Integration of Social Science & Atmospheric Research
All economic sectors, regions, and individuals are affected by weather and, thus, weather forecasts. Both improved weather forecasts and improved use of current forecasts can enhance personal safety, reduce property damage, and increase economic efficiency, saving multiple lives and millions of dollars each year. In order to realize the potential benefits associated with improved weather forecasts, we need to understand how individuals and socioeconomic sectors do and could use different types of weather information. However, few assessments of the benefits of weather information have been performed, and much of the knowledge available on the use and value of weather information has generally been difficult to locate and utilize.
To address this need, the National Center for Atmospheric Research, with funding from the U.S. Weather Research Program, established the Collaborative Program on the Societal Impacts and Economic Benefits of Weather Information, better known as the Societal Impacts Program (SIP), to create a dedicated focal point for assembling, coordinating, developing, and synthesizing research and information on the societal impacts and economic benefits of weather information.
SIP's overarching goal is to help society benefit from current and emerging weather forecasting capabilities by integrating social sciences knowledge and methods into the weather research and policy-making communities. More specifically, the program conducts research, develops infrastructure and outreach programs, and leads workshops aimed at developing and synthesizing knowledge on the use and value of current and improved weather information; building a community of researchers and practitioners engaged in developing knowledge on societal aspects of weather information; and applying the knowledge developed to enhance weather policy-making, weather research, weather information development, and weather information use.
The goal of most weather forecasting activities is to provide useful information for decision making, benefiting society. Yet weather products and services sometimes fail to have the intended impacts, as Hurricane Katrina illustrates. To help the weather community enhance the utility of weather forecasts, SIP conducts research to understand and improve how people interpret and use weather forecasts in both everyday and high–impact weather situations.
Uncertainty is an unavoidable part of weather forecasting, and thus of forecast communication, interpretation and use. Yet communicating forecast uncertainty effectively – in ways that enhance decision making and societal benefit ‐ remains challenging. SIP is addressing this understudied area by including examination of forecast uncertainty in the research on forecast interpretation and use, as well as by conducting research focused on investigating how people communicate, perceive, interpret, and use weather forecast uncertainty information.
Although weather and weather forecasts are known to have significant economic effects, little is known about the economic value of weather impacts and forecasts. Such knowledge is important both for quantifying economic impacts and for aiding decisions about allocation of societal resources for forecasting. To fill this gap, SIP research is examining the economics of weather impacts and forecasts both across the U.S. economy and in specific high–impact weather situations.
The weather community currently lacks specific knowledge about the societal relevance of weather products and services, hindering product and service improvements. For example, most current metrics for forecast verification focus primarily on monitoring forecast performance from a meteorological perspective and thus guide forecast providers towards improvements that may not provide the desired benefits to users. Over the next decade, SIP aims to build a program addressing this area through research evaluating the societal relevance of weather services and developing methodological infrastructure for user–relevant verification of forecast products.
Weather and Society * Integrated Studies (WAS*IS) is a capacity building effort to fully integrate social science into meteorological research and practice by:
- Building an interdisciplinary community of practitioners, researchers, and stakeholders who are dedicated to this vision
- Providing this community with ways to learn basic tools and social science concepts for conducting integrated work
The effort is primarily conducted through week–long workshops held during the summer in Boulder each year, as well as additional workshops help in other locations from time to time. In addition to conducting capacity building workshops, developing and suststaining WAS*IS–related research efforts, and planning ongoing educational activities, the SIP is partially supporting development of an edited volume of WAS*IS research projects, funded primarily from the NCAR Weather and Climate Impacts Assessment Science Program. This collection will highlight the methods, results, and cooperative efforts of successful integrated weather and social science projects. It has the potential to be used as part of undergraduate and graduate–level courses (e.g., in atmospheric science, geography, hydrology, environmental science) and as a reference for scientists and practitioners to apply in their own work. The compendium is slated for completion in FY09.
Weather and Society Watch is a newsletter published quarterly by SIP since October 2006. The purpose of Weather and Society Watch is to provide a forum for those interested in the societal impacts of weather and weather forecasting to discuss and debate relevant issues, ask questions, and stimulate perspective. The newsletter is intended to serve as a vehicle for building a stronger, more informed societal impacts community by connecting a diverse group of researchers, policymakers, meteorologists, emergency managers and other professionals and interested members of the general public. Content ranges from a guest editorial for each edition to research findings and ideas to historical perspectives, program updates, and conference and job announcements. The newsletter is published in print, PDF and online formats quarterly in October, January, April and July of each fiscal year. Interested readers can view the newsletter online and subscribe to it by visiting http://www.sip.ucar.edu/news/current.jsp.
The Extreme Weather Sourcebook is a collection of historical monetary loss data on severe weather events and presents a summary of damage suffered from hurricanes, floods, and tornadoes in the United States and its territories. The goal of the Web site is to educate viewers on the economic impacts of severe weather events and stimulate interest in the societal impacts of weather. Loss totals by state are presented alphabetically and by monetary rank with data adjusted for both inflation and inflation and wealth. The Sourcebook also displays aggregate monetary loss information for hurricanes, floods and tornadoes, as well as information on fatalities, casualties, injuries and damages for severe weather events such as lightning, hail and wind. Interested viewers can visit the Sourcebook at http://www.sip.ucar.edu/sourcebook/.
SIP and the NOAA Earth Systems Research Lab established a monthly seminar series focused broadly around societal impacts of weather and climate. This seminar series is meant to leverage the large hydrometeorological community in the Boulder–Fort Collins–Denver area and to build capacity among local physical scientists (meteorologists, hydrologists, etc.), social scientists, and stakeholders. The seminar series hosts monthly seminars focusing on topics such as communicating forecast uncertainty; visual communication in meteorology and hydrology; weather, climate and dengue fever; the recent decline in Arctic sea–ice; and sources, perceptions, uses, and values of weather forecasts. The seminars are also being broadcast in Webinar format so that non–local NOAA/NWS and other personnel can attend virtually.
In late 2003 NOAA and SIP formed the ad hoc HFSEWG. The group's goal is to identify social science research capabilities, needs, and priorities for the hurricane forecast and warning system. The ultimate objective is to recommend research initiatives and projects that can be supported through interagency cooperation, funding for public–and private–sector academic and commercial research enterprises, and partnerships with private–sector information consumers. White papers developed in support of the HFSEWG 2005 Pomona, CA workshop were published as a special issue of Natural Hazards Review in August 2007. A summary paper based on this effort was accepted at the Bulletin of the American Meteorological Society as an Inbox article.
In FY07 SIP staff developed an initial Economic Primer for Evaluating Hydrometeorological Products and Services, funded in part by NOAA's International Activities Office (IAO), elucidating economic theory and methods for application to hydrometeorological research and analysis for non–economics. This included a set of case studies. In FY09, SIP staff will continue to work with NOAA IAO staff to revise and update the Primer and arrange for publication of the document as a NOAA contribution to GEOSS. Interested viewers can see the economics primer in its entirety at http://www.sip.ucar.edu/primer.jsp.
Inspired by WAS*IS, the objectives of this workshop are to
- Build capacity among forecasters, media, emergency managers, and social scientists at the local National Weather Service (NWS) Weather Forecast Office (WFO)/County Warning Area (CWA) level
- Introduce integrated meteorology–social science research and ideas to forecasters, media, and emergency managers to help them better serve their users
The first workshop of this kind will be held for the Kansas City/Pleasant Hill CWA, as the idea originated from their Warning Coordination Meteorologist and associated funding is coming from local resources. However, this workshop will be conducted with the intent that it may serve as a model to be replicated at other WFOs.
SIP-funded primary research:
This project examines the sensitivity of state–level economic sector output to weather variability, using 24 years of economic data and 70 years of historical weather observations. To estimate sectoral sensitivity to weather impacts, such as temperature and precipitation, we use a transcendental logarithmic production function. We identify states more sensitive to weather impacts and rank the 11 non-governmental sectors based on their degree of sensitivity to weather variability. The aggregate dollar amount of variation in U.S. economic activity attributable to weather variability using 70 years of historical weather observations in calculated, finding that economic output varies by about $470 billion a year of 2007 gross domestic product.
Communication of Forecast Uncertainty (CoFU) / Forecast Sources, Perceptions, Uses, and Values (SPUV)
The goal of the CoFU/SPUV research effort is to support the effective provision of weather forecasts, including those incorporating uncertainty information, by collecting and analyzing primary data on the U.S. public's attitudes toward forecasts and forecast uncertainty information. More specifically, the CoFU study explores respondents'
- Uncertainty–related interpretations of a deterministic forecast
- Confidence in different types of forecasts
- Interpretations of probability of precipitation forecasts
- Preferences for receiving deterministic forecasts versus those expressing uncertainty information
- Preferences among formats for conveying forecast uncertainty
The results from the study provide insight into where people obtain forecasts, how they perceive and use forecasts in general and uncertainty forecasts in particular, and their interpretations and preferences related to uncertainty forecasts. This knowledge can help the NWS and other weather forecast providers better design forecast products for the public.
This project builds on the SPUV/CoFU work to examine how people's attitudes and behaviors for weather forecast information vary based on their socio–demographic characteristics, their experiences with weather and weather forecasts based on where they live, and their responses to other questions from the survey. The socio–demographic characteristics of the survey sample are comparable to the U.S. population, and there is representative geographic distribution with responses from every U.S. state. To provide data on respondents' experiences with weather and weather forecasts, respondents were matched with NWS climatological data and forecast verification measures based on their reported location. Statistical analysis is then performed to analyze how respondents' experiences, socio–demographic characteristics, and other factors relate to their responses to the SPUV/CoFU survey questions.
An important component of providing useful weather forecast information is understanding how people interpret uncertainty in weather forecasts and how forecast uncertainty affects people's weather–related decisions. The CoFU/SPUV survey included several questions in which respondents were asked to use temperature or precipitation forecasts to make hypothetical decisions to protect or not protect from a potential frost or flood. The latter decision–making scenarios include evaluation of respondents' understanding and use of quantitative precipitation forecasts (QPF) and probabilistic QPF (pQPF). The protection component of the scenario involves monetary costs, and the impact component (flood or frost) involves monetary losses. For each scenario, respondents were given deterministic forecasts and forecasts that conveyed uncertainty in different ways, and they were asked what decision they would make with the different information. These scenario questions are similar to experimental economics approaches that empirically assess how individuals use information. The analysis uses an expected value framework to examine the decisions individuals made with the different information. The results provide information about respondents' understanding of forecast uncertainty information, inferences of forecast uncertainty, and ability to use uncertainty information in decision making.
Storm Data aims to identify and evaluate how National Weather Service (NWS) employees enter damage data into the Storm Data database in order to better understand the Storm Data process, provide recommendations to improve NWS Storm Data training, and ultimately improve the quality, consistency reliability, and usability of NWS damage data. The Storm Data survey is divided into two parts, the first of which was deployed in July 2008 and focused on NWS Weather Forecast Offices' (WFO) perceptions of the data collection and entry process as a whole and asked each WFO what changes could be made to improve the quality of loss estimates, their training from NWS, and their access to useful resources. The second part of the survey, Part B, was implemented in December 2008 and surveyed NWS personnel who entered damage data into Storm Data for a specific weather event in the preceding year. This part of the survey asked about the sources, types of losses included, and the estimation methodology used to make the damage estimate.
The Super Tuesday tornado outbreak on February 5–6, 2008, resulted in 57 fatalities, the greatest number of deaths from an outbreak since May 31, 1985. The event spawned 82 tornadoes in nine states and caused fatalities in four states. There were five violent EF4 tornadoes, two each in Tennessee and Alabama, and one in Arkansas; the Arkansas EF4 tornado had a remarkable 123–mile continuous damage path. The NWS formed a Service Assessment Team, which included a SIP scientist, to evaluate its performance during this catastrophic weather event. For the first time, the Service Assessment charter included an emphasis on assessing the societal impacts of the event on members of the public. Two key components of the societal impacts assessment were to better understand why the large loss of life occurred and, more generally, to gather information about people's actual warning response behaviors. Specifically, there was an emphasis on ascertaining:
- What information people had about the severe weather situation and how they interpreted that information
- How people perceived their risk in this situation
- What decisions people made
These aspects were assessed through semi–structured interviews with 41 members of the public. This type of empirical data is critical to helping the meteorological community improve its understanding of how people assess risk and, potentially, to improve its communication of risk to the public.
Non-SIP-funded primary research:
Building on prior work (Lazo 2002 and Lazo 2004) on the Assessment of Total Household Benefits of Improved Hurricane Forecasting project, this project continues efforts to evaluate households' prior experience with hurricanes and hurricane forecasts, their perceptions of the reliability of hurricane forecast information, their behavioral responses to this information, and their values for current and potentially improved hurricane forecast information. Specifically this research addresses the value of hurricane forecasts. Estimates are derived for the willingness–to–pay (WTP) for improved forecasts from a small sample of survey respondents. Econometric methods were used to analyze choices of forecast scenarios. WTP is estimated by comparing the marginal utility of forecast attributes to the marginal disutility of the cost of the improved forecast.
Appropriate information dissemination and sound decision making during weather emergencies are critical to avoid disasters. Funded by NSF for a three year period starting April 2008, this project addresses these needs by developing an integrated understanding of warning systems and processes with a focus on hurricanes in Miami, Florida, and flash floods in Boulder, Colorado. The project, which addresses the HSD decision making, risk, and uncertainty emphasis area, will:
- Address the role of uncertainty throughout the warning process, including information dissemination and decision making
- Identify more completely the suite of factors influencing organizational and public decision making and action during extreme weather events
- Characterize public preferences for different attributes of forecast and warning information
The project leverages a multidisciplinary, multi-method approach to understanding weather warning systems, system components, and their interactions. The project includes six components:
- Face–to–face organizational interviews with forecasters, public officials, and media personnel
- Focus groups with public officials, media, and study area residents
- Face–to–face mental models and decision modeling with National Weather Service forecasters, through individual and group elicitation
- Face–to–face mental model interviews with public officials, media, and study area residents
- A bilingual (English/Spanish) stated preference survey of Miami area residents
- An interdisciplinary weather warning workshop with forecasters, public officials, and media personnel
Building on several post–Katrina assessments and other studies, NOAA's "Hurricane Forecast Improvement Project Plan: An Integrated 10–Year Plan to Improve 1–5 Day Hurricane Forecasts" defines a set of aggressive metrics related to research for hurricane track and intensity improvements for 1, 2, 3, 4, and 5 day lead times, with a focus on rapid intensity change. The work considered here is designed to meet the socio–economic research needs of the HFIP in coordination with ongoing research efforts at NCAR's Societal Impacts Program in a manner consistent with the broader social science research issues identified by the social science community. The project will focus on assessing the benefits – assessed qualitatively and if possible, monetized benefits – of achieving the HFIP's stretch goals. The research will focus on:
- Assessment of emergency managers' needs, uses, and decision–making primarily with respect to hurricane intensity forecast information
- Households' values for improved intensity forecasts
Funded by NSF/NOAA for a two year project starting October 1, 2008, this project investigates communication of hurricane forecast advisories and warnings. Through a multi–method approach, a multidisciplinary team will examine:
- The process through which advisories and warnings are developed, and the resulting content
- The communication channels used by participants in this process
- How at–risk coastal residents, including more vulnerable populations, comprehend and react to specific components of advisories and warnings
The ultimate goal is to improve communication of hurricane information in order to promote more effective public–protective decision making, thereby saving lives and property. The project is a collaboration of researchers from the social and physical sciences (communication, sociology, economics, management information systems, and meteorology) and an advisory group of consisting of key stakeholders. The research will focus around two geographical areas: greater Miami and Houston/Galveston. In each geographical area, the research team will conduct semi–structured interviews and observations with:
- National Weather Service forecasters
- Broadcast meteorologists
- Emergency managers
The research team will then examine how members of the public comprehend and react to sample messages using:
- A household survey
- Focus groups with vulnerable populations
- A laboratory test including direct physiological observation
A prototype Earth–gauging system integrating weather and health data to manage meningitis (Google-Africa)
The overarching goal of this proposal is to save lives and enhance livelihood in Ghana by integrating health and environmental data and using that integrated data in health–related decision–making. Specifically, we aim to build and implement a prototype decision–support system that integrates 2–14 day weather forecasts and epidemiological data to provide actionable information that is used to contain the spread of meningitis epidemics. By applying a preliminary economic evaluation of this decision support system, we will be able assess the potential benefit of environmental data to improve public health outcomes, help prioritize continuing investment in meningitis management in Ghana and throughout the Meningitis Belt, and determine the appropriateness of extending the prototype to other diseases, nations, and continents. This effort is a small piece of an overall Google and Moore foundation effort to develop an Earth–gauging System that will integrate environmental, health and development data into products that stakeholders and researchers could use to monitor variables, analyze trends and uncover relationships among different variables. The Earth–gauging System will support the prediction of emerging threats, and provide the basis for an integrated early–warning system that will improve health, food security, and development and conservation outcomes.
The broadcast meteorology community is a critical intermediary for providing weather information to members of the public. Previous research has shown that local and cable television are primary sources of weather information for people for everyday weather and for major weather events. The importance of broadcast meteorologists extends to communicating forecast uncertainty information, and they are uniquely positioned in this role. Broadcasters are interpreters and communicators of forecast uncertainty across multiple media (e.g., television, internet), and they receive constructive solicited and unsolicited feedback from their viewers. Moreover, broadcasters themselves have to receive, interpret, and utilize forecast uncertainty information. As both users and providers of uncertainty information, broadcast meteorologists have considerable knowledge, perceptions, and experiences that can inform the broader meteorological community. The exploratory project, funded by NOAA, begins to tap into this knowledge of the broadcast community. More specifically, we explored broadcast meteorologists'
- Use of and preferences for current and future information on forecast uncertainty
- Perceptions of the public's understanding of, use of, preferences for, and potential benefits from using forecast uncertainty information
These topics are sufficiently complex that a single project cannot provide definitive answers, but the research reported here can contribute to filling our knowledge gaps and will serve as important groundwork for future research. FY09 efforts will focus on developing a short BAMS Inbox paper on results of the FY08 efforts.