Will global warming increase El Niños?

Published February 1, 2001

During the height of El Niño mania in 1998, a few scientists, most prominently Kevin Trenberth of the National Center for Atmospheric Research, began pushing a theoretical El Niño-global warming linkage. Surely, people reasoned, El Niños must get worse or become more common as carbon dioxide levels increase. Mustn’t they?

It didn’t take long for some studies to appear showing strong linkages between warming patterns and El Niños, or evidence that, over the historical record, the spate of El Niños in the 1980s and 1990s couldn’t possibly have been caused by random variability.

But other analyses demonstrated that those recent El Niños paled in comparison with the whoppers that occurred hundreds of years before the carbon dioxide-spewing internal combustion engine was even conceived.

Now comes a newly published paper in Nature by the University of Colorado’s Frank Urban, Julie Cole, and Jonathan Overpeck. They seem to strengthen the link between warming and El Niños.

They extracted the El Niño signal from oxygen-18 isotope records of corals from Maiana in the tropical Pacific Ocean. Warm El Niño events raise temperatures and enhance rainfall in the region, while conditions are cooler and dry during La Niña events—differences that are reflected in the coral’s oxygen isotope signature.

Comparing the coral record and two different El Niño indices since 1950 shows a strong correspondence, suggesting that it might be possible to reconstruct El Niño conditions prior to the instrumental record. Doing so reveals that an abrupt shift appears to have taken place in 1976 toward more El Niño-like conditions.

The authors properly noted that similar changes occurred at other times in the record, and that those changes do not correlate well with atmospheric greenhouse gas levels or changes in solar output or volcanic eruptions. But they counter that the 1976 shift “has been attributed to the intensification of the hydrological cycle by anthropogenic greenhouse forcing, which has a strong regional signature in the central equatorial Pacific consistent with the delta18O shift in our record.”

At any rate, the paper leaves the impression that, over the last quarter-century, El Niños have suddenly become more common.

Contrast that implication, however, with a new paper by Matthew Collins of Britain’s Hadley Centre for Prediction and Research. Collins ran the latest, greatest version of the Hadley general circulation model (GCM), monikered HadCM3. That model purports numerous improvements over the previous incarnation—better horizontal resolution over the oceans, no need for a flux adjustment to account for wild “drifting” of model parameters, changes in the method to generate clouds in the model, and so on.

Collins examined the frequency, amplitude, and timing of El Niño/La Niña events by keeping track of changes in Pacific sea-surface temperatures. When the older Hadley Model (HadCM2) was forced with carbon dioxide levels four times the natural background (yes, I said four!), El Niños became more common, more severe, and occurred earlier (in autumn instead of winter).

But when Collins repeated the experiment using the fancy new HadCM3 model, he found no change. That’s right: Even quadrupling the level of carbon dioxide in the atmosphere relative to natural background levels generates no change in El Niños.

Should we trust the new HadCM3 any more than the previous version? Probably not, although the lack of “flux adjustments” is typically a good sign. In general, as GCMs have grown in sophistication over time, their prognostications of the rate of future warming have dropped significantly. It’s also worth remembering that the earlier HadCM2 was the less extreme of the two models used in developing the U.S. National Assessment report.

It does seem like El Niños have become fairly common over the last 25 years. But to claim, as NCAR’s Kevin Trenberth did in 1998, that there’s “got to be a connection” between this increase in El Niños and greenhouse gas requires a level of scientific “chutzpah” and precognition that has, at present, escaped the rest of the climate community.

Robert E. Davis, Ph.D., is an associate professor of environmental science at the University of Virginia.


El Niño and global warming: What’s the connection? UCAR Quarterly, 24, Winter 1997.

Collins, Matthew, 2000. Understanding uncertainties in the response of ENSO to greenhouse warming. Geophysical Research Letters, 27, 3509-3512.

Urban, F.E., J.E. Cole, and J.T. Overpeck, 2000. Influence of mean climate change on climate variability from a 155-year tropical Pacific coral record. Nature, 407, 989-993.