Chapter 8
Implications for the Future

Planning committee member, Marshall Shepherd, thanked participants for their contributions to the workshop and introduced the final activity. He explained that the committee intended the roundtable discussion to consider the workshop as a whole, synthesizing the rich material covered over the previous activities. Shepherd and fellow committee member Brad Colman served as moderators, and the panelists were Julie Demuth, National Science Foundation (NSF) and National Center for Atmospheric Research (NCAR), Brock Aun, HAAS Alert Systems, Gina Eosco, National Oceanic and Atmospheric Administration (NOAA) Weather Program Office (WPO), Sherman Gillums, Jr., Federal Emergency Management Agency (FEMA), and Rebecca Morss (NSF).

ROUNDTABLE DISCUSSION: WORKSHOP RETROSPECTIVE AND IMPLICATIONS FOR THE FUTURE

Shepherd opened the discussion by asking the panelists to share what surprised them about the workshop. Several panelists mentioned the challenges posed by increasingly complex data, tools, and approaches—together with the need to maintain simplicity in a complex context. Aun observed that many of the challenges to communication center on the interrelationship of message, medium, and audience, which become more complicated as people have more sophisticated understandings of the content, as well as more communications tools. These challenges will become more complex, he noted, and more and more robust approaches and sophisticated tools must be developed to adequately address the complexity. “I think that it’s an illusion to think that there is one message that we need to get out there and keep it simple.” Demuth highlighted the tension between the desire for hyper-localized information and efforts to communicate uncertainty, especially in relation to longer lead times. Although the ability to deliver hyper-local messages

Eosco raised the question of how to know when there is enough research to make changes to current systems, or recommend them to partners, in order to make systems more inclusive. She noted the usefulness of a “gaps analysis” of the research and development process, identifying what has been done and what is needed, and prioritizing the latter. In 2023, the Weather Program Office (WPO) focused on diversifying its portfolio of projects, she explained, and has considered risk communication in different contexts, such as the deaf and hard-of-hearing community, immigrant populations, pregnancy and heat, and other “unique situations.” Morss echoed and emphasized the necessity of identifying gaps and actively working to address them, stressing the need to identify who should be included. She added that listening to people once they are at the table is essential, and this includes attending to the full “complexity of their situation” and also revisiting conversations over time, as people, technologies, and situations change. Demuth added that “this qualitative work with people in the communities is so essential,” adding that “if we don’t know—really richly and in the complex ways that [Morss and Eosco] are talking about—how people are processing what some of those issues are that they’re facing, we can’t design some of the . . . quantitative work that is trying to measure that.”

Colman then asked panelists to share their takeaways from the workshop and their ideas for plans to move forward: “Where are we going from here?” Morss shared that her largest takeaway was the importance of bringing together diverse perspectives. She pointed to the recurring discussion about how a hazard fits into a person’s particular world: how meteorologists, social scientists, developers of new technologies, communications experts, and others all approach the various challenges and opportunities differently. Having each group understand the others’ perspectives and ensuring that the work of each is “contextually relevant” to the others is critical, she noted. Morss also highlighted the importance of toggling between “the real world” in all its complexity and research to develop understanding in more simplified contexts: “How do you take the questions from the real world and do research . . . or build systems to address them in a more focused way?” This effort requires collaboration across perspectives, and Morss said, “I think that we’re finally at that point where there are enough people in the room” to address these complex questions. Echoing Morss, Eosco noted the importance of bridging research and practice and to ensuring that knowledge gained in research is effectively shared with and used by partners and the general public.

Referring to discussions about localization and personalization of information, Eosco stated a question that framed one of her takeaways from the workshop: “How do we create an agile forecast to meet the needs of every user while still maintaining an official forecast?” Aun added that personalization and localization are “the entire future that we are going towards.” For him, this future entails not only customization opportunities and geographically tailored messaging, but also communication systems with feedback loops so that data can be gathered toward optimizing use, expanding accessibility, and boosting efficacy. Ideally, this system would be built with the aim of establishing a national approach based on best practices.

Demuth raised the idea of a “predictions challenge,” evident in the tension between clear and accurate risk communication, on the one hand, and the extent to which events and impacts simply cannot be predicted or known, on the other hand. Much of the discussion around risk communication assumes the presence of knowledge, she said, but this is not always the case. Rapid intensification is one instance of a prediction challenge, as are co-occurring compounding hazards. Predicting impacts is important, but in the case of the extremes, mapping impacts onto the meteorology can be very challenging, she said, reiterating a point made by Jeff Lindner in his presentation the previous day. Demuth also reiterated Lindner’s question about the role of the meteorologist and whether or not they are responsible for predicting impacts.

Gillums reminded the group that human beings are, and must be kept, visible and at the center of all of this work. In that vein, he urged attendees not to underestimate the importance of their own experiences, perspectives, emotions, and predictions about how they might react to certain situations or questions—and to leverage those insights as they conduct research and craft messaging.

Shepherd then asked panelists to comment on any gaps or significant opportunities for further work they observed over the course of the workshop. Eosco responded, offering four major touchpoints. First, there is no good way to evaluate the system of forecasts and impacts; with so many components within the system, how can an individual component be evaluated? Second, returning to the theme of localization versus predictability, she noted that many current tools are rooted in a one-size-fits-all mode of thought. One gap, then, is building new forecast systems that enable more agility and customization, which would involve learning the various needs of users. Third, when resources are limited, understanding and targeting hyper-focused audiences can be challenging. Finally, artificial intelligence (AI) should be considered a tool, and not a solution. Aun noted a gap between “the human element at the local leadership level” and the potential for new forms of technology that might help to enhance decision-making by local officials.

Several panelists offered other gaps in research, language, and perspective. The critical incident stress experienced by forecasters, emergency managers (EMs), broadcasters, officials, and community members is another gap in knowledge, Demuth observed. A broad frame of study might be useful here, she noted, and might explore whether these groups are at risk for moral injury when a lack of capacity causes them to be unable to protect the lives and livelihoods of the people they serve. Gillums observed that the pervasiveness of a “deficits perspective” constitutes a sort of gap. To this end, he advised, “assume what you’re doing is already working to an extent,” and voiced a hope that people will not underestimate how successful their work is and has been in helping people think through decisions during hazardous events. Information is trickling through and having a positive effect: “What you’re doing is saving lives.” Morss, offering a final response to this question, observed that the partnerships, learning, and conversations deepened over the course of the workshop will help to tackle the gaps mentioned above.

Colman noted that, although a great deal of uncertainty is inherent in weather information, other areas exhibit much more certainty. These areas include knowledge about communities, the built infrastructure, and the physical environment, which can be drawn on to mitigate or inform uncertainty about weather information, especially in the work of communicating information in personalized or localized ways. “Don’t let the uncertainty piece necessarily be the driver of personalization,” he commented.

An audience member shared the perspective of EMs who make and communicate about decisions around evacuation orders and other protective measures. They noted that panelists might learn a great deal from close study of the moments when a decision is made, attending to both how quickly EMs must make decisions and how deliberate they are in making those decisions. Communicating the decision-making process to the public has also been important in their community, the audience member explained: who is making decisions, who has the authority to do so, and what expert advice is informing the decision. Aun noted his own surprise at learning, through his work, how under-resourced first responders are, and how much is expected of them. Taking on more risk with fewer resources is an impossible task, he observed.

Bob Hershey, an audience member, wondered about the possibility of representing probability of damage and loss resulting from a storm in terms of monetary value. This exercise might be done ahead of time as a way to communicate how the cost of damage might increase or decrease depending on the path of the storm, he suggested. Gillums responded that the idea is interesting, but that people might still misunderstand the concept of probability that underpins those numbers. He added that uncertainty sometimes means the spread of probabilities is very large.

A third audience member asked about approaches to supporting partnerships—particularly because communities and local groups might already be overtaxed and under-resourced. Gillums noted that a FEMA remit is to assess local needs, which is shared by municipal, state, and tribal authorities often working with the federal government. He noted that FEMA’s reactive position is problematic: resources are available, but many communities do not know what they need. This gap could be mitigated though ongoing relationships that help communities prepare for such events, but because FEMA becomes involved after an event, these relationships and pre-event preparedness work often do not occur, so communities cannot take full advantage of the federal resources available. “I don’t think it’s a resource problem, at least from the federal government side,” he commented. Aun noted that federal funds to expand the HAAS Alert system and to provide for basic equipment for first responders are drawn from the same pool. Rather than compete for those limited resources, he said, “we had to partner with our customers to go to the federal government” and lobby for new funding sources.

Eosco, in response to the earlier comments about EMs, noted that partnerships could also enable real-time observation. Often research is conducted after the fact, which can yield different results than research conducted in the moment. She wondered about the possibility of joining a local office during an event—not