RETURN TO TOP OF PAGE
In response to recommendations made during the November 1995 International Forum on Forecasting El Niño, a planning process to further explore the concept and foster the development of the International Research Institute (IRI) has begun. The international planning process will proceed along two parallel tracks, as the design of a complex network of climate research, forecasting and applications activities with a centralized core facility requires the participation and commitment of a diverse collection of countries, institutions and disciplines.
Early this year, a meeting was convened in Vancouver, Canada (16-17 January 1997) to begin a realistic assessment of the prospects for multilateral support for activities at the "core" of this international network, the central facility of the IRI. It is envisioned that the IRI core will centralize functions associated with the regular production, dissemination and improvement of experimental forecast information and will work closely with an associated network of centers conducting a full suite of modeling and assessment activities. Since issues related to the IRI core facility was the primary focus of the planning effort launched in Vancouver, as opposed to the full suite of IRI network activities, the group involved in those discussions refers to itself as the "IRI Core Group". The second Core Group meeting will be hosted by Brazil, 29-30 April, at the Instituto Nacional de Pesquisas Espaciais (INPE), Sao Jose dos Campos, Sao Paulo.
In parallel, plans are underway to convene the first meeting of an IRI ad hoc working group in the second half of 1997, as expressed in the 1995 Forum Statement. As noted in the Forum Statement, there is a need to articulate the requirements for the integrative modeling and assessment activities of an "end-to-end" climate information system for seasonal-to-interannual time scales. It is intended that this first meeting of the ad hoc working group will begin an exploration of the requisite components of an IRI network of research centers and applications activities, their respective roles and the relationships among the component parts. In addition, the group will be encouraged to address the tangible tasks necessary to transform the concept and shared vision of an IRI network into a reality. We anticipate that the ad hoc working group will consist of individuals representing the institutions that will participate in the IRI network, states sponsoring the IRI core facility, researchers from the physical and social science, representatives of related international programs and participants from the user communities that the IRI will serve (e.g water and agriculture managers).
As the work of these parallel planning efforts moves forward, we will continue to provide the broader international climate community and potential users of IRI forecast products with updates via the ENSO Signal.
by Dr.Guillermo Berri, IRI Pilot Training Project
Climate variability on seasonal to interannual time scales has a profound effect on water resources and food production around the world. Water catchment and crops develop in cycles completely embedded in these time scales. Departures from average conditions often create situations in which society experiences major environmental problems. There are occasions when extreme events bring such disasters to society, such as drought, flooding, famine and disease outbreaks. Often, society is not prepared to respond with countermeasures because it is unaware of such events.
In many tropical and subtropical regions around the world, the El Niño/Southern Oscillation (ENSO) is responsible for much of the observed climate variability, and is second only to the seasons themselves. Advances in scientific knowledge made possible, more than 10 years ago, the first successful prediction of El Niño. Based on these results, the National Oceanic and Atmospheric Administration (NOAA) developed a Pilot Project in 1993 to demonstrate the operating concepts of an International Research Institute (IRI). The focus is to provide practical experience and generate products useful in implementing a multi-national institute dedicated to the practical applications of seasonal-to-interannual climate prediction to decision-making.
Activities of the pilot project include training courses for scientists of the climate, hydrology and agriculture communities, a series of workshops organized by the NOAA Office of Global Programs (OGP) to initiate dialogue in regions around the world, and a project to intercompare experimental simulations of several ocean atmosphere coupled climate models, carried out at the Scripps Institution of Oceanography (SIO), University of California. The purpose of the pilot project training courses is to expose scientists of the climate, hydrology and agriculture communities, who are familiar with seasonal to interannual climate variability and ENSO related impacts in agriculture and water resources, to state-of-the-art climate monitoring and short-term climate prediction. Particular attention is paid to practical applications linking climate model forecasts and regional impacts as revealed by observations. A training facility was established at Columbia University/Lamont-Doherty Earth Observatory (LDEO), Palisades, NY where each participant has a terminal dedicated to a multiprocessor computer for the exclusive use of the course.
Two 9-month training courses have been conducted to train climatologists and oceanographers from countries affected by seasonal and interannual climate variability and ENSO related impacts. In view of the experience gained in these courses, the pilot project expanded its training activities by organizing shorter courses aimed at exploring practical applications of climate predictions to decision making in agriculture, water resources management and hydro-electricity generation. Two 3-month courses were conducted at LDEO, one for agronomists and the other one for hydrologists. In addition to the local IRI training, a series of regional short courses targeting the needs of specific regions have been organized. The first 1-month course was conducted in Santa Fe, Argentina during July 1995 for hydrologists and hydraulic engineers from Central and South America. A 2-week course was conducted in San Jose, Costa Rica during August 1996, when for the first time climatologists, hydrologists, agronomists and economists from Mesoamerica and the Caribbean exchanged ideas in a multi-disciplinary environment.
The IRI training courses are theoretical and practical. The theoretical part covers one-third of the course and is aimed at providing an overview on climate analysis and ENSO dynamics, seasonal-to-interannual climate prediction techniques, current coupled models and data requirements. During the practical part, participants work under the continuous supervision and assistance of IRI training staff and other collaborators. This team of experts helps participants define a practical application project, primarily aimed at developing methodologies for tailoring seasonal-to-interannual climate predictions to practical use by decision makers in agricultural planning, water resources management and hydro-electricity.
Participants are requested to bring the longest possible records available on monthly averages of precipitation, temperature and river discharges, as well as crop statistics from their region. The availability of this data is a pre-condition for participation in the IRI training courses, as the free exchange of data and results among the participants have always guaranteed the success of the training activities. Participants incorporate their data bases into program Climlab, a software specifically developed by the pilot project for research on practical applications of short-term climate predictions to decision-making. Climlab can run on PCs and Unix systems and combines a powerful statistical package with a graphics package with geographical references and a historical data base of global sea surface temperatures. A user-friendly environment allows users to create simple statistical prediction models and validate these model predictions with the local data. Climlab has become a main tool of IRI training and in order to facilitate the future activities, participants are furnished with a copy of the program and are assisted in the development of communication links and protocols for the routine acquisition and transfer of data between their home institutions and IRI.
At the end of the course participants present a final report describing the experiments performed and results obtained, along with recommendations for future actions. These reports are compiled in a final comprehensive report on the training course which is edited and distributed soon after the activity. Every effort is made during the courses to prepare the participants to serve as a nucleus for initiating country specific demonstration projects and help establish the international network required to support a fully multi-national IRI.
The IRI pilot project has made a major contribution to the development of an international cadre of scientists working in the application and practical use of seasonal-to-interannual climate predictions for decision-making. Through its courses, a total of 76 climatologists, oceanographers, hydrologists, hydraulic engineers, agronomists, economists and application specialists from 27 countries and regions have received training. Participating countries and regions are: Argentina, Australia, Brazil, Chile, China, Colombia, Costa Rica, Cuba, Ecuador, El Salvador, Guatemala, Honduras, India, Indonesia, Japan, Kenya, Mexico, Nicaragua, Panama, Paraguay, Peru, South Africa, Taiwan, the U.S. Pacific Islands, Uruguay, Venezuela and Zimbabwe. The figure depicts the regions covered by the practical application projects and the different topics addressed by participants.
The International Forum on Forecasting El Nino, held in Washington DC in November 1995 with the purpose of launching IRI. As a result, a grant was issued by the NOAA/OGP to Columbia University/LDEO in consortium with the University of California/SIO, to prepare the provisional IRI. It is envisioned that the IRI will produce, interpret and deliver experimental forecast guidance products and will promote their practical application for decision-making throughout the world on a regular and continuing basis. Application specialists from around the world will use these model forecasts to tailor seasonal-to-interannual climate predictions to practical use in decision-making (i.e. policy-making by government officials, water resources management, crop planning, human health and prevention, and awareness of such natural disasters as droughts and floods). Thus, the IRI concept is s to build a multi-national network committed to advancing the understanding of climate variability in order to better manage resources and mitigate its adverse impact on society.
A L du Pisani, Namibia Meterological Service
Namibia is an arid country, probably the driest country in Africa south of the Sahara. The mean annual rainfall is about 270 mm and ranges from less than 20 mm in the Namib desert to more than 700 mm at Katima Mulilo in the Caprivi strip. The distribution of land area receiving different categories of rainfall is as follows:
Rainfall (mm) Percentage of Land Surface(%)
<100 22
100 - 300 33
300 - 500 37
>500 8
Namibia has a pastoral agriculture with crop production only being practiced in the northeast where >500 mm rainfall is received. Climate variations like droughts and to a lesser degree, floods, have a major impact on the agriculture, and since a large proportion of the population still lives in rural areas, drought usually has a devastating effect on the rural poor. Even in good years, Namibia is not self sufficient as far as cereals are concerned and has to import varying quantities from its neighbors, usually South Africa.
During the past eighteen years, Namibia has had a long run of below-normal rainfall with very few normal years. The 1994/95 season was a particularly bad year with low and erratic rainfall and unfavorable rainfall distribution. This is borne out of the fact that during late summer, flooding was a problem in the north and was especially unexpected because the catchment areas of the rivers that flooded are in southern Angola.
Namibia has been interested in long lead forecasting for some time and in 1994 started its own initiative in the Meteorological Service based on a moderately successful preliminary study on the subject. The Meteorological Service has already started to initiate discussions with South African scientists who were part of the first IRICP training program at the Lamont-Doherty Earth Observatory of Columbia University and have built up enviable experience during the past two seasons. Also, the relationship between Namibian researchers and those at the Climate Prediction Center in Washington, D.C. has led to an impending international collaboration in the Africa Desk initiative.
In a country like Namibia where the year-to-year variability of weather can be very severe with consequent devastating effects on the economy, it is important to take part in any endeavor which can help us predict on a seasonal or interannual time scale so as to be prepared for any emergencies. The farming community is very aware of the usefulness of these predictions and through the Namibian Agronomic Board has even contracted scientists from the University of Cape Town, South Africa to issue such forecasts. The Emergency Management Unit of the Office of the Prime Minister works in close cooperation with the Meteorological Service as part of the Early Warning Service which issues ten-day outlooks based on past weather, NDVI maps and the forecast, both on short and longer range. There is also close cooperation with the Southern African Development Community (SADC) Drought Monitoring Centre in Harare, Zimbabwe.
Namibia has also launched NAPCOD (Namibia's Plan to Combat Desertification) as a multi-agency programme in which all Non-Government Organizations can be involved. The Ministry of Environment and Tourism plays the lead role with the Ministry of Water, Agriculture and Rural Development, as well as the Meteorological Service in the development of this initiative. All of these efforts show how determined Namibia is to make use of all tools to combat the natural limitations of the climate in order to make Namibia prosperous.
by Dr. E.S. Sarachik, University of Washington and Eileen Shea, Center for the Application of Research on the Environment/Institute of Global Environment and Society, Inc.
Introduction
Whether we are aware of it or not, we all make predictions: we adjust our lives according to what we believe is going to happen in the future. When it comes to weather and climate, in the absence of any other information, we usually assume conditions will be the same next year as they were last year&emdash;we are usually wrong.
Over the past few years, the skill in predicting aspects of the climate system, especially sea surface temperature (SST) in the tropical Pacific characteristic of ENSO variations, has increased notably. Tropical Pacific SST is directly and robustly connected to climate conditions (especially temperature and precipitation) in some tropical regions of the world such as Peru, Ecuador, Chile, Australia, and Indonesia. For other regions, in particular, regions outside the tropics (such as the continental US), a connection between Pacific sea surface temperature and rainfall patterns can still be identified, albeit less reliably, other regional and local conditions and processes must also be accounted for in the development of useful forecasts. In both cases, however, scientists and decision-makers are faced with the same central question: what can we do with these forecasts? More specifically, how can these forecasts be used for the benefit of affected communities, regions and economic sectors. This question is at the heart of the vision for the IRI: scientists everywhere and decision-makers in the affected regions share a responsibility for finding an answer to this question.
The recent publication of the IPCC report indicates that a similar situation arises when considering the climatic response to the anthropogenic increases of radiatively active constituents in the atmosphere. There the answer is clear: an "integrated assessment" of the effects of long-term climate change is performed so that the presumed causes of this climate change, the emission of CO2 and other radiatively active constituents, can be altered in order to ameliorate future climate changes. In considering the short-term climate variations characteristic of ENSO, however, we take the point of view that nothing can be done to alter short-term climate variability and all that can be done is to predict the climate variations and make use of the predictions. Since the problem of the use of short-term climate predictions is so radically different from the long-term case, a radically different terminology is called for. We will therefore use the terminology "end-to-end" prediction to describe the suite of activities that proceeds from making the prediction to using it for societal or private benefit. In other words, we are compelled to discuss an "end-to-end" prediction program which provides for forecasts of changing climatic conditions, assessments of the consequences of those variations, and the application of the resulting information to support decision-making in the public and private sectors.
The Elements of End-to-End Prediction
We can divide the elements of end-to-end prediction into three separate categories: physical climate prediction; the consequences of climate variations; and the applications of these predictions to ameliorate unfavorable consequences and exploit favorable ones.
Physical Climate Prediction
Current short-term climate prediction uses coupled atmosphere-ocean models, initialized by data in the tropical Pacific, to predict sea surface temperature in the tropical Pacific a few months to a year in advance. While the skill of this prediction varies decadally, for reasons yet unexplained, approximately 60% of the variance of SST variations can be predicted a year in advance. The same models that predict SST also predict rainfall in the tropical Pacific region and those countries around the that region may be able to use the rainfall forecast directly. But for regions affected by ENSO but far removed from the tropical Pacific (e.g. the northwestern and southeastern parts of the US; Zimbabwe; the northeastern and southeastern parts of Brazil), the sea surface temperature conveys imperfect and, at present, less reliable information about the climate to be expected in these regions. To use the SST information, atmospheric models with fine resolution, possibly including imbedded mesoscale models, are run using the predicted SST as boundary conditions.
The objective of short term climate prediction is not to make perfect forecasts; it is rather to learn to use the skill of the forecasts that can be made.
Consequences of Short Term Climate Variability
It is not enough to say how the physical aspects of climate vary in a given region, for this tells us little about what its true consequences are. In addition to providing information on anticipated changes in physical conditions, such as rainfall and temperature in a given region, we must also address the nature of how those changing conditions are likely to affect sectors and communities which are sensitive to climate variability, and how adaptable these sectors are. The word "vulnerability" has been used to characterize the degree to which sectors can be affected by climate variations. Vulnerability has two aspects: sensitivity and adaptability. In general, vulnerability decreases with the degree of economic and cultural development of a region or sector and increases with the length and intensity of the stress applied to the system. Lest one think that climate consequences are all negative, we will also consider the "opportunity" afforded by favorable climate variations.
The consequences of climate variability on a given system are determined by the way that system normally works, and the way that it works is usually much different than obvious at first glance. Most systems of interest, such as agriculture, water resources, transportation, etc., have an intrinsic human and economic dimension. For example, we may think of agriculture as the act of putting seeds in the ground, waiting for rain, and harvesting the crop. But include the additional factors of price supports and government payments in kind, the availability of loans, the number of generations of the family living on the land, the cultural assets and norms of the community, the distribution of resources between large and small landholders, the existence of agricultural co-ops, the willingness of neighbors to help out in bad times, the availability of alternate work close by, the need for the children to work during harvest time, etc., and we realize that the system is far from simple and the successful workings of agriculture may depend on a number of factors not usually considered agricultural.
Climate variability works on the agricultural system as it really is and has consequences that depend both on the obvious factors of agriculture and on the added human and economic dimensions of agriculture. For example, in the USA, a low rainfall year may lead to crop failures and the imposition of a whole series of disaster relief efforts, most of which are designed to keep people from losing their farms. In the Northeast part of Brazil, however, the lack of such efforts combined with the total lack of alternate work for the peasant farmer has in the past led to mass migrations away from Northeast Brazil to the bigger cities of the south.
In general, therefore, physical climate variability is mediated by the vulnerability and opportunity characteristic of the system under consideration. An assessment of the consequences of climate variability must therefore consider the normal workings of the entire system in order to assess the vulnerability and opportunity afforded by the climate variations.
Applications of Predictions
Once the consequences of climate variability have been identified for a given community or sector, the question arises: can a prediction of climate a year or so in advance be of any use? The purpose of applying climate predictions is to exploit the opportunities and ameliorate the vulnerabilities presented by climate variability and this requires that some user be able to take some action that can affect the future.
First, a user of short-term climate predictions must be found. While research can help clarify the vulnerability of certain communities and sectors to climate variations and the opportunities that these climate variations present, the useful application of new forecasts requires the identification of relevant decision-makers and an exploration of the timing and characteristics of forecasts which they would find most useful. This requires the establishment and maintenance of a continuing interactive dialogue between the forecasting community and the potential users. The vision of an IRI has, from the beginning, recognized the need for such a continuing dialogue through which users help shape the forecasting system and the forecasters help the users to understand and apply the information.
A fundamental distinction must be made at this point: that between public and private users. The goal of private organizations (and we will mean commercial rather than non-governmental) is profit. We don't have to worry that private organizations won't seek their best interest and strive to accomplish their goal of earning higher profits-the growth of the weather industry in the US and its ability to find private applications for tailored weather information is a testament of the ability of private concerns to identify and exploit the opportunities involved with climate prediction.
The challenge of facilitating applications in the public sector is more complex: involving a more diverse set of decision-makers representing different, and sometimes competing, policy interests. In general, the goals of public-sector actions are complex and indirect, influenced as they are by politics, bureaucracy, lack of clear authority, institutional rigidity, fear of failure, and competing and conflicting demands on public resources. The very nature of the public enterprise, then, can present numerous barriers to the application of short-term climate forecasts, including among others: political considerations, existing bureaucratic structure and institutional rigidity, competing demands for common resources (e.g. water), and broad public consequences of failure.
Yet when conditions are right, the public sector can respond. In northeast Brazil, for example, the public sector has been able to respond to climate forecasts by building irrigation infrastructure, distributing drought resistant seed, and otherwise directly advising farmers so that agricultural output, normally devastated by poor rainfalls, can be leveled out. The mass migrations characteristic of poor rainfall years have ceased and agricultural output has become far less dependent on the vagaries of climatic variability.
The overall problem, then, is to find out when the right conditions for the applications of climate forecasts exist, to determine who the actors are and what motivates them, to discover what freedom of action these actors have and what are the correct actions under each circumstance for each sector, and what measure exists (or can be developed) to prove that the actions taken have indeed been beneficial.
Conclusion
We have described the elements of "end-to-end" forecasting in terms of physical prediction, the assessment of the consequences of climate variability on the normal workings of systems, and the applications of (and evaluations of the consequences of) these forecasts. It is important to emphasize that all three activities are part of end-to-end forecasting and none can be short-circuited. A perception has arisen in some circles that a direct interaction of physical scientists with end-users will be sufficient to lead to useful applications. We believe that, while such a dialog is essential and can be a useful way of starting the process, the ultimate benefits of the applications of seasonal-to-interannual predictions cannot be realized without a complementary effort to more thoroughly understand the human, social, and economic systems within which climate forecasts will be applied. Orchestrating efforts to address all three activities of an end-to-end system may be considered one of the integrating functions of the IRI.
The IRI is being established to demonstrate that end-to-end forecasting is possible and valuable. Implementing the shared vision of the IRI discussed at the November 1995 Forum provides a focus for addressing the scientific, technical and institutional challenges associated with the implementation of an "end-to-end" prediction program like the one described above. Physical and social scientists, governments, national and international organizations, and private-sector interests each have critical roles to play. By working together with the IRI as a focus, we can, as described in the Forum Statement, realize the true value of "collectively undertaking a focused research program to generate products based on experimental seasonal-to-interannual climate predictions and facilitate their application to practical problems in affected regions."
¥ The Pacific ENSO Update is a bulletin of the Pacific El Niño Southern Oscillation (ENSO) Applications Center in Honolulu, Hawaii. The bulletin is intended to supply information on climate variability related to the ENSO climate cycle for the benefit of those involved in such climate-sensitive sectors as civil defense, resource management, and development planning in the various jurisdictions of the U.S.-affiliated Pacific Islands (USAPI).
The Pacific ENSO Update can be accessed through the world wide web at: http://naulu.soest.hawaii.edu
¥ The mission of the Climate Prediction Center/NCEP in Camp Springs, MD USA is to maintain continuous watch on short-term climate fluctuations and to diagnose and predict them. These efforts are designed to assist agencies both inside and outside the federal government in coping with such climate related problems as food supply, energy allocatio, and water resources.
The Center produces an electronic version of their Monthly Climate Diagnostics Bulletin which can found at: http://nic.fb4.noaa.gov
¥ The Environmental Modeling Center (EMC, formerly the Development Division), improves numerical weather, marine and clmate predictions at the National Centers for Environmental Prediction (NCEP, formerly the National Meteorological Center), through a broad program of research in data assimilation and modeling. In support of the NCEP operational forecasting mission, the EMC develops, improves and monitors data assimilation systems and models of the atmosphere, ocean and coupled system, using advancec methods developed internally, as well as cooperatively with scientists from Universities, NOAA Laboratories and other government agencies, and the international scientific community.
Visit the website for the EMC/NCEP at: http://nic.fb4.noaa.gov:8000/
¥ The 8th Global Warming International Conference and Expo, 25-28 May 1997 in New York, NY USA. The focus will be on regional extreme climatic swings; extreme events and the extreme event index (EDI); industrial technology and greenhouse gas emissions; and global and regional natural resource management. For more information, contact the 8th Global Warming Conference & Expo, PO Box 5275, Woodridge, IL 60517-0275 USA. Tel: 630-910-1551. Fax: 630-910-1561.
¥ The International Climate Change Conference and Technologies Exhibition will be held 12-13 June 1997 in Baltimore, Maryland. The conference objectives are to demonstrate to the private and public sectors that the climate change policy process is moving forward in such a manner that will affect the global economy and therefore corporate business practices and to examine issues such as the status of technologies in existence today and technology options for the future. For more information, call Heather Tardel, The International Conference on Climate Change, 2111 Wilson Blvd., Suite 850, Arlington VA 22201 USA. Tel: 703-807-4052. Fax: 703-243-2874
¥ The 1997 Open Meeting of the Human Dimensions of Global Environmental Change Research Community will be held 12-14 June 1997 in Laxenburg, Austria. The intention of the meeting is to bring together the human dimensions research community to promote the exchange of information on current research and teaching and to encourage networking. The organizers hope to attract social scientists, humanists, and others not previously involved in human dimensions work. For more information, contact Claudia Heilig-Staindl, IIASA, A2361, Laxenburg, Austria. Fax: 43-22367131. E-mail: staindl@iiasa.ac.at
