The Why Files The Why Files -- whyfiles.org

The fact of global warming

POSTED 24 FEB 2005
Testing, testing
BGases belch out of factories, polluting the air.oth the concern and the skepticism about global warming has focused on computer climate models. We need climate models because we don't have an extra planet for experimental purposes.

Every time you burn fossil fuel, you produce carbon dioxide, the primary greenhouse gas. Photo: EPA

Critics have scoffed that the models couldn't even reproduce the actual climate from historic data, much less predict the future. Are they right -- are computerized climate models worth any hard-disk space at all? To find out, scientists are testing models against a flood of climate data from ground stations and satellites.

One technique starts by identifying the "fingerprint" -- the specific climatic impact -- of each natural and artificial factor, like changes in greenhouse gases, the sun, or reflective aerosols from volcanoes, that can affect climate. Using math of a complexity we couldn't begin to grasp, Ben Santer and others at the Program for Climate Model Diagnosis and Intercomparison at Lawrence Livermore National Laboratory scrutinized climate records for the fingerprints.

Graph of modeled and observed temperatures after volcanic eruptions.
When this climate model started from actual 1980 conditions, it accurately predicted the temperature of the lower atmosphere. During the period, two huge volcanoes spewed out aerosols that reflected sunlight back to space, cooling the planet. The model accounted for both the arrival and eventual removal of the aerosols. Graph: Ben Santer, Lawrence Livermore National Laboratory.

The fingerprint, he says, is "an educated guess, a physically based estimate about how climate might change in response to combined human and natural effects. Then you look for that pattern in the observed climate."

Fingerprint experts
The results point to a human influence on many aspects of climate: average temperatures at several levels, atmospheric pressure, and the amount of sea ice. And who, you may ask, did the fingerprints finger as the guilty party in climatic change? You and me. Unless you add the rising levels of greenhouse gases, the climate models don't match reality.Color coded map shows where temps have risen since 1951.

Map compares average surface temperatures in 2004 to a baseline of 1951 to 1980. Yellow and red indicate warming has occurred. Graph: NASA

The tests also showed -- as you might expect from the work of a gang of brainy nerds -- that the models are getting better. "There has been improvement as a whole in the endeavor of climate modeling," says Santer.

Early models may have relied on untested theories about the ocean-atmosphere system, but no longer prove, Santer says. "It bugs me when people say you use these complex models, but never compare them with reality. My place has as its reason for being to evaluate all the world's climate models."

The results still finger greenhouse gas accumulation as the key culprit in climate change, Santer adds. "The question is how much we influence climate, not whether we do. We're not just looking at surface temperatures alone. The climate system is telling us an internally consistent story. Natural explanations cannot explain the change we have seen over the last century."

A water planet warms
Most of the attention about global warming has focused on the atmosphere. That's where we live, and it's where the greenhouse effect occurs. But most of the planet is water, and water plays a huge role in climate. Is water warming?

The question may be simple, but the answer is not, because temperature records from the ocean are scarce. And the first climate modelers found it tough enough to deal with the atmosphere, so they ignored the oceans. Newer, "coupled" models treat atmosphere and oceans as one system, which reflects an intimate, real-world relationship: Oceans absorb heat and carbon dioxide from the atmosphere, while releasing water vapor into the air and distributing heat toward the poles.

Graphs show that natural climate variability does not explain warming in the ocean.
Can climate models predict ocean temps? Researchers at Scripps Institute of Oceanography tested models with (green) and without (blue) the rising level of greenhouse gases. Only the model runs with man-made levels of greenhouse gases matched actual ocean temperatures. "The agreement between what is observed and what is expected to arise from anthropogenic forcing is excellent in all ocean basins," says Tim Barnett, who directed this work. Notice how "standardized signal strength" moves deeper in the ocean with time. The Why Files tried mightily to understand what that means. We assume it has something to do with temperature, but our efforts to find out were met with... jargon about "standard deviations." Courtesy Tim Barnett

That's an el-simplifico description of an interlocked system that is hideously complex and difficult to model. But evidence that climate models realistically describe oceans is emerging from studies showing ocean warming. In a talk to the AAAS, Tim Barnett, a research marine physicist in the climate research division at Scripps Oceanographic Institute in San Diego, reported on a comparison of climate models and ocean data from the last 40 years. Barnett says he was "stunned" to see that the computer models reproduced the actual warming in every ocean. "The statistical significance of these results is far too strong to be merely dismissed and should wipe out much of the uncertainty about the reality of global warming."

What does ice tell us about global warming?

back more


Megan Anderson, project assistant; Terry Devitt, editor; S.V. Medaris, designer/illustrator; David Tenenbaum, feature writer; Amy Toburen, content development executive

©2023, University of Wisconsin, Board of Regents.