IPCC (Intergovernmental Panel on Climate Change) future projections ar
e based on a set of emission scenarios, IS92a to f, which are used to
calculate future atmospheric concentrations of greenhouse gases. These
, in turn, are used to calculate projections of radiative forcing, and
then projections of future temperature and sea level change to the ye
ar 2100, using computer climate models. The assumptions of these 6 IPC
C emission scenarios for the years 1995 and 2000 are compared with cur
rently available information on greenhouse gas emissions, world popula
tion trends, and trends in world coal production. All of the scenarios
exaggerate one or more of these quantities. Calculations of confidenc
e limits on the net human-induced contribution of carbon dioxide to th
e atmosphere show a very high level of inaccuracy. When added to the e
ven greater uncertainties connected with assumptions on the main green
house gas, water vapour, and also on clouds, plus the uncertainties of
the computer models themselves, the current IPCC future projections o
f global temperature and sea level must be regarded as extremely unrel
iable. Fossil fuel emissions assumed by the IPCC scenarios for the yea
r 2000 are plausible for scenarios IS92a, b, c and d, but not for e an
d f. The calculated rate of increase of atmospheric carbon dioxide con
centration since 1990 assumed by the IPCC is exaggerated by 13% for al
l scenarios. The calculated rates of increase in atmospheric methane f
rom 1990 to 2000 are exaggerated by 3 to 7 times, world population inc
reases by up to 5.5%, and world coal production increases by 60 to 510
%. The rate of increase of carbon dioxide in the atmosphere has been a
lmost constant, at 0.4% a year, between 1971 and 1996, despite a 54% i
ncrease in emissions from the combustion of fossil fuels over that per
iod. Currently suggested reductions from present emission levels are t
herefore unlikely to influence carbon dioxide concentrations, or globa
l temperatures. Since all of the IS92 scenarios exaggerate one or more
current climate and economic trends, the calculated future projection
s of atmospheric greenhouse gas concentrations are thus correspondingl
y exaggerated. A more realistic set of scenarios, which would include
a mechanism for continuous updating, needs to be developed, thus scali
ng down the current values. Even if this is done, however, the accumul
ated inaccuracies inherent in the final calculations of climatic effec
ts are so great as to render them unreliable as a guide to public poli
cy.