| The Magazine of the CALIFORNIA ACADEMY OF SCIENCES |
|
|
|
horizons El Niño's Tropical Punch I don't blame you for being weary of the media's Cassandra cries about the monster El Niño lurking of late in the Pacific. The warnings of heavy rainfall, though, did make me wary enough to get a new roof. And the prospect of parting with that much money got me thinking about how scientists predict El Niños, those Brigadoons that emerge from the tropics every few years with climatic consequences that reverberate around the globe. What techniques have been devised, and how reliable are they? By the early 1980s scientists knew about the cyclic behavior of El Niño and its sibling, La Nina, and some held out hope for prediction. At a major climate meeting in 1982, debate ensued over whether an El Niño would occur that winter. Even as one prominent expert was pooh-poohing the possibility, the smoking gun of rising sea-surface temperatures was already in place in the Pacific. Unfortunately, volcanic ash from erupting El Chichon had obscured satellites' views of the warming ocean. When the clouds cleared, climatologists collectively uttered, "Uh oh." The infamous El Niño of 1982-83 wrought thousands of deaths and $13 billion in damage worldwide. El Niño prediction was obviously high-stakes science, and researchers redoubled efforts to foresee future events. Stephen Zebiak was then an MIT graduate student, working with his advisor, Mark Cane, to devise an accurate model to explain how El Niños arise and, perhaps, to predict them. They distilled the tropical Pacific system down to a few essential ingredients, including wave motion in the upper ocean, water temperatures at the surface, and the atmospheric response. Interaction between air and sea powers the El Niño cycle. Imagine a large fan blowing the length of a swimming pool. The fan creates wind that shoves water from the near end toward the far end. At the far end, water that had been at the surface gets pushed down, while at the end closer to the fan, water wells up to replace what has moved across the pool. In the Pacific, tradewinds blowing westward along the equator cause the upwelling of cold water off South America, creating a temperature gradient of cold to warm surface water from east to west. El Niño upsets the usual climatic course when sea surface temperatures in the east begin to increase, weakening the east-west temperature gradient. The tradewinds slacken, and upwelling in the east gets stifled. The thermocline--the transition zone between warm and cold ocean layers--flattens, and more warm surface water in the west spreads eastward toward the Americas. The Cane-Zebiak model simulates two ocean layers, warm above and cold below. By feeding observed data about winds, currents, and surface temperatures into a computer, the model reconstructs the climatic cycle of events out in the Pacific. In 1985, Zebiak and Cane first put their model to the test with data from 1981. They weren't sure what to expect but found that the model did forecast the El Niño that snuck up in 1982. They made additional trial runs with data from previous El Niños, and in each case the warming event could be "seen" up to a year in advance. Buoyed by these results, Zebiak and Cane took current wind and temperature data and let the model project forward in time. That led them to forecast a warming event for 1986. Despite doubts from other climate researchers, the prediction held up, but three months later than the model anticipated. For the first time, an El Niño had been predicted successfully. Next, the Cane-Zebiak model scooped the National Weather Service by correctly predicting more than a year ahead that an El Niño expected in 1990 would not arrive until the following year. "Most of our colleagues were skeptics at the beginning of this," says Zebiak, now a researcher at Columbia University"s Lamont-Doherty Earth Observatory. But a sea change of opinion ensued during the past decade, with most confirmed skeptics converting to true believers in the emerging power of predictive models. "We're far ahead now of where we were in the early '80s," he says. At least ten models have been developed and are being tested and refined by various research groups. One model that showed particular promise in 1997 is called the Coupled Ocean Atmosphere or NCEP Model, developed by scientists at the National Centers for Environmental Prediction. Thanks to an array of buoys stationed throughout the tropical Pacific to measure ocean temperatures and currents as deep as 500 meters, the NCEP model can add accurate estimates of subsurface temperatures, an extra layer of complexity, to its simulated ocean. This model also uses patterns in surface temperature and physics equations to predict how wind patterns will change. For instance, the NCEP model saw the 1991-92 El Niño coming nine months beforehand. "Having a model in the first place was a huge advance," says climate researcher Lisa Goddard of the University of California's Scripps Institution of Oceanography. "The next step was getting this more realistic model." But the 1990s has been a befuddling decade for El Niño researchers. Both the Cane-Zebiak and NCEP models failed to predict rising temperatures that occurred in the winters of 1993 and 1994. Part of the problem stems from the fact that since 1988 the Pacific has been consistently warm, without the intervening cold periods, known as La Ninas, that typically follow El Niño. Some scientists contend that the 1990s have witnessed a single protracted El Niño, building up to what now looms in the Pacific. But a study by Goddard and colleague Nicholas Graham in the May Journal of Geophysical Research concludes that, until this year, the decade's weird weather cannot be blamed on El Niño. The strangest situation happened in 1994. Models had forecast a mild cooling for early that year, but instead a rewarming occurred and began to impact weather almost to the degree of a true El Niño. According to Goddard and Graham, this warming evolved from persistent ocean temperature anomalies unrelated to El Niño or La Nina. The predictive success of models may depend on how they account for the effects of past air-sea interactions which linger in the ocean like a memory that can be recalled. No one knows what caused the oddities earlier this decade, but all agree that this year's El Niño is not only the real thing, but probably the biggest of the century. "We knew at the beginning of this year that an El Niño was likely," says Goddard. "What we didn't know was that it was going to grow so rapidly." And no one knows for how long it will wreak havoc with our weather. If the current event behaves like the 1982 El Niño--and Goddard has run computer comparisons between Pacific conditions now and then that look alarmingly similar--both northern and southern California can expect up to twice the normal amount of rainfall this winter. This year, the NCEP model performed swimmingly and predicted a major El Niño several months before the typical signature of rising sea surface temperatures appeared. But despite conspicuous signs of this El Niño, the Cane-Zebiak model missed it altogether and, says Goddard, has been "dead in the water" this year. Zebiak agrees that his model has a problem and thinks that its clock is running too slow. He's been tinkering with it, and as of late October it may be back on track. Advances in the computer models and in sophisticated surface and satellite instruments to keep watch on the seas have raised hopes that El Niño and similar phenomena in other oceans can be understood and consistently forecast. Toward that goal, in April the National Oceanic and Atmospheric Administration (NOAA) earmarked $18 million to create the International Research Institute for Climate Prediction (IRI). With bi-coastal bases in La Jolla, California, and Palisades, New York, IRI combines scientists from Scripps and Lamont-Doherty and seeks to develop better forecasting tools, make the best predictions, and distribute that information from the realm of science into society. If forewarned is forearmed, then the IRI's efforts could help nations in Asia, Africa, or the Americas prepare for climate change. Or nationals from participating countries can learn the latest prediction techniques. This year, for example, says Nicholas Graham, director of the Institute's Climate Forecast Division, IRI assisted several national weather services in southern Africa in developing a consensus forecast for the region. Past El Niños have also warmed surface waters in the Indian Ocean, causing drought across southern Africa. Whether the specific forecast is for drought or deluge, the future looks bright for El Niño prediction. Now for that little problem of global warming. Blake Edgar is an Associate Editor of California Wild. His email address is bedgar@calacademy.org. |
Winter 1998
|
