3.More and Less Uncertainty
What do these findings really mean for the human future? How likely are
greenhouse gases to continue to rise? How much? One, 5 or 8 degrees? And
how soon? Fifty, 100, or 200 years? And what would a temperature rise of
thismagnitude mean for humans and ecosystems? Beyond global averages,
how would the effects be experienced in different regions and ecosystems?
These are contentious questions, but most fractious of all are the policy ques-
tions: Do we know enough to act, and if so, what should be done? And
when? Scientists can't answer many of these questions with certainty
because there are still too many missing pieces of the puzzle, Here are some
of the more important ones.
While the correlation between greenhouse gas concentration and temperature
. fluctuation works well fur geological history. itworks less well for shorter time
spans, particularly in the current period for which we have the best data.
Greenholl~egases have increased rather ste,ldily since the turn of the century,
but temperature increases have not: Most warming occurred in the 19205 and
the 1970s and. in fact. duringthe 19•10s and 19505 the world cooled slightly'
Given the current theory, this should not have happened and has been com-
pared to a "murder mystery in which the whereabouts of principal suspects are
unknown"(Schneider, 19QOb: 34-35).
There are other important missing suspects. Among these is the missing carbon
sink. About 45 percent of the total anthropogenic CO~ emissions since prein-
dustrial times are unaccounted for ("whereabo'utsunknown"). Scientists
believe that the ocean helps moderate tropospheric temperature by removing
about 29 percent of the excess CO" we pump into the atmosphere, but they
don't know if they can absorb more. If the oceans warm significantly, they may
no longer be able to act as great buffer systems (Miller, 1998: 371).
Beyond their role as carbon sinks, the role of the oceans in the warming process
is largely unknown with current methodologies, but it is likely to be large. The
oceans store most of the planet's heat and CO2and have deep circulation that
is not well known or modeled. Their enormous mass will act as a thermal
sponge slowing any initial increase in global warming while the oceans them-
selves heat up, but the magnitude of this increase in temperature wiII depend
on ocean circulation, which may itself change as the earth warms (Schneider,
1990c: 31).
Similarly, the inability of GCMs to factor in effect of vegetation and forests
means ignoring their effect on ground surface reflectiveness (or albedo), their
function as carbon sinks, or the significance of their release of water vapor and
cloud formation.
Likewise, the interactions between temperature change and cloud formation
and the resulting feed backs are unpredictable. Will heating of the atmosphere
create more or fewer clouds? And would more clouds trap more heat at the
earth's surface or reflect more solar radiation into space?
Most scientists believe that warming the atmosphere would melt polar ice
caps, causing sea levels to rise. But that effect is not beyond question. It has also
been hypothesized that warming the air would accelerate air circulation and
polar precipitation, snowfall, and the formation of polar icepacks. In fact, in
1991 one research team documented a significant accumulation of ice in eastern
Antarctica since 1960 (Sullivan, 1991).
The interaction between temperature change and the photosynthetic processes
of plants and the resulting feedback mechanisms is unclear. Warmer tempera-
turesare known to accelerate plant growth and hence the absorption of CO2
from the atmosphere. Would that be sufficient to dampen global warming at
some point? Or would greater cloud cover block enough sunlight to retard
plant growth even in warmer conditions? Flip a coin!
Perversely, the human production of pollution, smog, and soot may act to
absorb some of the radiation that would warm the atmosphere. Whether these
effects would be large enough to be significant no one knows.
In sum, human-cuoironnicntilllt:mt:li(lll~arc((llI/p/"x .uu! includ« I/IIl11!1 /I,'l//i'/car
rciatiauship« alldfeed/JOckII/edwl/i:;/II~_ Given the current imperfect state of
knowledge of these complex system connections, our ability to predict th?
timingand magnitude of global warming is impaired, .indin particular thl'
more concrete changes in windpatterus.j-ainfall, .iud humidity that would dif.
ferentiallyeffect regions and ecosystems (Schneider. 19)Oa; National Academv
of Sciences, 1991: 88-94)
With all of these uncertainties and unknowns, you may be wondering
how one can have any confidence in the threat of future global warming, [f
so, you have some respectable scientific company But just enumerating the
unknowns and anomalies understates the degree of consensus in the clirna.
tologicalscientific community about the global warming and its probable
consequences. That has have been summarized by the National Research
Council and other policy groups (National Research Council, 1987; Silver
and DeFries, 1990; Krause et al., 1992: 28-29; National Academy of Sciences,
1991: 94)_ Here are some conclusions, arranged from the Virtually certain to
the uncertain. Virtually certain means that there is nearly unanimous agree-
ment within the scientific community that a given climatic effect will OCCur.
Very probable means greater than about a 90 percent (9 out of 10) chance, and
probable implies more than about a 67 percent (2 out of 3) chance. Uncertain
refers to hypothesized effects but for which there is a lack of appropriate modelingor observational evidence.
Large stratospheric cooling (virtually certain).Upper atmosphere destruction of
ozone by chlorine and other gases will markedly increase the loss of infrared
radiative-heat in the upper stratosphere.
Global mean surface warming (very probable). For a doubling of atmospheric CO2
(or its equivalent from all greenhouse gases), which is expected sometime
during the next century, the long-term global mean surface warming is pre-
dicted to be in the range of 1.5-5.0 (centigrade)". Currently the most widely
used models predict a narrower range of 3-5.5°, but when a broader range of
possible feedback effects is considered the average warming from doubled CO2
could be as high as 6.3-80or more (Dickinson, 1986; Lashof, 1989). The most
important uncertainty arises from the difficulties in modeling the feedback
effects of clouds, and the actual rate of warming over the next century will be
governed by the slowly responding parts of the climate system, such as the
oceans and glacial ice.
Global mean precipitation increase (very probable), Increased heating of the
earth's surface will lead to increasedevaporation and subsequently to greater
global mean precipitation. Nonetheless, precipitation may well decrease in
many individual regions, which would be hotter and dryer.
Northern polar surface warming (very probable), Winter surface temperatures in
polar regions would be much warmer than they are now (three times the global
mean warming), with a greater fraction of open water and thinner sea ice as
well as a probable reduction in sea ice.
Northern high-latitude precipitation increase (probable). The increased poleward
penetration of warm, moist air may increase the annual average precipitation
andriver runoff in high latitudes (e.g.. such as in northern Canada: and
Scandinavia), '
Summer continental dryness/warming (probable), Soil moistures in the mid -latitude
continental interiors may decrease during summer, caused mainly by all earlier
termination of snowmelt and rainy periods and thus an earlier onset (It the
, normal spring-to-summer reduction of soil moisture,
Rise in global mean sealevel! (probable). Sea level is likely to rise asseawater expands ill response to the warmer-future climate. FM less certain is how much this will be affected by possible melting of glaciers,
Regionalvegetation changes(uncertain). Climactic changes in temperature and
precipitation must inevitably lead to long-terms change in surface vegetation,
but the exact nature of these and how they in turn might affect climate are
uncertain.
Tropical storm increases (uncertain). A warmer.wetteratmosphere has been
hypothesized to lead to more frequent and more intense tropical storms, such
as hurricanes. But this effect has not been satisfactorily addressed in the coarse-
resolution climate models because tropical disturbances are relatively small.
To grasp the enormity of these probable changes, you need to
compare them to the climate history of the earth. A global average
. warmingof 1.5° would represent a climate not experienced since the
beginning of agricultural civilization some 6,000 years ago; 3-50would
represent a climate not experienced since human beings appeared on the
earth some 2 million years ago. The last time the earth was this warm was
in the Pliocene period (some 3-5 million years ago), and more than 5°
warming would mean a climate not experienced since theEocene period
(40 million years ago), before the evolution of birds, flowering plants, and
mammals, when there were no glaciers in the Antarctic, Iceland, and
Greenland (Krause et al., 1992: 28). Furthermore, the projected rate of
warming is 15 to 40 times faster than the "natural" warmings after the
major ice ages and much faster than what most species living on the earth
today have ever had to face. Warming could far outstrip the ability of
ecosystems to adapt or migrate (Silver and De Fries, 1990: 71). A several-
degree warming over a lOO-year period would greatly exceed that natural
rates of change, pushing forests poleward by 2.5 km per year, compared
with the less than 1 km per year migration of even fast migrating tree
species (CEC 1986). The result would be a rapid dieback while new