Berlin,  Heidelberg, Mew York:

Mann,  K.H.  1973-    Seaweeds:  Their  productivity and  strategy  for  growth.
   Science.  182-975-981.

Marinaro, J.Y.,  and  M.  Bernard.   1966.   Contribution  a 1'etude des  oeufs at
   larves  pelagiques  de  poissons  mediterraneens  I.    Note preliminarie  sur
   1'influence lethale du rayonnement solaire sur les oeufs.   Pelagos.   6:49-

Murphy, T.  1975.   Effects  of UV  radiation on nucleic  acids.    In  Impacts of
   climatic change on  the  biosphere,  part I - ultraviolet  radiation  effects,
   eds.  D.S.  Nachtwey,  M.M.  Caldwell,  and  R.H. Biggs,   (1)3-(1)30.    U.S.
   Department of Transportation  Climatic  Impact Assessment Program, Monogr. 5.
   Springfield,  VA: NTIS.

NAS   1979.     Protection   against  depletion  of   stratospheric  ozone   by
   chlorofluorocarbons.  Washington,  D.C.: National  Academy Press.

NAS  1984.   Causes  and  effects  of   changes  in  stratospheric ozone:  Update
   1983.   Washington,  D.C.: National  Academy Press

Rothschild, B.J.  1981.  More food  from the sea.   BioScience.  31:216-222.

Rpbinson, M.A. 1979.   Prospects for  world  fisheries to 2000.  FAQ  Fish.  Circ.
   722. Rome: FAO.

Russell-Hunter,  W.D. 1970. Aquatic productivity.   New  York:  MacMillan

Ryther,  J.H.
1960.   Photosynthesis and  fish  production  in the sea.  Science
Setlow,  R.B.  1974.   The wavelengths  in sunlight effective  in  producing skin
   cancer: A theoretical analysis.  Proc. Natl.  Acad.  Sci.  (USA) 71:3363-3366.

Smith, R.C.,  and K.S.  Baker.    1979.    Penetration of UV-B and  biologically
   effective dose-rates in natural waters.  Photpchem.  Photobiol.   29:311-323.
Smith, R.C., K.S. Baker, 0. Holm-Hanson, and  R.  Olson.  1980.
   of photosynthesis in natural waters.  Photochem Photobiol.
Steemann Nielsen, E. 1964.  On a  complication  in  marine productivity work due
   to the  influence  of ultraviolet light.   J. Cons. Cons.  Int.  Explor.  Mer.

Sverdrup,  H.U.,  M.W. Johnson, and R.H.  Fleming.    1942.   The oceans,  their
   physics,   chemistry,   geology,   and   biology.     Englewood   Cliffs,   Mew

Thomson, B.E., R.C.  Worrest,  and  H. Van Dyke.  1980.   The growth response of an
   estuarine diatom  (Melosira nummuloides [Dillw.]  Ag.J to  UV-B  (290-320 nm)
   radiation.  Estuaries.  3:69-72.

Whittaker,  R.H., and G.E.  Likens.   1975.  The biosphere and man.   In Primary
   productivity of the  biosphere.   Ecological  studies 14,  eds.  H.  Lieth and R.
   H. Whittaker, 305-328.   Berlin, Heidelberg,  Mew York: Springer-Verlag.

Worrest, R.C.   1982.   Review  of  literature   concerning  the  impact of  UV-B
   radiation  upon  marine organisms.     In  The  role  ofsolar  ultraviolet
   radiation in marine  ecosystems, ed.  J. Calkins, 429-457.   New York: Plenum.

Worrest, R.C.  1983a.    Depletion  of  stratospheric ozone:   Impact  of  UV-B
   radiation  upon  nonhuman  organisms.   In  The analysis   of actual  versus
   perceived risks,  eds.  V. Covello et al.,  303-315.   Mew York: Plenum.

Worrest, R.C. 1983b.  Impact  of  solar  ultraviolet-B  radiation (290-320nm)  upon
   marine microalgae.  Physiol.  Plant.   58:428-434.
Worrest,  R.C., B.E.  Thomson,  and H.  Van Dyke.   1981.
   upon estuarine microcosms.  Photochem.  Photobiol.
                                       Impact of UV-B radiation
ZoBell,  C.E.,  and G.F. McEwen.   1935.   The lethal  action of  sunlight upon
   bacteria in seawater.   Biol.  Bull.   68:93-106.

                                                         £  b-
                                                         <»  !=
                                                         O)  (0
                                                    *-    "

   > 5


     O  0) o    -2

     if §    §
     ®  c o
     O  TO
     c  a, a)
     O — O)
     o .c co
     c O 0
     co -^ >
     a>  o co


                                                   Long i tude

                              TEMPERATURE  (FIVE YEAR RUNNING  MEAN)
               -so -








   Plate 2.  Observed rate of temperature change in the past century, estimated from meteorological station data.  The geographical
temperature trends in part (a) include estimates for remote regions with station records as short as 25 years. The zonal mean trends in part
(b) refer to the average for all longitudes with available records. A nonlinear scale is used in part (b) so that there is a similar area for each

Source: Hansen et al. (this volume)

                     Plate 3. Annual mean temperature changes
                   computed  with  the  Goddard  Institute for
                   Space Studies global climate model, which
                   has a sensitivity of about 4°C for doubled CO2.
                   Parts a and b, for the decades 1990-1999 and
                   2010-2019, respectively, were obtained from
                   the transient experiment with CO2 and trace
                   gases increasing according to Scenario A; the
                   indicated  temperatures are the difference
                   between the transient run and a control run
                   with 1958 atmospheric composition.  Part c is
                   the  equilibrium  warming  in the model for
                   doubled CO2.

                   Source: Hansen et al. (this volume)

                                               Black   - rivers
                                               Red     - forest cover; transplanted
                                               Orange - deepwater aman rice/boro rice
                                               Blue    - water
                                               White   - aman rice/fallow
                                               Grey    - deepwater rice/urban area (Dacca)
             Scale 1: 250,000
               6       10
                            it  V1..'-.'     *     "     »*.**  •
  Plate 4. The high scenario represents the shoreline retreat from a 2 to 2.5 meter global rise in sea level if dams built on
the Ganges prevent sediment from reaching the delta. The low scenario represents the impact of a 50 centimeter global
rise with only a partial disruption of the sedimentation process.
Source: Broadus et al. Volume 4


 The Greenhouse  Effect:  Projections of Global
 Climate Change

 J. Hansen1, A. Lacis1, D. Rind1, G. Russell1,1. Fung1, P. Ashcraft2
 S. Lebedeff3, R. Ruedy3,  and P. Stone4

     We present projections of global climate during the next  several decades,
based principally on  climate model  simulations in which atmospheric  CC^ and
trace gases  increase  steadily at rates estimated  from  observations.   This is
the first time that a global climate  model has been used for simulating the
climate  effects  of  the  transient growth of greenhouse gases, and as  such it
permits  an  estimate of when  the greenhouse effect should begin to be evident
above the level of natural climate variability.  We emphasize  that a number of
caveats  must be attached  to  the climate  model results.   But we  also stress
that the  climate sensitivity  of our model  has been extensively compared to the
sensitivity  of other  models  and  is  consistent with available empirical  evi-
dence from past climate  changes.

     Our  presentation is  organized  as  a  response to  a letter from Chairman
John Chafee  of the  United States Senate Subcommittee on Environmental Pollu-
tion of  the  Committee on Environment and Public Works  who  requested testimony
at a hearing on the  greenhouse effect on June 10, 1986.   Specifically, his
letter asked that the  following topics be  addressed:

     •  The nature of  our work in modeling  greenhouse climate  effects

     •  How we test the  models to determine their validity
2  NASA Goddard Space, Flight Center, New York, NY
2  State University of New York, Albany, NY
•j  M/A COM Sigma Data Inc., New York, NY
   Center for Meterology  and Physical Oceanographies, Massachusetts
     Institute of Technology, Cambridge,  MA

     •  The relative  contribution  of different  greenhouse gases  to  possible
        future climate change

     •  The temperature  changes  predicted for  the next  few  decades  and  the
        next century,  assuming some reasonable growth in trace gases

     •  How the  predicted temperatures compare  to past  temperatures  experi-
        enced on the earth

     •  How temperature  changes  of  the magnitude  predicted  might alter  the
        number of days with  temperatures  above a given  limit  for  Washington,
        D.C. and other U.S.  cities

     •  Further evidence  needed to  confirm and quantify the greenhouse theory.


     In principle,  global climate  models are  quite simple.   For  example,  in
our model (Hansen et al.  1983, hereafter  referred  to as paper 1) the earth is
divided into  'gridboxes'  as  shown  in Figure  1, and each  gridbox  is  divided
vertically into a number  of  layers,  typically  nine,  in the atmosphere.  Simi-
larly, the  ground  or  ocean  in each  gridbox  is divided  into  several  vertical
layers.   The  mathematical equations  describing the  fundamental conservation
laws of physics are solved numerically for each gridbox by a computer program,
which  calculates  the  transfer of mass, energy,  and momentum  from  one box to
another and also simulates physical processes within the boxes which represent
sources and sinks of these substances.

     Such global  climate  models,  or  GCMs, are able  to  reproduce  the  general
features  of the  earth's climate.    Climate  variables  such   as  temperature,
winds, and  storm tracks,  and their  variations from season   to season,  from
latitude  to  latitude,  and from continent  to ocean,  are  represented  realist-
ically, at least in a qualitative  sense.   But the  models are not sufficiently
realistic  to   portray  accurately  regional  patterns  of  precipitation,  ocean
currents,  and other processes that  are important for determining the practical
consequences of climate trends because of greenhouse warming.   Recent evidence
that the  large-scale ocean circulation may have undergone dramatic changes in
the past  (Broecker, Peteet,  and  Rind 1985)  is  of  special concern; the repre-
sentations of oceans in current GCMs are not  sufficiently realistic to predict
such phenomena.

     Improvement  of the  climate  models  will depend especially  upon better
knowledge of physical processes occurring  within the climate  model gridboxes.
These  processes are  represented  by  submodels or  "parameterizations."   For
example, Figure Ib schematically illustrates  convective clouds associated with
convective  transport  of  moisture,  heat,  and  momentum between  model  layers.
Although  there  have been major field experiments to  study convection clouds,
substantial work is  needed  to provide  more  realistic  submodels   for  use in
GCMs.   Another example,  one which is beginning to  receive greater attention,
concerns  the  role of  vegetation  and  soil  processes  in  the  transfer  of
moisture, heat,  and momentum between the earth's surface and the  atmosphere.
As a   final example,  we must have a better understanding of small-scale ocean

-60  -

TABLE 1  Fundamental equations
                                      Longitude (degrees)
Conservation of momentum dV _
(Newton's second law of ~ft = ~2U X V ~ p p
motion) + g -I- F (Tl)
Conservation of mass dp
(continuity equation) 4, ~ 0"'V + C D (T2)
Conservation of energy dl dp'
(first law- of J, = -P ~^~ + Q (T3)
Ideal gas law p = PRT (T4)
(approximate equation of
V velocity relative to rotating earth
/ time
— total time derivative = — + V- V
dl L dl J
fl planet's angular rotation vector
p atmospheric density
g apparent gravity [=true gravity - Q X (Q X r)]
r position relative to planet's center
F force per unit mass
C rate of creation of (gaseous) atmosphere
D rate of destruction of atmosphere
/ internal energy per unit mass 1=^7"]
Q heating rate per unit mass
K gas content
c, specific heat at constant volume.
•' . -. :. stratospheric

aerosols • .' y ••.••'••:•
.y^^- .** 1.T -^VZt— r~— . - -,
I large-scale supersaturation cloud >
convective cloud .
c ~y~
latent and
heat fluxes .
i t It
heat and .
— trace gases,
* »
run-off _,tt )CE
LAND T*~-Z.( diurnal)
s^ -*• Z 2 (seasonal 1
                                        Figure 1.

mixing before we can develop an accurate model of global ocean circulation  and

     A  principal  conclusion  is  that  decades  of  research  are  likely  to  be
required  to  improve climate  models to  the  point  that  they can  be used  to
predict  local and  regional  climate changes with a  high degree of  confidence.
Such  improvements  will be  possible only  if  appropriate observations of  the
climate  system  and  climate  processes are  carried  out.    In  the  meantime,
climate  models  can provide a  useful  indication of  the possible magnitude  of
future climate trends,  although the results must be accompanied by  appropriate
explanations and caveats, especially the results at smaller scales.

     Finally, we would  like to stress one key  characteristic of both climate
model  results  and real-world climate,  i.e.,   natural climate  variability.
Figure 2  illustrates the global mean temperature in a  100-year control  run  of
our 3-D  GCM.   It can be seen  that  the temperature  fluctuates, both  from year
to year and with decadal trends, even though the amount of atmospheric C02  and
other  climate  "forcings" are  unchanging  in  the  model.  Such variability  or
"noise"  is  a natural  characteristic  of  the climate system.   Although this
phenomenon is captured  by  the governing fundamental  equations in  the climate
model,  individual  fluctuations are  not  predictable.   Thus,  for  any climate
trend to be detected,  it must exceed the level of natural climate variability.








                       I      I      I      I      I      I      T
                        Global Annual-Mtan Surface Air Temptraturt
                               IOO Year GCM Control Run
                10    20    30    40    50    60
                                    Time (years)
        Figure.  2.   Global-mean annual-mean  surface air temperature  in
        the 100-year control run with the GISS global climate model  with
        predicted ocean  temperature.   Atmospheric CC>2,  trace gases, and
        aerqsols are unchanging  during  this run.  The ocean-mixed layer
        has the  observed seasonal  and geographical depth variations and
        no exchange of heat with the deeper ocean.


     A good overall test of the greenhouse effect is provided by examining the
climates of several planets—Mars,  Earth,  and Venus—which have  a  wide range
of abundances of atmospheric greenhouse gases.   These  planets are warmer than
they would  be if  they  were  simply  in  blackbody equilibrium with  the energy
absorbed from the sun.  The observed greenhouse  warmings  are a few degrees on
Mars, about 35  degrees on Earth, and several hundred degrees .on Venus.   The
magnitude of  these warmings  is  in  excellent agreement  with  the  greenhouse
theory and simple climate models.

     Another  test,  of  more  direct  relevance,   is provided by  prehistoric
climate  variations  on  Earth.    This  test is  only recently  beginning to  be
exploited.   The climate  of the Earth has  fluctuated between  ice ages  and
interglacial warm periods  several times during the past  few  hundred thousand
years.   It  has  been realized for more  than a  decade  that small variations in
the earth's orbital characteristics about the sun were the 'pacemakers' of the
10,000- to 100,000-year climate changes (Hays et al. 1976), but the mechanisms
by which global temperature changes  were produced  remain  uncertain.  Recently
it has  been discovered that  the atmospheric  C02  abundance  fluctuated along
with  ancient  climate,  and was  thus a  probable agent  producing  the global
temperature changes.  This inference of past  C02 warming allows an empirical
evaluation of climate sensitivity to a C02 change.

     In particular, the paleoclimate record indicates that a C02 change of 50-
100 parts per million (ppm) (Oeschger et al.  1984)  is associated with a global
mean temperature change of 4°-5°C (CLIMAP Project 1981).   The global radiative
forcing  ( ATQ)  due  to  this C02  change, i.e., the  surface temperature change
which  would  occur  if  there  were  no  climate "feedback"  processes,   can  be
accurately computed and is

      ATQ ~ 0.5°C.                                                   (1)

Thus, under the assumption that the C02 change is the dominant forcing for the
global temperature change, the total climate feedback factor f, defined by

      AT = fAT ,                                                    (2)

is   f ~  10 for  the  recent glacial-to-interglacial  climate changes.  Physical
processes contributing  to  f have been analyzed (Hansen et al. 1984, hereafter
referred to as  paper  2) and it  has  been shown that a  substantial part of the
total feedback factor is due to  the  growth and decay of large continental ice
sheets, a process  which is not  significant on the  time scale of the next few
decades.    The  feedback   factor inferred from  the  paleoclimate  data  for
processes that are relevant to a climate change on decadal time scales is
f ~ 2-4.

     The  paleoclimate  evidence  thus indicates  that  if  atmospheric  C02  were
doubled from, say, 300  ppm to  600 ppm,  and the climate system were allowed to
come to a new equilibrium, the earth would warm by

      AT(2*C02) ~ (2°-4°C) * ATQ ~ 2.5°-5°C,                         (3)

where   ATO ~  1.25    is   the  no-feedback  radiative  forcing  due  to  doubling
atmospheric COg.

     The climate  sensitivity  inferred from paleoclimate evidence is reasonably
consistent with the climate sensitivity estimated  by several National Academy
of Science studies  (Charney et al.  1979;  Smagorinsky et al. 1982),
      AT(2*C02) = 3° ± 1.5°C,                                        (4)

which was  based  mainly on inter comparison  and  analysis of  several different
climate model studies.  Thus  there is general agreement about the magnitude of
global climate  sensitivity,  but  the  uncertainty in its  value is  at  least a
factor of two.

     Despite  all  the  above  evidence, the  one   truly  convincing test  of the
models must  be  a  comparison  of the  models'  predictions  with  the  observed
response  of   the  earth's  climate to  the  present  anthropogenically  induced
growth of atmospheric CC^ and trace gases.  Thus, we illustrate in Plate 2 the
temperature changes that have occurred during the past century,  a time during
which CC>2 has increased from about 280 ppm  to 340  ppm and several other trace
gases have also increased.

     The  geographical patterns  of temperature  change (Plate  2a) contain a
large amount of natural variability, as well as  errors in regions where there
are few stations and  short records.   The  influence  of the climate  'noise' and
incomplete station  coverage  is reduced by  averaging  results over  all longi-
tudes.  Plate  2b  shows the resulting temperature trends  for the period 1880-
1984 as a function of latitude.

     Several conclusions follow from the  data illustrated  in Plate 2.   Most of
the earth  has warmed  in the  past 100 years, but  there is  a  great amount of
local variability.   The  global mean warming since  1880 is  about 0.6°C (1°F),
with both  hemispheres warming by about  that amount.    Although  the earth was
about as  warm  in  the  1930s  and  1940s  as  it  is  in  the 1980s,  the  earlier
warming was  more  concentrated  in the  high latitudes  of the Northern Hemi-
sphere,  while the recent warming is more  global.

     The  global  warming of  0.6°C in  the  past  century  is  of  the magnitude
expected as  a result of increasing CC>2 an<^ trace  gases.   It  is difficult to
make  a  more definitive  statement than that, in part  because  the greenhouse
forcing is time dependent (most of the growth coming in the past 25 years) and
the climate system has a finite response  time,  probably several decades (paper
2 and Hansen  et al.  1985), so  that  only  a part  of the eventual warming due to
these added gases has occurred as yet.  Also, a  warming of 0.6°C is not parti-
cularly large compared to  climate fluctuations that have  occurred in the past
millenium, as illustrated below.

     We  show below  that  the  expected greenhouse  climate  signal is rising
rapidly  and  should  soon  rise  well  above  the  level  of  natural variability.
This would provide strong empirical verification of the greenhouse effect.


     The  estimated  contributions  of different  greenhouse  gases to  climate
forcing  is  illustrated in Figure  3 for different  periods.   The  C02  contri-
bution  is  known accurately,  within about 10  percent.   The  contributions of
chlorofluorocarbons CCloF (F^) and CC12F2 (F^'  comPounds which are entirely
man made, are also known accurately.  The  CHjj  greenhouse contribution is less
certain; there is some  recent  evidence  suggesting that the CH^ growth rate we
assumed (0.6 percent per year  in the  1960s,  1  percent per year  in the 1970s,
and 1.5  percent  year  in the  1980s) may be too high  for  the  1980s and  perhaps
too low for the  1960s  (Rasmusson and  Khalil  1986;  Rinsland et al. 1985; Blake
and Rowland 1986).  The Oo and stratospheric H20  changes,  based on a chemical
model,  are  very uncertain  and are best described  as  hypothetical;  indeed,
recent spotty observations of Oo profiles do  not support a positive greenhouse
forcing due to  changing 0^  (Lacis  et al.  1986).   Despite  uncertainties about
the trends of some of the gases, two firm conclusions can be made:

     •  In the past few decades the rate of  increase of greenhouse forcing of
        the climate system has  increased rapidly;  it is now three to ten times
        greater than during the previous century,  1850-1960.

     •  Non-C02 greenhouse gases now  add to the  greenhouse  effect  at  a rate
        comparable to that of C02.

     We have  carried  out the  first GCM simulations of  climate  trends  due to
the current changes of atmospheric C02  and trace  gases.   One disadvantage of
presenting recent research  results is  that  we cannot claim  that the  results
have been confirmed by other climate modeling groups.   However,  the sensiti-
vity of our global climate model has  been  compared and found to be similar to
that  of other  climate  models at  the  NSF  National  Center for  Atmospheric
Research (Washington and Meehl 1984)  and the NOAA Geophysical  Fluid Dynamics
Laboratory (Manabe  and Broccoli 1985;  Wetherald  and  Manabe 1986).    Thus we
expect  that  the  general conclusions discussed  below would not  be changed if
the models of these other laboratories  were  used,  provided that similar basic
assumptions were made, for example, with regard to trace gas growth rates and
ocean heat uptake.

     The global  climate model  employed  for these simulations  is  our model II
documented in  paper  1, which  has   an equilibrium global mean  sensitivity of
4.2°C for doubled C02  as  documented in paper 2.   For these  transient  experi-
ments the ocean  mixed  layer  depth   is based on  climatology,  including  geogra-
phical and seasonal variations.  Heat capacity  of the ocean  beneath the mixed
layer is  included  by  treating  a heat perturbation as a passive  tracer which
can diffuse into the deeper ocean;  the diffusion coefficient varies geographi-
cally as  a  function of the  climatological  local  stability  beneath  the mixed
layer, according to the empirical  relation given in paper 2.   Horizontal ocean
heat transports are fixed as  described in paper 2.

               Decadal Increments of Greenhouse  Forcing

         1850 -I960
         (per  decade)
                                              £••••»*•*• u n
            Decadal  additions  to  global  mean greenhouse  forcing  of the
              ATO   is the  computed temperature  change at equilibrium for
     Figure  3.
climate system.
the  estimated  decadal  increase  in  trace  gas  abundances,  with  no  climate
feedbacks  included.   Multiply   AT0   by  the feedback factor  f  to  get  the
equilibrium  surface temperature change  including feedback  effects.   Most  of
the estimated trace gas  increases are  based on measurements.   However, the Oo
and stratospheric  f^O  trends (dotted bars)  are based principally  on 1-D model
calculations of Wuebbles et al.   (1983).

     Our GCM simulations begin  in  1958, when CC^ began  to  be monitored accu-
rately by  C.D.  Keeling, and  extend  through the  present into the  future.  We
consider two scenarios, A  and B,  to allow  for  uncertainties  about past trace
gas changes and  future CC^ and trace  gas changes.   Scenario  B  includes only
five greenhouse gases that have been measured reasonably well (C^, FH» ^12'
CHjj, N20) and it assumes that their growth  rates  will  decrease rapidly during
the next  few  decades.  Scenario  A includes an allowance  to  approximate the
greenhouse forcing of the  several  other hypothesized gases in Figure 3, and it
allows presently estimated growth  rates to continue.  We do  not contend that
Scenarios A and B represent extreme possibilities; however,  they do help us to
estimate how sensitive  the conclusions are  to  assumptions about  the climate

     Scenario  A  achieves  a radiative  forcing equivalent to  that  for doubled
C02 about  40  years  from  now, in  the  late 2020s.   Scenario  B  achieves this
level of  forcing  in  about 2060.   The  more detailed  trace  gas  scenario  of
Ramanathan et al.  (1985)  is  close to Scenario  A  (see Figure  4).   Finally,  we
note  that  both  scenarios  include some  additional  radiative  forcing  due  to
volcanic aerosols,  a forcing  that tends to  cool  the  surface.    The aerosol
forcing is  the  same in the two scenarios for   the  period 1958-1986,  with two
significant coolings:  1963-66 (Mt. Agung) and   1982-85  (El Chichon).  Scenario
B assumes that the mean aerosol forcing  in  the  future  will be similar to that
in  the  period  1958-1966,  a  period  of  active volcanism,  while  Scenario  A
assumes a negligible  future volcanic  forcing,   as was the case for the period

     Global maps of  temperature changes  obtained  in the GCM  climate experi-
ments are  shown in  Plate 3.   The  detailed geographical  patterns  of these
changes are not  to be  taken  seriously in view  of  the limitations of current
GCMs  mentioned  above and  the  gross  assumption about  ocean  heat transports.
However, some  of the large-scale features are more likely to be meaningful.

     The warming  in  Scenario  A at most midlatitude Northern  Hemisphere land
areas such as  the United States is typically 0.5°-1.0°C (1°-2°F) by the decade
1990-2000 (Plate 3a)  and  1°-2°C (2°-4°F) by the  decade  2010-2020 (Plate 3b).
Even  in  the  latter decade  the warming  is much  less  than  the equilibrium
response to doubled  C02 (Plate 3c) which has   a  warming of about  5°C in the
United States.   At  all future times  the largest temperature changes are  in
regions  of  sea  ice  and the  smallest are  at   low  and  middle  latitude ocean

     The global mean temperatures  computed by the  model are compared to obser-
vations in Figure  5  for the period 1958  to the present.   In  this period the
greenhouse forcing of  Scenarios A  and B  differ only by  about 10 percent.  It
is apparent that the  observed temperatures and the  model results are consis-
tent  during this period.   But the natural  variability of temperature in both
the  real  world  and the  model are  sufficiently  large   that  we  can neither
confirm  nor  refute  the modeled  greenhouse effect on  the basis  of current
temperature trends.

                  I '
                                  Greenhouse Forcing
                               for two trace gas scenarios
                              doubled .
                               CO,   V
                                                        Scenario B,
                                              Ramanathan etol. (1985)
                        1980   2000
                                    2020   2040
2060  2080   2100
     Figure 4.   Greenhouse  forcing for two  trace gas scenarios.    ATO is the
equilibrium warming with no  climate  feedbacks.   Scenario B includes only COp,
CHh, NpO,
                and  F*,  and it assumes  that their growth rates will  decrease
            ^       ^*,
rapidly in  the next  2.5  years.   Scenario A  includes an  allowance for  other
trace gasses  hypothesized  in Figure  3  and it allows  current growth  rates to
continue indefinitely.
     In Scenario A CO^ increases as observed by Keeling  for  the  interval 1958-
1984 and  subsequently  with  a  1.5  percent growth  of  the  annual  increment.
CCloF  and  CCl2Fp  emissions  are  from  reported  rates  to  date  and assume  3
percent per year  increased emission in the future, with atmospheric lifetimes
for the gases of 75 and 150 years, respectively.  CH^  increases  from 1.4 parts
per billion  in  1958  at a rate of 0.6  percent  per year  until 1970,  1  percent
per year   in the 1970s and  1.5  percent per year  thereafter.   N^O increases
according to  the semi-empirical formula  of Weiss  (1981),  the race  being 0.1
percent per  year"  in  1958,  0.2  percent per  year  in  1980,  0.4 percent per
year  in 2000 and 0.9 percent per year' in 2030.  Potential  effects  of several
other trace  gases (such as  Oo,  stratospheric HpO,  and chlorine  and fluorine
compounds other  than  F,,  and F«p)  are  approximated  by  multiplying  the CFC
amount by the factor 2.

     In Scenario  B  the growth of the  annual  increment  in COp is reduced from
1.5 percent  today  to  1  percent  in  1990,  0.5  percent in  2000  and  zero  in
2010.    The  growth in annual  emissions of CCloF  and CClpFp  is reduced from 3
percent  today to 2  percent  in  1990,  1  percent  in  2000 and  0.5  percent  in
2000.    N20  increases  are based on  Weiss'  (1981)  formula,  but  the parameter
specifying annual growth in anthropogenic emissions decreases  from 3.5 percent
today to 2.5 percent in 1990, 1.5 percent in 2000 and  0.5  percent in 2010.  No
increases  are included  for  other  chlorofluorocarbons, ozone,  stratospheric
water or any other gases.

     On the other hand,  it is also apparent  from  Figure  5  that the predicted
greenhouse warming for Scenario A rises above the level of natural variability
by  the 1990s.    The standard  deviation  of  the  five  year smoothed  global
temperature during the past century is   o  ~  0.2°C.    Thus a global warming of
0.6°C, compared to,  say,  the 1958-1980 mean temperature, would be a 3o effect,
significant at the 99 percent confidence level.   The model  predicts that such
a warming will be achieved in the 1990s.

     Indeed, the model predicts a temperature level  in the  next 15 years that
has not existed on earth  in  the  past  100,000 years,  as illustrated below.  In
view  of  the significance  of  such conclusions,  we stress here  the principal
caveats which must accompany the result:

     •  The model  sensitivity is  4.2°C for  doubled  C^.   Emergence  of the
        warming signal above the level  of  natural  climate variability will be
        delayed if the true sensitivity is  less than that.

     •  The  projection  is  based  on  Scenario  A.   If  Scenario  B  is  more
        realistic, emergence of the warming  signal will  be  delayed.  We esti-
        mate  that Scenario  B  would  delay  the emergence  of  the signal  by
        several years, but a more quantitative statement requires extension of
        the Scenario B simulation.

     •  Other major  climate forcings that tend  to  counteract  the greenhouse
        warming may  occur  during  the  next  several  years.   As  one  example,
        satellite measurements indicate that solar irradiance decreased in the
        period 1980-1984  (cf.   references  in Kerr 1986)  at  a  rate that would
        approximately cancel  the increase in greenhouse  forcing  during the
        same  period.    Although  decreases in solar irradiance  are  probably
        cyclical  and  must  be  balanced by comparable  increases  over  a  time
        scale of several decades, it  is possible that a decreasing trend could
        continue  for  several more years.    As a  second example,  an unusual
        increase in volcanic activity could  conceivably counteract the green-
        house warming  for as long as a decade or  so:  Scenario  B contains a
        substantial  amount of volcanic  aerosols, similar  to  the amount in the
        volcanically  active  period  1958-1985, but it  is  possible to  have a
        still greater level of volcanic activity.

     •  There may be crucial climate mechanisms  that  are  omitted or  poorly
        simulated in  current climate  models.   An  example is  changes  in ocean
        circulation,   such as  the  formation  of deep water  (Bennett  et al.
        1985).   If  the rate of  deep  water  formation is reduced,  it  is  even
        possible that the North Atlantic and Europe may cool while most of the
        world is warming.   A change in  ocean circulation would also raise the
        possibility  of  a  more  rapid  transition  to  a  warmer  climate,  if it
        reduced the  effective  thermal  inertia of  the ocean and  thus reduced
        the climate  response time (paper 2; Hansen et al. 1985).


     Global  temperature  changes  are  illustrated  in  Figure  6  for  the  past
century,   millenium,  and  25,000  years  and  in Figure  7  for the  past 150,000
years.   The global mean  temperature  has  varied by  about  0.5°C  in  the  past
century  and 5°C in   the past several  hundred thousand years (National Academy


                     (a) Seasonal Mean Global Temperature  Trends

                 SCENARIO A

                 SCENARIO B

                          I I
              V  V
             1958 I960
^ 0.6




1 1 1 1
b) 5


1 1 1

r Smc

>othed Glo


1 1 1


i i

1 y— •>

1 1 1 1

1 1 1 1
bal Temperature Trends

i i

1 1 i i

i i i

i i i i


I960 1970 1980 1990 2000 2010 201
     Figure 5.   Global temperature  trends from observations (solid  line)  and
from  calculations with  the GISS  global  climate  model.   Part  (a)   shows  the
temperature anomalies  plotted each season (December-January-February,   March-
April-May, etc.).   Part  (b)  shows the  5-year smoothed  results.  Simulations
have only been carried out  into  the  future for Scenario A.

                p    °
(o) past cintury
               annual mean
               5-year running mean
                              1900     1920    1940
                             (b) past millenium
                           I    i   I    i    I
                          1000    1200   1400    1600    1800
                             (c) past 25,000 years
                       30      25    20      15     10
                                 Thousands of Years Ago
     Figure 6.   Global temperature  trend for the  past century (a), millenium
(b), and  25,000  years (c).    (a)  is  based on  Hansen  et al.  1981,  updated
through 1981.   (b)  is  based on  temperatures in central England, the  tree  limit
in the White  Mountains of California,  and oxygen  isotope measurements in  the
Greenland  ice  (W.  Dansgaard,  private  communication),  with  the temperature
scale set by the variations in  the  last 100 years,   (c) is based on  changes in
tree lines,  fluctuations  of alpine  and  continental  glaciers and  shifts  in
vegetation patterns recorded in  pollen spectra  (National Academy of  Sciences
1975),  with the temperature  scale set by  the  3°-4°C cooling obtained  in  a  3-D
climate model   (Hansen  et  al.  1984)  with  the boundary  conditions  for  18,000
years ago.   Thus the  shapes  of curves (b) and  (c)  are based on only  Northern
Hemisphere data.

of Sciences  1975).   During   the peak of the current  interglacial  (Altithermal
period 5,000  to  10,000 years  ago)  the mean temperature  is estimated to  have
been 0.5°-1.0°C warmer than today (Figure 6).

     A  comparison of  the  temperature  changes projected  for  the  next  few
decades with  estimated temperatures  for the  past 150,000 years  is shown  in
Figure 7.   By the early twenty-first  century the global  temperature  should
have risen well above any level experienced in  the past 100,000 years.
                           Global Mton Ttmperaturc

                        I  (Dal«)

                                                           ( Sctnario A)
         150     125     100     75      50
                     Thousands of Years  Ago
     Figure 7.  Global  temperature  trend for the past 150,000 years  and
     as projected  for the next  few decades on  the  basis of the  global
     climate model simulation with Scenario A trace gas trends.

     How might temperature changes of the magnitude predicted alter  the  number
of days with  temperatures above a given  limit for Washington,  D.C. and other
U.S. cities?   We estimate  this by compiling  climatological  data for a given
city from a  long series of daily  observations (including maximum and minimum
temperatures  for each  day)  and  adding  to  this  record the  mean  (monthly)
increase  in  daily maximum  temperature and in  daily minimum  temperature  as
predicted by the climate  model  for the gridbox that  includes that city.  This
procedure tends  to minimize the effects  of any errors in the model's control
run climatology.  Although we should, in principle, also  examine  the effect of
changes  in  higher  moments of  the  temperature  distribution,  Mearns  et  al.
(1984) have  shown  that changes in the higher  moments have relatively  little
effect on the total number of days above a  temperature extreme.

     We have carried out  this procedure for several U.S. cities  for the equi-
librium change  in  climate  for  doubled CQ^-   Because of the climate  system's
finite response  time,  the results may  be most applicable to some time  in the
middle of the twenty-first century, if  the  trace  gas  growth rate  of  Scenario A
is approximately correct.

     The results of this  exercise  for  several  cities in the United States are
shown in  Figure  8.   The  number of days  per year in which  the  maximum daily
temperature  exceeds 38°C  (100°F)  increases  from  about  one   to  twelve  in
Washington and from three to twenty in Omaha.  The number of days with maximum
temperature  exceeding  32°C  (90°F)  increases  from  about thirty-five  days  to
eighty-five  days  in both cities.   The number of  days  per year  in which the
nighttime temperature does not fall below 27°C (80°F) increases from less than
one  day  in  both  cities  to nine  days in Omaha  and nineteen  in Washington,
D.C.  Analogous results for six other U.S. cities are included in Figure 8.

     There are  a number  of reasons why  these estimates may differ  from the
real world  response.    Principal  among these  are  the following.   First,  the
estimates are based on a  model  with a  sensitivity of 4°C for doubled CC^; the
real world  sensitivity is  uncertain by  about a  factor  of two.   Second,  the
model assumes that  the ocean will continue  to operate  essentially as it does
today;  if  North  Atlantic deep  water  formation and  the  Gulf  Stream should  be
substantially modified,   for  example,  that  could  change the  results  for  a
location  such as  Washington,   D.C.    And third,  there  are  many small-scale
processes  that  are not  resolved  by  the  model,   which could  cause  local
responses to vary.

     A number of expected  effects of the  greenhouse  warming  in  the United
States are  summarized  in Table 1.  The  summer  cooling  requirements  in some
United States cities will increase  several  times.   On the other hand, heating
degree days  in  the winter  decrease  on the  order  of 50  percent.   The winter
warming,  expected to exceed that in summer, may also affect such things as the
ability of indigenous  forests  to survive (Shantz and Hoffman 1986),   and the
amount, type, and distribution  of  precipitation.   Table  1 also  indicates that
the  growing  season will  increase  in  the United  States, typically  by fifty
days, and  it is expected that  increasing atmospheric C02  will  enhance plant
growth  (Lemon  1983).  Nevertheless,   the  high  midsummer temperatures  could
conceivably   result  in   lower   agricultural  productivity,   as  occurred  in
(regional)  midsummer   heat   waves  in  the United  States  in  1980 and  1986.
Quantitative  evaluation  of  beneficial and  detrimental  effects  will  require
greatly improved climate forecast capabilities as well as research in societal

     We  conclude  that the magnitude  of  temperature   changes   predicted  by
current climate models  is sufficiently great that, on the average, the warming
will be very noticeable to the populations at middle latitudes.   Other discus-
sions of the  practical impacts  of  greenhouse warming have focused on possible
indirect effects such  as  changes of sea  level,  storm frequency,  and drought.
We   believe,  however,   that   the   temperature   changes   themselves   will
substantially modify the environment and have a major impact on the quality of
life in some regions.


     From a  scientific point of view,  evidence  confirming  the  essence of the
greenhouse theory is already overwhelming.   However, the greenhouse  issue  is
not likely to receive the full attention it deserves until the global tempera-
ture rises  above the  level of natural  climate variability.   This  will  not
occur at  some sharp  point  in  time,   but rather gradually  over a period  of
time.  If  our model is approximately  correct, that time may be  soon, within
the next decade.


Days per Year with Temperature Exceeding  100°F

                         ••(950-1963 Climate
                         I  1 Doubled C02
    -L a
   Washington D.C.
                         New York
    Denver    Los Angeles     Dallas

                                                         Days  per Year with  Temperature Exceeding 90°F
                                                           Washington D.C.    Omaha
                                                                                   ••(950-1983 Climate
                                                                                   I   I Doubled C02
                                                                                 New York     Chicago


                                                                      Los  Angeles    Dallas
Days with  Minimum  Temperature Exceeding 80°F
                             1950-1983 Climate
                             Doubled COz
  Washington D.C.     Omaha
                              I 3

                         New York

00 01 4

Denver Los Angeles Dallas

               Figure  8.   Annual
        number    of    days     in
        several  U.S.  cities with
        (a)  maximum  temperature
        greater  than  100°F,  (b)
        maximum       temperature
        greater  than  90°F,  and
        (c)  minimum  temperature
        greater  than  80°F.   The
        results  for  doubled  CC>2
        were generated  by adding
        the warming in  a doubled
                                                                   experiment   to  recorded
                                                                   temperatures  for   1950-

L. Ol
3 01
4-> L.
1- Ol
01 TJ
O. •
01 C O
4-> — VJ

3 O Ol
E <-> —
— a
X »- 3
?0 0
X Ol t.
— E *-
TO 3
T) C T)
01 •- 4J
4J 0) E
4-1 ••-
U E l/>
^ — 01
"X u in

c CM
— 0 TJ
t. TO
Ol Ol Ol
X — 4J
>- 3 E
Ol O —
O.TJ —
in T)
X c r-i
T) Oi
m — i
U. CO 1
o 01 o
—• in
L. 1 Ol
01 3 *•*
.0 in
E 01 01
3 — • .c
C 4J
•-.C U
in 4J o
4J V-
C 0 C
§^» o
Ol 0 Ol
> o m
4-> Ol
4-> Ol •*•
Ol 0) O
in Ol <-
Ol u Ol
L. X
a. oi u_
Oi O
1- 01
L. f
1- 3 4J
O 4J 01
»- TO C
t- Ol
CM 01 *—
0 Q.
Ol C

— E— -
JD 3 LL.
3 E o
o— in

0^ 01


3 4-f *-^
.C Wl
t^t' 0
— Ol
-* Ol ^
TJ ' —
Ol O'
^ 01 m

t— Ol T)











































•yi *<^ CM *•* VD in m «^
r*» ^" m m o vo *^ ^^

m vo co 01 CM in r*> ^f
c*> m r^. r^. vo vo in m
CMCM-^ — — »ncMCM

CM m «• «• n _ _

— •COvD^-'-xeMr^oi
«-«TVOVOVO— 4!Mm

r^. CO Oi 01 *^ 3 ^f CTI
^^ CM ^M m ^^ ^^ n 01

Ol «• CM CO r^ CM
9 CM

Ol C*"l ^ Ol O ««4 33 v^
_ -. vo vn


»*^ —^

n — H — m m o vo

— i CM -- r~ rr

^^ o o *n o ** 01 ^^

^nrn vO^*OiCMr^^.
r^ vo ^^ CO Oi 01 CO QO

vor^CMCM— i in V i»»

r^ ^ co 01 co 01 oo o

0101 o^-'^-ncotn

C 01
o •—
U *L * Ol
en L. O 01 in
c o en u c m f-
^ u ^ > — a.
mx — TOCin— E






     When  that  point in  time  is reached people  will begin to  ask practical
questions  and  want quantitative answers.   We  are now totally  unprepared to
provide  information  of the  specificity that  will be  required.    Our  under-
standing of the climate system and  our ability to model climate must be vastly
improved.   Some of the tasks,  such as the  development of a  realistic ocean
model appropriately interactive with atmospheric climate changes, will require
major long-term efforts.

     The greatest  need,   in  our opinion,  is for  global  observations  of the
climate system over a period of at  least a decade.  Observations are needed to
document and  quantify climate  trends, to  allow  testing  and calibration of
global  climate  models, and  to  permit  analysis  of  many small-scale  climate
processes  which  must  be  parameterized  in  the global models.   The  data needs
will require  both  monitoring from  satellites  and in situ  studies  of climate
processes.   Prestigious scientific  groups,  such as the Earth  System Sciences
Committee  appointed  by  the NASA Advisory  Council, have defined  the required
observations in detail.

Bennett, T.,  W. Broecker,  and J.  Hansen.  1985.    North Atlantic  Deep  Water
     Formation.  NASA CP-2367, 84 pp.

Blake, D. R.,  and  F.  S.  Rowland.  1986.  World-wide  increase  in tropospheric
     methane, 1978-1983.  J. Atmos. Chem. 4, 43-62.

Broecker, W. S., D.  M.  Peteet, and  D.   Rind.  1985.  Does the ocean-atmosphere
     system have more than one stable mode of operation?  Nature. 315:21-6.

Charney, J., et al.  1979.   Carbon  Dioxide  and  Climate:   A Scientific Assess-
     ment.   Washington,  D.C.: National Academy Press,  22 pp.

CLIMAP Project, A.  Mclntyre, project leader. 1981.  Seasonal reconstruction of
     the earth's surface  at the  last  glacial maximum.  Geol.  Soc.  Amer.,  Map
     and Chart Series, No.  36.

Hansen,  J., D.  Johnson,  A.  Lacis,  S.   Lebedeff,  P.  Lee,  D.  Rind, and G.
     Russell,   1981.     Climate   impact  of   increasing  atmospheric  carbon
     dioxide.  Science.  213:951-66.

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 Climate Processes  and Climate Sensitivity,  eds. J. E.  Hansen and T.
     Takahashi, pp. 130-63.   Washington,  D.C.: American Geophysical Union.

Hansen,  J., G.  Russell,  A.  Lacis,  A.   Fung,  D.  Rind, and  P.  Stone  1985.
     Climate  response  times: dependence  on  climate  sensitivity and  ocean
     mixing.  Science. 229:857-9.

Hansen, J.,  G. Russell, D.  Rind,  P. Stone, A. Lacis,  S.  Lebedeff, R. Ruedy,
     and  L. Travis.  1983.    Efficient   three-dimensional  global  models  for
     climate studies: models I and  II.  Mon. Wea.  Rev. Ill, 609-62.

Hays, J. D., J. Imbrie, and N. J. Shackleton. 1976.  Variations in the earth's
     orbit:  pacemaker of the ice ages.  Science. 194:1121-32.

Kerr, R. A. 1986.  The sun is fading.  Science.  231:339.

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.  Submitted to Geophys.  Res. Lett.

Lemon, E. R.,  ed.,  CO? and Plants:   The Response of Plants to Rising Levels of
     Atmospheric Carbon Dioxide (Westview, Boulder, CO, 1983).

Manabe,  S., and  A.   J.  Broccoli.  1985.   A  comparison  of climate  model
     sensitivity with data  from  the  last  glacial  maximum.  J.   Atmos.  Sci.

Mearns, L. 0.,  R. W  Katz, and S. H. Schneider 1984.   Extreme high-temperature
     events:   changes in  their probabilities  with changes  in  mean  temper-
     ature.  J. dim. Appl. Meteorol. 23:1601-13.

National   Academy   of   Sciences   1975.      Understanding   Climatic   Change.
     Washington, D.C.: National Academy Press, 239 pp.

Oeschger,  H.,  J.,  Beer,  V.   Siegenthaler,  B. Stauffer,  W.  Dansgaard,  C.  C.
     Langway.    Late  glacial  climate history  from ice  cores.   In  Climate
     Processes and  Climate Sensitivity,  eds. J.  E.  Hansen and T. Takahashi,
     pp. 299-306.  Washington, D.C.:  American Geophysical Union.

Ramanathan, V., R.  J. Cicerone, H.  B. Singh, and J. T.  Kiehl. 1985.  Trace gas
     trends and  their potential  role in  climate change.   J. Geophys.  Res.

Rasmussen, R.  A., and M.  A. K. Khalil. 1986.  Atmospheric trace gases:  trends
     and distributions over the last decade.  Science.  232:1623-4.

Rinsland, C. P., J.  S. Levine, and T. Miles. 1985.  Concentration of methane
     in the troposphere  deduced from 1951  infrared solar spectra.   Nature.

Shantz, W.,  and J.  Hoffman,  eds.  1986.   Rising C02  and  Changing  Climate.
     Washington, D.C.: Conservation Foundation,  (in press).

Smagorinsky,  J.,  et  al.   1982.    Carbon  Dioxide  and	Climate:   A  Second
     Assessment, Washington,  D.C.:  National Academy Press, 72 pp.

Washington, W.  M.,  and G. A.  Meehl.  1984.  Seasonal cycle  experiment  on the
     climate sensitivity due  to a  doubling of COg with an atmospheric general
     circulation model coupled to a simple mixed-layer  ocean model J.  Geophys.
     Res.  89:9475-503.

Weiss,  R.  F.   1981.    The  temporal and  spatial  distribution  of  tropospheric
     nitrous oxide.   J. Geophys.  Res. 86, 7185-95.

Wetherald,  R.  T.,  and S.  Manabe.  1986.   An investigation of cloud cover change
     in response to thermal forcing.   Climatic Change.  8:5-23.

Wuebbles, D.  J.,    F. Luther,  and J.  E. Penner.  1983.   Effect of  coupled
     anthopogenic perturbations on stratospheric ozone.   J.  Geophys.  88:1444-

The Causes and Effects of Sea Level Rise
James G. Titus
U.S. Environmental Protection Agency
Washington, DC USA

     For  the  last several  thousand  years,  sea  level  has  risen so slowly that
for most practical purposes it has  been  constant.   As a result, people and
other maritime species have had  the opportunity  to extensively develop the
shorelines  of the world.   Whether one is talking about a vacation spot  in Rio
de Janeiro, swamps in Bangladesh, farmland in the Nile  Delta,  marshes along
the Chesapeake Bay,  or the merchants of Venice, life along the  coast is in a
sensitive balance  with the level of  the  sea.

     This balance would  be upset by the rise  in sea level that  could  result
from the global  warming projected  by  Hansen  et al.  (this  volume).   Such a
warming  could  raise  sea  level  one meter  or  more  in the  next  century  by
expanding  ocean  water,  melting  mountain glaciers,   and  perhaps eventually
causing polar ice sheets  to melt or slide into  the  oceans.   Sea level rise
would  inundate  low-lying  areas,  drown  coastal  marshes and  swamps,  erode
beaches,  exacerbate  flooding, and increase the salinity of rivers, bays, and
aquifers  throughout the world.

     This paper  provides  an overview of the causes and effects of sea level


Past Trends in Sea Level

     The  worldwide average sea  level depends  primarily on (a)  the shape and
size of  ocean basins,  (b) the  amount  of water  in  the  oceans,  and  (c) the
average density  of seawater.   Subsidence,  emergence, and  other  local factors
can cause  trends  in  "relative  sea   level" at  particular  locations to  differ
from trends in "global sea level."

     Hays and  Pitman  (1973)  analyzed fossil  records and concluded  that over
the last 100 million years,  changes in mid-ocean ridge systems have caused sea
level to rise and fall over  300  meters.   However,  Clark, Farrell,  and Peltier
(1978)  have pointed  out  that  these  changes  have  accounted  for sea  level
changes of  less than  one millimeter  per century.   No  published study  has
indicated that  this determinant  of sea level is likely  to  have a  significant
impact in the next  century.

     The impact of climate on  sea  level has  been more pronounced.   Geologists
generally recognize that  during  ice ages  the glaciation of  substantial por-
tions of the northern  hemisphere has removed enough  water  from the oceans to
lower  sea   level  one  hundred  meters  below  present  levels  during  the last
(18,000 years  ago)  and  previous  ice  ages   (Don, Farrand,  and  Ewing  1962;
Kennett 1982; Oldale 1985).

     Although the glaciers that  once covered much of the northern hemisphere
have retreated, the  world's remaining ice  cover  contains enough water to raise
sea level  over seventy-five meters  (Hollin and Barry  1979).   Table  1  shows
estimates by  Hollin and  Barry (1979) and  Flint  (1971) that  existing alpine
glaciers contain enough water to  raise sea level 30 or 60 centimeters, respec-
tively.  The Greenland and West Antarctic  Ice Sheets each contain enough water
to raise sea level  about seven  meters, while East  Antarctica has enough  ice to
raise sea level over 60 meters.

     There  is  no evidence  that  either  the  Greenland  or  East  Antarctic  Ice
Sheets have completely disintegrated  in the  last two  million years.  However,
it is  generally recognized  that  sea level was  about  seven  meters  higher than
today during the last interglacial period, which was one to  two degrees warmer
(Moore  1982; Mercer  1968).   Because  the  West Antarctic Ice  Sheet is marine-
based and thought to be vulnerable to  climatic  warming,  attention  has focused
on this source  for the higher  sea  level.   Mercer  (1968) found that lake sedi-
ments and other evidence suggested that summer temperatures  in Antarctica have
been 7° to 10°C higher than today at some  point  in  the last  two million years,
probably during the last  interglacial period  125,000  years  ago, and that such
temperatures could  have  caused  a disintegration  of the  West  Antarctic  Ice

     Tidal  gauges have been available to  measure  the change  in sea  level at
particular  locations over the  last century.   Studies combining these measure-
ments to estimate global trends have concluded that sea level has risen  1.0 to
1.5 mm/yr during the last century (Barnett 1983; Gornitz, Lebedeff, and  Hansen
1982; Fairbridge and Krebs 1962).  Barnett (1983)  found the rate of sea level
rise for the  last  fifty years  to  be  between  2.0  and 2.5 mm/yr, while in the
previous fifty years there was  little change; however, the acceleration  of the
rate of sea level  rise was  not  statistically  significant.   Emery and  Aubrey
(1985,  n.d.)  have  filtered  out estimated  land  surface movements  in  their
analyses  of tidal  gauge  records in  northern    Europe and  western   North
America,  and  have  found  an acceleration  in  the  of sea level rise over the
last century.'  Braatz  and  Aubrey  (n.d.)  have  found  that the rate  of
   This result  was  reported in the North  America  study.   The data also shows
   it  to  be true  in the  northern  Europe study,  but the result  was not re-
   ported.   David  Aubrey,  Woods  Hole  Oceanographic institute,  Woods Hole,
   Massachusetts, personal communication.


   Table  1.  Snow and  Ice  Components  (Modified  from Hollin  and Barry 1979)
(10b km3)
Land ice:  East Antarctica13               9.86        25.92        64.8
   West Antarctica0                       2.34        3.40          8.5
   Greenland                              1.7         3.0           7.6
   Small ice caps and mountain
     glaciers (Hollin and Barry 1979;     0.54        0.12          0.3
     Flint 1971)                                                    0.6

Permafrost (excluding Antarctica):                                  0.6

   Continuous                             7.6         0.03         0.08
                                                      to            to
   Discontinuous                         17.3         0.7          0.17

Sea ice:  Arctic
     Late February                       14.0         0.05
     Late August                          7.0         0.02

     September                           18.4         0.06
     February                             3.6         0.01
Land Snow Cover

   N. Hemisphere
     Early February                      46.3         0.002
     Late August                          3.7

   S. Hemisphere
     Late July                            0.85
     Early May                            0.07

3  400,000 km3 of ice is equivalent to 1 m global sea level.
   Grounded  ice  sheet,  excluding  peripheral,  floating ice  shelves  (which do
     not affect sea  level).   The shelves have a  total  area of 1.62 x 10  km
     and a volume of 0.79 x 10° knH (Drewry and Heim 1983).
°  Including the Antarctic Peninsula.
   Excluding the Sea of Okhotsk,  the  Baltic Sea,  and the Gulf of St. Lawrence
     (Walsh  and  Johnson 1979).   Maximum ice extents  in these  areas  are 0.7
     million, 0.4 million,  and 0.2 million km ,  respectively.
e  Actual  ice  area  excluding  open water  (ZwaJJy et al.  1983).   Ice extent
     ranges between 4 million and 20 million km .
   Snow  cover  includes that  on  land ice  but  excludes snow-covered  sea ice
     (Dewey and Heim 1981).

Source:  Glaciers, Ice Sheets, and Sea Level, National Academy Press, p. 272.


relative sea level rise on  the  east  coasts  of North America accelerated after

     Several  researchers  have  attempted  to  explain  the  source of  current
trends in sea level.  Barnett  (1984)  and Gornitz,  Lebedeff, and Hansen (1982)
estimate that thermal  expansion of  the  upper layers of  the  oceans  resulting
from the observed global warming of 0.4°C in the last century could be respon-
sible for  a rise  of 0.4  to 0.5 millimeter per year.   Roemmich and  Wunsch
(1984) examined  temperature and salinity measurements at Bermuda,  found that
the  4°C  isotherm had  migrated  100  meters  downward,  and concluded  that  the
resulting expansion of ocean water could be responsible for  some or all of the
observed rise in relative sea level.   Roemmich (1985)  showed that the warming
trend 700 meters below the  surface  was  statistically significant.   However,
Barnett  (1983)  found no significant trend based on  an examination of the upper
layers of the ocean.   Nevertheless,  Braatz  and  Aubrey  (n.d.)  note that long-
term steric changes  in  the ocean are not confined  to  the upper layers of the
oceans,  which implies  that  the Barnett analysis does  not necessarily contra-
dict the Roemmich and Wunsch conclusion.

     Meier (1984) estimates that retreat of alpine  glaciers  and small ice caps
could be responsible  for  a  current  contribution to sea  level  of between  0.2
and 0.72 millimeters per  year.   The National  Academy  of  Sciences (NAS) Polar
Research Board  (Meier  1985) concluded  that existing  information  is insuffi-
cient to determine whether  the impacts of Greenland and  Antarctica  are posi-
tive, negative,   or  zero.    Although  the  estimated  global warming  of the last
century  appears  at least partly responsible for the last century's rise in sea
level, no study  has demonstrated  that  global warming might  be responsible for
an accelerated rate of sea level rise.

Impact of Future Global Warming on Sea Level

     Concern  about  a  substantial  rise  in  sea level  as  a  result  of  the
projected global warming stemmed  originally  from Mercer  (1968), who suggested
that  the  Ross  and Filchner-Ronne ice shelves  might disintegrate,  causing  a
deglaciation of  the the West Antarctic  Ice  Sheet and a resulting six to seven
meter rise in sea level,  possibly within 40 years.

     Subsequent   investigations  have concluded  that  such  a  rapid rise  is
unlikely.   Hughes  (1983)  estimated  that  such a disintegration would  take at
least two hundred years, and Bentley  (1983), five  hundred.   Other researchers
have estimated that this process could take considerably longer (Fastook 1985;
Lingle 1985).

     Researchers have  turned their  attention to the  magnitude of  sea level
rise  that might occur  in the  next  century.   The best understood factors  are
the  thermal expansion  of  ocean water and the  melting  of  alpine glaciers.   In
the National Academy of Sciences  report  Changing Climate, Revelle (1983) used
the model of Cess  and  Goldenberg (1981)  to  estimate temperature increases at
various   depths   and  latitudes   resulting  from  a 4.2°C  warming  by  2050-2060
(Figure  1).   While noting  that  his assumed  time constant of 33 years probably
resulted  in a   conservatively  low  estimate,  he  estimated  that  temperature
increases would  result in an expansion  of the upper ocean sufficient to raise
sea level thirty centimeters.

             60° N     40° N     20° N

0      20°S     40°S     60°S     80°S














1-6 I






| 3.5






3.5 :

-"1 '














3 2
2.5 1.7

1.7 1.1

0.8 0.5

     Figure  1.   Computed near-equilibrium  changes  in ocean temperature
     for a doubling of  atmospheric  carbon dioxide and probable increase
     in  other  greenhouse gases,  about  2080.   The  surface temperature
     increase is based on Flohn's (1982) prognosis.
     Using a model  of the oceans developed  by   Lacis et al. (1981), Hoffman,
Wells, and  Titus (1985)  examined a  variety of possible  scenarios of future
emissions  of greenhouse  gases and  global  warming.    They estimated  that  a
warming of between  1° and  2.6°C  could result in a thermal expansion contribu-
tion to sea level of between 12 and 26 cm by 2050.  They also estimated that  a
global warming of 2.3°  to 7.0°C  by 2100  would  result in thermal expansion of
28-83 cm by that year.

     Revelle  (1983)  suggested that  while  he  could  not estimate  the future
contribution of alpine glaciers to sea level rise, a contribution of 12 centi-
meters through 2080 would be  reasonable.    Meier  (1984)  used glacier balance
and  volume  change  data  for  twenty-five  glaciers  where  the  available record
exceeded fifty years to estimate the relationship between historic  temperature
increases and the resulting negative  mass balances of the glaciers.  He esti-
mated that  a 28-mm  rise had  resulted from a warming  of 0.5°C,  and concluded
that a  1.5°  to 4.5°C warming  would result  in  a rise of 8-25  cm in the next
century.   Using these results, the  WAS Polar Board concluded that  the contri-
bution of glaciers  and small ice caps through 2100 is likely to be  10 to 30 cm
(Meier et al.  1985).   They noted that the gradual depletion of remaining  ice
cover might reduce  the  contribution of sea level rise somewhat.  However,  the
contribution might  also be greater,  given that the  historic  rise took place
over  a  sixty-year  period, while  the  forecast  period  is  over  one  hundred
years.  Using  Meier's estimated  relationship  between global  warming  and  the
alpine contribution, Hoffman,  Wells, and Titus (1986) estimated alpine contri-
butions through 2100 at 12-38  cm for a global warming of 2.3° to 7.0°C.

     The  first  published estimate  of the  contribution of Greenland  glacier
meltwater to sea level was Revelle's (1983)  estimate of 12 cm through the year
2080.   Using  estimates  by Ambach  (1980,   1982)  that  the equilibrium  line
(between snowfall accumulation and melting)  rises one  hundred  meters for each
0.6°C rise in air temperatures, he concluded that the projected 6°C warming of
Greenland would  be likely  to  raise  the  equilibrium  line  1,000 meters.   He
estimated that such a change in the  equilibrium line  would result in  a 12-cm
contribution to sea level rise  for the next  century.

     The  MAS  Polar Board  (Meier  et al.  1985)  noted that  the  large ablation
area makes Greenland a  "significant potential  contributor  of meltwater to the
ocean if  climatic  warming causes an  increase  in the rate  of  ablation and an
upward shift of the equilibrium line."  They found  that a 1,000-m rise in the
equilibrium  line  would  result  in a contribution  of  30 cm through  2100.
However, because Ambach  (1985) found  the  relationship  between  the equilibrium
line and temperature to  be  77  meters  per  degree (C), the panel concluded that
a 500-m  shift in  the equilibrium line would  be more  likely.   Based on the
assumption of a  6.5°C warming by  2050 and   constant temperatures thereafter,
the  panel  estimated  that such a  change would  contribute  about  10  cm to sea
level through 2100,  but also  noted that  "for an extreme  but  highly unlikely
case, with  the  equilibrium line  raised  1,000 m, the  total rise  would  be 26
centimeters." Although  Bindschadler  (1985),  had   treated  the  two cases  as
equally plausible, his analysis  was  conducted before  the results  of Ambach
(1985) were known;  he has since indicated  agreement  with the findings of Meier
et al. (1985).

     Available estimates  of the  Greenland contribution assume  that all melt-
water flows  into the  oceans and that the ice  dynamics  of  the  glaciers do not
change.  The NAS Polar  Board suggested  that some of the water would refreeze,
decreasing  the  contribution to  sea  level  rise.   Although a change  in ice
dynamics  might  imply  additional  deglaciation  and  increase the  rate  of sea
level rise, the panel  assumed  that  such changes were unlikely  to occur in the
next century.

     The  potential impact  of  a  global  warming on Antarctica  in the  next
century  is  the  least  certain  of all  the factors  by  which a global  warming
might contribute to sea level rise.  Meltwater from East Antarctica might make
a significant  contribution by  the  year 2100,  but  no  one  has  estimated it.
The  contribution of ice  sliding into  the  oceans,  known as "deglaciation," has
been the subject of several studies.

     Bentley  (1983)   examined  the  processes  by  which  a  deglaciation  of
Antarctica might occur.  First an accelerated melting of the undersides of the
Ross  and  Filchner-Ronne  ice shelves  would  occur  due  to  warmer  water circu-
lating underneath them.  The thinning of these ice shelves could  cause them to
become unpinned and their grounding lines to retreat.  Revelle  (1983)  suggests
that  the ice  shelves  might  disappear  in   100  years, after  which  time the
Antarctic  ice  streams would flow  directly  into the oceans, without the back
     Robert  Bindschadler,  Goddard  Space Flight Center,  Greenbelt, Maryland,
     personal communication.
     James  Hansen,  Goddard  Institute for  Space Studies, New  York, personal
     communication to J.S. Hoffman.

pressure  of  the  ice  shelves.   Bentley  suggests  that all  the ice  could be
discharged over a period of 500 years.

     Although  a West  Antarctic deglaciation  would  occur  over  a period of
centuries, it  is possible  that an  irreversible deglaciation  could  commence
before 2050.    If the  ice shelves thinned more  than  about one meter per year,
Thomas, Sanderson, and Rose  (1979) suggested  that  the ice  would move into the
sea at  a sufficient  speed  that even  a cooling back to the  temperatures of
today would not be sufficient to result in a reformation of the ice shelf.

     To  estimate the  likely  Antarctic  contribution for  the next  century,
Thomas (1985)  developed  four scenarios measuring  the impact  of a 3°C global
warming by 2050:

     •  A shelf  melting  rate of 1 m/yr with seaward  ice fronts  remaining at
        present locations—implies a rise of 28 cm by the year 2100.

     •  A shelf melting rate  of 1 m/yr with ice fronts calving back to a line
        linking  the  areas where  the  shelf is  grounded,  during  the  2050s—
        implies a rise of 1.6 m by 2100.

     •  Same  as  case  1  but  with a melt rate of  3 m/yr—implies a  rise of 1 m
        by 2100.

     •  Same  as case 2 but with a melt rate of 3 m/yr—implies a rise of 2.2 m
        by 2100.

     Thomas  concluded  that  the 28-cm rise  implied  by case  1 would  be most
likely.   He  also stated that even if  enhanced  calving  did  occur,  it  would be
likely to occur after  2050,  "suggesting that  probably associated sea level
rise would be closer to the 1 m of case 3 than the 2.2 m of case 4."

     The  MAS  Polar  Board (Meier et  al.  1985) evaluated the  Thomas study and
papers by Lingle (1985) and  Fastook  (1985).    Although  Lingle estimated that
the contribution  of West Antarctica through 2100  would  be  3  to 5 cm,  he did
not evaluate  the contribution  from  East  Antarctica, while  Fastook  made no
estimate  for  the year  2100.    Thus,  the  panel   concluded  that  "imposing
reasonable limits" on Thomas' model yields  a  range of 20 to 80 cm by 2100 for
the Antarctic  contribution.    However, they also  noted several factors that
could reduce  the amount of  ice discharged  into the sea:   the removal of the
warmest  ice   from  the   ice   shelves,  the  retreat  of grounding   lines,  and
increased lateral shear  stress.  They also concluded that  increased  precipi-
tation  over   Antarctica  might  increase  the size  of  the   polar   ice  sheets
there.   Thus,  the panel concluded that  Antarctica could cause a  rise in sea
level up  to  1 m, or  a drop  of 10  cm, with  a  rise between 0 and 30  cm most

     Table 2  summarizes  the  various  estimates of global sea level rise.  The
report Projecting Sea Level  Rise (Hoffman,  Keyes,  and Titus  1983) estimated
that the rise would be between 56 and 345 cm, with a probable rise between 144
and 217  cm.    Revelle (1983) estimated that the rise was likely  to be 70 cm,
ignoring the  impact of a global warming on Antarctica; Revelle also noted that
the latter contribution was  likely to  be  1  to 2 m per century after 2050, but
declined to add that to his estimate.   The NAS Polar Board (Meier et al. 1985)

          Table 2.   Estimates of Future Sea Level Rise (centimeters)
Year 2100 by
NAS (1983)
EPA (1983)
NAS (1985)1*
Thomas (1985)
Hoffman et al
Total Rise in
NAS (1983)
EPA (1983)
Hoffman et al
Cause (2085 in the
. 1986 28-83
Specific Years:**

low 8.8
high 13.2
. 1986
case of NAS


Greenland Antarctica
12 * 70
# # 56-345
10-30 -10-+100
6-27 12-220
2050 2075 2085

23 38
53 91
79 137
117 212

20 36 44
55 191 258




*   Revelle (1983) attributes 16 cm to other factors.

**  Only EPA reports made year-by-year projections for the next century.

#   Hoffman et al.  (1983)  assumed  that the glacial contribution  would be one
    to two times the contribution of thermal expansion.

##  NAS  (1985)   estimate   includes  extrapolation  of  thermal  expansion  from
    Revelle (1983).

Sources:  Hoffman et al.  (1986); Meier et al.  (1985);
          Hoffman et al.  (1983); Revelle (1983);  Thomas  (1985).

projected that  the  contribution  of glaciers would be  sufficient  to raise sea
level 20 to  160 cm, with a rise  of "several tenths of a meter"  most likely.
Thus,  if one  extrapolates  the   earlier  MAS  estimate of  thermal  expansion
through the year  2100,  the 1985  NAS report  implies a  rise  between 50 and 200
cm.   The estimates of  Hoffman,  Wells, and  Titus  (1985)  were similar  to the
estimates of Hoffman  Keyes,  and  Titus (1983)  for the year 2100,  but for the
year 2025,  they lowered their estimate from 26-39 cm to 10-21  cm.

Future Trends in Local Sea Level

     Although most  attention has  focused on projections of global sea level,
impacts on particular  areas  would depend on local  relative sea  level.   Tidal
gauge measurements  suggest that  relative sea level has risen  10  to 20 cm per
century more rapidly  than the  worldwide average along much of the U.S. coast
(Hicks, DeBaugh, and Hickman  1983). However, Louisiana is subsiding close to  1
m  per  century,  while  parts of Alaska  are emerging  10  or more  cm per century.
Bruun  argues  that  throughout  most of the  world,  sea level has  been rising.
However, Bird   (Volume  4) argues  that  sea  level appears  to  be  stable  in
Australia,  which may imply that future rates of sea level rise will also be 10
to 15 cm per century less than the worldwide average.

     Local   subsidence  and  emergence  are  caused  by  a variety   of  factors.
Rebound  from the  retreat of glaciers  after the last ice age  has resulted in
the  emergence  of Alaska  and parts  of Scandinavia.   The  emergence  in polar
latitudes has resulted  in subsidence  in  other  areas.   Groundwater pumping has
caused  rapid  subsidence around Houston,  Texas; Taipai, Taiwan;  and Bangkok,
Thailand, among other areas  (Leatherman  1984;  Kuo,  Volume 4).   River deltas
and other newly created land  subside as the unconsolidated materials compact.

     Although  subsidence  and  emergence  trends  may  change   in   the  future,
particularly where anthropogenic  causes are curtailed, no one  has linked these
causes to future  climate  change in the next century.   However, the removal of
ice  from Greenland and  Antarctica would deform  the  ocean  floor.   Clark and
Lingle have  (1977)  calculated the  impact  of a uniform  1-m contribution from
West  Antarctica.    They  concluded that  relative  sea level  at  Hawaii would
increase by an additional 25 cm,  and that along much of the U.S.  Atlantic and
Gulf Coasts there would be an  additional  15 cm.  On the other hand, sea level
would drop at Cape  Horn by nearly 10 cm,  and the rise along the southern half
of the Argentine and Chilean  coasts would be less than 75 cm.

     Other  influences  on  local sea  level  that might change as a result of  a
global warming  include currents,  winds,  and freshwater flow  into estuaries.
None of these impacts, however, has been estimated.


     A  rise  in sea level of  one  or  two  meters  would permanently inundate
wetlands  and   lowlands,   accelerate  coastal  erosion,   exacerbate  coastal
flooding, threaten  coastal structures,  and  increase  the  salinity of estuaries
and aquifers.   Substantial research has been  done  on the,  implications of sea
level rise for coastal  erosion and wetlands, while relatively little work has
been done in the other areas.

Submergence of Coastal Wetlands

     The most direct impact of a  rise  in  sea level is the inundation of areas
that had been  just  above the water level before  the  sea  rose,  also described
by Park, Armentano,  and  Cloohan  in Volume 4.   Coastal  wetlands are generally
found at  elevations below  the  highest tide  of the  year  and above  mean  sea
level.   Thus,  wetlands  account for most of  the land  less than  1 m above  sea

     Because a  common means  of  estimating  past  sea  level rise has  been  the
analysis of marsh peats, the  impacts of sea  level rise  on wetlands are fairly
well understood.  For the rates of sea level  rise of the last several thousand
years,  marshes  have  generally kept pace with sea  level  through sedimentation
and peat formation  (Emery  and Uchupi 1972;  Redfield  1972,  1967; Davis 1985).
As sea level  rose,  new  wetlands formed inland  while  the  seaward boundary  was
maintained  (Figure  2a,  2b).   Because the wetland area has  expanded,  Titus,
Henderson, and Teal  (1984) hypothesized that  one  would  expect a concave marsh
profile,  i.e.,  that  there  is  more marsh area than the  area found immediately
above the  marsh.    Thus,  if sea  level rose  more rapidly  than the  marsh's
ability  to  keep pace,  there would  be a  net  loss of  wetlands  (Figure  2c).
Moreover,   a  complete  loss  might  occur   if protection  of  developed  areas
prevented the inland formation of new wetlands (Figure 2d).

     Kana, Baca, and Williams   (1986)  and  Kana et al.  (n.d.)  surveyed marsh
transects in the areas of  Charleston, South  Carolina, and two sites near Long
Beach Island,  New  Jersey,  to evaluate  the  concavity  of  wetland profiles  and
the vulnerability of wetlands to  a  rise  in  sea  level.   Their  data from  the
Charleston area showed all the marsh to be between 30  and 110 cm above current
sea level,  an  elevation range of  80 cm.   The  area with  a similar elevation
range just above the marsh was only  20  percent  as large.   Thus, a rise in sea
level exceeding vertical  marsh  accretion  by  80 cm would  result  in an  80
percent  loss of wetlands.   In the New Jersey sites,  the  marsh  was  also found
within an  elevation range of  80  cm;  a rise  in sea level 80 cm in excess  of
marsh accretion would result in 67 to 90 percent losses.

     The future ability of marshes to accrete vertically  is uncertain.  Based
on field studies by Ward and Domeracki (1978), Hatton  et al.  (1983), Meyerson
(1972),  and  Stearns and MacCreary (1957),  Kana et al.  (n.d.)  concluded that
current  vertical  accretion rates  are  approximately 4  to  6 mm/yr  in  the  two
case study  areas,  greater than  the  current  rate  of  sea level  rise but less
than the  rates  of  rise projected  for the next  century.   If current accretion
trends continue,  then  87  and  160 cm  rises  by  2075  would  imply  50  and  80
percent  losses  of wetlands in the Charleston area.   Kana  et al. (n.d.)  also
estimated 80 percent losses  in the New  Jersey sites for a  160-cm rise through
2075.  However, because  the high marsh  dominates in that area,  they concluded
that the principal  impact  of  an  87-cm rise  by 2075 would be the conversion of
high marsh to low marsh.

     In both  cases,  the losses of marsh  could  be greater if inland areas are
developed and  protected  with bulkheads or levees.  Because there is a buffer
zone  between  developed  areas  and  the marsh  in  South  Carolina  protecting
development  from a  160-cm rise  would  increase the loss  from  80-90 percent.
Without the buffer,  the loss would be close to  100 percent.

                A.  5000 Years  Ago
                                       -z— Sea Level
                      B.  Today
    Sedimentation and
    Peat Formation
Sea Level

Sea Level
                     C.  Future
   Substantial Wetland Loss Where There is Vacant Upland
                                         -Z- Sea Level
                                            Sea Level
                     D.   Future

         Complete Wetland Loss Where House is Protected
                 In Response to Rise In Sea Level
                                          . Sea Level
                                          Sea Level
Figure 2.  Evolution  of Marsh as Sea Level  Rises

    . Louisiana, whose marshes and swamps account for 40 percent of the coastal
wetlands  in  the  United  States  (excluding  Alaska),  would  be  particularly
vulnerable to an accelerated rise in sea level.   The majority of the Louisiana
wetlands are less than one meter  above  sea  level,  and  are generally subsiding
approximately one meter per  century as its deltaic  sediments  compact (Boesch
1982).   Until  the  last  century,  the  wetlands  kept pace  with this  rate  of
relative  sea  level  rise,  because  of  the  sediment  the  Mississippi  River
conveyed to the wetlands.

     Human activities,  however, have largely disabled the natural processes by
which coastal Louisiana might keep pace with sea level  rise.  Dams, navigation
channels, canals,  and  flood  protection levees  have interrupted the  flow  of
sediment, fresh water, and  nutrients to the wetlands.    As  a  result,  over 100
square kilometers  of wetlands convert  to open  water every year (Gagliano  et
al.  1983).    A  substantial  rise  in  sea  level  would  further  accelerate  the
process of wetland loss in Louisiana.

     To  develop  an understanding of the potential  nationwide  impact of sea
level  rise  on  coastal  wetlands  in  the United  States,  Park,  Armentano,  and
Cloonan  (Volume 4) use topographic maps to  characterize  wetland elevations  at
fifty-two  sites comprising  4800  square kilometers  (1.2  million acres)  of
wetlands, over  17  percent of  all  U.S.  coastal  wetlands  (see  also Armentano,
Park,  and  Cloonan  n.d.).   Using  published  vertical  accretion  rates,  they
estimate the impact of 1.4 and 2.1  m rises  in sea level  through the year 2100
for each of the sites.  Park,  Armentano,  and  Cloonan (Volume 4) estimate that
these  scenarios  imply losses  of 40  and  76 percent of  the  existing coastal
wetlands in their  sample, which could  be reduced  to 22  and 58 percent if new
wetlands are allowed to form  inland.   However,  Titus (n.d.) found that if the
Park et al. sample is weighted according to  an inventory  of wetlands in parti-
cular areas (Alexander, Broutman, and Field  1986),  the resulting estimates  of
U.S.   wetland loss are somewhat  higher,  47-82  percent  of  existing wetlands,
with a potential for reducing those losses to 31 to 70  percent.

     Throughout the  world,  people have dammed,  leveed,  and channelized  major
rivers,  curtailing the amount  of  sediment that  reaches river deltas.   Even at
today's  rate of sea  level rise,  substantial amounts of land are converting to
open water  in Egypt  and Mexico (Milliman  and  Meade 1983).  Other deltas, such
as the Ganges in Bangladesh and India, are currently expanding seaward.  These
areas would require increased sediment, however, to keep  pace with an acceler-
ated  rise  in sea level.   Additional projects  to  divert the  natural flow  of
river water would  increase the vulnerability of these areas  to a rise in sea
level.   Broadus  et al.  (Volume 4) examine  this  issue  in detail for Egypt and

     Plate  4  (appearing  immediately before this section)  shows the projected
land loss for Bangladesh  for two scenarios:

     •   Low—a 50-cm global rise in sea level with current rates of subsidence
         and sedimentation

     •   High—a  200-  to  250-cm  rise  with  alterred   sedimentation due  to
         additional dams.

     Several options have been identified for reducing wetland loss due to sea
level rise.   Abandonment of developed areas inland of today's wetlands could
permit new  wetlands to form inland.   In some  cases,   it might  be  possible to
enhance the ability of wetlands  to  accrete  vertically by spraying  sediment on
them  or—in the  case  of Louisiana and  other  deltas—restoring  the  natural
processes that  would  provide sediment  to the  wetlands.   Finally,  some local
governments in  Louisiana have proposed  to artificially control water levels
through the use of levees and pumping stations  (Edmonson and Jones  1985).

     The need for anticipating sea level rise would vary.  Artificial means to
accelerate  wetland  accretion need  not  be  implemented  until  the  rise  takes
place (although a lead  time would  be necessary  to develop the required tech-
nologies).   Similarly, levees  and pumping stations could  be  delayed.   On the
other hand, a planned retreat would  require several   decades of lead  time to
permit the  design of  new mobile structures and  the  depreciation  of  the old
immobile structures.


     Although coastal wetlands are  found at  the  lowest elevations, inundation
of lowland  could  also be  important in some areas, particularly if  sea level
rises at least one  meter.   Unfortunately, the  convention of ten-foot contours
in the  mapping of  most  coastal  areas  has  prevented  a  general  assessment of
land  loss.   Although  a  few case  studies have  been  conducted  in  the United
States,  very   few studies  have   been  undertaken  to   quantify  the  potential
impacts on other countries, other than the paper by Broadus et al.  (Volume 4).

     Kana   et  al.  (1984)  used data from aerial   photographs to assess eleva-
tions  in  the  area  around Charleston.   They concluded  that 160- and 230-cm
rises would result in 30 and 46 percent losses  of the  area's dry land,  respec-
tively.  Leatherman (1984)  estimated  that such rises  would result  in 9 and 12
percent losses  of the land  in  the area  of  Galveston and Texas City, Texas,
assuming that  the elaborate network  of seawalls and   levees  were  maintained.
(Many of  the  summary  results  from Leatherman  1984;   Kana  et al.    1984; and
Gibbs 1984 appear in the appendix of Titus et al. 1984.)

     Schneider and  Chen   (1980)  conducted the  first  nationwide  assessment of
the  inundation  from projected sea  level rise.   Unfortunately, the smallest
rise  in  sea level  they  considered  was  a 4.5-m (15-ft)  rise,  in part  because
smaller contours  are  not generally available in  topographic  maps.   Neverthe-
less, their findings  suggest which coastal  states would  be  most  vulnerable:
Louisiana  (which  would  lose 28  percent of its  land and  51 percent  of its
wealth),   Florida   (24   and  52   percent),   Delaware   (16   and  18  percent),
Washington, D.C.  (15  and  15 percent),   Maryland  (12  and 5 percent),  and New
Jersey (10 and 9 percent).

     As with  wetland  loss,  the responses to inundation  fall  broadly into the
categories  of  retreat  and holding back  the  sea.   Levees are used  extensively
in the Netherlands  and New  Orleans  to prevent  the flooding of areas below sea
level  and   could  be  similarly  constructed  .around other  major cities.   In
sparsely developed  areas,  however,  the cost of a levee  might be greater than
the value of  the  property  being  protected.   Moreover, even where levees prove
to  be cost-effective,  the  environmental implications  of replacing  natural
shorelines with manmade structures would need to be considered.

Coastal Erosion

     Sea  level  rise  can also  result in  the loss  of land  above sea  level
through erosion.   Bruun  (1962)  showed that the erosion resulting  from a rise
in sea  level would depend upon the average slope of  the  entire beach profile
extending from  the dunes out  to the  point  where the water  is too  deep for
waves to  have a significant impact on the bottom (generally  a  depth  of about
10 meters).  By comparison, inundation depends only  on the slope  immediately
above the original sea level.   Because  beach profiles are generally flatter
than the portion of the beach just above  sea  level,  the "Bruun Rule" generally
implies that the  erosion from  a  rise in sea  level  is several  times greater
than the amount of land directly inundated.   (See  Figure  3.)

     As Bird (Volume  4)  emphasizes,  processes other  than  sea level rise also
contribute to erosion, including  storms, structures,  currents,  and alongshore
transport.  Because sea  level  has  risen slowly in recent centuries, verifi-
cation of the Bruun Rule  on the open coast  has been  difficult.  However, water
levels along the Great Lakes can  fluctuate over one meter in  a decade.  Hands
(1976,  1979, and 1980) and  Weishar and Wood  (1983)  have  demonstrated that the
Bruun Rule generally predicts the erosion resulting  from  rises in water levels

     The Bruun  Rule has  been applied  to  project erosion due to sea level rise
for  several  areas  of  the United States where it  is believed  to adequately
project future  erosion.   Bruun (1962)  found  that  a  1-cm rise in  sea level
would generally result in a 1-m shoreline  retreat,  but that the retreat could
be as great as  10  m along some  parts  of  the  Florida coast.  Everts (1985) and
Kyper and Sorensen (1985), however,  found that along the  coasts of Ocean City,
Maryland, and   Sandy  Hook,  New Jersey,  respectively,  the shoreline retreat
implied by the Bruun Rule would be only about 75 cm.   Kana et  al. (1984) found
that along  the  coast of  South  Carolina,  the  retreat  could be 2 m.   The U.S.
Army  Corps  of  Engineers  (1979)   indicated   that  along   the  coast  of  San
Francisco, where waves are generally larger than along the Atlantic Coast, the
shore might retreat 2-4 m for a 1-cm rise in  sea level.

     Dean  and   Maurmeyer (1983)  generalized  the  "Bruun  Rule"  approach  to
consider  the "overwash" of barrier islands.  Geologists basically believe that
coastal  barriers  can  maintain  themselves in  the face of slowly  rising sea
level  through  the  landward  transport of sand, which washes over  the island
during  storms,  building  the island upward and landward.   Because  this formu-
lation of the Bruun Rule extends the beach profile horizontally to include the
entire  islands  as  well  as  the active surf  zone,  it  always  predicts greater
erosion than the  Bruun Rule.    However,  the  formulation may not be applicable
to developed barrier islands, where the common practice of public officials is
to bulldoze sand back onto the  beach after a major storm.

     The  potential erosion  from a rise in  sea  level could  be particularly
important  to  recreational  beach  resorts,  which  include  some of  the world's
most  economically  valuable  and intensively used land.  Relatively few of the
most  densely  developed  resorts have  beaches  wider  than  about 30 m at high
tide.   Thus, the rise  in relative sea  level of 30 centimeters projected in the
next  40 to  50  years could erode most  recreational beaches  in developed areas,
unless  additional  erosion response measures are taken.

Inundation When
Sea Level Rises
Erosion Due  to
Sea Level Rise
                           Figure 3.   The  Bruun  Rule

     Bruun (Volume 4) examines potential responses  to  erosion  in considerable
detail  (see  also  Magness 1984;  U.S.  Army  Corps  of  Engineers  1977).   The
responses fall  generally into three  categories:    construction  of walls  and
other  structures,   the   addition  of  sand  to  the  beach,  and   abandonment.
Although seawalls have  been  used  in the past, they  are  becoming increasingly
unpopular among shore communities because erosion can  proceed  up to the wall,
resulting in a complete  loss  of beach,  which has happened in many areas (Kyper
and  Sorensen  1985; Howard,  Pilkey, and  Kaufman  1985).   A  number of other
structures have been used to decrease the ability of  waves  to cause  erosion,
including groins  (jetties) and breakwaters.   Bulkheads  are often  used where
waves are small (Sorensen, Weisman,  and Lennon 1984).

     A more  popular form of erosion  control has been the placement  of sand
onto the beach.   Although costs  can exceed one million  dollars  per kilometer
(U.S. Army Corps of Engineers 1980; Howard, Pilkey, and Kaufman 1985), it is
often justified by  the  economic  and recreational  value of beaches.   A recent
study of Ocean City, Maryland, for example,  concluded that  the cost of holding
back  the  sea  for  a  30-cm  rise  in sea  level would  be  about  25 cents  per
visitor,  less  than  1   percent  of  the  cost  of  a  trip  to  the  beach  (Titus
1985).  That community  also  provides an example of  the practical consequences
of sea  level  rise.  Until  1985,  the State of Maryland's  policy  for  erosion
control was the construction of  groins,  which curtail erosion caused  by sand
moving along the shore,  but  not  erosion caused by sea level rise.   Sea level
rise was  cited  as  the motivating concern for  the state to abandon  the groin
plan and use beach replenishment,  which can  effectively control erosion caused
by both types of erosion (Associated Press 1985).

     Although shore  protection is  often  cost-effective  today,  the favorable
economics might change  in the future.  A more rapid rise in  sea  level would
increase  the  costs of  shore  protection.   A  number  of  states  have  adopted
erosion  policies   that   assume  a  retreat  from  the  shore.    North  Carolina
requires homes  that  can be moved to be set back from  the  shore  by a distance
equal to shoreline  recession from 30 years of erosion,  while  high-rises must
be  set  back  60  years.     Maine  requires  people  to  demonstrate  that  new
structures will not erode for  100 years.    Other  jurisdictions discourage the
construction of bulkheads and seawalls (Howard, Pilkey, and Kaufman 1985).  As
Bird  (Volume  4) discusses,  in many undeveloped  countries,  small,  relatively
inexpensive houses  are  found very  close to  the shore.  Because  the  value of
these houses is less  than the cost of protecting them,  they must  be  moved as
the  shore erodes.   An  accelerated rise in sea level would speed this process
of shoreline retreat.

     The need for anticipating erosion caused by sea level  rise varies.  Where
communities are  likely   to adapt  to erosion,  anticipation can  be important.
The  cost  and  feasibility of  moving a house back depends  on  design decisions
made when the house is built.  The willingness of  people to abandon properties
depends  in  part on whether  they  bought  land  on  the assumption  that  it would
eventually  erode  away  or had assumed  that the  government would  protect it
   North  Carolina  Administrative  Code,  Chapter  7H,  1983.    Raleigh,  North
   Carolina:  Office of Coastal Management.
   Fred  Michaud,  Office  of  Floodplain Management,  State of  Maine,  personal

indefinitely.  Less anticipation  is necessary  if  the shore will be protected;
sand can  be added  to the  beach  as necessary.   Nevertheless,  some  advanced
planning may be necessary for communities to know whether retreat or defending
the shore would be most cost effective.


     A rise  in sea  level  could  increase  flooding  and storm damages in coastal
areas for three reasons:   erosion caused by sea level rise would increase the
vulnerability of  communities;  higher  water levels would  provide storm surges
with a  higher base to build  upon; and higher  water  levels  would  decrease
natural and artificial drainage.

     The  impact  of erosion  on  vulnerability  to  storms is generally  a major
consideration  in  projects  proposed  to  control  erosion,  most  of  which  have
historically been  funded through  the  Army Corps  of Engineers  in  the United
States.   The  impact  of  sea   level  rise, however,  has  not  generally  been
considered separately from other causes of erosion.

     The impact of  higher base  water levels on flooding has been investigated
for the areas  around  Charleston,  South  Carolina,  and Galveston,  Texas (Earth
and Titus  1984).    Kana et  al.  (1984)  found that  around  Charleston,  the area
within the  10-year  flood  plain  would increase  from 33 percent in 1980, to 48,
62, and  74   percent for  rises  in  sea  level of 88,  160,  and 230  cm,  respec-
tively, and  that the  area within  the  100-year  flood plain would increase from
63 percent  to  76,  84, and  90  percent  for the three  scenarios.   Gibbs (1984)
estimated  that even  an  88-cm rise  would double the  average annual  flood
damages  in   the  Charleston  area  (but that flood  losses  would not  increase
substantially for  higher  rises in  sea level  because shoreline retreat would
result in a large part of the community being completely abandoned).

     Leatherman  (1984)  conducted a   similar  analysis   of Galveston  Island,
Texas.    He  estimated that  the area  within  the  100-year flood  plain  would
increase  from  58  percent to 94 percent  for an  88-cm rise in  sea level,  and
that for  a  rise  greater  than one meter,  the Galveston  seawall  would be over-
topped during a  100-year  storm.  Gibbs  estimated  that  the damage  from a 100-
year storm would be tripled for a rise of 88 cm.

     A  wide variety  of  shore protection measures  would  be   available  for
communities  to protect themselves from  increased  storm  surge and  wave damage
due to  sea  level  rise  (Sorensen,  Weisman, and  Lennon  1984).   Many  of  the
measures  used  to address  erosion and  inundation,  including  seawalls,  break-
waters,  levees,   and  beach  restoration,  also  provide  protection  against
storms.   In the case  of Galveston, which  is  already protected on the ocean
side  by  the  seawall,  Gibbs   hypothesized  that  it might  be necessary  to
completely encircle the developed areas  with a levee to prevent flooding from
the bay side; upgrading the existing seawall might also be necessary.

     Kyper and Sorensen (1985 and n.d.)  examined the implications of sea level
rise for  the design of coastal .protection  works  at  Sea Bright, New Jersey, a
coastal community  that currently  is protected by  a  seawall and has no beach.
Because the seawall is vulnerable to even a 10-year storm, the U.S. Army Corps
of Engineers and  the State of New Jersey have  been considering  a possible
upgrade.  Kyper and Sorensen estimated that the cost of upgrading  the seawall


for current conditions would be $3-5 to $6 million per kilometer of shoreline,
noting that  if  designed  properly,  the seawall would  be  useful throughout the
next century.   However,  they estimated  that  a rise in  relative  sea  level of
30-40 cm would  be  likely to result in serious damage  to the  seawall  during a
major  storm, due  to  higher  water  levels  and  the  increased wave  heights
resulting from  the erosion  of  submerged sand  in front  of  the seawall.   To
upgrade the  seawall  to withstand a 1-m  rise  in  relative sea  level would cost
$5.7 to  $9  million  per  kilometer  (50  percent  more).   They  concluded that
policy makers would  have to weigh the tradeoff  between  the  cost of designing
the  wall  to withstand projected sea  level  rise and  the cost  of subsequent
repairs and a second overhaul.

     In addition to  community-wide engineering  approaches, measures  can also
be  taken  by individual  property owners to  prevent  increased flooding.   In
1968,  the  U.S.  Congress  created   the  National  Flood  Insurance  Program  to
encourage communities to avoid risky  construction  in flood-prone areas.   In
return for  requiring new  construction  to  be elevated  above  expected  flood
levels,  the  federal  government  provides  flood  insurance,  which  is  not
available from  the  private  sector.   If sea level  rises,  flood risks will
increase.    In  response,  local  ordinances  will  automatically  require  new
construction  to  be  further  elevated,  and  insurance  rates  on  existing
properties  will rise  unless  those  properties  are  further  elevated.    As
currently organized, the National  Flood  Insurance Program would  react  to sea
level rise  as  it occurred.  Various  measures to enable  the  program  to anti-
cipate sea  level rise have  been proposed,  including warning  policy holders
that rates may  increase  in  the future if sea level rises; denying coverage to
new  construction in  areas  that are expected  to  be  lost  to erosion within the
next 30  years;  and  setting  premiums  according  to the  average risk  expected
over the  lifetime  of the  mortgage (Howard,  Pilkey,  and  Kaufman  1985;  Titus

     Kuo (Volume 4) describes case  studies  in Charleston, South Carolina, and
Fort Walton  Beach,  Florida,  which examined  the implications  of sea level rise
for  rainwater flooding  and  the  design of coastal  drainage  systems.   Waddell
and  Blaylock (n.d.)  estimated  that a 25-year  rainstorm  (with no storm surge)
would result in no damages for the Gap  Creek watershed  in Fort Walton Beach.
However, a  rise in sea   level of 30-45  cm would result  in damages of $1.1 to
$1.3 million in this community of  4,000  residents during a 25-year storm.  An
upgrade costing $550,000, however,  would prevent such damages.

     LaRoche and Webb (n.d.), who had previously developed the master drainage
plan for Charleston,  South Carolina,  evaluated  the  implications of sea level
rise for  the Grove Street  watershed  in that  community.   They estimated that
the  costs  of   upgrading the  system  for current  conditions  would  be $4.8
million, while the cost  of upgrading the system for a 30-cm rise would be $5.1
million. If  the system is designed for current conditions and sea  level rises,
the  system  would be deficient and  the city  would face retrofit costs of $2.4
million.  Thus,  for the additional $300,000  necessary to upgrade for a 30-cm
rise, the city could ensure  that it would not have to spend an  additional $2.4
million  later.   Noting   that the decision whether  to  design  now for a rise in
sea  level depends  on  the probability  that sea level will rise, they concluded
that a 3 percent real social discount rate would imply  that  designing for sea
level rise is worthwhile if  the probability of a 30-cm rise by  2025 is greater
than 30  percent.   At a  discount  rate of  10 percent, they concluded, designing
for  future conditions is not worthwhile.


Increased Salinity in Estuaries and Aquifers

     Although most  researchers and  the  general  public  have  focused  on  the
increased flooding and shoreline retreat  associated  with a rise in sea level,
the inland penetration of salt water could be important in some areas.

     As  de  Sylva (Volume  4)  describes,  a  rise  in  sea level  increases  the
salinity of  an  estuary by altering  the balance between  freshwater and salt-
water forces.   The  salinity of an  estuary represents the outcome  of (1)  the
tendency for the  ocean salt  water to completely mix with the estuarine water
and  (2)  the tendency  of freshwater  flowing into the  estuary to  dilute  the
saline water and  push it back toward the  ocean.   During droughts,  the salt
water penetrates upstream, while during the rainy  season, low salinity levels
prevail.   A  rise  in sea level has an impact similar to  decreasing the fresh-
water inflow.  By widening and deepening the estuary, sea level rise increases
the ability of salt  water to penetrate upstream.

     The implications of sea level rise for increased salinity have only been
examined in detail  for Louisiana  and the  Delaware estuary.   In Louisiana and
other river deltas,  saltwater  intrusion is  causing  the  conversion  of cypress
swamps (which cannot  tolerate  salt  water)  to openwater  lakes,  and increasing
the salinity levels of fresh and  intermediate marshes.   Accelerated sea level
rise would speed up  this process.

     The impact of current sea level trends on salinity has been considered in
the long  range  plan of  the  Delaware River  Basin Commission  since  1981 (DRBC
1981).   The drought  of  the   1960s  resulted in salinity levels that almost
contaminated the water supply of Philadelphia and surrounding areas.  Hull and
Tortoriello (1979)  found  that  the 13-cm rise projected  between 1965 and 2000
would result  in  the  "salt front"  migrating two  to four kilometers farther
upstream during a similar drought.   They  found  that  a moderately sized reser-
voir  (57 million cubic  meters) to  augment  river  flows  would be  needed  to
offset the resulting salinity  increases.

     Hull, Thatcher, and Tortoriello  (1986) examined the potential impacts of
an accelerated  rise in sea level due  to  the greenhouse  warming.   They esti-
mated that 73-cm and 250-cm rises would result  in  the salt front migrating an
additional 15 and 40 kilometers,  respectively,  during a  repeat  of the 1960s
drought.   They also found that the health-based 50-ppm sodium standard (equi-
valent to 73 ppm  chloride)  adopted by New  Jersey would  be exceeded 15 and 50
percent  of  the  time,  respectively,  and  that  the  EPA drinking water 250-ppm
chloride standard would be exceeded  over  35 percent  of the time in the latter

     Lennon,  Wisniewski,  and  Yoshioka (1986)  examined  the  implications  of
increased estuarine  salinity  for  the Potomac-Earitan-Magothy  aquifer system,
which is recharged by  the (currently  fresh) Delaware River and serves the New
Jersey suburbs  of Philadelphia.   During  the 1960s  drought,  river  water with
chloride concentrations as high as  150 ppm recharged these  aquifers.  Lennon
et al.   estimated that a  repeat of  the  1960s drought with a 73-cm rise in sea
level would  result   in  river  water  with  concentrations as high as  350  ppm
recharging the aquifer,  and  that  during the  worst month of  the drought, over
one-half of the water  recharging the aquifer system  would have concentrations
greater  than 250  ppm.   With a 250-cm rise,  98  percent of the recharge during


the worst  month  of the drought  would exceed 250  ppm,  and 75 percent  of the
recharge would be greater than 1000 ppm.

     Hull  and  Titus  (1986)  examined the  options by  which various  agencies
might respond to  increased  salinity in the Delaware estuary.  They concluded
that planned but  unscheduled reservoirs would  be more than enough to  offset
the salinity  increased from  a  one-foot (30-cm)  rise  in sea  level,  although
those reservoirs  had  originally  been intended  to  meet  increased consumption.
They noted  that  construction of the  reservoirs would  not  be  necessary until
the rise became more  imminent.   However, they  also  suggested  that, given the
uncertainties,  it  might be advisable  today  to identify additional reservoir
sites,   to  ensure  that future  generations retained  the  option of  building
additional reservoirs, if necessary.

     A rise  in  sea level could  increase salinities in other  areas,  although
the  importance  of  those impacts  has  not  been  investigated.   Kana  et al.
(1984)  and Leatherman  (1984)  made preliminary  inquiries  into  the potential
impacts  on coastal aquifers  around  Charleston  and Galveston,  respectively.
However, they concluded that  in-depth assessments were  not worthwhile because
the aquifers around Charleston are  already salt-contaminated  because of over-
pumping, and pumping of ground water has been prohibited in the Galveston area
as it causes land subsidence.  The potential  impacts  on  Florida's Everglades
and the shallow aquifers around Miami might be significant, but these have not
been investigated.

Economic Significance of Sea Level Rise

     Only  two studies  have  estimated a dollar  value of  the likely impacts of
sea level rise for particular nations.  Broadhus et al.  (Volume 4) examine the
impacts  on Egypt  and  Bangladesh.    Schneider  and  Chen  (1980)  estimated the
economic  impact  on the  United  States  of what was once  (but  is  no longer)
thought  to be  a  plausible  scenario:   rises  of 4.6 to  7.6 meters  (15  to 25
feet) occurring with little or no warning during the early part of the twenty-
first  century.    They estimated  that  these  scenarios would  result in  real
property losses of $100 to  $150  billion,  representing 6.2 to 8.4  percent of
all real property in the nation.

     The only comprehensive attempt to  place  a  dollar  value on the impacts of
sea level rise for particular communities was the study by Gibbs (1984)  of the
Charleston  and  Galveston areas,  summarized  in Volume 4.   Gibbs'  analysis,
which considers scenarios  ranging  from 0.9- to 2.4-meter  rises  through 2075,
estimates  what  the economic  impact would  be if actions are  taken  in antici-
pation of sea level rise versus the cost of responding  to sea level rise as it
occurs.  Gibbs  also modeled how investment decisions might respond to  floods
and  erosion,  and  explicitly  considered community-wide  strategies  to limit
losses,  including shore protection and abandonment.

     In  the Charleston study, Gibbs assumed  that  in  anticipation of sea level
rise,  efforts  would be  made to avoid  developing some vacant suburban areas
likely  to  be flooded  in the  future; that a  partial  abandonment  would take
place; and  that the existing  seawalls protecting  Charleston would be elevated
to provide  additional  protection.   For  a rise of 28-64 cm through  2025, Gibbs
estimated  that  the present  value of  the cumulative  impact would be $280-1065
million  (5-19 percent  of economic  activity  in the area for the period), which


could be reduced to $160-420 million  if  sea level rise was anticipated.  Most
of this  impact  would  result from a 10-100  percent  increase  in expected storm
damages, although Gibbs also estimated $7-$35 million in losses as a result of
erosion.  For the period  1980-2075, Gibbs  estimated that the economic impacts
would be $1250  - $2510 million  (17-35  percent) and could be reduced to $440
-  $1100 million through  anticipatory measures.    Gibbs  performed  a  similar
analysis of the Galveston area,  concluding  that  the impacts  of sea level rise
through  2025 would  represent $115 -  $360  million (1.1 - 3.6  percent)  if not
anticipated, and $80 - $140 million if anticipated.

     Other  studies  can be  used  to  understand  the economic  significance  of
particular classes of  impacts.   As discussed above,  a thirty-centimeter (one
foot)  rise  in  sea  level  would  erode most  recreational  beaches back  to the
first  row of houses.   The  studies cited  in our section on  erosion indicate
that the typical  beach profile  extends  out about  1000  meters,  implying that
300,000  cubic meters of sand per kilometer  of shoreline are  required to raise
the beach profile 30 cm.   If sand costs  are typically $3-$10 per cubic meter,
the  beach  rebuilding  costs  of  a 30-cm  rise in  sea level would  be $1  -  $3
million  per kilometer.  If  the  United States has a few thousand kilometers of
recreational beaches,  then it would cost billions and perhaps tens of billions
of  dollars   to  rebuild these  beaches  in   response to  a 30-cm  rise  in  sea
level.    This estimate  considers  only  the beaches themselves;  raising people's
lots to avoid  inundation would  further  increase the  costs.   The U.S.   Army
Corps of Engineers (1971) estimated that in 1971,  25,000 kilometers of shore-
line (exclusive of Alaska, the  Great Lakes, and Hawaii) were  eroding, of which
17   percent   were   "critically   eroding,"   and  would   require  engineering
solutions.   If  17  percent  of  all shorelines  require erosion  control,  that
would  imply  protection of close to 10,000  kilometers  of shoreline.   Sorensen
(1986)   describes  dozens  of engineering options  for preventing  erosion,  the
least  expensive of which  costs  $300,000 per  kilometer,  implying a cost of at
least $3 billion for protecting shorelines.

Other  Impacts of the Greenhouse Warming

     The  impacts of   sea   level  rise  on   coastal  areas,  as  well  as  their
importance,  are  likely to depend  in  part  on other  impacts  of the greenhouse
warming.  Although future sea level is uncertain, there is a general consensus
that a global  warming  would cause sea level to  rise;  by contrast, the direc-
tion of most other changes is unknown, as Manabe and Wetherald describe in the
next paper.

     One of  the more  certain impacts is that most  areas  will be warmer.  For
coastal  resorts in mid  latitudes,  the  beach season  would  be  extended  by  a
number   of  weeks.   For densely  developed  communities  like Ocean  City  with  a
three-month peak season,  such  an  extension might  increase  revenues 10 to 25
percent, far more than the  estimated  cost  of controlling erosion.  Some areas
where  the  ocean is too  cold to swim today  might  have  more  tolerable water
temperatures in  the  future.  Warmer  temperatures  in  general  might encourage
more people to visit beaches in the summer.

     Warmer  temperatures  might  change  the ability  of wetlands to  keep pace
with sea level  rise.   Mangrove  swamps,  which are  the  tropical equivalent of
salt marshes,  generally  accrete differently  than  salt  marshes.    If  warmer
temperatures enable mangroves  to grow at  higher latitudes,   the loss of wet-


lands  to  sea  level  rise  may be  altered.   On  the other  hand, marsh  peat
formation is  generally  greater  in cooler climates;  warmer  temperatures might
reduce the rate of vertical accretion for these wetlands.

     De Sylva  (Volume  4)  suggests  that  changing  climate could  alter  the
frequency and  tracks  of storms.    Because  hurricane formation  requires water
temperatures of 27°C or higher (Wendland 1977), a  global  warming might result
in an  extension of the hurricane  season and in hurricanes  forming  at higher
latitudes.  Besides increasing the amount of storm damage, increased frequency
of severe  storms would  tend to  flatten the  typical  beach profile,  causing
substantial  shoreline  retreat  unless   additional   sand  was  placed  on  the
beach.  A decreased frequency of severe  winter  storms might have the opposite
impact at higher latitudes.

     Because warmer temperatures  would  intensify  the hydrologic  cycle,  it is
generally recognized that a global warming would result in  increased rainfall
worldwide.   Thus,  rainwater flooding  might  be  increased because  of  both
decreased drainage and increased precipitation.  The impact  of  sea level rise
on saltwater   intrusion  could be  offset by  decreased drought  frequency  or
exacerbated by increased drought  frequency  (Rind and Lebedeff 1984).


     The  studies reviewed  in this  paper  appear  to  support  the  following
conclusions regarding  the causes  and  effects of sea level  rise:


•  The projected global warming would accelerate the current rate of sea level
   rise by  expanding  ocean  water, melting  alpine glaciers, and eventually,
   causing polar ice sheets to melt or slide into  the oceans.

•  Global average  sea level  has  risen  10 to  15 cm over  the  last  century.
   Ocean and  glacial  studies suggest  that the rise is  consistent  with what
   models  would  project,  given   the  0.4°C  warming  of  the  past  century.
   However,  no cause  and effect  relationship has  been  conclusively  demon-

*  Projected global warming could cause global average sea level to rise 10 to
   20  cm by  2025 and  50 to 200 cm by  2100.   Thermal expansion could cause a
   rise of  25  to  80  cm  by  2100;  Greenland and  alpine  glaciers  could each
   contribute  10  to  30  cm  through 2100.    The  contribution  of  Antarctic
   deglaciation  is  likely to  be  between  0  and  100 cm;  however,  the possi-
   bilities cannot be  ruled out that (a)  increased snowfall could increase the
   size of  the Antarctic  ice sheet,  thereby offsetting part of the sea level
   rise from  other  sources;  or  (b) meltwater and  enhanced  calving of the ice
   sheet  could  increase  the contribution  of  Antarctica to as  much  as  two

•  Disintegration of  the West Antarctic Ice Sheet  might raise  sea  level an
   additional  six meters over the next few centuries.  Glaciologists generally
   believe that such a disintegration would take at least three hundred years,
   and probably as long as five hundred years.  However, a global warming

   might  result  in  sufficient thinning  of  the  Ross  and Filcher-Ronne  ice
   shelves in the next century to make the process irreversible.

   Local  trends  in subsidence  and  emergence must  be added or  subtracted to
   estimate the rise at particular locations.
   A substantial  rise in  sea  level would  permanently inundate  wetlands and
   lowlands,  accelerate   coastal  erosion,  exacerbate  coastal  flooding,  and
   increase the salinity  of estuaries and aquifers.

   Bangladesh and Egypt appear to be among  the  nations most vulnerable to the
   rise in sea level  projected for  the  next century.   Up to 20 percent of the
   land in  Bangladesh could be flooded with a two  meter rise in  sea level.
   Although less  than  1  percent  of Egypt's  land  would be threatened,  over 20
   percent  of  the Nile  Delta,  which  contains most  of the  nation's  people,
   would be threatened.

   A large  fraction of the world's coastal  wetlands  may  be lost, threatening
   some fisheries.   A rise  in sea level of one  to  two meters by  2100 could
   destroy  50-80  percent of U.S.   coastal  wetlands.   Although  no  study has
   been  taken  to  estimate  the  worldwide   impact,  this  result  is  probably

   Erosion  caused by sea  level  rise  could  threaten recreational  beaches
   throughout the world.   Case studies have concluded that a thirty-centimeter
   rise in sea level  would result  in beaches eroding twenty to sixty meters or
   more.   Because the first row of houses  or hotels  is  often  generally less
   than twenty meters from the shore  at high tide,  if available studies are
   representative, then  recreational beaches  throughout  the  world would  be
   threatened  by  a  thirty-centimeter  rise unless  major beach  preservation
   efforts are undertaken.

   Sea level rise would  increase the costs  of flooding, flood protection, and
   flood  insurance  in coastal areas.    Flood  damages would  increase  because
   higher water levels would  provide a higher base  for storm surges;  erosion
   would increase the vulnerability to  storm waves;  and decreased natural and
   artificial drainage would increase flooding during rainstorms.

   Future sea  level rise  may already be an  appropriate factor  to consider in
   the designing coastal  drainage  and flood protection structures.

   Increased salinity from sea level rise would convert cypress swamps to open
   water and threaten drinking water supplies.

   The adverse impacts of sea level  rise could  be ameliorated through antici-
   patory land use planning and structural design changes.

   Other impacts  of global  warming might offset  or  exacerbate  the impacts of
   sea level rise.   Increased  droughts might amplify the  salinity impacts of
   sea level  rise.    Increased  hurricanes  and  increased rainfall  in  coastal
   areas could amplify flooding from sea level rise.  Warmer temperatures

   might enable mangrove swamps—which accrete differently than salt marshes—
   to advance further north,  perhaps changing wetland loss caused by sea level

   River deltas  throughout the  world would  be  vulnerable to  a rise  in  sea
   level, particularly those  whose rivers are dammed  or leveed.

   Economic studies of  Bangladesh,  Egypt,  and the United States  suggest that
   sea level rise would be economically important to  coastal  areas.

Alexander, C.E.,  M.A.  Broutman,  and  D.W.  Field.    1986.    An inventory  of
   coastal wetlands   of the  USA.    Rockville,  Md.:  National  Oceanic  and
   Atmospheric Administration,  National Ocean  Service.

Ambach, W.  1985.   Climatic  shift of  the equilibrium  line—Kuhn's  concept
   applied to the Greenland ice cap.  Annals  of Glaciology.   6:76-78.

Ambach, W. (translated by G.P.,  Weidhaas).   1980.   Increased CC^ concentration
   in the atmosphere  and climate  change:   Potential effects  on  the  Greenland
   ice sheet.  Wetter  und  Leben  32:135-142,  Vienna.   (Available at  Lawrence
   Livermore  National  Laboratory,  Report,  UCRL-TRANS-11767,  April 1982.)

Associated Press. 1985.   Doubled erosion seen  for  Ocean City.  Washington Post,
   November 14th. (Maryland Section).

Barnett,  T.P.  1984.   The  estimation of "global"  sea  level  change:   A problem
   of uniqueness. Journal of Geophysical Research.  89(C5):7980-7988.

Barth, M.C.,  and J.G. Titus (eds).   1984.    Greenhouse  effect and sea level
   rise:   A challenge  for this  generation.   New York:   Van  Nostrand  Reinhold.

Bentley,  C.R.   1983-   West Antarctic ice sheet:  Diagnosis  and  prognosis.  In
   Proceedings:  Carbon Dioxide Research Conference:  Carbon dioxide,  science,
   and consensus, Conference 820970.   Washington,  D.C.:  Department  of Energy.

Bindschadler, R.A. 1985.   Contribution of  the Greenland ice  cap to  changing
   sea level:  Present and  future. In:   M.  F.  Meier  et al.  1985.

Boesch, D.F.,  (ed).   1982.  Proceedings of the Conference  of Coastal Erosion
   and Wetland Modification in Louisiana:   Causes,  consequences,  and  options,
   FWS-OBS-82/59.  Slidell, Louisiana:   National Coastal Ecosystems  Team, U.S.
   Fish and Wildlife  Service.

Braatz, R.V.,  and D.G.  Aubrey (n.d.).   Recent  relative  sea-level  movement
   along  the  eastern  coastline  of  North America.   Journal  of  Sedimentary
   Petrology. In press.

Bruun, P.  1962.   Sea  level  rise as  a cause  of  shore  erosion.  Journal  of
   Waterways and Harbors Division.  (ASCE)  1:116-130.

Cess, R.D.,  and  S.D. Goldenberg.   1981.   The  effect of ocean  heat capacity
   upon global warming due to  increasing  atmospheric  carbon  dioxide.  Journal
   of Geophysical Research.  86:498-502.

Clark,  J.A.,  W.E.  Farrell,   and W.R.  Peltier.    1978.   Global changes  in
   postglacial sea level:  Numerical calculation. Quarternary Research. 9:265-

Clark,  J.A.,  and C.S. Lingel.   1977.   Future  sea-level changes due  to  West
   Antarctic ice sheet fluctuations.  Nature.   269(5625):206-209.

Davis,  R.A.  1985.   Coastal  sedimentary  environments. New  York:   Springer-

Dean, R.G., and E.M.  Maurmeyer.   1983.   Models  for beach profile respose.  In
   Handbook of coastal  processes and erosion.   Fort Belvoir, VA:   U.S.  Army
   Corps of Engineeers,  Coastal Engineering Research Center.

Delaware  River  Basin  Commission  (DRBC).  1981.    The  Delaware River  Basin
   comprehensive  (Level  B)  study:   Final  report  and environmental  impact
   statement.  West Trenton,  New Jersey:  Delaware River Basin Commission.

Don,  W.L.,  W.R.  Farrand,  and  M. Ewing.  1962.   Plesistocene  ice volumes and
   sea-level lowering.  Journal of Geology.  70:206-214.

Edmonson, J., and R.  Jones.  1985.   Terrebonne  Parish barrier island and marsh
   management program.  Houma,  La.:  Terrebonne Parish Council (August).

Emery, K.O., and D.G. Aubrey.   1985.   Glacial  rebound and relative sea levels
   in Europe from tide-gauge  records.   Tectonophysics.  120:239-255.

Emery,  K.O.,  and  E. Uchupi.    1972.   Western North Atlantic  Ocean  Memoir 17.
   Tulsa, Oklahoma:  American Association of Petroleum Geologists.

Everts,   C.H.  1985.   Effect  of sea level  rise  and net sand  volume  change on
   shoreline  position at Ocean  City,  Maryland.   In Potential  impact  of sea
   level rise on  the beach at Ocean City, Maryland.   Washington,  D.C.:  U.S.
   Environmental Protection Agency.

Fairbridge, R.W., and W.S.  Krebs, Jr.  1962.  Sea level and the southern oscil-
   lation.  Geophysical  Journal.  6:532-545.

Fastook,  J.L.  1985.   Ice  shelves and   ice streams:  Three   modeling  experi-
   ments.  In Melier et  al. 1985.

Flint, R.F. 1971.  Glacial  and quarternary geology.  New York:  John Wiley and

Gagliano, S.M., K.J.  Meyer-Arendt,  and  K.M. Wicker.   1981.   Land loss in the
   Mississippi deltaic plain.   In Transactions of  the  31st  Annual  Meeting of
   the  Gulf  Coast  Association  of  Geological  Societies,   293-300,   Corpus
   Christi, Texas.

Gibbs, M.  1984.  Economic analysis of sea level rise:   Methods and results.  In
   Barth and Titus 1984.

Gornitz, V.S.,  S.  Lebedeff,  and J. Hansen.  1982.   Global sea level  trend  in
   the past century.   Science.   215:1611-1614.

Hands, E.G.  1980.  Prediction  of shore retreat and  nearshore profile adjust-
   ments to  rising water levels on the  Great Lakes.  Fort  Belvior,  VA:  U.S.
   Army Corps of Engineers,  Tech. paper 80-7.

Hands, E.G. 1979.   Changes in rates of offshore retreat:   Lake Michigan.   Fort
   Belvoir, VA:   U.S.  Army Corps  of  Engineers, Coastal  Engineering  Research
   Center,  publication TP79-4.

Hands, E.G. 1976.  Observations  of  barred  coastal  profiles  under influence of
   rising water levels:  Eastern Lake  Michigan.  Fort Belvoir,  VA:  U.S.  Army
   Corps of Engineers,  CERC publication TR76-1.

Haydl, N.C.  1984.  Louisiana coastal  area, Louisiana:   Water supply. Initial
   Evaluation Study.   New Orleans:   U.S.  Army Corps of Engineers.

Hatton, R.S., R.D. DeLaune,  and W.H. Patrick.   1983.   Sedimentation,  accretion
   and  subsidence in  marshes  of  Barataria  Basin,  Louisiana.    Limnol.  and
   Oceanogr.  28:494-502.

Hays, J.D., and W.C.  Pitman  III.   1973-   Lithsopheric plate motion,  sea level
   changes, and climatic and ecological consequences.  Nature. 246:18-22.

Hicks, S.D., H.A., DeBaugh, and  L.H.  Hickman.  1983.   Sea level variations for
   the United States: 1855-1980.   Rockville, MD:  U.S. Department of Commerce,

Hoffman, J.S.,  D.  Keyes,  and J.G.  Titus.   1983.   Projecting future  sea level
   rise. Washington,  D.C.:  Government Printing Office.

Hoffman, J.S., J. Wells, and J.G. Titus.   1986.  Future global warming and sea
   level   rise.     In   Iceland   coastal  and  river   symposium.    ed.   G.
   Sigbjarnarson.  Regkjavik/Iceland:   National Energy Authority.

Hollin,  J.T.,   and R.G. Barry.    1979-    Empirical  and  theoretical  evidence
   concerning the response  of  the earth's  ice and snow  cover to  a global
   temperature  increase.  Environment International.   2:437-444.

Howard, J.D., O.H. Pilkey, and A.  Kaufman.   1985.   Strategy for beach preser-
   vation proposed.  Geotimes.  30(12):15-19.

Hughes, T.  1983.   The stability of  the West  Antarctic Ice Sheet:  What has
   happened and what will happen.   In Proceedings:   Carbon Dioxide Research
   Conference:   Carbon  dioxide,  science,  and consensus,  Conference 820970.
   Washington,  D.C.:  Department of Energy.

Hull, C.H.J., and  J.G.  Titus  (eds).  1986.   Greenhouse effect, sea level rise,
   and  salinity in  the  Delaware Estuary.   Washington,   D.C.:   Environmental
   Protection Agency and Delaware River Basin Commission.


Hull,  C.H.J.,  and  J.G.  Titus.   1986.   Responses  to salinity increases.   In:
   Hull and Titus (eds) 1986.

Hull, C.H.J.,  M.L.  Thatcher,  and R.C.  Tortoriello.  1986.   Salinity  in the
   Delaware Estuary.  In Hull and Titus (eds) 1986.

Hull, C.H.J.,  and  R.C. Tortoriello.   1979.   Sea level  trend and  salinity in
   the Delaware Estuary.   Staff report.  West  Trenton,  N.J.:   Delaware River
   Basin Commission.

Kana, T.W., J. Michel, M.O. Hayes, and J.R.  Jensen.  1984.  The physical impact
   of sea  level  rise  in  the area of  Charleston, South Carolina.  In Barth and
   Titus (eds) 1984.

Kana, T.W.,  B.J.  Baca, and  M.L.  Williams.    1986.   Potential  impacts  of sea
   level  rise on  wetlands  around  Charleston,  South  Carolina.   Washington,
   D.C.:  Environmental Protection Agency.

Kana, T.W., W. Eiser,  B. Baca,  and  M.L.  Williams.    (n.d.).  Potential impacts
   of sea  level rise  on  wetlands  around  Little Egg  Harbor,  New  Jersey.  In
   Impacts  of  sea  level  rise  on  coastal  wetlands  in  the  United  States.
   Washington, D.C.:  U.S.  Environmental Protection  Agency.

Kennett,  1982.  Marine geology. Englewood Cliffs,  Mew Jersey: Prentice-Hall.

Kuo, C.   (ed) (n.d).   Potential  impact  of sea  level rise on coastal drainage
   systems.  Washington,  D.C.:  Environmental Protection Agency.

Kyper, T.,  and R.  Sorensen.  1985.    Potential  impacts of selected  sea level
   rise scenarios on  the  beach and  coastal  works at  Sea  Bright,  New Jersey.
   In Coastal  Zone, 85, eds. O.T. Magoon et al.   New  York:   American Society
   of Civil Engineers.

Kyper, T.,   and  R.   Sorensen.  (n.d.).  Potential  impacts of sea   level  rise
   on  the   beach   and  coastal  structures   at   Sea   Bright,  New  Jersey.
   Washington, D.C.:   Environmental  Protection  Agency and U.S.  Army Corps of
   Engineers  (in press).

Lacis,  A.,  J.E.  Hansen,   P.  Lee,  T.  Mitchell,  and  S.   Lebedeff.  1981.
   Greenhouse  effect  of   trace  gases,   1970-1980.    Geophysical  Research
   Letters.  81(10):1035-1038.

Laroche, T.B., and  M.K. Webb.  (n.d.).   In  Potential impacts  of sea level  rise
   on coastal drainage systems, ed.   C. Kuo.   Washington, D.C.:   U.S. Environ-
   mental Protection Agency (in press).

Leatherman,  S.P.  1984.    Coastal geomorphic  responses  to   sea  level  rise:
   Galveston Bay, Texas.   In Barth and Titus (eds)  1984.

Lennon,  G.P., G.M.  Wisniewski,  and G.A.  Yoshioka.    1986.  Impact of increased
   river salinity on New Jersey aquifier.  In Hull and Titus  (eds)  1986.

Lingle, C.S.   1985.   A model of a polar  ice  stream  and  future sea-level rise
   due to possible drastic  of the West Antarctic Ice Sheet.   In  Meier et al.

Magness,  T.  1984.   Comment.   In Greenhouse effect and sea  level rise:   A
   challenge for  this  generation,  eds.  M.C.  Earth and J.G. Titus.   New York:
   Van-Nostrand Reinhold.

Meier, M.F.    1984.    Contribution of  small glaciers  to  global  sea level.
   Science.   226(4681):1418-1421.

Meier, M.F., et al.   1985.   Glaciers,  ice sheets,  and sea level.   Washington,
   B.C.:   National Academy Press.

Mercer, J.H. 1968.   Antarctic ice  and Sangamon  sea  level.   Geological Society
   of America Bulletin.  79:471.

Meyerson,  A.L.   1972.  Pollen and  paleosalinity  analyses  from a Holecene tidal
   marsh sequence.  Marine Geology.  12:335-357.

Milliman,  J.D., and  R.H. Meade.   1983.   World-wide  delivery of river sediment
   to the oceans.   Journal of Geology.   9KD:1-21.

Oldale, R.   1985.   Late  quarternary  sea level  history of  New  England:   A
   review of published sea level data.   Northeastern Geology.   7:192-200.

Redfield,  A.C.   1967.    Postglacial change  in sea level in the western North
   Atlantic Ocean.  Science.  157:668-692.

Redfield,  A.C.   1972.    Development of  a New England salt  marsh.   Ecological
   Monograph.  42:201-237.

Revelle,   R.   1983.    Probable  future   changes  in  sea   level  resulting  from
   increased atmospheric carbon dioxide.   In Changing Climate.   Washington,
   D.C.:   National Academy Press.

Rind,  D.,  and  S.  Lebedeff.   1984.   Potential  climate  impacts of increasing
   atmospheric Cp2 with emphasis  on water  availability  and hydrology in the
   United States.   Washington, D.C.:   Government Printing Office.

Roberts,  D., D. Davis,  K.L. Meyer-Arendt, and  K.M.  Wicker.   1983.  Recommen-
   dations  for freshwater  diversion  to  Barataria  Basin,  Louisiana.   Baton
   Rouge:    Louisiana   Department  of  Natural  Resources,  Coastal  Management

Roemmich,  D., and C.  Wunsch.  1984.  Apparent changes in the climatic state of
   the deep North Atlantic Ocean.   Nature.  307:47-450.

Roemmich,  D.   1985.  Sea level  and thermal variability  of the ocean.  In M.F.
   Meier et al. 1985.

Schneider,  S.H.,  and R.S.  Chen.   1980.   Carbon dioxide flooding:   Physical
   factors and climatic  impact. Annual Review of Energy.   5:107-140.

Sorensen,  R.M.,  R.N. Weisman,  and G.P.  Lennon.  1984.   Control  of  erosion,
   inundation, and salinity intrusion.  In Earth and Titus (eds) 1984.

Stearns, L.A., and B. Mae Creary. 1975.  The case of the vanishing brick dust.
   Mosquito News. 17:303-304.

Titus,  J.G.,  M.C.  Barth,  J.S.  Hoffman, M.  Gibbs, and M.  Kenney. 1984.   An
   overview of the  causes  and effects of sea level rise.   In  Barth and Titus
   (eds) 1984.

Titus,  J.G.   1984.    Planning  for sea  level  rise before and  after a coastal
   disaster. In Barth and Titus (eds)  1984.

Titus,  J.G.    1985.   Sea  level rise  and  the Maryland  coast.   In Potential
   impacts  of  sea   level   rise  on   the   beach  at   Ocean  City,  Maryland.
   Washington, D.C.:  Environmental Protection Agency.

Titus,  J.G.   (n.d.).    Greenhouse  effect,  sea  level  rise, and  society's
   response.    In  Sea   Surface  Studies,  ed.  R.J.   Devoy.    Cork,  Iceland:
   University College.

Titus,  J.G.,  T.  Henderson,  and J.M. Teal.  1984.   Sea level rise  and wetlands
   loss in the United States.  National Wetlands Newsletters.  6:4.

Titus,  J.G.    (n.d.).   Greenhouse effect,  sea  level  rise,  and wetland loss.
   In:   Impacts  of sea  level  rise  on coastal wetlands in  the United States.
   Washington, D.C.:  U.S. Environmental Protection Agency (in press).

Thomas,  R.H.   1985.  Responses  of  the polar ice  sheets  to climatic warming.
   In Meier et al. 1985.

Thomas,  R.H.,  T.J.O.  Sanderson, and  K.E.   Rose.  1979.   Effect  of  climatic
   warming on the West Antarctic Ice Sheet.   Nature. 227:355-358.

U.S.  Army  Corps  of  Engineers.  1982.    Louisiana coastal area,  Louisiana:
   Freshwater  diversion  to Barataria  and Breton Sound Basins.   New  Orleans:
   Corps of Engineers.

U.S.  Army  Corps  of Engineers.   1980.  Feasibility report  and final environ-
   mental impact statement:  Atlantic coast of Maryland and Assateague Island,
   Virginia.  Baltimore:  Corps of Engineers.

U.S.    Army Corps  of Engineers.   1979.    Ocean beach  study:   Feasibility
   report.  San Francisco:  Corps of Engineers.

U.S.  Army  Corps  of  Engineers.  1977.   Shore protection manual.  Fort Belvoir,
   VA:  Corps of Engineers, Coastal Engineering Research Center.

U.S.  Army  Corps  of  Engineers.   1971.   National shoreline study.   Washington,
   D.C.:  Corps of Engineers.

Van Beek,  J.L.,  D.  Roberts,  D.  Davis,  D. Sabins,  and  S.M.  Gagliano.   1982.
   Recommendations for Freshwater diversion to Louisiana estuaries east of the
   Mississippi River.   Baton Rouge: Louisiana Department of Natural Resources,
   Coastal Management  Section.

Waddell, J.O.,  and  R.A.  Blaylock  (n.d.).   Impact  of  sea  level rise  on  Gap
   Creek watershed  in  the  Fort Walton  Beach, Florida  area.  In   Kuo,  C.Y.
   (ed).,  in press.

Ward,  L.G., and D.D.  Domeraki.  1978.   The stratigraphic significance of back-
   barrier tidal channel migration.  Geol. Soc. Amer.   10(4) :201.

Weishar, L.L.,  and W.L.  Wood.  1983.   An  evaluation  of  offshore and beach
   changes on a tideless  coast.  Journal  of  Sedimentary Petrology.  53(3) :847-

Wendland,   W.M.  1977.    Tropical  storm  frequencies  related   to sea  surface
   temperatures.  Journal of Applied Meteorology.   16:480.

Reduction in Summer Soil Wetness Induced by an
Increase in Atmospheric Carbon Dioxide1
S. Manabe and R. T. Wetherald
Geophysical Fluid Dynamic Laboratory/NOAA
Princeton University
Princeton, New Jersey USA
ABSTRACT.   The geographical  distribution of  the  change  in  soil  wetness  in
response  to  an increase in atmospheric  carbon  dioxide  was  investigated  by
using  a mathematical model  of  climate.   Responding to the  increase in carbon
dioxide,  soil moisture  in the model would be  reduced in summer over extensive
regions of the middle  and  high  latitudes, such as the North American  Great
Plains,  western  Europe, northern  Canada, and Siberia.   These  results  were
obtained  from the model with predicted  cloud  cover  and  are qualitatively
similar to the results from several numerical experiments conducted earlier
with prescribed cloud cover.

    In assessments  of  the possible change in  climate caused by the increasing
CC^ in the atmosphere,  major emphasis has been placed on estimating the change
in atmospheric temperature.   However, for agricultural planning, the change  in
soil wetness may be  just as important.  In this study, which is a continuation
of earlier studies (Manabe,  Wetherald, and Stouffer  1981; Manabe and Wetherald
1985),   CC^-induced  changes  in  soil   wetness  were  investigated  using   a
mathematical model of climate  in  which cloud amount is a predicted variable.
Because of the large temporal variability of  the model hydrology, it has been
difficult  to  distinguish the  CC^-induced change  from  the  natural variability
of soil wetness.   Therefore,  the  earlier reports  discussed mainly the  zonal
mean rather than  the geographical  distribution of  soil  wetness.   To overcome
this difficulty,  it is  necessary  to  determine  soil wetness  of  the  model
averaged  over a very long period.  The present study represents an attempt  to
extract  the  geographical  distribution  of  soil   wetness by  substantially
extending the average period.
1 Reprinted from Science 232:626-27, May 2, 1986, Copyright 1986 AAAS.

     The  mathematical  model  of  climate  used   for   this   research  is  an
atmospheric general circulation model coupled  with a  static  mixed-layer ocean
model .   The model has a' global computational domain,  realistic geography, and
seasonally varying isolation.

     Precipitation is  computed whenever  supersaturation  is   indicated  by the
prognostic equation for water  vapor  (Manabe, Smagorinsky,  and Stickler 1965).
It is identified  as  snowfall when the air  temperature  near  the surface falls
below freezing;  otherwise  it  is  identified  as rain.   The  moist  convective
processes are  parameterized by a moist convective  adjustment scheme (Manabe,
Smagorinsky, and  Strickler  1965).   Cloud cover  is  predicted whenever  the
relative  humidity exceeds  a  certain  critical value,  which  is  99$  in  this
case.   The  distribution  of  cloud  cover  thus determined   is  used  for  the
computation of solar and terrestrial  radiation  (Wetherald and Manabe 1980).

     A change  in  snow  depth is computed as a  net  contribution from snowfall,
sublimation, and  snowmelt  that is determined from the  requirement  of surface
heat balance  (Manabe  1969).   The  budget  of soil  moisture is  computed by the
so-called bucket method (Manabe 1969).   For simplicity it is  assumed that soil
can hold  15  cm of liquid water.   When soil is not saturated with  water, the
change  in soil  moisture  is  predicted  as a  net  contribution of  rainfall,
evaporation, and  snowmelt.    If  the  bucket  is   full,  the  excess water  is
regarded  as runoff.    The  rate  of evaporation  from  the   soil  surface  is
determined as  a  function  of  the  water  content of  the bucket  and  potential

     The  mixed-layer  model of  the  ocean  is idealized as  a  50-m  thick,
vertically  isothermal  layer  of  water  with  predicted  sea   ice (Manabe  and
Stouffer  1980).   The effects  of  horizontal heat   transport  by ocean currents
and heat  exchange between  the mixed layer  and the deeper layer of  the ocean
are neglected.

     Two  separate experiments  were performed,  one  with  the normal  atmospheric
concentration of CO^ (300 ppm), and the other with twice the  normal value (600
ppm).   By comparing  the  results  from the  two experiments,   the   CC^-induced
change  in hydrology  could  be determined.   In each  experiment a  numerical
40-year integration  of the  model was  conducted  starting from  an  isothermal
     The thickness of  the  mixed-layer model of the  oceans  was chosen so that
     the  amplitude  of  the  observed  seasonal  variation   of  sea-surface
     temperature  was  approximately  reproduced  by   the  model  (Manabe  and
     Stouffer 1980).
     The  moisture-holding   capacity  of   soil   is   assumed  to  be  constant
     everywhere  in  view of  our  ignorance of its geographical distribution.
     According to unpublished  results from a  recent  numerical experiment, the
     simulated distribution  of soil moisture expressed  as  a  fraction  of the
     moisture-holding capacity of soil  is  not very sensitive  to the magnitude
     of the field capacity.
     Potential evaporation is  the evaporation that would occur were there an
     adequate  soil  moisture  supply  at  all  times.    According to  M.  Budyko
     (1974)  potential   evaporation   is  approximately  equal  to  the  total
     radiative energy absorbed by continental surfaces.

initial condition.  Toward  the end  of this period, the temporal variation of
the  global  mean  sea-surface  temperature  of  the model  no  longer  had  a
systematic  trend,  indicating  that  the  model  had attained an  equilibrium
climate.  To  distinguish  the CC^-induced change from  the  natural variability
of the  model hydrology,  the results were  averaged over a  sufficiently long
period, i.e., the last 10 years of the 40-year integration.  It is encouraging
that,  in  the  experiment  with  the  normal  C02  concentration,  the  model
successfully  reproduced   the  broad-scale   features   in   the   geographical
distributions of precipitative and annual mean runoff.

     The geographical distribution of the  CC^-induced  change in  soil moisture
during June to August is illustrated in Figure 1A.   In summer the soil becomes
drier  in  the mid-continental region of  North  America,  western  Europe,  and
Siberia in response to the doubling of atmospheric CC^.

     The  reduction  in soil  moisture in  North  America,  western  Europe,  and
Siberia is statistically significant at the  10$  level  (Figure 1B).   In other
words, the probability of falsely rejecting the null  hypothesis  of no change
in soil moisture is 10/5 or less in these regions.

     To  demonstrate  the   practical   implication  of   the  CC^-induced  summer
dryness  identified  above,  the change  in soil  moisture was expressed  as  a
percentage of the soil moisture  from  the normal  CC>2  experiment  (Figure 1C).
This  result  suggests  that the  CC^-induced  reduction   in soil  moisture in the
mid-continental  regions of North  America,  Siberia, and western Europe amounts
to a substantial fraction of the soil moisture present in  the  standard CC^

     In general,  the soil moisture in the  model  continents in middle and high
latitudes is  reduced  from the  peak  level in  spring  to the  minimum  level  in
summer.   In high  latitudes,  this spring-to-summer reduction in  moisture  is
caused  by  intense evaporation  in late spring,  when   the  continental surface
absorbs a large amount of solar  energy because of strong  insolation and the
disappearance of  snow  cover  with  high albedo.   In middle  latitudes a similar
mechanism also operates over the model continents.   In addition,  the reduction
in soil  moisture from spring  to  summer  in middle  latitudes  is  caused by the
termination of the rainy period in spring that results from the poleward shift
of the rain belt.

     According  to the  comparison of  the surface water  budget  between  the
normal and high  CC>2 experiments, the CC^-induced reduction in soil moisture in
summer in Siberia and northern Canada is a result of the earlier disappearance
of snow cover in  the warmer  climate.   Since  the snow cover has a high surface
albedo, its  disappearance increases  the surface absorption   of  solar energy
and  accordingly increases the  rate of  potential   evaporation  (Budyko 1974).
Thus, the  earlier  termination of  the snowmelt season results in the earlier
     Student's t  test  was used.   For this  test,  ten samples  of summer soil
     moisture were obtained from the last 10-year periods of normal and above-
     normal  CC>2  experiments  conducted  in  this  study.   Because  each sample
     represents an average  of ninety daily soil moistures,  it is likely that
     these samples are distributed approximately as Gaussian (as  inferred from
     the central  limit theorem), justifying application the of this test.

                     I     I     I    I     I     I     I    I
          30 -
          60 -
            0    30    60   90   120    150    180   ISO   120   90    60    30    0
     Figure 1.   (A)  Geographical distribution  of  the change in  soil  moisture
(centimeters)   in  response to a  doubling of  C02  for the  period  from June  to
August.  Shading  indicates regions where  the C02-induced change  is  negative.
(B)  Statistical significance of the C02-induced change  in soil moisture shown
in (A).  Areas  shaded  with lines are regions where  the  negative  soil moisture
change is statistically significant at the  10$  level.   (C)  C02-induced  change
in soil moisture expressed as a  percentage of soil mositure obtained from the
normal C02 experiment.

commencement of  the spring-to-summer reduction in  soil  moisture,  causing the
COp-induced reduction in soil moisture in summer.

     In the Great  Plains the earlier termination of  the snowmelt season also
contributes to  the summer  CC^-induced  reduction in  soil moisture.   Another
factor  responsible for  the summer  reduction  is the change  in the  rate of
precipitation  in middle  latitudes.    The,  CC^-induced  warming  in  the  lower
troposphere in  the model increases  with  increasing latitude.   Therefore, in
the high  COp  atmosphere  the warm,  moisture-rich  air penetrates  into higher
latitudes than  in  the normal CC>2 atmosphere.   Thus the  rate of precipitation
increases markedly  in the  northern  half  of the middle  latitude rain belt of
the Northern Hemisphere, whereas it decreases in the southern half of  the rain
belt.   Because of the poleward shift of the rain belt from winter to summer, a
location  in middle latitudes  which  is  situated at the  northern half of the
rain belt  during winter enters  the  southern half  in  summer.   Therefore, the
COp-induced increase  in  rate  of  precipitation  at  this location  is rapidly
reduced from  spring  to  early  summer, contributing to  the  reduction in soil
moisture during this period.

     In western  Europe  the  CC^-induced reduction  in soil moisture  occurs in
summer  in a manner qualitatively similar to that  in North America.    However,
the contribution of snowmelt is much smaller.

     The  summer  dryness in North  America  and  western Europe  is  further
enhanced  by  the  positive  feedback  process  involving  the  change   in  cloud
cover.  When  soil  moisture  is reduced, a  larger  fraction of radiative energy
absorbed by the  continental surface is ventilated  through  the upward flux of
sensible heat rather  than through evaporation.  As a result, the temperature
of  the  continental surface and  the overlying layer increases,  causing the
general  reduction  in  relative  humidity  and  precipitation   in  the  lower
troposphere of the  model.   Accompanying  the reduction in relative humidity is
a  reduction   in  total  cloud  amount, causing  an  increase  in solar energy
reaching the continental surface.  Thus,  the radiation energy absorbed by the
continental surface also  increases,  raising the  rate of potential evaporation
(Budyko  197^).     Both  the  decrease in  precipitation  and  the increase in
potential  evaporation further reduce soil moisture  during early  summer and
help to maintain it at a low level throughout the summer.

     During summer over  the Great  Plains  total cloud  cover  is  reduced and
surface air temperature increases substantially in  response to the doubling of
the CC>2 concentration  in  the  model atmosphere  (Figure  2).   Qualitatively
similar but  smaller  changes in  cloud cover  and  temperature  also  occur over
western Europe.    In  summary,  the positive feedback process  involving  cloud
cover enhances the  CC^-induced summer dryness  in  the Great Plains and western

     The seasonal  variation of the change  in zonal mean soil moisture caused
by the COg doubling is illustrated in Figure 3.  In middle and high latitudes,
zonal mean  soil moisture  over  the  continents declines substantially during
summer  months  and  is  consistent with the  geographical   distribution  of soil
moisture change discussed above.

     Figure 2.   Geographical distributions of  the  changes in (A)  total  cloud
amount  (%)  and  (B) surface air  temperature  (°C)  in response to a  doubling of
CC>2 content for June to August.  Only the distributions  in  the neighborhood of
the North American continent are illustrated.
     Figure 3.   Latitude-time distribution  of the change  in zonal mean  soil
moisture  (centimeters)  over  the  continents of  the  model  in response to  a
doubling  of CC^.    Results  are not  shown  for  latitudes  where  open ocean

     In winter,  when the middle  latitude rain  belt is displaced  toward the
equator, soil wetness increases not only  in  high latitudes but also in middle
latitudes.   However,  it is  reduced  at about 25°N,  that is,  in  the southern
half of the  rain  belt.   Qualitatively  similar  changes  of zonal  mean  soil
moisture have been obtained  previously  (Manabe,  Wetherald,  and Stouffer  1981;
Manabe  and  Wetherald 1985;  Manabe and  Stouffer  1980;  Wetherald  and  Manabe

     The  results  described  above  were  obtained  by  using  the  model  with
predicted  cloud  cover.    Similar  experiments  are  also  conducted by  using
another  model  with  fixed  cloud   cover.    A  detailed  analysis  of  these
experiments  also  indicates  the CC^-induced reduction  of  soil  moisture  in
summer, although the magnitude  of  the reduction  is  considerably  smaller  than
that suggested  here.    These results  suggest that  the  summer dryness  in the
mid-continental  regions occurs  despite  the absence  of  the cloud feedback
process.  However,  the cloud feedback process enhances the dryness.

     In the  subtropical  portion  of the continents in the  Northern  Hemisphere
(such  as  the northern  coast of  Africa,  central  Asia,  India, and Southeast
Asia) the geographical distribution of the CC^-induced change in soil moisture
varies greatly from one  experiment  to another.  Furthermore, these changes are
not  always   statistically  significant,  partly  because   of the large  natural
variability  of   the  model  hydrology.    Therefore,  one  should not take the
changes  in  these  regions  too literally,  pending further  investigation.   A
similar caution  applies to  the  geographical distribution  of  the  CC^-induced
change in the soil moisture  in the continents of the model in the tropics and
the Southern Hemisphere.

     It is  tempting  to  speculate that the Dust  Bowl  drought of the 1930s may
have been  induced  by  a warm-climate  anomaly.   As  noted  by  Vinnikov and
Groismann (1982) the seasonal and latitudinal profiles of the positive anomaly
of surface  air  temperature  during  the 1930s  resemble the  CC^-induced  warming
obtained from the present experiments.

     In  the  model  the CC^-induced  global  mean  increase   in  surface  air
temperature  with predicted   cloud  cover  is about  4°C and  is  1.5  to  2 times
larger than  the  corresponding warming with prescribed cloud cover  (Manabe and
Stouffer  1980).  This  indicates that  the cloud  feedback  process  can  enhance
C02~induced  increase in global  mean  surface air  temperature,  as  suggested by
Hansen et al. (1984).   However,  a  recent  study  by Somerville and  Remer (1984)
suggests that the  increase  in the  liquid water  content  of clouds  in response
to the  warming of  air may act as a negative  feedback between temperature and
cloud  cover  by   increasing the  planetary  albedo,   thereby reducing   the
sensitivity of climate.   This effect  is not  considered  in the model.  In view
of the primitive state  of the art  for  the parameterization of cloud formation
and other  processes,  the quantitative aspect of the present  study should be
interpreted with caution.

     Despite  these   uncertainties,   it  seems   significant  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.   Furthermore, the  analysis of  the soil moisture
budget suggests  that these large-scale changes  in soil  wetness are determined
by the latitudinal  and seasonal  profiles of the CC^-induced warming and do not


critically depend on the details of the warming.   Therefore,  it is likely that
the basic conclusion is valid despite imperfections of the model.


We thank  A.J.  Broccoli and R.J.  Stouffer,  who gave excellent advice  for the
interpretation of the results from the numerical  experiments  conducted in the
present study, and T. Delworth,  N.C.  Lau,  and H.  Levy II  for  valuable comments
on the manuscript.


Budyko, M. 1974.  Climate and Life.  New York:  Academic  Press.

Hansen, J.E., A. Lacis, D.  Rind, and G. Russell,  1984.  Climate Processes and
     Climate Sensitivity.  Washington,  D.C.: American Geophysical  Union.

Manabe, S. 1969.  Mon. Weather Rev. 97:739.

Manabe, S., and R.J. Stouffer.  1980.  J.  Geophys. Res.  85 (C10):5529.

Manabe, S., and R.T. Wetherald,  1985. Adv. Geophys. XXVIII (part A):131.

Manabe, S., J. Smagorinsky, R.F. Strickler.  1965. Mon.  Weather Rev. 93:769.

Manabe, S., R.T. Wetherald, R.J. Stouffer. 1981.  Clim.  Change. Ill 347.

Somerville, R.C.J., and L.A. Remer. 1984.  J. Geophys. Res. 89:9668.

Vinnikov, K.Ya., and  P.Ya.  Groisman.  1982.  Izv.  A.S.  USSR Atmos.  Ocean. Phys.

Wetherald, R.T., and S. Manabe.  1980. J. Atmos. Sci. 37:1485.

Wetherald, R.T., and S. Manabe.  1981. J. Geophys. Res. 86 (C2):1194.

Effects  of Climatic Changes on Agriculture  and
Forestry: An Overview

Martin L Parry
Atmospheric Impact Research Group
University of Birmingham, UK

Timothy R. Carter
International Institute for Applied Systems Analysis
Laxenburg, Austria

     The purpose of this paper  is to give an overview of the potential effects
of climatic  change on agriculture and forestry.   We  focus largely on long-term
climate  changes,  particularly  those  likely  to result from  changes in  the
composition  of the atmosphere.  We do not distinguish climate change caused by
natural factors  from that caused by  greenhouse gases; but we  are concerned
with  evaluating  the potential  effects  of  the resultant climatic  change,
whatever its causes.   We  note at  the outset that our concern here  is  with
climatic change,  not with  the  "direct"  effects  of  changes in  composition of
the atmosphere  (e.g.,  the enhancing effects  of increased  atmospheric carbon
dioxide  (CC^) on plant  growth).   A strong case can  be made  that effective
assessment of the potential effects  of increased CC>2 requires  both direct and
indirect (climatic)  effects be considered,  but  that case  will  not be further
developed here.

     In this paper, we first  consider briefly the types of  climate change  that
may  affect  agriculture  and  forestry.   This discussion   is  followed by  an
outline of the impact that climatic  changes are  known  to have had in the  past
and what we  can  learn from this experience.  We  then examine the methods  that
have been developed,  quite recently  in  most  cases,  to assess  the effects of
climatic change.  There is  then some discussion  of the  results of a few recent
assessments, particularly  of the potential effects of (X^-induced  climatic
changes on  agriculture  and forestry.  The  paper concludes  with an indication
of the present gaps in our  knowledge and our future  research priorities.


     Because different  types of  climatic  change can  have  markedly different
ecological and economic effects,  it  is important to distinguish them.  First,
we  should   distinguish   between  climatic   noise,  climatic variability,   and
climatic change  [see Hare  (1985) for  further details].  We may use  the  term

climatic noise to refer to that  part  of the variance of the climate resulting
from short-term weather changes.   It  includes,  for  example,  anomalous weather
extremes  (such  as  floods  and droughts)  that  can  have a  serious  impact  on
agriculture.  Climatic  variability represents the manner of  variation of the
climatic parameters  (e.g.,  their standard  deviation,  frequency,  etc.) within
the typical averaging period  (normally 30 years).   Climatic  change refers to
those  differences  between   successive   averaging   periods   that  cannot  be
accounted for by noise.

     Secondly, we should distinguish three types of climatic change [see Parry
(1986) for further details]:

     •  Short-term,  frequent  changes  to which social and  economic systems are
        generally adjusted

     •  Medium-term climatic  changes  to which society probably  needs to adapt
        to avoid undesirable impact

     •  Long-term climatic changes  that  occur on time scales too large to be
        considered significant for the planning horizons of most societies.

     This classification may  seem  somewhat  unusual  because  it is based not on
the nature of the climatic changes, but  on  their impact.   It may, however, be
appropriate since our task is to assess  impacts,  not to explain them.  We may
pursue   this   line   of  argument   further   by   identifying  two  additional
perspectives  on  climatic changes,  once  again  distinguished by  their impact
rather than by their nature.   The first, termed by Warrich,  Gifford, and Parry
(1986)  as  the  "slow change"  view,  emphasizes the  significance  of gradual
increases  in  mean  surface  temperatures  likely to  result  from  increases  in
atmospheric CC>2.   These  can  be expected to  lead to  gradual,  long-term,  and
cumulative  changes  in  average  regional  climates.    Following  this  view,  we
might  conclude  that so  long  as we can  predict the  long-term  climate trend,
there need be no surprises in store for the farmer or forester who should have
time to modify their practices accordingly.

     The  alternative  view   emphasizes  the  changes  in  the  frequencies  of
unusually disruptive  (or  beneficial)  events  that may result from changes in
climate.   This  view assumes  that  impact from  climatic change  comes not only
from  the average  but  also  from  the extreme  event,  a point  that  has  been
developed at  some length elsewhere (Parry  1978,  1985).   To  illustrate,  few
farmers  plan  activities on  their  expectation  of  the average  return.   They
gamble on good years and  insure  against bad ones (Edwards 1978).   In general,
we might expect subsistence  farmers  to  tune  their activities  to bad years,
attempting  to minimize the  impact of  those  years, while  commercial farmers
(such as those on the Canadian prairies) may tune them to good years  [which is
why the  prairie farmer  is periodically stressed when bad years  prevail (McKay
and Williams  1981)].

     The impact from  extremes can alter  in  a number of ways.  To  illustrate,
let  us assume  that annual  rainfall  or  temperature has a  Gaussian  (normal)
distribution  about  the  mean  x  and  variance  k  .   We  also   assume  that
agriculture is  only  seriously affected if  rainfall  or temperature is outside
the range ( x  ± k) (Figure 1a).  In the  first  type  of change,  x   changes but

                                                               Impulsive change  of
                                                               central tendency
                                                            increasing variability
                                                        Change in centra! tendency
                                                        and increasing variability
     Figure  1.   Changes  in probability distributions  and stochastic  outcomes

for shifts in mean,  variance,  or both.   After  Fukui (1979) and Hare  (1985).

remains  unchanged  (Figure   1b).    The  frequency  of  negative  extremes  (as
originally defined  by  the agricultural  system  which we assume to  be slow to
adapt to the new climate regime) is increased, and positive extremes are
reduced.   If this change occurred  for rainfall on  the  Canadian  prairies, we
might expect  the  bad  years  to far exceed the good  years.   In  the second type
of  climatic   change,  k  changes but   x   does   not (Figure  1c).    There  is
increased risk  of both deficient and  excessive  temperatures  and/or rainfall.
Both  x  and k change in the third type (Figure 1d).


     Our  intention  here is  to  present  a  very  brief  summary  of some  of the
changes  in  climate  which we  know to have  occurred  in  the  past  [for more
information,  see Wigley, Ingram, and  Farmer  (1981)]  to provide a backdrop for
a later discussion of the effect of these changes on agriculture and forestry.

Changes on a Millenial Scale

     If  we   take  as  our  starting point  the  last glacial  maximum  (around
eighteen thousand years ago),  we may note that a  rapid global  warming began
around  thirteen  thousand years  ago,   was  interrupted  around  eleven thousand
years ago by a sharp and severe cooling episode around  11,000 BP,  and was then
resumed  (Wigley,  Jones,  and Kelly  1986)  (Figure 2a).    By  about  ten thousand
years ago (the  beginning of the so-called Holocene),  global  temperatures had
roughly reached their present-day mean.

Changes on a Secular Time Scale

     During the past two thousand years,  the major  changes that have occurred
include the medieval warm epoch (800 to 1200 A.D.) and  the Little Ice Age (ca.
1500 to 1800 A.D.) (Figure 2b).   In western Europe mean annual temperatures at
the  nadir  of  the Little  Ice  Age  were  about   2°C  lower  than  in  the  early
medieval period.   These estimates  are based on  a  wide array of  proxy data,
including oxygen  isotopes,  harvest  size and  harvest  dates, and  dates  of
freezing and thawing of navigable rivers and lakes.

Decadal Changes

     The instrumental  record  enables  us to be more  precise for periods after
1850.  A long-term warming trend appears to have occurred,  interrupted by cool
episodes  around  1900  and  during  the  1960s.    Present-day  mean  surface
temperatures have been broadly  restored to their medieval  levels.   Figure 2c
shows observed temperatures and the moving average from  1850 to the present.


     The  following  section  is  a brief  summary  of  a recent synthesis  of our
present knowledge [for further detail, see Dickinson (1986)].  Our aim here is
simply to indicate the direction and approximate scale of the changes that may
occur and how these compare with changes  experienced  in the past.   These may
be summarized as follows (WHO 1986):

         _ 45
         |  40

         |  35

                900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900
          -0.4°C- •
     Figure 2.  (a) schematic representation of  temperature  changes during the
past 20,000 years  (Wigley 1982);   (b)  50-year averages of  winter  temperature
for  England.    The  continuous  line  gives unadjusted  values  based soley  on
regression analysis  and  Winter  Severity  index  values prior to  1700.   Ninety-
five percent confidence limits based on  the regression  equations and using the
t-distribution are shown.   Values shown after 1700  (temperature) are observed
instrumental  data  from  Manley's central  England  temperature series  (Manley
1974).  The  long  dashed lines project  the unadjusted  estimates  back in time,
but only  100- to 200-year  means are given  based on sparse data.    The short
dashed  line  in the  temperature  curve  gives adjusted  values, accounting for
botanical evidence.  The dotted curve is the analyst's  opinion based on a wide
variety of evidence  (Lamb  1981);   (c) Northern  Hemisphere annual mean surface
temperatures from land-based record (Wigley et al.  1986).

     •  The change in global equilibrium temperature from increases of C02 and
        other trace gases equivalent to a doubling of COp is expected to be in
        range of  1.5  to 5.5°C (Dickinson 1986).  This  change  would be of the
        same order as that which occurred between the last glacial maximum and
        the Hypsithermal  (Figure 2a).   A doubling  of C02  concentrations  is
        expected to be reached between 2050 and 2100.

     •  The global  warming up  to  the present  suggested by models  to result
        from previous increases  of  COp and other trace  gases  is  in the range
        of  0.3°  to  1.1°C  (WHO  1986).    The  observed  increases  in  mean
        temperature  during  the  last  century   "cannot  be  ascribed  in  a
        statistically rigorous  manner  to the increasing  concentration of C02
        and other greenhouse gases,  although the magnitude is within the range
        of predictions"  (WMO 1986).

     •  There are indications that warming will  be  enhanced  in high latitudes
        and that  summer  dryness  may become more frequent  over  the continents
        at middle latitudes in the Northern Hemisphere (WMO 1986).


     In this section we  review  some  of the  effects  that climatic changes have
had on agriculture  and  forestry in the past.  Our aim  is  to indicate some of
the types of effects that  have  occurred  and  some of the adjustments that have
proven successful, which provides a basis for considering potential impacts in
the future  and  for  identifying appropriate  responses  to  them.    We  take  a
number of examples  to   illustrate  the  different  kinds  of climatic  changes
outlined  in  the  section  on  types  of climatic change,  from  single  extreme
events  to  long-term   trends.     The  examples  are   arranged  in  research
chronological order, moving from recent isolated extreme events to prehistoric
long-term climatic changes.

The Effect of Isolated Extreme Events

Drought in the  U.S.  Corn Belt,  August  1983.  Midsummer 1983 saw  a pronounced
drought in the  U.S. corn belt and  the  southeastern  United States.  U.S. maize
yields fell by  about  a  third, from  over  seven  thousand kilograms per hectare
to about  five  thousand  kilograms per  hectare  (Figure 3).   In  the same year,
however,   the  Payment-in-Kind Program  (PIK) had  encouraged large  numbers of
corn (maize) farmers not to plant in 1983 (as part of an effort by the USDA to
reduce the  national grain surplus).   As a  result,  the U.S. area planted to
maize also fell by  about a third,  from  thirty  to twenty-one million hectares
(Figure 3).   The combined  effect  of decreased  yield  and reduced  area was  a
fall in U.S. maize production by almost one half (from 210 million metric tons
in 1982 to 110 million metric tons in 1983) (Figure 3).

     The  effects  were  felt  not only  nationally,  but also  globally,  because
U.S. maize accounts for  about one-eighth  of the world's total marketed cereal
production.    World  total cereals  production  in   1983  fell  by  3  percent,
harvested area  by 1.4 percent,  and  yield by 2  percent, these reductions being
almost fully accounted for by the U.S. figures alone (Figure 4).

                  USA MAIZE PRODUCTION
                 USA MAIZE AREA
                 USA MAIZE GRAIN YIELD
             HECTARE I
    Figure  3.
         USA Maize  Production, Area Harvested and Grain Yield 1972-83


                  AREA AND GRAIN YIELD 1972-1983 (FAO)
         METRIC TONSi
                1300 +
         MILLIONS OF|

         PER HECTARE


    Figure 4.
1972-83 (FAO).
World Total Cereal Production, Harvested Area, and Grain Yield

Drought in the United Kingdom, 1976.  From May 1975 to August 1976 rainfall in
the southeast  of  England was about 60  percent  of normal (Doornkamp, Gregory,
and Burn  1980).   (Mo  drier  spell of comparable length appears on the 300-year
instrumental  record.)    Almost  all crops  were  affected.   Nationally,  potato
yields were down  25 percent of  the 1970-74 average and cereal yields 10 to 15
percent.   Since  planted  area had  not  been substantially  affected,  falls in
production  were  broadly  commensurate:    Potato  production down  2.25 million
tons  and  wheat  down  2  million tons.    The shortfall  in cereal  production
required  the   import  of  around  1  million  tons  of  cereal worth 80 million
(British)  pounds.

     But prices responded accordingly.  U.K. potato prices  increased threefold
over  1975-76  and  wheat  increased  one  third over  1976-77 from  the previous
year.  As a result, aggregate net income of the agricultural industry rose in
real  terms  by 7  percent  from 1975 to  1976,  with little change  from 1976 to
1977.   Certainly  some  farmers  suffered  (such  as those engaged  in  livestock
fattening in lowland areas where higher feed prices were not fully balanced by
increased livestock prices),  but  the general  message of the  1976  drought is
that market reactions combined  with sensitive government  response  in raising
guaranteed  prices  can   effectively  buffer even  the  severest  of  climatic
events.   The  problem  really  arises when government  policy is  insensitive to
climate change,  such  as the  U.S.  Department of  Agriculture's Payment-in-Kind
program, which could not  foresee that a massive  yield downturn as a result of
drought,  in combination with reduced  planting induced by  government policy,
could  lead  to reductions  in production  that  were almost  twice as  large as
those  intended.   Fortunately, U.S.  corn yields  were back up in the benevolent
weather-year of  1984.   But  if 1984 had  also been a drought year, the outcome
could  well  have   been  a  severe  drop   in  national  and  global  cereal  food
stocks.   This leads  us  naturally to  consider  the  dramatic  effects  that  a
series of climatic anomalies  can have on agricultural production.

The Effect of  Recurrent or "Back-to-Back" Climatic Events

     Runs or  sequences  of anomalous weather-types  tend to have  a  cumulative
and  compounding  impact,  the  net  effect being greater  than the sum  of  the
single extreme-year effects.  Over 1932-37, for example, persistent drought in
the U.S.  Great Plains helped bring about  approximately  two-hundred thousand
farm  bankruptcies or  involuntary  transfers and  the migration  of  more  than
three-hundred   thousand  people from the  region.   If  the same weather were to
occur  today, assuming  1975  technology and a  1976  crop  area,  the impact would
still  be  considerable.    For a  recurrence  of  the worst weather-year,  1936,
simulated production  shows  a drop  of  25  percent,  reducing national  wheat
production by  about 15  percent  (assuming average production  elsewhere  in  the
USA)  (Figure  5).   The cumulative  effect could be  substantial:  yearly yields
simulated for  the weather over  the period  1932 to  1940  average  about 9 to 14
percent below  normal and  amount  to a  cumulative  loss over the decade equal to
about a full year's production in the Great Plains (Warrich 1984).

The Effect of  Secular  Fluctuations of Climate

     To assess  the past effects of long-term changes of climate our subject
matter becomes historical,  and  one problem confronting us is that  there  are
few  proofs in history:   connections  between climatic  changes  and  economic

              • PREDICTED YIELDS
               PLANTED ACRE)
              •PERCENT OF
               EXPECTED YIELDS
     Figure 5.   Simulated Wheat Yields  on the Great  Plains;
1975 Technology (Warrick 1984).
                              1936 Weather,

changes in the past  depend  on skilled but essentially qualitative interpreta-
tion.   However,  to  illustrate  the  potential  effects  of secular  climatic
changes, we refer to one or two episodes during which the two events have been
plausibly connected by historical scholarship.

The Little Ice  Age  in Western Europe  1500-1800.   Between about 1500 and 1800
mean annual temperatures  were about 1.5°C below  that  of the present norm and
also that of  the preceding warm  epoch of A.D.  800 to  1200.   At the nadir of
this  cold episode,  the  accumulated  temperature of  the  growing season  in
southern Scotland was 10 percent less  than at present (Parry 1978).  There was
a permanent snowfield in the  Scottish highlands.   Icebergs were reported to
have drifted  ashore  carrying  polar bears  from  the  Arctic (Lamb 1982).   There
was  widespread   desertion   of  the   more  marginal   upland  settlements  in
Scotland.    In Norway about  half of  the  farms were abandoned in   the late
Middle  Ages   (Parry  1978).    In  Iceland the  cultivation of cereals  died out
before  1600  AD.   The period was  clearly one of  considerable  difficulty for
those farming near the northern limit  of agriculture, but few historians would
deny that even  under these unusual  climatic  conditions there  existed a range
of adjustments  that  farmers  could  employ to respond and survive.   Deserting
settlements and moving somewhere  else  more attractive  were simply some of the
more radical forms of one response.

Late Medieval Cooling,  1250-1500.  On  the  issue  of adaptation the experience
of Morse  settlers in Greenland  can be instructive.   The Norse  had  gained  a
foothold along  the  Greenland coast  in around 985, but their  settlements had
been completely  abandoned at least  by 1500,  probably  between  1350  and 1450.
In the  13th century  the population, which  then  totaled about  six thousand in
two settlements,  was  subjected  to a synergistic  interaction of stresses from
hostile Eskimos  (Inuit),  the  decline of the European market for walrus ivory,
and challenging spells of cool summers and stormy winters, particularly around
1270-1300, 1320-1360,  and  1430-1460.   The failure  to adapt  to the  changing
circumstances  is believed  to explain much  of  the  Norse  decline  (McGovern
1981).  The Norse continued to  emphasize  stock  raising in the face of reduced
capacity of the  already  limited pastures.  The  option  of exploiting  the rich
seas around  them, as  the Inuit  successfully did,  was not  taken up,  perhaps
because of  the  inflexible  attitude of priest-rulers  in a  highly stratified
religious society:

     "Baldly  put,  a  society  whose administrators  (as  well as  its  peasants)
     believe  that lighting  more  candles to St.  Nicholas will have as much (or
     more) impact on the spring seal hunt as more or better boats is a society
     in serious trouble" (McGovern  1981).

According to  McGovern, the Norsemen  had lost  some  of  the adaptiveness and
resourcefulness  that  they displayed some  three  hundred years  earlier  at the
time of the settlement.   It is  an extreme example of how governments can fail
to  identify   and  implement  appropriate  policies of  response,  not  only  to
climatic  change  but  to  the  synergistic  effects of a number  of concurrent
events, and it  serves to emphasize the value of  designing  policies  that are
responsive to the  "package"  of  environmental   issues  currently facing  us
today—of  increasing  concentrations  of  greenhouse gases,  acid  deposition,
desertification,  groundwater  pollution,  salinization,  deforestation,  etc.—a
package that requires an integrated set of policies.


     One  particularly  promising  way  of  studying  the  interactions  between
climate  and  society is  to  trace  the  effects  of a  climatic  event  as  they
cascade through  physical  and social systems, and are disguised and modified by
various sets  of intervening  factors.   An  example  of this approach is offered
by Warrich  and  Bowden  (1981), tracking  the impacts  of drought  occurrence in
the U.S. Great Plains (Figure 6).   These authors traced a variety of pathways
that drought  impacts could take, spanning a variety  of spatial  scales (from
local to global) and a variety of systems (from agricultural  to social).   The
cascade of  effects  of  an  anomalous climatic  event work  their way  through a
hierarchy of  responses (e.g.,  climatic anomaly, crop yield,  farm production,
farm income,  regional agricultural sector, regional economy,  national economy,
society).   To  understand  and evaluate  each  level of  response,  we  need  to
develop a hierarchy of models that can simulate these  impacts.  In particular,
we can  identify three  sets of models:   climatic  changes, climate impacts on
potential and actual yield, and the downstream economic and  social effects of
these (Figure 7).   Scenarios using outputs from  climate models  (e.g., atmo-
spheric general  circulation models) or data from instrumental climatic records
are used as inputs  to agroclimatic models  to predict potential or actual yield
responses  to  climatic  change.    To  trace  the downstream   effects of yield
changes, outputs from the  agroclimatic models  are  used as  inputs to economic
models  (farm  simulations,  regional  input-output models,  etc.).   It  is  then
possible to consider what  policies  best  mitigate certain  impacts at specified
points  in the system.  This  research  strategy has been  adopted in a number of
recent  climate impact assessments,  the results of which are  summarized in the
next section.
                                     System affected
                         Scale  AGRICULTURAL      ECONOMIC         SOCIAL
                             /	7	'	7
                            ,          ,    GRAIN PRICES. ___/   STRESS    ,
                            /   GLOBAL   / Jf DISTRIBUTION  S ' ""  eq      ,
                      GLOBAL  /   FOOD    / f           / !   lam.rw.   /
                           /   SUPPLY   /  V FOREIGN  /  /    social     /
                                           ECONOMIES    /     conflict /
                          /     .      /     ECONOMIES    ;     conflict  /
                          /     A      I       i         i          I
                          i	-f-	/	-f	-/	r
                         /    /     /   ..FOODPRICES. -   y STRESS    /
                      i   /  US FOOD   ' S DISTRIBUTION S '"~ «9 •      /
                      'L '  SUPPLY  —-f—C    *     / '  trad' "'"""• /
                        ,   SUPPLY    /  >^NAT,ONAL  S  /   ,nflit,on    '
                       /     A     I     ECONOMY    /          /
                             i     I       1

                      !           !       I       I   STRESS   /
                       npnnpjii   » _  FrnMOMtr      '   * a       /
[ I
REGIONAL 1 "oo^cTION "f"
i f
, — — — f 	
LOCAL / 4 f
1 1 management
1 1
fc ECONOMIC ' eg
~~ PRODUCTIVITY r^ migration

/' health effects. /
Farm bankruptcy /
Gov't policy vulnerability /
     Figure  6.   The  Hypothetical  Pathways  of  Drought  Impacts  on  Society
(Warrick and Bowden  1981).

  < cs
  u 2

LU ec
          o ,

             o t- u <"
             — oo *2* -i
             oc — Qtu
I  l
I  I

             U oo
           [DNIXVW-NOISI33Q 13A31 ^IMV J




                                      (U >0
                                      4_J OO
                                      cd o
                                      S '-

                                                  S-> CO
                                                  O 4-i
                                                  CM O
                                                  n cx
                                                  rH E
                                                  Q) >— t
Recent Assessments 1:   Effects on Crop Yields

     In this section we draw heavily from a recent review by Warrich, Gifford,
and  Parry  (1986)  who have  summarized  results from a  number of studies.   We
reduce here part of their summary,  updated  to include  recent results from the
Soviet Union  (Figure  8).    In  essence,  Warrich's conclusion  is  that,  despite
the  diversity  of modeling  methods  and scenarios  adopted in  these different
studies,  there  is a  remarkable  degree  of  unanimity  regarding the expected
direction  of  effects of  climatic  changes  on crop  yields.    Warming  appears
detrimental to  cereals  in  the  core wheat-growing areas  of  North  America and
Europe.  With no change in precipitation (or radiation), slight warming (+1°C)
might  decrease  average  yields  by about  1  to  9  percent while a 2°C increase
might  reduce  average yields  by  about  3 to  17  percent.    To place  this  in
perspective Warrick notes that a 10 percent decrease in wheat and maize yields
in  North  America  is  equivalent  to about   20.5  million metric  tons, or  10
percent  of global  trade in  cereals.    Reduced  precipitation also tends  to
decrease yields, implying that both higher precipitation and temperature could
have offsetting effects  on  yields,  as reflected  in  the negative slope to the
plots  in  Figure 8.   To  these  conclusions we  add first,  that quite different
effects on the  same crop may occur  in  different regions (note the differences
for  the  same  wheat  cultivar   in  Perm,  58°N latitude,  and  Volgograd,  48°N
latitude)  and  that different  cultivars may  respond  quite  differently.    We
shall see later that these variations can be important in suggesting the kinds
of  technological adjustments that may  be  most appropriate for responding to a
given climate change.

     If, for  the  present,  we ignore the adjustment  in agriculture, such  as a
switch  of  crops,   that  is  likely to accompany or follow a  long-term  climate
change, we may  conclude  that given  the kind of changes in climate most likely
to  result  from  increased C02 and other greenhouse gases, particularly enhanced
warming  in higher  latitudes,  with summer dryness becoming  more frequent over
the  continents  at  middle  latitudes  in  the Northern  Hemisphere, then decreases
in  yields  on  the  order  of  10 percent might occur in the core wheat  production
areas of North  America  and  the  USSR.   We should emphasize, however, that such
estimates are unrealistic in that they fail  to acknowledge that adjustments in
agriculture will occur.

Recent Assessments 2:  Spatial  Effects on Crop Location

     One of the major adjustments most likely  to occur  is the  spatial shift of
cropping  areas,  which  is  somewhat akin   to  the  shift of biomes that has
occurred as a response  to  long-term climatic changes in the past (see section
on  forestry  below).   We can map  this  effect through a sequence of  steps that
has  been summarized by Parry (1985) as follows:

     •   Identification of the climatic variables  that constrain crop growth in
         the region and limit its spatial extent

     •   Identification  of  critical levels  of those  that  limit  the  spatial
         extent  of  the crop

     •   Characterization of climatic changes as changes of the critical levels

     •   Mapping these as a  shift of extent  of  the crop.









1 i i r
i -i
4 	 j
! ri £
1 c. 9.b
I 0) U. U4
I "5 *
* 1 ^
w 1
Zl x
° 0
1 1 1 I

1 1 1 1
1 0
! I
1 O
1 «Q
1 Z
1 V
i 3 S
i I s
i o 2
t i f i

l*Egj 'w---^.-o '
' « O L
1 0.3 0 g>
1 Z -2*
i D a * x >
,2 * o o
|0 Z Q. Q
1 <« Z
1 Al ...
--f 0-
1 0
o *-
*- ' o
0 1 J£
jc i a
S.i Q
r i z
i i i i
0 O O O C
«J~ CVJ CO ^











                                                                              co o
                                                                              (1) •!-(
     1VW8ON JO lN3DM3d-S3Mni8Vd3a QT3IA


                  I	I 1C colder

     Figure 9.  Estimated shift of the Canadian Spring Wheat Belt and  the  U.S.
Corn Belt  in response  to a  1  degree reduction  in  annual temperature  (after
Williams and Oakes 1978; Newman 1980).
     This approach has been  used  to estimate the effects of climate  change  on
wheat-  and  maize-growing  areas  in North  America.    In Canada  the  isopleth
bounding the wheat-maturing zone is based on photothermal time scale  equations
which consider the date  of first  fall freeze,  growing season temperature, and
radiation conditions  (Williams and Oakes  1978).  For the  U.S.  corn belt the
limits have been expressed in terms of minimum frost-free period for  maturity,
minimum  and  maximum  thermal requirements,  and  moisture requirements.   Both
crop  zones  are  shown as  shaded  areas in  Figure  9.    In this  example,   an
arbitrary decrease of  1°C  has  been used to perturb the  crop zones.   In  Canada
the area of  spring wheat  is reduced  by  about  one-third.  In the corn belt,  a
1°C cooling with  no  change  in  precipitation would shift the belt in a  south-
westerly direction by about 175  kilometers since  moisture stress  would  be
reduced at  the drier SW margin while the growing season would  be reduced  at
the cool NW margin.  A logical development of this approach  is to consider the
shift of wheat-growing  regions for  climatic  scenarios predicted  by general
circulation models for doubled concentrations of  CC^  experiments.    The  model
projects a great extension of the  winter wheat belt into Canada, a switch from
hard  to  soft  wheat in the Pacific Northwest  due  to increased precipitation,
and an expansion of areas  in fall-sown spring wheat in  the  southern  latitudes
due to  higher winter  temperatures.   In Mexico,  wheat-growing  regions  would
remain the same but greater high-temperature stress may  occur (Figure 10).



                                                                                      O   S
                                                                                 =  2   oo
                                                                                 DC  =   O
                                                                                 °-  ?   OC
                                                                                 c/i  »•
                                                                                 _  oo
                                                                            cc    >  o
                                                                       $  o  <  o








                                                                                                                •H   •

                                                                                                                o  in
                                                                                                                S-  CO
 nJ  bo
 CD -H
J2  QJ
 3  3
 C  C
 03  (1)
 O  CO
•H  O
 £-, CC
 QJ -^

Ǥ  C
  •  £-.
•a o
 O C_>
 «- o

 S-, CO
 3 i-t

Recent Assessments 3:  Potential Effects on the Regional Agricultural

     Hierarchies of linked models, similar to those discussed above, have been
used  to  assess  the potential  effects  of  climatic changes  on the  regional
economy.    In  the  International  Institute  for Applied Systems  Analysis/UNEP
project  (Parry 1986)  climatic scenarios  using outputs  from climate  models
(e.g.,  model   experiments   for  doubled  CC^  climate)   provide   inputs  to
agroclimatic  models  to  estimate  potential   or  actual  yield  responses  to
climatic change.   Output from  the  agroclimatic models provide  inputs  to the
economic  models   (e.g.,  farm simulations,   regional  input-output  models,
etc.).   While each  set of models  attempts to simulate  a limited  number  of
feedback effects within its own subsystem, the form of analysis is essentially
sequential, allowing estimates  to be made of  a series  of effects  based on an
assumed and  essentially static set  of agronomic  and economic  responses.   We
may call calculations "impact experiments"; they record the impacts that would
occur if  there were no  changes in the  farming system.   However,  as  we have
seen, farming systems have been remarkably adaptive to climatic changes in the
past and we can reasonably expect them to be in the future.  The major task we
face  is  to  make  the  adaptations  in  an optimal  manner,  that  is, with  the
minimum of social  and  economic cost.   It  is  possible  to  evaluate  the various
options  that  are  available  for  either  mitigating the  negative   effects  or
exploiting  the positive  ones  by  altering  some  of the  assumptions  in  our
agroclimatic and economic models,  for  example,  by  experimenting with a switch
to a different crop, different  amounts of fertilizer,  or  different amounts of
irrigation.  For each  set of  these "adjustment experiments" we can generate a
new  set  of  impact estimates  that  can  be  compared with the  initial  impact
estimates  (on  the  unadjusted  system),  and  thus begin  to identify  the more
appropriate types of adjustment to various types of climatic change.

     In the summary that follows we outline results selected from a wide range
of  experiments for both  impacts  and  adjustments.   Only  a few  are presented
here.  These are drawn from studies conducted in  Canada,  the Soviet Union and
Japan.   Other studies  in the  IIASA/UNEP  project were conducted  in Iceland,
Finland, Ecuador,  Brazil, Kenya,  India, and Australia  (Parry  1986).   In each
case study region  a  series  of three or four impact experiments were conducted
along the lines outlined above  and  in a broadly compatible manner.   These were
then  followed  by  a   series  of  adjustment  experiments  to  evaluate  the
appropriateness  of various  responses.    In  the  example  summarized  here  the
emphasis  was on the  potential effects  of  climatic changes due  to increased
atmospheric  CC>2.   Scenarios  of these  changes were based  on  outputs from the
doubled-COo  experiment for  the GISS GCM.   To compare  these  effects with the
effects  of  short-term climatic  variations,   the  results were  compared with
those  for an  extreme  decade  and  extreme  year  scenario.   Because different
crop-weather models  require different  data  (ten-day,  monthly,  annual, etc.),
this array of  scenarios  varied somewhat between the case studies.   The Soviet
study,  for example,  includes  synthetic  scenarios  that  vary  temperature and
precipitation  by  arbitrary  increments.    In general,  however,  the   results
permit  comparison  of  the  potential  effects  of  a  doubled CC^ climate both
between  different  regions  and between the  effects of  a  long-term  climate
change and  short-term  climatic variability.   The  results are,  of course, not
predictions.    A  high  level  of  uncertainty  is  attached  both  to  model
predictions  of  regional climatic  change  and to  the  estimations  of their
effects on agriculture.


Regional Variations in Effects on Crop Yields.  As indicated above, the higher
temperatures that  may be  expected  to prevail  under conditions  of increased
atmospheric C02  tend  to  favor  higher  yields  of cereal  crops  now grown  in
regions where temperature  limits  the  growing  season.   In the central European
region of  the  Soviet Union,  for  example,  wheat yields  under  the doubled CC^
climate projected by GISS increase by about one third (Table 1).  If we assume
that these warmer conditions  do  not  create problems of water supply and that,
for  example,  Japanese rice  production  will  remain  fully irrigated,   we  may
expect average rice yields in central Japan (To Kohu) to increase by perhaps 5
percent  (Table  2).    However,  where  cereal  production  is already  drought-
prone,  increased  rates  of   evapotranspiration  may  well  place   a brake  on
output.   Simulations  of  wheat  yield in  Saskatchewan  point to  quite severe
early-summer  moisture  stress   on   ground   spring-sown  wheat   plants  with
consequent yield  reductions  of  between  one-fifth  and  a  third,  depending  on
soil  type (Table  3).   As  we  discuss  below,  one way  of mitigating these
negative effects is to switch from spring- to winter-sown wheat.  However, the
regional pattern of crop yield responses will probably be extremely varied for
a number of reasons.

     *  Spatial  Complexity of the  Climate Change.   This is  not  simply  the
        result  of  the  varying  degrees  of  absolute  change  in  temperature,
        precipitation,  cloudiness,   etc.,   and  other  factors   that could  be
        expected to occur  in  different places, but also stems  from the ratio
        of  this  change  to   the  existing  climate.    Thus,  while  effective
        temperature  sums   (ETS)  in  central  Japan  (To   Kohu)  increase by  27
        percent in  the GISS  doubled  C02 experiment,  in  the north at Hokkaido
        (where  they are already one  quarter lower  than  in the  center)  the
        increase is 37 percent.   Thus much of the variation in yield response
        estimated for  irrigated  rice in  Japan  is a function of the geography
        of existing agroclimatic  potential.   This  factor,  as  much as  greater
        increases  in  temperature expected  in Hokkaido,  explains  why  average
        rice yields in Hokkaido  increase  by  8 percent and in  To Kohu  by 5

     *  Spatial  Complexity   Introduced  by  Non-climatic  Factors.     In  the
        estimates  presented   above  we have  assumed  the same   terrain, soil,
        management,  etc.   When  regional  variations  in these  are introduced
        there can  result  quite localized  variations  in responses to climatic
        change.  To  illustrate,  while spring  wheat  yields in Saskatchewan are
        reduced by one-fifth under the GISS projection for doubled CO^ climate
        on  brown  soils,   on  black   soils they  are   reduced  by  one-third.
        Overlaid on these  variations are  differences  in infrastructure (farm
        size,  etc.)   and   management  (levels  of  fertilizer  applications,
        pesticides, etc.).  Even if we assume the same management for all farm
        sizes,  the  effects   of  similar  yield changes  on  farm   income  vary
        according   to  farm   size partly   because   of  varying  yield-income
        functions  and partly because   different  sized  farms are found  on
        different soil types  (Table 3).



o Hin
O E-i CO X
X Cd <-3 O
£ C> -3
CM M < O
OS 2
M X t-t Cd
O Cd :s E-

X Cd
OS 0 i
CO Id <
CO 0- E->
M X <
O Cd
i— i

Cd O
< O
£ 0
J 0
0 2
Id O
2 Cd
M 9-nCM
Cd X 2
< M OS
m s E-

Cd ,-
E~* £H
< X O
£ H O

J 3 O
0 2
,• — * rn
Id O O
2 CO Cd
M 1 H
Cd in O
co as oo
< T- OS
m -^ ,-

2 •
Cd •
o •
2 3





* —



, —

, —



OS E-1
M <
< OS
2 0.
< £
Id Id
£ H







0 0
1 +









< O
E-1 Cd
H Ou

, —





, —



1 —

%*. ^**
CM a-

^r CM
oo oo

« —


%* fe*.
00 t—
+ +

Os O
•> •*
c^- t —

2 E- Q
M < J
OS Cd Cd
o- x "-1
NOI03H W«3d

t —









OS E-«
M <
•t, OS
2 0,
< £
Id Id
£ E-








t —







< U
E- Cd
0 OS
E- Q_










I— I










, —





f —
t —









' — i










Cd H
E- <
2 Cd
i— i ac
* S











, —











, —











4 —








i —


t— i


                              NOTES TO TABLE  1.

1.   Unless otherwise stated (lakimets,  Pitovranov and Sirotenko).

2.   Technology  trend,  1980-2035,  unless  otherwise  stated  (lakimets  and

3.   Up to the year 2000 (lakimets et al).

4.   Goodard Institute for Space Studies  (GISS)  General  Circulation  model  2 x
     COp experiment;  output processed for impact experiments  (Bach),  unless
     otherwise stated.

5.   GISS model climate, fertilizer,  drainage or crop varieties adjusted.

21.  May-September mean air temperature at Perm (station adjacent to  Cherdyn),
     period undetermined.

22.  May-September mean precipitation at Perm, period undetermined.

23.  Arbitrary  warming  relative  to   1951-80 baseline  (May-September  mean)

24.  GISS model climate; May-September mean,  interpolated  from grid  points to
     Cherdyn (Sirotenko).

25.  Arbitrary variations in precipitation (May-September)  (Sirotenko).

26.  GISS model climate;  calculated  as  2  x COo  -  1  x  C02  precipitation;
     interpolated  from grid point data (SirotenkoJ.

21.  Simulated yield relative to modeled 1951-80 mean yield (Sirotenko).

28.  Precipitation 29 mm below baseline (Sirotenko).

29.  Baseline precipitation (Sirotenko).

30.  Precipitation 20 mm above baseline (Sirotenko).

31.  Present crop  variety and farming efficiency (Sirotenko).

32.  Simulation incorporating  direct  effects  of C02  on plant photosynthesis
     and water use (Sirotenko).

33-  New middle-maturing variety requiring an extra  50 GDD temperature  during
     the growing season (Sirotenko).

34.  New late-maturing variety requiring 100 GDD higher temperature during  the
     growing season (Sirotenko).

35.  Annual  mean  air  temperature  at  Moscow  (assumed representative of  the
     Central Region), period undetermined.

36.  Annual mean precipitation at Moscow, period undetermined.

37.  Arbitrary  temperature   change   relative  to  baseline   (annual  mean);
     precipitation unchanged, assumed to occur in 1995 (Kiselev).

38.  As in Note 37,  this is not the GISS scenario (Kiselev).

39.  Based on 1970-80 technology trend (Kiselev).

40.  Area of crop required to minimize total expenditure (Kiselev).

41.  The product of area and crop yield (Kiselev).

42.  Expenditure per hectare and resource supply function (Kiselev).

43.  Trend yield for 1980, quintals per hectare (Kiselev).

44.  Trend yield extrapolated to 1995, relative to 1980 trend yield (Kiselev).

45.  Estimates relative  to  1980 trend yield  using  crop  production simulation
     model (Kiselev)

46.  Cropped area in 1980 (Kiselev).

47.  Cropped  area under  improved technology,  assumed  the  same  as  in  1980

48.  Optimization model  for minimizing expenditure:   values  relative to 1980

49.  Area  of perennial  grassland  for  hay   production  fixed  at   1980  value

50.  Mix of crops chosen primarily to meet human demand for vegetable products

51.  Trend production (matching yield trend)  (Kiselev).

52.  Cost based on trend data (rouble/quintal), 1980 values (Kiselev).

53.  Costs assumed the same, under improved technology, as in 1980 (Kiselev).

4->  s.,  cd
.H  cd
3      >,
CO  CO  60

    cd  o
>> E  C

cd  -p  o
c  w  a;
•M  CD 4-5
•H   ~ 73
•H  T3  C
CD  CD  Cd
S-,  4->
Q-,  Cd 4J
    l_.^  ^
    CO  CD
C  CD  60
cd  w  cd
Q.-H  c
cd  3  cd
>-)  s-  S
c  .c xi
•H  4-5 4J
    o -H
C      3
o  n
•H  W   -
4J  CD  CD
T3  O  S
 O     -H
 L,     i-H
 Q.  •  O

rH  O 4->
 cd  CD  cd  3

4J  S-      J

 o  nil
•H  -u     .5
 L,  O  T3
 C  I-H

 £0  4->
 c  cd
 O  E
•rH  -r-l
4-5  rH
 Cd  O

 Cd  [d

        W  o

       "O  CD
       r—I  CD
        CD  to
 O  -\
•H  «=C
4J  CO
 O  CD
,   ^:
<*H  4J
        o vo
        4J CO
        T3 «~

        cd   •
        CD  to
        cd *>
        C  0)
    4J CO

    CX CD   .

    SS  w

    o  o  >
    •H  4J  CD

    Cd  CD
    C  > O

    o  "-"S
    O  4J CT>
    M  Cd «-

    CM  CD
    O  t  O


I — i
Cu E-i
CO X -(

2 0
O 1
Cd .-
co in
^ CTv
CD «—

i— 1 



, —


^ —
1 1

Cd M
CJ •£.


t —


, —




1— 1
Cd Cd
o a
I-H 2









CC E-i
I— 1 *£
< CC
2 Q-
< S
Cd Cd

t" —

t —







Cd 3
> E-i
t-H «£
E-I ce
O Cd
cd a.
Ct. 2
Cti Cd
td H





i —



t —
td M
0 S
l-l I-H


t —

, —
^ —


1 —


. T
Cd Cd
l-H 2
ce I-H

, —

, —




• c
«— 4J

l T
J a.
<: 3
2 CO
M Cd
«* I-H
2 CC

s w
0 O


s; co
T- 4J

«aj f— »
2 CO
l-t Cd
H 0

                               NOTES TO TABLE 2

1.   Unless otherwise stated.

2.   1980, unless otherwise stated.

3.   1974-83, unless otherwise stated.

4.   1957-66, unless otherwise stated.

5.   1978, unless otherwise stated.

6.   Goddard Institute for Space Studies  (GISS) General  Circulation  Model 2 x
     CC>2  experiment;  output  processed for  impact experiments  (Bach).   Air
     temperature  change  only; precipitation  change not considered  important
     for irrigated rice crops  in impact experiments.

7.   District average; July-August air temperature (T.  Uchijima).

8.   District average; July-August air temperature (Horie).

9.   Calculated  as  2 x  CC^  -  1   x  CC^  temperature;  spatial average  (T.

10.  Effective  temperature  sum  (ETS)  above  10°C  base;   values  based  on
     relationship  between  July-August air temperature and ETS over  all Japan
     (150 station) (Uchijima and Seino).

11.  Present altitudinal limit (meters) (T.  Uchijima).

12.  Using  temperature  lapse  rate and  ETS  threshold  for rice cultivation of
     2600 GDD  (above  10°C  base).   Values relate  to a  baseline  temperature of
     20.4°C (T. Uchijima).

13.  Rice yield regression index  developed  for  1888-1983 detrended  yields.
     Values relative to baseline index value of 100 (T. Uchijima).

25.  Integrated   Economic  and  Climatic  Model;   Simulation   period   1966-82

26.  Approximate value relative to 1966-82 baseline (Tsujii).

27.  July-August temperature change of 3.4°C (Tsujii).

28.  Using  temperature  lapse  rate and  ETS  threshold  for rice cultivation of
     2600 GDD  (above  10°C  base).   Values relate  to a  baseline  temperature of
     22.5°C (T. Uchijima).

  W 0) 4->
 4-i SH CO
 ^ a)
  3   >.
  CO CO bD
  CU  o
  C CO 

  . *°

   CO O

   W -


   t-H  S
•—i    QJ *u

«^5| «
^^•« fc
 0) CO o
CM 4-1 CO

 QJ 0) ^

jO CO Q;
O 4-5 CT>
w cd —
CM <1J  •
O £-. O



i— I
os in
Cd E-

M Cd
f^ , ^3
CO Cd <

O -^
t— I C—
Cd 1
r>4 CT*
o ^^

>H ,-
QS ^

J 1
W *-
co in


1— 1












<*. O
O E->
M 3
> Q
0 0
as as
0 £


i —




, —


1 —





M td
> 2:
0 0
OS 0
o- z
1— 1
o <









i —





Cd E-




t —
, —









E- H
Z 0
Cd Cd



< —









t — 1
1— 1
> E-
0 Z

< —


i —


1 —






a a
j a
o *-*
^3 S
ac o-

J 0
«* z

                               NOTES TO TABLE 3

1.   Calculations for  a northern station,  Uranium City, are  footnoted  where

2.   Unless otherwise noted.

3.   Unless otherwise noted.

4.   Goddard Institute for Space Studies  (GISS) General  Circulation  Model  2 x
     C02  experiment;  output  processed for  impact experiments  (Bach).    Air
     temperature and precipitation changes included in scenario.

5.   GISS model  outputs:   air temperature changes only.   Precipitation  fixed
     at 1951-80 baseline.

6.   May-August:  all crop districts:  precipitation values rounded to nearest
     5 mm (Stewart)
     Uranium City values:  12.8°C, 149 mm.

7.   Calculated as 2 x C02 - 1 x C02 temperature;  spatial average (Stewart).

21.  Estimates from farm level simulation models (Fautley).

22.  1980 value  (Fautley).

23.  Aggregated for all crop districts based on yield simulations (Stewart).

24.  Income aggregated by farm  size,  whole  province;  1980 values;  all cereals

25.  Aggregated  for  all farms,  whole  province;  1980  values;  all  cereals

26.  Employment model (Fautley).

27.  1980 data; whole province (Fautley).

28.  Input-out model (Fautley).

Differential Crop  Responses.   Because different  crops  have different
growing  requirements,  they  frequently respond  quite  differently  to
changes  in  their environment.   In  the  central-European part  of the
Soviet  Union,  for example, a moderate warming would  increase winter
wheat yields (which are at present limited by a short, relatively cool
growing  season  in  the  region)  but  decrease barley  yields  (since
spring-sown barley is suited to the cool conditions but is susceptible
to  the  early  summer  moisture stress  that might  accompany increased
temperatures).  Under a warming experiment ( AT =  +1.5°C) winter wheat
yields  in  this  region  increased  by  30  percent  and  barley  yields
decreased  by  4  percent   (Table  1).    As  might  be  expected,  cool
temperate crops  such  as oats  and  potatoes also  decrease,  while  corn
(maize) increases.

Resultant Changes in Crops' Potential.   Two consequences flow from the
geographic  complexity  described  above:   there are spatial  shifts  of
crop  potential and spatial  shifts  of  comparative advantage.   Areas
that, under present climatic  conditions, are  judged to  be most suited
to  a  given crop,  combination  of  crops,  or  a   specified level  of
management will change location.  This means  that a change will occur
in  the  range  of  crops  that  can be profitably grown at  a particular
place, as a result of the shift of the physiological limits for growth
of different crops.  For example, we might expect that  under a warmer
climate both  winter  and spring wheat will  expand northward, assuming
that terrain and soils permit.   This is  well-illustrated in Hokkaido,
where the  limits  of the "safely  cultivable" area  for  irrigated  rice
(defined  as   a  return  period  of critically   low  growing  season
temperatures that can cause failure of the rice crop)  fluctuate quite
markedly between warm and cool periods and could be expected to expand
under a climatic warming induced by increased atmospheric CC^ or other
greenhouse gases (Figure 11).

Spatial  Shifts of  ComparativeAdvantage.   More  important are  the
changes  in  area  under different   crops  resulting from  differential
changes  in  yield that,  in turn,  result  in  changes either  in their
relative  profitability  at a  particular place  or  in the comparative
advantage that one  crop may  hold over another.   For  example,  under a
market system, it is probable that,  if other factors remain unchanged,
crops that  exhibit a substantial decrease  in  yield would be displaced
by those that show an increase in yield.   An important assumption here
is  that  alternative  areas   are available   for  production  of  this
displaced crop.   The  relative  availability  of  potential  production
areas would affect their comparative advantage.   If alternative areas
were not available,  the crop might not be displaced and there would be
no radical  change in the pattern  of land  use.  The operation of these
forces of supply  and  demand  in determining how much  change occurs in
the  pattern of  land use  can be  more  clearly   seen  in  a centrally
planned economy such as that  of  the Soviet Union.  To illustrate this
point Table 1  includes results from experiments  with  an optimization
model that  allocates land  to  various crops to minimize  expenditure on
inputs  (e.g.,  fertilizers   and machinery)   while  meeting  certain
regional  production  targets.  Under  the    +1.5°C  experiment  this
approach  increases  the  allocation  of land  to  winter  wheat by  26
percent while  reducing that under barley  by  20 percent  reflecting the



                                                         0  cO
                                                         3 -U
                                                         bO (!)
                                                         tt,   -

        fact the wheat  yields increased by 30 percent  and  barley yields fell
        by 4 percent.

     •  Inter-regional Differences in Crop  Yield  Sensitivity.   The picture is
        further complicated by the fact  that  the  same crop  grown in different
        regions will  respond  differently to  the  same change  in  climate.   In
        northern Finland  under the  GISS scenario,  for example,  barley would
        relish  a  higher  ETS  without moisture  shortage,  whereas  in southern
        Finland  yields  could  actually  decrease  because   of  early  summer
        moisture stress  (for  details see Parry  1986).  Comparable sensitivity
        analyses of spring wheat yields  in  the  northern and southern parts of
        the European USSR indicate similarly  contrasting  responses.    In Perm
        (west   of   the   Ural   Mountains)   yields   increase  with  increasing
        temperature, while  in Volgograd (some 1,000  km further  south)  they
        decrease (Table 1).   These differing regional responses will, in turn,
        affect  the pattern  of  comparative  advantage  in  cropping and  thus
        influence the shift of cropping  patterns  that may result from changes
        in climate.

     •  Effects on  Crop  Responses  to Management.   One  of the difficulties of
        estimating  effects  of  climatic  change  on  agriculture  is that  the
        sensitivity of  yield  to inputs,  such as fertilizers  and pesticides,
        also varies with climate.   Generally  speaking, the  more  ideal  the
        climatic  conditions   are  for  plant  growth,  the  greater  the  plant
        response to fertilzer applications, which means that adjusting levels
        of  fertilization can be  an  effective  means  of  stabilizing  yield
        variability resulting  from short-term climatic changes.

     •  Effects on  Variability of  Production.    Even  if we  disregard  those
        changes in  interannual variability of  temperature  and  rainfall  that
        may occur as part of  a transition  to  a  warmer climate and assume that
        climatic variability remains unchanged,  the effect of a change in mean
        climate on  mean  yield can  be different from  its  effect  on the above-
        or  below-average yields.    Figure 12,  for  example,  shows  that  in
        southern Finland  quite different  increases  occur  in the  5-percent,
        mean, and 95-percent spring wheat yields simulated for the recent warm
        period  1966-73,   and  that similar  (though  perhaps less  pronounced)
        differences are  likely  to  occur with different wheat varieties  in  a
        the doubled C02 climate projected by GISS.

Recent Assessments 4: Downstream Economic Effects

     From experiments with  farm  simulation models and  input-ouput  models,  we
can  estimate  the   effects  of  climatic changes,  via  crop  yields,  on  farm
incomes,  employment,   and levels  of  economic  activity  in  nonagricultural
sectors.  The critical parameters here are the relationships between yield and
farm  income,  farm  income  and  on-farm  employment,  and   farm  income  and
expenditures by farmers  of  nonagricultural  goods.   These  relationships  are
described by regression  equations  that  pertain to a  particular  data set (for
example, the models  for  Saskatchewan-are based upon  data for  1974 and 1980).
They allow an estimation  of the effects  of a  specific climatic event,  such as
an extreme dry year or dry decade,  if that event were to occur now.  But these
background factors are constantly changing and,  indeed, would almost certainly
change  in  response to  a longer  term transient  change  in  climate resulting
from, for example,  increases in atmospheric CO^ or other greenhouse gases.


                          a) Warm period (1966-73) scenario
                                b) GISS 2 x C02 scenario

     Figure 12.  Effect of  climate  on  mean,  lowest (95 percentile) and highest
(5 percentile) spring wheat yields  in  Finland a) for warm period 1974-82, with
present-day variety; b) for GISS  "2 x  CC>2" climate with adapted variety having
thermal requirement  120 GDD greater than present varieties.  Isolines indicate
percentage of  yield simulated  for  the  period 1959-83.  From  0.  Rantanen,  in
Parry et al. 1986.

     With these  caveats  in mind, we  report that experiments  in  Saskatchewan
indicate that under the changes  in  temperature  and  precipitation  indicated by
the GISS  projections,  total provincial farm  income decreases by  26  percent,
on-farm employment by  3  percent, and provincial gross  domestic product by 12
percent.  Considerable margins of error embrace these estimates.   For details,
see Fautley  in Parry  (1986).    In  this experiment  the change in  climate is
treated as  a sudden,  step-like  event in which  no  adjustment is  allowed for
changes in technology, management, prices,  harvested area,  etc.  In reality we
can be sure  that these would  change substantially between  now and the time at
which levels  of  atmospheric COp  are  doubled.   Current estimates  for  the COo
doubling time lie between 2050 and 2100 (Bolin et al. 1986).

     Because  of  the  uncertainties  surrounding  future  changes  in  prices,
management,   and  technology,  the results  from these  experiments present an
extremely unreal picture  of the  effect of long-term climatic changes.   They
do, however,  provide a means  of testing a number  of  adjustments  to  climate
change by comparing the results  of  the  adjusted  experiments  with  those of the
initial, unadjusted ones.   The analysis can thus take  on  an interactive form
which  more  closely  resembles  the  dynamics  of  the  real  world.     These
adjustments are discussed in the following  section.

Recent Assessments 5:  Some Technological  Responses  to Climatic Changes

     In this  section  we  refer only to adjustments  that could  be  put  in place
now,  since  this  enables  us  to  parameterize  and  input  them  to the  linked
models.   Vague assumptions about future  changes  in technology,   demand,  and
prices  are  much  less  easy  to specify.   The adjustments  are of  four  types:
crop  variety,  soil management,  land  allocation,  and purchases to supplement

Changes in  Crop  Variety.   Changes  in crop  variety  can be made to respond to
climatic  changes.    Such  problems   include  changing  from  spring to  winter
varieties,  changing  to  varieties  with   higher  thermal   requirements,  and
changing to varieties that give less varied yields.

     •  Change from Spring to Winter Varieties.   Our investigations in Canada,
        Finland,  and the northern USSR indicated that spring-sown  crops (e.g.,
        wheat, barley and  oats) experienced  reduced  yields  under the  GISS
        projections  of future  climate  because  the  increased  frequency  of
        moisture stress early  in the growing  period  (see Tables  1  and  3).  A
        switch to winter wheat or,  in some  areas, winter rye might reduce the
        effects of high evaporotranspiration in  the early  summer a well as
        taking advantage of the longer potential growing season (assuming that
        snow  cover remained  sufficient to  protect  the  crop against  winter
        kill).      In   Saskatchewan,   the   Prairie   Farm   Rehabilitation
        Administration (PRFA)  has experimented  by switching 5 percent  of the
        cropped area from spring wheat to  winter wheat to assess  the effect on
        average yield and yield  security.   The initial work has  focused only
        on  dark  brown  soils   and  for a  single extreme year (the  1961-type
        drought).  This indicates that while overall wheat  production would be
        less  than  at  present  in a normal year as a  result of this change, it
        would be greater in a  drought year.   The  next step is to  evaluate the
        effects over five or ten years of  variable weather  to see  which mix of
        varieties offers  the  longest aggregate production or, alternatively,


        the most  stable income  over a  given period.   The  PFRA could  then
        recommend various planting strategies for  various  farming goals.

     *  Change to Varieties with  Higher  Thermal  Requirements.   A  logical  way
        of exploiting longer and  warmer  growing  seasons at high  latitudes is
        to use later maturing varieties with higher thermal requirements.   In
        some cases the yields  of present-day varieties such as spring wheat in
        northern USSR  and  early-maturing rice  in Northern Japan tend to be
        reduced under the GISS  climate  projections  for doubled C02 (see  Table
        2).  However,  spring  wheat  varieties with thermal requirements 50 to
        100 GDD greater  than  present-day varieties exhibit increased yields.
        Allowing both  for  the  direct  effects  of  atmospheric  COp  on  plant
        photosynthesis and water use and the change  in varieties, spring  wheat
        yields  in  the Perm  regions are  a  third  to  a half  greater  than at
        present (Table 1).  In  Hokkaido,  late-maturing varieties  of rice  (at
        present grown  in central not northern,  Japan) gave  yields  about 30
        percent greater under  the GISS scenario  than the present early variety
        grown in Hokkaido.

     •  Change to Varieties Giving Less  Variable  Yields.  We know little  about
        what changes could occur in the  interannual  variability of temperature
        and  precipitation.    But,  even  if  we   assume the  same  degree  of
        variability under an altered climate,  the  effects  on crop yields  would
        be  significant;  it is  possible  to test  a number of  varieties  for
        stability of yield.  Figure  13  illustrates  graphically  the effects of
        the climate projected by GISS on  two different  varieties (for further
        details see T.  Horie in  Parry 1986).

Changes in Fertilizing and Drainage.  Also can be  made to  respond to  climatic

     •  Altered Fertilizer Applications.  The IIASA/UNEP  project includes  two
        types  of  fertilizer  experiments.    The  first increases  levels  of
        fertilizer  application   to   optimize  yields  under  the  climate  for
        doubled COp  projected by  GISS.  The second  varies  applications to
        maintain, for example,  1980  production levels.  Variable applications
        of  fertilizer  to   stabilize yields  by  offsetting  the  effects  of
        anomalously  cool or  warm summers  are  at  present being tested  for
        feasibility by the Icelandic government (for  details,  see lakimets et
        al. and Bergthorsoon et  al.  in Parry 1986).

     •  Improvements in  Soil  Drainage.    Increased  precipitation predicted in
        the  GISS  experiment  for  doubled  COp  might  be  expected  to  cause
        increased soil erosion,  which might  offset the beneficial effects of a
        warmer  climate  and technological  improvements.   It  should  be  noted
        that precipitation  may  well be  considerably  exaggerated  by the  GISS
        model.   Improvements  in  soil drainage  are  therefore  tested in  the
        11 ASA study of the Leningrad region  of northern USSR,  which indicated
        slightly reduced yields,  presumably a result  of the  leaching of  soil
        nutrients.   This  effect would  have to  be  weighed  against  reduced
        erosion  and more  efficient  disposal of  nitrate  pollutants in  the
        regions  to  fully  assess  the  consequence of such  a  measure   (see
        lakimets et al. in Parry 1986).

                   GISS "2 x C02" climate:
                        late-maturing variety
                           GISS "2 x C02" climate:
                                present variety
                       Observed climate:
                         present variety
                        	I	I	I	i	I	I	I
     Figure 13.   Simulated  rice yield under present  (observed)  and GISS "2 x
COp" climate  applied  to  1974-83  in  Hokkaido, N.  Japan.   Responses  to  the
doubled CC^  climate of  two varieties of  rice are shown.   From T.  Horie in
Parry et al.  1986.
Changes  in Land  Allocation.   Finally,  land  allocation  can  be changed  to
respond to the effects of climatic change.

     •  Changes of Land  Use  to Optimize Production.   Because  different crops
        respond differently  to changes  of  climate and  to various  levels  of
        fertilizer application under  those  climates,  any   attempt to maximize
        output of  each crop  while  minimizing  production  costs  is  likely  to
        identify  quite different  allocations  of land  to alternative  crops
        under  different  climates.    In  the  central  region,   European  USSR,
        experiments for  arbitrary increases  in mean  annual  temperature,  for
        example a  1°C  increase,  indicate an optimal  land  use, which  increases
        the area under winter wheat,  corn, and vegetables while decreasing the
        allocation to spring-sown barley, oats,  and  potatoes  (Table  1.)  This
        pattern of land  use  begins  to resemble  that  at  present found further
        south in the USSR and points to the value of using regional analogs to

        identify  possible  responses  to  climate changes,  a  point we  shall
        return  to  later.     Broadly  similar  experiments  in  Iceland  have
        investigated  the  increase  in   carrying capacity  of  both  improved
        grassland  and  unimproved  rangeland,  assuming  unaltered  technology
        (carrying  capacity  increases by  two thirds  or more  under the  GISS
        projections).   Given  that  inputs  are unchanged, estimates  show  that
        increases  in  carcass  weight of sheep could  be obtained at the  1980
        stocking  levels, or  alternatively the  1980 capacity levels could  be
        maintained with reduced inputs (Bergthorsoon et al.  in Parry 1986).

     •  Changes of Land Use to Stabilize  Production.   Experiments by the PFRA
        in Saskatchewan have  tested the efficacy of removing marginal cropland
        from production as  a  means of drought mitigation.  Wheat crops  on this
        land tend to be profitable in years of normal or above-normal rainfall
        but  can  cause  major  losses in  dry  years.    The  present work  has
        identified previously  unimproved land  brought into wheat  production
        over  1951-81.    This  amounts  to about 14  percent of  all  present
        cropland.   In the experiments  this  is  converted  back from wheat  to
        pasture for beef cattle,  and total provincial output from the  new mix
        of  land  uses is compared  with  that  from the present  land use  for a
        variety of drought  and nondrought years (Williams and Oakes 1986).


     A  spatial  shift of  climatic  zones  resulting from  a change  of  climate
would,  all  other  things being  equal,  lead  to  a spatial shift  of vegetation
types.  But the rate of response would depend on the rates of migration of the
species.   These vary from ten to one  hundred meters per year  for  many cool-
temperate forest  species in  Europe (Shugart et  al.  1986).   There is evidence
that  considerable  changes  in the  distribution  of forests have  occurred  as a
result  of  climatic changes  in the past.  Shugart  et  al.  cite  the transforma-
tion from spruce-dominated  to jack pine-dominated forests in eastern Minnesota
about  10,000  years ago occurring  over  at most a few hundred,  and  perhaps  in
less  than  one hundred years.  They also note the  effect  of smaller climatic
episodes  such as  that  of the  Little  Ice  Age  in  the  prairie-forest  border
region  of  Minnesota  where  an increase  in precipitation  caused  a reduction in
the  frequency of  fires, thus encouraging a  forest with more  large hardwood
trees  than  the  oak forest  it  replaced.   More recently, we may note the effect
of  even a  small fluctuation  such as  that  of  a  series of warm  summers  in
northern Scandinavia in the 1930s  inducing a regeneration of the boreal forest
at its northern limit and a measureable advance of the timber line (Shugart et
al.  1986).    If  we ignore  for the present our  uncertainties  surrounding the
rates  of  climatic  change  that  might  occur  as  a  result  of  increasing
atmospheric C02  and  also the  rates of  migration of forest species in response
to an altered climate,  we can  consider the  equilibrium locations of  forests
that  can  be predicted for  a given scenario  of climatic conditions.  Emanuel,
Shugart,  and  Stevenson (1985)   have  mapped  the  changes   in  vegetation  as
classified  by Holdridge Lifezones that would  result  from altered temperature
and  precipitation  values for the  Manabe and  Stouffer (1980) 2 x C02 climatic
changes scenario.   Values  from the  Manabe and  Stouffer experiment, which was
for  quadrupled    C02,  were   halved to  approximate  values  for  doubled  C02.
Comparison  of  the base  climate  and   doubled  C02  maps shows  a  37  percent
decrease  in the areal  extent  of the boreal  forest,  which  is  replaced at  its
southern  edge by  cool  temperate  step and to a lesser degree  by cool boreal
forest  (Figure  14).


                •Ill   TROPICAL WET FOREST
                •MB   SUBTROPICAL WET FOREST
                        SUBTROPICAL MOIST FOREST
                        TROPICAL DRY FOREST
                        WARM TEMPERATE FOREST
                        DRY WARM TEMPERATE FOREST
                        COOL TEMPERATE FOREST
                        MOIST BOREAL FOREST
                        WET BOREAL FOREST
                        TROPICAL THORN WOODLAND
                        TEMPERATE THORN STEPPE
                        COOL TEMPERATE STEPPE
                        TEMPERATE DESERT BUSH
                        TROPICAL DESERT BUSH
                        BOREAL DESERT
                    TROPICAL WET FOREST
                    SUBTROPICAL WET FOREST
                    TROPICAL MOIST FOREST
                    TROPICAL DRY FOREST
                    WARM TEMPERATURE FOREST
                                      T-  »•>
                    DRY WARM TEMPERATE FOREST V*
                    COOL TEMPERATE FOREST
                    MOIST BOREAL FOREST
                    WET BOREAL FOREST
                    TROPICAL THORN WOODLAND
                    TEMPERATE THORN STEPPE
                    COOL TEMPERATE STEPPE
                    TEMPERATE DESERT BUSH
                    TROPICAL DESERT BUSH
                    BOREAL DESERT
      Figure 14.   Holdridge  Classification of N.  America:   a)  base case b) with
biotemperature    increased   to   reflect   climate   simulated   under   elevated
atmospheric C02  concentration.   (Emanuel et  al.  1985)

     More  insight  into how  forests  might respond  to changes of  climate can
probably  be  achieved  by   running   computer  simulations  of  forest  growth,
appropriately validated against independent data, under altered climates.  The
general results of the  study have been  summarized  by Shugart et al.  (1986) as

     *  Slower growth of most deciduous tree  species throughout  much of their
        geographical range;

     •  A  dieback of  many  dominant  trees  particularly   in  the  transition
        between boreal and deciduous forest;

     •  An invasion of the southern  boreal forest by temperate deciduous trees
        that is delayed by the presence of the boreal species; and

     •  A shift in the overall pattern of forest vegetation that resembled the
        pattern obtained from  the Holdridge  experiments by  Emanuel, Shugart,
        and Stevenson (1985) and with a time lag of up to 300 years.

     Some  preliminary studies have  recently  been  undertaken  to  consider the
economic implications  of these  kinds of large-scale  spatial  shifts  of forest
types.  As  in  the  case of  assessing  economic  impacts in agriculture, the task
is one  of  using  outputs  from  a general  circulation model  for  doubled C02
experiment with a  simple growth model  and  then  using the estimates of changes
in growth  and  forest area  as inputs to an economic  model.   In the first part
of the  study,  Kauppi and  Posch  (1985)  correlated the location  of the boreal
forest  with minimum and  maximum effective  temperature  sum (ETS)  require-
ments.   They  remapped the  ETS boundaries  for  the projected   climate  and
inferred the  northward displacement  of the  boreal  forest.    From these maps
Binkley (in Parry  et al.  1986)  has  calculated,  country by  country, the change
in  productive  areas as  taiga areas  become warm  enough  to support  forest
ecosystems and also  the increase in growth on  extant forest lands (Table 4).
Using  these new data,  the IIASA global  forest sector model,  which projects
production,  consumption,   prices,  and  trade  of sixteen  forest  products  in
eighteen countries  on the basis  of  forest growth,  timber  supply, processing
facilities, and final demand was solved for a fifty-year projection horizon to
provide  estimates  of  changes  in  price,  harvest,   and   income   from  timber
sales.  In general,  timber producers in northern regions benefit by the warmed
climate, although  the  size  of  the benefit  is  small except  in Finland.  Income
from  timber sales declines  for  all producing  countries  except  Finland and
Canada, as a  result of substantial  price falls.    To illustrate,  while the
quantity of coniferous  sawlogs  harvested  in Sweden increased by 24.5 percent,
prices  fell by  14.9  percent  and  income  by 7.6  percent.   For  Finland the
figures were +63.2,  -8.6,  and +22.4 percent, respectively.   Full details are
reported in Binkley.

            Table 4.  Impact of 2 x CCU on forest growth and area.
| Growth






Increase in
forest area (%}






(3-01 )
1 .82
(1 .60)






% increase






     m-vha/yr,   based  on  the  Kauppi
     temperature data  interpolated on
     Shutz and Gabes (1971 ).
a 4°
x  5°
 (1986)   growth  model  and
latitude/longitude grid  by
 b:   (),  Estimates of current growth,  m'/ha/yr (Binkley and Dykstra 1985,  Table


     We summarize  in  Figure 15 and its accompanying  notes the direction that
future research should take.   Four  main  needs emerge.  First, we require more
specific  and   user-oriented  information  regarding   climatic   change  (its
likelihood,  nature,  magnitude, areal  extent, duration,  and  most important,
rate  of  onset).   This  information needs  to be  expressed  in  terms  readily
applicable to the  user (and often  as  derived  parameters such as date of first
and  last  frost,  days of  heat  stress,  etc.),  which would  enable agroclimatic
measures, such as  length of growing season,  to  be more readily  assessed.  It
is  also  important  that  information on  variability  be available  to  provide
estimates of possible changes in those extreme conditions that we have seen to
be important in agricultural decision-making.

     Second,  an  important   path  of  climatic  impact  on  the most  climate-
sensitive sectors  of our  society (farming,  fishing,  and  forestry)  is  an
indirect one,  i.e., they occur through changes in other physical systems (soil
chemistry,  and  ocean  currents,  agricultural pests  and  diseases, and  their
vectors).   Although we  have barely begun  to grasp  these  interactions,  they
clearly  have a major influence  on some production  systems.  The collapse of
the  Peruvian anchovy  fishery as  a  result of  the  1972 El Nino is one example.
Probably more  widespread is the  effect  of  climatic conditions  on  disease.
Outbreaks of  many plant  diseases   (such  as potato blight,  wheat  rust,  etc.)
are  triggered by  specific  weather  conditions.   Climatic  changes  that  affect
the frequency of these conditions could alter  the incidence of such outbreaks.

   c >
   O W
   +j c


. 1

T .
                             [First-order impacts)



               Impacts at enterprise level
                (e.g. firms, farms etc..
                           Downstream effects at
                           regional & national level
Figure 15.  Some Research  Priorities for Climate Impact Assessment:

   1)  More appropriate  data
   2)  Better  understanding of relationships  between climate  and  changes
       in other physical systems
   3)  Better  understanding of first-order impact
   4)  Greater precision in tracing downstream effects
5),6)  Greater precision in simulating impacts at enterprise  and regional
       levels.  Better understanding of
   7)  Perception of  effects,
   8)  Adaptive responses  at enterprise level
   9)  Strategy  formulation  at   regional,  national,   and  international
       levels  (Parry  1986).

     Furthermore, we  need to specify  with greater precision  the  interaction
between  climate and  other  resources  in  the  primary  sector,  by  modeling
empirically or  by simulation  crop-climate and  timber-climate relationships.
It should  then  be  possible to trace, with  greater  confidence, the downstream
effects of these first-other  impacts on  other  sectors of economy and society,
by reference to  a  hierarchy  of the three  types  of  models  we have  considered:
climatic, impact, and economic.

     Finally and, perhaps most important, we need to explore in greater detail
the range of technological and policy  adjustments  available in agriculture in
order  to  evaluate   their   efficiency   in  mitigating  negative  impacts  or
exploiting new options offered by changes of climate.

Bolin, B.  et al.,  eds.  1986.   The greenhouse  effect,  climatic  changes  and
     ecosystems: synthesis of present knowledge.  Chichester:  Wiley.

Dickinson, R.W.  1986.  The climate  system and modeling of future climate.   In
     The  greenhouse  effect,   climatic  changes and  ecosystems:    synthesis  of
     present knowledge,  eds. Bolin, et al. Chichester:  Wiley.

Doornkamp, J.C.,  K.J.  Gregory,  and A.S.  Burns.   1980.   Atlas  of drought  in
     Britain, 1975-76.   London:   Institute of British Geographers.

Edwards,   C.   1978.     Gambling,   insuring  and   the  production  function.
     Agricultural Economics Research. 30:25-28.

Emanuel,  W.R., H.H. Shugart,  and M.P.  Stevenson.   1985.   Climatic changes and
     the  broad-scale   distribution   of   terrestrial   ecosystem  complexes.
     Climatic Changes.   7:29-44.

Fukui, H.  1979.   Climatic variability and agriculture  in  tropical moist  re-
     gions.    In WMOProceedings of  the World Climate Conference. Geneva,  12-
     13 February 1979.   Geneva:WMO.

Hare, F.K. 1985.  Climatic variability and change. In Kates et al.

Kauppi,  P.,  and M. Posch.   1985.    Sensitivity  of boreal  forest to possible
     climatic warming.   Climatic Change. 7:450-54.

Lamb, H.H.  1981.   An approach to the  study  of the  development  of climate and
     its  role in  human affairs.   In Climate  and  history,  eds.  T.M.L. Wigley,
     M.J. Ingram, and G. Farmers.   Cambridge:  Cambridge University Press.

Lamb, H.H. 1982. Climate,  history and the mordern world. London:  Methuen.

Manabe, S. and  R.J. Stouffer 1980.    Sensitivity of a global climate model to
     an increase of C02 concentration  in  the atmosphere. J. Geophys. Res.  85:

McGovern, T.H.  1981.    The  economics of  extinction in  Norse  Greenland.   In
     Climate and  history, eds.  T.M.L.  Wigley,  M.J.  Ingrain,  and G.  Farmer.
     Cambridge:  Cambridge University Press.

McKay, G.A., and G.D.V.  Williams.  1981.  Canadian climate and food production.
     Canadian Climate  Center  Report No.  81-3,  unpublished manuscript.

Newman, J.E. 1980.  Climate change  impacts on  the growing season of the North
     American Corn Belt.  Biometeorology.   7  (part 2):128-42.
Parry,  M.L.
1978.    Climatic  change,  agriculture  and  settelment.    Dover:
Parry, M.L. 1985.  The  impact of  climatic  variations  on agricultural margins,
     In Climate  impact assessment,  eds.    R.W.,  Kates,  et  al.   Chichester:

Parry,  M.L.   1986.     Some  implications   of  climatic  change   for   human
     development.  In Sustainable development of  the biosphere,  eds. Clark, W.
     and R.W.  Munn.  Cambridge:   Cambridge  University Press.

Parry, M.L.  1985.    Potential  CC^-induced  climate  effects  on  North  American
     wheat-producing regions.  Climatic Change.  7:367-89.

Shugart, H.H., M.J. Antonovsky,  P.G.  Jarvis,  and  A.P.  Sandford.   Assessing the
     response  of  global  forests  to  climatic  change  and  direct  effect  of
     increasing  C02.     In  The   greenhouse  effect,   climatic  changes  and
     ecosystems:   synthesis   of   present   knowledge,  eds.   Bolin  et   al.
     Chichester:  Wiley.

Solomon, A.M., and H.H. Shugart.   1984.  Integrating  forest  stand simulations
     with  paleoecological  records to  examine  the long-term  forest dynamics.
     In State and  change  of  forest ecosystems. Indictors  of  current research,
     ed. G.I.  Agren.  Report  13,  Uppsala,  Sweden:  Swed.  Univ.  Agricu.  Scie.,
     Dept. of Ecology and Environmental Research.

Wigley, T.M.L.   1982.     Personal  communication,  quoted  by  Jager,  J.  (1983)
     Climate and energy systems.  Chichester:   Wiley.
Wigley, T.M.L., M.J.  Ingram,  and G. Farmer, eds.
     Cambridge:  Cambridge University Press.
                                     1981.   Climate and history.
Wigley, T.M.L.,  P.D.  Jones,  and P.J. Kelly.   1986.   Warm world scenarios and
     the detection of a CO^ induced climatic change. In The greenhouse effect.
     climatic changes and  ecosystems, synthesis  of  present knowledge,  eds. B.
     Bolin et al. Chichester:  Wiley.

WMO.  1986.   Report of  the International Conference on the  Assessment of the
     Role  of COp  and  of  Other Greenhouse  Gases  in  Climate  Variations and
     Knowledge, Associated Impacts, WMO, No. 661. Geneva:   WMO.

Warrick, R.A.   1984.   Possible  impacts  on wheat production of a recurrence of
     the 1930s drought on  the U.S.  Great Plains.  Climatic Change.  6:27-37.

Warrick, R.A., and M. Bowden.   1981,  Changing  impacts of drought in the Great
     Plains.  In The Great Plains: perspectives and prospects,  eds. M. Lawson
     and M. Baker.  Lincoln, Nebraska:  Nebraska University Press.

Warrick, R.A., R.M. Gifford with  M.L.  Parry.   1986.   C02, climatic change and
     agriculture.     In   The  greenhouse   effect,   climatic   changes   and
     ecosystems:	synthesis   of  present  knowledge,   eds.   Bolin   et  al.
     Chichester:  Wiley.

Williams,  G.D.V.,  and  W.T.  Oakes  1978.   Climatic  resources for  measuring
     barley and  wheat  in Canada.    In Essays on meteorology  and  climatology,
     eds. Hare, K.D. and E.R. Reinelt.   In  honour  of R.W. Longley.  Monograph
     3. Alberta, Canada:  University of Alberta, Studies in Geography.

The Water Resource Impact  of  Future Climate
Change and  Variability
Max Be ran
Institute for Hydrology
Wallingford, Oxfordshire, United Kingdom

     An altered availability of water  will  be  one  of  the more obvious impacts
of climatic change.   This paper describes  methods and models  that have  been
used by  hydrologists  to  try  to forecast  the effect of climatic change on the
availability  of  water for  human  consumption,  irrigation,  power  production,
effluent  dilution, and  navigation.  This assessment  requires  two  steps;  the
first concerns  impact on hydrological variables such as annual  runoff;  the
second is  required to  assess  the impact  in  human terms and  expresses  the
hydrological  variables in terms that quantify  the  exploitable fraction of the

     This paper surveys the  literature on both parts of this process.  Causal,
conceptual, and empirical models have  been  used  to  translate GCM and analogue
climate  scenarios  into hydrological terms.   Conceptual  and  empirical models
have been applied  to  the further task of modeling the consequences to water
resource  outputs,  particularly the performance of storage  reservoirs.   In  an
alternative approach  an  empirical model based upon the storage yield diagram
is used  to quantify   the  impact on reservoir reliability  of  given changes  to
the mean and variability of reservoir inflows.   To  extend  this  approach  to
accommodate climatic  rather  than hydrological change, the use  of comparative
hydrology  techniques  is advocated in  which a future climate  at a  point  is
likened to  the current climate  elsewhere.

     The  advantages and shortcomings of the  various approaches are enumerated,
and  these  give rise  to a  list  of research  suggestions and  strategies  for
future impact studies.   The  few papers  and  reports  devoted  to  the subject
clearly,  indicate  that too little attention  has been  given  to  the problem  in
the past.   Interest   has begun to  increase  since the  scientific consensus  on
climatic  change  appears  to have  centered  on  a  global  warming  induced  by
radiatively active gases.  International  organizations have an important role
in coordinating the necessary effort.



Scope of Paper

     This  paper  consists  primarily  of  a   survey   of  the  literature  on
hydrological  and water  resource  impacts  of  climatic  change.    The  paper
comments  on  the  status  of  the  methods  and  suggests  further  necessary
research.   Water resource  impacts are  clearly a  relatively unworked  area:
only twenty-one papers and reports have been uncovered that concern themselves
with future changes.

     However,  hydrologists  have  not   ignored  the climatic  change  issue  and
considerable literature exists on paleohydrology in which actual and surrogate
hydrological   records   are   used   to   detect  past   climatic  fluctuations.
Hydrologists also contribute  to  the general study of  climatic change through
their mathematical  descriptions of  hydrological process in global atmospheric
circulation models  (GCMs).  These two  topics  are  not  covered in any detail in
this survey.

     The following  subsections distinguish hydrology  from water resources and
describe hydrological  measures that may  be used  to  quantify  water resource
problems.    The  section on  hydrological  modeling  for  impact  studies  is
concerned  with the  types  of  climate-hydrology  transfer functions  that  have
been used  in  studies of impacts on catchment  response.   Readers unaccustomed
to hydrological  modeling may  wish to  limit their  reading to introductory and
closing  paragraphs,   i.e.,   postscripts.     The   section on  sensitivity  of
reservoir reliability to climate change deals with those few papers that focus
directly  on  water   resource  impacts,   in each  case  through  its  effect  on
reservoir   performance.      The   section on   research  requirements   lists
requirements revealed  by  this survey  and describes the  role of international
organizations in coordinating climate  impact research.

The Distinction Between Hydrology and Water Resources

     The  subject areas  of  meteorology  occupy  neighboring  segments of  the
familiar water cycle diagram.  Hydrology is concerned with the near-land phase
of  the  cycle  and  follows  the progress  of  moisture  from  vegetation  canopy
level, through the soil, into aquifers and watercourses, while losing sight of
it as  it enters  the sea.   Analytical  techniques  span a range of complexities
reminiscent   of   the  zero   to   three  dimensional   atmospheric   models  of
meteorology.  There  is also a branch of the science equivalent to climatology
in  which  the  hydrological  regime is   characterized  statistically  through
measures  of  the  mean and  variability of  such  elements as  flood,  drought,
aquifer state, or soil moisture deficit.

     This climatological treatment rather than the description of processes is
of most  interest to the water resource planner.  Until recently the planner
and  the hydrologist have  not been distinct but  increasingly the  two  have
diverged as water resource  science  has introduced techniques drawn from other
disciplines such as  economics  and administration.   The planner's attention is
focused  strictly   on    the   fraction  of  the  water  that  is  capable  of
exploitation—most  critically  for human  consumption,  but also for hydropower,
cooling,  effluent  disposal, irrigation, navigation, and amenity.   Thus while
all natural water is of hydrological interest, only water that is available at


the time and place required, and in suitable quantity and quality, is of water
resources interest.

     It is helpful  to  imagine the hydrological system  as  a transfer function
whose outputs are  forced by input meteorological variables.   Similarly it is
useful to  think of water  resources  as a system  in  which  hydrological inputs
are transformed to water resource outputs.  Since water resource projects tend
to be site-specific, formulating general  relationships  between the inputs and
outputs  that  encompass  all projects  is not  possible.   However,  a  limited
number of characteristic hydrological  relationships  tend  to be influential in
determining the properties and  performances  of  water  resource schemes.   By
substituting  such  relationships  for  output  from  climate-hydrology-water
resource models quantifying  the water  resources impact  of  climatic  change
becomes possible.   The following subsection  describes  the most  important of
these characteristic  hydrological relationships  and considers  three  classic
water resource problems.

Examples of Water Resource Variables

Water Storage.   A  very  common  instance  is urban water  supply from  a local
catchment.   Often an incomplete  match  between  availability and need occurs in
that, although  in  annual average terms  the attachment runoff may exceed the
demand seasons  and occasionally  entire  years  occur during  which  the  river
cannot meet  the demand.   A reservoir stores  excess water from  periods when
runoff exceeds  demand  in  order  to meet  the  town's requirements  during such
periods of deficit.   The  relationship characteristic  of  this problem is the
storage yield  diagram Figure  1 which shows the reservoir volume  needed to
supply a given continuous yield with a given level of security, here expressed
as percentage failure rate.

Flood Risk.   A second case  concerns  flood-prone  land alongside a river.   An
upstream detention reservoir can contain  excess water or  the river's carrying
capacity can  be increased.  The objective  in either case  is  to  increase the
value of the land by reducing flood incidence.  Peak discharges and hydrograph
volumes are analyzed statistically as an initial step in the design process of
devising a scheme  that maximizes  the margin  of  benefit over  construction
costs.  Figure  2 shows  the flood frequency curve which is the characteristic
relationship for this class of problem.

River Abstraction.   Many  rivers receive and convey  waste,   so  knowledge of
their diluting  capabilities  is  important.   Interest  here  focuses  on  the
proportion of  time  that  the river  flow  drops  below  some  threshold  level
representing the limiting  condition for self-purification.   The flow duration
curve Figure 3  is  the characteristic  relationship for  this problem.  The same
relationship applies to many other problems where water is abstracted directly
from rivers such  as run-of-river hydropower,  and small irrigation  and water
supply projects.

     In all three  cases  the high and  low flows,  i.e.,  hydrological extremes,
determine  the  water  resource  decision.    These examples are  deliberately
simplified—real cases mix elements from each.  They may span more than single
sources and demand points, and  decisions often  must take  account of factors
outside the  domain of  scientific  hydrology.    Nevertheless,   for  purposes of
climatic  change  impact  studies  these  examples  and  their  characteristic
relationships can substitute adequately for water resource decision variables.


           0. 1

             99.9   98       90       70     50     30       10         2  1.00.5

     Figure 1.  Storage  yield diagram.   The reservoir  storage corresponding to
a  given constant demand and  risk  of failure can  be  read  from i;he  diagram.
Alternative representations  show failure  risk as  parameter  (McMahon  and Mein

                             • 58/2  Wy* it Bclmont
                              N * 58

, ,

i i i
25 100200
i i i



     Figure 2.   Flood frequency curve.  The  graph shows the average  recurrence
interval between exceedances of the given peak discharge.  It can be drawn for
flood volume as  well  as peak.
                               .03  QR£NIG
                                          POMT Y RHUOOFA
                                                             i DAr
                               10   10 30  40 50 80 70 80  9}  98  MM


                 Flow  duration curve.   The graph shows the  proportion  of time
     Figure  3.
the flow  in the river  exceeds the  given level.
distribution function of flows.
                                                    It is  the  complement  of the

                                               INPUT Program control parameters
                                        Micrometeorological characteristics of control site
                                            Characteristics of canopy to be modeled
                                            Characteristics of soil moisture storage
                                       Standard meteorological data at control site for hour
                                           METEOROLOGICAL MODEL (Control Site)
                                Computes balance of momentum, energy, and water; profiles of wind,
                                  temperature, and humidity, and resistances to turbulent transfers
                                                 Computes Geostrophic Wind
                                Computes wind speed, temperature, and humidity at reference height
                                                 MULTILAYER CROP MODEL
                                    Computes momentum balance below-canopy profiles of wind,
                                temperature, and humidity; above canopy turbulent resistances; within-
                                canopy profiles of wind, temperature, and humidity, soil heat flux, and
                                            thence complete energy balance of crop
                                        Computes water balances of crop and soil moisture
                                    giving interception, transpiration, throughfall stemflow, etc.
                 ONE TIME STEP
              ACCORDING TO STATE
              OF CANOPY WETNESS
                                                                   HYDROLOGICAL MODEL
Updates soil moisture storage
      Figure  4.    Illustration  of  water  balance  of  small  vegetated  catchment
(Lockwood   1985)   physical  model   tends  to  concentrate  attention  on  the  crop
and/or  soil  system.


Types of Hydrological Model

     Models of  the  land phase  of  the water cycle vary  considerably in their
complexity  and  in   their   approximation   to   physical   reality.     Although
categorization boundaries may be blurred, four types of model are recognized:

    •  At the leading edge are  causal  models  which  closely replicate physical
       processes within  small  segments of the cycle such  as moisture flux at
       vegetated surfaces, below ground level, and in river channels.

    •  Models that operate at  the  catchment or aquifer  unit scale—the rele-
       vant  one for water  resource  studies—tend  to   lump processes  into
       conceptual  relations  (the  hydrological  analogue  of  parameterizing
       sub-grid processes in GCMs).

    •  The  third  level  of  process  simplification  includes  empirical  models
       that exhibit  little direct dependence  on  physical  principles.   Often
       they consist  of  statistical associations,  e.g.,  regression equations,
       or of convolution relationships between input and  output variables.

    •  The simplest  type of  model  is  the water balance.   Although apparently
       based  upon  a  physical  principle—the  conservation  of mass—dynamical
       aspects  of  the  processes  which  can  be important  at the  annual  time
       scale are neglected.

     All four types  of  model have  been used  to assess the influence on runoff
of  climatic  change  and examples  from the literature follow.    In addition,
models  of the  second  and  third  type have  been applied  to water resource
sensitivity studies  and their  application  is  described  below.   Although not
directly  related  to  the problem  of impacts,  there  is   considerable  current
interest  in   incorporating   causal   models,   especially   of  soil  moisture
processes, within the structure of  modern  GCMs (Bauer et  al.  1984; Brandyk,
Dooge,  and  O'Kane 1985;  Carson 1981;  Hansen  et  al.  1983; Hunt  1985;  Mintz
1981, 1984).

Causal Models

     It is helpful here  to consider  the hydrological  phase of the water cycle
as a system whose inputs are precipitation and evaporation and whose output is
river runoff.   The  inputs  to  this   system  therefore interface    with  state
variables of  climate models.   The physics of the processes  in  the immediate
vicinity of this interface are reasonably well understood.  However, with very
few exceptions, "downstream" processes are  not readily expressible in physics
terms; consider for  example  the complexities of  the  passage of a particle of
water through vegetation toward a  watercourse.  As  a consequence,  models that
adopt the more  physical route  describe processes on the plot,  or even leaf,
scale.   Figure 4 shows one example,  the MANTA  model  (Sellers and Lockwood
1981; Lockwood  1985),  which closely  replicates  energy  and moisture  fluxes
between ground and canopy level but devotes much less attention to the details
of moisture movement within the "Hydrological Model" box.

     Nemec  (1983)  following  Lettau and Baradas  (1973)  has termed  the output
from this type of model as "climatonic runoff," and contrasts it with measured
runoff which  is  the integration of  the output of many  differing  spatial and
temporal  units  through  a  complex  of hillslope,  subsurface,  and  channel

The Briggs  Model.   An  estimate  of the impact of COp doubling  on  runoff was
made for  the  European  Community  area by Schnell  (1984).   His  approach was to
adapt the rain and evaporation component of the Briggs biomass potential model
in which runoff is computed as the rainfall remaining after evapotranspiration
and  soil moisture replenishment.    The  British Meteorological  Office  GCM
(BMO-GCM) provided the precipitation  and temperate  change inputs.   Schnell
recognizes  the distinction  between  measured catchment  runoff and  the value
generated from the model which is labeled "runoff potential."  Figure 5, which
maps the  predicted change in average  runoff due to COp  doubling,  reveals an
overall north to south trend in runoff reduction.

Evaporation Impact.  Shuttleworth (1983) contemplates the use of causal models
to describe the  response  of evaporation to  changes  in  climate variables.  He
sounds a  warning  note  to atmospheric  modelers  who  incorporate land-surface
interactions but neglect the certainty that vegetation will change in response
to  changes   in  precipitation  and  temperature.    A  quantitative  basis  for
describing transitions  between biomes  that occupy adjacent climatic niches is
needed  and   this  limits  predictions  to  small-scale  or short-term  climate

Conceptual Models

     Conceptual models  combine features of physics-based and empirical models
of the land phase  of the  hydrological  cycle.  They are the workhorses of many
water resource  investigations  because  they  incorporate  sufficient  physics to
answer "what  if"  questions  about  change of  land use,  and they  are flexible
enough   to   link   to   mathematical   models  of  water   resource   schemes.
Meteorological time  series are  input  to  the models which yield hydrological
time  series,  usually  catchment  runoff,  as  output.    When the  hydrological
output  is  in  turn  put  into  the project  description,  then water  resource
decision  variables, for example, a cash flow, can be  generated from climatic
inputs.  Schaake and Kaczmarek (1979) suggested the use of models of this type
to investigate climate  impacts  and two examples  of  their use in hydrological
sensitivity tests are described below.

Sacramento  Model.    Figure  6  shows  the  algorithmic  arrangement  of  the
Sacramento model used by  Nemec and Schaake (1982) in their studies of climate
change  impact.   Instead  of the  detailed  treatment of moisture  and energy
fluxes of causal models,  the "boxes" of  conceptual  models contain relatively
simple "step,  branch,  and ramp"  formulations.   The  Sacramento model has been
refined  by  the U.S. National  Weather  Service to serve  nationally  as a  river
flow  forecasting  tool.   As  a  consequence it  is  structured to  emphasize
processes  that  control hydrograph  geometry.    Most  of  its  parameters are
determined from field experience and the model is calibrated to match recorded
flows  by adjusting the values of a  few dominant parameters  controlling the
runoff response.

                                                BIO 8MO
                                                  Aunol t  « mm /a
                                                         ISO BM01
Figure 5.  Generalized  change in the runoff potential  (mm)
           of western Europe following CC^ doubling.
       Figure  6.   Flow chart for Sacramento  model.

     Results  concerning  runoff sensitivity  only are  discussed here:   water
resource sensitivity  is  included  in  the  next major section.   The Sacramento
model was calibrated on the  records  of  three  rivers,  two in the United States
and an  influent  stream to Lake Victoria in  the  Upper Nile basin.   Baseline
statistics,  primarily  long-term mean runoff,  were extracted from  the model's
output  time  series (see  Table  1).    The authors then considered a  range of
altered climate scenarios  ranging  from  a 25  percent decrease  to  a 25 percent
increase in rainfall,  and from a  4  percent decrease to  a  12  percent increase
in  potential  evaporation  (supposed   equivalent  to  a  1°C decrease  to  a  3°C
increase in  surface air  temperature).    These adjustments  were  affected by
scaling the input  series,  for example,  by multiplying each  rainfall value by
    Table 1.  Percentage Change in Runoff from Base Due to Climatic Forcing
Runoff Ratio for Scenarios
+25, +12
+ 154$
+ 49T«
+ 170$
-25. -4
    Taken  from  Figures 4 and  5 of  Klemes and  Nemec (1985)  or  inter-
    polated from the original paper.   First figures of scenario headings
    relate  to  the  precipitation  change  from  base,  the  second  to
    evapotranspiration change.
     The authors conclude that the changes in runoff fall outside the range of
modeling  error  of  the  rainfall runoff  process.   The  experiment  is  said to
reveal the  dominance  of changes to precipitation and  the  extreme sensitivity
of arid area catchments.

     Klemes and Nemec (1985) subjected Nemec and Schaake's results to detailed
statistical scrutiny.   The principal  concern  appeared to be  the statistical
significance of  the change,  that is,  whether  or  not the forecast runoffs lie
outside  the range  of  fluctuations  that  occur naturally  under  the  current
climate.   By  random sampling from a  log-normal parent it  was determined that
changes in  runoff  due to precipitation  changes of  less than about 15 percent
to 20 percent could not be  distinguished from random fluctuations in the base
record.  This result  is of course dependent on  the record length used.

SHOLSIM  Model.   Aston  (1984)  considered  the  effect  of C02 doubling  on the
hydrology  of  the  study area  in New  South  Wales,  Australia.    Although GCM
predictions  indicated   no   change   in  precipitation  for  the  area,  Aston
anticipated an  important  impact due  to the direct effect of C02 enrichment on
vegetation—an  increase in  stomatal  resistance with  consequent reduction in
transportation.  Evidence presented  in  the paper  shows that in many plants  a

doubling of  COo concentration  can lead to  more  than a  two-fold  increase in
resistance.  The scenario studied is for a doubling in resistance.  An initial
estimate based upon  a  combination  equation  for evaporation  pointed  to an
increase in (climatonic) runoff of 85 percent.

     Aston then pursued this finding through a conceptual model, SHOLSIM, of a
417  km  catchment.    This  model  uses  a  much-simplified description of the
evaporation process  in which the overall  effect  of  the  resistance change was
represented  by doubling  the gradient  of  the ratio  of  actual  to  potential
evapotranspiration (AE to PE) with soil moisture content, and also by reducing
the limiting rate at which transportation can occur.  The result of the second
influence  amounted  to a  nearly 60 percent  increase  in   runoff  in  an average
year;  the  two influences  together  cause  the  modeled   catchment  runoff to
increase by 90 percent, very much in accord with the causal model result.

Postscript on Conceptual Models.  At first sight, the two studies are not easy
to compare as  the first  is  strictly an exercise  in  model  sensitivity,  while
the  second  traces  the  effect  of  a   specific   climatic  scenario.    Aston
considered the changes due to the direct effect of CC^ enrichment to equate to
an approximate halving of PE, well outside the nominal range considered in the
Sacramento model  simulations.    However,  Klemes  and Nemec  (1985)   in  their
analysis of  Nemec  and  Schaake's  results  revealed  very strong  linearities
within the simulated  series.  Extrapolating (perhaps  unreasonably)  from these
one finds  a  1  percent increase in runoff  for  a  1  percent decrease  in ETP for
the Leaf river model, and by about twice this amount for  the Pease and Nzoia
river cases.    This suggests  that  SHOLSIM and Sacramento  model predictions are
in reasonable accord.

     It  is important  to remember  that  these results relate  to model,  not
prototype,   sensitivities.    A  good  fit  to  the  output series  within  the
calibration period is  no guarantee of a  continued close  match  beyond it, nor
of the correctness of the physical principles embodied in the model.   Likewise
statistical significance of the difference between the runoff records from two
climate scenarios do not prove the truth of either.

     However,  the  writer is unclear  on  the relevance of Klemes'  and Nemec's
(1985)  requirement   for  establishing   significance.     When  searching  for
discontinuities  in  past records it  is  clearly  essential  for  statistical
significance to  be  established.    But  in  a prognostic  study,  it  is  beyond
contention that  changes  in  input  create  changes  in  output and  changes, no
matter how small,  must be regard as  real  effects.  In this  context,  lack of
significance provides  an interesting sidelight on  the limitations  imposed on
an imaginary  analyst  searching  for evidence  of change  in  the past  data by
short records  but  serves no  useful  function in prognosis studies.   Here the
contentious issue is whether the model continues to replicate runoff correctly
through a climatic change.

     Neither Nemec's  nor Aston's studies  considered  the  compatibility of the
precipitation  with  the hypothesized  evaporation  pattern, nor  was  it allowed
that under the altered climate the rainfall may occur in  different seasonal or
storm patterns.  Catchment response  to  intense periods of rainfall tends to a
proportional rather  than  the  subtractive response  pattern  implicit  in the
"bucket" model approach (see the section on water balance models).

     Aston's study is more  immediately  helpful  to planners in that it follows
the logic of the  specific climate  change as far as current knowledge permits,
although  by  substituting  "resistance  doubling"  for  "C0?  doubling"  some
important issues  are sidestepped.   Contrary,  or at least moderating,  views on
the direct  CC^ effect  have  appeared  in  the  literature  (e.g.,  Kramer 1981,
quoted in Woodwell 1985) with the following implications and caveats:

     •  Plant species vary greatly in their reaction to CC>2 concentrations and
        it is reasonable to suppose  that a plant mix would eventually develop
        that would exploit available resources

     •  The impact reduces as the plant matures

     •  Catchments are not fully vegetated

     •  Published experience  relates to ideal  conditions,  e.g.,  greenhouses,
        where other stresses do not impinge

     •  "Greenhouse gases" other than CC>2 that may be responsible for  climatic
        warming are not beneficial to plant development

     •  Nutrient  availability  and light  conditions also  control  growth,  and
        stomatal  resistance  is  but  one  dependent state  variable within  a
        complex physiological process (Jones 1984).

Empirical Modeling

     This class  of model includes some that can operate in the  time domain
such as  regression  versions of  conceptual  models and others that consist of
associations between average  values but  imitate in an  empirical  fashion  the
causal models discussed above.

Time Series  Regression  Approach.   Nemec and Schaake  (1982)  reported  that in
1979,   two  researchers  named  Beard  and  Maristany  calibrated  a  regression
relationship between seasonal  runoff and climate data.   They then scaled the
precipitation and temperature input  data  to study  the  influence  on  runoff.
The conclusions they reached  are stated  to be comparable with those quoted by
Nemec  and  Schaake  for  their  conceptual  modeling approach.    This   can  be
explained by  Klemes'and Nemec's  (1985)  observation of  strong  linearities in
the  monthly  simulated  runoff   such   that,   although  ostensibly  nonlinear
(Figure 6),  the Sacramento model behaves as a regression equation with respect
to its primary inputs.

Analysis  of Historic  Records.    Palutikof,  Wigley,  and  Lough   (1984)  used
observed temperature,  pressure, and precipitation data to investigate possible
spatial and seasonal distributions of climatic change.  Noting that the effect
of  the   CC>2   increase   is   hemispherical  warming,   they   argue  that  some
appreciation of  its  effect  could be  gained  from  the  contrast  between  the
coldest (1901-20) and the warmest 20-year period  (1934-53) in this century.

     This type of analysis has been extended by this writer to consider runoff
from major European basins.   Table 2 shows the  difference, warm-cold,  between
catchment rainfall and runoff.

 Table 2.  Seasonal Rainfall and Runoff Differences for Some European Basins
                    (Difference  - Standard Deviation Units)
Winter P
Spring P
Summer P
Autumn P
     The rainfall  values  (P)  were obtained  by eye  interpolation from
     maps presented in Palutikof, Wigley, and  Lough  (1984).   The runoff
     (Q)  data  were  prepared   by  the  writer.    The  differences  are
     expressed as a multiple of the seasonal standard deviation.

     Total agreement  between  rainfall and  river  flow should not  be expected
because the latter is influenced by antecedent climatic conditions.  However,
agreement in terms of the sign of the change is good; only four of the twenty-
eight show  a  reversal,  and this  is  reduced further  if a half-season  lag is
incorporated into the comparison.  Apart  from  a small autumn increase, runoff
appears to have  reduced  during this  historical hemispherical warming;  in the
standardized units  shown  on Table 1, this  was most marked  in  the spring and
summer,  but in absolute  units  the winter reduction was more important.

     This  approach does   not  by  itself  provide quantitative estimates  of
climate impacts.   The most immediate question  concerns the extent  to which
non-CX^-induced warming can be represented, even in  muted form,  by contrasts
between historic sequences  in  which COp  warming  presumably  played  little
part.   Some  support comes  from the  GCM findings of  mid-latitude  cooling and
precipitation  reduction  (see   Section 6.3.4 of Wigley,  Jones,  and  Kelly,  in
     Another question  of interest  not  directly answered  by the  approach is
what  fraction   of  the  total  C^-induced  movement  is represented  by  this
historic  change?   In  terms  of  the  hemispherical   temperature  change  the
historical  warming  represented no  more  than  20 percent  of the  lower bound
estimates  of  the  C^- induced  movement.    The runoff reductions  are  very
variable—11 mm  in the Thames and  48  mm in the Rhone—overall of the order of
10 percent of the values shown in Figure 5.

Empirical Evaporation Model  Approach.   Cohen  (1986)  computes net  runoff from
the 766,000  km   catchment  area of the  Great Lakes  for current  and double (X>2
conditions.  The heart  of the model  is an empirical  formula,  a modified form
of the Thornthwaite model, from which monthly mean evaporation may be obtained
from  temperature and latitude.    A  soil  model  is also  included  from which

"climatonic" runoff  is  computed and then  calibrated to accord  with measured

     Temperature patterns from  two distinct GCMs  (GISS  and  GFDL) were used to
provide  alternative  sets  of  inputs for  the Thornthwaite  model.   Some  32
percent of the catchment area consists of  lakes,  so open water evaporation is
of great importance.   Alternative wind scenarios had to be superimposed on the
GCM results.   The effect of C02  doubling  varied  between a  15 percent  and 21
percent  reduction in  runoff  with  current  or GCM generated winds for  the
different models.  If  a reduced wind speed was allowed,  the runoff reduction
was diminished to only 4 percent.

     Cohen notes  the seasonal  differences  but annual  similarity  between  the
results  from  the two models.   Uncertainties  regarding the  future windfield
over the lakes and lack  of  information on  annual  variability preclude any but
very generalized conclusions about consequential water resource impacts.

Postscript on Empirical Models.   Close inspection of the empirical procedures
may leave an impression  of  "clutching at straws"  by the authors.  However, an
analysis  of  their weaknesses   reveals  few that  do not  equally  afflict  the
ostensibly more  "respectable"   causal and  conceptual approaches.   Those that
can yield a  time series,  the  regression,  and  Cohen's  approaches  have  the
advantage that they  can  provide inputs to water  resource  studies although in
the regression  case  the series will differ  from  the base series  in a purely
linear fashion, a very unlikely eventuality.

Water Balance Models

     The basic water  balance equation is:

             Q = P -  E

where  Q is runoff,  P  is  precipitation,  and  E  incorporates  evaporation  and
other  "losses."   A few authors have used  this relation to  either directly or
in modified form  quantify the  impact on Q  of climatic change acting through P
and E.

Average  Rainfall  Runoff Relations.  Stockton  and Boggess  (1979)  made  use of
relationships  for predicting   annual  average losses   from temperature  and
precipitation  that  had  been  derived by  the  U.S.  Geological Survey  (USGS).
They  investigated the  effect  on  runoff  of a GCM  predicted  2°C temperature
increase and 10 percent  precipitation decrease across  the  United States.  The
implied  sensitivities  were checked  by  Revelle and Waggoner (1983)  using an
interannual  regression   based  estimator  of  the  sensitivity of  the Colorado
River annual runoff at Lea Ferry to those same changes.  The agreement between
the approaches  was quite  good; 40  percent  reduction  by  regression compared
with  44  percent  by   the earlier  method.    Over  the  entire study  area with
average  precipitation of  30  cm and  4°C  average  temperature,   the expected
diminution  in  runoff  is 35 percent.   The most  severely affected  areas  are
those with lower rainfall and higher temperatures (see Waggoner,  Volume 3).

Transpiration Effect.  The conclusions of Revelle and Waggoner were  questioned
by Iso and Brazel (1984), who disputed the applicability of  the USGS empirical
relationships  in a  CC^-enriched  environment.   When  the  effect  of  reduced


 transpiration  (conceptual  models  section)  was  allowed  for  the  average  41
 percent runoff reduction on  five  study  basins predicted  by  the original method
 became  a  42  percent  increase.    The authors  list  shortcomings  in   their
 analysis,  though  for  the  most  part  shared  by  the original  analysis,  and
 conclude with a  salutory  phrase that "plausible additions  to state-of-the-art
 model studies of the  COo problem  can  dramatically  alter  current  perceptions of
 future consequences."

 Simple Water Balances.  Wigley and  Jones (1985) used  equation 1  to  investigate
 the  representativeness of  the apparently  differing results quoted above  as
 well  as  those of  Rind and Lebedeff  (1984)  and Aston  (1984).   This approach
 invokes  differential   calculus   to   derive  the proportional change  in runoff,
 AQ/Q,   due  to  a  given  proportional  change  in  precipitation,   AP/P,    and
 evaporation, AE/E.  The large proportional increases  quoted by  Idso and  Aston
 were  shown to occur  only in  regions  where P-E  is  small;  in  more temperate
 situations  the  rainfall   change  dominates  and  runoff  can be expected  to
 reduce.   The approach  has  been  extended by Glantz  and  Wigley  (in  press)  who
 separate  the contribution  of  deep  infiltration  loss  from the evaporation
 terms.   In the case  where  precipitation change  dominates, the amplification
 factor between  runoff and rainfall  change  is given  as  the reciprocal  of  the
 runoff  coefficient,  i.e.,  P/Q.    In  general  the  results indicate  greater
 sensitivity than those  predicted by Stockton and Boggess (1979).

 Non-subtractive Models.  The present  writer has investigated the source of  the
 difference between Wigley's and Stockton's conclusions.  It was suspected  that
 the main cause must be the  imposition of a purely subtractive formulation  for
 runoff formation in equation 1.   Examination of observed catchment precipita-
 tion  and  loss  values  reveals  tendencies  such  as those  shown on Figure  7  for
 Cypriot rivers.  The  reduction in  P-Q  as Q  increases can  be interpreted as a
 transition from a  proportional  response to  rainfall—runoff coefficient—to a
 subtractive response—constant losses.

     The United  States  Soil Conservation  Service  (USSCS) model  provides a
 suitable functional  form  for this  type of response  as  it  permits  the runoff
 coefficient  to  vary according to the state of a  notional  storage.   Although
 originally intended for "event"  analysis the USSCS  model  can  be modified  for
 use with annual rainfall and runoff data.

             Q = (P-I)2/(P+S-I)

 Parameter  I  represents  initial  abstraction  and parameter S can  be  thought of
 as a  maximum evaporation loss  such as would occur  if water supply were  not
 limiting.   A good  fit to  the Cypriot data of Figure 7 is achieved with S = 51
 = 1350 mm.

     Consider first the case of the  simple  water  balance  model of equation 1
with P =  600,  E =  535, and Q =  65  mm.  The runoff  ratio  is  therefore  0.11,
which, as explained above,  gives rise to a nine-to-one amplification in runoff
over rainfall change.   Because equation 2 is nonlinear, no simple  or general
 relationship exists between  the  sensitivity of an  individual year  to a given
climatic  adjustment  and   the average  over  all  years.    For  illustrative
purposes,  however,  it is sufficiently accurate to assume that the effect on an


    1000 ••
  •£ 400-
~  Kouri9
~  Zyghos
~  Dhiarizos
                 200       4OO        6OO       800       1OOO
                        Catchment  Annual  Rainfall  mm
     Figure  7.    Rainfall/catchment  loss  diagram  for  rivers  draining  the
Troodos mountains, Cyprus.  This  pattern is typical of semi-arid  regions  where
annual rainfall is seasonal and varies from 300 to 1000 mm.

individual year experiencing average  rainfall will  represent  the average over
all years.  Taking increments of equation 1  gives

                       AQ     2P(1-yei/ep)    P(1+Uyei/ep)
                       Q  "      (P-I)    +     (P-f-41)

where  y   is  the  ratio of I to P and  ei   and   ep  are  relative changes in I
and  P.    With P  =  600 as  before equation 3 yields  an  amplification  factor
between runoff  and  precipitation change of 3.3, very much smaller  than that
given  by   the  purely  subtractive  model  of  equation 1.   Under the  runoff
relation of equation 2  the  amplification  is controlled  by  \i   rather than by
the runoff coefficient.

Postscript on Water Balance Models.    The use  of time averages and  simple
formula have   the advantages  that  they  can  be   easily  calibrated  against
observed  data.    Nevertheless,  little  agreement exists between  the  results
obtained  by   the  various  approaches.   The disparities  are  as  much due  to
differences in  mathematical structure as to the uncertainties  that  accompany
extrapolation  to  a  different  climatic  regime.    This  was  demonstrated  by
contrasting  a  subtractive   and  a  partially  proportionate  model.    It  is
significant  that  as  well  as   equation 1  many  other models,   implicitly  or
explicitly, employ a  formulation in  which  runoff  appears as a residual from
rainfall.    This property is shared also by the structure of  the Sacramento
model (Figure 6) which may therefore also be expected to exaggerate the impact
on  low-runoff  catchments.   This  writer also  cannot agree  with Glantz  and
Wigley's   structure    that   extra    caution   is   required   "in   accepting
results ...  if these results  appear to  be at  odds  with simple water budget
calculations."    Equation 1 does  not  express  a  dynamical  truth  and  the
year-on-year  interaction  between  P  and E  needs  to be  considered in climatic
impact studies.



     Despite  its  popularity  as  a  paper or  report  title,  very  few  authors
consider impacts to water resource  variables in  the sense that  the section on
the distinction between hydrology and water resources describes.   Impacts may
be divided into the  altered requirement for water  due  to climatic change and
the altered availability  of water  to  meet  those demands.  Those authors who
have attempted  to quantify  water  resource impact have used the  reliability of
reservoir  supply (Figure 1)  so it is that measure that is discussed here.

Effect on  Demand

     The projection of future demand  for water  is  an  art practiced at various
levels of sophistication by water supply authorities.   Single purpose projects
such as irrigation or  power supply  pose a  relatively  simple problem in demand
forecasting.   However,  most planning problems are multidimensional and fraught
with uncertainties even within  the  existing climatic regime.   The imposition
of trend  on an already fluctuating climate was  considered by Schwartz (1977)
as  "additional  fuel  for  the   'let's  wait  and   see'  style  of  facilities

     In the same report Lofting and  Davis  (1977)  considered the definition of
demand as  it  relates  to water supply noting that it is  led by  supply and not
subject  to  the  same  checks  and  balances  encountered  with  conventional
industrial products and services.   In their paper, they analyze the demands of
various users separately and make  projections  to  the year 2020 for the United
States.  They consider the impact of climatic change through demand pattern in
the agricultural sector alone.

     Hargreaves  (1981)  considered  irrigation  project  planning  specifically.
He  reduces  the  problem  to  one  of  estimating  the  change  in  potential
evapotranspiration (PE), this being regarded as the sole determining influence
on water  demand.   His  analysis of  GCM results  leads  to  a  latitude related
increase  in  PE of  16  percent to  63 percent, which  is  to  be fed  into  U.S.
standard  procedures  for  computing  irrigation  need.     As  admitted,  this
assessment makes no allowance  for  changes  in clouds  (affecting  radiation),
precipitation, or photosynthesis.

     Cohen (1986)  presents  in a flowchart the many  interconnected components
of climate impact assessment  in his  Great  Lakes  study.   As well as the direct
effects on consumption  of an altered  temperature,  many  indirect  effects are
listed.   For  power  production these may take  the form  of a switch from water
resources  for  hydropower to  cooling water  for  coal,  oil, or nuclear power
stations.  Crops may require less  water  in a COp-enriched world,  and if the
study  area occupies  a zone  where  warming  occurs  then  enhanced  biological
activity requires a lesser dilution of effluents.

     Glantz and Wigley (in press) report briefly on demand impact and refer to
recent more  comprehensive work by  Callaway  and Currie  (1986)  who divide the
impact sector by sector.

Effect on Availability

Schwartz*  Methods.   Schwartz  (1977) offers  three  approaches to  the task of
evaluating the water supply impact:

     •  Case  studies  of  particular schemes  with special  reference  to their
        reaction to past anomalies

     •  Speculation  about impact  on broad  classes of  system attributes to
        types of change in runoff pattern

     •  Behavior  analysis  of  water  supply  facilities  with  synthetically
        generated inflow sequences.

     The  first  approach,  also used  by  Glantz and Wigley (in press), reveals
important  issues to be treated  in the second and  third approaches,  but does
not provide any quantitative estimate of impacts.

     The  result  from the second  approach  was an  "impact matrix"  showing the
direction of change in run-of-river, reservoir, and groundwater yield due to  a
decrease   in   mean   streamflow,   and  increase  in  variance,  skewness,  and
persistence.   Other attributes which were  considered  included water quality,
system reliability, operation costs, and flexibility to  change.

     In the  third approach Monte  Carlo  simulation was used  to generate 1000
year synthetic  traces.   Eight scenarios were  considered with  two  levels  of
decrease  in  the mean,  and increase  in  variance,  skewness,  and persistence.
The  flow  sequences  were  input  to  a   reservoir  whose  capacity  had  been
determined  to  give  5  percent  probability  of  failure  under  present-day
conditions.  The  most severe effect of  the  eight  was  the case of a reduction
in the log-mean of 20 percent but even this reduced the reliability only to 91
percent.   Restoring  the  reliability to  the original  95  percent  reduced the
yield by 20 percent.

Sacramento Model.   Nemec  and  Schaake  (1982) used their  adjusted  data series
(empirical modeling section) as recommended by Schaake and Kaczmarek (1979)  by
treating them as inflows to a storage reservoir.  They investigated the change
in yield  from  a given  storage  and  the change in storage  to  maintain  a given
yield.    In both  cases  the reliability  was  held  constant.   Klemes  and Nemec
(1985)  extended the results  to  consider  how  the reliability of supply from an
existing reservoir was affected by the same range of changes.

     Figure 8a shows some amplification  in the yield change over precipitation
change  but the more  dramatic  effect occurred  with  the storage change  for a
given  yield.    Figure  8b  shows  highly   nonlinear  results  with  up  to  20:1
amplification  over rainfall  change.    As  with  the  runoff experiments  the
precipitation change  was  more  critical  than the  evaporation change.    For a
drier climate,  the  drop in reliability  may  be  more  significant than the drop
in precipitation;  this  effect  is greatest on river  systems  with a high level
of  development.    In  one  of   the  experiments, a yield  equivalent  to  over
three-quarters of the  mean  flow was assumed.   To provide  95  percent supply
reliability this requires  a  600 million  nr storage.   Under the drier climatic
regime this reliability drops to below 25 percent,  i.e., in three out of every
four months the desired yield would be unavailable.

Empirical  Approach.     The United  Kingdom  Low Flow  Studies   (Institute  of
Hydrology  1980) demonstrated that the storage  yield diagram within a region
can be  stabilized if the  yield  is  expressed as a point  on  the flow duration
curve (FDC).   Figure 9  shows  such curves for  Britain, Malawi,  and Southeast
Germany.   Despite  considerable differences among  the  three  regions  it  is
notable that the curves are not entirely dissimilar.

     Beran  (1984)  considered  that  such   diagrams   permit  a  preliminary
estimation of  the impact  of climatic  change on reservoir  performance.   Like
Schwartz  (1977)  he assumed  that  climatic  change would  be  effected  through
changes  in  the  mean  and  variability   of   discharge.    Mo  change  in  the
standardized storage yield diagram was thought necessary.

     Consider a reservoir  in Malawi  designed for a yield  equivalent to the 80
percentile point on the FDC and supplied with 90 percent reliability, i.e.,  10
percent  failure   rate.    The  effect   of   a  change   in   mean   runoff  is
straightforward—a  decrease of 20  percent   leads  to  an  equal decrease  in
yield.    The effect  on  failure rate  is more  complex  as  it depends  on the
gradient of  the  FDC.   Figure  9  indicates  a storage  requirement equal  to 2
percent of the  annual runoff.    If  the average  runoff were  to  decrease by  20
percent then this reservoir would  appear to hold 2.5 percent  of  the  revised
annual  runoff.  For  most  influent streams to Lake Malawi the yield would now
correspond  to  between the 70 and 75 percentile  points on the revised FDC.


2 .a
!; 100
                             % changes In ETP
                               • -4%

                               A +4%

                               • +12%
                            ——— Leaf River
                            — —Pease River
N  400H
S  3OO
                                                               lncr*M*in ETP

                                                            4% lncr««*« In ETP

                                                            4% d»cr««»« in ETP
         -25-20   -10    0    +10   +2O+25
                % change in precipitation
     +35  +25      +1O    0    -K)
                % change in precipitation
             Figure 8.    Effect  of changes  in  precipitation and  evaporation on  (a)
       reservoir yield  for  a fixed  storage  equivalent  to  10?  of annual  runoff  and
       fixed  reliability  of  90?;  and   (b)  reservoir  storage  for  a  fixed  yield
       equivalent to 20? of  mean discharge and  fixed reliability.

     20 H






o 0-5-

 > 0-2
                                                Yield = 60 percentile flow
                                                Yield » 80 percentile  flow
                                                Yield = 90 percentile flow
                                                United Kingdom

                                       — — —  Malawi

                                       _ . mm  South-east Germany
            50  40   30    20       10     5       21
        Percentage of years  in which given reservoir would

            be inadequate to supply  the  stated yield.
     Figure  9.   Standardized  storage  yield  diagrams  for  United  Kingdom  and
Malawi.   This  differs  from Figure 1  only in  that  the  yield is expressed as  a
percentile on the flow  duration curve.

Reentering Figure 9  with the  revised  yield and  storage  values results  in a
postchange reliability of 70 percent equivalent to a  drop from a 10-year to a
three-year return period of failure.

     The effect  of  increased variability can  be  more dramatic.   If the same
reservoir  were   subjected  to  a 20  percent increase  in  the  coefficient  of
variation  (proportional  to the  gradient  of the  FDC  on  lognormal  paper)  the
percentile corresponding  to the  given yield  reduces from  80 percent  to  70
percent, which without the compensating effect  of  an  apparent storage increase
gives rise to a four-fold diminution  in reliability.

     The impact in the United Kingdom and Germany  is  more complex as the range
of FDC  types  is  much wider.  The largest impact  is on  rivers with steep FDCs
corresponding  to  catchments   of  impermeable  geology   and  low   base  flow
support.   Most  upland reservoirs in  the  United  Kingdom  occupy catchments  of
this type.  The  impact  is  aggravated  by  the relative  steepness of the storage
yield lines.

     Figure 9  applies  principally   to  smaller  reservoirs  where  over-year
storage  is   seldom   required.    McMahon   and  Mein  have   derived  simple
relationships expressing  reservoir  volumes  as  a  function of the  interannual
variability of  the  inflow  series  for use  in  more variable  climatic regimes
where over-year  storage  is  common.  These  indicate amplifications  between 1.5
and 2 between inflow variability and  storage requirements; in other words a 10
percent  increase in  variability leads  to  a  15  to  20  percent  increase  in
reservoir volume requirement (Beran  1984).

Use of  Comparative Hydrology Approach.  Of  the three  quantitative  procedures,
only the Sacramento model currently  uses climatic  inputs;  even so,  no specific
climate  model-based  scenario  was  considered.   The  other  quoted  procedures,
Schwartz'  and Beran's,  did not employ  climatic  inputs  but were  based upon
predefined changes  in the  runoff regime.    This  could  have  been  remedied  by
directly  relating  the  salient statistics  used  in those approaches—storage
yield, skewness, and persistence—to  controlling climatic statistics.

     The technique of "comparative hydrology"  is  one means  for deriving such
connections in  which hydrological  characteristics such as  those listed above
are  computed  from  the  available  hydrological data  and  then  compared with
corresponding  quantities  from different climatic  regions.   This  step will
introduce further uncertainty  and in  this  respect is  inferior to a model that
operates in the time domain and that uses climatic inputs directly.

     All  the  examples  described  in  the   previous  section assumed  that  the
inputs  would  change  but the internal model  structure and relationships would
be preserved  over the climatic change.   While  such a standpoint may serve for
a preliminary assessment any fuller  impact  evaluation  should assume that the
transfer function itself may change with the climate.

     Again  the  comparative hydrology  approach  may be  useful.    If  it  is
possible  to  express  climatic  change  scenarios in  terms such  as  "the  future
climate  of Britain  may be  similar  to the  current  climate  of (for example)
southeast  Germany"   then  the  current  data  for  that  region  (Figure 9) will
provide  the necessary answers.  This approach cannot readily  be applied  to the

conceptual  model approach  and  in  this  respect  the  empirical  approach  is

     This is an  approach that has already been  employed  in impact  studies of
food production by comparing climates within the North American midwest (Parry
1984).   Quantitative methods  of climate typology  such as  Litynski's  (1983)
procedure will be valuable in recognizing climatic analogs.


Introductory Remarks

     A  vast literature  has  accumulated  on  the topic  of climatic  change in
general and  impact  studies in particular.   The reference list at  the  end of
this paper  confirms  that surprisingly  little  of this  activity has concerned
hydrological and water resource futures.  Only twenty-one relevant papers have
been discovered,  the  majority of which treat  hydrological  aspects  only,  mean
annual  runoff  being  the most  commonly  treated variable.   Fewer   than  five
papers  consider  water  resource  oriented variables  in  a  quantitative  way.
Moreover, no consensus  on preferred methods of  attack  is  apparent  from these
few papers.

     An obvious  need  therefore exists  for more  research,  particularly  impact
studies  of  water resources,  and  the  following subsection  makes  suggestions
based upon  this  survey.   An  international framework  exists through the World
Meteorological Organization (WMO) and the United Nations Environment Programme
(UNEP) for impact studies and relevant aspects of their programs are described

Research Topics

     In the past, the main contribution by hydrologists to the climatic change
issue has been to paleoclimate studies and land-phase process descriptions for
GCMs.   Engineering hydrology  had tended  to  distance  itself from  the issue in
the  period  when  even  the sign  and time scale  of projected  change remained
uncertain.  Reaction was typified by one  correspondent to Rodda, Sheckley, and
Tan  (1977):  "water  management policy  could not be  based on models  with low
explained variance."   Since climate projections have  centered  on radiatively
active  gas-induced  warming a  little more attention  has  been given  to water
resource impact  studies.   The listed topics have  been  suggested  primarily by
shortcomings and uncertainties in the models that have been used so  far.

Climate-Hydrology Transfer Function.  This has been  a long-term preoccupation
of hydrological  science  for which  climatic change  studies  provide yet further
impetus.   Models of all  types  need   continued  development particularly  to
inject  increased  physical  content in conceptual models.   To date models have
tended  to  concentrate attention  either  on  hydrograph  generation or on  soil
processes.   Such models need  "balancing" to achieve a greater uniformity of
process description.

Hydrology-Water  Resource  Transfer  Function.    There  is  a need  for  "type
projects" on which  to  base  experiments  and  test  sensitivities.    The three
suggestions of  this  report  retain their  hydrological heritage  but  others are
desirable that  encompass more complex  schemes.   It  is important that  chosen


project  descriptions  be  realistic  and capable  of  transfer  to  a range  of
climatic and geographical regions.

Relationship  Between  Climatonic  and  Measured  Runoff.   Many  of  the  more
physically  based  models  equate  runoff with the  drainage  term  in  a  crop
model.   Were  general  relations  available  that  explained  the  spatial  and
temporal scale on which the two runoff  types  could  be equated,  or how one may
be translated into the other, then  greater  practical  use could  be made of the
causal model predictions.   An extra benefit from this  knowledge  would be the
ability to make direct use of GCM hydrological outputs.

Other Hydrology Variables.  There has been  a tendency to concentrate on runoff
as the  variable  of prime interest.   However, in certain  circumstances other
variables are of equal or greater  importance  and  some may be more amenable to
climate impact investigation.   Actual and potential  evaporation,  soil moisture
deficit, and groundwater recharge have  a good  claim for investigation.  Other
physical and chemical properties of  water should  also be studied,  for example
temperature and buffering capacity.

Effect  of  COp Enrichment  on Transpiration  Rate.   Although more  properly a
topic for  plant  physiologists rather  than  hydrologists, recent  research has
demonstrated  the importance  of  this  phenomenon on  water  resources  impact.
Statements  are  required  about  the  likely  vegetation  response,  and  these
(possibly alternative)  views  should be incorporated in  impact  models.   One
crucial  issue  is  the  transpiration response  to increased  (X^ of  different
vegetation  classes  growing in  an outdoor  competitive  environment.   Another
concerns  the  likelihood  and   time   scale   of  biome  adjustment   to  the
opportunities presented by a new regime.

Historical  Data.   Scenarios  based  upon past  periods of warming  and  cooling
will remain for  some time to  come  the  most  believable indicator of the likely
regional and  seasonal  (i.e.,  sub-grid) pattern  of  climatic change.   Further
research  is needed  to  extend  the  approach  to water  resource variables  in
different localities.   It  would  be useful for  standard regional sets  to  be
made available based  upon known contrasting  periods in order to  provide the
test bed for sensitivity experiments and intercomparisons.

Model  Intercomparisons.    At  a  later  time  it will  be  important  to  set  up
controlled  experiments  along the  lines of  previous WHO  intercomparisons  of
forecasting and  snowmelt models to  compare  the capabilities  of  various models
to forecast impacts on standard water resource schemes.

Empirical   Approach  and   Comparative   Hydrology.    Diagrams  such  as  the
standardized storage yield diagram will probably  retain their basic form over
a climatic  change.   Further work is needed to produce  such  diagrams (as well
as  other  diagrams  and  statistics  representing  different  water  resource
problems)  for climatic   regions,  to  identify  the  major   controls,  and  to
quantify  likely  changes  to  stated  climatic  variation.   Historic  data from
recent  past warm  and cool  periods can  provide temporal  validation  of the
approach.   There  exists  already  a program  within Unesco's  "International
Hydrology Pogramme" on comparative hydrology.

Water  Budget  Models.     Useful  insight  has   been   obtained   from  simple
formulations concerning  mean values and  annual, totals.  This  paper presents
ideas for generalizing their  form  to accommodate more  closely the statistical
dynamics  of runoff on  the annual  scale.  Further  development  is  needed to
apply  them  to   temperate  regions  and  to   incorporate   possible  reduced
transpiration and model inaccuracy.

Demand  Forecasting.    A  more  analytical  approach   is  needed  to  demand
forecasting through a period  of  climatic  change.   Direct and indirect effects
should be enumerated and a judgment made  about the direction of response.

Probabilistic  Impact  Forecasting.    Even  though  the scientific  community has
now arrived at a consensus  about  the likely direction  for  the  climate, much
will  remain uncertain  about  spatial  and  temporal  details.    By  attaching
weights to  alternative scenarios and models, probabilistic statements could be
made about  the range of water resource impacts.  Alternative design strategies
could  be  tested  in  this  way.   For  example,   an  important practical  issue
concerns  the period of  record that  should  be  used  for project  evaluation—is
it better to use  the  entire  period of record,  is it better to use a short but
recent run  of data,  or  should  designers project forward  in time  to  try to
account for possible  climatic  changes?    Data  requirements for recognizing
various levels of change should also flow from such an analysis.

Coupling  of Water  Resource  and Other  Impact  Variables.    Agricultural  and
energy  impacts have  a  similar socioeconomic   importance  to water  resource
impacts and are  also,  to some extent,  water  dependent.   Impact  studies which
could simultaneously encompass all  three  would  be of value in assessing their
relative  importance  in  particular  changed  circumstances,   and  in  ensuring
mutual compatibility in policy determination.

International Programs

     The  international community has recognized  the  crucial  importance of the
climate issue.   Perhaps  the clearest statement  yet  to emerge was that issued
by  the  UNEP/WMO/ICSU meeting  at Villach in  October  1985  (WHO   1986).   This
meeting asserted  that the  combined  concentrations of  "greenhouse"  gases will
be radiatively equivalent  to a  doubling  of CC^  concentration by as early as
2030 and  this will lead  to  a global warming of between  1.50 and 4.5°C.  There
was  little  residual  doubt  that there would  be profound  effects  on  global
ecosystems,  agriculture,  water  resources,  and  sea  ice, so  governments were
urged to support programs to improve knowledge and ameliorate effects.

     The  World  Climate  Programme   instituted  in  1979  following   the  World
Climate Conference  is  the  main  agency  for  coordinating  the  international
effort.   This  has four  components:  research, data,  applications, and impact.
This  last  component   is  being supervised  by   UNEP  but   is  also  strongly
represented  within  WMO's  Commission for Hydrology (CHy)  through  the  World
Climate Programme (Water).

     The WCP is  not  solely concerned with  the  climatic change  issue although
this has  become  increasingly central to  its  activities, especially  those of
the  WCRP.    UNEP's  programs  so far have  concentrated  on  food and  climate
variations,  and  more  recently on  ecosystem and societal impacts of climatic
change.    WCP  (Water),  largely  organized within  the WHO  Commission  for


Hydrology,  has  components  corresponding  to  the  main  divisions  of  WCP.
Research  topics  concern  the  use  of  long  term hydrological  data  in  the
detection of past changes as well as contributions to the HAPEX and other WCRP
experiments.   Nemec's,  Klemes',  and Beran's studies cited here  have all been
conducted under WCP (Water) auspices.


     Now  that  consensus  has  been reached on the  likely  direction  of climate,
at least  in  global  terms,  more effort must be  given to  projecting the likely
hydrological  and  water  resource  impacts.    In  this   paper,   hydrological
variables  have  been  described  which  may  substitute   for  water  resource
problems.   Models of varying  levels  of complexity have  been  presented  which
allow climatic information to be translated into hydrological terms.  Examples
from  the literature  illustrate  this  process  and  have   allowed  a  number  of
conclusions to be reached about fruitful avenues of future research.

     Scale and process shortcomings  in  climate modeling still  may radically
alter forecasts insofar as they affect individual regions.  For this reason it
is considered  that  historical data from the  study region should  provide the
base  data for  extrapolation  in  impact  studies.   In  this  approach  salient
hydrological properties  are  extracted from recently relatively  cool and warm
periods.  The  contrast  between  the  two  sets of statistics can be scaled up to
conform to the projected climatic change.

     No  preference  is  expressed between  models  as  all  have  strengths  and
drawbacks.   Conceptual  models  are  inevitably deficient  in their physical
description and  offer no greater certainty of correct performance than much
cruder  water  budget  or  empirical  formulations.    However,   it   is  easier
operationally  to allow for a climatic change through adjustments to  the inputs
to a  conceptual or  regression  model than empirical models such as  the storage
yield  diagram  or  water  budget  which  work   with time  average   quantities.
Nevertheless,  because the latter  can  make use  of climate scenarios based upon
analog  regions they  also offer,  through  comparative  hydrology,   a powerful
approach  to answering impact questions.

      Research  suggestions that  promote  all approaches  have been made together
with  a  framework for intercomparisons.   Researchers must remain  vigilant to
research  elsewhere  in climate science as  surprises that  may upturn forecasts
are still possible.   An example  from  the recent  past  is the possibility that
water  use by  vegetation may  be much   inhibited  in  a  CC^-rich environment.
Future surprises may  emanate from sea-level changes, and from new knowledge on
gas uptake  by  oceans and the  effect  of improved treatment of oceans, clouds,
and soil  moisture within GCMs.

      Research  findings  from paleoclimatology  and hydrology  should  also  be
studied  closely  for  information on  past analogs.   One obstacle  that  still
remains   to  acceptance  of   the CO-,  issue   is  that,   despite   an  already
considerable  increase in the  atmospheric  concentration  of radiatively active
gases  (possibly  20  percent, since preindustrial times), there has  been  no
unequivocal  and identifiable hydrological  impact  that cannot  be explained in
terms of  random fluctuations.

     This desirability for water impact studies to be part of the community of
climatic change  research  gives it the character of  "big science" in terms of
the  intellectual,  experimental,  and computational requirements.   While scope
exists for  individual effort  the  greater  dividends  will come  from concerted
research by networks of laboratories operating within the general framework of
the WCP.


     Thanks are due to colleagues at the Institute of Hydrology, especially to
Dr. Jim Shuttleworth  for  detailed comments on an early  draft  and to Dr. John
Roberts  for  advice  regarding  the possible  effect  of  CC^  enrichment  on
vegetation.   Dr. Tom  Wigley of  the  Climatic Research  Unit kindly provided
preprints and advance information on related studies.


Aston, A.R.  1984.   The effect of doubling atmospheric C02 on  streamflow:   A
     simulation.  Journal of Hydrology. 67:273-280.

Bauer, E.,  E.  Heise,  J.  Pfaendtner, and  V. Renner.  1984.  Development of an
     economical  soil  model  for   climate  simulation.   In  Current  issues  in
     climate research,  eds.  A.  Ghazi  and  R.  Fantechi,   219-27.   Dordrecht:

Beran, M.A.  1984.    Climate change:   New  problems for  water  resources  and
     hydrology.   In Proceedings  of International Conference on Hydrology  and
     the Scientific Bases for  the  Rational  Management  of Water  Resources,  pp.
     3-22.   Paris,   1982.  Unesco/WMO/IAHS.   SC-84/WS/53.

Brandyk, T., J.C.I. Dooge, and J.P.J.  O'Kane.  1985.   Soil moisture in general
     circulation   models.      Report   to    EEC/WEST  Project   CLI-038-EIR(H).
     Dublin:  Civil Engineering Department, University College.

Callaway, J.M., and J.W.  Currie.  1986.  Water  resources.   In State of the  art
     report on  the impact of  CO-induced  climatic change.  Washington, D.C.:
     U.S. Department of Energy (in press).

Callaway, J.M.,  and D.P.  Lettenmaier.  1986.  Effects  of hypothetical changes
     on  water  availability  in the Salt  River basin  (in press).   Richland,
     Washington:  Pacific Northwest Laboratory.

Carson,  D.J.  1981.   Current parameterizations of  land  surface  processes  in
     atmospheric global  circulation models.    In Land  surface  processes  in
     atmospheric   general   circulation   models,   ed.    Eagleson,   67-111.
     Cambridge:  Cambridge University Press.

Cohen, S.J.  1986.   Impacts of C02-induced  climatic  change  on  water resources
     in the Great Lakes basin.   Climatic Change.  8:135-53.

Glantz,  M.H.,  and  T.M.L.  Wigley.   In press.   Climatic variations  and their
     effects on  water resources.   In  Resources  and world  development,  eds.
     D.J. McLaren and B.J. Skinner. Berlin:  Springer-Verlag.

Hansen, J.,  G.  Russell,  D. Rind, P.  Stone,  A.  Lacis,  S. Lebedeff,  R.  Ruedy,
     L. Travis.  1983.  Efficient three-dimensional global models  for climate
     studies:  Models I and II.   Monthly Weather Review.  111:609-662.

Hargreaves,  G.H.  1981.   Water  requirements  and  man-induced climate change.
     Journal  of the  American  Society  of  Civil  Engineers,  Irrigation  and
     Drainage Division.  107 (IR3):247-55.

Hunt, B.C.  1985.  A  model  study of  some aspects of soil  hydrology  relevant to
     climatic  modeling.     Quarterly  Journal  of  the  Royal   Meteorological
     Society. 111:1071-85.

Idso,  S.B.,  and A.J.  Brazel,  1984.    Rising  atmospheric  carbon  dioxide
     concentrations  may increase streamflow.   Nature. 312:51-53.

Institute of Hydrology.  1980.   Low  Flow Studies Report,  Part 1.  Wallingford,
     U.K.: Institute of Hydrology.

Jones, M.B.  1984.  The effect of elevated  carbon  dioxide levels on the growth
     of crop plants:  An attempt to  predict the  consequences  for grassland and
     maize production in Europe.   In Current  issues in  climate  research,   eds.
     A. Ghazi and R. Fantechi,  294308.  Dordrecht:   Reidel.

Klemes, V.,  and J.  Nemec.   1985.   Assessing the impact  of  climate  change on
     the  development of surface  water  resources.   In  Sensitivity  of  water
     resource  systems  to  climate  variations,  ed. V.  Klemes.  Geneva:  World
     Meteorological  Office, WCP-98.

Kramer,  P.J.  1981.    Carbon  dioxide  concentration,   photosynthesis and  dry
     matter production.  Bioscience.  31:29-33-

Lettau, H.H., and M.W.  Baradas.  1973.  Evaporation climatonomy  II:   Refinement
     of  parameterization   exemplified  by  application  to the  Maabacan  River
     watershed.  Monthly Weather Review. 101:691-99.

Litynski,  J.K.  1983.  The  numerical  classification  of  the  world's  climates.
     Geneva:  World  Meteorological  Office,   WCP-63.

Lockwood, J.G.  1985.  World climate  systems,  292p.  London: Edward Arnold

Lofting,  E.M.,  and  H.C.  Davis.  1977.   Methods  for estimating  and projecting
     water  demands  for water  resource planning.   In Climate,  climatic change
     and  water  supply,   ed.  National Research Council, 49-69.   Washington,
     D.C.:  National Academy of Sciences.

McMahon, T.A., and R.G. Mein.  1978.   Reservoir capacity and  yield.   Amsterdam:

Mintz, Y.  1981.   The influence of soil  moisture  on  rainfall and circulation:
     a review of simulation experiments.   In Proc. Study Conference on Land
     Surface  Processes in  Atmospheric General  Circulation  Models.   Geneva:

Mintz,  Y.  1984.     The   sensitivity  of  numerically  simulated  climates  to
     land-surface  boundary  conditions.   In  The global  climate,   ed.  J.T.
     Houghton, 79-105.  Cambridge:  Cambridge University Press.

Natural  Environment  Research  Council.  1975.  Flood Studies  Report. 5  vols.
     Swindon, UK: NERC.

Nemec,  J.  1983.   The  concept  of  runoff  in the  global water  budget.   In
     Variations in theg global water budget, eds.  A. Street-Perott, M.  Beran,
     and R. Ratcliffe, 479-488.  Dordrecht:  Reidel.

Nemec, J., and  J.C.  Schaake,  1982.   Sensitivity of water resource systems to
     climate variation.  Hydrological Sciences Journal. 27:327-43.

Paultikof,  J.P.,   T.M.L.  Wigley, and  J.M.  Lough.   1986.   Seasonal  climate
     scenarios for Europe and North  America  in a high CO?, warmer world.   U.S.
     Department of Energy Report, TR012.   Washington, D.C.:   U.S. Department
     of Energy.

Parry, M.  1984.  Climate  impact  assessment  in cold  regions.   In Report of the
     study  conference on  sensitivity  of  ecosystems  and society  to  climate
     change, World Meteorological Organization, pp.  6-26.   WCP-83, Geneva.

Revelle, R.R.,  and P.E.  Waggoner,  1983.   Effects of a carbon  dioxide  induced
     climatic  change on  water  supplies  in  the  western United  States.   In
     Changing  climate:    Report  of  the carbon dioxide assessment committee.
     Washington, D.C.:  National Academy Press.

Rind, D.,  and S.  Lebedeff.  1984.   Potential climatic impacts  of increasing
     atmospheric COp with emphasis on water  availability  and hydrology  in the
     United States.  U.S. Environmental Protection  Agency Report.  Washington,
     D.C.:  U.S. Environmental Protection  Agency.

Rodda, J.C.,  A.V.  Sheckley,  and  P.  Tan.  1977.  Water  resources  and climatic
     change.   Journal of the  Institution of Water   Engineers  and Scientists.
     3276-83.  Communications 32:450-53.

Schaake, J.C.,  and Z.  Kaczmarek. 1979.  Climate variation and the design and
     operation  of  water  resource  systems.    Proc.  World Climate Conference.
     Geneva, Feb. 12-23.   WHO No. 537,  pp.  290-312.

Schnell,  C.   1984.    Socio-economic  impacts  of  a  climatic  change  due  to  a
     doubling  of  atmospheric  C02  content.    In Current issues  in  climate
     research,  eds.  A. Ghazi and R.  Fantechi, 270-87.   Dordrecht:  Reidel,

Schwartz, H.E.  1977.   Climate change and water  supply:   How sensitive  is the
     Northeast?  In  Climate,  climatic  change and water supply,  ed.  National
     Research  Council,   111-20.  Washington,  D.C.:     National  Academy  of

Sellers, P.J., and J.G. Lockwood. 1981.  A numerical simulation of the effects
     of  changing  vegetation  type   on  surface  hydroclimatology.    Climatic
     Change. 3:121-36.

Shuttleworth, W.J.  1983.   Evaporation  models in the global water  budget.   In
     Variations in  the  golbal  water budget,  eds. A. Street-Perott,  M.  Beran,
     and R. Ratcliffe, 147-71.   Dordrecht:  Reidel.

Stockton,  C.W.,   and  W.R.  Boggess.  1979.    Geohydrological  implications  of
     climate change on  water  resource  development.   Fort Belvoir,  Virginia.
     U.S. Army Coastal Engineering Research Center.

Thorn,  A.S.,  and  D.C.  Ledger.   1976.    Rainfall, runoff  and climatic  change.
     Proceedings  of the  Institution  of  Civil  Engineers,  Part 2.  61:633-52.
     Discussion, 63:721-31.

Wigley, T.M.L., and P.D. Jones. 1985.   Influences of precipitation changes and
     direct C02 effects on streamflow.   Nature.  314:149-52.

Wigley, T.M.L., P.D.  Jones,  and P.M.  Kelly.  In press.  Warm world scenarios
     and  the  detection  of  climatic  change  induced  by  radiatively  active
     gases.   In The greenhouse effect,  climatic change  and ecosystems,  eds.
     B. Bolin, B.R. Doos,  J. Jager, and R.A.  Warrick,  SCOPE  series. New York:

WHO.    1986.    The  role  of  carbon dioxide  and other  "greenhouse" gases  in
     climatic variations and associated impacts. WMO Bulletin.  35:130-34.

Woodwell,  G.M.  1983.   Biotic effects  on  the  concentration  of  atmospheric
     carbon dioxide:  A review and projection.  In  Changing  Climate, Report of
     the  Carbon  Dioxide  Assessment Committee.    Washington,  D.C.:  National
     Academy Press.


Determining the Options—The Role of UNEP in
Addressing Global Issues
Peter Usher
United Nations Environment Programme
Nairobi, Kenya
     UNEP  is the environmental consciousness  of the United Nations System.  It
is not an implementing  agency.  Research and application  of knowledge is the
responsibility of the specialized  agencies and for governments.   UNEP's role
is to  identify  environmental concerns,  assess the nature and  scale  of the
problem,  and as  necessary,  to develop and recommend methods or guidelines for
the sound environmental management  of  the  issue  concerned.   It does  so  by
means of  coordination and catalysis:

     0 Coordination of global activities and research  to  ensure the accurate
       analysis and efficient management of  the issues

     0 Catalysis  in  the   sense   of  encouraging  and  supporting  relevant
       activities necessary  to achieve  the goal of managing the  issue through
       avoidance, prevention, or mitigation  of the problem.

     UNEP  draws  the attention of  the world  to environmental  issues  and may
make  suggestions  that specific issues  be addressed.  The  decisions regarding
what  will be  done are those  of  the  UNEP Governing  Council, a representation
of member states of the United  Nations.  The  Executive Director of UNEP has
the  responsibility  to   satisfy  these  decisions,  and  my  task,  as  UNEP's
atmospheric specialist,  is  to ensure that the appropriate  assessment of rele-
vant  atmospheric  issues  is made in  a  timely and  efficient manner upon which
managerial decisions can be  based.   Thus the classic role of UNEP is problem
recognition, assessment,  and  then identification of policy alternatives.  UNEP
is currently  applying  this policy to two major global issues  -  the risks  to
the ozone  layer and the greenhouse gas/climate change issue.

     Some  may view this report as  a  collection of research results and infor-
mation on the impacts of modification of climate  or the ozone  shield.   I  am
obliged to look at  it  in  the  broader  context  of future global response  to
environmental concerns and in particular how this report and  the conference
aid me in making available to governments the  best assessment of ozone-layer

modification, its  impacts,  and the  implications  of climate change,  in order
that the most appropriate response can be formulated and applied.  This report
is to me  a single link, yet an absolutely essential  one  in the protection of
the ozone  layer and  in  dealing  with  a future climate never before experienced
by mankind.

     To  illustrate  this,  I  would like  to  describe  the  events,  these other
links in  the chain,  that have led us  to this point and  to  speculate what we
can hope to achieve.

     The hypothesis of Rowland and Molena of stratospheric ozone layer made in
1974 led to  worldwide speculation and  the  categorization  of the issue by UNEP
as an  emerging environmental  issue.   The  UNEP  Governing Council  asked the
Executive Director to convene an international conference to review the matter
and my  association  with the  Programme  began when  I was  seconded  from the
national meteorological  service for  whom I worked to organize  the conference,
which took place at the State Department building in Washington, D.C. in March
1977.   The outcome, as discussed by Golubev (this volume), was a World Plan of
Action on the Ozone Layer requiring research carried out or coordinated by the
appropriate specialized agencies and government and nongovernment institutions
under the  overall coordination of UNEP.   The initial task was  to assess the
issue and quantify the risk preliminary to determining the response.

     A Coordinating  Committee  (about twenty  active members, mostly government
representatives)  meets  regularly to  consider  recent research  results  and
assess ozone layer modification  and its impact.   Under  the  plan of action,
this  assessment  should consider  the  results  of  monitoring  and  modelling
changes of the  atmosphere,  and should estimate the  consequences for humanity
and the environment of  predicted  ozone layer change,  including UV-B radiation
effects on  human  health, on  aquatic  and terrestrial ecosystems, on plants and
agriculture, on climate  and  on  the  socioeconomic  implications  of such changes
or  of  mitigation  strategies,  a  menu  similar  to  this  report's   table  of

     The  Committee  first  met  in Geneva in  November  1977  and  has  met eight
times through June 1986.  Each year  its assessment has grown physically larger
as  knowledge improves;  this  year   it  incorporates the  results of  the  NASA
assessment  made by an  international  team over the previous year.  As a conse-
quence of having to  digest this voluminous  product,  it was  not  possible to
make an  assessment  of effects of ozone  layer  change  and,  also as a result of
the NASA  effort in addressing the physical  and  chemical  science of the ozone
layer, an  imbalance  in the quality  of the  respective parts of  the assessment
became apparent.  Effects science needed a boost and I decided on  the need for
a  scientific conference on effects  to inject momentum  into  the process to
estimation  of effects.   By  happy   coincidence,  EPA was travelling  the  same
road, and  we have combined  forces  to produce this  report  and the associated
conference.  My next job is to use the information presented here  and with the
help of  the Coordinating Committee  on the Ozone Layer (CCOL),  make an assess-
ment of effects to complement the already available physical assessment.

     UWEP,  in response  to governments' requests, also organized  workshops that
will, among other matters, assess the  costs  of some of the possible measures

suggested to  protect  the ozone layer.   Thus  the World Plan of  Action  on the
Ozone  Layer  is  being  implemented  and  the  assessments  made  are  the  best
possible under the circumstances.

     Nevertheless,  the issue is not  one that  can yet be confirmed by monitor-
ing  nor quantified  with certainty  by  research.    Unfortunately,  it  is not
possible to await these clarifications and still have time to avoid the conse-
quences of inaction.  UNEP has  been  obliged  at its Council's request to begin
the process of identifying management options for the issue.

     The Vienna  Convention  is now a reality and will  soon  enter  into force.
Soon an ad hoc legal and technical working group will be convened to decide on
a very significant document, a  protocol to the Convention on Chlorofluorocar-
bons, which will require  its signatories to  restrict  their  emissions  of CFCs
to  agreed  limits.    Observance of these limits  is  likely to exact  economic
penalties and  should not be overly  conservative.   UNEP's  Environmental Law
Unit will  convene  and service  these meetings.  My  responsibility will be to
provide  the  best available  scientific  and  technical  information, a  globally
accepted assessment of ozone layer modification, impacts of such modification,
and a socioeconomic analysis of the  consequences of  changes and  control.  The
program that I have outlined hopefully provides the lawyers and administrators
sufficient  factual  data  and realistic projections  on which  they  can  make
realistic policy  recommendations.   The  report  and  the  "Effects  Conference"
have been essential to that process.

     The option under  consideration  limits  production and emission of  CFC 11
and  12  as  an  initial  step.   Control of other CFCs and  additional  potential
ozone-depleting  substances   are  options  available  in the  future,  and  will
depend on  progress  in developing safer  substitutes  or technological  improve-
ments in manufacture  and use of  these  chemicals to prevent emissions.   UNEP
will have a  continuing  role as assessment coordinator  and will  act,  at least
temporarily, as secretariat to the Convention.

     The environment  should  not be manipulated beyond  its  capacity  to return
to its former  state.  Unhappily,  the limit has been exceeded beyond recall in
many  instances,  and  economic  demands  oblige  acceptance of  change  for the
greater good.

     Climate change is also under consideration by UNEP as a result of tropos-
pheric  trace  gas increases,  particularly carbon  dioxide,  and  the discussions
on  potential  impacts  have been most useful.    The  UNEP program in  this area
mirrors  that addressing the  ozone layer problem:  identification,  assessment,
and analysis of policy options.

     Two assessments  have been made  and a  consensus has  been reached  on the
probable physical  response.   However,   the  problem of response is far more
complex than for ozone.   Prohibition of nonessential uses of relatively small
amounts of  some  chemicals is one thing,  limiting  emissions  of carbon dioxide
from coal and  oil burning processes  is  quite another.   The 1985 assessment of
the  carbon dioxide/climate  question at Villach  provided  the  international
community and, one  hopes, governments  with  a broad program to  follow.  This
includes a comprehensive menu of research needs and a requirement to undertake
analysis of regional  socioeconomic   impacts  of climate  change.   Our  under-
standing of future regional  climates is currently limited and  a wide range of


possible as  well as  probable scenarios  will have  to be  considered  in  the
assessment process.   UNEP  will support  a limited  number  of  these  regional
exercises  and  will  encourage  governments  to  examine  the  consequences  of
climate change in their own regions.   The methods for carrying out such policy
exercises  are under  discussion.    One  can  be   sure  a  wide  range  will  be
considered.   Some idea  of the kind of thing  that can  be  done is  described by
Parry  (this  volume)  respecting  the vulnerability  of  agriculture  to  climate
variability and change in the climatic sensitive  regions.

     What  UNEP would  hope  is  that the  large  international  effort will  be
closely  coordinated  and that the  WMO/UNEP/ICSU  Advisory Group on  Greenhouse
Gases established in response to  a Villach recommendation  can  have a role to
play.   Only  after considerable research  and  assessment will we  be convinced
that a global response is or is not necessary.  The need for a climate conven-
tion  is not  as  obvious as  the need  for  a  convention  to  protect  the ozone
layer.   It  remains  an  option  that governments  will   have  to consider,  but
presently  we would  urge them not  to waste  the time  we have available  for
action,  whether  in  cooperation with  neighboring countries or  by themselves,
but to join in the policy analysis process.


Gus Speth
World Resources Institute
Washington, D.C. USA
     This report  is  another important step  toward  understanding and recog-
nizing the threats to  human health and the world environment posed by ozone
modification  and  climate  change  from a greenhouse  warming.  In  a remarkably
short period  of time, an  issue of  concern  to a small group of scientists has
become an  important  policy problem as  well.  The  focus of  scientists  and
policymakers  is no longer "whether", but  "when", and with what  effect.   We
have even begun to hear talk of the most  difficult question of all:  what can
we do?

     Consider what has  been  accomplished during  the past  15 months:

     •  In  March 1985,  twenty countries signed  the Vienna Convention for the
        Protection of  the  Ozone  Layer,  the first  international agreement
        calling for  action  to  prevent an  environmental  problem  rather than
        simply respond  to  it.

     •  In  October 1985, eighty scientists from twenty-nine countries met at
        Villach,  Austria to  assess  our  knowledge  of the greenhouse problem.
        They  concluded  that  the problem is not only real, but occurring much
        faster  than   had   been   thought,  and   recommended   an   "active
        collaboration"  between scientists and policymakers.

     •  In  January 1986,  the National Aeronautics  and  Space Administration
        issued a report  to Congress on processes  controlling ozone, based on a
        two-year  study  by  150   scientists  from  around  the  world.    While
        strictly a scientific study,  the report warns that "we are conducting
        one   giant   experiment   on  a   global    scale   by   increasing  the
        concentrations  of trace  gases  in  the atmosphere without  knowing the
        environmental consequences."

     e  In the  spring  of  1986,  the U.S.  Department  of Energy  released  four
        volumes summarizing  our  current understanding  of the  climate  change
        issue.  This effort  also  reflected extensive  international scientific
        participation and peer review.  The review of the global  carbon cycle
        concludes  with  the  recognition that  "human effects  on  atmospheric
        composition and  the size  and operations of the terrestrial eco-systems
        represent  major  excursions  that  may  yet overwhelm  a  life-support
        system crafted  in nature  over billions of years."

     •  On June 10 and  11,  1986,  a Senate subcommittee chaired  by Senator John
        Chafee held two days of  hearings on the  greenhouse  and ozone issues.
        Unlike prior hearings on  the subject,  this one gave  equal time to what
        can  and  ought  to  be done,  as  well  as  the   status  of  scientific
        knowledge.   Senator  Chafee  proposed  six  policy  initiatives,  and
        representatives  of government  agencies  were pressed  about their plans
        to address these problems.

     Government officials,  for the most part,  have responded to this explosion
of scientific concern with predictable and complacent calls  for more research,
but no action.  For example,  William Graham, the President's nominated science
adviser, told the Senate hearing  that "projections for the future have a large
uncertainty  to  them and  have  to be  reduced  before  we take  actions."   He
suggested we  need a "comprehensive understanding of  the global environment,"
which might take a decade.

     There  is no  denying  the  existence  of  major scientific  uncertainties.
However,  I  draw  a conclusion opposite  to Dr. Graham's.   When  dealing  with
risks that are grave but cumulative  and irreversible, uncertainty is a strong
argument in favor of action.   We  must take whatever steps are feasible as soon
as  possible   to  reduce  the  dangers.   Inaction  is  not a  neutral,  low  risk
policy,  but  rather a gamble that risks  much greater  harm.    As  the Villach
statement concludes, "While some  warming of climate now appears inevitable due
to  past actions,   the rate  and degree  of  future warming could  be profoundly
affected by governmental policies on energy conservation, use of fossil fuels,
and the emission of some greenhouse gases."

     A  wait-and-see approach also only makes sense if  we can  be certain that
the changes  will  be gradual  and  detected  in  time to  prevent catastrophe. The
recent  discovery  of an  "ozone hole" over the  Antarctic in  springtime demon-
strates that  natural systems  can  experience very  rapid,  marked changes almost
before  we have  time to confirm our measurements.  This phenomenon was not pre-
dicted  and cannot  be explained by the  best available models; nevertheless, it
happened over a period of less than a decade.

     There are other negative  consequences of delay.  Some strategies, such as
developing solar  energy  or  air  conditioning  equipment   that does  not require
CFCs, can only be  implemented over a long period of time.  Unless the develop-
ment  of these  technologies  is accelerated today,  these options  will  not be
available when rapid reductions in emissions become necessary.

     The  necessary policy  responses will  almost  certainly  become more severe
as  atmospheric  concentrations increase.  For  example, we can  limit  long-term
increases  in  the  atmospheric concentration of  CFCs  by  moderate action today,
such  as cutting  aerosol  uses and  capping total production.   Conversely, if


emissions are  allowed to  grow  unchecked, future  reductions will  have  to be
much more abrupt, without the benefit of time to search for acceptable substi-

     To allow the world's environment to deteriorate into a situation in which
we are dependent on rapid  and extreme  responses is irresponsible. This "cure"
may be as bad as the  disease. As  former EPA Administrator William Ruckelshaus
(1984) has  said, "The  ultimate danger  is  that by  remaining reliant  on the
'catastrophe theory of planning' in an  era  producing catastrophes of a magni-
tude  greater  than  in  the  past, we can  place our institutions  in situations
where  precipitate  action   is   the sole option—and   it  is  then  that  our
institutions themselves can be imperilled and individual rights overrun."

     It  is  also likely  to take far  longer  to  resolve the  major scientific
uncertainties than the  decade suggested by Dr. Graham.   In  all  likelihood we
will experience major impacts before we fully understand them.  The recent DOE
assessment of the greenhouse problem,  for example,  concludes that it is impos-
sible to predict when C02  induced  changes will  be  identified with statistical
confidence  with greater  precision  than "over  the  next few  decades."  The
problem,  they warn,  "may be too gradual  to detect until  it  is  well advanced,
too advanced  to stop  by  the  time it  is detected,  and  capable  of inducing
profound changes in the delicate environmental balances ..."

     The history of the ozone depletion issue reveals repeated assertions that
further research will  provide much better insights  on  the need  for policy in
three, five, or  ten  years  time.   When  the designated  period passed, the plea
for "just  a few  more  years"  has been  renewed.    Since research constantly
generates more questions,  the process can be repeated many times.

     Research will also not resolve the  inherent and  unavoidable uncertainty
attributable to human behavior,  the determinant of growth rates  for CC^, CFCs
and quite  likely  an  important  factor  in  emissions of  the remaining  trace
gases.  The  greatest  uncertainty  involved in predicting  CC>2 emissions is the
rate  of  growth  in  fossil  fuel use,  a  variable  that can  be influenced  by
government policy,  but not  predicted through scientific research alone.

     We must  assume  that  many  years will be needed to devise and implement
effective  international  solutions  to  these  problems.     To  avoid  future
political as well  as environmental crises,  we should  adopt  certain measures
that are feasible now,  recognizing that  they will be  only  first  steps.  While
multilateral and  international  cooperation  will ultimately  be  required, this
should not preclude  short-term  domestic action.  The  U.S. has  a  tradition of
international  environmental  leadership,   the  best  sources  of  scientific
understanding,  and the greatest resources. We can, and  should be, a model for
the world.

     Several steps should be taken immediately, among them:

     •  An international effort  to halt tropical deforestation,  which accounts
        for a  large  share  of  CC^ emissions  attributable to biotic sources.
        Emissions from  biotic sources  are currently estimated to be equal to
        20 to  40 percent  of  CC^  emissions from  fossil fuel use.  A detailed
        program for large-scale planting  and  forest  management  is outlined in
        a recent report of  the  World Resources  Institute,  the World Bank, and


        the  United  Nations  Development  Programme, Tropical  Forests:  A  Call  for
        Action.    This program  has  received  international  support for many
        reasons,  but it would  also  be highly beneficial  from  the  standpoint of
        climate  change.

     •   A  ban on nonessential  uses  of  CFCs  and a  production  cap to  limit
        future growth.   There  are proven, accepted substitutes for  almost  all
        aerosol  uses of CFCs  as  demonstrated  by the experience  of  the  United
        States,  Canada, and Sweden.   There  is  also no  reason  to continue using
        CFCs for retail food  packaging,  whipped cream stabilizers, and  other
        such uses.   Greater use  of  recovery and  recycling methods could  reduce
        CFC   use  for solvents,  foam  blowing,  and  refrigeration  and  air
        conditioning. Over  the  next decade,  we can  redesign  refrigeration
        equipment to operate on CFC-22 or other substitutes  that present much
        less risk of ozone  modification and climate  change.

     •   An increase  in energy efficiency.  Numerous  studies have  shown  oppor-
        tunities for large  reductions in the energy needed to power our cars,
        appliances,  lighting,  electric motors, and  industrial processes. This
        could reduce our dependence  on coal-generated  electricity, producing
        substantial  economic benefits while at the same time ameliorating  the
        global  acid  rain  problem.    Such efficiency  improvement would  be  of
        even greater benefit  in  the  developing countries, where the cost  of
        capital  is greater and the infrastructure requirements for  increasing
        coal production are enormous.

     •   Adoption  of  advanced technologies  for  producing  electricity from
        natural   gas.  This   could  increase power  supplies,  lower  costs,  and
        reduce   carbon  dioxide emissions  simultaneously.    Technologies like
        advanced combined-cycle  systems  and  steam-injected  gas   turbines  are
        technically feasible,  cheaper than coal, and  environmentally  superior
        in almost all respects.  Substituting them  for existing  gas  turbines
        would not increase  our  gas dependence and would significantly  reduce
        the  need for coal power.

     •   Tighter   regulations   to  limit  carbon  monoxide  emissions  from  auto-
        mobiles. This  would  reduce  a  suspected  indirect source  of  methane
        accumulation. Further reductions could  also be  attained  by  efforts to
        control  leakage  from gas pipelines  and abandoned coal mines.

     •   Preparation  of environmental impact  statements on all  projects that
        may   contribute   to or  be  affected   by climate  change  and  ozone
        modification, as recommended by Senator  Chafee.   This  is not  being
        done, although  we  believe  it  is  legally  required  by   the  National
        Environmental Policy  Act.    The  impact  statement  process is  valuable
        because   it  expands awareness of the  issue  throughout the  government
        and  would help to  identify further opportunities to  prevent and adapt
        to climate change.

     These steps can all  be justified not only because they reduce the dangers
of ozone modification and  climate  change but  because  they  are  environmentally
sound.   As the  Villach statement notes,  "Reduction  of  coal  and   oil  use,  and
energy  conservation undertaken  to  reduce  acid deposition  will also  reduce
concentrations of greenhouse  gases;  reduction  in emissions  of  CFC will help


protect the ozone layer and will also slow the rate of climate change." Imple-
mented today, they would substantially reduce the risks we face.

     More  aggressive action  will  be needed  in  the  long  run.    We  should
seriously study carbon taxes to discourage future use of coal and shale and to
finance reforestation,  solar  energy development, and  energy efficiency tech-
nologies.   We must  begin  the  transition  now or find  ourselves increasingly
dependent on fossil fuels for our prime source of energy.

     International cooperation  on these  issues must  be continued,  building on
the model  of the  Vienna Convention  for the Protection  of the Ozone  Layer.
China  and  the Soviet  Union  consume  as much  coal  as  the United  States  and
currently  plan   to  significantly  increase  their  use.    I  strongly  support
Senator Chafee's proposal for high  level government  meetings on these issues,
including the next US/USSR  summit.   It  is also  time to plan an international
convention to respond to climate change.  A convention could be a mechanism for
developing  common  policy  approaches  as  well  as  cooperation  on  scientific

     It would  also  be  a terrible  mistake if these problems were  viewed as
primarily  concerns  of  the  industrialized  countries.    In  the  developing
countries,  increased use   of   fossil  fuels  and CFCs could  largely  offset
reductions achieved  elsewhere.   Some  developing  countries, particularly those
with populations concentrated in low lying coastal areas, would be among those
to suffer most from  the effects of climate  change and ozone modification.  A
major focus of international efforts must  therefore  be to meet the legitimate
needs of the developing countries for energy and refrigeration without greatly
adding to the buildup of greenhouse gases.

     Lest  these  proposals  be  viewed  as  opposition  to further  scientific
efforts,  I  should  emphasize  my belief  that  increased research—particularly
research on  health  and environmental  effects—is also essential.  This  is so
not only  because we  need to  improve  our knowledge  of the risks we  face,  but
also in order to allow preparations for changes that are now inevitable.  Past
emissions of  trace  gases may have  already committed the  world  to  an average
warming of one degree or more,  and  therefore perhaps twice that in high lati-
tudes.   As a practical  matter, we  are  committed to  still more warming  and
climate change because  of  built-in  momentum in our  patterns of energy  supply
and use. It will take years to  develop and implement emission control strate-
gies and to replace the existing energy and industrial infrastructure.

     Research on effects is also important to help  define the risks of these
problems in widely understood  terms.   A major reason  for the extensive media
interest in  the  Senate hearings was  undoubtedly the  detailed  description of
potential impacts from ozone modification  and climate  change.   The  public was
told not only that  global average temperatures could rise several degrees,  but
that the  number  of Washington days hotter  than  100° could rise from  1  to 12
(see Hansen  et  al.,  this   volume).    Similarly,  graphic  terms  were used to
characterize  the  potential   devastation   of  U.S.   forests   and   shoreline

     We must take on these  problems as part  of  our  responsibility to conserve
and protect the  community  of  life on  the  planet.  Although  our dominion over
the earth  may be  nearly absolute,  our  right  to exercise  it  is  not.   The


responsibility for the greenhouse  and  ozone problems is ours.  As trustees of
the earth for future generations,  we must correct what we have wrought.


Ruckelshaus,  W.  1984.  Forward.   In Greenhouse effect and sea level rise, eds.
     Barth,  M.C.  and J.G.  Titus.   New York:  Van Nostrand Reinhold.

The Need for Responsible CFC  Policy
 Richard Barnett
 Alliance for Responsible CFC Policy
 Rosslyn, Virginia USA
     I am honored to present my views on the issues confronting us as scien-
tists, policy-makers,  business representatives,  interested participants,  and
most  importantly,  inhabitants  of  our earth.   The  tasks  we perform  in  our
efforts to better understand  the effects of  potential ozone depletion, global
climate warming,  and climate modification are important to  all as we assess
these issues.

     They have received a great deal  of attention  lately  as  a  result of the
hearings  held by   Senator  Chafee's  Environmental Pollution  Subcommittee.
Unfortunately, the  media  accounts  of those  hearings  could not  explain  the
enormous  scientific complexities   and uncertainties  surrounding the  ozone
depletion and climate change issues.   The  scientific  uncertainties  of these
issues are difficult for the general public to comprehend and they cannot be
adequately conveyed  on thirty seconds of network  news coverage or a brief news

     Hearings such  as these are useful  if  they help us to  remain alert and
responsive to the issues and the policy questions they engender; if they serve
to inform, not to  alarm;  and  if they promote rational  discourse and action,
rather than responses based on emotion.

     It is critical, therefore,  that  these meetings  are held and that reports
such as this are  generated to ensure the development of  international scienti-
fic consensus on  these important issues.

     Much has been  made  of the statement  from  the National Aeronautics and
Space  Administration (NASA)XWorld  Meteorological  Organization  (WMO)  Science
Assessment that  we're conducting an experiment "on  a global scale by increas-
ing  the  concentrations  of  trace   gases  in  the atmosphere."   Evolutionary
history tells us that the composition of the global atmosphere has naturally
been in a dynamic state.  The significant difference between the human race

today and  our  evolutionary predecessors  is  that today society  has  the capa-
bility  to  change  the  earth's atmosphere and also  has some  ability,  albeit
limited, to understand the consequences of our actions and to try to avoid any
adverse effects of our so-called "experiment."

     It is  somehow ironic that the  same  capacity that allowed  us  to  have an
Industrial  Revolution,  to develop new technologies, and  to land men  on the
moon, now leads us into the maze of complex questions we must now address.

     Such is the case with the development of chlorofluorocarbons,  or  CFCs as
they are  commonly  known.   These  synthetic compounds were  first developed in
1931 as  a result  of an intensive research  effort  to  identify  an  efficient,
safe refrigerant for home use.  Since that time they have come to be used in a
wide  variety  of  applications besides  refrigeration,  most  notably for  air-
conditioning,  foam product manufacturing,  cleaning solvent for the electronics
industry, and many  other  applications.   Their uses  have  become  so  widespread
because of their many desirable properties.  They are nonflammable,  noncarcin-
ogenic, noncorrosiva, extremely energy efficient, and they have low toxicity.

     The contribution of these substances to worker safety and consumer health
is  substantial  considering the products  in  which  they are utilized  and the
contribution of such products to  our  quality  of  life.  It is incumbent on us,
however,  to examine  these substances  and  their potential  for harming  the
environment in  the  long  run.   On the other hand,  we should not  rush  into
short-term  regulatory  decisions  that result  in  the use of  alternatives  that
present immediate,  real  threats  to worker and consumer safety with little or
no  environmental  benefit.   In this  case,  it appears  that the  penalties of
premature regulation could be real in terms  of an immediate increase in expo-
sure to more toxic substances or increased energy consumption.

     We  must  not  forget  that  other  substances  also  have  the   potential to
modify  the  ozone  layer and to  increase  the  earth's  average surface tempera-
ture.  Even if CFCs had never been discovered, these two issues would still be
with us today.

     The Alliance for Responsible CFC Policy dismisses  the notion of "wait and
see"  on the ozone  depletion  and  climate change issues  as unacceptable.   I
would hardly characterize  the activities over the  last twelve years as "wait
and see."

     The United States  and many  other countries  developed scientific programs
to understand the  ozone  layer and the processes  that control it.  Practically
all we know about the stratosphere has been learned in  the last ten to fifteen
years.  Furthermore, intensive programs  to study climate and possible climate
modification are  continuing.    In addition,  U.S. industry  strongly supported
efforts to develop the international framework for addressing the ozone deple-
tion  issues through the Vienna Convention for Protection  of the Ozone Layer.
Given  the  enormous complexities  of  the   issues we  are  confronting,  I  believe
that the progress  from  the scientific and policy development perspectives has
been remarkable.

     In  our view,  these  global  issues  can  only be addressed  at  the inter-
national  level.    Further unilateral actions  by individual  countries would
provide no  significant  environmental  benefit.   To this end, the progress made


to date  on international scientific  and  policy consensus  is  substantial but
much remains  to be done.   The science, as we  currently  understand it, tells
us, however, that  there  is  additional time in which to solidify international
consensus.  This must  be done  through discussion and negotiation, not through
further unilateral regulation  in the name of  "leadership."   Experience tells
us  that  other  countries do  not  regulate simply  because  the  United  States
does.   In fact, the opposite  may be  true.   Other nations may  view our uni-
lateral  actions  as  solving  the  problem;  thus,   giving  them  reason   to  do

     We are not saying,  however, that  industry  should  not  take precautionary
measures  while  the research  and  negotiations  continue at  the  international
level.    As Chairman  of  the CFC Alliance,  and as  the  Chief Operating Officer
for a major air-conditioning  manufacturer, I believe we  can,  we have,  and we
will  continue  to  examine  and  adopt  such  prudent  precautionary  measures,

    •  Recapturing, recycling,  and  recovery techniques, where cost-effective,
       during manufacturing processes to control CFC emissions

    •  Efforts  to  transition  to  existing  alternative  fluorocarbons that are
       considered to be more environmentally acceptable

    •  Equipment  change-out practices  to  replace,  where  possible, existing
       systems  at  the  expiration of their useful  life  to equipment utilizing
       other CFC formulations

    •  Technical practices in the field to prevent CFC emissions where practi-

    *  Encouragement of  CFC users to continue to  look  for  technologies, pro-
       cesses,  or  substances  that  are as  efficient,  safe,  productive, -or
       better than what is presently available.

     I  believe  such actions   respond  to  the  concerns raised  by  the  issues
discussed at the International Conference on the Effects of Ozone Modification
and Climate Change.    I  know,  for example, that many of  these  practices have
been put into effect already.   For example, unitary air-conditioning equipment
has converted  substantially over the  last  twenty years  to R-22  from R-12.
Many manufacturers have  instituted  recycling  and recapturing  techniques  in
their plants.    But as  the history of humanity has shown,  we are always striv-
ing to improve.   What we must guard against is a solution that could be viewed
as a step backwards.   Many of the CFC uses have no technically feasible alter-
native CFC formulations.  Other  substances discussed  as substitutes for CFCs,
including  sulfur  dioxide,   ammonia,  and  methyl  chloride  raise  immediate
concerns for environmental safety, and worker and consumer protection.

     The using  and producing  industries of the  CFC Alliance  are committed to
being active participants in the  exploration  and the  successful  resolution of
these issues.   We  would like to  think  that  the reasons our companies succeed
are because the products we offer contribute to  the  quality of life  for us
all, and we hope to continue to do so in a responsible fashion.

     The  potential  effects discussed  in  this  report are  serious.   Through
continued  pursuit of  scientific  research  relating  to  ozone depletion  and
climate change we can understand the nature and extent of any effects that may
actually occur and how  we can best deal with them.   I also  hope  that prudent
and responsible  actions on the  part of  government,  the private  sector,  and
industry  will  minimize  any deleterious  impacts on  society  as  a  result  of
premature or unwise regulatory decisions.

     As  I  indicated above,  we have  established the  framework of  an  inter-
national  plan   to address  these  important  environmental   concerns.    These
environmental  concerns   are serious,  but  their successful  resolution  will
require  greater   global  cooperation  in  conducting  the  necessary  scientific
research  and  monitoring,  and  in  developing  coordinated,  effective,  and
equitable policy decisions for all nations.

Cooling the Chemical Summer:
Some Policy Responses to
Ozone-Destroying and Greenhouse Gases
 David D. Doniger and David A. Wirth
 Natural Resources Defense Council
 Washington, DC USA
     This  exciting report has produced the most comprehensive  international
exchange  of knowledge  ever  seen  on  the known and  potential  consequences of
unabated  release  of ozone-destroying  and greenhouse gases.   We  will  not
attempt to  summarize  the papers or discussion,  except to say that  the picture
of our future being  painted by the world's best scientific minds is,  to use
the word  without  exaggeration, disastrous.   Unless  the private  and  public
institutions of the world respond, we can  look forward to steady increases in
UV-B radiation  due  to  ozone-destroying gases, including  CFCs,  halons,  and
other compounds.   Indeed,  Watson  (this volume) reports  that 2 to 3 percent
ozone depletion may already  have  been observed,  taking  into  account  the
contribution  of atmospheric aerosols.    As  a consequence,  we  can  expect
increases  in  fatal  and nonfatal  skin  cancer;  broad, systemic disruption of
human immune  responses;  cataracts and  other  vision  disorders; and possibly
many other  illnesses.   We can expect many serious environmental harms from the
increased  UV-B, including widespread  damage  to crops and  natural vegetation
(Teramura,  this volume), possible  disruption of microbial populations on which
the terrestrial ecology depends,  and possible  disruption of critical aquatic
food chains (Worrest, this volume).  We  can expect a certain amount of climate
change solely from  the  redistribution and overall loss of ozone.

     From  the greenhouse gases—carbon  dioxide,  methane,  nitrous oxide, CFCs,
and others—we  can expect a steady global warming  to temperatures that even
within a  decade may exceed the  range  experienced in  the last one hundred
thousand  years.   The  warming promises major  consequences for rainfall  and
other weather patterns  (Manabe,  this  volume)  and a  sharply  higher sea level
(Titus,  this  volume).    Agricultural  and  forest belts  are likely to  shift,
possibly  by hundreds of miles  (Parry,  this volume).   Even the major currents
and circulation patterns of the ocean may  change; northern Europe may plunge
into cold even  as  the  world as a whole heats  up.    The direct and indirect
health effects  of  the  warming are  likely  to  be massive,  ranging from large
increases  in  summertime urban  death-rates to death  and  disease  brought about
by drought, storm,  and  flood.

     These  catastrophic  prospects  should  bring  home  to  governments  and
ordinary  citizens  the  simple  concept  that  the  atmosphere  and climate  are
fragile.   The only  thing  objectively good  about the current  atmosphere  and
climate is that  they are the ones we are used to.   Life  and civilization  are
adapted   to   this    environment;   change   necessarily  will   be  disruptive.
Especially because so much of the  world's  population lives with but  a small
agricultural and economic  safety margin, almost any  stratospheric or climatic
change will be for  the worse.

     Yet,   except  in   parts  of the  world where the  weather  is traditionally
variable  and  where  life  is  lived close to  the margin, most  governments  and
most people take the  stability of climate for granted.  The common conception,
to this point, has been of an atmosphere resilient  enough to absorb stresses
of human origin without  danger to  existing  ecosystems and the economies which
depend on  them.  Even more dangerous is the  Panglossian  view,  encountered in
some governmental quarters,  that  greater  radiation and heat  will  be good  for
life on earth.  These illusions must be rapidly dispelled.

     Senate hearings  held in June 1986 on the greenhouse effect and threats to
the stratosphere went a long way to dispel  these illusions in this country and
to awaken  our  public and political system.   It is hoped  that the participant
in this  international conference  will  carry the  same  message to his  or  her
country.   The message,  however,  requires persistent  repeating  by scientists,
governmental  officials,  nongovernmental  organizations, and  ordinary citizens
around the world.

     To be  sure, much of the scientific data is  incomplete.   Yet the  tip of
the iceberg which is  visible argues  strongly  for  changing course now.   As EPA
Administrator  Lee  Thomas  recognized  in  his  Senate  testimony  and in  this
volume, with  the health  and environmental stakes  so high,  the  world  cannot
afford the luxury of waiting for certainty.   In fact, with the stakes so high,
uncertainty is an  even more powerful  argument for taking  early  action.   How
many unexpected surprises, like the Antarctic ozone hole,  can the world afford
to risk?  The appropriate standard of proof for this problem is aptly captured
in  the  American  Clean  Air  Act,  which   requires  action  to  protect  the
stratosphere when damage to  it,  and damage to health and welfare down below,
"may reasonably be  anticipated."

     Even though some degree of warming and  ozone destruction are now inevit-
able, the future extent  of the damage  is dependent on decisions to be made in
the  coming  few years.   The  magnitude and  timing of these  changes  are fully
under human control.

     In this  country, it  is most encouraging to see the  keen awareness  and
interest of key legislative leaders such as Senators Chafee and Gore, who were
joined  in  the landmark  hearings  by  Senators Stafford, Mitchell,  Baucus,  and
others.  We are also encouraged,  thus  far,  by EPA's conduct of the scientific
assessment  and decision-making process  pursuant   to  the  Stratospheric Ozone
Protection Plan, of  which this conference  is a part,  and by preparations for
resumption of CFC protocol negotiations next fall.

     Senators  Chafee and Gore have  laid out a number  of important, positive
policy initiatives  which NRDC strongly supports.   This report has demonstrated
to us,  however, that an  even  stronger policy  response  is required  by  the


magnitude  of  the threat  to global health  and welfare.   Our  purpose  in the
remainder of this paper  is  to  set forth some additional, though surely incom-
plete, proposals on CFCs and CC^.


     CFCs are  the  dominant chemicals  in  ozone depletion.   They also account
for  at least  one   sixth  of the current atmospheric  stock  of anthropogenic
greenhouse gases, and they are growing more rapidly than others.

     Domestic  and  international discussion  has  focused up  to  this  point on
alternatives for CFC controls  which  range  from  very modest  measures  to, at
most,  a  cap on production  at  current levels.   But there are now compelling
reasons to conclude  that  it is  no longer  sufficient to talk just about avoid-
ing  increases  in CFC production.   Even  with such a  cap,  CFC  levels  in the
stratosphere will  continue  to   increase.   Given  the gravity  of both problems
and the relative difficulty of rapidly affecting  emissions of other gases, CFC
reductions  clearly  present the  best  present  opportunity  to  stave  off ozone
depletion and to limit the global warming.

     Consequently,   MRDC  supports  actual,  rapid   reductions  in  production of
CFCs and  related  chemicals.  We offer the  outlines of  such  a proposal here,
acknowledging  that  many  details need  to  be developed.  We  propose,  both for
domestic regulation and international  agreement,  an 80 percent cut in permis-
sible  production of CFC^,  CFC^r halons,  and perhaps  other  compounds,  over
the next five years, and a total phaseout over the next ten years.
     We would exclude  from these production limits substances  such as
and CFC^oij^.  These are promising alternatives for air conditioning and refri-
geration.    They  are safer, at  least from a  stratospheric  ozone perspective,
because they break down in the troposphere.  At this point we are not sure how
to treat CFC22> which  could be  substituted in mobile  air conditioning.  It is
also an ozone  depleter,  but at  a  lower  rate  than CFC^o- Much  will depend on
whether CFC ^3 °r ^^134A become adequate substitutes and on whether the world
requires mobile air conditioning at all.

     In this  connection  it was most distressing  to  hear  the  major domestic
producers of CFCs state, at the March  1986 workshop and subsequently, that in
the early 1980s they essentially stopped  product development and toxicological
testing on  potential  alternatives  to CFC^  and CFCip-   Even  so, currently
available  information  indicates  that   CFC^Q   anc*   ^^134A  can   ^  reaclily
developed into  feasible  alternatives.   American industry  representatives to
the recent Rome meeting are reported  to  have  said that they can produce these
compounds for  three to six  times the amount of CFC^  and CFC12-   Tne cost
should drop off  with volume  and time; but even  if  it did not,  it still would
amount to less  than a $10 rise in the  price of a refrigerator.   It is also
most distressing,  given  the  industry's  prior opposition to  projections that
CFC production will  grow,  to  read  that the major American company, DuPont, is
planning to expand production of CFCs in  Japan.

     We recommend  consideration of  proposals to  give credits  against these
production  limits  for uses from which  producers  can prove, to  domestic and
international authorities, that  no  emissions  result.   Such  an  approach would
encourage closed systems and  product recovery  and  recycling.   Any such system


of exceptions,  however,  would  have  to be  based on  rigorous  demonstrations,
with the burden of proof strongly on the producer.

     A market-oriented phaseout should be supplemented  with  a  ban on specific
frivolous uses.  In the United  States,  substantial  strides have been taken to
eliminate non-essential aerosol uses.   But even  here certain  absurdities are
still cherished.  We ought to be able to reach reasonably quick agreement that
the modest virtues of using CFC^c as a whipped topping stabilizer or CFC^ in
chewing gum  remover,  boat horns, and  pressurized drain cleaners  should give
way to protection of the atmosphere and climate.

     A phase-out over  this  period would allow  time  for an orderly transition
to other  chemicals, other processes,  and other  end-products.   The gravity of
the  situation  demands nothing  less."    Cuts of  this  magnitude will  not even
prevent increases  in  the  level  of ozone-destroying CFCs reaching the strato-
sphere; that appears  to  require an 80 percent  cut  in today's  CFC production,
not production five years from now.


     We do not  pretend to comprehensively  address strategies  for C02 control
in this  paper.   We  want to  emphasize,  however, the enormous  opportunity to
stabilize  or even  cut CC^  emissions  from industrialized nations'  electric
power  generation  through conservation  and  efficiency  improvements.   We have
borrowed  heavily  here from  the  work  of our  colleagues  Ralph  Cavanaugh and
David  Goldstein in  NRDC's San  Francisco  office.    Here  are  some  startling
statistics they have compiled:

    •  The typical  American office  building  is  lit  by electric  lights that
       consume  six to nine  kilowatts of  electricity per square   foot  per
       year.    Commercially available  state-of-the-art technology  can reduce
       consumption to  1.5 kilowatt hours per square  foot  without sacrificing
       reading ability or other functions of lighting.

    •  Typical  residential  water heaters  use  4500-6000  kilowatt   hours  of
       electricity per year; state-of-the-art  technology only  uses 800-1200
       kilowatt hours per year.

The National Resources Defense Council estimates  that American residential and
commercial electricity use can  be cut  in  half through efficiency improvements
such as these.  NRDC estimates that industrial electricity use could be cut by
25 percent through conservation  and  efficiency  improvements.   In California
and in the Pacific Northwest, utilities have invested more than half a billion
dollars   in   residential   efficiency   improvements,   i.e.,  buying  efficient
refrigerators instead  of  building new  power plants.   These efficiency invest-
ments, moreover, are clearly cheaper.  Measures such as these,  applied nation-
wide,  could substantially trim projected C02 emissions.

     On the basis of this potential, we wish to offer the  outlines of a modest
proposal  for  establishing "C02 budgets"  for  the utilities  and industries of
the  United States.   For a start, Congress  should enact legislation  requiring
each utility  to develop  scenarios for  alternative C02  futures, in conjunction
with state regulatory authorities,  EPA, and other  agencies,  through an open,
objective  public  process.   These  scenarios  should  include low-  and no-C02


growth futures, and they  should  place  primary reliance on efficiency improve-
ments and alternative energy  sources,  without increased dependence on nuclear
power.   Similar  efficiency  analyses  should  be required  for major  energy-
consuming industries, including  the  transportation  sector.   Once the opportu-
nities  for  saving  energy  through  attractive  conservation  and  efficiency
investments are apparent, we will be able to take concrete steps toward reduc-
ing emissions of CC>2 and the global warming that results.

     We  have  no  illusions  that these  proposals  on CFCs  and COp  are either
fully  developed  or  sufficient  alone.    But  as  Senator  Gore  (this  volume)
suggests, the institutions of government, science, business, and other sectors
must not  be  paralyzed  by the enormity  of  the  total  problem.   Rather,  we must
break  the problem  down  and  come  to  grips  with as  many of the  pieces  as
possible, as  fast as  we can.   Even  if  significant  stratospheric and climatic
change still  results,  our efforts will be  justified,  for  surely we will know
that without them, things would be even worse.

Climate Change and Stratospheric Ozone Depletion:
Need for More Than the Current Minimalist Response

John C. Topping, Jr.1
Topping & Swillinger
Washington, DC USA
     Devotees of late night  television in the  United States may  have seen the
celebrated  W.R.  Grace and Company deficit trial  advertisements.  An elderly
man, presumably  of our generation,  is  on trial for failing to halt the mush-
rooming federal  deficit.    This  courtroom  scene  in  the  year  2017  has  the
defendant fading a jury of children  and an adolescent prosecutor  who asks,  "In
1986 the national debt reached $2 trillion.   Didn't that frighten you?"  Our
contemporary  lamely  replies,  "No one was  willing to  make the  sacrifice."
Stepping down from the witness stand,  he implores,  "Are  you  ever  going to
forgive us?"

     While  this  advertisement  may be  too strong for the sensibilities of the
television  networks,  it may be an  accurate representation  of how our grand-
children will  view our mortgaging the future  of  the  global environment.   By
2017 our grandchildren  will  be  struggling with  a radically  warming  world
climate causing  environmental  changes beyond  anything ever witnessed by  our
species.   Unless humanity  profoundly alters energy consumption and emission
practices, the global environmental  trauma could become the ecological equiva-
lent of the hyperinflation  in Weimar  Germany  in  the 1920s, with perhaps as
profound consequences for the health of democratic government.

     Mid-range projections  of  average global  surface temperature warming by
2100 are generally  at  or  above 5°C.    Virtually all   plausible energy-use
projections  (other than those  hypothesizing the destruction of  humanity in a
final world war) point toward a doubling  of the preindustrial levels of carbon
dioxide well  before  the end of the twenty-first  century.  We are told  by an
eminent climatologist that this C02 doubling alone would warm the  earth by an
average of  3°C.   As a result of  recent  calculations   in  some  of  the more
 The author served from September 1983 through  January 1986 as Staff Director
of  the  Office  of  Air  and  Radiation  of the  U.S.  Environmental  Protection

sophisticated climate models of various feedback effects associated with a CC>2
doubling,  the  likely warming  due  to  such growth  of  carbon  dioxide  in  the
atmosphere may prove to be at least 4°C (Hansen, this volume).

     Although fossil fuel burning alone could  produce  a doubling of preindus-
trial  CC>2  levels  in  the next  seventy-five  years,  such  a doubling  could be
accelerated by a decade or more if we continue to clear the world's forests at
current  rates.   The biosphere, especially  forests,  constitutes a  major sink
for taking up  carbon  dioxide so that it does  not build up in  the atmosphere.
Yet the  largest source  of  such uptake, the  tropical  rain forests,  is often
severely pressed by development pressures  in Third World  countries from both
indigenous sources and multi-national companies.

     Despite abundant evidence that the buildup of carbon dioxide levels alone
could  produce  an  average  global climate  warming of  5°C  by  2100,  the most
plausible  current  projections  indicate   that  this  COo-induced  temperature
warming  will  be roughly matched by a  temperature rise due to  a simultaneous
buildup  in  other  trace  gases  as  chlorofluorocarbons, carbon  tetrachloride,
nitrous oxide, carbon monoxide, methane, and compounds of bromine.

     The long-term  implications of  such a rapid climate  warming for humanity
and most other  animal and plant species on this planet are staggering.  These
involve  losing  tens of  millions of  acres of coastal land  to the  sea,  the
impairment of many of the world's port facilities,  the loss of countless lives
of coastal and  lowland dwellers to  increasingly destructive storm surges,  and
massive salt water intrusion on fresh water supplies.   The greatest human toll
is likely to come in famines that could ultimately claim the lives of hundreds
of millions in Third World countries.  The tragic famine we are now witnessing
in Africa may be but a  small harbinger  of  the catastrophe that will unfold as
desertification  claims  arable  land and shifts  in  rainfall patterns disrupt
already  marginal food  production   and  distribution systems.    Although many
plants will grow more rapidly in a higher C02 environment, it is unlikely over
the next century that such  potential gains in the  food supply would match the
losses caused  by migration  of  food belts.  Moreover,  the non-COn greenhouse
gases provide no such benefits but could disrupt the climate just as much.

     Besides  its potentially large  toll  in human  lives,  this  rapid climate
change will  probably  destroy  the  habitats of thousands of animal and plant
species,   thus  robbing our  planet   of much of  its  biological  diversity.   In
addition to  its direct human  and  ecological  toll,  adjustments  to this rapid
climate  change  could  be   expected  to  consume the   entire  growth  dividend
humanity would have used to  improve living standards of future generations and
to redress the  impoverishment of the Third World.  Huge sums would be required
for dikes  to  preserve  many  coastal  areas,  and much  of our  planet's great
investment in agriculture would be made obsolete by a rapid change  in climate.

     As  apocalyptic as the  consequences of such climate shock would appear to
be over  the long haul, people tend to glaze over at discussions that seem more
relevant to their  great  grandchildren.  Economists  have developed the concept
of discounting  the value of human  lives  saved  in  the future.   As some have
pointed  out, by  adopting a  high enough  discount rate,  one  could conclude  that
the destruction of human  civilization one  or two centuries  from now  is of
little or  no economic  consequence  in  present value  terms,   even  if such  a
certain  fate  results  from today's decisions.   Sadly  the calculations of most


public policy-makers  seem equally unmindful of  the injunction of  that great
political philosopher Edmund Burke that government owes a duty not only to the
living, but also to the dead and to those who are yet to be born.

     Yet,  as  some  eminent  NASA  scientists  have recently  projected,  climate
change is  no  longer merely a  concern for future generations.  By  1990,  they
suggest,  we are likely  to see  temperature changes outside of the normal range
of variation with an  increase  of  about 1°C in  average global surface tempera-
ture between now and the year 2000.  At the end of this century humanity could
face a warmer  world than at any  time  in  the last hundred thousand  years with
the prospect of a continued warming at an ever-accelerating rate.

     A major factor in  the  acceleration of these climate warming projections
has been a late-blooming recognition of the important  role of highly absorp-
tive greenhouse gases,  such as chlorofluorocarbons,  carbon tetrachloride,  and
various bromine compounds, which  remain in the  atmosphere  for decades if not
centuries  before  breaking  down.     Methane  growth  due  to agriculture,  and
nitrous oxide and carbon monoxide growth from fuel combustion have also caused
the warming projections to rise.

     Because of their powerful bonding and insulating qualities, chlorofluoro-
carbons are being used for a profusion of uses  well past refrigeration and air
conditioning to such diverse purposes as computer manufacturing and preserving
the taste  of  a  "McDLT"  sandwich from a fast-food  restaurant.    These  same
qualities  that make  chlorofluorocarbons  such  readily adaptable  commercial
compounds also make them  powerful  greenhouse gases  and effective  depleters of
our stratospheric ozone layer.

     Due to the rapid growth in CFC use during  the 1980s and the rapid buildup
of  bromine compounds,  the  danger  to the stratospheric  ozone layer  appears
greater  than  was  originally   thought when concern  first  arose  in  the  mid-
1970s.   Yet much more  important   than  the sizable increases  in  skin cancer,
potential weakening of  the human  immune system,  and likely crop and materials
damage resulting from a chlorine and bromine compound induced depletion of the
stratospheric ozone layer are  the  sizable boosts that such gases  will provide
to climate warming.

     Given present CFC-use projections, it is conceivable that these compounds
alone  will produce  an  amplification of  1°  to  1.5°  C  in the  total  global
temperature warming over  the  next several generations.   A small  increment in
the  greenhouse warming  can  also  be  anticipated  from  the  growth of  such
chlorine compounds as carbon tetrachloride and  methyl chloroform as well as by
a profusion of bromine compounds.

     These stratospheric  perturbants  would be  the first targets of  any effec-
tive  international  effort  to  slow  the  greenhouse  warming.    They are  all
industrially-produced compounds,  many virtually  unknown  two  generations ago.
Many have  some available  substitutes,  which  are  neither stratospheric pertur-
bants nor  greenhouse  gases  for most commercial  applications.  Regulation of
these compounds could afford  dual  benefits by  providing stratospheric protec-
tion and slowing the greenhouse warming.   Control of even small volumes of

these chlorine and bromine compounds can  have  a  highly  beneficial effect, due
to their  long persistence in the  atmosphere and their  tremendous absorptive
potential  (sometimes  several  thousand  times an  equivalent molecule  of carbon

     Any effective U.S. and international effort to  address  the dual problems
of greenhouse  warming and stratospheric  ozone depletion will  require  a  much
more tangible commitment of resources and attention by senior policy-makers in
the major  industrial  nations.   In the United  States there  recently  have  been
some encouraging  indications of  intellectual recognition  by  senior  policy-
makers  of  the  importance  of  the  problem  of  climate  change.    William
Ruckelshaus' strong foreword to the provocative work Greenhouse Effect and Sea
Level Rise  (Earth  and Titus  1984) and Lee Thomas' clarion  call in March  1986
for  stratospheric  ozone  protection pointed  out the  critical  importance  of
these issues.

     Yet  recognition   of  these  issues has  not  translated  into a  tangible
resource  commitment.    Despite  indications  that  prospective  climate  change
dwarfs  all  other  environmental  problems  combined,  fewer  than  nine   EPA
employees  out of  a  workforce of  about  12,000  work full-time on either the
issue of  greenhouse  warming or stratospheric  ozone  depletion.   Although the
proportion has grown considerably in the past year,  still only about one-tenth
of  one   percent  of  EPA's  total   resources  is  devoted  to  addressing  these
issues.    This  minimalist  resource commitment is not due  to  a lack of will on
the part  of EPA's senior  leadership but to the  statutory ambiguity of EPA's
role and  the deep-seated hostility  that  the Office of  Management and Budget
(OMB) has exhibited  toward  EPA's  addressing  such  first-order environmental
issues as  climate  change,  indoor  air,  and radon.  OMB has resisted giving EPA
resources  in apparent fear that the agency might  uncover problems with result-
ing budgetary  or  regulatory consequences.   While somewhat  greater  sums  have
been  directed  at  studying  aspects  of  climate   change   at  the  National
Aeronautics   and   Space   Administration,   the  National  Oceanographic   and
Atmospheric Administration, and the Department of Energy, no coherent federal-
wide focus exists  on  these problems,  which will ultimately  affect  virtually
every agency in government.

     With  a few  exceptions  the  resource commitment to  seriously addressing
these issues by U.S. environmental groups has been equally fainthearted.  Some
environmentalists, by reflexively opposing  all  forms of nuclear power,  have
helped  increase our  dependence on  fossil fuels.  Moreover,  both greenhouse-
induced  warming  and  stratospheric ozone depletion  are  global  problems, and
environmental   policy-makers   and  environmental   organizations   in   other
industrialized countries have often lagged behind those in the U.S. in addres-
sing these problems.

     Despite  the  clear inadequacy of the present response  to the prospect of
climate change, a  number  of  tangible steps  can be taken by the U.S. and other
countries  to make  the problem more manageable.  These include:

     •  Preservation  of  critical  satellite  tracking   research  and  climate
        modeling  concerning  ozone  depletion  and  greenhouse  warming.    The
        recent  turmoil at NASA,  the destruction at launch of  a key weather
        satellite,  and the  general environment  of  federal  budget  austerity
        have  placed  some critical  research  in jeopardy.    Its loss   could


greatly hinder our developing a clear understanding of the likely pace
and regional implications of climate change.

Development of a  serious  external  constituency for climatic stability
including  environmentalists,  scientists,  coastal dwellers,  farmers,
the nuclear and  the solar  industry,  and others.   If such coalitions
are organized  in  the major  industrial  democracies,  they  can  help to
ensure the attainment of broadscale international agreements essential
to responding to climate change.

Immediate phasing out of  such  relatively frivolous and easily substi-
tutable  uses  of  CFCs  as  aerosols,  egg  cartons,  and  fast  food
containers  through  a  combination  of  consumer  boycotts,  industry
voluntary  action, national  regulation  or  taxes,  and  international
agreement.   Concerted  activity by  consumers  to  avoid CFC  aerosols
preceded U.S.  regulatory action.  A similar focus on the threat to the
environment posed by such environmentally dangerous and frivolous uses
of CFCs  might trigger  a  similar  scramble  by fast  food  retailers to
point out that their products are  not only tastier,  but safer for the
environment, than those of their competitors.

Approaching regulation  of CFCs and other  trace gases such  as carbon
tetrachloride, methyl  chloroform,  and  bromine compounds  by  weighing
their greenhouse absorptive potential, their residence time, and their
ozone depleting capacity.  Thus, even among CFCs,  substitutions might
be encouraged of  longer residence  time  compounds by  shorter residence
time compounds.

Where appropriate,  amend environmental  statutes to  address  possibly
harmful substitution effects caused  by  proposed environmental regula-
tions.  EPA recently  issued a  notice of an intent to list perchloro-
ethylene,  a  significant  dry  cleaning   solvent,  as  a hazardous  air
pollutant under Section 112 of  the  Clean Air Act.   Such  listing and
subsequent  regulation  could conceivably  trigger a  rapid  movement in
the dry  cleaning  industry  to   CFCs,  on balance  likely to  pose  much
greater  environmental  dangers  than  perchloroethylene.     Language
specifically permitting consideration of greenhouse  warming and ozone
depleting  effects  of  available  substitutes would  facilitate  more
rational regulation.

Requirements  that  long-term  projects   reviewed under  the  National
Environmental Policy  Act  consider  climate  change and  sea  level rise.
This might  be accomplished through amendment of implementing regula-
tions or  by  a  suit directed  at  an  appropriate  project,  such  as  a
coastal highway  that has  failed to  consider likely sea  level rise.
Institutionalization of climate change considerations into the project
planning process  for  long term public and  private  sector  investments
should ultimately save taxpayers and investors billions of dollars and
ensure that we leave future generations  a sounder environment.

Insistence  by bilateral  and multilateral  lending  institutions  that
their funds not finance the destruction of  tropical  rain  forests and
other important  global  forest  resources.   Recognition of the global
stake in  preserving these  resources, located largely in developing

        countries,  may ultimately  require  the  industrialized  countries  to
        finance some of the economic  loss  developing  countries  would perceive
        occurring should they forego development in the tropical rain forests.

     •  Recognition, even in the face of the tragedy at Chernobyl,  that along-
        side energy  conservation nuclear  power  remains the  likely  principal
        alternative to  fossil  fuel  burning over  the  next generation  or  two.
        Besides strengthening  the powers  of  the  International  Atomic Energy
        Agency  to improve safety  standards and  ensure against diversion  of
        nuclear materials  to  nonpeaceful uses, recapturing  public confidence
        in  the  safety  of nuclear  power  will  require implementing  environ-
        mentally  sound  radioactive  waste  policies.    It  is  likely  that
        environmental  and energy  considerations  may  militate  toward  large
        scale nuclear plant construction before the end of  this century.   Yet
        to meet  deep public  concerns  about the  safety of nuclear  power  may
        require the  development of  a new generation  of  technology  that  is
        relatively   invulnerable   to   nuclear   meltdown.       Governments,
        environmental  groups,  and the  nuclear power  industry  have a common
        interest  in the success  of such  an effort.   Simultaneous satisfaction
        of  safety  concerns  and economic  considerations will  require  very
        rigorous  safety standards coupled  with  a  virtual  regulatory  certainty
        that  careful  adherence  to  such  standards  will  ensure  expeditious
        licensing and avoid protracted litigation.

     •  Renewed international  research focus  on  improving the  efficiency  of
        such alternative energy  sources as solar  power  and  perfection  of such
        potential breakthrough technologies as hydrogen powered  vehicles.

     •  A  concerted  public-private  sector effort  to  mitigate  the  costs  of
        climate change by  adjusting  land use  and  coastal  development policies
        to allow for the prospect of increasing sea level  rise,  by encouraging
        the  development  of  crops that  can flourish  under  rapidly  changing
        climatic  conditions,   by factoring  climatic  change and  rising  CC^
        levels  into  the  strategic planning of  industry,  and  by incorporating
        the prospect of a rapidly warming climate into calculations of utility
        peak load demand and necessary future capacity.

     Due  to the  enormous  momentum  in  the  earth's  climate system  already
present from our  massive  increase in fossil fuel  burning  and release  of many
manmade  greenhouse  gases,  adaptation  will  be   an  essential  part  of  our
response.   The  sooner we  incorporate  climate change considerations  into  our
societal planning  the  better  our prospect will be  for  turning  the greenhouse
warming  from an  unprecedented  catastrophe  to  merely an enormous  societal
challenge.   Yet,  even the best  adaptation and mitigation strategy  may be of
little  avail  if  present  energy  use and  gas release trends  remain  unabated.
Action by EPA and UNEP to sharply curtail emissions of the greenhouse gases at
issue may  afford  us the time  to fashion  the  more comprehensive response that
will ultimately be required.


Barth, M.C., and J.G. Titus. 1984.  Greenhouse effect and sea level rise.  New
     York: Van Nostrand Reinhold.

The Implications of Health  and Environmental
Effects for Policy
Daniel J. Dudek and Michael Oppenheimer
Environmental Defense Fund
New York, New York USA

     In approaching  the question of whether the protection of  stratospheric
ozone requires  the control of CFCs and/or other ozone-depleting agents, the
U.S.  Environmental Protection Agency  (EPA)  has  divided the  problem  into  a
consideration  of future emissions and atmospheric and  health  effects.   The
first workshop was devoted to the former topic.  From that session  it is  clear
that  in the absence of  policy action, CFC use  will  continue  to grow in the
future.  While the papers presented  provided a  wide  range  of  possible future
CFC levels,  it is  important to emphasize that no credence can be  given to zero
growth assumptions for either long term production or emissions.

     The various papers in this report  focus on the consequences  of  failing to
manage  emissions  of  ozone-depleting  substances.    These  environmental and
health effects are the damages that  can  be  avoided  by  controlling  emissions.
These impacts  can  be distributed  into direct  and indirect categories stemming
from  UV-B radiation or climate change.  The effects documented  involve primary
production sectors such as agriculture and fisheries, as well  as higher-order
damages to  human  health,  materials,   and  coastal  property.  Little  of the
research has  generated  information  in sufficient detail  to  estimate  dose-
response  relationships.   However,  when  we  compare our  current knowledge
against the information on  effects available in the late  1970s,  we  can see
that  there has been substantial learning on this issue.   Several  of  the papers
presented a sufficient range of  responses to allow a preliminary  examination
of the benefits of controlling emissions.

     Two trajectories  of ozone depletion  in response  to  alternative scenarios
of chlorofluorocarbon  (CFC) and trace gas emissions were selected to represent
the 'with'  and  'without1 action extremes.   These depletion  predictions formed
the basis  for predicting  effects in  the United States.   Control action is
simulated in  the  atmospheric model by  holding emissions  constant  at   1980
levels which is  broadly  similar  to a production capacity  limit.   The main

analysis focuses on nonmelanoma skin cancers and the damages to PVC.  Indirect
effects  of   increased   UV-B  radiation  are   illustrated   by  extrapolating
photochemical oxidant production and its yield-reducing  impacts  on major crop
commodities.   The avoided damages from now  until  the year 2025 were estimated
to be approximately 1.65 million cases  of nonmelanoma skin cancer and approxi-
mately $7  million  of PVC damage.   Yield reductions due  to  increased ambient
ozone  would   be significant  for   soybeans  and  cotton  beginning  in  2010.
Numerous effects not quantified are  described qualitatively.

     To   complete   the   analytical  illustration   of   stratospheric   ozone
management,  the focus  naturally shifts  to the  estimation of  the  costs  of
control action.  The changes in  economic surplus  associated  with the position
of a  production cap  were derived  from existing  research  results.   Average
annual present  value control costs  were estimated to  range  from $196 million
to $455  million, depending  upon the  ease of  substitution of  alternatives.
Transition costs might  also be  substantial if policy action is  implemented
with little  advance notice.  The safety implications  of  the substitutes also
were neglected.  However, given  the relative magnitude of  benefits and cost,
the prudent  course would be  to focus on  the design of efficient  and equitable
management strategies for limiting CFC  emissions.

     Consequently,  the current  focus of research resources should be to assess
consequences  of possible policy  choices.  What  are the effects of differences
in policies,   timing,  and force?  If we  delay  policy action, will  it be more
difficult  in the future?   Delay would  not provide  the  needed stimulus  for
research  into  substitutes or  reduced  emission  technologies.    What are  the
opportunity costs of deferring  control?  To address these policy  questions and
deliver  information  to a  rational policy process,  we  need  to focus  our
research resources  in the next  year  on  the articulation of our uncertainty and
its explicit  incorporation into policy  analysis.  What are the consequences of
being wrong  and what is the least  risky alternative?  The  real challenge is
the design of  policies  to limit CFC emissions  with  associated monitoring and
enforcement components that  will protect the environment but allow all nations
to share in  the benefits of  these chemicals.

     This  report  presents results of  scientific analyses of  various effects
associated  with  changes   in  stratospheric  ozone.    This  document  and  the
conference on  which it reports is one  part  of a process that  started  in the
fall  of  1985.    In approaching  the  question  of whether  the  protection of
stratospheric ozone requires the  control  of  CFCs and/or other ozone-depleting
agents, the U.S.  EPA has  divided  the  problem into a consideration of economic
and technical  issues associated with  emissions,  and  the resulting atmospheric
and health effects.   The process  is  designed  to  provide  the  best  possible
information on potential ozone-depleting substances,  their quantities and uses
now and in the  future,  and the  consequences  associated with  depletion.   In
organizing this massive amount  of information, economists divide  it  into the
broad  categories  of benefits and  costs.   Using this  framework,  then,  we can
establish a correspondence between the various conferences and reports and the
information that  economists believe is useful for policy analysis.

     In the fall of  1985,  there  was  a World Meteorological Organization (WHO)
Climate Conference convened "to  provide governments  around the world with the
best  scientific   information   currently   available  on   whether  any  human
activities represent a  substantial  threat to the ozone  layer" (Watson 1986).
The executive summary from the conference  report succinctly concludes that:

          Given  what we   know  about  the  ozone and  trace  gas-
          chemistry-climate problems  we should  recognize  that we
          are conducting one giant experiment on a global scale by
          increasing  the   concentrations  of  trace   gases  in  the
          atmosphere    without    knowing    the    environmental
          consequences.   (Watson 1986)

     The  next  set  of  workshops  attempted  to delimit  the  size  of  this
experiment by  assessing current and  expected  future  demands for CFCs.   The
effects and efficiencies of existing  control  policies  were also reviewed.   In
our  benefit-cost  language, this  information would  enable us to assess  the
costs of any control action.   The  international component  of this process was
concluded in a May 1986  meeting in Rome.  While there were disagreements among
the  delegations  about  the  precise   future  for  CFC  production,  use,  and
emission, there was a consensus  that  annual  growth  in  CFC  demand would likely
be above zero but less than five percent.

     This report represents an  articulation of our  best understanding of the
benefits associated  with   any  decision concerning  the management of strato-
spheric ozone.   These effects are  viewed as benefits because they  represent
the  consequences  that would be  avoided by pursuing a particular alternative
course of action.  Clearly, the range of alternatives includes the possibility
of doing nothing.

     This  paper  highlights  the  process by  establishing  a  correspondence
between seemingly diverse  activities;  summarizes the information presented in
this report within the  context of  that process; and previews the next step in
the  process,  which will  consider  general control  strategies.   The  paper is
divided  into  several sections  which  focus  on  the  general  methodology,  the
estimation of  costs associated  with  control action,  the evaluation  of  the
associated benefits, and  the  implications of this  preliminary assessment  for
further research.


Attributes of the Problem

     Before elaborating on  the  economist's world view of  policy  analysis, it
is important to briefly consider several  salient aspects of the stratospheric
ozone  depletion  problem.     One   of  the   most   interesting,   yet  vexing,
characteristics of the problem is its global nature.  This  attribute is impor-
tant  if  we think  of the   problem  as one  of  managing a global commons.   Mo
individual,  firm,   or   nation  can   capture  the  benefits   from   reducing
emissions.    We  cannot  exclude  others from  enjoying the benefits  from  our
emissions reductions even  if they do not contribute.  Consequently, no indivi-
dual  actor  bears the full consequences  of  their decisions  to emit  or  not.
This condition creates an  imperative for concerted global action.

     The lengths of lags  in the atmospheric  system  due to the very long resi-
dence  times  of  these   chemicals  result  in  two  additional  attributes  of
concern.   The  first  is  irreversibility:    if the  earth's  systems  are  being
severely perturbed, what  will the new equilibria be  like?  Can we ever regain
the old?  Will the simultaneous changes  wrought by  ozone depletion and climate
change  irrevocably alter the  biosphere?   Further,  what are  the  implications
for future  generations?  These long  lags  necessitate long planning horizons
and drive us headlong into the question  of  intergenerational equity.  What are
the responsibilities  of  the  present  to the future?   Mishan and  Page (1982)
have argued that we should adopt the  view  that:

          the means for  future well-being  (should)  be  at least as
          good  as  our own,  thus focusing attention  on  the  future
          condition  of   the   resource   base  and   its   ultimate
          renewability  and  the  portfolio  of  catastrophic  risks
          that are passed from one generation to another.

     As  a  current custodian  of  this  planet,  I  believe  that   this is  an
appropriate vantage  from which to review  the  information  presented  in  this

General Framework

     As  indicated  in  the introduction,  the general  framework of  organization
for  this  paper draws  from  the   policy  process  in  which  we  are  engaged.
Economists  view this  process  as  one  of  assessing the benefits  and  costs
associated with alternative courses of  action.   In  each  case, it  is important
to evaluate  the effects  of a policy with respect  to a  particular baseline.
This necessity  is  imposed upon  us because we  wish  to  ascertain which changes
can be  ascribed to the  action of the policy and because the  problem involves
chemical substances with extremely long  lifetimes.

Costs.   In the case  of stratospheric ozone,  costs  will be  the  net economic
effects  on CFC  producers,  consumers,  and  the  economy  as  a  whole.   These
effects  are  expected  to fall  under  several  categories:  surplus  losses,
efficiency losses, nonrecoverable  resource cost, safety  cost associated  with
alternatives,   and   administrative   costs    (Yarrow   1986).      Of   these,
administrative  and  transition costs   can  be  effectively  assumed  to  be  zero.
Administrative  cost  should be  small  because  the  number of  CFC  producers is
small.   Transition cost would be expected to  be a function  of the timing of
the policy action.   If  the  policy   intent  and its  implementation  date  are
announced well  in advance of  the proposed  action,  firms  producing CFCs should
have the opportunity to  alter  their resource  allocations.  A commonly employed
measure  of  economic  welfare   is the  sum of economic surpluses.   Graphically,
using  linear  supply  and  demand functions  we can illustrate  several of these
concepts (Figure 1).   In this  simplified example, consumers' surplus is repre-
sented  by  the  area  under the  demand curve  DD1 above the equilibrium  price
p*.  Because  the supply curve is  depicted  as a horizontal line and producers'
surplus  is  normally  the area  below the equilibrium  price level but above the
supply  curve,  producers' surplus  is zero.   For any  proposed policy  we  can
compute  the change in economic welfare as  the sum of consumers' and producers'

Price  _,
         Figure 1.  Conceptual Illustration Of Surplus Losses

     To more  effectively  describe the approach  pursued in this  paper,  it is
necessary  to  consider  the   specific  range  of  policies  to  be  evaluated.
Policies considered for the control of CFCs have either involved the  prohibi-
tion of specific  uses  or  limits on the total amounts  that  could  be produced.
Each policy  type  has  its adherents and  detractors.   For  expository purposes
only, we evaluate  a  production capacity  limit set  at  1980  production levels.
Because our intent is  to  summarize  the large  quantity  of existing research on
this topic into a useful integrated framework, the choice of policy for analy-
sis was conditioned on the available literature  (Palmer and Quinn  1980).   Such
a production capacity  limit can be  represented  as a vertical  line segment LL'
set at the existing 1980 production level.

     As demand for CFCs  continues to  grow, we would expect competition  among
users for  the  privilege  of using some of  the limited  quantity  of CFCs.   This
competition  would   likely  take   the   form  of  increased   prices  for  the
chemicals.   Such  intensified  competition is expected to result  in the alloca-
tion of  this  limited supply  to highly valued uses  only.  Thus,  early in the
adjustment process we  would  expect to see most  aerosol applications of CFCs
disappear  due  to their substantially  increased  cost.   Figure 1  illustrates
this process.  Demand shifts  to dd' but the quantity available cannot increase
beyond the 1980 production level LL'.   Depending on the feasibility of substi-
tution in  both technical  and  economic terms,  CFC users will  either switch to
alternatives or  compete for  the  limited  supplies.   The resulting  change in
economic surplus is measured as the shaded area  p*p**abc in Figure 1.

Benefits.    Noneconomists are  often  confused   by   the convention   that  the
benefits associated with any  environmental improvement are the damages avoided
as  a result  of  the  policy  action.    We  compute  the  production,  use,  and
emissions  of  CFCs  expected  in  the  future  in  the  absence  of any  policy
action.  These emission trajectories can  then  be transformed into  estimates of
ozone depletion over time through atmospheric  models.  The depletions are then
translated into the  effects that  are addressed  in the  various volumes of this
report.  Similarly,  a  trajectory of emissions associated with  the imposition
of  a production  capacity limit  can be  calculated along  with  its  expected
effects.   In  real  time,  we  would expect  the  time path  of emissions  to be
altered in response to the effects of policy as  indicated in the discussion of
cost.  Consequently,  it is important to evaluate  the behavioral adjustments of
the  economic  actors  prior to estimating  the  impacts  of those  decisions upon
the  atmosphere and  environment.    Thus we have  a "without  policy" set of
effects and  a "with policy"  set;  the differences  between  them represent the
effects avoided.   These are the benefits resulting from the policy action.


     Irrespective of whether  we actually  perform the benefit-cost computation
and whether discounting is applied and in what degree,   it is useful to identi-
fy the intertemporal distribution of benefits and cost.  There are a number of
theoretical and ethical issues  associated  with  the application of discounting
techniques across  long time scales, and  so it is important to take the inter-
generational  view alluded  to  earlier.    The practical  implication  of such
vantage is that most effects  will be left  in the  physical units of occurrence.

     As  indicated in  the  preceding section,  this  type of  analysis  requires
time  dependent  scenarios  describing  the evolution   of   the  atmosphere in


response  to  emissions.   This analysis  utilized Scenarios  A  and C  from the
time-dependent, two-dimensional  model presented by Stordal  and Isaksen  (this
volume).   Scenario  A  assumes that  emissions  of  chlorocarbons will  be held
constant at  1980  levels.   The specific  chlorocarbon  emissions considered are
methyl  chloride,  carbon tetrachloride,  methyl  chloroform,  F-11,  F-12,  F-22,
and F-113.  Detailed emission  rates  for  the individual constituents are  given
in Table  2 of Stordal  and Isaksen.   Scenario A is  assumed to approximate the
effect of imposing  a production  limit at  1980  levels.   This is an assumption
since there is not a strict equivalence between production and emissions.  For
example,  production  could  increase  over  time while  emissions  were  held
constant.  This could  be accomplished by reducing  prompt emissions,  improving
recovery  and  recycling.    However,   this  approximation  is  quite useful  in
allowing us  to simulate the effects  of  the  imposition of a production limit.
The specific  assumptions and  effects associated with Scenario A,  the  "with
action" alternative, are presented in Table  1.

     Extension  of  the status quo,   i.e.,  continued   chlorocarbon  production
growth,  is represented by Scenario  C.   Of  the four  time paths  presented by
Stordal and  Isaksen, this  is the most pessimistic  in that it projects a high
rate of  sustained  annual  growth  in emissions.   While emissions do not neces-
sarily  correspond  to production, this scenario  was specifically  derived from
the late market maturation and new market  development assumptions of Quinn et
al (1980).  These yield approximately a 4% annual rate of growth in production
which was translated  into  a 2.1% annual increase  in  the rate of emissions of
individual chemicals.   An  immediate  question arises  as  to the realism of the
production growth  rate.   From  the  UNEP  review of  CFC  projections  studies
presented  in  Rome, 4% is  an  upper bound within  the  range of  estimates
produced.  Thus,  the results  presented  in  this  paper  can be construed to be a
conservative upper  limit.  It  is conservative in the  sense that no account is
taken  of the  potential for  demand   increases  among  developing  countries or
among Eastern  Bloc nations.    This  "without action"  scenario  is  described in
Table 2.

     From the  UNEP  workshop  on demand and  controls held in  Rome  in May  1986,
several facts  clearly  emerged.   First,  chlorofluorocarbons have highly desir-
able physical  and  chemical  properties,  which  account for the wide  array of
beneficial applications  in  which they are found.   These characteristics have
resulted in substantial growth in demand and use.   Despite the slump in  sales
induced  by  controls on aerosol  applications   in some countries  in  the late
1970s  and a  worldwide  recession,   production   levels  have  increased  beyond
previous  peaks.   Projections  of future CFC demand  produced by  a  variety of
alternative methodologies  and authors all indicated  that demand  would  grow.
Major areas  of uncertainty were  identified  as  future population  growth, the
level  of economic  activity  in  the  world,  and the  growth  of sales  in the
developed countries.   From that  session it is  clear that in  the  absence of
policy action, CFC  use will  continue  to  grow in the future.  While the papers
presented provided a wide range of possible  future CFC levels, it is important
to emphasize that:

           Table 1.   Global Average Ozone Depletion With Constant
                     Emissions at 1980 Levels
constant 100
" 100
11 100
" 100
11 100
1 .40*
Source:  Stordal and Isaksen (this volume).
                Table 2.  Global Average Ozone Depletion With
                          Continued Production Growth


                   EMISSIONS INDEX
(growth of 4.0* per year)
    "                   155
    "                   227
    "                   299
    "                   421
                                                       OZONE DEPLETION


Source:  Stordal and Isaksen (this volume).

          Under no circumstances  do zero growth assumptions appear
          to  be  a  reasonable basis   for  calculating  long  term
          emissions profiles.  Moreover, growth rates in production
          and emissions of  these  chemicals  will probably be higher
          in  the   next  few  decades than  in  the  distant  future.
          Holding  other  things constant, these  findings  will tend
          to  increase  expected ozone depletion  relative  to models
          with  unrealistic   steady  state  emissions at  historical
          levels.   (Quinn et al. 1980)

     Further, it is important to remember that the results produced by Stordal
and  Isaksen  were  spatially,  as  well as   temporally,   distributed.    In
particular,   their  estimates   indicate   that  ozone   depletion  will   be
latitudinally  distributed  with  increased  depletion  at high  latitudes.   To
illustrate  the significance  of  these  spatial  differences,  Tables  3  and  4
present the ozone depletion effects associated with Scenarios A and C,  respec-
tively, for 60°N latitude.   Even  with  emissions held constant at 1980 levels,
depletion  is  projected to  reach 8%  by 2025.    If production  and  emissions
continue to  grow,  8% depletion would  be attained at 60°N by  2000,  rising to
M% in 2025.  Consequently,  the global  average depletion trajectories employed
in this analysis are conservative with  respect to the extreme experiences that
could result.


     As previously  indicated, the  social  cost of  any  control  action  can be
expected  to  fall  into  several  categories;   but  we  only  measure  surplus
losses.    The costs  of  controlling  emissions  through  the  imposition of  a
production capacity limit set at  1980  levels  are taken  from a study conducted
by Rand  (Palmer and Quinn  1981).   That study  considered  each of  the major
market outlets for chlorocarbon inputs  with a focus on  understanding both the
availability  and  ease  of substitution.   In  generic terms,  the alternatives
evaluated ranged from reduced input use from changes in product formulation or
production technology, improved recycling and recovery,  and  switching  to non-
CFC inputs.  For each major CFC user these  alternatives were arrayed according
to cost so  that substitutions would occur in  synchrony  with increases in the
price of CFCs.  These responses were then aggregated for each chemical and for
the industry as a whole.

     There are  three basic  areas  of uncertainty  in  this  analysis:   the speed
of user  responses,  the rate  at  which  new alternatives  would  be commercially
provided, and the precise mechanism for  market clearing.   These uncertainties
are accommodated by assessing the impact of these rates upon costs for various
assumed rates.   While the  Rand  study  evaluated  seven  alternative scenarios,
only two are  considered in  this paper.   The extremes of the Rand results were
selected to represent pessimistic and  optimistic extremes that should bracket
the actual  costs.   Their  results  were updated  to  1984 terms  to facilitate
comparison.   The  annual  cost of  a 1980  production cap  for the years  1981
through  1990  are  displayed  in Table  5.   Using  these  annual  estimates  and
converting  to  present  value  terms   at an   11%  discount  rate  results  in
cumulative  control  costs  for  this  10-year  period  ranging  from approximately
$1.9 to $4.5  billion.   From Table  5,  it is clear that costs  would  rise over
time as the demand for CFCs continues to grow, as users continue to bid up the
prices of the increasingly scarce chemicals, and as the easy substitutions are


       Table 3.   Ozone Depletion With Constant Emission at 1980 Levels
                 at 60° North Latitude
Source:  Stordal and Isaksen (this volume).
         Table 4.  Ozone Depletion With Continued Production Growth
                   at 60° North Latitude
(growth of 4.0$ per year)
1990 "
2000 "
2010 "
2025 "


Source:  Stordal and Isaksen (this volume).

                        Table 5.  Time Stream of Costs
   YEAR                          OPTIMISTIC                  PESSIMISTIC

                          (millions of 1984 dollars)
1981                                27.4                         49.2
1982                                65.5                        152.7
1093                               135.7                        317.2
1984                               181.0                        578.1
1985                               285.2                        873.2
1986                               425.9                       1427.6
1987                               571.4                       1453.2
1988                               664.8                       1480.9
1989                               926.7                       1511.0
1990                              1543.5                       1558.2
Present Value                     1865.3                       4543.0
Source:  Palmer and Quinn.
achieved.   In  sum,  for  this  period  average  annual control  cost  would  be
expected to range between $187 and $454 million.


     This report presents papers documenting effects upon human health and the
environment from  two  problems that are  linked through CFCs.   Increased UV-B
radiation throughout the  biosphere is expected as the total  ozone column is
reduced  through  chain  reactions  in the  stratosphere catalyzed  by chlorine.
CFCs are also radiatively  active, absorbing  infrared radiation  and contri-
buting to  the climate  warming due  to  the greenhouse effect.   The direct and
indirect effects  induced by  these changes  are the  damages  associated  with
emissions and  thus  are the benefits  that would result from controlling such

     Historically, concern with ozone depletion has centered on increased UV-B
radiation and  its direct implications  for human  health.    The  effects docu-
mented include  increased  skin cancer,  immune  system suppression,  and ocular
damage.  This paper focuses  on  the implications of  skin  cancer  research for
changes  in  human  health when coupled  with alternative trajectories  of ozone
depletion.   Epidemiological  studies conducted by the National Cancer Institute
formed  the  basis  for  our  effects  assessment  (Fears  and  Scotto  1983).
Estimates of  ozone depletion  can be translated into  increases in UV-B radia-
tion through  application of  an enhancement ratio.   In  this paper  that was
assumed  to  be a conservative   1:2  (National  Academy  of  Sciences  1984).


Coupling the  alternative  ozone depletion scenarios previously  discussed with
the enhancement ratio, trajectories of UV-B  increase  can  be predicted.   Using
the results  from  the Fears and  Scotto  (1983) power  model  (more conservative
than  its  exponential counterpart), UV-B  increases  are mapped  into  increased
incidences of skin cancer.  These  UV-B  increases  are  applied to the base case
of  400,000  cases per  year and  adjusted according  to the  projections of  a
Caucasian population presented in Rundel and Nachtwey  (1978).

     Time and policy  dependent  rates  and  numbers  of skin  cancers for  the
United States are shown  in Tables 6 and 7.  Table 6  indicates  that  even with
emissions held constant at 1980  levels,  the number of new  skin cancers would
rise  to  approximately  142,000 per  year  by 2025.    A  more dramatic  change is
associated with growth  in chlorocarbon emissions as  the number  of  new cases
for that year is estimated to be  roughly 256,000 (Table 7).   Our interest here
is  the difference between these trajectories,  because that is  the  number of
cases  of  skin   cancer  that   could  be  saved  by  constraining  chlorocarbon
emissions and thus  slowing the rate of  ozone depletion.   For  example, about
114,000 cases of skin cancer could be prevented in 2025 by controlling chloro-
carbon emissions  growth.    Table  8  presents this  component of  the benefits
trajectory as the differences over time  for each of the  projection  years.
However,   these  results are more  easily understood  graphically.   Figure  2
depicts  the  number   of  new cases  of skin  cancer expected  if  emissions  are
controlled (with action)  and  if  emissions continue to grow  (without action).
The shaded area  between these two trajectories is  the total number of cases
that would be avoided  as  a result  of policy action.   In this  analysis, 1.66
million skin  cancers could be prevented  if emissions are  controlled  at 1980
levels.  This result is all the more significant given  recent  reports citing
the lack of  progress in managing cancer and  the  need to  emphasize prevention
(Associated   Press   1986).     From  this   vantage,   controlling  chlorocarbon
emissions would  seem to be a judicious  investment.

     At a more mundane  level,  other direct  effects of increased UV-B include
materials damages, in particular impacts on  polymers  and  elastomers.  Andrady
and  Horst   (Volume   2)   identify  effects  ranging   from   discoloration   to
brittleness.   The specific damage depends upon  the  type of plastic  and  its
application.   For example, in the case  of rigid PVC  used for home siding, UV-B
causes  discoloration  which  results  in  an  increased  rate  of  rejection.
Increased  PVC brittleness due  to  UV-B  radiation   in  pipelines   can  cause
premature failure and more  serious  consequences.  In  each case, these effects
can be ameliorated  by changing the  plastic's formulation by adding titanium
dioxide.   Andrady and Horst (Volume 2)  project the amounts of plastics expect-
ed  to  be used in the future,  the  increased UV-B  this stock would be exposed
to, and the costs of ameliorating the effects of such exposure.

     Their results   are  interpolated for  the  two  ozone depletion  scenarios
analyzed in this paper.   Table 6  contains  the damage  estimates to U.S. PVC if
emissions are controlled.   At a maximum, these would be  $8 million annually.
In  the absence of control,  however,  this  figure would increase to $35 million
per year  (Table  7).   Consequently, the maximum annual  benefit  in the form of
foregone PVC damages would be $2? million (Table 8).   Using a planning horizon
extending to  2075 and  applying  a discount  rate  of  10 percent they estimate
that  the present value of PVC damages  would range  from $10  to $27 million
depending upon the strength of PVC demand.  While  these damages are  relatively
small  compared  with  the  costs estimated  for the production  cap policy, they
illustrate the pervasive nature of the effects of ozone depletion.


               280  -i
               240  -
               200  -
160  -
               120  -
                80 -
                40 -
            Figure 2.  The Incidence of Skin Cancers in the United States
                       Associated with Alternative Ozone Depletion Scenarios

          Table 6.  Effects of Global Average Ozone Depletion With
                    Constant Emissions at 1980 Levels



1 . 40$
2 . 00$
[% increase]
5.6/ 4.2
1.21 5.4
8.0/ 6.0
8.0/ 6.0
[new cases/yr]

[millions of
1984 $/year]

Based upon a 4x increase in the incidence of skin cancer for males and a
3x increase in incidence for females per unit depletion of ozone (Fears
and Scotto 1983).

Increase from a base level of 400,000 cases/yr.

Interpreted from Andrady and Horst (Volume 2).
          Table 7.  Effects of Global Average Ozone Depletion With
                    Continued Production Growth




[$ increase]
1.21 5.4
12. O/ 9.0
17. 6/ 13.2
30. O/ 22.5
[new cases/yr]

[millions of
1984 $/year]

Based upon a 4x increase in the incidence of skin cancer for males and a
3x increase in incidence for females per unit depletion of ozone (Fears
and Scotto).

Increase from a base level of 400,000 cases/yr.

Interpreted from Andrady and Horst (Volume 2).

          Table 8.  Difference Between Constant 1980 Emissions and
                    Continued Production Growth Scenarios
[% increase]
[new cases/yr]
[millions of
   1984 $/year]
 1.67   1.2
 4.87   3.6
 9.67   7.2
22.07  16.0
     A less  obvious  category of  effects  include those  generated  as indirect
consequences.   For  example, Whitten  and  Grey  (Volume  2) present research
indicating that  increased levels  of UV-B  irradiance  due  to  ozone depletion
would stimulate increased formation  of  photochemical  oxidants.   Their results
indicate  that ozone  concentrations  would peak  earlier  in  the  day and  at
greater distances from the source, with the result that a larger human popula-
tion would be exposed and at risk.   EPA is currently assessing existing ozone
standards  in  light  of  new  research  concerning  general  health  effects.
Further, as many as 24 cities across the nation will not be in compliance with
existing standards by  the December  30,  1987  deadline.   Increasing levels of
UV-B will  definitely exacerbate  this  problem and  increase substantially the
magnitude of investment required to maintain urban air quality.

     These effects are not limited  to urban areas.   Some of the most signifi-
cant damages  to  date associated with increasing ambient ozone concentrations
have accrued  in  agricultural areas.   As  described in  the  National Crop Loss
Assessment (Adams and McCarl 1985; Garcia et al.  1986),  yield reductions and
the  consequent  impacts upon profitability have  cost  between  $0.7  and $2.0
billion annually.  Increases in the ambient  concentrations of ozone in rural
areas  obviously  would  increase  the extent  of  these  damages.    Whitten and
Grey's results  were extended  to rural areas in an  attempt  to  identify the
possible magnitude of this impact.   The increases  for ground  ozone concentra-
tions in urban areas associated  with different levels of ozone depletion from
their research were applied to existing average rural ambient ozone concentra-
tions.   The  resulting concentrations were evaluated for yield  reductions by
using the dose-response relations developed by Heck et al.  These yield reduc-
tions for the "with action" and "without action"  scenarios are shown in Tables
9 and 10.

      Table 9.  Indirect Effects of Global Average Ozone Depletion
                With Constant Emissions at 1980 Levels
% yield reduction
                                      soybeans    cotton
2 . 00*
less than
1.0* for all

   Additional damages from dose response relations given in Heck et al.
      Table 10.  Indirect Effects of Global Average Ozone Depletion
                 With Continued Production Growth




* yield reduction
corn wheat
less than 1 .0*
less than 1 .0*
0.45* 0.45*
1.50* 1.40*

for all crops
for all crops


   Additional damages from dose response relations given in Heck et al.

     The  results   under  controlled  emissions  show   very  slight  effects
throughout the period  of analysis.   The  increasing  depletion  associated with
continued emissions growth would cause reductions in soybean and cotton yields
by the year 2010.  These would be more pronounced by 2025.  The differences in
percentage yield  reductions  that  could  be  attributed  to  policy  action  are
displayed in  Table 11.   No  attempt  has been  made  to  translate  these yield
reductions into  an economic  dimension  because  of the complex  interaction of
behavioral  and  market  responses.     The   four   crops   examined  are  major
commodities in  the world's  food and  fiber system,  as   illustrated  in Table
12.   When aggregated  across  the  immense acreages and values  shown for these
crops, even small changes would produce significant damages.

     This paper has been modest  in  the costs that have been estimated.  Table
13  provides  an  itemization  of  the  effects  identified  at  the  EPA/UNEP
conference which  have not  been  included.   Despite  the  lack of  attention to
this wider catalogue of  effects, this  analysis  suggests  that we have answered
the question of whether  to act.  Even  those effects  about which we have great
uncertainty represent  incentives  to  action.   Perhaps  the most  significant
category  of  such  uncertain  impacts  is  that  described  as  environmental
effects.  The changes stimulated by  the direct and  indirect  effects of both
increases  in  UV-B  radiation and  climate  change  will  be  most profound  for
natural ecosystems and their  inhabitants.   Consider,  for example,  the spatial
shifts  in agricultural production  regions  likely  to occur  from  the combined
stresses of increased UV-B and climate change.   Soil moisture changes, direct
yield  reductions,  changes  in the  temperature regimen,  and enhanced  carbon
dioxide concentrations will  alter  comparative  advantage.   Production regions
are  likely  to  encroach  on  currently  marginal   lands   devoted  to  wildlife
habitat, forestry,  and open  space.   The magnitude of these changes have the
potential to  eradicate the substantial  investment  that  we  have  committed to
maintaining species diversity and wildlands.

     A caveat is in order:   In valuing future outcomes,  we are conditioned by
current prices  that reflect  the existing pattern of resource and commodity
availabilities.   As population grows in the future and as increasing resources
are   devoted   to  maintaining   the   economic  well-being   of   those  future
generations,   the  availability  of  natural  resources and  wildlands  can  be
expected to decline.   This  change  in  physical  supply can occur at exactly the
time when demands for the nonconsumptive uses of such resources will be higher
if current tasks and preferences endure.   Consequently,  they may be much more
precious to future generations.

     Our real dilemma  in the evaluation  of these alternative  futures  is the
lack of a criterion by which to gauge them.   In  particular, policy formation
is hampered  by  the lack of  an  acceptable ozone depletion  management target
framed in terms of public health and environmental systems damage.  Until that
target is set and  its  implications  for allowable  atmospheric chlorine concen-
trations, CFC growth, and use levels are calculated,  we  don't really know the
potential implications of widely divergent alternative  futures.   Further, in
the absence of a policy  target for  a maximum allowable ozone depletion, it is
difficult  to   assess   the  importance   of  the  uncertainties  underlying  the
alternative  long-term  forecasts  for   CFCs.    Recall   that scientists  have
cautioned that  "once  the predicted  change  in  climatology or  ozone depletion
diverges  substantially from  current  conditions,  the validity of  the models
becomes suspect."  Consequently,  what is needed is a decision concerning


          Table 11.   Difference Between Constant 1980 Emissions and
                     Continued Production Growth Scenarios
  YEAR             DEPLETION                CROP LOSS
                                          % yield reduction
                                      corn     wheat     soybeans    cotton

1980              0.0$                 000           0
1990              0.4$                 -
2000              1.2$
2010              2.4$                0.45      0.45       1.20       1.80
2025              5.5$                1.50      1.40       2.80       3.30
           Table 12.  Principal Crop Commodity Statistics for 1984
CROP           WORLD ACREAGE           U.S.  ACREAGE          U.S. CROP VALUE
	— (thousands of acres)		(billions 1984$)

Cotton             85,067                10,380                    3.66
Corn              312,299                80,434                   20.50
Soybeans          130,646                67,735                   11.40
Wheat             570,795                79,213                    8.70
Source:   Agricultural  Statistics  1985,   U.S.  Government  Printing  Office,
                      Washington, D.C., 1986.

Table  13.   Effects of Ozone Depletion and Climate Change

       Direct Effects of UV-B Radiation
            immune system suppression
            ocular damage

       Indirect Effects of UV-B Radiation
            increased human exposure to  increased air  pollutants

       Direct Effects of Climate Change
            temperature thresholds and increased mortality

       Indirect Effects of Climate Change
            flooding severity from sea level  rise
            increased frequency of severe  weather


       Direct Effect of UV-B Radiation
            crop yield reductions

       Indirect Effects of UV-B Radiation
            changes in competitive balance between  plant species

       Direct Effects of Climate Change
            soil moisture changes
            changes in crop suitability

       Indirect Effects of Climate Change
            comparative advantage and spatial location of production
            equipment complements
            chemical input use
            pest timing and impact
            property values and financial  stress


       Direct Effects of UV-B Radiation
            phytoplankton mortality
            larval sensitivity

       Indirect Effects of UV-B Radiation
            changes in species diversity
            endangered species
            reductions in commercial catch


       Indirect Effects of Climate Change
            coastal property inundation
            salt water intrusion


            species diversity
            endangered species
            pesticide, herbicide, trace  element loadings
            habitat reductions
            wild and scenic rivers
            wetlands disappearance

acceptable public health and environmental impacts associated with alternative
levels of stratospheric ozone depletion.


     It  has   become   extremely  fashionable  to  highlight   the  scientific
uncertainties  surrounding  any  particular  environmental  problem  requiring
public decisions  to  be made.   As most of us  know,  uncertainty  is  a  fact of
life.  Its new found vogue in the policy arena belies this normal human exper-
ience.   For example,  when  making a corporate move,  new housing  is normally
acquired.   Such  moves  typically occur  across steep  housing gradients.   In
today's markets of rapidly appreciating prices and high  demand,  the potential
buyer must be prepared  to make  quick decisions or be foreclosed on the possi-
bility of making a decision.   Business  is not much different except the stakes
are higher.   Opportunities  present themselves and  then vanish.   To continue
the analogy,  the business  of  running  a global  environment  has  even  higher
stakes.  We, too, are presented with opportunities for action-decision making,
which may never be offered again.  To defer action when the opportunity arises
could mean  its irrevocable  loss.   So, there  is  uncertainty  in  the economic
future of CFCs, uncertainty  in the atmospheric science and effects, and uncer-
tainty in the opportunities  for decision making.

     The best  way to  represent this  problem  of  uncertainty  is to view  any
policy action  that might  be taken  as insurance.   After  all,  the point to
insurance is  the opportunity to manage  uncertainty  and its  attendant  risks.
Limiting the  emissions of CFCs  can  be viewed as an  insurance policy  against
the multifarious potential future damages that are explored by the authors of
this report.   What  prudent  person among us has not  purchased insurance?  The
consequences of ozone depletion alone should cause us to act,  but the possibi-
lity of preventing some of the  global  warming  is  further reason.  Research is
still  needed  but  it  should  be  focused.    The  following  are  some specific

     When one considers the  potential consequences of climate change and ozone
modification,   the relative  lack of  research on  ecological  consequences is
striking.   Not only is the  general  level of funding  for  research in  ecology
extremely limited, but  there  is  virtually no  focus on the consequences of the
anticipated changes for natural ecosystems.

     Other  ecological   stresses have  been  explored  in  some  detail.    For
example, we have  begun  to understand the  response of freshwater ecosystems to
acid deposition, and a rudimentary and  fragmented  picture is developing of the
effects  of  air  pollution  stresses on  mid-latitude   forests.    These  efforts
provide a good  launching  point for ecological studies of  the consequences of
climate  change  and  ozone  modification.   There  are  two  reasons for   this
argument.   First, adequate  baseline data  for the relevant systems are being
developed.   Second,  as Bruce  (this  volume)  points  out,  the  climate/ozone
reduction  stress will  act   synergistically  on these systems  with the  acid
deposition-air pollution stresses that are the focus  of current studies.

     A  few  examples  illustrate this  point.   Mid-latitude forests that  are
stressed by air  pollution in  Central Europe  experienced  rather rapid declines
after  1979.    Some  have suggested  that  this  decline occurred  as  a  natural
climate stress  (successive warm periods)  was superimposed  on a  long term air


pollution stress.   It  is possible  that  this large,  rapid change  in forest
status is a precursor to  the  effects of climate  change, acting in combination
with  air  pollution.   One  might suppose  that  this  synergistic  effect  will
accelerate the effect of climate change alone in altering forest distributions
on  the  earth's  surface.   This  particular  synergy  obviously links  with the
Whitten study;  the  enhanced  ultraviolet  flux  from  ozone modification  will
increase oxidant concentrations, which act as phytotoxins.

     From a policy perspective,  the  interactive  effects of these stresses are
very important.  By accelerating the consequences of climate change, they will
put  that  stress  in  a current  context, and signal  large  ecological changes
sooner  than  otherwise  expected.    Timing  becomes  especially  important  if
extreme temperature periods are ecologically significant because the number of
such periods  exceeding a  given temperature  will  increase  as  climate changes.
As an unfortunate consequence, heavily  polluted  areas  become  a testing ground
for  the  effects  of climate  change.   We  would urge  a  mid-latitude forest
"watch" as an ecological early warning area.

     Both  the nitrogen  and  the sulfur  cycles  will  be  affected by global
warming,   providing  another  set  of  synergisms and  critical  ecosystems.
Clearly,  we have massively disturbed these  cycles, and will do so more in the
future, without  knowing  either  the  implications  or  where  we  are  headed.
Increases in  global nitrogen and  sulfur deposition  and the  effects  of  such
deposition  are  already measured.    The interaction  of these increases  with
climate change and  enhanced ultraviolet flux should be  an important research

     These  interactions  also  underscore the linked  nature of climate change
and  other  air pollutant  stresses  both at  their sources  and at  the effects
levels.    They  emphasize the  efficacy  of  strategies  to  limit trace  gas
emissions in general and  add  a  note  of urgency to such action.  In a way, the
effects are already upon  us.   We should slow  them down and let  our knowledge
expand faster than the rate at which we're changing the earth.


     Before identifying the priority policy research  needs that this summariz-
ing  exercise  leaves us  with,  it  is  important  to make  a few  observations.
First, most  discussions  concerning   ozone depletion  focus on  CFCs 11  and 12.
It should be  clear by now that  we must include a wider range of ozone-deplet-
ing  chemicals  within  our  regulatory  framework.     The  remainder  of  the
commercially produced CFCs  plus the  halons  should  be  included.   The bromine
contained within  the  halons  is  many times  more  effective  at depleting ozone
than the chlorine of CFCs.  If only  F-11 and F-12 are managed and other chemi-
cals  are  produced  for  a  mature  market,   we  will  not  attain  the  desired
objective of protecting the ozone layer.  Furthermore,  consider  the implemen-
tation  of  the  production   capacity   limit  within   the  European  Economic
Community:  In that setting,  the cap is set to the sum of the nameplate capa-
cities for  the various existing  production  facilities.   There are a number of
swing facilities that primarily produce F-11 and F-12 but for a portion of the
year also produce F-22.   If the F-11 and  F-12 production capacity limits were
to become binding, then these facilities would be dedicated to F-11 and F-12,
and a F-22  facility could be  constructed, adding more  CFC emissions.  We need
a consistent and reliable method of  adding  these different chemical compounds


together to ensure  that  their  individual depleting effects are  accounted for
in any policy.

     We need  to  identify  an  internationally  and environmentally  acceptable
limit to stratospheric  ozone depletion.   Is  the  intent to limit  the  rate of
growth of CFCs  or is it to  constrain the  ultimate quantity discharged to the
atmosphere?   While  we may perceive  ourselves to be  currently  constrained in
our ability to either articulate or  implement  any policy objective,  we should
actively assess  the consequences of the  possible choices.  What  if CFCs are
not best conceived  as  a mature or saturated market?  What are  the effects of
differences in  policies,  timing and force?   If we delay  policy  action,  will
policy action in the future be  more difficult?  Delaying would not provide the
needed  stimulus   for   research   into  substitutes   or  reduced   emission
technologies.    What  are  the  opportunity  costs of  deferring  control?    To
address these policy  questions and  deliver  information to a rational policy
process, we  need to  focus  our  research  resources  in  the next year  on the
articulation  of  our uncertainty  and its  explicit  incorporation  into policy
analysis.    What are  the  consequences   of  being wrong  and what  is  the least
risky alternative?  The real challenge  is  the design  of policies to limit CFC
emissions   with   associated  monitoring   and enforcement  components  that  will
protect the environment and  yet allow all  nations to  share in the benefits of
the use of these chemicals.

     It  is  important  to  distinguish  between  irreversible  decisions  and
irreversible  actions.   In the case of stratospheric ozone,  an  irreversible
action  is  depletion and its attendant  effects which cannot  be  undone except
with the passage of very long periods of time.  By contrast,  policy actions to
limit the growth  in emissions  of CFCs and  other ozone-depleting substances do
not represent irreversible decisions.   Any agreement can  be  reconsidered and
changed.  The critical point is the irreversibility of the consequences, which
represent  an  imperative  quite  distinct  from whether  policy may  be reasses'sed
and changed.  This  places the  policy question  before us strictly in the form
of trading off  short-term economic advantages for an environment that can be
sustained  in the long run.


Adams,  Richard   M.  and Bruce  A.  McCarl.  1985.   Assessing  the  benefits of
     alternative  ozone   standards   on   agriculture:  The  role  of  response
     information. Journal of Environmental Economics and Management.

Associated  Press.  1986.   U.S. Seen as Losing Fight With Cancer.   New York
     Times, May8, 1986.

Camm,  F.,  et al.  1986.   Social cost of technical control options  to reduce
     emissions of potential  ozone  depleters in the United States: An update.
     Note N-2440-EPA, Rand, Santa Monica, 75 pp.

Fears,  Thomas R.  and Joseph  Scotto.    1983.    Estimating increases  in  skin
     cancer  morbidity  due  to  increases  in  ultraviolet  radiation  exposure.
     Cancer Investigation. 1(2):119-26.

Garcia, Philip, et  al.  1986.   Measuring the  benefits  of environmental change
     using  a duality  approach:  The  case of ozone and  Illinois  cash  grain
     farms.  Journal of Environmental Economics and Management.  13:69-80.

Heck,  W.  W., et al.  A  reassessment of crop  loss from ozone.   Environmental
     Science and Technology.  17:573A-581A.

Hoffman, John  S.,  John B.  Wells,  and James  G.  Titus.    1985.   Future global
     warming and  sea level rise.    In Iceland  Coastal and  River  Symposium:
     Proceedings,  ed. G. Sigbjarnarson.  Reykjavic,  Iceland:   National Energy

Kripke,  Margaret  L.  and  Warwick   L.  Morison.  1985.    Modulation  of immune
     function  by  UV  radiation.    The Journal  of  Investigative  Dermatology.

Mishan, Ezra J. and  Talbot Page.    1982.    The  Methodology  of  benefit-cost
     analysis with particular reference to the CFC  Problem,   in The economics
     of  managing   chlorofluorocarbons:    Stratospheric   ozone  and   climate
     issues.  eds.  John H.   Cumberland, James R.  Hibbs, and Irving Hoch,
     114-56.    Washington  B.C.:     Resources  for  the   Future,  Johns  Hopkins
     University Press.

National  Academy  of  Sciences.    1984    Causes  andeffectsof  changes  in
     stratospheric  ozone:  Update  1983,  Washington, B.C.:    National  Academy

Palmer,  Adele   R.   and Timothy  H.   Quinn.    Economic  impact  assessment  of  a
     chlorofluorocarbon production  cap.  Note N-1656-EPA,  Rand, Santa Monica,
     CA, 1981,  85  pp.

Palmer,  Adele  R.,   et   al.   1980.    Economic  implications  of  regulating
     chlorofluoro-carbon emissions  from nonaerosol applications.  Report
     R-2524-EPA, 292 pp.

Quinn,   Timothy  H.,  et  al.   1986.   Projected use, emissions, and  banks  of
     potential  ozone-depleting  substances.    Note  N-2282-EPA,  Rand,  Santa
     Monica, January,  140 pp.

Rundel,  R.   B.  and  B.  S.  Nachtwey.   1978.    Skin  cancer  and  ultraviolet
     radiation.  Photochemistry and Photobiology.   28:345-56.

Watson, Robert, 1986.   Braft assessment of  ozone layer modification   United
     Nations Environment Programme, Nairobi,  January, 133 pp.

Yarrow, George.   1986.   The economic  consequences of a  production  capacity
     limitation for chlorofluorocarbons.   Paper submitted to the UNEP  Workshpp
     on Demand and Control, May,  22 pp.