A Climatic Crystal Ball
It is one of the most ambitious programs in science--designed not only to predict climate throughout the world, but to help nations cope with weather’s devastating effects, such as droughts, floods and disease.
The goal is to erase a familiar cliche that is perhaps as old as civilization: Everyone talks about the weather, but nobody does anything about it.
The International Research Institute for Climate Prediction--known as IRI--on Columbia University’s Palisades campus stands at the intersection of highly complex physics and the social sciences.
Its existence is testimony to major advances in long-range forecasting technology and growing knowledge about how the Earth’s atmosphere works in tandem with oceans and land masses to produce changes.
“Climate forecasting has been going on for decades, but with the new tools, the ability to assimilate a lot of data and analyze it has greatly improved,” said Michael H. Glantz, a senior scientist at the National Center for Atmospheric Research in Boulder, Colo.
Researchers at other institutions say IRI’s scope is unique.
“IRI is the first attempt of the climate science community to create a capacity to make long-range climate forecasts and to relate those forecasts to the concerns of human beings,” said Charles Kennel, director of the Scripps Institution of Oceanography.
“We go from basic science to the basic needs of society,” added Antonio Divino Moura, IRI’s director, former head of Brazil’s National Meteorological Service. Experts say the potential economic, political and social consequences of accurately predicting climate months ahead are profound.
Forecasts of wet or dry seasons, or normal or abnormal temperatures, may influence financial markets, particularly the prices of commodity futures, and may affect what crops farmers plant.
Industries may realize energy savings by purchasing fuel in advance, if they know whether temperatures will be higher or lower than usual.
Faced with a forecast of a season of heavy rains, public health authorities may be able to prepare mosquito control programs for such diseases as malaria and West Nile viral encephalitis.
In nations with unstable governments and chronic food shortages, serious climate changes can foment unrest and amplify political problems.
Historians say climatic change has contributed to the collapse of civilizations. A sudden drought played a major role in the downfall of the empire stretching from Turkey to the Persian Gulf led by Sargon I of Akkad 4,200 years ego.
“The economics of climate takes you into everybody’s world,” said Glantz. “It’s food, water, energy, security.”
Recognizing the difference between weather and climate is central to understanding IRI’s mission. If weather is a highly perishable snapshot of the atmosphere, climate is a far bigger canvas--the average of conditions over a much longer period of time.
IRI, part of Columbia’s Earth Institute, was founded in 1996 to produce long-range predictions for the use of policymakers. In August it received a $42-million grant from the National Oceanic and Atmospheric Administration--the largest in the agency’s history to a private institution.
The decision to establish IRI stemmed from observations of El Nino and La Nina phenomena in the 1980s--variations of ocean surface temperatures in the equatorial Pacific that produce extreme climate changes elsewhere.
Predicting climate is far more complicated than ordinary weather forecasting.
“What has made climate forecasting work is not the weather data but the data about the ocean,” said Edward S. Sarachik, professor of atmospheric sciences at the University of Washington. “The ocean moves slowly, and it contains the information about the future of the atmosphere.”
Researchers look for above- or below-normal sea surface temperatures as clues to climate. These departures from the norm usually occur slowly.
“Departures from normal sea surface temperatures create specific atmospheric circulation patterns, which can impact climate--the frequency of storms and the tendency for warm or cold weather,” said Tony Barnston, IRI’s head of forecast operations.
Central to the process are observations of conditions in the tropical Pacific. Networks of buoys measure temperature, ocean currents and winds. Orbiting environmental satellites, weather balloons and other instruments contribute more data.
IRI uses information from at least three climate prediction models issued by groups around the world. Some of these models move beyond sea surface temperature changes to measure snow depth and soil moisture, which can contribute to the average weather over a season.
To check accuracy and their capacity to predict, researchers at the institute run the models to simulate forecasts for previous years. The results are compared with climate records stretching back 45 years to evaluate the models’ performance.
Trying to Make a Forecast Six Months in Advance
Using the models, the institute arrives at a consolidated forecast for as much as six months in advance. IRI’s predictions are expressed in terms of probabilities for above-, near- or below-normal temperature or precipitation.
The process strains some of the world’s most powerful computers because of the complexity of the math and physics necessary to reach conclusions.
IRI’s accuracy varies by season and region.
“We do best in areas impacted by El Nino and La Nina,” Barnston said. “The challenge not only is to express the probabilities correctly, but to find broadened sources of forecasting skills in years when the tropical Pacific is quiescent.”
“It is an improving science, and the challenge is to try to understand all the components that make the climate work,” Kennel added. “It is one of the most complicated problems that scientists are trying to attack.”
Issuing climate predictions is only half of IRI’s mission. What makes the institute unique is the emphasis it places on helping countries use its forecasts.
“It opens up a new way of thinking,” said Carolyn Mutter, IRI’s science program coordinator.
To aid in that process, social scientists, economists, oceanographers, public health specialists, anthropologists, water resource experts and trainers in the use of forecasts work alongside climate experts at the institute.
Sometimes, predictions can raise surprisingly complex issues, including the politics of weather. Institute officials are well aware that they can face a minefield of sensitivities.
In an age of intellectual property rights, some countries have been reluctant to share data necessary for climate forecasts.
Other potential problems include government officials’ decisions on whether forecasts will be widely or fairly disseminated. Also of concern are whether adequate resources exist to protect against losses from climate catastrophes and whether predictions couched in probabilities will cause action or be ignored through inertia or distrust.
IRI has been treading lightly. In an effort to avoid such problems, it has undertaken dialogues around the world with potential users and has formed a wide network of partnerships.
The institute has contracts with the government of Taiwan to provide data, particularly in forecasting the frequency of typhoons. In the Horn of Africa, it is working with 10 nations to provide early warning of droughts and potential famines.
It has projects in eastern Australia to help predict seasonal rainfall and the rain’s effects, and in northeastern Brazil, where staff members are working with policymakers to plan for better water management during droughts.
“The mission of IRI is climate prediction and application under the same roof in a very integrated fashion,” said Moura, the institute’s director. “. . . There are things you can do by knowing the future.”
Goldman can be reached at:
(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)
Precipitation Predictions
The International Research Institute for Climate Prediction has released its global precipitation map for January-March 2001. In North America, only the southeast region of the United States is expected to be drier than normal. On the map, the three-number groups indicate precipitation probabilities expressed in percentages. The top bar is above normal, the middle bar is near normal and the bottom bar is below normal.
*
Source: International Research Institute for Climate Prediction
