Playing the numbers with climate model accuracy

Published July 1, 2000

Those who proudly claim their climate models duplicate reality, provided [insert your favorite caveat here], have been irritated by one particularly nasty thorn. (And no, we’re not talking about Environment & Climate News.) When the atmosphere is filled with carbon dioxide (CO2), every model warms the Earth’s surface, usually in a pretty linear manner (i.e., a given slug of CO2 generates a proportional slug of temperature rise).

But how do we account for the decline in global temperatures from the end of World War II until the late 1970s? The models simply have a devil of a time producing cooling with all that CO2 in the mix.

A recent (unintentionally) humorous article in Science reported that a couple of federal climatologists have found the solution. And it really couldn’t be simpler. All you need to do is run your model many times, compare the output with the observed temperature history, and find one that matches it! Gee, why hadn’t they thought of that earlier?

Seriously. They ran their model five times. Each run had gradual increases in greenhouse gases (GHG) and cooling sulfates based on values from 1865 to the present. According to the authors, “in one of the five GHG-plus-sulfate integrations, the time series of global mean surface air temperature provides a remarkable match to the observed record, including the global warmings of both the early (1925-1944) and latter (1978 to the present) parts of the century. Further, the simulated spatial pattern of warming in the early 20th century is broadly similar to the observed pattern of warming.” They suggest that the big unknown is the natural variability of climate.

Getting only one out five right would flunk you out of math class. Yet it’s enough to get you published in the world’s most prestigious scientific journal under the title, “Simulation of Early 20th Century Global Warming.”

We’ve crafted a few more apt titles:

“Model Fails to Reproduce Reality 80 Percent of the Time”

“Four Out of Five Model Runs Are Remarkably Mismatched to Observations”

Or how about just a change in punctuation? “Simulation of Early 20th-Century Global Warming?”

The problem with global climate modeling groups is some actually believe their own P.R. Yes, it’s possible to get a model to reproduce anything you choose merely by tweaking a few parameters and running it enough times. But the model that reproduces the temperature history screws up precipitation, and the model that gets rainfall correct can’t generate the proper wind or pressure fields.

The reason is actually quite plain: We don’t understand the physics of the atmosphere well enough to model climate change.

That is the grim reality that at least four out of five climate models chose to ignore.


Delworth, T.L., and T.R. Knutson, 2000. Simulation of early 20th-century global warming. Science, 287, 2245-2250.