In January 2020, the U.S. National Climate Service’s annual outlook predicted much higher than normal probabilities for substantially reduced precipitation over the agricultural southeastern United States, as well as a continuation of the drought over the far West. The grain belt throughout the Midwest was projected to have near normal to above normal rainfall. The latest climate outlook from the European Climate Center agreed in general with the U.S. projection. The outlooks also predicted increased probabilities of much warmer than normal temperatures along the East Coast. Both the U.S. and European outlooks included a range of possible conditions and described the basis for the level of uncertainty. The outlooks included a historical context derived from the mining of the data archives. Analyses of the historical duration of similar past climatic events, their impacts, and successful mitigation strategies were incorporated as part of the risk evaluation process.
The annual outlooks serve as the basis for a wide array of planning activities and decision making. The utilities industries expanded their ability to support the increased power requirements of the elevated air conditioning demand that the summer was expected to bring and used a model of the distribution and timing of energy demand to purchase fuel during periods of anticipated reduced demand and hence lower price. The Department of Agriculture evaluated crop
projections with respect to overall food demands by the United States and the nation’s ability to support exports to other countries. With the expectation that drought throughout much of Africa would continue, the Department of State and the Department of Defense examined options for assistance that would be needed if social stresses increased throughout most of the continent. The Federal Emergency Management Agency (FEMA) issued warnings in regions prone to increased weather risk with the goals of improving mitigation efforts and increasing public awareness. U.S. industries guided the manufacture, marketing, and distribution of weather-sensitive products from clothing to home improvement goods and services. Travel agencies and the transportation industry planned for the expected demand for travel to specific winter and summer vacation destinations.
The hypothetical scenario above represents a potential for climate services two decades into the future if aspects of current capabilities and agency responsibilities are simply extrapolated forward in time. It illustrates the importance and impact of such a service, without even considering major breakthroughs in science, technology, or information management. The timely delivery of useful products through direct and accessible user interfaces can maximize the societal benefits and limit national risks. It represents directed efforts to translate knowledge of the climate system into a national service function.
Climate is an increasingly important element of the public and private decision-making process. Advances in monitoring and predicting variations in climate, coupled with growing concern over the potential for climate change and its impact, are yielding an increased awareness of the importance of climate information for enhancing economic vitality, maintaining environmental quality, and limiting threats to life and property (Changnon 2000). The importance of climate information has fostered the concept of a “climate service”.
Climate has a local- to regional-scale component, and information and expertise are needed to address these space scales. For example, the climate of urban St. Louis is markedly different from that of rural Missouri, and the climate of the Midwest is strikingly different from that of the High Plains and the Northeast. Most activities impacted by climate operate at such local to regional scales and so need information for those areas. To serve the many agricultural interests in the Corn Belt, the Midwestern Regional Climate Center
has developed an operational on-line climate-crop yield model. Users can call in any time during the corn (or soybean) growing season and get up-to-date information on the status of climate conditions for their area of concern—say, southeastern Indiana. They can then examine a variety of climatologically-based choices for weather conditions during the remainder of the growing season. And finally they can obtain crop-yield estimates based on what has already occurred and what might occur.
Historically, climate services have revolved around the statistical analysis of existing weather records. This information formed a useful basis for estimating future agricultural production and energy needs. Information on extreme events, such as “100-year floods” and peak wind speeds, continues to be critical in the design and construction of major facilities such as dams, highways, and buildings. In 1922, for example, the division of water allocations between states in the Colorado River Basin (the Colorado River Compact) was based on historical observations for a time of anomalously high flow, compared with later years. The allocations were based on a ten-year period to even out natural variability but could not incorporate longer time-scale variability. The Mexican Water Treaty of 1944 controls the allowable salinity and volume of water entering Mexico from the United States. The system of dams constructed along the Colorado River allows storage of four times the average annual flow as a hedge against low-flow years and as a means to maintain treaty requirements. Historical climate records are also used to set “safe” levels of water storage and prevent emergency water releases that would cause downstream flooding (e.g., Rhodes et al. 1984). The availability of the various sources of climate information and products summarized in this example has historically been supported through many federal, state, and local agencies and represented a climate service to the nation. In recognition of the growing importance of this service, the National Oceanic and Atmospheric Administration’s (NOAA) National Climatic Data Center (NCDC) in Asheville, North Carolina, was founded in 1951 as the U.S. Weather Records Center.
Today, however, the science and understanding of global and regional climate have gone well beyond a statistical analysis of historical records. Research efforts of the last two decades have resulted in a substantial improvement in our understanding of short-term climatic fluctuations such as El Niño and La Niña (NRC 1996). A number of groups around the world now regularly forecast aspects of the El Niño/Southern Oscillation (ENSO) phenomenon and its associated impacts. In many regions, the impacts of ENSO are increasingly anticipated by decision makers as a result of forecasts that are
taken seriously and serve as input for decision strategies and choices (Changnon 2000).
In April 1997, record flooding along the Red River of the North and in Grand Forks, North Dakota, resulted in $2 billion in damages. The magnitude of the damages was especially surprising given that the record flood was anticipated a season in advance. In February 1997, the U.S. National Weather Service issued a flood outlook for Grand Forks for the coming spring flood season warning residents to expect a flood of record, 49 feet, sometime in mid-April. While the 3-month outlook was within 10 percent of the actual flood crest—highly skillful by any measure of predictive accuracy—neither forecasters nor local residents fully understood the degree of uncertainty in the flood outlook (e.g., as to the timing and height of the water crest, the duration of the highest water, and the area covered by the flood). Consequently, preparations to fight the flood placed an improper precision on the 49-foot outlook, leading to surprise and then anger when the river’s crest exceeded that level (Hooke and Pielke 2000).
The Grand Forks experience and many similar experiences send a two-pronged message. On the one hand, they provide optimism that forecasters can produce skillful predictions and other information on seasonal (and longer) time scales. On the other hand, they suggest that the effective use of the information is not always straightforward. As knowledge of climate and the use of climate information have developed, both the challenge and promise of providing operational “climate services” have become readily apparent. It is time to focus on the challenge of providing climate services to realize the promise that they hold.
The combination of historical observations, paleoclimate data, and efforts to develop a hierarchy of coupled models is increasing the understanding of the causes and character of climate fluctuations, although the nature of decade to century climate variability remains an important topic of research (NRC 1995, 1998c). From such research, a growing appreciation has emerged that “climate” can no longer be considered stationary—climate changes, sometimes dramatically, over periods of years to decades. Within the last two decades, concern about the potential of human-induced changes in the earth’s climate has resulted in the development of new methods for observing the earth system and in better understanding of the possible courses of future climate. These developments and advances are in large part the result of core research in the atmospheric, oceanic, and biological sciences and are the research foci of the U.S. Global Change Research Program (USGCRP 2000). The increased
understanding of the climate system has provided greater knowledge of the variations recorded by historical measurements. The tools developed for experimental long-lead-time forecasts (a month to a year) now permit experimental predictions of seasonal to interannual climate patterns and a limited but growing capability to project future climate changes. The societal value of such information depends on many factors, including the following:
The strength and nature of the linkages between climate, weather, and human endeavors.
The nature of the uncertainties associated with climate forecasts.
The accessibility of credible and useful climate information by decision makers.
The ability of users and providers to identify each other’s needs and limitations.
The ability of users to respond to useful information.
Increasing realization of the importance of climate is stimulating user demand for improved information and substantially broadening the scope of climate services. Because of these factors, the subject of climate services was an agenda item at the fall 1999 meeting of the Board on Atmospheric Sciences and Climate (BASC), held jointly with the Federal Committee for Meteorological Services and Supporting Research (FCMSSR). Subsequent to the meeting, the federal agencies, through FCMSSR, asked BASC to review the status of climate services and to recommend directions for the future provision of climate services to the nation. In particular, BASC was asked to address the following items outlined in the statement of task:
Define climate services.
Describe potential audiences and providers of climate services.
Describe the types of products that should be provided through a climate service.
Outline the roles of the public, private, and academic sectors in a climate service.
Describe fundamental principles that should be followed in the provision of climate services.
This report summarizes BASC’s response to the task statement. The board’s efforts included a discussion of potential topics at its fall 1999 meeting; organizational discussions and preliminary briefings at its spring 2000 meeting;
an information-gathering activity during a workshop held August 8–12, 2000, in Woods Hole, Massachusetts; and additional discussion and review of materials at the fall 2000 meeting that followed the workshop.
CLIMATE SERVICES DEFINITION
Unlike climate services, the concept of a “weather service” is familiar to most citizens. A weather service focuses on the description, analysis, and forecasting of atmospheric motions and phenomena on very short times scales, extending up to a period of one week to ten days. The objective is to provide forecasts of continually changing weather and warnings of severe weather events. The benefits of this service are measured in lives saved, injuries avoided, or reduction in property damage. The National Research Council (1998a) has noted that this service is provided by a four-way partnership in which (1) the government acquires and analyzes observations and issues forecasts and warnings, (2) the government, newspapers, radio, and television participate in the dissemination of weather forecasts and severe-weather warnings, (3) private meteorological firms use government data and products to provide weather information for the media and special weather services for a variety of industries and activities, and (4) government, university, and private-sector scientists work to develop improved understanding of atmospheric behavior and turn it into new capabilities.
In contrast to weather, climate is concerned with the longer-term statistical properties of the atmosphere—ocean—ice-land system. Climate refers to the many statistical properties of such variables as temperature and precipitation over a specific region, the range of values of these variables, and the frequency of particular events as a function of location, season, or time of day. Climate variability and change are products of external factors, such as the sun; complex interactions involving the different components of the earth system; and human-induced (or anthropogenic) changes to the earth system. Following the World Climate Research Programme, BASC (NRC 1998a) adopted three general categories of climate variability and change: seasonal to interannual variability, decadal to century climate variability, and changes in climate induced by human activities, such as emission of greenhouse gases. A climate service must focus on very different types of activities to address all of these major categories of variability and change. Each of the activities is associated with different types of users or decision makers and with different types of needs and products, as is evident by the current use of climate information.
The earliest climate services consisted of descriptive information and statistical analyses of weather observations. The statistical analyses of the weather data evolved in response to the needs of engineers, designers, and insurers. Seasonal to interannual forecasts developed over the last few decades have generated additional demands for information and products. The value of the information is a function of the ability of the various sectors (e.g., agriculture, water, energy, transportation, and health care) to use climate information a season to a year in advance. The assessments of long-term potential changes in the global and regional climates as a result of anthropogenic factors add to the demands and variety of potential climate services. This information is being used by national and international policy makers and in a variety of long-range planning activities.
BASC reviewed the evolution of climate services in the nation (see Chapter 2 ) and examined the various definitions of climate services that have been used in the past. In reviewing the breadth of useful climate information—from historical analyses to century-scale predictions—BASC chose to be inclusive by adopting the following definition of climate services:
The timely production and delivery of useful climate data, information, and knowledge to decision makers.5
EXAMPLES DEMONSTRATING THE BREADTH AND DEMAND6 FOR CLIMATE SERVICES
The demand for and use of climate services has grown substantially over time. For example, NCDC now provides 120 major data sets and products online that describe a variety of climate elements. Since the initiation of the online data system in January 1998, orders have steadily increased to a rate of 1,800–2,000 requests per month. A profile of off-line orders reveals a customer base of about 77 percent business, 13 percent government, and 7 percent individuals (Karl 2000). Products of the NOAA Climate Prediction Center (CPC) include both extended weather outlooks and climate information. The most popular of the climate products include an ENSO Diagnostic Advisory, 30- and 90- Day Outlooks, long-lead-time (out to 13 months) outlooks, a U.S. Drought Assessment, and an Atlantic Hurricane Outlook. Users of NOAA’s specialized data products include FEMA; the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE); the Nuclear Regulatory Commission; and the National Association of Home Builders (NAHB). For example, NAHB uses temperature data when revising building codes. The air-freezing index can be used in conjunction with soil properties and surface ground cover to compute the ground freezing potential for particular areas and guide codes on building foundations. According to a U.S. Department of Housing and Urban Development study (HUD 1993), the revised building codes could result in annual savings of $300 million in building costs. ASHRAE is using NCDC temperature, humidity, wind speed, and solar insolation data with the goal of reducing user energy consumption by 50 percent. The U.S. Navy now uses monthly and seasonal climate outlooks to control the large heating systems on its installations, resulting in a savings of $65 million in utility costs (Cuff 2000). Previously, the dates for turning on the
heating and cooling systems were determined using the average climate conditions for a particular region.
The practical use of climate information extends well beyond the long lead time (13 months) of the NOAA products described above. For example, the Navy has entrained long-term climate-model projections with Arctic sea ice models in its long-range planning for fleet appropriations and construction. Other enterprises with long lead times—including construction of offshore drilling rigs, planting of timber, and development of transportation options— have also included climate change projections in their decision making.
A wide variety of cases demonstrate positive outcomes or reduced risk when climate information is incorporated into the decision-making process. However, because a comprehensive survey of such cases is not possible, this report cannot assess the magnitude of the value and importance of climate services. Instead, four examples have been selected from those investigated by BASC to illustrate of the breadth and character of the growing demand for climate services. The examples were selected to show the expansion of climate services from statistical analyses to more sophisticated products, including prediction as well as the range of partnerships involved in providing and accessing climate services. The first two examples, (1) coastal exposure and community protection from severe storms and (2) weather derivatives, describe services built directly on analyses of historical observations but for which there is a growing demand for enhanced predictive capability. The third example, based on ENSO forecasts, directly examines the importance of seasonal to interannual forecasts. The fourth example, based on the recent U.S. assessment of future climate impacts, considers the value of longer-term climate projections. In addition to showing both the breadth of and growing demand for climate services, the examples guide several of the recommendations in this report. Many more examples could be cited as well, involving, for example, waterresource management, water quality—climate connections, droughts and agriculture, and human health.
1. Climate information for the control of risks in coastal communities. Historical averages of weather data have been used in many applications throughout the nation. For example, the historical record shows that more hurricanes hit southern Florida than any other part of the country, prompting that region to write the nation’s most stringent building codes. However, Hurricane Andrew, which crossed southern Florida in 1992 as a powerful category 4 storm, dramatically changed the risk and vulnerability perceptions of
residents, governments, builders, and insurance companies. FEMA’s Federal Insurance Administration has identified outdated codes and a wide variety of inadequacies in construction, code enforcement, inspection, and training. Insurance industry experts estimated that up to 40 percent of the insured damages could have been avoided if building codes that were developed using more accurate climate information had been enforced (Pielke and Pielke 1997). The fact that damages were much greater than expected (an estimated $15.5 billion in insured damages and total damages about twice that) has also resulted in revised estimates of coastal damage potential, with numerous eastern seaboard cities’ insured-loss projections exceeding $50 billion for a category 4 or 5 hurricane (Insurance Research Council 1995). Substantial savings can be realized if building codes that take improved climate data into account are adopted by builders and enforced. However, an ability to anticipate risk and vulnerability also has potential value.
Both societal exposure (coastal population growth and wealth) and the climatological understanding of the frequency and intensity of weather hazards are of major importance to the insurance industry (e.g., Munich Re 1994, Berz 1993, Dlugolecki 1992). The industry cannot provide coverage blindly and therefore must base the price and scope of underwriting on an assessment of potential liability. The potential liability of the industry has typically been based on a statistical analysis of past events to determine the probability of future events and on a geographic assessment of a variety of various liability categories. Those factors are combined to determine loss ratios. Both likely loss and probable maximal loss are important. Over the last several decades there has been significant growth in insured losses. Typically, changes in premiums and in scope of underwriting do not occur until after a substantial payment of claims. However, the increased insured losses have prompted the insurance industry to examine all aspects of its liability, including those dealing with changing climate patterns—either natural variability or longer-term climate change. The key issues identified by the industry are how weather changes with seasonal to decadal variations, whether climate change will be associated with new extremes, and whether changes in weather patterns due to seasonal to decadal variability or climate change will result in changes in the geographic location of exposures or the intensity or frequency of events. Hence, the role of climate services has been extended from a statistical analysis of the historical record to one in which understanding of climate variability and change in the context of population and demographic shifts becomes of substantial importance to governments and industry. The insurance industry, particularly the
reinsurance portion of the industry, seeks better climate information and outlooks so it can better manage its exposure. In addition, the insurance industry has established, as a priority, programs to mitigate the loss of life and property. Improved climate information assists the industry in planning for those important activities.
2. Weather Derivatives. Weather derivatives are financial instruments designed to allow businesses sensitive to the vagaries of weather to protect themselves against loss from changes in costs and sales linked to variations in climate. They gained in popularity during the winter of 1998–99 when long-range forecasts called for warmer than normal weather during that period (due to the onset of El Niño). Weather derivatives can theoretically be designed for almost any weather variable (e.g., rain, snow, and wind), although most of the activity so far has been based on temperature. They are quite different from insurance in that insurance is based on the willingness of individuals or corporations to pay a premium to transfer risk. The premium is based on weather statistics and a potential damage function, and the payout from the insurance reflects actual damages incurred. However, for many enterprises weather risk is defined by its impact on performance rather than on loss of property (e.g., impact on production rates, product availability, pricing, and product demand). The number of industries whose performance is impacted by weather and climate is large (WeatherRisk 2000). For example, moisture availability is a key factor in agriculture; under adverse growing conditions, the largest corn-producing counties in the nation can experience 30 percent reductions in yield. Agrochemical industries are also impacted; for example, weevils cost cotton growers $300 million annually, and weevil severity is directly correlated with mild winter conditions. Viticulture has a strong climatological signal, with a lack of sunshine and cool temperatures from pre-bloom to maturation producing poor grape years and excess precipitation causing grape rot. The 1998 California grape harvest was 30 percent lower than the norm because of poor weather conditions. Beverage consumption and ice-cream sales are strongly correlated with warm temperatures. Clothing retailers are recognizing correlations between summer and winter sales volume and the magnitude of the departure from normal weather conditions. Amusement parks’ profit margins are defined by peak attendance days associated with good weather. The construction industry often experiences penalties of 10–15 percent of job costs for delays related to, for example, temperatures needed for concrete to set,
wind conditions, or rain-freeze events. Weather derivatives are designed to address such examples of climatic impacts on performance.
The key to the growth of a weather derivatives market is that weather risk has differential impacts in terms of costs and benefits—a two-way flow of risk promotes the concept of trading risk. High snow accumulation aids a ski resort by increasing profits, but it substantially increases the cost of snow removal in cities. Hot summer conditions can increase power demand by a factor of 20 at greatly increased costs. Heat stresses livestock and agriculture, but it aids the beverage, ice cream, and summer clothing industries. Large fluctuations in performance associated with weather and the differential nature of impacts are fostering a weather derivatives industry. Weather derivatives add stability because during optimal conditions profits are high, so paying for the derivatives is not a burden. During poor conditions, the payout from derivatives helps to maintain cash flow and profitability, so risk is managed.
The first transaction in the weather derivatives market took place in 1997. In 2000, between 2,500 and 3,000 weather-related contracts were issued in the United States, with a total value of $2.5 billion (PricewaterhouseCoopers 2001). The Weather Risk Management Association commissioned a recently completed survey by PricewaterhouseCoopers to determine the exact state of the weather derivatives industry (see <http://www.wrma.org>). EnronOnline (<http://www.enrononline.com>), an Internet weather derivatives leader, currently has a contract portfolio of $5 billion. The largest economic sector involved is energy because of the strong and obvious links between temperature extremes and energy demand, but both the number and scope of companies have been increasing since 1997 (including such new fields as leisure and home improvement). Some changes that are occurring in climate services are helping the growth of the weather derivatives industry. These include the availability of NCDC databases; the formation of private climate service companies that provide clean, standardized data directly for weather risk assessment at a large number of locations; the development of stochastic modeling tools as an aid in risk analysis; and the development of operational seasonal forecasts at the NOAA National Centers for Environmental Prediction (NCEP), particularly involving ENSO (WeatherRisk 2000).
3. The applications of regional ENSO forecasts. The 1925–26, 1972–73, and 1982–83 El Niños, which resulted in dramatic changes in rainfall and flooding, are examples of short-term climate events that had substantial societal impacts. The magnitude of ENSO and its impacts, a growing observational
and modeling capability, and a desire to better understand the role of the tropics in governing climate variability resulted in the development of the 10-year Tropical Ocean Global Atmosphere program. Important outcomes of that effort have been the development of a tropical observing system in the Pacific Ocean and some demonstrated skill in producing El Niño forecasts. NCEP’s CPC is now experimenting with the use of coupled ocean-atmosphere forecasts, together with several statistical techniques, to arrive at an official consensus El Niño forecast with a lead time of two seasons. The scientific community is far from demonstrating skill in predicting all the features of ENSO (e.g., onset, duration, amplitude, and decay). However, once the El Niño is under way, the tendency for it to persist into the boreal winter results in substantial capability to anticipate and plan for many of its effects (Landsea and Knaff 2000).
Such investigations are just beginning in the United States, but such El Niño events as those of 1997–98 provide the opportunity to demonstrate the potential value of an increased ability to anticipate climate phenomena. The CPC correctly forecasted six months in advance the increased precipitation and storms across California, the southern plains, and the Southeast and a warmer winter in the northern tier states. An analysis of the losses and benefits associated with the 1997–98 event is instructive (Changnon 1999). In response to the 1997 climate predictions and on the basis of knowledge gleaned from the 1982–83 El Niño, FEMA initiated several mitigation activities, particularly in California and Florida. For example, California spent $7.5 million on flooding-landslide preparedness efforts and alerting the public, and the roofing and home repair industry reported $125 million in completed mitigation activities (Changnon 1999). The U.S. Geological Survey and the Bureau of Reclamation used the forecasts to guide water management in western systems. A number of northern utility companies used the forecasts to guide oil and natural gas purchases. For example, three utilities in Iowa and Michigan saved $39, $48, and $147 million through fuel purchase planning (Changnon 1999). Although it is difficult to estimate, 161 fatalities and $2.2 billion in losses were attributed to the 1997–98 El Niño in the United States (Changnon 1999). Of note, the losses in California were estimated (Changnon 1999) to be half those of the 1982–83 El Niño, suggesting that the extensive mitigation efforts were extremely beneficial. The impacts of ENSO variability and the potential savings resulting from the effective use of forecasts of ENSO events cover many areas, including agricultural production, coastal fisheries, emergency preparedness, systems for early warning of droughts, strategies for hydro-
electric power generation, insurance concerns, and mitigation of forest fires (NRC 1996).
4. International and national assessments of the impacts of future climate change. The Intergovernmental Panel on Climate Change (IPCC) was established by the World Meteorological Organization and the United Nations Environment Programme to assess the science of climate change. The IPCC is charged with generating an assessment of the state of knowledge of climate change every five years. A critical product of this process is the Summary for Policymakers, which details the elements of the assessment deemed most significant for national and international decision makers. The IPCC has become a standard by which governments assess the implications of climate change, and IPCC reports underpin international negotiations on climate change and carbon emissions. In the Summary for Policymakers, the IPCC identifies the importance of the maintenance and improvement of the global observing system, the monitoring of key climate elements, the incorporation of prehistorical data into the examination of the climate record, the development of comprehensive climate models, and focused process studies as the foundation for advice to policy makers (IPCC 2001). The United States, as directed by the Global Change Research Act of 1990, has also undertaken a National Assessment of Potential Consequences of Climate Variability and Change, a landmark effort to examine the potential consequences for the regions of the nation, including coastal regions, and for forestry, water, agriculture, and human health. The historical record and the results from global climate models act as the foundation for stakeholders to assess the implications of climate change and then to consider adaptation strategies (USGCRP 2001).
Unlike earlier reports, the National Assessment does not analyze the potential for climate change. Instead, climate information is the basis for examining consequences of and vulnerabilities to climate change. A number of the consequences have important economic and quality-of-life implications. Climate modeling and analysis are the foundation for developing climate scenarios that describe alternative futures for analysis of potential consequences of climate change, potential adaptation options, and ultimately the vulnerability of communities, institutions, sectors, and regions. In addition, there is a call for greater emphasis on sustained quality-controlled observations. The assessment process reflects a growing need by decision makers and stakeholders for future climate information. In this important context, the climate observing
system and the national capability to provide nature climate projections represent a climate service to the nation (NRC 1998b, 2001b; USGCRP 2000).
Climate information is increasingly important to the decision-making needs of a wide variety of users. The range of applications is enormous and indicates the impact of climate services. Applications include such diverse entities as water set-asides for instream flows to protect ecosystems, state and national water compacts, building codes, insurance premiums, irrigation and power production decisions, beverage consumption rates, retail clothing volumes, and construction schedules. The applications are also growing with the increased understanding of how climate influences human endeavors. That a weather derivatives industry, designed to manage risk associated with climate variability, is growing rapidly and expanding to include diverse elements of commerce and industry is indicative of the importance of climate services to the nation. The extension of climate capabilities from a relatively straightforward, statistical analysis of historical observations to seasonal and interannual forecasts and to century-scale projections has enabled a broader set of applications, including enhancing productivity in weather-sensitive industries, managing weather risk, protecting life and property, and negotiating international treaties.
Public, private, and academic sectors have all played important roles in the extension. Much of the innovation for individual industry and economic sector use has been developed in the private sector. The academic community continues to improve understanding of climate variability and predictability through its research. The public sector has improved the accessibility of its data and information, including the output from the extensive forecast models now being run. Further, the scientific and user communities are increasingly articulating the improvements required in climate services to enable improvements in decision making.