Data sets of 1° latitude by 1° longitude CO2
emissions in units of thousand metric tons of
carbon (C) per year from anthropogenic sources
have been
produced for
1950, 1960, 1970,
1980, and 1990.
Detailed
geographic
information on
CO2 emissions can
be critical in
understanding the
pattern of the
atmospheric and
biospheric
response to these
emissions. Global,
regional, and
national annual
estimates for 1950
through 1992 were
published
previously. Those national annual CO2 emission
estimates were based on statistics about
fossil-fuel burning, cement manufacturing, and
gas flaring in oil fields, as well as energy
production, consumption, and trade data, using
the methods of Marland and Rotty (1984). The
national annual estimates were combined with
gridded 1° data on political units and 1984
human population to create the new gridded CO2
emission data sets. The same population
distribution was used for each of the years as
proxy for the emission distribution within each
country. The implied assumption for that procedure was that per capita energy use and fuel
mixes are uniform over a political unit. The
consequence of
this first-order
procedure is that
the spatial changes
observed over time
are solely due to
changes in national
energy consumption and nation-based fuel mix.
Increases in
emissions over
time are apparent
for most areas; for
example, from
1980 to 1990, a
63% increase in
CO2 emissions
(based on 1980
emissions)
occurred in mainland China and a 95% increase
in India. However, actual decreases from 1980
to 1990 occurred in Western Europe: 30% in
Sweden, 27% in France, and 23% in Belgium.
Latitudinal summations of emissions show a
slow southerly shift (in the Northern Hemisphere) in the bulk of emissions over time. The
large increases, from 1950 to 1990, in China's
and India's contributions to anthropogenic CO2
emissions compared with those by the United
States are, for example, very apparent at the
latitudinal band around 25.5°N.
