Responding to Climate Predictions
One of the most predictable climatic events is El Niño, the development of a warm pool of ocean water in the eastern and central Pacific Ocean that typically persists for 12 to 18 months and recurs approximately every two to seven years. Weather phenomena associated with El Niño have significant societal, economic, and environmental impacts. Moreover, these phenomena are somewhat predictable on a seasonal time scale and, therefore, present decision makers with opportunities to develop response strategies.
Prior to 1997, the strongest El Niño of record (i.e. largest sea surface temperature anomaly) for this century was recorded in 1982/83. Quite significant weather responses to that El Niño in the United States were observed that resulted in large societal impacts on safety, property, and economic development. The Financial Times (July 28, 1997) reported that the 1982/83 El Niño led to estimated U.S. losses of $2.2 billion with 161 human deaths. Several retrospective scientific investigations, including model research and development, showed some promise in the prediction of the anomalous climate of 1982/83.
In 1997/98, an even stronger El Niño occurred, and several seasonal predictions made prior to the onset of that event appeared to show statistical skill beyond climatological chance. Decision makers acted on those forecasts, taking measures that abated certain hazardous impacts, and enabled decision makers to capitalize on economic opportunities. The 1997/98 El Niño was the largest and warmest to develop in the Pacific Ocean in the past 100 years, and a milestone for seasonal forecasts. Analysis of 1997/98 El Niño impacts on the U.S. economy suggests that climate predictions have great potential for risk management in several climate sensitive industries, as well as in federal, state, and local disaster preparedness. For example, there is strong evidence that California saved over $1 billion in property damages due to better preparation by state and local officials in response to the 1997/98 El Niño forecast.1 Based on that estimate, taking advantage of climate predictions could result in potential savings of billions of dollars annually to the public and private sectors of the U.S. economy.
To cope effectively with climate impacts, institutions and organizations require accurate, timely seasonal and longer-term predictions. However, this is only a starting point. To respond effectively to climate predictions that involve significant seasonal variations, governments also require appropriate policies governing public and private decisions – at the local, national, and international levels. However, in most instances, the needed policies are not in place.
In response to this need, the Atmospheric Policy Program of the American Meteorological Society, in collaboration with Columbia University, developed and convened a forum. The Forum was designed to identify improvements in the development of strategies that benefit from effective application of climate information and seasonal climate predictions. The Forum brought together 96 participants representing weather and climate scientists, specialists in developing decision strategies, and policy makers. The resulting discussions developed findings, policy options, and recommendations necessary to achieve the improvements.
The General Findings
FINDING 1 - Anticipation and response to seasonal climate variability offer potentially significant societal benefits.
Although the accuracy of climate predictions is still limited, they can be applied in positive ways to benefit society. Application of forecasts of the strong 1997/98 El Niño demonstrated that significant societal benefits could result from anticipating and properly responding to predicted seasonal climate variations. Climate services can enhance many economic and societal sectors such as water resource management, agriculture, emergency preparedness, public health, and natural disaster reduction.
FINDING 2 – Regional and societal climate impacts are significant, but not yet fully understood.
The impacts of climate variations on society can be significant, particularly at the regional level. Some of the phenomena that El Niño affect are drought in northern Australia; temperatures and precipitation on the western coast of tropical South America; and tropical and extra-tropical cyclone frequency and intensity, as well as, precipitation quantity along the California and Gulf coasts, throughout the Pacific and in some interior portions of the United States. However, the regional climate impacts with regard to timing, duration, and specific locations are not fully understood.
FINDING 3 - Seasonal forecasts have a limited, but useful, level of skill and with enhanced infrastructure and research this level can be raised substantially.
Limited objective verification has demonstrated that seasonal forecasting has a small level of skill (accuracy above climatological chance). Improvements in observations, models, and understanding of climate variability hold great promise for higher levels of skill in seasonal forecasts. However, a balanced research effort to achieve these improvements is needed, to improve the forecasts and to make better use of the resulting forecast information.
FINDING 4 – Climate predictions and information are not provided in the most useful way to decision makers.
In many cases, climate predictions and information are not well suited for user needs. Much of the terminology associated with climate information is confusing. In addition, the timing requirements of users are not well understood by the information providers. New or improved methods of communicating climate information need to be considered, including formats and content.
Since climate predicting, particularly on a seasonal basis, is relatively new, decision makers need supporting evaluations of success before making major climate-related decisions that directly affect the public or stewardship of public resources (e.g., water supply). Finally, there is little information available on objective evaluations of successful applications of climate services.
For most decision makers and the public, statistical or probabilistic climate information is poorly understood. To be useful for decision makers, climate prediction information should be provided with certainty (or uncertainty) quantified in a clear, understandable presentation or format.
FINDING 5 – Climate information is most effectively developed and applied through partnerships between climate information providers and decision makers.
Use of climate information can be most effectively developed and applied when scientific data and information are placed in a context and format that is a part of the decision maker’s overall management system. Decision maker’s input regarding the most useful climate information, as well as timing and formats, must be a part of the overall development and application of climate prediction and information services.
FINDING 6 – Special training is needed to prepare professionals that can communicate information between the providers and consumers of climate information.
A growing number of application oriented scientists have begun to take on the role of “science integrator” in order to bridge a significant gap between the climate scientist and the consumer of climate information. To successfully gain an understanding of the consumer’s information needs and applications, education and experience in both applications of climate science and the use of analytic tools is necessary. Academic institutions should develop advanced interdisciplinary opportunities to educate students to undertake roles as science integrators.
FINDING 7 – The use of climate predictions by decision makers is limited by a lack of evaluation of the risks and benefits.
The uncertainty inherent in climate predictions naturally leads to probabilistic formulations and therefore, has the potential to be applied to risk management. For many decision makers risk management is one of the most important aspects of their job. Risk management decisions can involve relatively large investments in order to either avoid large future costs or to take advantage of future opportunities that will reap major benefits at low cost. Using good climate predictions can optimize those decisions that are affected by future weather and climate. However, very few assessments or evaluations have been made regarding climate risk management for most decision-making situations.
The evolution of climate prediction services will, in large part, depend on climate scientists working with users to demonstrate the relevance and utility of climate information by drawing on the results of forecast applications and/or retrospective analyses. Once systematic climate information has become part of an overall decision process, it is important to establish a set of performance metrics. The metrics should effectively quantify the impacts and benefits of applying climate information to the process of decision making. Maximum acceptance of the value of climate information will emerge from objective evaluations of applications that have been adequately documented over a sufficient period of time.
The General Recommendations
RECOMMENDATION 1 – The nation should increase investments in climate science research, climate impact assessments, and strengthening the supporting infrastructure to improve climate predictions resulting in significant societal benefits.
RECOMMENDATION 2 – Providers of climate predictions should include clearly defined uncertainty measures (probabilistic information) that are presented with mutually understood terminology.
RECOMMENDATION 3 – Representatives of government agencies, private sector organizations, and academia should establish a collaborative approach to develop and provide a national capability for climate prediction and information services that would foster mutual trust and useful applications of climate information.
RECOMMENDATION 4 –Academic institutions should establish educational programs to produce “science integrators” who understand how to communicate user needs to providers and facilitate the application of climate information for users.
RECOMMENDATION 5 – Climate service providers and decision makers should work together to develop measures of improved performance resulting from using climate information as part of the decision process.
There are specific recommendations regarding policy issues in responding to climate variability and improving the usefulness of climate prediction services detailed in Sections III and IV of this Report.
The Specific Policy Recommendations
1. Government decision makers should integrate the use of climate information into national and international planning. Tools, based on climate data and analysis, should be developed that help integrate climate information and understanding into planning at all government levels.
2. In partnership, the government, academia, and private sector should identify one or more “grand challenges” to improve climate prediction services and applications. These challenges will provide a strategic focus for organizing shared efforts to advance the mainstreaming of climate information in policy formulation in the near- and long-term. An example of such a grand challenge is an overall study to document and report on the impacts of climate variability on society.
3. Governmental and academic institutions should consider increasing educational opportunities for climate scientists to better understand society and climate. Understanding how society reacts to climate and related forecasts is an important undertaking that can help define needed meteorological research as well as improve the benefits from climate prediction services.
4. Government agencies, resource management institutions, and private sector enterprises should aggressively recruit and retain scientists with science integration skills as part of their programs in climate-sensitive sectors. Providers and potential users of climate information should recognize the important role of scientific integration, a relatively new profession. Scientific integrators possess not only knowledge and/or experience in climate science and forecasting, but also an understanding of the processes involved in user decision- making.
5. The government should make balanced investments in research to advance the chances of reaching the theoretical potential in climate predictions to meet national economic and social needs. The investments should be devoted to: enhanced understanding of climate variability; more accurate climate prediction modeling techniques and data assimilation methods; expanded quality and quantity of observations; supporting infrastructure of computers and information systems; and a collaborative modeling focus, both nationally and internationally, among institutions and governments.
6. The U. S. government, in partnership with academia and the private sector, should encourage the World Meteorological Organization (WMO) to provide mechanisms for sharing information on climate services and applications internationally. Ongoing discussions about the emergence of Regional Climate Centres in the WMO provide a special, near-term opportunity. To achieve sustainable development on a global scale, the nations of the world will need to effectively use climate information.
7. The AMS should provide opportunities for continued science-policy dialogues, since climate science policy is a relatively new area. There should be more opportunities for open dialogue that is focused on climate-sensitive sectors. The goals of these dialogues include: identification of critical climate information needs; enhanced understanding of vulnerability; improved assessment methodologies; exploration of response strategies; and increased awareness of current and emerging forecasting capabilities and decision-support tools.
The Specific Recommendations to Improve Climate Prediction Services
1. There should be regular discussions among representatives of the providers of climate services concerning enhancements in those services. In many cases, guidance information widely distributed by the NOAA Climate Prediction Center (CPC) is enhanced significantly and is tailored to be application-specific for users by climate scientists in the private sector and/or academia. By having regular discussions about potential service enhancements, the various providers can continue to build a cooperative partnership.
2. Providers of climate information and forecasts should ensure that the space scales, timescales, and variables of interest to end-users are addressed clearly. It is especially important to convert basic meteorological elements to variables, such as stream flow, that are of interest to end-users.
3. Climate prediction providers should ensure that uncertainty measures are clearly stated to the public and decision makers. The measures should include, at a minimum, information related to probabilities of occurrence of the forecasted variables or quantities, accuracy of recent forecasts, and general physical reasoning behind the forecast.
4. Climate information providers and decision makers should create partnerships that support shared learning (co-production of knowledge) and joint problem solving. These partnerships should be joint, interactive endeavors between participants in climate science/application and in societal decision-making. The partnerships should design and execute pilot projects, and jointly assess the benefits, reliability, and impacts of climate prediction applications in an attempt to “scale-up” the lessons from individual projects to more general, successful applications. In the course of carrying out these projects, the mutual trust necessary to make the partnerships effective will be developed.
Many aspects of the recommendations presented in this Report must be met by the private sector—including corporate decision makers and the value-added meteorological services industry that provides climate information. Other aspects require public sector investments and actions. There are also recommendations that require university actions and research activities. The recommendations developed in this Forum, although directed, for the most part, at specific portions of these interested communities, can be best implemented through cooperative efforts among those communities.
These cooperative efforts, therefore, require effective public-private-academic sector partnerships. As noted in Section III, the Forum participants recommended that the AMS should be proactive in organizing dialogues that could foster the development of effective partnerships between the climate information providers and users. The AMS is prepared to serve in that role.
Implementation of the Forum recommendations would enable the development of effective responses to climate variations, within the United States, as well as internationally.
1 “Improving El Nino Forecasting: The Potential Benefits,” ed. Rodney F. Weiher, U.S. Department of Commerce, National Oceanic and Atmospheric Administration, August 1999.