Geoscience Reference
In-Depth Information
For academic meteorologists, the key to understanding the atmosphere
lay in numerical weather prediction. The key to numerical weather predic-
tion, in turn, lay in computers. In order to describe the dynamic processes
of atmospheric motion, scientists had to solve long series of nonlinear
equations, and they had to solve them quickly. During the First World
War, British meteorologist Lewis Fry Richardson surmised that it would
take about sixty-four thousand human computers solving equations con-
tinuously to keep up with the earth's weather as it unfolded— and that's
to say nothing of predicting future weather.
30
During the Second World
War, however, two electrical engineers working for the U.S. Army Ord-
nance Corps, Presper Eckert and John Mauchly, devised a machine that
addressed this computing problem, the Electronic Numerical Integrator
and Computer, or ENIAC.
31
Recognized as the first digital computers,
ENIAC and its successor EDVAC (Electronic Discrete Variable Auto-
matic Computer) were designed to compute the differential equations of
ballistic weapons-firing tables more quickly, but they could also solve the
differential and partial differential equations that described meteorologi-
cal phenomena.
Following the suggestion of physicist Vladimir Zworykin (a pioneer
in television tubes), meteorologists like Weather Bureau chief Francis
Reichelderfer and his eventual successor Harry Wexler soon began to push
for a computer-generated numerical weather forecast.
32
The result, after
more than five years of research and development, was the interdisciplin-
ary Joint Numerical Weather Prediction Unit. In March of 1950, this group
used mathematician John von Neumann's new and improved computing
machine, along with the expertise of European and American meteorolo-
gists, physicists, and mathematicians, to produce the world's first 12- and
24-hour computer-assisted numerical weather forecast.
33
In the mid-1950s, as the IGY approached, scientists and their spon-
sors at the Weather Bureau and the National Science Foundation began
to consider using models as a way to go beyond forecasting to a deeper
understanding of the geophysical processes of the atmosphere. With 12-
and 24-hour forecasts continuing to improve, meteorologists like Joseph
Smagorinsky and Jule Charney at Princeton, as well as Victor Starr of MIT,
turned their attention to general circulation models (GCMs), mathematical
representations not just of the weather but of the movement of the atmo-
sphere as a whole.
34
In 1955, Smagorinsky and Wexler launched a spin-off
