6.
APPLICATIONS OF ENSO PREDICTION

As the skill for predicting sea surface temperature in the Pacific improved during the latter half of TOGA, attention was directed towards applying these predictions for economic and societal benefit. Physical and social scientists opened a dialogue during TOGA on how to develop and apply forecasts of short-term climate variations. Several of the tropical nations directly affected by ENSO variations have begun applying the fledgling ENSO predictions.

The scientific objectives established at the outset of the TOGA program focused solely on observing and modeling the coupled atmosphere-ocean system. Although it was anticipated in 1985 that observations and coupled-modeling efforts under TOGA would lay the foundation for operational prediction of short-term climate variations, the time when this might be practical could not be foretold. Assessment studies of the societal impacts of ENSO were not an integral part of TOGA at the beginning. By the mid-point of the program, however, scientists had demonstrated significant skill in predicting aspects of ENSO—sufficient skill for the forecasts to have beneficial value for society.

The demonstration of significant skill led the TOGA Panel to recommend the establishment of an experimental center for short-term climate prediction (NRC 1990). As the prospects for skillful and regular ENSO forecasts became more of a reality, a demand was generated for quantitative assessments of the socioeconomic impacts of ENSO. Similarly, it became necessary to understand how to apply ENSO forecasts and to formulate strategies for mitigating the more harmful aspects of climate variability. During the second half of TOGA, applications were emphasized as a necessary adjunct to the modeling, observational, and process-study components of the program. It was not a coincidence that this increased emphasis on ENSO applications evolved in tandem with a related emphasis within the USGCRP (CEES 1993) on the human dimensions of global climate change.

Prior to the start of TOGA, the most well-studied societal impact of El Niño was the collapse of the Peruvian anchoveta fishery in 1972–1973 (Glantz and Thompson 1981). The effects of El Niño along the west coast of South America received widespread attention. What was once thought to be a regional climate problem was seen to have global economic implications in the form of rising prices for fish meal. Examinations of lead and lag relations between the Southern Oscillation Index and regional climate anomalies, particularly



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program 6. APPLICATIONS OF ENSO PREDICTION As the skill for predicting sea surface temperature in the Pacific improved during the latter half of TOGA, attention was directed towards applying these predictions for economic and societal benefit. Physical and social scientists opened a dialogue during TOGA on how to develop and apply forecasts of short-term climate variations. Several of the tropical nations directly affected by ENSO variations have begun applying the fledgling ENSO predictions. The scientific objectives established at the outset of the TOGA program focused solely on observing and modeling the coupled atmosphere-ocean system. Although it was anticipated in 1985 that observations and coupled-modeling efforts under TOGA would lay the foundation for operational prediction of short-term climate variations, the time when this might be practical could not be foretold. Assessment studies of the societal impacts of ENSO were not an integral part of TOGA at the beginning. By the mid-point of the program, however, scientists had demonstrated significant skill in predicting aspects of ENSO—sufficient skill for the forecasts to have beneficial value for society. The demonstration of significant skill led the TOGA Panel to recommend the establishment of an experimental center for short-term climate prediction (NRC 1990). As the prospects for skillful and regular ENSO forecasts became more of a reality, a demand was generated for quantitative assessments of the socioeconomic impacts of ENSO. Similarly, it became necessary to understand how to apply ENSO forecasts and to formulate strategies for mitigating the more harmful aspects of climate variability. During the second half of TOGA, applications were emphasized as a necessary adjunct to the modeling, observational, and process-study components of the program. It was not a coincidence that this increased emphasis on ENSO applications evolved in tandem with a related emphasis within the USGCRP (CEES 1993) on the human dimensions of global climate change. Prior to the start of TOGA, the most well-studied societal impact of El Niño was the collapse of the Peruvian anchoveta fishery in 1972–1973 (Glantz and Thompson 1981). The effects of El Niño along the west coast of South America received widespread attention. What was once thought to be a regional climate problem was seen to have global economic implications in the form of rising prices for fish meal. Examinations of lead and lag relations between the Southern Oscillation Index and regional climate anomalies, particularly

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program droughts, had already established global connections. However, decision makers in Peru were not aware of this climate information that might be useful for their planning of agriculture and aquaculture. The global socioeconomic effects of ENSO did not attract intense interest until after the severe conditions of 1982–1983. Just as the 1982–83 climate anomalies mobilized the meteorological and oceanographic scientific communities under TOGA, so did the related human suffering draw the attention of the social, economic, and political communities. Drought in Australia, Indonesia, northeast Brazil, and southeast Africa, along with flooding in Ecuador, Peru, southeastern South America, and the western and southern tier of the United States, had adverse effects on social conditions and some economic sectors. Economists, social scientists, anthropologists, and politicians came together with the physical scientists to study the global societal effects of ENSO. One of the legacies of the TOGA program is the dialogue that was established between the physical and social sciences (Moura 1992). This exchange of ENSO information was critical to the ultimate success of the program. Prior to TOGA, most information on ENSO impacts was anecdotal and derived from the popular press. TOGA helped to sponsor and support several forums for those with a common interest in ENSO. Sociologists, economists, and politicians were informed of the advances in and limitations of experimental ENSO forecasts. Climate scientists learned of the needs and problems of policy and decision makers. Education and training programs were geared to the application of ENSO forecasts for participants from tropical and developing nations. Collaborative arrangements were initiated between potential providers and users of ENSO forecasts. The governments of countries that are most directly affected by ENSO (e.g., Peru, Brazil, and Australia) now routinely take experimental ENSO forecasts into account when making decisions. It was within this context of being able to provide societally relevant climate information to the countries affected by ENSO that a proposal (Moura 1992) for an International Research Institute for Climate Prediction (IRICP) arose. In June 1992, at the United Nations Conference on Environment and Development (UNCED), the Bush administration committed the United States to implementing “a pilot project to demonstrate the operating concepts embodied in the plan and invited government officials and scientists from all interested nations to join in developing an International Research Institute for Climate Prediction” (OMB 1992). DEVELOPMENT OF APPLICATIONS AND ASSESSMENTS ENSO forecasts are societally relevant because the agrarian sectors in tropical countries depend critically on the water received during the rainy season. In the

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program tropics, precipitation and sea surface temperature are tightly coupled, so useful precipitation information can be derived from forecasts of sea surface temperature. Information on the probability of decreased or enhanced rainfall on seasonal-to-interannual time scales offers some countries the opportunity to take intelligent action. The impacts of ENSO are varied. They depend on whether a region is rural or urban, whether the economic sector is agricultural or industrial, and whether the land holdings are small or large. The form of response to ENSO will vary according to the past history of the forecasts, the ability or desire of the government to intervene, and the acceptance of the information by the citizenry. Some societies are averse to risk and will implement mitigation strategies, while others may choose not to do so. For example, if upcoming precipitation patterns or tendencies are predicted with sufficient accuracy and a drought is forecast, a farmer can substitute drought-resistant crops, usually reducing yield but avoiding disaster. Similarly, a commodities broker has the option of obtaining supplies from another portion of the globe if a particular crop is in danger. During the course of the TOGA decade, physical scientists, social scientists, economists, and decision makers from industry and government came together to analyze a range of factors influencing responses to ENSO effects by country and region. This entailed assessing the consequences of ENSO, discussing the application of ENSO forecasts, and judging the economic value of ENSO predictions. These assessments included studies of agricultural production, coastal fisheries, emergency preparedness, systems for early warning of drought and famine, strategies for hydroelectric power generation, insurance concerns, and mitigation of forest fires. A series of workshops proved very successful in bringing these seemingly disparate communities together to work on topics of mutual concern. The series* included: Societal Impacts Associated with the 1982–83 Worldwide Climate Anomalies, November 1985, Lugano, Switzerland; Workshop on the Economic Impact of ENSO Forecasts on the American, Australian, and Asian Continents, August 1992, Tallahassee, Florida; Workshop on ENSO and Seasonal-to-Interannual Climate Variability: Socio-Economic Impacts, Forecasting, and Applications to the Decision Making Process, September 1992, Fortaleza and Florianopolis, Brazil; Pacific El Niño-Southern Oscillation Applications Workshop, October 1992, Honolulu, Hawaii; *   Informally published proceedings of most of these workshops were prepared and can be obtained from the NOAA Office of Global Programs.

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program Usable Science: Food Security, Early Warning and El Niño, October 1993, Budapest, Hungary; Inter-American Institute Workshop on ENSO and Interannual Climate Variability, July 1994, Lima, Peru; Regional Workshop on Research and Application of Climate Forecasts in the Decision Making Process in Northern and Northeastern South America, September 1994, Fortaleza, Brazil; Regional Workshop on Research and Application of Climate Forecasts in the Decision Making Process in Southeastern South America, September 1994, Montevideo, Uruguay; and Workshop on Short-Term Climate Forecasts and their Applications for Social and Economic Benefit and Sustainable Development, November 1994, Bali, Indonesia. The cross-disciplinary communication fostered by these meetings heightened awareness of the relationships between climate fluctuations and society. It became apparent that scientists in developing nations were on the verge of being able to advise their governments on the value of short-term climate predictions. Numerous examples and case studies made it clear that a well developed strategy would be needed to produce accurate and socioeconomically relevant climate information. These workshops, held during the last half of TOGA, helped to identify the needs of the economic sectors most affected by short-term climate variations, areas in need of further work if ENSO forecasts are to be useful for mitigating economic and social impacts, and methodologies for the systematic production and dissemination of climate-forecast information. Emerging from these exchanges was an appreciation of the issues encountered in applying ENSO forecasts, as well as several examples where governments were actually taking action on the basis of experimental short-term climate predictions. During the workshops, it was recognized that economic assessment exercises were being done on a case-by-case basis. A systematic quantitative evaluation of the global economic impact of ENSO was called for. It was noted that programs are needed to improve the confidence in ENSO forecasts by means of education, data exchange, improved regional-model accuracy, a focus on probability forecasts, and support for near-real-time, sustained environmental monitoring. The fact that the societal response mechanisms differ from country to country means that communication and education are key to the successful application of ENSO forecasts. There is a fundamental obligation to consult with, and understand the needs of, the users and consumers of the information. The education that is needed to accomplish this goes in both directions. In order to be useful, ENSO forecasts, bulletins, and advisories must be tailored for users on the basis of input from them. The end users must also be apprised of the

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program limitations of the forecasts. Regular production of forecasts is needed to build confidence and establish rapport between forecasters and users. Decision makers must be informed of the value of today's climate information and how it might be used in the future. In response to this need for education and training, and to the U.S. proposal to UNCED for an International Research Institute for Climate Prediction pilot project, an applications and training pilot project was established at the Lamont-Doherty Earth Observatory in 1993. Approximately 29 participants from 17 countries were trained as of the end of 1994. The trainees learned about current capabilities in climate modeling, and helped to develop practical applications for their home country that link model forecasts with regional impacts. APPLICATIONS OF REGIONAL ENSO FORECASTS At the end of TOGA, ENSO forecasts and their applications were performed by small research groups in developed countries on an ad hoc basis. These groups lacked the infrastructure required to incorporate all relevant information into forecasting schemes. Research groups did not have the resources to adapt and disseminate the forecasts on the regional scale. Notwithstanding these impediments, many countries now recognize the value of short-term climate forecasts. Countries such as Peru and Brazil provide the best examples of how this forecast information has been used by policy makers to mitigate the socioeconomic effects of ENSO. The successes achieved in South America and else-where have demonstrated that a strong partnership is needed between developed and developing countries if all are to benefit from the new knowledge of ENSO predictability. Peru In Peru, the government plays an important role in setting agricultural policy. Every year an agricultural plan must, by law, be developed. This plan encompasses irrigation practices, prices of seed and fertilizer, agrarian loans, crop subsidies, and technical assistance. Following the 1982–1983 El Niño, the gross national product and the gross value for the agricultural sector dropped drastically. However, because of the application of predictive information, agricultural production did not decrease following the 1986–1987 El Niño (Lagos and Buizer 1992). The growing season of 1986–1987 was the first time the agricultural plan had been influenced by the prediction of an upcoming warm event. In late 1986, government officials and nongovernmental agrarian organizations were advised by Peruvian scientists that the observation and modeling efforts of

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program TOGA investigators were forecasting the coming of El Niño. The prediction of a warmer and wetter growing season helped determine the appropriate combination of crops. Two of the primary crops of the northern coastal region of Peru, rice and cotton, are very sensitive to the amount and timing of rainfall. Rice requires a large volume of water and relatively warm conditions during its growing phase. Cotton has deeper roots and is capable of greater yields during years of light precipitation. By using ENSO-based forecasts as a basis for crop selection, Peruvians have been able to sustain the gross value of their agricultural sector, increasing it 3 percent in 1987 in spite of the moderate 1986–1987 El Niño. In contrast, it decreased by 14 percent in 1983 following the 1982–1983 episode before forecasts were available. Northeast Brazil Northeastern Brazil, or the Nordeste, offers another example of the advantageous use of interannual climate predictions for agricultural planning. This semi-arid region, home to 30 percent of Brazil's population, encompasses 18 percent of the country's area (Magalhaes and Magee 1994). The Nordeste has a long history, going back to the colonial era, of social upheavals and migrations tied to droughts. In this region of nearly 45 million inhabitants, one in five Nordestinos emigrates to another portion of the country, often the Amazon, to escape drought. During modern times, drought in the Nordeste has required the federal Brazilian government to devote considerable resources for ameliorating the drought conditions in order to forestall further emigration. In the Nordeste state of Ceara, the government has demonstrated that population displacements can be reduced by assisting subsistence farmers. The government in this state has become quite sophisticated and active in using forecasts related to ENSO. The state-supported Ceara Foundation for Meteorology and Hydrological Resources (FUNCEME) has the responsibility for advising the government on drought-related decisions. Short-term climate predictions from abroad and local monitoring information in meteorology, hydrology, and agriculture are used to guide state decisions on agricultural and water conservation. As in Peru, climate-forecast information has helped to alleviate the drastic consequences of ENSO in Brazil (Moura 1994). For example, the normal annual grain production in the state of Ceara is 650 000 tonnes, mostly grown during the rainy season from February to April. In 1987, even though an ENSO forecast had been made (Cane et al. 1986), policy makers were unaware of it and no action was taken. The precipitation was 30 percent below normal that year and the total crop production was only 100 000 tonnes. In 1992, the environmental conditions were as poor as at the time of the previous ENSO

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program warming, but the response of the agricultural sector was markedly different. Precipitation for 1992 was 27 percent below the long-term mean, yet the grain production for the year was 530 000 tonnes. The primary reason for the dramatic improvement was that in late 1991, FUNCEME acted on a combination of ENSO forecasts and information on conditions in the Atlantic, and alerted the government to an impending dry growing season. Then-Governor of Ceara, Ciro Gomes, traveled to the interior of the state with the Secretary of Agriculture and the President of FUNCEME to initiate a series of actions that culminated in the farmers planting crops appropriate for a shorter growing season. Potentially catastrophic conditions were averted for the general populace when FUNCEME combined observations of the state's water resources with the drought forecast, and advised that the capital city of Fortaleza would run out of drinking water by December 1992. The state acted to restrict water consumption and invested U.S.$13 million to construct a new dam. In 1993, actions were taken in anticipation of a second year of drought. Rainfall was 45 percent below the norm and grain production dropped to 250 000 tonnes, yet this level was still 2.5 times more than the total production in 1987. The responses encouraged by FUNCEME prevented what could have been the disastrous consequences of a second consecutive year of drought. Australia South America is not the only continent where short-term climate predictions have been put to regular use. The seasonal rainfall of Australia is characterized by significant year-to-year variability, much of which is linked to ENSO. Simpson et al. (1993) showed that annual flow variations for the River Murray, Australia's most extensive river system, are negatively correlated with sea surface temperature in the eastern equatorial Pacific. Encephalitis epidemics are known to occur in the Murray Valley during periods of high rainfall and flooding that coincide with cold phases of ENSO (Nicholls 1994). Because of the severe impact of ENSO-diminished rainfall on rural agricultural production, the Australian Bureau of Meteorology has, since 1988, been issuing seasonal climate outlooks. The primary basis for the outlooks is the status and forecast of ENSO. Information is disseminated by the national and state governments, and through the media. Going into the 1991–1992 ENSO warming, rainfall over much of Queensland and northwestern Australia was generally below to very-much-below average for the period November 1991 to January 1992. Warnings of the forthcoming shortfall in precipitation were issued up to three months in advance, permitting planning by the Queensland Department of Primary Industries. The use of the climate outlooks has tended to flatten variations in crop yields from year to year as a result of mitigating actions by farmers. The suc

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program cess of these outlooks has prompted the Queensland State Government to establish an impacts, monitoring, and decision support system to provide advice at the regional and even at the individual-farm level. Asia In Asia, experimental ENSO forecasts are a key input to the monsoon-forecast schemes of the Indian Weather Service. The relationship between ENSO and monsoon variability is indisputable (Webster and Yang 1992), but still an active research topic. In Japan, an ENSO-prediction division has been established within the Japanese Meteorological Agency (JMA). ENSO forecasts based on the predicted displacement of the jet stream are used to develop outlooks for summer and winter temperatures. JMA successfully issued long-range forecasts of a cool and less sunny 1991 boreal summer and a mild 1991–1992 winter, on the basis of a May 1991 prediction of the 1991–1992 warming. In the Philippines, a drought early-warning and monitoring system has been established. Seasonal rainfall advisories are based primarily on ENSO conditions. Other Foreign Regions In addition to the examples presented above, where ENSO forecasts are beginning to be used on a routine basis, TOGA research has pointed to many other regions and sectors that could benefit from systematic ENSO forecasts. The National Meteorological Service of Ethiopia has been issuing experimental seasonal forecasts based on ENSO predictions since 1987 (Bekele 1994). Advisories are released for the spring and fall growing seasons. The recommendations take into account land-use strategy, conservation policies, and economic assistance policy. The development of a seasonal forecast model has been hampered by a shortage of Ethiopian personnel trained in long-range forecasting. Southern Africa was one of the regions dramatically affected by climate variations during 1991–1992. The drought there was worse than any other in that region for the past hundred years. By December 1991, rainfall accumulation was still less than half of the long-term mean throughout Mozambique and southern Zimbabwe. By January 1992, precipitation was as much as 70 percent below normal throughout the southern portion of the continent. Nearly 100 million people were affected, and 12 million tonnes of commodities were imported by relief organizations. Cane et al. (1994) showed that there is a strong link between ENSO and interannual fluctuations of both precipitation and maize production in Zimbabwe. They also demonstrated that coupled-model predictions of ENSO could be used to provide an accurate forecast of the

OCR for page 123
Learning to Predict Climate Variations Associated with El Niño and the Southern Oscillation: Accomplishments and Legacies of the TOGA Program Zimbabwe maize yield, with lead times up to one year. This information could be a valuable component of famine early-warning systems for the region. Rosenzweig (1995) provided a global perspective on the effects of ENSO on crop yields. Her preliminary work indicated that yields of five major crops in about thirty countries appear to be correlated with ENSO variations. For most of the crops, yields tended to be lower in years with anomalously warm equatorial Pacific sea surface temperature. One-third of the world's production of maize, wheat, sorghum, soybeans, and rice was found to be correlated with phases of ENSO, although for many regions the signal was small. United States Investigations of links between agricultural production and ENSO are just beginning for the United States. Adams et al. (1995) used the historical temperature and precipitation data for the southeastern United States, together with a model of crop yield, to estimate that climate variations associated with ENSO could decrease crop yield in the southeast by up to 15 percent. The value of improving an ENSO forecast from correlations at the level of roughly 0.6 to a level of 0.8 at lead times of six months in advance of planting decisions was estimated to be more than $100 million per year for the southern agricultural sector alone. Skill for predictions outside the tropics is likely to be much lower than skill for the tropics because of the greater variability of the climate around the mean. Nevertheless, experimental prediction activities for the United States have begun (Ji et al. 1994a, Kerr 1994). Near the end of TOGA, the National Meteorological Center began to use coupled ocean-atmosphere forecasts, together with several statistical techniques, to arrive at an official consensus forecast with lead times up to two seasons in advance.