El Nino has for centuries been a regular, if somewhat reckless, climatic event. It typically makes its presence felt around midwinter, hence a name that loosely translates as “Christ child” in the Spanish of the Peruvian fishermen who first noticed it. El Nino begins, however, some months prior to that, when trade winds in the western tropical Pacific drop or even shift, switching from blowing westward to the east. When this happens, a body of warm water that normally pools in the ocean east of Australia begins to move toward the coast of Peru. Warm air rising from the surface of this mass—which is as much as 12 F warmer than normal acts like a paddle stuck into the southern jet stream, redirecting it northward and altering weather from Australia to Canada to Africa. The warm water itself, meanwhile, is like a cap on a bottle when it hits the coast of Peru, halting the rise of cold, nutrient-rich water that typically emerges along the South American coast from deep in Pacific. That drastically affects the food chain for marine mammals, birds and fish.
Scientists have grown steadily more familiar with El Nino in the past 20 years. “There’s been a fundamental change since the 1982-1983 El Nino,” the devastating event the 1997-1998 El Nino surpassed in size, says McPhaden. “We didn’t even know that one was happening until it was almost over. In the 1997-1998 El Nino, we could tell you day by day what was happening.” The reason? Two new powerful tools—instrumented satellites and buoys—now make it as easy for scientists to watch the ocean as if it were a wading pool in their backyard.
Still, scientists were disappointed by one significant aspect of the past year’s work on El Nino: their ability to forecast it accurately. While a few computer models suggested that an El Nino would develop in 1997, none came close to predicting its scope or the speed with which it developed. Even the gold standard of forecasting turned into lead. A model developed by Mark Cane and Steven Zebiak was considered the best among El Nino forecasting models. But it didn’t read the global tea leaves correctly, forecasting an El Nino that was much later and smaller than the one that actually hit.