News from
Office of Communications
Stanhope Hall, Princeton, New Jersey 08544-5264
Telephone 609-258-3601; Fax 609-258-1301

Contact: Steven Schultz (609) 258-3601

June 15, 2000

Recent climate changes probably not a fluke

Princeton, N.J. -- Recent changes in the frequency and intensity of the global climate event El Niño may be a trend and not just a statistical fluke, according to Princeton scientists.

Over the past 20 years, scientists have observed longer intervals between major El Niño episodes compared to those in the 1960's and the 1970's, as well as stronger intensity of the events. The century's two most powerful El Niño episodes, in 1982 and 1997, caused flooding and severe weather in the U.S. It has not been clear, however, if those changes are simply statistical "noise," a random clustering of long, strong cycles, or if they result from other long-term changes in global climate.

In the June 16 issue of Science, geoscientists Alexey Fedorov and George Philander review the evidence for both possibilities. The authors suggest that the changes, particularly the lengthening of the cycle, are a trend that stems from a long-term rise in ocean temperature in the Eastern Pacific and an associated weakening of trade winds in the region. That rise in ocean temperature may, in turn, be a statistical blip or the result of some natural decades-long oscillation in ocean temperature. But it may be, as the authors point out, the result of human-driven global warming.

El Niño, which often brings mild, rainy winters in some part of U.S, occurs when east-to-west trade winds relax, allowing warm water to accumulate in the Eastern Pacific. Its opposite, La Niña, tends to cause colder, dry winters and southern droughts. It occurs when strong trade winds drive warm water to the Western Pacific. The cycle between the two is usually called the Southern oscillation.

For the past 100 years, the system oscillated between warm (El Niño) and cold (La Niña) conditions with a natural frequency of roughly 3-4 years. In the past 20 years, however, the period has been closer to five years, with 4 major warm events in the 80's and the 90's.

At the same time, scientists have observed that the baseline sea-surface temperature in the Eastern Pacific - the average of the El Niño highs and the La Niña lows - has risen by about one degree in the last 20 years, while the trade winds have weakened. The average intensity of El Niño has also increased.

Ordinarily, making sense of these observations would be a matter of statistics. Is the trend established and pronounced enough to be more than a fluke? Twenty years, however, is too short a period to provide enough data for a positive answer.

So Philander and Fedorov turned to physics. They used a computer model that, based on the current understanding of the physical processes occurring in the tropical Pacific, simulates the winds and ocean temperature gradients that drive the Southern oscillation. When they put in the conditions of 20 years ago along with a rising sea-surface temperature and weaker trades, the model predicted an increase of the time between warm events from three years to five, as well as larger amplitudes of the events.

"We suggest that this is really what is happening," says Fedorov. "One may need to look beyond the simple statistics. Knowing the underlying physical processes can be most helpful, even if our conclusions can not be checked statistically because of the brevity of the observational data."

The scientists acknowledge that their model is highly simplified, leaving out several factors that could affect the oscillation. For example, it does not explain the asymmetrical nature of the oscillation, that La Niña tends to last much longer that El Niño.

Fedorov also noted that their model predicts only average cycles and can not say how long a particular cycle will be. Nonetheless, Fedorov and Philander believe that their approach may give climate forecasters the ability to make general predictions about the period of the Southern oscillation. The model suggests, for example, that the next one is more likely to occur in the year 2002 than during this year, for instance. That information could be beneficial both to developed countries including the US, and to nations with agriculture-based economies, which are often affected by El Niño-driven weather.