Questions Three for the Witches of Warming

S.Fred Singer is the head of the Science and Enviromental Policy Project (SEPP) and one of the principle scientific critics of anthropogenic global warming. He responds to an argument in World Economics (sorry, subscriber only) in the current issue:


Published in World Economics Oct. 2006


The exchange (World Economics, April 2006)
between Sir Nicholas Stern and Ian Byatt and
co-authors has concentrated on climate
science. I would like to add some arguments
based on logic, requiring little knowledge of climate science.

To do this I will pose three fundamental
questions: 1] Is there evidence for or against an
appreciable human contribution to current climate
warming? 2] Would a warmer climate be better or
worse than the present one? And 3] Realistically
speaking, can we do something about climate? Is
it possible to influence the climate by policy actions in an effective way?

1. To address the first question we need to
recognize that the climate has always been
changing. It is either warming or cooling – on
time scales ranging from decades to millions of
years -- though on average it has not changed
very much since the beginning of
time. Continents shift, mountain chains grow and
decay, ice ages come and go; but there has been
remarkable stability overall -- even with huge
variations in the atmospheric levels of
greenhouse gases like carbon dioxide. Since
1979, when weather satellite temperature
measurements began, there has been a slight
warming trend; its exact magnitude is in
dispute. How can we tell whether this warming is
due to human influences, such as the rise in
atmospheric greenhouse gases, or whether it is
simply another natural fluctuation?

It's no use asking the thermometers; they cannot
talk. The melting of glaciers and ice sheets,
the rise in sea level, severe storms, floods,
droughts -- all of these are interesting, to be
sure, but really irrelevant to our
question. They may well be connected to
a warmer climate -- or maybe not -- but they
cannot tell us what causes the warming. Nor can
a vote among scientists settle this scientific
issue. Nor can we argue that the rough
correlation with the rising level of greenhouse
gases proves a cause-effect relationship. World
climate cooled between 1940 and 1975 while
greenhouse-gas levels
rose sharply. Correlations have also existed in
the past; but in every case (for example, in the
sudden warming at the end of ice ages) the
increase in temperature preceded the increase in
carbon oxide. Clearly, carbon dioxide was not
the fundamental cause of the warming. In other
words, correlation is not proof a truth that is often forgotten.

So what is left? All working scientists agree
that one should compare the observed patterns of
warming with the patterns calculated from
greenhouse models: distinctive geographic and
altitude distributions of the temperature
trends. A just issued report, using the best
available data and models, gives an interesting
result: the patterns do not agree. [see
FTNT**]. Note that, logically, agreement by
itself cannot prove that the warming is due to
human causes; it only makes it plausible. But
when we find an instance of significant
disagreement, then we can argue that the
influence of human effects is minor compared to
the natural fluctuations of the climate.

**FTNT: The report to resolve this comparison,
published on May 2, 2006, is based on research
funded by the US government under the Climate
Change Science Program (CCSP). See

The crucial graph (Figure 5.4G) clearly shows the
disagreement in the sensitive tropical region
between observed temperature trends and those
calculated from greenhouse models. Note,
however, that the Executive Summary of this CCSP
report claims that there is "clear evidence" for
human global warming, disregarding the contrary
result within the report. I shall not address
this unexplained discrepancy here.
*********************************************************************

A separate but important question is: How
reliable are forecasts of climate models, which
indicate a major warming as a result of
increasing greenhouse gases -- a "climate
sensitivity" (CS) of 1.5 to 4.5 degrees C for a
doubling of carbon dioxide? Clearly, if we take
the disagreement with the data seriously, then
the observed patterns are consistent with a much
lower value of CS and therefore only a minor future warming. [see FTNT***]

***FTNT: Logically, however, one cannot exclude
the possibility that climate models are
completely in error and that the current warming
is anthropogenic after all. This assumption
leads then to the following calculation: If one
assigns all of the observed 0.6 degC temperature
increase of the 20th century to an anthropogenic
increase in GH gases (which together have gone
about 50% towards a doubling), then the
additional forcing from the next 50% will only
add a little additional warming. This is so
because the calculated temperature increases only
as the logarithm of CO2 concentration.
***************************************

This discrepancy between the observed and
calculated values of CS stems from the fact that
the models give very imperfect representations of
the real atmosphere. It is well known, for
example, that all models have a serious problem
in representing clouds. Some of the best work on
this problem has been done at the Hadley Centre,
first by John Mitchell, and more recently by
Murphy et al and Stainforth et al. at
Oxford. The microphysics of clouds is extremely
complex and has to be represented by
somewhat arbitrarily chosen
parameters. Stainforth has shown [Nature, 27
January 2005] that plausible assumptions about
cloud parameters can lead to temperature
increases of up to 11.5 degC for a doubling of
carbon dioxide. To some people these large
numbers suggest climate catastrophes; however,
this result may only show that models are
extremely sensitive to the assumptions put into
them and therefore cannot be used in a reliable
way to make predictions about future climate.

2. The second question is clearly in the realm of
economics. In the ongoing debate it seems to be
assumed, generally without much analysis, that a
warmer climate presents a "danger" or a "threat"
– or a similar prejudicial term, implying serious
consequences for economy, human health, ecology,
etc. In a proper analysis, various economic
groups disagree about the magnitude and even
the sign of the consequences. The IPCC report
(1995), for example, featured five publications,
all showing negative economic consequences. I
found, however, that in spite of similar total
damage numbers, they disagreed violently in
their numbers for different economic
sectors. On the other hand, a group headed by
Robert Mendelsohn (Yale University) showed
positive consequences; the GNP of the United
States would increase for assumed temperature
increases of 2 degC [Cambridge University Press, 1999].

William Nordhaus, a leading environmental
economist, asserts geographic evidence that
economic output decreases with increasing
temperature. In a paper in the Proceedings of
the National Academy of Sciences (7 March 2006)
he shows that economic activity decreases
strongly in the tropics. But many other factors
enter into the equation besides climate. For a
striking example one only needs to look at the
country of Zimbabwe before and after Mugabe.

Without going further into technical details, one
can look at historical evidence. We know from
voluminous records that human existence was good
and much more comfortable during the Medieval
Warm Period (ca. 1100 AD) than during the
following Little Ice Age, when crops failed,
people starved, and disease was rampant; life
then was nasty, brutish, and short. Another way
of looking at this question is to ask if things
are worse in 2006 than in 1956 when it was
colder. It becomes obvious that climate effects
are minor compared to everything else that
happened in the last 50 years; but that is
exactly the point: Technological progress and the
mobilization of capital far outweigh any climate
factor that one can think of in promoting
prosperity. It would be foolhardy to claim that
a slightly warmer climate, of say one degC in a
hundred years, will have serious effects on the
way we live, on the economy, human health, or
anything else. Indeed, since most agree that a
colder climate would damage the economy, one
might ask: What is the probability that the
present climate just happens to be the Panglossian optimum?

3. The third topic is perhaps the easiest to deal
with. What to do about climate change? People
often talk about "stabilizing the climate;" what
they really mean is stabilizing the
concentration of greenhouse gases in the
atmosphere. But even that is a daunting
task. But we know what it would take: As
published by the IPCC (and here everyone agrees),
we would have to reduce emissions worldwide by
between 60 to 80 percent in order to stabilize
the level of carbon dioxide in the
atmosphere. It is also agreed that the Kyoto
Protocol is a puny effort; at the very best it
only delays the increase in greenhouse gas levels
by about six years. The effect on climate, all
agree, would be minute: a calculated reduction of
temperature of only one-twentieth of a
degree. Such a change cannot even be measured by ordinary thermometers.

I noticed that the Stern report will include a
non-marginal dynamic cost-benefit analysis. We
are not told, however, what discount rate will be
chosen -- a crucial piece of information. I
am of course familiar with marginal cost-benefit
analysis, which is fairly standard. I've even
published some of the early work on dynamic
cost-benefit analysis in connection with
discovering the optimal path toward controlling
pollution from automobiles. (Here we consider
that the fleet of automobiles changes rather
slowly so that introducing new technologies that
reduce pollution has only slight effect on total
emissions -- and ambient air quality -- to begin with.)

But the real question for the
Stern Report: Since it is unlikely that the
current warming has much of a human component and
since it is unlikely that something substantial
can be done about reducing the growth in
greenhouse gases -- what is the point
to a cost-benefit analysis, when in fact it
seems most probable that a warmer climate
would produce positive benefits instead of damages?
***********************

Atmospheric physicist S. Fred Singer is professor
emeritus of environmental sciences at the
University of Virginia. He served as first
director of the US Weather Satellite Service.