The Baltimore Sun
(Copyright 2002 @ The Baltimore Sun
WASHINGTON - On July 25, at a hearing of
the House Oversight and Investigations Subcommittee (the same
folks who grill corporate executives), the nation found out
how little real science there is about global warming.
The hearing was prompted by the discovery that federal
scientists were using computer models that they knew could not
replicate U.S. temperatures. They appeared in two landmark
documents that have served as the basis for very expensive and
intrusive energy legislation.
What came out of the
hearing has people asking whether the same problems affecting
Enron, WorldCom, Global Crossing, etc., are now troubling
Puffy federal documents create
consequences. In this case, look no further than a recent
California law requiring the first statewide reductions on
global warming emissions from cars. It cites many findings
from the U.S. National Assessment of global warming, one of
the documents that provoked the House Oversight hearing.
Science operates by a hard and fast ethic. Theory must
conform to reality and be tested by reality.
case of climate science, our "theories" are huge computer
models that project various amounts of warming for the next
In my review of the Assessment, I
discovered two very disturbing facts.
team considered several computer models, but chose the two
that predicted the most extreme changes in temperature and
rainfall over the United States. That's prima facie evidence
for some type of bias, which isn't surprising. The team was
vetted through four Clinton administration committees, one of
which was headed by Al Gore.
Worse, these models
couldn't beat a table of random numbers, or two dice on a
craps table, when it came to predicting U.S. temperatures.
The Assessment team members were required to publicly
comment on the science reviews, and swept this criticism under
the rug. Behind the scenes they were much more fearful, and
replicated my experiment. They found out that the computer
models they were using couldn't even simulate 25-year
temperature averages over the United States as the greenhouse
effect changed in the last 100 years.
How, then, could
these models be used to assess climate change in the next 100
It gets worse. The temperature and
precipitation data these models spit out are then input to
other sectors of the U.S. economy, such as agriculture,
forestry and water supplies.
In science, random numbers
are garbage, and that's what went in. Refuse in science is
what we call "transitive." Start with it and you end with it.
Tom Karl, who heads the National Climatic
Data Center in Asheville, N.C., and co-chaired the production
of the Assessment, was the lucky person who had to come to
Washington to defend these shenanigans.
was about as painful as you would expect when a noted
scientist has to defend something so wrong.
defense of the federal establishment was that it was not
making "predictions" of future U.S. climate with these models.
Instead they were "plausible projections." Are the most
extreme estimates of U.S. temperature and rainfall changes
Can someone explain to a reporter, seeing
these dire results spit out by extremist computer models, that
there's a lick of difference between a "prediction" and a
Does anyone believe that the political
impact of either, if based upon bad models, isn't the same,
i.e., bad legislation?
This rings the same bell that
Bill Clinton did when he said, "That depends on what the
definition of the word `is' is." Predictions, projections,
forecasts - they're all the same to any reporter, or, for that
matter, to any professor. And when they're based upon computer
models that can't beat dice dancing in Atlantic City, they're
This story isn't going away. It surfaced
recently on MSNBC's Hardball, and it will bubble up every time
someone comes up with a new law on climate change in the
It's no accident that all of this wound
up in front of the Oversight and Investigations Subcommittee.
It recently targeted corporations that cook books and hide
things from shareholders.
Forecasting something as
important as future U.S. climate based upon models that do not
work is as deceptive as stating assets that a company does not
have. In this case, it is the federal science establishment
and we, citizens of the United States, who are the investors.
It's not just CEOs and CFOs who inflate results to
jack up their currency. The corruption has now spread to
Patrick J. Michaels is senior fellow
in environmental studies at the Cato Institute and an adviser
to the American Legislative Exchange Council in Washington,