EXPERIENCE has taught us that attempts to predict the weather far ahead in any meaningful way will be frustrated by chance or chaos. Until now, that is. If research published this week is correct, the arrival of the massive upheaval in Pacific weather patterns called El Ni帽o is predictable up to two years in advance.
Dake Chen and colleagues from the Lamont-Doherty Earth Observatory at Columbia University, New York, trained their predictive model on observations from the past 20 years, and then successfully 鈥渉indcasted鈥� every El Ni帽o for the past 148 years. Using nothing but historical records of sea-surface temperatures they were able to see signs of the biggest El Ni帽os two years before they happened (Nature, vol 428, p 709). The model generates a self-sustaining oscillation with a period of three to five years. And its remarkable accuracy suggests that El Ni帽o is driven by some internal rhythm, not by chance or chaotic events.
El Ni帽o brings storms to arid regions of the Americas and drought to south-east Asia and Australia. The effects ripple still farther afield, bringing snow and rain to Europe and drought to southern Africa. Short-term warnings of El Ni帽o already whiz round the world to planners in rich nations, giving farmers in South America, for example, enough of a heads-up to plant rice instead of cotton.
Advertisement
If reliable long-term predictions are now possible, the trick will be to make sure they reach peasant farmers in Asia and Africa as well. That way, El Ni帽o forecasting will have a chance of saving lives as well as farm profits.