CHAPTER 3

Assessing the Current State of Social and Behavioral Sciences Within the Weather Enterprise

Most discussions about integrating social and behavioral sciences (SBS) within the weather enterprise acknowledge that much progress has been made, but many barriers to progress still exist. This statement remains true, based on what we have learned from broadly surveying the recent history and current landscape and talking with a diverse array of scholars and practitioners who work at this intersection. This chapter provides a brief overview of recent and current activities related to SBS in the weather enterprise (see Section 3.1), using the categories of activity noted below. This is followed by discussion of key existing barriers to meaningful progress (see Section 3.2).

The SBS activities discussed herein are categorized into the following groups:

• Research activities: May include basic, applied, and development research efforts.
• Private-sector activities: May include marketing studies, product development research and evaluation, interface design, and assessments of workplace efficiency or team processes.
• Agenda-setting activities: Those for which the primary goal is to develop agendas for SBS-weather research.
• Research community programs and capacity-building activities: Those with a primary goal to help grow and sustain an SBS-weather research community.
• Communication and information-sharing venues: Formal standing groups and discussion forums, as well as ad hoc activities that allow researchers and practitioners to share ideas and research findings.

We also review the status of:

• Integrating SBS into research-to-operations efforts.
• Routine data collection efforts that help support SBS-weather research.
• Existing mechanisms for funding activities at the SBS-weather interface.

This overview focuses primarily on activities related to near-term weather hazards, however much related and useful SBS research addresses climate variability at sub-seasonal to seasonal timescales, as well as long-term climate change timescales. Likewise, SBS research can encompass “everyday” weather that is not hazardous. Drawing on insights from these related areas of research could help inform studies of near-term hazardous weather.

3.1 OVERVIEW OF RECENT/CURRENT SBS ACTIVITIES

Research Activities

Myriad studies have been conducted that have examined SBS aspects of weather hazards across a range of disciplines, investigating varying concepts and theories and using different methods. It is beyond the scope of this report to provide a comprehensive summary of all the relevant research that has been done over the past several years. Rather, here we provide a brief illustration of topics that have been investigated, including some key literature reviews.

Research on Weather Professionals

There is a growing body of research on weather enterprise professionals, including National Weather Service (NWS) forecasters, emergency managers and other public officials, and broadcast meteorologists. This research has focused on the individual, group, and system levels, and it has looked across weather hazards such as hurricanes, flash floods, and tornadoes. Some of this research has examined these professionals’ job roles, cultures, goals, and functions; their conceptualization of the risks of different extreme weather hazards; and the creation and dissemination of preparedness, forecast, and warning information (Bostrom et al., 2016; Demuth et al., 2012; Morss et al., 2015; Nagele, 2015). Some studies have focused on assessment and communication of uncertainty information as a lens through which to study forecasters (Daipha, 2012) and broadcast meteorologists (Demuth et al., 2009). Still others have taken an ethnographic approach to deeply understanding NWS forecasters’ cultures, socio-technical environments, and threat detection and communication (Daipha, 2015; Fine, 2007; Henderson, 2016). More recently, there has been an emphasis on examining the role of forecasters amidst rapidly developing and proliferating weather information, including how they interpret and use new weather radar information (Heinselman et al., 2012, 2015), how they interpret and use Geostationary Operational Environmental Satellite (GOES-R) imagery in operational

forecasting (Gravelle et al., 2016), how they manage cognitive task loads,¹ and how they use ensemble-based information from numerical weather prediction models to assess and warn for a threat.²

Among the many findings from this emerging research area are that professionals in the weather enterprise tend to be overtaxed and face challenging work environments, such as shift work (e.g., Daipha, 2015); and that even when there is significant coordination across weather forecast and warning systems (for example, in the form of coordination conference calls during major weather event) the involved professionals’ understanding of the overall forecast and warning system could be improved (e.g., Morss et al., 2015).

Weather Enterprise System-Focused Research

SBS research can examine, in many different ways, the weather enterprise as a system. This research includes the above-mentioned studies of weather professionals, as well as studies that examine people’s weather information sources, preferences, and uses (Demuth et al., 2011; Lazo et al., 2009; Sivle and Kolstø, 2016). It can include research about the influences of trust among different actors in the weather enterprise, for instance, related to the communication of hurricane risk information (Demuth et al., 2012) and flows of information across organizations during crisis management efforts (Militello et al., 2007). Economic assessments also have been conducted to examine how people value weather forecasts (Lazo et al., 2009), the economic sensitivity to weather variability (Lazo et al., 2011), time cost savings due to the transition from county-based to storm-based polygon tornado warnings (Sutter and Erickson, 2010), cost of avoided fatalities due to tornado safe rooms (Merrell et al., 2002), fine-scale hurricane damage losses (Czajkowski and Done, 2014), and financial hurricane risk mitigation strategies (Meyer et al., 2014b; Wilks and Horowitz, 2014).

Exciting advances are being made in interdisciplinary research that models and simulates interactions between infrastructure, communications, and weather-related behaviors such as evacuations (e.g., Quiring et al., 2014; Stephens et al., 2015; Ukkusuri et al., 2017; and the National Science Foundation (NSF) awards in Appendix A). Research to date, for example, illustrates the economic benefits of meteorological services that result from improvement of decisions made in numerous economic sectors

___________________

¹ See NOAA grant: “Probability of What? Understanding and Conveying Uncertainty Through Probabilistic Hazard Services” in Appendix A.

² See NOAA grant: “Refinement and Evaluation of Automated High-Resolution Ensemble-Based Hazard Detection Guidance Tools for Transition to NWS Operations” in Appendix A.

(e.g., Frei et al., 2014). It also demonstrates a continuing need for improved valuation and cost-assessment methods (e.g., Sutter and Ewing, 2016), for audience- and decision context–specific evaluations of graphical communications (e.g., Visschers et al., 2009); and for interdisciplinary studies of the dynamics of weather forecast, warning, and response processes.

Vulnerable Populations

Social vulnerability to disasters has been studied extensively in the hazards and disasters community. This body of work has resulted in several informative texts, including books that describe the roots and essence of vulnerability (e.g., Wisner et al., 2004), edited volumes that empirically summarize what is known about different vulnerability characteristics during disasters (Enarson and Pease, 2016; Phillips et al., 2009), and books that chronicle specific vulnerable populations following a given weather disaster, like Hurricanes Andrew (Peacock et al., 1997) and Katrina (Fothergill and Peek, 2015). Moreover, different indices have been developed as a means for quantifying vulnerabilities to hazards—among them the Social Vulnerability Index (Cutter et al., 2003)—with more recent related work conducted that compares the different indices (Bakkensen et al., 2016). Topics of weather vulnerability studies to date include, for example:

• vulnerabilities due to nocturnal tornadoes (Ashley et al., 2008);
• vulnerability of mobile home dwellers (e.g., Ash, 2017; Prasad and Stoler, 2016; Schmidlin et al., 2009);
• tornado vulnerabilities of Texas residents (Dixon and Moore, 2012);
• how Miami-area residents’ vulnerabilities intersect with receipt and use of hurricane risk messages (Lazrus et al., 2012);
• challenges that disabled people face in terms of physical safety and access to aid, shelter, evacuation, and relief during natural hazards (Hemingway and Priestley, 2014);
• vulnerabilities that result because people want to protect their pets (Edmunds and Cutter, 2008) or they fear they will not be able to return home if they evacuate (Siebeneck and Cova, 2008; Siebeneck et al., 2013); and
• communication strategies, crowd behavior, and event safety management when weather hazards affect large gatherings of people in outdoor venues such as at sports events, festivals, and on college campuses (e.g., Sherman-Morris, 2010; Zeitz et al., 2009).

These and many other studies illustrate that vulnerability is complex and results from interactions of multiple factors (e.g., Cuite et al., 2017; Demuth et al., 2012, 2016; Dewitt et al., 2015; Huang et al., 2012, 2016; Jaurnic and Van Den Broeke, 2016; Klowkow et al., 2014; Lindell et al., 2016; Morss et al., 2016a,b; Phillips and Morrow, 2007; Sutton and Woods, 2016; Sutton et al., 2014).

Message Design

There are myriad research studies that pertain to the design, interpretations, and effects of weather forecast and warning messages, focusing on different populations regarding weather hazards such as hurricanes, floods and flash floods, tornadoes, and high winds, as well as everyday weather. For example, past studies have addressed:

• trade-offs between message content and length to examine the effects of curtailing length in some channels such as Twitter and WEA messages (Bean et al., 2015; Sutton et al., 2015a,b);
• perceptions and interpretations of forecast uncertainty information (Fischhoff et al., 1994; Joslyn and LeClerc, 2012, Joslyn and Savelli, 2010; Morss et al., 2008a; Murphy et al., 1980; Savelli and Joslyn, 2013);
• effects of messages on how people assess and respond to weather risks (e.g., Ash et al., 2014; Broad et al., 2007; Morss et al., 2016b; Nagele and Trainor, 2012; Perreault et al., 2014; Ripberger et al., 2015; Taylor et al., 2009);
• dynamic responses to weather risks (Gladwin et al., 2001; Meyer et al., 2014a; Morss et al., 2017); and
• how people conceptualize and perceive weather risks (Hoekstra et al., 2011; Knocke and Kolivras, 2007; Peacock et al., 2005; Terpstra, 2011; Trumbo et al., 2016), including factors that influence their understanding (Demuth, 2015; Klockow et al., 2014; Trainor et al., 2015).

Some cognitive experimenters are now using eye-tracking technologies³ to better understand how people interpret weather messaging (e.g., Drost et al., 2015; Sherman-Morris et al., 2015). Wilson and colleagues (2016) discuss how such technologies also allow for the study of forecasters’ decision-making and cognitive processes to better understand how they use information during complex forecasting situations.

___________________

³ Also called gaze-trackers, these previously cumbersome but now more portable devices use cameras and sometimes headsets to track where and when a person’s gaze focuses when they look at a specific message or graphic. See Wilson et al., 2016, for an example of the use of this kind of research technology.

There is a large body of research on message design for many other, non-weather-related types of risks, which offer insights that are directly applicable to weather-related messaging. An extensive body of relevant message design studies with direct relevance for weather forecasts and warnings exists in areas such as health, safety, and risk (e.g., Cho, 2011; Fischhoff et al., 2011; Glik, 2007; Parrott, 2017; Wogalter, 2006). Many such studies provide guidance specifically relevant to designing warnings (Wogalter, 2006; Wogalter and Mayhorn, 2017), hazard communications (e.g., Huang et al., 2016; Thompson et al., 2017), or both (e.g., Sorensen and Mileti, 2017a,b,c). Also pertinent is an extensive body of research on health and risk communication program design (e.g., NIH⁴). The general takeaway from this broad area of research is that the effects of messages depend on the audience and many of their specific contextual characteristics. The empirical support for guidance on some types of message design has strengthened as the field has advanced over the past few decades, but gaps remain and continue to emerge as technologies evolve, suggesting that message design research merits a more systematic approach (NASEM, 2017a).

No discussion of SBS-weather research would be complete without mention of big data, which is generating increasing interest across many scientific domains in both the public and private sectors. See the later section on data collection efforts for discussion of the advances and opportunities arising on this front.

Literature Reviews and Syntheses

In addition to the many individual studies noted above, some important reviews of relevant literature have been conducted. Among these are extensive annotated bibliographies of public risk communication on hazard warnings, protective actions responses, and public education (Fischhoff et al., 1984; Mileti et al., 2006; Mileti and Sorensen, 1990). Lindell and Perry (2012) summarized studies of factors affecting people’s responses to different environmental disasters, including hurricanes and tornadoes. Literature reviews of how people assess and respond to specific weather risks also are available for hurricanes (Baker, 1991, Dow and Cutter, 2002, Huang et al., 2016; Lazo et al., 2015) and floods and flash floods (Kellens et al., 2013). There are also efforts under way by the Federal Emergency Management Agency (FEMA) to review the scientific base for protective actions that are recommended to the public in response to specific weather hazards (FEMA, 2017b).

___________________

⁴ Making Health Communication Programs work (NIH Pink Book), see http://www.cancer.gov/pinkbook.

As discussed further in section 5.1, reviews of related research areas also have important implications for the weather enterprise—including for instance, research on team science (NRC, 2015) and team performance (Cooke and Winner, 2007; Salas et al., 2008); judgment and decision making under uncertainty and risk (e.g., Bruch and Feinberg, 2017; Hastie, 2001; O’Connor et al., 2005; Shanteau, 2015; Stewart et al., 2004); and crisis and risk communication in a variety of related domains, such as health, climate, and environment (e.g., Fischhoff et al., 2011; Fitzpatrick-Lewis et al., 2010; Glik, 2007; McComas, 2006; Nisbet et al., in press; Parrott, 2017), and warnings (Wogalter, 2006; Wogalter and Mayhorn, 2017).

Systematic research reviews and literature syntheses on a wide array of SBS weather-related topics provide an essential basis for advancing the field. Like any endeavor, knowing the “state of the science” is the foundation for knowing how to more effectively move forward. This sort of periodic assessment can help to identify fits and starts of support, evolution in thinking, the different ways that SBS helps us know and understand people, what is known and where the knowledge gaps are, and who is involved in these efforts. The need for a central clearinghouse for collecting such information has been called for in several past reports, and we reiterate this need. If the National Oceanic and Atmospheric Administration (NOAA) and other partners support periodic assessments for SBS weather research, this would help to assure that those who plan, manage, and seek to apply SBS research are aware of what issues have already been well-addressed by the research community, and what issues remain as critical knowledge gaps.

Literature syntheses and assessments are useful to stakeholders across the weather enterprise in part because much of the important research being done appears only in specialized disciplinary journals that may, unless there is open access, be inaccessible to people who are not specialists in that discipline—for instance, operational or broadcast meteorologists generally do not have access to academic journals that require paid subscription. And just as physical scientists need accessible means to follow relevant developments in the social sciences, conversely, social scientists need accessible means to follow relevant developments related weather forecasting and meteorological sciences.

Research syntheses can be aimed at informing scientific audiences for shaping future research directions, or they may be aimed at aiding the work of operational practitioners. In the latter category, one encouraging example is the recent production by the NOAA Social Science Committee of the report Risk Communication and Behavior: Best Practices and Research Findings (NOAA, 2016). While not a comprehensive assessment, this report does provide a useful overview of risk communication

issues of direct relevance to NOAA. It is too soon to ascertain how effectively this overview is being promulgated and applied in NOAA’s operations and more widely across the weather enterprise. But as a start, this is a worthy example of the type of assessment that could, with sufficient financial support, be done on a regular basis, as systematically as possible, for a variety of SBS topics of importance within the weather enterprise.

Private-Sector Activities

A great majority of people in the United States receive their weather information from private-sector apps or websites, or from watching local or national television stations (Lazo et al., 2009). Given the large, diverse audiences that the private sector can reach—and the significant motivation to gain insights on how these audiences use and perceive the value of their services—the private sector has numerous potential opportunities to undertake SBS research and apply SBS insights to their operations. This includes the groundbreaking opportunities of some private-sector weather companies to pursue “big data” analyses, as discussed later in this chapter.

The challenge for this sector is determining how to most effectively take advantage of these opportunities, and to access and use SBS insights in ways that both increase the company’s value in the marketplace and advance our common goals of saving lives and protecting property. As part of the information gathering for this study, the Committee collected input via direct briefings and written exchanges from representatives of more than 20 private-sector companies that provide weather information services to a wide variety of customer bases. These private-sector representatives identified numerous ways that SBS insights can benefit their work, many of which would benefit the entire weather enterprise. Specific topics they mention included:

• to help present messages in ways that are understandable and actionable and that can influence people to take action during life-threatening weather,
• to determine the most valuable ways to communicate custom forecasts to respective clients with different requirements and decision support processes,
• to gain a deeper understanding of the cognitive decision processes individual engage in when digesting information and making decisions,
• to gain better understanding of how information influences decisions—to help drive product/feature development and marketing efforts to get the products and information into the hands of the right people, and
• to help identify the interests and common behaviors of the audience, thus helping to target ideal advertising partners.

While some of these goals relate solely to the commercial, competitive environment in which these companies operate, other goals align well with those of the weather enterprise overall. These private-sector representatives also helped to elucidate the activities they actually engage in to generate new SBS-related insights of value to their companies. For instance:

• Some companies use outside research firms and some companies conduct their own research to ascertain what types of information their audiences are interested in and how they expect this information to be delivered to them. These studies employ various methods such as surveys, usability studies, focus groups, custom research programs, and studies ranging in size from small (e.g., usability tests or surveys) to large (e.g., behavioral/attitudinal studies on a global scale).
• Some companies utilize access to big data from the billions of data requests they receive daily to develop business, economic, social, and artificial intelligence analytics to understand trends where value can be created.
• Many companies use analytics to measure how many members of digital audiences are looking at a particular story or video, the time of engagement, location, etc. This helps them determine which themes are working for a particular audience. This information is also used by ad sales departments as they help identify the interests and common behaviors of an audience and target ideal advertising partners.
• Some companies reported using behavioral science–trained interviewers to determine the most valuable ways to communicate custom forecasts to respective clients with different requirements and decision support processes. Others have used panel groups to understand product usage and perception.
• Many utilize SBS research to identify need states that are then woven into product strategy. Activities that pinpoint preferences, attitudes, and behaviors are used to fine tune offerings and to inform data implementation, architecture, and user experience practices.
• Some also have advisors who help on-air meteorologists develop effective communication styles for giving viewers the information they need. This includes both verbal and nonverbal communication, along with use of onscreen graphics.

Overall, it appears that most SBS-related activities in the private sector can be characterized as audience surveys, marketing research, and product research and development (R&D), all focused primarily on expanding viewership and market share for products and services. Most of these efforts are considered to be proprietary in nature; particularly any information gathered that directly informs a company’s R&D and product roadmap is thus unlikely to be shared with the broader weather enterprise.

Yet, some aspects of these activities hold the potential to contribute to fundamental new SBS insights. Examples of fundamental social science topics that are relevant to broader weather enterprise concerns include research on trust and source credibility (e.g., Hayden et al., 2007; Hoffman et al., 2009; Mayer et al., 1995; Rayner et al., 2005; Sherman-Morris, 2005) especially in the context of risk communication (Löfstedt, 2005; Siegrist et al., 2012), and research on creating virtuous feedback cycles that enable continuous product improvement (e.g., Van Doorn et al., 2010).

Most weather-related companies recognize that they share some common goals that reach across the weather enterprise, and some weather-related companies are open to exploring new opportunities for public–private partnerships—either for supporting actual SBS research, or for advancing research agenda-setting, research community- and capacity-building, and information-sharing activities. These partnership possibilities are discussed further in Chapter 6.

Agenda-Setting Activities

On several occasions over the past few years, groups of scientists and other stakeholders have convened workshops and other gatherings focused on identifying an agenda for SBS activities within particular key areas. Some of these were stand-alone activities and some were events held as part of the planning for ongoing research programs/projects. (Research recommendations from these and other similar activities are summarized in Section 5.1.)

A sampling of these activities is summarized below.

• Pomona workshop. In 2005, scientists gathered in Pomona, California, to develop a research agenda for SBS activities centered on the hurricane forecast and warning system. The Pomona workshop included participants from research organizations (e.g., the National Center for Atmospheric Research [NCAR]) and federal agencies (e.g., FEMA, NSF, and numerous NOAA divisions). Multiple types of social science expertise were represented including geography, sociology, economics, anthropology, and risk and decision management. Drawing from the discussion at this and other workshops, Gladwin and colleagues (2007) issued a “call to action for the appropriate agencies and organizations to support social science research on the high-priority issues in the hurricane forecast and warning system to meet societal goals of protecting lives and property in the face of the ever-present threat of hurricanes.”
• North American THORPEX Societal and Economic Research and Applications (SERA) group. In this regional effort within the international THORPEX

• program (The Observing System Research and Predictability Experiment; Morss et al., 2008b), more than 40 scientists from a range of disciplines such as meteorology, behavioral science, psychology, economics, atmospheric research, and communication research gathered in Boulder, Colorado, to outline SBS research priorities that address the overall international THORPEX goal to “accelerate improvements in the accuracy of 1-day to 2-week high-impact weather forecasts for the benefit of society, the economy, and the environment” (WMO, 2017). Scientists identified five priority themes for SERA research, with the recommendation that these be addressed first within an ongoing and sustainable SBS research effort.

• Weather-Ready Nation (WRN) meetings. In response to the large number of tornado-related deaths in the spring of 2011, NOAA developed the Weather-Ready Nation initiative, starting with a cross-disciplinary conversation held in Norman, Oklahoma, “to identify, prioritize, and set in motion actions to improve the nation’s resiliency against severe weather, especially tornadoes, to protect lives and property” (UCAR, 2012). The ~175 attendees included physical and social scientists, emergency managers, forecasters, government officials, and TV broadcasters. For many of the physical scientists, this was their first experience in such an interdisciplinary conversation. This initial conversation was followed by an “Imperatives for Severe Weather Research” workshop held in Birmingham, Alabama, in March 2012 with 65 participants representing numerous disciplines (Lindell and Brooks, 2013). Eight white papers were developed prior to the meeting, and the ensuing conversation resulted in twelve specific research recommendations.
• World Meteorological Organization High-Impact Weather (HIWeather) project. Following on the THORPEX efforts, a workshop was held in Karlsruhe, Germany, “to define a high-impact weather project . . . emphasizing the improvement of the predictions of high impact weather on the hours-to-weeks timescales with a stronger focus on shorter time and space scales” (PANDOWAE, 2013). Another workshop was held in 2014 in Silver Spring, Maryland, to develop an implementation plan which established HIWeather goals to “promote cooperative international research to achieve a dramatic increase in resilience to high impact weather, worldwide, through improving forecasts for timescales of minutes to 2 weeks and enhancing their communication and utility in social, economic and environmental applications” (Jones and Golding, 2014). The plan indicates the need for the physical and social science communities to work together to achieve these goals. A kickoff meeting for HIWeather was held in Exeter in April 2016.

• NOAA Flash Flood Summit. The new Water Center in Tuscaloosa, Alabama, served as the venue for a 2014 multidisciplinary summit designed to refine the vision of flash-flood forecasting in the United States. A significant component involved “exploring the intersection between science and social science requirements to help inform priorities” (NOAA, 2015). The summit resulted in three high-level recommendations spanning several disciplines.
• Living With Extreme Weather workshop. The goal of this May 2015 workshop was to “bring together researchers from the social, behavioral, and economic sciences, as well as physical sciences, engineering, technology and operational domains, to chart a bold and innovative course for addressing one of society’s greatest challenges: reducing mortality associated with extreme weather” (Droegemeier et al., 2016). This workshop played a major role in developments leading to the creation of “The Alliance,” which is discussed in the next section.
• VORTEX-SE workshops. In 2015, Congress provided funds to NOAA to carry out the Verification of the Origins of Rotation in Tornadoes EXperiment-Southeast (VORTEX-SE) program with the goal of understanding the tornado problem in the southeast, including both physical and social science aspects (Rasmussen, 2015). NOAA held a workshop in Birmingham, Alabama, in November 2015 to help define an interdisciplinary research agenda for the project. The output from the workshop was a set of research problem statements that informed the eventual Science Roadmap from which the NOAA federal funding opportunity was developed. A portion of the VORTEX-SE allocation facilitated a grants program for non-NOAA researchers. After the first field project phase (held in March-April 2016), a second workshop was held in conjunction with the American Meteorological Society (AMS) Severe Local Storms conference in Portland, Oregon, in November 2016 to revise the (renamed) Science Plan.

Research Community Programs and Capacity-Building Activities

Fostering the growth of a new field of research, policy, and practice—especially one that is highly interdisciplinary in nature—requires dedicated efforts to inspire, build, and sustain a true community among countless individuals who may be widely dispersed in terms of disciplinary background, geographic location, and sector. These efforts can provide opportunities for developing professional networks and relationships, for sharing knowledge and perspectives, for identifying research needs and opportunities, and, perhaps most importantly, for training, mentorship, and encour-

agement of people who are new to thinking and working at this interdisciplinary intersection.

It is not possible to characterize all activities that are helping build the SBS-weather community, so only a few are briefly described below. Some are longstanding programs that have played an indispensable role in building the SBS community such that it exists today, whereas others are nascent activities that hope to build on the foundation of efforts to date. These efforts are varied in their scope and in the ways that they support the community, but broadly speaking, they can be categorized into two groups:

• capacity-building efforts that focus on shaping and growing the community through such activities as developing research agendas and frameworks, training and mentoring, catalyzing partnerships, convening workshops and conferences, and so forth; and
• community programs that focus on conducting research (both fundamental and applied) that integrates SBS within the weather enterprise, and that contribute to community capacity-building.

A capacity-building movement that has left an indelible, continued mark on the SBS-weather community is WAS*IS (Weather and Society * Integrated Studies).⁵ WAS*IS began in 2005, envisioned as a one-time workshop to introduce social science concepts and methods to interested meteorologists. Due to sustained interest and support primarily from the Societal Impacts Program (SIP, discussed below), as well as from NWS, WAS*IS evolved into 10 workshops, including 2 international ones. The workshops were attended by 276 participants total, comprised of public, private, and academic sectors and backgrounds in atmospheric science and several social science disciplines. As such, the WAS*IS movement changed the culture of the weather enterprise by integrating social science into meteorological research and practice in comprehensive and sustained ways, not by funding research (although several WAS*IS participants did conduct research inspired by the workshop), but by building exposure, commitment, and capacity (Demuth et al., 2007). The WAS*IS workshops were discontinued after 2011 due to lack of funding. Yet, the movement continues virtually through a WAS*IS Facebook page with 895 members and a WAS*IS student Facebook page with 265 members as of this writing.⁶ The Facebook pages serve as focal points for discussing topics, jobs, policy, conferences, and so forth that pertain to SBS in the weather community.

___________________

⁵ See http://www.sip.ucar.edu/wasis for additional details.

⁶ Some people are members of both Facebook pages.

In 2016, a new capacity-building effort The Alliance for Integrative Approaches to Extreme Environmental Events⁷ was announced. The mission of The Alliance is “to serve as an organizing mechanism among a wide array of sectors and stakeholders in facilitating rapid and sustained progress toward mitigating the societal impacts of extreme environmental events.” The effort includes 10 initial programmatic elements in support of a broad SBS-weather community. Among them are elements to synthesize relevant funding opportunities, help identify collaborators, assist with proposal preparation, provide electronic resources for facilitating interactions, and provide travel funding. The Alliance eventually will have a core staff as well as a volunteer Steering Committee. The Alliance was seeded initially by a $3 million private gift with a goal of acquiring sustained funding from multiple sources.

Over nearly the last 15 years, there have been a handful of community programs dedicated to integrating social and behavioral sciences within the weather community. These programs have evolved, and some have ended due, in part, to funding changes. For instance:

The Collaborative Program on the Societal Impacts and Economic Benefits of Weather Information (SIP) was founded at the National Center for Atmospheric Research (NCAR) in 2004 to “improve societal gains from weather information by infusing social science research, methods, and applications throughout the weather enterprise” (NCAR, 2017). With support from NOAA (through the U.S. Weather Research Program) as well as from research grants, SIP currently conducts fundamental and applied research and supports community activities with a focus on economic assessments of existing and improved weather information. SIP previously also supported multiple capacity-building efforts, including costs to convene the WAS*IS workshops and associated staff time, the Weather and Society Watch newsletter, and a community discussion board.

Also at NCAR, the Weather Risks and Decisions in Society (WRaDS) program builds fundamental understanding of how weather- and climate-related risks intersect with society, including how they factor into decision making. WRaDS research integrates social sciences approaches and methods with knowledge about weather and climate research, prediction, and predictability. It includes studies with members of various publics, forecasters, public officials and public agency personnel, broadcast meteorologists, and other stakeholders, drawing from and contributing to the atmospheric science, natural hazards, risk communication, and environmental anthropology communities. WRaDS also helps build capacity for weather-society research and activities

___________________

⁷ The current webpage is at http://alliance.ou.edu; it is also described here: https://hazards.colorado.edu/article/the-alliance-for-integrative-approaches-to-extreme-environmental-events.

at universities and colleges, in public agencies, and in the broader community. WRaDS support comes from NSF and research grants.

Among past programmatic efforts is Social Science Woven into Meteorology (SSWIM). SSWIM was founded in 2008 at the National Weather Center in Norman, Oklahoma, to promote collaborative research and partnerships between the social sciences and the physical sciences to enhance societal relevance and to reduce the human risk from atmospheric and related hazards. SSWIM was supported by NOAA and the University of Oklahoma through 2012. Another effort, Weather for Emergency Management (WxEM), was a collaboration among NOAA, NWS, and the East Carolina University Renaissance Computing Institute, which focused on helping the emergency management community more effectively use weather information in their planning and in real-time decision-making.

Figure 3.1 provides a summary of the timeframe over which these various programs and activities have operated.

The programs discussed above focus on SBS specifically for weather hazards. There are also numerous other programs that conduct relevant SBS research and capacity building within the broader scope of societal risk, hazards, and disasters. We do not attempt here to offer an exhaustive list of all relevant institutions and programs, but we offer the following as some examples of key institutions:

• Center for Disaster and Risk Analysis at Colorado State University
• Department of Homeland Security (DHS) Centers of Excellence

• Hazard Reduction & Recovery Center at Texas A&M University
• Hazards and Vulnerability Research Institute at the University of South Carolina
• The Center for Advanced Public Safety at the University of Alabama
• The Center for Risk and Crisis Management at the University of Oklahoma
• The Disaster Research Center at the University of Delaware
• The Natural Hazards Center at the University of Colorado
• The Nurture Nature Center
• The Risk and Disaster Communication Center at the University of Kentucky

Communication and Information-Sharing Venues

A key requirement for advancing any field of research is the availability of venues where research ideas and results can be shared and discussed—typically through periodic conferences, peer-reviewed scientific journals, and standing committees and other groups. Advancing these standard communication mechanisms at the interface of SBS and meteorology requires deliberate efforts to work across traditional disciplinary stovepipes. There has been considerable progress made on this front in recent years, due largely to the efforts of scientific societies described below.

The American Meteorological Society (AMS) has, for nearly a century, worked to advance “the atmospheric and related sciences, technologies, applications, and services for the benefit of society.” For most of that period, the AMS and its members have seen improvements in physical understanding of the Earth system as the most effective pathway to realizing greater societal benefit. But in recent years, the AMS has turned to social science as providing the keys to further progress and has taken several steps to foster SBS research as applied to weather, water, and climate science and services. In particular, the Society has worked to build the infrastructure needed to support such research in two important ways, described below.

• The AMS journal Weather, Climate, and Society. Following a few years of study and preparation, the AMS launched a new peer-reviewed journal, published quarterly, that provides SBS researchers investigating applications in the Earth sciences with a platform for publishing their work. The quality, quantity, and diversity of the articles continue to grow, as does the journal’s reputation in the SBS community.
• AMS symposia and conferences. AMS Annual Meetings are structured as a collage of specialized conferences and symposia, and each year since 2006, this has included a Symposium on Societal Applications: Policy, Research, and Practice. Initially just a few dozen people attended 1 day of sessions. Today several hundred people attend 4 full days of panels, presentations, town halls,

The National Weather Association (NWA) is another professional society playing important roles at the interface of SBS and meteorology. The NWA hosts a Societal Impacts Committee, which undertakes efforts such as: (i) participating in the NWA Annual Meeting and other NWA-sponsored conferences and workshops to share recent research and applications and enhance dialogue about societal impacts of weather and climate; (ii) engaging in educational activities concerning the societal impacts of weather and climate and its application to decision-making processes; (iii) interacting with and serving as a resource for other NWA committees regarding activities and initiatives that involve societal impacts (e.g., design, implementation, and analysis of surveys; development of conference sessions, web page content, and outreach projects and materials); (iv) facilitating partnerships between meteorology, climatology, and social science communities to advance applied research on the societal impacts of weather and climate and the application to hydro-meteorological forecasting and decision support; and (v) providing advice, information, and policy statements to the NWA Council on matters concerning societal impacts of weather and climate.

Professional/scientific societies such as AMS and NWA bring together operational forecasters, service providers, and equipment manufacturers, as well as researchers. Their conferences, standing committees, and publications have thus far proven to be one of the most effective mechanisms available for weaving social science expertise and insights into the advancement of weather preparedness (readiness), forecasting, and communication. Chapter 6 discusses further roles these different groups can play in advancing enterprise-wide efforts.

There are also a variety of relevant conferences and other venues available among the broader hazards research community. For instance, since 1975 the University of Colorado’s Natural Hazards Center has hosted the annual Natural Hazards Research and Applications Workshop, a meeting point for exchange among researchers and practitioners, including federal, state, and local emergency officials, representatives of nonprofit and humanitarian organizations, hazards researchers, disaster consultants, and others. In addition, the International Research Committee on Disasters has for several years co-organized a set of research presentations as an addition to the main meeting that provide a forum for SBS-weather research. While too many to fully discuss here, there are numerous other agency- and disciplinary-focused conferences that offer potential as useful forums for discussing weather-related SBS research. For instance, this includes meetings of the following:

• American Planning Association
• American Psychological Association (Disasters and Terrorism interest area)
• American Society for Public Administration (see Section on Emergency and Crisis Management)
• American Sociological Association (see Section on Environment and Technology)
• International Association of Emergency Managers (communication/weather section)
• International Research Committee on Disasters of the International Sociological Association
• National Emergency Management Association
• Society for Risk Analysis (risk communication specialty group)

Integrating SBS in Research-to-Operations Efforts

The concept of transitioning research advances into the routine operations of NWS and other institutions of the weather enterprise is a long-standing challenge—one that is well-explored in earlier National Academies reports such as From Research to Operations in Weather Satellites and Numerical Weather Prediction: Crossing the Valley of Death (NRC, 2000). Over the past decade or so, NOAA has developed a variety of concepts and programs for fostering more rapid, effective research-to-operations transitions for new meteorological technologies and tools.⁸ Such concepts, however, can be difficult to apply meaningfully to SBS research, given the hugely varying social contexts in which this research is utilized (e.g., see Klockow, 2017). The Committee’s

___________________

⁸ See https://www.weather.gov/sti/R2O for examples.

outreach discussions did not identify any simple solutions to this challenge overall, but did point to a couple of fronts on which notable progress is being made—and where the potential for much greater progress exists—for effectively integrating SBS insights into meteorological research-to-operations efforts. This includes (i) the use of Testbeds and Proving Grounds, and (ii) the development of “Living Labs,” each discussed below. These endeavors exemplify creative approaches to evaluating the usability and usefulness of weather information and tools.

NOAA Testbeds and Proving Grounds

NOAA has multiple Testbeds and Proving Grounds⁹—for aviation weather, severe weather, hurricanes, and hydrometeorology, including for winter weather and flash flooding—which are important mechanisms for evaluating the utility of transitioning innovative research into NWS operations (Ralph et al., 2013). There are recent overviews of some specific efforts, for instance, for the joint hurricane testbed (Rappaport et al., 2012); the hazardous weather testbed (Clark et al., 2012); the GOES-R Proving Ground (Goodman et al., 2012); and the hydro-meteorological testbed’s Flash Flood and Intense Rainfall Experiment (Barthold et al., 2015) and multi-radar multi-sensor experiment (Martinaitis et al., 2017). There are also myriad examples of specific research that has been evaluated in the testbeds, such as probabilistic forecasting of severe convection (Karstens et al., 2015), use of new GOES-R satellite imagery (Gravelle et al., 2016), and tropical cyclone intensity forecasts (DeMaria et al., 2014; Kossin and DeMaria, 2016), including wind speed probabilities (DeMaria et al., 2009, 2013).

As the lists above illustrate, the testbeds’ research-to-operations focus primarily has been on atmospheric science. Yet, they also offer great potential for integrating SBS expertise and research, in particular opportunities to investigate questions about how originators and mediators of forecast information (e.g., NWS forecasters, broadcast meteorologists, emergency managers) access, interpret, and utilize new information being developed by the atmospheric science research community.

Some integration of SBS expertise into testbed activities is beginning to occur. For instance, SBS research has examined forecasters’ use of new systems for forecasting severe hail (Ling et al., 2015), and the effect of new rapid-scan radar information on severe weather warning decisions (Heinselman et al., 2012). The FY 2017 joint funding announcement for the Hurricane, Hazardous Weather, and Hydrometeorology Testbeds¹⁰ included SBS research topics. Fostering these and other opportunities to more deeply

___________________

⁹ See additional information at http://www.testbeds.noaa.gov.

¹⁰ See http://www.nhc.noaa.gov/jht/NOAA-OAR-OWAQ-2017-2005004.pdf.

and broadly integrate SBS into such activities will help realize the full potential of the NOAA Testbeds and Proving Grounds.

Living Labs: A Platform for Multi-Disciplinary Research and Transition to Practice

A growing convention across the weather enterprise is the development of integrated warning teams: locally focused groups that integrate NWS forecasters, emergency managers, and broadcast media in order to improve operations during severe weather hazards. To understand what occurs within these teams, researchers need concepts and methods to study the decision-making processes and interactions among forecasters, emergency managers, weather observing spotters, and other stakeholders during severe weather events (Bass et al., 2011). One model for meeting such needs is the “Living Lab” (Eriksson et al., 2005)—a platform for multidisciplinary integration to address user-driven research challenges.

A recent example of this approach can be found in efforts of the Dallas-Fort Worth (DFW) Urban Demonstration Network, also called the CASA¹¹ DFW Living Lab (Philips et al., 2011). This effort brings together collaborators from across the weather enterprise, including the NWS, the North Central Texas Council of Governments, Dallas Fort Worth Airport, more than 20 local communities, and a group of academic researchers. Together, they have created an enterprise that aims to:

• develop high resolution boundary layer observations and forecasts of wind, tornados, floods, hail, and ice;
• create impacts-based warnings and forecasts for a range of public and private decision-makers that result in measurable benefit for public safety and the economy; and
• demonstrate the value of multi-sector local, federal, and private partnerships to operate observing networks and fund research.

The effort is distinguished by features such as a focus on co-creation of knowledge with multiple information users, rather than treating users solely as subjects of study; experimentation in real-world environments and in contexts that influence innovation and uncover tacit knowledge and behavior of users; private–public partnerships that govern and structure interaction among stakeholders and the public and fund the experimentation; and a research and innovation arena where interested parties can collaborate.

___________________

¹¹ Center for Collaborative Adaptive Sensing of the Atmosphere, an NSF Engineering Research Center headquartered at the University of Massachusetts.

In recent work focused on decision making related to urban flash floods, the CASA team is developing mobile applications that provide weather alerts based on contextual information and user-controlled alerting thresholds such as distance from the event. The users not only receive alerts but can report events in order to provide verification data and can provide other information about their experiences and responses during the weather event.

This Living Lab approach is an innovative way to foster cross-disciplinary learning. It is also challenging work, and requires leadership and sustained, sufficient resources, time, and diverse expertise to plan and execute such studies, and to analyze the resulting data. The approach provides an effective “forcing function” for cross-disciplinary learning and respect. It also fosters greater research-to-operations development, given that live systems must actually work for all the stakeholders involved. As a result, this approach offers a rich opportunity for growth.

Data Collection for SBS-Weather Research

Advancing the social and behavioral sciences requires the regular collection and sharing of high-quality data, including ongoing observations that may need to be sustained over periods of months, years, or even longer. This data collection serves many purposes, for instance, to better understand how key factors within a given population or organization vary over time, locations, and across different groups; to help detect gradual trends or abrupt changes in those factors over time or in response to particular events; and to explore possible correlations and causal relationships with other observed variables of interest.

There is a growing recognition of the importance of integrated data collection and analysis for understanding disasters, including those stemming from hazardous weather. In recent years, there have been some significant advances in better integration of physical science data from multiple disciplines, but much more needs to be done among weather enterprise core partners to improve the collection and integration of data that provide a social and behavioral context for understanding disasters. To this end, there are currently numerous activities managed by government agencies that can provide opportunities to collect information of great value to those studying the weather/society interface, in particular if the synergies among these different efforts are better exploited. Some examples of these relevant efforts are described below.

• NOAA/NWS Service Assessments. Conducted by teams that can include experts from outside as well as within the Weather Service, NWS Service Assessments “evaluate activities before, during, and after events to determine

• the usefulness of NWS products and services. The team generates a report, which serves as an evaluative tool to identify and share best practices in operations and procedures, and identify and address service deficiencies. The goal of the activity is for the NWS to continuously improve its services to the nation.” (NWS, 2017b) Service assessments have included social science expertise and evaluations with increasing frequency over the past several years, but going forward it would be helpful for NOAA to carefully evaluate how to do this most effectively.

• NOAA Natural Hazard Statistics. NOAA compiles annual injuries and fatalities from weather events (NWS, 2016), which provide an important basis for social science analyses. For example, an analysis of flood fatality data (Ashley and Ashley, 2008) evaluated the different reasons why people intentionally walked through flood water, thus helping to identify which of these deaths may have been preventable. Storm Data Reports contain information on storm paths, deaths, injuries, and property damage. NWS collects the data from a variety of sources, including county, state, and federal emergency management officials, local law enforcement officials, skywarn spotters, NWS damage surveys, newspaper clipping services, the insurance industry, and the general public, among others.¹² Information about behaviors is available in some of these reports.
• FEMA Mitigation Assessment Team Program. This program draws on partnerships and combined resources of federal, state, local, and private sector to assemble and rapidly deploy teams of investigators. The teams, which are drawn from disciplines such as structural and civil engineering, architecture, building construction, natural hazards research, and code development and enforcement, are deployed to evaluate the performance of the buildings and related infrastructure in response to the effects of natural and man-made hazards (FEMA, 2016). Based on the evaluations, the teams develop recommendations (e.g., for code development, mitigation activities for greater resistance to hazard events), which are reported through a variety of methods, including publications and training. In addition, the “OpenFEMA” system provides a number of relevant datasets, including disaster declarations, grants, and disaster assistance.
• FEMA National Household Survey. This survey has been conducted biannually from 2001 to 2007 and annually since 2012 to identify the effective drivers of preparedness for specific hazards and specific populations, to enhance understanding of factors that influence preparedness behavioral

___________________

¹² See https://www.ncdc.noaa.gov/stormevents/faq.jsp for examples.

• change, and to measure and track progress toward national preparedness. The report Preparedness in America, Research Insights to Increase Individual, Organizational, and Community Action (FEMA, 2014), which summarizes findings from several recent surveys, highlights critical factors of risk communication (particularly the relevance of the risk to the individual), and the importance of communication networks such as the workplace and school. Since 2013, the National Household Survey has been designed to better understand how key factors vary by hazard in order to improve the effectiveness of outreach for specific weather-related hazards, including tornados, floods, hurricanes, extreme heat, winter storms, and wildfires.

• CDC Public Health Surveillance During Disasters. Centers for Disease Control and Prevention (CDC) surveillance is “the systematic collection, analysis, and interpretation of deaths, injuries, and illnesses which enables public health to track and identify any adverse health effects in the community.” It is the base for assessing the human health impacts of a disaster and evaluating potential problems related to planning and prevention (CDC, 2012). The Disaster Surveillance Workgroup (DSWG) includes experts from across CDC who set standards for data collection, sharing, and reporting during a public health disaster. Morbidity and mortality surveillance tools and training are developed based on the standards developed by the DSWG.
• CDC Community Assessment for Public Health Emergency Response (CASPER). The CDC Division of Environmental Hazards and Health Effects has developed the CASPER toolkit to assist personnel from any local, regional, state, or federal public health departments in conducting community assessments during disasters, in order to standardize assessment procedures in U.S. disaster response. The CASPER toolkit provides guidelines on data collection tool development, methodology, sample selection, training, data collection, analysis, and report writing.¹³
• CDC National Center for Health Statistics (NCHS) Mortality and Injury Data. Mortality data from the National Vital Statistics System are an authoritative source of demographic, geographic, and cause-of-death information used as a source for research on causes of death related to hazards. In 2016, the NCHS launched a project to modernize these efforts and improve the data collection for understanding deaths related to natural hazards. This includes an effort to establish and promulgate the use of case definition for disaster-related deaths, and to increase the medical examiner/coroner’s (ME/C) use of this designation in death certificates. CDC has worked to adapt the contents

___________________

¹³ See details at https://www.cdc.gov/disasters/surveillance.

There is a growing recognition of the importance of including SBS in these and other post-disaster data collection efforts, as evidenced by a number of recent reports that discuss new data collection efforts and call for greater coordination and uniformity in such efforts (e.g., Aitsi-Selmi et al., 2015; NIH, 2017; NRC, 2012; Oliver-Smith et al., 2016). Specific social science or interdisciplinary approaches and research methodologies are recommended in some reports, such as the Forensic Investigations of Disasters (FORIN) approach described in Oliver-Smith et al. (2016). Although some efforts are under way in this direction, widely agreed-upon standards or instruments for social science disaster reconnaissance data collection have yet to be developed.

In addition to the need for survey research on experiences and survival of weather hazards, there is a need for continued development of in-depth and holistic anthropological and ethnographic research on weather hazards and the weather enterprise, as illustrated by, for example, Daipha (2015), Henderson (2016), Lazrus (2016), and Orlove et al. (2010). Recent innovative interdisciplinary weather research on which anthropologists are collaborating (e.g., Meyer et al., 2013) illustrates ways in which this rich body of qualitative knowledge can contribute to understanding human experiences with weather. In addition, social scientists from a variety of backgrounds have conducted informative content analyses of qualitative weather-related data, for example, of individual interview or focus group transcripts (Lazrus et al., 2016) and of situational reports (Tierney and Trainor, 2004). Furthermore, there is a long tradition of storytelling, illustrated by projects such as the NPR StoryCorps project (StoryCorps, 2017), the Field Innovation Team’s Stories from the Field (FIT, 2014), and the Hurricane Digital Memory Bank (HDMB, 2017). Extracting from these resources large-scale systematic data on weather decision making, behavior, and risk reduction is theoretically feasible but challenging, given that content analysis is still often conducted manually and is resource-intensive, requiring multiple coders and extensive training. As machine learning and algorithmic approaches to content analyses improve, applying social and behavioral theories to such data sources may reveal new insights.

Personal vehicle event data recorders, analogous to flight recorders in airplanes (see Box 3.1), have been used for several decades by some automotive manufacturers to evaluate factors such as the performance of airbags in crashes, for traffic enforcement, and for research purposes (Kowalick, 2005; PROSPER, 2006; Wouters and Bos, 2000). Assessments suggest that event data recorders are highly beneficial, for example, by reducing crashes by an estimated 20% and providing training and event reconstruction data (Schmidt-Cotta, 2009). Journey data recorders, used both for monitoring driv-

ing behavior and for managing traffic and cargo, have potential to be a rich source of research information on driving behavior under hazardous weather conditions (Malta et al., 2007; Toledo et al., 2008).

Most research fields must grapple with questions about developing data standards, to facilitate validation and replication of earlier studies and inter-comparison among studies that collect data at different places and times. In the social sciences, such questions can be particularly challenging, given the diversity of research methodologies that are used (see Section 5.2) and the inherent trade-offs that can exist between conducting precisely tailored place-based studies and fostering coordination among different research projects to enable data inter-comparisons and meta-analyses. There are no easy answers to such questions, and we do not suggest imposing rigid new data collection protocols for research at the SBS-weather interface, given the great benefits that can accrue from pursuing triangulation among a diversity of research

approaches. However, SBS and interdisciplinary researchers can facilitate efforts to mirror past research approaches and/or to compare among different studies by diligently documenting the research approaches that are used. For instance, by:

• making available (in published journals, in supplemental material, or to those who request it):
  • data collection instruments, including all open- and close-ended questions for surveys, interviews, focus groups, etc.,
  • coding schemes and, as appropriate, inter-coder reliability results,
  • thorough descriptions of data collection (sampling approach, implementation) and data analysis, including of quantitative and qualitative data,
  • model code, e.g., for agent-based modeling efforts;
• making datasets publicly available as possible (after they are de-identified, in accordance with Institutional Review Board [IRB] human subjects approvals); and
• as possible and appropriate, utilizing past data collection instruments and measures.

Harnessing of “Big Data”

The exact definition of big data varies but is broadly considered as datasets whose size, complexity, and heterogeneity preclude conventional approaches to storage and analysis (NASEM, 2016b). Within meteorology this concept existed well before the label became popular, given the countless atmospheric observations collected and numerical weather prediction model outputs and ensembles. Big data also is playing an increasing role—and offering increasing research and operational opportunities—in SBS-weather research, particularly through user-generated content and platforms. Such platforms may be used to elicit information, for instance, by crowdsourcing weather reports through NOAA’s mPING (mobile Precipitation Identification Near the Ground¹⁴), an app which allows anyone to report precipitation information, or by encouraging people to use specific hashtags to share weather reports with NWS offices.

Numerous sources of novel big data now being used in other domains hold promise to be useful in studying weather-related decisions and behaviors, such as big data on purchases and information searches (Choi and Varian, 2012), security camera footage (Lambie et al., 2016, 2017), or smartphone data, which can include metadata such as time, place, and weather (Bogomolov et al., 2014).

___________________

¹⁴ See http://mping.nssl.noaa.gov.

A growing source of big data that is already highly relevant to the SBS-weather interface is use of social media when hazards threaten or occur. Social media offer new opportunities for people to obtain, combine, create, and share information in creative ways during hazardous weather (Morss et al., 2017; Neeley, 2014). Crisis informatics is a burgeoning, multidisciplinary field that combines expertise in computer and information sciences with social sciences to study how people use new communication technologies during disasters, especially in the face of uncertainty (Palen et al., 2010; Palen and Anderson, 2016). For example, Hughes and colleagues (2014) found that during Hurricane Sandy, the use of online media for public communications by fire and police departments was limited, with 25% using Facebook, 7% using Twitter, and 5% using Nixle (a subscriber-based notification service).

Social media are changing the types and amount of hazard-related information that people can interact with and respond to while also offering a window to study these behaviors. Because Twitter data are publicly available, researchers are utilizing the opportunity to investigate social and behavioral aspects of a multitude of hazards and disasters (Houston et al., 2015; Hughes and Palen, 2014; Palen et al., 2010). For weather hazards, the growing body of Twitter research includes, for instance, macro-analyses of relevant keywords used and correlations between number of tweets and geographic areas at risk (e.g., tornado warned areas) or affected (Lachlan et al., 2014b; Ripberger et al., 2014; Shelton et al., 2014). Other research includes analyzing the content of tweets qualitatively and quantitatively to characterize and understand what people are tweeting about (Anderson et al., 2016; Lachlan et al., 2014a; Spence et al., 2015).

Big data is also quickly becoming an important analytical tool for private-sector weather companies. For instance, a recent article titled, Why the Future of Social Science Is with Private Companies (Schrage, 2015), stated “Google, Facebook, Amazon, Microsoft, Netflix, Alibaba and scores of other global enterprises conduct literally thousands of experiments on their networks every day. No doubt, many or most of them are marginal or incremental in design. But with literally billions of measurable customer, client, and channel interactions a year, be sure they’re also testing hypotheses that could lead to profitably disruptive innovations.” Having this capacity provides some weather companies with the potential to serve as unique “user laboratories” that government and academic organizations cannot easily replicate. This capacity promises to revolutionize weather information services.

Big data poses novel analytical and ethical challenges, however. For instance, some uses of big data may require specialized statistical and computer science expertise (e.g., in natural language processing, machine learning, and specific computing hardware systems and software languages); thus meaningful application of big data analyses to

SBS research is likely to require interdisciplinary training or interdisciplinary teams of researchers and practitioners. Big data also raises new ethical challenges, in particular with regard to issues of anonymity. Although SBS researchers are trained in privacy protection of human subjects, there have been instances where people were identifiable by their social media posts, despite researchers having taken steps to anonymize the data (e.g., Zimmer, 2010), thus raising new and evolving privacy concerns (e.g., Boyd and Crawford, 2012). Ethical questions also arise regarding privacy and ownership of big data that go beyond the confidentiality issues that SBS researchers typically face; however, approaches are being developed to address such concerns (e.g., Xu et al., 2014).

Funding Support of SBS in the Weather Enterprise

Like any field of research, SBS research on weather-related questions requires funding support. We discuss here some of the key federal agency support provided over the past several years, focusing on the investments of NOAA, NSF, and DHS.¹⁵ This is not a comprehensive analysis, given that not all funding information is available or in a form that can be easily compared among programs and agencies, but it offers a sense of the different mechanisms and relative size of the investments being made.

NOAA Research Support

NOAA’s Office of Oceanic and Atmospheric Research/Office of Weather and Air Quality (OAR/OWAQ) and National Weather Service (NWS) are critical sources of support for SBS activities for weather enterprise, encompassing a mix of efforts such as customer satisfaction surveys, development of social science curricula for meteorologists, development of societal performance measures, workshops, and research of various types that is funded through different vehicles.

Drawing on overview information provided to the Committee by NOAA about SBS-related awards made from 2004 to 2016, it appears that NWS-funded research has primarily, and increasingly, been through Cooperative Institutes, indefinite delivery/indefinite quantity (IDIQ) contracts, and blanket purchase agreement contracts. Many of these funds have been used to support applied research endeavors, largely focused around stakeholder engagement and assessment of operational forecast products (e.g., hurricane storm surge products, hazard simplification, convective weather outlooks) and services and tools (e.g., fire weather services, NOAA Weather Radio). Other

___________________

¹⁵ Federal Highway Administration (FHWA) support for SBS research specifically related to road weather issues is discussed in Chapter 4.

efforts have focused on economic assessments such as case studies of specific events and valuation of improved forecast information.

OWAQ-funded research is comprised of efforts supported through open competitions, restricted competitions, and Cooperative Agreements. For instance, in 2012 there was an open competition for Social Science Weather Research, which provided approximately $400,000 to support two to four projects. And for 2015-2017, there was a yearly open competition associated with the VORTEX-SE project, providing annual support in the range of $100,000-$250,000 total for a variety of 2-year grants. These open competitions have primarily focused on severe convective weather, although in 2016 NOAA announced a broader funding opportunity for their Hurricane, Hazardous Weather, and Hydrometeorological Testbed. The OWAQ also has funded efforts competed within a restricted context, including a Research-to-Operations grant (internal to NOAA, but could include non-NOAA collaborators), and a supplemental award opportunity in collaboration with NSF. The cooperative agreement funds have supported a mixture of foundational and applied research efforts—as well as the NCAR Societal Impacts Program (SIP), FACETs, and Social Science Woven Into Meteorology (SSWIM) efforts discussed earlier. See Appendix A for details.

A more quantitative analysis of these research investments is precluded by a lack of information about NOAA funding on an individual project level. Some general insights can be gleaned however, from the Federal Plan for Meteorological Services and Supporting Research (OFCM, 2016). This report suggests that for FY15 and FY16 together, NWS invested a little over $37 million in research support overall. In comparison, for this period of time NWS invested approximately $2.9 million in activities related to SBS-weather research. For OAR/OWAQ, overall research investments for FY15 and FY16 together were roughly $29 million. In comparison, for this period of time OWAQ invested approximately $3.3 million in activities that could be characterized as SBS-related. These comparisons, however rough, illustrate that SBS-related efforts are a relatively small part of the overall portfolio of NOAA’s weather-related research.

It is encouraging, however, that NOAA’s funding for SBS-related activities has generally been increasing over the past decade or so, and likewise encouraging to see support both for baseline initiatives as well as specific funding opportunities, as this sort of mix is important for allowing social science research in the weather community to deepen and grow. Yet, the availability and scope of these different funding opportunities have varied considerably over time. This irregularity of support can constrain scientific progress, particularly for the fledging weather-social science community. Progress of the research community would also be aided by greater transparency, for instance, regarding who is eligible to compete for funds, which proposals are being

funded, the funding topics, and whether and where results from prior funded efforts are available.

NSF Research Support

The National Science Foundation (NSF) does not keep records to specifically track funding for research at the nexus of SBS and weather, but a search of awards on its public databases yields some useful insights. A search of awards made from 1989 to 2016 on the term “weather” resulted in more than $1.7 billion in awards. Restricting this search to awards that included SBS-related terms such as “perceptions, behavior, communication, decision, action, human” (with some additional hand-screening) produces a total of $113 million in funding, much of which is interdisciplinary. It is impossible to assess from the public data how much of a total award goes specifically toward SBS research, and when the research is truly interdisciplinary, it is not feasible to truly disentangle SBS research from the other sciences involved. Thus, this analysis can at best be considered an approximation. Two additional searches were also conducted (see Figure 3.2):

• A search for weather-related awards specifically from the Directorate for Social, Behavioral, and Economic Sciences (SBE) suggests that approximately $60 million was awarded from 1989 to the present for weather-related activities.¹⁶ This includes numerous standard awards, and other types of awards such as quick-response grants (i.e., the collection of time-sensitive ephemeral data for disaster reconnaissance research) and some large cooperative agreements.
• A search for weather- and SBS-related awards from Program Directors of what is now the Infrastructure Management and Extreme Events (IMEE) program in the NSF Engineering Directorate (ENG), which has for several decades funded quick-response grants and other hazards research. Throughout the 1970s and 1980s, IMEE led much of NSF’s social science funding for research on warnings and risk communication, but the program reduced its focus on such issues during the subsequent decade. Our search suggests that approximately $63 million was awarded from 1989 to 2016 for weather- and SBS-related awards directly from ENG through IMEE and related programs.

While these are very rough estimates, the analyses illustrate that only a small portion—likely less than 10%—of all weather-related funding from NSF since 1989 has directly concerned human behavior, decision making, perceptions, or communications.

___________________

¹⁶ This is defined as awards for which “weather” appears in the award title or abstract after manually redacting awards with spurious references to weather (e.g., rock weathering) from the search results.

Weather-related awards from the SBE Directorate over the past several years are diverse—ranging from small dissertation awards on farmer decision making, to large collaborative awards examining perceptions and responses to extreme weather, to prestigious CAREER awards to junior faculty that focus on fundamental theoretical work with the potential to contribute to understanding of the human dimensions of weather in a variety of ways. A couple of notable examples of the larger awards to highlight:

• NSF Award 1331399 funds the development of an Integrated Hazard, Impact, and Resilience Model by an interdisciplinary team of researchers with expertise in individual and organizational behavior, economic modeling, climate

• science, infrastructure engineering, hazard modeling, public health, and spatial landscape analysis. The project will use systems modeling to advance understanding of the impacts repeated hurricanes and heat waves have on regional vulnerability and resilience in the mid-Atlantic region, and to develop approaches for improving resilience to these repeated weather hazards.

• NSF Award 1444755 funds the development of a multi-city network of scientists and communities across nine geographically dispersed cities—the Urban Resilience to Extremes Sustainability Research Network—to explore how the social, ecological, and technological dimensions of urban systems interact to generate vulnerability or resilience to extreme weather related events. The network will also investigate how these dynamics can be guided along more resilient, equitable, and sustainable trajectories.

Tables A.1 and A.2 in Appendix A list additional examples of funding awards from SBE and ENG for research that are relevant to weather concerns. Table A.3 shows the total current funding by directorate from across NSF for weather-related projects related to “perception, behavior, communication, decision making, or action,” along with one sample active award from each directorate.

As is evident from the examples collected, NSF has funded a variety of useful research on human perceptions of and communications about weather hazards (with a small amount of this jointly funded by NOAA), and research on modeling and simulation of resilience and responses of integrated social and physical systems to repeated weather hazards. Awards target fundamental advances in understanding of underlying social and behavioral processes and dynamics, as well as practical application of resulting models and understanding, for example, to increase community resilience to extreme weather events. Although the estimates resulting from our analyses should be considered approximations at best, they suggest that funding levels for weather-related research are quite variable, but have increased over the past three decades¹⁷ (see Figure 3.2).

DHS Research Support

While focused on hazards more generally, a great deal of relevant research is supported through the Department of Homeland Security (DHS) Science and Technology Directorate programs, which aim to “deliver effective and innovative insight, methods, and solutions for the critical needs of the homeland security enterprise” (DHS, 2017).

___________________

¹⁷ These trends persist, even accounting for inflation. Note that the trends are affected by a few large awards given in some years.

The S&T Directorate’s Office of University Programs (OUP)¹⁸ builds partnerships with researchers and educators at numerous U.S. colleges and universities, in part through the establishment of DHS S&T Centers of Excellence (COEs). The COEs are a consortium of hundreds of universities conducting research to address homeland security challenges by developing customer-driven tools and technologies. COE partners include academic institutions, industry, other federal agencies, state, local, tribal and territorial homeland security agencies, and first responders.

A COE is established through a competitively bid 5-year cooperative grant. COE grants cover a range of funding levels depending on the challenge area, but all grant requirements include a systematic review of the delivery of value to end users and active planning for the transition of COE work products to self-sustaining operating status. The COEs are responsible for developing and managing individual project-level grants, which are relatively small grants, with funding reviewed annually.

Many of the COEs support research of direct relevance to the weather enterprise; they support SBS research on public attitudes and behaviors related to risk assessment, perception, and communication applied to multiple hazards, including weather hazards. Appendix A lists some specific examples of research that is relevant to weather concerns being supported at four COEs: the Coastal Resilience Center of Excellence led by the University of North Carolina at Chapel Hill (see Table A.6); the National Center for Risk and Economic Analysis of Terrorism Events led by the University of Southern California (see Table A.7); the National Consortium for the Study of Terrorism and Responses to Terrorism led by the University of Maryland (see Table A.8); and the Critical Infrastructure Resilience Institute led by the University of Illinois at Urbana-Champaign (see Table A.9).

In conclusion, the Committee’s exploration of funding opportunities illustrates a promising upward trend in funding for SBS weather-related research over the past several years, but also illustrates that funding opportunities have been highly variable and are limited by comparison with overall agency budgets for weather-related research. Much, if not most, SBS weather-related research funding appears to have been awarded in the context of interdisciplinary research projects.

3.2 BARRIERS TO INTEGRATING SBS WITHIN THE WEATHER ENTERPRISE

The many agenda-setting, research community- and capacity-building, and communication and information-sharing activities discussed in the prior section have,

___________________

¹⁸ See examples at https://www.dhs.gov/science-and-technology/office-university-programs.

in different ways, contributed to introducing SBS to the meteorological community, introducing weather hazards and associated challenges to the SBS community, and providing forums for collaborating and exchanging ideas. Yet, despite this progress, multiple barriers still limit how SBS research is conducted and applied in the weather enterprise. Some barriers are not specific to SBS and weather (in that they are present whenever disparate research and practitioner communities come together), while other barriers appear to be particularly problematic for this realm of research. Several of these barriers and their implications are discussed below. Although the different barriers are parsed for exposition, in practice most are closely interlinked. This section draws on the expertise and direct experiences of the Committee members themselves and of many people who provided input to this study. It also echoes the findings of recent studies on team and interdisciplinary science (see Box 3.2).

Working at the Intersection of Research Communities with Different Knowledge, Goals, Capacities, and Roles

There are profound differences in the knowledge, roles, goals, and capacities of people who comprise the SBS and weather communities. As in other interdisciplinary research contexts, such differences can present significant challenges, both for individuals who deliberately seek to conduct research at the SBS-weather interface and for individuals who may inadvertently find themselves in this space. These differences can exist among participants within SBS, between the SBS and the weather enterprise, and across public, private, and academic sectors; and they can manifest through specialized languages and terminologies for programs, projects, and job positions, and by research instruments, tools, theories, concepts, and methodologies.

For instance, meteorologists work with numerous instruments to collect in situ data (e.g., thermometers, hygrometers, anemometers), remotely sensed data (e.g., satellite, radar, lidar), and numerical weather prediction model forecasts—all of which form the basis of empirical analysis for weather forecasting and research purposes. Social and behavioral scientists likewise have numerous instruments to collect data, including surveys, structured interviews, experiments, and direct observations in the case of participant observation and ethnography. Meteorologists are often unfamiliar with how social scientists collect data and make meaning of it in ways that are reliable and valid.

This lack of understanding can be exacerbated by the fact that most meteorologists have access to a wealth of freely available atmospheric data collected by agencies such as NOAA and NASA. In contrast, the types of social science data related to weather hazards that are freely available are relatively limited in the type of information they can provide to test and shed light on key concepts and theories.

Another common challenge when groups with differing disciplinary knowledge interact is the potential for miscommunication because vocabulary is not shared or terms are actually defined differently among disciplines. For instance, words as seemingly straightforward as “model accuracy” (defined by most meteorologists as an objective gridpoint-to-gridpoint comparison of forecast to observations) can for social science disciplines mean something different (e.g., a measure of overall effectiveness, which is much broader). These terminology differences can lead to frustration and unnecessary confusion that derail productive discussions. Such tensions can be minimized by defining key terms up front and allowing participants to learn and grow from the interaction. There also has to be an openness to put ego aside and ask for definitions and provide them without judgment. Only when everyone is comfortable asking the questions they need to ask—especially within the context of a specific research or application—can collaboration truly thrive. Ultimately, there is a need both for individual-level efforts to facilitate an environment in which everyone is comfortable asking the questions they need to ask, and for more systematic, collective efforts to advance the development of a shared language and vocabulary.

Differences between SBS and weather research communities reflect the ways in which scientists, scholars, managers, and practitioners are trained, both within their disciplines and on the job. The differences further manifest through job cultures, including organizational missions and visions, specific job roles, and incentive and reward structures. These challenges exist in both directions—that is, for SBS scholars who want to work with meteorologists, and for meteorologists who wish to engage with SBS scholars. Disciplinary and organizational stove-piping can limit incentives and opportunities for groups to work together. Studies of university

support for interdisciplinarity have found widespread barriers, including limited resources, academic reward structures that only reward efforts within the promoting department, and cultural differences between institutions (NRC, 2005b; see also Box 3.2 and the additional references noted therein).

Such challenges can deter those who might otherwise see weather as a rich domain in which to ask interesting questions about human cognition, behavior, and culture that can advance SBS theory. Similarly, meteorologists who wish to work with SBS and stakeholders often are not compensated or credited for such collaborations, or they may feel pressure to contract back to their atmospheric science scope when writing proposals or doing research efforts, despite the potential for such interdisciplinary collaborations to enrich the research done. These challenges are potentially exacerbated by a lack of any kind of social science training for meteorologists, particularly at the graduate level. Furthermore, few social and behavioral scientists are aware of the meteorological contexts in which their skills could be applied.

Such divisions lead to constraints in the number and nature of interdisciplinary research questions that are articulated and investigated. These limitations also are felt in the workforce, affecting whether or not there is any SBS expertise within different weather organizations, and where and how this expertise is situated within a given organization. NOAA/NWS has very little social science expertise within its line office, for example, and extremely few other academic, research, or operational meteorological organizations have SBS expertise either.

Growing and supporting the next generation of workers requires developing ways to train people with interdisciplinary knowledge and skillsets. Just as importantly, it requires ensuring there are employment opportunities where they can meaningfully apply their knowledge and skills. Within the workplace, this requires creating new ways to value, evaluate, and promote people; in the broader community, it requires developing mechanisms and venues for fostering collaborations and for sharing training resources and research across disciplines and sectors. The foundation of strong team performance, in team science as well as other environments, is that each person has a strong area of expertise that the other team members recognize, not that everyone knows about everything (NRC, 2015). The critical goal is to help people understand and learn to function in project-based team environments; both individual expertise and well-coordinated team information sharing are essential to team effectiveness (Cooke et al., 2013).

Problem Identification and Framing

Another major barrier pertains to how SBS-related research questions in the weather domain are identified and framed. Often SBS topics are discussed at conferences, in policy arenas, and through online discussion forums populated solely or largely by meteorologists—with very limited representation of different SBS disciplines, research approaches, and theories. This imbalance in what expertise is brought to the table drives how research and application problems are framed, what research agendas are set, how funding priorities are determined, how success is defined, and even what time frames are considered—for instance, by focusing only on the forecast and warning processes rather than preparedness and mitigation opportunities.

One interesting example of a framing challenge comes from the growing concern among meteorologists about 7-10 day forecasts of weather hazards such as winter storms being posted and shared via social media, sometimes with little to no context of uncertainty information, and in some cases these might be worst-case scenarios. Some meteorologists are concerned that such information can lead to over-warning and false alarms, leading to general distrust of the source and poor decision making. However, this assumes that there is active public attention to such information and that people make protective response decisions based on it. Furthermore, it does not consider possible positive outcomes, such as raising people’s awareness of a possible threat and spurring their information seeking. In other words, such practices tend to be seen by meteorologists as universally bad, and something that must be reduced. Yet, people have always been active consumers of information, especially during times of threat, and this is true in the social media world as well (Neeley, 2014). Although meteorologists’ concerns about such potentially misleading forecasts are valid, it would be useful if the questions were reframed to explore whether, to what extent, when, how, and why these forecasts may have a negative effect, and to better understand the balance between negative and positive effects.

This example also illustrates that often problems identified within the weather enterprise are narrowly focused and reactive to a recent weather event; and they are often driven by paternalistic perceptions that there is “correct” information that people “should” want, and a “correct” response that people “should” take. Thus, the goal is to persuade people to behave accordingly, and people who do not follow these expected ways are labeled as “irrational” and deemed to act in ways that are not aligned with their own interests and well-being. (See further discussion of this issue in Box 2.3.)

Funding Constraints

Another major barrier is the limited funding for SBS-weather activities, including for research, application, and community and capacity-building programs. Funding constraints particularly inhibit high quality social science data collection (including sampling, developing instruments, and conducting field work) and analysis, and research approaches that require significant time to complete, such as longitudinal data analysis and certain qualitative research approaches such as ethnography.

The limited SBS work that is funded tends to be focused on development projects, rather than on basic or applied problems. NOAA’s heavy emphasis on contractor-funded projects, particularly on “stakeholder engagement,” further reflects this tendency. Another major challenge with how NOAA approaches its SBS funding decisions is a lack of transparency in whether and how research priorities are set, how funds are competed, and how funding decisions are made.

A related concern is the lack of mechanisms to support the study of “end to end” processes that link different parts of the weather enterprise. NOAA primarily supports research about NWS processes and products, and individual weather companies focus largely on improving their own capabilities. It is much harder to find support for research about information flow across the “weather communication chain” or about the critical interactions happening across different sectors of the weather enterprise. In Chapter 6 we discuss some possible options for more collective support of these sorts of cross-cutting research needs.

Limited Understanding and Misperceptions of SBS by the Weather Community

The barriers discussed above exemplify a general problem that impedes progress in meaningfully integrating SBS within the weather enterprise—that is, the limited understanding of SBS by many members of the weather community. This problem manifests in many ways, for instance:

• Interdisciplinary insights are lost when just one person—often found opportunistically—is selected to be “the social scientist” and tasked to provide answers, often drawing on just one theory or one set of studies.
• The potential for effective outcomes is limited when social scientists are asked to advise on how to improve an idea or product (e.g., for communicating warning information) that has already been approved for use—and thus the social science input affects only limited aspects of how the product will be used.

• Opportunities for advancing broadly applicable knowledge may be lost when research questions of large, often fundamental scope are framed as narrow projects for contractors to bid on and to “answer” within a very short timeframe and minimal financial resources.
• Valuable workforce development opportunities may be overlooked when graduate student and job positions in meteorological organizations are open to social and behavioral scientists, but are defined only through the lens of what meteorologists think this role should be.
• Progress for the weather enterprise overall may be limited when there is a lack of substantive collaboration with social scientists across the stages of agenda-setting, research, applications, workforce development, training and mentoring, and so forth.
• Risks of problem oversimplification are heightened when new communication products or policies are based solely on meteorologists’ intuitive beliefs about human behavior, rather than more systematically examined professional social science perspectives.

Effective cooperation among professionals with different sets of skills and knowledge requires that all sides involved have a clear understanding of the nature, scope, and limitations of everyone’s expertise, as well as clear expectations about the variety of interventions and solutions that the other experts can offer. Collaborators must be sensitive to these differences and anticipate the implications in order to avoid frustrations and disappointments.

For example, some meteorologists wanting to know how to best issue warnings and/or evacuation recommendations in anticipation of a severe weather event may seek advice from a social scientist (for example, a risk communication expert), hoping for a direct, unequivocal recommendation. Yet, in a social science framing the answers to such a question would likely depend on many factors, such as intended audience, context, dissemination method, geographic and temporal specificity (e.g., does the warning apply to a small, densely populated urban area or to a large rural area with a widely distributed population), and the type, severity, and timescale of the weather event. Some types of social scientists could offer general advice in line with their areas of expertise, but would not necessarily be able to answer the questions posed at the level of specificity requested without consulting with experts in other areas and possibly conducting additional research. Social scientists, just like meteorologists, have diverse areas of disciplinary expertise, thus they may only be able to provide partial answers to broad questions that encompass areas of specialization that diverge from their own.

As another example, meteorologists sometimes express frustration with people’s failure to comply with recommendations for responding to hazardous weather warnings based on highly accurate data and forecasts. This is often rooted in a lack of appreciation of the multiple factors that drive human behavior and the spectrum of activities that lead people to take response actions (e.g., information seeking, understanding and confirmation of information gathered, consideration of whether recommended actions are feasible for a particular individual). Social science helps shed light on the tremendous variance that exists within any given population—in terms of differing contexts, experiences, beliefs, attitudes, cultural values, and other individual attributes—all of which can affect people’s responses to risk (e.g., see Fischhoff et al., 1984). Failure to understand this diversity can create false expectations regarding the effectiveness of weather warnings; and it can increase the risks of operationalizing communication practices that have unintended negative consequences, and that could even place people at increased risk.

Addressing the barriers discussed above will help ensure that SBS research can effectively advance the goals of the weather enterprise. Providing the level and consistency of support that can sustain a growing body of thoughtful, rigorous SBS research will help bolster perceptions of this research across the weather enterprise. It will also help mitigate frustrations that lead scientists working at this interface to pursue alternate research domains, thus helping the field maintain valuable talent, institutional knowledge, and champions of interdisciplinary collaboration. None of the systematic impediments discussed here are insurmountable, but real progress in integrating SBS within the weather enterprise will require sustained efforts to overcome these challenges.

This page intentionally left blank.